download "Translations" - MDC

Note
This is a story about people, science, and ideas, written for anyone who is curious about today’s research and its
practitioners.
Society is in the early stages of a revolution which is already changing our lives and will have a much greater impact in
the years to come. Until recently it was happening quietly, in the cell cultures and laboratory flasks of research institutes
all over the world. But now not a day goes by without important new discoveries that shed light on fundamental
questions about life and give insights into the processes that lead to cardiovascular and neurodegenerative diseases,
cancer, AIDS, malaria, and the other major afflictions of mankind. While these discoveries are as much a cultural treasure
as great works of literature or art, they are harder to appreciate because they usually concern details of biological
processes that can’t really be put into context in soundbytes or newspaper articles.
Sometimes the best way to understand a whole is to observe a small part in great detail and show how it is related to
everything else. The campus in Berlin-Buch is a microcosm of what is going on in experimental and clinical research
laboratories all over the world. By focusing on the work being done there, this book aims to open a window on that much
larger context. At the same time, the campus is home to unique people with a special relationship to history, culture and
the arts. Their stories are fascinating in their own right.
As big as this book is, it presents only a small part of the research on campus. Other groups have equally interesting
stories that are waiting to be told. These stories are intertwined with those of many others who quietly work in the
background to make the science possible: technicians, administrators, the teachers of the Gläsernes Labor, secretaries,
architects, cooks, cleaners, drivers, etc. This book is dedicated to them. It is also dedicated to the dozens of scientists who
directly participated in its production and permitted me to write about their lives in a very personal way. I apologize for
the mistakes that remain despite their immense patience in explaining their work.
I would particularly like to thank Walter Birchmeier, Jens Reich, Heinz Bielka, Walter Rosenthal, Friedrich Luft, Björn Maul,
Gudrun Erzgräber, and Detlev Ganten for many stimulating discussions and help with the manuscript. Okan Toka very
kindly spent a lot of time sharing his adventures in the Muskateer story, for which Silvia Bähring provided most of the
photos; some of them were taken by Hakan Toka. Ewan St. J. Smith provided images of the naked mole rat, and Catherine
Adamidi photographed the planaria. Peter Himsel and David Aussenhofer also generously provided photos. Elizabeth
Kujawa-Schmeitzner, Christina Quensel, Maxine Saborowski and Wolfgang Groll provided invaluable organizational and
moral help. Jane Holland helped with the interviews. Thanks also to Andreas Maetzold and Ulrich Scheller for many
discussions about the campus and its future. Wolfgang Uckert had been saving some old laboratory bottles for many
years, knowing that they would eventually be needed for something, and they finally found their destiny on the cover of
this book. The Ivics family invited us into their home for pictures at the dinner table. Caroline Hadley did an excellent job
of critically reading and editing the manuscript. Thanks to Dr. Engelsing and his colleagues at the museum of the City of
Konstanz for a superb image of the Leiner collection. My appreciation to Hus No. 37 (the Swedish coffee shop in Pankow),
and Star Coffee in Heidelberg, where large portions of this book were written. Thanks to Rainer Bender, Ingrid André, and
the team of ColorDruck for their sharp eyes and good ideas and all of their input at many stages of this project. Maj Britt
thanks in particular Manuela Brunner, Anita Glanz and Andreas Schmidt. She also thanks Klaus and the Vorgrimmler
family for all their support and help. Nicola thanks her family: Jürgen, Lynn and Franka for their patience and support.
I would like to thank my family - Gabi, Jesper, Sharon, and Lisa - for the many sacrifices they have made so that this book
could exist.
Russ Hodge
May, 2008

translations
from today’s science
to tomorrow’s medicine
in Berlin-Buch
by Russ Hodge
Maj Britt Hansen, photography
Nicola Graf, design

Table of contents
Science and the city: a drive to Berlin-Buch............................................................. 4
Part One: The lives of nomads .................................................................................................. 20
Songlines and shapeshifters ......................................................................................... 22
The archeology of blood ................................................................................................. 40
Breakouts in the brain..................................................................................................... 52
Stem cells in a jar .............................................................................................................. 62
Capturing a firefly............................................................................................................. 70
A very quiet cure................................................................................................................ 78
A brief incident of teleportation.................................................................................. 86
Interlude: A glass heart.................................................................................................................. 88
Part Two: Identity crisis.................................................................................................................. 100
Where slow rivers meet.................................................................................................. 102
Lost anchors and wrecked vessels............................................................................... 112
A pact with the devil........................................................................................................ 118
The decorators of Versailles........................................................................................... 132
Waking a Sleeping Beauty and other tales of ancient genes............................ 140
Table of contents
2

Interlude: When rats don’t dance to the red-hot chili peppers......................... 150
Part Three: Frontiers and ferrymen........................................................................................ 158
The Swiss army knife of the cell .................................................................................. 160
Bad origami......................................................................................................................... 174
Trouble in the waterworks............................................................................................. 184
Bribing the ferryman ....................................................................................................... 194
The electrician’s toolbox................................................................................................. 200
A kiss, interrupted ............................................................................................................. 210
Interlude: The case of the short-fingered musketeer............................................... 218
Part Four: From the protein village to the cosmopolitan cell ............................. 238
Interviews:
A hub in the city of the cell ........................................................................................... 240
The pleasures and powers of green tea.................................................................... 250
Playing the piano of planaria........................................................................................ 262
A rat in a tree and a wolf on the loose ...................................................................... 274
Walter Birchmeier, Scientific Director, MDC ............................................................ 284
Walter Rosenthal, Director, FMP................................................................................... 288
Detlev Ganten, Charité.................................................................................................... 292
Gudrun Erzgräber, BBB .................................................................................................... 296
Index.............................................................................................................................................................. 300
Further reading.......................................................................................................................................... 302
Credits........................................................................................................................................................... 305
3 Table of contents

Science and the city:
a drive to Buch
This book was born during a ride through
Berlin on a luminous autumn day in 2006.
The train had come in from the west, gliding over a
landscape of parks so spacious and green that it was
often hard to believe we were entering the heart of a world
capital. We passed above the zoo, where a long line of children
waited eagerly for their first look at a newborn polar bear. “Knut”
could be seen everywhere, in gift shops, the city’s newspapers and on
the title pages of magazines. He had become a symbol of the zoo, of the
city, of environmental campaigns. Global warming was high on everyone’s
minds; people had been talking about it on the train. November
had been so mild that the trees were only now changing color.
The train was full of people on their way to Berlin – tourists, students,
half an orchestra, scientists, business people talking loudly into cell
phones. It was always interesting to strike up a conversation with a
neighbor; people were being drawn to the city for a thousand different
reasons. The conversations stopped as the train slid into Berlin’s new
main station, a cathedral of glass and girders that decorated several levels
of platforms in light and shadows.
Science and the city a drive to Buch
4
Berlin main train station

At the top of the stairs a man in a leather jacket
asked people for used S-Bahn tickets.
Andreas Hollman, a driver from the Max
Delbrück Center, was waiting at the bottom of
a labyrinth of escalators. There would be a
short walk to the car, for which he apologized.
The capital city had built a fantastic Bahnhof,
he said, but it was miserable for drivers.
Hundreds of people needed to be dropped off
and picked up every hour, and the one small
street in front couldn’t handle the traffic.
Leaving the station, we caught a glimpse of
the glass dome of the Bundestag. Around the
corner stood the tall brick buildings of the
“Charité”, one of the world’s largest and oldest
hospitals, founded in 1710 in anticipation of an
outbreak of the bubonic plague that never
came. Andreas pointed out the television
tower that looms over the Alexanderplatz, in
the distance. “The round part at the top
Knut and Thomas Dörflein
at the Berlin Zoo
looked like a soccer ball for the World Cup.
They just removed the decorations.” He
paused. “They should have left it.”
The scientific campus of Berlin-Buch lies on
the northeastern outskirts of the city and is
home to a number of important insitutions:
the Max Delbrück Center for Molecular
Medicine (MDC), the Leibniz Institute for
Molecular Pharmacology (FMP), clinics of the
Charité, and a number of biotech companies.
If we had made this drive in 1988, the first
time I had visited Berlin, we would have had
to cross the border into East Germany. By the
time of my next visit, in the mid-1990s, the
Wall was gone but navigation through the
city was hindered by hundreds of construction
sites. Now there were few obvious signs
of either phase of the city’s recent past. The
drive took us past a huge building complex of
the Bayer Schering Pharmaceutical company,
through Wedding, along the Schönhauser
Allee into Pankow, then into the quiet neighborhoods
of the suburbs.
Andreas often picked up new staff members
like me or people coming for interviews and
was happy to play tour guide. If most of Berlin
didn’t have the monumental personality of
London or Paris, he said, the reason lay with
history. The city had not begun as a capital but
rather as separate villages that grew together.
That explained the low buildings, the green
spaces. Now those villages had become
neighborhoods with distinct personalities.
Buch has its personality as well. In the late 19th century Berlin – like most of industrialized
Europe – was afflicted by an epidemic of
tuberculosis. Hospital wards were so full that
patients spilled out into the hallways; beds
were set up wherever space could be found.
Science a nd the city a drive to Buch
6
Infectious diseases love a crowd, and health
officials decided to build new hospitals outside
the center to limit their further spread.
One of Germany’s most distinguished architects,
Ludwig Hoffmann, was hired to begin a
massive construction project in Buch.
Beautiful brick buildings with balconies and
gables were erected on several sites around
the village. Buch quickly became known as

the “Hospital city,” or “the city of the ill.” That
tradition has played an important role in the
current incarnation of Buch as one of
Germany and Europe’s most important centers
for biomedical research.
The campus institutes are devoted to “molecular
medicine,” reflecting the promise that
what has been learned about the basic units
of life – DNA, proteins, and the other mole-
cules within our cells – can be transformed
into new methods of diagnosing and treating
diseases. Fifteen years ago the campus in
Buch was one of the earliest centers in Europe
to devote itself to this theme. At the time it
was a bold move. Laboratories and clinics
belonged to different scientific cultures, concerned
with different questions, using different
methods, and training their students very
7 Science and the city a drive to Buch
Roland Schmidt
differently. To arrange a marriage between the
cultures, they first had to be brought together.
With its existing clinics and a strong biology
institute, Buch seemed to be the ideal
place to build bridges between the treatment
of patients and research labs devoted to fundamental
(and often arcane) questions about
the functions of genes and the inner lives of
cells.

In the meantime many institutes are defining
(or redefining) themselves around this theme.
That reflects progress in science, but also
some realities about science politics. Most
research in molecular biology is carried out in
publicly-funded universities or other academic
institutes. They regularly have to compete
for funding, and one source is the large sums
that politicians are willing to spend on medical
research. So many institutes are redefining
themselves as centers for molecular medi -
cine, in so many contexts, that it is often hard
to understand what the term means.
I didn’t really know what it meant, although I
had been working alongside scientists for
more than a decade. Maybe the campus
would provide an answer. If anyone could
define molecular medicine, it ought to be the
scientists of Berlin-Buch, who have been practicing
it for 15 years. This book reflects what
many of them have to say.
Science and the city a drive to Buch
8
��
During the drive Andreas Hollman talked
about modern Berlin and its spectrum
of cultural offerings: museums, theaters, concerts,
lectures and literary events. He pointed
out traces of history: high-rise buildings that
have sprung up in the empty zone that used
to border the Wall; houses still bear scars from
shells fired during the Second World War. He
didn’t say much about Berlin’s scientific histo-

y although many of the world’s greatest scientists
have lived and worked in Berlin, and
their tracks can be found throughout the city.
Following their trail is worthwhile; one way to
understand what is so unique about today’s
science is to take a brief look back at its roots
in the 19th century.
Within just a few decades in the mid-1800s, a
number of thinkers – almost entirely from
Europe – completely revolutionized the life
sciences. They created modern medicine, cell
biology, evolution, genetics, embryology,
chemistry, physics, and other fields. There is a
similar revolution going on today. It also
involves changes in our concepts about
nature and life, but its main feature is how all
of these fields are rapidly coming together in
new ways.
There was time to drive through town, so
Andreas followed the Friedrichstrasse to
Unter den Linden. Most tourists come here to
see the Brandenburg Gate and to stroll down
9
Science and the city a drive to Buch
the spacious, tree-lined boulevards. Farther
down the street is a marble statue of
Alexander von Humboldt, the namesake of
Berlin’s largest university, on a pedestal outside
some of the buildings where he gave lectures
in 1827 and 1828. By that time Humboldt
was rapidly becoming one of the most
famous men in the world. He was a universal
genius and author of articles on every conceivable
scientific topic – from mineral formations
on the Rhine to the vegetation that

grew in mines, the plants and animals of South
America, the movement of continents, and the
composition of the atmosphere. His accounts
of his expeditions through South America combined
exact observations with the romance of
discovery and inspired generations of young
naturalists, including Charles Darwin and
Alfred Russel Wallace. Their own journeys in the
tropics led to the theory of evolution.
Humboldt’s generation and the next saw the
birth of most of the fields that make up
today’s biology and medicine. One of those
disciplines, cell biology, began with a chance
meeting at a Berlin train station. A biology
student named Theodor Schwann struck up a
conversation with Mathias Schleiden, an
eccentric student from Heidelberg. There
Schleiden had studied law, even opening a
practice when he moved to Hamburg, but
bouts of depression led to an unsuccessful
attempt at suicide. As a result he carried a bullet
in his brain for the rest of his life. Now he
had moved to Berlin to start over, this time as
a botanist.
Alexander von Humboldt Hermann von Helmholtz
Schwann had arrived from Bonn with one of
his professors, Johann Müller, who pushed his
students to use a new type of microscope just
invented by the Englishman Joseph Lister. The
instrument was made of two lenses mounted
in a tube, a construction that solved some
problems of distortion and dramatically
increased its resolution. It gave scientists their
sharpest view ever of the microscopic world.
Schleiden’s observations led him to realize
that plants were built of fundamental units –
single cells – which somehow formed from
the nuclei of other cells. He had developed a
close friendship with Schwann and mentioned
the discovery to him one night over
dinner. Schwann immediately realized that
the idea might explain what he had been
observing in animal tissues. The two men left
their meal half-eaten and rushed over to
Schwann’s laboratory. Previously, anatomists
had believed that animal tissues were made
of fibers, grains, tubes, and other objects. They
had been looking at cells, Schwann said, without
knowing it.
Science and the city a drive to Buch
10
Many of Müller’s students went on to become
pioneers in biology and medicine. Today the
Charité’s institute of pathology is named after
one of the greatest of them, Rudolf Virchow.
After his studies Virchow received a double
professorship from the university and the
Charité, where he treated patients and continued
investigating basic questions about cells.
He took Schleiden and Schwann’s observations
a step further by proving that a cell
could only arise from another cell. Previously
many scientists had believed they could arise
by themselves, somehow crystallizing from
more basic substances. Virchow’s simple new
idea had a huge impact because it changed
the way scientists thought about all sorts of
questions, from the growth of embryos to the
nature of disease.
For example, it gave Virchow a new view of
cancer. He realized that tumors arose from
small pools of cells that divided too often, in
the wrong places. Removing the source might
stop the spread of the disease. He developed
new laboratory methods to diagnose the seri-

ousness of cancer and new surgical procedures
to treat it. Some of his ideas were ahead
of their time: his experience with patients
suggested that the disease often arose at
sites of injuries or infections. This made him
think that tumors might be linked to inflammations
– the body’s response to injuries. For
over a century most scientists rejected this
idea, but today scientists in Buch and elsewhere
have gathered convincing evidence
that he was right. There are clear links
between some types of cancer and autoimmune
diseases in which the body has trouble
distinguishing between its own and foreign
cells. Virchow’s discoveries likewise exposed
the origins of other diseases. For example, he
was the first to understand how blood vessels
could become blocked by clots that formed
elsewhere in the body, a process which he
called embolism.
One of Virchow’s colleagues, Robert Koch, laid
another cornerstone of modern medicine. He
discovered that a bacteria caused anthrax and
– with Louis Pasteur – proposed the theory
Old train station at
Pankow-Heinersdorf
that all infectious diseases were caused by
microorganisms. Over the next few years
Koch supported the theory with the discovery
of the sources of tuberculosis and cholera,
also bacteria. By 1891 his reputation earned
him the directorship of a new Institute of
Infectious Diseases in Berlin, which soon
moved to a new building in Berlin-Wedding.
Now called the Robert Koch Institute, it can
still be found at Norduferstrasse 20. One of
Koch’s first acts as director was to hire Paul
Ehrlich, who discovered substances that could
treat sleeping sickness and syphilis. His work
helped lay the foundations of modern pharmacology.
Today the development of new drugs lies
mainly in the hands of chemists, and that was
the subject chosen by a young Scotsman
named Archibald Couper when he enrolled at
the University of Berlin in 1854. Four years
later he drew the first map of the positions of
atoms inside a molecule. Prior to this time
most chemists believed that molecules
(assemblies of many atoms) could be under-
Diagrams of molecules by Archibald Couper
stood simply by listing the proportions of
each element they contained. Couper showed
that to understand their behavior, you also
had to know how their atoms were arranged.
(A German biochemist named Auguste
Kekulé, working in Heidelberg and Ghent, had
come to the same conclusions and published
atomic models the same year.) This principle
underlies the field of structural biology, which
exposes the physical and chemical structure
of molecules such as proteins and DNA to
explain their functions.

12
The same year that Couper published his
drawings, Charles Darwin and Alfred Wallace
announced a new theory to explain the origins
of species and their relationships to each
other. Evolution quickly began to change the
way research was being done in many fields.
Ernst Haeckel, who had studied medicine in
Berlin under Virchow and Müller and then
moved to the university in Jena, had decided
that he preferred the quiet life of the laboratory
to dealing with sick patients. His main
interest was the grand mystery of how the
bodies of animals developed from single cells.
As he studied this process in a variety of
organisms, he noticed that the embryos of
different species went through phases in
which they looked remarkably similar,
although they ended up quite different as
adults. Evolution offered a new perspective on
the problem, and Haeckel devoted himself to
understanding the links between an individual
animal’s life history and the long evolutionary
history of its species. One of his own
students, Oskar Hertwig, eventually returned
to Berlin to become professor of anatomy at
the university. A main theme of his work was
embryonic development. He was the first person
to describe the fertilization of egg cells by
sperm and some of the early steps of animal
development.
The work of Haeckel and Hertwig underlined
the fact that creatures don’t only inherit features
like eyes, brains, and hair color from their
parents; they also inherit the processes by
which those features form in an embryo.
Understanding why this was so would require
a solution to one of science’s greatest mysteries:
heredity. In the 1850s and 1860s that problem,
too, was being solved – in an unlikely
place, a quiet abbey in Moravia. Here a monk
named Gregor Mendel was carrying out
breeding experiments with mice until a visiting
bishop found the animals mating in his
room. The mice had to go. Mendel moved to
the monastery garden to work with peas
and other plants. No one objected to that
because, as Mendel remarked ironically, “The

ishop did not understand that plants also
have sex.”
The monk was an excellent statistician and
became one of the first to apply mathematics
to a biological problem, which is echoed in the
huge role that mathematics and computers
now play in the modeling of biological
processes and the analysis of genomes.
Mendel’s painstaking experiments revealed
the basic laws of heredity. His work was published
in a small journal, but few people
noticed – including Charles Darwin, who was
carrying out experiments on peas at the same
time.. Bad luck prevented Mendel from publishing
further important papers on the topic.
He had been corresponding with the famous
Swiss botanist Karl von Nägeli, who recommended
that he repeat the experiments with
a plant called the hawkweed. Cross-pollinating
the plant was very difficult, and no one
realized that there were some unusual factors
in the way it reproduced. This meant that
Mendel’s laws didn’t hold up, and the monk
became unsure of his results. Soon afterwards
he was appointed abbot of the monastery
and became immersed in administration and
political squabbles. As a result, his work was
almost completely forgotten for nearly four
decades.
The link to Berlin lies in Mendel’s rediscovery,
which happened simultaneously through the
work of three scientists working independently
in three different countries. One of the
men, Carl Correns, was studying heredity in
plants at the University of Tübingen. In
January 1900 Correns published an article
that appeared almost simultaneously with
studies by Hugo de Vries and Erich von
Tschernak. Taken together, they proved that
Mendel’s laws were almost universal. As well
as ensuring Mendel’s place in history, the
work gave a boost to Correns’ career. In 1913 he
was appointed director of the new Kaiser-
Wilhelm-Institut für Biologie in Berlin-
Dahlem. He stayed there until his retirement,
continuing to work in the laboratory but
rarely publishing his work. As a result, most of
it was lost during a 1945 bombing of Berlin,
when his manuscripts were destroyed. In
Berlin, science and history have always
been deeply connected. For better and for
worse.
��
O ne neighborhood blurs into the next.
Just as Berlin’s villages have grown
together to produce a cosmopolitan city, the
13 Science and the city a drive to Buch
villages of science have been growing together
to produce modern biomedicine. Cell biology,
genetics, evolution, embryology, chemistry,
physics, pathology, and medicine have steadily
become intertwined to produce a set of
powerful concepts and tools by which today’s
scientists are investigating life. The process of
fusion is not yet complete, but the closer it
comes, the closer we will be to understanding
the causes of the greatest health problems

facing our society today and finding ways to
treat them.
Until very recently, the greatest worldwide
threats to human health were infectious diseases.
Modern sanitation, antibiotics and vaccines
have changed that situation in many
places – but not everywhere. Malaria, AIDS,
and other infectious diseases continue to decimate
huge regions of the globe, and new
plagues will continue to arise and cut a swath
of destruction before cures are found. But in
developed and developing countries, the
major killers have become “old age” conditions
like cardiovascular and metabolic diseases,
cancer, and neurodegenerative diseases.
The culprits in these cases are partly
our own genes, partly the environment.
The last revolution in medicine could take
advantage of the body’s own defenses in com-
bating disease. Vaccines stimulate our existing
immune system, which arose under the
pressures of evolution. Natural selection
works to protect organisms as long as they
reproduce and care for their young, and viruses
and bacteria attack the young, so animals
developed a sophisticated immune system to
protect them from such diseases. But we do
not have natural defenses against today’s
major health threats, which arise late in life,
often as a result of normal aging processes.
Fighting these diseases – as well as very complex
infectious diseases like AIDS – will require
a new kind of medicine that reaches into cells
and reprograms the machines that drive
them. In doing so, it must avoid disrupting
normal operations. Most of the machines are
needed throughout our lives, in many different
contexts. They have to continue to func-
Science and the city a drive to Buch
14
tion in all of those situations in spite of the
changes brought about by drugs. First all of
their functions have to be known, and our
knowledge of what goes on in each of our
cells is still far from complete.
��
Athorough scientific tour of Berlin would
take weeks; the snapshot given in the
last few pages covers just a tiny fraction of
discoveries made in the 19th century, and those
traditions have continued to the present day.
The much fuller story has been captured in
other books, including several written by the
scientists of Berlin-Buch. Heinz Bielka, who
worked on the campus for over four decades,
has devoted his retirement writing about its
history. Helmut Kettenmann, whose ongoing
research at the MDC is described later in this
book, has also written extensively on the

topic. The two men collaborated on a Medical
History of Berlin, and Helmut co-wrote a second
book called Brains of Berlin, written to
accompany a 1998 exhibition by the
European Neurosciences Forum. These volumes
give a much fuller account of some of
the personalities of Berlin science, particularly
in the area of brain research.
I had the Brains of Berlin book open on my lap
as we drove north, alongside the elevated S-
Bahn tracks of the Schönhauser Allee. Ten
minutes later came the ruined round train
station at Pankow-Heindersdorf, ringed by
shattered windows; any intact glass was
cover ed in graffiti. The city gave way to countryside,
still a lush green from the warm
autumn. I leafed through the pages of the
book and tried to get a feeling for Berlin-
Buch.
A few names jumped out because of their
direct connection to the campus. If anyone
could be counted as the heir of Alexander von
Humboldt, the universal genius, it might be
Hermann von Helmholtz, whose statue
stands near that of Humboldt at the university.
Helmholtz was another pioneer of interdisciplinarity,
combining a career in medicine
with research into physics, and the link to the
campus in Buch lies in his name. The MDC
belongs to one of the largest research infrastructures
in Germany, the Helmholtz
Association of National Research Orga -
nizations.
As early as the 1850s Helmholtz was bringing
physics and medicine together in new ways.
One of his early projects focused on muscles,
trying to show that tissues contained enough
energy to drive their activity. This countered a
15
Science and the city a drive to Buch
Heinz Bielka
very old idea that substances required the
addition of some sort of special “vital force” to
become alive. In Europe vitalism was a strong
philosophical tradition, bordering on mysticism;
it was an inspiration for Frankenstein
and later a number of bad films in which mad
scientists zap dead bodies with electricity to
bring them back to life. Vitalism began to give
way in the late 19th century when chemists
managed to create organic substances out of

Jeanne Mammen: Trompetender Hahn (ca. 1943) Tempera on cardboard
inorganic matter. The philosophy was buried
entirely (at least in biology) with the discovery
of DNA’s role in heredity and cell chemistry.
But it had its function: a number of important
physicists became interested in biology, hoping
to discover new physical forces. They
brought along a new way of looking at life
that inspired an entire new generation of
young people to study science.
Helmholtz’s later work focused increasingly
on how the brain perceives sensory information
– color, motion, and sound. He made
groundbreaking measurements of the velocity
of impulses as they traveled through nerves
and even tried to develop theories of sart and
music appreciation based on what he was
learning about the nervous system.
Helmholtz was a gifted pianist himself. It is
not unusual for scientists to cultivate artistic
skills – that, too, is a tradition in Buch. Several
of the scientists working there are gifted
musicians or artists, and the campus is
adorned by sculptures and paintings.
One of those works, a Cubist portrait of the
great geneticist Max Delbrück, hangs in the
office of MDC Scientific Director Walter
Birchmeier. It was made by Jeanne Mammen,
a lifelong friend of Delbrück, who supported
her and amassed a large collection of her
paintings. More than a dozen of her works can
be found in the “Jeanne Mammen Saal,” a conference
room located in the gatehouse on the
Buch campus.
This small building is the first that visitors see
when they arrive at the site. It’s an unusual
entrance for a scientific campus, the product
of another curious bit of Berlin’s scientific history.
Originally city planners intended to
Science and the city a drive to Buch
16
develop a cemetery on the land. However,
most of Berlin sits on a thin layer of soil atop
a marshland. Any hole that is dug quickly fills
with water. That ruled out a cemetery,
although the gatehouse, a gardener’s house
and a chapel had already been built.
��
Over the past 50 years a revolution has
been happening in science, mostly a
quiet one, carried out in the biology departments
of hospitals, universities, and independent
institutes like the MDC and FMP. A
look at any day’s headlines shows that the
revolution is quickly expanding far beyond the
laboratory. Scientists have sequenced the
genomes of humans and many other animals;
they have learned to clone mammals,
and tried new “gene therapies” on human
patients. They hope to cure disease by training

the cells of the immune system to kill tumors,
or by transplanting stem cells that will regenerate
damaged tissues. These projects are still
in their infancy, and it may be many more
years before they become standard tools in a
doctor’s black bag. But most scientists expect
this to happen.
Science is becoming both highly interdisciplinary
and highly specialized. Most research
involves detailed events within cells or organisms
that are hard to understand without
expert knowledge. While the biomedical revolution
increasingly affects us all, few people
have daily contact with the world of
researchers. The picture of scientists presented
by the media is often unbalanced and
unfair. So it is no wonder that the rapid
changes have raised concerns and even fears.
In this environment, institutes have a respon-
17

sibility to open their doors and explain their
work in a public dialog.
Ideally, everyone should have the chance to
visit a place like Buch and talk directly to its
scientists. Each year thousands of school children
get that chance through trips to the
campus’ teaching facility, called the
“Transparent Laboratory,” and many more
visit during Berlin’s “Long Night of the
Sciences.” This book has been written for
those who have not yet made the trip and for
those who would like to know more.
��
In the novel Life, a User’s Manual the French
author Georges Pérec meticulously examined
an apartment building from top to bottom
as if it were made of glass, describing
every room, every stairwell, every occupant. If
we could achieve a complete understanding
of one place, he implied, we would gain
insights into the entire world.
Pérec’s book came to mind the first time I visited
the campus. The exterior of the new
Center for Medical Genomics consists entirely
of windows; from across the street, you can
see which scientists are in their offices, which
are meeting with students, and who is
thumbing through the newspaper in the coffee
room. The same is true of the MDC administration
building. You don’t need to call to see
if a colleague is in his or her office; you can
simply go outside.
This book does not visit nearly every laboratory
of the Buch campus, nor does it explore all
the stairwells (although some of them, such
as the spiral staircase in the Medical
Genomics building, are worth seeing). And if
you were to walk through the hallways and
knock on doors, you would encounter many
people who do not appear here – the secretaries,
the Hausmeister, the conference office,
the computer group, the people who order
and deliver and inventory supplies... The list is
much longer. All of these people play a vital
role in the life of the institutes, and without
their contributions none of the stories in this
book could have happened.
So even in a work of this size, there are
inevitably many gaps. But the intent is similar
to that of Pérec. There are stories to be found
everywhere. The institutes of Berlin-Buch are
among thousands across the globe devoted
to biomedical themes. The work being done
here has some unique features because of the
individuals who are doing it, the history of the
campus, and the character of the institutes
that have been assembled. Some of that work
is among the best in the world, but the purpose
of this book is not to boast or advertise.
Instead, a deep look at the work being done in
Buch offers a window onto this unique
moment in the history of science, and a
glimpse of what it is likely to produce in the
future.
Science and the city a drive to Buch
18

Part One: The lives of nomads

Songlines and
shapeshifters
Most days in the early afternoon, Walter Birchmeier leaves his
research lab in the Max Delbrück Center and traipses across
the street past the bus stop, a statue of a blue bear, and a bronze bust
of Hermann von Helmholtz. Arriving at the administration building, he’s
about to enter his second life as Scientific Director of the MDC, and
sometimes he sighs as he opens the door.
It’s a stunningly beautiful building. The front door opens onto a cathedral-like
space that rises five stories. The ceiling is reminiscent of the
bottom of a ship – if the ship were made of glass – it even has large, propeller-like
fans on either end. The far wall is a vast surface of white
brick, decorated with tall paintings. Behind that wall is one of
Germany’s most advanced facilities for the care and study of laboratory
animals. Sometimes there is a faint,
zoo-like scent (which you don’t notice after
the first month) and the fans kick in to draw
it out. A series of bridges link the offices to
long flights of stairs, and colorfully
painted alcoves look down from
the upper floors. It would be
a perfect place for theater
22

Walter Birchmeier

performances; a few years ago one of the
German television stations filmed part of a
murder mystery here. The acoustics are magnificent,
and short concerts are sometimes
given here by some of the musicians on the
staff.
Walter is one of them, as evidenced by three
tall organ pipes mounted on the wall just
inside his office. He inherited the pipes from
the former director, Detlev Ganten, but they
are appropriate decorations for Walter’s office
as well. Long ago, in what seems like another
life, he studied church music. “I have the diploma
here somewhere,” he says, rummaging
through a stack of personal papers. “If I can
find it, I’ll show it to you.”
Before entering the office he has to get past
his secretary, Elisabeth Kujawa-Schmeitzner,
who greets him with a list of his appointments,
a folder of paperwork waiting for his
signature, and other business. Occasionally
the sight of the list evokes another sigh. “No
appointments today,” he says hopefully, but
it’s a lost cause; the calendar is already full.
Administrative duties are the price that has to
be paid when leading a major scientific institution.
One reward is getting to develop a
vision for the future and steer several hundred
scientists toward it. Not that they are
always easy to steer.
Today’s most pressing issue is money. The
deadline is drawing near for three major
funding applications. Every five years the
MDC’s research programs have to resubmit
budgets and plans, a process involving fascinating
but hard discussions about how each
line of research is likely to develop over the
next few years, about which topics should be
pursued further, and which should be
dropped. It also involves a detailed scrutiny of
the progress of the research groups.
One of Walter’s goals as scientific director has
been to push the quality of science at the
MDC. That begins in his own lab. He has just
come from a meeting with the head of another
group who is trying to finish a manuscript
that they collaborated on. They have made an
important discovery: the appearance of a protein
in tumor cells may indicate whether they
will turn into metastases that wander
through the body, a process which often
transforms a benign tumor into a deadly one.
“It’s a good paper,” Walter says. “It might have
been better if I’d had more time to work on it...
But we have to submit it for publication
quickly so that we don’t get scooped.”
Other groups across the world work on similar
themes, collaborating and competing. So
much depends on an institute’s reputation.
It’s important to attract the best staff and the
24
brightest students, and they are drawn to the
places with the best track records.
��
Walter’s accent in English and German
gives away his Swiss origins. “When I
give talks, I don’t have to tell jokes,” he says. “I
just open my mouth and people laugh. They
hear my voice and they think of Emil, the
famous Swiss comedian.”
For a scientist, independence and a tenuretrack
position usually only come after study
and work in several laboratories that might lie
anywhere in the world. This means a move
every three or four years, repeatedly uprooting
the family. Walter’s history has been typically
nomadic. He worked as a school teacher in
Switzerland (confronting a class of 49 unruly
Elisabeth Kujawa-Schmeitzner

fifth-to-eighth graders) to put himself
through the university, then received his PhD
in biology from Zürich University. He joined a
laboratory in New York, came back to Basel,
left again for San Diego in the U.S., returned to
Zürich, then moved on to Tübingen and Essen.
He has been in Buch since 1993, right after the
MDC was founded, becoming scientific director
when Detlev Ganten was asked to become
head of the Charité in 2004.
So migrations are a personal theme as well as
a focus of his scientific work. He has always
been curious about what prompts cells to
leave their homes in tissues and organs and
set off on voyages through the body. This happens
all the time as embryos form: cells grow
in sheets that slide along each other, stretching
and folding to form organs; stem cells
migrate to the building sites of muscle, blood
vessels, and bone. By adulthood most cells
have settled into a sedentary lifestyle, embed-
ded in a tissue. Blood cells are the most familiar
exception, but there are many others. Skin
cells are born several layers below the surface
and move upwards to replace relatives that
die and are shed. Wounds are patched up by
cells that move in from the edges. Muscle
injuries suffered on the football field may heal
when the body calls up generic stem cells that
travel to the site of the injury and specialize to
become new muscle.
Cell migrations are also of intense interest
because of their role in cancer. Solid, benign
tumors can often be removed before they disrupt
the functions of vital organs, but all too
often they spawn metastatic cells which
escape and start to grow again in different
parts of the body. A surgeon may not be able
to remove such cells because they have
already escaped the tumor, or they may look
just like healthy cells. If even one is left
behind, the cancer may reappear and spread.
25
Part one: The live of nomads
A metastatic cell can’t be distinguished with
the eye, but its unusual behavior means that
its chemistry must be different than that of
other cells. Identifying any unique molecules
it contains might give a clue about why the
cell sets off on a dangerous path, and might
even reveal how it could be stopped.
But that’s a long-term prospect. The field of
molecular medicine straddles a frustrating

gap between knowledge and cures, Walter
says. Knowing that something is broken doesn’t
mean you can fix it. But it’s where you have
to start.
��
In his book The Songlines, world traveler
Bruce Chatwin recounts his experiences
among the nomadic natives of Australia. The
indigenous people of Australia believed that
the world was born in the Dreamtime, a period
in which formless spirit-beings assumed
the shapes of plants, animals and human
beings. Each feature of today’s landscape has
a mythical and spiritual meaning which is
recorded in stories and songs. Children may
be born anywhere during the wanderings of
the family, and the birthplace takes a special
significance in their lives. Each child learns the
traditions associated with that place.
Part one: The live of nomads
26
Everyone’s collection is unique because new
stories are added and old ones are reshaped
by time and the telling.
A cell also arises from a vaguer form, and its
identity is defined by the landscape in which
it is born. By tasting a particular neighborhood’s
molecules, it learns what to become.
The tasting is done by proteins called receptors
which float on the surface of the cell.
Each type of cell has a unique combination of

eceptors which sensitize it to different flavors.
And each environment tastes different
because it contains a unique set of stimuli,
called ligands. Combinations of receptors and
ligands give the cell its identity.
Tasting is the first step in a process that tells
a cell what to become and how to behave.
Receptors send information into the cell via
signaling pathways. They work a bit like the
telephone list that my children’s teachers use
to get information to parents. The teacher
calls one parent, who calls me, and then I call
the next parent on the list. Each bit of news
reaches the same parents in the same order.
In the cell information is passed from a receptor
(teacher) to proteins (parents) using
chemistry rather than telephone calls. When a
signal triggers the same proteins in the same
order it is called a pathway. Just as there are
many receptors, there are many pathways.
27 Part one: The live of nomads
Signaling pathways cope with the same types
of confusing situations that sometimes happen
with telephone lists. One teacher may
need to send out information to several classes.
(Some receptors can trigger multiple pathways.)
I have three children, so I am on three
telephone lists. (One protein may belong to
several pathways.) Whom I call depends on
who called me. (Depending on where a signal
comes from, a protein may react in different

ways.) If I can’t get in touch with the next parent,
the chain may be interrupted (a broken or
missing protein may block the signal). And as
a child may use the list to make prank phone
calls, viruses often use receptors to talk their
way into cells.
Eventually the information makes its way to
the genes in the cell nucleus, like customer
orders arriving at the kitchen of a restaurant.
DNA contains the recipes for other molecules
needed by the cell (RNAs and proteins).
Information from signaling pathways switches
on some genes and switches off others.
This changes the set of molecules present in a
cell, the way the arrival of new customers
changes the food on restaurant tables.
Tracking those orders – identifying what
receptor sends a certain signal, what proteins
pass it along, and what genes are affected by
the information – is the subject of a great deal
of today’s biological research.
Signaling pathways are involved in nearly
everything that happens in the cell – good
and bad. They give positional information
that helps stem cells develop, and they set
the cell’s clock that tells it when to divide. But
pathways can also break down. A protein may
go missing, or it may get stuck in a “transmitting”
mode where it broadcasts signals all the
time, even when it hasn’t received the call
from the receptor. If it tells the cell to divide
all the time, rather than taking breaks, the
result may be cancer.
��
In adult animals, sedentary cells sometimes
get a signal to run sprints, such as when a
wound needs to be healed. “The skin is the
body’s first line of defense and its integrity is
crucial,” Walter says. “Cuts need to be sealed
off fast to prevent bleeding and infections.
Cells move into the wound area and build
new skin, after which things have to return to
28
normal. So we thought this might be a good
system to use to look for the factors that
switch migrations on and off.”
In 2007 Walter’s group made some important
discoveries about the signaling pathway that
controls these effects. The story involved
some old friends (molecules that Walter has
been studying for over 15 years) and an even
closer friend – his wife. Carmen Birchmeier is
a coordinator of the MDC’s neuroscience program
and runs a large lab with wide-ranging
research interests. Sometimes the work of the
two Birchmeier labs overlaps.
Both Carmen and Walter have been studying
a signaling pathway that begins with a protein
called Hepatocyte growth factor/scatter
factor, or HGF/SF for short. The name is cumbersome
but revealing. A growth factor is
something that stimulates cells to divide (in
this case, liver cells called hepatocytes), and a
scatter factor prompts them to move. The scientist
who discovers a new molecule gets to
name it, usually picking something that
reflects its function in an organism. (Often
having a bit of fun in the process – there are
genes called sonic hedgehog, lunatic fringe,
teashirt, dreadlocks, rolling stones, and Swiss
cheese.) A lot of molecules have double or
hyphenated names because different scientists
start working on the same gene in different
organisms, and by the time people realize
they are talking about the same thing, the
two names have become entrenched in the
literature.
Carmen started working with HGF/SF to learn
how it influences the development of tissues
in the embryo, particularly the liver and muscles.
But when she developed strains of mice
to study its activity, she found several other
systems of the body in which it plays a role.
Walter has mainly been interested in aspects
relevant to the behavior of cells. Carmen,
studying large-scale processes at work in
whole animals, was mostly thought of by her
colleagues as a developmental biologist and
Walter was more focused on cell biology. In

the meantime it has become much easier to
slide from one scale of things to another, so
the fields (and a few of the people working in
them) are undergoing a sort of marriage.
It has been known for about 20 years that
HGF/SF prompts cells to divide and move.
When the protein is fed to laboratory cultures
of epithelial cells (which make up the skin and
the surfaces of many organs), they start to
replicate. Tumor cells raised in a dish in the lab
release each other and crawl away.
In the body HGF/SF is tasted by a receptor
called Met. Years of work by the labs of
Carmen and Walter have revealed some of the
molecules that transmit the information
from Met on to genes. Carmen has linked this
signaling pathway to several important
events in the body. “The correct development
of the liver and muscle depends on Met signals,”
she says. “They are needed to build
these tissues in the first place and to rebuild
them when they have become damaged.”
Usually HGF/SF and Met are made by cells in
different places, which is one way to keep the
pathway under control – the signal and its
receiver only meet up when conditions are
right. But cells in the heart, kidneys, lungs, and
skin begin producing high levels of both proteins
when there has been an injury. As if,
Walter says, you work to improve your
chances of transmitting a radio signal by
increasing the power of the transmission and
installing a larger antenna. The times and
places where the pathway was active made
him think that the proteins might be a master
control switch for healing. They might be con-
29 Part one: The live of nomads
trolling when cells divided and when they
moved into the wound area. Walter knew of a
way to find out, but first he needed a new
type of mouse.
��
Carmen’s group makes extensive use of
the MDC’s animal house, a state-of-theart
center for the care and study of mice, rats,
rabbits, fish, and a few more exotic organisms.
Animals have been key figures in biological
and medical progress and their role is increasing.
With the discovery that genes caused some
diseases and contributed to others, scientists
needed a way to study their effects over an
organism’s entire lifetime. A great deal had
already been learned by studying mutations

that occurred naturally in laboratory organisms
such as the fly. But this meant waiting
for random events to change random genes.
Experiments could only be conducted on
mutations that had already been discovered;
there was no way to choose a gene that you
were interested in and change it.
In the 1930s the situation began to change
thanks in part to events on the Berlin-Buch
campus. The American Hermann Muller had
discovered that X-rays and chemical agents
caused mutations by damaging genes. The
techniques still didn’t permit scientists to
interfere with a particular molecule, but they
provided plenty of new animals to study.
Muller was immersed in his X-ray studies
when he arrived in Berlin-Buch in 1932 to work
with Nikolay Timofeeff-Ressovsky. At that time
the campus was home to the Brain Research
Institute of the Kaiser Wilhelm Society, headed
by Oskar Vogt. He was half of another scientific
couple: his wife Cécile was an excellent
researcher and wrote or coauthored many of
his scientific papers.
It was a period of strong ties between
research in Berlin and the Soviet Union, which
was emerging from a phase of cultural and
scientific isolationism. Vogt had spent two
years in Moscow helping the Russians set up a
Brain Institute – whose main occupation was
to study Lenin’s brain. Vogt trained scientists
in specialized techniques of brain dissection,
worked with the brain in Moscow, and evaluated
many of his findings in Buch.
In Moscow Vogt had met Timofeeff-Ressovsky
and invited him to Buch to head the
Department of Genetics. When political difficulties
forced Vogt to leave Buch in the mid-
1930s, this department was transformed into
a separate institute devoted to Genetics and
Biophysics and Timofeeff-Ressovsky was
named its director.
Hermann Muller’s visit to Buch coincided with
another encounter that Timofeeff-Ressovsky
described many years later:
Part one: The live of nomads
30
In the beginning of the 1930s I became
a friend of Max Delbrück and, so to
say, involved him in our work. He was
a pure theoretical physicist, the pupil of
Max Born and Niels Bohr. Actually I
won him over to theoretical biology.
Now he is a major virologist and theoretical
biologist in America, a Nobel
Prize winner, and on the whole a very
remarkable person. At that time he was
a young man, and as are all great theorists,
was somewhat insolent, but this is
excusable. We also treated him
insolently, and he acquired our manners
very quickly and became quite an
acceptable young man. Henceforth he
was invited to our Buch group.
Timofeeff-Ressovsky and Delbrück collaborated
on one of the most important papers
about genes written in the first half of the
20th century. Called the “Green Pamphlet”
because it was reprinted in a green cover, the
document assembled what was known about
genes and made some suggestions about
how to find them. Twenty years before James
Watson and Francis Crick solved the riddle of
what genes were made of, the scientists of
Berlin gave an extremely accurate description
of how genes would have to be organized to
behave the way they do. The Green Pamphlet
drew a lot of physicists to the question; eventually
their study of DNA using X-rays gave
Watson and Crick information that they had
to have to understand the molecule.
Within a few years of the discovery of the
genetic code, scientists found proteins in bacteria
and other types of cells that could cut
DNA from chromosomes and paste them
back in again. The molecules were adapted so
that they worked the same way in other
species. Now instead of breeding huge numbers
of animals and waiting for mutations to
happen by chance, scientists could pick a gene

and remove it (knocking it out), transplant it
between species, or change it in other ways.
This was a huge step in the study of genetic
diseases. If a human disease was the result of
a mutation in a gene, researchers could
remove the mouse version of the gene and
see whether the mouse developed a similar
problem. The same approach is being taken
with cancer and other health problems that
are linked to defects in genes.
Genetic engineering can now be carried out in
a wide range of organisms, and scientists in
Buch and elsewhere are continuing to the
tools so that they can be used in new animals.
The unique features of other species can provide
valuable insights into human problems.
Carmen’s work with mice has provided
insights into neurodegeneration, muscle diseases,
and other health problems. A tiny
worm called planarium, studied in Nikolaus
Rajewsky’s lab, may help us understand how
wounds heal – or don’t – in humans. A blind
and deaf creature from East Africa, the naked
mole rat, is helping Gary Lewin understand
how our nervous system detects and transmits
sensations of touch and pain.
But those stories come later.
��
When Carmen’s lab knocked out the
gene for Met, the embryos experienced
fatal problems. The liver failed to develop
and many other tissues did not form in the
normal way; the mice died before birth. Such
animals couldn’t provide insights into Met’s
role in later stages of development or
tissue regeneration, so the lab had to
use another method of shutting
down genes called a conditional
knockout. The technique
These images show how cells respond to a skin wound. In the upper row, cells have functioning c-Met;
they reproduce quickly and begin moving into the wound area (right side). After 48 hours, a large
number of cells have moved into the area. In the lower row, most of the cells do not have c-Met. Cells
respond much more slowly, and only those with c-Met move into the region of the wound.
was developed by the laboratory of Klaus
Rajewsky (more on him in the story “Playing
the piano of planaria”), then at the University
of Cologne. The method involves attaching a
gene to a switch that allows scientists to shut
it down at precise times and places in the animal.
This usually requires mating two strains
of mice – one outfitted with the switch, and
another with a trigger that can shut it down
only in muscle, the skin, or another specific
tissue.
Carmen’s lab developed an animal with a
switchable form of Met and began using it to
investigate regeneration. The same mouse
could be used to investigate the skin, so
Walter looked for someone to carry out the
project. One of the group’s PhD students,
Jolanta Chmielowiec, had been doing similar
work with another strain of mouse but hadn’t
gotten anywhere. She had the skills necessary
to work on Met.
I met Jolanta as she was packing up to leave
for a new position at the Harvard Medical
31
Part one: The live of nomads
School. She is a blond, thoughtful woman
who often pauses to think before she speaks.
Making the decision to leave was hard, she
said. “The projects are going well now – finally.
It wasn’t always that way. Now that things
are moving it would be easy to stay, but the
longer you do, the harder it gets to move. I’m
not ready to settle down in one place yet.” She
admitted she was a little nervous about
Harvard. “But working with the two groups of
Walter and Carmen Birchmeier has been the
best preparation you could get.”
Jolanta arrived from Poland several years ago
after finishing her master’s degree on the
topic of cancer. Her work had involved comparing
the DNA of lifetime smokers with nonsmokers.
The topic was interesting, she
admits, but it was all carried out in the test
tube and the computer; she wanted to work
with animals. Buch offered that chance.
Walter proposed two projects, and she decided
to try to develop a strain of mouse with a
conditional knockout of a gene called Gab1.

But two years later the work had stalled, and
she desperately needed results for her PhD
dissertation. So it was a good time to take on
a new project.
Her first task was get to mice with a version of
Met that could be shut down only in the skin.
Jolanta did this by mating Carmen’s mice with
another strain, developed earlier in Walter’s
lab, whose skin cells (called keratinocytes) contained
a trigger that would shut down the
gene. A test of the newborn mice showed that
some had inherited both components. At first
Jolanta hoped the hard work was now finished
– but that didn’t turn out to be the case.
“You always hope things will go easy,” Jolanta
says. “I looked at the mice at various stages of
development – before birth and at other
stages of their lives – hoping to find something
wrong with the skin. But there was
nothing there. It’s the worst result you can
have, to do months and months of work only
to find out that your gene doesn’t do what
you expected it to. Maybe Met didn’t have any
effect on the skin at all.”
Time pressure was on to get results – she didn’t
want to be a PhD student forever. She still
had to look at wound healing, so she turned
to the scratch test, where the scientist makes
a small wound through several layers of the
mouse’s skin.
She closely followed what happened to the
injury over several days. “In both normal and
mutant mice, the wounds healed,” Jolanta
says. “I thought, ‘Here we go again,’ another
case where we wouldn’t find anything. That
was a really rough time. But in fact there was
a small difference between the two types of
animals. Mice with no Met in the skin healed
more slowly.”
The findings didn’t seem very promising, but
she had a discussion with Walter in which
they toyed with the idea of writing a paper
based on them anyway. Instead, they decided
to take a closer look at the wounds under the
microscope to try to find the cause of slow
Part one: The live of nomads
32
healing in the mutant mice. It meant going
back to Carmen’s lab for help with the
microscopy; that turned out to be a good
decision.
The scientists removed small patches of skin
from the scratch site using a laser. Jolanta’s
first discovery was that fewer cells were moving
into the wound. Next she took samples
from the cells to see if there were any obvious
differences. After that experiment, thing suddenly
made sense.
“In a conditional knockout, you rarely are able
to remove 100 percent of a molecule,” she
says. “A few cells slip through and produce it.
In this case, a few of the skin cells in the
mouse were still making Met. Those were the
ones that migrated into the region; they were
the only one to participate in healing. Cells
without the gene didn’t – either they were
unable to hear the signal to move, or they
couldn’t respond to it.”
Again with the help of Carmen’s lab, she
attached fluorescent markers to Met, which
let her track the cells that produced it as they
divided and moved. Some of the wound repair
comes from cells at the very edge of the
wound, but they need reinforcements. A distress
signal is heard by keratinocytes farther
away which divide quickly and then migrate.
This process is also influenced by the loss of
Met. The mouse has fewer cells that are able
to participate in healing, which is why the
process takes longer to get started. But once it
does, the few cells with Met divide more and
for a longer time. After a few days they make
up for what has been lost.
This shows that the body must have sophisticated
systems to start healing, to check on its
progress, and then shut it down once the
injury has been repaired. “Losing control over
any of these processes is potentially very dangerous,”
Walter says. “It’s an important motivation
for our work. Healing is a local event in
which cell division and migrations take place
in a highly controlled way. The hallmark of
cancer is a disruption of those two processes.

Hook-like cadherin proteins (left) allow cells to snag onto each other. When Walter’s
group interfered with cadherins using antibodies, or interfered with β-catenin signaling,
cells released each other and began to migrate.
In some cases that might be due to misbehavior
by the HGF/SF signal or the Met receptor
or other parts of the pathway. It may explain
why we find high levels of HGF/SF and Met –
a sign that the pathway is switched on – in
some types of tumors as well as wounds.”
If so, he says, the system might be especially
likely to break down at sites of injuries. Walter
recalls the hypothesis of the famous Berliner
Rudolf Virchow, who long ago observed a connection
between cancer and inflammations
in his patients. Jolanta’s project provides a
new link. “In a chronic situation where there is
a long-term wound or injury, the system has
to be switched on and off many times,” Walter
says. “Cells might become confused by the
signals, or they might forget how to switch
the pathway off again.”
��
Most of Walter’s work has been devoted
to another pathway, one which starts
with a molecule called Wnt. Receiving a Wnt
signal sets off a chain of interactions that
change gene activity and the cell’s behavior. In
the meantime the pathway, which involves
over 100 molecules, has been recognized as
one of the most ancient and important signaling
systems in animals.
The paths of Walter and Wnt first crossed 25
years ago, by accident, when he was working
at the Max Planck Institute in Tübingen. He
was using cell cultures to try to understand
why cells sometimes broke free of their tissues
to become nomads.
“We were using cultures of MDCK cells,” he
says. “These come from the surface layer of a
dog kidney. In cell cultures they form an
epithelium, a skin-like layer, which is one of
the things that makes them interesting in all
Jeanne Mammen: Damengespräch (ca. 1949) oil on cardboard
33 Part one: The live of nomads
kinds of experiments. We were trying to identify
proteins on their surfaces that tied them
to their neighbors.”
One way to find them might be to block some
of those proteins and see what happened, he
says. You might be able to do that with antibodies,
which work by gluing themselves to
proteins. In doing so, they might interfere
with the connections between cells. To get
antibodies against the proteins on MDCK
cells, Walter and his collegues injected the
cells into mice. The mouse immune system
recognized that the cells were foreign and
began producing antibodies against molecules
on their surfaces. Walter and his colleagues
harvested the antibodies, purified
them, and began putting them into the cell
cultures one at a time. If they found the right
one, he hoped, it would make the MDCK cells
lose their grip on each other. That’s what happened
– one antibody broke the connections.

The cells changed shape and begin crawling
around.
The antibody was binding to a surface molecule
called E-cadherin. “If you’ve ever seen a
microscope picture of Velcro, that’s a bit what
cadherins look like,” Walter says. “Each of these
proteins is a long fiber with a loop on the end.
They stretch away from the cell and become
entangled with identical cadherins on a neighboring
cell. There are lots of different types, Ncadherins,
T-cadherins... each can only hook
onto an identical kind on the neighbor. It’s one
of the things that make specific types of cells
stick onto each other without gluing themselves
to every other tissue in the body.”
The antibodies were somehow stopping cadherins
on one side from snagging those on
the other. It was an exciting finding. What was
happening in cell cultures mimicked something
that occurred all the time in embryos –
and in cancer. The group had transformed
sedentary cells into nomads. Maybe E-cadherin
and the same mechanism were responsible
for this switch in behavior in animals,
too.
One way to find out would be to check cells
from epithelial cancers, taken from patients,
so postdoc Jürgen Behrens and other members
of Walter’s lab went out and bought 20
to 30 cell lines. The lab grew the cells in cul-
Part one: The live of nomads
34
tures and used their antibody to look for Ecadherin.
“Some of the tumors grew in sheets,
like epithelia, and others were mobile,” Walter
says. “There was a direct correspondance. If
there was E-cadherin on the surface, they
formed an epithelium. If there wasn’t, they
crawled around.” If the scientists treated the
epithelial type with the antibody, the “skins”
broke up and the cells began to invade surrounding
regions.
There were two explanations. “Either the antibodies
were directly ‘gumming up’ the hooks
somehow,” Walter says, “or they were sending
a signal into the cell that changed its architecture
and shape. Or maybe both mechanisms
Manipulating Wnt signaling in cultures of MDCK cells causes them to leave their sheet-like structure (upper left) and begin crawling over each other (lower right).
Zwei Figuren, graphisch (ca. 1951) oil on cardboard
Krankes Kind (ca. 1946/7) oil on cardboard

Kleiner Kopf (ca. 1945)
were at work. We couldn’t immediately tell.
But we were also excited about the method –
we had shown that you could use antibodies
to study and manipulate the behavior of cancer
cells in cell cultures.”
Soon Walter and his lab had answered the
question. Their antibody was indeed interfering
with E-cadherin, but also with one of the
cell’s signaling systems. The switch between
adhesion and migration was really being controlled
from inside the cell, just below the
membrane. Each E-cadherin has a tail that
passes through the membrane to the cell
interior. There it is affixed to other molecules,
including a protein called β-catenin. If this
molecule is released there is no anchor, and
cells can no longer hold on to each other.
“With Jürgen and Jörg Huelsken, a new PhD
student, we did all kinds of experiments
mutating β-catenin and interfering with the
signals it received,” Walter says. “It was acting
as a switchboard for cell adhesion and migrations
in cancer, and we were beginning to look
at its role in migrations in the embryo.
Needless to say we were ecstatic.”
As it turned out, they were only holding the
tail of the tiger.
��
Asongline, in the world of the Aboriginal
Australians, is a chain of stories that are
sung to retrace the routes traveled by ancestral
gods as they created the world. Bruce
Chatwin called them “the labyrinth of invisible
pathways which meander all over
Australia.” The elements of the songs are
places, and by stringing stories together and
singing them in a certain order, a tribe can
navigate across the entire continent.
Important geographical features are intersections
in many of these routes, and they figure
in a lot of different songs.
β-catenin was a prominent feature in the
songlines of cellular information pathways,
but this fact wasn’t immediately obvious.
“Originally we were thinking only in terms of
the adhesion between cells,” Walter says. “We
were thinking of it as a structural element, an
anchor or a staple, holding onto things.”
But more labs had become interested in the
molecule. On the one hand that was good;
35 Part one: The live of nomads
information about β-catenin was coming in
quickly. On the other hand it meant more
pressure to keep from being scooped. In 1993
Walter was sitting down with Jörg, trying to
write a paper about the latest experiments,
when they heard about an important new
discovery. The labs of three scientists in the
USA – Kenneth Kinzler, Bert Vogelstein, and
Paul Polakis – had learned that β-catenin
docked onto another protein called APC.
That was interesting because APC had been
identified as a tumor suppressor gene. The
healthy forms of such molecules keep tumors
in check, but if they undergo mutations, the
cell loses control of an important process and
the result is a tumor. Aberrant forms of tumor
suppressors are often found in the tissues of
cancer patients.
“Jörg told me, ‘This is the best thing that could
have happened to us,’” Walter says. “Tumor
suppression was a new concept at the time.
We knew of mutations in other molecules
that caused cancer, but they were usually
thought of as tumor promoters – the defects
made them active; they were overstimulating

cells, sending them the wrong signals at the
wrong time. Tumor suppressors were more
like security guards, who needed to be around
to keep the status quo.”
Scientists knew that mutations in APC could
cause cancer, but since APC’s functions in
healthy cells were unknown, the information
didn’t mean very much. That began to change
through new findings, which Walter was
keeping his eye on as he moved his group to
Berlin-Buch. Jürgen had been looking for additional
binding partners for β-catenin using a
method called a yeast two-hybrid screen
(described in “The pleasures and powers of
green tea”), and came up with another protein
called Lef. Hans Clevers’ group at the
University of Utrecht discovered that
β-catenin also docked onto Tcf, a related
protein.
Lef belongs to a family of molecules which are
transcription factors – they control whether
genes are active or not. Transcription factors
are powerful proteins that affect important
aspects of cells’ behavior by changing the set
of genes that are active and thus the chemistry
of the cell. They are the final links in signaling
pathways, transmitting information to
the cell’s executive decision-making machinery
in the nucleus. When Jürgen examined
how the two molecules fit together, he discovered
that Lef alone couldn’t activate genes
– to do so it needed help from a particular
module of the β-catenin molecule, a “gene
activation domain.”
“In retrospect this shouldn’t have been that
much of a surprise,” Walter says. “We knew
that there was a form of β-catenin found in
the fruit fly, and we knew that it also entered
the nucleus. But we didn’t know what it was
doing there. Suddenly we had connected it to
transcription factors. Such factors often have
to work in groups to activate genes, so it
immediately suggested a completely new
function for β-catenin. We had been focusing
on the cell adhesion side of β-catenin. It was
like being at a magic show and watching the
wrong hand. Now we knew we had to start
looking at the other hand, too.”
This completely changed the status of βcatenin,
Walter says. “It had gone from a sort
of cellular ‘staple’ to an important gene-activating
device,” he says. “It was now an important
link in cell signaling pathways.”
Scientists had already identified the main signal
that controlled whether β-catenin headed
for the nucleus. That originated with a small
protein called Wnt, a signal known to control
developmental processes in embryos. “Things
kept getting more exciting,” Walter says. “The
molecule we were working on was now implicated
in three processes: cell adhesion, cancer,
and development.”
Studies in the fruit fly and other organisms had
shown that disrupting Wnt signals could have
serious effects on embryonic development. But
experiments kept turning up new kinds of
Wnts – the human genome alone contains 19
closely related forms of the proteins. Why so
many? What did they do in the body’s tissues?
How were the signals related to cancer?
Most Wnt signals influence β-catenin on their
way to genes, so the protein gave Walter a
good tool to find answers. That has been his
major preoccupation over the last 25 years.
��
Discoveries started to come fast. Jürgen’s
fishing expedition had turned up
another binding partner for β-catenin, a protein
called conductin. By docking onto both
APC and β-catenin, it played a major role in
the stability of β-catenin, which is controlled
by Wnt signals.
“By the end of the 1990s we had helped
establish that how much β-catenin is free in
the cell is central to Wnt’s effects,” Walter says.
“Labs everywhere were turning up new binding
partners for β-catenin, and we were all
working hard to figure out what these different
combinations were doing – what genes
were being activated, and what that meant
Part one: The live of nomads
36
for the embryo. Sometimes those effects were
pretty startling, as you can see when the system
is perturbed.”
He isn’t exaggerating – too much Wnt in the
wrong place, for example, causes a mouse or
frog embryo to develop two heads. With too
little signaling, there is no head at all. That
made sense because it was becoming increasingly
clear that locations and amounts of
Wnt/β-catenin signaling were helping to set
up the major directional signals in the early
animal body. One of the earliest things that
happens in the early embryo is a decision
about which end is the head and which is the
tail, which side is the front and which the
back. Wnt signals are like surveyors, laying
down the coordinates.
Context was everything, Walter says. “It wasn’t
as if everything sends β-catenin into the
nucleus. In fact, frequently just the opposite
happens. If it links up to Axin and APC and
another protein called GSK, then the cell
destroys it.
“What we had understood by 1998 was that
Wnt signals prevent it from being linked to
these molecules. That means the Wnt signal
stops β-catenin from being destroyed. Then it
can enter the nucleus and team up with Tcf
and Lef.”
Tcf/Lef normally sit on DNA and blocks the
use of nearby genes. Since some of those
genes trigger cycles of cell division, they need
to be shut off most of the time – especially in
differentiated cells that shouldn’t divide any
more. But if β-catenin gets to Tcf, the brake is
released and the genes get switched on.
“If that happens at the right time, when the
cell should divide, fine,” Walter says. “If it happens
at other times, you can get a tumor. So a
Wnt signal at the wrong time can cause cancer
by preventing the destruction of βcatenin.
And this model explains some other
causes of cancer. Anything that stops βcatenin
from binding to APC, actin, or GSK can
allow it to survive and do damage.”

Fontaine de Vaucluse (ca. 1955) oil on cardboard
��
Wnt and β-catenin are found in all animals,
in all sorts of tissues. “What we
started looking at as a simple clamp for cell
adhesion has turned out to be an all-purpose
tool for building bodies,” Walter says.
Over the past decade the lab has shown how,
along with a handful of other powerful signals,
Wnt directs the fates of cells and sculpts
tissues. “There are five or six really major pathways
like Wnt that nature uses over and over,
in different animals and tissues. The body produces
a variety of cells, tissues and organs
because of subtle differences in when pathways
get switched on in different places, and
they are activated in different sequences.”
β-catenin and Wnt are so crucial to the
growth and structure of the body that their
roles could only be understood through the
development of conditional knock-out animals.
With the help of Carmen, Walter’s lab
made animals with a switch attached to βcatenin.
They mated the animals with other
strains containing triggers that became
active in specific tissues and cell types. This
has given them the tools to begin a systematic
study of Wnt signaling throughout the
body.
In each tissue the system works differently,
but ten years of these studies have given
Walter a feeling for the special nature of Wnt
signaling. “It is usually not the pathway that
gets switched on first in a major building project,”
he says. “It doesn’t, for example, trigger
early cells to become precursors for all types
of blood or muscle; it doesn’t start the construction
of a limb. Instead, it controls finer
decisions. In the skin, for example, we found
that it tells stem cells to develop into hair fol-
37 Part one: The live of nomads
licles, but not the skin itself. It controls the
development of two types of cells – out of a
dozen – that develop into cells of the brain
and spine.”
Wnt steps in when the major foundations of a
tissue have been laid. The lab’s recent study of
heart development provides a good example.
The signals that shape the heart have been
tricky to unravel. It’s the first organ to form in
vertebrates, and it can only take shape if different
types of cells migrate to the right
places in a coordinated way. Wnt plays a role
in that process, Walter says, but there have
been conflicting accounts of its role. “In cell
cultures from frog and chick embryos, Wnt
activity has to be very low in order for cells to
begin to form the heart. But in mice, Wnt
needs to be active – if you block it, the heart
doesn’t form. It seemed like a contradiction –
how could we reconcile these two stories?”

Within the first days of an embryo’s life, cells
grow and separate into three distinct layers
that then go on to differentiate into specific
organs. The heart arises from the middle layer,
called the mesoderm. Cells in various areas of
this tissue receive different signals from their
neighbors that tell them how to specialize.
Some of them get instructions to become the
heart. Those cells divide into two regions
called the first heart field and second heart
field, and now they are ready to receive their
marching instructions.
The next stages happen relatively quickly, and
Walter sketches a map of the embryo on a
piece of paper. Cells from the first heart field
move into the chest area. Tagged with a fluorescent
marker and seen from the front, they
form a crescent, like a large horseshoe. Cells
from the second heart field move into the
area and line the inside of the shoe. When
everything is in place, the crescent starts to
fold inwards, as if the horseshoe has started
to melt. When the sides come together they
form a tube like the long, skinny balloon that
clowns mold into animals at carnivals. What
happens next resembles balloon art, too –
cells in different parts grow at different rates,
which makes the tube bulge. Since there isn’t
much space to grow it begins to bend, as if
the balloon has been threaded into a small
Part one: The live of nomads
38
bottle and then inflated. These patterns of
growth and bending create the chambers of
the heart, the way the clown might tie his balloon
into a bow.
Wnt may be involved in migrations and the
shaping, Walter says, but there were those
conflicting results from earlier studies... And
the picture is made more confusing by contributions
from other signaling molecules such
as the bone morphogenetic proteins, or Bmps.
“As you might guess from their name, these
molecules are mostly known for their role in
the development of bone and cartilege,” he
says. “But they also help build the front-to-

ack structure of the body of the early embryo,
and at least in the chicken they are needed for
the heart. So Alexandra Klaus, one of our PhD
students, tried to make sense of all this.”
She discovered that Bmp signals were active
in both heart fields. The pathway from Wnt
through β-catenin, on the other hand, was
only switched on the second. It was interesting
to find a difference between the two tissues;
it suggested that Bmp might be necessary
for the first heart field. Alexandra developed
a knockout strain of mouse in which
Bmp was switched off only in future heart
cells. As the embryos formed she investigated
the structure of their hearts. She discovered
that first heart field never formed the crescent.
But a few second heart field cells moved
into the chest to form a thin rim.
At what stage were things going wrong? “The
mice made progenitor cells that were supposed
to become the heart,” Walter says, “but
they didn’t stay around to finish their development.
We took this to mean that that the
Bmp signal was crucial to finish their ‘programming’
in the first heart field.”
But the second heart field could still form, at
least partially, without it. Since Wnt was
active in this region, Alexandra developed a
second mouse strain in which β-catenin was
shut down in the heart fields. This had a quite
different effect: cells were able to form the
simple heart tube, but the heart didn’t form a
loop. It was as if the clown had forgotten how
to twist and shape his balloon.
In mouse number three, Alexandra took the
opposite approach: she raised the volume of
the Wnt/β-catenin signal, turning it on in all
the future heart cells – even in the first heart
field where it normally isn’t active. This made
things go wrong fast; the basic heart tube
never formed at all. When Alexandra looked at
the two heart fields she discovered that cells
were straying off course and clustering in the
wrong places.
Cells not only have to specialize – they also
have to get to the right places. Both processes
are dependent on signals. All that was needed
to block the development of the heart was to
activate one of these signals in the wrong
place, or to silence it where it needed to be
heard. By watching what happened in these
different conditions, Alexandra showed that
Bmp was overseeing the behavior of future
heart cells at an early stage, in the first field.
Wnt and ββ-catenin were active in the second
field, stepping up later to sculpt the tube and
form heart structures.
��
The formation of the heart shows how in
the embryo, cycles of growth and migra-
39 Part one: The live of nomads
tions are a normal state of affairs. Cells continually
leave the structures where they are
born and move off to build new organs. As
they do so they specialize. Eventually most
cells settle down, leading to a long period of
adulthood for humans and most other animal
species. An inappropriate signal can lead to
disease by changing their developmental programs.
If the cells get so confused that they
leave their tissues and become nomads, the
result may be metastasis.
That often leads to disaster – but perhaps not
always. The earliest animals didn’t have
sophisticated organs or any genetic plan for
building them. Far back in evolution, the intricate
structures that serve our bodies must
have started off as deviations from the blueprint,
like tumors. If a signal were delayed, or
accidentally switched on a second time, the
shapes of tissues would be changed and new
structures would arise.
How did organs arise in the first place? If a
change in a gene caused cells to migrate and
build an odd new tissue, and if it was even
slightly useful, it could be passed on and
refined by natural selection. That wouldn’t
have to happen very often to have a huge
impact on evolution.
It’s an odd way to think of a process that normally
disrupts a life and may even lead to
death. But the work of Walter and his colleagues
support this view by showing that
the same mechanisms needed for growth and
migrations in embryos are also involved in
cancer. Maybe none of the tumors that have
appeared in humans – so far – have done anything
useful for their victims. On the other
hand, despite all the controls that make sure
the body gets built right, tumors still arise.
The system shouldn’t be so perfect that it
eliminates all freedom. It may be evolution’s
way of trying out new building projects from
time to time, a way of saying that the story is
not over yet, that change is still possible.

The archeology
of blood
Iwish there were a photograph of that moment in March, 1879, when
two Germans stood on the plains of western coastal Turkey, staring
up at a mound of dirt and rubble. One of them, the archeologist
Heinrich Schliemann, was having trouble convincing the world that this
unassuming hill held the remains of Troy. How could it contain the magnificent
city immortalized by Homer? Most scholars at the time had
become convinced that the ancient Greek poet had never existed and
the war described in the Iliad was a fantasy. Most of Schliemann’s contemporaries
regarded him as a rich, obsessed, and misguided amateur.
If so, he was a lucky one. His archeological digs in Mycanae and Troy had
turned up magnificent golden artifacts that confirmed, he claimed, the
fabulous stories he had read as a child.
The second man was a true amateur when it came to archeology, but
was held in high regard by his peers in his chosen profession. Rudolf
Virchow was recognized as one of Berlin’s best physicians and
researchers. His 20-year-old theory that “every cell arises from a preexisting
cell” and his explanation for embolisms (blood clots) had worked
their way into the textbooks. Yet Virchow found other conflicts to keep
him busy. After witnessing an outbreak of typhus in Silesia, he had
Part one: The live of nomads
40
Achim Leutz

ecome convinced that poverty and unclean
living conditions were major causes of disease.
He pushed Berlin to build a modern
sewage system, then became Germany’s
most forceful advocate of social medicine and
a founder of the German liberal party.
Eventually this made him such a thorn in the
side of the government that Chancellor Otto
Bismarck once challenged him to a duel.
(Virchow’s response was to laugh.)
Virchow’s interest in archeology had prompted
him to found the German Society for
Anthropology, Ethnology and Prehistory. (One
of its concerns was the study of the “Aryan
41 Part one: The live of nomads
race” – Virchow concluded that no such thing
existed, and that the rising trend toward
“Nordic mysticism” was nonsense). The position
put him into contact with Schliemann,
with whom he had exchanged letters for
years. Troy was the first time he took part in
one of the archeologist’s digs. Upon returning

Heinrich Schliemann
Part one: The live of nomads
42
Sophia
Schliemann
wearing
“Priam’s
treasure”
home he engaged in Herculean efforts to get
the city of Berlin to buy a large part of
Schliemann’s collection and put it on display.
Some of the archeologist’s greatest finds
remained in the city until 1945, when the
Russian Army removed them from a bunker
under the Berlin Zoo. Their whereabouts were
unknown until 1993, when they turned up in
the Pushkin Museum in Moscow.
The importance of the encounter at Troy is
symbolic. Mankind “discovered” history in the
19th century – in both the natural and social
sciences. James Hutton and Charles Lyell, two
Scottish geologists, had shown that the Earth
was thousands of times older than people
had believed – perhaps millions. Such a long
period, Charles Darwin and Alfred Russel
Wallace proposed, would be enough for natural
processes to produce life from inanimate
substances and spin off all of the Earth’s life
forms.
While evolution created a history for the
species, Virchow provided one for the individual.
The idea that every cell arose from another
cell dramatically changed the way scientists
thought of embryos. Each structure in
the body also had a history, one that could be
traced back to a single fertilized egg. Each

stage of development was built on the last,
the way that Troy was rebuilt at least nine
times on the foundations of earlier settlements.
While changing historians’ view of ancient
Greece, Schliemann made another inadvertent
contribution to the modern concept of
history. He published an autobiography which
is a confusing mixture of mix of facts, halftruths,
and probably downright lies. When it
fell into the hands of Sigmund Freud, it influenced
the young doctor’s views of the development
of an individual. A person’s psyche
arises as a mixture of facts and myth, much as
the true history of Troy is hidden under thousands
of years of art and story-telling.
Resolving the conflicts in a person’s life
requires a sort of archeology.
��
Virchow’s “archeology of the embryo” has
been extended by molecular biologists
to the life histories of cells, and this is the central
interest of Achim Leutz, researcher at the
MDC. Achim has been studying how generic
stem cells in the bone marrow specialize to
become many types of blood cells.
“A cell’s identity is determined by its pattern
of active and inactive genes,” Achim says.
“That’s a pattern spread across the whole
genome and it may involve hundreds of molecules.”
Activating a gene involves hooking it up to a
very complex machine made of many molecules.
This machine reads the base pairs on
one strand of the double helix and transcribes
the information into a similar molecule, called
RNA. Often the process begins when a key
protein, a transcription factor, binds to DNA or
does something else to it – for example, it
may remove another protein that blocks the
way. The change provides a site for the transcription
machinery to dock on and do its job.
Achim’s lab and many others have intensively
studied particular transcription factors
because they are so powerful. By switching on
Activating β-catenin at the wrong time in blood cell progenitors can cause them to develop in unusual
and dangerous ways. This sample, taken from the bone marrow, reveals unusually high numbers of
immature myeloid precursors.
one or more important genes, these proteins
can activate programs that cause cells to specialize,
divide, or die. Removing a key factor
from an animal’s genome may prevent particular
types of cells, tissues, and even organs
from developing. If a gene suffers mutations
or become defective, the result is often disease
and sometimes cancer or leukemia.
“The body maintains some hematopoietic
stem cells which are able to produce all types
of blood,” Achim says. “They reproduce, at a
slow rate. Usually when one of these cells
divides it makes one copy that is an exact
duplicate – it will become another stem cell –
and another that will first divide many times
but eventually specialize.”
43 Part one: The live of nomads
This second-generation cell is a precursor cell,
equipped with a new genetic program (a pattern
of active and quiet genes). It reproduces
much faster, generating offspring that usually
specialize quickly. The last generation is the
largest, and it is made up of cells that are fully
committed to becoming a particular type,
such as red blood cells or the T and B cells that
play an important role in the immune system.
Once they reach this phase, they stop reproducing.
When they wear out (red blood cells
have a lifespan of about 120 days, for example),
they are replaced by new precursors.
Transcription factors play a role in moving the
cells from one phase of their lives to the next.
For example, the proteins SCL and Myb are
needed at early phases. Later they are joined

Christiane Calliess
by combinations of factors such as C/EBP proteins
to make cells take the last step.
“Mutations that affect these molecules usually
have very bad effects,” Achim says. “If a transcription
factor doesn’t work properly, it may
stimulate the cell to keep reproducing longer
than it should, instead of specializing. Or it
may just block the process of differentiation,
which leads to the same thing, because the
cell can get stuck in a mode where it keeps
dividing. This is how a lot of kinds of
leukemias start.”
So taking a developmental step requires the
coordinated activity of a network of many
genes. “We’ve learned a lot about how the cell
controls the production of single molecules,”
Achim says. “But we’re much further behind in
seeing how it regulates complex networks.”
A cell’s identity is determined by the pattern
of genes it uses and those it doesn’t. While
each cell in the body carries a complete copy
of the genome, no cell uses it all to create proteins.
Some proteins are made only in the
brain; others are unique to blood cells.
Patterns begin to form early in life. The first
few cells in an embryo are nearly identical, but
they quickly take on individual characteristics.
Some are programmed to become skin, others
parts of the nervous system, others internal
organs. Usually once a cell starts down a particular
road, there is no going back.
“We understand these processes quite a bit
differently than we did a decade ago,” Achim
says. “The pattern that cells inherit is a whole
network. It’s not a case-by-case situation, a lot
of individual decisions about whether single
genes should be on or off.”
That network, he says, consists of combinations
of molecules that work together to coordinate
various processes, such as shutting
down cell division on the one hand while
pushing cells to develop and to mature. Not
only do cells need to keep producing the key
Part one: The live of nomads
44
transcription factors that control what’s
going on – they also have to manage the signaling
pathways that alert the factors to
what they should be doing. Often those pathways
do other things as well; understanding
blood development has caused Achim to
cross paths with several other groups, including
Walter Birchmeier and Claus Scheidereit
(“A hub in the city of the cell”). Walter’s
favorite pathway begins with a signal from a
molecule called Wnt, which is then passed to
a molecule called β-catenin. This pathway
plays a role in the development of limbs and
other tissues – and blood. The meaning of its
signals also depends on the context.
So the cell’s guidance of development is not
like people checking into a hotel one-by-one,
picking up their keys, and going into their
rooms. It’s more like a huge hotel where there
are several big conferences going on at the
same time. The people attending the same
conference all have the same schedule, and so

they enter and leave their rooms at about the
same time. They have a different program
than people attending another conference,
and a lot of coordination has to go on behind
the scenes to ensure that there are enough
rooms for the talks, the equipment gets to the
right place, and the restaurants can handle
the crowds.
There’s also a parallel to archeology: other
archeologists called Schliemann a treasure
hunter, a digger of “rabbit holes.” During his
first excavations he chose a promising spot
and dug straight down, rather than carefully
exposing and cataloguing the artifacts of
each layer before moving to the next. In
archeology, everything is context. For Achim,
the same thing is true of blood.
“Large developmental patterns are maintained
as the cells divide – they aren’t reset;
they don’t start from zero when a new cell is
born. That means the cell carries a sort of history
and memory through the process of dif-
ferentiation, and our research is devoted to
discovering where it lies and how it works.”
��
The DNA in a cell’s nucleus doesn’t look
much like the neat double helix drawn
by James Watson and Francis Crick in 1953. You
can’t magnify the image enough to see how
they intertwine. Even if you could, there are so
many proteins attached to DNA that for a
long time, most of the world’s best chemists
believed that the genetic code was made of
proteins.
Even under the most powerful electron microscope,
DNA appears as such a thin thread that
you almost have to imagine it is there. During
most of the cell cycle it is loose and spread
through the nucleus like a ball of yarn the cat
has been playing with. Just before cell division
it undergoes a process called condensation,
which packs it into huge, tight bundles – the
chromosomes.
45 Part one: The live of nomads
Achim shows me an electron microscope
image of the cell nucleus. In one the DNA
thread is loose. It looks like a thin string on
which a lot of beads have been strung. They
act like tiny spools, Achim says, and they are
called nucleosomes. Each one is made of eight
proteins called histones. This structure helps in
the process of packing when it comes time for
the DNA to condense. It also plays a very important
role in gene activation and silencing.
On the average, one nucleosome is found at
intervals of about every 200 base pairs along
the DNA strand. It wraps 1.65 times around
each of these beads – that takes 146 base
pairs. The remaining 50 or so base pairs form
a linker region that separates one nucleosome
from the next.
Histones, the building blocks of nucleosomes,
have long tails that stick out of the nucleosome.
Other proteins can attach themselves
here, or stick groups of atoms onto parts of
the tail so that it behaves differently. One

effect is to slide the nucleosome to a new
position. That may be important in activating
a gene. The molecular machines that read
information in genes to transcribe it into RNA
often have a hard time getting to DNA that is
wrapped around a nucleosome. By sliding the
spool to a new position, proteins can clear the
way for the transcription machinery.
Two chemical tags that are placed on the tails
of histone proteins often have strong effects
on gene activity. Often if a tail is tagged with
methyl groups, a nearby gene is silenced.
Another type of tag, acetyl, often activates the
gene. Sometimes tags are applied in a “blanket”
way to large stretches of DNA. That
would be one way of coordinating a lot of
nearby genes to behave the same way. It
might also work as a memory device.
Activating or silencing blocks of genes is one
way to control several things at the time,
Achim says, but the story is more complex.
The codes and patterns that tell cells what to
become are subtle. A methyl tag doesn’t
always turn a gene off; sometimes it has the
opposite effect, and the same histone tail may
even contain both types of tags. This makes it
very hard, Achim says, to decipher the code
and the interplay between the many factors
that control each gene.
By whatever means genetic programs are
coordinated, transcription factors must play
an essential role. So Achim’s group has been
looking into the links between these proteins
and histones. Recently they have received help
from an unlikely source: a virus.
��
Part one: The live of nomads
46
IIn the early 20th century Peyton Rous, a scientist
at the Rockefeller Institute for
Medical Research in New York, made a startling
discovery: he proved that a type of cancer
could be transmitted from one animal to
another by a virus.
From the reaction of his colleagues, you would
think he had claimed to discover Troy. No one
had trouble with the idea that a virus might
cause a disease – the microbe theory of Louis
Pasteur and the Berlin researcher Robert Koch
were known to every schoolchild. But cancer
was different. Rudolf Virchow had said that
cancer arose spontaneously from defects in
cells, and that was the accepted dogma.
Rous was working with chickens, which commonly
suffered from a type of cancer called a
sarcoma. These tumors grow in tissues such

as bone and muscle. After watching sarcomas
spread like infections through rural farmyards,
he began searching for the agent that was
transmitting the disease. He removed a sarcoma
from the leg of a chicken, prepared and filtered
it until there were no more cells, and then
injected it into a healthy bird. Transplantees
developed the same types of tumors, and by
repeating the process, Rous could pass them on
to new chickens in the same way.
Because he had carefully filtered the extracts
before injecting them into a new animal, Rous
hypothesized that the agent had to be something
smaller than a cell – a virus. Even after
Rous tried everything he could think of to disprove
his own hypothesis and failed, his colleagues
remained skeptical. He tried to find
cancer-causing viruses in mice and failed,
eventually becoming frustrated and abandoning
the line of work.
Two decades later, however, a colleague asked
for his help in the study of warts that arose in
mice because of a virus; the warts frequently
developed into cancer. The idea of cancercausing
viruses gradually became accepted
and finally in 1966, a few years before his
death, Rous was presented with the Nobel
Prize in Physiology or Medicine.
“It took molecular biology to build the bridge
between viruses and cancer,” Achim says.
“Viruses sometimes capture genes from cells.
Eventually those molecules mutate. When
they come back to infect a new person or animal,
they act differently. But the cell may not
be able to tell the difference; it may interpret
the molecule as one of its own.”
47 Part one: The live of nomads
Many viruses that cause cancer do so by interfering
with transcription factors or molecules
that interact with them. That was the case
with a bird virus called AMV. It once kidnaped
the gene for the transcription factor Myb
from an ancient host, added the molecule to
its own viral genome, and carried it along in a
mutated form as it invaded new hosts. When
the infected cells start making proteins from
this viral form of the gene (known as v-Myb),
it interferes with the cell’s own version
(c-Myb).
“Remember that Myb is one of the important
transcription factors that step in early in
blood development,” Achim says. “Later it
combines with other proteins to help them
differentiate. It has to work perfectly to do
both those jobs. v-Myb does fine at telling

A chart of blood cell development.
The progenitor of all types of blood cells, hematopoeitic stem cells, divide to create copies of themselves
so that the body will always have blood stem cells on hand. Some of the new cells begin to specialize
into ever-more specific types, passing through intermediate stages of development along the way. The
fully differentiated cells at the bottom of the chart no longer divide.
cells to reproduce, but it fails at the differentiation
step. This creates partially differentiated
cells that keep dividing, and that means
leukemia.”
What happens in such cells, he says, is a failure
to coordinate the programs that guide
division and differentiation. Maybe comparing
the two types of Myb would show what
was going wrong, and that might reveal
something about the coordination machinery.
There were more reasons to think that Myb
was crucial to decisions about the fates of
cells. “The viral form of the molecule isn’t the
only one that is linked to cancer,” Achim says.
“In many kinds of human leukemia, cells make
too much c-Myb. It makes you think that the
problem might be overactive Myb, maybe it
sends signals that tell the cell to divide too
often. On the other hand, high levels of Myb
alone aren’t enough to cause disease. A group
in Japan developed a strain of mouse that produces
too much of the molecule, and they
don’t develop tumors.”
But if you think of Myb as a coordinator of cell
division and specialization, he says, some of
these things make sense. He put his group to
work comparing the effects of v-Myb and c-
Myb on cells.
The basic differences between the two molecules
were well known, Achim says. “Myb is a
complex protein with many different modules.
The viral form of the protein is missing
segments at both ends, the head and tail.
Besides that, there are 11 spelling errors that
change the chemistry of the protein.” Nearly
20 years ago, when Achim had worked as a
staff scientist at the European Molecular
Biology Laboratory in Heidelberg, his colleagues
in Thomas Graf’s group showed that
three of these errors were the most important
Part one: The live of nomads
48
in causing the problems related to cell division
and differentiation. If they fixed the
spelling errors in these three places – by
changing the code of v-Myb – its ability to
cause leukemia dropped to almost nothing.
“With a transcription factor, your immediate
thought is that mutations might change the
way the molecule binds to DNA to activate a
gene,” Achim says. “But here that wasn’t the
case. Just a few years ago a Japanese laboratory
showed that when v-Myb docks onto DNA,
the part of the protein affected by the mutations
faces away from the DNA. So something
else is going on – we thought the change
might stop it from interacting with other proteins.
It was another hint that some sort of
coordinating activity was being disturbed. The
mutations might change how healthy Myb
docks onto other proteins, which stop it from
launching a cancer.”
Experiments by Achim and his colleagues previously
showed how some of the changes
affected a module of Myb that normally binds
to the transcription factor C/EBP. This interaction
is needed to stimulate the differentiation
of some types of blood cells.
But the spelling errors in v-Myb seem to have
another effect, Achim says. “One of Myb’s
modules is built like a machine we have seen
in many other proteins. In those molecules,
the module helps change the nucleosomes –
the spools that DNA is wrapped around. That
changes the accessibility of certain genes and
affects whether they can be activated.”
Xianming Mo, a postdoc in the lab, decided to
find out. He showed that the healthy form of
Myb can bind to the tail of one of the eight
histones in the spool, called H3. v-Myb can’t
duplicate that feat.
“Very interesting but not conclusive,” Achim
says. “Histones play a vital role in activating
genes. But finding out that things can bind
doesn’t mean that they do bind. Xianming
and Elisabeth Kowenz-Leutz had to prove that
they did. They started correcting the spelling

of v-Myb and discovered that changing any of
the three critical letters allowed it to bind to
the histone tail. And at the same time its cancer-causing
ability to block the activation of
the differentiation genes dropped off.”
On the other hand, he says, making v-Myb
more like c-Myb doesn’t seem to affect how it
binds to DNA. So the mutations don’t mean
that the two molecules target different genes
– it’s what happens afterwards that counts.
And what was that? “We knew that Myb could
bind to a protein called p300, which is an
‘acetyl-loader,’” Achim says. “Proteins like
p300 can tag histone tails with acetyl. That
chemical change often causes the activation
of a gene.”
The scientists combined different forms of
Myb and p300 in cells and checked whether
acetyls were loaded onto the tail. That didn’t
happen with the viral form of Myb. But they
could start the tagging by correcting any one
of the three crucial spelling errors.
So its presence on a gene allows healthy Myb
to coordinate a second process; that ability is
lost when the molecule is replaced by the viral
form of the molecule. The take-home message
from the study, Achim says, is that the
chemistry of the histone is a key player in
coordinating Myb’s effects on the division and
differentiation of blood cells.
And it’s the place to look if you want to know
how cancer arises.
��
The finding is theoretically interesting,
Achim says. It’s good archeology – the
difference between finding a bit of gold at
Troy and offering it for sale on E-bay, versus
learning where the metal in it was mined, and
what it says about the trade and culture of
the people who made it. Events shouldn’t be
seen through rabbit holes; good interpretations
require a context. Transcription factors
don’t act alone. What they mean depends on
the global condition of the cell’s DNA.
49

50
“And let’s not stop there,” Achim says. “It suggests
a new approach to cancers that involve
c-Myb. The problems might arise because H3
isn’t getting its dose of acetyl. That blocks the
activation of genes that tell the cells to stop
dividing and start differentiating. Those genes
aren’t active enough in leukemia cells. Well,
we could test that. We took leukemia cells in
laboratory cultures and fed them with a substance
that promotes more tagging of H3
tails with acetyl. Suddenly the key specialization
genes were switched on again and the
leukemia cells differentiated.”
So there may be medical applications. Both v-
Myb and c-Myb may behave the same way in
terms of binding to genes. The changes in
Myb that cause leukemia – even if there is no
virus involved – may be doing the same thing.
If that’s the case, Achim says, we need to
change the focus of therapy. Aberrant forms
of Myb may be sitting on the right genes,
ready to do their jobs. But if they can’t communicate
with the tail of histone 3, they can’t
help it get loaded with acetyl.
“Find another way to deliver that tag,” Achim
says, “and you may patch the problem with
Myb.”
��
One of the half-truths in Heinrich
Schliemann’s autobiography may be his
explanation for how he became interested in
the myth – and history – of Troy. Like so many
scientists, he claims to have been inspired by a
book. In his case, it was a world history for
children, illustrated with an engraving of Troy
and its massive stone walls.
Achim has one of those stories about a book.
“My father was a locksmith and my parents
didn’t much understand my interest in science,”
Achim says. “But I had an uncle who
was always giving me things. An old microscope,
a telescope. And one day he gave me an
old children’s book from 1938. Chemistry
experiments that work, by Hermann Römpp.
You could get everything you needed to do

the experiments from a normal pharmacy.
That began my experimental career, which
was mostly focused on producing bad smells
and blowing things up.”
From what I can tell after a little research, just
about every German scientist born after 1938
has at one time or another had his hands on
Chemistry experiments that work and cites it
as an inspiration to become a scientist. The
list includes Gerhard Ertl, winner of the 2007
Nobel Prize in Chemistry, and former director
of the Fritz Haber Institute of the Max Planck
Society here in Berlin.
So Achim knows the importance of a good
book and an education in science; maybe
that’s one reason he has spent so much effort
helping to build the MDC’s PhD program. As
Dean of the graduate school he coordinated
the program from its conception to the present
state: a student body of about 170,
increasingly oriented toward international
students. The MDC receives far more applications
from students than it can accept. That’s
unfortunate for the ones we have to turn
down, he says, but the end result it increases
the quality of the students and the program.
You can judge that for yourself by considering
the fact that the main investigators on many
of the stories in this book are PhD students.
Maybe we could collaborate, I say. “How about
a popular book: Experiments with histones
that work? Where you can get everything you
need from your back yard, and some tissue
samples taken from your brothers and sisters?”
He laughs.
“You mean to excite the next generation of
scientists?” he says. “You have to have the
explosions and the smells. If you can’t blow
anything up, and your mom doesn’t get mad
at you for smelling up the house, it won’t be
as inspiring to kids.”
51

Breakouts in the brain
If at any time during the first half of a Western film the bandit is
caught and locked up in a jail cell, a breakout is sure to follow. There
are a dozen variations on the theme. A wall may be blown up with dynamite.
A guard may be lured into the cell, or at least close enough for the
prisoner to snatch his keys. Or the sheriff might be paid off, becoming a
co-conspirator whose only job is to look away during the escape.
Researchers who study the brain are familiar with some of these scenarios
– a type of brain cancer called a glioma is a master breakout artist.
Part one: The live of nomads
52

Surgery almost never succeeds in completely removing these tumors
because cells escape all the time, and each one that leaves is like a
seed, able to give rise to a new tumor.
Many of the breakouts are witnessed by healthy brain cells that rush
to the site of a tumor, including some that you would think would act
as guards. Helmut Kettenmann and his laboratory at the MDC are trying
to discover their roles in containing tumors or helping them break
out. This work is part of a larger plan to uncover the functions of a
huge group of cells in the brain that have been underappreciated for
far too long. Most brain research has been focused on neurons. Their
smaller and much more numerous companions, a wide range of glial
cells, have been waiting patiently on the back burner.
Scientists got their first real look into the structure and functions of
brain cells in the late 19th century. The Italian researcher Camille Golgi
had developed a new method of staining tissues which, when applied
to the brain, revealed a intricate web of majestic, branching cells. Golgi
believed that they formed a continuous, fused network, like blood vessels.
One of his contemporaries, a Spaniard named Santiago Ramón y
Cajal, promoted a contrary view – neurons were single cells, separated
from their neighbors by tiny gaps called synapses. He made fantastically
detailed drawings of what he saw under the microscope and concluded
that neurons communicated by receiving impulses at their
bush-like dendrites, then passing them along their axons to other
cells.
Helmut knows quite a bit about Cajal – he has visited the scientist’s
historic lab in Zaragosa and followed some of his tracks through
Berlin. Cajal was a medical doctor with an indomitable passion for cell
biology, a field virtually unknown in Spain. He was a gymnast and an
artist – which helped when it came time to draw what he was seeing
under the microscope; his drawings are still used in textbooks to illustrate
the types and structures of nerve cells. Under the pseudonym
53 Part one: The live of nomads
Helmut Kettenman

Santiago Ramón y Cajal
and his drawings
Part one: The live of nomads
“Dr. Bacteria” he wrote science fiction stories but didn’t publish them
until 1905, fearing that his scientific work wouldn’t be taken seriously.
He also wrote novels, but later said they had been lost in Cuba, where
he spent two years as an army doctor.
A copy of the book of stories sits on the shelf behind Helmut’s head.
One describes an astronaut on Jupiter who encounters giant creatures,
enters their bloodstream, and watches battles between white
blood cells and parasites. Scientists in the stories are not only heroes.
“One concerns an old bacteriologist who suspects that his wife is having
an affair with one of his lab assistants,” Helmut says. “To prove that
they are meeting secretly in the lab, he rigs the couch with a seismograph.
When the results confirm his fears, he infects the assistant with
tuberculosis, who passes it along to his wife. The assistant dies, but at
the last minute the scientists administers an antidote to his wife.”
After nearly dying of malaria in Cuba, Cajal returned home and was
given a full professorship at the University of València in 1881. Later he
held professorships in Barcelona and Madrid. His vocation was also his
hobby; any extra money he made went toward the purchase of microscopes.
He learned of Golgi’s staining methods from a colleague in
Madrid and rushed home to try them out. He was a precise and meticulous
worker and his own improvements to the staining process
allowed him to see details of nerve structure that had been invisible to
Golgi. But he had few peers to appreciate his work in Spain, and trying
to get his research published was a frustrating experience. Finally he
decided to publish his own journal, the Trimonthly Review of Normal
and Pathological Histology, but he could only afford the printing and
postage costs of 60 copies, which he sent off to scientists around the
world. It was like mailing them into the void.
Cajal decided that the only solution was to head for one of the world
capitals of science – Berlin – and establish direct contact with other
scientists. He joined the German Anatomical Society and in 1889 managed
to find enough money to make a pilgrimage to their annual
meeting, held at the University of Berlin. At first his presence must
have seemed simply odd. Spain wasn’t known for its scientists, and he
didn’t speak a single word of German – he had to communicate in broken
French. And he had brought along his own microscope, a good
Zeiss model he was familiar with, not trusting the instruments that
might be available the conference. He set up his stand among all the
others, and then went through what every PhD student and postdoc
experiences at their first big conference: few people came by for a look.
Everyone seemed much more preoccupied with his own work.
But then some people did stop by to peer through his Zeiss, including
Albrecht von Kölliker, the “patriarch of German histology.” What
Kölliker saw made him stay, and that drew a larger crowd. For Caval it
was a life-changing moment. He wrote,
54

As to be expected, these savants, then world celebrities,
began their examination with more skepticism than curiosity.
Undoubtedly they expected a fiasco. However, when there had been
paraded before their eyes...a procession of irreproachable images of
the utmost clearness, the supercilious frowns disappeared. Finally,
the prejudice against the humble Spanish anatomist vanished and
warm and sincere congratulations burst forth.
Kölliker immediately took Cajal under his wing and whisked him away. “He took
me in a splendid carriage to the luxurious hotel where he was staying; entertained
me at dinner; presented me afterwards to the most important histologists
and embryologists of Germany, and, finally, made every effort to render
my sojourn in the Prussian capital agreeable.” Kölliker returned to his own lab
where he reproduced Cajal’s results, then began promoting the Spaniard’s reputation
in Europe. Almost overnight Cajal became a celebrity, receiving invitations
to conferences in London and even the United States – despite the fact
that the two countries had just been at war.
The biggest recognition, however, came in the form of the 1906 Nobel Prize in
Physiology or Medicine. Ironically, the prize was jointly awarded to Camillo
Golgi. Although by 1906 most of the scientific community had become convinced
that Cajal was right – neurons are separated from their neighbors by
synapses – Golgi remained stubborn and bitter. Instead of devoting his acceptance
speech to his own prize-winning research, he attempted to raise his old
theory from the dead.

��
For brain science, discovering neurons and their functions was like
emerging from the dark ages. Ever since, neurons have held the
spotlight of most research; they are given most credit as the architects
of thought and the storehouses of memory. But as important as they
are, they make up only ten percent of the cells in the brain. This figure
led to the urban legend that people only use ten percent of their
brains.
Helmut and his lab have devoted themselves to the other 90 percent.
Formerly regarded simply as assistants, or nuts and bolts that keep
things in place, glial cells are now being recognized for playing active
roles in nearly everything that happens in the brain. Including the
growth and spread of brain tumors.
Glial cells come in several types. There are shaggy oligodendrocytes,
which help carry out electrical maintenance. The long, cable-like axons
of neurons have to carry electrical charges along great distances, and
without some sort of insulation, the charge would quickly leak away.
Oligodendrocytes coat the axon in a sheath of fats and proteins
that act as insulators. Several diseases, including multiple
sclerosis, are caused by a loss of the sheath.
The functions of another type of glial cell, astrocytes, are
less clear. “We know they provide guidance during the
development of the brain,” Helmut says. “They have extensive
contacts with synapses and they respond to neuronal
activity; they can be stimulated by neurotransmitters. One
of our interests has been to understand their role in communication
between brain cells.”
Another type of cell, microglia, was described and
named by one of Santiago Cajal’s students, Pio del Río-
Hortegas, in 1919. A few years later he noticed that the
cells are drawn to sites of brain injury. This suggested
that they might have functions in the immune system.
The idea couldn’t be proven, however, until the late
1980s. Working at the University of Pennsylvania, William
Hickey and Hiromitsu Kumura found immune system molecules
called MHC proteins on the surfaces of microglia.
Microglia respond to tumors as well as injuries.“They
are attracted by small molecules released by the
tumors,” Helmut says. “Their response is so strong that
a typical glioma may consist of up to 30 percent
microglia.”
What hasn’t been clear is why they come –
whether they are fighting the tumor or somehow
helping it. Most of what is known about the cells
Part one: The live of nomads 56
comes from watching how microglia respond to disease. Their roles in
the healthy brain are less clear. Helmut says there is increasing evidence
that they constantly survey tissue and step in to repair minor
damage to blood vessels and to remove diseased neurons. In many situations,
such as strokes, microglia congregate at the site of damaged
tissue. A 2007 study carried out by Mélanie Lalancette-Hébert and
Jasna Kriz of the Université Laval in Quebec showed that strokes led to
significantly more damage in mice which lacked microglia than in
their normal counterparts. And when gerbils suffering from a strokelike
condition were given extra microglia, the cells homed in on the
damage site. These animals lost fewer brain cells than gerbils that had
not been treated. In these cases the role of microglia is clearly protective.
Their normal duties ensure health, but when faced with tumors or
another serious problem, microglia can aggrevate the problem.
Microglia help remove α-β, a protein fragment that accumulates
between cells and leads to the death of neurons in Alzheimer’s disease.
So their activation and protein-clearing activity has healthy
effects. But if there is an inflammation, damaged tissue releases molecules
that slow them down. So in Alzheimer’s disease
and other situations, such as gliomas, it might be best to
keep microglia away – or at least to control how they
respond to problems.
In 2005 Darko Markovic, a postdoc in Helmut’s group,
headed a study which showed that the number of
microglia in a glial tumor in mice corresponds to the
severity of the disease. More cells are a sign of a larger
tumor that is more likely to develop into metastases. This
made it very unlikely that the microglia were there to
fight the tumor. If so, they were doing a poor job. In fact,
the opposite might be the case. They might be helping
it along.
Helmut’s lab had developed a method of using thin
slices of brain tissue to watch the behavior of brain cells
and the growth of tumors under the microscope. One
discovery was that microglia appeared to be helping
tumors to grow and escape.
As this book was being written, Darko and his colleagues
had just found an important mechanism by which this
happened. The article they wrote is now in the mail to a
scientific journal, which means that everyone is hanging
on waiting to see if it will be accepted and published.
Journals have a strict rule that information in an article
can’t be made public before it appears in print, so
Helmut won’t tell me the name of the molecules
involved.

“Let’s call them X and Y,” he says. “What we discovered seemed to
revolve around the behavior of protein X. It’s an enzyme involved in the
escape of glioma cells, helping break down the ties between a cell and
its neighbors. That’s an essential step in releasing the cell from the
tumor so that it can migrate. Somehow this molecule was being activated
in tumor cells by the arrival of the microglia.”
Researchers at the University of Pittsburgh Cancer Institute in the US
had detected very high amounts of X and also Y, a molecule that it
interacts with, in tumors from glioma patients – levels far higher than
those found in healthy brain cells. X plays a crucial role in the activity
of its partner. Normally Y is attached to the membranes of cells in an
inactive form. X cuts off a part of the molecule and releases it. This
activates Y, which begins chewing at the protein fibers that link cells.
That was interesting, but in itself it didn’t say anything about why the
arrival of microglia stimulated tumors to grow or cells to escape.
Darko and his colleagues began trying to overhear the molecular conversations
that were taking place between the glioma and microglia.
First the scientists marked the cells so that they could tell the difference
between the two types. Their images suggested that most –
maybe all – of the X was being produced by microglia. Those that come
into contact with the tumor were producing much more of the protein
than microglia elsewhere in the brain. To be sure that this was the
case, the scientists returned to their brain slices, this time using tissue
whose microglia had been removed. They added tumor cells, which
grew and invaded the tissue. But this time they saw no rise in levels of X.
“All of this suggested that tumors might be causing microglia to produce
more of the molecule,” Helmut says, “so Darko did a number of
experiments to find out. He extracted the medium that had been used
to grow cancer cells and put it into cultures with microglia. This medium
contains molecules that have been secreted by the tumor cells.
After three to six hours, the microglia began producing much higher
amounts of X.”
Normally cells are told to start making X by a pathway of signaling
molecules – like the telephone lists described in the previous story. The
signals belong to a well-studied circuit called the MAPK pathway. Were
these also the molecules used by gliomas to talk to microglia? Darko
and his colleagues blocked the pathway with a drug called SB201290
and found that cells didn’t produce nearly as much X. An antibiotic
called minocycline, which interferes with MAPK signals, had even
stronger effects.
Microglia might be assisting in the breakout of tumor cells by bringing
along the tool that cut free Y on the tumor. Gliomas didn’t seem to
make this tool themselves; they didn’t produce much – if any – X. The
scientists wondered why not, so they transplanted the molecule into
tumor cells. Glioma cells which produced both molecules died. So to
activate Y and escape, they had to obtain the keys from microglia.
If escape was the topic of the dialogue between microglia and
gliomas, then X might make a good place to jam the conversation.
Once again using the brain slice model, Darko allowed cancer cells to
invade tissue whose cells couldn’t produce the protein. Without X,

tumors grew to only half the size. If the tissues also contained no
microglia, tumors only reached 16 percent of their normal size. “This
shows,” Helmut says, “that X is an important tumor-promoting factor
in the cells. But it isn’t the only one.”
Tumors could be slowed down by removing the protein entirely –
would blocking its activity do equally well? The scientists tried different
scenarios using the SB201290 drug and the antibiotic minocycline.
The antibiotic cut the growth of tumors by nearly half, specifically
because it interrupted the production of X in microglia. By itself
SB201290 had a smaller impact on tumors, but when used in tissues
without microglia, it cut their size by two-thirds.
“The results showed that minocycline might have a therapeutic
impact, so we administered it to mice with brain tumors,” Helmut says.
“Here we noticed a very significant impact. Over two weeks, tumors in
the treated mice grew to a size of about one cubic millimeter. In control
mice that had not received the antibiotic, they were more than
four times as large. Even when we started the treatment later, after the
tumor had already started to grow, we saw a significant reduction in
tumor size.”
Part one: The live of nomads
Minocycline may have additional tumor-fighting properties. A 2005
study by researchers at the University of British Columbia in Canada
showed that the antibiotic interferes with the activity of microglia in
animal models of Alzheimer’s disease and brain injury. “We don’t yet
completely understand how the substance works,” Helmut says.
“Besides blocking the MAPK signaling pathway, and stopping X, it has
other effects. It seems to interfere with some of the molecules secreted
by the tumor, which likely block signals that the microglia use to find it.”
��
Microglia are not the only cells that take notice of brain tumors
and respond. They are joined by neural precursor cells – a population
of cells that have not yet completely differentiated. For a long
time it was not known that adults had such cells. Most brain-building
activity occurs in the embryo and young animals and is lost in adults.
But a few regions of the brain protect and keep precursor cells through
adulthood. In some situations they are pulled from storage and triggered
to develop to take the place of injured neurons.
One of the difficulties in studying the development and functions of
these cells has been that they could not be kept alive and cultured in
the laboratory. Recently members of Gerd Kempermann’s lab and
Helmut’s group developed a new method for doing so. It’s a crucial step
in learning about the chemistry of the cells and their funcions.
“The fact that neuronal precursors target gliomas has made the
cells interesting from a therapeutic point of view,” Helmut says. “If
58
Neural precursor cells (green) migrate to the site of a tumor (dark
area in the center). The red dye helped Helmut and his colleagues
establish where in the brain they migrated from.

The Leiner room in the City of Konstanz is a "museum within a museum." It was established in 1870 and today has the status of a protected historical monument. The
closely arranged displays and hand-written descriptions reflect the museum style of the time.
their activity and development could be controlled, it might be
possible to use them to repair other types of damage. And they have a
second quality which has attracted interest – they are drawn to
gliomas.”
In 2000 this led Karen Aboody in Evan Snyder’s lab at the Harvard
Medical School to suggest that precursor cells might make good vehicles
to deliver gene therapies to tumors. The same year scientists at
the Istituto Nazionale Neurologico Besta in Milan modified the cells to
produce extra interleukin-4, a molecule that powerfully triggers
immune reactions. When they injected the cells into mice with
tumors, the animals survived longer. Interestingly, when unmodified
precursor cells were injected, the mice survived longer, too. So far the
studies had shown that cells injected into animals could move to the
tumor, but what about those produced by the animal’s own body? The
cells might have beneficial effects on their own. Helmut had several
members of his lab – Rainer Glass, Michael Synowitz, and Golo
Kronenberg – look more deeply into the relationship between gliomas
and precursor cells.
It was important to be able to tell the difference between cells attracted
to the tumor and any new ones that might arise from the tumor
itself – gliomas often arise as mutations in precursor cells. The lab of
Gerd Kempermann, a colleague in the MDC’s program on Functions
and Dysfunctions of the Nervous System, had developed a line of mice
whose precursor cells that could be identified because they produced
a protein with a green fluorescent tag. The tumor cells had been engineered
to emit a red fluorescent signal.
After two weeks, Rainer and his colleagues discovered that a large number
of green precursor cells had been drawn to the tumor. Not only had
they penetrated it – they had it surrounded, coating it in several layers
of cells. “In the border zone they were intermingled with cells from the
glioma, a few of which had escaped,” Helmut says. “Interestingly, most
of them were also accompanied by green precursor cells. So the precursors
are also strongly attracted to cancer cells outside the main mass of
the tumor.” This attraction was much stronger than to other types of
injuries, where migrations of precursor cells were also found, but at
lower levels and in a much looser structure.
59 Part one: The live of nomads

Steffi Siefert
If precursor cells somehow reduced the effects of tumors, it might
help explain why children rarely develop gliomas; most cases appear
when people reach their mid-fifties. While the brain retains precursor
cells its whole life, there aren’t as many with advancing age, and those
that are still around may not function as well. The study by Rainer and
his colleagues confirmed that the number of precursors attracted to
tumors in mice dropped sharply with age
In young mice the cells envelop the tumor, and their appearance
restricts its growth. In older mice, fewer cells are attracted, and the
gliomas grow more vigorously. To show that this was a direct effect of
the precursor cells, Rainer injected adult mice with a mixture of precursor
and tumor cells. The mice survived much longer than animals
that had not received the precursors. In fact, their prospects for survival
were just as good as those of much younger mice.
Another round of experiments helped to show why. Tumor cells reproduce
rapidly in cell culture, and few of the cells die. When grown alongside
precursor cells, four times as many suffered that fate.
The experiments of Helmut and his colleagues show that the body
takes notice of growing brain tumors; they attract the attention of
microglia, precursor cells, and many other types. Once they arrive on
the scene, relationships become complex. The precursor cells clearly
act as guards, whereas microglia might even be helping in their
escape. There are a lot of molecular conversations going on in the border
zone of the tumor. Keeping gliomas contained may not require listening
in on all of them – but it will certainly require defining the role
of those that are present, distinguishing those who are fighting the
tumor from those who are helping it.
��
The city of Konstanz, situated on the great lake in southwestern
Germany, is home to a unique museum within a museum. In 1870
Part one: The live of nomads 60
the pharmacist and city councilman Ludwig Leiner moved his enormous
collection of fossils, human artifacts, minerals, and stuffed animals
into an old house belonging to the butcher’s guild. Displays were
mounted on the wall or placed in glass display cases, waist-high. Each
object was labeled with a descriptive card, made out by hand in a perfect
cursive script.
Today one large room of the Rosgartenmuseum holds Leiner’s exhibits,
perfectly preserved – except for smudges on the glass where parents
have lifted their children to peer at dinosaur bones and trilobites. Now
that the site has been declared a World Heritage landmark, nothing
can be moved or loaned out. Entering the exhibit is like a stroll into the
19th-century mind, with its fascination for collecting and classifying
and its endless curiosity regarding the natural world. The museum
was established only 11 years after the publication of On the Origin of
Species, and it reveals how eagerly Darwin’s theories were embraced in
Germany by natural scientists and fanatical collectors such as Leiner.
Fossils and stones are enthusiastically dated as being tens or hundreds
of millions of years old.
I think Helmut Kettenmann would feel perfectly at home in the
Rosgartenmuseum. Visiting his office is like time travel (just ignore
the computer). Helmut has become one of the city’s preeminent historians
of science. His shelves are lined with leather-bound volumes,
many of which date to the 19th century; they give off the faint and lovely
smell of old books that have been well cared for. Most of them deal
with the brain, and many of them have a connection to the history of
Berlin. From the bottom shelf he produces an enormous folio with
reproductions of Ludwig Hoffman’s original architectural drawings for
the hospitals of Buch. As he talks about one of the diagrams, he seems
to be seeing it in his mind, the way it used to be.
The Berlin-Buch campus doesn’t have a Leiner room, but a small historic
laboratory has been recreated in the Oskar and Cecilia Vogt

house, complete with instruments that have been used at various
phases in the campus’ history. It’s one stop on the tour when the Long
Night of Science comes every year to campus, an event that attracts
thousands of visitors. Helmut has often been deeply involved in planning
the event and giving talks to the public. Sometimes when he
gives scientific talks he brings along some of his old books.
He also collects antique scientific instruments; several of his microscopes
are on display outside the lecture rooms of the Max Delbrück
Communications Center. And then there are antiques of a completely
different nature – I have heard that he drives in historic automobile
rallies. I ask him about it.
“Oh, that’s private,” he smiles.
As I pack my things to leave, I take a last look at the office. A hundred
years ago, every office would have looked this way. A book was a precious
thing, and the phases of your career could be read from the titles on your
shelves. Your professor gave you one and you would carry it along for the
rest of your life, from station to station of a scientific career. You would
keep it close at hand, where it could be consulted at any time.
I think of 60 copies of Santiago Ramón y Cajal’s small journal, spreading
slowly to the world by carriage and ship and train. I think of the
fact that Charles Darwin’s library held a copy of a scientific journal
which mentioned some experiments on heredity in peas carried out by
a monk named Gregor Mendel. In the 19th century, journals often
arrived with their pages uncut – they had to be sliced apart with a letter
opener to be read. The pages of Darwin’s copy had never been cut.
Mendel never found an Albrecht von Köllinger to launch him into
fame. Darwin might have played that role, but he never witnessed the
birth of genetics. He exchanged letters with Karl von Nägeli, whom
Mendel looked to for advice, but Nägeli never mentioned the monk’s
work.
In the 19th century every scientist had to know where his science came
from and see it as part of a much longer story. Things are not the same
today. Landmark discoveries are flying by much faster; there is so much
to learn about the science of this year, and maybe last year, that few
have time to go deeper.
This is another surprising thing about the campus: in contrast to so
many modern scientific institutes, history is omnipresent here, from
the buildings (which represent diverse architectural styles from most
decades of the 20th century), to the sculpted heads of people who have
defined the the campus, mounted outside the buildings: the Vogts,
Max Delbrück, and many others. There is a Holocaust memorial along
one of the wooded paths – a grim reminder of what can happen when
science becomes the slave of politics, or when it fails to impose ethical
boundaries on itself. The brains of victims of the Holocaust were
examined here in Buch. It is a good thing to walk by the statue from
time to time, to think, and remember.
I am standing in front of it when group of young PhD students walks
by on their way to lunch, heads bent, deep in discussions about their
work or the upcoming weekend. Today they take no notice of the
sculptures. But at some time during their stay here, I know they will
hear of this and the rest of the history of the campus, the story that
they are becoming a part of. Helmut Kettenmann, Heinz Bielka, Jens
Reich and a few others will see to it.
61 Part one: The live of nomads

Stem cells in a jar
Berlin, like most industrial cities, has become one of those places
where you can get almost any type of food at any time. In this culture
of instant grocery gratification, the art of canning food has almost
disappeared. Elena Vasyutina’s mother still knows how to preserve
fruits and vegetables by cooking them, sealing them in jars, then laying
up reserves to be doled out slowly over the winter months, but it’s
becoming a lost art. The best Elena can do, she admits, is to read the
expiration dates on packages.
Our bodies, on the other hand, remain experts at conserving, stockpiling,
and rationing cells, and Elena and her colleagues in Carmen
Birchmeier’s lab at the MDC are helping to reveal how this is accomplished.
“Some of our cells are more exposed and prone to injury than
others, and various types have different lifespans,” Elena says. “Shortlived
cells needed to be replaced. To solve this problem our bodies keep
reserves of cells which have specialized part of the way but which have
not taken the last developmental steps.”
Red blood cells, for example, survive only about 120 days and are then
replaced by fresh cells from the blood marrow. Those stockpiles are
Part one: The live of nomads
62

Carmen Birchmeier

already committed to becoming blood cells,
but there are several types, and the cells are
still flexible about exactly what they will
become. Other tissues draw on other kinds of
stem cells to regenerate themselves. The system
isn’t perfect; the body can’t replace all of
its cells, and the ability to heal decreases with
age. It may be that the reservoirs of intermediate
stem cells are used up, or maybe they
have forgotten the secrets of rejuvenation. It
would be useful to find all the types and learn
the genetic programs that cause them to specialize.
The answers might teach us how to
cope with injuries that the body doesn’t repair
very well because the stem cells are lacking, or
possibly because they have forgotten how to
make repairs.
��
Elena was born “in a small village in Russia
you’ve never heard of,” moved to the city
of Novosibirsk as a teenager, then came to
Germany seven years ago. She joined
Carmen’s lab at the MDC to do her doctoral
degree. When she arrived the group was
working on several projects – including molecules
involved in the development of the liver,
Alzheimer’s Disease, and the origins of
muscle.
Why such diverse projects? It turns out they
are related; the molecules that Carmen’s
group set out to investigate several years ago
have turned out to have diverse functions. “It’s
always been hard to do a gene knockout in an
animal,” Carmen says. “Even with all the
improvements that have been made, it usually
takes at least a year between your decision
to look at a particular gene and getting animals
that permit you to study its functions.
Then there can be a disaster – you may not
see anything odd about the animal at all.
After so much work I was always so glad to
find something that we pursued the leads
wherever they took us.
Elena says she wanted something “active”, so
she chose muscle. The muscles of our arms
and legs are born in ball-shaped clusters of
cells called somites which form in the early
embryo. Some of the cells in somites receive
signals that tell them to develop into
nomadic muscle. They travel a very long way,
along established routes, as if embarking on a
pilgrimage. Eventually they arrive at bumplike
structures that are the building sites for
future limbs. There they begin fusing to each
other to make long fibers.
Carmen knew that signals which passed
through Met (the wound-healing pathway of
the last chapter) were needed to send muscle
cells on their way. “Without these signals,
there was no migration,” she says. She shows
me a slide of an early mouse embryo, with
Met protein stained in purple. In the normal
mouse, the cells start off in a cluster and then
disperse. In the knockout mouse, they stay
clumped together. That made sense because
Met was known to make travelers out of
many different types of cells, but it raised
other questions. The cells have to travel an
amazing distance and find very specific tar-
Part one: The live of nomads
64
gets. How did the cells know what route to
take – what signs showed them the way? At
about the time Elena arrived in the lab,
Carmen had started following a promising
lead. Cells that later form muscle in the legs
produce a molecule called Lbx1. It is a transcription
factor: a protein that helps switch
genes on and off, and it is produced by these
particular cells as they specialize to become
muscle.
Carmen’s lab made that discovery when they
removed the gene from the mouse in hopes
of discovering its functions. “There was something
wrong with the mice, but we weren’t
sure what it was,” she says. One of her postdocs
carefully listed the tissues which needed
Met to develop properly. Some types of muscle
were on the list. That rang a bell; Carmen
remembered having heard of similar results
from work on another molecule. When the lab
began looking at muscles in different parts of
the embryo, she had an “Aha” moment.

“There were no muscles in the limbs,” she
says. “No cells were migrating there and so
the muscles never formed.” But cells were still
able to travel to other locations to build muscle.
Losing Lbx1 was sending only this particular
type of muscle astray.
“To know where to go, cells have to read cues
along the way, like following signs,” Elena
says. “Because cells take different routes, they
read different signs and respond to them in
individual ways. Without Lbx1 that wasn’t
happening for these particular cells, but we
didn’t know why. We didn’t know what signal
activated the molecule, or what pathway was
involved. That’s what I decided to try to find
out during my PhD work.”
A few years ago it would have been much
harder to find other molecules involved in the
Lbx1 pathway, but Elena could draw on some
new technology to start her search. Decoding
the genomes of mice, humans, and many
other organisms has given scientists a recipe
book for these species’ genes. Those recipes
have been turned into probes that can compare
gene activity in cells. The technology is
called DNA chips, or microarrays. It plays an
important role in the research of Norbert
Hübner’s laboratory at the MDC, and his
group offers a service that helps other labs
use it.
Norbert shows me one of the “chips”. It looks
like a normal glass microscope slide; there’s
nothing unusual about it until he holds it up
to the light. Then row after row of very faint
spots become visible – DNA printed onto the
slide. Each of the spots contains billions of
molecules that act as a probe for one gene.
The spots work as detectors because the
chemistry of a DNA molecule makes it able to
bind to a matching RNA – an RNA molecule
that has been made from that gene. Scientists
extract all the RNAs from cells, wash them
over the slide, and then check to see what has
been trapped. A single DNA chip like this one,
containing probes for all of a mouse’s genes,
can trap all the RNAs made by the cell. This
65

Diana Lenhard
66
reveals which genes are active and which are
silent.
The genes that helped muscle migrate and read
signs ought to be active in the cells that produced
Lbx1. Elena used DNA chips to compare
them to other types of cells and obtained a large
amount of data. A microarray experiment often
has a bewildering outcome, showing that dozens
or hundreds of genes behave differently even in
very similar cells. That was the case this time, and
the only way to find answers was to sift through
the data. Here, too, Norbert could help; his group
has become experts at distinguishing important
genes from false leads.
Many of the genes that had turned up in Elena’s
experiment had already been studied by other
labs for other reasons. One molecule on the list
was particularly intriguing, a protein called
CXCR4. Several laboratories, including that of
Martin Lipp at the MDC, had been working on the
protein because of its role in the immune system
– and in AIDS.
��
Viruses usually need the help of a conspirator
to invade cells – like an insider who provides
codes to the alarm system. That’s the case with
HIV, the virus that causes AIDS. It can can only
enter human cells that are outfitted with a receptor
protein called CD4. Since this molecule
appears mainly on cells of the immune system,
those are the first victims of the virus, and that
helps explain why AIDS causes a breakdown of
the body’s defenses against other diseases.
But when scientists mounted the receptor on
animal cells, the virus couldn’t enter. This probably
meant that it needed help from a second
human protein. In 1996 Edward Berger’s lab at
the National Institutes of Health in Maryland,
USA, discovered the identity of this partner –
CXCR4.
CXCR4 didn’t evolve in order to make things easy
for HIV. It’s an accident that the virus takes
advantage of the molecule, which has an important
function in healthy cells. In 1998 Reinhard
Förster, a PhD student in Martin Lipp’s lab,

showed that the CXCR4 receptor is produced,
used, and recycled by many types of immune
cells. These are the body’s hunters; they
migrate through tissues, recognize invaders,
and destroy them. CXCR4 is part of the guidance
and tracking system. It is attracted to
SDF1, a protein produced by other cells.
Elena had to go back and read up on the two
molecules. She discovered that SDF1 acts like a
powerful signaling beacon for cells with
CXCR4. “It’s used by the body in several different
ways. A lot of things go wrong in animals
that have a mutant form of either molecule.
Cells don’t migrate properly to form parts of
the brain, and the parents of egg and sperm
cells – which begin in one part of the body
and have to migrate a long way to the reproductive
organs – also get lost.”
The two proteins also seem to play a role in
cancer. CXCR4 has been spotted on many
types of tumor cells that undergo metastasis.
When these cells migrate, they behave a lot
like white blood cells, and they even seem to
use the same tracking system. Interactions
with SDF1 may explain where these cells go
and why they settle in particular places. Cells
that begin in tumors in the breast, for example,
typically settle in the lymph nodes, the
bone marrow, the lungs, and the liver. When
Albert Zlotnik’s lab at the DNAX Research
Institute in Palo Alto, California, blocked contact
between SDF1 and its receptor in mice,
breast cancer cells had a much harder time
taking up residence in the lymph nodes and
lungs.
Were muscle cells using the same system to
move to the limbs? “SDF1 is produced in the
area of the legs right as developing muscle
cells begin to migrate,” Elena says. “That’s
what you would expect to see in a beacon. A
good test would be to see what happened if
we moved the beacon somewhere else – so
we added SDF1 to locations that don’t normally
produce it.”
The decoy worked; migrating cells were pulled
off their normal route and attracted to the
artificial signal. So the two molecules were
acting as a directional signal for muscle, too.
The discovery gave Elena enough material for
her doctoral dissertation – and a paper in the
journal Genes and Development – but she still
had the feeling she was only touching the tip
of the iceberg when it came to understanding
the migration and development of muscle
cells. A lot of questions remained to be
answered. For example, how did cells interpret
multiple signals? Met was also needed to
prompt migrations. Did the two pathways
work together?
Carmen had been working on a protein called
Gab1, a molecule discovered in Walter’s lab in
1996. Gab1 receives the signal from Met and
passes it on to other molecules. Removing
Gab1 from mice also disturbed muscle migration
and development. When Elena studied
mice that lacked both CXCR4 and Gab1, she
found that cells no longer migrated to form
some types of muscle in the tongue. But
when only one of these molecules was missing,
the cells behaved properly. This showed
that there was some kind of connection
between the two pathways.
Once the cells arrive at their destinations,
they reproduce to create enough muscle. They
also develop further, changing their structure

Muscle tissue from normal mice (left) and animals lacking the molecule RBP-J (right).
Normal mice have muscle stem cells called satellite cells (red) but mutants do not.
so that they can fuse and make muscle fibers.
These processes are also controlled by signals,
and Carmen had been collecting evidence
that the Gab1 and Met pathways were
involved.
��
The switch-like behavior of stem cells
became the topic of Elena’s next project,
which she worked on with Diana Lenhard, a
PhD student who had recently joined
Carmen’s group. This time the theme wasn’t
migration but repair: how muscles recover
from injuries. And it involved a signaling system
that is used again and again, throughout
the body, over an entire lifetime.
Notch is a receptor protein like Met. It floats
like a fishing cork on the cell surface. Part of it
rises above the surface, where it can receive
signals; another part passes through the
membrane, and there is a tail which hangs
inside the cell. When the outer part comes in
contact with a signal on another cell, the protein
is cut. This frees the tail, which travels to
the cell nucleus and works with other mole-
cules to activate genes. Which genes it affects
– and what happens to cells – depends on its
partners.
A new Notch partner, a protein called RBP-J,
was discovered a few years ago by a Japanese
group. Usually RBP-J is bound to other molecules
that help it keep genes inactive. But
when Notch’s tail is freed, it bumps off these
other proteins. RBP-J can then link up with
other partners, and the new combinations
become gene activators.
“Several things got us interested in the relationship
between Notch and muscle,” Carmen
says. “First, Notch is very important in the
development of somites – remember, those
are the ball-shaped clusters of cells where
some types of muscle cells are born. Secondly,
as we age, our muscles heal poorly. That’s
partly due to changes in the protein which
activates Notch. This molecule is called Deltalike-1,
and it appears on the surface of neighboring
cells. When there has been an injury,
muscle cells produce Delta-like-1 and this triggers
Notch in their neighbors. But as we get
older, injured cells produce less and less of the
Part one: The live of nomads
68
protein. The signal is much weaker and so
there is less healing.”
Elena and Diana developed a new strain of
mouse that shut down RBP-J only in migrating
muscle cells. They did this by attaching
switches to the RBP-J gene in one type of
mouse and then adding a trigger to the Lbx1
gene, the gene that is only active in migrating
muscle cells, in another type. Mating the two
strains would give them mice in which Notch
could no longer use RBP-J to get signals to the
genes of migrating muscle cells.
But getting such offspring sometimes takes
months. “That’s a long time to wait, especially
if there is no guarantee that the animals will
shed light on the process you are interested
in,” Diana says. “It’s why Carmen likes us to be
working on two or three projects at a time; if
one of them doesn’t work out, you have something
to fall back on.”
Finally the litter was born, and some of the
mice pups had both elements needed to control
the gene. Elena and Diana looked carefully
at the mice for changes in muscle. Losing

RBP-J didn’t have any effect on cell migrations;
half-way through embryonic development,
the normal number of muscle stem cells had
traveled from the somites to the limbs. But
once they were in place, something unusual
began to happen. After one day there was a
sharp drop in the number of stem cells. At the
same time, the scientists found a higher number
of developed muscle cells.
“This seemed to mean that the stem cells
were developing earlier than usual,” Diana
says. “At first that might not seem like such a
big thing, but it soon started to cause problems
for the mice. A short time later the mice
without RBP-J had considerably less muscle in
the limbs, back, and diaphragm than normal
embryos.”
Just as canned foods have to be rationed to
last through the winter, stem cells need to be
rationed to build muscles and repair them
later in life. “Muscle keeps a stock of stem
cells, called satellite cells, in reserve,” Diana
says. “When we investigated the muscle in
our mutant mice, we found that these cells
were gone. The animals could no longer maintain
the cells; they specialized wildly, too
quickly, so they were all used up.”
RBP-J was acting as a “gene brake,” blocking
the final specialization of the cells so that the
animal keeps some reserves and form satellite
cells. Without the protein, the cells aren’t
around long enough to build proper muscles
in the embryo. This shows that the signal
passed along by the Notch protein is crucial in
setting aside a pool of cells that can be used
to build muscle in embryos and repair it
throughout the animal’s lifetime.
Work by other labs supports the idea that
RBP-J may act widely, in many types of tissue,
as a brake that sets aside supplies of stem
cells for future use. That’s probably the case in
the brain, where nerve cells produce Deltalike-1
as they mature. This activates Notch,
which triggers RBP-J. It docks onto genes and
prevents neurons from taking their final
developmental steps. Things slow down and
the nervous system preserves a population of
stem cells.
This system, too, wears down with age. After a
child has reached the age of two, the body has
a hard time making new neurons. The reason
may be similar to what happens in muscle:
there may be defects in the braking system. If
stem cells develop too quickly, the reserves are
used up. Then if the brain is damaged in a
stroke, Parkinson’s disease, or many other diseases
in which neurons die, they can’t be
replaced.
Could RBP-J be used to change that situation?
Knowing that brakes exist doesn’t mean you
can build them or install them in a car, Elena
says. But it’s crucial to know how they work.
Researchers hope to develop therapies in
which they transplant stem cells into damaged
brains, spinal cords, or muscle. The new
immigrants will need to understand the signals
that tell them how to behave, but they
may face the same problems as native cells.
The body may no longer produce the signals
needed to store cells. An alternative – if scientists
can learn how – might be to fit the wanderers
with their own brakes.

Capturing a firefly
IIt is an unusual place – a cool, empty room of muted colors. The walls
are covered with pale green tiles. In the middle of the floor is a hole,
a small crater, as if a tiny meteorite has struck. More likely a drain has
been dug open for repairs. Above the hole hangs an odd lamp with large
bulbs, mounted to the ceiling on a mechanical arm.
To me it is just a strange, abandoned space. But when Peter Schlag
stands in the middle of the room he sees the ghosts of the past. This is
the theater where, as Chief of Surgical Oncology of the Charité, he performed
thousands of operations. A few months ago the Robert Rössle
Cancer Clinic bustled with doctors and patients; now it is almost completely
deserted. Everything has been moved across the street into the
new HELIOS Clinic, a modern hospital complex that will soon host over
a thousand beds.
Peter snaps off the lights and leads the way through long, quiet hallways,
down a flight of stairs, into another corridor. Our footsteps echo
in the silence. He knocks on a white door and enters. A colleague sits at
a computer beside a blank television screen and a row of video cassettes.
“Films of operations,” Peter explains. He gestures at the other
equipment in the room. “This was a communications center. From here
70

Peter Schlag
we remotely directed operations all over the world, on oil rigs and other
hard-to-reach places.”
Down the hallway he tries another door; it’s locked and none of his keys
work. The locks have just been changed. A colleague produces a huge
ring of keys and tries one after another until the door opens. Inside we
find a table mounted with lasers and lenses, an elaborate microscope.
“We’re right below the operating room,” Peter says. “This installation is
why I decided to come here.” The idea, he explains, was to give the surgeon
an extra pair of eyes. The microscope would be able to illuminate
fluorescent molecules that had been attached to markers for deadly
cancer cells. Usually those remain invisible, indistinguishable from
healthy cells. Left in a patient, they may spawn tumors and metastases,
sometimes years later.
Peter dreams of rendering those cells visible. “At the moment, for many
types of tumors, it is extremely difficult to give a patient a prognosis
about whether his cancer will return,” he says. “A prerequisite to completely
removing a cancer is to see all the tumor cells, particularly
those that will metastasize. That would be a huge step forward.”
��
Asurgeon necessarily has a different perspective on “molecular medicine”
than other researchers on campus. Peter gets to see things
from both sides. His position itself is a bridge between the worlds of
clinical and basic research; he is professor at the Charité, the hospital of
Berlin’s university, and a researcher at the MDC. Several other scientists
have joint appointments at these two campus partners. Others split
positions at the MDC and the third research institute on the Buch campus,
the Leibniz Institute for Molecular Pharmacology (FMP).
Back in his office, Peter orders tea and begins to initiate me into a new
world view. “Molecular biology breaks bodies and bodily systems down
to their basics, that’s what it’s for,” he says. “It’s a set of techniques used
to understand fundamental questions. We need the techniques.
Knowledge is essential, but you can’t treat a patient only with knowledge.
A patient is a whole. You don’t do only oncology – you do diagnosis,
you do surgery, you do psychooncology.”
71 Part one: The live of nomads

The molecular sciences have a great potential to improve diagnostics,
he says. “Surgeons will always be needed, but as Theodor Billroth said,
‘Surgery is the worst of all kinds of treatment.’ Have you heard of
Billroth? He was born close to Berlin, in Bergen, lived and worked in
Vienna, where he developed a great number of new techniques – he’s
known as the father of modern general surgery. He was also an excellent
amateur pianist and violinist by the way, and one of the closest
friends of Johannes Brahms.”
We talk about history and music for awhile. Billroth studied in
Göttingen and Trieste, where he
learned dissection by studying the
nerves of an odd creature called the
torpedo fish. (One of the MDC’s scientists
has worked on the animal
and it reappears later in this book,
in the chapter called “the electrician’s
toolbox”.) Billroth received his
doctoral degree in Berlin and
worked in the university’s surgical
clinic for several years before moving
to Vienna. There he developed
successful new operative techniques
to cope with a wide range of
diseases – becoming the first to
remove a colorectal cancer.
“Billroth was a great surgeon, but
he very clearly recognized that surgery
is the front line of an undesirable war,” Peter says. “It’s a last resort,
it’s what you need if there’s been a car crash. It’s not really a solution
for cancer, which is harder than most other diseases. Cancer will never
be like tuberculosis, which is caused by a bacteria. Fighting tuberculosis
is hard, but there’s an identifiable enemy. Cancer is more heterogeneous;
we need to understand the pathways, the mechanisms, the preconditions.
We’re finding new drugs that influence tumor growth, and
in some cases it’s possible to block metastases for months. But as
promising and helpful as these discoveries are, they are not a real solution.
The only way we will learn to block metastases for years is to
intervene in the mechanisms that underlie them.”
Peter came to the MDC from Heidelberg 15 years ago and alongside his
duties in the clinic, has run a research lab the whole time. “Even during
surgery I have tried to see the ‘science’ aspect of cancer,” he says. “The
operating room is our research lab. Having this facility on campus has
given us unique opportunities. One of the major questions in cancer
research is what changes cells undergo when they migrate to new
locations in the body and establish new tumors there. Well, we could
collect tissue from tumors in different places in the same patient at
Part one: The live of nomads
different stages of the disease. We’ve had direct access to the samples
you need to answer those types of questions.”
The publications he has worked on reveal both aspects of his career.
Alternating with articles on surgical techniques (“Sentinel lymph node
biopsy for gastrointestinal cancers,” and “Image-guided surgery of liver
metastases by three-dimensional ultrasound-based optoelectronic
navigation”), he has worked on the basic processes that underlie
metastases, and reviews that combine the perspectives (“Clinical, biological,
and molecular aspects of metastasis in colorectal cancer”). The
more basic projects are carried out
by three groups in his department,
headed by Ulrike Stein, Wolfgang
Walther and Wolfgang Kemmner.
Conducting such a wide range of
projects requires interdisciplinary
expertise, so the groups include
doctors, basic researchers, computer
specialists, mathematicians, and
laser specialists.
One project has been the development
of new technology to help
surgeons recognize and cope with
dangerous cells. The group invented
a new “jet injection” device, a
method of dispersing DNA directly
into tissues. The idea for the technology
began during a trip that Ulrike and Wolfgang Walther made to
the United States over a dozen years ago. They spent two years at the
National Cancer Institute in Maryland, collaborating with Priscilla
Furth at the University of Maryland in Baltimore.
“Priscilla was very interested in gene transfer applications – ways of
delivering foreign DNA to cells,” Ulrike says. “The purpose is to get the
cells to absorb a new gene and use it to make proteins. There are several
reasons you might want to do this. If a person had a mutation that
caused a genetic disease, you could give them a healthy copy of the
gene. Another use would be to mark cancer cells so that they could be
seen and removed during surgery. Or you could feed tumor cells a ‘suicide
gene’ that would make them destroy themselves.”
Upon their return, Ulrike and Wolfgang began working on the apparatus,
a needle-less jet-injection device built by the company EMS
Medical in Switzerland. It recently passed phase one trials in humans,
in which a harmless marker gene was jet-injected into tumor tissues.
“We sat back and held our breaths,” Ulrike says. “There were a lot of reasons
why it might not have worked. But not only did the cells take up
the gene and use it to produce proteins – those molecules were func-
72

tional, and there were no side effects.” Soon the method will undergo
its next test at the Charité downtown, in an attempt to deliver a therapeutic
gene to cancer cells.
Meanwhile, the groups have been digging into the biology of metastases.
For most types of cancer, it is still impossible to tell the difference
between these dangerous nomads and benign tumor cells, and sometimes
even healthy cells. But their unique behavior must be guided by
a unique set of molecules. Identifying them, Peter says, is the first step
toward seeing them, and maybe finding ways to destroy them.
I think of summer evenings as children, when we spent hours catching
fireflies and putting them in jars, then carrying them around like
lanterns in the night. Even if a firefly was sitting still on a branch, you
couldn’t see one, or catch it, until it glowed. The same thing will probably
be true of metastases – they will have to be marked to be trapped,
studied, and conquered.
��
Ulrike attended the university in Halle when it still belonged to the
GDR. Only two universities offered degrees in biochemistry, and
there were places for only 40 students a year; excellent grades won her
one of those places.
Ulrike’s son is now studying at the Technical University of Berlin. She’s
a professor and she can’t help comparing her own university experiences
to those of her son. “Students have so many more opportunities
now,” she says. “They’re free to take any courses they want. Our time
was very strictly managed; there was no time to do anything extra. On
the other hand, there was much more in the way of social structure. We
were a small, tight-knit group; we did everything together – we helped
each other both academically and psychologically. Today things are
very competitive and individualistic and it isn’t always very healthy.
Students have to manage their own studies – they even have to take
care of a lot of organizational problems that the university used to
arrange for us.”
She is also in an ideal position to study the campus’ marriage of medicine
and fundamental research. She teaches biology courses for med-
73 Part one: The live of nomads

Ulrike Stein

ical students and always has some in her research group. They have
been particularly motivated and she has rarely had problems integrating
them. Ulrike admits that she has been a bit spoiled, because as a
rule, the physicians of the Robert Rössle Clinic have been very interested
in research.
But she acknowledges that there have been a few culture clashes –
sometimes amusing ones. “Writing a dissertation for a PhD in molecular
biology often takes years,” she says. “MDs sometimes finish their
theses in a semester. Once an MD student applied to the group and I
asked him how much time he
would be able to devote to his dissertation.
He thought about it a
moment and said, ‘I can work on it
every Thursday afternoon, for a
year.’” She laughs. “It’s a running
joke in our lab now. If a really hard
problem comes up, somebody says,
‘Sure, I’ll take care of it on Thursday
afternoon.’”
Working in a medical clinic with
Peter Schlag has been a tremendous
experience, she says. “It’s
exceptional to find a surgeon who
takes such a passionate interest in
the biology underlying what he’s
doing,” she says. “I’m very thankful
I could work in this environment.
As a biologist you dream that what
you are doing can be turned into
therapies, but working with physicians, you start to learn what makes
sense and what doesn’t, what can be applied to patients and what
can’t. In biology we often don’t think things through to the end in clinical
terms; you can spend years pursuing something that may not be at
all applicable to patients in the way you imagined.”
Significant differences in the education systems of physicians and
researchers can lead to misunderstandings, she says. “But these things
are usually easy to solve if there is direct contact, and if both sides
make an effort to understand each other.”
��
Recently she and her colleagues have made a breakthrough with
the discovery of a molecule that helps transform tumor cells into
metastases. Ulrike was under pressure to finish the paper. She looked
tired when she came by to talk about the manuscript.
“Until recently it was difficult to define the difference between a nonmetastasizing
cell in the primary tumor and one that would metasta-
size,” she says. “The old view was that a cell’s DNA suffered damage;
there was a mutation that caused a tumor. Then when the tumor had
grown for awhile, there was a second event, probably another mutation,
that caused one of the cells to metastasize. If that is really how
things work, you can analyze patient tissues all you want and you may
never find the one cell in a tumor that will turn into a metastasis. It
may be only one among billions of cells, and it might change just
before it casts off and wanders to a new tissue. The chances of seeing
that happen would be very small.”
But Ulrike and many scientists are
no longer sure that this scenario is
correct. “We’re starting to think
that there are different types of
cells in the tumor from the very
beginning. They may begin as a
sort of stem cell that moves into
an area, replicates quickly, and
then undergoes a process of development.
Some of those cells
become migrating cells that move
away. That happens over and over
during normal embryonic development,
and in that situation it isn’t
mutations that make cells migrate.
Rather, it’s a change in signaling
and the activation or repression of
genes. If you look at the question
that way, it may not be only a mutation that causes a metastasis.
Instead, we’re looking for signals and a pathway that become active in
some of the tumor cells.”
In other words, metastases may behave like muscle stem cells which
arise in a distant part of the body and then migrate to the limbs, the
process described in the previous chapter. The difference is that cancer
cells are defective, and upon arrival they don’t develop into a useful
structure. Instead they form a mass of tissue that eventually interferes
with the functions of healthy organs. And the process may start all
over again at the new site. Cells may detach themselves and travel
elsewhere to form further tumors.
The focus of the lab’s work has been colorectal cancers, a disease which
is on the rise in Western countries. “About fifty percent of these tumors
can be treated successfully through surgery, but it is notoriously difficult
to tell whether the cancer will return or metastasize,” Peter says.
“Diagnostic tools that help in other forms of cancer don’t work well
with this type. Since metastases are the major cause of death, what
75 Part one: The live of nomads

we’d like is a way to scan tumors for signs that they will develop this
way. If we find such markers, we’ll know which patients need careful
monitoring and further treatment and which don’t.”
Peter’s surgical clinic gave Ulrike and her colleagues unique access to
samples and data. “He has strongly supported us from the clinical side
the whole time,” she says. “He provided us with tumor tissues taken
from patients at various stages of the disease. We had access to even
more samples in the ‘tumor banks’ that were started by Peter, the
Robert-Rössle Clinic and the MDC. This unique repository of tissues
from cancer patients allows the validation of experimental hypothesis
in physiological situations and creates an important link to perform
translational research. We could examine colorectal tumors that
seemed benign, those that were not metastasizing, and cancerous tissue
that had already spread to other parts of the body. It was really
valuable to be able to look at several samples from the same patient.”
Peter also provided a biostatistician, Markus Niederstrasse, who
worked hard on the analysis. After comparing the complete gene activity
of different types of cells, one gene stood out; metastatic cells were
producing much higher amounts of a particular protein. The molecule
had been discovered when comparing the RNA molecules produced by
non-metastasizing and metastasizing tumors, but it hadn’t been stud-
Part one: The live of nomads
ied and didn’t yet have a name. Ulrike and her colleagues could now
give it one: they called it Metastasis-Associated in Colon Cancer 1, or
MACC1. Ulrike calls it “Maxie.”
The protein has turned out to be a very reliable sign that a colorectal
tumor will metastasize, or that it has already done so. Patients whose
tumors did not produce the protein usually had good prospects for
a lasting recovery. But when their tissue showed high levels of the
molecule, the prognosis was bad; metastases were almost sure to
develop.
Was MACC1 a key to tumor cell migrations, or just a side effect of some
other process? Ulrike and her colleagues fed the gene to cells in laboratory
cultures; when the cells began to produce the protein, they cast
off ties to their neighbors and began to crawl away from each other.
When the scientists added the human gene to tumors in animals,
metastases formed in the liver. So the protein was directly involved in
the cell migrations.
In another experiment the scientists removed parts of MACC1 that
allowed it to communicate with other proteins. “A signaling protein is
like a machine with a transmitter and receiver,” Ulrike says. “If you take
away either part, it can no longer hear messages or pass them along.
When we removed these components of Maxie, it no longer stimulat-
76

ed crawling or caused metastases. We had successfully blocked whatever
signal was making the cells migrate. That convinced us that the
protein was something like a master regulator of metastasis.”
The way the cells migrated in culture dishes resembled what Walter
Birchmeier had seen in his work with another signaling protein – Met.
Maybe there was a connection – maybe Met was passing information
to MACC1, or vice versa. “We had indirect evidence for this because cells
that produce MACC1 also make Met,” Ulrike says. “So the next step was
to prevent the cells from making any MACC1. Immediately they reduce
their production of Met as well.”
The recipe for the Met protein is contained in one of our genes, but few
cells use it. Like other genes it is usually silent until another protein
called a transcription factor binds to it and activates it. Experiments
showed that the amount of Met produced by the cell directly depended
on how much MACC1 was being made. A careful study of MACC1
showed that it could dock onto the Met gene – it was acting as a transcription
factor.
��
The next step will be to try to find something to block MACC1’s signaling
activity in tumors in mice, and eventually human beings.
This is the point at which basic research might become transformed
into real help for patients. With a great deal of luck, an existing drug or
substance might disrupt the protein’s activity. But the chances of an
easy success are very slim. An effective drug has to be precise – it
should target only the disruptive molecule, only in cells that are misbehaving.
The first problem is hard enough, although Ulrike says that
MACC1 has features that might make it susceptible to drugs. A second
hurdle is delivery: getting the substance to just the right cells in sufficient
quantities. Many believe that some sort of new conceptual breakthrough
will be necessary to solve it.
That may be on the verge of happening. In the past few years scientists
have discovered that you don’t necessarily have to rewrite the recipe in
a gene to keep a cell from producing a particular protein. There’s an
intermediate step; genetic information is first rewritten into a form
called RNA before proteins are made. The idea is explained in more
detail in the next chapter, but it basically involves building an artificial
RNA molecule that interferes with the one the cell has made. Ulrike
and her colleagues have tried this with MACC1 in laboratory cell cultures.
With a single treatment they have managed to block cell migrations
for months.
The group is now using genetically engineered mice to see if MACC1
RNA can be blocked in animals, and if so, whether that prevents the
development of metastases. Even if the strategy works, Ulrike admits,
it will be a long road to doing something similar in humans. Once
again, the hurdle will be to shut down the molecule only in tumor cells,
because blocking it everywhere in the body would surely have unwanted
side effects. “We don’t know what the healthy functions of Maxie
are,” Ulrike says. “But it is produced in a wide range of healthy tissues,
where it probably has important jobs.”
Pure knowledge isn’t a cure, as Peter Schlag says. Yet it may still have a
powerful impact on health. At the very least, MACC1 may help accomplish
something that has been impossible so far. It will probably make
an excellent tool to predict whether a patient’s colorectal tumor is likely
to spread and take on a much more deadly form. Metastases can’t
yet be caught and trapped in a jar, like fireflies. But maybe soon surgeons
will be able to see where they alight.
77 Part one: The live of nomads

A very quiet cure
In 1952 composer John Cage amused, mystified, and enraged the public
with the first performance of his work Four minutes and thirtythree
seconds of silence. David Tudor came onto a stage in Woodstock,
New York, sat down at a piano, and closed the lid of the keyboard. Then
he sat there, staring at his music and a stopwatch. He briefly opened the
lid and closed it again between each movement of the piece. Not that
there was complete silence during the performance; one commentator
said, “Mostly what you could hear was people getting up and walking
out.”
Two events inspired the composition. Cage’s friend Robert Rauschen -
berg had just painted a series of pure white canvases that never looked
completely white because of changing light conditions, the shadows of
viewers, and other factors. (The series may also have inspired the last
painting of Buch artist Jeanne Mammen, which is also white – as it
aged, she said, the surface would gradually turn gold, and underlying
colors would come through.) The second influence was Cage’s visit to a
unique room that had been built at Harvard University, a cork-lined
chamber that absorbed all sound. Upon entering, he hoped to
Part one: The live of nomads
78

experience absolute silence. Instead, he
wrote, “I heard two sounds, one high and one
low... the high one was my nervous system in
operation, the low one my blood in circulation.”
There was no such thing as complete
silence, he realized, and one aim of the new
composition was to demonstrate the impossibility
of achieving it.
In January 2004 the work was performed by
the BBC Symphony Orchestra and broadcast
live on Britain’s Radio 3. (Beforehand the station
had to switch off an emergency system
which breaks in automatically if a long silence
occurs during a broadcast.) Although the
piece has been performed many other times,
it never became wildly popular. That might
have happened if Cage had succeeded in his
plan to sell it to the Muzak company – people
in elevators and department stores probably
would have preferred it to the alternatives.
The popularity of silence has been growing
among biologists, who hope to use it as a
method of treating genetic diseases, cancer,
and a range of other illnesses. “Obviously we
don’t mean silence in music,” says Thomas
Christély. He is Chief Executive Officer at the
79 Part one: The live of nomads
Thomas Christély
company Silence Therapeutics AG, one of the
many private companies on the Berlin-Buch
campus. “Silencing genes means preventing
them from being used to make proteins.
Proteins are usually what disrupt cells and
ultimately cause the symptoms of disease, so
learning to block their production would give
us a handle on cancer, genetic diseases, and a
wide range of other health problems.” As the
last chapter showed, if tumor cells didn’t
make the protein MACC1, they might not set
off on dangerous journeys to other tissues. In
other cases, things break down because
mutations give a protein the wrong shape

and chemistry and it behaves badly. Switching
it off might solve the problem.
One of the challenges in the approach has
been to develop treatments that work as long
as possible. Four and a half minutes of silence
might seem almost unbearably long during a
radio broadcast, but you wouldn’t want to
have to treat a patient thirteen times every
hour. Most therapeutic molecules survive
longer than that, but not long enough, so biologists
have been trying to find ways to
increase their lifespans.
Silence Therapeutics AG is working on a solution.
They are building new kinds of molecules
and packing them in new ways. Their products
don’t have to achieve the length of John
Cage’s longest work, called As slow as possible.
(The piece lasts 639 years and is currently
being performed in St. Burchardi, a church in
the German town of Halberstadt.) But the
company hopes to make therapeutic molecules
that last a very long time.
��
Important discoveries in biology often arise
in unlikely places. In the late 1980s, scientists
in a biotech company in California were
using genetic engineering to try to give pale,
purple petunias a more intense color. Rich
Jorgensen’s lab inserted an extra copy of the
gene responsible for purple pigment, expecting
that more genes would produce more pigment
protein. Instead, the result was a completely
white flower without any pigment at
all.
A look into the plant’s cells showed that they
were actually using both genes to produce
RNAs – the molecules which bridge the gap
between genetic recipes and their final form
of proteins. Somehow RNAs from the two
genes were interfering with each other and
blocking the synthesis of purple pigment.
Labs across the world were discovering the
same phenomenon in other plants, and soon
Part one: The live of nomads
80
it would be observed in animals as well.
Finding out why took several more years.
A DNA molecule is a string of building blocks
called nucleotides, or bases. The cell builds two
strands of DNA with complementary bases –
that means the units in one strand chemically
attract those in the other. The two strands
wind around each other and form a shape
called a double helix. RNAs are made of the
same types of building blocks, but they are
usually found as single strands in the cell.
They could form double strands if the cell produced
two RNAs with complementary
sequences.
That’s what was happening in Jorgensen’s
petunias. The RNAs produced by the two purple
pigment genes were complementary.
When they doubled up, they came to the
attention of a cellular defense system. Many
viruses contain double-stranded RNAs, and
the cell perceives them as alien and dangerous.
When one is spotted, it is attacked by pro-

teins that chop up and destroy RNA molecules.
The first examples of this process turned up
in genetic engineering projects like
Jorgensen’s, but in the meantime this type of
RNA interference was found to happen naturally
in cells. New analyses of the human
genome show that it contains the recipes for
a huge number of RNAs, usually very small
ones, whose sole purpose seems to be to lock
onto other RNAs and prevent them from
being used to make proteins. This gives the
cell yet another system of control over how
the information in genes is used – like giving
a car an extra set of brakes. The discovery was
recognized as so important that Andrew Fire
and Craig Mello, who discovered the process
by which cells dismantle interfering RNAs,
were awarded a Nobel Prize in 2006.
Scientists quickly saw that this could be
turned into a new method of controlling
genes in the lab, and possibly in patients. It is
now routinely used in experiments with ani-
mals. Now Silence Therapeutics and others
are beginning to use the method in the treatment
of human diseases.
��
Surgeons have a different perspective
on molecular medicine than basic
researchers; biotech companies have yet
another. To survive they have to create and sell
products, provide useful services, and convince
investors that they will continue to do
so. Thomas Christély admits that it is a
tremendously competitive field. Success
requires finding an original approach to a
problem that can be realistically addressed
and which fills a niche in the market.
“Our company is doing that by focusing on
two crucial aspects of RNA interference therapies,”
he says. “The first is to make long-lasting
small interfering RNA (siRNA) molecules
that shut down genes involved in important
diseases. The second is to develop a new strat-
81 Part one: The live of nomads
egy for delivering those molecules to target
cells.”
Thomas says the company’s approach is the
result of a lot of hard work by several people.
“The Chief Scientific Officer of Silence
Therapeutics AG is Klaus Giese, who founded
the company in 1998 under its previous name,
‘Atugen AG.’ Other people who have been really
important are his colleagues Jörg
Kaufmann and Anke Klippel, who all studied
for many years on the West Coast in the US,
then worked for Chiron Corporation in
California before they joined the company.
With Oliver Keil, who joined later in 2002, they
jointly invented our proprietary siRNA and
delivery technologies.”
Several factors make RNA interference particularly
interesting, he says. One reason that
drugs are so expensive is that the development
process typically takes an incredibly
long time – ten years or more. “That process
usually starts by identifying a protein that is

Britta Dieckhoff
crucial to the development of an illness – one
that would stop the disease if you could block
it,” Thomas says. “That in itself usually takes
years and for many diseases it hasn’t happened
yet. If you get to that stage, many more
years are usually required to find and refine a
compound that influences the disease
process – if one exists. If you’re successful it
has to be studied for toxic side-effects, and
then undergo a number of other tests before
you can try it in humans. RNA interference is
based on a natural mechanism. It works with
just about every gene, and once you have
identified the one you’re interested in, you can
produce the siRNA in two or three months.”
One big advantage of siRNA molecules is that
the total development time, compared to
conventional drug technologies, can be shortened
by a few years – which is extremely
important from the commercial point of view
of a pharmaceutical company. “If the revenues
from the sale of siRNA drugs are generated
a few years earlier, this will significantly
Part one: The live of nomads
82
improve the cash flow for such drugs” he says.
“In addition, development costs will be significantly
lower due to the shorter development
time. Our siRNA technology will make it easier
to earn back the huge investment involved
in creating a new therapy.”
Companies with conventional therapeutic
technologies usually focus almost exclusively
on cures for diseases that affect huge numbers
of people. Bringing a drug to the market
faster would reduce the price and also entice
companies to work on serious diseases that
strike fewer people.
��
Asmall interfering RNA molecule looks like
a bit like a comb, or a ladder that has
been sawed in half from top to bottom, leaving
one side bar and half-rungs. When it binds
to its partner in the cell, it forms a complete
ladder. The RNAs designed by the chemists of
Silence Therapeutics have a different structure
than most of those made elsewhere.

Jens Endruschat
“A conventional siRNA doesn’t completely
match up to the RNA made by the cell,”
Thomas says. “It’s like a ladder with a longer
bar on one side and a few extra rungs that
don’t connect to the other side, which are
called overhangs. My colleagues Jörg
Kaufmann, Klaus Giese and Anke Klippel have
developed a special kind of molecule called
AtuRNAi which is a small interfering RNA molecule
with ‘blunt ends’ – in other words, without
any overhangs. When it docks onto its
partner, the ends match.” He lists some other
differences: AtuRNAi contains only naturally
occurring RNA building blocks, whereas most
other companies build molecules out of a
mixture of elements from RNA and DNA. The
AtuRNAi molecule can also be produced faster
and less expensively than other types of
siRNAs.
The most significant difference, though, may
be that AtuRNAi is more durable and still
highly potent. It works longer before the cell
breaks it down and a new treatment is
required. The short lifetimes of most RNA
molecules limits their use in therapies. The
lifetime of company’s AtuRNAi molecule has
been extended from a few minutes to a week
or so, which means that less of the molecule
is needed for therapeutic applications.
Another of the company’s innovations confronts
the problem of delivering therapeutic
molecules to diseased cells. “An siRNA has to
survive while it travels through the bloodstream
and it also has to be taken up by cells,”
Thomas says. “This means attaching it to a
delivery vehicle that will both protect it and
then be absorbed once it encounters the target.”
Most solutions that have been tried involve
attaching the molecules to liposomes, small
bubble-like compartments. They are built of
lipids, the fat molecules which make up cell
membranes. In the body, molecules are often
packed into liposomes when they have to be
shuttled between cells, so the approach takes
advantage of the fact that cells know how to
83
Part one: The live of nomads
absorb liposomes. Silence Therapeutics started
with the same basic structure but then
modified it to create something new.
The company’s calls its delivery vehicle the
AtuPLEX. “It’s a combination of AtuRNAi
attached to a specific lipid sphere,” Thomas
says. “The size of our sphere is about one-anda-half
times larger than the liposomes that
are typically used, but since different tissues
usually cope with liposomes of different sizes,
we can alter its dimensions to fit the situation.
One advantage of a large vessel is that
it can be used to deliver several different
small interfering RNAs to the same cell.
That’s important in the many cases where
you would like to shut down several
proteins. For example, it would allow you
to block multiple components of a signaling
pathway.”
Another difference has to do with the liposome’s
electrical charge. Lipid molecules carry
a charge that play an important role in
whether they are absorbed. Cells often let

Britta Dieckhoff
them enter – or block the process – because
they need to balance their own charge with
that of the environment. The AtuPLEX has a
lower charge than most liposomes, which
makes delivery more likely in most situations.
The company has several specific projects
under development, usually with major
industrial partners like Pfizer/Quark and
AstraZeneca. An AtuRNAi built to interfere
with a protein called RTP801 is currently
undergoing Phase I clinical studies in human
patients; the goal is to combat a common disease
called macular degeneration. The macula
is a small yellow spot in the retina that plays
an important role in the sharpness of our eyesight
and how well we perceive colors. In
many people it degenerates with age, a
process that can lead to blindness. Many of
the earliest therapies involving interfering
Part one: The live of nomads
84
RNAs have focused on this disease as a sort of
proof of principle; the protein culprit is clear;
the eye is a closed system.
By the end of 2008, clinical trials of five other
AtuRNAi molecules are scheduled to begin;
two more should follow soon after that. The
new therapies will target cancers of the gastrointestinal
tract, lungs, prostate, and liver, as
well as other very serious diseases.

Some of the preliminary results have been
very promising, Thomas says. The laboratory
of Prof. Bertram Wiedenmann of the Charité
recently used one of the company’s small
interfering RNAs, called Atu027, in a study of
pancreatic tumors in mice. The RNA blocks the
production of a protein called PKN3, whose
role in cancer was discovered in 2004. A company
research group headed by Anke Klippel
discovered that the protein passes along
information from a signaling pathway that
causes metastases in laboratory cell cultures.
Samples taken from patients with prostrate
tumors also showed high levels of PKN3. All of
these factors suggested it might make a good
target for an anti-cancer therapy.
In Wiedenmann’s study, mice with tumors
received ten injections of the interfering RNA
molecule over a three-week period. The
tumors of untreated animals grew at their
normal, devastating pace, whereas their
growth rate in treated animals was significantly
reduced, as was the metastatic spread
of cells. “This is strong evidence that our system
very effectively blocks the production of
cancer-causing proteins and has a huge
impact on the further course of the disease.
It’s also proof that the delivery system works
in live animals. We expect it to be equally
effective in humans,” Thomas says.
��
Silence Therapeutics has labs and offices on
the first and second floor of an elegant
white building on the west side of campus.
Aside from the security downstairs and the
firm’s logo printed on the glass doors, there is
little to indicate this is the domain of a company
rather than a research lab. But the world
of business has its own customs and
etiquette which surface from time to time as
Maj Britt Hansen tries to take photos –
some rooms are off limits – and during
my meeting with Thomas. When we sit down
in a quiet conference room, he is prepared
to talk me through a standard company presentation
that he gave last week at a
BioEurope meeting – a mixture of science,
performance indicators, logos of impressive
partners, and of course money matters.
All outlined in 16 impeccable Powerpoint
slides.
When I tell Thomas I am trying to get a feeling
for the personality of the company and understand
how it fits into the puzzle of a very complex
campus, he shifts gears. He talks a bit
about his own background. You would never
85 Part one: The live of nomads
guess it from talking to him, but he never
studied science – although his majors in high
school were chemistry and biology. He
received degrees in business and law from the
University of Hamburg, worked in mergers
and acquisitions of a Swedish investment
bank in London, then became managing partner
of an investment firm in Moscow. In 1996
he began working as Senior Vice President
and Chief Financial Officer at the Swiss pharmaceutical
company OXO Chemie AG and
founded a subsidiary called OXO Chemie Inc.
in San Francisco. He has been with Silence
Therapeutics since 2001. His career has taken
him from Belgium to England, Switzerland,
Russia, and the US. It’s not only scientists who
are nomads, but the many people associated
with it.
Thomas sees long-term perspectives for
Silence Therapeutics in Buch. “We’re not in the
city center,” he admits. “But being on the campus
has forged relationships with groups who
might discover new targets for therapies, who
can look at the behavior of those molecules in
diseases, and who can carry out clinical trials.
There are opportunities for partnerships at
each stage of the process of turning a target
into a marketable product. It’s rare to find that
range of expertise clustered so that they are
in continual, direct contact with each other.”

Teleportation for beginners
Part one: The live of nomads
It’s not only scientists and cells that undertake epic journeys –
sometimes artifacts from institutes become nomads as well. There
are the leather-bound books on Helmut Kettenmann’s shelves, and the
postal scale in Martin Lipp’s office (which once belonged to Nobel
Prize-winner Feodor Lynen). The farthest-roving travelers are surely the
cult objects of Thomas Jentsch, which sailed the Pacific before settling
in Berlin. But the strangest case is that of the cornerstone of the FMP,
which traveled hundreds of kilometers in a feat that any magician
would admire.
July 13, 1998, saw the groundbreaking ceremony for the new FMP
building on the Berlin-Buch campus. Staff of the institute and the
architects prepared a time capsule, a metal cylinder to be placed inside
the cornerstone. Among the contents were:
· plans of the new building
· a collection of coins (German marks)
· the FMP’s 1996-97 annual reports
· recent Berlin newspapers
· samples of proteins and DNA related to FMP research
The cylinder was lowered into the foundations – all of this observed by
over a hundred witnesses, including reputable representatives of the
local and national government – and later encased in concrete.
Jump to almost exactly five years later and 300 km away. On July 17,
2003, a team from the regional environmental office of Niedersachsen
86

was carrying out a routine investigation of an abandoned gravel pit.
They discovered a mysterious metal container – a bomb? Radioactive
waste? An artifact from a forgotten (but technically advanced) civilization?
They immediately called the police and the fire department, who
came with equipment to investigate toxic waste. An initial investigation
revealed no traces of explosives or drugs. The cannister was
shipped to various laboratories where it was X-rayed and submitted to
a gamut of other tests. Experts attempted to penetrate the container
with an endoscope but were unsuccessful.
After a month someone came up with the bright idea of simply prying
the thing open. Inside were the contents of the FMP time capsule, in a
deteriorated condition. The institute received a call from a press officer
in Niedersachsen, saying, “We think we have something that belongs
to you.”
The time capsule and its contents were returned to Berlin-Buch, and
until recently they could be seen in a display on the ground floor of the
FMP building. If you want to see them now, you’ll have to talk to Björn
Maul.
There are various hypotheses about how the cannister disappeared
from the campus and rematerialized somewhere else. The most likely
explanation – although not the most interesting one – is that the cornerstone
was accidentally removed when the FMP built a new
entrance to give access to the handicaped. (My personal favorite is
that one day the FMP switched on all their NMR machines at the same
time, briefly disturbing the Earth’s magnetic field, and accidentally
invented teleportation.)
You can’t have an institute without a cornerstone, so a new one was
made and outfitted with a time capsule which was filled with, well,
updated versions of the old stuff. (Euros, for example, instead of
German marks.) On June 1, 2004, it was packed into a stone and
embedded in the ground at the FMP. There it remains to this day.
At least, that’s where we think it is.
(freely adapted from an article by Björn Maul)
87 Part one: The live of nomads

Ludwig Thierfelder

Interlude:
A glass heart
Sometimes the history of medicine reads like a long line of unlikely
stories, and this is one of them. It goes like this: a group working on
one side of the campus removes a gene from a strain of mouse and discovers
that the animal is now unable to build a proper heart. The fate of
this small creature has large repercussions: it sheds light on work being
done with human patients on the other side of the campus, leading to
a new way of diagnosing a rare type of heart disease and a simple surgical
procedure that is saving people’s lives as you read this book. It’s
exactly what people mean when they talk about molecular medicine; it
is the way science is supposed to work. Yet it rarely happens so directly,
and when it does, it seems like a minor miracle.
The fragility of the heart has been the topic of many curious stories in
science. One of them dates back 80 years and involves a famous aviator
who decided to dabble in medical research. In May 1927 Charles
Lindbergh became an instant celebrity when he completed the first
non-stop flight across the Atlantic, taking off in New York and landing
in Paris. Four years later he published a small paper in the journal
Science, making his first contribution to the field of cardiology.
89 Interlude: A glass heart

The Carrel-Lindbergh pump
Charles Lindbergh (second from right)
Lindbergh Picture Collection. Manuscripts and Archives, Yale University Library
Interlude: A glass heart
90
Lindbergh’s quiet work on heart research only became public in
July 1935, when his second Science paper caught the attention
of the New York Times. By that time he had been collaborating
for several years with co-author Alexis Carrel, a Nobel prize-winning
surgeon and pioneer in the science of tissue cultures.
Carrel, who was working at the Rockefeller Institute for Medical
Research in New York, had a vision of a day when organs could
be removed from the body and kept alive, possibly even transplanted
from one person to another. This could only be accomplished
if tissues removed from the body could be supplied
with blood, which would require some kind of machine and an
engineer to build it, and there Carrel was stumped. His colleagues
at Rockefeller built device after device. Every machine
capable of feeding blood to organs also pumped them full of
bacteria, leading to massive infections and death.
Charles Lindbergh’s interest in the heart began with a personal
tragedy. In 1929, the year of his marriage, his wife’s sister was
diagnosed with rheumatic heart disease. The condition soon
turned fatal because doctors could not yet operate on an open
heart. What was lacking was a machine that could temporarily
take over the heart’s job and pump arterial blood to tissues. It
was, Lindbergh reflected, essentially an engineering problem.
While he knew nothing about medicine, and had never finished
his academic degree in mechanical engineering from the
University of Wisconsin, he could repair his own airplane; he
knew about pumps and valves and vacuum seals. As he began
taking notes and drawing plans of devices, he needed to talk to
physicians. This soon brought him into contact with Carrel, and
one of the most unlikely collaborations in scientific history was
born.
Originally Lindbergh planned to build some sort of heart-lung
bypass machine. Carrel quickly convinced him that the idea was
too ambitious. Instead Lindbergh should focus on a way to keep
tissues alive outside the body, which would permit researchers
to carry out a more thorough basic study of the biology of tissues
and organs. His first invention was a culture flask in which
liquid medium could circulate continuously. He moved on to
methods that could separate blood serum from plasma, and his
final invention was a system of glass vessels and pumps that
could supply whole organs with blood. The device permitted
the handling, examination, and transfer of organs without danger
of contamination. In one experiment Carrel kept epithelial
tissue alive for over 100 days. When the two men finally gave a
demonstration in a Copenhagen theater in 1936, their work
received enthusiastic reviews from scientists and a wild
response from crowds outside. The public misunderstood the
findings and began placing orders for “Lindbergh’s artificial

heart,” as a replacement when their own biological
one wore out.
The tissue culture system became part of the
standard repertoire of the Rockefeller
Institute and other labs, but the Carrel-
Lindbergh pump was difficult to operate and
fell out of use within a few years. It was no
longer necessary because most of the work
biologists wanted to do didn’t require whole
organs. They could work equally well with
cells or thin tissue slices which could be kept
alive through much simpler means. With
Carrel’s death and the beginning of World War
II, the project came to an end. But scientists of
Lindbergh’s time regarded his contributions
as important. His work on the heart, some
said, would be remembered far longer than
33-hour ride in an airplane. Crossing the
Atlantic, after all, was little more than a symbolic
gesture.
��
As long as we are healthy, it is easy to
think of the heart as a solid, engine-like
pump that will keep beating as long as it is
exercised and provided with fuel. But those
who are born with a heart defect know that
this is an illusion, and the rest of us start to
learn the lesson with middle age. The heart is
more delicate than any of us would like to
believe.
91 Interlude: A glass heart
One person who knows this very well is
Ludwig Thierfelder, Acting Chief of Cardiology
at the Franz Volhard Clinic of Charité/HELIOS
and head of a research lab at the MDC.
Ludwig is tall and youthful and to talk to him
you wouldn’t think he has a stressful job. To
get to his office you follow the road past the
Robert Rössle Clinic, walk past the oldest laboratory
building on campus, where the busts
of Oskar and Cécile Vogt keep vigil. (Cécile is
the taller one.) Across the Lindenberger Weg
is what seems like a different world: the new
HELIOS complex. The buildings are spacious
and quiet. Color-coded signs act as guides to
the Cardiology section. As you pass, patients
are being transported through the wide halls
on beds and wheeled into elevators. There are
open waiting rooms with large round reception
desks that would look equally at home in
one of Berlin’s ultramodern hotels.
Ludwig’s relationship with the heart began
during his medical studies in 1988 at the
University Clinic in Freiburg, where he treated
cardiology patients while carrying out clinical
research. “I had already done some work on
blood pressure in rats during a six-month fellowship
in the United States,” he says. “But it
was in Freiburg that I really started to get
interested in the genetics of heart muscle diseases.
The first project I participated in was a
collaboration with Hans-Peter Vosberg, of the
Max Planck Institute for Medical Research in
Bad Nauheim. Our aim was to study a large
family with inherited hypertrophic cardiomyopathy.
Cardiomyopathy is another way of
saying ‘heart muscle disease,’ and hypertrophy
means that the heart becomes dangerously
enlarged. In general, these diseases are
often connected to heart failure and a risk of
sudden death. This particular problem was a
rare condition caused by a dominant gene
that nobody had found yet.”
I have been wondering about the attraction of
such rare diseases – wouldn’t it be better to
spend your time working on problems that

Brenda Gerull
affect huge numbers of patients? Especially
in heart disease, which claims so many
victims?
“In this case it’s a bit deceptive to think of the
problem as rare,” Ludwig says thoughtfully.
“Congestive heart failure caused by various
types of hypertrophy is a major cause of
death. What we’re finding is that there are
many genetic ‘roads’ that lead to these conditions.
In most cases, each family has its own
unique mutation. Why so many different
defects lead to similar problems is an interesting
question in itself.
“Common cardiovascular diseases probably
arise because of the influence of several
genes and environmental factors. Most of the
time they probably don’t involve what we
would consider mutations. In the population
you can find many versions of each gene
involved in building the heart and operating it
throughout a person’s lifetime. Some of these
variants may make it more likely for you to
develop a disease, if you happen to inherit it
with a certain combination of other genes.
These cases are extremely complex, and figuring
out the contributions of each molecule
and the environment is very difficult. That
may not be easy to do, either, in a disease
caused by a single mutation. But it’s certainly
much easier.
“Most such single-gene diseases are rare,
that’s just the way evolution has worked. But
the genes that cause them are often molecules
whose ‘healthy’ forms may contribute in
a weaker way to these more common diseases.
So by looking at rare conditions we discover
genes that will help us understand common
ones. And it also tells us a lot about basic
processes at work in the heart – which helps
us understand the difference between
healthy functions and unhealthy ones.”
At the time of the Freiburg study, Ludwig says,
only one gene had been linked to cardiomyopathies,
a gene that encoded a muscle protein
called the β-myosin heavy chain. Ludwig
worked intensively with local physicians to
find families with the condition in Freiburg. By
the end he had recruited a family with over a
hundred members and completed their pedigrees
– filling out a chart showing the family
relationships of affected and non-affected
people, over several generations.
Interlude: A glass heart
92
“I still have those somewhere,” he says, and
goes to his bookshelf. He pulls down a binder
and spreads one of the pedigrees out on the
desk. Today pedigrees can be done on the laptop,
with programs that instantly turn everything
into neat colorful charts, but this one has
been painstakingly filled out by hand. He walks
me through one of the families. “Since the disease
involves a dominant gene, if one parent
has one copy of the gene, then half their children
will inherit it, statistically speaking.”
Comparing the DNA of affected and nonaffected
family members might reveal the
location of the gene. “On the basis of this data
and our samples, we proved that the disease
wasn’t caused by β-myosin,” he says. “That
was exciting because it meant another gene
was involved.”
He received a postdoctoral fellowship from
the German Research Council (DFG) to follow
up on the case – in the United States. He
packed up his family and moved to Boston,
joining the group of Christine Seidman, a
geneticist at the Harvard Medical School.
“We had a great time in the US, both professionally
and personally,” he says. “The other

people in the group were very nice; we’ve
maintained contact and are still friends, still
see each other every few years. One of our
children was born there, so we have an
American citizen in the family.
“In Boston the project that had started in
Freiburg moved along very successfully,” he
says. “First I pinned the families’ problem to
chromosome 15, and then specifically to a
mutation in the α-tropomyosin gene. This
molecule plays a key role in muscle contraction.”
(How scientists track down such genes
is described in “The case of the short-fingered
muskateer.”)
The results linked hypertrophic cardiomyopathy
to a problem with small structures called
sarcomeres, the key movers in muscle contraction.
“That meant we ought to be looking for
other defects in sarcomeres as potential causes
for heart muscle diseases.”
��
Most proteins carry out their jobs in cells
as parts of “molecular machines.” It’s a
metaphor based on the fact that the shapes
and chemistry of proteins let them snap onto
each other to carry out jobs like moving other
molecules, cutting them, tying them up, or
transmitting signals. Sometimes these
assemblies exist for just a split-second before
being dismantled; then the pieces can be
used for other things.
Sarcomeres more closely resemble what we
normally think of machines. They are built of
many parts that last for a long time. They
carry out a mechanical job: they are shaped
like miniature pistons and that’s how they
work as muscles expand and contract. The
whole structure is shaped like a tube. At the
center is a thick fiber made of the myosin protein,
surrounded by sheath-like actin fibers
that slide back and forth to lengthen and
shorten the structure. The pieces are held
together by the long, spring-like molecule
titin. Thousands of sarcomeres, lined up endto-end,
give muscle its striped appearance.
Over the past 20 years scientists have worked
out many of the chemical and physical details
that allow these machines to work. Sticking
out of the myosin fiber are small knob-like
“heads”. When they dock onto the nearby
actin sheath, the shape of the head changes.
That pushes the actin filament a short distance
and it frees up part of the head to grab
small energy molecules called ATP. When that
happens, the head lets go of the filament and
snaps back to its original shape and position.
It docks onto a new spot further down the
actin filament and the process starts over
again.
For this system to work, all the molecules
have to have the right shapes and chemistry.
If one of the proteins has a mutation, it may
function slightly differently, which can make
the sarcomere work less efficiently. Magnified
billions of times, over the length of a muscle
fiber, the effect is amplified. That easily causes
problems for organs such as the heart.
α-tropomyosin, the gene in which Ludwig had
found mutations, lies alongside the myosin
fiber and fine-tunes the contact between
myosin heads and actin. Mutations affect its
performance and eventually result in heart
disease.
If one defect in sarcomeres could cause disease,
so could others. Ludwig and his Harvard
colleagues soon discovered a new defect in
sarcomeres that caused hypertrophic cardiomyopathy.
A project headed by Hugh
Watkins, another postdoc in Christine
Seidman’s lab, turned up mutations in a protein
called MyBP-C. The abbreviation stands
for myosin binding protein C, and that’s what
the molecule does: it forms a bridge between
the myosin filament and other parts of the
sarcomere.
��
After Harvard, what next? Ludwig wanted
to return to Germany, where he could
practice medicine. From Boston he began
looking for jobs, occasionally flying back for
93 Interlude: A glass heart
interviews. One stop was Berlin-Buch, where
he visited the clinics of the Charité and the
new laboratories of the MDC. Its scientific
director, Detlev Ganten, was encouraging. The
institute needed young experts at home in
worlds of clinical and basic science.
“There were very few places in the world
where I could pursue both parts of my career,”
Ludwig says. “In Buch, that would be strongly
supported.” He toured the Franz-Volhard
Clinic tucked into a stand of trees on the spacious
grounds of Area 1, not suspecting that in
eight years he would be walk through the
doors as Chief of Cardiology
When his family moved to Berlin in 1994, he
had few illusions about what faced him. At
least two years of intensive medical practice
lay ahead before he completed his specialty
training. On top of that, he had to set up a
new research group, normally a full-time job
in itself. At Harvard he had begun working
with animal models of heart disease, particularly
rats. He wanted to continue that, which
meant establishing new lines of animals at
the institute. He needed to hire clinicians who
could learn molecular genetics. He would
need to establish relationships with other
physicians to recruit families with various
types of heart muscle diseases. And all of this
needed to lead to scientific results within just
a few years, if he hoped to turn his initial fiveyear
contract as a junior group leader at the
MDC into a long-term career.
Over the next few years the group studied a
variety of inherited heart muscle diseases,

making pedigrees and taking samples from
large families in Germany, Canada, Australia
and elsewhere. Their investigation of a heart
muscle disease called dilated cardiomyopathy
revealed yet another mutation in a sarcomere
component.
Titin, the largest protein made by human cells,
is a long spring-like molecule. It is lodged in
the ends of the sarcomere and stretches
toward the middle, linking all the pieces. As
the myosin and actin fibers crawl along each
other, the entire sarcomere is stretched. Titin
helps keep things anchored. But it doesn’t
only do structural jobs. The healthy form of
the protein consists of hundreds of modules
and many different types of domains. Some of
them help pass signals; others are important
in muscle development, probably because the
protein acts as a scaffold to attach things to
as sarcomeres are assembled from individual
proteins.
Brenda Gerull, an MD who has been working
on cardiomyopathies in Ludwig’s group, found
the mutation while studying a large
Australian family with an inherited form of
the disease. “They have just two additional
letters in the DNA code for this huge molecule,”
she says. “That has big effects, because
it scrambles some of the information needed
to make titin protein. The cell ends up making
titin proteins that are too short. You wouldn’t
expect a machine to work right if you shortened
its springs, so it’s no surprise that
changes in this molecule disturb the sarcomere
and the heart.”
Ludwig is frank about the meaning of these
discoveries. “Our original focus was to identify
genes,” he says. “Knowledge may be very helpful
– particularly in diagnosis. If you have a
gene that causes a problem in adulthood and
you find it in a young person, you may be able
to take precautions that will prevent the prob-
Interlude: A glass heart
94
lem from becoming really serious. But we’d
like to do more, and so increasingly the lab is
turning to identifying and studying the
processes by which defective genes cause
problems for the heart and other organs.”
Yet the power of knowledge and diagnosis
shouldn’t be underestimated, as the lab
was about to find out from an unexpected
source.
��
The first chapter of this book describes
Walter Birchmeier’s work on proteins
that govern cell migration. Cells in most tissues
are firmly tied to each other, but when
they migrate the links have to be dissolved.
This happens all the time as embryos develop,
and it happens as cancer cells free themselves
from a tumor. To work out how cells control
these processes, Walter and his colleagues
have been systematically removing genes in

mice and observing their effects on development
and other processes.
One of the molecules they are interested in is
plakophilin 2, which appears in tissues such as
growing skin, tumors, and heart cells. It is part
of the desmosome, a sophisticated set of set
of proteins which hook neighboring cells to
each other. Fibers that tie cells together have
to be anchored somewhere, and that’s what
plakophilin 2 does. Just under the membrane
at the fringes of the cell, it acts as a peg.
“Mutations in many desmosome proteins
have been linked to disease,” Walter says. “But
until until we knocked out the gene in the
mouse, no diseases had been linked to
plakophilin 2, and we didn’t have much of an
idea of its functions in development.”
Katja Grossmann, a PhD student in the group,
discovered that mice with no plakophilin 2
died mid-way through embryonic develop-
ment. The problem lay with the heart. “Under
the microscope you could see that the fibers
that were supposed to join cells were lying
loose and scattered around,” Walter says. “As if
someone had forgotten to secure the lines
when they tied two boats up to each other.”
Connections between heart muscle cells are
crucial, he says, to ensure that they can communicate
with each other and pass signals
telling them when to contract.
Walter says that many of the key proteins that
establish these connections were originally
uncovered by Werner Franke, a researcher in
Heidelberg at the German Cancer Research
Center (DKFZ) of the Helmholtz Association.
“Over the past 30 years, Werner has discovered
many of the proteins that play a crucial
role in connections between cells,” Walter
says. “And he has carried out a number of elegant
studies that show how the pieces fit
together.”
95 Interlude: A glass heart
One of those studies was published in 2002.
As Walter’s group made knockouts of some of
the pieces of desmosomes, one-by-one in the
mouse, Franke and his colleagues were putting
the pieces together. They were interested
in the fact that cells in different tissues used
many of the same components to attach
themselves to their neighbors, but there were
subtle differences. One way to understand the
details would be to extract whole desmosomes
and other types of junction complexes
from cells, but that approach had failed
because of the complexity of the structures
and the fact that they were bound to membranes.
Franke’s lab decided to investigate the question
by adding proteins to cells piece-by-piece
and watching what happened. They began
with a laboratory cell line that produces one
protein of the desmosome, called desmoglein
2, but doesn’t produce most of the other com-

ponents or use them to build proper links to
neighboring cells. The group began inserting
various combinations of other desmosome
molecules. One finding was that plakophilin 2
plays a key role in organizing many of the
other proteins, helping them get to their
proper places. But properly building the structures
required three proteins: desmoplakin,
plakoglobin, and plakophilin 2.
��
Walter knew only too well that mutations
in heart muscle proteins could
cause disease. A few years earlier his group
had knocked out plakoglobin, one of the
essential pieces of the desmosome.
“Removing it also caused fatal heart ruptures
in the mouse, but we didn’t immediately follow
up to see if there was a connection to dis-
Interlude: A glass heart
96
ease,” he says ruefully. “The result was that we
got scooped.”
The lapse meant that William McKenna’s
group from St. George’s Hospital Medical
School in London was faster in linking mutations
in human plakoglobin to Naxos disease.
This is one of several diseases called arrhythmogenic
right ventricular cardiomyopathies, or
ARVCs. (The translation for the non-cardiologist
is, “diseases of heart muscle in which
there are defects in the right ventricle and the
heart beats irregularly.”)
Naxos takes its name from the Greek island,
where a recessive copy of the gene is very
common. People who inherit two copies of
the mutation suffer from sudden, wild fluctuations
of cardiac rhythm that often lead to
heart failure and sudden death. ARVC is much

less common in other parts of the world, but
still it is the leading cause of sudden death for
cardiac reasons among young people.
Vowing not to make the same mistake twice,
Walter immediately took the plakophilin 2
findings to the MDC’s experts on heart muscle
diseases. Ludwig, Brenda Gerull and their
colleagues identified ARVC patients, enlisted
them in the study, and began checking for
mutations. In 120 unrelated patients, they
found 32 people with mutations in the
plakophilin 2 gene. In all, Ludwig says, about
30 percent of patients with the condition
have shown mutations in plakophilin 2. Many
of them may have a high risk for sudden cardiac
death.
“To give you an idea of how bad this condition
can be,” Ludwig says, “in a study of ARVC
which we completed in 2005, 50 percent of
the men in the high-risk group were dead by
the age of 39. Women at high risk survive
much longer – the average age of death was
71 years. We don’t yet know why there is a difference
between the sexes, and we also don’t
know what gives some people with the mutation
better chances of survival than others.”
In a collaboration that has been going on for
many years with a group at Memorial
University in Newfoundland, Canada, the scientists
investigated an extended Canadian
family with an inheritable form of ARVC. Here
defects in another gene were responsible for
ARVC – this time, on the third human chromosome.
In early 2008 the scientists finally
traced the problem to a mutation in a gene
called TMEM43. They do not yet know what
the molecule does, but it may have something
to do with deposits of fat that build up in
muscle fibers.
“We knew we couldn’t cure the disease, but
we could try to do something about the irregularities
in heartbeat rhythm that caused the
fatalities,” Ludwig says. “ARVC patients do not
reliably respond to drugs that stabilize the
heart, but they can be surgically implanted
with a small defibrillator. This device detects
An early sketch of Lindbergh’s pump
when the heart begins to beat erratically and
then administers an electric stimulation to
bring it back to a normal pattern. It also
records the events so we can later check to see
how the heart – and the devices – have functioned.”
The team carried out a clinical study in which
the defibrillators – called ICDs – were implanted
in 30 young, high-risk men and 18 women
who had mutations in the region of DNA that
held TMEM43. The results of the intervention
were very clear. “In five years after the operation,
70 percent of the devices had sprung
into action at least once,” Ludwig says. “But
there were no deaths among the patients.
Compared to men who could not receive the
defibrillator (because they died long before
the connection between the heritable heart
disease and TMEM43 was made), 28 percent
died within five years. Within ten years they
were all dead. So by identifying people with
AVRC with a genetic test, and using the defib-
97 Interlude: A glass heart
rillator as therapy, we have been able to save
several lives.”
One of those cases was local, here in Berlin.
The Franz Volhard Clinic also performs the
operation, implanting the devices in high-risk
patients. One person to receive the implant
was a young local woman who carried the
plakophilin 2 mutation. As she was outfitted
with the defibrillator, her husband underwent
first-aid training to learn what to do in the
event that his wife suddenly experienced
problems. That happened more quickly than
anyone expected. Shortly afterward, she experienced
a heart attack during the night.
Thanks to the implant and her husband’s
preparation, her life was saved.
Ludwig expects to find more families with
other mutations that cause AVRC. Linkage
studies should help scientists find the genes –
or at least regions of DNA – that are responsible.
Then new types of tests can be made to

identify those who have the highest risk of
sudden cardiac death. And then he knows
what to do.
You can tell it’s a very good feeling.
��
One Christmas when I was eight or nine
years old, my parents gave me a
Transparent Man, a human body that was 50
centimeters tall, with see-through plastic
skin, a skeleton, and organs that could be
removed. I gave it a name, which I can’t
remember. I remember being unhappy that
the intricate network of veins and arteries
were stamped into the plastic and painted
blue and red rather than threading their way
from the lungs to the heart and into the
organs. I remember removing the cap of the
skull and taking out the brain. And lifting the
ribcage to expose the heart. Later my little sister
hid one of the lungs and we never found it
again.
In high school I got the chance to watch surgery
for real, and remember that the inside of
the body was nothing like the tidy inner cavity
of my toy. Little wonder that doctors and
surgeons had to study so long – how could
you learn to see structure in the chaos of
blood and tissue under the skin?
Ludwig was never given a Transparent Man,
and the first time he saw surgery was during
his medical studies. And that was an opportunity
he almost never had.
“I was a bad student in school,” he laughs. “I
just didn’t get it; I had bad grades. I never
would have been admitted to medical school
in Germany.”
Interlude: A glass heart
98
He was born and grew up in this country, but
his father was Austrian and that gave him
Austrian citizenship. It also entitled him to
study there. So at the age of 20, he applied to
the University of Vienna and was accepted.
He loved the city and the medical school and
its great teachers. Suddenly, he says, “I got it.”
Medicine appealed to him very much. “I liked
the mix of practical skills and intelligence. I
seem to have been good at it,” he smiles.
It’s a mighty understatement for someone
who went from the back row of biology class
to Harvard, and then to one of the most important
medical positions anywhere: Head of
Cardiology at one of Germany’s best hospitals.
“There are a lot of kids in Germany today who
won’t have the chance I had,” he says. “I’ve
seen it with my own children. The current

educational system puts too much emphasis
on great performance at too early an age. At
the age of ten or eleven, children get pushed
into tracks that are hard to escape from. Most
of them won’t be able to go abroad, like I did.
It’s a shame, because Germany is losing a lot
of young people with great potential.” It’s a
subject we could talk about for a long time,
but there are patients waiting.
As we shake hands I think that for our generation,
this is the real molecular medicine. The
miracle cures – the molecules that will slip
into cells and replace defective genes, the
immune system cells that will be taught to
kill tumors – those are still somewhere on the
horizon, or just beyond. For now what can certainly
be done is to take the best of molecular
biology and use genetics and other methods
to probe the mechanisms that cause disease,
then turn that knowledge into diagnostic
tools. If the problem can be clearly identified,
it can be attacked with the best of today’s
medical techniques and a long line of new
drugs. That’s not an answer that will satisfy
everyone; it won’t let us live forever. The
crowds are still outside, still placing orders for
the artificial heart that Lindbergh dreamed of
but never managed to build.
From across the street I look back and realize
that Ludwig and his colleagues are making
the new transparent man, a better one. It will
be different than the model I used to have.
There will be far more parts. You’ll be able to
zoom in as far as you like, zoom through the
transparent skin and into transparent organs,
transparent cells, transparent genes. You’ll be
able to wind it up and watch the smallest
parts dance and whirl.
99 Interlude: A glass heart
A lot of us are too old for such a toy. But that’s
all right. We’ll give it to our children to play
with. They’ll grow up with it, learn where all
the parts go, and give it a funny name. And
that will be their vision of a human being as
they invent the medicine that Lindbergh
dreamed of, that we dream of.

Part Two: Identity crisis

Martin Lipp
In the year 1248, the city of Cairo constructed what
was probably the largest hospital that has ever
existed – a palace-like complex with high walls, open
courtyards, and graceful towers. Al-Mansouri hospital
held more than 8,000 beds, in specialized wards
for surgery, fractures, fever, eye diseases, mental illness
and other medical problems. Musicians and
storytellers were hired to entertain the ill. The hospital
accepted patients of all races, religions, and social
classes and kept them until they either died or were
cured. Patients had to prove they were ready to be
discharged by eating a whole chicken. Those who
passed were sent on their way with a new set of
clothes and the equivalent of about ten Euros in
pocket money. As well as a full belly.
The first Chief of Physicians at Al-Mansouri was the
great healer Ibn al-Nafis, author of an 80-volume
medical encyclopedia that became the definitive
103 Part Two: Identity crisis
Where slow
rivers meet

text for generations of physicians. Its ornately-scripted
texts and gilded drawings contain
history’s first correct description of the functions
and structure of the human circulatory
system, one of the accomplishments that
secured al-Nafi’s place in history. (A copy of
some of his manuscripts turned up in Berlin in
1924.) But his teachings contradicted those of
the great Roman physician and pundit Galen,
who did not believe that blood recirculated, or
that the heart pumped it. Because Western
physicians still preferred the opinions of the
ancient scholars over experiments and direct
experience, it took more than 350 years for
William Harvey and a group of Italians to rediscover
circulation and introduce it in the West.
It’s odd to imagine a culture ignorant of the
role of blood vessels, which are as near as the
next cut or scratch – and what could be more
obvious than the pumping activity of the
heart? But today’s society has its own blind
spots when it comes to the body. People of the
future might find it equally strange that we
were almost totally ignorant of a second riverlike
system that sends tributaries through all
of our tissues and organs. It is rarely noticed
until something goes wrong; then the lymphatic
system manifests itself as swollen
nodes, or tonsils, or much worse problems.
William Harvey would have done well to
know more about it; he and several other
physicians were barely escaped execution
when they failed to prevent the death of King
James I, who had been under their care. One
of the king’s ailments was arthritis. Many
autoimmune diseases, including some forms
of arthritis, are now being traced back to
defects in the lymphatic system.
Through this network of vessels and nodes
flows a slow river, carried along by the contractions
of muscles that line its tubes rather
than a central pump. But it is by no means a
uiet place. It carries white blood cells and
other police of the immune system throughout
the body, and serves as a training ground
where cells are taught to fight pathogens –
Part Two: Identity crisis
104
viruses, bacteria and other invaders. Pirate-like
cancer cells that have detached from a tumor
also use this trafficway. They take advantage
of the route to spread to new tissues, causing
deadly metastases and often taking up residence
in the lymph system itself.
��
Martin Lipp of the MDC is a quiet-spoken
man with blond hair that is becoming
silver ("It's all blond!" he says) and enough
patience to give me a beginner’s tour of the
lymph system, which he has studied intensively
for over two decades. We sit down at his computer,
where he runs through a set of slides
with neat, clear graphics. I ask who made them.
“Me,” he says. “I like to do illustrations; it’s
relaxing. A good change.”
One slide shows a huge network of lymph
vessels spread throughout the body. At regular
intervals are small round nodes, like rest
stops along a highway. “More like meeting
points and training centers,” he says. “Those
are two of the nodes’ functions.”
Martin and his colleagues have found new
kinds of cells that travel through lymph vessels,
identified signals that guide their migrations,
made important discoveries about how
the cells work together, and helped untangle
the intricate structure of lymph nodes. In the
process he has learned a great deal about
how the system arises in embryos and how it
contributes to disease.
Lymph consists of blood and other liquids that
collect between cells. The fluids drain into tiny
lymph capillaries that flow into larger and
larger vessels and finally into a vein near the
heart. This connects the lymphatic and circulatory
systems. The routes are traveled by
white blood cells and other immune system
cells that help fight off foreign invaders, are
responsible for rejecting transplanted organs,
and sometimes turn against the body’s own
tissues. It has taken decades to work out what
these cells are, how they work with each
other, and how they distinguish self from not

self. In other words, how the immune system
is able to recognize what belongs to your
body and what doesn’t.
“That process might not be so complicated if
all the cells in your body carried something
like an identical I.D. card, but they don’t,”
Martin says. “Each kind of human cell is
unique and looks different to the immune
system. Each foreign virus or bacteria is also
unique. Yet the body usually still manages to
tell the difference.”
That task belongs to white blood cells called B
cells and T cells. Both types are born in the
bone marrow, but T cells leave at an immature
stage. They move to the thymus, a lymph
gland near the heart, to finish their development.
B cells stay behind to mature. That
involves producing antibodies – Y-shaped proteins
that decorate their surfaces. Antibodies
are created in an unusual way, by cutting and
pasting genes together in random arrangements.
The result is that the body can probably
produce about ten billion different types
of antibodies – as if a locksmith were to take
billions of blank keys and cut them in different
shapes. Chances are, one of them would
fit and open the door of a particular house. In
this case, an antibody needs to fit an antigen
– a protein on the surface of a bacteria, virus,
or another invader. If it succeeds, the cell that
carries it can trigger a reaction from the
immune system.
This usually happens as a cooperative effort
involving B and T cells and another type called
105 Part Two: Identity crisis
dendritic cells. This latter kind of cell is stationed
in the skin and other tissues that constantly
come in contact with the environment
– the lungs, stomach, and intestines – where it
is usually first to come in contact with foreign
organisms and substances. It chews them up
and glues fragments of their molecules to its
surface, bound to molecules called MHCs,
which stands for the major histocompatibility
complex. Slight differences between the structures
of these molecules in different people
are important in helping the body distinguish
between native and foreign cells. They are
largely responsible for the immune system’s
rejection of transplanted cells or organs.
As they develop, T cells are outfitted with
receptors that can recognize combinations of

MHCs and foreign molecules. As with antibodies,
these receptors are made in a random way.
If a T cell encounters a dendrite that one of its
receptors can recognize, and it then meets a B
cell with a matching antibody, the B cell
begins to divide very quickly. This creates a
huge number of identical daughter cells that
produce and secrete the antibody. They glue
themselves onto the surface of the invader,
acting as a sort of alarm beacon. This summons
macrophages and other types of white
blood cells that digest the foreign object and
break it down. (Anyone who has seen the
movie Fantastic Voyage will remember
macrophages as the huge cells that swallow a
miniature submarine.)
“Why don’t B and T cells attack the body?” I
ask Martin. “What keeps B cells from making
‘keys’ that fit proteins on the surfaces of their
neighbors?”
That happens all the time, he says; billions of
the receptors and antibodies are made that
would behave that way, if they were released
into the body. But B cells that bear self-recognizing
antibodies are usually destroyed before
they leave the bone marrow.
“The process for T cells is quite different,”
Martin says. “During their development in the
thymus, they have to pass two types of tests.
The first is to make sure that they can bind to
MHCs. Cells that can’t do this wouldn’t be
much help to the immune system, so they are
destroyed. Those that do survive undergo a
second test, to see if they bind to combinations
of MHCs and human proteins. If a T cell
that did this got loose, it would probably start
an autoimmune reaction. So in the second
round, the cells that ‘succeed’ are trapped and
destroyed. In the end only about two percent
of T cells pass both tests.”
The B and T cells that survive these rites of
passage are now mature. They circulate
through the bloodstream and travel to lymph
nodes, meeting points that bring them into
contact with dendritic cells. There the T cell
may meet a foreign protein it recognizes,
which makes it reproduce and develop. Some
of the T cells leave the node to find and digest
the invader. Others become the helper cells
that activate B cells. B and T cells that aren’t
activated exit the node and pass back into the
lymph system. They will continue to cycle as
long as they live, until they meet a cell they
recognize. New dendrites continue to arrive in
the lymph nodes. Like scouts sent ahead
down slow rivers, they continually look for
threats, then bring the news back home.
��
While working in Uganda in the 1950s,
the Irish physician Denis Burkitt
noticed a disturbing pattern of disease
among children. One day a mother brought in
a very ill child with swollen lymph nodes
under the jaw. Burkitt’s diagnosis was cancer
of the lymph system. It was a tragedy for the
child, and the physician had an additional reason
to be worried – he had seen the same
thing before. He began digging through medical
records and made the alarming discovery
that there seemed to be an epidemic of such
cancers. The victims were almost always children,
and the tumors nearly always appeared
in the same place, under the jaw. In 1958
Burkitt published his findings in a medical
journal, and the world became aware of the
disease now known as Burkitt’s lymphoma.
Three years later the British Medical Council
gave him a grant – 250 British pounds, which
today seems ridiculously small – to study the
extent of the disease. Burkitt and two other
physicians set off on an epic journey in which
they traveled over 15,000 kilometers. At the
end the picture was clear: the tumors
appeared almost exclusively along the equator,
in regions that were also threatened by
malaria and yellow fever. Over time Burkitt
figured out the connection. Other infections
had weakened the children’s immune systems,
making them easy targets for a very
common type of herpes called the Epstein-
Barr virus.
Part Two: Identity crisis
106
Most people across the world are infected
with this virus, which invades B cells. Usually
it lives inside the cells without causing any
harm. If it threatens to break out, other
immune cells can handle it. But that can
change quickly – especially if the immune system
has been challenged by another disease.
So this type of cancer does appear in other
parts of the world that are constantly beset
with epidemics. Fortunately, Burkitt and his colleagues
were able to develop a cure, a type of
chemotherapy using a drug called cyclophosphamide.
It is so effective that a single treatment
sometimes causes the tumors to vanish.
In the 1970s Friedrich Luft (see the next chapter
and “the Case of the Short-fingered
muskateer”) heard Burkitt give a talk about
one of his other interests: the connection
between diet and cancer. Burkitt began a
campaign to get people to eat more fiber. “He
observed that colon cancer was inversely
related to stool bulk,” Friedrich says. “This
observation formed the ‘roughage’ hypothesis

to combat colon cancer. Needless to say, he
showed some slides of things you normally
don’t see in a scientific talk.”
B cells are especially susceptible to cancer
because of their ability to divide rapidly –
always a risky business because the process
may get out of control. In the early 1990s,
when Martin was working at the Ludwig-
Maximilian University in Munich, he decided
to try to find out what made healthy B cells
develop cancer. With his colleagues he compared
the genes activated in healthy and diseased
cells and discovered three candidate
genes that were being used to create proteins
at much higher levels in the tumors. “We had
to pick one,” he said, “so we picked the most
promising one.”
The choice was a good one: later the team discovered
that the protein’s structure closely
resembled a common type of receptor called a
G protein-coupled receptor (GPCR). These proteins
are usually involved in passing crucial
signals into cells. In the eyes, GPCRs transform
light into signals that can be passed to the
brain.
“At the time only a handful of GPCRs had been
cloned molecularly,” Martin says.“ Most were
known as hormone and neuropeptide receptors,
which transmit signals that trigger cardiac
and metabolic functions. Other GPCRs
were thought to regulate the movement of
white blood cells from the blood into tissues
under inflammation, but they had not yet
been identified.” The protein that Martin had
picked and provisionally named BLR1 (for
Burkitt Lymphoma receptor 1) was the first
GPCR identified in lymphocytes and might
have some of these functions in B cells. To find
out, Martin and his colleagues developed a
strain of mice without the blr1 gene. He
brought the animals along when he moved to
the MDC in 1994.
“We had knocked it out and discovered that
animals without the molecule had several
unusual features,” he says. “They were lacking
some of their lymph nodes entirely. Those
107 Part Two: Identity crisis
they had were not organized correctly – cells
didn’t migrate properly within them. This
made it clear that the protein was playing a
role in at least two processes: directing cells to
their proper locations and helping the lymphoid
tissues form.”
The group wrote up a paper and sent it off to
the prestigious journal Cell. The editors were
hesitant to take the paper. “Everything was
fine – they were convinced that we had found
an important signaling molecule, guiding B
cells into and through lymphoid organs,”
Martin says. “But we didn’t know what external
signal triggered it. Their problem was they
had never published a paper where the receptor
was known, but not its ligand. We wrote
back, ‘The changes we see in the bodies of
these mice won’t be any different once the
ligand once is identified.’ And then they sent
the paper out for review.”
Later the protein would be called CXCR5, for
chemokine receptor 5. “But it only got this
name in 1998, when we knew what it was

inding to,” he says. Independently, two
groups were able to identify its partner, a
chemokine called CXCL13, which triggers
migration of B cells through CXCR5.
But understanding exactly what the
chemokine receptors did would take several
more years of work and the discovery of a second
receptor protein called CCR7. “We wanted
to see if T cells contained similar receptors, so
we looked for a gene similar to CXCR5,” Martin
says. “CCR7 was what we found.”
When the scientists developed mice without
this second receptor, they found further
abnormalities in lymph nodes and the lymph
system. Putting the two scenarios together
gave them a new view of how cells move
through the lymph nodes and how this
affects the immune system. To explain, Martin
needs to give me an overview of what normally
goes on inside lymph nodes, and to do
that he reaches to the top of one of his filing
cabinets and pulls down a diagram that looks
forbiddingly complicated. He promises to
keep things as simple as possible.
He reminds me that lymph nodes are
the meeting points for dendritic cells that are
usually the first to encounter foreign substances
and organisms, and the T and B cells
that actually mount the immune response. “To
understand how these cells interact
and their importance, we have to look at the
structure of a node,” he says. “First, there is
the border of the node itself. To cross over, all
three kinds of cells have to have the
CCR7 receptor. It’s attracted to a powerful
signal, a chemokine called CCL21, which is
produced by cells in the node. Without
the receptor there’s no attraction and no entry.”
Inside the perimeter is a region called the T
zone. While it is mostly populated by T cells, all
Part Two: Identity crisis
108
three types enter here. There is a second compartment
inside, predictably called the B cell
follicle because it is a breeding ground for B
cells. A few T cells and dendrites also make it
inside. The entry key for this compartment is
the CXCR5 receptor.
Most immune responses depend on all three
kinds of cells, he reminds me, and they usually
meet up in the T zone. “When B and T cells
enter, they are usually ‘naive’ – they haven’t
yet encountered a substance that matches
their antibodies or receptors,” he says. “But
some of the dendrites that come in bear fragments
of foreign molecules. In that case, the
first thing that may happen is for the dendrite
to be recognized by a T cell. That triggers the
cell to start making copies of itself and also to
develop in various ways. Some of the T cells
leave the lymph node and start hunting whatever
it is that bears the antigen. If they find it,

they have various ways of destroying it. Other
T cells, called T helpers, move to the edge of
the B cell follicle. If they find a matching B cell
– one with an antibody that fits what their
own receptors have found on the dendrite –
the B cell is activated.”
It moves into the follicle, where it begins to
reproduce. Here it gets assistance from another
type of T cell – one which Martin discovered
and christened as a B follicular helper T cell.
“These cells are special because once they
have been activated, they lower the amount
of CCR7 they produce and build CXCR5 receptors,”
he says. “This allows them to cross the
compartment boundary. They have a quality
control function. Sometimes the antibody on
a B cell isn’t a very good match for the antigen.
The T cell sorts those out and they are
destroyed before they begin making defective
antibodies.”
Cells that pass the test reproduce very quickly.
Now, instead of making unique, individualized
antibodies, the cells turn into copying
machines for molecules that can bind to the
invader. They leave the follicle and the lymph
node and secrete the antibodies. These attach
themselves to antigens on the intruders,
which are now attacked and destroyed by T
cells and other white blood cells.
The array of different types of cells is bewildering.
Are there more to be found? I ask.
“I’m sure there will be,” he says. “This is our
current view of how things work. There are
still questions. But our lab has filled in several
of the key steps in what we know, and those
findings have become widely accepted. It’s
gratifying to see that some of the things you
discover become incorporated into the textbooks.”
��
In the corner of Martin’s office is a palm tree
that seems to have designs on the whole
room – it has stretched leafy branches over his
printer and half his desk. Next to his computer
monitor is a small Buddha, the size of two
clasped hands, sitting placidly in equilibrium
atop an old mechanical postal scale. The statue,
he explains, was a gift from two former lab
members from India who married and left.
The scale once belonged to Feodor Lynen, a
Nobel prize-winner who headed the Institute
of Biochemistry at the University of Munich. It
was still standing there years after his departure,
when Martin joined the institute in 1982,
and it was about to be replaced by a digital
scale. So it has accompanied Martin ever
since.
One urgent reason to study the development
of the immune system and how its cells
behave is to understand how – in spite of the
body’s safeguards – they participate in
autoimmune diseases. An increasing number
of diseases are now recognized to have an
autoimmune component. Martin also hopes
that the work will help explain cancers like
Burkitt’s lymphoma which affect the lymph
system. Here, too, there is mounting evidence
for a connection. More than a century ago
Rudolf Virchow postulated a connection
between chronic infections and cancer. The
link may well be the immune system.
Martin’s studies underline the connection
between the behavior of cells, the development
of structures in the body, and disease.
Without key receptors such as CXCR5 and
CCR7, mice did not develop proper lymph
nodes. Without the complex architecture of
the nodes, B and T cells might not develop and
interact properly. If that didn’t happen, some
of the cell types might be lacking, and crucial
parts of the immune system might fail.
These themes have come together, for example,
in a study headed by Uta Höpken, a senior
member of Martin’s group. She is now a
Helmholtz fellow with her own PhD students.
One of her interests is the behavior of T cells
outside the lymph system.
“We have learned quite a bit about the guidance
of immune cells in the nodes and secondary
lymph organs,” she says. “But these
cells spend part of their lives outside, in the
109 Part Two: Identity crisis
bloodstream and surrounding tissues, before
reentering the lymph system. That is an
important part of their surveillance functions
– it’s where they usually find foreign
microbes. And this type of circulation is also
known to help the T cells of newborns learn to
distinguish self from non-self. Yet we don’t
know much about what guides their migration
and behavior outside the lymph organs.”
Thinking that chemokine receptors might be
involved, Uta and her colleagues looked carefully
at the mice without CCR7. They discovered
massive accumulations of T cells outside
lymph vessels. In the stomach, intestinal tract
and lungs the cells collected in follicle-like
bulbs. “These are tissues that come in contact
with substances from the environment – food
and air – and so they play a special role in the
immune system,” Uta says. “Normally the
stomach and gut have a few follicle-like
lymph structures, but not as many, and not as
large. The animals still have some immune
defenses, but the cells don’t migrate properly.”
As the mice aged, they developed symptoms
that looked very much like a human illness
called Ménétrier’s disease. In this condition,
masses of wrinkled tissue form on the walls
of the stomach. Eventually the stomach
grows much larger and becomes covered with
large folds – until it almost looks like a brain.
The disease is rare and usually affects middleaged
men; it often leads to ulcers, and people
who suffer from it have a higher risk of stomach
cancer.
“We think this means there is a strong connection
between the formation of these
unusual follicles and the way the lymph system
normally develops,” Martin says. “Our
main approach to investigate this has been to
remove the receptors and block the signals.
But we’ve also learned from studies of other
groups that take the opposite approach –
increasing the strength of signals.” For example,
two laboratories in the US forced cells in
unusual places to produce high amounts of
CXCL13 or CCL19, the molecules that attract

A CD 3
CXCR5 ond B cells and CCR7 on T cells, repectively.
As a result, masses of the cells accumulated
and formed follicle-like structures similar
to those seen by Martin’s lab. But they
were not active. The cells organized themselves,
but the stimulus from the immune
system was missing.
These ectopic (“out-of-place”) follicles also
resemble structures that are seen in several
human autoimmune diseases, including
rheumatoid arthritis. “About half the people
who suffer from arthritis develop follicle-like
lymphoid structures in unusual places,” Martin
says. “They grow ectopically in the joints, and
most likely trigger aberrant immune responses
causing pain, bone destruction and finally
immobility. If they really are similar to the follicles
we have seen in mice, they might arise for
some of the same reasons. The mice could give
us a good system to investigate the mechanisms
that cause the disease.”
Previous mouse models of arthritis have not
accurately reflected the chronic course of the
disease, he says. A substance called mBSA
causes chronic inflammations of the animals’
joints, but it is artificial and the mechanisms
by which it operates probably do not reflect
what happens in the real disease. Another
strategy has been to inject a molecule called
collagen, derived from cattle, which provokes a
systemic arthritis lasting for only three weeks.
B B220 C PNAd
D CCL21 E CXCL13 F BrdU
But this method only works in a strain of
mouse which is hard to investigate because of
a lack of genetic tools.
PhD student Antje Wengner started working
on the theme using the mBSA method of creating
arthritis in animals that lacked CXCR5
and CCR7. She was hoping to peel apart the
links between inflammations, immune
responses, and arthritis. Antje followed the
animals’ health for a very long time – over
nine months. All the genetically normal mice
treated with the substance develop ectopic
follicles that are typical of arthritis, and it
takes a long time – three to six months – for
them to arise.
“In rheumatoid arthritis patients, large numbers
of T cells, B cells, and other immune system
cells build up in the regions around the
joints,” Martin says. “At different times each of
these cell types has been accused of being
responsible for the joint destruction seen in
arthritis. B cells are particularly good suspects
– therapies that eliminate them have had
impressive successes.”
Another sign of their involvement is that
inflammed tissues in many patients produce
high levels of the protein CXCL13. This is the
target molecule for CXCR5 receptors on B cells,
drawing them to B follicles and other places. If
B cells collect around the joints, they can stim-
Part Two: Identity crisis
110
These images show eptopic follicles (circular
structure in the middle) obtained from a mouse
with artificially induced arthritis. In the image
on the top left, T cells are tagged in red. They
gather in a different region of the follicle than
B cells (tagged red in the image at the top, in
the middle).
ulate the formation of local lymph follicles
with clear zones for T and B cells, which is a
hallmark of a functioning lymph node-like
structure. “Building a lymph structure on-site
might allow it to cope with a prolonged, local
inflammation. Instead, it develops into a continuous
autoimmune problem.” Martin says.
“We wondered what would happen if that
structure didn’t form. Maybe it was a necessary
step on the way to the prolonged inflammation
and tissue damage that happens during
rheumatoid arthritis.”
Since the loss of receptors disrupts the structure
of follicles, Antje wondered what would
happen if she induced arthritis in mice that
didn’t have CXCR5 and/or CCR7. Using both
methods of stimulating arthritis in the mouse
allowed her to watch the immune system’s
response to inflammations in the early phases
as well as later, chronic stages of the disease.
The first phases were similar in mice with and
without the receptors; the researchers discovered
local inflammations and an influx of
immune system cells. But mice lacking the
receptors had fewer T cells that recognized
the foreign substances. And over the long
term, there was much less bone damage in
animals lacking the CXCR5 receptor.
Additionally, the lack of the receptors prevented
the formation of functioning lymph follicles
around the joints.

Martin’s lab is continuing the project by challenging
the mice with multiple methods of
inducing arthritis. “In the novel model, it only
takes four to six weeks for the mice to develop
the full-blown symptoms of the chronic
disease, and 95 percent of them develop
active ectopic follicles,” Martin says. “This will
make a much more efficient model to study
the progress of arthritis and to grasp the contributions
of different receptors and cell
types.”
But the recent results are already quite promising.
“This suggests that in many cases,
CXCR5 may play a central role in some of the
most chronic destructive effects of rheumatoid
arthritis,” Martin says. “Finding a way to
shut it down locally might be a very effective
strategy to treat patients who suffer from this
and other autoimmune diseases.”
��
There is a diagram of the lymph system in
my daughter’s eighth-grade biology
book. When we look at it together, she says it
reminds her of the streams and rivers whose
names she has been learning in geography
class. Soon there will be a test and she will
have to draw and label the Rhine, the Neckar,
and dozens of other waterways on a blank
map.
We normally think of our bodies as well-built,
finely tuned machines that have to be maintained
through diet and exercise. We watch
ourselves grow and age, but just as we seldom
see rivers change their course, we aren’t
usually very aware of the dynamic nature of
our inner world. Martin and his colleagues see
the lymphatic system as a evolving structure
that grows and changes through its activity
and challenges from the outside. Maybe it
shouldn’t be any surprise that new lymph follicles
spring up at the site of a chronic inflammation.
After all, the same thing happens on
a map, over long periods of time. Villages also
arise at the places where rivers or roads run
together, where the train stops, where the
111 Part Two: Identity crisis
mail is delivered. Battles are fought there, and
the future is often decided at such places.

Lost anchors
and wrecked vessels
Iheard that Sibylle von Vietinghoff, a house-staff physician at the
MDC and the Charité, was investigating a disease involving B cells
and other white blood cells called neutrophils, and an autoimmune
reaction in which the body builds antibodies against one of its own
molecules. In this case, the target is a protein found in and on the neutrophils.
She worked in the groups of Ralph Kettritz and Friederich Luft,
clinicians who treat patients while running laboratories concerned with
the mechanisms that underlie disease. It seemed like an ideal position
for a resident in internal medicine to get an exposure to both worlds,
good preparation for a day when medicine and basic research work
much more closely together. But it’s a challenge to serve both masters.
Budding clinical scientists like Sibylle have the double burden of spending
time in the lab as well as working long, intensive shifts in hospital
wards, so it was hard to find time to meet. In March Sibylle was finishing
up a round of experiments while getting ready to start a six-month
rotation in the intensive care unit.
In spite of the stress, she found a spare hour on a beautiful day in early
spring. I took a ten-minute bus ride through Berlin-Buch to reach her
lab in the Franz Volhard Clinic, which specialized in hypertension and

Ralf Kettritz and Sibylle von Vietinghoff
systemic vascular diseases. The clinic has now
moved into the Helios Krankenhaus Berlin-
Buch complex, directly across the road from
the Berlin-Buch campus. But in the spring it
still resided in House 134 of “Area 1,” one of five
large hospital grounds scattered throughout
the village. Most of the other clinics in Area 1
had already moved, leaving behind an array of
elegant, aging buildings, like a ghost town of
the history of medicine.
The buildings are separated by large, park-like
spaces with gazebos and ornate fountains.
Children of some of the staff were running
around, climbing on things, peeking into
buildings. Some of the houses had evidently
been empty for a long time. Sunlight cast odd
shadows through the broken windows, into
old hospital wards, nurses’ stations, cafeterias.
Other structures have been carefully maintained.
An elderly couple was having lunch on
a sun-washed balcony of one house. They
reminded me of an anecdote from one of
Heinz Bielka’s books on the history of Berlin-
Buch: the town was home to the first senior
citizens’ home in Germany that permitted
husbands and wives to move in together.
I shouldn’t have been surprised – but was – to
come through the front door of the Franz
Volhard Clinic and find myself in the waiting
room for patients. A nurse at the front desk
provided directions to the labs, which were in
the basement. Another surprise was to find

Sibylle in the white shirt and slacks of a physician.
There’s no dress code for molecular biologists,
unless you count the white lab coat
they pull on to do some types of experiments.
Sibylle’s outfit – and the stethoscope around
her neck – were more reminders of a culture
gap. From time to time she looked at her
watch. She apologized. After we talked she
had to make rounds on a ward upstairs.
One of the projects of Ralph Kettritz’s lab
involves the way that white blood cells defeat
pathogens, and how this process sometimes
leads to an autoimmune disease. White blood
cells, or leukocytes, squeeze through the lining
of blood vessels to seek out viruses, bacteria,
and infected cells that have been marked by
antibodies. When a cell finds one, it destroys
the pathogen by releasing a burst of oxygen
atoms and toxic enzymes.
Sibylle compares the effect to a molecular
hand grenade. “The oxygen atoms are highly
charged with energy and that overwhelms
the invader. Its molecules become overloaded
and start to carry out too many reactions;
they spin out of control. This tears the
pathogen apart.”
The reactions have to happen at the right
time and place to avoid injuring the body.
There are nearly always healthy cells around
when the bursts happen, so they have safety
devices: they possess molecules that can
break down the overloaded oxygen, or they
enlist the help of antioxidants such as vitamin
C. But in a rare disease called systemic vasculitis,
the bursts are triggered too early, when
the neutrophils are still in the vessel or in the
vessel wall, and have not reached yet the site
of inflammation. This premature reaction
injures the blood vessels and eventually
destroys the organs they supply with blood.
“This is one thing that happens in autoimmune
diseases,” Sibylle says. “If a leukocyte
Part Two: Identity crisis
114
has been taught to target a healthy human
cell, it will treat it as a pathogen and try to
destroy it. That will obviously lead to inappropriate
bursts and collateral damage.”
One form of vasculitis is an auto-immune disease
that occurs when the body builds antibodies
(called ANCA) against proteins which
appear on the leukocytes themselves. Sibylle
and her colleagues in Ralph Kettritz’s lab
traced the problem back to bad behavior on
the part of two proteins.
Ralph knew that ANCA-vasculitis patients
develop antibodies against a protein called
PR3, which is produced only in neutrophils.
Most PR3 is kept out of reach, glued to internal
compartments of these cells, but occasionally
it appears on the outer surface. In
ANCA-vasculitis, much more of it moves from
the interior to the outer membrane. If the
body builds antibodies against it, the result is
a cycle of autoimmune reactions that eventu-

ally become deadly because of severe organ
damage. Neutrophil activation by ANCA is
central to this tissue injury.
“The therapies we currently use to treat
ANCA-vasculitis block the body’s ability to
manufacture antibodies against PR3, but they
also shut down the production of a wide
range of other antibodies,” Ralph says. “That’s
a very aggressive tactic, because it weakens a
person’s defenses against other diseases. The
thing to do was to try to figure out why PR3
was migrating to the surface of the cells.”
Samples of white blood cells were obtained
from healthy people and then compared to
samples from patients with ANCA-vasculitis.
As expected, the patients had more PR3 on the
surface of leukocytes. “Then we caught a lucky
break,” she says.
One of the patients on her clinical ward had
an unusual blood disease called polycythemia
rubra vera (PRV) – basically a form of
cancer – in which certain types of precursor
cells in the blood marrow divide too often.
One sign of the disease is an overabundance
of a protein called NB1 on the surface of blood
cells.
“At about the same time another patient
turned up with an equally obscure disease
called paroxysmal nocturnal hemoglobinuria,
or PNH,” Friedrich Luft says. “These patients
have red blood cells that burst in the night – a
tell-tale sign is blood in their urine in the
morning. Well, the white blood cells of these
patients never have NB1 on their surfaces. So
here was one patient with too much NB1,
another with none. We thought that was
interesting, so Ralph and I demanded that
Sibylle learn everything ever known about
NB1. And as things turn out, NB1 would end up
being pivotal in a third disease, the one Sibylle
wanted to study.
115 Part Two: Identity crisis
“This vignette is important because it reveals
the ingredients that are required for clinical
research to work,” Friedrich says. “You need to
start with curious, industrious young clinicians
who want to be scientists. Add to that
mentors who have a successful (and funded)
laboratory that can embrace young minds.
And occasionally there is an old chief of service
who has seen almost everything and
made a lot of academic blunders, who can
guide decisions and make things happen.
Sibylle made a ‘triple play’ in this regard.”
Healthy cells have to defend themselves from
the bursts of white blood cells, and part of the
protection comes from proteins on their surfaces
that can deal with excess oxygen. Many
of these proteins are attached to the membrane
by anchors made of fat molecules
called lipids. NB1 is one such protein. In PNH a
lipid anchor called GPI is defective, particularly
in red blood cells. It can no longer hold onto

NB1, and without an anchor the protein drifts
away. When that happens the cell loses an
important piece of protection against damage
from leukocytes and it is chewed apart by
the immune system.
The patient with PNH had, as expected, no
NB1 in the membranes of cells. Then the scientists
thought to check for the presence of PR3,
the protein involved in ANCA-vasculitis – and
didn’t find any. Because the two proteins were
both involved in autoimmune diseases, there
might be a connection. Maybe they worked
together in the cell.
No one had checked NB1’s expression in
ANCA-vasculitis, so Sibylle and her colleagues
did so. “We discovered that NB1 was behaving
just like PR3; in the disease, both molecules
move to the membrane in higher amounts
than normal. Particularly interesting was that
the percentage of neutrophils that express
both molecules on the outer cell membrane
was higher than in controls. We marked the
molecules with fluorescent tags and watched
them under the microscope. You can’t see
individual proteins under the microscope, but
you can see the patterns that lots of them
form, and they seemed to end up in the same
positions on the surface of the cell.”
Such close associations often mean that two
molecules are bound to each other. Did the
migration of PR3 somehow depend on NB1, or
vice-versa? It was hard to tell. In diseases
many things go wrong, and what looked like a
connection might be a coincidence caused by
something else. Interfering with NB1 might
give some answers, so Ralph’s laboratory used
an enzyme to cut NB1’s anchor in white blood
cells. Without this protein, PR3 disappeared as
well. Then they tried the opposite approach.
They took a cell that didn’t produce either protein,
and made it express NB1 and PR3, together
and separately. PR3 did not appear on the
membrane when it alone was made by cells.
However, when PR3 was co-expressed with
NB1, suddenly both proteins appeared on the
surfaces of cells.
“This was great evidence to support a direct
connection between the proteins,” Ralph says.
Part Two: Identity crisis
116
Microscope images revealed that NB1 (green) and
PR3 (red) take up the same positions when they
move to the surface of leukocytes. PR3 can only
move there with the help of NB1. Normally PR3 is
kept within the cell, but in ANCA-vasculitis patients
the two molecules appear on the surface at a high
rate. The body builds antibodies against PR3,
causing a dangerous autoimmune reaction.
“PR3 apparently needed NB1 to get to the
membrane and stay there, because when we
cut the anchor both proteins disappeared.
This finding explained what we had seen with
the PNH patients. They had lost their NB1, just
as if the anchor had been cut, so their mambranes
had no traces of either NB1 or PR3.”
ANCA-vasculitis is an unusual disease, but a
very serious one, and the connection between
PR3, NB1 and its lipid anchor gives scientists
new starting points to look for treatments. If
PR3 is directly responsible for the dangerous
symptoms of the disease, maybe scientists
can find a way to cut NB1’s anchor, or break
the connection between the two molecules.
Most autoimmune diseases are hard to treat
because it is hard to separate the healthy
functions of the immune system from
unhealthy ones. One strategy has been to use
drugs that shut down parts of the immune
system, but this leaves the body open to
attacks from pathogens. Ralph and Sibylle
know that finding the fundamental causes of
a disease is only the first step toward a thera-

117 Part Two: Identity crisis
py. But people who are equally comfortable in
two cultures are always the best to solve
problems of translation.
And chance plays a role – if a scientist and her
institute are prepared to take advantage of it.
“Uncovering a crucial mechanism in this case
stemmed from the fact that Sibylle cared for
both the PRV patient and the PNH patient,”
Friedrich says. “This clinical experience will
surely guide her future work. Ironically,
Sibylle’s findings are not likely to help patients
with either PRV or PNH. Both patient groups
can now look forward to acceptable and effective
therapies. Instead, her work may benefit a
completely different group, patients with PR3
disease. The link to NB1 may break open PR3
research.”

photo by Wayne McLean
A pact with the Devil
It’s not often that a cartoon character steps up to shed light on the
immune system, cancer, and the potential of new therapies involving
T cells. But the Tasmanian devil, best known as a star of children’s
television, has suddenly caught the interest of scientists across the
world. The real marsupial lives only on the island nation of Tasmania,
where its odd shrieks startle tourists and keep people awake at night.
Several hundred years ago the devil was also widespread on mainland
Australia, but it became extinct there with the arrival of the dingo. Now
it faces another threat. In 1996 scientists discovered an epidemic of cancer
that was killing the animals in a northern region of the island. Since
then the disease has spread to most of the island, so rapidly that
researchers estimate it may have destroyed half the entire species.
Tumors arise in the animal’s mouth and on its face, eventually making
it impossible or too painful for the devil to eat, and it starves to death.
What has been so puzzling about the disease is that it seems to spread
directly from animal to animal. The devils often have fierce fights in
which they bite each other on the face. This explains how cells might
move between animals – but not why they cause a tumor in the animal
Part Two: Identity crisis
118
Tasmanian Devil in defensive stance,
at Tasmanian Devil Conservation Park,
Tasman Peninsula.
Use under the Creative Commons Attribution
ShareAlike 2.5 License.

Thomas Blankenstein

that has bitten. Cancer cells aren’t like parasites,
able to jump between hosts and directly
cause a new disease.
Thomas Blankenstein of the MDC and the
Charité speaks of the devil during talks on
cancer and the immune system, because the
behavior of its cancer is an almost absolute
exception. “Everything happens in nature,” he
says. “But as far as we know this is only the
second case ever of a directly transmissible
cancer. Normally you could inject billions of
tumor cells into animals – if they didn’t come
from populations that had been heavily
inbred in the lab – and those cells would be
immediately rejected by the immune system,
like any other foreign cell.”
Several other types of cancer are thought of
as “transmissible” for a different reason. They
are linked to infections, the way the Epstein-
Barr virus leads to Burkitt’s disease (described
in the story “Where slow rivers meet”) or the
way the human papilloma virus triggers
tumors in the cervix. In those cases viruses
bring along genes that disrupt the cell’s own
molecules, causing it to spin out of control.
The opposite scenario – that cancer cells
themselves behaved like parasites, colonizing
a new host and then reproducing – had only
been seen once before, in a sexually transmitted
disease affecting dogs.
So the community was skeptical about the
transmissible cancer hypothesis until 2006,
when two Tasmanian scientists, Anne-Maree
Pearse and Kate Swift, examined cells from
animals with the devil-facial tumor disease.
They discovered some extremely odd characteristics:
cancer cells were missing five of the
normal devil chromosomes and had four
Part Two: Identity crisis
120
extra ones. It was extremely unlikely that such
massive changes in the genome could spontaneously
occur more than once, but tumor
cells taken from animals all across the country
had the same features. This virtually proved
that all of the cases could be traced back to
one source, rogue cells that evolved in a single
animal. Like parasites, they were being passed
from one devil to the next in a bite.
��
The immune system has evolved to fight
bacteria, viruses, and other infectious
agents. Likewise, animal immune systems
normally recognize cells from other animals
as foreign and reject them, whether those
cells come from tumors or not. But devils
don’t reject the tumors or many other types
of cells transplanted from other members of
their species. Usually this only happens

etween identical twins, and Pearse and Swift
proposed that heavy inbreeding within the
population had produced devils whose tissues
were behaving like those of twins. A
2007 study by other labs in Australia and
Tasmania confirmed this and showed what
aspect of the immune system was at fault.
Most devils are so closely related that their
cells produce very few types of MHC molecules.
(These are the “vacuum sweeper” proteins
on the surfaces of cells that combine
with fragments of foreign molecules,
described in “Traps along slow rivers.”)
Receptors on T cells recognize combinations
of MHCs and these fragments.
Most species have three types of MHC 1 proteins,
each of which picks up a different kind
of fragment. And mutations and evolution
have produced an enormous number of “flavors”
of each type. Normally this variety plays
an important role in the robustness of a
species’ immune system. Since inbreeding has
impoverished that system in devils, they can’t
protect themselves from invasions by foreign
cells.
A normal healthy immune system is so critical
that it sometimes even targets “self” and produces
autoimmune diseases. “This raises the
question,” Thomas Blankenstein says, “of
whether the immune system recognizes cancers
which arise spontaneously. The idea that
Anna Kruschinski
it should was a very attractive hypothesis
when it was first proposed. A problem with
very attractive hypotheses is that sometimes
they hang on even when there is evidence to
the contrary.” Or as another scientist named
Thomas – Thomas H. Huxley – once put it,
“The great tragedy of science – the slaying of
a beautiful theory by an ugly fact.” He meant
it ironically.
Thomas Blankenstein is a lanky man with
short-cropped black hair and glasses. If I didn’t
know what he did, I could imagine him roaming
gullies and desert ridges, digging for
dinosaur bones. Instead he spends his days in
the laboratory digging into the immune
system.
The themes he is working on stretch back over
a century, to Paul Ehrlich’s realization that
human cells likely underwent mutations all
the time. Why didn’t more of them lead to
cancer? He reasoned that the body had to
have a system for dealing with them. When T
cells were discovered in the 1950s, their func-
121 Part Two: Identity crisis
tions were unknown, but their behavior suggested
that they were involved in immunity.
Frank McFarlane Burnet and Lewis Thomas
proposed that the role of T cells was to block
tumors. The idea caught on and persisted as
researchers unraveled the many functions of
T cells in fighting infectious diseases.
Most tumors begin when a human gene
undergoes mutations. That happens more
often the older a person gets, and it usually
happens long after the immune system has
learned the difference between self and notself.
It is logical to suppose that the mutated
form of a gene (and the new protein it
encodes) would look foreign to the body’s surveillance
system and thus trigger an immune
response the way viruses or bacteria do. This
was the standpoint of the “attractive hypothesis.”
Since the body doesn’t reject all tumors,
however, the hypothesis needed a second
part: a mechanism that allowed cancer to
escape immune control. A lot of laboratories
began working hard to find it.
The work of Thomas and many others suggests
that the body treats tumors caused by
viruses as infections; it handles those that
arise spontaneously, usually late in life, differently.
“Work by Georg and Eva Klein carried
out 30 years ago clearly demonstrated that
the immune system can cope with some
kinds of tumors,” he says. “For example, T cells
can very efficiently control B cell lymphomas
that arise in association with the Epstein Barr
virus. They also control other tumors associated
with infections. This hinted that it should
be possible to vaccinate animals against

tumors if a clearly defined target – an antigen
specific to the tumor – could be found. And
this is the case.”
These findings make sense in the light of evolution,
he says. The immune system is the
product of natural selection, which helps
shape organisms up to the point that they are
finished bearing offspring. The immune system
has evolved to fight infectious diseases –
including those related to cancer – because
they attack young mammals. But there is no
pressure to protect organisms from old-age
diseases, like most types of cancer that arise
from spontaneous mutations. So these are
probably fundamentally different kinds of diseases,
and it makes sense that they should
evoke different reactions from the body. That
was what the Kleins’ experiments showed,
and it might have put the Burnet hypothesis
to rest for good.
In 2006 Georg Klein wrote to Thomas and
said he would like to visit the MDC as part of
a trip he was making to Berlin. It was a rare
opportunity; Klein is in his eighties and no
longer travels very much. “We quickly organized
a seminar,” Thomas said. “It was such
short notice that the only time we could find
was early on a Saturday morning. You can
imagine how eager people are to come back
to the campus at such a time, after a long
work week. But the room was completely full.”
The Kleins’ hypothesis received further support
in the mid-1970s with the discovery of
the nude mouse. This strain of animal had
arisen through a chance mutation. As well as
having no hair, the animal had no thymus –
thus no T cells, and an immune system that
couldn’t cope with infections. The animals
must be kept in a carefully controlled environment
or they quickly die.
“Here was another chance to pit the Burnet-
Lewis hypothesis against that of the Kleins,”
Thomas says. “Nude mice undergo just as
many cancer-causing mutations as other
strains. Now suppose that the adaptive
immune system played any role in suppress-
Part Two: Identity crisis
122
ing such tumors. You’d expect, obviously, that
nude mice would experience far more cancer
than other strains. But that doesn’t happen.
The conclusion that we draw is that T cells
don’t fight spontaneous tumors.”
A higher rate of tumors has sometimes been
found in animals that have been immunesuppressed
their entire lives. But this may well
be due, Thomas says, to unknown “latent”
infections or other environmental factors that
play a role in the growth of tumors. “This
wouldn’t be surprising,” he says, “because it
supports a hypothesis of Rudolf Virchow, who
discovered that cancer arises from defective
cells. Virchow noticed a connection between
long-term infections and cancer, and he was
absolutely right.”
In spite of such evidence, however, the idea
that cancer provokes and then escapes an
adaptive immune response has once again
reared its “attractive” head – mostly on the
basis of experiments with mice that had deficient
immune systems. And in many of these

models, researchers used chemicals to induce
cancer. Those results have to be interpreted
with caution, Thomas says.
“To demonstrate that the immune system is
responding specifically to a tumor you need to
find T cells which have been triggered by an
antigen that belongs to the tumor,” he says.
“Since no unique antigens had been identified
in those tumors, you cannot prove this.”
So many of his lab’s recent efforts can be
summed up this way: Does our immune system
offer any specific protection from tumors
that arise spontaneously, from random mutations
in genes? And do tumors escape by
interfering with mechanisms of immunity? It
has been challenging to phrase these questions
in the language of rigorous experiments,
particularly since most laboratory
models of cancer immunology have relied on
chemicals, or transplants of tumor cells
between mice. With any transplant, there is a
risk that the immune system will treat cancer
like an infection. The results might not say
SV40 virus
123 Part Two: Identity crisis

Martin Textor
anything about how it responds to tumors
that develop spontaneously.
Another issue is secondary infections.
Burkitt’s lymphoma, described at the beginning
of this part of the book, demonstrated
that a human loses the ability to cope with
the Epstein Barr virus – and the tumors that it
leads to – if its immune system is simultaneously
being challenged by the malaria parasite
or other infectious agents. This means
that cancer studies should ideally be carried
out in an environment where animals are free
from infectious agents.
Despite these challenges, Thomas’ lab has
found ways to approach the question. They
have been working with a mouse bearing a
gene called Tag, which is an oncogene – a molecule
that promotes the development of
cancer.
Tag comes from a virus called SV40, which can
infect rodents, primates, and humans.
Researchers believe that Tag blocks the functions
of at least two genes in mammals. One
of these is p53, whose job is to launch a selfdestruct
program in damaged cells. If there
are mutations in p53, or if it can’t do its job
due to interference by Tag or another molecule,
aberrant cells may survive and cause
cancer.
These factors have made SV40 a tool to study
cancer for 40 years. In the 1960s Galina
Deichman, of the Cancer Research Center in
Moscow, did a series of experiments that used
the virus to create tumors in mice, in hopes of
learning how the immune system responds.
SV40 could infect cells in her rodents but
could not reproduce itself. It functions by
sneaking its genes into the cell’s genome.
When the cell produces Tag it may become
cancerous, divide, and build a tumor.
Giving adult animals the virus would cause an
infection that would provoke an immune system
response. Instead, Deichman gave SV40
to newborn hamsters whose immune system
hadn’t yet been “switched on.” At this early
stage their bodies could not treat it as an
infection. Most of the hamsters infected as
newborns developed tumors after a long
latency period, similar to the slow pace at
which spontaneous tumors arise in humans.
Deichman discovered that the animals could
be vaccinated against the molecule. Giving
the virus to adult hamsters with a healthy
immune system triggered a strong, immediate
response. SV40 was fought like any other
infection. T cells wiped it out. They even wiped
out the disease in hamsters that had been
given the virus as newborns. This meant that
the immune system found cells that had been
infected much earlier and destroyed them as
it coped with the current infection. The hamsters
given an SV40 “vaccine” never developed
tumors. All the control animals, who were
infected but didn’t receive the immunization,
died from tumors.
Vaccines are preventative tools; they don’t
work if given too late in an infection. The
same was true here. If Deichman waited too
long with the second round of infection, until
about two weeks before cells had grown into
Part Two: Identity crisis
124
detectable tumors, the “vaccination” no
longer helped. Either the body became tolerant
when a certain number of cells became
cancerous, or something else was going on.
This shows, Thomas emphasizes, that giving
the virus to newborns didn’t make them tolerant
to SV40; they weren’t mistaking its
genes for “self.” Even so, tumors weren’t being
eliminated by the immune system.
With the development of new genetic engineering
techniques in the 1980s, scientists
could directly implant Tag into the genome of
mice; it was no longer necessary to use viruses
to introduce cancer-causing genes. The
results were the same. In young animals
whose genomes contained Tag, very few cells
– if any – produced Tag, so the mice did not
learn to recognize it as “self.” If the animals
were vaccinated early enough, they could be
protected even from tumors arising from
their own cells. This was getting very close,
Thomas says, to a good model for human
cancers.
Another study helped set the stage for
Thomas’ current work. In 1962, while working
with cancers induced by chemical carcinogens,
Lloyd Old and his colleagues at the
Sloan-Kettering Institute for Cancer Research
in New York took tumor cells from one mouse
and transplanted them into another. They discovered
that mice which received fewer transplanted
cells were more likely to develop cancer
than mice that received a lot of cells. You
might expect things to be the other way
around – unless high numbers of cells resembled
an infection, while lower numbers

escaped. By staying under the radar of the
body’s first wave of rejection, the tumors
seemed to be avoiding the second.
��
Inserting the Tag gene into the mouse
genome made a laboratory model that
came close to imitating human cancer, but it
could be refined further. Thomas and a postdoc
in his group, Gerald Willimsky, added
another feature to the artificial Tag gene. They
intended to make a conditional form of Tag,
which would allow them to switch it on in a
specific tissue, at any time they chose. (This
type of conditional gene is described in the
story “How to roust a resting cell.”) In the
process, something unplanned happened –
one of those lucky accidents that sometimes
make great science.
Building the artificial Tag gene meant inserting
it into the mouse genome with some additional
instructions. One bit of code added to
the gene was a promoter sequence, telling
cells to make the molecule. A second bit of
code acted as a brake, preventing that from
happening. The combination meant that Tag
wouldn’t be produced. The mouse needed to
be mated with a second animal outfitted with
another gene, one which released the brake in
a specific tissue. Any offspring of the mice that
inherited both genes could be used to study
cancer in that tissue. By mating the Tag mouse
with various other strains, the researchers
could study tumors throughout the body.
Then the accident happened: they noticed
that all of the Tag animals were developing
tumors even without being mated to the second
strain. “At first this was very disappointing,”
Thomas says. “It looked as if something
had gone wrong, and because it takes months
or years to develop such a mouse, it meant we
might have to start all over again.”
A closer look at the animals, however, showed
that something else was going on.
“Occasionally a rare random event, in random
cells in the body, would release the brake and
activate the Tag gene,” Thomas says. “It was
impossible to predict when or where these
mutations would happen. Most of the mice
developed one tumor; a few had more than
one, and in some cases there were metastases.
This happened in all sorts of tissues: in
the kidneys, liver, bone, spleen, and elsewhere.”
In fact, the situation was close to what happens
in spontaneous human cancers: somewhere
in the body, a mutation in a gene leads
a tumor to arise. Not completely by accident –
but with a little luck – Thomas and Gerald had
created a new mouse model that closely imitated
the development of human tumors.
They also had a perfect situation to study
interactions between the tumors and the
immune system. When cells became cancerous
and produced Tag, they can potentially be
noticed by the immune system. Are they? “We
now had a way to check,” Thomas says, “by
125 Part Two: Identity crisis
looking to see if the body produced antibodies
against Tag. And we found them.”
The scientists repeated the types of vaccination
experiments that Deichman had carried
out with the SV40-infected mice. None of the
mice immunized with Tag developed tumors
over a two-year period. All of the mice in the
control group, which had not received the vaccine,
died from tumors. Additionally, when
tumor cells were transplanted into young,
tumor-free transgenic mice, they were rejected.
It was further proof that the body treats a
transplant like an infection, and that cancer
cells themselves don’t come equipped with
special characteristics that help them evade
the immune system. They can be easily fought
if the mouse body becomes aware of them in
the right way. But tumors that arise spontaneously
in the body are different. The immune
system sees them, but then nothing happens.
T cells don’t take further action.
Thomas shows me a copy of the paper he
wrote with Gerald, who has now left the MDC
and holds a post at the Charité’s Institute of
Immunology, where Thomas has a second lab.
“The move happened for formal reasons, and
it was not always easy for him,” Thomas says.
“Today to be a real success as a scientist, you
have to publish often, and quickly. This project
didn’t fit well into that scheme – it took seven
years.”
It was worthwhile, he says, because it is the
strongest proof yet that there is a fundamen-

tal difference between the way the body
rejects a spontaneously generated tumor and
one which is transplanted – even if it is the
same tumor. This is at least a partial answer
to one of Thomas’ questions: cancer studies
based on transplants do not necessarily shed
light on some of the critical issues in immune
responses to cancer.
“Our results open a new set of questions
because they show that tumors aren’t escaping
by just hiding from the immune system,”
Thomas says. “ The body sees Tag long before
tumors appear and makes lots of copies of T
cells that recognize it, just as humans produce
T cells that recognize some kinds of tumors –
you find them in the tumor itself, in lymph,
and in blood – although there are no signs of
the body rejecting the tumor. The body does
not behave as if it has been immunized.
Tumors grow by somehow creating a tolerance
in the immune system, and this is something
we are investigating quite intensively
now.”
��
Iam having trouble getting used to this new
way of thinking. If anything in today’s medical
culture should be regarded as a devil, it is
The Berlin Wall
cancer. Strange that our bodies are able to
come to a truce with this terrible disease. Too
bad that it doesn’t last longer. On the other
hand, the pact lasts longer than the 24 years
given to Faust – long enough for most of us to
have and raise our children. After that, the
evolutionary peacekeeping force packs up and
leaves.
The last revolution in medicine began with an
understanding of the causes of infectious diseases.
Fighting those illnesses could take
advantage of an immune system that had
already been primed to cope with them. It
seems there will be no such natural help for
spontaneous cancers and the other threats
that strike in old age. It’s the reason “molecular
medicine” is necessary – if evolution hasn’t
given the body defenses, they will have to be
invented.
On the other hand, human intelligence – and
science – are also the products of evolution, so
in a way biotechnology and modern medicine
can be thought of as humans’ way of evolving
new defenses. This idea was expressed eloquently
a few years ago in a talk given by biologist
Hubert Markl, former President of the
German Research Council and of the Max
Planck Society. He said:
We should see ourselves not as some kind of
fallen angel, alien intruder, some aberrant
or deranged scourge of nature, but as its
constituent and heir. And not only as one
constituent part of nature among many
others, just an arbitrarily chosen biological
species, but as a unique, a quite extraordinary
kind of natural species, through which
nature entered into an entirely new stage of
evolution; a species that not only participates
in its future evolution like any other
species, but that increasingly commands
and determines this future, for better or
worse. In evolving the human species,
nature, as it were, began to take control of
its own future, to give it purposeful direction,
to assume responsibility for its own
future development... From such a comprehensive
evolutionary perspective, human
technological and economic inventiveness is
nothing other than nature’s way of intentionally
acting upon itself and forming its
own future.
��
Jens Reich

Thomas’ work suggests that it could very
well be possible to retrain the immune
system to fight spontaneous cancers – if
markers can be found, and diseases are
caught early enough. Putting this into practice
is the goal of most of the efforts of
Thomas, his lab, and some of his peers.
One of those colleagues is Wolfgang Uckert, a
former postdoc in Thomas’ lab, now head of
his own group at the MDC. Although he’s still
relatively young, Wolfgang has been a firsthand
witness to more of the history of the
Buch campus than most of the scientists who
work here. He earned his PhD here when this
area of Berlin belonged to the German
Democratic Republic, doing most of his work
in the oldest laboratory building on campus. It
was originally built to serve as Oskar Vogt’s
Institute for Neurological Research; it was this
building that Vogt studied Lenin’s brain, looking
for traces of genius; later it was integrated
into one of East Germany’s prestigious
Academies of Science.
Practicing science was different in the GDR,
Wolfgang says. Today’s researchers are world
travelers, expected to attend conferences
across the globe, often working as postdoctoral
fellows in two or three countries before
attaining a long-term job as an independent
researcher. Things were much different in the
East. It wasn’t necessarily impossible to travel
outside the Iron Curtain under the East
German regime, but it was extremely complicated.
“At one point I was offered an opportunity
to go to Belgium for three months,”
Wolfgang says. “By that time I had a family.
They stayed here, which was probably the reason
I was allowed to go.” But when he was
127 Part Two: Identity crisis
Wolfgang Uckert and a piece of the Berlin Wall
offered a longer-term position in California,
the state wasn’t as forthcoming.
Jens Reich, MDC emeritus and former head of
the MDC’s biocomputing groups, had extensive
first-hand experience with the Stasi. He
remembers a case when a postdoctoral fellow
in the group went abroad and failed to return.
“This was considered extremely serious, and it
had huge effects,” Jens says. “The lab was
somehow thought to be responsible – how
could this happen without our knowledge?
Had he said anything? Had any plans been
discussed?” The event led to months of questioning.
Jens received regular summons for
sessions of hour-long questioning.
“The summons was always disguised as an an
invitation to give an administrative report, in
the 1970s to the Ministry of Science and
Technology, and in the 80s to the headquar-

ters of the Akademie der Wissenschaften. You
only learned that they were Stasi once you
arrived at the first meeting and they introduced
themselves.”
“Things were also quite ‘exciting’ when foreign
researchers visited the campus,” Jens
says. “Once we had the exceptional opportunity
to host an international conference here
in Buch. The entire event, which lasted for several
days, was ‘scripted’ and rigorously
observed. Yet it had to be done in such a way
that the visitors didn’t notice anything. This
led to some ridiculous situations. It was a
circus.”
The issue, he says, was meetings between foreigners
and East German citizens. “You were
allowed to have contact with foreign guests
only with explicit permission, and you had to
report afterwards. The ‘inofficial staff’ had to
report directly to the Stasi; everyone else sent
their reports to the directorate, and if the
Stasi had questions they came back the same
way. The local authorities tried to keep things
from getting too ridiculous by ‘not seeing
everything,’ but if the Stasi got wind of this,
they were in for rough times. A prominent
incident happened in 1978 when the
European Biochemistry Congress was held in
Dresden. A young co-worker had to accompany
a prominent guest from the United States
on an excursion to the museums of Weimar.
During the trip he bared his soul about the
restrictions on international exchange and
other nuisances. The American turned this
into a very critical report in Nature, and that
led to a major scandal. The organizers had to
confess and repent and write a long, tortuous
readers’ letter to Nature disclaiming the
report.”
Under those circumstances, people stuck
together. The lab wasn’t only a workplace; it
had an important social function. Wolfgang
tells how members of his lab kept a collection
jar for money. “You had to pay a little bit if you
forgot to clean up or if you broke something,”
he says. “At the end of the year we always took
what was in there and went out to dinner in
Part Two: Identity crisis
128
downtown East Berlin. There were really nice
restaurants that were open late.”
One year he got home from the event at
around 11pm. He turned on the television – “It
was easy to get the West Berlin stations here
in the city” – and was amazed to see people
dancing on top of the Berlin Wall. “At first I
couldn’t believe it,” he says. “When I realized
what was going on I told my wife that we had
to go back downtown; we had to be there.”
History was in the making in the city, an event
not to be missed. He quickly asked a neighbor
to come watch their small children and headed
for the Kurfürstendamm. “Later I came
back with my daughters and a hammer to get
a piece of the Wall,” he says. “The nice part
was on the Western side, where it had been
painted with graffiti. We wanted to send
pieces to all our friends and relatives. It turned
out we didn’t get a big enough piece, and we
had to go back again.”
��

Can the immune system be taught something
it hasn’t evolved to do – to combat
spontaneous cancer? A major theme of
Wolfgang’s work has been to develop “designer”
T cells that might be able to do this. He has
been working on the problem in collaboration
with Thomas Blankenstein. The two men met
as the MDC was established, as Thomas set
up his new lab. He was looking for staff members
and interviewed some of the scientists
who had worked in the Academy on campus.
“It was a bit of an odd situation at the time,
hiring someone who was a year older than
myself,” Thomas says. “But it turned out to be
a very good fit. I was very pleased when
Wolfgang received his professorship and
could set up his own group, and we have
worked closely all of these years.”
“Adaptive immune therapy” aims to remove T
cells from a patient’s body and outfit them
with a new receptor protein that allows them
to recognize cancer. A major hurdle is to find
the right target for the receptor – a protein
that only appears on the surface of tumor
cells (the chapter “Capturing a firefly” shows
how difficult that can be). But even before
that happens, there are a number of technical
challenges to solve in rebuilding T cells and
making sure that they behave properly.
To explain, Wolfgang shows me a greatly oversimplified
drawing of a T cell. The receptors
look like the letter “Y” and are made of two
proteins. These are produced by the cell’s
genes and then mounted on its surface with
the assistance of other proteins called CD3s.
The two parts of the receptor are called alpha
and beta chains, he says, but it’s fine if I want
to think of them as “left” and “right” arms. An
artificial receptor also needs two arms, which
means adding two new genes to the cell.
“In real life things are much more complex,”
Wolfgang says. “A typical T cell uses its genes
to make about 40,000 identical copies of its
receptor, that means about 40,000 left arms
and 40,000 right. When we give that cell the
genes to make a second receptor, it doesn’t
129 Part Two: Identity crisis
stop making those. Instead, it starts producing
a lot more arms.”
The cell assembles the proteins into pairs and
moves them to the membrane. “Remember
that it now has four ‘arms’ to make receptors,
and that it can’t recognize where the different
arms have come from. It starts pairing them
up, a left and a right. A fourth of those pairs
will consist of just the new left and right arms
we have added. Another fourth will consist of
the cell’s native left and right arms. The rest
will be mixtures, one of our arms paired with
one of the cell’s.”
It is hard to predict how these different sets of
receptors will behave. “Suppose you put one
of these cells into the body and the new
receptor recognizes a protein on a tumor,”
Wolfgang says. “That will trigger the T cell to
make billions of copies of itself. That’s what
you want, because it creates the cells needed
to fight the tumor. But at the same time,
you’re making cells that may be looking for
extra targets whose identity you can’t predict.
The extra receptors on the cell might not recognize
anything, in which case you’re safe. But
if they come across a molecule anywhere in
the body that they can dock onto, there’s likely
to be trouble.”
For example, the “hybrid” receptors that combine
one foreign and one native arm might
recognize one of the body’s proteins and trigger
an autoimmune reaction. The native
receptor might also recognize a “self” molecule.
“Usually this is no real problem, because
the body won’t produce many of these cells
and they can be kept under control,”
Wolfgang says. “But if you suddenly start having
billions of them, because the body is using
the cells to fight a tumor, things might get
out of hand.”
Is there any way to force the T cell to produce
only the artificial receptors? I ask.
“What we’ve discovered is that some receptors
are very successful at moving to the surface
of the cell and others are not,” Wolfgang

Martin Textor
says. “In some cases when we engineer the T
cells, they produce almost only the new receptor,
and very little of the native one. In other
cases it’s the other way around. One of the
major activities of the group right now is to
try to figure out what determines how successful
a receptor will be. Because theoretically
this means we can build a molecule that
pushes aside the native one, and then we’ll
have eliminated these kinds of concerns that
we have about interference.”
Successful receptors may be the ones that are
best at recruiting partner proteins like CD3, he
says. And in this regard, mice have amazingly
potent molecules. If a T cell receptor is transplanted
from a mouse into a human cell, it
wins out over human receptors every time.
His lab is now trying to figure out which features
of the mice molecules make this happen.
If they succeed, they should be able to
custom design receptors on the surfaces of T
cells that will replace the cell’s molecules.
They may even be able to add more than one
type of receptor – in some cases it might
make the T cell more effective – and control
the “dosage” at which each type is made by
the cell.
Even then, he admits, biological processes are
so complex that there’s always a chance
something unforeseen will happen. Wolfgang
is well aware that making the jump from the
lab to humans always involve risks. The first
gene therapy experiments carried out in the
U.S. and elsewhere – mostly using viruses to
transport healthy genes into cells – have led
to both successes and tragedy.
“So while we have been designing these new
types of cells, we thought we would also build
in some safeguards,” he says. “We’re learning
to equip our T cells with an emergency switch
that allows us to shut things down fast if
something unexpected happens.”
PhD student Elisa Kieback has been pursuing
that project, adding an extra bit of code to the
receptor. The code allows the molecule to be
Part Two: Identity crisis
130
recognized by antibodies that a doctor can
give a patient if the cells start to misbehave.
The antibodies tag the engineered T cells and
call up a rapid immune system response,
which destroys the cells.
��
ne conclusion from our work is that
“Oif specific markers for tumors can be
identified, and we intervene at an early stage
in the disease, it is likely that patients can be
immunized,” Thomas Blankenstein says.
This may not be possible in the case of most
spontaneous cancers that arise because of
mutations in a random gene. But some types
of tumors behave in a stereotypical way, produced
by well-known oncogenes. If scientists
can identify them, it may be possible to detect
problems in patients at an early stage and
prepare a defense.
It may be easier in cancers associated with
viruses, but for most of these diseases it has
also been difficult to develop vaccines against

the viruses, and once they have settled in, the
cancers they cause are difficult to treat. “Once
classical therapies such as chemotherapy,
radiation and surgery have failed, there are
few options,” Thomas says. “Some recent successes
using T cells trained against the viruses
have encouraged us to continue to work in
that direction, but developing T cells as therapeutic
tools is a long-term, labor-intensive
project.”
In spite of that fact, clinical trials are currently
underway with some of the T cells that
Thomas and Wolfgang have engineered. Yet
the real goal, making T cell therapies a standard,
working tool in the doctor’s arsenal, is
obviously beyond the work of a single lab. If
you had a whole institute at your disposal, I
ask him, and could devote it entirely to T cell
therapies, what would you do?
Thomas’ eyes light up. He doesn’t have a
whole institute, but maybe he’s got something
even better. In 2006 he made a proposal
to the German Research Council to create a
special, national research area dedicated to
just this theme. The project unifies the work
from laboratories in Berlin and Münich,
including those of Thomas and Wolfang,
other groups from the MDC and Charité, the
Helmholtz Institute for Molecular
Immunology, and a number of universities.
Together they will try to develop specific
therapies for some forms of leukemia, renal
cell carcinomas, and other forms of cancer, as
well as pursuing basic issues in the use of T
cells.
The ultimate aim, Thomas says, is to develop
systems that can be used to produce T cells
targeting any molecule that the researchers
choose – ideally within a week. “The strategy
would be to have a stock of T cell receptors on
the shelf, which can be taken down and used
to produce T cells targeting any molecule we
choose – ideally within a week. They will be
used to target life-threatening viral diseases
or diverse forms of cancer, including some
solid tumors.”
131 Part Two: Identity crisis
How long will it be until customized T cells
become a widely-used tool in the fight
against cancer? What will have to happen?
“I’m always asked this question,” Thomas
sighs. “I refuse to answer with a specific time
frame. Reality will always be different than
what we hope.”
Still, he says, he strongly believes – as does
nearly everyone working in the field – that in
the long term, the strategy will work. When
that day comes, it will be possible to add
clauses to the body’s pact with cancer, or to
rewrite the agreement, extending it by many
more years. Or maybe the pact can be discarded
entirely, in favor of therapies that prevent
cancers from ever arising in the first place.

The decorators and
engineers of Versailles
It’s an amazing place to visit, but few would want to live there. To
modern taste, the palace of Versailles on the outskirts of Paris is
overwhelming. The French of the 17th and 18th centuries indulged themselves
in an orgy of ornamentation that today seems oppressive.
Decoration reaches into every cranny, every piece of furniture and picture
frame, the vaulted ceilings, and even empty surfaces. Extravagent
golden picture frames steal attention from the paintings they enclose.
Walls, tabletops and floors are inlaid with intricate patterns. Gilded
flowers are printed on wallpaper and stitched into cloth. Fireplaces are
crafted from mottled marble rather than plain stone. Thousands of bits
of glass in each of the chandeliers catch splinters of light and diffract
them in every direction, where they are caught by thousands of mirrors
and infinitely reflected.
How could you walk in the shoes, or get anything done in those clothes?
The lace on the sleeves would eventually drive you mad. In the court of
Louis XIV, heads were wigged and faces were powdered; legs were
stockinged and hands gloved, as if to protect the body from any contact
with the real world. Historical sources say that the perfumes of
Part Two: Identity crisis
132

Hans Baumeister

Versailles were equally extravagent; it makes
you glad that there have been two centuries
since the Revolution for the air to recirculate.
In the French Baroque, ornaments were as
pervasive as bacteria in the ecosphere.
Neither literature nor music escaped the
infection. Nearly every note written by
François Couperin is ornamented. Many of his
trills and vibratos are unique; he included a
handbook at the beginning of some of his
works to explain how to play them. In this
music, however, ornaments are more than
style. They create textures; they glide between
awkward intervals; they are fundamental to
the harmony, leaving a listener hanging in dissonances
that ache to be resolved. French
Baroque compositions can’t be played without
them; the music has no depth and makes
no sense without the ornaments.
The escape of art and life into an artificial
world couldn’t go on forever. In music, popular
taste shifted toward the Italian style of Vivaldi,
whose virtuosity was based on acrobatics
rather than subtlety. The Revolution stripped
off the wigs and powders and forced the nobility
to give up their cake for breads. It’s no sur-
prise that a lifestyle so overburdened with
style itself came crashing to an end.
��
Is there anything in the living world that is
simply ornamental? Charles Darwin threw
this idea into doubt. Variation is everywhere,
but it is a motor of evolution: sooner or later,
even tiny details come under the scrutiny of
natural selection. A patch of color on an animal
may attract a mate or a predator; the rich
palette of flower petals serves as signposts for
bees and butterflies, who see more colors
than humans do.
A type of chemical decoration found in cells –
the attachment of sugars (carbohydrates) to
proteins – probably began as an accident in
ancient unicellular organisms. Over the
course of evolution, this has expanded to a
level of ornamentation worthy of Versailles. In
the French Baroque, ornamentation rarely had
a function. But in the cell, combinations of
sugars and proteins (called glycoproteins)
became useful; many glycoproteins, in fact,
don’t function without carbohydrates.
Organisms evolved hundreds of specialized
molecules to do the decorating.
Part Two: Identity crisis
134
Anything found so widely throughout nature
has to be important, and every day
researchers discover new ways that sugars
influence the behavior of proteins. “We’re only
now getting an idea of how versatile glycoproteins
are,” says Hans Baumeister, of the
company Glycotope on the Berlin-Buch campus.
“In some cases carbohydrates’ functions
are still not known, but in other cases they are
essential, sometimes even vital for the development
or survival of an organism.”
Hans is the Chief Operating Officer of Glycotope,
a company founded in 2001 by CEO Steffen
Goletz and Andreas Eckert. Andreas is CEO and
founder of another well-established business on
campus: Eckert & Ziegler Strahlen- and
Medizintechnik. Steffen was a group leader at the
MDC until he and Andreas formed the two startups,
first NEMOD and one year later Glycotope,
with Steffen as CSO of both companies. Later
Steffen became CEO of Glycotope. Hans joined
NEMOD as a scientist and was one of the first
employees at Glycotope. Over the past few years
he has gravitated toward the business side of
the company. Now he heads Glycotope's service
activities as COO, and Franzpeter Bracht
serves as the company's CFO/CBO.

The company fills an important niche in the
biotech market, Hans says. “Most research and
medical applications require the production
of high quantities of proteins,” he says. “That’s
usually done in cultures of bacteria or animal
cells, sometimes in the animals themselves.
These cells have to make human proteins
properly – especially when those molecules
are destined for medical applications.
Achieving that hasn’t been as easy as people
hoped, and that has partly been due to glycosylation.
The company is based on unique
technology that helps us fix this problem.”
A century ago a lot of biochemists were
deeply interested in sugars, but with the coming
of molecular biology the focus shifted
largely to DNA, RNA and proteins. One reason
is the complexity of sugar molecules and the
way they are built by cells. Sugars are not
directly encoded in the genome but are constructed
by other molecules in a complex
series of chemical transformations. They are
made from many different simpler sugars
(monosaccharides) which form a much wider
variety of sequences than are found in DNA,
RNA and proteins. This makes their physical
structures are very complex. All of these
things have been obstacles to their investigation.
“Basically this means that in most cases
you know the protein you have in hand but
you don’t know exactly what sugar is
attached to the protein, or in some cases even
where it is attached,” Hans says. “There aren’t
the shortcuts you find with proteins – you
can’t look at a DNA sequence and produce a
complete catalogue of the sugars that a certain
organism can make.”
The surface of an animal cell is covered with
carbohydrates – just about every molecule
found on its surface or secreted from it has
been modified through the attachment of
sugars. “That includes protein hormones,
growth factors, cytokines, blood factors and
antibodies,” Hans says. “You know how important
antibodies are in research and medical
applications. Well, most antibodies for medical
applications have been made by inserting
genes into hamster cells, turning them into
antibody factories. As the molecules are
made, the cells attach sugars. But those are
different sugars than the ones you find in
human cells. The result is that the molecule
you’ve made for therapy often works differently
than it should. Even worse – it might
induce an unwanted immune response
against the therapeutic agent.”
��
Each organ and tissue is a composer, developing
its own unique language of ornamentation;
each has its own specialized types
of sugars. The molecules change with every
135
new context. Cells glycosylate proteins differently
as they specialize and when other
changes happen – for example, if they start to
build a tumor.
Most antibodies are decorated with just one
sugar, but it’s an important one. It is attached
to the part of the antibody that does not bind
to its antigen, but to receptors that activate
effector cells such as macrophages or natural
killer cells; these home in on the target cell,
which is recognized by the antibody. This is an
important mechanism for tumor therapy. “So
an optimal glycosylation of an antibody is
necessary for the therapeutic application of
that antibody,” Hans says. Sugars are also
important in the functions of hormones,
growth factors and cytokines, powerful signaling
molecules released by glands throughout
the body.
“An important part of their sugar decoration
are sialic acids, the only sugars that are negatively
charged,” Hans says. “Different numbers
of these acids are added during the process of

glycosylation. Having a lot of sialic acids
increases the molecule’s lifespan, because
hormones that don’t have any are grabbed by
the liver and removed from circulation. On the
other hand, having too many of them may
decrease the activity of the hormone. So it’s
another thing that has to be done right when
you try to produce a glycosylated molecule in
a cell line.”
Glycotope has found a way to address many of
these problems through the development of
new cell lines. The company’s scientists began
with human cells and modified them to carry
out different kinds of glycosylation. One type of
cell sialylates molecules to a high degree.
Another adds none, unless the cell is fed with
with sugar. How much is added to molecules
depends on how much the scientists feed it.
“This gives us a very wide spectrum of possibilities,”
Hans says. “Using human cells is
already an improvement over animal systems
that often glycosylate in the wrong way. And
this system can be adapted to produce the
molecules needed, with a range of carbohydrates
attached to them.”
There has been a huge demand for the company’s
technology, called “GlycoExpress,” he
says. Most of the clients are pharmaceutical
and biotech companies developing their own
therapeutic antibodies or other biotherapeutics;
Hans says they can significantly improve
their products with GlycoExpress by reducing
the necessary doses to be injected and avoiding
unwanted side effects.
Glycotope has several products of its own that
are currently in development. One is a new
“therapeutic antibody” which recognizes and
binds to target the carbohydrate of a protein
on the surface of tumor cells. Scientists have
known for over 40 years that a protein called
mucin-1 acts abnormally during cancer and
that sugars attached to that protein change
during tumorgenesis.
Mucin-1 was discovered when scientists
examined a gelatinous substance that accu-
mulated around breast tumors. An article
written in 1965 by J.B. Adams of the Prince of
Wales Hospital in Sydney, Australia, reported
that the substance contained high levels of
sugars and a protein. In the meantime, scientists
have discovered that the protein is heavily
loaded with sugars, and that large
amounts of it accumulate inside and around
many types of cancer cells.
Scientists at Glycotope discovered a new antibody
that very specifically recognizes the sug-
Part Two: Identity crisis
136
ars on mucin-1 of diseased cells. It was an
exciting find, Hans says, because only tumors
seemed to glycosylate the protein in this particular
way – not normal cells. “Any time you
find a unique marker, it opens therapeutic
possibilities. In this case the idea was to develop
this antibody as therapeutic agent to
destroy the tumor.”
Antje Danielczyk and the other scientists of
Glycotope know how to turn an antibody into
a therapeutic device: they attached a radioac-

tive isotope to the antibody which will destroy
the tumor cell when delivered to it.
Alternatively, the antibody may be used as a
therapeutic device as is (it’s called a “naked”
antibody as opposed to the radioactively
labeled antibody). That works if the antibody
is able to mediate killing of the tumor cells
through macrophages or natural killer cells.
The next step was to test PankoMab, as the
new drug candidate was called, in both the
radiolabeled and “naked” format in tumor-
loaded mice. (The name was inspired by
Pankow, a neighborhood of Berlin that lies
between the city center and Buch.) The tumor
quickly absorbed PankoMab and was very efficiently
killed using either form.
In parallel, the scientists conducted tests of
PankoMab in the test tube and discovered
another important feature: it binds much
more strongly to mucin-1 on the surface of living
cells than to mucin-1 molecules that have
been released by cells that have already died.
137 Part Two: Identity crisis
That is often a problem in antibody therapies,
because loose protein fragments may suck up
most of the injected antibodies.
“So PankoMab has passed all the tests we
could throw at it to determine whether it
could be used in therapies,” Hans says. “First
the antibody has been humanized. That’s an
important step during development of therapeutic
antibodies that are originally synthezised
as mouse antibodies – they would
induce strong immune reactions against the
drug itself if you applied them in humans.
Now it’s being produced under ‘good manufacturing
conditions’ (GMPs) to give it the
quality needed to be tested in clinical trials in
patients. If it passes that hurdle, we’ll be getting
closer to the day it can be used in any
patient who is waiting for treatment. It has
also been an important proof of principle that
our approach – focusing on improving the
sweet part of a protein – will lead to a much
more powerful therapeutic drug.”
��
It’s a Friday afternoon in the spring and I am
standing in the sun outside Hans’ office. He
passes two coffee cups through the window,
locks up his room, and joins me outside.
Glycotope is housed in the Erwin Negelein
building, which was built a few years ago as
the campus set up the biotechnology park. It’s
fitting that Hans works in a building named
after Negelein, a biochemist and departmen-

138
tal director during the GDR years. Negelein
was a biochemist who had developed cell
lines from mice into tools for the investigation
of cancer.
I ask if the campus research groups take
advantage of the company’s services.
“We have had a few collaborations with
groups here,” he says. “Martin Lipp’s lab needed
a human version of an antibody; the versions
they had obtained from mice would have
caused immune reactions. On the average I’d
guess that we work with campus groups two
or three times a year. Of course we would welcome
more business directly from the institutes,
but they’re not our typical target group.
Making humanized and glycooptimised antibodies
is usually too expensive for them. Most
of our clients are pharmaceutical and biotech
companies around the whole world.”
There are many other advantages to being on
campus. “Berlin itself is a good location,” he
says. “The people we’d like to hire are generally
eager to come; they like the city. Setting up
labs is not expensive compared to other
places – that’s very important for young companies
trying to keep costs down.”
The scientific life of the campus also has
advantages. “You get into fruitful scientific
discussions, and can get access to instruments
and expertise that you might need.
Obviously it’s a good place to be in terms of
recruitment. A lot of the people who worked
from the beginning in the company started
off in one of the campus institutes.”
We talk for a while about the differences
between working in a basic research institute
and working for a company. Is it hard for
someone to switch to industry? I ask.
“I guess I’m a good person to ask,” he says. “I
began as a basic researcher and have now
completely moved into the business side of
science.”
Hans grew up in Berlin – in the southwestern
part of the city, where he still lives with his

family. After receiving his diploma from the
Freie Universität of Berlin, he worked on his
PhD in Hamburg, with a focus on cell receptors.
His introduction to industry began with
his next move, to New Jersey, where he
worked for the Swiss pharmaceutical company
Hoffmann-La Roche.
“At that time the company was still heavily
supporting fundamental science,” he says.
“They had an institute which was really just
doing basic research. It was an interesting
place – the bosses were Americans, and just
about everybody else was European or Asian.
They were doing work on things like transcription
factors, and I don’t think they had produced
anything in the way of products for a
market for about thirty years.”
Eventually the institute went the way of so
many research laboratories funded by pharmaceutical
companies – they have been
closed down, or the work has been pushed in
a much more practical direction. This has left
a gap that institutes like the MDC or FMP
need to fill – if medical science hopes to take
advantage of the wealth of new knowledge
emerging from basic science.
Maybe an increasing movement of students
and other scientists into industry will help.
The transition isn’t as hard as you might
think, Hans says, although there are important
differences between the cultures.
“You have to get used to working on projects
that are much more focused,” he says. “The
projects we do don’t have the breadth of
things going on in places like the MDC or FMP.
And they have a different aim. In science the
measure of success is publications; in industry
it’s patents – at least for researchers.
Publications come thereafter. And research is
organized in milestones. When a project or
potential product to be developed misses too
many milestones, it is skipped.”
139 Part Two: Identity crisis
Companies have more of an engineering
mentality, he says. It’s a different kind of creativity.
“We’re less looking for some completely new
thing than to take something that exists, fiddle
with it, and improve it until it works,” he
says. “It’s enormously satisfying to see that
something you have worked hard on is used.
The high point is when it is developed into a
product that makes a difference in patients’
lives.”
One of the biggest contrasts, Hans says, is
that the science carried out in companies has
clients.
“When you begin a project you immediately
imagine its potential clients,” he says. “It’s not
all about a patent and a product, but who will
be using it. How much will it cost to develop
something? Is there enough money and are
there enough potential clients? Can you
excite an investor about an idea?” One of the
benefits when a company offers services, he
says, is that income can be rolled into the
development of new ideas.
Later in the process come the real clients.
“They’re trying something and having problems,
and then your job becomes communication,”
he says. “It’s a part of the job that I like
very much. Sitting down with someone and
figuring out how to solve technical issues.”
From sugars on proteins to powerful therapeutic
tools. Too bad the French never found a
way to use their extravagent ornaments to
heal the sick or feed the poor. The vast gardens
devoted to plants for perfumes might
have been used to raise medical herbs. That
might not have changed anything significant
in history, or the way science is practiced
today. On the other hand, some elements of
French style might have survived. Instead of
the black suit and colored tie that Hans sometimes
wears when he represents the company,
he might have to wear clothing that
looked like wallpaper. And a high powdered
wig.

Waking a sleeping
beauty and other tales of
ancient genes
Good stories, like living things, evolve over time. Nuances are
added or taken away; tales become confused and merge; some
are forgotten entirely, while others seem to live forever. The tale of the
sleeping beauty has gone through most of these transformations. An
early form of it appears in The Book of 1001 Nights, a collection of Arabic
tales from the tenth century. Giambattista Basile, a soldier from Naples,
inspired the modern story with a fable in his book Pentamerone,
published in 1634. Charles Perrault revived it in the Mother Goose
Tales (1697).
The version collected by the Grimm brothers a hundred years later is the
one most of us have heard. A young princess pricks her finger on a spindle,
fulfilling a curse placed on her at birth. This sends her and everyone
else in the castle into a deep sleep. They are frozen in time for a century,
until a prince fights his way through the thorny thicket surrounding
the castle to deliver his kiss. Everyone wakes up, and they all live happily
ever after.
The Grimms left out some of the darker details of the story. In Basile’s
account the princess doesn’t wake up until she has borne two of the
Part Two: Identity crisis
140

Zoltán Ivics
prince’s children. And the German version
leaves out the entire second half, told by
Basile and Perrault: one day the prince rides
off to battle. He leaves his young family in the
care of the queen mother, who is a descendant
of an ogre, creatures with a taste for
human flesh. She develops an insatiable
hunger for the children and the princess and
demands that they be cooked for her dinner. A
quick-thinking cook switches the children for
tender lambs. When it comes time to prepare
the princess, he has a hard time finding an
appropriate meat. (Perrault writes, “The young
queen was twenty years old, not counting the
hundred years she had been asleep, and the
cook was hard-pressed to find a meat that
would be tough enough.”) He finally settles
on a cow. Naturally the queen discovers she
has been deceived; naturally, in the nick of
time, the prince returns to save the day.
Genomes, too, are a sort of storybook, and
some of the genes they contain have been put
into a deep sleep by mutations. Even a prince
can’t wake them – now it will take a genetic
engineer.
��
Zsuzsanna Izsvák
Mythopoeisis is the art of making and
reshaping myths, and two labs at the
141 Part Two: Identity crisis
MDC are engaged in a sort of mythopoesis of
genomes. Zsuzsanna Izsvák and her husband
Zoltán Ivics head neighboring research groups
on the second floor of the Walther Friedrich
House. They are studying and waking ancient
genes for use in research and medical applications.
Genetic engineering is a collection of methods
that allow scientists to remove genes,
rebuild them, or transplant them from one
organism to another. Usually this involves
cutting parts from an existing DNA molecule
and pasting them in somewhere else. In the
1950s scientists found naturally-occurring

enzymes
in bacteria able to
do this cut-and-paste job.
Genetic engineering was born when
these molecules were adapted for use in
other situations. In 1972 Janet Mertz and Ron
Davis, of Stanford University, used them to
remove a gene from one bacteria and insert it
into another. A year later, their colleagues
Stanley Cohen and Annie Chang, with Herbert
Boyer of the University of California in San
Francisco, used the procedure to transplant a
gene between different species.
Today these techniques are biology’s most
powerful methods to learn what genes do
and how they work. But they aren’t perfect,
and they can’t be used as widely as scientists
would like. “We need more tools, particularly
to work with vertebrates,” Zoltán says. “At the
time we started, the zebrafish was becoming
an important model organism in the lab, but
we didn’t have tools to manipulate its
genome. Things are much better in the
mouse, but some tools for mice don’t work in
fish or mammals such as rats, or in many
other important model organisms.”
Zsuzsanna explains that they got started in
the field almost by chance, under tragic circumstances.
Zoltán had been awarded a sixmonth
fellowship to work on zebrafish genetics
in the United States, at the University of
Minnesota. Zsuzsanna was working on her
dissertation at the time – “which I could write
anywhere,” she says – and she came along
anticipating getting a job in the same lab. But
immediately after getting off the plane at the
airport, they were greeted with a piece of
devastating news: the boss of their laboratory,
Kevin Guise, had just died. He was
one of many early victims of the AIDS
epidemic, having received a transfusion
of blood containing the virus. It
had not yet become standard practice
to screen blood supplies for
signs of HIV. “We were shocked,”
Zsuzsanna says, “because we didn’t know
how seriously ill he was, and up until the last
minute he had been arranging things for our
stay. Apparently he had taken care of my visa
paperwork from the hospital.”
Arrangements were quickly made for them to
join the nearby lab of Perry Hackett, who had
helped introduce zebrafish at the university,
and who hoped to find ways to do genetic
engineering with them.
“At that time we were getting our first real
look at whole animal genomes and the huge
amount of sequence – in humans, about 98
percent – that doesn’t contain genes,” Zoltán
says. “The amazing thing was that about half
the genome consisted of the same little bits
of material, repeated over and over again.
Such repetition can be so substantial that
particular sequences, such as one called the
Alu repeat, are represented over a million
copies in the human genome. A similar phenomenon
is seen in fish. Over the course of
evolution, fragments of sequence had somehow
copied and pasted themselves all over
the genome, like little viruses.”
Some viruses were, in fact, known to behave in
a similar way. Retroviruses such as HIV reproduce
by squeezing their own genes into the
DNA of the host cell. The cell treats these molecules
like its own genes, using them to create
new RNAs and proteins. But rather than working
for the cell, these become the raw materials
to make new copies of the virus.
Cells had genes of their own that behaved
this way. In the 1930s geneticist Barbara
McClintock observed them in plants and animals
and called them transposons, or jumping
Part Two: Identity crisis
142
genes. One type of transposon cuts itself out
of the DNA sequence and pastes itself in at a
new place, sometimes after being copied. A
huge amount of extra DNA has accumulated
this way – apparently about half the human
genome. Transposons were probably a major
reason that the small genomes of the first
cells eventually became so huge. This creates
DNA sequences that may eventually be used
in genes, which has been a major factor in the
largest scale of evolution.
But for an individual, or a single species, transposons
are usually dangerous. “The fact that
they are usually pasted into random places
means they may land in the middle of another
gene,” Zoltán says. “That scrambles up its
information and usually destroys a molecule
that the cell needs.”
One goal of their research, Zoltán says, is to
find out how cells protect themselves. The

most straightforward defense involves random
mutations, which are just as likely to
affect transposons as anything else. In most
cases mutations have a negative effect on a
gene, damaging its functions. Damage to a
transposon would actually have a positive
effect because it would protect other genes.
The evolutionary pressure to deactivate jumping
genes is so strong that nearly all of the
transposons in animals have stopped working.
In the mid-1990s active jumping
genes had been found in bacteria, flies,
worms and plants, but none were known in
vertebrates.
Mutations would probably wipe out all transposons,
Zoltán says, if they didn’t find an
escape route. That seems to involve jumping
to other species, probably by riding as a hitchhiker
inside a bacterium or virus. The new
host offers fertile new DNA to colonize. The
transposon will copy itself over and over until
mutations catch up and disable it. Certain
types of transposons can be found in a wide
range of species, which is strong evidence for
the hitchhiker hypothesis.
Mutations don’t always wipe out a transposon;
sometimes they spin one molecule off
into a whole family of related forms. An
ancient family called Tc1/mariner can be found
throughout the living world, from bacteria to
fish, frogs, and humans. But, at least in vertebrate
species all of these transposons have
been inactivated.
“It would be nice to have one that works,”
Zoltán says. “They make up so much of animal
genomes that they need to be studied. And
some transposons encode all the tools needed
for cutting, copying and pasting DNA.
They’re like entire genetic engineering toolboxes
packed into the simplest form, into just
one gene.”
143
��
If nature hadn’t left any intact toolboxes
lying around, Zsuzsanna and Zoltán began
to wonder if they couldn’t make their own.
There were instructions to be found everywhere
in the transposon sequences that littered
every genome. Something was wrong
with nearly all of them – but each one had different
things wrong. By comparing all the versions,
the two scientists thought they might
be able to reconstruct the sequence of the
original gene, or at least a version that
worked.
Since the goal was to create tools that could
be used to transfer genes from one species to
another, it was best to start with a transposon
that had already proven its ability to spread
widely. The Tc1/mariner transposons seemed
to be the perfect choice.
“Mutations had inactivated all the versions
we found,” Zoltán says. “But they still contained
enough information to show how they
had worked. Originally the gene encoded
everything necessary to cut itself out and to
paste itself back in elsewhere.”
One goal of the analysis was to find the most
recent working version of the transposon. It
would have undergone fewer mutations,
which would mean having to repair fewer
broken parts to get it to jump again. A computer
analysis of the genes of various types of
fish identified one family of transposons in
which the average difference was small – only
five percent, despite the large evolutionary
distance between their hosts. This means
that this transposon family was very successful
in invading several fish species and has
been inactivated thereafter. The gradually
accumulating average difference, five percent,
functions as a molecular clock and the ”golden
age” of this invader family was estimated
to be around 10 million years ago.
By comparing several copies, Zoltán and
Zsuzsanna could make a hypothesis about
the identity of the chemical “letter” that

Andrea Schmitt
occurred most often in each position of the
DNA code in the transposon family. This consensus
recreated the genetic information of a
hypothetical invader which had successfully
colonized several salmonid fish species. Some
parts of the genes had undergone more
mutations than others. Evolution usually conserves
the most crucial features of an animal
– also of a gene – while less important characteristics
rapidly change. Zsuzsanna and Zoltán
found five regions that had been well conserved
in the protein-encoding part of the
transposon. These represented domains – the
main functional modules of the transposase
protein. As the lab reconstructed each module,
they tested their functions one by one.
As in all genes, the code in the transposon is
used to make a messenger RNA molecule,
which is then sent out of the cell nucleus to
be made into protein. In this case the protein
was the transposase, the scissors used to cut
their own gene out of the genome. Once it
has been made, the transposase has to get
back into the nucleus. One of the transposase
domains acts as a delivery code, telling the
cell where to deliver it. Inside the nucleus, the
transpose has to grip the DNA strand where it
will make a cut. Another domain handles this
job.
When the labs had completed the new version
of the molecule with the corrected
spellings, it gave them a new, artificial transposon.
They named it Sleeping Beauty –
“because it had been sleeping in the fish
genome a very long time” – and began testing
whether it could be used to transfer genes
into laboratory cultures containing human
cells. After allowing the cells to grow, they
checked to see if Sleeping Beauty had been
inserted into their genomes. The results were
positive. Not only had the delivery and insertion
parts of the molecule worked, there was
evidence that the gene had copied itself and
jumped to new positions in the chromosomes.
Part Two: Identity crisis
��
The couple brought along transposons
when they joined the MDC in 1999.
Zoltán was given a position as a group leader,
and Zsuzsanna took some time off to give
birth to twin boys – they both give a tired
144
smile (now) about the difficulty of managing
a family and their careers. Shortly afterwards,
Zsuzsanna won a position of her own as part
of a prestigious award from the European
Union: the European Young Investigator
Award. Competition was tough, she says; of
880 applications, only 25 researchers were
ultimately given funding. The grant helped
her establish her own laboratory at the MDC.
The successful resurrection of Sleeping
Beauty suggested that it might be possible to
develop a palette of transposons which could
be used in different organisms. This would
widen their potential as genomic tools. The
labs went on to try to isolate new active
transposons from vertebrate species. They
found that Tc1/mariner elements in the
genome of the Northern leopard frog were
unusually well conserved and contained only
a few mutations. Csaba Miskey, a postdoctoral
fellow in Zoltán’s group, needed to fix only
two mutations in one of the transposon
copies to make it jump again. The new transposon,
keeping the Grimm brothers’ nomenclature,
was named Frog Prince.

In 2006 the two labs produced yet another
working transposon called Hsmar1. “We think
that one of our distant ancestors, a primate,
originally acquired this gene about 50 million
years ago, probably from a virus,” Zoltán says.
“It reproduced itself heavily for about 13 million
years until mutations had disarmed all
the copies. It had already spread quite far. The
human genome alone contains about 200
inactive copies.”
Csaba decided to try to rebuild the original
Hsmar1 gene. Csaba and Balázs Papp, a specialist
in biocomputing at the University of
Manchester (UK), began analyzing the
“spelling” of human Hsmar1 genes.
Comparing different versions revealed common
features that had probably belonged to
the first Hsmar1, and this allowed them to
reconstruct the DNA sequence of the original
gene. On the average, they discovered that
about eight percent of each gene copy had
changed through mutations.
One of the molecules, however, had undergone
far fewer mutations: a gene called SET-
MAR had changed less than three percent.
SETMAR is a member of the Hsmar1 family. It is
an unusual case where cells still use it to
make RNA and proteins, but it no longer possesses
the cut-and-paste tools to allow it to
jump. “Some transposons took a different evolutionary
course,” Csaba says. “When they suffered
mutations, instead of just becoming
junk, they became domesticated – they
remained genes and took on other functions.
We’re not completely sure what those are in
the case of SETMAR, but there’s some evidence
that it helps repair breaks in DNA.”
Recently a laboratory at the University of
Oxford had shown that the SETMAR protein
retained at least partial transposon activity.
This suggested that rebuilding the original
might create a functional transposon.
Csaba and his colleagues repeated the strategy
that had been used with Sleeping Beauty
and Frog Prince: analyzing Hsmar1 copies in
search of letters that had changed the least,
a.
b.
The Sleeping Beauty transposon system.
a. The components: Orange = the target gene that scientists want to move. It is engineered to lie
between DNA sequences (white arrows) that can be recognized by the DNA-cutting transposase
molecule (purple).
b. When the transposase and target gene are produced in the same cells, the transposase recognizes
the binding sites and docks on. It draws them together and cuts the ends, releasing the target gene.
Cellular mechanisms repair the break in DNA and insert the transposon into a new site.
then building a new molecule with the most
frequent spellings. But their new version of
the gene didn’t jump.
Returning to the drawing board, Csaba and
Zoltán thought they knew why. The situation
was a bit like going to a city and counting the
number of blue-eyed and brown-eyed people,
finding a majority of citizens with blue eyes,
then coming to the logical conclusion that
the founders of the city must have had blue
eyes. But that might not be true. If in the
intervening generations, people with blue
eyes had a lot more children, this might skew
the Tobias results. Jursch The same thing could be true of
Hsmar1. A particular mutant version of the
145 Part Two: Identity crisis
molecule might have reproduced more than
others. If so, the “average” DNA sequence
would be biased in that direction.
Csaba tried a new type of computer analysis
in which, instead of building an average
sequence from all the Hsmar1 molecules, he
reconstructed their family tree. This revealed
four spelling differences between the “averaged”
DNA sequence and the original. When
the scientists incorporated the older spellings
into their artificial gene, they now had a working
transposon.
Along with the 200 inactive versions of
Hsmar1, the human genome also contains

about 2,500 copies of a fragment of the gene
called MITE. Further experiments with the
reconstructed gene showed that the ancestral
Hsmar1 was probably also responsible for
copying and shuffling these fragments
through the genome. The new version of the
gene made MITEs jump at a much higher rate
than the entire gene, supporting a hypothesis
that smaller elements can be moved easier
than larger ones.
The new transposon shows a high level of
activity, Zoltán says; it is just as efficient as the
two other transposons that the group has
rebuilt. This means that it should be possible
to turn the gene into a new tool to investigate
gene functions in mammals and other organisms,
just as Sleeping Beauty has been made
into one.
��
Transposons have surely been a big force
in the history of life. By expanding the
size of genomes, they have introduced a lot of
raw material for evolution to work on.
Mutations have turned some of the
sequences into new genes; over 100 human
molecules are thought to have been domesticated
from parts of transposons. One of them
is HARBI1, a molecule that evolution has conserved
in all vertebrate species. It originally
evolved long ago as part of a transposon in an
early ancestor, probably a fish.
“HARBI1 stems from a family of genes called
PIF/Harbinger that is found widely in vertebrates
and plants,” Zoltán says. “There are
copies in maize and rice which still work, but
as far as we know there aren’t any active
forms in vertebrates.”
Reactivating the molecule might shed light
on how it and other transposons have evolved
into useful genes like HARBI1. There was
another reason Zoltán, Zsuzsanna and their
colleagues were interested in the transposon:
it encoded a second molecule with unknown
functions. Its architecture closely resembled
another human protein called NAIF1. Any part
of a transposon that wasn’t understood
might give new insights into their behavior
and evolution.
HARBI1 was taken on by a French postdoc
named Ludivine Sinzelle, whom the scientists
had met in the course of an EU grant. “There is
not a very large transposon community in
Europe, and we are quite scattered about –
such grants are great opportunities to bring
people together,” Zsuzsanna says. “Ludivine is
a wonderful person – extremely nice and
cheerful. She has just left and everyone was
sorry to see her go. I met her at a conference
and we decided that she should come to the
MDC. I found two months of money from one
grant, and Zoltán put together four more
months, and finally we had nine months and
she could come.”
The nine months stretched into two years;
although she wanted to return to France, the
project was going so well that Ludivine couldn’t
bring herself to leave. Her reward was the
completion of a working version of the
Harbinger transposon. The labs have used it to
analyze the functions of its two genes.
One result is a better understanding of where
the transposon becomes inserted into the
genome. Some of the genes jump to random
places; others insert themselves at specific
DNA sequences. Vladimir Kapitonov, who had
done most of the computational work on the
project, had looked at the neighborhoods of
the fish transposons and found a 17-letter
code that seemed to be the target. But it was
hard to be sure; these sequences, too, had
undergone mutations.
The new, intact Harbinger allowed the scientists
to test the insertion code in living cells.
The gene had jumped 46 times in their experiments,
and Ludivine carefully inspected the
sites where it landed. Her computer analysis
revealed a sequence of 15 letters that told the
gene where to insert itself. “Twelve of those
are the same in the 17-letter code in
zebrafish,” Zoltán says. “This is interesting
because it means that even when you trans-
Part Two: Identity crisis
146
plant the gene from one species to another, it
looks for the same target.”
This fulfills one aim of the project: to produce
tools that can be used to insert foreign or
altered genes into new species. It also puts
them one step closer to another of the scientists’
goals – to turn transposons into tools for
medical therapies.
“Molecular biologists all over the world hope
to develop gene therapies – methods of delivering
healthy versions of genes to people with
genetic diseases or other illnesses,” Zoltán
says. “Perhaps the greatest barriers we have to
overcome are the problem of delivery – getting
foreign DNA or RNA into the right types
of cells – and then getting molecules to
behave the right way once they have arrived.
This means that healthy molecules need to be
inserted into the genome at specific places, so
that they don’t interfere with the activity of
another gene.”
A clinical trial conducted in hospitals
throughout the world in 2002 and 2003
showed how crucial this is. Researchers were
using a retrovirus to deliver a protein called
the interleukin 2 receptor to patients suffering
from a severe immune system disease
called X-SCID. People with defective versions
of this molecule are unable to produce B and
T cells, which allows even trivial infections to
run rampant in their bodies. They have to
spend their life in a sterile “bubble” unless a
bone marrow transplant from a close relative
takes hold.
But in several patients the virus inserted the
gene’s DNA into the worst possible position,
near the control region of a gene called LMO2.
This molecule is a proto-oncogene, which
means that it is known to cause cancer if it
undergoes changes. Somehow the arrival of
the gene triggered an uncontrollable wave of
replication among T cells – basically a
leukemia-like form of cancer.
“The only way to avoid these problems is to
know exactly where a gene will land in a

patient’s DNA,” Zoltán says. “That could be difficult
with a virus, which is very complex – transposons
are much simpler.” Ideally, a transposon
could be “programmed” to insert itself into precise
places in the genome by rewriting the tools
that told it where to go. This is a crucial step in
turning the molecules into therapeutic tools. In
June 2007, Zoltán and Zsuzsanna used Sleeping
Beauty to prove that this could be done. That was
an important step toward obtaining a new grant
from the European Union, which is uniting the
efforts of transposon researchers from several
countries to develop the genes into therapeutic
tools.
In a completely different approach, transposons
are not simply used as harmless delivery vehicles,
but are designed to be highly mutagenic. Indeed,
transposable elements can insert themselves
into many places in the genome that can cause
mutations. This is how Barbara McClintock discovered
them in maize: jumps of transposons
resulted in changes of the color of the kernels.
After the completion of several genome sequencing
projects, the new challenge is to understand
the function of each gene. “Transposons can be
exploited as discovery tools,” Zsuzsanna says,
“and we are adapting them to do systematic
studies of the genome.”
Most of the studies in this book involve disturbing
a gene’s functions through knock-outs and
then studying an organism to see how it gets
along without the molecule. Many of the largescale
methods for doing so involve chemicals that
cause DNA to mutate; when a gene is non-functional,
researchers can study its effects on the
animal. But chemical mutagens often alter only a
few letters in the code of a gene, which makes
these approaches time-consuming, Zsuzsanna
says. “It takes a lot of work in the laboratory to
identify the mutated gene– and often you need a
very large screening facility to look at the animals.
That is different if you disrupt the gene
with a transposon, because the mutated gene
will be tagged and you can use this label to find
out where it has gone just by scanning the
genome for the tag.”
147 Part Two: Identity crisis

Andras, Zuzsanna, Zoltán, and Zsombor
Once the transposon has been inserted into
an organism’s genome, it can be used to spin
off new mutants. It can be made to jump to a
new location simply by breeding an animal
that carries the transposon with another that
has the molecules needed to cut it out and
paste it somewhere else. Each offspring that
inherits both parts of the toolbox will be like a
new genetic experiment: the transposon may
disrupt a new gene by jumping somewhere
else. One can imagine the power of this
approach: each sperm can carry a new mutation.
Extra information can be packed into the
transposon so that it becomes activated only
in certain tissues or at particular stages of an
animal’s development. Zsuzsanna and Zoltan
are adapting their transposon tools for systematic
discovery of genes that are thought
to play a role in human disease. They are concentrating
their efforts on genetic screens to
uncover genes involved in different types of
cancer and cardiovascular disease.
��
Even for a couple working on common
projects in neighboring labs, life is not all
about science at the bench. In the little spare
time they have, Zoltán and Zsuzsanna have
Part Two: Identity crisis
148
been encouraging the development of molecular
biology in their native country. In 2007
the journal Nature printed a letter they wrote
about the current situation in Hungarian science.
The letter was inspired by a news article
about a government plan to cut budgets,
restructure its science academies, and privatize
parts of the national research infrastructure.
Both scientists feel a strong attachment to
their home country. “On the positive side, we
received an excellent education there,”
Zsuzsanna says. “Hungary continues to produce
very good students and postdocs. But
then they all leave to take jobs abroad. The

country does not offer good opportunities for
senior scientists, and the proposals that are
circulating now will further cut the already
limited amount of money that is spent on
basic research. This is amazing in a country
which has received the highest number of ‘per
capita’ Nobel prizes.”
The Hungarian academy was founded in 1825
to ensure the long-term scientific interests of
the nation. The new policies run directly
counter to that aim. “With its short-sighted
outlook,” their letter says, “the government
does not realize that its ‘reform’ activities
undermine the development of a healthy
innovation chain, and drown the personal creativity
that has been an asset and a source of
pride for Hungarian science.” Zoltán and
Zsuzsanna were among 2,500 scientists who
signed an open letter to the prime minister
urging that budgets not be cut.
Running labs, watching out for Hungary’s
interests, and taking care of twins leaves very
little time for anything else. Still the couple
tries to keep a balance. We talk for a while
about the difficulty of managing a family life
and two scientific careers.
“We never discuss science after ten p.m.,”
Zsuzsanna says.
149 Part Two: Identity crisis
“Actually it’s earlier than that,” Zoltán says.
“Usually.”
“Yes, if we talked about it all the time, the
twins would be left out.”
I try to imagine sitting at the dinner table
with two seven-year-olds when the discussion
turns to the nuances of jumping genes.
As good as Zoltán and Zsuzsanna are at
explaining their work in simple terms, that
would surely make the twins do some jumping
of their own.
Fairy tales would surely work better.

Interlude:
These rats don’t dance
to the Red Hot Chili Peppers
If you’re ever on the road in Central Eastern Africa with a shovel and
feel the need to dig a hole, you’re likely to strike a tunnel or burrow
made by one of the strangest creatures in the world: the naked mole rat,
which is neither a mole nor a rat. Its hairless, wrinkled appearance is
only the first of its oddities. Where you would expect to see an eye there
is only a small black hole – the rodent is blind. Ears are knobbed bumps
with no openings. It darts around frenetically on thin legs that carry it
equally quickly whether moving forward or backwards.
The naked mole rat prefers to live out its life huddled with its brothers
and sisters in huge masses. It has a social life like that of bees. A colony
has one queen and two or three fertile males; the rest are drone workers.
It is the world’s only cold-blooded mammal. It is capable of shutting
down its metabolism to a low rate, which is probably responsible for its
very long lifespan; the animals have been known to live for up to 25
years. The time that isn’t spent sleeping, or huddling, is usually spent
digging. That’s how the colony finds the roots and fleshy tubers that it
feeds on. (It also eats its own feces, and nourishes its offspring on
them.)
Interlude: These rats don’t dance to the Red Hot Chili Peppers
150

Gary Lewin

Underground, in a confined space that may be
inhabited by up to 300 densely-packed
rodents, the air is quickly exhausted. Huge
amounts of carbon dioxide accumulate,
enough to kill any other mammal. But the
naked mole rat survives. This suffocating
atmosphere may contribute to one the
strangest things of all about the animal. It
lives in a world without sight – and largely
without pain.
Pain is so fundamental that sometimes it
seems the world is divided into two categories:
things that cause it and those that
don’t. Interestingly, most animal species find
the same things painful – a pin-prick, a hot
flame, or acid. We are all too familiar with the
sharp pains of wounds and inflammations.
And then there is capsaicin, the active ingredient
in chili peppers. When it comes in contact
with the tongue, the skin, or any other tissue,
it burns. Birds don’t notice it, though, which is
good for the plant – they propegate its seeds.
Christiane Wetzel, M. Benkovic, Jing Hu, Gary Lewin
And the naked mole rat barely notices it at all,
even when the substance comes in contact
with an open wound.
��
Gary Lewin’s interest in chili peppers is
more scientific than culinary. The main
topic of his work at the MDC is how nerves
detect and pass along sensory information
like pressure and pain. “One reason that capsaicin
is interesting is that we think it stimulates
the same system, at least partly, which
transmits the pain of inflammation,” he says.
“We knew that some of the elements of that
system were lacking in the naked mole rat. So
alongside our work with more traditional laboratory
animals such as mice, we’ve recently
been developing this new unusual creature as
a model to study pain.”
So if you want to see one of the animals, you
don’t have to go to Africa. There are some in
the basement of the Max Delbrück house.
Gary leads us down long, dimly lit corridors, a
labyrinth of grey walls and doors. Somehow
he finds the right one and lets us into a small,
cramped room. Two naked mole rats scurry
around in a glass cage. It’s warm in the room
and warmer in the cage – a prerequisite to
keeping the animals alive. He reminds us that
they are cold-blooded.
Seven or eight of us have accompanied him
downstairs, a spontaneous mid-day field trip
organized by Elisabeth Kujawa-Schmeitzner
of the director’s office. There are two administrators
from the PhD program, two more from
the conference office, and the grants officer.
None of them scientists, but they have all
heard Gary talk about the animals and their
curiosity has been piqued.
After having seen pictures of the creatures –
which don’t conform to any standards of
human beauty I know of – it’s a surprise to
meet them in the flesh. They are about eight
centimeters long and, well... sort of cute.

“Oh, they’re great,” Gary says. “You can watch
them all day. Here, want to hold one?”
They dart around in the cage fast, somehow
reminding me of hyperactive lizards because
of their thin legs and their odd way of dashing
forward, stopping, and zipping off backwards
at the same pace. He reaches in and grabs
one, picking it up unceremoniously by the tail
and letting it scrabble up and down his
sweater. When he puts it on my hand it darts
for an open sleeve. Gary grabs it by the tail
again.
“They’d make great pets,” he says. “People have
considered it. But they’re not easy to take care
of because of the temperature issue.”
They would also never come when you call –
being almost deaf. On the other hand, they
would have no problem with my son’s stereo,
which has been getting louder and louder
over the past few months. He’s trying to make
me enthusiastic about the new band he likes.
Ironically, they’re called the Red Hot Chili
Peppers.
��
Understanding why the rodents have lost
their eyes and some of their pain
requires a bit of time travel. It was almost
exactly 150 years ago that a group of English
scientists shook the water off their umbrellas
and shed their soggy coats as they entered
the Royal Geological Society in London. They
were gathering for a meeting of the Linnean
Society, the prestigious British science club,
whose members included some of the world’s
best naturalists. They met every month to
read papers to each other and brag about
new species of beetles they had discovered
under moldy logs. Few of them had any idea
of what they were about to witness at the
meeting held on July 1, 1858. It took place in
the Burlington house on Picadilly, in an ornate
upstairs room whose walls and high ceiling
were decorated with gilded carvings. Charles
Darwin was not there – he had retired with
his family to mourn the death of a young son
– and Alfred Russel Wallace was chasing the
“bird of paradise” on the opposite side of the
globe, hoping to save himself from financial
ruin by selling it to some rich British collector.
In their absence, friends read their papers to
the Linnean Society, and that is how evolution
was announced to the world.
One of the things said that night was that
animals don’t acquire new hereditary traits
because they need them, or lose features like
eyes because they live in the dark. The naked
mole rat has not become blind because it
spends most of its life in an underground burrow.
Instead, in a cramped world occupied by
hundreds of expert diggers with sharp claws,
packed close together, eyes are easy targets
for scratches and infections. Ears are also likely
sources of infection, more trouble than they
are worth in the underworld. An animal in
which the organs had disappeared or become
covered over with skin would probably have
better chances of surviving to pass along its
genes.
The lifestyle would also confer benefits, Gary
Lewin says, to an animal that had lost its sensitivity
to certain types of pain. “The problem
was the air,” he says. “When so much carbon
dioxide builds up in the environment, the
body becomes very acidic. Acid is painful to
every other known species of mammals, fish,
and birds. But living the way the naked mole
rats do, there would be strong evolutionary
pressure to tune down or even eliminate the
systems that cause such pain.”
The same mechanisms that made the naked
mole rat insensitive to acid might reduce
other types of pain. “There are two main types
of pain associated with inflammations,” Gary
says. “You know how it is if you get an infected
finger that gets red and swells up? That
finger becomes extremely sensitive to pressure
and heat, much more than the other fingers.
That’s missing in the naked mole rat. It
feels pressure and the pain of being pinched.
But heat doesn’t hurt, and it doesn’t become
hypersensitive.”
Finding out why hasn’t been easy. There are a
lot of steps between the skin, where pain originates,
and the brain, where it is registered.
Gary’s research shows that evolution has tinkered
with the system at several levels.
��
Pain begins when wounded cells release
molecules that dock onto neighboring
neurons. This usually opens channels in their
membranes that let in charged atoms like calcium
or sodium. The result is a change in the
balance of charges between the inside and
the outside of the cell, turning the neuron’s
membrane into a fantastic conductor of electricity.
An electrical impulse shoots along the
nerve, sometimes traveling fantastic distances
to the other end of the cell. When they
reach the synapse, a small gap that separates
the nerve from its neighbor, the cell releases
small molecules called neurotransmitters. By
docking onto the next nerve, they cause it to
open and close its own channels, allowing it
to forward the message on to the brain.
Cells that receive pain are called nociceptors.
The main body of each cell, the part containing
the nucleus, lives near the spinal cord in a
cluster of nerves called the dorsal route ganglion,
or DRG. It is wired into the nervous system
in two directions. It extends an axon into
the spine, where it transmits signals to nerves
that carry them on to the brain. Another
extension grows huge distances to the surface
of the skin – sometimes a meter or more
153 Interlude: These rats don’t dance to the Red Hot Chili Peppers

to reach a person’s toes. (The nervous system
is built the same way in a giraffe, where single
nerves can be over four meters long.) At the
tips of these cells are receivers which sense
the molecules that cause pain.
Most of what is known about pain in the
naked mole rat has come from comparing this
system to a more traditional laboratory model
– the mouse. Gary’s group has been equally
successful at uncovering new sensory mechanisms
in this other small mammal – mechanisms
that are usually just as important in
humans. At the end of 2007 a collaboration
with the lab of Alistair Garratt turned up an
important mechanism behind the pain detection
systems in mice and other “normal”
mammals.
Alistair is a British scientist who came to the
MDC in 1996, as a postdoctoral fellow with
Carmen Birchmeier. “I had studied in
Cambridge and was looking to leave the UK. I
happened to see an advertisement that
Carmen had published in the journal Nature –
she was moving from Cologne to the MDC
and was looking for people to join the new
lab.” Coincidentally, Alistair had studied
German – inspired by a Swiss girl he used to
Alistair Garratt
take dancing lessons with. And Berlin seemed
like a good place to be.
Alistair has known he wanted to be a scientist
for a long time. One of his first memories is of
an experiment done by his father, a chemistry
and biology teacher. “He combined lithium
and water – it fizzed and caught fire,” Alistair
says. “I couldn’t have been more than three
years old, but I couldn’t take my eyes off it.”
That first spark of interest was nurtured
through his childhood by, among other
things, television programs. The Open
University of London broadcast lectures about
science. After graduating from high school,
Alistair went on to study science at the
University of Cambridge, then got his PhD in
Manchester.
In 2002, after several very successful years
working with Carmen, Alistair received one of
the much sought-after Helmholtz fellowships.
This allowed him to set up a small, independent
group in the MDC’s program on
Functions and Dysfunctions of the Nervous
System. Part of independence meant finding
new projects to work on. He had been interested
in the body’s pain circuitry and talked to
Gary Lewin and Carmen about how to find
new genes involved in the sensory system.
One possibility was to use DNA chips
(described in the story “Stem cells in a jar.”)
The technology would compare the genes
active in nociceptors to those in other nerves,
hopefully revealing those that were important
in pain sensitivity.
There were a few standouts. When Alistair
brought the list along one day to lunch, Gary
immediately recognized one of them: a receptor
protein called c-Kit. “There’s a story behind
c-Kit – it actually has to do with a cat,” Alistair
says. “A cat from the state of Missouri in the
U.S. – someone’s pet – that had cancer. This is
one of those forms of cancer caused by a
virus. Among other things, the virus brings in
its own, defective form of the Kit gene. That
pushes out the cell’s version of Kit, which
causes problems, because the virus molecule
can’t understand signals that the cell needs
to receive.”
c-Kit especially interested Alistair because it
closely resembled other receptors in nerve
cells that he had worked on; it might function
the same way. Gary was interested because it
might be another link in the pain network. In

1993, as a postdoc at Stony Brook University in
New York, he had discovered that nociceptors
had to be stimulated by a protein called NGF
(which stands for nerve growth factor) to survive
as the embryo developed. Later in life NGF
switched functions and seemed to be
involved in making animals hypersensitive to
pain. This condition, called hyperalgesia, is
what happens during inflammations.
Gary discovered that there were two very similar
types of nerves in the dorsal root ganglion.
NGF was one of the few differences:
only half of the nerves produce it. These
seemed to be the same nerves that transmit
hyperalgesic sensations. Identifying the molecules
and pathways that make such similar
nerves behave differently has been high on
Gary’s research agenda ever since.
Alistair suggested that one of his PhD students,
Christina Frahm, take on the project. She was
working on a protein that was a lot like c-Kit,
and it would make a good backup in case her
findings weren’t important enough to write up
as a dissertation. Gary also put one of his students,
Nevena Milenkovic, on the project. One
of the first things the lab needed was a strain
of mouse that didn’t produce c-Kit.
“We could have made the knockout ourselves,”
Alistair says. “But that takes time, and
strains already exist. They don’t survive too
well, because signals from c-Kit have several
different functions. The molecule that activates
c-Kit is called SCF, for stem cell factor,
and you can imagine from the name that it
has important jobs. For example, SCF triggers
c-Kit to guide the development of some types
of blood cells. Animals that don’t have it
become quite anemic and usually die by the
time of birth or very shortly thereafter.”
But Max Gassman’s group in Zürich had been
working with these mice and found a way to
keep them alive. The animals could be “doped”
with a molecule called erythopoeitin, a hormone
that stimulates the production of new
red blood cells. The treatment would disqualify
the mice for the Tour de France, but it
extended their lifespans. Alistair had recently
invited Gassman to the MDC to give a talk;
now he contacted the researcher again to ask
to borrow some mice. Gassman was willing to
help out.
“Sometimes the animals that we use are
more international than the scientists,”
Alistair says. “The mice we got came from the
University of Ulm, who had obtained them
from a group in Japan, who had originally
received them from the Jackson Laboratories
in Maine.”
��
One job of SCF, the trigger for the c-Kit
receptor, is to help guide the development
of the skin. So the protein is produced by
cells in the skin which lie right alongside the
nerves that sense pain. Everything seemed to
suggest that the two molecules were involved
in pain transmission, so Christina put
Gassman’s mice to the test. She found that
they didn’t respond strongly to heat pain, but
were oversensitive to pressure.
The next step was to have a look at the nerves
themselves. They were structured the same
way in mice with and without c-Kit, so the loss
of the molecule didn’t seem to be causing any
rewiring of the nervous system. By staining
cells in normal mice, Christina discovered that
the nerves which produced c-Kit were the
same as those that required NGF signals. And
those cells also bear a molecule called TRPV1 –
a channel protein that lets charged atoms
into a nerve. This is what creates the impulse
that turns the cell into a pain transmitter.
Gary had been working on TRPV1 for a long
time. Without the molecule, mice aren’t bothered
by heat or capsaicin, the substance from
chili peppers, and they don’t become hypersensitive
to pain.
Nevena and Christina examined the animals
and discovered that they didn’t seem to feel
heat pain caused by inflammations. On the
other hand, they were more sensitive to pressure
caused by inflammations than normal
mice. These effects might be due to any of a
number of things. The loss of c-Kit might
change the types of nerves that developed or
cause other structural changes to the nervous
system. But dissections under the microscope
revealed that this was not the case. The
nerves that transmit pain are called C-fibers.
There seemed to be the same number in the
mutant mice, and they seemed to be wired
into the nervous system the same way.
“If those things were true, what we needed to
check was whether the nerves were able to
155 Interlude: These rats don’t dance to the Red Hot Chili Peppers

transmit signals at the same strength,” Gary
says. One of his group’s specialties is the ability
to extract nerves and test how strongly
they conduct particular types of impulses, like
determining the strength of a battery by
measuring the current that passes through a
wire. When Nevena performed this experiment,
she discovered that heat signals in mice
without c-Kit were much weaker than those
in normal mice.
The data suggested that the scientists were
dealing with two types of pain receivers. Some
responded very strongly to heat and others
didn’t. The behavior of the mild responders
didn’t change much without c-Kit, but there
was a big change in the strong responders.
“This offered a possible explanation for the
difference between mild reactions to heat
and the strong pain that is felt in hyperalgesia,”
Gary says. “There are two types of C-fiber
nerves that transmit heat information. Both
are activated the same basic way, but one
responds much more strongly. It’s a bit like
having the same CD player hooked up to two
sets of speakers. Both get the same signal. The
cheap speakers never get very loud, but on the
expensive set, turning up the sound just a little
bit boosts the noise so high that it keeps
the neighbors awake at night.”
Was c-Kit turning on an amplification system
in the second set of nerve fibers? One way to
test the hypothesis was to shut down c-Kit in
normal mice. Gary knew of a drug that could
accomplish this. When it was administered to
normal animals, they no longer experienced
heightened pain from heat during infections.
A second test took the opposite approach. SCF
seemed to be passing the signal during an
inflammation. What would happen in a
mouse whose nerves detected high amounts
of SCF even without an inflammation? When
the experiment was performed, mild
amounts of heat caused intense amounts of
pain.
The results raised more questions – for example,
whether SCF and c-Kit directly provoked
heightened pain. “We already knew that
hyperalgesia depended on the TRVP1 receptor,
the sensor for capsaicin. We had a strain of
mouse that didn’t have the receptor, so we
started to look for a connection.”
The researchers repeated the experiments
with mice that lacked TRVP1. If the receptor
were lacking, the pain level didn’t rise, even
with more SCF. This meant that the signal
through c-Kit only worked if the receptor did
its job.
“The experiments have shown several new
links in the chain that causes intense pain
associated with heat,” Gary says. “In healthy
animals and humans, those links are intact.
Each one of them is a point at which it might
be possible to use a drug to intervene and
block the signals.”
He is quick to point out that SCF and c-Kit are
not the holy grail of pain research. “Removing
the receptor greatly reduces sensitivity to
heat, but it has another effect that seems
contradictory,” he says. “Mice actually become
Sören Markworth

much more sensitive to pain caused by pressure.
This is equally interesting. It shows that
parts of the same system regulate a second
type of pain in a different way. We’re still
working on how it does that.”
��
How much of what had been learned in
mice could explain what was going on
in the naked mole rat? Thomas Park and his
colleagues in Gary’s group have discovered
that evolution has worked at several levels to
reduce the rodent’s sensitivity to pain.
They are the only mammals known to lack
two proteins called SP and CGRP in pain receptor
nerves in the skin. Other labs had shown
that if these molecules were removed from
mice, the animals were much less sensitive to
some types of pain. “This convinced us to take
a systematic look at the behavior and biology
of pain in the rats,” Gary says. “We discovered
that they are the only mammals that don’t
sense acid; they don’t become oversensitive to
Cells stimulated by capsaicin, the
ingredient that makes chili peppers
hot, lie near the surface of the dorsal
horn in the mouse. In the naked mole
rat most of the nerves like deeper.
Where the stimulation takes place
may determine whether a sensation
is painful or not.
heat during inflammations, and they don’t
respond to capsaicin.”
But naked mole rats have TRPV1, the capsaicin
receptor, and it works – the nerve that is stimulated
by the substance receives and passes
along a signal. The lab’s measurements
showed that the strength of the impulse is
just as strong as in mice. This meant that
pain from the peppers – and from inflammations
– was being interrupted somewhere
higher in the network.
This was different from the situation in birds.
They also have TRPV1, but its architecture is
somewhat different. The small changes
means that capsaicin doesn’t dock on – the
nerve isn’t told to feel pain, and it doesn’t generate
an electrical impulse.
So where does the rodent’s insensitivity come
from? Thomas followed the signal to the
spinal cord – maybe the sensory nerves
weren’t wired up properly. Dissections
showed that there was a difference in the way
sensory nerves linked to their partners. In
157
These images show a cross section
of the dorsal horn region of the
spine. Synapses that bear the
pain-transmitting molecule TRPV1
are stained in red. Left: In the mouse,
nearly all of these synapses from
sensory neurons are found near the
surface of the dorsal horn. Right:
Cells producing TRPV1 are found
much deeper in the naked mole rat.
mice most nociceptor nerves – the pain
receivers – lie near the surface of the spinal
cord. In naked mole rats, most are much
deeper.
Previous studies have shown that when most
mammals experience pain, the greatest stimulation
of these nerves happens close to the
surface. Gary hypothesizes that balance may
be the crucial factor: if more signaling happens
at the border of the spinal cord, information
from the skin will be interpreted as pain.
If the most activity is deeper inside, as with
the naked mole rats, it will not.
All of us would welcome a method to reduce
pain from inflammations and other sources.
These studies reveal several different junctions
at which pain is perceived and passed
along; any of them might make good places
to interrupt its transmission. Understanding
why the naked mole rat doesn’t hurt, Gary
thinks, may reveal why the rest of us do.
Interlude: These rats don’t dance to the Red Hot Chili Peppers

Part Three:
Frontiers and ferrymen

The Swiss Army knife
of the cell
My personal favorite is the CyberTool 34, with the “translucent sap-
Part Three: Frontiers and ferrymen
phire” handle. It is outfitted with two knife blades (small and
large), and a corkscrew, can and bottle openers, screwdrivers, a wire strip-
per, a reamer, toothpick, sewing eye, bit wrench/case, tweezers, nine sock-
ets and bits, a ball point pen, pliers, a wire cutter/crimper, scissors, hook,
160

Thomas Willnow
mini screwdriver, straight pin, and key ring.
What else could you possibly need?
Admittedly, the laser pointer and flash memory
stick, which come with the “Swiss memory”
knife, might be useful.
The tool has come a long way since Karl
Elsener made the first one in 1891; he was
upset that his country’s army had outsourced
its knives to Germany. His original model
came with two screwdrivers, a can opener and
an awl. (Officers were given an enhanced version
with a corkscrew.) Today it has become a
standard part of the repertoire on expeditions
to Mount Everest and the space shuttle,
although you have to wonder how astronauts
open the knife while wearing those thick
gloves. Art historians apparently find them
useful because they’re standard equipment at
the Museum of Modern Art in New York City.
And then there is the case of John Ross, doctor
in Uganda, who had to use one to perform six
emergency amputations when someone stole
his surgical saw.
The Wenger company, which shares the manufacture
of Swiss Army knives with Elsener’s
company Victorinox, recently produced a collector’s
edition called the Giant, which is all
versions wrapped into one. It contains all 85
tools ever found in a Swiss army knife, including
instruments to clean the cleats of golfers’
shoes and the heads of their clubs. It’s 22 centimeters
thick, weighs nearly a kilogram, and
costs about a thousand Euros. You’d need a
wheelbarrow to carry one around. I think I’ll
stick with the CyberTool.
161 Part Three: Frontiers and ferrymen
��
As large as the Giant is, it’s a fraction of
the size it would be if Karl Elsener hadn’t
invented a special spring mechanism that
allowed him to pack tools into both sides and
both ends of the knife. That innovation is
what allows a soldier (or an art historian) to
walk around with one knife instead of five or
six. And it’s one reason that Thomas Willnow,
researcher at the MDC and professor at the
Charité, sees the Swiss army knife as a
metaphor for his favorite family of proteins.
The molecules Thomas has been working on
are related to a protein called the low-density
lipoprotein receptor. So they are called lowdensity
lipoprotein receptor related proteins
(LRPs). It’s a daunting name, Thomas admits.
One of the functions of the low-density
lipoprotein receptor and its LRP relatives is to
regulate the amounts of cholesterol and
other lipids (fats) in the bloodstream by drawing
them into cells.
“Lipids are the major component of cell membranes
and are used for many other things,”
Thomas says. “Cholesterol, for example, is
modified to make vitamin D and several types
of hormones, including the sex hormones
estrogen and testosterone. It’s also used as a
glue to hold molecules in the membrane
together. And you know of its potential negative
effects: if your cells don’t remove it from
the bloodstream, it builds up in the linings of
blood vessels and causes a clogging of your
arteries and many other problems. So lipid

The LDL receptors (left) and Sortillin gene families have many members.
Evolution has modified their structures by adding on and removing
modules, giving the proteins a range of functions in many different tissues.
and cholesterol metabolism are fundamental
to a wide range of biological processes.”
Most of the body’s cholesterol is synthesized
by cells in the liver and a few other organs;
some is absorbed through food. It is usually
transported from these sources to the rest of
the body through the blood. Because fat does
not mix with water, cholesterol and other
lipids have to be wrapped into specialized
cargo complexes for transport. The cargo
complexes for fat are low-density lipoproteins
(LDLs) or similar types of protein carriers.
A single LDL can hold about 1500 cholesterol
molecules. When the package arrives, it is recognized
by a receptor on the target cells – the
LDL receptor or an LRP. The cell engulfs the
protein with its cargo, and carries it to internal
compartments called endosomes, where it is
unpacked. This process of drawing in molecules
is called endocytosis.
Part Three: Frontiers and ferrymen
162
“Endocytosis is the main way that cells control
the entry of molecules they need,”
Thomas says. “A receptor binds to an outside
molecule and draws it inside. The cell builds a
wide range of receptors that behave this way.
One reason there are so many is that various
tissues need to import different types of molecules,
and specialized receptors help them
do so. But we don’t yet have a detailed understanding
of what many of these proteins do.”
For instance, human cells, as well as those of
other mammals, make nine types of LRPs. They
are all spin-offs of a common ancestor, arising
through evolution and taking on new functions.
When Thomas arrived at the MDC 12
years ago, the first LRPs just had been identified,
but little was known about its functions.
“Each of the receptors is able to bind to combinations
of proteins and lipids such as cholesterol,”
Thomas says, “but it wasn’t known
where in the body and under what circumstances
they did so. Like the LDL receptor, they
may play an important role in recycling and
using fats. That’s a crucial process throughout
our bodies, and it is disturbed in a number of
diseases. So we decided to take a systematic
look at the whole family.”
He has a diagram of the nine LRP molecules
shown side-by-side. Each is lodged in the cell
membrane, pushing small, root-like tails into
the cell. Outside, they stretch away from the
cell like beanstalks. The LDL receptor is the
smallest of the family; some of its cousins
have much larger external modules. But they
have similar features, Thomas points out –
evolution has removed some subunits,
stretched others, and copied yet others, like
adding new tools or changing the shapes of
existing ones so they perform different jobs.
Some regions of the molecules help them
bind to the cargo. Others have roles to play
while escorting their partners into the cell.
One module lets them release a partner once
it has been delivered inside to endosomes.
It’s another parallel to the Swiss army knife,
although evolution didn’t start with the sim-

plest form of the molecule and develop
increasingly complex types. Instead, the oldest
LRP seems to be a huge protein called
megalin, which is more like the Giant. Parts of
it have been removed or rearranged to make
specialized tools for particular types of cells,
the way there are specialized knives for mountain
climbers and golfers. “That’s also how cell
types evolved,” Thomas says. “Things started
with a primitive unicellular organism that
gradually developed into a multicellular
organism with many specialized types of cells.
“As our lab and other groups have knocked out
the LRPs in mice, we discovered that they have
many jobs in addition to binding combinations
of lipids and proteins. That multi-functionality
might seem strange because the
LRPs are so similar. The versatility comes from
the fact that each part of the protein can bind
to various partners, with different effects on
the cell. The external region might bind to
lipid carriers, or signaling molecules, or a virus.
Or the LRP might bind to another receptor in
the cell’s membrane, changing the external
partners they can dock onto. And the tail
region of the molecule, inside the cell, can also
bind to different proteins. Each combination
has unique effects.”
I know of one lab that has spent six years
working on just the tail region of one receptor.
Isn’t it a daunting task to take on a whole
family of molecules?
“It is,” he admits. “We couldn’t do it without
the help of colleagues with expertise in various
systems, many of whom are on campus.
We’ve done most of our projects in collaboration
with other groups. Everything we’ve
learned tells us that to deeply understand one
of these molecules, we’ll have to understand
several of them, and their functions in various
parts of the body. It’s easier to understand the
CyberTool 34 if you’ve seen other types of
Swiss army knives.”
��
163 Part Three: Frontiers and ferrymen

Thomas and his colleagues have found
that some of the molecules have an
impressive range of jobs. Megalin is a good
example. High numbers of the receptor are
found in the proximal tubules, a crooked
pipeline in the kidney. Since this region of the
organ is responsible for capturing sugar,
water, amino acids, and a range of proteins,
then recycling them for use by the body, it
made sense that an LRP should be found
there. Exactly what the receptor recycled,
however, was unknown.
In 1996 the lab developed a strain of mouse
without the molecule in hopes of finding out.
They expected to see defects in the absorption
of molecules in the kidney. That turned
out to be the case, but first another issue had
to be dealt with.
“For some reason, without the protein the
brains of the mice didn’t develop properly,”
Thomas says. “The forebrain was so seriously
malformed that most mice died within a few
minutes of birth. This specific type of deformi-
ty is called holoprosencephaly – in which the
two hemispheres of the brain and the features
of the face become fused. Sadly, this
defect is quite common in humans. The most
extreme form of this condition is cyclopism, in
which an organism develops a single eye and
often no nose.”
A project carried out by Robert Spoelgen,
Annette Hammes, and Uwe Anzenberger in
Thomas’ lab has offered a possible explanation
for the problem. The brain defects arise
partly because of megalin’s ability to recognize
combinations of proteins and cholesterol,
and partly from its binding to other
receptors in the membrane.
“To make a proper brain with two hemispheres,
cells in the early embryo have to
receive a signal from a small protein called
sonic hedgehog,” Thomas says. “Sonic hedgehog
is secreted by cells. For some reason, cholesterol
has to be attached to it for the successful
transmission of a signal. The reason
hasn’t been completely clear, but it makes
sense that an LPR protein might be involved in
Part Three: Frontiers and ferrymen
164
recognizing such a signal from a cholesterolmodified
protein. After all, these are the
favorite cargoes of the receptors. Robert discovered
that the sonic hedgehog signal
requires the presence of megalin to work in
the embryonic brain. We think that megalin
teams up with another molecule – perhaps
the main receptor for sonic hedgehog, called
patched – to allow cells to recognize and act
on the signal.”
The study showed where the molecules meet
up – in a surface layer of the developing brain.
Signals from sonic hedgehog help tell cells
whether they are at the front or back of the

animal and where they should migrate.
Another function of the signal is to tell some
cells to die, which sculpts and separates layers
of brain tissue. Without the signal, these
processes don’t happen normally, and the
brain doesn’t form two hemispheres.
The capture of sonic hedgehog by the cell is
only the first step in sending a message to
cells’ genes, telling them to change their
behavior. Megalin’s tail may pass the signal
into the cell, and in 2004 Helle Petersen, a
postdoc in Thomas’ group, figured out
how that might happen. To do so she had to
“fish” for proteins that might bind to this
region of megalin by presenting it with cellular
molecules and watching what it grabbed
onto.
One partner she found was a new protein she
called megalin-binding protein or MegBP (for
obvious reasons) “That was interesting
because MegBP can interact with other regulators
in the cells, transcription factors. We
suspect that megalin has gene-activating
functions, and Helle showed that this is probably
the case. Its tail binds to MegBP. When
megalin receives a signal, its tail module
releases MegBP. The factor is now free to
move to the nucleus, where it may bind to
165 Part Three: Frontiers and ferrymen
Kristin Kampf
transcription factors and regulate activation
of a specific set of genes.”
��
Not all animals are equally affected by
the loss of megalin. In about five percent
of the knockout mice, the effects on the
brain are mild and the animals live to adulthood.
The survivors gave the lab a way to
study the receptor’s functions in the kidney.
Work in the test tube had shown that megalin
could bind to many molecules that are
supposed to be absorbed by the proximal
tubules. In 1999 PhD student Jörg-Robert

Salim Seyfried
Leheste and other members of the group
showed that megalin is the major receptor in
this part of the kidney that draws in hormones,
vitamins and other molecules bound
to carrier proteins. Many of these molecules
accidently slip through the filter of the kidney
because they are small – but the body invented
a backup mechanism to capture them
before they are lost in the urine.
That finding also suggested a link between
megalin and disease. If megalin is defective,
mice are unable to absorb crucial vitamins and
proteins which then leave the body in the animal’s
urine. Soon vitamins and hormones are
depleted from the body. The consequence is
severe vitamin deficiency. The same sequence
of events is seen in patients suffering from a
condition called Fanconi’s syndrome.
“But how can you identify the true binding
partners for megalin in the kidney in the first
place if you have no clue what they could be?”
Thomas asks. He draws a diagram in which he
represents the kidney as a simple tube.
“Substances enter on one side from the bloodstream
and leave on the other in the urine. The
receptors line the inside of the tube. Anything
removed by the receptors doesn’t come out
the other side. So we thought of a very simple
experiment to do with our knockout mice. We
compared their urine to that of healthy mice,
expecting to find extra substances which
would normally be removed by megalin. We
discovered that vitamin D3 and its carrier were
no longer being absorbed.”
Vitamin D3 is a hormone which helps cells
control their levels of calcium. “At the time it
Part Three: Frontiers and ferrymen
166
was still thought that most hormones simply
slip through the membrane on their own,”
Thomas says. “Our experiments proved that
vitamin D3 needs help. At first this finding was
considered an exception.” Follow-up studies
by another student, Regina Burmeister,
showed that megalin was binding to a second
receptor, called cubulin, to carry out some of
its jobs in the kidney.
One thing that might help clarify the receptors’
functions would be to compare their
behavior in the mouse and other animals.
Thomas turned to Salim Seyfried, a junior
group leader at the MDC who has introduced
the zebrafish to the lab. Over the past 15 years,
this tiny animal has become one of the most
important laboratory organisms for the study
of developmental processes. It is almost com-

pletely transparent, which allows scientists to
watch organs form over time without having
to dissect the animal. It is so small that its
cells can feed themselves for a long time by
absorbing nutrients directly from the environment.
This means that the fish can survive
even if they have defects in the heart or circulatory
system. It is an ideal subject for Salim’s
lab, which focuses on heart development, and
needs to be able to study animals whose
organs grow in a defective way.
Uwe Anzenberger and colleagues from
Salim’s and Thomas’ lab carried out a study to
see how easy it would be to compare LRPs in
fish and mice. They found that zebrafish have
close relatives of megalin and cubulin, that
the molecules are produced together in the
fish’s kidney, and that they play an equally
important role in drawing hormones into
cells.
“Many of the molecules that are captured by
LRPs are delivered to internal compartments
in cells called endosomes,” Thomas says. “Uwe
found that without megalin, substances were
not drawn into cells in the fish – and there
weren’t any endosomes. So the system closely
parallels what happens in mammals and
will allow us to get a closer look at how the
kidney works.”
These varied studies have changed Thomas’
mind about the textbook view that hormones
simply slip into cells. That is surely true at
times, because some of the tissues that need
them do not produce LRPs. But where the
receptors are present, the lab kept discovering
exceptions. Studies in cell cultures revealed
that megalin was also needed to take up sex
hormones. That didn’t prove it also happened
in animals, but it was likely to. Megalin is also
found at high levels in the target cells for sex
hormones.
In 2005 this led postdoc Annette Hammes,
collaborating with scientists from the
University of Aarhus in Denmark, to take a
new look at the megalin knockout mice. She
discovered abnormalities in the development
of the reproductive organs of both males and
females, as if cells were no longer being stimulated
by sex hormones. That didn’t fit with
the hypothesis that hormones could simply
slip through membranes; it only made sense
if megalin was detecting them and escorting
them inside.
“The vast majority of cholesterol-derived hormones
and lipids are bound to proteins or
packaged in lipoproteins as they move
through the bloodstream,” Thomas says. “In
those cases, we think that their entry into the
cell is managed by receptors, usually the LRPs.
That hypothesis is giving us a handle on many
diseases in which what is broken – and what
needs to be fixed – is an endocytic receptor.
So, for example, we’ve found that signals from
another LRP called the ApoE receptor 2 are
needed to prompt cells to develop into
mature sperm. That signal never arrives when
this LRP is defective. Infertility is the consequence.
Seven percent of all human males
have fertility problems. LRPs may well play an
important part in this condition. Under -
standing why the signal goes missing may
one day help to restore it.”
��
The flat topography of Berlin and its spacious
streets make the city ideal for bikers,
if you’re willing to risk getting caught in
the tempests that sometimes arise without
warning in the fall and winter. One of them
flashed over the campus in January 2007 with
such ferocity that it stripped the bright red
façade off the Max Delbrück Commu nications
Center, or MDCC. The timing couldn’t have
been worse – in less than 24 hours the campus
would hold its New Year’s party – in the
MDCC. At two a.m. residents of the campus
guest houses had heard what sounded like an
explosion. When security personnel came to
investigate, they were greeted by a sight so
alarming that they immediately called
Gudrun Erzgräber and told her that the entire
roof had collapsed. She called Walter
Birchmeier.

Ulrike Förster
By three a.m. Walter was surveying the damage
with Stefan Schwartze, the MDC’s administrative
director. Things were bad enough, but
not as bad as they looked. The roof hadn’t fallen
in. What appeared to be stone rubble littering
the ground – and the bust of Max
Delbrück, at the front door – was actually a
layer of light, foam-like blocks that had
covered the building. Workers swept the
debris out of the way and the New Year’s celebration
went on as planned, with a talk by
Nobel laureate Aaron Ciechanover and a
per formance by jazz pianist Rich Bairach to
take our minds off the roof. The MDCC spent
most of the year covered in plastic and scaffolding.
Thomas is one of those who braves the Berlin
weather; his family doesn’t even own a car.
Every day he rides his bike from home to the
S-Bahn, catches a ride to Buch, then gets back
on the bike for the last bit to campus. We met
once at the S-Bahn station in Pankow, where
there is a shop that sells Swiss army knives.
The front window displays a larger-than-life
knife, even larger than the Giant, with rotating
blades.
“I’ve been using this metaphor for quite a few
years, and the more we learn the better it fits,”
he says. “LRPs can do so many things because
of their complexity – single molecules contain
several tools – and their ability to link up with
other tools. Besides proteins bound to lipids,
these molecules recognize viruses, antibiotics,
toxins, signaling molecules, and a range of
other partners. The result may be to draw the
partner into the cell, or to allow an influx of
calcium, or trigger cell migration, or change
the behavior of synapses... I could go on, but
you get the point.”
Thomas is willing to follow the LRPs wherever
they lead him. “Sometimes it is helpful not to
be preoccupied with a certain theory but to
watch what nature tells you,” he says. It’s one
of the advantages of being at the MDC. “If you
knock out a molecule and it leads to a problem
in the brain, we can find partners in the
neurobiology groups to help us follow up. Our
own expertise lies in the cardiovascular
system, but these systems are controlled
by fundamental processes which occur
throughout the body. So the processes and
the molecules involved in them have to
be looked at in a more holistic way. You have
Part Three: Frontiers and ferrymen
168
to look at many facets of a molecule to
truly understand it. You have to explore how
it has changed through evolution by observing
its functions in different model organisms.
“Everyone is talking about ‘systems biology’
these days. Well, hormone signaling is a good
example of a ‘systems’ problem. Hormones
are made and released by individual cells, and
taken in by individual cells, but in a way that
has to be coordinated all over the body. It’s
what raises blood pressure when there has
been an injury and blood has been lost, for
example. Hormone signals change the behavior
of blood vessels, the heart, and a number
of other bodily systems. So it’s the type of
problem that you have to try to approach in a
global way.”
Recently the lab has begun studying a second
family of proteins, the sortilins. Although the
members of this family evolved independently
from LRPs, they have a similar structure.
“Many regions of the brain don’t produce
LRPs, and they don’t have access to the bloodstream
where the fats circulate, but they need
cholesterol and other lipids,” he says. “So

there’s been an idea that sortilins might act
as lipid receptors as well.”
You never know what will happen when you
open a new biological box. The lab’s studies of
a sortilin called sorLA brought them faceto-face
with an entirely new problem: the
development of Alzheimer’s disease. If there’s
a biological equivalent of the roof falling in,
that’s it.
��
In 1901 a 51-year-old woman named
Auguste Deter was admitted to the State
Hospital for Epileptics and the Mentally Ill in
Frankfurt. At home her behavior had become
progressively more bizarre. She had paranoid
fantasies in which she accused her husband
of having affairs; she was losing her ability to
read, write, and reason. Finally the family was
unable to cope and admitted her to the
hospital.
She was examined by a 37-year-old neurologist
named Alois Alzheimer. He conducted a
medical and behavioral examination. His case
notes, which were found in the archives of the
University of Frankfurt in 1995, report:
She sits on the bed with a helpless expression.
What is your name? Auguste. What is
your husband’s name? Auguste. Your husband?
Ah, my husband. She looks as if she
didn’t understand the question. Are you
married? To Auguste. Mrs D? Yes, yes,
Auguste D. How long have you been here?
She seems to be trying to remember. Three
weeks. What is this? I show her a pencil. A
pen. A purse, key, diary and cigar are
identified correctly. At lunch she eats cauliflower
and pork. Asked what she is eating
she answers spinach. While eating meat she
answers potatoes and horseradish. After a
short time she forgets objects that have
been shown to her. She constantly speaks of
twins. When she is asked to write, she
holds the book in such a way that one has
the impression that she has a loss in the
right visual field. Asked to write Auguste
D., she tries to write Mrs and forgets the
rest. It is necessary to repeat every word...
By the time Deter died in Frankfurt in 1906,
Alzheimer had moved on to Munich, to a posi-
169 Part Three: Frontiers and ferrymen
tion at the Royal Psychiatric Clinic. He had not
forgotten the case, and when he learned of
her death, he wrote to the Frankfurt hospital
to ask permission to examine her brain.
Alzheimer had become one of Germany’s foremost
experts in the pathology of the nervous
system, and the hospital director granted the
request.
Alzheimer discovered bundles of “a peculiar
material” in the cortex of the brain. These
dense clusters of fats and proteins are called
amyloid plaques. In Alzheimer’s patients they
collect between neurons, blocking their communication
with each other and causing the
cells to die. Over time, this leads to the symptoms
of Alzheimer’s disease: a loss of cognitive
functions and a person’s control of his or
her body, and finally death.
In the 1980s, Konrad Beyreuther’s lab at the
University of Heidelberg identified the main
component of the plaques: a small peptide
called β-amyloid. This turned out to be a fragment
of a much larger molecule called APP, for
amyloid precursor protein. APP is a membrane
protein found on the surface of everyone’s
neurons. Although its functions in the healthy

ain are not well understood, it seems to play
an important role in memory and learning.
For most of people’s lives they produce β-amyloid
fragments without causing much harm.
However, by old age so much of it has accumulated
in the brain that it forms the conspicuous
deposits – first encountered by Alois
Alzheimer in his autopsy studies. Similar to the
build-up of “debris” in the vessels in cardiovascular
disease, the built of these plaques in the
brain harms the nerve cells, causing mental
problems typical of Alzheimer’s disease. If scientists
could find out why this happens, they
might be able to interrupt the process and
block the development of the disease.
Could it be a coincidence that diseases of
blood vessels and diseases of the brain share
so much similarity? Thomas suspects a connection,
and the connection may be receptors
for lipid-loaded lipoproteins. That’s why his
group was investigating sorLA, one member
of the sortilin family found in neurons, when
he read a study by James Lah’s laboratory at
Emory University in Atlanta, USA. Lah and his
colleagues had conducted a survey of the
molecules produced in the brains of
Alzheimer’s patients, looking for molecules
that behaved differently than in healthy peo-
ple. They found that people with the disease
had abnormally low levels of sorLA.
“This happened at a time when a flurry of discoveries
had been made connecting
Alzheimer’s disease and cholesterol,” Thomas
says. “A study by Elizabeth Corder from Duke
University in the US showed that sequence
variations in the protein ApoE were the highest
single predictor of a person’s developing
Alzheimer’s disease. I’ve already told you
about our work on ApoE – it’s a lipid-carrying
protein that is recognized by LRPs – and by
sortilins as well. So finding that some forms of
ApoE are highly likely to lead to Alzheimer’s is
one link between LRPs and the disease.”
An important piece of evidence for the
hypothesis came from the discovery of a
direct link between the receptor LRP1 and APP.
Brad Hyman’s lab in Harvard and others discovered
that a third protein was able to grab
both molecules, gluing them together.
“When this happens on the cell surface, the
two molecules are drawn into the cell and
delivered to endosomes,” Thomas says. “These
compartments contain the molecule that
slices up APP into the β-amyloid fragment. If
APP isn’t delivered, it won’t be cut that way.”
Part Three: Frontiers and ferrymen
170
What prevents LRP1 and APP from coming
together in every cell? I ask him.
“Part of it seems to have to do with another
LRP receptor called LRP1B,” Thomas says.
“There are only so many copies of APP on the
surface of the cell, and other molecules compete
with each other to bind to them. LRP1B is
drawn into the cell at a much slower rate than
LRP1. So if it binds to APP, it holds it on the cell
surface, where it eventually gets processed in
other ways, releasing ‘healthy’ fragments.”
SorLA comes in, he says, during the process of
delivery. “SorLA has the classic features of an
LRP. The main difference between it and other
LRPs is an extra module. This unit acts as a
sort of global positioning system that directs
the molecule toward specific addresses within
the cell. It’s a bit unusual for an LRP because
it operates mainly inside the cell, rather than
on its surface.”
Olav Andersen, a postdoc in Thomas’ lab from
Denmark, worked on the connection between
sorLA and Alzheimer’s disease for about five
years. “What we wanted to understand was
the ‘mechanical’ connection between the
activity of the two molecules, sorLA and APP,”
Thomas says. “We needed to prove that they
actually met in the cell and could influence
each other’s activity.”
Olav came to the lab along the same route as
several other talented young scientists;
Thomas recruited him from the University of
Aarhus, where he has long-standing collaborations.
“It’s a good partnership because
Anders Nykjaer’s group there is strongly
focused on the biochemistry and cell biology
of these proteins,” he says. “We concentrate
more on the physiology side. So if their postdocs
start getting interested in looking at
organs and animals, I can attract them here.”
Thomas says that he had to get adjusted to
Olav’s work habits. “He left at three or so in
the afternoon,” Thomas says. “I’d go into the
lab and the only thing there would be his
Birkenstock sandals and an empty bench.

Scientists have flexible work hours, but I
thought leaving at three was a bit extreme.
Then I found out he was coming into the lab
at five in the morning. He’s always done that.
It’s just the way he works.”
Olav’s first experiments proved that the modules
of sorLA and APP lying outside the cell
could bind to each other. Next he tagged the
molecules with fluorescent markers and
tracked their movement through the cell
under the microscope. The proteins seemed to
be moving together to certain internal compartments.
An additional experiment gave
cells a mutant form of sorLA which sat in the
cell membrane but could stay in the cell. This
left APP stranded on the surface as well.
The opposite approach – giving healthy cells
much higher amounts of sorLA than normal –
had an equally interesting outcome. “APP is
kept in the cell along with the receptor and
gets trapped in large compartments,” Thomas
says. “But this time it is a different compartment
than the endosome. It doesn’t get
chopped into the harmful β-amyloid fragment.
So sorLA has an interesting job. It delivers
APP to places in the cell where it is locked
up and remains harmless. It is no longer delivered
to the cell membrane to meet the culprit
LRP1.
“That’s what happens in ‘normal’ neurons, at
least as they get old, and it happens more
often when there is not enough sorLA. This
made us think that the cause of the accumulation
of β-amyloid fragments might be a
drop in the production of sorLA in patients.”
The scientists put the hypothesis to the test
by examining tissue samples from ten
Alzheimer’s patients. All of the patients – and
none of the healthy subjects – showed low
levels of sorLA in the brain.
The findings led Thomas’ lab to develop a
strain of mouse without sorLA. If the hypothesis
was correct, the mice’s APP ought to
behave like that of Alzheimer’s patients. By
ten months of age, the animals’ brains experi-
171

Daniel Militz
Part Three: Frontiers and ferrymen
enced a 30 percent increase in levels of the dangerous beta-amyloid
fragment, and they developed those notorious plaques.
The way APP is shuttled through the cell directly influences whether it
is processed into the harmful fragments that have been linked to
Alzheimer’s disease. Several molecules are involved in this process, and
pharmaceutical companies are attempting to find drugs that can alter
their activity. “SorLA looks like it would make a good target,” Thomas
says. “A drug that increases its activity could have the same effects in
patients we have observed in the lab: it might lock up APP and prevent
the formation of amyloid fragments and plaques.”
��
Another Danish alumnus of the Willnow group, Anders Nykjaer, is a
professor at the University of Aarhus. He spent almost two years
on a sabbatical in Berlin; while there he introduced the lab to sortilins.
Like many former Danish colleagues who spend time in Berlin, he continues
to collaborate with Thomas on themes started at the MDC. In
2006 a collaboration between the two labs and Weill Cornell Medical
College in New York City revealed new functions for sortilin, a protein
closely related to the sorLA receptor. They found that sortilin binds to
another membrane receptor called p75NTR . When that happens, the two
proteins build a platform that can dock onto a signaling molecule
called proNGF. The combination of the three molecules triggers a selfdestruct
program in the cell. This signal likely participates in the widespread
death of cells that follows strokes or other types of brain injury.
Blocking this signal prevents the cells from dying, as the scientists have
shown by knocking out sortilin in mice.
Megalin, sorLA, and sortilin all look like promising leads in the search
for drugs to curb diseases. A few years ago Thomas and Anders founded
a company called ReceptIcon to bring the studies to the point that
pharmaceutical companies might step in.
“Megalin is particularly interesting because it plays a role in the toxic
side-effects that arise from antibiotics and a huge range of other therapeutic
drugs,” Thomas says. “As the receptor filters vitamins passing
through the proximule tubules of the kidney, it sometimes also draws
in substances it should not – such as therapeutic drugs. It does so
because the receptor accidentally confuses them with its true binding
partners. Thus, instead of being excreted, these drugs build up in the
kidney and cause organ failure and death.”
One of ReceptIcon’s projects has been to help find substances that will
block megalin’s activity – at least temporarily – until drugs have passed
through the kidney. One such inhibitor has now passed animal trials
and is scheduled for clinical trials in humans next year.
It’s gratifying, Thomas says, to make a discovery with clear potential
applications and health benefits. “This is a real example of molecular
172

Anne-Sophie Carlo Chandresh Gayera
medicine,” he says. “It happened because we
found a molecule which plays a direct role in
processes of disease and organ damage. But
in many cases, for example the sortilin story, a
solution will probably have to be more subtle.
We have a good track record of uncovering
new and surprising functions of previously
unknown receptors in processes like the
uptake of binding partners or organ development,
but all these pieces have to be understood
and put into place before we understand
certain organs as a whole and how they
are connected to other body systems.”
The “genome age” has brought a mass of data
to the table, he says, but it is not well integrated.
“Results of experiments are most useful
when we can see their place in the whole –
from simple things like where particular molecules
are expressed in the kidney to more
complex questions like what signaling pathways
they participate in, and ultimately how
that perturbs the physiology of the kidney,” he
says. “But there haven’t been tools to ‘navigate’
the data that way. This prompted us to
propose an EU project that would capture
data from groups throughout Europe and
elsewhere to be plugged into a ‘Kidney Atlas.’
In a way, it is a virtual organ that we’re building
on the web now.”
The atlas is compiling information from laboratories
working with a variety of model
organisms including mice, fish, frogs, and rats.
Direct participants include 18 research groups
from universities and other institutes, plus six
university clinics, spread across nine European
countries. The goal is to create a four-dimensional
map of the kidney, showing how molecules
guide its development and activities
from its origins in the early embryo to adulthood.
During a disease all of these levels are
affected, Thomas says, sometimes over an
organism’s entire lifespan. Solving the health
problems related to kidneys will require scientists
to develop an integrated view of the kidney
as a whole and its links to the rest of the
body.
173
You’d need a fat user’s manual to keep track of
all the parts and possible functions of the
Giant Swiss army knife. “And that’s a knife
with only 85 tools,” Thomas says. “Imagine the
manual you would need to describe the thousands
of molecules that build and operate the
kidney, or other organs for that matter. At the
moment that information exists in bits and
pieces, not assembled into a coherent picture.
That’s the main task of the atlas. It’s a big job,
but every bit of knowledge added to this virtual
kidney will pay off in understanding the
real one.”

Bernd Reif

Bad origami
Even with step-by-step instructions, some of us would never be
able to fold a sheet of paper into an elegant origami flower or
crane. Let alone the model of an ankylosaurus, a dinosaur, which was
designed by Ronald Koh of Singapore and can be found on the Internet.
It has 61 steps, and that’s not really counting right, because most of
some of them tell you things like, “Repeat the fold you just made 20
times.” I gave up on the third step.
But suppose you had to learn out how to make the origami ankylosaurus
by unfolding a finished one. That would give you an idea of
how Bernd Reif feels when he looks at β-amyloid plaques, the culprits
in Alzheimer’s disease, introduced in the previous story. In trying to figure
out how they form, the situation is like that of the crane: even
knowing a lot about the starting material (paper) and seeing the finished
product (a paper flower) doesn’t explain how to get from one to
the other. Some of us have wastebaskets full of failed origami projects
to prove it.
“Amyloid plaques accumulate between nerve cells, leading to cell death
and the symptoms of Alzheimer’s,” Bernd says. “But we understand very
175 Part Three: Frontiers and ferrymen

The Reif group at the NMR spectrometer.
Mangesh Joshi, Rasmus Linser, Bernd Rief
and Veniamin Chevelkov

Part Three: Frontiers and ferrymen
178
Mangesh Joshi
little about what causes those effects. We think
that protein fragments assemble into amyloids in
an orderly way, going through many steps in
which they take on different structures. Some of
those structures are more toxic than others,
although they are made of the same subunits.
Finding out what makes the plaques so dangerous
will probably require getting a step-by-step look at
how the protein fragments assemble.”
Bernd’s laboratory at the FMP is trying to uncover
some of the basic principles underlying
Alzheimer’s, Huntington’s Diseases, prion diseases,
and other neurodegenerative conditions
that are caused by accumulations of proteins.
They all begin when a protein needed by the
healthy body begins to behave in an unusual way.
In the case of Alzheimer’s disease, brain cells produce
protein fragments all the time, but normally
the body can get rid of them.
“The same small bits of protein – made of just 40
amino acids – are sometimes completely harmless,
sometimes fatal,” Bernd says. “If you watch
them bind to each other in the test tube, you get
a variety of forms: single fragments that don’t
accumulate, or stringy fibers, or clumps that are
not very organized. What you get depends on the
experimental conditions.”
The plaques may assemble differently in the test
tube than in the brain, and even under controlled
conditions it has been difficult to get a glimpse of
the stages. The battle, he says, has been to find or
invent methods to get a look at single stages of
how protein fragments interact with each other
and other molecules found in the clumps. Call it a
folding manual.
��
If β-amyloid isn’t yet a household word, it probably
will be soon. More and more families are
confronting Alzheimer’s disease, the most common
age-related neurodegenerative disease in
the world.
The previous story describes how β-amyloid
begins as part of a much larger molecule called
APP, found in high numbers in the membranes of
nerve cells. What it normally does there is still a

mystery. Mice born without the molecule develop smaller brains, problems
of body coordination, and impaired memory and learning, but
scientists do not know how APP contributes to these processes in
healthy animals or humans.
Like most proteins, APP is processed by other molecules in several
ways. First it is cut by an enzyme called β-secretase, or BACE. Another
protein called γ-secretase cleaves the chain for a second time, releasing
the β-amyloid fragment. This is a normal, life-long process that
undoubtedly has a role in the healthy brain, but the fragment’s function
is not yet known. Normally it is broken down and disposed of, but
with age our cells and molecules behave differently. Later in life the
fragments accumulate to form fibers and plaques in the space
between cells.
“There are a few theories about why the aggregations are dangerous,”
Bernd says. “One is that they create holes in cell membranes that allow
charged calcium atoms to pass through. That would be bad because
179
calcium ions have to be carefully regulated for the cell to behave properly
and survive. Another idea is that plaques stabilize or produce free
radicals, highly reactive atoms that are dangerous because they can
modify a wide range of molecules. Whatever the answer is, β-amyloid
only does this in a specific type of cluster, or what we call an oligomer.
That’s simply a complex in which more than one copy of the same
molecule is bound together. For example, if copies assemble in
oligomers of six or eight they might not be dangerous, whereas
groups of ten could be fatal. Determining which of these forms is the
deadly one has been extremely challenging.”
And amyloid plaques contain more than β-amyloid. (“The pleasures
and powers of green tea,” at the end of the book, describes an MDC
group’s efforts to find other molecules that dock onto the fragment
and influence its folding.) Recently another laboratory discovered that
a chaperone protein called α-B-crystalline could bind to the fragments
in the test tube as well as in living cells.
Part Three: Frontiers and ferrymen

Chaperones are “assistants” whose normal job is to help keep proteins
folded in their proper forms – if they become unfolded, the cell usually
destroys them. Folding gives a protein the shape it needs to interact
with other molecules and snap into “molecular machines.” A protein
begins as a long string; folding happens because of chemical interactions
between the amino acids that make it up. Sometimes it needs
help to achieve or maintain the right form, which is where chaperones
come in. Experiments showed that the brains of some Alzheimer
patients contained chaperones, including α-B-crystalline, in abnormal
places.
“That’s interesting because α-B-crystalline is disturbed in people who
suffer from Down Syndrome,” Bernd says, “and nearly all of them
develop Alzheimer’s disease. They develop it early, at around 40 years
of age, which is a fairly dramatic acceleration of the disease process.”
Experiments in the test tube revealed that α-B-crystalline blocks the
formation of amyloid fibers. Fragments can still bind to each other in
the molecule’s presence, but they don’t take on the long, orderly shape
of fibers. Instead, copies of β-amyloid cluster together, in groups of
twos (dimers), threes (trimers), or higher combinations. Some remain
alone.
“It is unclear why only one oligomeric structure should be toxic,” Bernd
says. “To find out, we’ll have to watch the process of assembly. And that
has been a huge technical challenge. What is very interesting is that
the combination of β-amyloid with α-B-crystalline is more deadly to
Part Three: Frontiers and ferrymen
β-amyloid fibers like those found in the
brains of Alzheimer’s patients (upper left).
The presence of certain proteins or other
substances can change the types of larger
structures they form (shown in the other
images). Most of these aggregations would
probably not have the deadly effects seen in
the disease.
cells than β-amyloid on its own. Maybe if fibers could form, the situation
wouldn’t be so dangerous; α-B-crystalline is keeping that from
happening.”
Were fibers a transition state – were they involved at all in the development
of disease symptoms? Wondering why α-B-crystalline
increased the deadliness of β-amyloid, Bernd’s lab applied several
innovative techniques to study the nature of the contact between the
two molecules and its relationship to the formation of fibers and
plaques. The first question to answer had to do with basic architecture:
what regions of β-amyloids were linked to each other, and which
regions could bind to α-B-crystalline? Next the scientists wanted to
know how many copies of β-amyloid could be found linked within a
single cluster, and whether this changed because of the presence of α-
B-crystalline. If changes occurred, they might happen in several steps.
Sometimes in origami, a fold is made and then undone, so that the
crease can serve another function.
��
Saravanakumar Narayanan, a PhD student from India, used a combination
of methods to find some of the answers. First, β-amyloid
and α-B-crystalline were mixed together at changing concentrations.
Presumably a small number of copies of α-B-crystalline would have
different effects on the fragment than a high number. As the amounts
changed, the structures were examined by NMR, which provided an
atom-by-atom view of some regions of the molecules.
180

The experiments showed Saravanakumar and his colleagues the docking
sites of the two molecules. A tiny region of β-amyloid permits two
or more copies to bind to each other. The same region serves as the
binding site for α-B-crystalline. In fact, α-B-crystalline binds much
more strongly to β-amyloids than the fragments bind to each other.
“That is extremely interesting,” Bernd says, “because it suggests a way
by which α-B-crystalline can change the groupings and the function
of β-amyloid fragments. Without α-B-crystalline, there is nothing to
keep the fragment from forming large groups, including structures
like fibers. When the molecule is present, however, it competes for the
binding site. Fewer copies of β-amyloid will bind.”
The findings closely echo what the group found while investigating
another protein, Sup35, using similar methods. Sup35 is a yeast prion
protein, similar to the human protein that causes prion diseases (like
Mad Cow Disease). Both yeast and human prion proteins have functions
in healthy cells that they can no longer perform when they have
accumulated in clumps. Saravanakumar and Bernd discovered that a
chaperone protein called Hsp104, known to interact with Sup35, links
to clusters of Sup35 molecules and tosses out some of them, leading
to smaller clusters.
α-B-crystalline might be doing something similar, but the scientists
needed to know if it was changing the number of β-amyloid mole-
cules attached to each other. While investigating the yeast prions, the
scientists had developed a method to filter different sizes of molecules
by placing them in a dialysis bag and allowing them to leak out. Such
bags are used to filter the blood of people with kidney diseases; they
are full of holes so tiny that they trap large molecules while allowing
small ones to pass through. By using bags with holes of different sizes,
the scientists could test amounts of Sup35 molecules that were
detached from Sup35 fibrils. They could also track Hsp104’s effects on
the size of these groups.
Similar experiments revealed that α-B-crystalline binds weakly and
only for a very short time to single copies of β-amyloid. But it binds
very strongly to multiple copies of the fragment that have formed
fibers or are in the process of doing so. “When that happens, it causes
a structural change,” Bernd says. “We believe that α-B-crystalline shifts
the preference of the fragments – instead of linking up into long,
stringy fibers containing thousands of copies of the fragment, they
collect in smaller clusters.”
The association between the molecules has an additional effect: one
of the amino acid subunits of β-amyloid becomes oxidized. This means
that the molecule picks up additional electrons, changing its overall
charge and the way it interacts with other molecules. That, Bernd says,
might help explain why amyloid plaques are so deadly.
181 Part Three: Frontiers and ferrymen

“Oxidation seems to destabilize fibers so that they can dissolve,” he
says. “That might also explain why this form of β-amyloid is toxic to
neurons, because it produces highly reactive forms of oxygen.”
To test the idea, the scientists used a probe that could measure
α-B-crystalline’s oxidizing activity. They put α-B-crystalline into a mixture
of single molecules called GSH. On its own, the oxygen that is dissolved
in the solution provides an electron that oxidizes GSH. Adding
another protein may slow down the process, because it also undergoes
oxidation and captures some of the electrons that would otherwise
be used by GSH. That’s what happened when Bernd and his colleagues
mixed the two types of molecules. Mixing GSH with β-amyloid
fragments, however, had almost no effect. This meant that the
fragments alone weren’t producing the oxidizing reactions.
To “turn up the heat,” Bernd and his colleagues added copper to the
mixture of GSH and α-B-crystalline. “Copper doesn’t change the oxidation
state of GSH on its own,” Bernd says, “but in combination with α-
B-crystalline, it triggers much faster reactions. Our interpretation is
that copper binds tightly to α-B-crystalline and pushes its oxidizing
activity into higher gear.”
��
Bernd is just as helpless as I am when it comes to origami; his artistic
talents lie elsewhere. He has always loved singing, but things
started to get serious when he began his undergraduate degree in
physics in Beyreuth, Germany. (If you aren’t familiar with that town,
you’ve never heard of Richard Wagner, who established an opera
dynasty there.)
Bernd won a Richard-Wagner scholarship and took singing lessons
from Manfred Hegen, a school teacher and singing coach. Some of
Hegen’s students have made it into the choirs of the Beyreuth music
festivals or gone on to have impressive careers. Hegen founded Musica
Vocalis, a concert choir. First Bernd joined the choir, then graduated to
singing arias. He has performed in excerpts from operas and Bach’s
Christmas Oratorio, and recently Creation, by Haydn. There is a rumor
that he sometimes performs at scientific conferences, so we’ll have to
Part Three: Frontiers and ferrymen
follow up on that. And as if there weren’t already enough music in his
life ... he is married to a violinist in the Berlin Opera orchestra.
Bernd is an example of the possibility of combining a scientific career
with a love for music, but it isn’t always easy, and there have been
some dry spells. When he spent three years at the Massachussetts
Institute of Technology in the USA on a postdoctoral fellowship, it was
hard to find a choir.
Back in Germany that would be less of a problem, but at the beginning
he had his hands full setting up his own group in Munich, at the
Institute for Organic Chemistry and Biochemistry II of the Technical
University. He began thinking about ways to combine methods to get
a look at the way proteins cluster. Most questions about protein structure
were being solved by crystallography, a process that teases highly-purified
proteins into a lattice-like array, like stacking shoeboxes
onto shelves. If the molecules formed neat enough rows, they could be
examined by bombarding them with X-rays. But not all proteins
behave this way. Some, particularly membrane proteins like APP, don’t
form tidy crystals. Besides, the process locked molecules into stiff
arrangements – not a very good way to watch a delicate process that
might involve many proteins. While in the USA, Bernd had been introduced
to another method, NMR, which might provide answers.
��
We still have a long way to go in understanding the molecular
processes that lead to Alzheimer’s disease, Bernd says. “We’re
asking questions that tax all of our current methods. Finding answers
has required combining techniques in new ways.”
For example, the laboratory would like to get a look at how molecules
bind to β-amyloid in the fiber form. NMR, the technique of choice at
the FMP for getting structural information, works best with molecules
that float freely in solution. It applies a strong magnetic field to protein
samples. This alters the energy levels in the nucleus of every atom
in the sample in a way which is slightly different for each proton in a
protein. The slight energy difference allows scientists to determine the
182

identity of each proton – in other words, to tell what type of atom it
belongs to. When the strength of the magnetic field is lowered, or
“relaxed”, they can probe the distances between protons, permitting
the creation of atom-by-atom maps and thus a structure of the protein.
But for the method to work, molecules need to swim around
freely, which they can’t do if they are locked into a fiber.
In addition, Bernd says, one can try to get a lot of molecules in the
sample to align themselves in the same direction. He knew that some
of his colleagues had been using magnetic fields to force fibers into a
more organized structure, allowing them to obtain higher-resolution
images using X-rays. Maybe the same thing could be done using the
powerful magnetic field of the NMR machine.
“The amino acid subunits of proteins have different magnetic properties,”
Bernd says. “Think of a fibril as a shoestring that several small
magnetic beads have been strung onto – representing the most magnetic
subunits of the protein. If you lay the string alongside an iron
pipe, it will attach itself and hold on. This is an analogy of how the protein
fiber can be oriented in a certain way by the magnetic field. Now
if other proteins come along and dock on, they can only attach themselves
in certain ways. Those constraints give us enough information
to see some of the details.”
It often takes months or years to get such ideas to work, Bernd says,
and in many institutes scientists don’t have the freedom. “Methods
development is dry compared to some of the science stories,” he says,
“and it is sometimes difficult to publish interdisciplinary papers where
the focus is on methods. But the β-amyloid studies are a good example
of how techniques have to be adapted or used in creative new
ways to solve the kinds of questions we are asking right now. Walter
Rosenthal, the FMP’s director, has been very supportive of this
approach.”
��
With method developments, a wide range of scientific interests
and his love of music, Bernd clearly has his hands full, but he
recently took on another major project with the creation of a new
graduate school in Molecular Biophysics at the FMP. The program has
now been launched in collaboration with universities and other institutions.
The Leibniz Association has financed the creation of several of
graduate schools, on a competitive basis, over the past few years.
Walter Rosenthal says that Bernd’s proposal to create a department
devoted to biophysics – specifically the investigation of protein interactions
– was successful for several reasons.
“In the first place the topic fills a gap in the landscape of German academic
programs,” Walter says. “And it was a very strong proposal.
Bernd developed the concept, convinced all of us and the partner institutes,
and wrote the grant application.”
Currently there are 15 students enrolled in the program, about half
doing their work in Berlin and the rest at other institutions. Each
receives a healthy, competitive stipend – to which some expectations
are attached. The students present their work once a year in an annual
report, and participate in two major “block” courses whose purpose
is to give them a broader background than students normally receive.
“Biophysical techniques are acquiring a growing role in the investigation
of biological processes,” Bernd says. “We’re also in the midst of a
very exciting time in which new methods are being invented all the
time. Young scientists will need a very wide exposure to the palette of
methods that are available and should think creatively about new
ones.”
The proposal captures the best experts in particular areas from the
region, Bernd says. It will draw on campus expertise in X-ray crystallography,
NMR, and signal transduction; institutes within the Charité will
contribute in areas such as electron microscopy and biochemistry; the
Institute for Experimental Physics and other departments of the Freie
Universität will offer spectroscopy and other types of expertise, and
the Humboldt-Universität adds expertise in membrane biophysics
and several other fields. The University of Potsdam is another partner
in the areas of CD spectroscopy and thermodynamics.
“In putting together this program, we’ve collected a wide range of
techniques practiced by groups that have a lot of experience working
together,” Bernd says. “Scientific themes will include the biophysics of
protein aggregations, such as our work on Alzheimer’s disease; signal
transduction in processes like visual perception; the molecular
machines that build proteins, and viral infections. The questions in all
of these fields are extensive and require a combination of approaches.
Our plan is that each student in the program will be mentored by two
group leaders, to increase interdisciplinarity.”
What about a PhD program choir? That would certainly be interdisciplinary.
“Maybe someday,” he laughs. “We’d need a pianist. And we might have
to change the criteria by which we select our students.”
183 Part Three: Frontiers and ferrymen

Trouble in the
waterworks
It’s hard to imagine an advanced society without electricity. Hard to
imagine that progress and luxuries could be achieved without this
magical river that stems from the turbines of coal plants, dams and
windmills, flows into conduits underneath our streets and the engines
of factories, into secret spaces inside the walls of our houses. Yet without
electricity, the ancient Romans achieved a standard of living – at
least their rich did – that seems sumptuous even by today’s standards.
They managed it by building a society based on water.
Visitors to Pompeii and Herculaneum can view a film in which, thanks
to computer animation, the towns are restored to their full glory in the
days before Vesuvius erupted and covered them in meters of ash and
lava. Everywhere in this world is water: springing up through fountains,
pouring into public baths, irrigating colorful gardens. Without it, Roman
society would have crumbled.
Robert Harris’ novel Pompeii takes us underground in this ancient world.
In one scene a Roman engineer climbs the slopes of Mount Vesuvius
and enters a vast subterranean cavern, trying to discover why water has
stopped flowing. The problem stems from an earthquake that has
Part Three: Frontiers and ferrymen
184

Enno Klußmann

Walter Rosenthal and Enno Klußmann
shifted the ground and shoved aqueducts out
of alignment. As the water slows to a trickle in
the towns around the volcano, people begin
to panic. Life is utterly dependent on water,
and its loss seems much more threatening
than rumblings on the mountain.
��
Our own existence is completely dependent
on water, mostly on recycling it.
While a human body is 70 percent water, only
a fraction of that comes from our daily intake
of liquids and food. The rest – 99 percent – is
won through recycling. The kidneys process
about 180 liters per day. Recovering nearly all
of that water is the job of cells in the body’s
waterworks – the kidney – and they only manage
because they are leakier than other cells.
If it weren’t for proteins called aquaporins, we
would quickly die of dehydration even without
any disruptions in the water supply. That
partly explains the interest of two researchers
at the FMP: Enno Klußmann and Director
Walter Rosenthal.
In the 1920s researchers figured out that the
membranes of cells were made of fat molecules
called lipids, put together in such a way
that small quantities of water could simply
slip through. But that amount was clearly not
enough to supply the body’s big drinkers, like
kidney and red blood cells, so there had to be
another way for water to enter. Fifty
years later, this “water channel” still hadn’t
been found. Then along came Peter Agre, a
young hematologist who was working at
the John Hopkins School of Medicine in
Maryland.
Part Three: Frontiers and ferrymen
186
“The field was essentially stuck,” Agre says,
“but following the well known scientific
approach known as ‘sheer blind luck,’ we
stumbled upon the protein that is the answer
to the question: do water channels exist?”
Agre and his colleagues had found the first
aquaporin, a discovery so important that he
was awarded the Nobel Prize in 2003.
He was also the FMP’s first choice as a speaker
when the institute was thinking about who
should give the main talk at the campus New
Year’s party; in 2008 it was the FMP’s turn to
host the event. In his introduction, Walter
Rosenthal commented on the man as well as
his work, mentioning the fact that Agre was
an Eagle scout and a cross-country skier. He
also dwelt for a moment on Agre’s political
activities:

“I have also been deeply impressed by your
political activities, which are inspired by your
strong belief in humanity. You jumped to the
defense of Thomas Butler, who voluntarily
reported to the university safety office that 60
vials of plague bacteria were missing and had
probably been autoclaved. You defended his
refusal to plead guilty, which ultimately led to
him receiving a two-year prison sentence. You’re
also a founding member of a movement called
Scientists and Engineers for Change and the
slide shows one paragraph of the Bill Of Rights
of this initiative. It reads: ‘The federal government
shall not support any science education
program that includes instruction in concepts
that are derived from ideology and not science.’”
Agre is widely known as a modest man, and
when he took the podium he talked about his
career in science and its relationship to the
rest of his life, particularly the support he had
received from his family and colleagues. A
high point of the talk was his solo performance
of “the Elements song,” by Tom Lehrer,
dashing through the entire periodic table of
the elements in rhyming verse.
��
Since Agre’s key discovery in the 1980s, scientists
at the FMP and elsewhere have
learned a lot about how aquaporins work.
These proteins float in the plasma membrane
– the barrier between the cell and the world –
where they create passageways. Proteins are
folded in specific ways as they are made, and
the folds of aquaporins leave a gap through
the middle that permits the passage of water
molecules. When the body needs water, aquaporins
are added to the plasma membrane of
cells; when enough has been absorbed, they
are removed again. Cells know how much
water to let in because of hormone signals
that get released when the supply gets low.
The hormone doesn’t directly communicate
with aquaporins; first it has to trigger another
signal, called cAMP, which does that job.
Enno Klußmann began looking at aquaporins
as a postdoctoral fellow in the research group
of FMP Director Walter Rosenthal. He joined
the lab in 1997 – “on loan” from the Günther
Schultz’s group at the Freie Universität Berlin,
where he had attempted to identify a receptor
for steroid hormones in the cell membrane.
University regulations permitted postdoctoral
fellows to remain no longer than five
years at a time, and Schultz recommended
that he join the group of another former student,
Walter Rosenthal, at the FMP. “The idea
was that I would leave for a few months and
then go back,” Enno says. “But Walter’s lab was
working on some fascinating things, and ten
years later, I’m still here.”
One of the group’s projects was the study of
aquaporins. “The way that aquaporins behave
in the kidney is important and interesting,”
Enno says, “but it opens the door on some
187 Part Three: Frontiers and ferrymen
even more interesting questions. cAMP is a
strong signal, found almost everywhere in the
cell. It triggers a wide range of other processes
often at the same time, in the same cell.
That situation raised the question of how the
cell controls the signal – how it manages to
respond in specific ways, in specific places.
Without controls, everything would be happening
everywhere, all the time.”
Many processes triggered by cAMP are linked
to diseases – including hypertension, gastric
ulcers, and heart disease. It turns out that
some common drugs work by preventing
other proteins from “hearing” cAMP signals.
That’s desirable if the drug blocks only one
kind of activity. But there may also be unintended
side effects. In the 1990s doctors discovered
that lithium, commonly used to treat
people with a psychiatric problem called bipolar
disorder, was interfering with aquaporin.
Patients on the drug developed a form of diabetes
in which their bodies lost their ability to
recycle most of the water. Even though they
drank all the time, most liquids just passed
right through without being absorbed. This
was bad news because lithium was helping
many bipolar patients, and it was also being
considered as a supplement in the treatment
of other diseases. A better understanding of
aquaporins might help researchers find a way
to prevent the side effects.
A drug that blocked all of cAMP’s activity
would be fatal, but maybe there were ways to

stop a single process. Learning to do that
might show how to block other processes.
Enno and Walter hoped that an aquaporin
called AQP2 would make a good place to start.
��
Things went so well for Enno’s work at the
FMP that he could set up his own lab in
2001. He brought along the aquaporin project,
and has continued to work on it in collaboration
with Walter and a network of researchers
across Europe. He heads a project funded by
the European Community to identify or create
small molecules that can control cAMP signaling
in processes related to human diseases.
Enno and Walter have also been awarded a
Part Three: Frontiers and ferrymen
188
number of grants from the German Research
Council (DFG) to study particular aspects of
cAMP signaling and aquaporins.
Enno says that understanding how cells control
water channels – and thus the body’s
water supply – boils down to several questions.
Those have to be tackled one by one,

and over the past few years the FMP has
made progress on them all. The first question
is how the molecules are moved to the locations
where they are needed.
“The molecules triggered by cAMP have to get
to the right place to respond properly to a signal,”
Enno says. “One way to do that is to tie
them up to particular membranes. In addition
to the plasma membrane that surrounds the
cell, there are a lot of internal membranes
that often serve as duty stations. Tethering is
how the system that activates aquaporins
works. AQP2 is first stored in internal membranes
before it is moved to the cell surface.
Inside the cell it undergoes a kind of pre-fabrication;
it is installed alongside the machinery
that controls it. That includes a switch, which
is the actual receiver of the cAMP signal,
called a PKA protein. It also includes the
anchor that attaches the PKA to the membrane,
and other molecules. There are lots of
types of anchors – so far, about 50 have been
found. The anchors for PKA are known as
AKAPs (for A-kinase anchoring proteins). The
cell has so many because each one is used to
anchor a particular type of molecule to its
proper place.”
In 2001 it wasn’t certain that aquaporin functions
depended on the interaction of PKA with
an anchor, so Enno and Walter decided to find
out. For their experiments they used cells
taken from rat kidneys, raised in laboratory
cultures, which produced aquaporins.
Microscope studies had shown that water
channel proteins were held in the interior of
the cells, in small membrane-wrapped compartments
called vesicles, until they were
needed. When the cAMP signal was received,
the aquaporins moved to the plasma membrane.
The scientists began shutting down
proteins in the cells, hoping to block this
movement. Success would mean that they
had found a protein that had something to do
with positioning the aquaporin. Since it was
likely that an AKAP was involved, one thing
they tried was to prevent the anchor from
binding to PKA. To do this they used a small
molecule, a synthetic peptide, known to block
the contact between the two molecules.
Cutting an anchor sets a boat adrift, and if an
AKAP is disturbed, signaling molecules float
away from their duty stations – or never
become attached there in the first place. This
renders the anchor an interesting target for
drugs. “You wouldn’t want to block cAMP
itself to stop a process – that would disrupt a
lot of other important things going on in the
cell that rely on the signal,” Enno says. “The
anchor is a much better point of attack.”
Enno and Walter introduced the synthetic
peptide that disrupts anchors into the kidney
cells; they discovered that aquaporins were no
longer arriving at the membrane. An anchor
was obviously involved, but which one? In
2006 Postdoc Volker Henn, PhD student
Bayram Edemir found a candidate: a protein
called AKAP18-δ. Microscope studies and
other methods showed that AKAP18-δ was
installed alongside AQP2 in the cell interior
and moved along with water channels when
they moved to the plasma membrane – “guilt
by association.”
Another version of the AKAP18-δ protein influences
the contraction of the heart; there,
cAMP signals control a different type of channel
which allows the passage of calcium ions
into the cell. These channels are stimulated
through another hormone. “The situation is
similar to that of aquaporins,” Enno says. “The
hormone signal triggers a wave of cAMP signaling
that opens channels in the plasma
membrane. But only if the channel proteins
are connected to anchors.”
Beta-blockers, which are commonly used in
the treatment of cardiovascular disease, work
because they tune down the heart’s response
to the hormone. The same thing might be
achieved by blocking the connection between
anchors and their partners, such as calcium
channel proteins. “This makes the anchor a
very good target for drugs,” says Christian
Hundsrucker, another PhD student in Enno’s
group. “The peptides we had been using to
189 Part Three: Frontiers and ferrymen
block AKAP-PKA interactions don’t do so very
effectively. So we began trying to make new
ones.”
The scientists began by taking a closer look at
the docking site. Normally two PKAs form a
pair when the DD domain, a module in one
protein, binds to its counterpart in the tother.
Two domains joined in this way create a surface
called RII that plugs into the anchor.
Structural pictures of anchors had already
been obtained using NMR and X-rays,
providing a model of one of the surfaces. This
could be used in “docking” simulations –
twisting and turning the RIIs until the molecules
snapped together on the computer
screen.
Christian says that the anchor has a coiled
structure which fits into a pocket formed by
RIIs. Whether the coil fits at all and how
strongly it binds depend on the shape and
chemistry of the two structures, which are
determined by chemistry of the the amino
acid building blocks that make up the proteins.
Christian hoped to make a molecule
that would fit better and bind even more
strongly to RII, which would then lock onto
the domains of PKAs and prevent them from
acquiring an anchor. Christian began with a
fragment of AKAP18-δ and started to change
its recipe.
Understanding how the surfaces of two proteins
bind is an enormous chemical puzzle
that would probably be impossible to solve
without the help of sophisticated computer
techniques. Christian and his colleagues
enlisted the help of the lab of Gerd Krause, a
colleague at the FMP; they were using software
to study protein binding. By scanning
features of several different RII domains and
their binding sites on AKAP proteins, the scientists
could detect specific amino acids that
seemed to be crucial to linking the molecules.
Michael Beyermann’s group, experts in building
small proteins, helped design new versions
of the binding domain by replacing those subunits
with others that might make a better fit.

The resulting artificial molecules were
screened in heart muscle cells. The strategy
had worked – they were taking the place of
anchors, breaking the connections between
AKAPs and their targets. “What this showed
us is that we have identified the key points
that permit binding between the anchor and
its PKA partner,” Enno says. “This gives us a
much clearer strategy as we develop new synthetic
molecules.”
The dozens of known AKAPs may have evolved
from a single ancestral molecule, inheriting a
common “docking station” that allows them
to bind to their main partner, PKA. The situation
is a bit like the way a boarding ramp at an
airport fits the hatches of different airplanes.
You could design something that would prevent
a plane made by one manufacturer from
docking onto the ramp, but it would probably
Michael Gomoll
also interfere with other types of planes. In
the same way, a drug that stopped one anchor
from binding to its PKA would probably stop
others. So the binding site between anchors
and PKA proteins probably wouldn’t be the
best choice to interfere with.
But the experiment is important for another
reason, Enno says. “The approach is valuable
as a proof of principle to verify that pharmacological
interference with AKAP function is a
feasible concept. We know a lot about the
function of AKAP-PKA interactions. Less is
known about functions of AKAP interactions
with other proteins.”
Instead of disrupting the contact between the
anchor and PKA, the scientists figured it
would be better to aim a drug at another part
of AKAP18-δ – its targeting region. “This is the
‘address label’ that sends the anchor to the
same location as aquaporins,” Enno says.
“Interfere with that and the anchor gets lost,
so it can’t help aquaporins move.”
One goal is to produce potential drugs; another
is to give scientists tools to dissect other
aspects of the molecules’ activity. Aquaporins
don’t work forever, even when the cell is flooded
with cAMP. A study begun by PhD student
Eduard Stefan revealed one of the reasons. As
the anchor links to an aquaporin, it also brings
other molecules on board. One of them, called
a PDE, is a sort of volume control for cAMP. It
works by cutting a chemical link inside the
cAMP molecule – like disconnecting a wire
inside a radio – leaving pieces that can no
longer pass the signal.
PDE molecules are found throughout the cell,
and they help tune down the signal. This fact,
plus the cell’s ability to install different sen-

sors for cAMP in different places, is what
keeps the cell from becoming overstimulated
by the signal all the time. That they can bind
to an AKAP means that the anchor collects an
entire toolbox of molecules that give the cell
a fine level of control over the water channels.
��
One of the lab’s ongoing projects is to
clear up yet another question about
how aquaporins take up their proper positions
in the cell. cAMP delivers the signal to
move the water channel protein to the cell
membrane, but what actually carries the
aquaporin there? This is really a two-part
problem: if you want to move something
through the cell you need a track to move
things along and a motor to do the pulling.
Enno says that scientists had been focusing
on microtubules, a network of fibers that act
as a scaffold to give the cell its shape and also
function as tracks along which molecules are
delivered throughout the cell. Motor proteins
act as the locomotives, traveling down the
microtubules with other proteins in tow.
“There was experimental evidence that artificially
breaking down microtubules stopped
the transport of aquaporins,” Enno says. “And
motors that use microtubules had been
found in the same vesicles that contain aquaporins.
This looked like another case of preassembly,
in which aquaporins and the molecules
that were needed to transport them
were being wrapped up in the same package.”
However, another lab had recently done a
study of all the proteins found in aquaporin
191 Part Three: Frontiers and ferrymen
vesicles, and they failed to turn up the motor
proteins. Enno began to wonder whether
microtubules were really the delivery route –
after all, they weren’t the cell’s only transport
system. In 2005 a Japanese group did an
experiment in which a drug was used to break
down the microtubules in cells taken from
kidneys. They discovered that even without a
microtubule railway system, aquaporins could
be delivered to the membrane. Pavel
Nedvetsky, a postdoctoral fellow in Enno’s
group, hoped to find an explanation.
“Some papers from 20 or 30 years ago had
already hinted that another type of filament
called actin fibers might be involved,” Pavel
says. “So we began looking for another transport
system, one that could move along these
filaments.”

Working with other members of the group
and international partners, Pavel made an
important discovery: a motor protein called
myosin Vb, known to shuttle other types of
vesicles back and forth along actin fibers, was
consistently found in the neighborhood of
aquaporins. Motors often require adaptor proteins
and other helpers to find and attach a
cargo, and one of myosin Vb’s usual assistants
is a protein called Rab11. This gave the scientists
two ways to investigate whether myosin
Vb was acting as a transporter for aquaporins
– by interfering with the motor and with the
adaptor.
Pavel first used a non-functioning version of
the motor – it didn’t work because it was
missing the regions that normally attach it to
filaments. He introduced it into the kidney
cells and watched what happened when they
were stimulated with the “thirst” signal.
Normally they would take in extra water. Now
they didn’t, and a look under the microscope
showed why: aquaporins were trapped in
internal vesicles, and were no longer being
transported to the cell membrane. In another
set of experiments, Pavel interfered with
Rab11’s ability to bind to the motor. The same
thing happened; aquaporins became stranded.
“These experiments gave us an idea of some
of the proteins responsible for moving aquaporins
around,” Enno says, “but they still didn’t
tell us exactly what was happening. To understand
what we found next, you have to realize
that most of us believe that aquaporins get
recycled. That means that an aquaporin isn’t
finished once it has moved to the plasma
membrane. It helps absorb water as long as
the signal to do so continues, but when the
signal stops it has to be removed from the
membrane. We think that it gets transported
to an internal compartment, where it gets put
on hold, and then it can be called up for duty
again.”
Returning to the microscope, the scientists
found that some aquaporins managed to
Control
AVP
MyoVb tail AQP2 ZO-1 Overlay
Interfering with the tail of the “motor protein” myosin Vb (green) affects the normal location of
aquaporins (red) and how they are recycled in the cell. The top row shows cells with normal
myosin Vb, and the bottom row shows how the locations of the motor and aquaporins if the
motor doesn’t function. Second column: Aquaporin 2. Third column: labeling ZO-1, a protein found
in tight junctions, clearly shows where the borders of the cells are.
make it to the cell membrane without myosin
Vb, but once there, they couldn’t be shuttled
back inside or out again. The cell handles new
aquaporins a bit differently than those it recycles
– other motors and helpers probably lend
a hand. But myosin Vb and Rab11 appear to be
essential in handling the recycling of the
motor channels. Myosin Vb is thought to have
a similar role in other recycling events.
“The aquaporin story is interesting when you
think about diseases in which the body
retains too much water, like hypertension or
congestive heart failure,” Enno says. “In those
cases the body still needs to be able to absorb
water – just not so much. Maybe you could
temporarily shut down the recycling center
alone, without also turning off all the intake
valves. To do that you’d have to get a good look
at the recycling machinery, and find something
you could perturb without disturbing
the rest. That’s where these experiments are
taking us.”
193 Part Three: Frontiers and ferrymen

Bribing the ferryman
In Greek mythology, the dark river Styx separates the world of the living
and the dead. It is traversed by a ferryman named Charon, a
gloomy, glowering giant who floats up and down the river, guarding the
entrance to the underworld. To appease him the Greeks buried their
dead with a coin under the tongue – a bribe, or a fare – to secure a passage
to the afterlife. Those that didn’t have the coin, or who had been
improperly buried, were cast out of the boat.
Smugglers throughout history have used bribes to move their wares
cross borders, so it shouldn’t be surprising that scientists are trying
something like a bribe to slip drugs and other substances into cells. In
this case the barrier that has to be crossed is also a sort of river: the cell
membrane, two liquid layers of fats and proteins. One of its key functions
is to keep out toxins, parasites, and a wide range of other foreign
substances.
“It wouldn’t do much good to design the perfect drug,” says Sandro
Keller of the FMP, “if you couldn’t get it to a cell that was diseased, and
then get it inside the cell to do its job. Managing that means designing
a delivery vehicle that can get substances through the cell membrane.”

Sandro Keller

Once inside the cell, drugs have to survive other cellular defenses that
target and break down foreign substances. And ideally the substance
should last a long time; otherwise, the drug would have to be administered
continually, in such high doses that it would probably be toxic.
The task is tricky enough when the cargo to be delivered is a small protein
or chemical compound, like most of today’s drugs. Scientists are
now hoping to insert other things into cells: DNA, RNA, large proteins,
or complex molecules that can serve as probes of the inner workings
of the cell. Getting these past the membrane will require understanding
how it is built and how it functions, and that’s the task Sandro and
his laboratory have devoted themselves to.
��
It’s not often that a young scientist goes directly from receiving his
PhD to his first independent research position, without moving
abroad and passing through a stint or two as a postdoctoral fellow,
but that’s the case with Sandro. It has been two years now since the
Part Three: Frontiers and ferrymen
transition, and he seems entirely at home in his office at the FMP; he
has brought in some plants, and the bookshelves are full. But those
who don’t know him – including the organizers of some of the conferences
he gets invited to speak at – still mistake him for a student. He’s
young, and looks young for his age, which might make it difficult to
pick him out in pictures of the group. Sandro’s fast track to leading his
own team is the result of a success story: while working on his PhD,
Sandro began taking an interesting approach to the study of the cell
membrane and how to get cargos inside.
The membrane has two liquid layers, like a soap bubble. Both layers are
built of fat molecules called lipids, interspersed with proteins. Many
membrane proteins have a head region that hangs outside the cell, a
tail that dangles in the interior, and a complex folded region that
spans the two layers of the membrane. What happens in this zone has
been difficult – often impossible – to observe, but that will probably be
necessary to solve the problem of drug delivery.
Sandro would like to get a look at the regions of proteins embedded in
the membrane, to understand how they are linked to each other and
to lipids, and to watch what happens to those membrane structures
as molecules pass through on their way into the cell. The problem is
that the inner space of the membrane is a bit like the zone of radio
silence that astronauts pass through as they reenter the earth’s
atmosphere. It is hard to get any direct information from the zone.
“The classical methods used to study the structures of proteins don’t
work with proteins that span the membrane. We have to be very creative
and use less direct, biophysical techniques.”
One method that the group is using involved creating artificial molecules
that can function as probes. The goal is to understand how the
subunits of proteins interact and cause the protein to fold in the presence
of lipids. The folds determine the molecule’s shape and functions.
“We started with a simple bacterial protein called Mistic, which has
four domains that pass through the membrane,” Sandro says. “We
have created four separate, artificial molecules – one corresponding to
each domain – that can be used as a model to study the behavior of
the protein. These regions are called -helical domains; they are found
in many different types of proteins, and they play a crucial role in the
196

structure and organization of the membrane. But we know very little
about how internal forces in the molecules cause them to fold and
interact in certain ways. We can manipulate these artificial molecules
to test hypotheses about these processes and to expand what we
have learned from Mistic to many other molecules.”
The project is the subject of a grant proposal that Sandro submitted to
the German Research Council (DFG) and which was approved in Spring
2007. The grant has given Sandro two extra positions to pursue the
project in his lab. Getting a glimpse of protein structures and functions
in the membrane would help solve questions about how other
molecules enter the cell. Most methods used to observe this process
have drawbacks, Sandro says. Some leave “artifacts” that make it seem
as if a protein has crossed into the cell when it hasn’t. Other techniques
are restricted to one aspect of the problem; for example, they
either show how a protein binds to the cell surface, or give hints about
how it passes through – but not both. Different kinds of experiments
give contradictory results.
“All of this confusing data has led to debates about how molecules
enter,” Sandro says. “There has been quite a bit of discussion about the
peptide penetratin, which manages to get through the barrier.
Penetratin is a small module of another protein that has to move into
the cell during the development of the fruit fly. Penetratin is like a key,
and if you attach it to another molecule, it sometimes will escort that
into the cell as well. What was unclear was whether penetratin somehow
entered the cell on its own, or whether it came wrapped in a vesicle
– sort of a small membrane bubble. Cells often swallow up vesicles
in a process called endocytosis. The process by which they absorb single
molecules is different, and it’s important to know which one is
happening.”
In 2006 Sandro and collaborators at the FMP (Michael Bienert and
Margitta Dathe) and at the Martin Luther University Halle-Wittenberg
helped to demonstrate that the latter was the case, using a method
called isothermal titration calorimetry. The technique monitors and
compares the temperature of two cells: one a control, and one which
is subjected to experiments. In this case the researchers injected pen-
197
etratin into the experimental cell. They monitored its interactions with
lipid molecules by measuring the heat released upon adding lipid vesicles
into the cell. The instrument is so precise that it can distinguish
between different types of chemical reactions based on the heat
released or consumed in the reaction cell. The study showed that penetratin
was not entering on its own – suggesting that endocytosis
plays a key role in the cellular uptake of the peptide.
“Calorimetry is very good for tracking the movement of substances
into the cell,” Sandro says, “but it is not easy to use, the results are difficult
to analyze – and most researchers simply don’t have access to it.”
Searching for a simpler method, Sandro and his colleagues began
using fluorescence spectrometry to watch the passage of proteins
through membranes. “This approach monitors lipid vesicles with ultraviolet
radiation and measures light instead of heat,” he says. “It can be
used with any molecule whose fluorescence properties change when
it comes in contact with the cell membrane.”

Natalie Bordag
The method uses ultrasound (sound that is inaudible to the human
ear) to force the compound of interest into the interior of lipid vesicles.
The vesicles are then diluted to “loosen up” the interactions between
the compound and the lipid molecules. If the compound can cross the
lipid membrane and leave the vesicles, its fluorescence properties
change since it is no longer in contact with the membrane. If, however,
the compound is not able to get out of the vesicles, it will remain
bound to the membrane, and its fluorescence properties won’t
change. The technique passed a proof of principle test by confirming
the results of the earlier calorimetry studies of penetratin, and the laboratory
is now using it to study other molecules.
��
Anew delivery vehicle for drugs will need a key – a cell-penetrating
peptide like penetratin – ideally, one giving access to specific,
well-defined types of cells. Whether a vesicle can dock onto a cell and
deliver its cargo depends on the proteins and lipids on its surface. But
having the right ingredients isn’t always enough; they also need to be
put together the right way.
“One of the simplest sorts of transport vehicles that people use to try
to introduce substances into the cell is a small fat droplet called a
micelle,” Sandro says. “This is a small greasy drop that self-assembles
because of its chemistry. It’s the same phenomenon that you see
when you put oil into water – it forms a drop. That’s because each oil
Part Three: Frontiers and ferrymen
molecule has a hydrophobic tail that wants to avoid contact with
water and a hydrophilic side; the oil molecules cluster together and
turn the hydrophobic tails inward, away from the water, toward the
inside of the droplet.”
Even when accompanied by a cell-penetrating protein like penetratin,
micelles don’t always work well as transporters, he says. Micelles
assemble as a single layer, and the cell membrane is a double layer of
lipids. A membrane protein in a single-layer micelle probably won’t
fold or behave the way it would in its normal environment of a doublemembrane
layer.
Some proteins dock onto lipids and are able to assemble membranes
around themselves. The best-case scenario would be to have a delivery
protein that did this, as well. If it could collect a high number of lipids,
they might form a double layer. After a search, Margitta Dathe and colleagues
at the FMP decided to work with a small molecule called A2.
They discovered that attaching this peptide to a vesicle caused it to be
absorbed by capillary cells in the brains of rats. Sandro wondered
whether A2 could be prompted to assemble lipids around itself.
That’s where the bribe came in; something could be added to A2 to
make it more likely to build vesicles. “We knew from the literature of a
way to do this that might work,” he says. “If you create an artificial
molecule by combining a small protein and fatty acids, it tends to
accumulate a lot more lipid molecules – maybe enough to make a
double-layered membrane. So we modified A2, giving it binding sites
198

that lipids could dock onto, and another element that should serve as
a binding site for glycoproteins. These proteins are found in membranes
everywhere and when they bind to their partner molecules,
they cause them to be taken up into cells.”
In the test tube, the new hybrid A2 molecules bound to each other in
clusters. Isothermal titration calorimetry revealed that the groups contained
a “record-setting” number of lipids, compared to other hybrid
molecules. A positive sign, Sandro says, because it made it more likely
that delivery vehicles with double-layered membranes would form.
Next it was time to see how cells would respond to the peptide. The
scientists “doped” vesicles with A2 and watched what happened using
the laser scanning microscope and other techniques.
They discovered that vesicles with A2 readily docked onto cells and
were carried through the membrane very efficiently; once inside, they
dissolved slowly. This hints that in addition to being better at entering
cells, vesicles with A2 might protect a drug long enough for it to be
effective.
All in all, the new version of A2 seems to work well as a “bribe” to get
cells to take up drugs and other substances. “The A2 vesicles and
micelles seem to be a significant improvement over some other types
of delivery vehicles,” Sandro says. “It suggests that you might be able
to add A2 to other transporters and improve their characteristics.
We’re trying that now. We also hope to use the vesicles to solve other
kinds of problems – for example, to insert complex probes that can
monitor cellular processes.”
��
The River Styx and its ferryman are images from the dark corners
of Western culture; they figure prominently in Dante’s Inferno
and centuries of retellings of the myth of Orpheus. Michelangelo
depicted Charon and his boat among the writhing figures of the Last
Judgment. Despite their gloominess, the literary references appeal to
Sandro, who considered doing many other things before committing
himself to science.
Berlin doesn’t have the Styx, but there are places along the Spree
downtown where at night the water turns black and takes on a bleak,
industrial sort of feeling; the mythological river might look that way if
you transported it into the 21st century. Maj Britt, this book’s photographer,
had found a good place along the river to take Sandro’s portrait,
and he liked the idea. On the way into the city we talk about literature.
At the moment he is rereading Candide, by Voltaire, and is
working his way through the books of Sartre and Camus.
He considered going into literature or linguistics, he says. He grew up
in Grisons in Southeastern Switzerland – a breeding ground for polyglots,
with three official languages. His father was a chef in a great
199
Mistic is a small membrane protein from bacteria which is built of four small
helices that cross the membrane. Because the protein folds correctly in water
and membranes, Sandro and his lab can use it to compare the process of
folding in different environments.
restaurant, Sandro says, who comes from the German-speaking part of
Switzerland; his mother is a native of Austria, a kindergarten teacher,
and a practitioner of extreme sports. “We spoke German at home and
as a result I was the only kid who couldn’t speak Romansch when I
went to kindergarten,” he says. “That changed quickly; by the fourth
grade I had completely lost my German. We had it as a foreign language
one hour a week, and I remember I couldn’t spell anything.”
He took the “language track” in school, where he says no subject was
boring. He couldn’t imagine focusing on only one thing. But he was
eager to get out in the world, so he left school early, studied hard at
home for a year and a half, and passed his school-leaving exams at the
age of 18. After working as a meteorologist during his military service,
he entered the university and committed himself to biophysical chemistry.
That led to a PhD on the biophysics of membranes, and now his
own group. Which now has nine members. More people means that
the lab gets more things done, but it also costs time.
“Coaching is important,” he says. “You try not to push, not to be too
directive, but to be supportive.” He talks about a student who was
struggling after six or seven months in the group. “He was getting
frustrated because his project wasn’t working out. You have to sit
down and help them learn to be patient. Not to lose the big picture.
And to keep their eyes open. Sometimes when an experiment doesn’t
work out, it’s still giving you an answer – just not to the question you
asked.”
Part Three: Frontiers and ferrymen

The electrician’s toolbox
The earliest recorded cure for a migrane involves standing on a
fish. It sounds like the beginning of a bad joke, but it’s true. The
Torpedo, or electric ray, is a round, flat fish that lives on the sandy bottom
of the ocean. It stuns its prey by delivering a massive electrical
shock. This fact has been known since the ancient Greeks, although the
first reference to the animal comes from philosophy rather than medicine.
In Plato’s dialog on the nature of virtue, Meno says to Socrates,
“You seem to me both in your appearance and in your power over others
to be very like the flat torpedo fish, who torpifies those who come
near him and touch him, as you have now torpified me, I think. For my
soul and my tongue are really torpid, and I do not know how to answer
you.”
The first medical use of the fish comes from Scribonius Largus, court
physician to the Roman emperor Claudius. Although he was not the
emperor’s personal doctor (that post was held by Xenophon, who later
poisoned Claudius with a feather), Scribonius was regarded highly
enough to accompany Claudius on a campaign to Britain in 43 ce. Later
he was asked to record his medical knowledge and wrote a book called
the Compositiones, which contains 271 prescriptions. Its preface contains

Thomas Jentsch
the earliest recorded reference to the
Hippocratic Oath, in which doctors are
advised to do no harm.
One of the worst things a doctor can do,
Largus writes, is not to give a drug when he
knows it works. But one must be highly educated
and critical, and Largus denounces
some of the remedies proposed by other
physicians. He points out that drinking the
blood of a freshly killed gladiator, for example,
will not cure epilepsy. Standing on a Torpedo,
on the other hand, can cure the pain of gout
in the feet (which become numb from the
shocks). For migranes, a patient can immerse
his hands in a basin of water containing one
of the creatures.
Some of Largus’ prescriptions may have
worked well. For swollen tonsils he described
a complex mixture containing:
2 drachmas each of costus, celery seeds,
anise seeds, oil of camel grass, and
cinnamon-cassia, one-half drachma of
cardamom, 2 drachmas of the wild rue
(two-thirds of which is the seed), onehalf
ounce of fissile alum, 5 mediumsized
ground-up oak galls, 2 drachmas
of saffron, one-half drachma of the
refined residue of the oil of saffron,
one-half drachma of myrrh, 4 drachmas
of Cretan birthwort, 3 drachmas of
cinnamon, one ounce of the ashes of
a young wild swallow, and one-half
drachma of spikenard

202
Several of the ingredients contain biologically
active compounds; the pharmacological
screening library at the FMP (described in the
next story) likely includes extracts from some
of them. (Well, maybe not the ashes of a
young wild swallow...)
��
The FMP doesn’t have a Torpedo, either,
although a researcher who splits his
time between the FMP and MDC is interested
in the animal. For nearly 20 years Thomas
Jentsch has been conducting a systematic
study of an ancient and now widespread family
of molecules called chloride channels.
When they first evolved in one-celled organisms,
their function was to sense changes in
the environment and help the cell survive by
adjusting its chemistry. With the arrival of
multicellular life, they allowed cells to communicate
with each other and coordinate the
activities of the body. Dozens of types of chloride
channels are known. Thomas’ work has
focused on an important subset of these molecules,
called CLCs, and other ion channels and
transporters.
“The human genome contains nine CLC chloride
channels and transporters, all copies of a
single gene that evolved long ago,” Thomas
says. His laboratory found the first example
over a decade ago and has systematically
uncovered many more. “The extra copies arose
because of errors that were made as DNA was
replicated, and over time they have evolved
different functions.”
Many of these molecules float in the plasma
membrane on the fringes of the cell, where
they act as doormen. Their job is to control the
flow of negatively-charged chloride atoms
(ions) through the membrane. Sometimes
this traffic of ions balances the overall charge
of a cell with that of its environment. At other
times an imbalance builds up, creating a positive
charge on one side and a negative charge
on the other. In between is the thin cell membrane,
which suddenly becomes a superb conductor
of electricity. In neurons, this causes an

Different ways that proteins of the CLC family assemble to form channels in membranes.
electric charge to race huge distances down
the surface to the far end of the cell, where it
opens and closes channels. That causes a
release of molecules which drift over to the
next cell, open its channels, and start the
process all over again.
The movement of ions across the cell membrane
is crucial to a wide range of cellular
processes, for example moving salt and water
into and out of the cell, or coordinating the
activity of muscles and neurons. Some CLCs
are lodged in membranes inside the cell. One
of their jobs is to recycle cellular waste. But
scientists still don’t have a full view of the
functions of channels and transporters in the
body’s tissues.
The main method of finding out has been to
delete genes one-by-one in model organisms
like the mouse and then to closely watch the
animals’ development and behavior. “Another
approach has been to guess that a CLC gene
might be mutated in certain human disorders,”
Thomas says. “If we find CLC mutations
in certain types of patients, then diseases can
also provide powerful clues toward the normal
function of the protein.”
Over the last two decades Thomas has taken
on the CLCs one-by-one to uncover their func-
tions. Some of these projects have yielded
insights into disease and have promising
therapeutic implications. Thomas’ experiments
have shown that removing one channel
leads to deafness; when another is deleted,
kidney stones form; yet another is needed
to maintain the normal density of bones, and
others seem to be connected to conditions
like epilepsy.
To understand why a defect in a chloride
channel or transporter has these effects,
Thomas says, you need to understand how it
works in healthy cells, but that’s not easy. It
requires a sophisticated array of techniques –
from structural investigations of the building
plans of the molecules, to their roles in mouse
cells and human diseases. Thomas’ new home
at the Timoféeff-Ressovsky building on the
Berlin-Buch campus has been an ideal place
to put together an interdisciplinary team.
Some methods have been imported from his
last post at the Center for Molecular
Neurobiology at the University of Hamburg.
Others are available at the FMP. He can also
draw on resources from the MDC, his second
employer.
��
203
Part Three: Frontiers and ferrymen

Lena Wartosch
Like many of Buch’s scientists, Thomas’
interests extend to art and culture. His
office is adorned with cult objects he
obtained in Papua New Guinea: carvings of
the heads of crocodiles and other creatures.
They all have a triangular shape, which
reflects their source: originally, each was an
ornament carved into the prow of a boat.
When a boat falls into disuse, Thomas says,
the carvings are removed. Then, at least in
some cases, they are sold to scientists from
Berlin.
They look over his shoulder with fierce expressions
as he talks about his 20 years of
research into membrane proteins.
“Our cells contain an astounding variety of
channels for the transport of ions,” he says.
“There are over 60 different genes that
encode channels to transport potassium ions.
Why should there be so many molecules that
have about the same function? The diverse
roles of some of them have been discovered
through studies of knockout animals, or
through examination of tissues of patients
suffering from diseases, but a great number
remain puzzles. I believe that as they are
investigated, we will uncover many more connections
to disease.”
In the late 1980s, when Thomas began studying
the molecules, this work was in its infancy.
Scientists knew that chloride channels had to
exist but had not yet identified any genes
encoding them. A good place to look, Thomas
thought, was the Torpedo.
His interest stemmed from the fact that
Torpedo cells require an unusually high number
of chloride channels to produce the massive
electric discharge, which can reach levels
Part Three: Frontiers and ferrymen
204
of 100 volts, with currents up to one ampere.
Everyone knew the channels had to exist, but
attempts to isolate a CLC from the fish’s cells
had failed.
“To find a protein you needed a molecule that
would bind to it quite specifically, and we
hadn’t found one for the chloride channels,”
Thomas says. “So we took another approach;
we decided to clone Torpedo RNAs that might
contain the channel and put them into
Xenopus egg cells – a frog which is widely
used as a model organism.
The genes of these eggs are inactive, but all
the machinery is there to transform RNAs into
proteins. The cell used the Torpedo RNAs to
make proteins. By watching its electrical
behavior, we could tell whether we had
imported a molecule that functioned as a
channel.”

This could be checked using a voltage clamp
technique. In this method an electric current
is “injected” into the egg and scientists measure
the output – changes of voltage over the
egg’s membrane. If the cell has extra chloride
channels – because it has made them from
Torpedo molecules – the charge will leak out
through them, and more current needs to be
injected into the cell to reach the same voltage.
The measurements allowed Thomas and his
colleagues in Hamburg to identify the first
CLC and pin down specific regions of the protein
that were responsible for opening the
channel and moving ions. Over the next
decade, Thomas and his laboratory identified
several new CLCs and began removing them
one-by-one from mice to study their functions.
The work has revealed numerous connections
between defects in the channels and
disease. In several cases, the scientists first
discovered patients who carried mutations in
the genes, and then later created mouse models
of the diseases.
For example, a lack of ClC-1 leads to a condition
known as inherited myotonia. A person’s
muscles don’t relax after voluntary activity
like shaking someone’s hand or squinting in
bright sunlight. The condition was first
described in 1876 by Danish physician Julius
Thomsen, who suffered from it himself, as did
several members of his family. It took more
than a century, but Thomas and his colleagues
finally traced the disease to the chloride channel
gene and even pinpointed the specific
mutation that caused the disease in Dr.
Thomsen himself.
“We think that by allowing chloride ions to
flow into the cell, ClC-1 helps muscles relax
205 Part Three: Frontiers and ferrymen
and ‘reset’ so that they are ready for the next
stimulation,” he says. “Without the molecule,
external potassium levels build up and cells
begin to stimulate themselves. This leads to a
sort of seizure in which muscles no longer
relax.”

It wasn’t hard to imagine something similar
happening in other tissues. Thomas was
intrigued by a report from a clinical group
which had discovered mutations in ClC-2, a
version of the channel found in the brain and
other organs. The family that had the problem
suffered from epilepsy. Was the loss of channel
proteins causing an overstimulation of
neurons, leading to brain seizures? He was
skeptical. A couple of years earlier, his lab had
developed a strain of mouse without ClC-2.
Although mice can suffer from epilepsy, the
knock-out animals didn’t. Instead, they experienced
degeneration of the testes and the
retina.
Judith Blanz, a student in Thomas’ group at
Hamburg, decided to conduct a more careful
examination of the nervous system of mice
lacking this chloride channel. She discovered
that while individual neurons still had their
normal forms, the animals’ brains and spines
developed severe defects. “They took on a
sponge-like appearance because of holes that
formed in the white matter of the brain,”
Thomas says. “The only real behavioral difference
was related to blindness, caused by the
retinal degeneration we had noticed in the
previous study. And the mice responded
somewhat more slowly to stimuli, caused by
holes in the white matter. Signals in the brain
were not being transmitted as quickly.”
The sponge-like gaps in the brain were familiar;
they also appear in a human disease
called leukodystrophy. Researchers had
Part Three: Frontiers and ferrymen
206
already linked several genes to different forms
of human leukodystrophy, but some families
suffering from the disease didn’t have mutations
in any of these genes. So other, unknown
genes might be involved, including ClC-2. The
scientists obtained DNA from a group of 150
leukodystrophy patients and began looking
for a pattern of mutations. “We didn’t find
them,” Thomas says. “That was disappointing;
on the other hand, the damage in the brains
of humans who suffer from this condition is
mostly more severe than what we observe in
the mice. We may have picked the wrong subset
of patients; we don’t know of human
patients where the symptoms are combined
with degeneration of the retina. Such a case
would more closely resemble what we see in

the mouse. We are currently collaborating
with groups in human genetics to try to find
such patients.”
��
Amutation’s effects on health are often
strange and unpredictable. While looking
at a potassium channel called KCNQ2, the
scientists found mutations which cause
epilepsy in human newborns. The discovery
suggested that KCNQ proteins would make a
good target for anti-epileptic drugs. Later it
was found that a known substance, currently
in phase III clinical trials for use in treatments
of epilepsy, had its effects by docking onto
KCNQ potassium channels. Pharmaceutical
companies are now trying to develop
207

more specific anti-epileptic drugs by targeting
the channels. It’s another example of the
beneficial side-effects of basic research,
Thomas says: finding a gene responsible for a
very rare genetic condition can sometimes
lead to therapies for much more common
diseases.
Another serendipitous discovery arose from
studies of mice that lacked ClC-7. “Removing
this gene leads to unusually thick and brittle
bones in mice,” Thomas says. “So we had a
look at human bone diseases. We found a
defective form of the gene in some people
who suffer from osteopetrosis, a rare bone
disease also characterized by thick bones. One
aspect of this study is that it may give us hints
about how to treat a much more common
disease – osteoporosis, in which bones
become too thin, particularly in elderly
women.”
What’s the connection between thick bones in
mice and thin ones in aging humans? “ClC-7 is
active in osteoclasts, cells which degrade
bones,” he explains. “You wouldn’t want that
degradation to happen in elderly people, especially
women. A drug that partially inhibited
ClC-7 might help preserve bone density. And
that could make a treatment for osteoporosis.”
��
Recently the lab has extended its attention
to a family of molecules called KCCs
that transport both chloride and potassium in
a coupled way. “We call these molecules cotransporters,
and they were first discovered in
studies of red blood cells,” Thomas says. “As
these cells are squeezed through tiny capillaries,
and have to go through cycles of oxygenation
which are associated with fluxes of salt
and water, they have to precisely regulate
their volume. At least two types of KCCs help
carry out this job in the membrane of red
blood cells.”
A channel is relatively passive – allowing ions
to pass “downstream”, for example by shipping
negatively charged particles into an area
where the charge is positive. But transporters
like KCCs are able to push chloride ions
“uphill” – by moving them at the same time as
potassium ions.
In general, Thomas says, the KCCs are mainly
responsible for the outward passage of chloride
and fluid from the cell. Seth Alper of the
Harvard Medical School, who has been collaborating
with Thomas on the KCC project,
showed that there is too much transport of
potassium and chloride out of the cell during
sickle cell anemia. In this disease, red blood
cells lose too much volume and take on a
deflated, sickled shape. The cells then become
wedged in tiny capillaries, obstructing the
flow of blood and causing irreversible damage
to organs. A mutation in sickle cell anemia
causes hemoglobin proteins to form clumps
that do not dissolve inside the cell when
things get too crowded – and that happens
when red blood cells can’t control their volume
properly.
Seth has done a great deal of work on the
relationship between hemoglobin, cell vol-
Part Three: Frontiers and ferrymen
208
ume, and the symptoms of the disease.
“Clumps form when there is a high concentration
of mutant hemoglobin in the cell,” Seth
says. “This somehow causes cells to take on
their odd, sickle shape. If we could prevent the
aggregations, we could stop their harmful
effects. That means keeping the concentration
of hemoglobin low, and having enough
water in the cell is an important part of the
process. If KCC channels are too active, the cell
loses ions and water. If that didn’t happen, we
think the course of the disease would be
greatly slowed.”
It sounded good, but proving the hypothesis
would require studying KCCs in the blood
more carefully. Thomas and his colleagues
from Hamburg developed strains of mice
lacking KCC1 and KCC3, the two types of
potassium-chloride transporters found in
red blood cells. Removing KCC1 had no obvious
effects on the mice. But taking away
KCC3, or both molecules, reduces the transport
of chloride and potassium ions out of the
cell.

Would this offer the animals protection from
sickle-cell disease? The scientists crossed the
mice with another strain which has defects in
hemoglobin and develops a condition very
much like human sicke-cell anemia. “On the
whole,” Thomas says, “the blood cells of these
animals had more volume, and there was less
ion transport. But it didn’t help the cells that
were most dehydrated and deformed. One
hypothesis is that this subpopulation of cells
lose their water a different way, earlier in the
process by which red blood differentiates.
Maybe a combination of inhibiting KCCs and
other ion transporters like potassium channels
could help.”
��
Currently Thomas is most excited about
his work with chloride transporters at
work inside the cell, including other members
of the CLC family. “These molecules are found
in membranes along the endocytotic pathway,”
Thomas says. “That’s a route used by the
cell to absorb molecules from the outside, to
capture toxins and other foreign substances,
Ioana Neagoe
and to recycle proteins to the outer membrane
or to degrade them.”
Substances are wrapped in membranes so
that they can be processed. These membrane
bubbles are absorbed by internal cellular compartments
that contain high levels of acid and
enzymes that break things down. Raising the
acid level means bringing high numbers of
protons into the compartments. Vesicles are
equipped with proton pump molecules that
do this, but they can’t do the whole job – once
the acidity reaches a certain level, the pump
can no longer push protons “upstream” into
the compartment. “The solution is to neutralize
the positive charge going in by importing
an equal negative charge,” Thomas says. “That
happens through the CLC transporter, and it
allows the vesicles to continue to increase
their acidity.
“A whole branch of the CLC family – five molecules
in humans – are found mostly in internal
membranes, and we didn’t know what
they were doing,” he says. “In 2005, we
showed that ClC-4 and ClC-5 were also acting
209 Part Three: Frontiers and ferrymen
as coupled transporters, and not as channels –
inside the cell.”
Here, too, there were links to disease: ten
years ago, Raj Thakker in London and Thomas
had found that mutations that inactivate ClC-
5 cause Dent’s disease, an inherited disorder
that causes people to develop kidney stones.
Nils Piwon, a graduate student in Thomas’ lab,
then generated a mouse without ClC-5 and
found that the animals’ kidneys were severely
impaired in their ability to take up proteins,
which then entered the urine. The scientists
developed a hypothesis to explain why this
defect in protein uptake causes kidney stones:
without the ClC-5 chloride transporter, kidney
cells are unable to absorb another small protein,
a hormone called PTH. It therefore accumulates
at too high levels in the primary
urine, triggering abnormal signals in the cell
that limit the reabsorption of phosphate, .
Through a complex series of signaling events,
this also increases vitamin D levels. This molecule,
along with PTH, is the main hormone in
regulating body calcium. Patients with Dent’s
disease take in too much calcium from food,
which the body has to get rid of through the
kidney. The combination of high calcium and
high phosphate in the urine then leads to precipitation
of calcium crystals that finally build
up to the terribly painful kidney stones.
Thomas lists other diseases linked to internal
CLCs: neurodegeneration associated with ClC-
3, lysosomal storage disease (ClC-6) and now
osteopetrosis (ClC-7). The genome encodes 60
potassium channels alone, I remember. Are all
of them linked to disease?
You can tell that Thomas isn’t finished, that
he’s somewhere in the middle of the research
roadmap that he designed 20 years ago. It’s
still a good plan and he talks about it eagerly.
Behind him the lab is filling with new methods,
postdocs, students, and technicians.
There is still an immense amount to do with
ion channels. He’s building a laboratory for
the next 20 years.

Hartmut Oschkinat
A kiss, interrupted
It’s probably dangerous to compare a pair of proteins to a sculpture,
especially one that has been called a “sensual masterpiece,” a “mixture
of idealism and eroticism” – Auguste Rodin’s kissing lovers sit on a
rock that is only half-hewn from the marble block; above, their glossy
bodies entwine to bring their lips together. But having this image in
mind might be helpful in imagining the encounters between molecules
that drive processes within cells. Proteins also kiss, sometimes in brief
encounters that last for picoseconds. They enjoy a fleeting contact that
changes one or both molecules, transmitting information, loading each
other with energy, or carrying out other types of transformations.
Sometimes there is only a small peck on the cheek, precisely and neatly
delivered. In other cases, the molecules fold and wrap around each other.
Many of the encounters in the cell are the pecking type; they take place
between a tiny partner such as a hormone and a much larger protein.
Others involve whole-body embraces between two proteins, huge surfaces
that scientists have had trouble understanding and manipulating.
Achieving this understanding is important, says Hartmut
Oschkinat of the FMP, if researchers want to find drugs that control
interactions between proteins.
211 Part Three: Frontiers and ferrymen

Peter Schmieder and Barth Rossum
The active components of drugs are often small molecules that have
been extracted from much larger substances and then purified, leaving
only the pair of lips – only what is needed to deliver a kiss or block
someone else’s attempt to do so. Often they snap into a pocket or a
groove in one of the cell’s proteins, taking the place of another molecule
that is somehow linked to the symptoms of a disease. If
researchers find such a pocket, it may not be hard to find or design
something else that fits. That doesn’t mean it will be easy to turn their
invention into a drug, for a variety of reasons. Even so, it’s easier than
blocking contacts between large surfaces.
“Such interactions are crucial in most cellular processes,” Hartmut
says. “Finding ways to inhibit them is usually very challenging. In the
first place, these surfaces are large and flat, and it’s hard for a small
Part Three: Frontiers and ferrymen
molecule to get a grip – the way that you can’t free-climb on a surface
that’s too smooth. And there are very few natural small molecules that
bind to these surfaces, which would also give you a place to start in
designing new drugs.”
That’s not to say that it can’t be done; some existing drugs work on
larger surfaces. Taxol, which is widely used to fight lung and breast
cancer, was isolated from the bark of Pacific yew trees growing near
the Mount St Helens volcano in Washington. It latches onto a protein
called tubulin and prevents this molecule from binding to its partners,
thus triggering a cellular self-destruct program called apoptosis.
Another drug, cyclosporine A, was derived from a fungus that thrives
in the soil in Norway. It has been widely used for over 20 years as an
immune system suppressor to reduce inflammations and prevent
212

organ rejections. It works by blocking interactions between proteins in
T cells, reducing their ability to properly stimulate an immune
response.
“These examples show that in principle, protein-protein interactions
can be inhibited, and the molecules that do so might make powerful
drugs,” Hartmut says. “But we lack a lot of basic knowledge about the
interactions. I thought that if we could find the right pair of proteins,
it might give us a good model system to use to get a handle on the
problem.”
��
The publication of the human DNA sequence and other complete
genomes has given scientists a full catalogue of proteins and
other molecules that can be produced by our cells. A lot of them are
cousins; long ago they started off as one gene in one of our ancestors.
Then duplicates arose through errors that were made as DNA was
copied. Some of them were preserved through evolution, gradually
undergoing mutations and taking on different functions. One structure,
called a PDZ domain, has now been found in 250 proteins.
“This is one of the most important modules in cells that govern interactions
between proteins,” Hartmut says. “It helps organize networks
of signaling proteins; it participates in the delivery of parts to ‘molecular
machines,’ and one of its most important jobs is to control events
at the cell membrane. We began looking at this module because of its
importance, and also because of a feature found on its surface. PDZ
domains contain a shallow groove – not a real pocket, but something
that an inhibitor might get a grip on.”
Mangesh Joshi, a PhD student in the group, took on the project and
focused on a protein called AF6, which contains a PDZ module. One
reason for the choice was the role of AF6 in the development of a disease
called chronic myeloid leukemia. This type of cancer arises
through a flaw in another protein, called BCR, which binds to the PDZ
domain in AF6. Recently Yunyu Shi’s laboratory at the University of
Science and Technology in Hefei, China, discovered that BCR docks
onto just the surface groove that Mangesh intended to look at.
Each PDZ domain docks onto a specific module in one other protein.
Ten years ago, Hartmut’s group helped figure out how PDZs recognized
the right targets: they latch onto a loose, tail region of a domain
in the partner molecule. Because the amino acid recipe of each tail is
slightly different, it has different binding properties. Evolution has tailored
a specific tail to fit the groove of a particular PDZ.
“Because of what we know about how these molecules bind, it
seemed likely that plugging the groove with another molecule might
block access of the normal partner,” Mangesh says. “But to show that
to be the case, we had to find an inhibitor.”
��
Since 2002 the FMP has been creating a “library” of molecules –
ranging from natural substances to small artificial peptides – that
can be used in projects like the one planned by Hartmut and Mangesh.
The library is housed in a high-tech screening facility in the newly constructed
Genomics center, a sleek, black building on the east edge of
the campus. The building was opened in 2006 by the FMP and their
campus neighbor, the Max-Delbrück Center for Molecular Medicine.
Jens Peter von Kries heads the facility. His office has large windows
that look over the campus, which he knows well; from 1995-2000 he
was a scientist in Walter Birchmeier’s group at the MDC. He spent the
next two years of his career as the scientific head of the screening unit
213 Part Three: Frontiers and ferrymen

for a Berlin pharmaceutical company before
being recruited by the FMP to set up a similar
facility. Today his foot is in a temporary cast
from a football accident that happened while
he played with his son. He places it on his
desk like a paperweight as he talks about the
mission of the facility.
“Our aim is to help scientists search for
inhibitors for specific molecules and processes,”
Jens says. “The point may be to find a lead
compound that can be developed into a drug
– or the goal may be to develop a new tool to
study a cellular process.”
Sometimes, as in the case of Mangesh’s project,
the role of the facility is simply to open its
library for scientists to use. Mangesh had
decided to carry out the screen itself in
another part of the institute, using the FMP’s
NMR machines. Hartmut’s group has helped
turn NMR into the institute’s major tool used
in studying protein structures, and it also has
other applications such as screening.
“There were two reasons to do the screen
with NMR,” Mangesh says. “First, because of
the nature of the PDZ binding region, we
anticipated that our first ‘hits’ might bind
quite weakly. NMR is good at detecting weak
interactions. Secondly, it would give us a
detailed look at the structures of the molecules
as they interacted. That would be
important for the second stage of the project:
taking a substance and rebuilding it so
that it would bind better.”
At other times the screening facility takes a
much more active role in projects, offering a
full range of screening services and assistance
in chemistry for in-house researchers
and visitors. Compounds from the library are
incubated with samples, either purified proteins
or live cells, and then examined to see if
there have been effects. All of this work used
to be done by hand, but the entire process
has been automated. The day of our tour,
final tests were being run on a new automated
screening machine, so massive that it
takes up half of a lab. Packed inside are robots

that handle samples in a “hotel”, driving their containers to a central
corridor where they encounter substances from a large library of
chemical compounds.
The FMP’s library currently includes about 50,000 substances, mostly
purchased from other collections. “Some major pharmaceutical companies
have collections of millions of substances,” Jens says.
“Obviously the FMP couldn’t afford to build a library of that size, so we
had to be selective. That was an intensive process – making decisions
on the best representatives of different classes of molecules, so that
we would have a good spectrum of substances to work with.”
FMP Director Walter Rosenthal says the screening unit is necessary to
help the institute and the wider research community carry out better
research as well as fill in some of the gaps in the drug discovery
process. “The point is to take a tool that is commonly used in the pharmaceutical
industry and give academic scientists access to it,” he says.
“We’re not trying to compete with pharmaceutical companies – they
are much better equipped to ‘try everything’ in the search for new
drugs. The point is that academic researchers need access to a screening
platform, the way they have access to other mass technologies.
Here we’ve built a platform – unique to Germany and Europe – that
they can use; we’re making it as open as possible. Your everyday scientist
hasn’t had that kind of access because the price of having a screen
done may be prohibitive, or a major pharmaceutical company may
simply not be interested in a particular project. Academic researchers
may need an inhibitor simply to study a process that they’re interested
in, which is a much too basic problem for most companies.”
If a screen turns up a hit, researchers can turn to the FMP’s chemical
groups for help in improving it. New groups such as that of Jörg
Rademann have been brought in to strengthen its chemistry program.
“The farther we make it along the process of trying, testing, and
improving an inhibitor that works in a disease process,” Jens says, “the
more likely it is that a company will take interest and move it into
development and clinical trials.
There have already been some interesting success stories. One project
of the unit involves Mycobacterium tuberculosis (Mtb), the organism
that causes the dreaded lung disease, and is a good example of how
modern molecular science is being applied to infectious diseases. The
incidence of tuberculosis is on the rise – epidemiologists estimate that
one-third of the world’s population is infected, although only a small
percentage show symptoms – and about 20 percent of the bacteria
have developed resistance to the most common anti-tuberculosis
drugs. So national and international research campaigns have
branched out, searching for new ways to fight the disease.
Tuberculosis projects across the world have been looking for good
drug targets among the molecules in the tuberculosis bacterium. The
ideal target would be a protein with unique characteristics, which
might allow it to be blocked without interfering with one of the
healthy proteins in the cell. Recently Jens and Larissa Podust, now at
the University of California, San Francisco, carried out a screen to find
inhibitors for a protein called CYP51. This molecule is found in all kinds
of cells, ranging from bacteria to humans, where one of its jobs is to
synthesize cholesterol. That is vital to creating membranes, and without
it the bacteria cannot survive. The version of CYP51 found in bacteria
is different than the human form, with unique features that might
make a good drug target.
“Another advantage is that when CYP51 binds to another molecule, it
undergoes a change,” Jens says. “You can observe it with a spectrometer,
so it gave us a good method to observe the results of the screen.”
215 Part Three: Frontiers and ferrymen

A region of the AF6 protein called the PDZ domain changes when a small
binding partner attaches itself. This suggests that the site can be used as
a target for the development of small molecules or substances that
inhibit the activity of the many proteins that contain PDZ domains.
Part Three: Frontiers and ferrymen
High-throughput experiments turned up three substances that
bound to CYP51, including a known antifungal drug called EPBA. It
locked onto the bacterial protein strongly and very specifically – not
disturbing other molecules. EPBA is also used widely as a preservative
in cosmetics, which means that its effects on humans have been carefully
studied. If EPBA does turn out to be effective against tuberculosis,
that will save time when it has to be tested on patients.
The scientists discovered that EPBA strongly resembled the structure
of another molecule, called BSPPA, which they also tested; the second
compound behaved exactly the same way.
“Larissa Podust went to the synchrotron at the Argonne National
Laboratory in Illinois, in the US,” Jens says. “The purpose of the trip was
to use X-rays to get detailed structural pictures of how both of these
molecules dock onto CYP51. The images show us how the compounds
might be adapted to target other types of CYP51s. We’re hopeful that
EPBA can be used as a structural scaffold to spin off specific new substances
to do that. The fact that the molecule has already been
through clinical trials in another context – as an antifungal drug –
could dramatically speed up its development into an anti-tuberculosis
therapy.”
The collaboration with a lab in the United States shows that the facility
is ready to take on long-distance partnerships in interesting projects.
Jens is already working closely with a network in Scandinavia
called ChemBioNet, and other collaborations are underway with
Oxford University.
One ongoing project involves a group at the University of Oslo,
Norway, that works on specialized types of stem cells. Since the screen
involves live cells, they had to be shipped to Berlin from Scandinavia.
It’s not as far as Illinois, but in this case, distance proved to be a
problem.
216

“Jo Waaler tried to send the cells three times using express shipping
services,” Jens says. “Each time they got hung up somewhere at the
border, and the cells died. Finally Jo gave up, packed the cells in a rental
car, and drove them down himself. All the way from Oslo.”
This time the cells – and the project – survived.
��
When Mangesh began his screen on the PDZ module of AF6, no
synthetic binding partners were known. He began working
with a subset of the FMP’s collection, 5,000 small molecules particularly
suited for screening with NMR. This included representatives of
all the basic kinds of molecules in the collection, including the types
most frequently found in drugs.
“There are three critical stages for the design of inhibitors in drug discovery,”
Mangesh says. “First you try to find lead compounds – these
are substances that show moderate activity toward the target. Then
you go to the test tube and try to optimize the compound. Here is
where it pays off to know how the molecule binds to its target, to have
this structural pictures. That points out regions that are crucial to the
interaction. Finally, the system has to be put to work in cells and organisms.”
The screen turned up three different types of compounds that could
bind to the PDZ domain of AF6. The scientists chose the one with the
strongest effect, 2-thioxo-4-thazolidinone, for further work. Carolyn
Vargas, another student in Hartmut’s group, began systematically synthesizing
and purifying new spin-off compounds from it. To do so she
had help from Volker Hagen, head of the Synthetic Organic Chemistry
group at the FMP.
The structural pictures from the NMR screen revealed specific points
that might help strengthen the bonds between the compound and
AF6. After some work, Carolyn and Volker had come up with a new sub-
stance with the awe-inspiring name (2R,5R)-2-sulfanyl-5-[4-(trifluoromethyl)benzyl]-1,3-thiazol-4-one.
It’s not the kind of thing you would
want to have to say in the lab every day, or spell for someone on the
telephone, so the compound was nicknamed 7i. Further NMR experiments
showed that 7i was about as good at binding to the PDZ
domain as AF6’s natural partners in the cell.
“Although the AF6 PDZ itself is highly similar to other PDZ domains,”
Carolyn says, “the way that it bound to the new compound was surprising.
Somehow 7i had docked itself into a subpocket of the PDZ
domain that had never been seen before, in any structural picture of a
PDZ. Maybe the pocket wasn’t even there before. We think this means
that as 7i and the PDZ come into contact, the domain undergoes
rearrangements that create the new pocket.”
You’ll never see a statue do that – not even one as sensual as Rodin’s
carving of a kiss – but proteins are anything but static objects. Their
highly dynamic nature sometimes makes it hard to design a new drug,
or predict how a substance will affect a protein. But if AF6 really
undergoes a structural change, it might simplify things.
“When we originally thought about finding a way to make drugs to
disturb interactions between proteins,” Hartmut says, “we chose the
PDZ domain because of its shallow groove. Now it appears that a class
of molecules like 7i can create new pockets that allow them to get a
better grip. This makes PDZs even more promising candidates for
manipulations by drugs – especially if it turns out that these structural
rearrangements are a general characteristic of PDZ domains, that
they also happen in modules of other proteins. That’s likely to be the
case; if so, it may give us a new strategy to design drugs that can influence
a huge range of important cellular molecules.”
217 Part Three: Frontiers and ferrymen

Friedrich Luft

Interlude:
The case of the
short-fingered musketeer
This story could easily be made into a book of its own. It has all the
elements of a good adventure: a scientific mystery that dates
back three decades and then some, travels to remote regions on four
continents, a strange mutation that causes disease in humans, an
attack by wild boars, and some gunfire. And a physician became a
guinea pig on his own operating table. There was almost a jailbreak,
but it was made unnecessary by a last-minute reprieve.
A word of warning from the outset: most of the stories in this book
have a conclusion, but this particular case has not yet been solved. A
solution could be just around the corner thanks to the recent surpris-
ing discovery of a new type of code written into the human genome.
On the other hand, the new code may be a false lead, and the real cul-
prit may be a biological process that has not yet been identified. In the
end the case may stay open in the files, waiting for a new generation to
come along with a fresh perspective and new technology.
��
219 Interlude: The case of the short-fingered musketeer

Franz Volhard
Nearly every day until the winter of 2007
you could see Friedrich Luft on the
Lindenberger Weg, the main road that connects
the scientific campus to the former hospital
complex that included the clinics of the
Charité on the other side of Berlin-Buch. He
was hard to miss – you looked for the only person
with salt-and-pepper hair, dressed in
white doctor’s garb, riding a bicycle. In 2007
he still had an office on both sides of town.
The one on the MDC campus is at the top of
the Helmholtz house, with large windows
overlooking the campus. He always takes the
stairs. If he sees students waiting for the elevator,
he’ll tease them: “What, are the stairs
broken again?”
Up on the fourth floor he’s planning the new
Experimental and Clinical Research Center
(ECRC), a project which has taken up a lot of
his time over the past few years. So far it’s a
virtual center, of which he is director. If things
go as planned, the ECRC will soon take on
Interlude: The case of the short-fingered musketeer
physical form as a new building on campus.
More on that topic later.
For our first meeting Friedrich invited me to
his office on the other side of Berlin-Buch, the
hospital grounds known as Area 1, before the
the Franz Volhard Clinic made its move to the
Helios complex. His role there was head clinician,
devoted to nephrology (the study and
treatment of kidney diseases) and cardiovascular
disease that arises from hypertension.
As I walked over from the bus stop, he zoomed
by. “Get a bicycle,” he yelled.
The corridor to Friedrich’s office in the clinic
bore portraits of many of the personalities
who had worked there, and one who was
present solely in spirit: Franz Volhard as a
young man, holding a violin. The photograph
was taken in 1893 while Volhard was a student
at the University of Halle. He maintained
a life-long love of music; throughout his
career, as he moved from clinic to clinic, he
encouraged his staff to brush up their musical
220
skills and perform together. The tradition was
carried on in the clinic that bore his name;
until the day of the move, the Volhard Clinic
had a Bechstein piano in the lobby.
“Franz Volhard was the preeminent German
cardiovascular clinician of the early 20th century,”
Friedrich says. “His many contributions
concerned the kidney and the heart, oftentimes
both. Nephrologists and cardiologists
can equally lay claim to him as a role model
and therefore the Franz Volhard Clinic was
named after him.”
Early in his career Volhard became interested
in the fact that kidney diseases were almost
always accompanied by hypertension – chronically
elevated blood pressure. He was particularly
intrigued by a condition called Bright’s
disease, named after Richard Bright, the
physician who discovered chronic kidney disease
at Guys Hospital in London in 1827. The
illness was puzzling because it affected several
bodily systems: the kidneys became dysfunctional
and shrank; blood pressure
increased; the heart became enlarged;
patients suffered brain damage and inflammations
in the retinas of their eyes. Volhard
believed that all of the symptoms might be
due to an more basic problem related to
hypertension, and he set off on a search for
the causes of high blood pressure. While he
didn’t succeed, Volhard and his students managed
to convince the medical world that
hypertension was a systemic problem linking
some types of kidney and cardiac disease.
Researchers at the Volhard Clinic – including
Friedrich Luft – continue to pursue the underlying
causes. Friedrich is particularly interested
in essential hypertension, and I ask him
what this means.
“Essential hypertension is what everybody
has – it occurs in 25 percent of the entire population.
The name comes from a hypothesis of
fifty years ago. At that time, it was believed
that if blood vessels became constricted
somewhere in the body, likely the kidneys, it
would be ‘essential’ to raise pressure so that

lood could still be squirted through the
arteries. It was a natural conclusion. But after
all these years of research, we still haven’t
pinned down the true causes of the kind of
high blood pressure that affects individual
patients and the general population. Multiple
factors are probably responsible.”
In time, visionaries such as Franz Volhard recognized
that hypertension was not “essential”
– but rather a major health menace. The
name, however, stuck.
A few forms of hypertension are simple inherited
conditions that run through families, due
to defects in individual genes that are passed
from parent to child. This type of hypertension
is inherited according to the patterns
first described by Gregor Mendel in his studies
of peas. In these families, a single genetic
mechanism is responsible for a trait – in this
case high blood pressure. Studying those conditions
and making models of them in mice
have given insights into each particular dis-
ease. But most of these situations are different
from essential hypertension, where many
genes – rather than just one – are likely to be
responsible.
“The genes in these cases are involved in the
reclamation or reabsorption of sodium and
chloride in the kidneys,” Friedrich says.
“They’re all related to people’s intake of salt –
you see some pretty drastic changes in blood
pressure among these patients, depending on
the salt in their diet. However, these mechanisms
don’t seem be responsible for essential
hypertension. The majority of people who suffer
from it do experience an increase in blood
pressure if they eat more salt, but it’s not
much.”
��
William Bateson, a classical geneticist of
the early 20th century, advised his disciples
to “Treasure your exceptions.” Bateson
was referring to unusual discoveries in plants,
221 Interlude: The case of the short-fingered musketeer
but the same is true of exceptional families
harboring mutations that lead to disease. If
“classical” disease hypertension genes didn’t
say much about blood pressure in the general
population, Friedrich thought there might be
other rare Mendelian genetic syndromes that
did. “When I first arrived here in Berlin-Buch, I
was supposed to be working on genetics and
cardiovascular disease – but I could barely
spell genetics. Thomas Wienker, a real geneticist
who knew all about William Bateson,
showed me an interesting paper that had
been published in 1973 in the Journal of
Medical Genetics. The topic was an extended
family living in Turkey, in the region around
the Eastern Black Sea coast. Members of the
family suffered from an unusual genetic condition
with two main phenotypes – brachydactyly,
or shortened fingers and toes and
overall stature, and severe hypertension. If the
hypertension went untreated, the affected
family members all died of strokes before they
reached the age of 50. What particularly inter-

The region of Turkey where Nihat Bilganturan found an extended family with
brachydachtyly and hypertension.
ested us was the fact that their hypertension
wasn’t particularly dependent on salt. This
might mean that the mechanisms that
caused it were closer to essential hypertension
than other known genetic conditions. So
we thought it would be worthwhile to identify
the gene that caused it.”
One of Friedrich’s students, Hakan Toka, had
some background in medical genetics. He also
came from a Turkish family who had settled in
Germany a few decades ago, when the father
had been hired to work in a BMW plant.
Hakan had been born in Germany but spoke
Turkish. He volunteered to call his uncle, a
general in the Turkish army, to see if the
researcher who conducted the original study
was still alive. “Nihat Bilginturan was 68 years
old, and had retired from his position as professor
of pediatrics and endocrinology at the
University of Ankara,” Friedrich says. “He was
still living in Ankara, so several of us flew over
to see him. We met in a café. He was a pleasant,
rather formal fellow. Initially he was hesitant
about participating – afraid the people
would be commercialized or compromised
somehow. But we managed to convince him
that our motives were pure.”
Bilginturan introduced the clinician-scientists
to the families. They were spread among
three clans that didn’t get along with each
other, but the researchers managed to convince
them to undergo some initial tests.
Friedrich and his team stayed for a week conducting
interviews, carrying out over 50 physical
exams, measuring blood pressure, drawing
Interlude: The case of the short-fingered musketeer
blood to obtain DNA samples, and photographing
fingers. By the end of the stay the
family had agreed to participate in a yearlong
clinical study of anti-hypertensive drugs. There
were five German participants in the first
Turkish adventure. Hakan Toka and Okan Toka,
brothers and medical students who were fluent
in Turkish, Thomas Wienker, human geneticist
from the MDC, Herbert Schuster, internist
and medical geneticist from the Franz Volhard
Clinic and an associate of Friedrich’s. Not to
forget Friedrich himself, who was responsible
for anything that went wrong.
“We had no funding for any of this – so far
everything was being funded by my VISA
card,” Friedrich says. “Then I heard about a
NATO grant program from a captain I knew in
the air force – I had also been in the U.S. military.
They gave us 20,000 dollars, which was
used to pay the VISA card debt, and that’s how
the project started.”
��
Back in Germany, the group faced two
challenges. The first was to conduct additional
phenotyping, namely to get at the clinical
mechanism underlying the patients’
hypertension. The second was to find out
where the gene might be located. So a few
months after the visit, several members of the
family with the disease – and a few without –
were flown to Germany. They were checked
into the Franz Volhard Clinic, where they
underwent a week of extensive clinical testing.
At the time the process was unusual. “The
222
Turkish patients were obviously not insured
and could lay no claim to any health care in
Germany,” Friedrich says. “So who paid for
these studies and the patients’ time in the
hospital? That issue, which was addressed in
the U.S. more than 50 years ago, is one that
has occupied quite a bit of my time over the
past 15 years.”
Numerous tests were necessary to find out
whether the family’s problems resembled
essential hypertension. “What you do is measure
how the body responds to certain things
that are normally signals for an elevation or
drop in blood pressure,” Friedrich says. “If the
volume of the blood increases – if the system
is carrying more liquid – then pressure needs
to drop, and this is managed by changes in
the levels of certain hormones. If there’s more
salt in the diet, it tells the system to absorb
more water. We also checked other typical
kinds of problems that accompany hypertension
– we checked arteries in the kidneys and
eyes to make sure they were okay, and looked
at levels of things like calcium and lipids in
the diet. What we found was pretty normal,
which meant that this condition had a lot in
common with essential hypertension.”
The major challenge was to find the culprit
gene. The task fell mainly to Herbert Schuster
and Silvia Bähring. Herbert was a clinician-scientist
that Friedrich had recruited to assist in
genetic research; Silvia had a PhD in molecular
biology and substantial technical experience.
The two began a linkage analysis, a
method to identify the region of DNA that

carried a particular gene responsible for a disease.
Linkage is the road to answering the
question “Where is it?” – namely, what part of
the genome is responsible for a trait that scientists
want to study. Thomas Wienker was in
charge of the data analysis, making mathematical
models of the data to prove the location
of the responsible gene.
The strategy goes back to the early days of the
20th century, when the great geneticist
Thomas Hunt Morgan launched a new science
now known as “classical genetics.”
Morgan traced the locations of hundreds of
genes in the four chromosomes of the fruit fly
Drosophila melanogaster – in a day when no
one knew what genes were and there were no
molecular techniques to investigate them. His
students spread throughout the world, starting
new laboratories and leaving a trail of
Nobel prizes in their wakes. One member of
his laboratory, Herman Muller, spent some
time in Berlin-Buch. Muller, a Nobel prize winner
from Indiana University in Bloomington,
was a mentor to Nikolai Wladimirowitsch
Timofejew-Ressowski, who in turn worked
with Max Delbrück to put Berlin-Buch on the
scientific map. But that is another story.
Friedrich is undoubtedly one of the most
widely read people on campus, and it shows
when I mention Morgan’s name. “His grandfather
was a famous horse thief and confederate
general named John Morgan from
Kentucky. If you’re interested in the history of
the U.S. Civil War, he shows up as a bandit – in
the South he was called a general.” This
Bilganturan and
Thomas Wenker
discussing the
Turkish pedigree
Thomas Morgan
vignette isn’t the only colorful aspect of the
lineage of Morgan, who was also a direct
descendant of governors, diplomats, and
Francis Scott Key, the composer of the U.S.
national anthem.
Morgan started his career at a pivotal
moment in the history of science. Three
researchers in different European countries,
working independently, were about to rediscover
the principles of heredity that had been
found by the monk Johann Gregor Mendel
and then lost for 40 years. One of the rediscoverers
was Hugo de Vries, who coined the
term “gene”. In 1900 Thomas Morgan dropped
in on de Vries in Amsterdam and wrote, “No
one can see his experimental garden, as I have
had the opportunity of doing, without being
greatly impressed.”
The visit was the beginning of a life-long
friendship between the two men and for
Morgan, an introduction to genetics. His initial
interest was the connection between
heredity (concerned with how existing genes
were shuffled about and combined to give
organisms their physical features) and evolution
(which focused on the origins of new
223 Interlude: The case of the short-fingered musketeer
species). For the first decades of the 20th century,
a fierce debate raged about whether or
not the two ways of thinking were at all compatible.
Genetics could explain two of the
basic concepts of evolution – why individuals
within a species were different from each
other, and how they passed down characteristics
to their offspring. But evolution obviously
involved more than simply shuffling around
existing genes; new traits had to come from
somewhere.
De Vries came up with a proposal: organisms’
genes might undergo changes in a process
called mutation. He hoped that his studies of
plants might catch evolution in the act – if he
watched enough pea plants give rise to seeds
that were round or wrinkled, green or yellow,
a puffy purple one might suddenly appear.
Back in the U.S., Morgan decided to try the
same thing with animals. Finding mutations
would require a lot of animals and a lot of
generations, so he needed an organism that
reproduced quickly, had a lot of offspring, and
was easy to care for (his students would need
to learn to handle them). A colleague gave
him some fruit flies, which could be kept alive
on a diet of mashed bananas, and the stage

was set for over three decades of pioneering
research. Within two years Morgan began
finding mutations in flies. Over the course of
his career he found hundreds. He didn’t learn
much about evolution – his flies didn’t suddenly
develop into a new species – but he discovered
a great deal about the nature of
genes.
Mendel’s laws stated that genes were inherited
independently of each other, but Morgan
and his colleagues quickly began to find
exceptions. Genes on different chromosomes
behaved independently, but two genes on the
same chromosome were usually inherited
together. Even that principle had its exceptions,
which were revealed by statistical studies
of tens of thousands of flies over generations.
Within the same chromosome, genes
that were near each other were inherited
together more often than those farther apart.
The reason is that pieces of chromosomes
occasionally break as they are copied. The
pieces can be glued back into the chromosome,
but sometimes they are reinserted in
the wrong one. This separates genes that
were formerly on the same chromosome.
Genes that are close together are more likely
to be in the same fragment when it moves.
This concept, called linkage, allowed Calvin
Sturtevant and other members of Morgan’s
lab to use simple math and statistics to map
the locations of genes on chromosomes.
Today the same general strategy allows scientists
to discover genes responsible for human
diseases. They no longer do linkage studies
Interlude: The case of the short-fingered musketeer
quite the way Morgan did – which would
require huge, long-term studies with an enormous
number of participants. In the 1980s a
British researcher named Alec Jeffries discovered
a shortcut. He found small regions of
DNA called microsatellites, identical units
spread throughout the genome of every
human. Satellite DNA consists of “repeat”
sequences which – like at least 97 percent of
the genome – do not encode proteins. The
function of most of this non-coding DNA is
unknown, leading some to call it “junk” (a designation
which is “dangerous and premature,”
Friedrich says).
Microsatellites are very small pieces of satellite
DNA. They do not encode genes but are
inherited in much the same way that genes
are. Their locations is monotonously similar
from person and family to family. Within
months of Jeffries’ discovery, a huge number
of uses had been discovered for microsatellites,
including “DNA fingerprinting,” which
can match DNA from a crime scene to a suspect,
and paternity testing.
��
Everyone on Earth (as well as anyone currently
on the space station, Friedrich
quips) carries the same microsatellites at the
same locations in their genomes. They are
made of long strings of repeated sequences,
spelled from the four letters of the genetic
code (C, A, T, and G). A common microsatellite
repeat is CACACACACACA... (it goes on like
that for a while). What makes microsatellites
224
useful, Friedrich says, is that they come in different
lengths. Over time, some microsatellites
collect extra repeats through random
errors that happen as DNA is copied. They are
then passed along from parent to child, and
can be followed like footprints down a family
tree, over time.
Each person inherits a double copy of the
“DNA library” – 23 paired chromosomes, one
from each parent – and thus two copies of
each microsatellite. So there is one version of
a particular microsatellite on the “short arm”
of chromosome 12, and another version of the
same microsatellite (inherited from the other
parent) at the same location on the second
copy of chromosome 12.
“The idea is to identify specific versions of
microsatellites that are always inherited with
a particular disease, and never in non-affected
family members,” Friedrich says. “When we
find such sequences and know precisely
where they are located, we can conclude that
the disease gene must be very close to the
one or more microsatellites that have turned
up.”
Herbert and Silvia were going to use
microsatellites to home in the location of the
disease that caused the Turkish family’s
hypertension and brachydactyly. “We were
beginners at this,” Friedrich says. “And it was
expensive – at the time you needed radioactively
labeled microsatellite markers. Who
was going to pay for a bunch of beginners to
do a study like this? Our bosses were pes-

simistic and suggested we should turn over
the project to “experts”. I didn’t see things
that way – we were good enough to do the
dirty work, gathering the samples and working
with the patients, and now we were going
to let professionals get the glory? We said, ‘To
hell with them; we are doing this ourselves.’
Smart money was bet against us but the jury
was still out.”
To the rescue came Applied Biosystems Inc.
(ABI), a biotech company based in California.
“ABI had just developed non-radioactive fluorescent
microsatellite markers that covered
the entire genome,” Friedrich says. “‘We’ll give
you the markers and show you how to use
them if you acknowledge that effort in your
paper,’ they said, and our response was, ‘Done
deal.’”
Herbert traveled to ABI headquarters in
California and began working up through the
chromosomes, starting with chromosome 1.
Silvia stayed in Berlin and started with chromosome
22, working her way backwards. “It
would have been nice if the gene locus had
been at one of the ends; as things turned out,
it lay on chromosome 12, almost exactly in the
middle,” Friedrich says. “So there were no
shortcuts. Thomas Wienker shouted “Eureka”
somewhere on the short arm of chromosome
12. Even so, the whole process was accomplished
in six months – not bad for a bunch of
beginners.”
��
The original team that flew to Turkey.
Ithink of a trip my family took to the great
San Andreas Fault in California, source of
the terrible earthquakes that regularly rock
the coast, not far from ABI headquarters. The
fault passes through a park with a hiking trail.
You don’t see a great crack in the earth, but at
one point there is a picket fence that is no
longer straight – the earthquake of 1906 shifted
half of it five or six meters. If you were to
study the damage done to fences, houses, and
other landmarks, you could trace the line of
the fault on the map.
The Turkish family has undergone a sort of
genomic earthquake; Silvia and Herbert’s
study pinpointed its site of origin: a specific
region on the “short arm” of the twelfth
human chromosome. The next step was to
study the area and search for a potential gene
which had undergone mutations. That was
much easier said than done. No genes were
known to exist there. And the Human
Genome Project was not yet completed,
which meant that Silvia had to try to find a
gene the old-fashioned way, through a laborious
process called positional cloning. This
required selecting pieces of human DNA
packed in yeast or bacteria and sorting them
along the linkage interval. It allowed her to
identify a number of genes within the region
and sequenced them, but none had mutations.
The genes were identical in family
members affected by the disease and those
without it.
This outcome isn’t so unusual; a large number
of human proteins and the genes that encode
225 Interlude: The case of the short-fingered musketeer
them remain to be found. “It wasn’t necessarily
a dead end,” Friedrich says. “Now that the
genome has been completed, you can scan
regions like this by computer and look for bits
of code that are hallmarks of genes. Well, we
found sequences that had some of these
characteristics, but other features were missing.
So we were stuck.”
Members of the lab were keeping their eyes
open for other reports in the medical literature
that might shed light on the case. Silvia
found a report on a Japanese child with
brachydactyly whose condition had also been
pinned down to a region in the twelfth chromosome.
The Japanese child had a deletion
syndrome – namely it was missing an entire
portion of the chromosome 12 short arm.
Silvia asked the lead researcher for samples of
the child’s DNA and an X-ray picture of the
child’s hands.
There are many types of short fingers, probably
caused by defects in different parts of the
genome. The Turkish family had a particular
form termed “Type E brachydactyly.” Infor -
mation from the Japanese child would only be
helpful if it had the same type.
The scientists were ecstatic to find that the
fingers matched. Now Silvia used the same
microsatellites to study the child’s DNA. She
discovered that the abnormalities weren’t identical
to those of the Turkish family, but the portion
of the child’s genome that had been lost
partially overlapped with the linkage region
that had been identified from the family.

“It seemed very unlikely to us that there
would be two separate genes so close to each
other that caused brachydactyly,” Friedrich
says. “That assumption allowed us to narrow
down the region where the responsible gene
must lie by half.” It was a big step, but it didn’t
solve the major problem – there was still no
gene.
��
The next stage of the study would require
someone to go to Turkey for an extended
period, and the team found a candidate –
Hakan Toka’s younger brother Okan, a medical
student at the University of Munich, one of
the members of the team during the first visit
to Turkey. Okan needed a thesis topic and was
excited at the prospect of doing a clinical
study on hypertension.
“When you do this type of thing, you also have
a responsibility to take care of the patients,”
Friedrich says. “Okan was going to have a big
job on his hands – alongside taking regular
blood pressure measurements of 50 people
scattered across a large region, prescribing
medications, and managing the rest of the
clinical part of the study, he was responsible
for their basic health care and played social
worker. We got him an apartment but he
Interlude: The case of the short-fingered musketeer
needed a car. You can’t write a grant for a car,
but we received a little money from a drug
company to help us do the study. We bought
him a brand new car for 2,000 dollars – at the
end of the study we gave it to the family. We
talked to some local physicians who promised
to keep an eye on Okan, people he could call if
he needed any help. And then we left him
there. We talked regularly by cell phone, but I
don’t recall any panic calls from him. He managed
to solve his problems by himself.
For Okan, the story began in 1994. “I was a
young third-year medical student,” he says. “I
was looking for a challenging doctoral thesis
when Professor Luft told me he was looking
for someone to spend about a year in the
mountain villages of the Black Sea Cost in
Turkey to perform a quite sophisticated crossover
drug study with first-line-antihypertensive
drugs. I would get all the equipment and
resources needed. I remember he stressed
that the job would require a lot of dedication
because it wouldn’t be ‘routine research.’ I had
to do the clinical research but would also have
to serve as the family’s primary care physician
and social worker, too.”
The villages of the eastern coast of the Black
Sea are perched on the slopes of lush, forested
mountains that begin at the water’s edge.
226
Okan says that most have a poor infrastructure.
“There is no steady supply of warm
water, heat or power. People there are usually
farmers and live from plantations where they
grow black tea and hazelnuts. I had to visit
each of my patients every two weeks to measure
blood pressure over 24 hours, to take
blood and urine samples, and to try to detect
the effects – possibly adverse effects – of the
new medications that were being tried in the
study. Most of the patients lived in this coastal
region. However, others were scattered all
over Turkey, in Istanbul, Bursa, and Ankara. It
took me about a week to see all the patients
in the mountain villages, driving off-road from
one to the next in my vehicle. Afterwards I
usually flew to Istanbul, then on to Bursa and
ended my rounds in Ankara, before flying back
to the coast. One round took me about 12
days. Then I had two days to do all the analyses
before starting the next round. I flew
about 60 times and drove 30,000 kilometers
– mostly off-road – during that year.”
Sometimes it was a rough trip. One visit to a
patient required a drive into the mountains,
taking Okan to an altitude of about 1,500
meters. He finished late and the family told
him to stay overnight – descent in the dark
would be dangerous due to snow-covered
roads and wild boars. But he decided to leave
anyway. Halfway down, after midnight, in the
middle of nowhere, his car had a flat tire.
“I had to change the tire while one of the family
members who had escorted me held a flashlight
in one hand and a gun in the other. Yes, we
got some visitors – a pack of wild pigs. My escort
was afraid of an attack. I was mostly concerned
about the precious blood sample that I had just
obtained. He fired a shot and scared the pigs
away, and I finally got the tire changed and we
could continue on down the mountain.”
Okan’s duties often extended beyond medical
care. All hands were needed – including some
of his patients – when harvest time came
around. “I didn’t think it was a good idea to
send somebody out to do hard field work

while I was recording 24 hours of blood pressure
values to evaluate the benefits of a new
antihypertensive drug,” he says. What was a
doctor to do? He spent most of the month of
August harvesting hazelnuts, taking his
patients’ shifts in the field.
��
The second gunshot of the story had more
serious consequences.
“A lot of funny but also awkward things happened
during this time. I remember one particular
incident where one older member of
the family was put in jail. He was the head of
the clan we were investigating. He is a very
friendly, sweet old man; he is about 5 feet tall
and he wears huge glasses. The problem
started when I got an excited call from his
son. The father had gotten into some trouble
with his neighbors and when things got
threatening, he fired his gun in the air to scare
them away. They ran off, all right, but then
they called the police. That wouldn’t have
been a problem in itself. He wasn’t afraid of
the police because he had the right to defend
his land and scare them away. The problem
was the gun. He had fired an old musket he
had inherited from his grandfather and he
didn’t have a license for it. When he saw the
police coming he looked for a place to hide it
and in the heat of the moment, he simply
threw it out the window. It didn’t take them
long to find it, and when they did, they arrested
him.
“The son wanted me, as the family doctor, to
help his father out of prison – and that’s what
I tried. I visited the scared old man. He told me
the whole story all over again. I asked the
police what the accusations were and what
sort of punishment he should expect. The officer
told me that things looked bad – in the
Black Sea Coast area, using an unlicensed gun
was a serious offence. Judges usually imposed
harsh penalties. In any case, the old man
would probably be in jail for two or three
months until the trial. The only thing I could
do, the officer suggested, was talk to the dis-
Hakan Toka, Okan Toka
trict attorney. I might be able to obtain his
release until the date of the trial.
“So I went to the district attorney. I told him
that the old man was an honest person and
promised that he wouldn’t run away and that
he would appear for the trial. More importantly,
I said, he had a congenital disease with
severe high blood pressure. He had already
had one heart attack and from a medical
point of view, the stress of being in jail could
kill him. The attorney said he couldn’t be of
any help, but the judge might be able to do
something.
“I went to the judge and told him the whole
story again. At first he didn’t show any inclination
to be helpful. I was getting desperate so I
promised I would vouch for the old man, and I
would also write a medical report to his superiors
in Ankara, testifying that the old man
was too ill to be in prison. I dropped a lot of
names, trying to impress him. The study, I said,
was a joint collaboration between the famous
Hacettepe University Hospital in Ankara and
the famous Max-Delbrück Center for
Molecular Medicine in Berlin and was receiving
support from the Department of Health in
227 Interlude: The case of the short-fingered musketeer
Ankara. I told him that the old man was a key
subject in our study, and that we hoped we
would learn about the genetic mechanisms
behind primary arterial hypertension, a common
disease worldwide.
“The name-dropping didn’t seem to impress
him very much. What did get his attention
was therapy for hypertension – it had been
the cause of a heart attack that his mother-inlaw
had suffered. He asked, ‘What is all of this
about high blood pressure? Don’t you have a
cure for it?’ I started to tell him about primary
and secondary hypertension and about
‘exceptional families’ who can help to evaluate
the genetic causes of primary hypertension.
He ordered some black tea from his secretary.
We sat and talked at least for an hour.
At the end, he ensured me that I would have
his full support for the study and entrusted
the old man in my care.
“The old man was excited and very grateful to
leave the prison. He told everybody in the clan
that ‘the German Doctor’ had secured his
release from jail. After that I wasn’t the foreign
investigator anymore – I was adopted, and
regarded as a respected member of the clan.”

The year was full of stories, Okan says. “But
the main thing is that this year was an exceptional
period of my life. I learned how truly different
scientific fieldwork is from lab experiments.
It requires a great deal more spontaneity
and improvisation and certainly lots of
dedication to overcome difficulties in the ‘protocol.’
And I’m very grateful to Professor Luft,
who is an extraordinary personality as well as
a brilliant scientist and physician, for giving
me this opportunity.”
One important aspect of the year for Okan
was to establish contact with his “roots”. It is
one thing for a Turkish-Bavarian to have contact
with his heritage through his parental
home. It is quite another thing to experience
that heritage directly, alone, as a primary care
physician with immense responsibilities for
strangers, in a rural setting.
Okan obviously wasn’t the only one to have
profited. His work ensured that the family
received dedicated medical care, and one
result of the year was the discovery that their
most serious medical problems could be kept
under control with anti-hypertensive drugs. It
means a temporary reprieve for people who
previously had a life expectancy of about 50
years. Since the study began, Friedrich says,
none of the family members have died.
��
Finding a mutation in a gene would be a
start – but probably only a start – in
understanding why people develop brachydactyly
and hypertension. The symptoms of
Interlude: The case of the short-fingered musketeer
the Japanese and Turkish patients show that
wherever the genetic defect lies, it alters the
way the body develops. One effect is a change
in the growth of fingers. Hypertension might
be the result of another anatomical change,
somewhere else.
Some cases of high blood pressure could be
linked to the structure of arteries in the brain.
Ramin Naraghi, a neurosurgeon at the
University of Erlangen-Nürnberg, had used
magnetic resonance tomography (MRT) to
chart the course of blood vessels in patients
with essential hypertension. In 20 of 24
patients with essential hypertension – and only
three of his control patients – Ramin found that
blood vessels were squeezing the left side of a
small structure in the brain called the rostral
ventrolateral medulla. This tissue helps regulate
autonomous body functions like the rate of the
heartbeat and blood pressure.
“What happens is a bit like putting a kink in a
garden hose,” Friedrich says. “The twist in the
artery causes a build-up of pressure, and that
puts stress on the medulla. The compression
might make the tissue send more signals out
through nerves, which could explain high
blood pressure. Ramin discovered this right
about the time we had started our study, and
we hadn’t done imaging in the brain. So we
thought we’d better have a look at our
patients.”
The facilities were available in Turkey, so
Ramin, the Toka brothers, and Silvia – who
needed a break from all that lab work – took
228
off for Trabson. Trabson is a bustling city on
the eastern Black Sea coast with a university
hospital and a magnetic resonance facility. In
all of the 15 family members with hypertension
that were examined – and none of those
without it – Ramin discovered a loop in the
brain’s blood vessels which would put pressure
on the medulla.
As a physiological explanation for the patients’
high blood pressure, it made sense. Mutations
often lead to changes in the body – in this
case, shortened fingers and an abnormal routing
of a brain artery. It’s the kind of thing that
is seen all the time in animal models in the lab:
remove a gene and change the body.
All that would have been fine – even the
geneticists would have been happy – if the lab
could have pointed the finger at a mutated
gene. It might have permitted them to reproduce
the condition in mice and then begin
looking for possible treatments.
But so far, they still hadn’t found one. And
that was likely to be crucial to finding a treatment
for the condition.
“With serious developmental defects, even if
you’ve identified a problem such as the route
of a blood vessel, the solution for the patient
isn’t usually going to be to go in with surgery
and try to rebuild the body,” Friedrich says.
“Instead we need to pin down the mechanisms.
But without a gene, we were
stumped.”
Friedrich says that the hypothesis regarding
the kinked artery is technically termed a “neu-

ovascular compression syndrome.” This
mechanism is known to be involved in some
types of brain-stem diseases, and it can be
partially dealt with through surgery.
“However, operating on people’s heads to
lower their blood pressure is not a generally
accepted strategy to cope with hypertension
– although several neurosurgical groups volunteered
to do so. It wouldn’t occur to me to
operate on the family’s fingers to lower their
blood pressure, so why would I want to operate
on their heads?” asked Friedrich. “You certainly
wouldn’t go that way without a lot
more clinical evidence.”
So the family members were once again invited
to Berlin. By this time, a more sophisticated
Clinical Research Center (CRC) had been
established at the Franz Volhard Clinic. A CRC
is a unit dedicated to research in human subjects
and its mission is different from a hospital,
whose efforts are focused on treatments
rather than research.
The CRC is headed by Jens Jordan, who was
trained in medicine and clinical pharmacology
in Berlin-Buch, and cardiovascular regulatory
physiology at Vanderbilt University in
Nashville, Tennessee. Jens reasoned that the
neurovascular compression might increase
the activity of the sympathetic nervous system
in the hypertensive family members. To
test that notion, he needed a “readout” from
the sympathetic nervous system. This
involved implanting a tiny tungsten needle in
the peroneal nerve, which serves the leg
below the knee joint.
“The needle was inserted in sympathetic
nerve fibers that can then be monitored with
a recorder,” Friedrich says. “Now we had to
repeat our tests of the cardiovascular system,
challenging it with drugs. As we artificially
induced changes in blood pressure, we were
now watching how the nervous system
responded. One of the final tests involved
temporarily paralyzing the entire autonomic
nervous system with a drug. This procedure
was a major intervention for the patients and
the university ethics committee reviewed the
study protocol with immense care.”
When the procedure was approved, the team
found that family members did not develop
increased activity of the sympathetic nervous
system. Instead, the team learned that family
members with the condition had almost no
ability to cope with events that changed
blood pressure. In other words, their bodies
weren’t adapting and responding to changes.
Mechanisms that step in to lower blood pressure
in healthy people weren’t working.
The finding is an important clue, and it shows
that pressure on the medulla is connected to
the problem of hypertension. “But we won’t
offer any operations until we understand the
molecular mechanisms that underlie the
problem,” Friedrich says. “There may be more
going on, and until we have pinpointed the
defect in the genome, we can’t use animal
models such as mice to analyze and study the
disease.”
��
229
The study was only one of many other
things going on in the doctors’ lives.
Hakan Toka, Okan’s brother, went off to
Harvard for a clinical rotation in pediatric
nephrology. The rotation was arranged with
Friedrich’s help: Harvard students pay huge
tuition fees and to arrange a senior student
rotation there tuition-free is not a part of the
standard German medical education. While
there, Hakan was asked to give a talk on his
research, and he chose this hypertension
study as his topic.
Interlude: The case of the short-fingered musketeer
When he finished a man in the back raised his
hand. It was James Melby, an endocrinologist
from Boston, who has known Friedrich for
over 30 years. “I’m treating a person with
brachydactyly and hypertension, in my outpatient
clinic,” he said. “Maybe we should talk.”
The family of the patient that Melby was
treating fit the pattern perfectly. Four members
had the symptoms, four had died of
strokes, and three had normal stature and
blood pressure values. The same cause had to
underlie both the Turkish and American cases,
Hakan reasoned; DNA samples from Melby’s

Group picture in the hospital in Trabzon. The “muskateer” of the story is on the left.
patients might help narrow the search for a
defective gene.
Another adventure ensued when Herbert,
Silvia, Hakan, and Friedrich traveled to
Johannesburg, South Africa. They had been
given a lead on yet another group there with
brachydactyly and hypertension. When they
examined the family, they discovered the
same fingers, namely type E brachydactly.
However, not all affected persons were hypertensive.
“Essential hypertension is common,” Friedrich
reminds me. “Short fingers and hypertension
could coincide by chance alone. But when I
was asked to present a talk on the syndrome,
once again a hand went up in the back. The
hand belonged to a staff pediatrician at the
university hospital. She had a patient, a small
child with type E brachydactyly and hypertension.
Silvia went back to the lab and proved
that the child had our syndrome. This patient
is the only isolated case – an individual where
the syndrome doesn’t run in the family – that
we have identified so far. Such cases can happen
through spontaneous mutations in a single
person, if they affect a region of genome
that is known to cause disease.”
Then a third family turned up, in Canada. They
had been discovered by David Chitayat, a
physician and genetic counselor who is now
Interlude: The case of the short-fingered musketeer
head of the Prenatal Diagnosis and Medical
Genetics Program at Mount Sinai Hospital in
Toronto, Canada. At the time he was at the
Hospital for Sick Children in Toronto. Chitayat
got involved when a seven-year-old child was
admitted with persistent headaches that
were followed a few days later by seizures. The
child was short for his age, with the shortened
fingers and toes typical of brachydactyly
– and high blood pressure. A family history
revealed that his brother, father, uncle, cousin,
and grandmother (all on the father’s side)
were also affected. They were all being treated
for hypertension.
Chitayat had read the 1973 study by Nihat
Bilginturan (in his own paper he called the
Canadian family’s condition “Bilginturan
Syndrome”) and had also read the article in
which Friedrich’s group pinned the source of
the problem to a region of chromosome 12. In
1998 Hakan, Chitayat, Melby, and Friedrich’s
lab put data from the three families together.
Microsatellite studies of the Canadian and
American patients showed a genetic change
within the same region of the chromosome.
Because the problems did not completely
overlap with the Turkish data, it allowed the
group to narrow down the search even more.
There were still no clear genes within the target
region, but there was one nearby: a
sequence encoding the molecule L-SOX5. At
230
the time, little was known about the functions
of human L-SOX5. But Veronique Lefebre,
a French scientist working in Cincinatti, Ohio,
was investigating the version of L-SOX5 found
in mice. She sent an email to Silvia telling her
that she had discovered one of the functions
of the molecule. L-SOX5 protein encoded a
gene-activating protein called a transcription
factor. In mice it activated a molecule called
collagen type II. “That was interesting,”
Friedrich says, “because L-SOX5 is produced in
the tips of the toes, the long bones, and vertebra
of developing mice. Collagen type II is a
tough, fibrous protein that is the main component
of bone and cartilage. This made L-
SOX5 a great candidate gene in terms of
explaining the skeletal anomalies in patients.
So we thought we’d better sequence the
entire human gene.”
In the days before the completion of the
human genome, this often required a huge
effort – a bit like searching for a small bit of
text in the Encyclopedia Britannica and then
reconstructing the pages that come before
and after it, letter by letter. The gene was likely
to be spread across a large area of chromosome
12, probably broken into pieces, and
interrupted by bits of garbled text called
introns. Far from being irrelevant “junk”, these
bits of nonsense within a gene can disrupt a
gene’s functions and make the protein that it
encodes useless.
“It took Silvia a year to sequence the entire
human L-SOX5 gene,” Friedrich says. “It turned
out to be a monster, over 500,000 base pairs
long. The cause of our syndrome might be a
change in any one of those letters – we’d only
know if we found the same flaw in every
member of the family with the disease, and in
none of those who didn’t have it. In such a
huge gene you find all kinds of variations. So
every time Silvia found a single letter that had
changed, she checked the pattern of inheritance
in the family. There wasn’t a consistent
pattern. It allowed her to rule out L-SOX5.
There were lots of tears in the beer after this
setback! Our best candidate was gone.”

Silvia Bähring

Friedrich Luft plays guinea pig.
L-SOX5 lay just outside the target region on
chromosome 12. “But mutations weren’t the
only possibility,” Friedrich says. “We knew there
had to be something going on in that region
and started thinking about other kinds of
changes that DNA can undergo. By that time,
Silvia had already sequenced most of the coding
genes in the region. If nothing had become
misspelled, or none of the letters of the code
had been lost, then maybe the order of things
has just been rearranged.” (Some of the ways
that DNA sequences become rearranged are
described in the chapter “Waking a Sleeping
Beauty and other tales of ancient genes.”)
To discover whether this had happened, Silvia
used bits of human DNA grown in bacteria
(called BACs) to make fluorescent probes of
various colors. Silvia attached red, green, and
yellow fluorescent markers to specific locations
in the DNA from the BACs – like planting
flags in the ground to track the course of an
earthquake. If there had been no rearrangements,
all the flags would always appear in
the same order, in every person’s DNA. That
was the case for all the healthy members of
the family. But those that had inherited
brachydactyly had also inherited a jumbled
Interlude: The case of the short-fingered musketeer
region on one copy of chromosome 12 A piece
of DNA had been cut out, flipped around, and
reinserted into a new place. This had changed
the order of the flags.
“The real test was to look at the same
sequences in the Canadian and American
families, as well as the child from South
Africa” Friedrich says. “There, too, we found
rearrangements. Interestingly, each family
and the child had shuffled things around a bit
differently. But in each case, the same region
was disturbed.”
Even though the changes didn’t directly
involve a gene, they still might affect a gene
someplace else. For example, DNA sequences
near a gene often contain instructions telling
the cell when and where in the body it should
be switched on and off. In this case there
might be instructions important to the construction
of fingers, toes, or blood vessels.
Scrambling that information could well lead
to fingers that were too short, or to the
rerouting of an artery.
Such instructions are most likely to affect
nearby genes, and in this case there were four.
“I don’t need to go into detail,” Friedrich says,
232
“but from what we knew, any of them might
be involved in regulating blood pressure. We
had four pretty hot candidate genes We reasoned
that if we had tissue from affected and
nonaffected family members, we could test
these genes. That meant taking a probe of a
likely tissue – a blood vessel.”
Obtaining samples would mean another trip
to Germany for several members of the
Turkish family. A small section of blood vessel
would be removed in a simple procedure.
Once again the team relied on the CRC, run by
Jens Jordan at the Franz Volhard Clinic.
“We could obtain a sample of blood vessel
through a relatively simple procedure to remove
one from the patient’s buttock,” Friedrich says.
“Still, it was surgery – and elective surgery, not
directly aimed at improving the person’s health.
You never want to do that unless the whole procedure
has been perfectly ironed out. We needed
a guinea pig. So don’t ever say I haven’t put
my ass on the line for this study.”
One of the candidate genes included an ion
channel (see the story called “The electrician’s
toolbox”) and a second protein that controlled
its behavior. Ion channels sit in cell membranes
where they open and close, allowing
charged particles to pass into and out of the
cell. This changes the cell’s charge and creates
electrical impulses that travel along nerves –
which could well be involved in the kinds of
nervous system problems that had been
detected in the patients. Maik Gollasch,
internist, nephrologist, and electrophysiologist
– with credentials ranging from Moscow
to Vermont, but nonetheless a clinical investigator
from Friedrich’s department at the
Franz Volhard Clinic – relied on a sophisticated
set of experiments using a technique called
patch clamp, which can detect whether or not
ion channels are opening and closing properly.
(See “The electrician’s toolbox” for a closer
description of the patch clamp technique and
more of its uses.)
Unfortunately the results were negative. The
researchers couldn’t detect any differences in

the way the four genes were being used in the
tissues of patients and their healthy relatives.
The candidate genes were innocent.
More tears in more beers. Back to the drawing
board.
��
At some point or another, Norbert Hübner,
a basic MDC scientist whose specialty is
complex genetics, has been involved in several
of the stories in this book. Besides carrying
out several of their own research projects, his
group has been running the MDC’s DNA
microarray facility and helping other groups
look for diseases caused by combinations of
genes. That’s a huge task involving vast quantities
of data which can only be grasped using
sophisticated statistics and biocomputing
methods. Friedrich had been talking to
Norbert about the project for a while, and in
2003 he jumped in to make a major contribution
to the project.
“Up until that time there was still the possibility
that we were deluding ourselves about
the connection between this genetic
condition and ‘normal’ essential hypertension,”
Friedrich says. “It was time to try to
bridge the gap, and the way we did it was to
involve yet another continent in the project –
China.”
The idea was to try to repeat the type of study
that had pinpointed chromosome 12 with the
Turkish family, but this time with a “normal”
population suffering from “normal” essential
hypertension. “Such studies have been carried
out before, but generally the results have not
been clear,” Norbert says. “The problem is that
in any typical population you find a huge
range of genetic variation. People migrate and
genes from all over the world get mingled. It’s
hard to sift through and find single things
that contribute to disease. If there are multiple
factors involved, which is the case in
essential hypertension, the job gets much
harder. You’re not looking for a needle in a
haystack, you’re looking at a whole field of
haystacks, each of which has lots of needles in
it. You’re searching for the haystack with two
identical needles.”
The best solution is to find a large group of
people who live in a fairly isolated region,
where family relationships are clear so that
diseases can be tracked accurately through
233
Interlude: The case of the short-fingered musketeer
many generations. Norbert’s lab had extensive
collaborations in China, involving families
that had participated in clinical studies
before. This meant that an enormous amount
of preparatory work had already been done.
The project was carried out under the leadership
of Norbert and Maolian Gong, at the

Sino-German Laboratory of FuWai Hospital in
Peking. The group specializes in molecular
genetics. An active collaboration has developed
between the Fu Wai hospital and the
MDC. Maolian is currently “on loan” in Berlin,
Friedrich says. “Her endless enthusiasm and
capacity for hard work has made her a valued
partner for the research teams of Norbert and
myself.”
“We started with a group of 94 people from a
large family group,” Norbert says. “The analysis
showed a strong link between hypertension
and alterations in chromosome 12. Then
we added 32 more families – a total of 174 parents
and their children – from the Shijingshan
district. There was a high rate of hypertension
in this group. When we looked at the distribution
of this larger group, there was an even
stronger trend. There was a nearly perfect
Interlude: The case of the short-fingered musketeer
overlap between the region found in this
study and the one identified by Friedrich in his
Turkish patients. These groups live thousands
of kilometers from each other in different cultures.
This is very strong evidence that we’ve
found a part of the genome that – somehow
– plays an important role in regulating essential
hypertension.”
��
The best way to get an impression of
Friedrich Luft’s sense of humor, and his
finely-tuned ironic world view, is to sit next to
him at a meeting. He doesn’t hesitate to voice
his opinions when he thinks he’s hearing nonsense.
(“Never suffer fools gladly,” he grumbles.)
If you never have that pleasure, you can get a
sense of the man by scanning the titles of the
review articles he writes on a regular basis for
234
the Journal of Molecular Medicine, whose editorial
offices are located on the Berlin-Buch
campus. “A fat attack occurred in fat city,” one
title reads. “Rocking around the clock, while
time is relative.” “Is imprinting in printing or in
press?” “Phosphate’s fate made easier.” “Blue
acid blues,” and “Escalator-driven research.”
And that’s just going back to 2004. (All right, I
have to mention a classic from 2003: “Baa,
baa, black sheep, are your kidneys full?” which
discusses a link between poor diet in mother
sheep, diabetes, cardiac disease and hypertension.)
Alongside writing, caring for patients, and
heading research projects, Friedrich has had
another hobby over the past few years: has
been fighting for the construction of a new
type of institute on the Berlin-Buch campus.
“With ongoing changes in the health care sys-

tem and the way hospitals are administered,
it is becoming more difficult to carry out basic
research projects that involve patients,” he
says. “Today if we needed to check a family
into a hospital for a week, in pursuit of some
important but basic question, there’s no clear
structure to fund that, and the hosptials
aren’t set up for that kind of arrangement.”
And Friedrich knows first-hand how difficult it
is to bridge the gap between clinical research
and the laboratory. New structures are needed
as points of contact for collaborative projects,
to inspire the kind of interdisciplinary
thinking that is needed to carry them out, and
to give young scientists training from both
perspectives.
“There aren’t many good models for doing
this in Europe, but the United States have
been successful at it for many years,” Friedrich
says. “Exactly these issues led them to set up
clinical research centers, or CRCs. The contribution
that the Franz Volhard Clinic CRC has
made to the Turkish project is self-evident.
What we’re now planning in Buch is called the
ECRC, the Experimental and Clinical Research
Center. If things go as we hope, this will
include a significant amount of laboratory
space, support, and infrastructure for common
projects, as well as a more elaborate inpatient
CRC on the Charité side. That will be fully
equipped with examination rooms and diagnostic
tools to carry out a wide range of tests
with people on an outpatient basis. Patients
that need to be admitted can be cared for in
so-called ‘scatter beds’ in the hospital.”
A CRC must be multidisciplinary, Friedrich
says. “When my oncology friends finally devel-
235 Interlude: The case of the short-fingered musketeer
op that gene-therapy or T-lymphocyte cure,
we need to have the beds, laminar flow
rooms, etc. ready for them almost without
notice. The health care system cannot take
the responsibility for research; its not their job
and they cannot do it anyway.”
The new structure will have access to some of
the most sophisticated new instruments in
the world for studying the whole human
body. “Under the ECRC umbrella, with funding
from the BMBF, the MDC, and the Charité, we
have just made two major purchases,”
Friedrich says. “One magnetic resonance
imaging scanner that can carry out imaging
of the entire human body, and a second MRI
machine for use with small animals. The
buildings to house them are being constructed
as we speak. These instruments are important
because to extend what we are learning

to patients, we need to extend what we have
learned about cellular processes to the whole
organism. And we have to connect what we
learn from model organisms to humans – our
patients.”
Interlude: The case of the short-fingered musketeer
Accomplishing all of this will require a new
type of scientist, so one of the main activities
of the ECRC will be training clinician scientists.
The investment requires as much as four
years’ bench research that should culminate
236
A comparison between blood vessels from
patients with the hypertension condition
(right) and those without. An expert can
tell the difference!
in a PhD degree for these clinicians, obtained
with the same rigor as the PhD earned by the
basic scientist candidates.
��
Until the late 1990s, the failure to find a
gene in the key region of chromosome
12 would have probably been the end of the
road in the search for the causes of hereditary
brachydactyly and hypertension. Today things
have changed through the discovery that
genomes encode more than just genes. For
example, a great deal of the sequence is used
to make small molecules called microRNAs.
Unlike other RNA molecules, they aren’t used
to make proteins. Instead, they influence
whether other RNAs are allowed to do so. If a
cell produces a microRNA whose chemistry
allows it to bind to another RNA, the two molecules
dock onto each other. This prevents the
other RNA from being used to make proteins.
(The process is described in more detail in the
chapter “A very quiet cure.”)
The discovery of this “code within the code”
presents new mechanisms by which cells regulate
their affairs and diseases might arise.
Silvia, Friedrich and Norbert think the problem
arising from chromosome 12 could well
have to do with a microRNA. Some of these
small molecules have been shown to have
important functions. By preventing the production
of a protein, they can change how a
cell behaves and how it develops. “It’s entirely
reasonable to think that the result could be a
change in the building plan of fingers or
blood vessels in the brain,” Norbert says.
Rearrangements that have been found in the
chromosome may stop the cell from producing
an important microRNA or scramble the
sequence of another molecule that it needs to

ind to. Proving that this is the case will
require discovering the microRNAs encoded in
the region and identifying the target molecules
they bind to. These small bits of code are
hard to find; Norbert and his colleagues are
working on the problem.
“More needles and more haystacks,” he says.
“This time the needles are a lot smaller. But
we have some leads.” Silvia’s work to date is
consistent with a mutated microRNA, a regulator
of regulators as responsible. To bring in
the evidence to convince the scientific community
is a major challenge.
Fortunately, the team has a new strong partner.
Nikolaus Rajwesky recently joined the
MDC from the Rockefeller University in New
York. Nikolaus is a world’s authority in
microRNAs. (His work is described in “Playing
the piano of planaria.”) He is also bringing socalled
“deep sequencing” to the MDC – a strategy
to discover what the genome is doing
beyond simply turning out proteins based on
its genes. With all the amassed expertise,
Friedrich says, there is real reason to believe
this “rearrangement syndrome” can be solved.
“Had we found the thing quickly, we would
have written one fancy paper and gone on to
other things,” Friedrich muses. “There are so
many blessings from this project that are less
tangible.” In the laboratory, Atakan Aydin
moved from technician to full-fledged scientist
with a “summa cum laude” PhD degree.
Yvette Neuenfeld attended the university on
the side and obtained a Masters degree. In the
clinic, Hakan Toka is a nephrologist currently
working at the Massachusetts General
Hospital at Harvard University. Okan Toka is a
pediatric cardiologist at the University of
Erlangen. Herbert Schuster became professor
of medicine and has moved on to other challenges.
Thomas Wienker is professor of
human genetics at the University of Bonn.
Jens Jordan has just assumed the chairmanship
of the Department of Clinical
Pharmacology at the University of Hannover.
Silvia and Friedrich remain embroiled in the
The campus guest house provided lodgings when the families came to Berlin.
project. “We will never give up,” they say – and
then, why should they?
The culprit has been chased for more than
three decades, across four continents.
Although its identity remains a mystery, the
culprit has been surrounded. Getting this far
has required plodding detective work as well
as the use of the most sophisticated technology
science has to offer.
That’s where the scientific story ends, for the
moment. But stay tuned. A definitive solution
to the mystery may be close at hand. Or it may
require the development of new concepts and
methods. Ten years ago no one would have
guessed that a microRNA might be responsible
for disease. Who knows what the genome
will be whispering to us a decade from now?
237 Interlude: The case of the short-fingered musketeer

Part four: From the protein village
to the cosmopolitan cell

A hub in the city of the cell
By February the city had grown weary of the strikes. For two weeks
there were no S-Bahns, the main link between Buch and central
Berlin. That was an inconvenience, but with creative use of streetcars
and buses and some quick reconstructive surgery on an old bicycle,
most of us without cars could still get to work. When the trains began
rolling again, the bus drivers and U-Bahn staff took their turn. Suddenly
many parts of the city, including the campus, could only be reached by
taxi or on foot or that decrepit bike. Service briefly resumed for Easter,
but the first thing we heard after the holidays was that more strikes
were planned.
At the height of the strike, long lines of disgruntled passengers formed
at the service desk at the Friedrichstrasse, a central hub in Berlin’s transportation
network. They began to share strike stories. Did you hear
about the fistfight at the airport? Four people fighting over a taxi? At
the front of the line a man raised his voice. No reaction from the man
and woman behind the counter – they had heard it all before. The
phone rang and the woman picked it up. “Where are you coming from,
sir?” she asked. “Where do you need to go?” She frowned at her

Claus Scheidereit
computer. Even the most familiar routes had
to be replanned. Finally we got tired of waiting
and went to check a map of the transportation
system – maybe we had overlooked
a connection somewhere that would get us
downtown.
The map reminded me of a frustrating discussion
I’d had that afternoon with Claus
Scheidereit, a scientist at the MDC. The frustration
was all mine. Claus’ work concerns
details of another type of network: signaling
pathways in the cell. These are the information
routes by which signals from the environment
are passed to genes, a theme that
appears over and over in this book because of
the crucial role pathways play in most biological
processes.
Sometimes they resemble subway maps – but
they are far more complex. When a biochemist
begins talking about signaling it’s
easy to get lost in the details. To help me keep
track, Claus drew a map – it seemed endless. I
had to remember that nearly every feature on
this chemical map represented years of work
on the part of someone’s laboratory. For an
important purpose: nearly everything that
goes right in a cell – and nearly everything
that goes wrong – can be attributed to how
information travels along the chemical S-
Bahn of signaling pathways.
Although it takes thousands of molecules to
pass the signals that control the cell’s business,
some of the routes are more crucial than
others, and occasionally they cross at
Friedrichstrasse-like hubs. Claus’ laboratory
focuses on one of them, a sort of service desk
that receives information from all over the cell
and routes it to the right genes. Most of his
work concerns a complex of proteins called
NF-κB, spoken “N-F-kappa-B.”
These molecules have dominated his entire
career, partly because of their important functions.
And there’s another reason: if you want
to get an overview of Berlin’s transportation
system, you should go to the Friedrichstrasse.

All the transportation systems meet up here.
People from all corners of the city pass
through on their way to all other parts of the
city. Likewise, if you want to get a deep look at
the basics of how cells manage signals, NF-ΚB
is a good place to start.
��
In 1988, while working as a postdoc at
Rockefeller University in New York, Claus and
his colleagues purified a complex of proteins
containing NF-ΚB from human cancer cells. In
the test tube the factor could activate genes
brought into cells by HIV, the AIDS virus. The
tumor cells originally came from a patient suffering
from Burkitt’s lymphoma (described in
the story “Where slow rivers meet”).
“NF-ΚB had been discovered two years earlier
in the lab of David Baltimore at MIT,” Claus
says. “At the time it was only known to be
Part four: From the protein village to the cosmopolitan city
present in immune system cells. NF-ΚB helps
program the cell to release molecules called
cytokines, which alert the body to the presence
of infectious agents. Another function is
to help switch on the most complex genes in
the body – the sequences used to make antibodies.”
These genes are spread out over a huge
region of the genome. Making them requires
slicing out great chunks of that information,
throwing it away, and pasting the broken
ends of DNA back together. The process
begins when proteins bind to DNA. But until
the late 1980s, very few of the molecules
involved in making antibodies were known.
That changed with the discovery of NF-ΚB.
Baltimore and his colleagues discovered that
it could bind to regions of DNA called Kappa
sequences found in antibody gene sequences
and many other genes. If NF-ΚB didn’t bind
242
there, crucial parts of some antibodies couldn’t
be made. B cells failed to develop properly.
“Once we knew that NF-ΚB bound to Kappa
sequences, we started finding such sites all
over,” Claus says. “They occurred near many
other genes and in the genomes of several
viruses.” HIV had such a site. Another was
found in the monkey virus SV40 (introduced
in the story “A pact with the devil.” ) Both of
these infectious agents attack immune system
cells and insert their genomes into that
of the host. That information has to be turned
into RNA and proteins in order for the viruses
to reproduce, and they get the cell to do it for
them. By including a binding site for NF-ΚB,
they take advantage of a powerful gene-activating
molecule.
What was missing, Claus says, was a better
overview of what genes NF-ΚB might be activating.
And what kept it from switching on all

of those genes all the time? He moved to
Germany and started a group at the Max-
Planck Institute for Molecular Genetics in
Berlin. Six years later he started his lab at the
MDC. NF-ΚB has accompanied him on each
move.
After two decades and 70 research articles on
the topic, he’s still at it.
��
Part of the lab’s work has been devoted to
taking the NF-ΚB machine apart to see
how it works. NF-ΚB proteins are made in the
main compartment of the cell, the cytoplasm.
To activate genes they have to move into the
nucleus. This is a critical step, Claus says, and
understanding it has required learning how
NF-ΚB is built and how it interacts with other
molecules.
“The more we have worked on NF-ΚB, the
more functions we’ve found,” Claus says. “I
know you’ve written about DNA chips – the
technology we use to get a look at the entire
gene activity of a cell. A few years ago Daniel
Krappmann, a postdoc, used the method to
compare cells with active NF-ΚB to very similar
cells in which the pathway was quiet. He
found a ‘signature’ of about 70 target genes
that are affected by the signal. Many of those
genes encode proteins that go on to activate
other genes.”
NF-ΚB can set off cascades of events that lead
to the division, self-destruction, or specialization
of cells. Or they cause disease. “So it would
be a dangerous thing to have the signal
switched on for very long,” Claus says. “The status
quo in the cell is for NF-ΚB to be switched
off. It’s kept in check by keeping it away from
genes, keeping it out of the nucleus.”
This is accomplished by a sophisticated protein
“lock”. By finding the molecules that NF-
ΚB binds to – taking apart the machine –
Claus’ lab contributed to identifying the parts
of the lock and determining what kind of key
the cell uses to open it.
NF-κB can activate genes in different ways, depending on how it is activated itself. The pathway should not be
switched on all the time, but that happens in Hodgkin lymphoma disease. Claus hopes to understand why by
untangling the very complex interactions in the pathway.
“The lock is built of proteins called IKBs,” Claus
says. “They fasten themselves directly onto
NF-ΚB and have to be released before it can
activate any genes. Releasing the brake is the
job of a complex of proteins called IKK. So the
main signaling pathway that activates NF-ΚB
does so via IKK, which then unlocks IΚB.”
The lock isn’t just removed, he says; it gets
destroyed. An incoming signal changes the
chemistry of IΚB proteins. This brings it to the
attention of enzymes that come along and
break it down. Once that happens, NF-ΚB is
free to go.
IKK, NF-ΚB and the lock have to work to establish
the right pattern of active and silent
genes in cells. In 2007 PhD student Meike
Broemer and Michael Hinz, a postdoc in Claus’
group, discovered that a protein complex
called Hsp90-Cdc37 helped to give IKK the
shape it needs so that the signal – the key –
fits. Hsp90 and Cdc37 are chaperones, molecules
that help fold proteins into the right
structure. If they don’t do that for IKK, the
complex can no longer release the lock on
NF-ΚB.
Closing the Friedrichstrasse station for a day
would inconvenience a lot of people moving
through Berlin, but wouldn’t be a disaster.
Disrupting a major hub in a signaling pathway,
on the other hand, usually has devastating
consequences for the cell. As the functions
of NF-ΚB became clearer, Claus expected
to find diseases caused by defects in the
mechanisms that keep it under control.
��
Alois Alzheimer’s chance encounter with a
patient led brought him in contact with
the disease that now bears his name. A coincidence
brought Bernd Dörken, a clinical
243 Part four: From the protein village to the cosmopolitan city

Bernd Dörken
researcher at the MDC and Charité, in contact
with Hodgkin’s disease research.
Nearly two decades ago, while working in the
clinic at the University of Heidelberg, Bernd
was treating a terminally ill patient. “He had
enormous lymph nodes and a swollen
abdomen where fluid had collected,” Bernd
says. “While we were still trying to find a way
to help, we drained some of that fluid and put
it in cell cultures, planning to save it and look
at it later. The cells began growing wildly.”
A look through the microscope revealed that
the sample contained some large, very odd
cells. They contained two bulging nuclei,
which made them look like owls’ eyes. They
are known as Hodgkin/Reed-Sternberg (HRS)
Part four: From the protein village to the cosmopolitan city
cells and are a characteristic of Hodgkin’s disease,
a cancer of the lymph system.
In 1991 Bernd moved to Berlin and brought
the cells along, hoping they would yield
insights into the causes of the disease. “At the
time labs only had seven lines of cells derived
from Hodgkin’s patients,” he says. “An eighth
would be useful. And it turned out these cells
had some unusual characteristics. They didn’t
produce a protein called CD30, which had
been found in every other Hodgkin’s cell line.”
That fact led to a small scientific skirmish
when the group published their findings in
the Journal of Experimental Medicine in 1993.
A scientist at another institute in Berlin
refused to believe the results. He warned
Bernd to retract the paper – if not, he would
challenge it in print. But further work confirmed
the findings.
“One characteristic of HRS cells is that they
produce huge quantities of cytokines – small
signaling molecules that exert a powerful
attraction on other cells,” Bernd says. “It
explains the swelling of the lymph nodes during
the disease; so many blood cells collect
there. Very few of these cells – often less than
one percent – are actually dangerous tumor
cells. The rest are healthy cells that have been
attracted there. HRS are the only tumor cells
known to release cytokines this way. We
thought that we should look at a transcrip-
244
tion factor that caused the release of the
cytokines, and the logical candidate was NF-
ΚB. That got us talking to Claus Scheidereit
here at the MDC.”
It was the beginning of a collaboration that
has now been going on for more than a
decade. One step in finding out what causes
Hodgkin’s disease, Bernd says, means figuring
out what makes blood cells become so curious
that they don’t fit in any normal blood lineage.
That was difficult because the HRS cells
had characteristics of both T and B cells (introduced
in part two, “Identity crisis”).
In the late 1990s Klaus Rajewsky’s lab in
Cologne managed to isolate single cancer
cells from the tumors. When Klaus looked at
the cells’ antibody genes, part of the story
suddenly became clear. He discovered that the
genes had undergone two types of changes
that only happen in B cells. First, the DNA had
been rearranged in a complex cut-and-paste
operation that creates unique sequences. As
new cells are created, each inherits one of the
types.
Bernd compares them to the many types of
key forms found in a locksmith’s shop. “This
already makes a huge number of types of
cells, but things aren’t finished,” he says.
“There’s still fine cutting to do, to make random
changes here and there that really give
each cell its own identity – like taking one of

the blanks and cutting it to fit a specific door.
That’s a process called somatic hypermutation,
and it happens during the last stages of
the specialization of B cells.
“Klaus Rajewsky found that HRS cells had
undergone both types of changes,” he says.
“That means they must have started out as
mature B cells. But rather than finishing their
specialization, they now start off on a strange
developmental path. They start to produce
molecules found in other types of blood cells,
in T cells and macrophages.”
The findings change how researchers have
been thinking about both cell differentiation
and cancer. “A common feature of these blood
tumors is that cells reproduce at a high rate,”
Bernd says. “Since that’s a feature of stem
cells, a common assumption has been that
the process of cancer starts high in the tree of
development, in the early stem cell stages.
Then mutations or other problems push the
cells down a aberrant path. That’s the conclusion
you come to if you assume that specialization
is a one-way road, that once a cell has
taken a step, there’s no going back.
“But look at what we’ve found here. There is
plasticity! HRS cells start at the end of B cell
specialization, but they turn around and come
part of the way back. Somewhere along the
route they become pregenitor cells.”
The implications go beyond cancer, he says. If
a specialized cell can undo part of its programming
and revert to a more generic state
in cancer, could healthy cells be taught to do
the same thing? “Think of what that would
mean for regenerative medicine,” Bernd says.
“We have assumed that in order for the body
to repair or renew its tissues, it needs a stock
of stem cells. But with age, those stocks get
used up or die out, and some types of tissues
can’t be rebuilt. The only solution would be to
replace them. Unless you could find a way to
take existing specialized cells, make them
more generic again, and program them to go
down a new pathway.” Understanding this
escape from “terminal differentiation” would
require a deeper look at the genetic programs
at work in HRS cells.
In 2001 Michael Hinz of Claus’ group used
DNA chips to look at their total gene activity.
Of the many thousands of genes that are
active in the tumor cells, he identified a pattern
of about 50 molecules which were under
control of NF-ΚB. This “NF-ΚB signature” has
been a big help as the researchers try to figure
out the effects of the pathway in cancer cells.
“Overactive NF-ΚB signaling has several
effects that lead to cancer,” Claus says. “One
characteristic of tumors is that cells aren’t
able to stop dividing. The cell cycle is controlled
by signals, some of which are passed
along by NF-ΚB, as Michael Hinz found out
earlier. Those signals should only switch on at
the right times, but if something breaks
down, the pathway can get stuck in ‘transmitting
mode,’ and it keeps sending the message
245 Part four: From the protein village to the cosmopolitan city
Michael Hinz

to divide. With Stephan Mathas, an MD from
Bernd’s group, we then searched for further
signal-regulated factors that are out of control
in the tumor cells. This led to the discovery
of another growth promoting and gene-activating
protein, called AP-1, which was always
switched on in Hodgkin tumor cells. That molecule
has been linked to control of cell division
and it is known to be modulated by NF-ΚB.”
Loss of this control is one of the worst things
that can happen to cells, so they have a backup
system to protect themselves. Tumors and
other serious malfunctions often trigger a cellular
self-destruct program called apoptosis.
“That, too, is managed by signals,” Claus says.
“As you can probably guess, some of them
pass through NF-ΚB. So defects in this network
can have two different effects that lead
to the same end. They push the cell to divide;
simultaneously they rob it of its ability to stop
the process when it gets out of control.”
A look at patient tissues by Bernd’s group
revealed cases in which IkB proteins, the NF-
ΚB brake, had undergone mutations. This
echoed what Claus and his lab had found in
the test tube: without the lock, cells lose control
of NF-ΚB and become cancerous. Next,
human tumor cells were transplanted to mice
and Claus’ lab could prove that NF-ΚB was one
player in triggering the development of
tumors. Uwe Kordes, an MD who worked in
Claus’ lab and now a specialist in children’s
oncology, discovered the same basic problem
with NF-ΚB in a frequent form of leukemia,
and in 2005 Stephan found it in yet another
type of lymph node tumor.
NF-ΚB’s status as a signaling hub, Claus says,
makes it attractive when trying to figure out
how to cope with diseases such as these.
“Most things that go wrong involve the braking
system. That makes IKK proteins an interesting
drug target. If you can keep them from
receiving signals, they won’t release NF-ΚB to
enter the nucleus. Arsenic is poisonous
because it interferes with IKK, so we looked at
what effect it would have on Hodgkin’s cells.

We also tried drugs that block the activity of
Hsp90, the chaperone that helps shape IKK
proteins so that they can receive signals. Both
of these strategies prevented the activation of
NF-ΚB. As a result a lot of the cancerous cells
self-destructed.”
Until Claus and Bernd turned their attention
to the problem, the signals and pathways that
changed gene activity in Hodgkin’s disease
were unknown. Treatments for the disease
have been based on radiation and chemotherapy.
Uncovering the role of NF-ΚB and finding
mutations linked to its activation were huge
steps forward that suggested a clear strategy
for new therapies involving IKK. In recognition,
Claus and Bernd were awarded the 2005
German Cancer Prize.
Their findings suggest a more effective method
of treating Hodgkin’s disease and other NF-ΚB
associated cancers. That’s the goal, Claus says.
That’s what we are all aiming for.
��
It has now been 18 years since Claus moved
to Berlin, and when he talks about the city
it sounds like he’s at home. He lives in
Charlottenburg, an area of restaurants and
cafés that he enjoys when he has the time.
His other pasttime is sailing. Usually this
means renting a boat on the Wannsee, but he
has ventured much farther than that.
Claus grew up in Schleswig, an old coastal
town between the seas in Northern Germany.
In college he joined a rowing club, and then
didn’t have much time for water sports for a
few years. When he moved to Berlin a colleague
at the university, an MD, reminded him
that the world consisted of more than just science,
and they sneaked out for an afternoon
every week to take sailing lessons. Eventually
Claus received a license to sail on the high
seas, and that led to trips to the Canary
Islands and other parts of the Atlantic.
Now he tries to help his own students keep
perspective. “It’s a tough stage in their
careers,” he muses. “I see young people here,
spending 12 hours a day in the laboratory.
Hurrying from one building to the next without
noticing what a nice day it is. They have to
work hard right now, but you still somehow
have to keep a balance. Nobody teaches them
how to do that. One of the jobs of the group
leader is to help them through the tough
times and to keep up their excitement. ”
Does it take a certain kind of person to
become a biochemist? If so, Claus admits he
may be the archetype. His interest in the topic
goes back a long way. “A few years ago I went
to a school class reunion,” he says. “Somebody
there reminded me that in high school I had
given a talk on signaling. On synapses.”
At the university he considered medicine but
decided on chemistry. “It was about the most
complicated thing I could find,” he laughs. “I
wanted to do something hard. I got interested
in molecules and their structures. I liked the
sophistication of the methods.”
He seems like the type of person who would
learn the timetables of trains by heart, who
would enjoy intricate puzzles. You couldn’t ask
for a more complicated puzzle than unraveling
the NF-ΚB network, and to work on it he
has called in a wide range of expertise.
Alongside biochemists, cell biologists, and
chemists, the group now regularly works with
experts in bioinformatics and mathematics.
One reason is simply to keep track of all the
components of the network and their interactions.
“As long as we were dealing with a relatively
small number of molecules, we found ways to
represent their interactions,” he says. “But the
hub-like nature of NF-ΚB and the fact that it
can activate so many genes require much
more complex models – it’s like going from a
multiplication table to advanced calculus. Our
theoretical partner in this is Jana Wolf, who
has a small group that focuses on mathematical
modeling of cellular processes.” Jana’s
group has recently received funding under an
initiative launched by the MDC’s parent
organization, the Helmholtz Association, to
support projects in “systems biology.” As well
as collaborating with Claus, Jana is studying
the Wnt/β-catenin signaling system that
Walter Birchmeier’s group has worked on for
so many years.
Claus has known Jana for some years, and
when they first met she told him they had a
mutual acquaintance. “When she was in high
school, she won a big math competition,” he
says. “The first prize was a chance to meet this
famous person at the Humboldt University. It
was Jens Reich – now an emeritus of the
MDC.”
A factor that makes building a model of the
NF-ΚB network so hard is the pace of new discoveries.
“I used to be able to read every paper
on NF-ΚB,” he says. “Now if you look at the scientific
literature, you find over 20,000 papers
on the subject – there are more than 2,000
just on IKK alone. So it’s a major task to keep
track of the data and to integrate it. To try to
integrate it.”
��
One rainy afternoon in February, during a
short respite from the strikes, Claus and
I ride together into town on the S-Bahn. We
talk about sailing for a while; then the talk
turns inevitably to NF-ΚB. I’m getting a better
idea of the molecule’s place in the cell signaling
network, and its role in cancer. It’s such a
central node that it is probably involved in
guiding the development of embryos. I ask
him about it.
“Hair,” he laughs.
Hair?
247 Part four: From the protein village to the cosmopolitan city

Part four: From the protein village to the cosmopolitan city
“At first it seemed trivial. Ruth Schmidt-Ulrich is our mouse specialist. She
developed a strain in which NF-ΚB is generally supressed. We call it a
‘super-repressor’ mouse. It has a version of IkB that can’t be broken down,
so the lock is never taken off NF-ΚB. This mouse is interesting because it
can be used to test hypotheses about NF-ΚB’s normal physiological activities.
It is also useful in studying what happens when NF-ΚB is blocked –
that’s necessary to discover harmful effects that might appear during
treatments, so it serves as a model for therapeutic inhibition.”
As Ruth studied the animals, she found predictable problems – defects
in the immune system because of NF-ΚB’s role in immune cell regulation,
and a high rate of cell death in some parts of the liver. Then she
began noticing problems with the hair.
“We had been hoping to find major developmental defects,” Claus says.
“We already expected that NF-ΚB was needed for the development of
lymph nodes and other structures formed by immune cells which
makes sense because many of their biological functions depend on NF-
ΚB. And there is a strong effect if NF-ΚB signaling breaks down. But otherwise
there was nothing seriously wrong with the animals’ major
body structures. A big disappointment.
“Then Ruth brought me this mouse that had slanted eyes and sort of
‘punk’ hair. My first reaction was, ‘We are now getting into hair
research?’ I told her, ‘We’re not a molecular barber shop.’ But it was well
worth taking a closer look. There were other problems with structures
in the skin. Something was wrong with the follicles; it meant that the
mice only produced one type of hair. And sweat glands didn’t develop
properly. There were also problems with the fine structure of teeth –
the structure of cusps of the molars.”
A search of the medical literature revealed a similar combination of
symptoms in humans called ectodermal dysplasia. The disease had
been known for a long time – Charles Darwin was even familiar with it.
One of his many correspondants, a Scottish civil servant working in
India, had called it to his attention. In 1875 Darwin wrote:
“I may give an analogous case, communicated to me by Mr. W.
Wedderburn, of a Hindoo family in Scinde, in which ten men, in the
course of four generations, were furnished, in both jaws taken together,
with only four small and weak incisor teeth and with eight posterior
molars. The men thus affected have very little hair on the body, and
become bald early in life. They also suffer much during hot weather
from excessive dryness of the skin. It is remarkable that no instance has
occurred of a daughter being affected...though the daughters in the
above family are never affected, they transmit the tendency to their
sons: and no case has occurred of a son transmitting it to his sons. The
affection thus appears only in alternate generations, or after long intervals.“
248

Without knowing it, Darwin was describing a
rare mutation of a gene on the X chromosome,
which explained the pattern by which
it was inherited. But it would be nearly 40
years before fly geneticist Thomas Morgan
found sex-linked traits and figured out why
they occurred.
Ruth’s discovery meant that NF-ΚB played a
role in the development of hair and gland
structures – as well as ectodermal dysplasia –
in mice. And as the scientists talked about it,
they realized that the condition might offer a
new way to explore NF-ΚB’s functions.
“If you think about it, you have in this tiny
space in the skin a mini-organ with a complex
structure involving stem cells that specialize
to create multiple types and structures such
as the hair follicle and glands,” Claus says.
“That happens because of signals. We could
use our tools to manipulate the system and
watch the effects on the skin.”
Further studies by Ruth and the rest of the
team have shown that NF-ΚB and other components
of the machine work together in subtly
different ways to create different parts of
skin structure. In the early embryo NF-ΚB is
activated to produce placodes – thickened tissue
that will later form structures in the skin.
It is also needed later, as hair follicles form, to
stimulate cells to divide and push downward
through other layers of skin.
“It has been very hard to determine NF-ΚB’s
role in any developmental process in mammals,”
Claus says. “This is one of the very few
cases in which that could be done in detail. By
studying this mini-organ, Ruth and the lab
have identified the signals upstream that
guide this process and what part of the
machine they act on. For example, we know
that the downward movement of the placode,
which creates the follicle, is controlled
by signals that come through β-catenin.
That’s an important topic of Walter
Birchmeier’s group, so we’ve been able to pursue
it with him. At the same time, close by in
other parts of the skin, IKK proteins direct
some processes that don’t involve NF-ΚB.”
As the S-Bahn rolls toward the city center, he
talks about his lab’s future plans. Working out
the functions of the NF-ΚB pathway will take
many more years. Most of the work will
involve hammering out intricate details of cell
signaling. Sometimes it’s hard to see the big
picture, he admits. But if you want to deeply
understand how cells manage the huge variety
of signals that cross paths on their way to
genes, NF-ΚB is an excellent place to start.
One reason is the pathway’s involvement in
Hodgkin’s disease and other forms of cancer. If
you want to reroute the signals that are
responsible, without disrupting something
that healthy cells need, you’ll have to know
the details.
My stop comes first and we say goodbye.
From the platform I watch the train slide on to
the city. Claus will take the S2 to Friedrich -
strasse, then switch to a westward line for
Charlottenburg. That will work today, anyway.
If the strike begins again tomorrow, he’ll have
to find another route. On my way out of the
station I take another look at the transportation
map, with its crossing S-Bahn and U-
Bahn lines. Today, for some reason, it doesn’t
seem so complicated.
249 Part four: From the protein village to the cosmopolitan city

The pleasures and
powers of green tea
In 1973, a year after receiving his law degree, businessman Philippe
Neeser left his home in Geneva to study in Osaka, Japan. A few years
later he experienced a life-changing moment when he was introduced
to the traditional Japanese tea ceremony. Over the next three decades,
Neeser became an internationally recognized master of this tradition
which is part art, part philosophy, part religion.
History says that green tea had been known in Japan for several centuries
but remained a drink of aristocrats until 1191. In that year the
priest Myoan Eisai returned from a trip to China, bringing with him the
Zen Buddhist religion and tea seeds. He used the brew as a stimulant to
keep monks awake during their long hours of meditation. Three hundred
years later Marata Shuko developed the custom of serving tea as a
means of establishing spiritual harmony between a host and a guest.
Sen-no Rikyu, a disciple of one of Shuko’s disciples, developed the practice
into a highly ritualized ceremony. Rikyu wrote poems about tea in
the deceptively simple language of Zen:
Part four: From the protein village to the cosmopolitan city
250

Eric Wanker

Tea is not but this:
First you make the water boil,
Then infuse the tea.
Then you drink it properly.
That is all you need to know.
Thus the history of tea in Japan is inseparable from religious tradition.
It is equally intertwined with medicine. Eisai wrote a two-volume book
whose title translates as “How to stay healthy by drinking tea.” The
first sentence reads, “Tea is the ultimate mental and medical remedy
and has the ability to make one’s life full and complete.” Among the
curative powers he cites: improving the functions of the brain, bladder
and heart, easing thirst, and curing hangovers.
Part four: From the protein village to the cosmopolitan city
If anything, the ancient Japanese may have underestimated the medical
value of green tea. At the beginning of the 21st century, scientists
like Erich Wanker of the MDC are learning that its active substances
may have many more beneficial effects on health, and they are beginning
to understand why.
252
��
That a poet and priest should also be a healer is no contradiction
in Buddhist tradition; it is much more unusual in the West, in
modern times. Yet a few people like Philippe Neeser manage to fuse
radically different facets of modern existence. When he is not serving
tea, he is head of legal affairs for Ciba Chemical Specialties, a Swiss
company with offices in Japan. In 2005 he guided Japanese Emperor
Akihito and his wife through the high-tech Swiss pavilion of the World
Exposition, which was being held in Japan. Three years earlier Neeser
had received another tremendous honor when he was selected to
carry out a ceremony reserved for the greatest tea masters. He served
tea to the great Buddha of Nara during celebrations of the statue’s
1450th anniversary. The Buddha weighs 500 tons and has fingers the
size of a human being. Building the statue and its temple required the
efforts of over two million people, wrecked the Japanese economy, and
used up most of the country’s copper.
Erich Wanker’s path to the MDC is not quite that unusual, but it has
still involved a big cultural leap for the young Austrian researcher. “I
come from a small village in the Alps,” he says. “I had virtually no contact
with real science as I grew up. Of course I had biology, physics and
chemistry in high school. But there were no role models – scientific
research was happening ‘out there’ in some higher, remote sphere. In
spite of that, alongside sports, I always had an interest in nature.”
He says that at the age of 16 or 17, that interest blossomed into an
interest in “hard-core” sciences like engineering and chemistry, accompanied
by a desire to make an impact on the world. At the university
he started off with chemical engineering and then became attracted
to biochemistry and molecular genetics.
“I was fascinated the first time I heard of enzymes,” he says. “There
were people in the world who were making real discoveries, really finding
out how life works. That was for me – I had to do that. But it wasn’t
easy to explain this to my family and friends. There had never been
a doctor or a professor in our family, or even in the village. I had no idea
whether I was doing the right thing. But at some point there was no
turning back.”
He laughs as he remembers a conversation with his parents. “This was
after I had had some successful publications and it was clear I was
going to make it as a scientist,” he says. “They were worried, the way
parents always worry about their kids. They shook their heads and
asked, ‘Are you ever going to be able to make a living doing this?’”

He ticks off the stages of his career. As he speaks his hands move rapidly
and his blue eyes flash. It’s not hard to imagine that a small
Austrian village would have trouble containing all that energy.
He received his PhD in Biochemistry from the Technical University of
Graz in 1992, then it was off to the University of California in Los
Angeles to do postdoctoral work in the lab of David Meyer. One theme
of the lab concerned the processing of proteins that cells secrete or
put in their membranes. These molecules are synthesized from RNAs
at the surface of a structure called the endoplasmic reticulum, or ER.
Getting there requires transport molecules that recognize and dock
onto a receptor protein. Erich and his colleagues helped to identify the
receptor and show how it docked onto the protein-synthesizing
machinery. It was an important finding, but he had his sights set on
bigger things.
“I had decided to move out of the basic cell biology I had done at the
Meyer lab and into disease research because I felt a need to work in an
area that had a direct impact on more urgent human questions,” he
says. “Coincidentally, the gene that causes Huntington’s disease had
just been discovered. In late 1994 I started looking for a group leader
position, and heard that Hans Lerach – one of the people who had
found the gene – was looking for a group leader in Huntington
Disease research.”
Huntington’s is a brain disease that develops because of the premature
death of particular types of neurons. Symptoms include jerky
movements and a general loss of control of the body, accompanied by
changes in mental abilities and behavior. The problems usually appear
in middle age and grow progressively worse over time. As in
Alzheimer’s disease, the problem seems to stem from protein fragments;
they form clumpy fibrils that do not dissolve. Whereas in
Alzheimer’s the fibrils form in the cell cytoplasm and the space
between cells, in Huntington they mainly appear in the cell nucleus,
with some clusters in the cytoplasm.
Only a small percentage of Alzheimer’s cases are clearly hereditary,
linked to variants of genes which are described in the story “A Swiss
253 Part four: From the protein village to the cosmopolitan city

Alexandra Redel
Army knife of the cell.” But Huntington’s follows a strict hereditary pattern.
It is caused by a dominant mutation in a gene, which means that
a person needs to inherit only one copy of the molecule to develop the
disease. For a long time scientists had been trying to identify the
culprit.
Hans Lehrach’s group at the Imperial Cancer Research Center Fund in
London and other laboratories had steadily been closing in on a region
of the fourth human chromosome. The gene was finally identified in
1993 by the Huntington’s Disease Collaborative Research Group, an
international collaboration between six laboratories, including that of
Lehrach. They named it huntingtin, with an “i” – a common ending for
proteins – ensuring decades of spelling problems for biology students.
In 1994 Lehrach moved to the Max Planck Institute for Molecular
Genetics in Berlin to head a new department and began hiring new
groups. Erich applied and was initially given a one-year position, which
was then extended. “At the beginning it was just me and one student,”
he says. “That’s about the smallest group you can get.”
As Erich began setting up his new lab, scientists were discovering
some of the differences between the dangerous and healthy forms of
the huntingtin protein. Like all proteins, huntingtin is spelled of chemical
subunits called amino acids. At the “head” of the molecule is a long
stretch consisting of just one letter, glutamine, repeated over and over,
as if a key on the computer has become stuck. In healthy people this
string is between 11 and 23 letters long, but if the molecule has 40
repeats, a person is highly likely to develop the disease. Because the
chemical abbreviation for glutamine is “Q”, researchers call the region
a polyQ sequence.
While the gene responsible for Huntington’s disease has been known
since the early 1990s, no one knew why it caused problems until 1997.
Part four: From the protein village to the cosmopolitan city
Pablo Porrasmillan
That year Erich and his lab were the first to discover that huntingtin
accumulated in clumpy aggregates. “That discovery fundamentally
changed the way HD is classified,” Erich says. “From then on, it could be
grouped with amyloid diseases like Alzheimer’s. And it spawned an
enormous amount of research into the aggregation of proteins with
polyQ motifs, leading to the discovery that nine further diseases that
could be classified as polyQ diseases.”
The discovery of the aggregates was very intriguing, but it didn’t say
anything about why one form of the molecule led to problems for cells
or the brain. And it said nothing about why that problem starts to
become serious when a person reaches middle age – until then, the
body somehow copes.”
There were many possibilities, he says. The defective form of the molecule
might have an altered architecture, allowing many copies of
huntingtin to cluster in a way that couldn’t be dissolved. Or different
proteins might dock onto the two types of huntingtin, folding one of
them into a dangerous form. Scientists weren’t even sure why the protein
was toxic. The questions were similar to those posed by Bernd Reif
in the story “Bad origami:” the protein fragments that cause
Alzheimer’s and Huntington’s disease can accumulate in different
ways, forming clusters of twos, threes, fours, or much larger complexes
that go on to form super-complexes. The most dangerous form
might be a grouping that appeared somewhere along the way and
then adopted another form.
“The fact that symptoms normally arise between the ages of 40 and
50 means that something has changed at that point in life,” Erich says.
“With age, just as with some kinds of stress, the cell begins losing its
ability to cope with the continuous challenge of dealing with a misfolded
protein. This suggested we ought to look at other players – the
proteins that interacted with huntingtin – to discover the mecha-
254

nisms of misfolding and toxicity. But 15 years ago we didn’t have great
tools to discover what proteins interacted with each other. We were
going to have to spend time developing that technology.”
Erich’s personal history has taken him from a small village in the Alps
to a major world capital. Now science was about to take him from a
single molecule to the big city of cell networks.
��
During the 1990s improvements in DNA sequencing technology
had suddenly given researchers a huge list of new genes in
humans and other organisms. Most of these molecules had never
been seen in experiments, and no one knew what jobs they performed
in cells. “We were all looking for shortcuts to determining their functions,”
Erich says. “Figuring out what molecules they interacted with
would go a long way toward that goal.”
In 1989 Stanley Fields and Ok-Kyu Song of the State University of New
York invented a method called a two-hybrid system that used yeast
cells to “fish” for proteins that can dock onto each other. They began by
removing a gene that the cells needed to survive. They replaced it with
a new version of the gene that could be triggered by an artificial protein.
The protein had two modules: one that docked onto the gene and
activated it; another that acted as a bait. The trigger would only be
pulled if another protein (the prey) took the bait, and only then would
the yeast cell activate the gene and survive.
Fields’ and Songs’ clever strategy was to give one strain of yeast the
combination of bait and trigger, and another strain the prey. Then they
allowed the two strains to mate. If this brought together two proteins
that could bind, the yeast colony survived, and the results could be
seen with the naked eye.
By doing this across the entire genome, the method would theoretically
give you a map of all the possible interactions between proteins. But
creating tens of thousands of yeast clones containing baits and others
with the prey would be a huge amount of work – unless the procedure
could be automated. So Erich and his group began building the robots
and tools needed to carry out yeast two-hybrids on a massive scale.
This was crucial for the creation of the first map of protein interactions
across the whole human genome. Today his group has become so proficient
at the methods that when they have the time, they offer them
as services to other labs.
The resulting map is an important first step in charting interactions
between proteins. “Whether a pair actually binds in cells is another
question,” Erich says. “Even if their chemistry makes them a perfect fit,
two molecules may never get the chance to interact because they
occupy different regions of the cell – or are never made together in the
same type of cell. So at the end of a two-hybrid experiment you have
a lot of data that needs to be verified through other types of experiments.”
The mid-1990s saw the development of another technology that
could help out, mass spectrometry. It can be used to identify the proteins
in small samples taken from cells. While part of Erich’s growing
lab developed the two-hybrid system, he put others to work on mass
spectrometry.
The two-hybrid system provided a list of candidates that could bind to
huntingtin, but it didn’t answer other questions: how huntingtin
assembled into fibrils and what made them toxic. Other methods
developed in the lab were beginning to provide insights into these
questions. A filter method helped them trap large fibers that didn’t
dissolve; Erich hoped that these large clumps would contain other
molecules that played a role in their formation.
255 Part four: From the protein village to the cosmopolitan city
Anne Wagner

And what was it in huntingtin itself that allowed it to snap together
in groups, rows, and fibers? At a meeting he heard a talk from Max
Perutz, the great pioneer of protein structures. The clustering of huntingtin
proteins posed interesting questions that had attracted Perutz’s
attention.
While working in Cambridge in the late 1940s and 1950s, Perutz had
found a way to determine the shapes and architecture of proteins
loading purified proteins with atoms of metal and exposing them to
X-rays. The method allowed Perutz and his student John Kendrew to
obtain structures of the first two proteins and earned the men a Nobel
Prize in 1962. It was the same year that their close colleagues in
Cambridge, Francis Crick and James Watson, received the award for
deducing the double-helix structure of DNA. The two discoveries
launched modern molecular biology.
Perutz tells the story of his colorful life in a collection of autobiographical
essays called I Wish You’d Made Me Angrier Sooner. A native of
Austria, like Erich, his family had fled the country during the Nazi
takeover in 1938. During the war, he was briefly interred in a Canadian
prison camp until a group of important scientists intervened on his
behalf. Upon his release the British government put him to work on
the war effort. Because Perutz had once carried out a study of the
behavior of ice crystals in glaciers, he was enlisted in a crazy project to
Part four: From the protein village to the cosmopolitan city
build artificial icebergs to be used as massive floating aircraft carriers.
It didn’t work, Perutz said later, because the combination of wood pulp
and ice used to make the icebergs... melted.
“When I met him, Max Perutz was in his eighties but was still going
strong working on proteins,” Erich says. “At the conference he gave a
talk about huntingtin and figuring out what disease mutations did to
its structure. He had the idea that if two molecules lay side by side but
inverted, so that the head of one protein lay next to the tail of the
other, there would form a kind of zipper between the glutamine
repeats – the polyQ sequence. That’s why these big clusters would form.”
Erich approached Perutz and said that some of the data his group had
been collecting might support the hypothesis. That encounter led to a
long-term collaboration that continued until Perutz’s death in 2002.
Erich is a co-author on some of the scientist’s last papers. One of them
expanded the huntingtin findings to other molecules.
“There are many proteins that accumulate in and between cells and
cause disease,” Erich says. “Several of them contain repeats of amino
acids. Even though the recipe of each of these proteins is different,
Max wondered whether a single underlying principle might be
responsible for making them behave the same way. For example, a protein
in yeast – similar to the prion molecule that causes mad cow dis-
256

The current map of the “Human Interactome,” as depicted by a database that Erich’s
group has developed. Each dot represents a protein. Lines connect the molecules that
directly interact with each other.
ease – contains repeats of the amino acid asparagine. Prion diseases
also involve the accumulation of molecules between cells. Max
showed that adjacent asparagines can zip up to each other in much
the same way as glutamines.”
The paper also showed how regions of the β-amyloid protein fragment
– which collects between cells in Alzheimer’s disease – can link
up to other proteins into β-sheet structures that lead to fibers.
The work shows how sensitive some proteins are to changes.
Increasing the concentration of a molecule in the space between cells,
or a switch in one letter of its recipe, can have huge effects on its
behavior. “A protein folds in a certain way because the amino acid
string is loaded with energy,” Erich says. “Chemical interactions inside
the molecule stabilize it. But it achieves that stability because of its
complex equilibrium with everything else in the environment. If something
changes, you may quickly see a sort of chain reaction of misfolding
and aggregation – at least that’s how we imagine the process.”
��
So ultimately, understanding the behavior of a protein requires
understanding its place in the whole, a network of protein interactions
that change all the time as cells carry out their functions and
produce new molecules. The connections on the map are redrawn as
cells grow and develop, when they come under stress from the environment,
and as a natural result of growth and aging. The protein
interaction maps of healthy cells are different than those affected by
a disease. If you knew how the network normally behaved, you could
monitor the impact of things like mutations or drugs.
“That’s important because if you want to fix a problem with the drug,
you have to fix the network and not just what’s happening to one molecule,”
he says. “You’re not just asking, ‘Will protein X bind to protein Y
and stop some sort of bad behavior?’ ...You have to be sure that by fixing
one thing, you’re not disturbing something else.”
Experiments from Erich’s lab and other groups throughout the world
were beginning to fill in the map of protein interactions – like going
from the ancient Greeks’ map of the world to that of Vasco da Gama,
Erich says. “Those maps completely changed navigation and provided
a basis for exploring the world,” Erich says. “Having a complete chart of
human protein interactions – if it were high quality – would be the
equivalent, it would be a fantastic tool for every kind of biological exploration.
So developing this map has been a major focus of our efforts.”
By 2005, systematic studies with the yeast two-hybrid system had
given the lab a list of about 3,200 interactions between proteins, most
of which had never been detected in other experiments. Within the
257 Part four: From the protein village to the cosmopolitan city

next two years, 3,000 more interactions were added to the map. All of
the information went into a database that the lab maintains on the
Internet. The data is only the beginning, however; a major effort of the
group has to verify that the proteins actually come into contact in
cells, using other types of experiments.
Erich shows me one of the maps, on which proteins are shown as
round balls, connected by lines which reveal their interactions. “Can
anyone read this?” I ask him.
“That’s a big issue,” he agrees. “You can’t use it if you can’t read it. A lot
of our effort has gone into designing tools to help scientists query the
information in various ways. That begins with simple things, like looking
up a protein you’re interested in and looking up the molecules it’s
known to interact with. You can see whether it belongs to more than
one signaling pathway in the cell. You can put in the names of two
molecules and ask whether they bind to any common partners. You
can zoom in from a protein’s role in the whole network to its functions
in one tissue, at one developmental stage, in one disease.”
Part four: From the protein village to the cosmopolitan city
Ideally, Erich says, following a protein’s interactions will suggest new
links to a cell signaling pathway or a disease process. “It’s a hypothesisgenerator,”
he says. “Once you know where to look, you can confirm a
protein’s role through other experiments, such as checking for mutations
in patient tissues.”
The value of the tool depends on the quality of its contents, for example
by cutting down on the number of predicted protein interactions
that turn out to be false. Erich and his colleagues are trying to teach
the computer to recognize some of these cases itself, by presenting it
with “gold standard” lists of proteins that interact and those that
don’t.
258
��
After a decade of work on protein interactions, everything is coming
together in Erich’s current work on huntingtin. The group has
found intriguing new leads to the causes of disease and points of
attack for potential therapies. Initial experiments with the protein

showed that it might bind to hundreds of molecules. By 2004 postdoc
Heike Göhler and other members of the lab had whittled the list down
to 86 proteins that could link to huntingtin in at least 188 different
ways. Most of the connections had never been seen in other experiments.
Did these interactions have anything to do with the formation of fibrils?
Heike and her colleagues did another screen in mammalian cells.
This time they looked for proteins which, when brought together with
huntingtin, made it aggregate.
The story “A Swiss Army knife of the cell” recounts how Thomas
Willnow’s group found that a molecule played a role in Alzheimer’s
disease by drawing a membrane protein into the cell and sending it to
a particular internal compartment. One of the proteins investigated by
Heike, called GASP2, has a similar role in sorting membrane proteins. It
binds to huntingtin in mammalian cells and moves through the cell
with it.
Another interesting molecule has emerged from a collaboration with
Hitoshi Okazawa’s group at the Tokyo Medical and Dental University in
Japan. “The Okazawa lab has established a cell culture model of
Huntington’s disease,” Erich says. “I told you that mutant forms of
huntingtin move into the nucleus, where they form fibrils. As they do
so they trap molecules that are necessary for the cell to go about its
normal business. Okazawa and his colleagues have been trying to
understand how huntingtin gets into the nucleus and what happens
once it gets there. We jumped in to look at other proteins that bind to
huntingtin and to see if they have any influence on these processes.”
The scientists compared cells with mutant forms of the protein to
those with healthy huntingtin, and cells in which fibrils accumulated
and others in which they did not. One pattern emerged over and over:
whenever mutant proteins were present, levels of other molecules
called HMGB1 and HMGB2 dropped.
“These are incredibly important proteins,” Erich says. “They’re the most
common molecules in the nucleus and you find them everywhere on
259 Part four: From the protein village to the cosmopolitan city

the evolutionary tree. They are huge multi-functional tools that can
bind to DNA. They also act as adaptor plugs for all kinds of proteins
that need to attach themselves to DNA to repack it, repair it, activate
genes... the list goes on and on. When mutant huntingtin is around,
HMGB molecules seem to get stuck, and there are far fewer of the
molecules floating around in the nucleus to do their jobs.” When the
scientists checked cells taken from normal and diseased brains, they
found the same connection.
The results were exciting because they suggested that HMGBs might
make good drug targets. When the scientists added extra copies of the
molecules to cells with mutant proteins, cells normally affected by the
disease were able to survive as well as healthy ones.
The study revealed that HMGBs have another role that may play a
direct role in Huntington’s disease. In healthy cells the proteins seem
to pass signals that tell the cell when to stop dividing and repair its
DNA. If mutant huntingtin proteins stop the proteins from doing this
job, the timing may be thrown off and the cell will die. Adding HMGBs
to the cells can restore the proper timing of the signals.
��
As Erich’s lab has learned more about the mechanisms that underlie
Huntington’s disease, they have been searching for drugs or
other substances that change the way cells handle the defective huntingtin
protein.
One strategy involves trying to bolster the cell’s natural defenses
against problems. “When cells experience heat or other types of stress,
one thing they do is produce more chaperone proteins,” Erich says.
“These are ‘folding assistants’ whose jobs are to make sure that proteins
are being made properly. Proteins are very sensitive to heat or
other types of stress, and one of the first things that is likely to go
wrong is for them to misfold.”
An antibiotic called geldanamycin (GA) binds to a chaperone called
Hsp90 and triggers the cell to respond to heat. “In one of our earlier
studies we showed that treating cells with GA causes the cell to produce
chaperones,” Erich says. “This stops the aggregation of huntingtin
– the more antibiotic you add, the fewer the big clusters produced
by the cell. But GA is so toxic that using it in a therapy would
cause more harm than good.” Postdoc Martin Herbst has been working
with chemically altered forms of GA that might be more useful.
“Our studies turned up about 300 substances that reduce the clustering
of huntingtin in the test tube,” Erich says. “One of these belonged
to a family of molecules called the flavonoids, which are natural substances
known to have a variety of health benefits.”
Which brings us back to green tea.
Part four: From the protein village to the cosmopolitan city
The plant contains a flavonoid called epigallocatechin-gallate (EGCG).
Alongside centuries of use as a traditional medicine in China, Japan
and elsewhere, there is mounting new evidence that EGCG and other
substances in green tea have anti-tumor effects. They may also help
the body with other problems including microbial infections, allergies,
and heart disease.
Originally the health benefits of flavonoids were attributed to their
“anti-oxidant” properties. They soak up overcharged atoms of oxygen,
which are found in cancer and other diseases. These atoms undergo
frenetic interactions with biological molecules, often breaking them
apart. But some recent findings suggest that this effect may be much
stronger in the test tube than in living organisms.
Erich wondered if EGCG might interfere with the clustering of huntingtin
proteins. That’s what his lab found in experiments in the test
tube, cells and fly models. “EGCG potently blocks the process by which
the repeat sequences in huntingtin ‘zip up’ to each other,” he says. “You
remember Max Perutz’s hypothesis that such repeats in other proteins
might also make them cluster and cause disease? That made us think
that if EGCG was helpful in one case, it might be helpful in others. So
we have tested β-amyloid, responsible for Alzheimer’s, and a protein
called α-synuclein, which is found in Parkinson’s disease. EGCG prevents
them all from accumulating into large fibers. So this natural substance
recognizes a variety of proteins that might collect in large intracellular
clumps and prevents them from doing so.”
How does it manage this? “It doesn’t make the proteins go away,” Erich
says. “And they do still cluster. But when we used the electron microscope
to look at aggregation reactions that had happened in the test
tube, we saw that α-synuclein and β-amyloid had grouped into small
particles and then stopped. They didn’t go on to make the big, cumbersome
particles that don’t dissolve. We followed up with a second
experiment in which we added the different types of particles to cells.
EGCG produced small particles that didn’t harm the cells, while the
bigger fibril structure did.”
The proteins go from a single molecule to a cluster containing various
numbers of proteins and then big fibers. Somewhere along the way
they kill neurons. “Early in the clustering process, EGCG steps in and
groups the proteins in a new way,” he says. “That’s a detour from the
normal path. EGCG binds to the unfolded molecules and prevents
them from lining up in a β-sheet. That is a prerequisite for the formation
of larger, toxic clusters.”
It’s not a miracle cure, Erich cautions. Like most natural substances,
EGCG and other flavonoids will have to be purified, developed, and
extensively tested before they can be used in therapies.
But EGCG has a lot going for it. In the first place, people have been
drinking it for thousands of years – apparently with mostly positive
260

effects. Secondly, it has already been approved for clinical testing in
other diseases. Frauke Zipp of the MDC has recently started a clinical
trial using EGCG in the treatment of multiple sclerosis. And maybe
best of all, EGCG can slip out of the bloodstream and into the brain,
which means it can reach cells in distress. Most other drugs fail at this
because of a bodily defense mechanism called the “blood-brain
barrier.”
Erich and his colleagues aren’t just standing by, waiting to be discovered
by pharmaceutical companies interested in turning their findings
into new therapies. Once he felt the science was solid enough, Erich
put together a major proposal and submitted it to the German Federal
Ministry of Education and Research (BMBF). The program “GO-Bio” was
created in 2006 when the BMBF realized that the relationship
between basic research, pharmaceutical companies and clinics was
changing. Scientists were having to work much harder to attract interest
in promising projects and bring them to the marketplace. One
option that researchers have is to start a company, but many small
biotechnology firms fail because it usually takes many years to trans-
form a discovery into a product. As a result, a huge number of promising
leads are never taken beyond the lab.
As a sort of mix between a research grant and a company start-up
investment, GO-Bio supports very promising projects at an early stage
so that researchers will have more time to develop their findings.
Erich’s project was regarded as so promising that his lab was one of
just a few selected for funding in the initial launch of the program. The
money goes into an intensive search for drugs and substances – such
as the molecules of green tea – that can be used as therapeutic or
diagnostic compounds in the treatment of Alzheimer’s or
Huntington’s disease.
With all of this going on, you might expect his lab to resemble the
Berlin train station on the first day of school vacations, or a Starbuck’s
coffee shop on a Saturday afternoon. But the atmosphere in the group
seems calmer, somehow. Maybe it’s the steadily falling levels of
caffiene. Erich’s team hasn’t started drinking only green tea – not yet.
But you see a lot of teapots and teacups. They seem to have become
standard equipment at the bench.
261 Part four: From the protein village to the cosmopolitan city

Playing the piano
of planaria
It seems to take more than ten fingers to play an étude by
Alexander Scriabin – a pianist’s hands are constantly at work,
evoking turbulent shadows that sweep by in grand gestures,
exposing broken moments of light. Nikolaus Rajewsky is playing
Scriabin to get to know the grand piano upstairs in the MDC
Communications Center. Watching his hands, it’s impossible for a
non-pianist to imagine the feeling in those fingers, the combination
of discipline and wildness needed to play such a piece.
Moving on to Chopin, Nikolaus reaches most of the instrument’s
88 keys. This is the standard number, but a few pianos have more;
the Austrian firm Bösendorfer has made instruments with up to
nine more tones in the bass. To keep from confusing performers,
on some models the extra keys can be covered with a small lid. On
others the colors of the lowest notes are reversed, replacing white
ivory keys with black ebony ones. While modern music exploits the
full range of the classical piano, few pieces require the extra notes.
The long strings are there to add depth to the sound as they resonate
with other tones.
Part four: From the protein village to the cosmopolitan city
262
Nikolaus Rajewsky

Most biological processes make use of the full
keyboard of the cell, Nikolaus says – meaning
that they often involve dozens or hundreds of
genes, RNAs and proteins. But until very
recently, scientists haven’t been able to hear
all the notes or see the keys that produced
them. The discovery of microRNAs in the
1990s was like suddenly discovering that
pianists had been leaving out notes while
practicing their scales.
“For 50 years molecular biologists have been
scrutinizing the process by which the information
in genes is used to produce proteins,”
Nikolaus says. “At every step along the route
we have discovered ways that cells intervene
to modulate the production of a molecule or
alter it to change its functions. So it was
amazing to find that there was an entire level
of control in the cell that we had never been
aware of. The question now is how widely
Part four: From the protein village to the cosmopolitan city
cells use this mechanism in processes like
building a body, and what role it plays in disease.”
MicroRNAs have been invisible to most of the
tools used to investigate molecules. Rather
than encoding proteins, like genes, they dock
onto other RNAs and signal for them to be
silenced. This process, described in the chapter
“A very quiet cure,” represses the target
RNAs from being used to make proteins. It’s
hard to detect something by its absence – it
would be like listening for the fingers that
aren’t used during a particular measure of
Scriabin’s études.
But miRNAs have other functions which are
even harder to grasp. Sometimes they block
the production of a protein without destroying
an RNA; instead, they obstruct the work of
the protein-building machinery. Trying to
observe this activity of the tiny molecules
264
across the level of the genome, Nikolaus says,
has been a huge methodological challenge.
Another problem is that the computer analysis
of genomes, which has successfully identified
a huge number of new genes, was initially
not very successful at finding microRNAs.
“In a way, they were too successful,” Nikolaus
says. “Most of these original attempts led to a
huge number of ‘false positives.’ At least half
of the sequences predicted to be microRNAs
in these early studies probably aren’t.”
Finding the molecules and understanding
their functions will require a combination of
experiments in the test tube, cells, and organisms
– and completely new ways of analyzing
huge amounts of data. When Nikolaus arrived
at the MDC, he brought along a vision for a
new type of lab that would marry the
approaches. That fit in well with the campus’
vision for the Center for Medical Genomics.

He found himself in good company – one floor
down is Norbert Hübner, who is taking an
interdisciplinary approach to the problem of
finding multiple genes responsible for disease,
and upstairs is Thomas Jentsch, who is
doing pioneering work on membrane proteins
(see “The electrician’s toolbox”).
Part of Nikolaus’ group is doing experimental
work on cells and organisms. Another part is
creating innovative computer algorithms to
detect microRNAs. One way that they overlap
is that a hypothesis about the identity and
functions of microRNAs that has been generated
in the computer can be tested in the lab.
“But one doesn’t have the upper hand over
the other,” Nikolaus says. “I’m just interested
in problems, and in bringing the right methods
to bear on them.”
One project within the group is to establish a
new, unusual model organism at the MDC: a
small freshwater flatworm called the planarium.
This animal has amazing regenerative
powers that have fascinated scientists for over
200 years. Cut into halves, or even smaller bits,
the fragments can grow until they have reconstructed
the entire body. Charles Darwin collected
them during his five-year voyage on the
Beagle, and Thomas Morgan wrote at least a
dozen papers on planaria before devoting his
laboratory to the fruit fly. Morgan discovered
that it only took 1/279th of the worm’s body to
rebuild an entire planarium.
Over the past year Nikolaus’ interdisciplinary
team has been settling into the newest building
on campus, the Medical Genomics Center,
a sleek black structure funded by the MDC
and FMP. When people work at night – which
is not unusual – lights from the large office
windows make it look like a futuristic vending
machine.
Science does not keep regular hours.
��
a very wide effort to discov-
“Despite
er and profile microRNAs, there’s
no consensus about the number that might
exist in the human genome,” Nikolaus says.
“Estimates range from a few hundred to tens
of thousands.”
Theoretically these molecules can be captured
in experiments, he says, but there are
some problems. One is that cells may only
produce a few copies of some microRNAs.
Those may have important jobs in the cell, but
they are easily overlooked when ten thousand
copies have been made of another molecule.
Computer scans of the genome might be
another way to identify them, but even if the
machine can be programmed to find their
sequences, it won’t say what types of cells (if
any) produce the molecules or what functions
they might have. Finding that out requires
experiments... which may not see microRNAs
that are made in small amounts... “You get the
point,” he says.
New technology might offer a partial solution.
Advances are so fast, he says, that even
the most forward-thinking scientists find
them hard to believe. “Technology is moving
to the point that it will be possible to
sequence a person’s entire DNA in an hour,”
he says. “The first time that was done, it took
15 years, and the job was just more or less
completed five years ago. Obtaining individual
genomes will not only help to answer a lot
of fundamental questions about humans –
for example, how much difference is there
really between one person and the next –
but it will also change a lot of aspects of
medicine, from new diagnostics to new therapies.”
The increasing speed and falling costs of
sequencing makes it possible to solve some
stubborn problems in new ways. “If you’re
looking for something small and relatively
265 Part four: From the protein village to the cosmopolitan city

are, one solution is simply to do a lot more
sequencing,” Nikolaus says. It’s a bit like conducting
a census to count the population of a
city over and over again, in hopes of finding
people who aren’t at home. If you do it
enough times, eventually you’re likely to find
almost everyone.”
The method, called deep sequencing, is being
applied to a variety of fascinating problems:
to detect mutations in tumors, to sequence
DNA from Neandertal homonids and the
woolly mammoth, to study the genetics of
social behavior in wasps, and to complete the
genome of the grape Pinot Noir.
Part four: From the protein village to the cosmopolitan city
“There’s one catch to the power of deep
sequencing,” Nikolaus says. “It’s heavily
dependent on computing. In order to make
sense of what you find, you have to be able to
sort through immense amounts of sequence
data and map them to specific locations in
the genome.”
The task is a bit like being given a fragment of
text and having to find everywhere it appears
in the Encyclopedia Britannica. That wouldn’t
necessarily be a huge problem if you had the
whole text of the encyclopedia on your computer,
in a single file that could be searched.
But the text of most genomes isn’t available
in that form.
266
This has to do with the way most genomes
have been sequenced and assembled. Instead
of decoding human DNA letter by letter, from
beginning to end, a “shotgun approach” was
used. This would be like downloading the
encyclopedia from the Internet onto your
computer, in bits and pieces a few hundred
letters long. If you did that (and you shouldn’t),
it would be possible to reassemble the
articles into a complete encyclopedia by
matching up the beginnings and endings of
texts and alphabetizing them. “Shotgun” DNA
sequences can also be plugged into their correct
positions in human chromosomes. But
small entries might have been missed while

downloading the Encyclopedia Britannica,
and one of those might contain the fragment
you’re trying to find. That’s the situation for
the human genome and other complex
organisms that have been sequenced using
the shotgun method. In fact, the only animal
for which scientists have a full, beginning-toend
DNA sequence is the worm C. elegans.
So one of the computing challenges is to
identify microRNAs by matching them to
their proper places in the genome. Another is
to deal with the “artifacts” that often arise in
experiments. What may look like a microRNA
may be a fragment of another molecule, an
alternative version of part of a gene, or simply
a mistake made while sequencing. If a computer
program can’t tell the difference, it will
find far too many false positives while looking
at deep sequencing data.
PhD student Marc Friedländer and other
members of Nikolaus’ lab have found a way to
address this problem in a program called
miRDeep. “Basically we have taught the computer
something about the biology of a
microRNA,” Nikolaus says. “The cell processes
premature microRNAs into active, mature
micoRNAs. While doing so, it leaves a signature
on which parts of the premature
microRNA survive and which parts don’t.
miRDeep detects these signatures in the deep
sequencing data.”
Each microRNA begins as part of a larger RNA
molecule, which consists of a string of subunits
called nucleotides. Some of the subunits
are complementary to each other – they
chemically attract each other. They are separated
by a few nucleotides that don’t bind,
instead making a hairpin-like fold. The shape
of this fold is unique in microRNAs, which
allows them to be recognized by a protein
machine called the Dicer complex. It cuts the
RNA into fragments. This leaves behind a
microRNA, the loop at the top of the hairpin,
and the short complementary sequence that
the microRNA used to bind to. This last fragment
is normally destroyed.
Why can’t you find microRNAs through the
chemistry of these folds? I ask. By scanning
the DNA sequence for the right patterns?
“False positives again,” Nikolaus says. “There
are at least eleven million sequences in the
human genome that would form hairpins if
they were transcribed into RNA. Only a fraction
of those are actually made into micro
RNAs. So you can’t make predictions based on
sequences alone. On the other hand, if you
look at the small RNAs that have turned up in
experiments, and find their positions in the
genome, you can work backwards. You can ask
if they belong to the right kind of hairpin.
That’s an important indication that you’re
dealing with a real active microRNA, and not
just a random fragment of some other molecule.”
As well as the microRNA, the other fragments
left over by Dicer ought to appear in experiments.
By comparison there ought to be more
copies of the microRNA – after it is made, the
cell protects it until it can be used.
miRDeep uses all of these criteria and several
others – for example, the fact that microRNAs
are usually conserved through evolution, so
the same sequences are found in many different
species – to score the small RNAs found in
deep sequencing experiments.
When Marc and his colleagues finished the
program, they put it through three tests. “We
used deep sequencing data from the worm C.
elegans because its genome is relatively small
and the sequence is of high quality,” Nikolaus
says. “MicroRNAs were discovered in that
species and extensive studies have been carried
out. We figured that if an analysis using
miRDeep yielded the microRNAs that had
already been found, it would validate the
method. Then we carried out deep sequencing
ourselves on white blood cells obtained
from a dog, and a human cell line. We had different
laboratories carry out the deep
sequencing experiments and then ran them
through miRDeep.”
The tests had very positive results. The program
revealed 89% of the microRNAs in C. elegans
and found several new ones that were
shown to be real by further experiments. It
only proposed a handful of false positives.
Marc and Nikolaus anxiously awaited the
results from the other tests. They didn’t have
anything to worry about; miRDeep proved
itself in both.
“The human genome is 30 times larger than
that of C. elegans, but it has also been extensively
studied in search of microRNAs,”
Nikolaus says. “At the time of the study, 555
microRNAs had been verified in the genome.
No one kind of cell produces them all; the type
we studied were known to produce 213. Well,
miRDeep found 154 of them, nearly threequarters.
If you work out what that means –
in a single experiment with a single cell type,
the program correctly identified 28 percent of
all known human microRNAs. That’s pretty
amazing.”
The scientists were just as eager to see the
results of the experiments on the dog cells.
The genome of the domestic dog was completed
in 2005, and it has become an increasingly
important model for the study of
human disease. Yet comparatively little work
has been done on its genome, and only six
microRNAs were known.
The experiment with miRDeep has changed
that picture dramatically. The program found
three of the known microRNAs (four were present
in the sample), and 203 more sequences
passed the test. They will have to be verified in
experiments, but so far miRDeep has done an
excellent job of picking real microRNAs from
other small molecules. Marc and Nikolaus are
confident that nearly all of the sequences will
turn out to be real. If so, they will confirm that
the method works even for genomes whose
microRNAs have not been studied. The dog is
an important test case, Nikolaus says, because
researchers hope to use it to understand
microRNAs’ contributions to disease in an animal
that is closely related to humans.
267 Part four: From the protein village to the cosmopolitan city

He couldn’t resist trying the method out on
planaria as well. “It’s an extreme case because
of its very distant evolutionary relationship to
humans and other animals with well-studied
microRNAs. If you assume that the program
has some built-in biases because it is based
on familiar genomes, then data from planaria
ought to point that out.”
Quite a bit of experimental data was available
on planaria microRNAs – 61 were known.
miRDeep found 86 percent of them and proposed
39 new ones. Of those, Nikolaus and his
colleagues have tested 19 to see if they really
function as microRNAs. So far 16 of them have
turned out to be real.
��
None of these molecules could have been
found without the marriage of laboratory
and computational science that Nikolaus
has drawn together in the Center for Medical
Genomics.
“A program can predict as many microRNAs as
you want, but those predictions only have any
value if they turn out to be real,” Nikolaus
Part four: From the protein village to the cosmopolitan city
says. “And search algorithms have to be finetuned
by the biology. When the two
approaches work together, we can learn a lot
about microRNAs from deep sequencing
experiments, even if the subject is a relatively
unexplored genome like that of planaria.”
Computer programs have become adept at
looking for the genes in genomes; they are
large sequences with complex patterns. But
the lack of equally high-quality methods to
identify microRNAs and their target genes
has made it hard for researchers to get a grip
on the problem. That’s something Nikolaus
would urgently like to rectify, since experiments
suggest that microRNAs play an
important role in all kinds of biological
processes – likely in the development of many
diseases.
Recently Nikolaus helped take a global look at
the effect of microRNAs in a particular developmental
process: the maturation of B cells,
the white blood cells that produce antibodies.
(They are a central theme in the story “Where
slow rivers meet”) Making a B cell’s antibody
is a complex process that involves cutting the
268
left: A fully-grown planarium
right: a worm in the process of regrowing its head
far right: The smaller worm has just regenerated its tail
DNA strand, discarding long regions, and
reassembling the sequence to form a new
antibody gene. There are a huge number of
pieces to choose from. Because different
regions are removed and combined in each
cell, it acquires a unique gene and builds a
unique antibody. Parts of this process are not
yet understood, and experiments suggested
they might involve microRNAs. A group at
Harvard decided to find out. They needed
computational expertise in microRNAs, and
Nikolaus was the logical choice. Besides being
one of the world leaders in the field, he had a
connection to the lab – it was run by his
father.
Klaus Rajewsky is one of Germany’s bestknown
scientists. Before relocating to the
Harvard Medical School in 2001, his lab carried
out a great deal of pioneering work on the
biology of B cells at the University of Cologne.
Along the way the group created one of the
most powerful tools in genetic research: the
conditional knockout. Until the method was
developed, researchers could not knock out
many of the cell’s most crucial genes to study
their functions, because removing them

caused embryos to die at a very early stage –
often before they had even undergone the
first cell division. Conditional knockouts
allowed a gene to be shut down in a single
type of cell at a precise time. It is now a standard
tool in labs across the world and is the
basis of several of the projects described in
this book.
Did microRNAs play an important role in the
maturation of B cells and the creation of antibodies?
Sergie Koralov and Stefan Muljo, two
members of Klaus’ group, thought of a way to
find out. They used a conditional knockout to
delete the Dicer gene from immature B cells
in mice. Since this is the key enzyme in the
Dicer complex, its removal meant that RNA
hairpins would not be cut to create
microRNAs.
The experiment had a dramatic impact on
the mice. They failed to develop any mature B
cells; the cells died in an early stage of development.
Klaus and his colleagues suspected
that this was due to the presence of molecules
that would normally be blocked by
microRNAs. They compared activity in the
immature B cells to those of normal mice.
Nikolaus and his PhD student Azra Krek analyzed
the results.
“We found 411 different molecules that were
being produced at significantly higher levels
in the experimental animals,” Nikolaus says.
“The levels of some of them – maybe most of
them – had probably changed because they
were no longer being held in check by
microRNAs. The only way to tell was to check
the molecules to see if they had the features
of microRNA targets, and dozens of them did.
And the computation predicted that a single
microRNA cluster had the largest effect in
the Dicer knockout experiment. Subsequent
analysis of potential targets of this cluster
identified bim as a very likely target that
should be specificially important for B cell
maturation regulated by microRNAs.”
Another lab had shown that a protein called
bim is a key player in the survival of B cells.
Cells that produce it die. They are supposed
to – this keeps the body from making cells
that will attack itself and cause autoimmune
diseases. But not all B cells die; they are
allowed to survive when they block the production
of bim.
“This is an essential step in making B cells
which can participate in healthy immune
responses,” Nikolaus says, “and the study
strongly suggests that it’s partly the work of
microRNAs.”
It’s not the only role that the tiny molecules
play in the immune system. A paper last year
by Klaus’ lab – also with the help of Nikolaus
– showed that microRNAs are equally important
in the development of another type of
white blood cell – T cells.
��
The Medical Genomics Center, home to
Nikolaus’ group, has spacious rooms at
the corners of the floors. The groups have put
in comfortable sofas, coffee machines, and
old copies of journals – not all about science.
There’s a motorcycle magazine on a table. It’s
a place to relax, talk, and think about other
things. As Nikolaus lounges on a sofa his fingers
drum a light pattern on the armrest.
Maybe it’s Chopin, some difficult passage
269 Part four: From the protein village to the cosmopolitan city

that he once rehearsed for hours and hours,
until he had it in his fingers – no brain
required.
We have been talking about one of his main
interests, a new style of science called systems
biology. The expression is a fad, so widely used
these days that it seems to mean everything
and nothing. Nikolaus has a firm idea of what
it means; it has been guiding the way he has
put his group together. He says that his ideas
and approach have been heavily shaped by his
experiences in the United States, where he
worked for nine years after obtaining his PhD
from the University of Cologne.
But a systems approach to science requires
more than one group, and he has a bigger
vision of things; while getting settled in, he
proposed unifying the work of laboratories
and medical specialists throughout Berlin
into an institute that will be devoted to systems
biology.
Today he’s excited because he just received
word that the vision will soon take on very
concrete form. The German Ministry of
Education and Research (BMBF) had
announced a major new program to fund
projects in research and innovation in the
new German states. It was an opportunity not
to be missed. Nikolaus and the directors of the
institute put together a joint proposal on
behalf of the MDC and the Charité and submitted
it. Now he has been told that the
BMBF has pledged 7.5 million Euros of support
over the next three years. The idea is no longer
just virtual; it will have a real roof and walls.
“Over the last 50 years, molecular biology has
been breaking things down to their most fundamental
level, learning what the units are
and how to control them,” he says. “Now we’re
starting to make a serious attempt to put
things back together. That doesn’t only mean
figuring out how proteins work together in
networks or machines. That’s part of it – to
see the complexity of interactions at the scale
of molecules. But molecules are embedded in
Part four: From the protein village to the cosmopolitan city
structures of larger scales – cells, tissues,
organs, animals, even the environment.
Ultimately all of those levels work together to
read the information in a genome and turn it
into an organism, with a lot of input from the
environment. They also work together to
determine whether you’re healthy or sick,
whether over the long term you’ll develop
Alzheimer’s – or whether you can learn to play
the piano.
Until very recently, Nikolaus says, technology
has generally restricted a biologist’s view to a
particular slice of things. “We have methods
to look at the structures of molecules, which
is the smallest scale; then you jump up to the
level of biochemistry, which tells you whether
proteins are interacting with each other.
Another jump gives you the light microscope
view of the world, which is good at watching
major cell structures. Other techniques are
used to examine larger structures such as tissues
and organs. Eventually you reach the
domain of the doctor and the surgeon.
“Most scientists now have access to highthroughput
tools that give them a very broad
overview of what’s happening at a particular
270
level, like the DNA microarrays we use in deep
sequencing,” he says. “It’s giving them the
feeling that they’re looking at whole organisms.
But obtaining a snapshot of the entire
set of RNAs in a cell isn’t in itself systems biology,
unless that information can be woven
into higher levels. And it only gives you a small
piece of the picture of what is going on at the
molecular level.
“One of our long-term goals is to be able to
zoom in and out at different scales and to
make realistic predictions like, ‘if we change
one letter of the genetic code, what’s likely to
happen at all of those larger scales?’ Or, ‘if a
person takes this drug, how will that find its
way into cells, change biochemical processes,
and then work its way back up to influence his
overall health?’ Or, ‘what does this disease in a
mouse tell me about something similar that
happens to humans?’”
The marriage between laboratory and computational
science has already paid off in the
projects on microRNAs: turning the most
basic level of information in the genome –
tiny bits of DNA sequence – into crucial
insights about immunity and other systems
in the body.
Another recent project, Nikolaus says, makes a
good example of how everything comes
together. “We’ve known that microRNAs act in
two ways to block the synthesis of proteins,”
he says. “One method is by triggering the
destruction of their target RNAs. In the
other way, the RNA stays around, but it’s prevented
from being used to make proteins.
How do you tell which RNAs are being handled
in each way? You can’t do it molecule-bymolecule,
because you don’t know what the
targets are.
He teamed up with Matthias Selbach, who
can quantify protein levels of thousands of
genes using the latest advances in mass spectrometry.
In addition, by changing the procedure
used to make measurements, they could
for the first time quantify how many proteins

Catherine Adamidi

were synthesized after activating or repressing
microRNA in human cells. Three months
of pure measurement time and a year of computational
and experimental follow-up analyses
were carried out by PhD students Bjoern
Schwanhaeusser from Matthias’ lab and
Nadine Thierfelder from Nikolaus’ group, with
Minnie Fang, also in Nikolaus’ lab. They arrived
at a detailed picture how cells react, on the
proteome, transcriptome, and sequence level,
to microRNAs.
The result, Nikolaus says, is the first ever picture
of how proteins in cells respond – in
totality – to changes in levels of an miRNA.
“We saw that raising levels of a particular
miRNA has an effect on the protein production
of thousands of genes,” he says. “That
was interesting in itself, but the way we had
set up the study gave us answers to some
other questions that have been very hard to
answer. For example, we knew that most
miRNAs don’t work like an on-off switch, com-
Part four: From the protein village to the cosmopolitan city
pletely shutting down a protein. Are they
more like a volume control, adjusting how
much gets produced? And we knew that
miRNAs worked in these two ways – by causing
the breakdown of their targets, or blocking
translation. But it was completely unclear
how widely this second mechanism was used
by cells, and how important it was in volume
control.”
I’m suddenly reminded of the piano again.
Once you’ve played a note you can’t take it
back, but you can dampen it or change its
sound by pressing one of the three pedals.
Translational control was like a new pedal;
Nikolaus wanted to understand how it
worked and when it was used as the cell
played its daily recitals.
Answering his questions required a way to
monitor the behavior of the cell’s entire set of
RNAs and proteins as the activity of an miRNA
changed. “We raised the production of an
miRNA and studied its effects. The method
272
allowed us to watch the cell’s response, in
terms of which proteins were being produced,
and at what amounts. Then we did exactly the
opposite – we lowered the levels of the
miRNA. It was like using a volume control; the
amount of the miRNA determined the
amount of proteins. We saw direct effects on
hundreds of genes, and indirect effects on
thousands.”
This gave Nikolaus and his colleagues the first
global look ever at the importance of miRNA
translation control in the cell. “Hundreds of
target RNAs were blocked at the step of translation,”
he says. “But the effects were relatively
week. Usually they cause a less than two-fold
drop in the protein output. So it’s a volume
control, but a dial that is stuck in the middle –
you can’t turn it all the way up or down.”
Is there anything special about the RNAs that
are controlled by blocking translation, as
opposed to those that are simply destroyed
when an miRNA docks onto it?

“The first type seems to happen most in the
cases of molecules that are to be secreted, or
inserted into cell membranes,” he says. He
may even know why. But he’s not ready to
explain, not today. “That’s for your next book.”
��
One of Nikolaus’ colleagues comes in and
he introduces us. It’s Catherine Adamidi,
a technician from France. “She’s in charge of
the planaria,” he says. “Want to see some
worms?”
“Oh, that’s good, I was just about to feed
them,” she says.
Catherine leads the way to the lab. The planaria
are slinking along the bottom of white
plastic boxes, half-filled with water. I peer over
the edge as she gives them fresh water.
They’re easy to care for, she says, but the containers
have to be kept clean.
“You can find them almost anywhere there is
fresh water, in ponds and in lakes,” she says.
“These come from a fountain in Barcelona. I
went wading.”
The largest of the worms are about the length
of my fingernail. They are thin and very flat,
with triangular heads and two eyespots that
make them look cross-eyed and slightly
ridiculous, like cartoons. The spots, she says,
are some of the most primitive eyes found in
nature: two cups lined with cells called photoreceptors,
which transform light into electrical
stimuli. A planarium usually swims
away from the light. They also don’t like to be
disturbed very much. When she jostles the
container, they curl up into tiny balls.
Some species of the flatworm reproduce
through sex, but most types create offspring
without the help of a mate. From time to
time, particularly when the worms are under
some sort of stress, they spontaneously split
into pieces. Each part grows into an entire
new flatworm.
“This ability to regenerate seems to come
from the fact that they have one type of stem
cell, and it’s able to do everything,” Catherine
says. “Adult humans, you know, have many different
types that are limited in what they can
become. If you had only one type, too, it would
need to be flexible enough to regrow any part
of your body, the way an embryonic stem cell
does.”
It’s no wonder that planaria are a favorite of
groups studying stem cells and regeneration.
Maybe human cells could be taught to behave
the same way, to regrow new organs when
they fell prey to injuries, old age, or disease.
This isn’t the only odd aspect of the flatworms’
lives. “If you starve them, they don’t
die,” Catherine says. “They shrink. And they
regress – they repeat their process of development,
only backwards. When you feed them
again they develop and grow.”
And then there’s memory. Planaria are simple
animals, but for their body size they have relatively
large brains and synapses. This means
they can be trained to do things like run a
maze, using light and mild electroshocks.
What happens if the worm is trained and
then cut in half? You’d expect the head half to
remember the maze, but interestingly, when
the tail grows a new head of its own, that
worm seems to remember the maze, too.
There isn’t much in the literature about this;
researchers have shied away from the topic
since the 1960s. Back then a researcher
named James McConnell tried to prove that
planaria could learn through cannibalism, by
eating the ground-up remains of other flatworms
that had learned the way through his
mazes. It turned out to be a case of “what you
want to see is what you get;” when the experiments
were repeated with proper controls,
the results could never be reproduced.
It’s comforting to find out that cannibalism
won’t make the worms any smarter.
Otherwise Catherine would probably come in
one day to find them all gone – except for one
very fat, very smart planarium in the middle
of the tank, plotting its escape. But I wonder
about memories stored in tails, in cells. The
planarium’s strange method of reproduction
means that all of the worms are clones,
regrown from the cells of some original
ancestor that was born a long time ago, not
too long after the origins of the first animals.
So they never really die. It’s strange to be looking
at an animal which, for all practical purposes,
is immortal. Even stranger to think that
it still might have tiny splinters of memories
of some of those former lives.
273 Part four: From the protein village to the cosmopolitan city

A rat in a tree and a
wolf on the loose
It was a bad year to be a British citizen traveling in the United States.
The colonies were waging war against England, and spies were as
thick as mosquitoes. In 1778 John Berkenhout, an English physician,
arrived in the state of Pennsylvania pretending to be sympathetic to
American independence. But by the fall he had been exposed as a spy
and was languishing in a Philadelphia prison.
Berkenhout was eventually released and lived to stake another claim to
fame. Like many physicians of his day, he was an avid naturalist, devoting
his free time to the search for new species of insects and plants.
Europe – and particularly Britain – was caught up in a fever of collecting
and classifying. This had started with the work of the great (and
slightly crazy) Swede Carl Linneaus, who invented a magnificent system
that catalogued the world of plants according to the structure of their
sexual organs. It turned out to be a brilliant insight, compared to other
attempts to classify species based on color or the shapes of their leaves.
(Linneaus had the advantage of knowing that plants reproduced sexually,
a fact which had only been discovered at the end of the 17th century.)
Linneaus’ scientific work was based on precise observations, but his

Norbert Hübner

prose was full of poetic rhapsodies with heavy sexual overtones. The
following passage, for example, comes from the opening of his dissertation
submitted to the University of Uppsala in 1730:
Words cannot express the joy that the sun brings to all living
things. Now the blackcock and the capercailzie begin to
frolic, the fish to sport. Every animal feels the sexual urge.
Yes, Love comes even to the plants. Males and females, even
the hermaphrodites, hold their nuptials (which is the subject
that I now propose to discuss), showing by their sexual organs
which are males, which are females, which hermaphrodites...
The actual petals of a flower contribute nothing to generation,
serving only as the bridal bed which the great Creator
has so gloriously prepared, adorned with such precious bedcurtains,
and perfumed with sweet scents in order that the
bridegroom and bride may therein celebrate their nuptials
with the greater solemnity. When the bed has thus been made
ready, then is the time for the bridgegroom to embrace his
beloved bride and surrender himself to her...
Ah, if only dissertations could be written that way today.
The tradition of mixing botany and sex was enthusiastically continued
by the Englishman Joseph Banks, who sailed with James Cooke to the
South Pacific as a naturalist. Bank’s amorous adventures with Tahitian
women became the talk of high-society London. (Less known was the
fact that the islanders were willing to trade sex for bits of metal, such
as nails, which led Cooke’s crew to remove so many nails from the ship
that it started to fall apart.) In any case, Bank returned to England and
began preaching the classification system of Linneaus. This influenced
generations of young English scientists, from John Berkenhout to
Charles Darwin.
One of the questions that interested naturalists was how species
spread from place to place. Berkenhout’s book Outlines of the Natural
History of Great Britain, published in 1769, attempted to provide an
answer for a species of rat that infested Britain. Believing that the animal
had been carried to the island 50 years earlier on Norwegian
ships, he named it the Brown Norway rat. As it turns out, he was
wrong; modern studies have showed that the animal probably originated
in central Asia and arrived in England much earlier. But the
name has stuck.
Part four: From the protein village to the cosmopolitan city
In the 20th century the Brown Norway and other strains of rats were
brought into the laboratory, where they have become some of the
most important modern organisms for medical research. Their value,
says Norbert Hübner of the MDC, lies partly in the fact that they are
more closely related to humans than the mouse is, and are larger and
easier to dissect. “There have been exquisite studies of the animal’s
anatomy and physiology,” he says. “Structures that are tiny and difficult
to examine in the mouse are much easier to investigate in rats.
That’s especially important if you’re trying to determine how genes
affect development or cause disease.”
276
Henricke Maatz, Judith Fischer
and Norbert Hübner

��
Heavy inbreeding in the laboratory has led to the development of
hundreds of strains of rats, many of which have problems that
closely resemble human diseases. Often the symptoms appear to be
the work of multiple genes. In nature, disease-causing genes brought
together in individuals usually scatter again as animals find diverse
mates. Dangerous combinations don’t entirely disappear, but they
remain rare. That changes if there is heavy inbreeding in a population.
Both parents are much more likely to carry disease-causing combinations
of genes and pass them along to their offspring.
All of this has made the rat useful in traditional medical research and
pharmacology, Norbert says. “It’s a great model to study how body systems
respond to diseases and drugs. Now we need to bring it into the
modern age. That means developing the knowledge and tools to
manipulate its genes as we do with the mouse.”
If that can be done, he says, it should give scientists insights into a
much more fundamental biological question. “There is an almost oneto-one
match between the genes of humans and rats, but we’re still
very different,” he says. “Small differences between genes must be
responsible for that. But which genes cause which differences? That’s
277 Part four: From the protein village to the cosmopolitan city

the question you ask if you want to compare two species. It’s also the
issue if you’re comparing two individuals. One has high blood pressure;
one does not. One responds to a drug; the other does not.”
Most biological processes are collaborations between many genes, so
most health problems are also likely to be due to combinations of particular
varieties of molecules. The best strategy to discover them,
Norbert says, is to find an animal for which many strains exist and for
which there are already good models of complex human diseases –
the rat is a perfect example. The next step is to carry out linkage studies
in the animals, to identify genes or blocks of DNA that are inherited
together in animals that have similar problems, and then to see if
those patterns appear in people suffering from the human form of the
disease.
Before that could be done with the rat, however, scientists needed to
get a look at the complete rat genome. This was accomplished in 2004
by an international consortium of researchers who finished the
genome sequence and analyzed it. Norbert’s lab participated. As other
labs collected sequences, the MDC group helped plug them into their
proper positions on the rat chromosomes. The process was like assem-
Part four: From the protein village to the cosmopolitan city
bling a puzzle with million pieces, each showing a tiny fragment of a
map of Europe. The end result was a complete, gapless genome
sequence.
With this in hand, Norbert and his colleagues were ready to begin
measuring diversity – in other words, to see how individuals and
strains differ from each other and the “norm” of the genome. Norbert
says there are two main kinds of variation.
“The first type is a simple ‘spelling’ difference between the same
sequences in two animals,” he says. “Often this involves just one letter
– two regions of the code are identical except for a single base pair. The
technical term for this is a single nucleotide polymorphism, or SNP.
Within a species, there are normally several possible spellings for each
gene – it usually contains a number of SNPs. Often this has nothing to
do with disease – it’s just the normal variation that makes each of us
unique.”
Other differences are more significant. The story “Waking a Sleeping
Beauty and other tales of ancient genes” tells how blocks of DNA
sometimes jump to new positions in the genome. The blocks may con-
278

Henricke Maatz, Judith Fischer and Katharina Grunz
tain just a few letters, a whole gene that is thousands of bases long, or
even longer strings of DNA. The positions of these blocks provide
another way of comparing and classifying rats. Animals with the
blocks arranged in the same way are said to belong to the same haplotype.
Finding out what we want to know about the rat genome, Norbert
says, means studying enough animals to discover the main haplotypes
and listing the SNPs found at various positions in their genetic codes.
The result will be a family tree or another type of diagram that shows
how all the strains are related. That would be easier if scientists knew
the origins of the rats first brought into the lab. But that is not the
case, and nor do lab records describe how the early animals were interbred
and inbred. Some of these questions can be cleared up by studying
SNPs and haplotypes
Wouldn’t you need a complete genome sequence from each individual
animal? I ask.
“That would be ideal,” he says. “In the future, as it gets easier and less
expensive to sequence DNA, it will be feasible. For people as well as
rats. Someday you’ll probably carry around a copy of your genome on
something like a credit card or memory stick and give it to your doctor
when you go in for a checkup. It will be helpful for diagnoses and other
aspects of your medical care.”
��
For the moment, collecting and analyzing sequences at a somewhat
lower “resolution” has given the lab enough to do. The rat
genome gave the group and their colleagues across the globe an outline
to begin cataloguing SNPs and haplotypes. Kathrin Saar, a ...
in Norbert’s group, headed the project – which involved coordinating
the work of 12 laboratories in Germany, the United Kingdom,
France, Japan, Spain, and the Czech Republic. Their findings will
go to press at about the time this book is printed.
The study revealed that the strains can be clustered into ten clades –
or ten main types. Nine of these are closely related, which means that
their ancestors were similar to each other – or they arose from a common
pool of animals. The Brown Norway rat turns out to be an exception,
a more distant cousin. It’s good to have such an “outlier”, Norbert
279 Part four: From the protein village to the cosmopolitan city

280
says, when you’re trying to understand the causes and effects
of diversity. “Just as if you wanted to get a feeling for human
variety, you might start by looking at people from one region,
but you would certainly want to include a few from far away.”
Originally Norbert hoped that the study would permit a
detailed reconstruction of the rats’ family tree, but that hasn’t
been possible. “This probably means that the first rats brought
into the lab were quite diverse, already coming from mixed
strains, and that they were interbred in the early years, mixing
things up even more.”
The project was able to plot three million SNPs, obtained from over
150 strains of rats, onto the genome. “That works out to about one
variation per every 800 base pairs,” Norbert says. “We were also
able to make predictions about the effects that roughly a tenth of
these variations probably have on cells and tissues. In 56 cases we
predict that the change has an effect on the function of a protein.”
Seven of the variations involve rat molecules that are closely
related to genes involved in hereditary diseases or cancer in
humans. One is ALDH2, which has been linked to acute alcohol
intolerance. Another is AFF4, which causes cancer if it doesn’t
function properly. Several hundred of the variations probably
affect the way cells cut and paste RNAs to allow different forms
of proteins to be made from single genes. Others affect regions of
DNA that control when and how genes are activated.
The study has permitted the lab to identify parts of the genome
that change faster than others. Every genome has such areas,
and sometimes they reveal features that are being rapidly
changed by evolution. Here they discovered an interesting parallel
to the mouse: a great deal of variety was found in genes
involved in the sense of smell. Scent is so central to a rodent’s
life that even small changes in the processes that govern it can
have a big influence on the animal’s ability to survive and pass
along its genes.
The haplotype map remains incomplete, but the data gathered
so far confirms something that Norbert has suspected for a
long time: a rat is not always a rat. “While the substrains can be
clustered into ten groups with similar genomes, they aren’t
nearly identical,” he says. “There’s still quite a bit of diversity. It’s
not that obvious if you look at the genome as a whole, but if
you look at regions that are changing the fastest, there is a lot
of variation between the substrains. In a rat known as the Long-
Evans strain, 29 percent of the base pairs in these areas have
different spellings than you find in other strains.
“This happens because until now, it hasn’t been possible for
labs to cleanly identify what type of rat they started out with,”

Norbert says. “Then when they breed the animals with other rats –
whose origins also aren’t clear – the offspring inherit genomes that
are unique to each place. It means you have to be careful when you try
to extend what you’ve found in one animal to others. A genetic problem
may be unique to one strain because it has a particular genetic
makeup. You may not be able to generalize to another strain of rat – let
alone a human being. These projects should help labs ‘clean up’ their
strains. At least now we have some standard reference points that can
help you define the animal you’re working with.”
Norbert quickly points out that all of this work is descriptive.
“Successfully linking a SNP or haplotype to a disease is only the starting
point,” he says. “The next step is to determine why one form of a
gene – or a set of genes – lead to disease. That requires like conditional
knock-outs that remove the gene in one tissue at a specific time.
While those tools are well-developed in the mouse, most of them
don’t yet work in the rat. There’s a lot of catching up to do.”
��
Norbert appeared in “The case of the short-fingered muskateer”
for his role in helping Friedrich Luft’s group look for the genetic
causes of brachydactyly and hypertension. That story shows how difficult
it can be to pin down the causes of a disease, even when the problem
seems to lie in a single region of DNA. With multiple genes, the
task becomes staggering. I wonder what has motivated Norbert to
tackle one of the hardest questions in today’s science.
“I don’t really know,” he says. “I like complicated things. But my being
here right now, doing this, is a bit of an accident.”
In 1989 he was a medical student at the University of Heidelberg. He
was looking for a part-time job in a laboratory and was hired on as a
technical assistant by Detlev Ganten, who would soon move to Berlin
to fully dedicate himself to the creation of the MDC. At about the time
Norbert finished his first degree, another member of the group named
Klaus Lindpaintner was getting ready to move to Harvard in the USA
to set up his own group. He wanted to hire good people and invited
Norbert to go along. So that’s what he did.
After two years of work on his doctoral thesis at Harvard, Norbert
moved back to Germany, this time to Berlin, to finish his medical studies
at the Charité. That made him one of the few people to have
earned both an MD and PhD. It also made him just the right type of
person to find a job at the MDC – once again, he laughs, in the group
of Detlev Ganten. This time as a postdoc.
In 2003 he was given his own group, followed a year later by his
appointment as director of the MDC’s “Gene Mapping Center.” Since
2005 he has held a Professorship in Medical Genomics and Genetics at
the Charité.
One red thread that ties together all these phases of his life has been
the rat.
“Some complex processes are hard or impossible to explore in the
other models we have,” Norbert says. “For example, mice don’t
develop arteriosclerosis. We don’t have mouse models of natural types
of hypertension or heart insufficiency. It may be that the tiny anatomy
just makes some of these things hard to find – on the other hand,
something in mouse biology may protect them. Another thing is
that when the mouse was brought into the lab, it was mainly used
to watch development. The rat, on the other hand, has been t
he favorite in looking at human disease. These two ways of looking
at an animal are deeply connected, and we need to bring them to -
gether.”
Norbert admits that he and his colleagues are just opening the book
on diversity in rats and their link to human disease. But the first pages
of that book have already shown how an intimate knowledge of rat
strains and their defects can be a powerful tool in establishing such
connections.
“One of the problems in human disease is that it’s not only genetic
factors which are intertwined – so are the symptoms and various body
systems,” he says. “Think about one of the most common causes of
death in humans: heart failure. That’s somehow connected to other
things, like high blood pressure and hypertrophy (enlargement of the
heart). High blood pressure is one of the major risk factors for heart
failure. But is it the cause? Are both things caused by the same genes?
Or are they separate problems, influenced by different sets of genes?
In some patients they all seem to be connected, one symptom following
on the heels of another. But other people suffer from hypertension
and hypertrophy without suffering any of the clinical signs of
heart failure. So we may actually be talking about different diseases.”
One way to answer the question, he says, is to look at available rat
models of disease and try to peel the problems apart by studying different
strains. That’s what Jan Monti and Judith Fischer, two members
of his group, have been up to with colleagues from the MDC and
Charité, from the pharmaceutical industry, and from other institutes
in the UK and the USA.
They began with a strain known as the spontaneously hypertensive
heart failure (SHHF) rat, which develops a disease that mirrors most of
what is known about human heart disease linked to high blood pressure.
“The animals not only develop heart failure after high blood pressure
and hypertrophy of the heart,” Jan says, “but when you look at the
activity of their genes, you discover patterns that strongly echo what
happens in human tissue.” Another strain, the stroke-prone spontaneously
hypertensive (SHRSP) rat, has similar problems of high blood
pressure but does not suffer from heart failure.
281 Part four: From the protein village to the cosmopolitan city

Patterns of heredity showed that the symptoms were based on genes.
“That meant that if you mated the two types of rats, you would get different
types of offspring,” Jan says. “Some would have all of the problems
of both parents, and others would have various combinations.
We thought we could use this to peel apart the genes and their symptoms
and try to figure out their connections.”
At the same time the lab crossed the SHHF (heart failure) rats with a
control strain that had normal blood pressure. This would lead to different
sets of genes and symptoms in the offspring – some of which
would still suffer heart failure. If the strategy worked, Jan and Judith
would have a variety of different animals with different sets of genes.
It ought to be possible to find the molecule or molecules responsible
for heart failure, and to see whether the disease only happened in animals
that also had high blood pressure.
The laws of genetics and the behavior of dominant and negative
genes meant that the researchers had to wait two generations to get
the right animals. They sorted the offspring into groups based on
blood pressure and other factors, closely watched their health, and
began looking for the common denominator between heart failure
and DNA.
While the rats’ health problems were made worse by high blood pressure
and the enlargement of the heart, some mice experienced heart
failure without the other symptoms. This hinted that a separate gene
was at work. The scientists pinpointed the heart failure defect to a
region of the rat chromosome 15, and eventually to a gene called
Ephx2. The laboratory sequenced the gene in the animals and found
several variations – SNPs – that would change the way cells made proteins
from the Ephx2 gene. That in itself was promising. But things
really got exciting when the scientists found out what the gene did.
Part four: From the protein village to the cosmopolitan city
“A search of the literature showed that Ephx2 works in the heart,”
Norbert says. “It modifies a type of hormone called EET found in cardiac
muscle. EETs play a role in relaxing blood vessels and they also step
in to help protect the heart after some types of heart attacks. When
we checked levels of these molecules in SHHF animals, we found high
amounts of Ephx2 and significantly lower amounts of EETs.”
All the evidence suggested that the researchers had found a new
heart failure gene, but there was more to do. If Ephx2 was the culprit
in heart failure, they might be able to protect the animals by shutting
it down. But that couldn’t be done in rats.
Norbert sighs. “Here was the classical case – a model for human disease
that was working in the rat, but we didn’t have the genetic tools to
follow up.”
The only alternative was to turn to mice, which have a very similar version
of Ephx2 that seemed to have the same functions in the heart.
But switching model organisms always makes things a bit complicated.
In this case, there was no such thing as a spontaneously hypertensive
heart failure mouse. The researchers would have to delete the
gene in healthy mice and then find a way to simulate the type of heart
damage seen in rats.
Removing the gene caused no obvious problems for the mice – they
had normal blood pressure, and their hearts were the usual size. But
differences appeared when the animals were given a drug that caused
the muscle cells of the heart to beat irregularly.
Normal mice experienced symptoms resembling a heart attack.
Things were different in the animals without Ephx2.
“It was much more difficult to cause an irregular heartbeat,” Norbert
says. “That was true when we used drugs, also with the use of electrical
stimulations. Even when you subjected the heart to irregular electrical
pulses, it kept beating on in a steady rhythm.”
One more thing had to be done to prove that the case had any connection
to human heart disease. The scientists needed to see how Ephx2
behaved in human patients. They obtained samples from the cardiology
units at the Charité to check whether there was anything unusual
about the protein.
“Jan and Judith found that Ephx2 levels were lower than usual in
patients recovering from heart failure,” Norbert says. “That makes a lot
of sense if you realize that EET hormones are needed for the recovery
process. What you would want to do in that case is lower the amount
of Ephx2 – so that it stops repressing the activity of EETs.”
This fits with what scientists believe about the functions of the hormones,
Norbert says. They seem to control channels that let charged
atoms pass into and out of heart cells. That flow is the major factor in
coordinating the activity of billions of cells, making them contract at
282

the same time to create a heartbeat. It also fits neatly with a recent
study in which drugs were used to control the activity of Ephx2 in mice
with enlarged hearts. A group at the University of California-Davis
showed that treatment with molecules called sEH inhibitors could
prevent the condition from developing. An unexpected side effect of
the study was that the same procedure protected the mice from
arrhythmic heartbeats.
Now Norbert – and the rest of us – know why.
��
If you’ve been waiting for the wolf in this story – the one in the title
– it’s probably not what you think. Norbert hasn’t proposed using
wolves as model organisms in the search for disease genes. At least
not yet. In this case the wolf is a disease called Systemic lupus erythematosus,
or SLE, or simply lupus.
The disease was first described in the Middle Ages, and there are various
hypotheses about why it was given the Latin name for the wolf.
One explanation is that patients often develop a reddish rash across
their noses and cheeks – supposed to resemble either patterns of fur
on the face of a wolf, or scratches or bites made by the animal. These
are the mildest symptoms of the disease.
Sometimes the survival of the body calls for the sacrifice of some of its
own cells, such as those infected by viruses. White blood cells track
them down and prompt them to self-destruct in a process called
apoptosis. If for some reason healthy cells become targets, the result
may be an autoimmune disease such as lupus (systemic lupus erythematosus,
or SLE). Lupus is dangerous because it can spark spontaneous
inflammation anywhere in the body, eventually leading to permanent
tissue damage and death. Most often affected are the circulatory
and nervous systems, joints, and the skin. That’s why the unusual
rash often develops on the face.
Norbert’s group became involved through a collaborator. Min Ae Lee-
Kirsch, a clinical researcher at the Technical University of Dresden, was
investigating a family that suffered from a rare form of hereditary
lupus. Working with Norbert’s lab, she carried out a linkage study that
revealed defects in a gene called TREX1.
That made them wonder whether other types of lupus might involve
mutations in TREX1. They examined samples provided by clinical partners
throughout Europe and found changes in the genes of patients
suffering from SLE, the most common form of lupus. None of the samples
from non-lupus patients had these types of mutations.
Why should defective TREX1 cause an auto-immune disease?
“The gene encodes a protein that can enter the nucleus and chew up
the cell’s own DNA – one of the steps of normal apoptosis,” Norbert
says. “Normally it doesn’t do so because it is tethered to structures
outside the nucleus. It stays there until it’s time for the cell to die. And
there’s evidence that it may have a second job; it may act as a sensor
that monitors the cell for DNA brought in by viruses.”
The mutation seems affect the part of the protein that ties it to its
duty station, so animals with defective TREX1 may not be able to hold
it there. The molecule escapes and spreads to other places, including
the nucleus. Healthy cells die, releasing DNA and other molecules from
the nucleus at the wrong time. The immune system misinterprets
them as foreign and builds antibodies against them, which then leads
to attacks against healthy cells.
TREX1 is not responsible for all cases of SLE. Defects in other DNA-digesting
proteins have also been linked to the disease. But the study may
explain doctors’ long-standing suspicion that some cases of lupus are
triggered by viral infections. If TREX1 really is a DNA sensor, it may
detect viral molecules and launch its own defense by killing the cell.
Because this is not the normal way the body becomes aware of an
infection, the result may be an autoimmune disease.
Norbert’s office is on the ground floor of the Medical Genomics building.
The floor-to-ceiling windows on the north side look out onto a
grassy field and a stretch of woods. He peers outside, lost in thought,
and I instantly know what’s on his mind.
“Wolves,” Norbert says.
“Don’t even think about it,” I say.
“Dogs do make excellent models for human diseases,” he muses. “And
there’s definitely enough space on campus...”
Don’t even think about it.
283 Part four: From the protein village to the cosmopolitan city

Walter Birchmeier,
Scientific Director of the MDC
What makes the Max Delbrück Center unique?
For 15 years the MDC has been devoted to doing excellent science
around a concept of molecular medicine, primarily in collaboration
with the Charité and the universities of Berlin, more recently with the
arrival of the Leibniz Institute for Molecular Pharmacology on the
campus. What is happening now in science makes this a very fertile,
attractive theme for basic researchers and clinical scientists, and the
campus is appealing because of our experience.
I think we handle interdisciplinarity in a unique way – our researchers are
not typically limited by borders of technology, models, or model systems.
Our scientists have been exploring new model organisms such as the
naked mole rat, which Gary Lewin has introduced in order to get a grip on
sensory mechanisms in the nervous system, and Nikolaus Rajewsky has
brought in planaria, which he is using to look at microRNAs and issues
related to regeneration. Alongside the very classical systems of the mouse
and the fly, we are also doing original work with more established organisms
such as zebrafish and the rat. There genetic tools have lagged
behind, but groups like that of Norbert Hübner have been doing some
creative work to try to overcome those obstacles.
This willingness to branch out into new territory allows researchers
such as Thomas Willnow, whose original focus was on the cardiovascular
system, to make a foray into a neurobiological theme when a
knockout animal takes him there. Or scientists like Claus Scheidereit,
who began as a classical biochemist, to make real contributions to our
understanding of the progression of specific types of cancer. Carmen
Birchmeier is co-head of our Neurodegenerative Diseases program,
Interview
but her work has increasingly taken her into muscle development. And
the context of the MDC has led to valuable collaborations between my
own lab and clinically-oriented groups such as those of Ludwig
Thierfelder and Peter Schlag. It’s a very satisfying feeling for a cell biologist
to participate in work that has a clinical relevance. These are only
a few examples that I could give. There are many others in this book.
One unique aspect of our existence is the generous funding that we
receive through the BMBF and the State of Berlin. It allows us to concentrate
on science – mostly! – and helps a great deal in making outstanding
recruitments. The MDC can push for a high quality of science,
and our regular external evaluations have confirmed that we seem to
be on the right path.
Obviously there are many more unique features to the MDC, including
the historical and cultural aspects of the campus and its situation in
Berlin, which I think are well represented in this book and some of the
other publications we have produced over the years. Our staff have
taken the initiative in writing about the history of the campus and science
in Berlin – one publication has just come out on some of the
great figures in genetics that have had a connection to Buch. There
have been other books and pamphlets on the campus’ many artworks,
biographies of the scientists who have worked here, and even the
campus flora. Then there are our activities with schools, and a very
strong campus participation in the Long Night of the Sciences...
There have been tremendous changes in science over the past
15 years. How has this affected what people mean by
molecular medicine and how it is practiced?
284

Stefan Schwartze and Walter Birchmeier
One huge change has come through the explosion of new technologies,
including high-throughput platforms that have moved us from a
very narrow focus to a much broader one. When I started at the MDC,
the norm was for one professor to work on one gene. Now, potentially,
everyone can work on every gene. That’s a danger because to make
progress you need to focus. Everyone is solving this problem in his own
way. In the case of our lab, we have concentrated on Wnt/β-catenin
and Met-Gab1 signaling. Claus Scheidereit has focused on NFκB;
Thomas Jentsch on channels and transporters... I could go on.
In this environment some traditional areas of focus no longer really
make sense. There have been discussions about whether institutes, for
example, should focus on one disease type – whether the German
Cancer Research Center should “take over” cancer, while we concentrate
on cardiovascular diseases. Well, everything we are learning
shows that common pathways and mechanisms are implicated in
embryonic development, cancer, autoimmune diseases, and so on.
Understanding a disease will require looking at a process from many
different angles – and understanding a process will require looking at
the role it plays in making a wide variety of phenotypes.
As well as broadening our themes, labs have expanded the repertoire
of methods and techniques that they use on a daily basis. We have so
many more means at hand now to perturb and thus understand bio-
logical systems. It’s amazing to remember that the knock-out technologies
that we use in virtually every paper are relatively new.
Conditional knock-outs are newer than that, and now at the frontier
are small interfering RNAs. We have also built a platform for smallmolecule
screening in which we actively seek inhibitors for the pathways
and processes that we are working on. Those are crucial tools as
you develop animal models of disease and want to manipulate them,
as well as in the search for new drugs.
Taking advantage of these technologies means that groups have to
integrate new types of expertise – mastering mass technologies,
learning to cope with data on a genomic scale, working with bioinformatics
and new ways of modeling processes. Not every group can do
everything, which means we need to flexible about the question of
whether platforms are maintained centrally or by groups. The MDC
has this flexibility, and here once again the context of the Helmholtz
Association is very helpful.
An explicit mission of “molecular medicine” raises hopes that
there will be some fast translation of basic science into therapeutic
tools. Sometimes a lot of “hype” is generated around a
particular approach – for example, gene therapies. There have
been a few attempts at trying out virally delivered gene therapies
in human patients, and some have led to unexpected
285 Interview

Laying the cornerstone of the new building to house the 7-Tesla MRT machine. Left to right: Prof. Ernst Otto Göbel, Prof. Walter Rosenthal, Dr. Peter Lange, Dr. Hans-Gerhard
Husung, Prof. Walter Birchmeier, Dr. Siegfried Russwurm.
side effects or have been disappointing in other ways. On the
other hand, your own work has contributed to new diagnostic
tools that have saved lives. Do you see “molecular therapies”
becoming common tools in the treatment of patients
anytime in the near future?
Molecular approaches have already been extremely valuable in diagnostics,
and I’m convinced that molecular therapies will become common
tools in medicine. Yet we have to be humble and do things we
can. It’s worthwhile to screen small-molecular-weight compounds
with the FMP, but you have to be realistic – only one in 500 lead compounds
make it into the clinic. On the other hand there have been
some impressive successes – Herceptin from Genentech/Roche,
Gleevec from Novartis. Many more are in the pipeline: antibodies, vaccines,
therapeutic proteins, and therapeutic cells.
You have to beware of hype and fads. They come and go, and I’m
always a bit relieved when they go. They’re often associated with unrealistic
expectations and pressure to apply technologies that may not
be ready. Stem cells are a good current example. It’s obvious that they
Interview
have a powerful potential – that doesn’t mean they can be injected
right away in a desperate measure to cope with neurodegenerative disease.
That would be dangerous – in animals stem cells operate operate
in very controlled environments, in small “stem cell niches.” Their behavior
in other contexts is unpredictable and may be dangerous.
What are the most important things clinicians need to
understand about basic research?
That it’s hard work and takes time to understand a biological problem.
Although there is still a gap and a difference in mentality, I must say
that we have some excellent MDs at the MDC: Ludwig Thierfelder,
Bernd Dörken, Friedrich Luft and others. And many excellent young
ones. I hope that the ECRC will attract very high-quality new MDs,
both seniors and juniors.
What do you consider to be the main bottlenecks and
challenges for the campus in terms of achieving its ambitions
in molecular medicine?
Basic science is doing well. Many of the collaborative projects could be
improved, and we need to start new projects. All of the health-related
286

Helmholtz Centers do molecular medicine now and have “translational”
centers. So to stand out we will need to be more successful, and
develop new projects. But we are in an excellent position. Things are
moving forward with the ECRC, and just this week we have taken a
major step forward with plans to create a new institute in Berlin for
systems biology. I could also envision a major push in the direction of
stem cell biology. What we have to learn in this field is still immense –
on stem cell niches, on cancer stem cells, on the pathways that control
the various participating cells.
Institutes become marked by the character and style of their
directors. You are only the second director of the MDC, having
taken over the reins from Detlev Ganten, who was obviously
very strongly associated with the MDC. How have you put
your own stamp on the MDC? What do you think your legacy
will be?
I was an intermediary, an internal appointment for five years, and during
a lot of this time a search for a new director has been going on.
That has limited my maneuvering room. I think that if I weren’t a scientist,
the situation would have been very frustrating. But this situation
has given us the chance to search intensively for an excellent new
person to direct the institute for the future, someone with experience
in the outside world, an excellent, widely respected scientist, who
understands molecular medicine.
One of my own goals was to bring more outstanding scientists to the
MDC. For a while in the 1990s, this was a concern. I think here we have
had excellent success in the recruitment of Thomas Jentsch, Nikolaus
Rajewsky, Norbert Hübner, and Gary Lewin, as well as youngsters like
Matthias Selbach, Oliver Daumke, and Ines Ibanez-Tallon. I could list
many others.
287
I also wanted to push for excellent science, and the recent review of
our proposals for the next five-year funding period was extremely positive.
All of the programs had improved – in some cases very significantly
– over the past five-year period. Overall, the MDC will certainly
rank among the best – and will possibly have the best overall rating –
of all five of the Helmholtz health-oriented centers. I am particularly
pleased about the performance of the program on neurobiology,
which had the highest rating of “seven” in the review. One consequence
will be that nobody will try to restrict the institute’s focus to
only cardiovascular disease.
What advice will you have for your successor?
To keep practicing his or her science.
How will you spend all of your “new free time”?
To rebuild my own lab. We have had some nice stories – Jolas’ work on
wound healing, Martas’ work with NOMA-GAP, Ute’s Gab1 mutants,
Klaus’ work on a Shp2 inhibitor, Dietmar’s studies of β-catenin in the
brain, and the role of Wnt and Bmp in heart development from
Alexandra. Getting such papers published in the best possible journals
is vital to these young scientists’ careers, and it’s something that I can
help with, if I have the time.
What have been the hardest decisions you have had to make
as director?
To tell to some of our young group leaders that they will not be
tenured. It’s heartbreaking, particularly since some of them have been
so outstanding previously.

Walter Rosenthal,
Director of the Leibniz Institute
for Molecular Pharmacology
What makes the FMP unique?
The FMP is unique partly because of its scientific vision and the people
and tools we have assembled to accomplish that vision; it is also special
because of the way in which we operate as a member of the
Leibniz Association, which gives us a great deal of scientific and organizational
freedom.
I came to the FMP in 1996. Like many other scientific institutes that
had formerly operated under the GDR, the FMP was still in a process of
evolution. Before the “Wende” it had been called the “Institut für
Wirkstofforschung.” At that time there had been a good mixture of life
scientists and chemists among the staff, because it had worked in conjunction
with pharmaceutical companies. When a scientific council
was appointed to guide the formation of the FMP, they originally
wanted to make it into an institute with a focus on cell signaling. But
the major development that we started in 1996 was to significantly
expand structural biology within the FMP. My predecessor had already
begun to push this area. Our vision was to establish a strong program
in structural methods centered around NMR. That has worked out very
well: scientists of the FMP have made very important contributions,
not only in NMR applications, but also in the development of methods
like solid-state NMR.
In a subsequent phase, we wanted to strengthen the chemistry activities
again. Michael Bienert was here, doing peptide chemistry. To that
we have added a Medicinal Chemistry group focusing on combinatorial
chemistry and the Screening Unit. Soon we’ll have a new group
here that makes artificial proteins. What we have achieved is to estab-
Interview
lish a good mixture of cell biologists, signal transduction people,
chemists and structural biologists. This mix represents a potential we
should use to work on rather difficult projects. For example, several of
our groups are working on membrane proteins such as receptors and
channels; these are among the most important areas where structure
elucidation and drug design are concerned, yet they are also among
the most difficult. Finally, we now have groups working on
protein-protein interactions, hoping to develop approaches to interfere
with them pharmacologically. A research group funded by the
German Research Council (DFG) has now been established with that
focus.
We work without restricting ourselves to particular organs or diseases.
The ideal projects require experience from groups from all the
subdisciplines represented here at the FMP. Teams from these areas
already work well together under our roof, but many scientists are still
unaccustomed to working this way. Our aim is that everybody in the
house – from PhD students to established scientists – should take
advantage of the resources of the entire institute. When we hire we
look for people who will be able to do that. And interdisciplinarity is
also a hallmark of our new Leibniz graduate school, which has just
been established thanks to an application submitted by Bernd Reif. It
has been set up so that the students get real exposure to a spectrum
of platforms and scientific themes. It’s a good mirror of our overall
?concept.
You have said that institutes like the FMP are evolving a new
role in the process of drug design...
288

Today we have this vision of a “rational” drug design which will operate
not on a basis of blind trial-and-error, but on our knowledge of
how molecules are structured and how they behave. The plan is to find
better and easier ways to identify good targets for drugs, to quickly get
a look at the details of their surfaces, and to engineer artificial molecules
that will change their behavior. Achieving this will require a great
deal of work collecting basic knowledge about the structures and
functions of molecules and processes in the cell. Large pharmaceutical
companies will not invest in such high-risk basic issues, although it’s
clear that this type of research is essential to drug design. It is better
carried out in an academic setting. So we’re moving towards new
types of partnerships between these communities.
Science has been changing rather dramatically over the past decade.
Most research groups have access to a wide range of technologies that
have enabled them to bring their work much farther along the road to
applications. But there are still some important gaps. For example, as
far as I know, there are no open platforms in Germany where academic
researchers can carry out screens for small molecule inhibitors of
the molecules they are interested in. They might be looking for something
that can be developed into a drug, or they may simply need
some sort of a probe that can be used for basic investigations of the
cell. They will only get access to a screening platform by paying for it –
which is usually too expensive – or by getting a company interested in
the project. In most cases that’s unlikely to happen.
We’re helping to fill that gap by setting up a major new screening unit
that will cater particularly to academic users. This will give a boost to
their basic work; in the case of those interested in drug design, they
will be in a better position to hook a pharmaceutical company. They
may leave the facility with several hits – compounds that bind to
potential drug targets. There are several more steps that have to be
289 Interview
Walter Rosenthal

taken after that, including optimizing the compounds and testing
them in cell culture systems or laboratory animals. The chemistry
groups we have put into place can help there. Again, the farther along
a group brings a project, the more likely they are to catch the interest
of a company.
We’re working in a context where, over the last 20 years, Germany has
steadily lost its position as the world’s major producer of pharmaceuticals.
We hope that the platforms and expertise we have assembled
will make the region attractive to those who want to invest in
Germany and Berlin.
How does the FMP fit into the research landscape in Buch and
Berlin? In Germany? In Europe?
Going back to the 1970s, there was the Institut für Wirkstofforschung,
the IWF, a part of the GDR Academy of the Sciences, which was housed
in Berlin Friedrichs?felde. After the reunification scientific reviewers
came in to evaluate its activities and to think about its future. They
were impressed by some of its hallmarks, especially the interdisciplinarity.
There was already a tradition of working with animal models,
doing chemistry, research into signal transduction, and clinical pharmacology
work. There was even a small company alongside the institute,
a unique “academic industrial complex.” That type of combination
was a new approach for the West. The recommendation of the
review was to transform the institute into the FMP, with a focus of
molecular pharmacology. It was opened in January, 1992.
It was clear that Friedrichsfelde wouldn’t do over the long term; the
old buildings weren’t adequate for our needs. It was very difficult to
set up cell biology labs there, for example. The only place big enough
to house our NMR machines was a big open building with a huge iron
beam running across the ceiling. When asked why
it was there, they said it had been used to hoist dead elephants or
something, as once the buildings were used by a veterinary institute
that worked for the animal park nearby.
The move to Buch was a very good decision. We have very strong partners
on the campus that we can support and get support from, and
are working closely with them on several very interesting projects. Our
most important partner in Buch is the MDC. The research concepts of
the two institutes complement each other: while the molecular medical
research at the MDC is particularly dedicated to diseases or clinical
symptoms and their molecular explanations, the FMP investigates
the functional and structural characterization of proteins as well as
the development of strategies for influencing them pharmacologically.
The close connection to the MDC extends to the organizational level.
Thus, large equipment is shared and jointly operated. Guest scientist
contracts make it possible for scientists of one institute to use equip-
Interview
ment belonging to the other. Both establishments send representatives
to important committees of the other. The planning of costly and
long-term research projects as well as the appointment of leading scientists
takes place in joint agreement. The MDC and the FMP arrange
and finance joint events for those studying for their doctorates.
In terms of Germany, the FMP is a member of the Leibniz Association,
a national research organization similar to the Max Planck Society, the
Helmholtz, and the Fraunhofer Associations. Leibniz institutes retain a
great deal of independence but also profit from the association. They
perform problem-oriented research of national interest and strive for
scientific solutions for major social challenges.
Another hallmark of Leibniz is the fact that we promote very strong
links to the university system. At the moment the FMP has six full staff
with joint appointments (three full, two associate and one honorary
professorships).
One particular feature of Leibniz is that its institutes are not guaranteed
a permanent existence. Every seven years, the FMP – like every
other Leibniz institute – undergoes a thorough scientific evaluation.
The question is whether the institute is needed – whether it is fulfilling
its mission, and whether the same scientific work could be accomplished
without it. This challenges us to perform at a very high level
and to be responsive to the shifting landscape of science. In some
cases, of course, institutes are closed – or they leave the organization
and find another mode of funding. But new institutes are joining. Very
soon we will add a research laboratory in Berlin whose focus is
rheumatology, and another new member will be the Berlin Natural
History Museum.
This system also allows for expansion of institutions that are doing an
excellent job. Despite a period of financial problems in Berlin, we
haven’t suffered – just the opposite. We have received budget increases
and hope for more, particularly in investments into NMR and other
major equipment. Recommendations for growth and other types of
changes arise from the evaluations. We also have had quite strong
support from the scientific council in achieving our plans.
As far as our relationships to Europe are concerned, the FMP has participated
in several European research networks over the last years.
Since October 2006 Enno Klußmann has been the first FMP scientist
to coordinate a project funded by the European community (STREP
“Identification of the therapeutic molecules to target “compartmentalized
cAMP signaling networks in human disease (thera-cAMP),”
under the 6th framework programme).
Facing the European deficit in Chemical Biology, the FMP coordinates
activities in the field of small molecule screening. One of these activities,
the infrastructure initiative “EU-OPENSCREEN”, will provide access
to screening platforms for european academic groups. Thus, local
290

expertises of research institutions like the Screening
Unit of the FMP can be used throughout the entire
European research area.
What’s life like for scientists here?
We have junior and senior groups. As of November
2006, we standardized the system so that junior
group leaders are not initially hired into tenure-track
positions. They are very well funded for an initial period
of five years; that can be extended for another four
years. All of our junior group leaders have gone on to
find positions, becoming chairs of departments or
receiving C3 professorships. About two-thirds of our
alumni move into academia; the remaining third go
to work for drug and biotech companies.
The fact that we use tools similar to the pharmaceutical
industry means that people who receive training
here should be interesting candidates for biotech
companies. To promote this link, the campus has set
up a summer school – “From target to market” – to
help prepare postdocs and predocs for careers in
industry. The next course will be offered in September.
What do you consider to be the most significant
bottlenecks for the institute and the field
in general?
Right now I would like to add a team of synthetic
chemists who can work on hits which we have identified
in our Screening Unit. There is also a general bottleneck
in academia when it comes to support for cell
biologists and biochemists to synthesize new tools.
We have one excellent group here, but that’s not
enough. As an important step toward a solution,
we’re setting up a national network of chemists and
biologists (ChemBioNet). Our ambition is to further
expand those activities and to become a national
platform to promote this symbiosis of chemistry and
biology.
In classical academia, you characterize a target and
usually stop there. However, while maintaining the
focus on basic research, we strive to extend our
pipeline. We take projects to the point that we have
small molecule inhibitors for a protein, including
small molecules to disrupt protein-protein interactions.
This is a significant contribution to narrowing
the gap between pharmaceutical companies and
academia.
291

Detlev Ganten,
CEO of the Charité,
Universitätsmedizin Berlin and
Founding Director of the MDC
The MDC was established with a vision of blending fundamental
and clinical research in a new way, in hopes of making
it easier to translate findings from basic science into something
that would be medically useful. How did that vision
evolve while you were at the MDC, and what is your take on
the progress of molecular medicine today?
When I came to Berlin-Buch in 1991, genomics was the name of the
game. I was a pharmacologist who was already starting to take a
“genomics” approach to hypertension in the work of my own lab, and
many others were doing likewise – even though we didn’t yet have a
national genome program in Germany at that time. But we were
already beginning to expand our focus beyond small numbers of molecules
and single processes, trying to see things in terms of complex
relationships within genomes. Almost at the same time, in 1992, the
idea of gene therapies started to generate a lot of enthusiasm. The
idea was that we could exchange a bad gene for a good one, using a
virus or some other type of delivery vehicle, a “gene taxi.” Well, the
results of this work have been mixed, which is to be expected considering
that the methods are in their infancy. But we had very stimulating
and successful scientific meetings on gene therapy in Berlin-Buch.
The focus has now shifted as we realize that genome-wide phenomena
crystallize at the level of cell biology and systems biology and many
of the things we want to do will require a thorough understanding of
what happens at this level. I remember when experiments from
Walter Birchmeier’s lab on a signaling protein called the hepatic scatter
factor started us thinking about the mechanisms that underlie
Interview
metastases in liver cancer. At the time I wasn’t really sure how this
would develop and whether it would be a basic mechanism that
would be of interest to many people. But you see how that has turned
out – everything we discover about the fundamental cell biology of
these crucial processes has the potential of becoming something very
powerful in diagnostics and possibly treatments. Another lab in which
this relationship is very clear is that of Thomas Jentsch, who is clearing
up basic issues about the functions of membrane proteins that play
important roles in several diseases.
The term “molecular medicine” implies that there will be detailed, indepth
studies on basic molecular mechanisms, and there will be a
patient-oriented medical outcome of your work. It gives you an orientation
when you discover that a molecule such as β-catenin or NFKB
controls migrations or the cell cycle. You don’t stop at saying, “What
does this do during embryonic development?” – you go on to see
whether the process also takes place in cancer cells, in cardiovascular
tissue or in the brain. It might seem like an accident that a scientist
like Thomas Willnow finds a connection between receptor proteins
and Alzheimer’s disease, but it’s not, not really. If it weren’t Alzheimer’s,
it would be something else. Today’s biggest killers – in the developed
world, and increasingly in the rest – are diseases that stem from fundamental
problems in cells. So to deal with them we’re going to have
to bridge the gaps between molecules, cells, organ function and the
whole body in health and disease.
The idea was to gather clinicians and basic researchers on one site, to
have them share students so that there would be an exchange of
292

Detlev Ganten
293
methodologies and questions between basic research and
the clinics. While the approach has been successful, we still
haven’t completely solved the culture gap between patient
care in the clinic and laboratory work. It’s especially gratifying
to see how many clinicians have really become interested
in the mechanisms of the diseases they confront every
day, and unraveling connections between those and much
rarer diseases. Maybe it’s easier for clinicians, who are concerned
with practical problems and real patients, to
become interested in the theoretical side of things. For
researchers the goal has always been to publish in great
journals. That has been the goal so long that it’s harder to
escape. They may respect clinicians for their ability to treat
patients, but that’s not the main criterion for success in science.
So while it’s already a great step forward to collect
people under one roof – and that’s essential – it is not a
guarantee that they will develop the right mentality to
really collaborate.
But in Buch it works, and I can cite many examples. When
Claus Scheidereit was at the Max Planck Institute he was
working on NF-ΚB, doing basic biochemistry, without thinking
much about any real applications. Then he came to the
MDC and is now working with Bernd Dörken and Friedrich
Luft, trying to understand how lymphomas develop and
how blood vessels are damaged by hypertension and what
kind of new therapies might be possible. The same thing
happened with Achim Leutz, who converted from being a
cell biologist to a person interested in leukemia. From the
clinical side – it was the constellation of combining things
that attracted people like Ludwig Thierfelder, a cardiologist
who has become interested in experimental genetics. Peter
Schlag, who arrived with a high-tech perspective on surgery,
realized what could be gained through collaborations
with basic science and cell biology.
Those interactions should be pushed even more
through the ECRC, which is now being created in
conjunction with a Clinical Research Center on the
Charité side...
Yes. Clinical and experimental science have to be under one
roof. They have to have labs next door to each other – even
better is to have clinicians and basic scientists working
together, within one lab. Originally, the two academic clinics
in Buch were in danger after German reunification
because they were small and science-oriented; they cost
money. Fortunately, they became part of the Charité university
hospitals. When HELIOS took over, there was a big dis-

cussion about where the new hospital would be built – one plan was
to put it near the highway, which would have had some advantages in
terms of infrastructure. But we fought for two years to get the hospital
right next to the science campus where it is now. The clinics and
research labs have to be one community.
One thing you were involved in was the development of the
Helmholtz Association itself as a national organization; in
fact, you served as its president.
You have to understand the history of Helmholtz, where it came from.
After World War II, in the 1950s Germany was rebuilding its science
infrastructure, and one key decision was to start working again on
atomic and nuclear physics for peaceful purposes. Other types of scientific
institutes that had been strong before the war – such as in aviation,
energy research, etc. – also had to be rescued or restructured.
Several institutes were founded; they were completely independent of
the academic university structure and each of them was directly supported
through government funding. Of course that caused problems
– they were too strongly linked to politics, and they needed scientific
and administrative freedom. Since a number of institutes were in the
same boat, they decided to organize themselves as an association to
represent their joint interests. Before I became president, they decided
they needed a corporate identity of their own. That required long discussions,
and we decided it would be the “Helmholtz Association of
National Research Centers.” The government was not really amused by
this move – they were afraid of losing control of the programs – and
didn’t want to leave us alone. We agreed that we were bigger and scientifically
as good as the Max Planck Institutes and needed academic
independence.
Then we formalized the institutions of the Helmholtz association, and
created a senate. The ministry was present in the senate but didn’t
preside over it, so we slowly developed independent structures. The
Interview
next step was to organize ourselves across the centers in scientific
programs such as health, energy, and the environment. Each of the
institutes was still independent and whenever something important
happened we would have to go directly to the ministry because they
provided the money. Now the situation is different. A major part of funding
comes through grant money and third-party funding from applications
made by groups working on common themes across different
institutes – so for example a certain amount of program funding will be
devoted to cardiovascular and metabolic research, most of it going to
the MDC but some to the DKFZ and other groups. This is still an ongoing
process; Helmholtz wants to develop a corporate identity of its own, and
ensure its academic freedom and independence from politics.
I took up the job of the Helmholtz presidency because I felt that in
Germany the many research organizations and their separation from
the universities leads to fragmentation and should be better integrated.
Things are still evolving; it will be interesting to see how everything
turns out. But the new president of Helmholtz is a strong integrator,
bringing the centers and insitutions together.
You invested a lot of effort in trying to build up Buch itself as a
community, and to draw new institutes and businesses here...
Everybody needs a long-term vision. After German reunification, there
was enormous emotion going through the country. Buch is a nice
place with a strong tradition and a lot of potential; it would have been
a tragedy to lose that. Such places needed to be protected and developed.
The town of Buch itself hasn’t evolved as well as we would have
liked, but I believe that will change as the huge HELIOS complex develops
its infrastructures, and as projects like the Biotechnology Park and
the proposed new Life Sciences Center develop. The constellation we
have achieved here is interdisciplinary and very special and is also
receiving support from the Senate of Berlin and other directions. The
current president of Humboldt University is a theologian, and under
294

his presidency the interdisciplinary life sciences have become a major
focus of the university’s development, which is also good for Buch.
At the beginning things were different, and there was quite a bit of
anxiety about how things would turn out. In the very beginning I
talked to the architect Prof. Wischer, who did the general planning of
the campus. We had meetings and walked around on the campus,
looking at leaking roofs and we produced a big list of things that needed
to be done. He said, “Don’t get too focused on small problems; think
about long-term perspectives.” Prof. Wischer organized a meeting with
architects and engineers, landscapers and gardeners. He asked them,
“What should this campus look like in 50 or 100 years? Don’t think
about money, think about possibilities.” We drew up a plan and then
began to realize it piece by piece. At that time we would not have
believed that our dreams would ever become a reality. But within the
last 15 years we got the FMP, the communications center, the new animal
facility, and the administrative building, the mensa, the new buildings
of the technology park, and the medical genomics center. The
biotechnology park is not to be underestimated because it has the
potential of systematically linking basic research, clinical applications,
and the economic and commercial use of research results, patenting
licenses and so on. The Gläsernes Labor is a relatively small activity but
with a huge impact in communicating with the public. The plans for a
new “Life Sciences Center” are a logical consequence of it.
You have been very engaged in promoting the public understanding
of science; the Gläsernes Labor is a fantastic
platform for reaching school children.
Everything you do in this area is important and eventually returns benefits
to you. Such a lab has a high value for the campus, and not only for
children. I’ll give you an example: Walter Jens is a friend, the President of
the Academy of Arts, and he is a very influential person in Germany. I had
known him a long time. He is very interested in medicine, but he has
been extremely critical of genomics, for sort of classical reasons. He felt
that it would lead to genotyping and exposure of intimate private
details about people’s lives. Well, I invited him to come with some of his
people. He visited the Gläsernes Labor and learned what a genotype and
phenotype meant and then we had lunch. He said, “I understand now –
it’s just a methodology. I largely overestimated the conclusions that can
be drawn from genome analysis.” The visit to the Gläserne Labor, and the
hands-on experience with gene technology and modern genome
methodology, converted him from a critic to a supporter. Convincing just
two or three people like Walter Jens, who write books and for journals
and are influential, can change German science politics.
Regarding the industry side of the campus... Wasn’t the idea
originally that the companies would be spin-offs of the lab, or
provide services for them?
In the beginning that was more the case. The first companies were
spin-offs of the academy institutes or the MDC. The most prominent
example is Eckert and Ziegler, who started up with intellectual property
from the MDC and established a company and got on the market.
Several others such as Mr. Bendzko and Dr. Fichtner started that way.
Later there was an influx of companies from the outside who liked the
campus, found the conditions there very attractive for start-ups, and
were also enthusiastic about being in an environment with the
research institutes and clinics. The Biotech campus is flourishing and
there is still space for new spin-offs. Like always, success such as this
depends on opportunities and institutions, but especially on people.
Dr. Gudrun Erzgräber still is the energetic and creative motor behind
the campus management group.
Now you have moved to the Charité – which must give you a
unique perspective on the relationship between the clinical
and research worlds. What has been the biggest adjustment
you’ve had to make?
If you are long enough at an institute like the MDC, you know almost
everybody. You have hired important people, you know the institute in
depth. Maybe you haven’t read every paper produced by the institute, but
you know people’s work, their reputations, the grants they have. A scientific
institute like the MDC is managed in a very personal way by interactions
directly with the people and helping to solve their problems.
Well, then you come to the Charité and find yourself confronted by
15,000 people on four different sites, across the city of Berlin, covering
themes from basic research to clinics. There is a huge amount of
administration and technical support to manage. You can’t even know
all the important people. This type of organization is completely new
for me and I learned a tough lesson: you can only lead strategically. You
can define a direction and then you have to select a few, able people
who are able to translate your ideas into the institution. If you don’t
assemble the right team then you are lost.
Fortunately the Charité is an excellent, unique institution with a great
tradition and a fantastic potential and really excellent people in
research and in the clinics, as well as in the administration. The Charité
is a very important institution for Berlin and thus has a lot of political
support. In the fourth year after the merger of the medical faculties of
the Free University (West) and the Humboldt University (East), the
“Charité-University Medicine Berlin” has made great progress. We have
a development plan above 500 million Euro for the next seven years
and Nature has called the Charité “the beacon of reform” in the
German University system. So it appears we are reaping the fruits and
it has been worth all the efforts.
295 Interview

Gudrun Erzgräber,
Managing Director,
BBB Management GmbH
What were the circumstances surrounding the founding of
the BBB?
The BBB was founded in the summer of 1995 as an offshoot of the
Max-Delbrück Center. Why? There was a recommendation from the
Scientific Council of FRG that, alongside its other functions, the campus
should host related industrial research activities - that meant we
needed to make an effort to bring in companies. The Council was well
aware that a considerable amount of applied research had been carried
out in the GDR institutes that had been located here, and saw that
as a potential area for growth. In 1992 the MDC began operations on
campus. At the time it was the only research institute here, and it
acquired responsibility for managing the campus. To handle that it
founded a local management group, and I was put in charge of organizing
it. At the beginning the group was responsible for the site
management, in terms of supply and the technical infrastructure –
what we’re doing today, on a much smaller scale. At the time we still
belonged to the MDC.
Then came a point where we said, if our goal is really to attract companies,
then we’ll have to be able to apply for external financial support.
The GEBA – which is responsible for providing such support on
behalf of the state of Berlin – said that these activities were outside
the normal domain of a research institute and to be eligible, the MDC
would have to establish a company. So in 1995 the Senate passed a
resolution that permitted the MDC to found the BBB. This was a new
public mission, it led to the creation of the BBB as a spin-off firm, and
the necessary start-up capital came from the state of Berlin through
Interview
the MDC. I brought along two colleagues from the MDC and that’s
how the BBB got its start.
On January 1, 1996, I was officially appointed business manager. At the
time I had no idea what I was getting into! But there are two things
about the job that I have really enjoyed: solving difficult problems –
and there have been a few of those! The other thing is the opportunity
to communicate with a wide variety of people – with the companies,
with the scientific institutes, with politicians, and many others.
We knew that things wouldn’t be easy.
The biggest challenge for me as manager has been to handle the
financial aspects of the company. If we were going to apply for financial
support, it meant we would also have to bring our own money to
the table, and that meant borrowing it and assuming debts. But we
weren’t per se eligible for credit – of course we had backing from the
MDC, but that wasn’t a usual situation for a firm. After complicated
negotiations with the state, we obtained 5.5 million. We had to pay
that back ourselves, including the interest, from rent income. We did
obtain start-up money between 1996 and 1998, because it was obvious
that there wouldn’t be much income in the initial period. Since then
– since the beginning of 1999 – we have been financially independent.
My goal was to see that we broke even, and maybe even do a little better.
Not to make a profit – the BBB wasn’t there to make money for
itself; that wouldn’t have worked anyway, since the buildings were 80to-90
percent funded with state support. Instead, what we earned
could be rolled back into the campus, to improve the infrastructure
and make things easier for the companies. We’re proud that we have
296

Gudrun Erzgräber
managed – there was a very difficult period from 2003 to 2004. That’s
when we opened the building we’re sitting in now, and at first there
was only one occupant!
I can remember when State Secretary Strauch came for the opening
ceremony... I was standing in front of the building, it was a beautiful
day in June, and I thought – “If they only knew how I feel right now!
With this empty building...” The technology boom had crashed; there
hadn’t even been any expressions of interest from companies... Very
difficult. We had to stretch our accounts to the limit just to pay our
bills.
Then there was a revival of the incubation business. A succession of
new businesses moved here. 2005 was our best year, an amazing year
in which 12 new companies moved onto campus. In 2006 it was four,
and 2007 four, to the point that today 88 percent of the available
space is occupied. And new requests are coming in. Of course from
time to time one of the companies leaves; some shrink, some grow –
an example is Silence Therapeutics, which is currently in a growth
phase. We have a very good anchor in Eckert and Ziegler. In addition to
their own building, they rent space from us.
What have been the major phases in the development of
the campus?
The first companies that came here in many cases worked with equipment
that was left over from the former institutes; many of them worked
part-time, in lab space that belonged to the MDC. I don’t know if
you have seen pictures from those days, but everyone was in barracks
– the campus was full of barracks. You can imagine that biotechnology
companies in barracks, without proper lab equipment – that just
doesn’t work. So our first goal was to improve the infrastructure and
297 Interview

uild laboratories. The first stage was in 1998, with the renovation of
the Oskar and Cécile Vogt House as well as house 79. Then we began
building here on the “industrial strip.” In 2001 we opened buildings 80
and 82, and in 2003 building 85 was finished. Those were the three
major phases. Another important step was the establishment of the
Gläsernes Labor, which now does such a wonderful job with school
children. For this we made use of one of the oldest houses here.
Professor Ganten had the idea of setting up these teaching activities,
and he said, “See how you can finance this!” We renovated the house
and rebuilt the upper floor, and set up an exhibit room as a meeting
point for businesses, scientists, and clinical research. In doing so it
became part of the state supported incubator facilities. Otherwise
that would never have existed here – as valuable as it has been, it was
simply too expensive in the start-up phase. There, too, we depended on
public funding; the MDC could never have financed it on its own.
When the FMP established itself on campus, in 2000, it “bought into”
the BBB, by taking a share of 20%. Another very important moment
came during the national BioRegio competition, and that brought
Schering, which came in at another 20 percent – that’s now the phar-
Interview
maceutical concern Bayer Schering. That’s the status today – the MDC
has a 60 percent share and the other two partners the rest. The
Charité has applied to become another shareholder. There are still formalities
to be faced before that happens. I’m hoping it will come before
the end of the year. But according to the rules, the MDC will remain
majority shareholder.
Why is it important today for there to be a significant company
presence on campus, even if their primary mission is not to
provide services to the research groups?
It was fully clear to us that firms on campus would not subsist on services
they provided to research groups, or be mainly dedicated to developing
products that come from the research laboratories. It was never
intended just as the incubator for firms arising from the MDC and
FMP. There would never be enough intellectual property at the right
stage of development to launch start-up companies on this scale.
Some of the firms do have collaborations with the institutes, but overall
the biotechnology park was intended to have a regional and superregional
function. From the very beginning the companies would
serve the clinics in Buch, but they are global actors, but could also
298

serve local partners. The campus has a greater potential than just playing
host to excellent research institutes. And as we develop that
potential, it pushes the development of the institutes.
The campus offers good infrastructures to companies. If you’re seeking
collaborations, you can find them here. You can go to talks, to workshops,
there is an excellent library, and an IT infrastructure...
It’s a bit harder to say what the typical scientific group gets out of a
strong company presence on campus. From my point of view, there
could be more interest in this interaction. I understand that perfectly
well – I have a research background myself. You tend to get very focused
on getting good results and publishing good papers and you often
forget to look at what’s going on around you. The most important
thing is that when businesses express an interest in collaborations,
that there’s a resonance from the side of the laboratories. And an
important point, obviously, is that many of the scientists here don’t go
on to academic careers. Those that go into industry have a good
launch pad through campus firms.
The ECRC may become a good platform for the integration of companies.
I am a member of the council and know the concept. At the
moment the project hasn’t yet evolved to the point that it would be
really simple for companies to dock on. But as things progress, if the
ECRC is open, and they are prepared to sit down with visitors and talk,
it could function that way. Another point of contact will be core facilities,
if they have extra capacity that can be used by companies.
When a company expresses interest in coming, how do you
make the campus attractive to them?
There are the scientific strengths of the MDC in cardiovascular and
metabolic diseases, cancer, and neurodegenerative diseases. That’s
complemented very well by the structural and pharmacological work
of the FMP. The presence of HELIOS with the possibilities of clinical
testing. We also have the Academy of Health here, which offers a very
extensive training program. About a thousand students are enrolled
there now, being trained for a broad spectrum of health-related professions.
It’s an association which carries out training as a service for
hospitals, rehabilitation centers, and similar institutions. The students
will become lab technicians, massage therapists, people who work in
retirement homes, and so on. The Academy operates in an association
including clinics, who co-finance its activities.
We also show them the potential of Buch – there are other attractive
campuses in the village, such as the area where the Franz Volhard
Clinic used to be – that are now empty. We hope very much that they
will be developed as partners in this complex of activities related to
health.
Buch does have a few disadvantages. The town isn’t in the
city, which doesn’t always make things easy...
During the first two phases of development on the campus, as the
FMP was moving in, as everything was being renovated and built
anew, we were worried that the discrepancy between a very progressive
scientific campus and a village from the former GDR would become
wider. We were getting more and more questions like, “Where can
we go shopping here? What are the schools like?” For ten years, with
the exception of this campus, Buch had been experiencing an exodus.
Clinics were moving out; people were leaving, and there were empty
houses. Then when HELIOS bought the state clinic in 2001, things suddenly
started to move ahead. Several of us got together and decided to
found a “Buch management association.”
We convinced the region to do something similar; they created a regional
management association. Their main responsibility was to identify
the interests of the various partners and then to speak to the Senate
with a unified voice about what has to be done for the region. One of
their functions was marketing – to show outsiders the positive
aspects of Buch. They have now received sponsoring for about six
years and have done a lot for the town. These haven’t always been
things that directly affected the campus, but they have been important
for the infrastructure of the community and that has had important
indirect effects. There have been historical walks through the
Hoffmann campuses, business establishment workshops, and public
meetings. We have met with potential investors and shown them the
empty campuses and discussed possible uses. Some of these areas
have now been bought and will be developed in a variety of ways –
many of them will be related to the region’s overall health mission.
One of them will house the new Life Sciences Center. I think all of
these things together have helped Buch make a big step forward – and
that in turn helps the campus.
You’ll be retiring at about the time this book appears...
Yes! Andreas Mätzold and Ulrich Scheller will take over my functions in
the BBB. But I won’t completely disappear. One project I’ll be involved
in is working on the Life Sciences Center – something I’m really pleased
to be involved in, because this project has been an immense
amount of work for almost eight years now. We have the other things
we need: an excellent scientific campus, an enormous new hospital.
While all the other projects that are being planned are important –
new living areas and rehabilitation centers, housing for the families of
patients – we also need something special. The Life Sciences Center
has the potential of bringing a huge new public to Buch. As well as
being something good just in itself, that will also give a strong boost
to our local infrastructure. We think it will make an ideal new meeting
point for all of the players here – from physicians and scientists to
industry, children, and the general public.
299 Interview

300
Index
Adamidi, Catherine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271
Agre, Peter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
Alzheimer, Alois . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
Andersen, Olav . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170
Anzenberger, Uwe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
Bähring, Silvia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
Bairach, Rich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168
Baumeister, Hans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132
Behrens, Jürgen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
Bendzko, Peter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295
Berkenhout, John . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .274
Beyreuther, Konrad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
Bielka, Heinz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14
Bienert, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197, 288
Bilginturan, Nihat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
Bilroth, Theodor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Birchmeier, Carmen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28, 62, 154, 284
Birchmeier, Walter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22, 94, 168, 213, 284
Blankenstein, Thomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119
Blanz, Judith . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206
Bordag, Natalie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198
Bright, Richard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220
Broemer, Meike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243
Burkitt, Dennis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .106
Burmeister, Regina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166
Cage, John . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
Cajal, Santiago Ramón y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53
Caliess, Christiane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44
Carrel, Alexis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90
Chevelkov, Veniamin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176
Chitayat, David . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230
Chmielowiec, Jolanta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31
Christély, Thomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79
Ciechanover, Aaron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168
Clevers, Hans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
Correns, Carl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Couper, Archibald . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
Crick, Francis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Danielczyk, Antje . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .136
Darwin, Charles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10, 248
Dathe, Margitta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .197
de Vries, Hugo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
De Vries, Hugo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
Deichman, Galina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124
Delbrück, Max . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
Deter, Auguste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
Diekhoff, Britta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82
Dörken, Bernd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243, 286, 293
Eckert, Andreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .134
Edemir, Bayram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
Eisai, Myoan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .250
Elsener, Karl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .161
Endruschat, Jens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82
Ertl, Gerhard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
Erzgräber, Gudrun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167, 296
Fichtner, Iduna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295
Fischer, Judith . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276
Förster, Reinhard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
Frahm, Christina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
Franke, Werner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
Friedländer, Marc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .267
Furth, Priscilla . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Ganten, Detlev . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .281, 292
Garratt, Alistair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154
Gassman, Max . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
Gerull, Brenda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
Giese, Klaus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Glass, Rainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Göhler, Heike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .259
Goletz, Steffen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .134
Golgi, Camille . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53
Gollasch, Maik . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .232
Gomoll, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190
Gong, Maolian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .234
Grossmann, Katja . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .95
Grunz, Katharina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279
Guise, Kevin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142
Hackett, Perry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142
Haeckel, Ernst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
Hammes, Annette . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
Harris, Robert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .184
Helmholtz, Hermann von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
Henn, Volker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
Hertwig, Oskar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
Hinz, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243
Hoffmann, Ludwig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
Hollman, Andreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
Höpken, Uta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109
Hübner, Norbert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65, 233, 274
Huelsken, Jörg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
Humboldt, Alexander von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9
Hundrucker, Christian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189
Ivics, Zoltán . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
Izsvák, Zsuzsanna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
Jeffries, Alec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .224
Jens, Walter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295
Jentsch, Thomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86, 201
Joshi, Mangesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176, 213
Kapitonov, Vladimir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146
Kaufmann, Jörg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Index

301 Index
Kekulé, Auguste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
Keller, Sandro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .194
Kemmner, Wolfgang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Kempermann, Gerd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58
Kettenmann, Helmut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14, 53, 86
Kettritz, Ralph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
Klaus, Alexandra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Klein, Eva . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121
Klein, Georg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121
Klippel, Anke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Klußmann, Enno . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
Knut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
Koch, Robert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11
Koralov, Sergie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269
Kries, Jens Peter von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .213
Kronenberg, Golo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Kujawa-Schmeitzner, Elisabeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
Leheste, Jörg-Robert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166
Lenhard, Diana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
Leutz, Achim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40, 293
Lewin, Gary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31, 151, 284
Lindbergh, Charles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
Linser, Rasmus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176
Lipp, Martin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66, 86, 103
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .106, 112, 218, 281, 286, 293
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .218
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .281
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .286
Luft, Friedrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .293
Maatz, Henricke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276
Mammen, Jeanne . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16, 78
Markovic, Darko . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56
Mätzold, Andreas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299
Maul, Björn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
Melby, James . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .229
Mendel, Gregor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12
Milenkovic, Nevena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
Miskey, Csaba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144
Mo, Xianming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48
Monti, Jan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .281
Morgan, Thomas H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
Muljo, Stefan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269
Muller, Hermann . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30, 223
Müller, Johann . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
Naegoe, Ioana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .209
Nägeli, Karl von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13, 61
Naraghi, Ramin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228
Narayanan, Saravanakumar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .180
Nedvetsky, Pavel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
Neeser, Philippe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .250
Nykjaer, Anders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172
Oschkinat, Hartmut . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .210
Pérec, Georges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
Perutz, Max . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .256
Petersen, Helle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .165
Podust, Larissa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .215
Porrasmillan, Pablo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254
Rajewsky, Klaus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31, 244
Rajewsky, Nikolaus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .237, 262
Redel, Alexandra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254
Reich, Jens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247
Reif, Bernd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .174
Rio-Hortegas, Pio del . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56
Rodin, Auguste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211
Rosenthal, Walter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186, 288
Rossum, Barth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212
Rous, Peyton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46
Scheidereit, Claus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .240, 284, 293
Scheller, Ulrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .299
Schlag, Peter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70, 284, 293
Schleiden, Mathias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
Schliemann, Heinrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40
Schliemann, Sophia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Schmidt, Roland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
Schmidt-Ulrich, Ruth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .248
Schmieder, Peter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212
Schmitt, Andrea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144
Schuster, Herbert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
Schwann, Theodor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
Schwartze, Stefan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168, 285
Scriabin, Alexander . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262
Seyfried, Salim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166
Siefert, Steffi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .60
Sinzelle, Ludivine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .146
Spoelgen, Robert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
Stein, Ulrike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Synowitz, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Thierfelder, Ludwig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88, 284
Timoféeff-Ressovsky, Nikolay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Toka, Hakan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
Toka, Okan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222
Tschernak, Erich von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
Uckert, Wolfgang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127
Vargas, Carolyn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217
Vasyutina, Elena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62
Vietinghoff, Sibylle von . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
Virchow, Rudolf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10, 40
Vogt, Cécile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Vogt, Oskar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Volhard, Franz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220
von Kölliker, Albrecht . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54
Waaler, Jo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .217
Wagner, Anne . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .255
Wallace, Alfred . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
Walther, Wolfgang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
Wanker, Erich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251
Wartosch, Lena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204
Watson, James . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
Wengner, Antje . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110
Wiedenmann, Bertram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85
Wienker, Thomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .221
Willimsky, Gerald . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
Willnow, Thomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160, 259, 292
Wolf, Jana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .247

Further reading
Andersen, OM, Reiche, R, Schmidt, V, Gotthardt, M, Spoelgen, R, Behlke, J, von Arnim, CA F,
Breiderhoff, T, Jansen, P, Wu, X, Bales, KR, Cappai, R, Masters, CL, Gliemann, J, Mufson, EJ,
Hyman, BT, Paul, SM, Nykjær, N and T E Willnow. (2005) SorLA/LR11, a neuronal sorting receptor
that regulates processing of the amyloid precursor protein. Proc. Natl. Acad. Sci. USA
102, 13461-13466.
Atanasoski, S, Scherer, S, Sirkowski, E, Leone, D, Garratt, AN, Birchmeier C, Suter U. (2006).
ErbB2 signaling in Schwann cells is largely dispensable for maintenance of myelinated
peripheral nerves and proliferation of adult Schwann Cells following injury. J. Neurosci. 26:
2124-2131.
Bähring, S, Kann, M, Neuenfeld, Y, Gong, M, Chitayat, D, Toka, HR, Toka, O, Plessis, G, Maass, P,
Rauch, A, Aydin, A, Luft, FC. (2008) The inversion region for hypertension and brachydactyly
on chromosome 12p features multiple splicing and patient-specific expression of non-coding
RNA. Hypertension (in press).
Ball LJ, Kühne R, Schneider-Mergener J, Oschkinat H (2005) Recognition of proline rich
motifs (PRMs) by protein-protein interaction domains. Angew Chem Int Ed 44, 2852-2869.
Bárány-Wallje E, Keller S, Serowy S, Geibel S, Pohl P, Bienert M, Dathe M (2005) A critical
reassessment of penetratin translocation across lipid membranes. Biophys J 89, 2513-2521.
Begay, V, Smink, J, and Leutz, A. (2004). Essential requirement of CCAAT/ enhancer binding
proteins in embryogenesis. Molecular Cell Biol, 24, 9744-9751
Blankenstein, Th (2007). Do autochthonous tumors interfere with effector T cell responses?
Sem. Cancer Biol. 17: 267-274. Weinhold, M, Sommermeyer, D, Uckert, W and Blankenstein,
Th. (2007). Dual T cell receptor expressing CD8+ T cells with tumorand self- specificity can
inhibit tumor growth without causing severe autoimmunity. J. Immunol., in press.
Böddrich, A, Gaumer, S, Haacke, A, Tzvetkov, N, Albrecht, M, Evert, BO, Futschik, M, Müller, EC,
Lurz, R, Breuer, P, Schugardt, N, Plaßmann, S, Morabito, LM, Warrick, JM, Suopanki, J, Wüllner,
U, Frank, R, Hartl, FU, Bonini, NM, Wanker, EE. (2005) An arginine/lysine-rich motif in ataxin-
3 is responsible for the interaction with the molecular chaperone VCP/p97 that modulates
aggregate formation and neurotoxicity. EMBO J. 25, 1547-1558.
Cambridge SB, Geißler D, Keller S, Cürten B (2006) A caged doxycycline analogue for photoactivated
gene expression. Angew Chem Int Ed 45, 2229-2231.
Chaurasia G, Iqbal Y, Haenig C, Herzel H, Wanker EE, Futschik ME. (2007). UniHI: an entry
gate to the human protein interactome. Nucleic Acids Research, 2007, 35, D590-4. Herbst,
M, Wanker, EE. (2007): Small molecule inducers of heat shock response reduce polyQ-mediated
huntingtin aggregation. A possible therapeutic strategy. Neurodegener Dis. 4(2-3),
254-60.
Chen, K, Rajewsky, N. (2006). Natural selection on human microRNA binding sites inferred
from SNP data, Nature Genetics 38, 1452-1456.
Chen, K, Rajewsky, N. (2007). The evolution of gene regulation by transcription factors and
microRNAs. Nature Reviews Genetics 8, 93-103.
Chmielowiec, J, Borowiak, M, Morkel, M, Stradal, T, Munz, B, Werner, S, Wehland, J, Birchmeier,
C, Birchmeier, W. (2007). c-Met is essential for wound healing in the skin. J. Cell Biol. 177,
151-162.
Ehrnhoefer, DE, Duennwald, M, Markovic, P, Wacker, JL, Engemann, S, Roark, M, Legleiter, J,
Marsh, JL, Thompson, LM, Lindquist, S, Muchowski, PJ and Wanker, EE. (2006). Green tea (-)epigallocatechin-gallate
modulates early events in huntingtin- misfolding and reduces
toxicity in Huntington’s disease models. Hum Mol Genet. 15(18), 2743-2751.
Engels, B, Nößner, E, Frankenberger, B, Blankenstein, T, Schendel, D, Uckert, W. (2005).
Redirecting human T lymphocytes towards renal cell carcinoma-specificity by retroviral
transfer of T cell receptor genes. Hum. Gene Ther. 16, 799-810.
Further reading
Engels, B, Uckert, W. (2007). Redirecting T lymphocyte specificity by T cell receptor gene
transfer – A new era for immunotherapy. Mol. Aspects Med. 28, 115-142.
Gerull , B, Heuser, A, Wichter, T, Paul, M, Basson, CT, McDermott, DA, Lerman, BB, Markowitz,
SM, Ellinor, PT, MacRae, CA, Peters, S, Grossmann, KS, Drenckhahn, J, Michely, B, Sasse-
Klaassen, S, Birchmeier, W, Dietz, R, Breithardt, G, Schulze- Bahr, E, Thierfelder L. (2004).
Mutations in the desmosomal protein plakophilin-2 are common in arrhythmogenic right
ventricular cardiomyopathy. Nat Genet. 36, 1162-4
Glass R, Synowitz M, Kronenberg G, Markovic D, Wang LP, Gast D, Kiwit J, Kempermann G
and Kettenmann, H. (2005). Gliomainduced attraction of endogenous neural precursor
cells is associated with improved survival, J. Neurosci., 25, 2637-2646.
Grego-Bessa, J, Luna-Zurita, L, del Monte, G, Bolós, V, Melgar, P, Arandilla, A, Garratt, AN, Zang,
H, Mukouyama, Y, Chen, H, Shou, W, Ballestar, E, Esteller, M, Rojas, A, Pérez-Pomares, de la
Pompa, JL. (2007). Notch signalling is essential for cardiac ventricular chamber development.
Developmental Cell 12: 415-429.
Haas B, Schipke C G, Peters O, Söhl G, Willecke K. and Kettenmann H. (2006) Activitydependent
ATP-waves in the Mouse Neocortex are Independent from Astrocytic Calcium
Waves. Cereb Cortex, 16, 237-46.
Hammes, A, Andreassen, TK, Spoelgen, R, Raila, J, Hubner, N, Schulz, H, Metzger, J,
Schweigert, FJ, Luppa, PB, Nykjaer, A, Willnow, TE. (2005) Role of endocytosis in cellular
uptake of sex steroids. Cell 122(5): 751-762.
Hanisch U-K and Kettenmann K (2007) Microglia – active sensor and versatile effector cells
in the normal and pathologic brain, Nature Neuroscience, 10,1387-1394.
Heidemann A C, Schipke C G, Kettenmann H. (2005) Extracellular application of nicotinic
Acid adenine dinucleotide phosphate induces Ca2+ signaling in astrocytes in situ, J. Biol.
Chem., 280, 35630-35640.
Henke, N, Schmidt-Ullrich, R, Dechend, R, Park, JK, Qadri, F, Wellner, M, Obst, M, Gross, V, Dietz,
R, Luft, FC, Scheidereit, C, and Müller DN. (2007). Vascular endothelial-specific NF-κB suppression
attenuates hypertension-induced renal damage. Circ Res 101, 268-276.
Hinz, M, Broemer, M, Çöl Arslan, S, Dettmer, R, Scheidereit, C. (2007). Signal-responsiveness
of IB kinases is determined by Cdc37-assisted transient interaction with Hsp90. J. Biol.
Chem. in press.
Höpken, UE, Wengner, AM, Loddenkemper, C, Stein, H, Heimesaat, MM, Rehm, A, Lipp, M.
(2007) CCR7 Deficiency causes ectopic lymphoid neogenesis and disturbed mucosal tissue
integrity. Blood 109, 886-95 [Epub 2006 Oct 3]
Hübner, N, Wallace, CA, Zimdahl, H, Petretto, E, Schulz, H, Maciver, F, Muller, M, Hummel, O,
Monti, J, Zidek, V, Musilova, A, Kren, V, Causton, H, Game, L, Born, G, Schmidt, S, Müller, A,
Cook, SA, Kurtz, TW, Whittaker, J, Pravenec, M, Aitman, TJ. (2005). Integrated transcriptional
profiling and linkage analysis for disease gene identification. Nature Genetics. 37, 243-253.
Hübner, N. (2006). Expressing physiology. Nature Genetics 38, 140-141.
Humphries, JD, Schofield, NR, Mostafavi-Pour, Z, Green, LJ, Garratt, AN, Mould, AP,
Humphries, MJ. (2005). Dual functionality of the anti-1 integrin antibody, 12G10, exemplifies
agonistic signalling from the ligand-binding pocket of integrin adhesion receptors. J. Biol.
Chem. 280: 10234-10243.
Hundsrucker C, Krause G, Beyermann M, Prinz A, Zimmermann B, Diekmann O, Lorenz D,
Stefan E, Nedvetsky P, Dathe M, Christian F, McSorley T, Krause E, McConnachie G, Herberg
FW, Scott J, Rosenthal W, Klußmann E (2006) Highaffinity AKAP18-delta-protein kinase A
interaction yields novel protein kinase A anchoring disruptor peptides. Biochem J 396,
297-306.
302

Ivics, Z, Izsvák, Z. (2006). Transposons for gene therapy! Curr. Gene Ther. 6, 593-607.
Ivics, Z, Katzer, A, Stüwe, EE, Fiedler, D, Knespel, S, Izsvák, Z. (2007). Targeted Sleeping Beauty
transposition in human cells. Mol. Ther. 15, 1137-1144.
Janz, M, Hummel, M, Truss, M, Wollert-Wulf, B, Mathas, S, Jöhrens, K, Hagemeier, C, Bommert,
K, Stein, H, Dörken, B, Bargou, RC. (2006). Classical Hodgkin lymphoma is characterized by
high constitutive expression of activating transcription factor 3 (ATF3) which promotes viability
of Hodgkin/Reed- Sternberg cells. Blood. 107, 2536-2539.
Jentsch TJ. (2008) CLC chloride channels and transporters: from genes to protein structure,
pathology and physiology. Crit Rev Biochem Mol Biol. 43(1): 3-36.
Joshi M, Vargas C, Moelling K, Boisguerin P, Diehl A, Schmieder P, Krause G, Hagen V, Schade
M, Oschkinat H (2006) Making protein-protein interactions drugable: discovery and 3D
structure of low-molecular-weight ligands complexed with the AF6 PDZ domain. Angew
Chem Int Ed 45, 3790-3795.
Jundt, F, Raetzel, N, Müller, C, Calkhoven, CF, Kley, K, Mathas, S, Lietz, A, Leutz, A, Dörken, B.
(2005). A rapamycin derivative (everolimus) controls proliferation through down-regulation
of truncated CCAAT enhancer binding protein and NF-κB activity in Hodgkin and
anaplastic large cell lymphomas. Blood. 106, 1801-1807.
Jüttner S, Wissmann C, Jöns T, Vieth M, Hertel J, Gretschel S, Schlag PM, Kemmner W, Höcker
M. (2006) Vascular endothelial growth factor-D and its receptor VEGFR-3: Two novel independent
prognostic markers in gastric adenocarcinoma. J. Clin. Oncol. 24, 228-240.
Kasper D, Planells-Cases R, Fuhrmann JC, Scheel O, Zeitz O, Ruether K, Schmitt A, Poët M,
Steinfeld R, Schweizer M., Kornak U, Jentsch TJ (2005) Loss of the chloride channel ClC-7
leads to lysosomal storage disease and neurodegeneration. EMBO J. 24, 1079-1091.
Kaufman, CD, Izsvák, Z, Katzer, A, Ivics, Z. (2005). Frog Prince transposon-based RNAi vectors
mediate efficient gene knockdown in human cells. Journal of RNAi and Gene Silencing 1,
97-104.
Keller S, Heerklotz H, Blume A (2006) Monitoring lipid membrane translocation of sodium
dodecyl sulfate by isothermal tritation calorimetry. J Am Chem Soc 128, 1279-1286.
Keller S, Heerklotz H, Jahnke N, Blume A (2006) Thermodynamics of lipid membrane solubilization
by sodium dodecyl sulfate. Biophys J 90, 4509-4521.
Keller S, Sauer I, Strauss H, Gast K, Dathe M, Bienert M (2005) Membrane-mimetic nanocarriers
formed by a dipalmitoylated cell-penetrating peptide. Angew Chem Int Ed 44, 5252-
5255.
Kharkovets T, Dedek K, Maier H, Schweizer M, Khimich D, Nouvian R, Vardanyan V, Leuwer R,
Moser T, Jentsch TJ (2006) Mice with altered KCNQ4 K+ channels implicate sensory outer
hair cells in human progressive deafness. EMBO J 25, 642-652.
Klaus A, Birchmeier W. (2008) Wnt signalling and its impact on development and cancer.
Nat Rev Cancer, 8(5), 387-98.
Klaus A, Saga Y, Taketo MM, Tzahor E, Birchmeier W. (2007) Distinct roles of Wnt/β-catenin
and Bmp signaling during early cardiogenesis. Proc Natl Acad Sci 104(47), 18531-6. Epub
2007 Nov 13.
Klaus, A, Saga, Y, Taketo, MM, Tzahor, E and Birchmeier, W. (2007). Distinct roles of Wnt/catenin
and Bmp signaling during early cardiogenesis. Proc. Natl. Acad. Sci. USA. 104, 18531-
18536.
Krek, A, Gruen, D, Poy, MN, Wolf R, Rosenberg L, Epstein EJ, MacMenamin P, da Piedade I,
Gunsalus KC, Stoffel M, and Rajewsky N (2005). Combinatorial microRNA target predictions.
Nature Genetics 37, 495-500.
Krutzfeldt J, Rajewsky N, Braich R, Rajeev KG, Tuschl T, Manoharan M, Stoffel M. (2005).
Silencing of microRNAs in vivo with ‘antagomirs’. Nature 438, 685-689.
Lall, S, Grun, D, Krek A, Chen K, Wang Y-L, Dewey DN, Sood P, Colombo T, Bray N,
MacMenamin P, Kao H-L, Gunsalus KC, Pachter L, Piano F, Rajewsky, N. (2006). A genome-
Wide Map of Conserved MicroRNA Targets in C. elegans. Current Biology 16, 460-71.
Lange PF, Wartosch L, Jentsch TJ, Fuhrmann JC (2006) ClC-7 requires Ostm1 as a beta-subunit
to support bone resorption and lysosomal function. Nature 440, 220-223.
Lee-Kirsch MA, Gong M, Chowdhury D, Senenko L, Engel K, Lee YA, de Silva U, Bailey SL, Witte
T, Vyse TJ, Kere J, Pfeiffer C, Harvey S, Wong A, Koskenmies S, Hummel O, Rohde K, Schmidt
RE, Dominiczak AF, Gahr M, Hollis T, Perrino FW, Lieberman J, Hübner N. (2007). Mutations
in the gene encoding the 3’-5’ DNA exonuclease TREX1 are associated with systemic lupus
erythematosus. Nature Genetics 39, 1065-7.
Lietz, A, Janz, M, Sigvardsson, M, Jundt, F, Dörken, B, Mathas S. (2007). Loss of HLH transcription
factor E2A activity in primary effusion lymphoma confers resistance to apoptosis. Br J
Haematol. 137, 342-348.
López-Bendito, G, Cautinat, A, Sánchez, JA, Bielle, F, Flames, N, Garratt, AN, Tagmale, D, Role,
LW, Charnay, P, Marín, O, Garel, S. (2006). Tangential migration controls axon guidance: a
role for Neuregulin-1 in thalamocortical axon navigation. Cell 125, 127-142.
Martinez-Salgado C, Benckendorff AG, Chiang LY, Wang R, Milenkovic N, Wetzel C, Hu J,
Stucky CL, Parra MG, Mohandas N, Lewin GR (2007) Stomatin and sensory neuron mechanotransduction.
J Neurophys 98, 3802-8.
Mathas, S, Janz, M, Hummel, F, Hummel, M, Wollert-Wulf, B, Lusatis, S, Anagnostopoulos, I,
Lietz, A, Sigvardsson, M, Jundt, F, Jöhrens, K, Bommert, K, Stein, H, Dörken, B. (2006). Intrinsic
inhibition of E2A by overexpressed ABF-1 and Id2 is involved in reprogramming of the neoplastic
B cells in classical Hodgkin lymphoma. Nature Immunol. 7, 207-215.
Mathas, S, Jöhrens, K, Joos, S, Lietz, A, Hummel, F, Janz, M, Jundt, F, Anagnostopoulos, I,
Bommert, K, Lichter, P, Stein, H, Scheidereit, C, Dörken, B. (2005). Elevated NF-{kappa}B p50
complex formation and Bcl-3 expression in classical Hodgkin, anaplastic large cell, and
other peripheral T cell lymphomas. Blood. 106, 4287-4293.
Mehrhof, FB, Schmidt-Ullrich, R, Dietz, R, Scheidereit, C. (2005). Regulation of Vascular
Smooth Muscle Cell Proliferation: Role of NF-κB Revisited. Circ Res 96, 958-964
Milenkovic N, Frahm C, Gassmann M, Griffel C, Erdmann B, Birchmeier C, Lewin GR, Garratt
AN (2007) Nociceptive tuning by Stem Cell Factor/c-Kit signaling. Neuron 56: 893-906.
Miskey, C, Papp, B, Mátés, L, Sinzelle, L, Keller, H, Izsvák, Z, Ivics, Z. (2007). The ancient mariner
sails again: Transposition of the human Hsmar1 element by a reconstructed transposase
and activities of the SETMAR protein on transposon ends. Mol. Cell. Biol. 27, 4589-600.
Mo X., Kowenz-Leutz E, Xu H, Leutz A. (2004) Ras induces mediator complex exchange on
C/EBP. Molecular Cell, 13: 241-250
Mo, X, Kowenz-Leutz, E, Laumonnier, Y, Xu, H, and Leutz, A (2005) Histone H3-tail positioning
and acetylation by c-Myb but not the v-Myb DNA binding SANT Domain. Genes &
Development, 19: 2447-2457
Moeller, H, Jenny, A, Schaeffer, H-J, Schwarz-Romond, T, Mlodzik, M, Hammerschmidt, M, and
Birchmeier, W. (2006): Diversin regulates heart formation and gastrulation movements in
development. Proc. Natl. Acad. Sci. USA 103, 15900-15905.
Narayanan S, Kamps B, Boelens W, Reif B (2006) B-crystalline competes with Alzheimer´s
disease beta-amyloid peptide for peptide-peptide interactions and induces oxidation of
Abeta- Met35. FEBS Lett 580, 5941-5946.
Narayanan S, Reif B (2005) Characterization of chemical exchange between soluble and
aggregated states of β-amyloid by solution state NMR upon variation of the salt conditions.
Biochemistry 44, 1444-1452.
Nedvetsky P, Stefan E, Frische S, Santamaria K, Wiesner B, Valenti G, Hammer III JA, Nielsen
S, Goldenring RJ, Rosenthal W, Klußmann E (2007) A role of myosin Vb and Rab11-FIP2 in the
aquaporin-2 shuttle. Traffic 8, 110-123.
Oeckinghaus, A, Wegener, E, Welteke, V, Ferch, U, Çöl Arslan, S, Ruland, J, Scheidereit, C,
Krappmann, D. (2007). TRAF6 mediated Malt1 ubiquitination triggers IKK/NF-κB signalling
upon T cell activation. EMBO J., in press
Park TJ, Lu Y, Jüttner R, Smith ES, Hu J, Brand A, Wetzel C, Milenkovic N, Erdmann B,
Heppenstall PA, Laurito CE, Wilson SP, Lewin GR. (2008) Selective inflammatory pain insensitivity
in the African naked mole-rat (Heterocephalus glaber). PLoS Biol. 6: e13.
Pocock, J M and Kettenmann H. (2007) Neurotransmitter receptors on microglia, Trends
Neurosci., 30, 527-535.
Poët M, Kornak U, Schweizer M, Zdebik AA, Scheel O, Hoelter S, Wurst W, Schmitt A,
Fuhrmann JC, Planells-Cases R, Mole SE, Hubner CA, Jentsch TJ (2006) Lysosomal storage
disease upon disruption of the neuronal chloride transport protein ClC-6. Proc Natl Acad
Sci USA 103, 13854-13859.
303 Further reading

Poet, M, Kornak, U, Schweizer, M, Zdebik, AA, Scheel, O, Hoelter, S, Wurst, W, Schmitt, A,
Fuhrmann, JC, Planells-Cases, R, Mole, S, E, Hübner, CA, Jentsch TJ. (2006). Lysosomal storage
disease upon disruption of the neuronal chloride transport protein ClC-6. Proc. Natl. Acad.
Sci. USA 103, 13854-13859.
Qi, M-L, Tagawa, K, Enokido, Y, Yoshimura, N, Wada, Y-ichi, Watase, K, Ishiura, S-ichi, Kanazawa,
I, Botas, J, Saitoe, M, Wanker, EE and Okazawa, H. (2007). Proteome analysis of soluble nuclear
proteins reveals that HMGB 1/2 suppress genotoxic stress in polyglutamine diseases.
Nature Cell Biology 9(4), 402-414.
Rasheed, AU, Rahn, H-P, Sallusto, F, Lipp, M, and Muller, G (2006) B helper T cell activity is confined
to CXCR5hiICOShi CD4 T cells and independent of CD57 expression. Eur. J. Immunol.,
36, 1892-1903
Rat Genome Sequencing Project Consortium. (2004). Genome Sequence of the Brown
Norway rat yields insights into mammalian evolution. Nature. 428, 493-521.
Rogaeva, E, et al. (2007) The neuronal sortilin-related receptor SORL1 is genetically associated
with Alzheimer’s Disease. Nat Genet 39, 168-177.
Rust MB, Alper SL, Rudhard Y, Shmukler BE, Vicente R, Brugnara C, Trudel M, Jentsch TJ,
Hubner CA (2007) Disruption of erythroid KCl-cotransporters alters erythrocyte volume and
partially rescues erythrocyte dehydration in SAD mice. J Clin Invest 117, 1708-1717.
Sasse-Klaassen, S, Probst, S, Gerull, B, Oechslin, E, Nurnberg, P, Heuser, A, Jenni, R, Hennies, HC,
Thierfelder, L. (2004). Novel gene locus for autosomal dominant left ventricular noncompaction
maps to chromosome 11p15. Circulation. 109, 2720-2723
Schaeper, U, Vogel, R, Chmielowiec, J, Huelsken, J, Rosário, M, and Birchmeier, W. (2007).
Distinct requirements for Gab1 in Met and EGF receptor signaling in vivo. Proc. Natl. Acad.
Sci. USA, 104, 15376-15381.
Scheel O, Zdebik AA, Lourdel S, Jentsch TJ (2005) Voltagedependent electrogenic chlorideproton
exchange by endosomal CLC proteins. Nature 436, 424-427.
Scheel, O, Zdebik, AA, Lourdel, S, Jentsch, TJ. (2005). Voltage dependent electrogenic chlorideproton
exchange by endosomal CLC proteins. Nature 436, 424-427.
Schmidt-Ullrich, R, Tobin, DJ, Lenhard, D, Schneider, P, Paus, R, Scheidereit, C. (2006). NF-B
transmits Eda A1/EdaR signalling to activate Shh and cyclin D1 expression, and controls
post-initiation hair placode down-growth. Development 133, 1045-1057.
Schneider. MA, Meingassner, JG, Lipp, M, Moore, HD. and Rot, A (2007) CCR7 is required for
the in vivo function of CD4+ CD25+ regulatory T cells. J. Exp. Med., 204, 735-745.
Sieber MA, Storm R, Martinez-de-la-Torre M, Muller T, Wende H, Reuter K, Vasyutina E,
Birchmeier C. Lbx1 acts as a selector gene in the fate determination of somatosensory and
viscerosensory relay neurons in the hindbrain. J Neurosci. 2007 May 2;27(18):4902-9.
Singh RK, Kefala G, Janowski R, Mueller-Dieckmann C, von Kries JP, Weiss MS (2005) The
high-resolution structure of LeuB (Rv2995c) from Mycobacterium tuberculosis. J Mol Biol
346, 1-11.
Sommermeyer, D, Neudorfer, J, Weinhold, M, Leisegang, M, Charo, J, Engels, B, Nößner, E.
Heemskerk, M, Schendel, DJ, Blankenstein, T, Bernhard, H, Uckert, W. (2006). Designer T cells
by T cell receptor replacement. Eur. J. Immunol. 36, 3052-3059.
Sommermeyer, D, Neudorfer, J, Weinhold, M, Leisegang, M, Engels, B, Noessner, E, Heemskerk,
M, Charo, J, Schendel, D, Blankenstein, Th, Bernhard, H and Uckert, W. (2006). Designer T cells
by T cell receptor replacement. Eur. J. Immunol. 36, 3052- 3059.
Stefan E, Wiesner B, Baillie GS, Mollajew R, Henn V, Lorenz D, Furkert J, Santamaria K,
Nedvetsky P, Hundsrucker C, Beyermann M, Krause E, Pohl P, Gall I, Maclntyre AN, Bachmann
S, Houslay MD, Rosenthal W, Klußmann E (2007) Compartmentalization of cAMP-dependet
signalling by phosphodiesterase- 4D is involved in the regulation of vasopressinmediated
water reabsorption in renal principal cells. J Am Soc Nephrol 18, 199-212.
Stein U, Arlt F, Walther W, Smith J, Waldman T, Harris ED, Mertins SD, Heizmann CW, Allard
D, Birchmeier W, Schlag PM, Shoemaker RH. (2006) The metastasis-associated gene S100A4
is a novel target of b-catenin/T-cell factor signalling in colon cancer. Gastroenterology 131,
1486-1500.
Stein U, Walther W, Stege A, Kaszubiak A, Fichtner I, Lage H. (2007) Complete in vivo reversal
of the multidrug resistance (MDR) phenotype by jet-injection of anti-MDR1 short hairpin
RNA-encoding plasmid DNA. Mol. Ther. (Epub ahead of print).
Further reading
Synowitz M, Glass R, Faerber K, Markovic D, Kronenberg G, Herrmann K, Schnermann J, Nolte
C, van Rooijen N, Kiwit J, and Kettenmann H. (2006) A1 Adenosine Receptors in Microglia
Control Glioblastoma-Host Interaction. Cancer Res 66, 8550-8557.
Thirunarayanan N, Cifire F, Fichtner I, Posner S, Benga J, Reiterer P, Kremmer E, Kölble K, Lipp
M. (2007) Enhanced tumorigenicity of fibroblasts transformed with human herpesvirus 8
chemokine receptor vGPCR by successive passage in nude and immunocompetent mice
Oncogene. 14, 523-532 [Epub 2007 Mar 23]
Vasyutina E, Lenhard DC, Wende H, Erdmann B, Epstein JA, Birchmeier C. RBP-J (Rbpsuh) is
essential to maintain muscle progenitor cells and to generate satellite cells. Proc Natl Acad
Sci U S A. 2007 Mar 13;104(11):4443-8.
von Vietinghoff, S, Busjahn, A, Schonemann, C, Massenkeil, G, Otto, B, Luft, FC, Kettritz, R.
(2006) Major histocompatibility complex HLA region largely explains the genetic variance
exercised on neutrophil membrane proteinase 3 expression. J Am Soc Nephrol. 17, 3185-3191.
von Vietinghoff, S, Tunnemann, G, Eulenberg, C, Wellner, M, Cardoso, MC, Luft, FC, Kettritz, R.
(2007) NB1 mediates surface expression of the ANCA antigen proteinase 3 on human neutrophils.
Blood. 109, 4487-4493.
Walisko, O, Izsvák, Z, Szabó, K, Kaufman, CD, Herold, S, Ivics, Z. (2006). Sleeping Beauty transposase
modulates cell-cycle progression through interaction with Miz-1. Proc. Natl. Acad.
Sci. USA 103, 4062-4067.
Walisko, O, Izsvák, Z, Szabó, K, Kaufman, CD, Herold, S, Ivics, Z. (2006). Sleeping Beauty transposase
modulates cell-cycle progression through interaction with Miz-1. Proc. Natl. Acad.
Sci. USA 103, 4062-4067.
Walther W, Arlt F, Fichtner I, Aumann J, Stein U, Schlag PM. (2007). Heat-inducible in vivo
gene therapy of colon carcinoma by human mdr1 promoter-regulated tumor necrosis factor-alpha
expression. Mol. Cancer Ther. 6, 236-243.
Walther W, Minow T, Martin R, Fichtner I, Schlag PM, Stein U. (2006) Uptake, biodistribution
and time course of naked plasmid- DNA trafficking after intratumoral in vivo jet-injection.
Hum. Gene Ther.17: 611-624.
Wengner, AM, Höpken, UE, Petrow, PK, Hartmann, S, Schurigt, U, Bräuer, R, and Lipp, M. (2007)
CXCR5- and CCR7-dependent lymphoid neo-genesis in a murine model of chronic antigeninduced
arthritis. Arthritis & Rheum. 56, 3271-3283
Wetzel C, Hu J, Riethmacher D, Benckendorff A, Harder L, Eilers A, Moshourab R, Kozlenkov A,
Labuz D, Caspani O, Erdmann B, Machelska H, Heppenstall PA, Lewin GR (2007) A stomatindomain
protein essential for touch sensation in the mouse. Nature 445:206-209.
Wildner H, Muller T, Cho SH, Brohl D, Cepko CL, Guillemot F, Birchmeier C. dILA neurons in the
dorsal spinal cord are the product of terminal and non-terminal asymmetric progenitor cell
divisions, and require Mash1 for their development. Development. 2006 Jun;133(11):2105-13.
Willem, M, Garratt, AN, Novak, B, Citron, M, Kaufmann, S, Rittger, A, Saftig, P, De Strooper, B,
Birchmeier, C, Haass, C. (2006). Control of peripheral nerve myelination by the gamma-secretase
BACE1. Science 314: 664-666.
Willimsky, G and Blankenstein, Th. (2005) Sporadic immunogenic tumours avoid destruction
by inducing T-cell tolerance. Nature 437,141-146.
Willimsky, G and Blankenstein, Th. (2007). The adaptive immune response to sporadic cancer.
Immunol. Rev., in press.
Willnow, TE, Hammes, A, Eaton, S. (2007) Lipoproteins and their receptors in embryonic
development – more than cholesterol clearance. Development, in press
Zimdahl, H, Nyakatura, G, Brandt, P, Schulz, H, Hummel, O, Fartmann, B, Brett, D, Droege, M,
Monti, J, Lee, YA, Sun, Y, Zhao, S, Winter, E, Ponting, C, Chen, Y, Kasprzyk, A, Birney, E, Ganten, D,
Hübner, N. (2004). A SNP map of the rat genome generated from transcribed sequences.
Science. 303, 807.
304

Credits
Written by Russ Hodge
Photography by Maj Britt Hansen
Graphic design by Nicola Graf
Paintings by Jeanne Mammen, from the collection on the campus in Berlin-Buch
Additional photography by
Silvia Bähring and Hakan Toka pages 226-234, 236, 237, 241
Peter Himsel pages 289, 290, 297
David Aussenhofer page 119
Catherine Adamidi pages 268-269
Ewan St. J. Smith page 153 (naked mole rat)
Russ Hodge pages 75, 95, 100, 103 (Helios), 137
Sharon Hodge
Lindbergh Picture Collection,
Manuscripts and Archives,
page 6 (Knut)
Yale University Library, images on
Dr. Tobias Engelsing and Carolin
Schulz, the Museum of Konstanz,
provided the image of the unique
page 90
Leiner Collection in Konstanz page 59
Robert Schröter page 126 (Berlin Wall)
For more information about the campus:
www.mdc-berlin.de
www.fmp-berlin.de
www.charite.de
www.bbb-berlin.de
Exposure and printing by ColorDruck GmbH, Leimen
www.colordruck.com
305 credits