primary prevention of coeliac disease - Associazione Italiana ...

Perspectives on Coeliac Disease
Volume I
PRIMARY PREVENTION
OF COELIAC DISEASE
THE UTOPIA
OF THE NEW MILLENNIUM?
EDITED BY
CARLO CATASSI
ALESSIO FASANO
GINO ROBERTO CORAZZA
AIC Press

PRIMARY PREVENTION
OF COELIAC DISEASE
THE UTOPIA OF THE NEW MILLENNIUM ?

The AIC Meeting was held in the Aula Scarpa of the University of Pavia, Italy, October 12,
2001.
Meeting participants (left to right, from back to front): P. Ciclitira, F. Koning, L. Sollid, R.
Anderson, R. Troncone, M. Mäki, A. Fasano, G. Gasbarrini, F. Cucca, M. Stern, C. Feighery,
G.R. Corazza, C. Catassi, P. Howdle, G. Holmes.

Perspectives on Coeliac Disease
Volume 1
PRIMARY PREVENTION
OF COELIAC DISEASE
THE UTOPIA OF THE NEW MILLENNIUM?
Editors
Carlo Catassi
Department of Pediatrics
University of Ancona
Ancona, Italy
Center for Celiac Research,
Baltimore, MD, USA
Alessio Fasano
Division of Pediatric
Gastroenterology,
Center for Celiac Research,
University of Maryland
Baltimore, MD, USA
Gino Roberto Corazza
Department of
Gastroenterology,
University of Pavia,
Pavia, Italy
AIC Press

Italian Coeliac Society, Via Picotti, 22 - 56124 Pisa, Italy
ã2003 Italian Coeliac Society. All rights reserved. This book is protected by copyright.
No part of it may be reproduced, stored in a retrieval system, or transmitted, in any form or by
any means, electronic, mechanical, photocopying, or recording, or otherwise, without the
written permission of the publisher.
Made in Italy
The material contained in this volume was submitted as previously unpublished material,
except in the instances in which credit has been given to the source from which some of the
illustrative material was derived.
Great care has been taken to maintain the accuracy of the information contained in the
volume. However, neither the AIC nor the Editors can be held responsible for errors or for
any consequences arising from the use of the information contained herein.

Foreward
It is a great honour for me to introduce the Proceedings of the Meeting on Coeliac
Disease that was held in Pavia on October 12, 2001.
First of all, I would like to thank and to remind the memory of Mr. Sergio Spinelli, a
man who strongly wanted and appreciated this meeting during its organization, who
unfortunately passed just a few months before the event took place.
I would also like to thank Prof. Corazza and Dr. Catassi for their hard work and
collaboration during the organization of the scientific programme and for selecting the
most important scientists who participated as speakers and chairmen.
Finally, I cannot forget Mrs. Marina Marengo, Mrs. Katia Pilo and the whole
editorial staff, especially Mr. Franco Lucchesi and Mrs. Giusy Cappellotto, for
professionally organizing the whole meeting and preparing all the informative material,
including this book.
The Meeting held in Pavia was one of the most important scientific events on
coeliac disease that have ever been organized, not only in Italy but also in Europe.
The Italian Coeliac Society (AIC) has always given a strong support to research and
felt it was time to organize a meeting for giving an up to date on the most important
recent findings on coeliac disease, both in Italy and other countries.
The aim of doing this was twofold: (a) to evaluate the level of the current knowledge
on the main aspects of coeliac disease by comparing different scientific working groups
and (b) to identify the possible topics on which the research should be focused on in the
next future. Should these goals be reached, the life of the coeliacs will be easier
wherever.
I think that the quality level of both the speakers and the audience allowed to entirely
achieve these aims.
Adriano Pucci
President of the AIC
viii

Preface
Coeliac disease (CD) is a unique example in medicine of a genetically-based,
immune-mediated condition entirely curable. Treatment with the gluten-free diet
(GFD) is indeed followed by a full clinical and histological recovery (the true restitutio
ad integrum), with the patient shifting from a diseased state to a normal condition of
life. In typical cases, the response to treatment is sometimes prodigious, so that a
severely malnourished and miserable patient is rapidly transformed into a sturdy and
healthy subject. Even in atypical cases, such as those presenting with anaemia,
behavioural changes, or infertility, the diet treatment may miraculously be effective.
However, there is a price to pay for this success. For many populations glutencontaining
foods make a substantial contribution to daily energy intake and are
enjoyable to eat. Bread had, and still has, such a basic role in human nutrition to rise to a
symbolic value in the Holy Communion. The changes needed to begin and maintain a
GFD are substantial and have a major impact on daily life. It is therefore not surprising
that patients ask if a cure is on the horizon, wondering whether they will ever be able to
tolerate a normal diet.
Recently, some gliadin peptides that seem to have a primary role in activating the
cascade of immune responses leading to the celiac enteropathy were characterized by
different groups worldwide. These findings echoed in the lay media and were
enthusiastically interpreted as a clear-cut step toward the development of a coeliac
“vaccine”. Is this science fiction or a realistic hope? In the year 2000 the Board of the
Italian Coeliac Society (AIC) ripened the concept that it was time to give members a
correct information on this hot topic. Therefore, the AIC Scientific Committee
organized an innovative conference on primary prevention of CD, having two basic
ideas in mind: (1) invite all the experts working in the field (in order to hear the different
opinions); (2) translate the scientific language of the researchers into plain information
that could be understood by the lay members of the Society. On October 12, 2001 a
high-level scientific conference took place in Pavia, Italy, in the extraordinary scenario
of the old Aula Scarpa of the University. On October 13, 2001 the technical concepts
were summarized to a large lay audience in a meeting in Milan and society members
had plenty of time for asking questions to the international experts.
This book contains all the lectures presented at the Pavia meeting in order to give an
up-to-date scenario on the recent developments on treatment strategies alternative to
the GFD. In the first part of the book leading experts in the field report their results on
the identification of toxic gluten peptides. In the second part, the role of environmental
xi

xii
risk factors that can trigger the onset of CD, particularly infant feeding, is discussed in
detail. Finally, new strategies of treatment, based on either the introduction of
genetically detoxified grains or induction of oral tolerance to gluten, are presented. The
reader will appreciate that the “holy grail” for a possible coeliac cure is still far away.
Biochemically speaking, the complexity of gluten toxicity is disarming. Furthermore,
the lack of an animal model of disease greatly hampers the investigations on the CD
pathophysiology at the mucosal level. In spite of these problems, the route has been
traced and primary prevention of CD should not be considered an utopia anymore at the
beginning of the third millennium. We hope that this volume will stimulate researchers
and clinicians alike to pursue this goal.
We express our gratitude to all speakers and chairmen at the Pavia meeting, and to
the academic authorities of the University of Pavia. We gratefully acknowledge the
support of the AIC National Board, particularly of Ing. Spinelli, who was one of the
promoters of this meeting that unfortunately passed away. The publication of this
volume would not have been possible without the expertise of the AIC Editorial Board,
particularly of Dr. Franco Lucchesi and Ms. Giusy Cappellotto. We also thank Dr.
Elisabetta Fabiani for her valuable help with several aspects of this book.
C. Catassi A. Fasano G.R. Corazza

Contents
Current understanding of the basis for coeliac disease ............................................
1
Ludvig M Sollid
The identification of toxic T cell stimulatory gluten response peptides early
in coeliac disease ...................................................................................................
13
Frits Koning
Toxic gluten peptides in coeliac disease identified by in vivo gluten challenge:
a single dominant T cell epitope ? .........................................................................
17
Robert P Anderson
Role of A-gliadin 31-49 peptide ........................................................................... 27 27
Paul J Ciclitira
The association of the HLA-DQ molecules with coeliac disease in the
Saharawi: an evolutionary perspective on coeliac disease ....................................
31
Francesco Cucca, Carlo Catassi
Primary prevention of coeliac disease by favourable infant feeding practices .....
43
Anneli Ivarsson, Lars Åke Persson, Olle Hernell
Mechanisms of oral tolerance: lessons for coeliac disease ? ................................
61
Conleth Feighery
Genetically detoxified grains in coeliac disease ....................................................
75
Federico Biagi, Antonio Di Sabatino, Jonia Campanella, Gino Roberto Corazza
Immunotherapy of coeliac disease: where do we stand? .......................................
83
Carmen Gianfrani, Mauro Rossi, Giuseppe Mazzarella, Francesco Maurano,
VirginiaSalvati, Delia Zanzi, Salvatore Auricchio, Riccardo Troncone
The most recent advances on gluten toxicity in coeliac disease ............................
89
Gino Roberto Corazza, Carlo Catassi, Alessio Fasano
xiii

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives
on Coeliac Disease, vol. 1, AIC Press, pp 1-11
Current understanding of the basis for coeliac disease
Ludvig M. Sollid
Institute of Immunology, University of Oslo, Rikshospitalet, N-0027 Oslo, Norway
lmsollid@labmed.uio.no
Introduction
Coeliac disease (CD) has received increased attention in recent years. The disease is
more common than previously thought with a prevalence of about 1:130-1:300 in
1
Western societies . It is an acquired disorder; it may be diagnosed in early childhood
with classical symptoms like diarrhoea and malabsorption, but it may also be diagnosed
1
later in life often with symptoms that do not directly allude to a gut disease .
CD develops because of intolerance to ingested wheat gluten (consisting of the
subcomponents gliadin and glutenin) or related proteins from rye and barley. There is a
2
chronic inflammation in the small intestine with resultant flattening of the mucosa .
The current treatment is a life-long gluten exclusion diet which often impairs the quality
of life of those who are affected. For this reason many coeliacs ask for novel treatment
modalities and methods to prevent the disease.
CD belongs to the group of chronic inflammatory diseases with multifactorial
aetiology where genetic and environmental components are involved. Among these
disorders CD stands out as a particularly good model. This paper will briefly cover
some of the recent advances in the understanding of this disorder.
Genetic factors
3
A high prevalence (10%) among first degree relatives of CD patients and a high
4
concordance rate of 70-100% in monozygotic twins indicate that susceptibility to
develop CD is strongly influenced by inherited (genetic) factors. Both HLA and non-
HLA genes contribute to the genetic predisposition, and assuming a multiplicative
model of disease genetics it has been estimated that the overall importance of non-HLA
5-6
genes is greater than that of HLA genes . These figures should be interpreted with
caution, however, as increased sharing of environmental factors by the sibs would tend
to overestimate the role of the non-HLA genes.
The majority of CD patients carry the DRB1*0301-DQA1*0501-DQB1*0201
1

2
THE BASIS OF COELIAC DISEASE
haplotype (the DR3, DQ2 haplotype) or are DRB1*11/12-DQA1*0505-
DQB1*0301/DRB1*07-DQA1*0201-DQB1*0202 heterozygotes (carry the DR5-
7-11
DQ7/DR7-DQ2 haplotypes) . The chains encoded by DQA1*0501 and DQA1*0505
differ by one residue in the leader peptide, and the DQbchains encoded by DQB1*0201
and DQB1*0202 differ by one residue in the membrane proximal domain. These
substitutions are highly unlikely to have any functional consequence. CD patients with
the above mentioned DR-DQ combinations share the same functional DQ molecule on
the cell surface, encoded by genes carried in cis (e.g. DQA1*05 and DQB1*02 carried
on the same haplotype) or trans position (e.g. DQA1*05 carried on a different
12
haplotype to DQB1*02) . Accumulating evidence suggests that the DR3-DQ2, the
DR7-DQ2 and the DR5-DQ7 haplotypes have a close evolutionary relationship.
Fragments of DNA flanking the DQA1 gene of the DR3-DQ2 haplotype have been
identified on the DR5-DQ7 haplotype, and fragments of DNA flanking the DQB1 gene
13-14
of the DR3-DQ2 haplotype have been identified on the DR7-DQ2 haplotype . The
genetic information in the DQ subregion of the DR3-DQ2 haplotype is thus reestablished
in DR5-DQ7/DR7-DQ2 heterozygotes, although the sequence information
is split between two chromosomes. Susceptibility to CD therefore likely depends on an
interaction between at least two genes on the DR3-DQ2 haplotype that are reunited in
DR5-DQ7 / DR7-DQ2 heterozygous individuals. The DQA1 and DQB1 genes are the
primordial candidates since their products interact to form an HLA class II heterodimer
and since they are situated close to a putative recombination site.
Almost all the CD patients who are DQA1*05 and DQB1*02 negative bear the
DRB1*04, DQA1*03, DQB1*0302 haplotype (i.e. DR4-DQ8 haplotype) and it is likely
that these patients have an HLA association which is different to those who are DQ2
positive. Although it is less clear what the primary disease susceptibility determinant of
15
the DR4-DQ8 haplotype is, most data favour DQ8 .
Overall the existing data suggest that the susceptibility to develop CD is primarily
associated with two conventional DQ molecules DQ(a1*05,b1*02) (=DQ2) and to a
lesser extent DQ(a1*03,b1*0302) (=DQ8). DQ molecules bind peptides and present
these to CD4+ T helper cells carrying the abT cell receptor (TCR). The genetic
evidence thus points towards a central role of CD4+ T cells in controlling the disease
development in CD.
Genome wide linkage studies in CD have indicated numerous susceptibility regions
with weak genetic effects, and the indications are strongest for susceptibility genes
16-17
located at 5qter and 11qter . However, no disease associations have been established
so far for any gene of these regions. The only gene for which there are relatively
18-20
consistent reports on disease association is the CTLA4 gene on chromosome 2q33 .
CTLA4 is involved in down regulating T-cell responses, and the A allele of the +49
dimorphism is associated with increased CTLA4 expression and enhanced control of T
21-22
cell proliferation . Thus the +49 dimorphism is a prime suspect for the observed
genetic effect, although other polymorphisms in linkage disequilibrium cannot be ruled
out. In this respect it is worth noting that the CD28 and ICOS genes whose gene
products are central players in T and B cell activation, are located very close to the
CTLA4 gene.

THE BASIS OF COELIAC DISEASE
3
Environmental factors
Gluten is obviously a critical environmental factor in CD. Whether other
environmental factors are also involved is still an open question. Gut infections may
well be involved. Adenovirus 12 has been in the forefront among the candidate
microorganisms. This virus was originally proposed as a candidate because of partial
23
linear homology over 12 amino acids in a virus protein and a-gliadin . Based on our
current understanding of the pathogenesis of coeliac disease, the rationale for the
candidacy of this virus is weak, and there is in fact very little epidemiological data
24
supporting a role for this virus .
Peptide binding to the CD associated DQ2 and DQ8 molecules
Both HLA class I and class II molecules bind peptides in a groove located in their
25
membrane distal part . Stable binding is achieved by multiple hydrogen bonds
between amino acids of HLA and peptide main chain atoms. Many polymorphic
variants of HLA molecules exist. Amino acid residues which differ between the
polymorphic variants are clustered around the peptide binding site where they
contribute to the formation of specific binding pockets. Side chains of amino acids of
the peptide (so-called anchor residues) fit into these pockets and their interaction with
HLA contribute to the binding of the peptide. The binding site of HLA class II
molecules, in contrast to HLA class I molecules, is open at both ends allowing the
bound peptides to protrude. The class II peptide ligands thus vary in length. The
interactions with HLA mainly take place in a core region of nine residues. Within this
region side chains of amino acids in positions P1, P4, P6, P7 and P9 dock into pockets of
the class II binding site. The chemistry and size of the various pockets vary between the
different class II alleles so that some amino acids are preferred and some are not
preferred. DQ2 has a unique preference for binding peptides with negatively charged
26-28
side chains at the three middle anchor positions . The binding motif of DQ8 is
different from that of DQ2, but DQ8 also displays a preference for binding negatively
29-31
charged residues at several positions (i.e. P1, P4 and P9) . Hence, both the DQ2 and
DQ8 molecules share a preference for negatively charged residues at some of their
anchor positions.
Preferential T cell recognition of gluten peptides presented by DQ2 and Dq8
The observation that gluten reactive CD4+ TCRabT cells can be isolated and
propagated from intestinal biopsies of CD patients has been instrumental for recent
32
achievements . Strikingly, such T cells of patients carrying the DR3-DQ2 haplotype
were found to recognise gluten fragments presented by the DQ2 molecule rather than
32-33
by other HLA molecules of the patients . Both DR3-DQ2 positive and DR5-
DQ7/DR7-DQ2 positive antigen presenting cells (i.e. carrying the DQA1*05 and
DQB1*02 genes in cis or in trans configuration) are able to present the gluten antigen to
these patient T cells. Likewise, T cells isolated from small intestinal biopsies of DQ2
negative, DR4-DQ8 positive patients predominantly recognise gluten-derived peptides
32-33
when presented by the DQ8 molecule . Taken together, these results allude to

4 THE BASIS OF COELIAC DISEASE
1
presentation of gluten peptides in the small intestine as the mechanism by which DQ2
1
and DQ8 confer susceptibility to CD. HLA molecules are also important for
determining the repertoire of peripheral T cells during maturation in the thymus. A
thymic effect of the same DQ molecules on the TCR repertoire selection is, however,
not excluded by these results.
Importance of gluten deamidation for T cell recognition
Wheat gluten is a mixture of a large number of gliadin and glutenin polypeptides.
Generally, gluten proteins are rich in proline and glutamine residues while many other
amino acids, including glutamic and aspartic acid, are unusually rare. Proteins of the
gliadin fraction can be subdivided according to their sequence into the a-, g-, and w-
34
gliadins . Initially it was difficult to reconcile the DQ2 (and DQ8) binding motifs with
presentation of gluten peptides, as gluten proteins have an unusual scarcity of
negatively charged residues. A clue to help explain this paradox came from the
observation that the stimulatory capacity of gliadin preparations for gliadin specific
intestinal T cells was significantly enhanced following treatment at high temperatures
35
and low pH . These conditions are known to cause non-specific deamidation of
glutamines to glutamic acid and may thus convert gliadin from a protein with very few
peptides with the potential to bind to DQ2/DQ8 into one with many. An important and
general role for deamidation of gluten for T cell recognition was sustained by analysis
of the response pattern of a panel of polyclonal, gliadin specific T cell lines derived
36
from biopsies . All the cell lines responded poorly to a gliadin antigen prepared under
conditions of minimal deamidation (chymotrypsin-digestion), when compared to the
same antigen that had been further heat-treated in an acidic environment.
The characterisation of gluten epitopes recognised by intestinal T cells has extended
the knowledge about the importance of deamidation for their T cell recognition. Of the
epitopes characterised until now, most of them (DQ2-g-gliadin-I, DQ2-a-gliadin-I and
DQ2-a-gliadin-II) fail to stimulate T cells in their native form, but are potent antigens
when a single glutamine residue is exchanged with glutamic acid in certain positions.
For one DQ8 restricted epitope (DQ8-a-gliadin-I), the T cell recognition is augmented
37
by introduction of negatively charged residues whereas this is not seen for another
38
DQ8 restricted epitope (DQ8-glutenin-I) . These data demonstrate that most, but not
all gluten specific intestinal T cells from CD patients recognise gluten proteins only
after they have undergone deamidation.
Tissue transglutaminase deamidates gluten peptides in vivo
There is accumulating evidence that the deamidation in vivo is mediated by the
39-40
enzyme tissue transglutaminase (tTG) . tTG is expressed in many different tissues
and organs; in the small intestine it is mainly expressed just beneath the epithelium in
39
the gut wall . The activity of tTG in the small intestinal mucosa in CD patients with
41
untreated disease is elevated compared to controls . The enzyme is present both
intracellularly and extracellularly, and in the extracellular environment tTG has been
demonstrated to play a role in extracellular matrix assembly, cell adhesion and wound

THE BASIS OF COELIAC DISEASE
5
42
healing . The calcium dependent transglutaminase activity of tTG catalyses selective
43
crosslinking or deamidation of protein-bound glutamine residues . In contrast to the
non-enzymatically mediated deamidation that results in a near random deamidation of
the often numerous glutamine residues in gliadin peptides, tTG appears to carry out an
ordered deamidation of some few specific glutamines. In all of the known major DQ2
and DQ8 restricted gluten epitopes recognised by gut T cell of adult patients, there are
36-37-44
glutamic acid residues modified by tTG which is important for T cell recognition .
Interestingly, the deamidation of glutamine residues that are not targeted by tTG (e.g.
40-45
by acid treatment) can be deleterious for T cell recognition . This suggests that
deamidation in vivo is mediated by tTG. This idea is further supported by the results of
experiments where T cell lines have been established from biopsies challenged with a
minimally deamidated gliadin antigen (chymotrypsin-digested). In all but one of 18
adult patients, the established T cell lines only barely responded to the chymotrypsindigested
gliadins, but efficiently recognized the in vitro tTG-treated variants of the
46
same gliadins .
Normally we do not mount immune responses to edible proteins. Moreover,
experimental animal models have demonstrated that oral administration of antigen
47
usually results in systemic hyporesponsiveness to the same antigen . This
phenomenon, which is termed oral tolerance, is believed to occur because of active
tolerization towards edible proteins. In keeping with this thinking, oral tolerance to
gluten in CD patients may not have been established properly or is broken. Given the
preferential intestinal T cell response to deamidated gluten fragments in CD patients, it
may be that deamidation is involved in the perturbation of the oral gluten tolerance.
Deamidation increases the binding affinity of gliadin peptides for DQ2 from poor but
36-
significant binders, to epitopes with reasonable, but by no means exceptional affinity
44
. The moderate binding affinity of these epitopes concurs with the finding that they do
not carry optimal anchors in all the anchor positions. Interestingly, T cell clones specific
for the DQ2-g-gliadin-I, DQ2-a-gliadin-I and DQ2-a-gliadin-II epitopes generally fail
to recognise native peptides at higher concentrations that should compensate for their
lower binding affinity for DQ2. Thus, concurrent with the increase in the binding
affinity for DQ2 caused by deamidation of the gliadin peptide, there likely is a change in
the conformation1of the gliadin/DQ2 complexes. Oral tolerance to antigens ingested in
47
high doses is usually established by T cell anergy or deletion . It is possible that T cells
specific for native gluten sequences are usually anergised or deleted, and that phlogistic
T cell responses are effectively mounted to novel gluten epitopes being created in an
inflamed environment by the help of tTG.
How many gluten T cell epitopes?
There exist several epitopes in gluten that are recognized by small intestinal T cells
35
of CD patients . Recent results of the author's laboratory and the laboratory of Frits
Koning in Leiden indicate that there may be more than ten distinct DQ2 restricted
epitopes. The existence of multiple epitopes raises several interesting questions: are
only some of the epitopes pathogenic and thereby relevant to explain the HLA
association? Are responses towards some of the epitopes generated during the early
phases of disease development, while the responses to others are a result of epitope

6 THE BASIS OF COELIAC DISEASE
spreading? Are different epitopes recognized by distinct groups of patients (e.g.
children vs. adults)? Are some epitopes more relevant to disease as responses to them
are found in the majority of the patients or because there is a higher precursor frequency
of T cells in the lesion specific for these epitopes? The answers to most of these
questions must await further investigations. At present we know that for the DQ2-agliadin-I
and DQ2-a-gliadin-II epitopes, intestinal T cell reactivity is found in most if
44
not all adult DQ2+ patients , whereas for the DQ2-g-gliadin-I epitope intestinal T cell
36
reactivity is found in fewer DQ2+ patients . Less is known about the DQ8 restricted
epitopes because few DQ8 positive patients have been tested so far. However, the DQ8-
37
a-gliadin-I appears to be frequently recognized . What causes the variance in
responsiveness to the different epitopes and whether this reflects qualitative or
quantitative differences between the patients are presently unclear.
Formation of the coeliac lesion
The evidence discussed above provides strong evidence that CD4+ TCRab+ T cells
in the lamina propria are central for controlling the immune response to gluten that
produces the immunopathology of CD. The knowledge of the events down-stream of T
cell activation is, however, still incomplete. Knowing how the immune system usually
utilizes a multitude of effector mechanisms for fighting its opponents, it is reasonable to
believe that there may well be multiple effector mechanisms involved in the creation of
the coeliac lesion. Adding to the complexity, recent in vitro organ culture studies have
indicated that gluten exerts additional immune relevant effects that are independent of T
48-49
cell activation . Some of these effects have rapid kinetics and conceivably the direct
effects of gluten may facilitate subsequent T cell responses.
Cytokines produced by lamina propria Cd4+ T cells may be involved in the
increased crypt cell proliferation and the increased loss of epithelial cells. IFN-g
induces macrophages to produce TNF-a. TNF- activates stromal cells to produce
50
KGF, and KGF causes epithelial proliferation and crypt cell hyperplasia . IFN-gand
51
TNF-acan jointly have a direct cytotoxic effect on intestinal epithelial cells . It is also
conceivable that IELs and in particular gdT cells play a role in the epithelial cell
52
destruction by recognizing MIC molecules induced by stress .
Alterations of the extracellular matrix can also distort the epithelial arrangement as
the extracellular matrix provides the scaffold on which the epithelium lies. Enterocytes
adhere to basement membrane through extracellular matrix receptors so that
modification or loss of the basement membrane can result in enterocyte shedding.
There is evidence for increased extracellular matrix degeneration in CD, and this may
53
be an important mechanism for the mucosal transformation found in CD . The
increased production of metalloproteinases by subepithelial fibroblasts and
macrophages is likely to be directly or indirectly induced by cytokines that are released
from activated T cells.
Coeliac patients on a gluten containing diet have increased levels of serum
54-55
antibodies to a variety of antigens including gluten and to tTG . We do not as yet
know whether the autoantibodies play a role in the pathogenesis of CD. The tTG
antibodies can inhibit the activity of tTG. This can cause villous atrophy by blocking

THE BASIS OF COELIAC DISEASE
7
interactions between mesenchymal cells and epithelial cells during the migration of
56
epithelial cells and fibroblasts from the crypts to the tips of the villi . Moreover, the
tTG antibodies may modulate the deamidating activity of tTG either in an inhibiting or
57
promoting fashion . Further research will tell us what role these antibodies play.
Translation of the new knowledge into therapy
The increasing insight into the molecular and cellular basis of CD should give
benefits to the patients. The knowledge on which gluten epitopes are recognized by gut
T cells should allow the methods by which gluten free foods are assessed to be
improved. Moreover, the new knowledge should uncover novel targets for therapy.
There are already some attractive possibilities. Activation of CD4+ gluten specific T
cells appears to be a critical checkpoint in the development of CD, and interference with
this step in the pathogenesis should be an effective way to control the disease. One
possibility, which is basically an extension of today's treatment with a gluten free diet, is
to produce wheat that is devoid of T cell epitopes, either by breeding programs or
transgenic technology. Success by use of classical breeding already seems unlikely as
epitopes are found in both a-gliadins and g-gliadins which are encoded by the Gli-1 and
Gli-2 loci located on chromosomes 1 and 6, respectively. The classical breeding
approach is further complicated by the hexaploid nature of wheat. tTG is a target for
intervention because of its critical role in generating gluten T cell epitopes. Inhibitors of
tTG activity exist and likely inhibitors suitable as drugs can be developed. The biggest
problem with this approach is that tTG inhibitors may have unacceptable side effects.
tTG is involved in many different physiological processes including programmed cell
42
death . Another strategy would be to aim at the gluten specific T cells. If coeliacs have a
normal oral tolerance to native gluten proteins, but a broken tolerance to deamidated
gluten peptides, exposing the gut immune system to already deamidated peptides may
establish oral tolerance to these deamidated gluten peptides as well. This approach will
utilize the body's own mechanism to silence T cells. Alternatively, one could try to
directly silence gluten specific T cell by using soluble dimers of HLA/ peptide
complexes, which have been demonstrated to induce antigen specific apoptosis
58
because of inappropriate T cell stimulation . The central role of DQ2 and DQ8 in
presenting gluten peptides offers yet another target for intervention. Blocking the
binding-sites of these HLA molecules would prevent presentation of disease inducing
gluten peptides. The challenge with this approach will be to find an efficient way to
target and block the binding sites of DQ molecules, which are continuously synthesized
by antigen presenting cells. This approach, blocking of peptide presentation, has also
been suggested as therapy for other HLA associated diseases. CD should be better
suited to this approach than many other HLA associated diseases because drug delivery
in the gut is easy compared for example to joints in rheumatoid arthritis or islet cells in
type 1 diabetes.
Whatever new therapeutic modality is introduced in CD, it will have to prove better
than the current gluten free diet regime, also when coming to its long-term safety. This
fact must be taken into consideration when devising new treatments. Although there are
already interesting therapeutic principles with a good, rational basis that can be tested, it
may for this reason take some years before a new treatment becomes reality.

8 THE BASIS OF COELIAC DISEASE
Acknowledgements
Studies in the author's laboratory are funded by grants from the Research Council
of Norway, the European Commission (BMH4-CT98-3087, QLRT-1999-00037, QLRT-
2000-00657), Medinnova, the Jahre Foundation and EXTRA funds from the Norwegian
Foundation for Health and Rehabilitation.
References
1. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac
disease: an evolving spectrum. Gastroenterology 2001; 120: 636-51.
2. Sollid LM. The molecular basis of celiac disease. Annu Rev Immunol 2000; 18: 53-
81.
3. Ellis A. Coeliac disease: previous family studies. In: McConnell RB, editor. The
genetics of coeliac disease. Lancaster: MTP press, 1981: 197-9.
4. Polanco I, Biemond I, van Leeuwen A, Schreuder I, Meera Khan P, Guerrero J et al.
Gluten sensitive enteropathy in Spain: Genetic and environmental factors. In:
McConnell RB, editor. The genetics of coeliac disease. Lancaster: MTP Press,
1981: 211-31.
5. Risch N. Assessing the role of HLA-linked and unlinked determinants of disease. Am
J Hum Genet 1987; 40: 1-14.
6. Petronzelli F, Bonamico M, Ferrante P, Grillo R, Mora B, Mariani P et al. Genetic
contribution of the HLA region to the familial clustering of coeliac disease. Ann
Hum Genet 1997; 61: 307-17.
7. Keuning JJ, Pena AS, van Leeuwen A, van Hooff JP, van Rood JJ. HLA-DW3
associated with coeliac disease. Lancet 1976; i: 506-8.
8. Ek J, Albrechtsen D, Solheim BG, Thorsby E. Strong association between the HLA-
Dw3-related B cell alloantigen -DRw3 and coeliac disease. Scand J Gastroenterol
1978; 13: 229-33.
9. Mearin ML, Biemond I, Pena AS, Polanco I, Vazquez C, Schreuder GT et al. HLA-
DR phenotypes in Spanish coeliac children: their contribution to the understanding
of the genetics of the disease. Gut 1983; 24: 532-7.
10. Trabace S, Giunta A, Rosso M, Marzorati D, Cascino I, Tettamanti A et al. HLA-
ABC and DR antigens in celiac disease. A study in a pediatric Italian population.
Vox Sang 1984; 46:102-6.
11. Tosi R, Vismara D, Tanigaki N, Ferrara GB, Cicimarra F, Buffolano W et al.
Evidence that celiac disease is primarily associated with a DC locus allelic
specificity. Clin Immunol Immunopathol 1983; 28: 395-404.
12. Sollid LM, Markussen G, Ek J, Gjerde H, Vartdal F, Thorsby E. Evidence for a
primary association of celiac disease to a particular HLA-DQ a/bheterodimer. J
Exp Med 1989; 169: 345-50.
13. Lin L, Jin L, Kimura A, Carrington M, Mignot E. DQ microsatellite association
studies in three ethnic groups. Tissue Antigens 1997; 50: 507-20.
14. Lin L, Jin L, Lin X, Voros A, Underhill P, Mignot E. Microsatellite single nucleotide
polymorphisms in the HLA-DQ region. Tissue Antigens 1998; 52: 9-18.
15. Sollid LM, Thorsby E. HLA susceptibility genes in celiac disease: genetic mapping

THE BASIS OF COELIAC DISEASE
9
and role in pathogenesis. Gastroenterology 1993; 105: 910-22.
16. Greco L, Corazza G, Babron MC, Clot F, Fulchignoni-Lataud MC, Percopo S et al.
Genome search in celiac disease. Am J Hum Genet 1998; 62: 669-75.
17. Greco L, Babron MC, Corazza GR, Percopo S, Sica R, Clot F et al. Existence of a
genetic risk factor on chromosome 5q in Italian coeliac disease families. Ann Hum
Genet 2001; 65: 35-41.
18. Djilali-Saiah I, Schmitz J, Harfouch-Hammoud E, Mougenot JF, Bach JF, Caillat-
Zucman S. CTLA-4 gene polymorphism is associated with predisposition to coeliac
disease. Gut 1998; 43: 187-9.
19. Naluai AT, Nilsson S, Samuelsson L, Gudjonsdottir AH, Ascher H, Ek J et al. The
CTLA4/CD28 gene region on chromosome 2q33 confers susceptibility to celiac
disease in a way possibly distinct from that of type 1 diabetes and other chronic
inflammatory disorders. Tissue Antigens 2000; 56: 350-5.
20. Holopainen P, Arvas M, Sistonen P, Mustalahti K, Collin P, Maki M et al.
CD28/CTLA4 gene region on chromosome 2q33 confers genetic susceptibility to
celiac disease. A linkage and family-based association study. Tissue Antigens 1999;
53: 470-5.
21. Kouki T, Sawai Y, Gardine CA, Fisfalen ME, Alegre ML, DeGroot LJ. CTLA-4 gene
polymorphism at position 49 in exon 1 reduces the inhibitory function of CTLA-4
and contributes to the pathogenesis of Graves' disease. J Immunol 2000; 165: 6606-
11.
22. Ligers A, Teleshova N, Masterman T, Huang WX, Hillert J. CTLA-4 gene expression
is influenced by promoter and exon 1 polymorphisms. Genes Immun 2001; 2: 145-
52.
23. Kagnoff MF, Austin RK, Hubert JJ, Bernardin JE, Kasarda DD. Possible role for a
human adenovirus in the pathogenesis of celiac disease. J Exp Med 1984; 160:
1544-57.
24. Mahon J, Blair GE, Wood GM, Scott BB, Losowsky MS, Howdle PD. Is persistent
adenovirus 12 infection involved in coeliac disease? A search for viral DNA using
the polymerase chain reaction. Gut 1991; 32:1114-6.
25. Madden DR. The three-dimensional structure of peptide-MHC complexes. Annu
Rev Immunol 1995; 13: 587-622.
26. Johansen BH, Vartdal F, Eriksen JA, Thorsby E, Sollid LM. Identification of a
putative motif for binding of peptides to HLA-DQ2. Int Immunol 1996; 8: 177-82.
27. van de Wal Y, Kooy YMC, Drijfhout JW, Amons R, Koning F. Peptide binding
characteristics of the coeliac disease-associated DQa1*0501, b1*0201. molecule.
Immunogenetics 1996; 44: 246-53.
28. Vartdal F, Johansen BH, Friede T, Thorpe C, Stevanovic S, Eriksen JA et al. The
peptide binding motif of the disease associated HLA-DQa1*0501, b1*0201.
molecule. Eur J Immunol 1996; 26: 2764-72.
29. Godkin A, Friede T, Davenport M, Stevanovic S, Willis A, Jewell D et al. Use of
eluted peptide sequence data to identify the binding characteristics of peptides to
the insulin-dependent diabetes susceptibility allele HLA-DQ8 (DQ 3.2). Int
Immunol 1997; 9: 905-11.
30. Kwok WW, Domeier ML, Raymond FC, Byers P, Nepom GT. Allele-specific motifs
characterize HLA-DQ interactions with a diabetes-associated peptide derived from

10 THE BASIS OF COELIAC DISEASE
glutamic acid decarboxylase. J Immunol 1996; 156: 2171-7.
31. Lee KH, Wucherpfennig KW, Wiley DC. Structure of a human insulin peptide-HLA-
DQ8 complex and susceptibility to type 1 diabetes. Nat Immunol 2001; 2: 501-7.
32. Lundin KEA, Scott H, Hansen T, Paulsen G, Halstensen TS, Fausa O et al. Gliadinspecific,
HLA-DQa1*0501,b1*0201. restricted T cells isolated from the small
intestinal mucosa of celiac disease patients. J Exp Med 1993; 178: 187-96.
33. Lundin KEA, Gjertsen HA, Scott H, Sollid LM, Thorsby E. Function of DQ2 and
DQ8 as HLA susceptibility molecules in celiac disease. Hum Immunol 1994; 41:
24-7.
34. Shewry PR, Tatham AS, Kasarda DD. Cereal proteins and coeliac disease. In:
Marsh M, editor. Coeliac disease. Oxford: Blackwell Scientific Publications, 1992:
305-48.
35. Lundin KEA, Sollid LM, Norén O, Anthonsen D, Molberg Ø, Thorsby E et al.
Heterogenous reactivity patterns of HLA-DQ-restricted small intestinal T-cell
clones from patients with celiac disease. Gastroenterology 1997; 112: 752-9.
36. Sjöström H, Lundin KEA, Molberg Ø, Körner R, McAdam SN, Anthonsen D et al.
Identification of a gliadin T-cell epitope in coeliac disease: general importance of
gliadin deamidation for intestinal T-cell recognition. Scand J Immunol 1998; 48:
111-5.
37. van de Wal Y, Kooy YM, van Veelen PA, Peña SA, Mearin LM, Molberg Ø et al. Small
intestinal T cells of celiac disease patients recognize a natural pepsin fragment of
gliadin. Proc Natl Acad Sci U S A 1998; 95: 10050-4.
38. van de Wal Y, Kooy YM, van Veelen P, Vader W, August SA, Drijfhout JW et al.
Glutenin is involved in the gluten-driven mucosal T cell response. Eur J Immunol
1999; 29: 3133-9.
39. Molberg Ø, McAdam SN, Körner R, Quarsten H, Kristiansen C, Madsen L et al.
Tissue transglutaminase selectively modifies gliadin peptides that are recognized
by gut-derived T cells. Nat Med 1998; 4: 713-7.
40. van de Wal Y, Kooy Y, van Veelen P, Peña S, Mearin L, Papadopoulos G et al.
Selective deamidation by tissue transglutaminase strongly enhances gliadinspecific
T cell reactivity. J Immunol 1998; 161: 1585-8.
41. Bruce SE, Bjarnason I, Peters TJ. Human jejunal transglutaminase: demonstration
of activity, enzyme kinetics and substrate specificity with special relation to gliadin
and coeliac disease. Clin Sci 1985; 68: 573-9.
42. Aeschlimann D, Paulsson M. Transglutaminases: protein cross-linking enzymes in
tissues and body fluids. Thromb Haemost 1994; 71: 402-15.
43. Folk JE. Mechanism and basis for specificity of transglutaminase-catalyzed e-g-
(glutamyl) lysine bond formation. Adv Enzymol Relat Areas Mol Biol 1983; 54:1-
56.
44. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Vader W, Kooy YM et al. The
intestinal T cell response to a-gliadin in adult celiac disease is focused on a single
deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000; 191:
603-12.
45. Quarsten H, Molberg Ø, Fugger L, McAdam SN, Sollid LM. HLA binding and T cell
recognition of a tissue transglutaminase - modified gliadin epitope. Eur J Immunol
1999; 29: 2506-14.

THE BASIS OF COELIAC DISEASE
11
46. Molberg Ø, McAdam S, Lundin KEA, Kristiansen C, Arentz-Hansen H, Kett K et al.
T cells from celiac disease lesions recognize gliadin epitopes deamidated in situ by
endogenous tissue transglutaminase. Eur J Immunol 2001; 31: 1317-23.
47. Komagata Y, Weiner HL. Oral tolerance. Rev Immunogenet 2000; 2: 61-73.
48. Maiuri L, Picarelli A, Boirivant M, Coletta S, Mazzilli MC, De Vincenzi M et al.
Definition of the initial immunologic modifications upon in vitro gliadin challenge
in the small intestine of celiac patients. Gastroenterology 1996; 110: 1368-78.
49. Maiuri L, Auricchio S, Coletta S, De Marco G, Picarelli A, Di Tola M et al. Blockage
of T-cell costimulation inhibits T-cell action in celiac disease. Gastroenterology
1998; 115: 564-72.
50. Bajaj-Elliott M, Poulsom R, Pender SL, Wathen NC, MacDonald TT. Interactions
between stromal cell--derived keratinocyte growth factor and epithelial
transforming growth factor in immune-mediated crypt cell hyperplasia. J Clin
Invest 1998; 102: 1473-80.
51. Deem RL, Shanahan F, Targan SR. Triggered human mucosal T cells release tumour
necrosis factor - alpha and interferon-gamma which kill human colonic epithelial
cells. Clin Exp Immunol 1991; 83: 79-84.
52. Groh V, Steinle A, Bauer S, Spies T. Recognition of stress-induced MHC molecules
by intestinal epithelial gdT cells. Science 1998; 279: 1737-40.
53. Daum S, Bauer U, Foss HD, Schuppan D, Stein H, Riecken EO et al. Increased
expression of mRNA for matrix metalloproteinases-1 and -3 and tissue inhibitor of
metalloproteinases-1 in intestinal biopsy specimens from patients with coeliac
disease. Gut 1999; 44: 17-25.
54. Mäki M. The humoral immune system in coeliac disease. In: Howdle PD, editor.
Coeliac disease. London: Bailliére Tindall, 1995: 231-49.
55. Dieterich W, Ehnis T, Bauer M, Donner P, Volta U, Riecken EO et al. Identification
of tissue transglutaminase as the autoantigen of celiac disease. Nat Med 1997; 3:
797-801.
56. Halttunen T, Mäki M. Serum immunoglobulin A from patients with celiac disease
inhibits human T84 intestinal crypt epithelial cell differentiation. Gastroenterology
1999; 116: 566-72.
57. Sollid LM, Scott H. New tool to predict celiac disease on its way to the clinics.
Gastroenterology 1998; 115: 1584-6.
58. Appel H, Seth NP, Gauthier L, Wucherpfennig KW. Anergy induction by dimeric
TCR ligands. J Immunol 2001; 166: 5279-85.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 13-16
The identification of toxic T cell stimulatory gluten
response peptides early in coeliac disease
Frits Koning
On behalf of: Willemijn Vader*, Yvonne Kooy*, Peter van Veelen*, Arnoud de Ru*,
#
Diana Harris*, Willemien Benckhuijsen*, Luisa Mearin , and Jan Wouter Drijfhout*
#
Departments of *Immunohematology and Blood Transfusion and Paediatrics,
Leiden University Medical Centre, Leiden, The Netherlands
It is generally accepted that coeliac disease (CD) is caused by uncontrolled T cell
responses to gluten peptides that are presented by HLA-DQ2 and/or -DQ8 molecules.
In recent years five gluten peptides have been identified that stimulate T cell clones
1-5
derived from small intestinal biopsies of CD patients . An important breakthrough has
been the demonstration that deamidation of the gluten peptides by the enzyme tissue
transglutaminase (tTG) is either required for, or enhances, T cell recognition of four of
1-8
these peptides . The conversion of glutamine into glutamic acid by deamidation
generates negative charges in gluten peptides that facilitate binding to HLA-DQ2 and -
9-11
DQ8 molecules , thus providing a molecular basis for the well established association
between CD and HLA-DQ2/8.
Two major issues, however, remain unsolved. First, all the studies so far have
investigated the gluten specific T cell response in adult patients. It is unclear, therefore,
whether the identified gluten peptides are also involved in T cell activation earlier in the
disease process. Second, it is not known whether deamidation of gluten peptides is
required for the breaking of oral tolerance or that it merely enhances T cell reactivity
towards gluten.
To investigate these matters we have now carried out an extensive investigation of
the gluten specific response in children with recent onset CD. Twenty-two Caucasian
DQ2-(DQA1*0501/DQB1*02) positive patients with a confirmed diagnosis of CD
were included in the present study. Their age at diagnosis (first small bowel biopsy) was
between 1 and 9 years (average age 3.6 years ± 1.8). Biopsies were collected from these
patients and used to generate gluten reactive T cell lines. Subsequently, gluten specific
T cell clones were generated from T cell lines of nine patients. The T cell lines and
clones were tested for reactivity against gluten and gluten that has been treated with tTG
(tTG-gluten hereafter). Three patterns of reactivity were observed:
1) T cells that did not respond to gluten but did respond to tTG-gluten;
2) T cells that responded to both gluten and tTG-gluten;
3) T cells that did respond to gluten but not to tTG-gluten. In 8 out of 9 patients tTGdependent
clonal T cell responses were found while in 7 patients specific responses to
13

14
T CELL RESPONSES TO GLUTEN PEPTIDES
non-deamidated gluten were also observed. These results indicate that a large
proportion of the gluten specific responses are directed to deamidated gluten but that
responses to non-deamidated gluten are also common.
Extensive testing of the T cell clones against the 3 known HLA-DQ2 restricted T
cell stimulatory gliadin derived peptides indicated that the large majority of the T cell
clones did not respond to these peptides, and were thus likely reactive towards yet
unidentified gluten peptides. To characterize these novel peptides we have used two
different methods. First, we purified and characterized T cell stimulatory gluten
epitopes from pepsin/trypsin digests of (tTG-) gluten by rpHPLC and mass
1-2
spectrometry as described . This method led to the characterization of 3 novel T cell
stimulatory peptides. Second, we tested the T cell clones against a set of 250 synthetic
gluten peptides, representing gliadin and glutenin sequences. This method led to the
identification of 3 additional novel T cell stimulatory peptides.
Subsequently we tested the response of the T cell clones to the identified peptides in
deamidated and non-deamidated form. The T cell response towards 3 peptides required
prior deamidation. In contrast, the response to 2 peptides was found to be largely
indifferent to deamidation. Finally, deamidation abolished the response to the sixth
peptide. Thus, the effect of deamidation by tTG on gluten specific T cell stimulation is
heterogeneous and can be positive, neutral and negative.
Subsequently the gluten specific T cell clones of all patients were tested against the
3-5
previously characterized HLA-DQ2 restricted gluten peptides as well as against the
peptides reported in the present study. While responses to some peptides were found in
one patient only, responses to other peptides were found in various patients and these
may thus represent more immunodominant peptides. T cell responses towards the a-
gliadin peptides which have been reported to be immunodominant in adult patients
were found in three paediatric patients, among whom the identical twins. In these 9
patients we observed 8 distinct reactivity patterns towards the gluten peptides (Table).
Altogether, these results indicate a highly diverse response against the gluten peptides.
These results indicate a discrepancy between the specificity of adult and paediatric
gluten specific T cell responses. While immunodominant responses to a particular a-
gliadin peptide are found in adult patients, the response in paediatric patients appears
more diverse. Our results also indicate that T cell responses to peptides other than the
immunodominant a-gliadin peptide can lead to disease. Moreover, our results
demonstrate that responses to non-deamidated peptides are frequently found,
suggesting that native gluten peptides are immunogenic in celiac disease patients.
The discrepancy between the specificity of the adult and paediatric gluten specific T
cell response could be explained by a deamidation driven narrowing of the gluten
response towards immunodominant T cell stimulatory peptides after initiation of
disease by responses towards a diverse repertoire of gluten peptides.
Acknowledgements
This study was financially supported by the European Community project no. BMH
CT-98, a grant from the Dutch Digestive Disease Foundation and the University of
Leiden.

T CELL RESPONSES TO GLUTEN PEPTIDES
15
DB
JB
NB
SB
JP
NP
MS
NV
SV
Glia- Glia- Glia- Glia-
2a 20a 9a 1g
Glia-
30g
Glt-
17
Glt-
156
Glu-
5
Glu-
21
Table. Overview of paediatric T cell responses to gluten peptides.
References
1. Van de Wal Y, Kooy YM, Van Veelen PA, Pena AS, Mearin LM, Molberg O, et al. Small
intestinal T cells of celiac disease patients recognize a natural pepsin fragment of gliadin.
Proc Natl Acad Sci U.S.A 1998; 95: 10050-4.
2. Van de Wal Y. Kooy YM, Van Veelen P, Vader W, August SA, Drijfhout JW, et al. Glutenin
is involved in the gluten-driven mucosal T cell response. Eur J Immunol 1999; 29: 3133-
9.
3. Sjostrom H, Lundin KE, Molberg O, Korner R, McAdam SN, Anthonsen D, et al.
Identification of a gliadin T-cell epitope in coeliac disease: general importance of gliadin
deamidation for intestinal T-cell recognition. Scand J Immunol. 1998; 48: 111-5.
4. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Vader W, Kooy YM, et al. The
intestinal T cell response to alpha-gliadin in adult celiac disease is focused on a single
deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000; 191: 603-12.
5. Anderson RP, Degano P, Godkin AJ, Jewell DP, Hill AV. In vivo antigen challenge in
celiac disease identifies a single transglutaminase-modified peptide as the dominant A-
gliadin T-cell epitope. Nat Med 2000; 6: 337-42.
6. Molberg O, McAdam SN, Korner R, Quarsten H, Kristiansen C, Madsen L, et al. Tissue
transglutaminase selectively modifies gliadin peptides that are recognized by gut-derived
T cells in celiac disease. Nat Med 1998; 4: 713-7.
7. Van de Wal Y, Kooy Y, van Veelen P, Pena S, Mearin L, Papadopoulos G, et al. Selective
deamidation by tissue transglutaminase strongly enhances gliadin-specific T cell
reactivity. J Immunol 1998; 161: 1585-8.
8. Quarsten H, Molberg O, Fugger L, McAdam SN, Sollid LM. HLA binding and T cell
recognition of a tissue transglutaminase - modified gliadin epitope. Eur J Immunol 1999;
29: 2506-14.
9. Van de Wal Y, Kooy YMC, Drijfhout JW, Amons R, Koning F. Peptide binding
characteristics of the coeliac disease-associated DQ(alpha1*0501, beta1*0201)
molecule. Immunogenetics 1996; 44: 246-53.
10. Vartdal F, Johansen B.H., Friede T, Thorpe CJ, Stevanovic S, Eriksen JE, et al. The
peptide binding motif of the disease associated HLA-DQ (alpha 1* 0501, beta 1* 0201)
molecule. Eur J Immunol 1996; 26: 2764-72.
11. Kwok WW. Domeier ML, Raymond FC, Byers P, Nepom GT. Allele-specific motifs
characterize HLA-DQ interactions with a diabetes - associated peptide derived from
glutamic acid decarboxylase. J Immunol. 1996; 156: 2171-7.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 17-26
Toxic gluten peptides in coeliac disease identified by
in vivo gluten challenge: A single dominant T
cell epitope?
Robert P. Anderson
The Royal Melbourne Hospital Autoimmunity and Transplantation Division, Walter and Eliza
Hall Institute c/o Post Office RMH, Victoria, Australia 3050
b.anderson@wehi.edu.au
The realization that as many as 1% of Europeans and North Americans are affected
by coeliac disease adds urgency to understanding the immunopathogenesis of coeliac
disease and developing a rational therapy without the inconvenience of a gluten free
diet.
Antigen-specific immunotherapy is highly effective in animal models of antigendriven
immune-mediated disease. Because human immune-mediated diseases are
usually only available for study when the immune response is well established, it has
been impossible to be sure what the initiating antigens are, or whether there are critical
dominant T cells epitopes that trigger disease.
Coeliac disease is unique among human immune-mediated disease since gluten is
known to maintain disease, and treatment is successful when dietary gluten is excluded.
If coeliac disease is triggered by one critical gluten component it may be possible to
develop antigen-specific therapies or preventive “vaccines”. This chapter discusses the
rationale for in vivo strategies that have allowed the identification of a critical gliadin
peptide that is the dominant coeliac-specific T cell epitope in a model alpha-gliadin
protein. A molecular model for the interaction of HLA-DQ2, dominant gliadin epitope
and T cell receptor is presented, and preliminary data indicating the potential of altered
peptide ligands to antagonize this interaction is discussed.
CD4 T cells have the potential to initiate immunopathology
Gluten and HLA-DQ2 are definitively implicated in the aetiopathogenesis of
1
coeliac disease . Since HLA-DQ2 presents peptides to CD4 T cells, there has been
intense interest in defining the specificity and phenotype of T cells in coeliac disease
that recognize gluten peptides restricted by HLA-DQ2.
CD4 T cells play a pivotal role in coordinating immune responses. CD8 T cell and
many B cell responses require CD4 T cell “help” provided directly or through CD4 T
cell activation of antigen presenting cells (particularly dendritic cells). In coeliac
disease, CD4 T cells predominantly secreting Th1-like cytokines appear in the small
17

18
A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
intestinal lamina propria within hours of gluten challenge in subjects on gluten free diet.
Initial damage to the intestine following gluten challenge in treated coeliac disease is
likely to be due to cytokines (for example interferon gamma [IFNg] and tumour
necrosis factor [TNF]) secreted by CD4 T cells that activate macrophages causing
2
release of inflammatory mediators and other cytokines that activate other cell types . In
addition to immunopathology induced by gluten-specific CD4 T cells, increasing
3
evidence suggests that gluten may also directly activate macrophages , possibly
facilitating induction of T cell responses to dietary antigen.
Disease chronicity and epitope spreading: relevance to T cell epitope mapping in
coeliac disease
Experimental autoimmune diseases, such as experimental allergic encepahalitis
(EAE), can be initiated by immunization of susceptible mice with adjuvant together
with myelin basic protein (MBP) or a specific peptide derived from this protein. This
4
peptide corresponds to the “dominant” T cell epitope of MBP . Exactly what qualities,
in addition to affinity for HLA and resistance to proteolysis during processing, that lead
5
to one peptide in an antigenic protein to be dominant are not fully understood . As EAE
progresses, a variety of other peptides (sub-dominant epitopes) derived from MBP and
other myelin-associated proteins are recognized by specific T cells in a process termed
4, 6
epitope spreading (also seen in human multiple sclerosis ). Specificity of T cell
responses can also be shaped by preferential presentation of peptides by B cells
following antigen uptake via B cell receptor or Fc receptor-mediated uptake of
7
antibody-antigen complexes . Post-uptake processing of antigen may also be altered by
the cytokine milieu in established immune responses giving rise to a different repertoire
8
of peptides being presented by antigen-presenting cells .
Interestingly, when MBP-specific T cells are transferred to a healthy recipient,
MBP-specific T cells are initially abundant in spleen and blood. Just before onset of
EAE, MBP-specific T cells disappear from spleen and blood, and become abundant at
9
the site of antigen (central nervous system) . Consistent with this, oral administration
of antigen is initially followed by proliferation of antigen specific T cells in gut-
10
associated as well as systemic lymphoid tissue . Furthermore, T cells with identical
11
specificity are found in murine gut epithelium, lamina propria and the thoracic duct .
Taken together, these studies in mice indicate that T cells specific for dominant or
subdominant epitopes appear at different time points, and may be located in different
anatomical sites according to the chronicity of the immune response. Contrary to the
12
widely held view in coeliac disease research , T cells with identical specificities are
present in gut and extra-intestinal sites such as blood in the early phase of immune
responses caused by gut antigen.
Multiple toxic peptides in coeliac disease identified by multiple methodologies
In coeliac disease, a variety of in vivo, ex vivo and in vitro methods have been

A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
19
1, 13
exploited to search for “toxic” gluten peptides . A-gliadin, the first fully sequenced
14
wheat gliadin protein , has been used as the archetypal “toxic gluten” protein source
for most of these peptides (see Fig. 1). More recently, recombinant gliadins and
peptides corresponding to parts of cDNA-derived gliadin sequences have been studied
15
. In view of the diversity of model systems, some with no relationship to coeliac
disease (for example fetal intestinal explants), it is not surprising that a variety of
different peptides have been defined as toxic. In other studies, non-specific markers of
toxicity (for example epithelial cell height) have been used as surrogates for
immunological toxicity. If the primary hypothesis is that coeliac disease is a T cell
mediated disease, rational identification of “toxic” gluten peptides in coeliac disease
should ideally begin with assays designed to detect HLA-DQ2 restricted CD4 T cells
that secrete Th1-associated cytokines (for example interferon gamma).
Intestinal T cell clones for identification of toxic gluten peptides
More recently, intestinal T cell clones from duodenal biopsies of coeliac patients on
gluten free diet pulsed with protease-digested gliadin have been used to search for
15-17
gliadin-specific T cell epitopes . These studies have revealed that intestinal T cell
clones raised from gliadin-pulsed coeliac intestinal biopsies predominantly recognise
12
deamidated gliadin epitopes . T cell epitopes generally correspond to gliadin peptides
deamidated by tissue transglutaminase (tTG). tTG is induced with inflammation and
apoptosis, and irreversibly cross links proteins and peptides via glu-lys isopeptide
18
bonds, or directly deamidates glutamine residues to glutamate . Intestinal tissue,
particularly if inflamed, is rich in tTG. Introduction of glutamate in gliadin T cell
12, 15
epitopes by tTG greatly enhances their binding to HLA-DQ2 .
T cell epitope mapping in immune-mediated diseases
In human immune-mediated diseases, T cell epitopes have been mapped using
synthetic 15-20mers overlapping by 10 residues spanning known antigenic proteins in
assays of peripheral blood T cells. In coeliac disease, it has been contended that gliadin-
1
specific peripheral blood T cells are qualitatively different from intestinal T cells .
Gliadin-specific intestinal T cell clones are generally HLA-DQ2 restricted while
19
peripheral blood T cell clones are HLA-DR, -DQ or -DP restricted . However, these
studies were performed before the realization that deamidation of gliadin was
important for T cell recognition. Hence, contemporary understanding of toxic gluten
peptides in the context of coeliac disease as an HLA-DQ2 associated CD4 T cellmediated
disease has relied upon identification of epitopes of intestinal T cell clones
using protease-digested gliadin with or without deamidation by tTG.
Unfortunately, it is impossible to know whether epitopes of T cell clones are
dominant or subdominant, or whether a clone is specific for only a part of a polyclonal T
cell response in vivo. Furthermore, recent studies to identify epitopes of gliadin T cell
15
clones have used chymotrypsin-digested gliadins even though chymotrysin
selectively cleaves peptide bonds following bulky hydrophobic aminoacids (the same
aminoacids known to be anchor residues at the N - and C-terminal end of the HLA-DQ2
1
binding motif ), raising the possibility that bioactive peptides may be artefactually

20 A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
truncated or destroyed.
Hence, a variety of data support the hypothesis that coeliac disease could be
initiated by a Th1-like HLA-DQ2 restricted CD4 T cell response focused on particular
(probably deamidated) gluten peptide/s. However, current methods relying upon T cell
clones, or challenge of coeliac tissue in vivo or ex vivo are incapable of defining
whether peptides are dominant or subdominant in the immunopathogenesis of coeliac
disease.
In vivo gluten challenge and peripheral blood epitope mapping to define T cell
20
epitope hierarchies
Antigen challenge of sensitized subjects results in activation of cognate memory T
cells. Memory T cells tend to reside at anatomical sites where their cognate antigen was
previously encountered. Recently, memory T cells have been divided into “effector”
- + 21
(cytokine-secreting, CCR7 ) and lymph node-homing subtypes (CCR7 ) . We
+
reasoned that gluten challenge in healthy HLA-DQ2 coeliac subjects following a strict
gluten free diet would reactivate gluten-specific memory T cells. The initial intestinal
inflammation documented by others may be driven by effector memory T cells, but
subsequent appearance of T cells in peripheral blood, perhaps homing back to the gut,
may reflect proliferation of gluten-specific T cells in lymphoid tissue during the days
after antigen exposure. Hence, the kinetics and frequency of epitope-specific peripheral
blood T cells might reveal dominant versus subdominant gluten epitopes.
Synthetic 15mer peptides overlapping by 10 aminoacids corresponding to the
composite aminoacid sequence (rather than cDNA-derived sequence) of A-gliadin
(Fig. 1) with or without in vitro deamidation by tTG were studied in overnight ex vivo
A 1. VRVPVPQLQP QNPSQQQPQE QVPLVQQQQF PGQQQ QFPPQ QPYPQPQPFP SQQPYLQLQP
B
P
A 61. PQ PQLPYPQ PQ SFPPQQPY PQPQPQYSQP QQPISQQQ AQ QQQQQQQQQQ
B [..] P R Q Q[..]
A111. QQQILQQILQ QQLIPC MDVV LQQHNIAHAR SQVLQQSTYQ LLQELCCQHL WQIPEQSQCQ
B R GS Q Q
A171. AIHNVVHAII LHQQQ KQQQQ PSSQVSFQQP L QQYPLGQGS FRPSQQNPQA
B [..] [..] Q S
A221. QGSVQPQQLP QFEEIRNLAL QTLPAMCNVY I APYC TIAPF GIFGTN
B P [..]
14
Fig. 1. Aminoacid sequence of A-gliadin used to derive overlapping 15mer peptides in
20
gluten challenge studies (A), and residues that deviate from the consensus sequence derived
from 61 Genbank Triticum aestivum a - and a/b-gliadin cDNAs using ClustalW (B) ([..]
indicates polymorphic insertion).

A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
21
interferon gamma (IFNg) ELISpot assays using peripheral blood mononuclear cells
(PBMC). IFNgELISpot assays define the frequency of cells secreting IFNg, and are
capable of detecting peptide-specific T cell frequencies as low as 5-10 per million
+
PBMC. Blood was collected from HLA-DQ2 healthy non-coeliac and coeliac subjects
after gluten free diet for at least 4 weeks, and then in the 12 days after commencing
gluten challenge (200 g white bread daily for ½ day [n=1], 3 days [n=10], or 10 days
[n=1]). Short-term gluten challenge was generally well tolerated by coeliac subjects,
11/12 subjects were able to consume 200g gluten bread for 3 or more days, and 8/12 had
only mild symptoms. One subject had abdominal cramps and vomited within 3 hours of
the first 2 slices of bread. All symptoms resolved within 1 to 3 days after ceasing gluten
challenge.
Prior to gluten challenge, no A-gliadin peptide elicited responses in the IFNg
ELISpot. However, in 11/12 coeliac subjects and 0/4 healthy control subjects there was
induction of IFNgELISpot responses on day 4-8 for one pool of overlapping peptides
only when treated with tTG. In all cases, IFNginduction was attributed to 2 peptides
overlapping by 10 aminoacids (A-gliadin 56-75). Aminoacid sequencing demonstrated
that only one glutamine residue in A-gliadin 56-75 was susceptible to tTG-mediated
deamidation (Q65). Truncations of A-gliadin 56-75 treated with tTG indicated that
residues 64-68 (PQLPY) were critical for bioactivity, and that the 17mer
QLQPFPQPQLPYPQPQS (57-73) was the optimal peptide. Bioactivity of
QLQPFPQPELPYPQPQS was identical to tTG-treated QLQPFPQPQLPYPQPQS,
demonstrating that bioactivity of this peptide was dependent upon a single deamidated
glutamine residue (QE65). Immunomagnetic bead depletion of PBMC prior to addition
of peptide showed that this peptide specific immune response was due to CD4 T cells
E
expressing the b,
7
but not the aintegrin protein, indicating that A-gliadin 57-73 QE65
specific T cells express the abintegrin
4 7
associated with homing to the intestinal lamina
22
propria . Pre-incubation of PBMC from HLA-DQ2 homozygous subjects with
antibody specific for HLA-DQ (but not HLA-DR or HLA-DP) blocks A-gliadin 57-73
QE65 responses, indicating HLA-DQ2 restriction.
In one subject who consumed only 2 slices of bread, IFNg-secreting T cells specific
for A-gliadin 57-73 QE65 were induced on day 6 and persisted until day 12. In another
subject who consumed 4 slices of bread daily for 10 days, A-gliadin 57-73 QE65-
specific T cells were also present on day 6-8. IFNgsecretion was only induced by A-
gliadin 57-73 QE65 in subjects who consumed bread for 3 days. However, PBMC
collected on day 11 from the subject who consumed bread for 10 days secreted IFNgin
response to 6 out of 10 pools of tTG-treated A-gliadin peptides (one pool included A-
gliadin 57-73). Hence, T cell epitope spreading occurs as early as 10 days after
commencing antigen challenge.
These studies indicated:
1. In A-gliadin, there is a hierarchy of T cell epitopes with only one dominant T cell
epitope.
2. T cells specific for the dominant epitope are present in peripheral blood only

22 A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
transiently after gluten challenge in vivo, but have predictable kinetics.
3. Gluten challenge induces T cells specific for the same dominant A-gliadin
epitope in all HLA-DQ2 coeliac subjects.
Gliadin-specific intestinal T cell clones are focused on two peptides closely related
to A-gliadin 57-73 QE65
15
Arentz-Hansen et al. have shown that two peptides from a panel of 11 recombinant
a-gliadins are common epitopes for gliadin-specific intestinal T cell clones. T cell lines
specific for one or both of these peptides (QLQPFPQPELPY and
+
PQPELPYPQPELPY) were raised from intestinal biopsies of 17/17 HLA-DQ2
coeliac subjects on gluten free diet. These peptides closely resemble A-gliadin 57-73
QE65 and contain the core sequence PELPY, supporting the concept that intestinal and
peripheral blood T cell responses induced with gluten challenge are qualitatively
similar. More importantly, if intestinal and peripheral blood T cell responses share the
same specificity it is likely the ex vivo polyclonal peripheral blood T cell response will
be more informative than T cell clones or lines for studies of molecular specificity and
definition of T cell epitope hierarchies in coeliac disease.
Comparison of HLA-DQ2 restricted epitopes of T cell clones with A-gliadin 57-73
QE65
We have studied various HLA-DQ2 restricted gliadin epitopes (Tab. 1) in IFNg
ELISpot assays using PBMC from HLA-DQ2 coeliac subjects on day 6 of gluten
23
challenge (3 days, 200g gluten-containing bread daily) . A-gliadin 31-49 and GDB2
do not induce IFNgresponses above background levels, while A-gliadin 57-68 QE65 is
generally 25 % and a2 gliadin 62-75 QE65 QE72 60% as bioactive as A-gliadin 57-73
QE65 at optimal concentrations (25 µ g/ml). To determine whether T cells specific for
a2 gliadin 62-75 QE65 QE72 are part of the polyclonal T cell response to A-gliadin 57-
73 QE65, IFNgELISpot responses to a2 gliadin 62-75 QE65 QE72 or A-gliadin 57-73
QE65 (25 µ g/ml) alone or mixed together were compared. There was no difference
between A-gliadin 57-73 QE65 alone or mixed with a2 gliadin 62-75 QE65 QE72,
suggesting that in vivo T cells specific for a2 gliadin 62-75 QE65 QE72 and/or A-
gliadin 57-68 QE65 are simply part of the polyclonal T cell response targeting A-gliadin
57-73 QE65.
A-gliadin 57-68 (QE65)
QLQPFPQPELPY
a2 gliadin 62-75 (QE65 QE72) PQPELPYPQPELPY
A-gliadin 31-49
LGQQQPFPPQQPYPQPQPF
g-gliadin (GDB2)
QQLPQPEQPQQSFPEQERPF
Tab. 1. HLA-DQ2 restricted epitopes of gliadin-specific intestinal T cell clones
A-gliadin 57-73 QE65 is the optimal a-, a/b-gliadin polymorphism
23
A-gliadin 57-73 spans a highly polymorphic region of the a-, a/b-gliadins (Tab.2) .

A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
23
We have compared IFNgELISpot responses of all the a-, a/b-gliadin polymorphisms of
A-gliadin 57-73 found by SwissProt using the search sequence
XXXXXXXPQLPYXXXXX. Amongst these polymorphisms, bioactivity of
QLQPFPQPQLPYPQPQ[P,L] is identical to QLQPFPQPQLPYPQPQS after
d e a m i d a t i o n b y t T G o r s u b s t i t u t i o n Q E 6 5 . t T G - d e a m i d a t e d
PQLPYPQPQLPYPQPQ[P,L] is generally 80% as bioactive as
QLQPFPQPQLPYPQPQS, but substitution of P69 for serine or leucine reduces
bioactivity by 60%. Other polymorphisms are generally less than 20% as bioactive as
A-gliadin 57-73. Hence, tTG-deamidated or QE-substituted at position 9,
QLQPFPQPQLPYPQPQ[P,L,S] is the a-, a/b-gliadin polymorphism of A-gliadin 57-
73 with optimal bioactivity.
QLQPFPQ……………… PQLPYPQPQP
QLQPFPQ……………… PQLPYSQPQP
QLQPFPQ……………… PQLPYSQPQQ
QLQPFPQ……………… PQLSYSQPQP
QLQPFPQ……………… PQLPYLQPQP
QLQPFSQ……………… PQLPYSQPQP
QLQPFLQ……………… PQLPYSQPQP
QLQPFLQPQPFP……… PQLPYSQPQP
QLQPFPQPQLPYPQPQLPYPQLPYPQPQP
QLQPFPQPQLPYPQ… PQLPYPQPQP
QLQPFPQPQPFPPQLPYPQPQLPYPQPQP
QLQPFPR……………… PQLPYPQPQP
QLQPFPQPQPFP……… PQLPYPQPPP
QLQPFPQPQPFL……… PQLPYPQPQS
QLQPFPQPQPFP……… PQLPYPQPQS
QPQPFP…………………PQLPYPQTQP
QPQPFPPQ…………… PQLPYPQTQP
Tab. 2. Polymorphisms in the region of A-gliadin 57-73 among the 61 Triticum aestivum a-,
a/b-gliadin cDNA-derived protein sequences in Genbank
One reason for the selection of QLQPFPQPQLPYPQPQ[L,P,S] and
PQLPYPQPQLPYPQPQ[P,L] as potent epitopes may be that these sequences are
resistant to proteases. Chymotrypsin and pepsin both cleave peptide bonds after
hydrophobic residues. In vitro digestion of these peptides and less potent
polymorphisms with tyrosine at position 12 followed by proline, serine or leucine,
indicates that proline at position 13 prevents susceptibility to both chymotrypsin and
pepsin. Hence, it is possible that the specificity of the T cell response is shaped by the
susceptibility of gliadin peptides to proteases in the gut (and presumably in antigen
presenting cells).
Immune toxicity of wheat, rye and barley due to peptides cross-reactive with A-
gliadin 57-73, or epitope spreading initiated by B cells specific for sequences
adjacent to A-gliadin 57-73 QE65
The data we have gathered using PBMC following in vivo gluten challenge in

24 A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
coeliac disease are consistent with A-gliadin 57-73 deamidated by tTG being the
dominant a-, a/b-wheat gliadin T cell epitope. Searches for peptides including the core
sequence PQLPY have not revealed other wheat, rye or barley gluten sequences outside
the a-, a/b-gliadins of wheat. One possibility is that epitope spreading initiated by the
dominant epitope leads to other epitopes in gluten being recognized. Interestingly, B
cell epitopes have been identified that are immediately adjacent or including the C - and
24-25
N-terminal portions of A-gliadin 57-73 . The other HLA-DQ2 restricted epitopes, g-
gliadin GDB2 and A-gliadin 31-49 are also adjacent to or include sequences identical to
or very similar to the same gliadin B cell epitope (QXQPFP). Hence, B cell mediated
epitope spreading may be initiated by B cells given “help” by T cells specific for the
dominant epitope. Alternatively, wheat, rye and barley may contain prolamin
sequences that are cross-reactive with A-gliadin 57-73 but do not include PQLPY, or
there are other unique dominant epitopes.
Fine molecular specificity of A-gliadin 57-73 QE65-specific peripheral blood T
cells
Residues in A-gliadin 57-73 QE65 that determine bioactivity were mapped by
comparing bioactivity of A-gliadin 57-73 QE65 variants substituted with lysine at each
22
aminoacid . Bioactivity was abolished if lysine was substituted at position 8-11
(PELP), and substantially reduced with lysine at positions 4-7 (PFPQ) and 12-13 (YP).
Single aminoacid substituted variants of A-gliadin 57-73 QE65 with all naturally
occurring aminoacids except cysteine at positions 4-13 were synthesized and their
bioactivity compared to the parent peptide. Positions 4-7 were highly sensitive to
substitution, no more than three aminoacids at each position conveyed bioactivity
greater than 50% of the parent peptide. Positions 4-7 and 12-13 were less sensitive to
substitution but certain aminoacids such as proline, lysine and arginine tended to
abolish bioactivity. The fine molecular specificity of peripheral blood T cells for A-
gliadin 57-73 QE65 was similar amongst all of the eight coeliac subjects tested.
Antigen-specific therapy using A-gliadin 57-73 Qe65
A-gliadin 57-73 QE65 is clearly “the dominant” or “one of the dominant” gliadin T
cell epitopes in HLA-DQ2-associated coeliac disease. Therefore it is reasonable to
consider peptide therapeutics based on A-gliadin 57-73 QE65. Peptide delivered orally
or nasally would be simple and easy to formulate. Another possibility is design of
altered peptide ligand (APL) antagonists that differ from the parent peptide by one or
more residues and subtly alter T cell receptor signaling. APL antagonsists have the
26
potential to “switch off” or “skew” TH1 to TH2 responses in vitro . We have shown
that at least 5 single aminoacid-substituted variants of A-gliadin 57-73 QE65 with weak
agonist properties also significantly reduce IFNgELISpot responses to A-gliadin 57-73
QE65 when incubated in 5-fold excess with A-gliadin 57-73 QE65 (unpublished
observations). It is likely that multiple substitutions or “cocktails” of APL antagonsists
will be required for complete blockade of the polyclonal T cell response to A-gliadin
57-73 QE65.

A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
25
Conclusions
Gluten challenge in coeliac disease allows gliadin-specific T cell responses to be
measured and epitope hierarchies defined. This method has significant advantages over
T cell clones. One dominant T cell epitope has been defined in A-gliadin and is likely to
be the dominant a-, a/b-wheat gliadin T cell epitope. Whether there are other distinct
dominant T cell epitopes in other classes of gliadin in wheat, rye and barley, or whether
toxicity of these proteins is due to cross-reactivity with A-gliadin 57-73 QE65 is crucial
to the design of antigen-specific therapeutics. In either case, gluten challenge and
testing the bioactivity of peptides and proteins using PBMC will allow the importance
of particular epitopes to be determined. The efficacy of altered peptide ligand
antagonists targeting A-gliadin 57-73 QE65 provides proof of principle that peptide
therapeutics may be a practical approach to the treatment of coeliac disease without
resort to gluten free diet.
References
1. Sollid LM. Molecular basis of coeliac disease. Ann Rev Immunol 2000 ; 18: 53-81.
2. Nilsen EM, Jahnsen FL, Lundin KEA, Johansen F-E, Fausa O, Sollid LM, et al.
Gluten induces an intestinal cytokine response strongly dominated by interferon
gamma in patients with celiac disease. Gastroenterology 1998; 115: 551-63.
3.Tuckova L, Flegelova Z, Tlaskalova-Hogenova H, Zidek Z. Activation of
macrophages by food antigens: enhancing effect of gluten on nitric oxide and
cytokine production. J Leuk Biol 2000; 67: 312-8.
4.Tuohy VK, Yu M, Ling Y, Kawczak JA, Kinkel RP. Spontaneous regression of primary
autoreactivity during chronic progression of experimental autoimmune
encephalitis and multiple sclerosis. J Exp Med 1999; 189: 1033-42.
5. Yewdell JW, Bennink JR. Immunodominance in major histocompatibility complex
class I-restricted T lymphocyte responses. Ann Rev Immunol 1999 ; 17: 51-88.
6. Tuohy VK, Yu M, Weinstock-Guttman B, Kinkel RP. Diversity and plasticity of self
recognition during the development of multiple sclerosis. J Clin Invest 1997; 99:
1682-90.
7. McCluskey J, Farris AD, Keech CL, Purcell AW, Rischmueller M, Kinoshita G, et
al. Determinant spreading: lessons from animal models and human disease.
Immunol Rev 1998; 164: 209-29.
8. Drakesmith H, O'Neill D, Schneider SC, Binks M, Medd P, Sercarz E, et al. In vivo
priming of T cells against cryptic determinants by dendritic cells exposed to
interleukin 6 and native antigen. Proc Natl Acad Sci USA 1998; 95: 14903-8.
9. Flugel A, Berkowicz T, Ritter T, Labeur M, Jenne DE, Li Z, et al. Migratory activity
and functional changes of green fluorescent effector cells before and during
experimental autoimmune encephalitis. Immunity 2001; 14: 547-60.
10. Gutgeman I, Fahrer AM, Altman JD, Davis MM, Chien Y-h. Induction of rapid T cell
activation and tolerance by systemic presentation of an orally administered
antigen. Immunity 1998; 8: 667-73.
11. Arstila T, Arstila TP, Calbo S, Selz F, Malassis-Seris M, Vassalli P, et al. Identical T
cell clones are located within the mouse gut epithelium and lamina propria and

26 A DOMINANT EPITOPE IN GLUTEN PEPTIDES ?
circulate in the thoracic duct. J Exp Med 2000; 191: 823-34.
12. Molberg O, McAdam SN, Korner R, Quarsten H, Kristiansen C, Madsen L, et al.
Tissue transglutaminase selectively modifies gliadin peptides that are recognized
by gut-derived T cells in celiac disease. Nat Med 1998; 4:713-7.
13. Weiser H. The precipitating factor in coeliac disease. Balliere's Clin Gastroenterol
1995; 9:191-207.
14. Kasarda DD, Okita TW, Bernardin JE, Baeker PA, Nimmo CC, Lew EJ-L, et al.
Nucleic acid (cDNA) and amino acid sequences of type gliadins from wheat
(Triticum aestivum). Proc Natl Acad Sci USA 1984; 81: 4712-6.
15. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Van der Wal Y, Kooy YMC, et
al. The intestinal T cell response to agliadin in adult celiac disease is focused on a
single deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000;
191: 603-12.
16. Sjostrom H, Lundin KEA, Molberg O, Korner R, McAdam SN, Anthonsen D, et al.
Identification of a-gliadin T-cell epitope in coeliac disease: general importance of
gliadin deamidation for intestinal T-cell recognition. Scand J Immunol 1998 ; 48 :
111-5.
17. van de Wal Y, Kooy YM, van Veelen P, Vader W, August SA, Drijfhout JW, et al.
Glutenin is involved in the gluten-driven mucosal T cell response. Eur J Immunol
1999 ; 29: 3133-9.
18. Aeschilmann D, Paulsson M. Transglutaminases: protein cross-linking enzymes in
tissues and body fluids. Thromb Haemost 1994; 71: 402-15.
19. Gjertsen HA, Sollid LM, Ek J, Thorsby E, Lundin KEA. T cells from the peripheral
blood of coeliac disease patients recognize gluten antigens when presented by HLA-
DR, -DQ, or DP molecules. Scand J Immunol 1994 ; 39: 567-74.
20. Anderson RP, Degano P, Godkin AJ, Jewell DP, Hill AVS. In vivo antigen challenge
in celiac disease identifies a single transglutaminase-modified peptide as the
dominant A-gliadin T-cell epitope. Nat Med 2000; 6: 337-42.
21. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. Two subsets of memory T
lymphocytes with distinct homing potentials and effector functions T cell memory.
Nature 1999; 401: 708-12.
22. Anderson RP, Jewell DP, Hill AVS. Peripheral blood T cells induced by gluten
challenge in coeliac disease target a specific molecular motif and express a guthoming
integrin. Gastroenterology 2001; 120: A-684.
23. Anderson RP, Jewell DP, Hill AVS. Bioactivity of peptides homologous to the
coeliac-specific dominant A-gliadin T cell epitope. Gastroenterology 2011; 120: A-
683.
24. ten Dam M, Van de Wal Y, Mearin ML, Kooy Y, Pena S, Drijfhout JW, et al. Antialpha-gliadin
antibodies (AGA) in the serum of coeliac children and controls
recognize an identical collection of linear epitopes of alpha-gliadin. Clin Exp
Immunol 1998; 114: 189-95.
25. Osman AA, Gunnel T, Dietl A, Uhlig HH, Amin B, Fleckenstein B, et al. B cell
epitopes of gliadin. Clin Exp Immunol 2000; 121: 248-54.
26. Bielekova B, Martin R. Antigen-specific immunomodulation via altered peptide
ligands. J Mol Med 2001; 79: 552-65.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 27-29
Role of A-gliadin 31 - 49 peptide
Paul J Ciclitira
The Rayne Institute, St. Thomas’ Hospital, London, UK
Coeliac disease was first described in 1888. Samuel Gee who gave a classic
description of a wasting disease mostly affecting children and causing diarrhoea,
concluded at the end of his treaties, "if the patient can be cured at all, it must be by means
of the diet". It is unfortunate that his treatment at that time comprised thin gruel which
contained wheat flour. It remained until 1953, for Professor Dicke and colleagues to
publish their findings that wheat, rye and barley exacerbated the condition. They also
noted that oats was toxic, which was probably due to contamination of flour with wheat.
Dicke's experiment at that time compromised of making a slurry of the individual flours
in water, pouring them down an eight year old child 's throat and noting that the child
collapsed with vomiting, diarrhoea and near death. They extended this work to show
that the most toxic component of wheat resides in the endosperm or flour fraction.
However, the germ and husk were mildly toxic. This was probably due to
contamination of these fractions with the gluten proteins.
Flour contains proteins and water soluble starch fractions. Purified wheat starch is
often used for the production of gluten-free products. Wheat flour proteins can be
divided into four groups of which gliadins and glutenins together comprise 90% of the
protein. The albumins and globulins are minor protein fractions. The glutenins have
molecular weights ranging from 50 kilodaltons to many millions and entrap carbon
dioxide in the dough enabling it to prove and rise. It was in early studies that the alcohol
soluble or gliadin fraction of gluten was shown to be toxic in coeliac disease. The
gliadin proteins were sub-divided into four fractions termed a, b, gand waccording to
their relative electrophoretic mobility. More recently, they were classified as a, g, and w
fractions according to their N-terminal amino acid sequences. We previously purified,
using column chromatography with 20 litres eluent, gliadin into its a,b, g, and w
fractions. The complexity of the individual gliadins was shown in further experiments.
Having made polyclonal anti-sera against each gliadin fraction, we looked for
precipitin lines on Ouchterlony gels which revealed four precipitin lines to gliadin, with
an anti-gliadin serum. Precipitin lines were shared between gliadin subfraction
suggesting that they share antigenic determinants. We subsequently used an in vitro
27

28 ROLE OF A-GLIADIN 31-49 PEPTIDE
organ culture model. Small intestinal biopsies were placed on an organ culture grid and
incubated for 16-18 hours in the presence or absence of a putatively toxic fraction. We
measured enterocyte surface cell heights on sections of the biopsies. Using this
technique, we reported that each of the gliadin fractions, that is a, b, gand was
enterotoxic to coeliac small intestinal mucosa. Dr Kasarda subsequently went on to
sequence gliadin. The now classic sequence of A-gliadin is known to be 266 amino
1
acids long . This is an unusual protein in that it is composed by 10-15% proline and 30-
35% glutamine residues; this is an efficient method of storing nitrogen for the growing
plant.
We subsequently went on to develop an in vivo method of assessing the toxicity of
gliadin fractions. This involved passing a Quintron multiple biopsy capsule into the
proximal small intestine to which was taped an infusion tube. This allowed test
fractions to be infused and multiple small intestinal biopsies to be taken over eight
hours. These could then be sectioned and assessed blindly for villous height, crypt
depth ratio, epithelial surface cell height and intraepithelial lymphocyte count. Using
this system, we tested on different days 10 mg, 100 mg followed by 500 mg, and 1g of
gliadin. We showed that 10 mg produced no histological relapse, 100 mg minimal and
500 mg moderate histological relapse. With 1 g there was gross flattening of the mucosa
which commenced between 1-2 hours, was maximal at 4 hours and started to recover by
6 hours. We observed flattening of the mucosa, with complete loss of the normal villous
2
architecture . We have also used electron microscopy to show increase in the number of
lysosome-like bodies below the brush border, which occurs 1-2 hours after
commencing gluten challenges, suggesting that these may be relevant to the
pathogenesis of the condition. After a matter of hours, there is apoptosis and shedding
of the surface enterocytes. We hypothesize that there is absorption of gluten protein
between and through the enterocytes. Toxic gliadin peptides are presented by antigen
presenting cells to gluten sensitive T cells which exacerbate the condition through the
production of TH1 cytokines. We subsequently went on to perform in vivo experiments
assessing the toxicity of peptide fractions corresponding to amino acids 31-49, 202-
220, 3-21 of A-gliadin. We used, as a positive control, unfractionated gliadin. We tested
four subjects and showed that a fraction corresponding to amino acids 31-49 was
3
enterotoxic and caused flattening of the mucosa while the other two peptides did not .
In a variety of different experiments we showed that there was an induction of TH1
cytokines and inducible nitric oxide, infiltration of the mucosa with T cells, a change in
4
the villous morphology and enterocyte surface cell height .
Since gliadin is presented by antigen presenting cells to gluten sensitive T cells by
HLA DQ2 or DQ8, we established a cellular binding assay to study the binding of
gliadin peptides to DQ2. We demonstrated that the peptide, corresponding to amino
acid residues 31-49 of A-gliadin which caused in vivo toxicity, bound to DQ2. The two
other peptides which we tested in vivo, but which did not cause toxicity, did not bind to
5
DQ2 . To date, no group has demonstrated the presence of a small intestinal T cell clone
that is sensitive to gliadin 31-49. However, a peripheral blood clone which responds to
6
this peptide has been demonstrated . Interestingly, the peptide has also been shown to
7
induce CD4 T cell infiltration into the buccal mucosa of coeliac disease patients . We
therefore felt that there is good evidence that A-gliadin 31-49 is a coeliac toxic epitope,
although it may be a minor one.

ROLE OF A-GLIADIN 31-49 PEPTIDE 29
References
1. Kasarda DD, Okita TW, Bernadin JE, Backer PA, Nimmo C, Lew E, et al. Nucleic acid
(cDNA) and amino acid sequences of type gliadins from wheat (Triticum aestivum). Proc Nat
Acad Sci 1984; 81: 4712-6.
2. Ciclitira PJ, Evans DJ, Fagg NLK, Lennox ES, Dowling RH. Clinical testing of gliadin
fractions in coeliac patients. Clin Sci 1984; 66: 357-64.
3. Sturgess R, Day P, Ellis HJ, Lundin KE, Gjertsen HA, Konakou M, et al. Wheat peptide
challenge in coeliac disease. Lancet 1994; 343: 758-61.
4. Kontakou M, Przemioslo RT, Sturgess RP, Limb GA, Ellis HJ, Day P, et al. Cytokine mRNA
expression in the mucosa of treated coeliac patients after wheat peptide challenge. Gut 1995;
37: 52-7.
5. Shidraw RG, Parnell ND, Ciclitira PJ, Travers P, Evan G, Rosen-Bronson S. Binding of
gluten-derived peptides to the HLA-DQ2 (alpha 1*0501, beta 1*0201) molecule, assessed in
a cellular assay. Clin Exp Immunol 1998; 111: 158-65.
6. Gjertsen HA, Lundin KEA, Sollid L, Eriksen JA, Thorsby E.. T cell recognition of gliadin
peptides presented by the coeliac disease associated HLA-DQ (a1*0501, B1*0201)
heterodimer. Human Immunology 1994; 39:243-52.
7. Lahteenoja H, Mäki M, Viander M, Raiha I, Vilja P, Rantala I, et al. Local challenge on oral
mucosa with an alpha-gliadin related synthetic peptide in patients with coeliac disease. Am J
Gastroenterol 2000; 95:2880-7.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 31-41
The association of the HLA-DQ molecules with
coeliac disease in the Saharawi: an evolutionary
perspective
1 2
Francesco Cucca and Carlo Catassi
1
Department of Biomedical and Biotechnologies Sciences, University of Cagliari,
2
Cagliari, Italy. Department of Pediatrics , University of Ancona, Ancona, Italy.
The familial clustering of CD
Coeliac disease (CD) is an immune-mediated enteropathy caused by the ingestion
of wheat and other gluten-containing cereals (rye, barley) in genetically predisposed
1
individuals . CD is strongly clustered in families as illustrated by the sibling lifetime
2-3
risk/population prevalence ratio, lequal s to 30-60 in different studies . These
epidemiological observations indicate the strong influence of genetic factors in the
disease predisposition. This is also strongly suggested by the recurrence risk of 70-
4
100% observed in monozygotic twins (MZ) . Twin studies are particularly important
for human geneticists because the risk in MZ twins provides a direct estimate of
penetrance (the effect of the genotype on the phenotype) for the whole complement of
susceptibility alleles at multiple loci. While it is necessary to extend these preliminary
studies in larger cohorts of twins, the fact that the recurrence risk might not be 100%,
and thus that the penentrance is incomplete, might indicate that other modifying
epigenetic factors, such as the stochastic rearrangements of the T-cell receptor, and/or
other environmental factors (in addition to the gluten in the diet) could also be involved
in CD.
The contribution of genes located in the HLA region to CD predisposition
The major histocompatibility complex (MHC) human leukocyte antigen (HLA)
region on chromosome 6p21 contains the major and, so far, the only consistently
confirmed genetic component for CD predisposition. The HLA/MHC contains a
densely packed array of at least 150 genes in 3,500 kb of DNA. Proteins encoded within
the HLA region determine the way in which antigens are processed, translocated and
presented to T-lymphocytes. During development of the T cell repertoire, HLA class I
and class II molecules control positive and negative selection of the T-lymphocytes in
the thymus. Not surprisingly, therefore, genes in the HLA region influence
susceptibility to a large number of disorders, particularly autoimmune diseases.
31

32 COELIAC DISEASE IN THE SAHARAWI
A long series of association studies from different Caucasian populations provided a
convincing case that within the HLA region, the combined presence of the DQA1*0501
and DQB1*0201 alleles, encoded by genes located in the class II sub-region, is the
5
primary predisposing factor for CD . These alleles are found either in phase within the
DR3-DQA1*0501-DQB1*0201 haplotype or in trans in individuals carrying the DR7-
HLA-DQA1*0201-DQB1*0201/DR5-HLA-DQA1*0501-DQB1*0301 genotype.
Additionally, although to a much lesser extent, another molecule, HLA-DQA1*0301-
5
DQB1*0302, found on DR4 haplotypes, is associated with CD . More recently, linkage
analysis in families with multiple cases, in particular tests of allele sharing by affected
sibpairs (ASPs), proved that the HLA region encodes the main disease locus
predisposing to CD. From these linkage data it was possible to compute that this region
contributes about 40 % of the observed familial clustering of the disease assuming a
2-3.
multiplicative model of inheritance
The role of these DQ molecules is further substantiated by the observation that
antigen-presenting-cells positive for the DQA1*0501-DQB1*0201 (DQ2) and
DQA1*0301-DQB1*0302 (DQ8) molecules are able to recognize and present gluten
derived peptides to T cells obtained from biopsy specimens from the small intestine of
6-7
CD patients . The gluten-reactive T cell clones isolated from biopsies were CD4+,
CD8- and inhibition studies with anti-HLA antibodies demonstrated predominant
6
antigen presentation by the dimers encoded by the HLA-DQ loci .
Non-HLA genes and CD
The linkage data from affected sib pair (ASP) families from different populations
indicates that the locus specific ls for the whole HLA region is equal to ~3. This value,
assuming a multiplicative model of epistasis (which fits with the rapid fall off in
recurrence risk from first to second to third degree relatives) is equal to 40% of the
familial clustering of the disease (log3/log15). These results imply that 60% of the
familial clustering of the disease must be accounted for by non-HLA genes. To detect
these non-HLA loci , some whole genome scans have been performed but, with the
exception of the major disease locus in the HLA region on 6p21, they have given weak
8-10
and conflicting results . These results illustrate the difficulties of detecting (or
replicating the detection of) genes of low to moderate effects using ASP-based
methods. Low penetrance, compounded by small individual genetic effects of most of
the non-HLA disease loci create major limitations on the power of a study (and all the
linkage scans so far performed were severely under-powered). Additionally, locus and
allelic heterogeneity further complicate the search. Nevertheless, the generation of
primary hypotheses deriving from linkage analysis remains an invaluable step when
embarking on an effort to define the genetic risk factors involved in an inherited disease
especially since whole genome scans of association are still technically and
economically impractical and the presence of a disease model with strong epistasis
between the disease loci could severely impair association studies but not linkage
11 -5
analysis . Prior evidence of linkage dramatically lowers the threshold (P value to 10 )
required in an association/LD study of the linked region to be confident that the result is
12
not a false positive . Some of the aforementioned factors, in particular those related to
the statistical power of the study, could be alleviated by further increasing the number of

COELIAC DISEASE IN THE SAHARAWI
33
ASP families, for instance in a consortium type study addressed to analyse all the CD
ASP collectable worldwide on the order of 2,000 or more (ASPs). However, this
strategy based on the analysis of very large sample sets of ASP families that have to be
necessarily collected from different populations cannot provide a significant benefit
regarding the confounding effects of genetic heterogeneity. An alternative approach
could be to concentrate on an isolated population showing an high disease prevalence,
to reconstruct all the relationships between the various affected cases and scan with
parametric and non-parametric linkage tests the whole genome to detect non-HLA gene
effects. It is likely that such a study design, already applied in the case of type 1
13
diabetes , would provide sufficient statistical power to detect significant results across
the genome.
The high prevalence of CD in the Saharawi population
14
CD is typically found in individuals of Caucasian origin . In Europe and North
15-16
America the disease prevalence is around 0.5-1 % of the general population .
Recently a tenfold higher CD frequency (5.6 %) was reported in the Saharawi, a people
of Arab and Berber origin with high degree of consanguinity, living in the Sahara desert
17
. We analysed the genetic association of the main CD locus in a group of Saharawi
children affected with CD and their first-degree relatives. The aims of this study (which
is described in details below) were:
1) to evaluate whether the degree of association at the main disease locus was similar to
that observed in other populations;
2) to investigate how the frequency of the predisposing HLA molecules in the general
Saharawi population relates to the high prevalence observed in this ethnic group.
Patients and methods
This work was performed on Saharawi refugees living in the camps near Tindouf,
Algeria. Blood samples were collected with the informed consent from a group of
previously diagnosed subjects with CD and their first-degree relatives. The diagnosis of
CD was based on the finding of serum IgA class antiendomysium antibodies (EMA)
positivity on at least two consecutive blood samples and, in a subgroup of patients, the
17
typical enteropathy at the small intestinal biopsy, as previously described . EMA were
detected by indirect immunofluorescence on serum diluted both 1:5 and 1:50, using
monkey oesophagus as the antigenic substrate (Antiendomisio Eurospital, Trieste,
Italy). Patients were 79 subjects (33 males and 44 females) deriving from 69
independent families, with a median age of 8 years (range: 2-37) and including 69
probands, 8 siblings and 2 mothers. Overall 136 healthy subjects were also investigated
with a median age of 30 years, range 2-70 years. These included 29 individuals from 9,
non-CD, unrelated families and 107 first-degree relatives of the CD patients (53
mothers, 17 fathers and 37 siblings). Additionally, in 12 parents of CD patients for
which DNA samples were not available, the DQ genotypes were unequivocally
deduced based on the available data from the other parent and the offsprings.
DNA was extracted using the chelex method starting from dried blood spots
18
(Guthrie cards) . Amplification of the polymorphic second exon of the HLA-DRB1,

34
COELIAC DISEASE IN THE SAHARAWI
DQB1 genes and dot blot analysis of amplified DNA with sequence specific
19-21
oligonucleotide probes (SSO) were carried out as previously reported . The DQA1
alleles were inferred using the known patterns of linkage disequilibrium from the
DRB1-DQB1 haplotypes.
Parental haplotypes were reconstructed using a program written by Frank
Dudbridge available at ftp://ftp-gene.cimr.cam.ac.uk/pub/software/.
A possible drawback of case-control analyses is their sensitivity to problems related
to genetic stratification. In order to circumvent this problem, 82 affected family-based
control, (AFBAC) DQA1-DQB1 haplotypes have been selected from all the families in
22
which the parents were available, as described for single alleles by Thomson . The
AFBAC frequencies are based on the chromosomes that are never transmitted from the
parents to affected children and provide a source of controls which is not sensitive to
population stratification unless very recent. The control frequencies were also enriched
by 18 chromosomes deriving from 9 unaffected and unrelated Saharawi individuals.
The pseudo-controls (AFBAC) and control frequencies were nearly identical and
therefore were merged (total controls equal to 100 chromosomes). Using these control
haplotype frequencies and assuming Hardy-Weinberg equilibrium, we also established
23
the control genotype frequencies as previously described . The frequencies of the HLA
class II haplotypes and genotypes obtained in patients and in controls, were then
compared using a 2 x 2 chi-squared Pearson test. The odds ratios ORs) were calculated
using the Wolf formula [a x d)/(b x c)]; when one element of this equation was 0 we used
the Haldane formula: RR=[2a + 1) 2d +1)]/[2b +1) 2c +1)], where a is the number of
patients possessing the HLA antigen; d is the number of controls lacking the particular
HLA antigen; b the number of patients lacking the particular antigen; c the number of
controls possessing the particular HLA antigen. In the association analyses performed
in this study, we considered only the probands in the 8 families with more than one
affected sibling. A correction for number of tests performed was applied by multiplying
the P values for the tested markers present in more than 1 patient or 1 control
(respectively 10 haplotypes and 8 genotypes).
We also evaluated the evolutionary relationship between the Saharawis and the
24
other human groups. To this aim we performed a multi-dimensional scaling analysis
carried out using the correlation matrix of the haplotype frequencies reported in the
various populations and determined with a gene counting procedure.
Who are the Saharawis?
In order to obtain a visual output of the DRB1-DQA1-DQB1 haplotype distribution
worldwide and obtain information about the evolutionary relationships between the
various human groups we performed a multi-dimensional scaling analysis. The results
of this analysis are shown in Fig.1. The first dimension splits Eastern and Western world
populations. The Asian and the African appear to be the most differentiated human
groups, with the Kogi and Cayapa native American populations as the most
differentiated within the Asian group. The Caucasians are closely grouped and are
located between the African and the Asian range of variability. The second dimension
explains the diversity of the Caucasian group, which includes the North-African and the
Saharawi samples. More specifically, the Saharawi population appears to be located in

COELIAC DISEASE IN THE SAHARAWI
35
the border between the European and the African range of variability. Thus, the genetic
structure of this population is consistent with its geographical location.
The association of the DQ molecules with CD in the Saharawi population
We first established the sibling lifetime risk/population prevalence ratio, ls in our
Saharawi sample set, that was equal to 3 (16.7%/5.6%). Because of the high disease
prevalence, the degree of familial clustering was lower in the Saharawi than in
Caucasian populations ls = ~30).
Next, we established the association of the haplotypes and genotypes at the main
disease superlocus HLA-DQB1-DQA1 in this population. The distribution of the
various haplotypes among the Saharawi CD patients and controls is shown in Tab. 1.
The DQA1*0501-DQB1*0201 haplotype was positively associated with CD (OR:
-6
4.5, Pc = 1.4x10 ) while the DQA1*0102-DQB1*0604, haplotype showed some
-3
degree of negative association with the disease (OR: 0.4, Pc = 4.7x10 ). To our
knowledge the latter finding has been never reported in other populations and therefore
needs confirmation in independent sample sets.
Tab. 2 shows the distribution of the genotypes among patients and controls. The
(DR3)DQA1*0501-DQB1*0201/(DR3)DQA1*0501-DQB1*0201 encoding two
putative a*0501 - b*0201 heterodimers was positively associated with CD (OR: 6.4,
-2
P= 2.4x10 ). The (DR3)DQA1*0501-DQB1*0201/(DR4)DQA1*0301-DQB1*0302
and the (DR3)DQA1*0501-DQB1*0201/(DR7)DQA1*0201-DQB1*0201 genotypes
showed a trend toward a positive association (OR: 8.6 and 2.4 respectively) without
reaching the level of significance.
Overall, 91.3% of the patients, but only 38.9% of the controls were found to be
positive for the combined presence of the DQA1*0501 and DQB1*0201 alleles
encoding in cis or in trans at least one putative a*0501 - b*0201 heterodimer. The
patients positive for the DQA1*0501-DQB1*0201 or DQA1*03-0302 heterodimer
HAPLOTYPES Patients Controls
DQA1 DQB1 N N OR 95% C.I. P

36 COELIAC DISEASE IN THE SAHARAWI
were instead 95.6% in comparison with 41.6% of the controls. Finally, all but one
patient (98.6%) and 42.5% of the controls were found to be positive for the
DQB1*0201 allele.
The CD associations observed in the Saharawi population were then contrasted with
5
those observed in other populations . While the (DR3)DQA1*0501-
DQB1*0201/(DR3) DQA1*0501-DQB1*0201 genotype seems to be strongly and
GENOTYPES
DQA1 DQB1 DQA1 DQB1
Patients Controls
N N OR 95% C.I. P

COELIAC DISEASE IN THE SAHARAWI
37
confirmation of the reliability of the CD diagnosis (based on both the serum EMA
testing and the small intestinal biopsy) in these North African children. Most
importantly, these results are also clear evidence that CD has a similar HLA association
in distantly related populations. The aetiological mechanism of the class II association
involves binding and T cell recognition of gluten derived peptides deaminated in the
25
small intestine by the enzyme tissue transglutaminase , and these properties of class II
molecules are very likely conserved among different ethnic groups.
In a manner consistent with the high disease prevalence, the main CD susceptibility
haplotype DQA1*0501-DQB1*0201, detected nearly always within a DR3 haplotype,
exhibited one of the highest frequencies in the world in the general background
Saharawi population. This could partially explain the very high CD prevalence
observed in the Saharawi population. However, it should be noted that a similar
frequency of DQA1*0501-DQB1*0201 has been found also in other populations, such
26
as the Sardinians where the disease is around 5 times less frequent (1%) . Our results
therefore suggest that other non-DQ genes or environmental factors are also operative
in determining the high CD prevalence observed in the Saharawi.
From an evolutionary perspective, such disease frequency might be related to the
high prevalence of genes such as the Dq2, that have been selectively neutral or even
beneficial for several hundred generations, that became disadvantageous following
quick and dramatic changes of the environmental context in which the Saharawis lived.
In the traditional Saharawi diet, the intake of gluten was very low and its introduction
usually occurred after the second or third year of life. During the last century, in a span
of few generations, wheat-based products, especially bread, have become the staple
food after the Spanish colonization. Nowadays, gluten is introduced early in the
Saharawi infant's diet, occasionally as early as the first month of life. Since in CD the
27
degree of intestinal mucosal damage is related to the amount of ingested gluten , the
enteropathy of the “ancestral” CD-prone individuals was presumably milder. As such,
in terms of genetic fitness, it did not represent a major selective disadvantage or might
have even exerted some protective effect against intestinal infections because a
moderately atrophic jejunal mucosa partially lacks the membrane receptors required
28
for microorganisms adhesion . Moreover, a mild CD and the accompanying small
degree of inflammation of the jejunal mucosa, can also provide a source of reactive
lymphomonocytes, thus further increasing the protective effect against intestinal
infections. In contrast, the severe celiac enteropathy currently found in Saharawi
children is associated with an early and high gluten intake and it is no longer neutral or
“protective” but instead harmful, as it causes early intestinal malabsorption and
29
malnutrition, potentially leading to death .
From these data a more general theory can be proposed. High disease frequency of a
complex trait in a given population might be related to the high prevalence of genes that
have been selectively neutral or even beneficial for several hundred generations, but
became disadvantageous following quick changes of the environment. According to
this interpretation, complex traits can be regarded as “residual traits“ caused by
dramatic transformations of the environmental context in which a given population
used to live across evolutionary time periods. Some environmental transformations can
occur too quickly to be accompanied by changes in the genetic structure of the human
populations. Indeed, the evolutionary time periods necessary to modify the genetic

38 COELIAC DISEASE IN THE SAHARAWI
Dimension 2
1.4
Sub Saharian Africa
SEN
FIL
Asia
1
+ Europe and N. Africa
+
+
Gre
+
+
0.6
+
Sard
Bulg
+
+
+
HONG KONG
0.2
+
+ + Cre
ETH
+
-0.2
+ ++ ++ +
+ Rom
CHI
JAP
+
+
Ita
+
Tun
-0.6
Tur
POL
+
Mor
+
+
Fra
Spa +
-1
S AFR
+
GAB
+ Cze
Alg+
Nw Rus
Saha
+
+
Nor + Gb
-1.4
+
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Dimension 1
Fig. 1. Multidimentional scaling of 26 populations tested for HLA DRB1-DQA1-DQB1
haplotypes. The first dimension splits Eastern and Western world populations. The
Caucasians are closely grouped and located in the African range of variability. The second
dimension explains the diversity of the Sub-Saharan African and Caucasian groups, which
include the North-African and Saharawi samples.
On the right box, European and North African patterns were enlarged.
30 31 32-33 34
Population samples: ALG, Algerian ; BULG, Bulgarian ; CHI, Chinese ; CRE, Cretan ;
35 36 37
CZE, Czech ; ETH, Ethiopian ; FIL, Filipino ; FRA, French; ITA, Italian; RUM, Rumanian;
38 39 40
S-AFR, South-African; SEN, Senegalese; SPA, Spanish ; GAB, Gabonese ; GB, British ;
41 32 42 43
GRE, Greek ; HONG KONG, Hong-Kong ; JAP, Japanese ; MOR, Moroccan ; NOR,
44 45 46 47
Norwegian ; NW-RUS, North-Western Russian ; POL, Polynesian ; TUN, Tunisian ; TUR,
48 49
Turkish ;SARD, Sardinian and in circle font the Saharawi (current work).
asset of a given population are on a much larger scale in comparison to those required to
change environmental factors, i.e. the socalled “gene-environment evolutionary tempo
mismatch”.
Finally there is another important lesson coming from CD to understand the bases of
other multifactorial traits. CD tells us that some of the environmental factors involved
in triggering these complex traits could be nearly ubiquitous (gluten). It might be
extremely difficult to identify them in diseases in which the target organ is not able to
regenerate when the trigger factor is removed.
Acknowledgements
We wish to thank Stefano De Virgiliis, M.Doloretta Macis, Mauro Congia, Michael
Whalen , Laura Morelli, Patrizia Zavattari and Antonio Cao for advice and support.
References
1. Mäki M, Collin P. Coeliac disease. Lancet 1997; 349: 755-9.
2. Risch N. Assessing the role of HLA-linked and unlinked determinants of disease. Am

COELIAC DISEASE IN THE SAHARAWI
39
J Hum Genet 1987; 401:1-14.
3. Petronzelli F, Bonamico M, Ferrante P, Grillo R, Mora B, Mariani P, et al. Genetic
contribution of the HLA region to the familial clustering of coeliac disease. Ann
Hum Genet 1997; 61: 307-17.
4. Polanco I, Biemond I, van Leeuven A, Shreuder I, Meera Khan P, Guerrero J, et al.
In: mcConnel RB, ed. The genetics of coeliac disease. Lancaster 1981: MTP, 211-
31.
5. Sollid LM, Thorsby E. HLA susceptibility genes in celiac disease: genetic mapping
and role in pathogenesis. Gastroenterology 1993; 105: 910-22.
6. Lundin KE, Scott H, Hansen T, Paulsen G, Halstensen TS, Fausa O, et al. Gliadinspecific,
HLA-DQ alpha 1*0501,beta 1*0201 restricted T cells isolated from the
small intestinal mucosa of celiac disease patients. J Exp Med 1993; 178: 187-96.
7. Lundin KE, Scott H, Fausa O, Thorsby E, Sollid LM. T cells from the small intestinal
mucosa of a DR4, DQ7/DR4, DQ8 celiac disease patient preferentially recognize
gliadin when presented by DQ8. Hum Immunol 1994; 41: 285-91.
8. Greco L, Corazza G, Babron MC, Clot F, Fulchignoni-Lataud MC, Percopo S, et al.
Genome search in celiac disease. Am J Hum Genet 1998; 62: 669-75.
9. Zhong F, McCombs CC, Olson JM, Elston RC, Stevens FM, McCarthy CF, et al.An
autosomal screen for genes that predispose to celiac disease in the western counties
of Ireland. Nat Genet 1996; 14: 329-33.
10. Houlston RS, Tomlinson IP, Ford D, Seal S, Marossy AM, Ferguson A, et al. Linkage
analysis of candidate regions for coeliac disease genes. Hum Mol Genet 1997; 6:
1335-9.
11. Culverhouse R, Suarez BK, Lin J, Reich T. A perspective on epistasis: limits of
models displaying no main effect. Am J Hum Genet 2002; 70: 461-71.
12. Dahlman I, Eaves IA, Kosoy R, Morrison VA, Heward J, Gough SC, et al.
Parameters for reliable results in genetic association studies in common disease.
Nat Genet 2002; 30: 149-50.
13. Verge CF, Vardi P, Babu S, Bao F, Erlich HA, Bugawan T, et al. Evidence for
oligogenic inheritance of type 1 diabetes in a large Bedouin Arab family. J Clin
Invest 1998; 15: 1569-75
14. Trier JS. Celiac sprue. N Engl J Med 1991; 325: 1709-19.
15. Catassi C, Fabiani E, Rätsch IM, Coppa GV, Giorgi PL, Pierdomenico R, et al. The
coeliac iceberg in Italy. A multicentre antigliadin antibodies screening for coeliac
disease in school-age subjects. Acta Paediatr 1996; Suppl 412: 29-35.
16. Not T, Horvath K, Hill ID, Partanen J, Hammed A, Magazzù G, et al. Celiac disease
risk in the USA: high prevalence of antiendomysium antibodies in healthy blood
donors. Scand J Gastroenterol 1998; 33: 494-8.
17. Catassi C, Rätsch IM, Gandolfi L, Pratesi R, Fabiani E, El Asmar R, et al. Why is
coeliac disease endemic in the people of the Sahara? Lancet 1999; 354: 647-8.
18. Singer S, Tanguay R. Use of chelex to improve the PCR signal for a small number of
cells. Amplifications 1989; 3: 11.
19. Horn GT, Bugawan TL, Long CM, Erlich HA. Allelic sequence variation of the
HLA-DQ loci, relationship to serology and to insulin-dependent diabetes
susceptibility. Proc Natl Acad Sci 1988; 85: 6012-6.
20. Wordsworth BP, Allsopp CEM, Youg RP, Bell JI. HLA-DR typing using DNA

40 COELIAC DISEASE IN THE SAHARAWI
amplification by the polymerase chain reaction and sequential hybridization to
sequence-specific oligonucleotide probes. Immunogenetics 1990; 32: 413-8.
21. Bugawan TL, Erlich HA. Rapid typing of HLA-DQB1 DNA polymorphism using
nonradiactive oligonucleotide probes and amplified DNA. Immunogenetics 1991;
33: 163-70.
22. Thomson G. Mapping disease genes: family-based association studies. Am J Hum
Genet 1995; 57: 487-98.
23. Noble JA, Valdes AM, Cook M, Klitz W, Thomson G, Erlich HA. The role of HLA
class II genes in insulin-dependent diabetes mellitus: molecular analysis of 180
Caucasian, multiplex families. Am J Hum Genet 1996; 59: 1134-48.
24. Kruscal, J.B. Multidimensional Scaling by Optimizing Goodness of fit to a
Nonmetric Hypothesis. Psychometrika 1964; 29: 1-27.
25. Schuppan D. Current concepts of celiac disease pathogenesis. Gastroenterology
2000; 119: 234-42.
26. Meloni G, Dore A, Fanciulli G, Tanda F, Bottazzo GF. Subclinical coeliac disease in
schoolchildren from northern Sardinia. Lancet 1999; 353: 37.
27. Marsh MN. Gluten, major histocompatibility complex, and the small intestine: a
molecular and immunobiologic approach to the spectrum of gluten-sensitivity
"celiac sprue". Gastroenterology 1992;102: 330-54.
28. Kerneis S, Chauviere G, Darfeuille-Michaud A, et al. Expression of receptors for
enterotoxigenic Escherichia coli during enterocytic differentiation of human
polarized intestinal epithelial cells in culture. Infect Immun 1992;60: 2572-80.
29. Rätsch IM, Catassi C. Coeliac disease is a primary, potentially treatable, health
problem of Saharawi refugee children. Bulletin of the World Health Organization
2001; 79: 541-5.
30. Arnaiz-Villena A, Benmamar D, Alvarez M, Diaz-Campos N, Varela P, Gomez-
Casado, et al. HLA allele and haplotype frequencies in Algerians. Relatedness in
Spaniards and Basques. Hum Immunol 1995; 43: 259-68.
31. Ivanova R, Naoumova E, Lepage V, Djoulah S, Yordanov Y, Loste MN et al. HLA-
DRB1, DQA1, DQB1 DNA polymorphism in the Bulgarian population. Tissue
Antigens 1996; 47: 122-6.
32. Chang YW, Hawkins BR. HLA class I and class II frequencies of a Hong Kong
Chinese population based on bone marrow donor registry data. Hum Immunol
2002; 56: 125-35.
33. Fernandez-Viña M, Gao X, Moraes ME, Moraes JR, Salatiel I, Miller et al. Alleles
at four HLA class II loci determined by oligonucleotide hybridization and their
associations in five ethnic groups. Immunogenetics 1991; 34: 299-312.
34. Arnaiz-Villena, A., Iliakis, P., Gonzalez-Hevilla, M., Longas, J., Gomez-Casado, E.,
Sfyridaki, K., et al. The origin of Cretan populations as determined by
characterization of HLA alleles. Tissue Antigens 1999; 53: 213-26.
35. Cernà,M, Fernandez-Vina M, Ivaskova E. and Stastny P. Comparison of HLA class
II alleles in Gypsy and Czech populations by DNA typing with oligonucleotide
probes. Tissue Antigens 1992; 39: 111-6.
36. Fort M, de Stefano, G.F., Cambon-Thomsen, A., Giraldo-Alvarez, P., Dugoujon,
J.M., Ohayon, E., et al. HLA class II allele and haplotype frequencies in Ethiopian
Amhara and Oromo populations. Tissue Antigens 1998; 51: 327-36.

COELIAC DISEASE IN THE SAHARAWI
41
37. Mack SJ, Bugawan TL, Moonsamy PV, Erlich JA, Trachtenberg EA, Paik YK, et al.
Evolution of Pacific/Asian populations inferred from HLA class II allele frequency
distributions. Tissue Antigens 2000; 55: 383-400.
38. Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojobori T. Allele and haplotype
frequencies for HLA and complement loci in various ethnic groups. In, Tsuji, K.,
Aizawa, M. and Sasazuki, T. eds.:, HLA 1991. Oxford University Press, Oxford, UK,
Vol. 1, pp. 1065-1220.
39. Migot-Nabias F, Fajardy I, Danze PM, Everaere S, Mayombo Y, Minh TN, et al.
HLA class II polymorphism in a Gabonese Banzabi population. Tissue Antigens
1999; 53: 580-5.
40. Doherty DG, Vaughan RW, Donaldson PT, Mowat AP. HLA DQA, DQB, and DRB
genotyping by oligonucleotide analysis, distribution of alleles and haplotypes in
British Caucasoids. Hum Immunol 1992; 34: 53-63.
41 Papassavas EC, Spyropoulou-Vlachou M, Papassavas AC, Schipper RF, Doxiadis
IN, Stavropoulos-Giokas C. MHC Class I and II phenotype, gene, and haplotype
frequencies in Greeks using molecular typing data. Hum Immunol; 61: 615-23.
42. Hashimoto M, Kinoshita T, Yamasaki M, Tanaka H, Imanishi T, Ihara, et al. Gene
frequencies and haplotypic associations within the HLA region in 916 unrelated
Japanese individuals. Tissue Antigens 2000;44: 166-73.
43. Izaabel H, Garchon HJ, Caillat-Zucman S, Beaurain G, Akhayat O, Bach JF et al.
HLA class II DNA polymorphism in a Moroccan population from the Souss, Agadir
area. Tissue Antigens 1998; 51: 106-10.
44. Rønningen KS, Spurkland A, Markussen G, Iwe T, Vartdal F, Thorsby E.
Distribution of HLA class II alleles among Norwegian Caucasians. Hum Immunol
1990; 29: 275-281.
45. Kapustin S, Lyshchov A, Alexandrova J, Imyanitov E, Blinov M. HLA class II
molecular polymorphisms in healthy Slavic individuals from North-Western Russia.
Tissue Antigens 1999; 54: 517-520.
46. Gao X, Zimmet P, Serjeantson SW. HLA-DR,DQ sequence polymorphisms in
Polynesians, Micronesians, and Javanese. Hum Immunol 1992; 34: 153-161.
47. Hmida S, Gauthier A, Dridi A, Quillivic F, Genetet B, Boukef K. HLA class II gene
polymorphism in Tunisians. Tissue Antigens 1995; 45: 63-8.
48. Saruhan-Direskeneli G, Esin S, Baykan-Kurt B, Ornek I, Vaughan R. Eraksoy M.
HLA-DR and -DQ associations with multiple sclerosis in Turkey. Hum. Immunol.
1997; 55: 59-65.
49. Lampis R, Morelli L, Congia M, Macis MD, Mulargia A, Loddo M, et al. The interregional
distribution of HLA class II haplotypes indicates the suitability of the
Sardinian population for case-control association studies in complex diseases.
Hum Mol Genet 2000; 9: 2959-65.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 43-60
Primary prevention of coeliac disease
by favourable infant feeding practices
1,2 1,3
Anneli Ivarsson , Lars Åke Persson and Olle Hernell 2
1 2
Department of Public Health and Clinical Medicine, Epidemiology and Department
3
of Clinical Sciences, Pediatrics, both at Umeå University, Umeå, Sweden, and ICDDR,B:
Centre for Health and Population Research, Dhaka, Bangladesh.
Coeliac disease is now recognised as a common health problem throughout a large
1-12
part of the world . An effective treatment is available through adherence to a strict
gluten-free diet. Screening studies have revealed that most cases are undiagnosed,
which indicates the need for active case recognition, and possibly also a need for
13-14
screening efforts . However, even when coeliac disease has been diagnosed, the
widespread use of gluten-containing foods makes compliance with the treatment
15
difficult . Clearly, then, it is desirable to explore additional strategies such as primary
prevention.
Family clustering
The risk for coeliac disease is higher in first degree relatives compared to the general
population. In family members of coeliac patients, a prevalence of 10% is often
16
mentioned , and in monozygous twins the concordance rate has been estimated at
17-18 19
70% , or even higher .
Family clustering is generally considered as evidence for a strong genetic influence
in coeliac disease susceptibility. However, the fact that family members share not only
genetics but, in many respects, also the same environment, should be taken into
consideration. This is especially true for monozygous twins, who have a similar
environment not only during childhood, as is the case for most siblings, but also have a
similar environment in utero.
A complex aetiology
The aetiology of coeliac disease is not fully understood, largely as a consequence of
its complexity, although our understanding of different aspects of the disease process is
rapidly increasing. Based on present knowledge, an immunological pathogenesis for
the disease seems most likely. In some genetically susceptible individuals, exposure to
gluten, or related proteins in rye and barley, triggers an errant immune response, which
43

44
INFANT FEEDING PRACTICES AND COELIAC DISEASE
20-23
by means of a cascade of events results in a chronic enteropathy .
The human leukocyte antigen (HLA)-DQ2 molecule is expressed by more than
90% of coeliac disease patients, compared to 20-30% of healthy controls, and the
22
majority of the remaining patients express the DQ8 molecule . It has been estimated
that the HLA-complex confers 40% of the sibling family risk, and that non-HLA genes
24-25
are the most important . However, efforts to identify these contributing non-HLA
26-27
genes have thus far not resulted in conclusive evidence . In assessing the impact of
genetics, the ratio lis crucial. This ratio is constructed by the prevalence in relatives of
28
an affected individual over the prevalence within the general population . Thus, if
causal environmental exposures are aggregated in the family, the impact of genetics
29
will be overestimated .
Most common diseases have a multifactorial aetiology, and they thus develop
through an interaction between an individual's genetic predisposition and various
30
environmental exposures . This has been shown for insulin dependent diabetes
mellitus (IDDM), which is an autoimmune disease with many similarities to coeliac
disease. The immune system has a key function in the pathogenesis of both diseases,
31
and there is an increased risk for coeliac disease in persons with IDDM .
It has recently been suggested that each genetic risk factor, taken separately, may
frequently be present in the general population, and it is the combination of some of
26
these and their interaction with environmental factors that induces coeliac disease .
Thus, in addition to the mere presence of gluten in the diet, environmental exposures
may also be expected to be part of the causal pattern responsible for coeliac disease.
Failure of oral tolerance
Environmental exposures, including infant feeding, influence the immunological
32-34
process resulting in oral tolerance - or intolerance - to a food constituent .
Considering recent knowledge of the aetiology of coeliac disease, this may be viewed at
least hypothetically as a failure in the development of, or the later loss of, oral
35
tolerance .
In most individuals, oral tolerance towards gluten develops and prevails throughout
life. If oral tolerance fails to develop, however, or is later broken down, then gluten may
act like a “dangerous” foreign antigen, resulting in the development of coeliac disease.
The Swedish epidemic
Sweden has experienced an unusual epidemic of symptomatic coeliac disease in
36-37
children . The incidence reached levels higher than ever reported, and the decline
that followed was amazingly abrupt (Fig. 1). Only children below two years of age were
affected, and most of the cases had symptoms as severe as those that had been observed
38-39
earlier . These findings are based on our population-based prospective incidence
register established in 1991, covering 40% of the Swedish child population, and
36
retrospective data collected back to 1973 from 15% of the child population .
Recently it was shown that this was indeed an epidemic of coeliac disease
enteropathy, and not only a shift in the proportion of cases with symptomatic

INFANT FEEDING PRACTICES AND COELIAC DISEASE
45
Cases per 100 000 person-years
250
200
150
100
50
0-1.9 years
2-4.9 years
5-14.9 years
0
1975 1981 1987 1993 1999
Year of diagnosis
Fig. 1. Annual incidence rate of coeliac disease in different age groups of children from
37
1973 to 1999. From with permission.
enteropathy that were thus more readily diagnosed. A screening of 2½-year-old
Swedish children including birth cohorts both from the epidemic and post-epidemic
years showed this. In the cohort of 1992-1993 (epidemic years) with 3004 children,
there were 22 cases diagnosed due to symptomatic coeliac disease, and out of 690
40
screened children, as many as 9 had previously unrecognised coeliac disease . In the
cohort of 1996-1997 (post-epidemic years), only about half as many coeliac disease
cases were identified, and this was true for symptomatic cases as well as cases detected
by screening, with the studies being comparable in all other respects (Carlsson A,
personal communication). Obviously, the epidemic cannot be explained by genetic
changes in the population, as it occurred over such a short time period.
This epidemic of coeliac disease indicates an abrupt increase and decrease,
respectively, of one or a few causal factors influencing a large proportion of Swedish
infants over the period in question. As only children below two years of age were
affected, changes in infant feeding practices were suspected to have contributed. If
these causal exposures could be solidly identified, this might open doors to the
development of a primary prevention strategy.
Infant feeding
The infant feeding pattern is one of the few environmental exposures other than
dependence on gluten proteins in the diet that has been approached as possibly
contributing to coeliac disease aetiology. Taking into consideration established
guidelines for causation (table), some aspects of infant feeding will be discussed as
potential risk determinants for development of coeliac disease.

46 INFANT FEEDING PRACTICES AND COELIAC DISEASE
Temporal relationship
Plausibility
Consistency
Strength
Dose-response relationship
Reversibility
Study design
Coherence of evidence
41
Adapted from .
Breast feeding
Table. Guidelines for the evaluation of causality.
Does the possible cause precede the effect?
Is the association consistent with other knowledge?
Have similar results been shown in other studies?
What is the strength of the association between the
possible cause and the effect?
Is increased exposure to the possible cause associated
with increased effect?
Does the removal of a possible cause lead to reduction
of disease risk?
Is the evidence based on a strong study design?
How many lines of evidence support the conclusion?
42-45
Based on increasing knowledge about the immunological impact of breast milk ,
it seems plausible that introduction of a dietary antigen while the child is still breast-fed
might increase the likelihood of developing oral tolerance to that antigen. Whether or
not this relationship holds true for dietary gluten and the risk for development of coeliac
disease has not been conclusively established.
Based on observations of coeliac disease patients, it was suggested as early as the
46
1950s that breast-feeding delays onset of the disease , a view supported by later similar
47-48
studies . Furthermore, an increase in breast-feeding was suggested as a possible
factor contributing to the declining incidence of coeliac disease in the early 1970s in
49-51
England, Scotland and Ireland .
52
In the 1980s, the Italian case-referent studies by Auricchio et al (216 cases/289
53
siblings) and Greco et al (201 cases/1949 referents) demonstrated that coeliac disease
cases were breast-fed for a shorter duration than the referents. This was confirmed in a
54
Swedish case-referent study by Fälth-Magnusson et al (72 cases/288 referents), and
55
recently in a German study by Peters et al (143 cases/137 referents). In contrast, in a
56
family study Ascher et al did not find a difference in breast-feeding duration when
comparing screening-detected cases with their siblings (8 cases/73 siblings). This study
was comparatively small, and the design involved overmatching with respect to dietary
factors, constraints discussed by the authors.
Taken together, the case-referent studies demonstrate that coeliac disease cases in
general have been breast-fed for a shorter period than other children. However, these
studies could not clarify whether breast-feeding had a direct causal effect, or if the
protective effect was indirect as a consequence of the postponed introduction of infant
formula (more specifically cow's milk protein), or if it occurred through reduction of
the amount of dietary gluten ingested at an early age.
We have recently reported results from a case-referent study (491 cases/781
57
referents) , which could eliminate some of the above-mentioned constraints through a
somewhat different design in comparison with the previously mentioned studies. Our

OR (95%CI)
INFANT FEEDING PRACTICES AND COELIAC DISEASE
47
study was population-based, with a high participation rate, and the results should thus
be representative for Sweden at large. Since our findings are consequences of
biological phenomena, they should be valid for infants in general. Only incident cases,
i.e. newly diagnosed cases, were included, which reduced the recall period. A
comprehensive questionnaire concerning children's diet and health in general was
mailed to the families; it did not reveal our special interest in coeliac disease. A semiquantitative
food frequency questionnaire was used to assess both the consumption of
gluten-containing cereals at the time when these cereals were introduced into the diet
and the total amount consumed by seven months of age. The amount of glutencontaining
flour in homemade foods was calculated based on standard Swedish recipes
and the amount in industrially produced foods was obtained from the manufacturers.
Multivariate analyses were used to adjust risk estimates for confounding and to suggest
57
causal relationships .
Our main finding was that the risk for coeliac disease was reduced if the child was
breast-fed during the time period when gluten-containing foods were introduced [Odds
57
Ratio (OR) = 0.59, 95% Confidence Interval (CI) 0.42-0.83] . This protective effect
was even more pronounced if the child was also breast-fed beyond the period of gluten
introduction (OR=0.36, 95% CI 0.26-0.51) (Fig. 2). These risk estimates were adjusted
for the age of the infant when gluten was introduced into the diet and the amount of
gluten that was given.
4
2
1
½
¼
Discontinued Continued Continued beyond
Breast feeding status at introduction of gluten
Fig. 2. Breast-feeding status (BF) at introduction of gluten-containing flour into the diet
and risk (OR, 95% CI) for coeliac disease before two years of age.
A protective effect of breast-feeding was supported by our ecological study of the
Swedish epidemic, i.e. using aggregated data to explore any temporal relationship
between the changes in incidence rate and changes in infant dietary patterns. Both the
rise and later fall in incidence had a temporal relationship to a change in the proportion
36
of infants introduced to gluten while still being breast-fed .
It is important to note that at the time of these studies the majority of Swedish

48 INFANT FEEDING PRACTICES AND COELIAC DISEASE
infants were breast-fed for six months or longer. As a result, most of the infants were
introduced to cow's milk products and other food products while still being breast-fed.
Also, for most infants the termination of breast-feeding did not coincide with
introduction of infant formula, but rather with increased ingestion of other foods.
Moreover, in the case-referent study we could show that a reduced risk for coeliac
disease still remained when the age of the infants at introduction of gluten into the diet
was considered, along with the amount of gluten given at that time (fig. 2). Thus, our
findings strongly supported breast-feeding as directly reducing the risk for coeliac
disease, and not merely influencing the risk indirectly through changes in other
exposures.
It has recently been suggested that the intestinal lesion in coeliac disease might be
caused by an uncontrolled production of pro-inflammatory IFN-gby intraepithelial T
cells that is not sufficiently counteracted by an increased production of downregulating
TGF-b1 (Forsberg G, personal communication). It is thus tempting to
35
speculate that the TGF-b1 in breast-milk compensates for this although this is only
42-45
one of several possible mechanisms for a protective effect of human milk .
Thus, a protective effect of being breast-fed when dietary gluten is introduced is
supported by several epidemiological studies of different design. Moreover, the
protective effect is biologically plausible when taking into account our present
knowledge of breast-milk composition and the impact on immune responses, and the
aetiology of coeliac disease.
Amount of dietary gluten
The dose of dietary antigen ingested may influence whether or not oral tolerance
32-34
develops . It is not yet clear if this applies to the amount of gluten given in relation to
the risk of developing coeliac disease.
Interestingly, a larger consumption of wheat gluten was reported for healthy infants
in Sweden and Italy as compared to Finland, Denmark and Estonia, and the former
58-61
countries also reported a higher occurrence of coeliac disease . The study designs,
involving aggregated data, did not, however, allow adjustments for differences in other
exposures.
54
The Swedish case-referent study by Fälth-Magnusson et al indicated that the
coeliac disease cases, more often than the referents, were introduced to gluten by means
of gluten-containing follow-on formula. In infancy it is clear that bottle-feeding,
compared to feeding by cup and spoon, more readily contributes a larger amount of
food, and thus also a larger amount of gluten. In the Fälth-Magnusson study, however, it
was not possible to estimate the amount of gluten given to the infants.
In our case-referent study it was possible, for the first time, to assess the
consumption of gluten-containing cereals on an individual basis by use of a semi-
57
quantitative food frequency questionnaire . We could show that introduction of glutencontaining
foods in larger amounts, as compared to small or medium amounts, was an
independent risk factor for coeliac disease development (adjusted OR=1.5, 95% CI 1.1-
2.1). Also, at seven months of age the cases consumed larger amounts of glutencontaining
flour than the referents. It is important to note that through the use of
multivariate analyses we adjusted for differences in breast-feeding practices and the

INFANT FEEDING PRACTICES AND COELIAC DISEASE
49
age of the infant when gluten was introduced into the diet. Thus, our findings clearly
indicate that introduction of gluten in larger amounts increases the risk for coeliac
disease. Furthermore, we found that the type of food used as the source of gluten, i.e.
57
solid foods or follow-on formula, was not important as an independent risk factor . The
relevance of our observations is therefore not only applicable to Sweden with its custom
of using gluten-containing follow-on formula from the age of six months.
The daily amount of gluten consumed during infancy as a risk factor for coeliac
disease is further supported by our ecological study of the Swedish epidemic, where we
used aggregated data to explore any temporal relationship between changes in
36
incidence rate and changes in infant dietary patterns . The rise in incidence was
preceded by a twofold increase in the average daily consumption of gluten through the
use of follow-on formula, and later, the fall in incidence coincided with a consumption
36
that was decreased by one-third .
62
Marsh et al concluded that gluten-sensitised individuals respond in a time-related
and dose-dependent fashion to gliadin, which is an observation supported by several
63-66
other studies . These experimental studies do not, however, clarify whether the
amount of gluten is also crucial when infants are introduced to this antigen for the first
time.
Taken together, there is evidence to suggest that consumption of a large amount of
gluten-containing flour (increased antigen dose) during infancy increases the risk for
coeliac disease. It is, however, not clear whether there is a direct dose-response effect or
a threshold effect. Furthermore, it seems likely that the amount of gluten tolerated
varies with the genetic predisposition of the individual, other environmental exposures,
and the age of the individual.
Age at introduction of gluten
There might be an age interval during which humans have decreased ability to
develop oral tolerance to a newly introduced dietary antigen. Hypothetically, the age of
the infant upon introduction of gluten into the diet might thus influence the risk for
coeliac disease.
In a comparison of English coeliac disease patients in the 1950s and 1960s, it was
suggested that earlier introduction of dietary gluten resulted in earlier presentation of
67
the disease . Some clinical studies in which differences in breast-feeding duration
47-48
have been taken into account did not show such a relationship . In fact, a delayed
introduction of gluten into the diet of infants was suggested as contributing to the
49-51
decline in incidence of coeliac disease in England, Scotland and Ireland in the 1970s .
However, at that time comparable dietary changes occurred in Sweden without any
68-69
observed change in incidence . Furthermore, the increased incidence in Swedish
children in the middle of the 1980s was preceded by a delayed introduction of dietary
36,38-39
gluten from four until six months of age . Thus, these ecological observations
resulted in contradictory findings. However, a study design based on aggregated data
cannot by itself provide conclusive evidence.
However, the case-referent design based on individual data allows for adjustments
for differences in other exposures. Such studies in which adjustments have been made
for differences in breast-feeding duration have indicated that the age of the infant at

50 INFANT FEEDING PRACTICES AND COELIAC DISEASE
52-54
introduction of dietary gluten is of no importance with respect to coeliac disease risk .
In our case-referent study we could move one step further by also adjusting for the
differences in amount of gluten-containing foods given at various ages of introduction.
Our results also suggested that the age of the infant at introduction of dietary gluten was
57
of no importance for coeliac disease risk , although with the limitation that only the
first year of life was evaluated.
Thus, present evidence does not support age of the infant at the time of gluten
introduction as an independent risk factor for coeliac disease development. However,
even when considered together these studies do not encompass all possible ages for
introduction of gluten, e.g. after one year of age or even later. Thus, this possibility also
needs to be evaluated.
Other possible contributing factors
Infections
70 71
It has been proposed that adenovirus type 12 , and also other adenoviruses , could
initiate coeliac disease as a consequence of sequence similarities between a protein
produced in conjunction with the viral infection and A-gliadin. This has been
72-73
questioned , but not convincingly excluded. Gastrointestinal infections cause a
74
disruption of the barrier function of the small intestinal mucosa , which theoretically
could result in an increased antigen penetration and unfavourable immune responses.
We found a higher risk for coeliac disease in children born during summer as
75
compared to winter, but only in children below two years of age . These findings
indicate that environmental exposure(s) with a seasonal pattern may have a causal
effect. A temporal relationship suggests that this might be due to a causal effect of
infections during foetal life and/or an interaction between infections and introduction
of gluten into the diet. However, any exposure with a seasonal pattern might be the
explanation, and non-infectious exposures should also be explored.
In our case-referent study we found that children who experienced three or more
infectious episodes before six months of age had an increased risk for coeliac disease
76
before two years of age (adjusted OR=1.4, 95% CI 1.0-1.9) . This was true even when
episodes of gastroenteritis were excluded, and after adjustments were made for
differences in infant feeding patterns and family socio-economic group. Infectious
episodes later in infancy were not included, as these could be secondary to the disease
itself. The risk for coeliac disease increased considerably if in addition to having many
infections the child was also introduced to gluten in large amounts, as compared to
76
small and medium amounts .
Socio-economic background
In our case-referent study children in families belonging to the lower as compared to
middle and upper socio-economic strata of Swedish society had an increased risk for
76
coeliac disease (adjusted OR=1.4, 95% CI 1.0-1.8) , but again only in the age group
below two years. Our reported estimate was adjusted for differences in infant feeding
practices and infectious episodes. This indicates that there are also other as yet
unidentified environmental exposures contributing to the causal pattern of coeliac
disease.

INFANT FEEDING PRACTICES AND COELIAC DISEASE
51
A multifactorial aetiology
A simplified model of the multifactorial aetiology of coeliac disease is outlined
below (Fig. 3). It is based on the results of our own studies put into the context of present
knowledge of coeliac disease aetiology. The emphasis in the model is on the influence
of environmental exposures on coeliac disease risk.
Throughout life, including foetal life, an individual's genetic makeup interacts with
the continuous and varying exposures of the environment. Theoretically, both genes
and the environment, as well as the interaction between them, confer either increased or
reduced disease risk. The exposures of causal importance most likely vary throughout
life, and the range of potential risk determinants is wide. More specifically, with regard
to coeliac disease the immunological response to dietary gluten is shaped.
Genetic susceptibility and presence of dietary gluten are considered necessary
casual factors, i.e. without these factors coeliac disease will not develop. However,
component causal factors also contribute to whether or not coeliac disease develops.
Combined, the necessary causes and one or several component causes produce a
77
sufficient cause, i.e. development of the disease is unavoidable .
As component causal factors in coeliac disease aetiology we suggest: i) duration of
breast-feeding, ii) whether breast-feeding is ongoing or not when gluten is introduced
into the diet, iii) amount of gluten given to infants when introduced into the diet, and iv)
repeated infectious episodes early in life.
Associated factors are not considered to have a causal effect by themselves, but act
through other directly causal factors. When identified, however, they may be used as
markers for an increased disease risk, and thereby focus the search for causal factors.
Such a factor is seasonality in births, with an increased risk for coeliac disease in
Season
Dietary recommendations
Food contents
Socioeconomic conditions
Structural
factors
Associated
factors
Infections
Breast feeding
?
?
?
Component
causal
factors
Amount
DIETARY GLUTEN
NECESSARY
CAUSAL
FACTORS
GENETICS
sex
Immunopathogenesis
Fetal life Infancy Childhood Adulthood
Fig. 3. Causal model on the multifactorial aetiology of coeliac disease.
37
From , with permission.

Age (years)
52 INFANT FEEDING PRACTICES AND COELIAC DISEASE
children born during summer as compared to winter. It might be that seasonality reflects
a varying exposure to infectious episodes throughout the year, but other possible
explanations should be explored. Another associated factor is the increased risk for
coeliac disease in Swedish children from the lower as compared to middle and upper
socio-economic strata, suggesting that there are further component causes yet to be
identified.
Structural factors are any change on a societal level influencing the risk for coeliac
disease. However, these must of course exhibit their effect through component causes
close to the individual, as illustrated in the depicted model (Fig. 3). These could include
dietary recommendations influencing how gluten is introduced into the infant diet, and
changes in the gluten content in industrially produced infant foods.
What can be learnt from the Swedish epidemic?
What caused the epidemic?
Changes in infant feeding practices have been suspected to contribute to the
Swedish epidemic of coeliac disease in children. To clarify whether or not this was the
case, we used an ecological approach comparing estimated yearly changes in infant
feeding practices with the yearly incidence rate of coeliac disease in children below two
36
years of age .
We collected national data for the years 1980 to 1997 on duration of breast-feeding
36
and intake of gluten-containing cereals in Swedish infants . The latter was estimated
by changes in gluten intake by means of industrially produced follow-on formulas as an
estimate of changes in total gluten intake. These follow-on formulas, combined with
59
porridge, provide about half of the total intake of gluten proteins , and of that half,
follow-on formulas account for about 90%, according to the manufacturers.
It should be noted that the incidence rate for a particular year in children below two
years of age might also be influenced by the exposure pattern of the two preceding
years. The reason for this, as illustrated by the Lexis diagram, is that the incidence rate
in children below two years of age in 1997, for instance, is based on cases diagnosed
that year, and these children were born during the period 1995 to 1997, which is thus the
period during which they might have been exposed (Fig. 4).
2
Incidence rate in children 0-2 years of age in 1997
1
1995 1996 1997 1998 1999
Exposure period
Fig. 4. A Lexis diagram illustrating the population and exposure periods on which the
incidence rate by age is based.

INFANT FEEDING PRACTICES AND COELIAC DISEASE
53
36
Our results showed rapidly increasing incidence rates between 1985 and 1987 .
The period of interest with regard to exposure was characterised by i) about half of the
infants being breast-fed at six months of age, ii) doubling of the average daily
consumption of flour through use of follow-on formula from 1981 to 1983 with regard
to the total amount of wheat, rye and barley, while the amount of oats decreased, and iii)
a national recommendation at the end of 1982 to postpone introduction of gluten from
four to six months of age (Fig. 5).This latter change was in accordance with European
300
250
Incidence rate
Breastfeeding index
Wheat/Rye/Barley index
200
150
100
50
0
Changed national recommendations
1980 1985 1990 1995
Fig. 5. Annual incidence rate of coeliac disease per 100 000 person years for children 0-1.9
years of age from 1980 to 1997, in relation to breast-feeding habits, flour consumption and
c h a n g e d n a t i o n a l re c o m m e n d a t i o n s . T h e a r ro w s i n d i c a t e c h a n g e d
recommendation on introduction of gluten to infants. Breast-feeding index based on the
proportion of all children breast-fed at six months of age (1980; index 100 = 37%).
Wheat/Rye/Barley index based on the estimated average daily consumption of these cereals
-1 -1
provided by follow-on formulas (1980; index 100 = 16 gram x child x day ).
36
From with permission.
78
recommendations .
A rapid decline in the incidence rate began in 1995, and was still ongoing through
36
1997 . The period of interest with regard to exposure was characterised by i) a
continuous increase (from 54% to 76%) in the proportion of infants still breast-fed at six
months of age, ii) the average daily consumption of flour through the use of follow-on
formula decreased by one third starting in 1995 with regard to the total amount of
wheat, rye and barley, while the amount of oats increased, and iii) the national
recommendation was changed in the autumn of 1996 to introduction of gluten into the
diet in smaller amounts from four months of age, and preferably while the child is still
breast-fed (Fig. 5). However, starting at the end of the 1980s, and particularly during a
one-year period starting in the autumn of 1995, the media focused much attention on the
increased incidence of coeliac disease in relation to the current dietary pattern. This
most likely promoted a change in dietary patterns that was more extensive than revealed

54 INFANT FEEDING PRACTICES AND COELIAC DISEASE
by our study, and that took place before the new national recommendation was
launched.
A certain delay between changes in exposure and disease occurrence would be
expected. Indeed, in the 1980s the change in gluten intake with regard to both amounts
consumed during infancy and age at introduction preceded the increase in incidence. In
contrast, the decrease in amounts of gluten consumed was first noted in 1995, i.e. the
same year as the decline in incidence started. Further, in 1997 the estimated amount of
gluten consumed had decreased by one third, while the incidence rate had already
returned to the level of the early 1980s. It should also be noted that the proportion of
infants breast-fed at six months of age continuously increased during the 1990s without
any sharp increase before, or coinciding with, the decrease in incidence. Hence, the
changes in exposure caused by each of these factors alone cannot explain the entire
epidemic.
Considering the combined effects of changes in infant feeding practices over time, it
is clear that both the rise and fall in incidence were accompanied by a change in the
proportion of infants introduced to gluten in small amounts while still being breast-fed.
Thus, Swedish infant feeding practices have shifted over time from a favourable to an
unfavourable and then back again to a favourable pattern with respect to coeliac disease
risk.
Public health impact
The public health impact of a change in causal exposures can be estimated by the
population attributable fraction (AF), which takes into account the prevalence of
different exposures among the cases (Pc), and the adjusted odds ratios (OR) of these
exposures [AF = Pc (OR-1)/OR]. We used our case-referent study, performed during
the peak of the epidemic, for such estimates. Our analysis revealed that about half of the
coeliac disease cases during the epidemic might have been avoided if all infants had
been introduced to gluten in small amounts while still being breast-fed (Fig. 6).
120
100
= Preventable cases
No of cases
80
60
40
20
0
A B C D
Continued
Small-medium
Continued
Large
Discontinued
Small-medium
Discontinued
Large
Breast-feeding status at introduction of flourand amount of flour given
Fig. 6. Preventable coeliac disease cases below two years of age with respect to breastfeeding
status at introduction of gluten-containing flour into the diet and the amount of
flour given. Risk estimates were based on conditional logistic regression with 392 matched
sets of cases and referents, and adjusted for age of the infant when flour was introduced.
37
Odds ratio (95% CI); A) 1.0, B) 2.0 (1.4, 3.0), C) 2.8 (1.9, 4.0), D) 3.3 (2.3, 4.8). From
with permission.

INFANT FEEDING PRACTICES AND COELIAC DISEASE
55
Thus, other casual exposures changing over time must also have contributed to the
epidemic.
However, it is likely that the proportion of coeliac disease cases that might be
avoided by appropriate measures varies among different populations, being higher in
populations with unusually unfavourable environmental exposures, as was apparently
the situation for Swedish infants during the high incidence years of the epidemic.
What will follow after the epidemic?
In a Swedish study based on symptomatic adult coeliac disease patients, the highest
79
age-specific prevalence was reported for people born between 1927 and 1936 , and
interestingly, the majority of cases in our population-based adult screening study were
5
also born during that time period . Thus, these studies indicate a cohort effect, i.e. the
life span of certain cohorts coincides with environmental exposures that have resulted
in an excess risk for coeliac disease throughout life.
Accordingly, it might also be that as a consequence of an unfavourable exposure
during their first years of life, the birth cohorts of the epidemic years will carry an
excess risk for coeliac disease throughout their lives. If so, the cohorts of the postepidemic
period might have a decreased lifetime risk for coeliac disease. Thus far, at
comparable ages the cohorts of this later period actually have a lower cumulative
6
1992
6
Cases per 1000 births
5
4
3
2
1
1989-90
1991
1987-88
1985-86
1983-84
1978-82
1973-77
5
4
3
2
1
1993
1994
1995
1996
1997
0
0 2 4 6 8 10 12 14
Age (years)
0
1998
0 2 4 6
Fig. 7. Cumulative incidence of coeliac disease by age for the birth cohorts from 1973 to
1998. To reduce the graph complexity, the cohorts of 1973 to 1982 are aggregated in groups
of five, the cohorts of 1983 to 1990 in groups of two, while the cohorts from 1991 to 1998
are reported separately.

56 INFANT FEEDING PRACTICES AND COELIAC DISEASE
incidence than the cohorts of the epidemic period (Fig. 7).
A longer follow-up will reveal to what extent this lower risk continues, and to what
extent new cases develop later in life.
Large screening studies of cohorts from both the epidemic and post-epidemic
periods, with repeated screening for a considerable number of years, would increase our
understanding of the natural history of coeliac disease development in relation to
exposures early in life. These exposures would be complicated, of course, by the
addition of exposures later in life.
An option for primary prevention
Primary prevention of coeliac disease is a clearly desirable option, as it would
contribute substantially to the health of the general population. By definition, it aims at
intervening before the disease has developed.
Genetic modification of the gluten-containing cereals so that they lose their ability
to trigger celiac disease might someday become an alternative, although this is
something for the distant future. Another option, at least theoretically, would be
induction of immune tolerance to gluten peptides by giving a vaccine. Even today,
coeliac disease could be effectively prevented if the use of gluten-containing cereals
was abandoned, but in most countries such a suggestion would be considered
unacceptable.
However, if coeliac disease has a multifactorial aetiology, which is likely, then
primary prevention would be possible, at least in some individuals, without completely
abandoning the use of dietary gluten. This might be attained through a change in
component causal exposures, thereby increasing the chance for infants to develop oral
tolerance to gluten, and possibly also promoting the maintenance of tolerance
throughout life (Fig. 3).
When knowledge is consolidated from the clinical, epidemiological and basic
sciences, it seems likely that infant feeding practices, in addition to the mere presence of
gluten in the diet, have an important role in coeliac disease development. Thus, a
significant contribution to primary prevention seems possible through a gradual
introduction of gluten-containing foods into the infant diet before breast-feeding is
discontinued. The duration of breast-feeding in itself seems to reduce the risk even
further, and a long duration should thus be promoted. The risk for coeliac disease is then
clearly reduced, at least up to two years of age. This infant feeding pattern most likely
also contributes to a decreased lifetime risk of coeliac disease, although this remains to
be established.
Only a few of all the potential causal environmental exposures have thus far been
explored, and the search for such factors, which exhibit their effect during different
periods in the life span, should be intensified. This approach most likely will lead to the
identification of other entry-points for primary prevention.
Acknowledgements
We thank Don Kasarda, US Department of Agriculture, Albany, California, for
valuable comments on the manuscript.

INFANT FEEDING PRACTICES AND COELIAC DISEASE
57
References
1. Grodzinsky E, Franzen L, Hed J, Ström M. High prevalence of celiac disease in healthy adults
revealed by antigliadin antibodies. Ann Allergy 1992; 69: 66-70.
2. Catassi C, Rätsch I-M, Fabiani E, Rossini M, Bordicchia F, Candela F, et al. Coeliac disease
in the year 2000: exploring the iceberg. Lancet 1994; 343: 200-3.
3. Not T, Horvath K, Hill ID, Partanen J, Hammed A, Magazzu G, et al. Celiac disease risk in the
USA: High prevalence of antiendomysium antibodies in healthy blood donors. Scand J
Gastroenterol 1998; 33: 494-8.
4. Johnston SD, Watson RGP, McMillan SA, Sloan J, Love AHG. Coeliac disease detected by
screening is not silent--simply unrecognized. QJM 1998; 91: 853-60.
5. Ivarsson A, Persson LÅ, Juto P, Peltonen M, Suhr O, Hernell O. High prevalence of
undiagnosed coeliac disease in adults: a Swedish population-based study. J Intern Med
1999; 245: 63-8.
6. Catassi C, Rätsch IM, Gandolfi L, Pratesi R, Fabiani E, El Asmar R, et al. Why is coeliac
disease endemic in the people of the Sahara? [letter]. Lancet 1999; 354: 647-8.
7. Csizmadia CGDS, Mearin ML, von Blomberg BME, Brand R, Verloove-Vanhorick SP. An
iceberg of childhood coeliac disease in the Netherlands [letter]. Lancet 1999; 353: 813-4.
8. Gandolfi L, Pratesi R, Cordoba JC, Tauil PL, Gasparin M, Catassi C. Prevalence of celiac
disease among blood donors in Brazil. Am J Gastroenterol 2000; 95: 689-92.
9. Rabassa EB, Sagaró E, Fragoso T, Castañeda C, Gra B. Coeliac disease in Cuban children.
Arch Dis Child 1981; 56: 128-31.
10. Sher KS, Fraser RC, Wicks AC, Mayberry JF. High risk of coeliac disease in Punjabis.
Epidemiological study in the south Asian and European populations of Leicestershire.
Digestion 1993; 54: 178-82.
11. Rawashdeh MO, Khalil B, Raweily E. Celiac disease in Arabs. J Pediatr Gastroenterol Nutr
1996; 23: 415-8.
12. Demir H, Yuce A, Kocak N, Ozen H, Gurakan F. Celiac disease in Turkish children:
presentation of 104 cases. Pediatr Int 2000; 42: 483-7.
13. Marsh MN. Screening for latent gluten sensitivity: questions many, but answers few
[editorial; comment]. Eur J Gastroenterolog Hepatol 1996; 8: 3-6.
14. Catassi C, Fabiani E, Ratsch IM, Coppa GV, Giorgi PL. Celiac disease in the general
population: should we treat asymptomatic cases? J Pediatr Gastroenterol Nutr 1997; 24:
S10-2.
15. Lohiniemi S. Tricky to find, hard to treat, impossible to cure. Lancet 2001; 358: 14.
16. Ellis A. Coeliac disease: Previous family studies. In: McConnell RB, ed. The genetics of
coeliac disease: International symposium on the genetics of coeliac disease, 1979,
Liverpool, England. MTP, Lancaster: 1981; 197-9.
17. Polanco I, Biemond I, van Leeuwen A, et al. Gluten sensitive enteropathy in Spain: Genetic
and environmental factors. In: McConnell RB, ed. The genetics of coeliac disease:
International symposium on the genetics of coeliac disease, 1979, Liverpool, England. MTP,
Lancaster: 1981; 211-31.
18. Bardella MT, Fredella C, Prampolini L, Marino R, Conte D, Giunta AM. Gluten sensitivity in
monozygous twins: a long-term follow-up of five pairs. Am J Gastroenterol 2000; 95: 1503-5.
19. Hervonen K, Karell K, Holopainen P, Collin P, Partanen J, Reunala T. Concordance of
Dermatitis Herpetiformis and Celiac Disease in Monozygous Twins. J Invest Dermatol 2000;
115: 990-3.
20. Ferguson A. Coeliac disease research and clinical practice: maintaining momentum into the
twenty-first century. Baillieres Clin Gastroenterol 1995; 9: 395-412.
21. Walker-Smith J, Murch S. Coeliac disease. Diseases of the small intestine in childhood. 4 ed.
Isis Medical Media; Oxford: 1999; 235-77.

58 INFANT FEEDING PRACTICES AND COELIAC DISEASE
22. Sollid LM. Molecular basis of celiac disease. Annu Rev Immunol 2000; 18: 53-81.
23. Sollid LM, Lundin KEA. An inappropriate immune response. Lancet 2001; 358: s13.
24. Petruzzelli F, Bonamico M, Ferrante P, Grillo R, Mora B, Mariani P, et al. Genetic
contribution of the HLA region to the familial clustering of coeliac disease. Ann Hum Genet
1997; 61: 307-17.
25. Bevan S, Popat S, Braegger CP, Busch A, O'Donoghue D, Fälth-Magnusson K, et al.
Contribution of the MHC region to the familial risk of coeliac disease. J Med Genet 1999; 36:
687-90.
26. Clot F, Babron MC. Genetics of celiac disease. Molecular Genetics and Metabolism 2000;
71: 76-80.
27. King AL, Ciclitira PJ. Celiac disease: strongly heritable, oligogenic, but genetically complex.
Molecular Genetics and Metabolism 2000; 71: 70-5.
28. Risch N. Assessing the role of HLA-linked and unlinked determinants of disease. Am J Hum
Genet 1987; 40: 1-14.
29. Salvetti M, Ristori G, Bomprezzi R, Pozzilli P, Leslie RDG. Twins: mirrors of the immune
system. Immunol Today 2000; 21: 342-7.
30. Sing CF, Haviland MB, Reilly SL. Genetic architecture of common multifactorial diseases.
Ciba Found Symp 1996; 197: 211-32.
31. Cronin CC, Shanahan F. Insulin-dependent diabetes mellitus and coeliac disease [see
comments]. Lancet 1997; 349: 1096-7.
32. Strobel S, Mowat AM. Immune responses to dietary antigens: oral tolerance. Immunol Today
1998; 19: 173-81.
33. Constant SL, Bottomly K. Induction of Th1 and Th2 CD4+ T cell responses: the alternative
approaches. Annu Rev Immunol 1997; 15: 297-322.
34. Kay AB. Allergy and allergic diseases. First of two parts. N Engl J Med 2001; 344: 30-7.
35. Hernell O, Forsberg G, Hammarström S, Ivarsson A, Hammarström M-L. Coeliac disease: A
model to study oral tolerance. In: Hart AL, Stagg AJ, Graffner H, Glise H, Falk P, Kamm MA,
eds. Gut Ecology. Martin Dunitz, London: 2002; 73-9.
36. Ivarsson A, Persson LÅ, Nyström L, Ascher H, Cavell B, Danielsson L, et al. Epidemic of
coeliac disease in Swedish children. Acta Paediatr 2000; 89: 165-71.
37. Ivarsson A. On the multifactorial aetiology of coeliac disease. Scand J Nutr 2001; 45: 184-5.
38. Ascher H, Krantz I, Kristiansson B. Increasing incidence of coeliac disease in Sweden. Arch
Dis Child 1991; 66: 608-11.
39. Cavell B, Stenhammar L, Ascher H, Danielsson L, Dannaeus A, Lindberg T, et al. Increasing
incidence of childhood coeliac disease in Sweden. Results of a national study. Acta Paediatr
1992; 81: 589-92.
40. Carlsson AK, Axelsson IEM, Borulf SK, Bredberg ACA, Ivarsson SA. Serological screening
for celiac disease in healthy 2. 5-year-old children in Sweden. Pediatrics 2001; 107: 42-5.
41. Beaglehole R, Bonita R, Kjellström T. Basic epidemiology. World Health Organization;
Geneva, 1993; 1-175.
42. Cummins AG, Thompson FM. Postnatal changes in mucosal immune response: A
physiological perspective of breast feeding and weaning. Immunol Cell Biol 1997; 75: 419-
29.
43. Hanson LA. Breastfeeding provides passive and likely long-lasting active immunity. Ann
Allergy Asthma Immunol 1998; 81: 523-37.
44. Garofalo RP, Goldman AS. Expression of functional immunomodulatory and antiinflammatory
factors in human milk. Clin Perinatol 1999; 26: 361-77.
45. Goldman AS. Modulation of the gastrointestinal tract of infants by human milk. Interfaces
and interactions. An evolutionary perspective. J Nutr 2000; 426S-31S.
46. Andersen DH, Di Sant'Agnese PA. Idiopathic celiac disease: I. Mode of onset and diagnosis.
Pediatrics 1953; 11: 207-22.

INFANT FEEDING PRACTICES AND COELIAC DISEASE
59
47. Greco L, Mayer M, Grimaldi M, Follo D, De Ritis G, Auricchio S. The effect of early feeding
on the onset of symptoms in celiac disease. J Pediatr Gastroenterol Nutr 1985; 4: 52-5.
48. Mäki M, Kallonen K, Lähdeaho ML, Visakorpi JK. Changing pattern of childhood coeliac
disease in Finland. Acta Paediatr Scand 1988; 77: 408-12.
49. Kelly DA, Phillips AD, Elliott EJ, Dias JA, Walker-Smith JA. Rise and fall of coeliac disease
1960-85. Arch Dis Child 1989; 64: 1157-60.
50. Logan RFA, Rifkind EA, Busuttil A, Gilmour HM, Ferguson A. Prevalence and "incidence" of
celiac disease in Edinburgh and the Lothian region of Scotland. Gastroenterology 1986; 90:
334-42.
51. Stevens FM, Egan-Mitchell B, Cryan E, McCarthy CF, McNicholl B. Decreasing incidence of
coeliac disease. Arch Dis Child 1987; 62: 465-8.
52. Auricchio S, Follo D, De Ritis G, Giunta A, Marzorati D, Prampolini L, et al. Does breast
feeding protect against the development of clinical symptoms of celiac disease in children? J
Pediatr Gastroenterol Nutr 1983; 2: 428-33.
53. Greco L, Auricchio S, Mayer M, Grimaldi M. Case control study on nutritional risk factors in
celiac disease. J Pediatr Gastroenterol Nutr 1988; 7: 395-9.
54. Fälth-Magnusson K, Franzén L, Jansson G, Laurin P, Stenhammar L. Infant feeding history
shows distinct differences between Swedish celiac and reference children. Pediatr Allergy
Immunol 1996; 7: 1-5.
55. Peters U, Schneeweiss S, Trautwein EA, Erbersdobler HF. A case-control study of the effect of
infant feeding on celiac disease. Ann Nutr Metab 2001; 45: 135-42.
56. Ascher H, Krantz I, Rydberg L, Nordin P, Kristiansson B. Influence of infant feeding and
gluten intake on coeliac disease. Arch Dis Child 1997; 76: 113-7.
57. Ivarsson A, Hernell O, Stenlund H, Persson LÅ. Breast-feeding protects against celiac
disease. Am J Clin Nutr 2002; 75: 914-21.
58. Mäki M, Holm K, Ascher H, Greco L. Factors affecting clinical presentation of coeliac
disease: Role of type and amount of gluten-containing cereals in the diet. In: Auricchio S,
Visakorpi JK, eds. Common food intolerances 1: Epidemiology of coeliac disease. Karger,
Basel: 1992; 76-82.
59. Ascher H, Holm K, Kristiansson B, Mäki M. Different features of coeliac disease in two
neighbouring countries. Arch Dis Child 1993; 69: 375-80.
60. Weile B, Cavell B, Nivenius K, Krasilnikoff PA. Striking differences in the incidence of
childhood celiac disease between Denmark and Sweden: A plausible explanation. J Pediatr
Gastroenterol Nutr 1995; 21: 64-8.
61. Mitt K, Uibo O. Low cereal intake in Estonian infants: the possible explanation for the low
frequency of coeliac disease in Estonia. Eur J Clin Nutr 1998; 52: 85-8.
62. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular
and immunobiologic approach to the spectrum of gluten sensitivity ('celiac sprue').
Gastroenterology 1992; 102: 330-54.
63. Doherty M, Barry RE. Gluten-induced mucosal changes in subjects without overt smallbowel
disease. Lancet 1981; i: 517-20.
64. Leigh RJ, Marsh MN, Crowe P, Kelly C, Garner V, Gordon D. Studies of intestinal lymphoid
tissue. IX. Dose-dependent, gluten-induced lymphoid infiltration of coeliac jejunal
epithelium. Scand J Gastroenterol 1985; 20: 715-9.
65. Ferguson A, Blackwell JN, Barnetson RS. Effects of additional dietary gluten on the small
intestinal mucosa of volunteers and of patients with dermatitis herpetiformis. Scand J
Gastroenterol 1987; 22: 543-9.
66. Catassi C, Rossini M, Ratsch IM, Bearzi I, Santinelli A, Castagnani R, et al. Dose dependent
effects of protracted ingestion of small amounts of gliadin in coeliac disease children: a
clinical and jejunal morphometric study. Gut 1993; 34: 1515-9.
67. Anderson CM, Gracey M, Burke V. Coeliac disease. Some still controversial aspects. Arch

60 INFANT FEEDING PRACTICES AND COELIAC DISEASE
Dis Child 1972; 47: 292-8.
68. Lindberg T. Coeliac disease and infant feeding practices [letter]. Lancet 1981; i: 449
69. Stenhammar L, Ansved P, Jansson G, Jansson U. The incidence of childhood celiac disease in
Sweden. J Pediatr Gastroenterol Nutr 1987; 6: 707-9.
70. Kagnoff MF, Paterson YJ, Kumar PJ, Kasarda DD, Carbone FR, Unsworth DJ, et al.
Evidence for the role of a human intestinal adenovirus in the pathogenesis of coeliac disease.
Gut 1987; 28: 995-1001.
71. Lähdeaho ML, Parkkonen P, Reunala T, Mäki M, Lehtinen M. Antibodies to E1b proteinderived
peptides of enteric adenovirus type 40 are associated with celiac disease and
dermatitis herpetiformis. Clin Immunol Immunopathol 1993; 69: 300-5.
72. Howdle PD, Blair Zajdel ME, Smart CJ, Trejdosiewicz LK, Blair GE, Losowky MS. Lack of a
serologic response to an E1B protein of adenovirus 12 in coeliac disease. Scand J
Gastroenterol 1989; 24: 282-6.
73. Vesy CJ, Greenson JK, Papp AC, Snyder PJ, Qualman SJ, Prior TW. Evaluation of celiac
disease biopsies for adenovirus 12 DNA using a multiplex polymerase chain reaction. Mod
Pathol 1993; 6: 61-4.
74. Holm S, Andersson Y, Gothefors L, Lindberg T. Increased protein absorption after acute
gastroenteritis in children. Acta Paediatr 1992; 81: 585-8.
75. Ivarsson A, Hernell O, Nyström L, Persson LÅ. Children born in the summer have increased
risk for coeliac disease. J Epidemiol Community Health [in press].
76. Ivarsson A. On the multifactorial etiology of celiac disease: An epidemiological approach to
the Swedish epidemic. Department of Clinical Sciences, Pedatrics, Umeå University,
Sweden. [Dissertation] 2001.
77. Rothman KJ, Greenland S. Causation and causal inference. In: Rothman KJ, Greenland S,
eds. Modern epidemiology: 2nd ed. Lippincott - Raven, Philadelphia: 1998; 7-28.
78. ESPGAN committee on nutrition. Guidelines on infant nutrition III. Recommendations for
infant feeding. Acta Paediatr Scand Suppl 1982; 302: 16-20.
79. Hallert C, Gotthard R, Jansson G, Norrby K, Walan A. Similar prevalence of coeliac disease
in children and middle-aged adults in a district of Sweden. Gut 1983; 24: 389-91.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 61-73
Mechanisms of oral tolerance - lessons
for coeliac disease?
Conleth Feighery
Department of Immunology, Trinity College Dublin and St. James's Hospital, Dublin, Ireland
Introduction
Coeliac disease (CD) is an inflammatory disease of the upper small intestine and
1
results from gluten ingestion in genetically susceptible individuals . Gluten is the
essential environmental factor in the development of CD: this is demonstrated by the
clinical and mucosal recovery which follows the institution of a gluten-free diet.
Genetic factors also play a critical role in the development of CD and a strong
association with certain Major Histocompatibility Complex (MHC) class II genes
2
(HLA-DQ2 and HLA-DQ8) is well established . However, CD is a polygenic disorder
and many additional genes, yet to be identified, also contribute to the pathogenesis of
CD. There is no evidence to suggest that these genes are in any sense “abnormal” and
similar genetic polymorphisms are well represented in the normal population.
Since gluten is part of the normal stable diet, and the genes which permit the
development of CD are present in the normal population, the question can be
reasonably posed - what makes a particular individual susceptible to gluten induced
damage? It is speculated that additional factors, for example infectious agents and sex
hormones, may be involved. Other environmental factors may also contribute but
currently there are few clues as to what these are. This leads to the broader issue how
does the gut immune system “view” foods? Foods of their essence are foreign proteins
and as such might be targeted for elimination, as is the case with any other foreign
antigen.
CD is a common disorder. According to current evidence, it may affect 0.5 to 1% of
3-4
the population . This suggests that gluten has unique, but ill-understood, properties
which frequently results in a T cell, delayed-type hypersensitivity response to this food
antigen. There is little evidence that other food proteins cause a similar reaction. In
contrast, many foods stimulate food allergy, a type I hypersensitivity reaction with
5
specific IgE production to the offending antigen . In food allergy, a systemic reaction to
61

62
MECHANISMS OF ORAL TOLERANCE
the food is quite common and the reaction remits promptly after the food has been
eliminated from the diet. Food allergy often develops in early childhood but may no
6
longer be clinically problematic in later years .
Concept of immune tolerance
The absent or low immune reactivity of the gut to food antigens raises the concept of
7
immune tolerance . Essentially, it would appear that a limited, local reaction develops
to ingested foods and this response is confined to the bowel mucosa. This permits the
antigenicity of food to be recognised but does not allow a damaging immune response
to develop. In this situation, immune tolerance of the food is said to have taken place.
From the viewpoint of an immunologist, tolerance is defined as the inability or
failure of lymphocytes to respond to an antigen. The importance of tolerance is best
observed in the concept of self-tolerance: in this, an individual's lymphocytes fail to
react with self-antigen and as a consequence the development of autoimmune disease is
prevented. Self-tolerance is a central paradigm in immunology, first proposed some
8
100 years ago by Paul Erlich who predicted that “horror autotoxicus” might develop if
such a control system was not in place.
Central and peripheral tolerance
7
Immune tolerance can be divided into central and peripheral forms . Central
tolerance relates to events which take place in organs such as the thymus and bone
marrow. In this, it is envisaged that as T cells mature in the thymus, those cells which
might be capable of self reactivity (by binding strongly to self-antigens presented by
self-MHC molecules) undergo apoptosis and are thereby deleted. These events prevent
the development of T cell mediated autoimmunity. There is considerable experimental
evidence in support of such thymic events. It is also proposed that a similar mechanism
9
may cause deletion of B cells in the bone marrow .
Despite the evidence for lymphocyte deletion in the thymus, it is not precisely
known how universal this process is. Indeed, it is increasingly being accepted that a low
level of autoreactivity is physiological and may even be important for normal immune
10
function . Since it is improbable that all T lymphocytes capable of self-reactivity are
removed by this process, a second mechanism has been invoked: this is referred to as
peripheral tolerance (Fig. 1). In this, the potential self-reactivity of circulating
lymphocytes is curbed and these cells are rendered tolerant in the periphery. The
mechanisms involved in peripheral tolerance are not well understood.
7-9
There is evidence that some lymphocytes are deleted in the periphery . A more popular
concept is that lymphocyte anergy develops.
One suggested basis for the development of anergy is that when lymphocytes encounter
self-antigen, they do so in the absence of the appropriate additional signals provided by
co-stimulatory molecules. These molecules include in particular the B7/CD28
interaction, which markedly enhances interleukin (IL)-2 production and hence full T
9
cell activation . If such co-stimulation is absent, the lymphocyte enters a dormant state,

MECHANISMS OF ORAL TOLERANCE
63
Peripheral tolerance
•Clonal deletion
T cell
apoptosis
•Clonal “anergy”
T cell
Inactive / asleep?
•Regulatory clone
T reg
T reg
cytokines
Fig. 1. Mechanisms of peripheral tolerance. T reg = regulatory T cell
incapable of reacting with a specific antigen. It is possible that anergic T cells can be
rescued from this state, thereby permitting such a cell to display auto-reactivity.
The presence of circulating, regulatory T cells is a further mechanism that has been
proposed to explain the phenomenon of peripheral tolerance. Following appropriate
activation, these cells release cytokines which inhibit the auto-reactivity of the immune
system. In the 1970s, considerable interest surrounded the concept of T suppressor cells
performing this task. However, since no specific cell markers or specific suppressor
factors could be reliably identified, the concept became disreputable and it was
damaging to a research grant application or publication to mention the “S” word!
However, fashions change and there is a saying that “what goes around, comes around”
and there is now a growing belief that regulatory T cells play a central role in the
11
prevention of auto-immunity . According to this concept, regulatory T cells can be
12
considered to conduct the orchestra of the immune response .
Finally, a fourth mechanism whereby peripheral tolerance may operate is the
7
concept of antigen “ignorance” . According to this concept, antigen internalised within
a tissue or present in only very small quantities, is unable to interact productively with
lymphocytes. The immune control mechanisms described above do not develop for
such antigens. If tissue damage develops, antigen is released and a potentially
damaging immune response may develop.
Mucosal immune system
Exposure to external antigens in humans commonly takes place at two different
sites: the skin and the mucosal surfaces. Although antigen can enter via the skin, the

64 MECHANISMS OF ORAL TOLERANCE
potential for antigen interaction at mucosal surfaces is even greater. The nature of skin,
a multi-layered relatively impermeable structure, inhibits the penetration of harmful
antigenic structures such as bacteria. In contrast, mucosal surfaces are typically
protected by a single layer of epithelial cells. This makes the task of antigen access all
the easier. Furthermore, in sites such as the small intestine which have a major role in
absorption of nutrients, constant sampling of luminal material takes place. Thus, the
mucosal immune system needs to differentiate between harmful antigens (such as
pathogenic bacteria) and innocuous antigens (such as food).
The organisation of the mucosal immune response is structured to permit a local
response to the majority of antigens which penetrate to the lamina propria cells of the
mucosa. However, this immune response is typically of low intensity and remains local.
Systemic reaction to such antigen is either minimal or totally absent. Precisely how this
type of response is achieved is not fully understood. Secretory IgA produced in the
mucosa may play an important role but other immune components are doubtless also
13
involved . This controlled mucosal immune response could be considered as a form of
peripheral tolerance and the term “oral tolerance” is frequently used.
Oral tolerance
The phenomenon of oral tolerance has been extensively investigated in various
14-16
animal model systems over the past 30 years . As with all experimental models, they
may not precisely replicate the normal in vivo situation but may nonetheless be
informative about how local mucosal immune responses are regulated. The essential
feature of experimental oral tolerance is that by feeding an antigen orally, the induction
of a systemic immune response to that antigen is prevented.
A typical experiment examining oral tolerance is to study what happens to an animal
after oral administration of ovalbumin: when the animal is later exposed to ovalbumin
systemically (by injection), circulating lymphocyte responsiveness to that antigen is
16
either absent or significantly diminished . Animals not pre-fed ovalbumin have a
normal systemic response to the antigen. This effect of oral feeding causing systemic
tolerance could be regarded as a form of induced peripheral tolerance. In oral tolerance,
although hyporesponsiveness of both specific T and B lymphocytes is noted, T cells
seem to be more profoundly affected. There is no well validated equivalent experiment
confirming the induction of oral tolerance in man.
Relevance of oral tolerance to coeliac and other gastrointestinal diseases
The essence of the concept of oral tolerance is that local gut mucosal immune
responses are regulated and prevent systemic reactions to orally ingested antigens. If,
for whatever reason, a breakdown in oral tolerance develops, an excessive
inflammatory mucosal reaction may take place. Thus, such a breakdown may be
regarded as central to the development of inflammatory gastrointestinal disorders.
Examples of these include coeliac disease, Crohn's disease and ulcerative colitis. In the
latter two diseases the responsible gut antigens are not known, although bacterial
antigens are suspected. In the case of CD, it is well established since the 1950s that
17
gluten is the causative agent . It is postulated that immune tolerance of gluten, found in
normal subjects, is lost in CD patients and the consequence is inflammation at the major

MECHANISMS OF ORAL TOLERANCE
65
site of gluten exposure, the small intestine.
Experimental models of oral tolerance differ from CD in certain respects. In CD, the
most prominent evidence of inflammation occurs locally, where antigen exposure takes
place. This is not the case in animals in which the effect of abrogation of oral tolerance
predominantly affects systemic immune responses. However, in some animal models,
18
oral challenge with the antigen was shown to cause a mild mucosal lesion . Of interest,
there is also evidence that gluten intolerance is associated with systemic damage in
some individuals, as represented by diseases such as dermatitis herpetiformis, epilepsy
19-21
and cerebellar ataxia . Moreover, there is some evidence that CD is also associated
with inflammation in the distal intestine and rectal histological changes have been
22
reported . Finally, it is worth noting that systemic antibody responses to gluten are
found in a range of disorders, in the absence of local gut damage or evidence that these
23
antibodies have a systemic pathogenic significance .
Interest in oral tolerance
Over the decades, an interest in the topic of oral tolerance has persisted, in the belief
that its study could lead to a more fundamental understanding of homeostasis of the
immune system. Furthermore, it is postulated that systemic inflammatory diseases
14
might be controlled by inducing oral tolerance to the stimulating antigen . According
to this concept, the administration of oral antigen (such as the autoantigenic target of
systemic disease) could lead to a reduced or aborted systemic response to this antigen.
Thus various feeding studies with myelin basic protein (in patients with multiple
24 25
sclerosis ) and collagen (in patients with rheumatoid arthritis ) have been described.
Little therapeutic benefit has been noted in these studies to-date.
Finally, an understanding of oral tolerance mechanisms is pertinent to the
development of oral vaccines, since this route of vaccination is designed to give not
only a local protective immune response, but also systemic protection.
Factors involved in control of the mucosal immune response
A clear understanding of the homeostatic mechanisms responsible for controlling
the mucosal immune response has yet to be reached. Undoubtedly, many interrelated
factors play a role (Fig. 2) and these will be considered briefly here.
The epithelial lining layer “outpost” of mucosal immunity
The epithelial cell sitting on its basement membrane, in a sense guards the mucosal
tissue from full, free exposure to antigen. However, the epithelial cell is not an absolute
barrier and it permits rapid absorption of soluble antigen. Thus, epithelial cells, like
many other cells in the body, constantly sample the external milieu. We know that food
antigens can rapidly penetrate this epithelial layer, since these antigens are found in the
26
circulation even within minutes of food ingestion .
In addition to antigen absorption, epithelial cells have also been shown to process
and load antigen into MHC II molecules. This was demonstrated in the case of gliadin,
27
using immuno-electron microscopy . Furthermore, it is known that MHC class II
28
molecules are expressed or can be induced on intestinal epithelial cells . These
findings raise the possibility that epithelial cells might function as antigen presenting

66 MECHANISMS OF ORAL TOLERANCE
Epithelial
cells
IELs
Anatomical
structure
Antigen
structure
Lamina
propria
T cells
Environment
(cytokines)
Antigen
presenting
cells
Fig. 2. Components involved in control of mucosal immune response. IELs =intraepithelial
lymphocytes.
cells. However, there is little evidence that conventional antigen stimulation takes
29
place and it may be that epithelial cells interact with T lymphocytes using non-
30
classical MHC molecules such as CD1d . Since epithelial cells do not express the costimulatory
molecules CD80 and CD86, it might be expected that these cells play a role
31-32
in inducing anergy of T cells rather than an active T cell response .
The importance of the epithelial cell barrier is emphasised by what happens when
barrier function is disrupted. This may occur during the course of local inflammation in
33
which tight junction function is impeded . This in turn would allow ready access of
antigens to immunocompetent cells in the lamina propria. This may explain the
presence of high levels of antibodies to dietary antigens (including gliadin) in Crohn's
34
and peptic ulcer disease . A further demonstration of the importance of the epithelial
cell barrier is found in certain gene knockout mouse models in which epithelial cell
adherence or their cytoskeletal structure is disrupted: in these models, genes for either
E-cadherin or keratin 8 (respectively) are knocked out, with the spontaneous
35-36
development of inflammatory bowel disease as a consequence .
Antigen factors
The size, structure and nature of antigens affects the response of the mucosal
immune system. Thus in health, soluble dietary antigens induce a limited, local immune
response and often a barely discernible response in the systemic circulation. Likewise,
the gut flora induces a local immune response only. In contrast, particulate antigen is
37
absorbed by specialised epithelial cells called microfold or M cells . M cells are found

MECHANISMS OF ORAL TOLERANCE
67
interposed between the lining epithelial cells and following uptake of antigen,
particulate antigen is processed and presented to local T helper cells. This in turn results
in a systemic immune response to the antigen.
As discussed earlier, in contrast to other food antigens, gluten appears to have an
unique capacity to stimulate a local T cell mediated immune response in CD patients,
leading to damage of the local mucosa. The chemical properties of gluten involved in
eliciting such a response are unknown, since no essential structural difference is
10
thought to exist between self-antigens and foreign antigens .
Intraepithelial lymphocytes
Intraepithelial lymphocytes (IELs) are T cells located within the surface epithelium
of the mucosa. Despite intense research of these cells over the past two decades, their
38
principal function remains enigmatic . The majority of IELs express the CD8
molecule and employ the abT cell receptor (TCR). It is well established that increased
numbers of IELs are found in certain intestinal inflammatory states such as coeliac
disease. Moreover, in coeliac disease the number of IELs expressing the gdTCR
39
increases, with as many as 30% belonging to this phenotype . According to recent
evidence, IELs may include heterogeneous populations of T cells, natural killer T cells
40
and also natural killer cells . It has been speculated that IELs play a role in oral
tolerance and there is some experimental evidence to support a role for gdT cells in
41
tolerance induction .
Lamina propria T cells
The major population of lamina propria T cells express the CD4 molecule and
employ the abTCR. These T helper cells are thought to have initially encountered
mucosal antigen in organised lymphoid structures such as Peyer's patches and after
passage through the systemic circulation have specifically homed back to the intestine.
In this location, lamina propria T cells play a major role in orchestrating other cells of
12
the immune system and act as conductors of this orchestra . These T cells possess a full
range of surface molecules to allow productive interaction with many other cell types.
Moreover, individual populations manufacture the variety of cytokines required to
influence B cell production of specific antibody isotypes, to induce specific populations
13
of antigen presenting cells and to activate other populations of T cells . Thus,
regulation of immune events in the intestine is largely a function of these cells, and
hence maintenance of “oral tolerance” or mucosal immune homeostasis is likely to be
achieved by these cells.
Dendritic and other antigen presenting cells
The activation of intestinal T cells requires the presentation of antigen by local
populations of antigen presenting cells (APCs). This task may be performed by a
variety of APCs including dendritic cells, macrophages, B cells and, as discussed
12-13
earlier, possibly by enterocytes . Populations of all these cell types have been
identified in the mucosal tissue. In individual situations, it is likely that the presentation
of antigen is principally performed by specific cell types. Dendritic cells, in particular,
may play a central role in presenting antigen and influencing the evolution of specific

68 MECHANISMS OF ORAL TOLERANCE
types of T cell populations in the intestine.
Much has been learned about dendritic cells in recent years: some types particularly
42-43
influence the development of Th1 cells and others Th2 cells . These dendritic
populations can be identified on the basis of expression or absence of the adhesion
molecule CD11c and also by the type of cytokines (such as IL-12) which they produce.
The heterogeneity of dendritic cells is increasingly being recognised and it is now
appreciated that the major function of some populations is to induce T cell tolerance,
44
rather than a productive immune response .
As with all events in the immune system, a circular feedback system (so-called
inter-cellular cross-talk) operates between dendritic cells and the T cells which they
help stimulate (fig. 3). Thus, the cytokine products of dendritic cells influence the T cell
populations which they generate and in turn, the cytokines produced by T cells
42
influence the type of dendritic cells which occupy that locale .
A further population of cells, likely to be involved in antigen presentation, which
express the macrophage scavenger receptor (CD163) has been identified by
45
immunohistochemistry in the central core of individual villi . The cytokine products of
this population of macrophage-like cells, including secreted CD163, may play a role in
46
down-regulation of the local immune response .
Cytokine environment of the mucosa
Virtually all cells in the mucosa, including enterocytes, produce a wide range of
cytokines and the profile of cytokines determines the net effect on the type of immune
response elicited. It is presumed that a major, final determining influence is the type of
cytokines produced by the local T helper populations. Since the principal default setting
of the mucosal immune response is towards tolerance of encountered antigen, the
cytokines which could determine tolerance have been investigated. The three cytokines
which have received most interest for their potential involvement are IL-4, IL-10 and
TGF-b.
For some time it was considered that T cell tolerance in the mucosa was achieved by
a predominant Th2 response to local antigens. Major products of Th2 cells include the
cytokines IL-4 and IL-10. However, in some animal model experiments, it was shown
that these cytokines are not required for tolerance induction. Moreover, in experimental
oral tolerance, inhibition of IgE production is a characteristic feature: the reverse would
14
be predicted in an IL-4 dominant environment . Hence, it is now proposed that
tolerance in the intestine may be achieved by a further T cell population, referred to as T
47
regulatory cells . The major cytokine products of these cells are IL-10 and
transforming growth factor-beta (TGF-b).
Although IL-10 is not essential in some models of oral tolerance, the importance of
this cytokine in maintaining tolerance is emphasized by the spontaneous development
48
of colitis in mice when the gene for this cytokine is “knocked out” . Of great interest,
49
TGF-bcan reverse the colitis . It is also known that IL-10 may be required for the
production of TGF-band that the latter cytokine inhibits the IL-12/interferon-gamma
50
(IFN-g) pathway, likely to play a central role in many inflammatory diseases .
It is currently hypothesised that control of the mucosal immune system is
maintained by regulatory T cells, but at present, other than studying their cytokine

MECHANISMS OF ORAL TOLERANCE
69
IEL
Lamina propria
M cell
CD4
DC
Cytokines of each cell type
influence each other
Epithelial
lining
Fig. 3. Interaction between CD4 + T cells and dendritic cells in lamina propria of the
mucosa. DC = dendritic cell; IEL = intraepithelial lymphocyte.
products, we do not have reliable, specific markers for these cells. It is probable that
TGF-band IL-10 are major products of these cells. However, it should be emphasised,
that even when a cytokine is shown not to be essential for experimental oral tolerance,
this does not mean that in vivo, this cytokine has no role to play, since a level of
redundancy is a common feature of the immune system.
Mucosal immune tolerance and coeliac disease
The preceding discussion of mucosal immune tolerance raises issues which may be
pertinent to the pathogenesis of coeliac disease. Is the essential defect in coeliac disease
a failure of the normal tolerance mechanisms, which allow gluten exposure but no
mucosal damage in normal individuals? If so, might tolerance of gluten be restored and
thereby permit CD patients to eat a normal diet? If this were possible, how might we go
about achieving it? Obviously, these are major issues and it could be argued that
restoring tolerance to gluten is a fanciful solution, perhaps unlikely ever to be achieved.
Nonetheless, with the gathering pace of modern knowledge and technology, it is a goal
which should at least be examined.
Adolescent coeliac disease
Some 30 years ago, one of the most frequent inquiries about the nature of CD, was
whether gluten intolerance was permanent. Of course, the dogma now is that permanent

70 MECHANISMS OF ORAL TOLERANCE
51
intolerance is a central feature of the disorder . However, in the past, cases of possible
52
transient gluten sensitivity were described . Is it possible that this phenomenon of
53-54
transiency is more common than currently appreciated? .
The events that happen to pediatric cases of CD, when they reach adolescence, may
be informative concerning the permanency of gluten intolerance. Detailed information
about these events is absent. Nonetheless, it is recognized that many children adhere
less well to a strict gluten free diet when they reach adolescence. This may represent the
general “rebellion” of teenagers! Despite eating gluten containing foods, few
symptoms may develop. This in turn may lead to increasing carelessness about their
diets. Some of these patients eventually have intestinal biopsies performed and the
mucosa can appear surprisingly normal or even completely normal. Doubts about the
accuracy of the initial diagnosis of CD may then be raised. Some of these patients are
then given a gluten challenge and it is reported that it can take many months or even
55-56
years for histological evidence of CD to return . This contrasts with gluten challenge
in patients with an adult diagnosis of CD when symptomatic and histological evidence
56
of disease relapse often only requires one or two months of gluten intake .
Although categorical data about adolescent remission of CD is lacking, the
information available strongly suggests that this really does occur. If this is the case, it
may indicate that restoration of tolerance to gluten is possible. The majority, if not all of
these patients, eventually develop symptomatic gluten intolerance in adulthood,
demonstrating the essential permanency of CD. However, it is possible that
investigation of adolescent CD patients, during their period of remission, may give
insight to T cell tolerance mechanisms towards gluten.
Conclusions
The gut mucosal immune response to the majority of food and other antigens is
typically of low intensity and remains confined to the local environment. The term ²oral
tolerance²may be applied to this controlled immune response. However, in the case of a
single food type, gluten, an enhanced T cell immune response develops in a substantial
number of individuals and features of coeliac disease develop. In this disorder, immune
tolerance of gluten is apparently lost. With an improved understanding of the
mechanisms of oral tolerance, it is possible that manipulation of the immune response
to gluten in coeliac patients could restore the normal physiological state of low
reactivity to this food antigen. This would remove the requirement of gluten-free diet
treatment of coeliac disease.
References
1. Feighery C. Coeliac disease Fortnightly review. BMJ 1999; 319: 236-9.
2. Kagnoff MF. Genetic basis of Coeliac Disease: role of HLA genes. In: MN Marsh
(Ed) Coeliac Disease, Oxford, Blackwell Scientific Publications, 1992; pp215-38.
3. Catassi C, Fabiani E, Rätsch IM, Coppa GV, Giorgi PL, Pierdomenico R, et al. The
coeliac iceberg in Italy. A multicentre antigliadin antibodies screening for coeliac
disease in school-age subjects. Acta Paediatr Suppl 1996; 412: 29-35.
4. McMillan SA, Watson RPG, McCrum EE, Evans AE. Factors associated with serum

MECHANISMS OF ORAL TOLERANCE
71
antibodies to reticulin, endomysium, and gliadin in an adult population. Gut 1996;
39: 43-7.
5. Furukawa CT. Adverse Reactions to Food. In: Allergy in Primary Care. Eds. LC.
Altman, JW. Becker, PV Williams. Publishers. W.B. Saunders Company 2000; 18:
259-64.
6. Sampson HA. What should we be doing for children with peanut allergy? J Pediatr
2000; 137: 741-3.
7. Kamradt T, Mitchison NA. Tolerance and Autoimmunity. N Engl J Med 2001; 344:
655-64.
8. Bibel DJ. Milestones in Immunology: a historical exploration. Science Tech
Publishers, Wisconsin, USA. 1988.
9. Janeway CA, Travers P, Walport M, Capra JD. Immunobiology: the immune system
in health and disease. Churchill Livingstone, 1999.
10. Davidson A, Diamond B. Advances in Immunology: Autoimmune Diseases. N Engl
J Med 2001; 345: 340-50.
11. Shevach EM. Certified Professionals: CD4+ CD25+ Suppressor T Cells. J Exp Med
2001; 193: F41-5.
12. Feighery C. Intestinal Immune System - A Hidden Symphony. In: Gastrointestinal
Immunology and Gluten-Sensitive Disease. Proceedings of the Sixth International
Symposium on Coeliac Disease held at Trinity College, Dublin in July 1992. Eds. C.
Feighery, C. O'Farrelly, Publishers Oak Tree Press, Dublin.
13. Brandtzaeg P, Baekkevold ES, Farstad IN, Jahnsen FL, Johansen FE, Nilsen EM, et
al. Regional specialization in the mucosal immune system: what happens in the
microcompartments? Immunol Today 1999; 3: 141-51.
14. Garside P, Mowat AM, Khoruts A. Oral tolerance in disease. Gut 1999; 44: 137-42.
15. Weiner HL. Induction and mechanism of action of transforming growth factor-betasecreting
Th3 regulatory cells. Immunol Rev 2001; 182: 207-14.
16. Mowat McI A. The regulation of immune responses to dietary protein antigens.
Immunology Today 1987; 8: 93-8.
17. Dicke WK. Coeliakie. PhD thesis. University of Utreacht, Utreacht, The
Netherlands, 1950.
18. Lamont AG, Mowat AM, Parrott DM. Priming of systemic and local delayed-type
hypersensitivity responses by feeding low doses of ovalbumin to mice. Immunology
1989; 66: 595-9.
19. Fry L. Dermatitis herpetiformis: from gut to skin, contributions to the nature of
coeliac disease. In: Coeliac Disease, Mäki M, Collin P, Visakorpi JK (eds).
Proceedings of the Seventh International Symposium on Coeliac Disease, Finland.
Publishers. Coeliac Disease Study Group, Tampere, Finland,1996; pp67-74.
20. Cronin CC, Jackson LM, Feighery C, Shanahan F, Abuzakouk M, Ryder DQ, et al.
Coeliac disease and epilepsy. Q J Med 1998; 91: 303-8.
21. Hadjivassiliou M, Grunewald RA, Chattopadhyay AK, Davies-Jones GA, Gibson A,
Jarratt JA et al. Clinical, radiological, neurophysiological, and neuropathological
characteristics of gluten ataxia. Lancet 1998; 352: 1582-5.
22. Ensari A, Marsh MN, Loft DE, Morgan S, Moriarty K. Morphometric analysis of
intestinal mucosa. V. Quantitative histological and immunocytochemical studies of
rectal mucosae in gluten sensitivity. Gut 1993; 34: 1225-9.

72 MECHANISMS OF ORAL TOLERANCE
23. McCormick PA, Feighery C, Dolan C, O'Farrelly C, Kelleher P, Graeme-Cook F, et
al. Altered gastrointestinal immune response in sarcoidosis. Gut 1988; 29: 1628-
31.
24. Weiner HL, Mackin GA, Matsui M, Orav EJ, Khoury SJ, Dawson DM, et al. Doubleblind
pilot trial of oral tolerization with myelin antigens in multiple sclerosis.
Science 1993; 259: 1321-4.
25. Barnett ML, Combitchi D, Trentham DE. A pilot trial of oral type II collagen in the
treatment of juvenile rheumatoid arthritis. Arthritis Rheum 1996; 39: 623-8.
26. Husby S, Jensenius JC, Svehag SE. Passage of undergraded dietary antigen into the
blood of healthy adults. Further characterization of the kinetics of uptake and the
size distribution of the antigen. Scand J Immunol 1986; 24: 447-52.
27. Zimmer KP, Poremba C, Weber P, Ciclitira PJ, Harms E. Translocation of gliadin
into HLA-DR antigen containing lysosomes in coeliac disease enterocytes. Gut
1995; 36: 703-9.
28. Kelly J, Weir DG, Feighery CF. Differential expression of HLA-D gene products in
the normal and coeliac small bowel. Tissue Antigens 1987; 31: 151-60.
29. Mayer L, Schlien R. Evidence for function of Ia molecules on gut epithelial cells in
man. J Exp Med 1987; 166: 1471-83.
30. Canchis PW, Bhan AK, Landau SB, Yang L, Balk SP, Blumberg RS. Tissue
distribution of the non-polymorphic major histocompatibility complex class I-like
molecule, CD1d. Immunology 1993 ; 80: 561-5.
31. Byrne B, Madrigal-Estebas L, McEvoy A, Carton J, Doherty DG, O'Donoghue D, et
al. Human duodenal epithelial cells constitutively express molecular components of
antigen presentation but not costimulatory molecules. (2002, submitted).
32. Mowat AM. Regulation of Immune Responses to Dietary Protein Antigens. In:
Gastrointestinal Immunology and Gluten-Sensitive Disease. Proceedings of the
Sixth International Symposium on Coeliac Disease held at Trinity College Dublin,
July 1992. Eds. Feighery C, O'Farrelly C. Publishers: Oak Tree Press, Dublin.
33. Madara JL. Regulation of the movement of solutes across tight junctions. Annu Rev
Physiol 1998; 60: 143-59.
34. Feighery C, Weir DG, Whelan A, Willoughby R, Youngprapakorn S, Lynch S, et al.
Diagnosis of gluten-sensitive enteropathy: is exclusive reliance on histology
appropriate? Eur J Gastroenterol Hepatol 1998; 10: 919-25.
35. Hermiston ML, Gordon JI. Inflammatory bowel disease and adenomas in mice
expressing a dominant negative N-cadherin. Science 1995; 270: 1203-7.
36. Baribault H, Penner J, Iozzo RV, Wilson-Heiner M. Colorectal hyperplasia and
inflammation in keratin 8-deficient FVB/N mice. Genes Dev 1994; 8: 2964-73.
37. Trier, J. S. Structure and function of intestinal M cells. Gastroenterol Clin North Am.
1991; 20: 531-47.
38. Hayday A, Theodoridis E, Ramsburg E, Shires J. Intraepithelial lymphocytes:
exploring the Third Way in immunology. Nat Immunol 2001; 2: 997-1003.
39. Viney J, MacDonald T, Spencer J. Gamma delta T cells in the gut epithelium. Gut
1990: 31: 841-4.
40. Eiras P, Leon F, Camarero C, Lombardia M, Roldan E, Bootello A, et al. Intestinal
intraepithelial lymphocytes contain a CD3 - CD7+ subset expressing natural killer
markers and a singular pattern of adhesion molecules. Scand J Immunol 2000; 52:

MECHANISMS OF ORAL TOLERANCE
73
1-6.
41. Ke Y, Kapp JA. Oral antigen inhibits priming of CD8+ CTL, CD4+ T cells and
antibody responses which activating CD8+ suppressor T cells. J Immunol 1996;
156: 916-21.
42. Banchereau J, Briére F, Caux C, Davoust J, Lebecque S, Liu Y-J, et al.
Immunobiology of dendritic cells. Annu Rev Immunol 2000; 18: 767-811.
43. Lanzavecchia A, Sallusto F. Regulation of T cell immunity by dendritic cells. Cell
2001; 106: 263-6.
44. Jonuleit H, Schmitt E, Steinbrink K, Enk AH. Dendritic cells as a tool to induce
anergic and regulatory T cells. Trends in Immunol 2001; 22: 394-400.
45. Whelan CA. (personal communication).
46. Hogger P, Sorg C. Soluble CD163 inhibits phorbol ester-induced lymphocyte
proliferation. Biochem Biophys Res Commun 2001; 288: 841-3.
47. Groux H, O'Garra A, Bigler M, Rouleau M, Antonenko S, de Vries JE, et al. A CD4+
T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature
1997; 389: 737-42.
48. Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. Interleukin-10-deficient mice
develop chronic enterocolitis. Cell 1993; 75: 263-79.
50. Neurath MF, Fuss I, Kelsall BL, Presky DH, Waegell W, Strober W. Experimental
granulomatous colitis in mice is abrogated by induction of TGF-b-mediated oral
tolerance. J Exp Med 1996; 183: 2605-16.
50. Fishman-Lobell J, Friedman A, Weiner HL. Different kinetic patterns of cytokine
gene expression in vivo in orally tolerant mice. Eur J Immunol 1994; 24: 2720-4.
51. Weijers HA, Lindquist B, Anderson CM, et al. Round table discussions. Diagnostic
criteria in celiac disease. Acta Paediatr Scand 1970: 59: 461-3.
52. Walker-Smith JA. Transient gluten intolerance. Arch Dis Child 1970; 45: 523-6.
53. Schmitz J. Is celiac disease a lifelong disorder? Clin Invest Med 1996; 19: 352-6.
54. Ansaldi N, Tavassoli K, Faussone D, Forni M, Oderda G. Clinical-histological
behavior of celiac patients after gluten load following the definitive diagnosis.
Pediatr Med Chir 1988; 10: 3-6.
55. McNicholl B, Egan-Mitchell B, Fottrell PF. Variability of gluten intolerance in
treated childhood coeliac disease. Gut 1979; 20: 126-32.
56. O'Mahony C. Immunological and Clinical Studies in Gluten Sensitive Disease.
University College Dublin. MD Thesis 1986.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 75-82
Genetically detoxified grains in coeliac disease
Federico Biagi, Antonio Di Sabatino, Jonia Campanella, Gino Roberto Corazza
Department of Gastroenterology, University of Pavia, IRCCS Policlinico San Matteo, Pavia, Italy
Coeliac disease (CD) is a chronic enteropathy triggered in genetically predisposed
individuals by gluten, the storage proteins of wheat, barley, and rye. So, from a
“coeliacologist” point of view gluten represents the coeliac offending agent. More
precisely, gluten can be defined as the rubbery mass that remains when wheat dough is
1-2
washed to remove starch granules and other soluble constituents .
Classification of gluten proteins
According to plant taxonomy, all the species (i.e. wheat, barley, and rye) toxic for
patients affected by CD are members of the Gramineae grass family. However, it
should be noted that other Gramineae grass family species, such as rice, sorghum,
maize, and probably oats, are safe for CD patients. Since wheat, barley, and rye all fall
in the tribe Hordae/Triticeae, it is possible that only species found in that tribe have
proteins active in CD. Moreover, Hordae proteins are all rich in proline and glutamine,
which is the reason why they were named prolamins. Although non toxic prolamins,
such as avenins and zeins are also proline - and glutamine-rich, they have a lower
3
proline content compared to Hordae tribe grains .
On the basis of their different solubility in a 70% ethanol solution, wheat proteins
1-2
can be divided into gliadins and glutenins . Gliadins are soluble single-chain
polypeptides. On the basis of different electrophoretic mobility, gliadins component
were classified into a-, b-, g-, and w-gliadins. Glutenins are insoluble proteins crosslinked
one to another by intrachain disulphide bonds. Components were obtained after
the reduction of the disulphide bonds, and grouped into low - (LMW) and highmolecular
weight (HMW) subunits. However, it was shown that the extraction of gluten
with aqueous solution does not lead to clear-cut fractions. Very often, glutenin subunits
can be found in the soluble gliadin fraction and gliadins are also present in the insoluble
glutenin fraction. So, it has recently emerged that the most important criteria to classify
4-5
gluten components is the primary structure . Using analysis of amino acid sequence
and molecular mass, gluten proteins have been divided into high-, medium - and low-
75

76
GENETICALLY DETOXIFIED GRAINS
molecular weight proteins. High-molecular weight proteins (x - and y-type subunits)
comprise the HMW subunits of glutenins; glutamine, glycine, and proline represent
70% of total residues. Medium-molecular weight group consists of w-gliadins; 80% of
the amino acids is due to glutamine, proline, and phenylalanine, but cysteine and
methionine, i.e. sulphur containing amino acids, are almost absent. Low-molecular
weight proteins consist of a-, b-, g-gliadins and LMW subunits of glutenins; they are
rich in sulphur containing amino acids.
Functional properties of gluten proteins
Although the only known function of these proteins is to act as storage proteins, by
releasing nitrogen for the germinating seed, their biochemistry structure has a crucial
2,6
role in providing the unique baking quality of wheat . Both LMW and HMW have not
only intrachain but also interchain disulphide bonds allowing them to crosslink other
glutenin and gliadin peptides, which results in the formation of beta-spirals. During the
baking process, CO
2
is produced by yeast fermentation and is trapped in the beta-spiral
three dimensional net. Glutenins spirals become extended but maintain the capacity to
return to the colloid state. So, elasticity and cohesivity of the dough system are due to
the cross-linked nature of the glutenin subunits, mainly the high molecular weight
glutenins. On the other hand, soluble monomeric gliadins guarantee extensibility and
viscosity to the dough (Fig. 1).
Coeliac toxicity of gliadins and glutenins
Since the demolition of these proteins to single amino acids results on loss of coeliac
C
C
N
N
N
TECHNOLOGICAL
PROPERTIES
N N C C C
HMWx HMWy LMW g a
Elasticity & Cohesivity
Extensibility & Viscosity
Fig. 1. Schematic representation of gluten protein components. Black segments unique
sequence; white segments repeating sequence; ] intrachain disulphide bond; - interchain
disulphide bond; C C-terminal sequence; N N-terminal sequence; HMWx high-molecular
weight glutenins x-type subunit; HMWy high-molecular weight glutenins y-type subunit;
LMW low-molecular weight glutenins; g-gliadins; a-gliadins.

GENETICALLY DETOXIFIED GRAINS
77
toxicity, several studies investigated which one of the several gluten components is
responsible for CD intolerance. Briefly, a-, b-, g- and w-gliadins are all toxic for CD
patients. Although glutenins were considered to be non dangerous for coeliac patients,
1,7
their safety is at least controversial . First of all alcohol extraction does not lead to
clear-cut fractions. So, gliadins have actually been found within the insoluble glutenins,
and that is an obstacle in understanding pure glutenin toxicity. Moreover, it has very
recently been shown that glutenin peptides can induce an in-vitro specific activation
and proliferation of T cell from coeliac small bowel. Such a response is virtually
identical to that seen with gliadin peptides and is nowadays considered to be an indirect
8-9
evidence of coeliac toxicity .
Although a major effort needs to be done to understand which gluten components
are toxic in CD, this knowledge would be extremely important in order to develop
genetically detoxified grains (GDG). The development of wheat strains lacking those
genes coding for toxic peptides is hopefully regarded as a new and profitable tool for the
treatment of CD. GDG could provide the solution to problems such as the poor
palatability and high costs of commercially available gluten-free foods. GDG may be
theoretically achieved with the modern techniques of genetic plant engineering which
makes the deleting or silencing of one or more genes a possible approach.
Genetics of wheat proteins
Bread wheat is a hexaploid species containing three different but related genomas
10-11
(A, B, and D), each consisting of seven chromosome pairs (Fig. 2) . Information for
storage proteins is encoded by clustered or dispersed gene families located on
chromosomes 1 and 6 of the three different genomes. In particular, genes for a-gliadins
are located on the short arm of chromosomes 6A, 6B, and 6D, tightly clustered at the
Gli-A2, Gli-B2, and Gli-D2 loci. Genes coding for gw
-and -gliadins are on the short arm
1A
1B
1D
6A
Glu-1 (HMW)
Glu-B2
Gli-A3
Gli-B3
Glu-3
(LMW) Gli-1 (gw)
Gli-2 (ab)
Gli-A5
Gli-B5
Gli-D5
6B
6D
Long Arm
Short Arm
Fig. 2. Chromosomal position of the genes coding for storage proteins of bread wheat
endosperm.

78 GENETICALLY DETOXIFIED GRAINS
of chromosome 1 at three homologous loci, named Gli-1. However, some w-gliadin
peptides have been found to be encoded by additional dispersed genes located on the
same arm of the same set of chromosomes. Genes controlling LMW glutenin subunits
are clustered at Glu-A3, B3, and D3 loci and only the genes coding for HMW glutenin
subunits occur on the long arm of the group 1 chromosomes, at a major locus named
Glu-1. So, the multigene family nature of the wheat genes and the spread of these genes
over six different chromosomes entail a very high degree of genetic polymorphism.
By applying a Southern analysis to the common cultivar Cheyenne, Anderson et al.
found that up to 150 genes code for a-gliadin, although only half of them are expressed
12-13
. An Italian study showed that there is a high degree of variability in the genomic
DNA between different cultivars. A genetic polymorphism at the Gli-2 locus encoding
14
for a-gliadin was evident among different subspecies of Triticum aestivum .
Naturally occurring detoxified grains
This huge genetic variability let people hope to individuate a “naturally occurring”
GDG. This was the case of the nullisomic 6A-tetrasomic 6D variety of Chinese Spring
wheat, in which chromosome 6A is missing but compensated for by two extradoses of
chromosome 6D. Preliminary results were encouraging. The administration for two
weeks of 65 grams of bread made with this wheat variety did not deteriorate absorptive
15
function . However, it was later proven that these variants do contain a-gliadin, coded
by B and D chromosomes, and that they are toxic, on both histological and clinical
16
ground, for patients affected by CD .
More recently, the toxicity of two lines of cultivar Raeder, one lacking the Gli-A2
encoded gliadin components and the other lacking both Gli-A2 and Gli-D1 gliadin
components was tested with an organ culture system based on coeliac duodenal biopsy
17
specimens . Although the Gli-A2 components were not compensated for by extradoses
of chromosomes B and D, those lines were deficient but not devoid of gliadins.
Although the toxicity of these two mutant lines was reduced compared to the original
cultivar Raeder, yet it was still present. In conclusion, all these studies show that a
naturally-occurring non-toxic line has not been found so far.
Technologies to develop genetically detoxified grains
Technologies allowing genetic plant transformation do exist. For example, a
fragment of foreign DNA cut with a restriction enzyme, and plasmid, cut open with the
same enzyme, are mixed together. DNA ligase is added to stitch the complementary
base sequences so that the recombinant plasmid contains foreign DNA that, under
appropriate conditions, is taken up by the Agrobacterium, a naturally occurring plant
pathogen. When host plant cells are exposed to Agrobacterium, plasmid DNA is
transferred into plant chromosomes. At the end of the process transgenic plants are
18
regenerated from single transformed cells (Fig. 3A) . An alternative method is the
biolistic “gene gun technique”.
A suspension of thousands of tiny gold or tungsten particles coated with transgenic
DNA are fired into the target tissue, using compressed helium as propellant. Particles
19
penetrate the nucleus and incorporate into the host genoma (Fig. 3B) .

•
•
GENETICALLY DETOXIFIED GRAINS
79
Foreign DNA
EcoR1
DNA
ligase
Plasmid from
Agrobacterium
EcoR1
Agrobacterium
Recombinant plasmid
containing foreign DNA
•
•
Gene
suspension
Sheath
fluid
•
Carbon
dioxide
3A
Cultured cells
from host plant
Whole plants
regenerated from
single cells
Plasmid DNA
transferred into
plant chromosomes
A) agrobacterium technique B) gene gun technique
Fig. 3. Genetic plant transformation technologies.
Additionally, it may be possible to switch off specific plant genes to eliminate or
reduce natural toxins or allergens. This can take place both at posttranscriptional and
transcriptional level. In the first case, which is the most common, injected doublestranded
RNA is cut into small fragments of 20-30 nucleotides, amplified and then
dissociated into single strands complementary, but in reverse to the target messenger
RNA which is thus inactivated. In other words, silenced genes are still active but
20
messenger RNA is degraded before it can be translated into proteins .
It should be kept in mind that these techniques have been applied to wheat only with
the aim of improving dough rheological properties by increasing the level of high
molecular weight glutenins, resulting in improved functional properties. Moreover,
transformed plants show Mendelian segregation of the transgenes and strategies have
21-24
been developed to ensure transgene stability through generations . However, these
techniques have not been directed to the treatment of CD so far.
Genetically detoxified grains for coeliac disease
Petri dish with
receiving cells
Power source
Electric
field
A project aiming at the production of a genetically detoxified wheat still suitable for
mixing and baking should be set up taking into account a few points:
(a) a minimal amount of gluten can be toxic to the coeliac small intestine;
(b) all the different classes of gliadins have been shown to be harmful for coeliac
patients;
(c) upwards 150 genes codify for the a-gliadins;
(d) identification of all possible toxic sequences is still lacking;
(e) bakery properties are mainly due to HMW glutenins.
Obviously, the first thing to be checked is whether HMW glutenins are really safe
8
for CD patients. Despite the most recent findings on T cell assays , it should be noted
that pure glutenins have not been tested on coeliac small bowel mucosa so far. So, once
pure glutenins would be obtained, their safety should be tested first in-vitro and then in-
3B

80 GENETICALLY DETOXIFIED GRAINS
I STEP
Production of HMW &
LMW glutenins without any
gliadin contamination
III STEP
Testing the rheologic
functions of pure glutenins
(in in vitro experiments) to
confirm that elasticity is mainly
ensured by them
II STEP
Testing HMW
& LMW
glutenin toxicity in in vitro
& in vivo feeding studies
IV STEP
Integrating genes
of HMW glutenins and
promoters into maize
by the biolistic method
Fig. 4. A multistep approach to obtain genetically detoxified wheat suitable for baking.
vivo on coeliac volunteers (Fig. 4). If the hypothesis of safety of HMW glutenins is
confirmed, experiments should be planned to confirm the rheologic fractions of these
pure fractions and that elasticity is mainly ensured by them. In that case viscous
function could be maintained by proteins other than gliadins. Accordingly, the final step
would be the introduction of HMW genes to improve baking quality in grains safe for
CD patients, such as maize, rice and sorghum. This aim is not an easy one. However, it
could be attempted in parallel with targeted inactivation of gliadins genes, a procedure
requiring more preliminary information and less likely to lead to the production of
acceptable baked goods.
Acknowledgements
The authors are in debt with Dr. Donald D Kasarda (US Department of Agriculture,
Agricultural Research Service, Albany, CA), Dr. Herbert Wieser (German Research
Institute of Food Chemistry, Garching, Germany), Dr. Frederik W Janssen (Insp. of
Health Protection, Zutphen, The Netherlands), Dr. Rita Redaelli (Istituto Sperimentale
per la Cerealicoltura, Bergamo, Italy) for their helpful suggestions and with
Associazione Italiana Celiachia for financial support.
References
1. Wieser H. The precipitating factor in coeliac disease. Bailliere Clin Gastroenterol
1995;9:191-207.
2. Shewry PR, Tatham AS, Kasarda DD. Cereal proteins and coeliac disease. In: MN
Marsh, ed. Coeliac disease. London: Blackwell Scientific 1992;305-48.

GENETICALLY DETOXIFIED GRAINS
81
3. Kasarda DD. Gluten and gliadin: precipitating factors in coeliac disease. In:
Coeliac disease. Mäki M, Collin P, Visakorpi JK (eds). Proceedings of the Seventh
International Symposium on Coeliac Disease. September 5-7, 1996, Tampere,
Finland. Tampere: Coeliac Disease Study Group; 1997 pp. 91-9.
4. Shewry PR, Tatham AS, Forde BG et al. The classification and nomenclature of
wheat gluten proteins: a reassessment. J Cereal Sci 1986;4:97-106.
5. Wieser H, Seimeier W, Belitz H-D. Klassifizierung der Proteinkomponenten des
Weizenklebers. Getreide Mehl Brot 1991;45:35-8.
6. Kasarda DD. Glutenin structure in relation to wheat quality. In: Pomeranz Y, ed.
Wheat is unique. St. Paul: AACC, 1989:277-302.
7. Wieser H, Belitz H-D. Coeliac active peptides from gliadin: large-scale
preparation and characterization. Zeitschrift für Lebensmittel-Untersuchung und-
Forschung 1992;193:428-32.
8. Vader W, Kooy Y, van Veelen P, de Ru A, Harris D, Benckhuijsen W. The gluten
response in children with celiac disease is directed towards multiple gliadin and
glutenin peptides. Gastroenterology 2002;122: 1729-37.
9. Martucci S, Corazza GR. Spreading and focusing of gluten epitopes in celiac
disease. Gastroenterology 2002;122:2072-5.
10. Payne PI. Genetics of wheat storage proteins and the effect of allelic variation on
bread-making quality. Ann Rev Plant Physiol 1987;38:141-53.
11. Pogna N. Genetic improvement of plant for coeliac disease. Probiotics & Prebiotics
New Foods, Rome: Università Urbaniana, 2001: 155-9.
12. Anderson OD, Litts JC, Greene FC. The a-gliadin gene family. I. Characterization
of ten new wheat a-gliadin genomic clones, evidence for limited sequence
conservation of flanking DNA, and Southern analysis of the gene family. Theor Appl
Genet 1997;95:50-8.
13. Anderson OD, Greene FC. The a-gliadin gene family. II. DNA and protein sequence
variation, subfamily structure, and origins of pseudogenes. Theor Appl Genet
1997;95:59-65.
14. D'Ovidio R, Tanzarella OA, Masci S, Lafiandra D, Porceddu E. RFLP and PCR
analyses at Gli-1, Gli-2, Glu-1 and Glu-3 loci in cultivated and wild wheats.
Hereditas 1992;116:79-85.
15. Kasarda DD, Qualset CO, Mecham DK, Goodenberger DM, Strober W. A test of
toxicity of bread made from wheat lacking a-gliadin coded for by the 6A
chromosome. In: Mc Nicholl B, Mc Carthy CF, Fottrell PF, eds. Perspective in
celiac disease. Lancaster: MPT Press, 1978:55-61.
16. Ciclitira PJ, Hunter JO, Lennox ES. Clinical testing of bread made from nullisomic
6A wheats in coeliac patients. Lancet 1980;ii:234-6.
17. Frisoni M, Corazza GR, Lafiandra D, De Ambrogio E, Filipponi C, Bonvicini F, et
al. Wheat deficient in gliadins: promising tool for treatment of coeliac disease. Gut
1995;36:375-8.
18. Jones L. Genetically modified foods. BMJ 1999;318:581-4.
19. Vasil V, Srivastava V, Castillo AM, Fromm ME, Vasil IK. Rapid production of
transgenic wheat plants by direct bombardment of cultured immature embryos. Biol
Technology 1993;11:1553-8.
20. Voinnet O, Pinto YM, Baulcombe DC. Suppression of gene silencing: a general

82 GENETICALLY DETOXIFIED GRAINS
strategy used by diverse DNA and RNA viruses of plants. Proc Natl Acad Sci
1999;96:14147-52.
21. Blechl AE, Anderson OD. Expression of a novel high-molecular-weight glutenin
subunit gene in transgenic wheat. Nat Biotechnol 1996;14:875-9.
22. Barro F, Rooke L, Békés F, Gras P, Tatham AS, Fido R, et al. Transformation of
wheat with high molecular weight subunit genes results in improved functional
properties. Nat Biotechnol 1997;15:1295-9.
23. Altpeter F, Vasil V, Srivastava V, Vasil IK. Integration and expression of the highmolecular-weight
glutenin subunit 1Ax1 gene into wheat. Nat Biotechnol
1996;14:1155-9.
24. Srivastava V, Anderson OD, Ow DW. Single-copy transgenic wheat generated
through the resolution of complex integration patterns. Proc Natl Acad Sci. USA.
1999;96:11117-21.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 83-88
Immunotherapy of coeliac disease:
where do we stand?
1 1 1 1
Carmen Gianfrani , Mauro Rossi , Giuseppe Mazzarella , Francesco Maurano ,
2 2 2 2
Virginia Salvati , Delia Zanzi , Salvatore Auricchio and Riccardo Troncone
1 2
Institute of Food Sciences and Technology, CNR, Avellino, Italy; Department of Pediatrics
and European Laboratory for Food-Induced Disease, University Federico II, Naples, Italy
Introduction
Coeliac Disease (CD) is with little doubt an immune-mediated disease, triggered by
the ingestion of wheat gliadin and related prolamines from other toxic cereal such as
barley and rye. The hallmark of CD is the enteropathy characterized by an altered
morphology of villous architecture and severe malabsorption, but it is now very clear
1
that the disease affects not only the small intestine .
An important feature of CD is its strong genetic association with HLA class II genes:
more than 90% of patients carry the HLA-DQA*0501/DQB*0201 (DQ2) alleles,
whereas the majority of DQ2-negative patients are DQ8-positive
2
(DQA*0301/DQB*0302) . Among the non HLA genes that have been suggested to
contribute to the CD susceptibility so far, only the CTLA4 gene has been found
associated to coeliac disease in several populations examined. The strong genetic
association with HLA molecules is strengthened by the finding that gliadin-derived
peptides are recognized by coeliac intestinal lymphocytes when presented by the
disease predisposing DQ2 or DQ8 molecules. In fact, from the immunological point of
view, it is now generally accepted that a recognition of gliadin peptides in association
with HLA Class II molecules by Th1 T cells secreting proinflammatory cytokine leads
3
to the overt CD lesions at the intestinal level .
CD, in the great majority of cases, is efficiently cured by the gluten-free diet;
nevertheless, for some patients the diet is unpalatable and for others the compliance is
quite poor. For these reasons a more acceptable form of therapy would be desirable:
many efforts from our laboratory are focused on the development of conceptually
amenable therapeutic strategies. In particular, considering that CD is an immune
mediated pathology, any immunomodulatory strategy that could specifically block the
adverse immune reactions triggered by gliadin, represents a valid alternative to the
exclusion diet. Moreover, because the dietetic approach has a proven efficacy and
safety, the immunomodulatory therapy must be strictly antigen-specific, not affecting
the immune response to other antigens, and devoid of risks.
83

84
IMMUNOTHERAPY OF COELIAC DISEASE
CD shares several immunological features with other less benign conditions, for
example, insulin dependent diabetes or multiple sclerosis, for which
immunomodulatory strategies have been studied and found to ameliorate the disease.
When devising immunomodulatory strategies it is of fundamental importance to have
systems where to test the success of the strategy implemented. Most of the attempts
done so far have been realized using the in vitro organ culture system, which represents
the only available tool since an animal model of CD is at moment still lacking.
The organ culture system of the treated coeliac mucosa is an experimental system to
explore the immunological events triggered by the contact of the antigen gliadin with
the jejunal mucosa. Changes suggesting an activation of T cell immunity have been
shown at the lamina propria level, including increased density of monuclear cells
expressing IL2 receptor or enhanced expression of costimulatory molecules or
adhesion molecules. The epithelium is also involved becoming infiltrated by
4-5
lymphocytes and expressing Fas molecules . As far as cytokines are concerned, a
6
strong enhancement of mRNA expression for g- interferon and IL2 has been observed .
Block of costimulation
B7/CD28 is one of major costimulatory pathways for the T cell activation providing
important signals for complete T cell activation; in fact, as in the absence of
costimulation T cells become unresponsive (anergic), the blockage of CD28/B7
7
costimulation represents a potential target for immunotherapeutic strategies . In
addition, the inhibition of costimulation has been reported to block specific antibody
response, to prolong transplants survival and to inhibit animal model of autoimmune
8
diseases .
CTLA4 (CD152) is a CD28 homologue that binds to CD80/CD86 with higher
affinity. CTLA-4 is expressed on activated T cells and down-regulates T cell activation,
9
providing a feedback negative signal . Recently, Maiuri and co-workers observed that a
fusion protein of CTLA-4 (CTLA-4-Ig) specifically blocked gliadin-induced immune
10
activation in the treated coeliac mucosa challenged with gliadin . They observed that
the addition of CTLA4-Ig in the organ culture system of treated mucosa, had a clear
effects on T cell activation in terms of decreased percentage of CD25 positive cells and
reduced expression of ICAM-1 on mononuclear cells after 24 hours of culture.
Interestingly, they found that almost 10% of lamina propria mononuclear cells showed
apoptotic phenomena. Nonetheless, two other features, such as T cell infiltration and
FAS expression by epithelial cells, were not affected. As a whole, these data suggest
that not all phenomena triggered by gliadin in the mucosa are the result of T cell
activation, and warn us that probably not everything may be modulated by affecting T
cell activity.
Gliadin peptides as tolerogens
The identification of gliadin epitopes has a crucial importance to understand the
mechanism leading to CD mucosal damage, but also to offer an important tool for a
strictly focused immunomodulatory strategy based on the modification of the structure
of antigenic peptide. In fact, it has been reported that is possible to modulate the specific

IMMUNOTHERAPY OF COELIAC DISEASE
85
11
immune response to antigenic peptide, altering the peptide structure . Peptides could
be enginereed so that they would bind HLA molecules but not TCR, or bind TCR with
the result of a switching from a proinflammatory Th1 to a Th2 or protective Th3
12
responses . Of course, the possibility of using peptides in immunotherapy and chances
of success depend on the identification of immunodominat epitopes.
Regarding CD, a number of gluten peptides have been implicated in the
pathogenesis of the disease. At least five different antigenic peptides, three of them
13-14 15 16
obtained from a-gliadin , one from g-gliadin , and one from glutenin have been
13-15 14-16
identified so far. Three peptides are DQ2-restricted , two are DQ8-restricted . It is
noteworthy to mention that the complexity of the gluten antigens, the variability of the
responses of CD patients to the identified epitopes and the role of each peptide in the CD
3
pathogenesis make the available data difficult to interpret .
Nevertheless, the possible therapeutic use of such peptides is hampered by many
unsolved problems, that are not only technical, as the poor bioavailability of synthetic
peptides and the choice of the administration route. First of all, it is still undefined the
number of such epitopes: how many they are and which are the immunodominant ones.
Moreover, among the different peptides it is also difficult to identify those which are
primarily disease eliciting and those which are created by antigen spreading.
Furthermore, it is possible that different groups of patients show different pattern of
reactivity. To this respect, differences have been hypothesized between adults and
children with CD. Again, Dq8- and DQ2-positive coeliac patients may recognize
different sets of peptides. We have recently examined in the organ culture system one of
14
the epitopes characterized in Dr Koning's laboratory as one DQ8-restricted peptides .
It was clearly recognized in vitro only by DQ8 positive coeliac patients. Biopsy
fragments from DQ8 positive, but not from DQ2 positive coeliac patients, showed a
significant increment of CD25 positive cells in the presence of such peptide, while both
groups of patients reacted to the whole mixture of gliadin peptides. As a matter of fact,
one of the problem with gliadin peptides is the posttranslational modifications they
undergo. Although many evidences have been produced on the importance of
deamidation of some peptides by tissue transglutaminase, it remains to be defined the
13,17
extent of such phenomenon . Incidentally, we found that both native and deamidated
18
forms of the peptide were equally recognised .
Finally, and this is probably the most important issue, it is still unclear if structures
reacting with T cell receptors are really the ones which trigger the disease; in other term,
the relationship between toxicity and immunogenicity is still to be clarified. As already
mentioned, in vitro and in vivo challenge studies have pointed to peptides (e.g. A-
gliadin 31-43 sequence), which instead have shown a scarce if any reactivity with
mucosal T cells.
IL10 and regulatory T cells
Among cytokines with regulatory properties, there is no doubt that IL10 is a
molecule with a strong immunomodulatory activity. IL10 is a potent immunoregulatory
cytokine that has been found to suppress T cell-mediated immune response by either
19
inhibiting the expression of costimulatory molecules on antigen presenting cells or
20
directly inducing a long term unresponsiveness of specific T cells . Regarding the

86 IMMUNOTHERAPY OF COELIAC DISEASE
hypothetical use of IL10 in the treatment of intestinal disorders, there are many
evidences both in vitro, in the fetal organ culture system, and in vivo, in a mouse model
of inflammatory colitis, that IL10 or cells producing it, such as Tr1 cells, are able to
downregulate T cell activation and to control the pathogenic process. Moreover several
reports showed that IL10 may have a central role in the development and homeostasis
of gut associated immune system (GALT). We know that in the mucosa of untreated
coeliac patients IL10 mRNA levels (and protein level) are higher than in controls, but
they face, and cannot efficiently contrast, enormously higher levels of proinflammatory
6
cytokines as gamma interferon . Also the addition of further exogenous IL10 to the
organ culture of biopsies from untreated coeliac patients fail to control the
inflammatory process (Salvati et al, unpublished observations). Nonetheless, the
addition of IL10 in the organ culture system of treated coeliac mucosa challenged with
gliadin is able to downregulate the specific mucosal immune response to gliadin in
terms of reduced densities of CD25+ cells, reduced expression of CD80/CD86
21
costimulatory molecules and of mRNA for inflammatory cytokines .
Moreover, preliminary evidences show that the production of g-interferon upon
gliadin stimulation is still absent 3 weeks later the isolation of T cells from biopsies
cultured with gliadin in presence of IL10. Interestingly, the effect is partially reverted
by the addition of anti-IL10 receptor and of anti-TGF b, this further suggesting the
presence of an active suppressive mechanism, maybe mediated by regulatory T cells.
In conclusion, our results showed that immunomodulation by IL10 on gliadin
reactive T cells could be a consequence of downregulation of costimulatory molecules
but also suggest that IL10 may operate through induction of Tr1 cells and inhibits the
migration into the intestine of pathogenic Th1 cells. Given that, the induction or the
expansion of specific T regulatory cells remains one of the possible
immunomodulatory strategies to implement in CD.
Mucosal tolerance
A very efficient route to induce tolerance to a specific antigen is the intravenous
administration, but the induction of tolerance via mucosae is a strategy more often
implemented in the experimental therapy of autoimmune diseases. Nevertheless, in
CD, being the small intestine the target of the disease, mucosae other than those of the
small intestine are candidate to elicit mucosal tolerance. Despite its peculiar
physicochemical properties, gliadin has been shown to be a good oral tolerogen.
We studied the potential for tolerance induction in gluten-free diet Balb/c mice by
intranasal administration of whole antigen, as it has been demonstrated that this
22
strategy is effective in a variety of different experimental systems . Our results showed
that intranasal administration of native gliadin before immunization abrogated the
specific T cell response. Moreover, a marked decrease in g-interferon and IL-2 levels,
and normal levels of IL-4 expression were detected, suggesting that intranasal
23
administration of gliadin can induce tolerance in Th1-cells . These data were produced
in mice parenterally immunized with gliadin: it is a model quite far from CD, but yet it
offers a series of questions, in particular concerning the ability of the nasal
administration of antigen to tolerate also gut T responses and to control a Th1 mediated
enteropathy.

IMMUNOTHERAPY OF COELIAC DISEASE
87
In conclusion, CD has a strong T component that renders it susceptible of
immunomodulation. There are still a number of significant theoretical problems to be
solved, first of all the relevance of the non-T pathogenetic component. In any case the
strategy eventually chosen must be really "competitive" with the safety of the glutenfree
diet.
Acknowledgements
Most of the studies reported in this paper have been carried out with financial
support from the commission of the European Communities, specific RTD programme
"Quality of Life and Management of Living Resources", QLK1-CT-1999-00037,
Evaluation of the prevalence of coeliac disease and its genetic components in the
European population.
References
1. Auricchio S, Troncone R, Maurano F. Coeliac disease in the year 2000. Ital J
Gastroenterol Hepatol 1999; 31: 773-80.
2. Kagnoff, MF. HLA genes in Coeliac Disease. In: Auricchio S, Greco L, Maiuri
L, Troncone R, eds. Coeliac disease. Naples: Jean Gilder Editions, 2000; 5-14.
3. Sollid LM. Molecular basis of coeliac disease. Ann Rev Immunol 2000; 18: 53-
81.
4. Maiuri L, Picarelli A, Boirivant M, Coletta S, Mazzilli MC, De Vincenzi M, et al.
Definition of the initial immunologic modification upon in vitro gliadin challenge
in the small intestine of celiac patients. Gastroenterology 1996; 110: 1368-78.
5. Maiuri L, Ciacci C, Raia V, Vacca L, Ricciardelli I, Raimondi F, et al. FAS
engagement drives apoptosis of enterocytes of coeliac patients.Gut 2001; 48:
418-24.
6. Salvati VM, Troncone R, Mazzarella G, et al. High levels of IFN-gmRNA in
untreated celaics and in treated mucosa in vitro cultured with gliadin. Ital J
Gastroenterol Hepatol 1999; 31: 559.
7. Bluestone JA. New perspective of CD28/B7-mediated T cell costimulation.
Immunity 1995; 2: 555.
8. Levisetti MG, Padrid PA, Szot GL, Mittal N, Meehan SM, Wardrip CL, et al.
Immunosuppressive effect of human CTLA4Ig in a nonhuman primate model of
allogeneic pancreatic islet transplantation. J Immunol 1997; 159: 5187.
9. Lenshow DY, Zeng J, Thistlethwaite JR, Montag A, Brady W, Gibson M, et al.
Long term survival of xenogeneic pancreatic islet grafts induced by CTLA4Ig.
Science 1992; 257: 789.
10. Maiuri L, Auricchio S, Coletta S, De Marco G, Picarelli S, Di Tola M, et al.
Blockage of T cell costimulation inhibits T-cell action in celiac disease.
Gastroenterology 1998; 115: 564-72.
11. Whindhagen A, Scholz C, Hollsberg P, Fukaura H, Sette A, DA Hafler.
Modulation of cytokine patterns of human autoreactive T cell clones by a single
aminoacid substitution of their peptide ligand. Immunity 1995; 2: 373.
12. De Magistris MT, Alexander J, Coggeshall M, Altman A, Gaeta F, Grey HM, et

88 IMMUNOTHERAPY OF COELIAC DISEASE
al. Antigen analog-major histocompatibily complexes act as antagonist of the T
cell receptor. Cell 1992; 68: 625.
13. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Vader W, Kooy Y, et al.
The intestinal T cell response to a-gliadin in adult celiac disease is focused on a
single deamidated glutamine targeted by tissue transglutaminase. J Exp Med
2000; 191: 603-12.
14. van de Wal Y, Kooy Y, van Veelen PA, Pena SA, Mearin ML, Molberg O, et al.
Small intestinal T cells of celiac disease patients recognize a natural pepsin
fragment of gliadin. Proc Natl Acad Sci U S A 1998; 95: 10050-4.
15. Sjostrom H, Lundin K, Molberg O, Korner R, McAdam S, Anthonsen D, et al.
Identification of a gliadin T-cell epitope in coeliac disease: general importance
of gliadin deamidation for intestinal T-cell recognition. Scand J Immunol 1998;
48: 111-5.
16. van de Wal Y. Kooy Y. van Veelen P.A. Vader W. August SA. Drijfhout J. Pena
S.A. Koning F. Glutenin is involved in the gluten-driven mucosal T cell response.
Eur J Immunol 1999; 29: 3133-39.
17. Molberg O, McAdam S, Lundin KEA, Kristiansen C, Arentz-Hansen H, Kett K, et
al. T cell from coeliac disease lesions recognize gliadin epitopes deamidated in
situ by endogenous tissue transglutaminase. Eur J Immunol 2001; 31: 1317-23.
18. Mazzarella G, Maglio M, Paparo F, Tardone G, Stefanile R, Greco L, et al. An
immunodominant DQ8-restricted gliadin peptide activates small intestinal
immune response in in vitro cultured treated coeliac mucosa. J Pediatr
Gastroenterol Nutr 2000; 31: S18.
19. de Waal Malefyt R, Haanen J, Spits H, Roncarlo MG, te Velde A, Figdor C, et al.
Interleukin 10 (IL-10) and viral IL-10 strongly reduce antigen-specific human T
cell proliferation by diminishing the antigen-presenting capacity of monocytes
via downregulation of class-II major histocompatibilty complex expression. J
Exp Med 1991; 174: 915-24.
20. Groux H, Bigler M, de Vries J, Roncarolo MG. Interleukin-10 induces a longterm
antigen specific anergic state in human CD4+ T cells. J Exp Med 1996;
184: 19-29.
21. Salvati V, Troncone R, Mazzarella G, Stefanile R, Borrelli M, Riccio F, et al.
Recombinant IL-10 down-regulates T-Cell activation and mRNA expression of
proinflammatory cytokines in treated celiac mucosa in vitro cultured with
gliadin. Scand J Immunol 2000; 4: 438 C25.
22. Myers LK, Seyer JM, Stuart JM, Kang AH. Suppression of murine collageninduced
arthritis by nasal administration of collagen. Immunology 1997; 90:
161-4.
23. Rossi M, Maurano F, Caputo N, Auricchio S, Sette A, Capparelli R, et al.
Intravenous or intranasal administration of gliadin is able to down-regulate the
specific immune response in mice. Scand J Immunol 1999; 50: 177-82.

Catassi C, Fasano A, Corazza GR (eds):
Primary prevention of coeliac disease. The
utopia of the new millennium? Perspectives on
Coeliac Disease, vol. 1, AIC Press, pp 89-92
The most recent advances on gluten
toxicity in coeliac disease
1 2 3
Gino Roberto Corazza , Carlo Catassi , Alessio Fasano
1
Department of Gastroenterology, University of Pavia, Italy;
2
Department of Pediatrics, University
3
of Ancona, Italy; Division of Pediatric Gastroenterology, Center for Celiac Research, University
of Maryland, Baltimore, MD, USA
After the Pavia meeting held on October 12, 2001, significant advances on the
possibilities of a primary prevention of coeliac disease have been achieved. While a part
of these have been recently published, most issues were discussed at the 10th
International Symposium on Coeliac Disease held at the Pasteur Institute in Paris, last
June.
The majority of these works deals with a further characterization of the gluten
peptides that trigger a T cell response. Although it is worth reminding that T cell
activation does not always mirror gluten toxicity in vitro or in vivo, the reportoire of
gluten peptides that are able to induce T cell activation seems now wider and more
heterogenous than reported at the Pavia meeting. At that time two different groups
found that a region of a-gliadin corresponding to aminoacids 57-75 acted as powerful,
immunodominant epitope if a glutamine residue (Q65) was deamidated by tissue
1-2
transglutaminase (tTG) to glutamic acid . Those findings disclosed new exciting
perspectives on possible alternatives for the treatment of coeliac disease, such as the
achievement of a non-toxic wheat by removal or modification of the above mentioned
antigenic sequences or specific immune therapy to induce oral tolerance.
Looking at the T cell profile in coeliac children with recent onset of disease,
3
Koning's group found that 6 novel additional gliadin and glutenin peptides were able
to induce T cell responses in the early phase of the pathogenesis of this condition. The
observation that immune activation was at some extent independent of deamidation
indicated that T cell responses can also be initiated by native gluten peptides. On the
basis of their results they estimated that gluten may contain up to 50 T-cell antigenic
peptide sequences (gliadin plus glutenin). No doubt that these results weakened the
4
expectations for new therapeutic approaches for this condition . In particular, as far as
the production of genetically detoxified grains is concerned, to the technical difficulties
listed by Biagi et al (pp 75-82 of this volume), the demonstration that glutenin peptides
are able to induce T-cell activation in coeliac disease makes the suggestion of
introducing high molecular weight glutenin genes to improve baking quality of maize
no longer tenable. On the other side, the observation that the number of peptides
89

90
RECENT ADVANCES ON GLUTEN TOXICITY
involved in the pathogenesis of coeliac disease is higher than previously thought further
reduces the technical feasibility of targeted inactivation of gliadins (and glutenins)
genes.
The most recent results of the Sollid's group are on the same line of Koning. Even in
adults, DQ2-restricted mucosal T-cells recognize an additional a-gliadin epitope and
several new g-gliadin peptides, confirming also that deamidation is not an absolute
prerequisite for T-cell activation. Interestingly the new and the previously identified
peptides are not randomly scattered accross gliadin molecules but cluster in regions of
5
the proteins with a high content of proline residues .
In apparent contrast with these recent studies indicating a wide array of a,/g-gliadin
and glutenin peptides playing a complementary role in coeliac disease activation, a
6
recent study identified a 33-mer peptide having several characteristics suggesting it is
the primary initiator of the inflammatory response in this condition. First of all, the 33-
mer peptide appeared to be stable toward digestion by gastric, pancreatic and brushborder
proteases. Secondly, it reacted with tTG with substantially greater selectivity
than known natural substrates of the enzime. Finally, it turned out to be a potent inducer
of gut-derived coeliac T-cell lines and its homologs were absent in cereal that are not
toxic for coeliac patients. Besides these important features, the aminoacid sequence of
the peptide included the residues 57-89 of a-gliadin, largely overlapping with the 57-68
1 2
and the 62-75 Arentz-Hansen and the Anderson 57-73 peptides. In addition it should
be noted that the 33-mer peptide is particularly rich in proline residues (n=13) and that
its exposure to a bacterial prolyl-endopeptidase catalyzes the breakdown of the peptide
diminishing its toxic effect. Should the 33-mer peptide be the dominant epitope in
coeliac disease, this would offer a possible strategy for oral peptidase supplement
therapy for the treatment of this condition.
The apparent discrepancy between multiple and a single toxic peptide in the
7
pathogenesis of coeliac disease has been addressed in a recent editorial . It has been
hypothesized a multi-step sequence in which, in an early phase, T-cell response is
directed to a wide array of native gluten (glutenin plus gliadin) epitopes as a
consequence of an epitope spreading due to the ubiquitous presence of the same highly
repetitive sequence. In a more advanced phase the immune response becomes polarized
to an immunodominant peptide (the 33-mer ?) as a consequence of an epitope focusing
catalized by tTG increasingly released from the tissue lesions.
It should be underlined that this view does not preclude a selective immunotherapy.
Several lines of evidence suggest that the primary problem is understanding the
hierarchy of epitope spreading in coeliac disease, since in other animal and disease
models it has been shown that immunodominant T-cell epitopes behave as potent
8
tolerogens when ingested by alternative routes and that this process may induce
tolerance to the whole antigenic molecule in a mechanism of intramolecular epitope
9
suppression . Alternative strategies may be represented by DNA vaccines based on the
10
construction of plasmids containing multiple epitopes and by the use of antigen
presenting cells engineered to present both the dominant gluten peptide and the FAS
11
ligand molecule to target and kill FAS positive antigen-specific coeliac T-cells .
Toxic peptides must reach the sub-epithelial environment to exert their pathogenic
role. At the Paris meeting new data have been presented on the interaction between

RECENT ADVANCES ON GLUTEN TOXICITY
91
gliadin and the intestinal mucosa leading to functional changes of the gut barrier
function. Zonulin represents a novel eukaryotic protein that reversibly opens intestinal
12
tight junctions (tj) . It has been recently demonstrated that zonulin expression is
13
increased during the early stage of CD , suggesting that the reported opening of tj at the
early stage of the disease could be mediated by zonulin.
Recent studies indicate that gliadin activates the zonulin-signaling pathway both in
14
normal and celiac disease-derived intestinal tissues . In normal intestinal epithelial
cells in vitro, the cellular response to gliadin was observed only a few minutes after
incubation and was characterized by a zonulin-dependent cytoskeleton reorganization
with a redistribution of actin filaments mainly in the intracellular subcortical
14
compartment . Spectrofluorimetry experiments revealed that such cytoskeleton
reorganization was associated to an increment of F-actin amount secondary to an
increased rate of intracellular actin polymerization. Experiments performed in Ussing
chambers showed that the addition of gliadin peptides to the intestinal mucosa in vitro
causes in a few minutes a significant increased of intestinal permeability mediated by
zonulin. Considering the results of this study and preliminary data generated by using
15
intestinal tissues from both celiac patients in remission and healthy controls , it is
conceivable to hypothesize a possible gliadin mechanism of action leading to zonulinmediated
increase in actin polymerization and intestinal permeability. These new
findings lead to innovative lines of research possibly relevant for the prevention and
treatment of the disease.
References
1. Arentz-Hansen H, Korner R, Molberg O, Quarsten H, Vader W, Kooy YMC, et
al. The intestinal T cell response in adult coeliac disease is focused on a single
deamidated glutamine targeted by tissue transglutaminase. J Exp Med 2000;
191: 603-12.
2. Anderson RP, Degano P, Godkin AJ, Jewell DP, Hill AVS. In vivo antigen
challenge in coeliac disease identifies a single transglutaminase-modified
peptide as the dominant A-gliadin T cell epitome. Nat Med 2000; 6: 37-42.
3. Vader W, Kooy Y, van Veelen P, de Ru A, Harris D, Benckhuijsen W, et al The
gluten response in children with recent onset celiac disease. A highly diverse
response towards multiple gliadin and glutenin peptides. Gastroenterology 2002;
122: 1729-37.
4. Kagnoff MF. Coeliac disease pathogenesis: the plot thickens. Gastroenterology
2002; 123: 39-43.
5. Arentz-Hansen H, Mcadam SN, Molberg O, Fleckenstein B, Lundin KEA,
Jorgensen TJD, et al. Celiac lesion T cells recognize epitopes that cluster in
regions of gliadins rich in proline residues. Gastroenterology 2002;123: 803-9.
6. Shan L, Molberg O, Parrot I, Hausch F, Filiz F, Gray GM, et al. Structural basis
for gluten intolerance in celiac sprue. Science 202; 297: 2275-9.
7. Martucci S, Corazza GR. Spreading and focusing of gluten epitopes in celiac
disease. Gastroenterology 2002; 122: 2072-5.
8. Hoyne GF, O'Henir, Wraith DC, Thomas WR, Lamb JR. Inhibition of T cell and
antibody response s to house dust mite allergen by inhalation of the dominant T

92 RECENT ADVANCES ON GLUTEN TOXICITY
cell epitope in naïve and sensitized mice. J Exp Med 1993; 178: 1783-8.
9. Hoyne GF, Jamicki AG, Thomas WR, Lamb JR. Characterization of the
specificity and duration of T cell tolerance to intranasally administered peptides
in mice: a role for intramolecular epitope suppression. Int Immunol 1997; 9:
1165-73.
10. Steinman L. Despite epitope spreading in the pathogenesis of autoimmune
disease, highly restricted approaches to immune therapy may still succeed (with
a edge on this bet). J Autoimmun 2000; 14: 278-82.
11. Wu B, Wu J-M, Miagkov A, Adams RN, Levitsky HI, Drachman DB. Specific
immunotherapy by genetically engineered APCs: the "guided missile" strategy. J
Immunol 2001; 166: 4773-9.
12. Wang W, Uzzau S, Goldblum SE, et al. Human zonulin, a potential modulator of
intestinal tight junctions. J Cell Sci 2000; 113: 24, 4435-44.
13. Fasano A, Not T, Wang W, et al. Zonulin, a newly discovered modulator of
intestinal permeability, and its expression in coeliac disease. Lancet 2000;
355:1518-1519.
14. Clemente MG, De Virgiliis S, Kang JS, Macatagney R, Congia M, Fasano A.
New insights on celiac disease pathogenesis: gliadin-induced zonulin release,
actin polymerization, and early increased gut permeability. Gut (in press).
15. Drago S, DiPierro M, Giambelluca D, Iacono G, Catassi C, Fasano A. Gliadin
induces occludin down-regulation and tight junctions (tj) disassembly in human
intestine. 10th International Symposium on Coeliac Disease Paris (France),
June 2-5, 2002.

Index
A
Antiendomysium antibodies (EMA), 33-37
Antigen
presenting cells, 3, 17-18, 67
specific T cells, 18, 90
-specific therapies, 17
B
Barley, 1, 23-24, 27, 31, 43, 53, 75, 83
B cell, 17-18, 62
Breast feeding, 46-48
C
Cytokines, 6, 17-19, 68
IL-10, 68, 85-86
interferon gamma (IFNγ), 6, 18, 21, 48,
68, 86
TGF-β1, 48
Th1 associated, 19
Th1-like, 17
tumour necrosis factor (TNF), 6, 18
D
Deamidation, 4-5, 14, 19, 90
Dendritic cells, 17, 67-68
Detoxified grains, 78-79, 89
E
Environmental factors
gut infections, 3
Infant feeding, 45, 52, 54, 56
role of, 36-37, 43
Epidemic of coeliac disease, 44-45, 52, 54
Epitope, 4-7, 14, 22, 24, 84, 90
dominant, 90
hierarchies, 20-21, 90
toxic, 28
Extracellular matrix degeneration, 6
F
Familial clustering, 31-32, 35, 43
G
Genetic predisposition, 36, 61
CTLA4, 2, 83-84
DQ2, 2-4, 6-7, 13, 17, 28, 32, 37, 44, 61,
83, 85
DQ7, 2-3
DQ8, 2-4, 6-7, 13, 28, 32, 44, 61, 83, 85
DR3, 2-3, 35-37
DR4, 2-3
DR5, 2-3
DR7, 2-3
5qter, 2
11qter, 2
Gliadin, 1, 4, 75-77, 91
A-gliadin, 19-23, 25, 27-28, 89
31-49, 28
57-73 QE65, 22-24, 90
57-68, 90
57-89, 90
62-75, 90
deamidation of, 19
epitopes, 84, 90
HLA-DQ2 restricted, 22, 24
fractions, 27-28
toxicity, 28
genes, 90
intranasal administration, 86
peptides, 28, 91
Gluten, 1, 3-4, 6, 17, 43-44, 48, 51, 54, 61,
64, 67, 75
age at introduction of, 49
amount of, 47-49, 54, 56
challenge, 18, 21, 25
-containing cereals
consumption of, 48
genetic modification of, 56
intake, 52
consumption of, 48
deamidation, 4
-derived peptides, 3
epitopes, 4-7, 14
free diet, 7, 17-18, 21, 61, 83
immune tolerance of, 64
in follow-on formula, 48
in the Saharawi's diet, 37
intolerance, 65
native, 90
peptides, 4, 37, 85, 89
deamidation, 5, 13-14
dominant, 90
immune tolerance to, 56
"toxic", 19
33-mer, 90
proteins, 4
high-molecular weight, 76
medium-molecular weight, 76
93

94
INDEX
low-molecular weight, 76
functional properties of, 76
specific response
in children, 13
in adults, 14
toxicity, 89
-specific memory T cells, 20
Glutenin, 1, 4, 27, 75-77, 89-90
Gut barrier, 91
H
HLA genes - see genetic predisposition
I
Immunotherapy, 83-88, 90
Infant feeding, 45, 52-54
Insulin dependent diabetes mellitus (IDDM)
risk, 44
Intestinal mucosa, 91
Intraepithelial lymphocytes (IELs), 6, 67
γδ, 6
M
Mucosal immune
response
control of, 65-69
system, 63, 66
tolerance, 69-70
N
Non-HLA genes, 32, 44
O
Oats, 27, 53
Organ culture, 6, 28, 84-85
P
Pathophysiology of coeliac disease, 6, 43,
69, 90
Aetiology, 1, 17, 43
multifactorial, 51
Prevalence of coeliac disease, 1, 33
in the general pediatric population, 1, 36
in the Saharawi, 33, 37
in the Sardinians, 37
in family members, 1, 43
Primary prevention, 43, 56, 89
R
Risk factors, 31-32
associated, 51
component causal, 51
necessary causal, 51
contributing, 50
infections, 50
socio-economic background, 50
Rye, 1, 23-24, 27, 43, 53, 75
S
Saharawi population, 33-37
Small intestinal biopsy, 33, 37
Susceptibility haplotype, 37
T
T cell, 2-4, 7, 62, 64. 90
activation, 6, 13, 17, 62, 89-90
antigen specific, 18
CD4, 17-19, 21
CD4+ TCRαβ, 2-3, 6-7
CD8, 17
clones, 5, 14
intestinal, 19, 22
peripheral blood, 19
delayed-type hypersensitivity, 61
DQ2-restricted, 90
epitope, 7, 17, 19
dominant, 24
gliadin-specific, 19
hierarchy, 21
gluten reactive, 3
helper, 67
gliadin-specific, 21
receptor, 17
regulatory, 63, 68, 85
response, 14, 89-90
suppressor, 63
Th1, 83
tolerance, 68
TGF-β, 68
Tissue transglutaminase (tTG), 4, 6-7, 13,
19, 37, 90
activity, 4-5
antibodies, 6-7
Tolerance, 61-63, 90
central, 62
immune, 56, 62
mucosal immune, 69, 86-87
oral, 5, 7, 13, 44, 46, 48, 64-65, 67-68, 70
experimental models of, 65
peripheral, 62
Treated coeliac disease, 18
Twins, 1, 31, 43
V
Vaccine, 17, 56, 65, 90

INDEX 95
W
Wheat, 7, 23, 27, 31, 53, 75, 83
and gluten, 1
flour proteins, 27
gliadin, 83
protein, 75
genetic of, 77-78
gliadins, 75, 77
glutenins, 75, 77
starch, 27
Z
Zonulin, 91