RESEARCH REPORT - Peter MacCallum Cancer Centre

the first-ever CrYstaL struCture of a perforin-Like pore-forMinG protein
This exciting work was carried out in collaboration with the structural biology group headed by Professor James Whisstock at Monash University. It has long
been appreciated that perforin shares structural similarity with members of the complement membrane attack complex (circulating proteins of the immune system
that are important for killing certain bacteria). However, the mechanism used by this family of proteins to attach to and pierce target cell membranes has been
a complete mystery. We are now far closer to solving this important problem. Firstly, advanced bio-informatics approaches were used by Dr Michelle Dunstone
(Monash) to demonstrate that the family of proteins that includes perforin and complement is far more broadly represented in nature than previously appreciated,
with more than 500 members. This group of proteins has now been termed the MACPF (Membrane Attack Complex/Perforin) superfamily. In fact, MACPF
proteins are expressed by many species of bacteria, other human pathogens such as plasmodium (the parasite that causes malaria) and by higher organisms such
as the fruit fly Drosophila where the MACPF protein torso has an important role in promoting segmentation, a critical phase of development and growth of the
organism. Other MACPF proteins such as astrotactin are critical for accurate neuronal development and communication in animals and humans.
Lining up the amino acid sequences of such a diverse group of proteins for the first time enabled us to show that a small number (a few percent) of amino acids
had remained invariant over billions of years of evolutionary time. Remarkably, this finding also meant that a class of proteins used by bacteria to invade and
colonise hosts (including humans) are also used by the human immune system in its defence against those invading pathogens. We were also able to make the quite
remarkable observation that some of the key residues that undergo mutation in children born with the immune-deficiency disorder HLH are conserved all the way
back in evolutionary time to primordial organisms, including bacteria! This strongly implied that the mechanism of action of the MACPF proteins had not changed
materially over many hundreds of millions of years.
We then went on to determine the crystal structure
of a bacterial MACPF protein, Plu-MACPF from
Photorhabdus luminescens, to 2.0 angstrom
resolution. The MACPF domain revealed structural
similarity with pore-forming cholesterol-dependent
cytolysins (CDCs) from gram-positive bacteria, a
group of toxins whose mode of action is reasonably
well understood. This suggests that lytic MACPF
proteins may use a CDC-like mechanism to form
pores and disrupt cell membranes. Following
membrane attachment, the MACPF domain is
predicted to undergo a marked change in its 3D
conformation, enabling membrane spanning helices
to adopt beta sheet conformations and pierce the
target cell membrane.
This work is highly significant as it enables us
to predict how perforin functions. We now have
beautiful images and a model of how individual
perforin monomers bind together to form a pore in
3D. Ultimately, this information will be crucial for
understanding how perforin functions, leading to
new drugs to augment its anti-cancer activity.
[Reference: Rosado C. et al., (2007) Science 317:
Figure 1: Mechanisms of perforin-mediated cell death pathways.
1548–51, see publication number 240.]
manifested by abnormalities at both the lymphocyte
(presynaptic) and target cell (postsynaptic) levels.
However, the severe intrinsic defect in activity can
be partly rescued by expression in the physiological
setting of an intact CTL. These findings provided
the first direct evidence that hPRF-A91V is
functionally abnormal and a rationale for why
it may be responsible for disordered immune
homeostasis if inherited with another dysfunctional
perforin allele. Why such an abnormal allele has
persisted in human populations is now under
further investigation. [Reference: Voskoboinik I et
al., (2007) Blood 110: 1184–90, see publication
number 296.]
Apoptotic signal transduction by NK cell
granzyme B.
Human granzyme B (GraB) preferentially induces
apoptosis via Bcl-2-regulated mitochondrial
damage but can also directly cleave caspases
and caspase substrates in cell-free systems. How
hGraB kills cells when it is delivered by cytotoxic
lymphocytes (CL) and the contribution of hGraB
to CL-induced death is still not clear. We showed
that primary human NK cells, which specifically
used hGraB to induce target cell death, were able to
induce apoptosis of cells whose mitochondria were
protected by Bcl-2. Purified hGraB also induced
apoptosis of Bcl-2-overexpressing targets but only
when delivered at 5 to 10 times the concentration
required to kill cells expressing endogenous Bcl-2.
Caspases were critical in this process as inhibition
of caspase activity permitted clonogenic survival of
Bcl-2-overexpressing cells treated with hGraB or
hNK cells but did not protect cells that expressed
only endogenous Bcl-2. Our data therefore showed
that only hGraB triggers caspase activation via
mitochondria-dependent and mitochondriaindependent
mechanisms that are activated in a
hierarchical manner and that the combined effects
of Bcl-2 and direct caspase inhibition can block
cell death induced by hGraB or primary hNK cells.
[Reference: Sedelies K et al., (2008) Cell Death and
Differentiation 15: 708–17.]
11