ica69 autoreactive t cells in type diabetes: molecular and ... · acknowledgements 1 would iike to...
TRANSCRIPT
ICA69 Autoreactive T cel ls in Type 1 Diabetes: Molecular Mimkry and Precipircrfion of Disease
Lakshman Gunaratnam
A thesis submitted in conformity with the requiremenü for the degree of Master of Science Graduate Department of Immunology
University of Toronto
8 Copyright by Lakshman Gunaratnam 1999
National Library m*B of Canada Bibliothèque nationale du Canada
Acquisitions and . Acquisitions et Bibliographie Services seMces bibliographiques
395 Wellington Street 395. rue Wellington Ottawa ON K1A OPJI OitawaON KlAON4 Canada canada
The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of this thesis in microfom, paper or elecîronic formats.
L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/frlm, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantiai extracts fiom it Ni la thèse ni des extraits substantiels may be printed or othenvise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.
ICA69 Auto reac t ive T c e l l s in T y p e 1 D i a b e t e s :
Molecular Mimicry and hcipitation of Disease
Lakshman Gunaratnam, Master of Science
Department of Irnrnunology , University of Toronto, 1999
Abstrac t
ICA69 is a target autoantigen in type 1 diabetes. Our laboratory previously provided indirect
evidence for antigenic mimicry between the ABBOS epitope of BSA and the Tep69 epitope of
ICA69. We generated a NOD mouse T ce11 hybridoma that recognizes both ABBOS and Tep69
and mapped the structural prerequisites of rnirnicry to the KAxyKK motif shared by both
peptides. My data provide formal proof for antigenic mimicry between these two epitopes.
To test the functional significance of Tep69-specific T ce11 pools, we used the adoptive
transfer and the cyclophosphamide models of diabetes development. Intravenous administration
of Tep69 into adoptive transfer recipients or NOD rnice 5 days pnor to cyclophosphamide
(250mg/kg) dramatically precipitated the development of diabetes. My present data imply that
Tep69-specific T cells can play a direct role in disease progression and suggests that caution
should be used in designing antigen-based irnmunotherapeutic trials in humans.
Acknowledgements 1 would iike to thank Dr. Michael Dosch for his supervision, advice, and support throughout
my graduate training. His wealth of knowledge and love of science were an inspiration to me.
Thank you for beiieving in me, for k i n g my mentor and a good &end.
1 thank the members of my thesis cornmittee, Dr. Pamela Ohashi and Dr. Philippe Poussier, for their valuable input and guidance.
1 would also like to thank al1 the members of the Dosch lab for their help, advice, and moral
support. 1 especially would like to thank Roy Cheung for generating the E12.3 hybridoma, Shawn
Wiener for assisting me with the E12.3 studies, and Denise Hammond-McKibben for her help
with some of the in vivo studies. 1 am grateful to al1 of them for making my stay in the lab
mernorable.
Last, but not least, 1 would like thank my family and my fiancée, Narrnatha, for al1 their love
and support.
Table of Contents
................................................................................................................. Abstract i ... ............................................................................................... Ac know ledgements iii
................................................................................................. Table of Contents iv ........................................................................................................ List of Tables vi . . ....................................................................................................... List of Figures VII ... ....................................................................................................... Abbreviations vi i i
............................................................. ................ Chapter 1 . Introduction ..... 1 ............................................................................................... 1.1. Overview -1
1.n. Clinicai Manifestations of IDDM ........................ ....... ..................... 2 ....................................................... I.III. Animal models for type 1 diabe tes 3
....................................................................................... The BB rat 3 ............................................................................ The NOD mouse -3
........................................................... Additional models of IDDM 4 1 . N . Etiology of type 1 diabetes ................................................................... 4
......................................................... I.IV.1. Genetic predisposi tion -5 .............................................. . . 1 N .II Role of environmental factors 6
Cow miik protein ................................................................ -7 ........................................................................ Controveny -7
............................................................. The Toronto mode1 -8 ......................................... My role in the process ... ............. 8
............................................................. I.V. Pathogenesis of type I diabetes 9 ............................................................................. Human Evidence -9
............................................ ............. Experimental Animals ..... 1 0 1.W. ICA69
.......................... . . . An autoantigen in type 1 diabetes mellitus ...... 1 .................................................................. Molecular propenies 1 1
............................................................................ Functional studies 12 ......................................................... Chapter II . Purpose of the present studies 13
................................................ II . 1 . In vitro studies -13
II.II. In vivo studies ..................................................................................... 13 ................................................................................... Chapter III . Specific Aims 16
.................................................................. Chapter IV . Materials and Methods 17 ................................................................ . IV.1 Peptide and protein antigens 17
............................................................................................. W.D. Animals 1 8 IV.III. Isolation of spleen cells for adoptive transfer ........................ 19 W.IV . Adoptive transfer of diabetes ............................................. 19
.................................................. . IV.V Induction of diabetes by CY -19 ................................................................. N.W. Peptide injections 1 9
............................................................................... N .VII . Tissue culture -19
IV.VIII. Generation of ABBOS-specific T cell hybridomas ............ 20 .................................... N.E. E12.3 hybridoma proliferation assay 21
............................................................ W.X. E 12.3 cell death assay I ............................................. N.M. NOD T cell proliferation assay 21
IV.XIII. Flow cytometric analysis .................................................... 22 .......................................................................... . . N W Molecu!ar biology -22
................... N . W . Cloning of TCR a and I3 variable region genes 22 ......................................................................................... N .XV . Statistics -24
Chapter V . Results ............................................................................................... 25 .................................................................................... V.I. In v i m studies 25
.................... V.I.I. Antigenic mimicry between ABBOS and Tep69 25 .............................. V.I.II. MHC Restriction of Mimicry Response -27
.................. V.1.m. Mimicry Response is Antigen Dose Dependent 27 ................................... V.I.N. Structural prerequisites for mimicry -28
......................... V.I.V. Immunization generates E12.3-Like T cells 2 9
........................................................................... V.I.VI. E12.3 TcR 30 ................................................................................... V.1 I. In vivo studies 1
.................... V.II.1. Modification of adoptively transferred diabetes 32 V.n.II. Calibration of adoptive diabetes transfer .............................. 32
.............. V.II.111. Tep69 precipitates adoptively transferred diabetes 33 .......................................... V.II.N. Tep69 & CY-induced diabetes -35
............................... V.II.V. Conventional immunization with Tep69 36 ....................................................................................... Chapter VI . Discussion 38
................................................ VI.1. Molecular mimicry at the Clonal Level 39
. .................................................... W.II Tep69 Peptide precipitates diabetes 42 ........................................................................... Chapter VII . Future Directions 46
.............................................................................................................. References 47
List of Tables Table 1 . Test antigens used in this study .......................................................................... 17
Table 2 . List of peptides & ALA replacement peptides .................................................... 18
List of Figures Figure 1 . TCR a and $ cDNA cloning strategy ................................................................ 23
Figure 2 . Antigen-specific growth inhibition of E12.3 cells ............................................... 25
Figure 3 . Antigen-specific activation-induced ce11 death of E12.3 cells .............................. 26
Figure 4 . Antigen-specific growth inhibition of E12.3 cells ............................................... 27
Figure 5 . Dose-dependence of mimicry response in E12.3 cells ...................................... ..28
Figure 6 . Replacement of critical amino acid residues within ABBOS and Tep69 ............. 28
Figure 7 . T ce11 responses to alanine replacement peptides in NOD mice ............. .. ......... ..30
Figure 8 . Expression of C D ~ E on NOD spIeen and E12.3 cells ........................................ 31
Figure 9 . Cumulative incidence of adoptively transfemd diabetes is dependent
on the number of diabetic spleen cells transferred .............................................................. 32
Figure 10 . Treatment strategy for peptide treatment after adoptive transfer ........................ 33
Figure 1 1 . Effect of various i.v. peptide treatments on adoptively transferred diabetes ....... 33
Figure 12 . Pre-transfer and pst-transfer i . v. Tep69 peptide injection ............ ...,. ............... 34
Figure 13 . ERect of i.v. Tep69 peptide on diabetes incidence following adoptive
transfer of 107spleen cells from diabetic females ............................................................... 35
Figure 14 . Effect of i.v. Tep69 treatment on CY-induced diabetes ...................................... 36
Figure 15 . Cumulative incidence of CY-induced diabetes in mice immunized i.p. with
Tep69 and incomplete Freund's adjuvant (FA) or OVA152 and IFA ................................ 37
Abbreviations aa: Amino acid
Ab: Antibody
ALA peptide: Alanine replacement peptide
APC: Antigen presenting ce11
BB: BioBreeding
BCECF: 2,7-bis-2-carboxyethyI-5-6c~xyfluorescein
BCS: Bovine calf serum
BSA: Bovine semm albumin
Ca: T ce11 receptor alpha chain constant region primer
Cp: T ce11 receptor beta constant region primer
CD: Cluster of differentiation
CFA: Complete Freund's adjuvant
CTL: Cytotoxic T lymphocyte
CY: Cyclophosphamide
D'IT: 1 +Di thiothrei tol
FITC: Fluoroscein isothiocyanate
GAD: Glutamic acid decarboxylase
HAT: Hypoxanthine-aminopterin-thymidine
HLA: Human leukocyte antigen
HS Medium: RPMI l6N medium supplemented with 10% equine serum, 2rnM L- glutamine, 1 ûûU/mL penicillin, 100pg/mL streptomycin, and 5OW 6-2- nercaptoethanol
HSP: Heat shock protein
ICA69: Xslet ce11 antigen 69
IDDM: Insulin-dependen t diabetes me1 litus
IFA: Incomplete Freund's adjuvant
LCMV: Lymphochonomeningitis virus
IL-2: Interleukin 2
IL-4: Interleukin 4
IL-10: Interleukin 10
IL-12: Interieukin 12
IFN-y: Interferon garnrna
IPTG: Isopropyl-B-D-Thiogalactopyranoside
MHC: Major histocompatibility complex
2-ME:&3-mercaptoettranoi
NOD: Nonobese diabetic
NOR: Nonobese normal
PBS: Phosphate buffered saline
PEG: Polyethylene glycol
recICA69: Recombinant ICA69
RT-PCR: Reverse transcription-polymeme chain reaction
TCR: T ce11 receptor
Tep69: T ce11 epitope 69
Chapter 1. Introduction Autoimmune diabetes is a chronic and eventually debilitating disease with strong genetic and
environmental roots. Autoimrnunity as an entity remains as one of the big 'black boxes',
unresolved challenges in irnmunology. While we do understand, at least in broad terms. the
acquisition of self-tolerance through thyrnic positive and negative selection, its maintenance and
breaches continue to puzzle. Some question the presence and inherent need for self-tolerance in
favor of a somewhat mystenous 'peace and war' or 'danger' concept that appeals but lacks
support by fundamental data. In this context, my studies focused on very specific elements of
diabetic autoimmunity: the possible role of one identified target self-antigen in diabetes. Whether
the data generated resemble events central to autoimmunity in diabetes is not certain, but my results
define new and definitive potentials of one autoantigen-and by extension perhaps other - diabetes
antigens in driving specific beta ce11 destruction. Knowledge such as this will be critical for
immunotherapeutic strategies that curtail autoirnmunity and interdict its progression to terminal
beta ce11 loss and diabetes.
1.1. Overview
IDDM results from the autoimmune destruction of the insulin-producing Bcells in the islets
of Langerhans (1). Insulin deficiency is the hallmark of diabetes mellitus. The term diabetes
denves from the Greek tenn for "siphon" to reflect the volurninous urine formation characteristic
of the disease (2). Mellitus is derived from the Greek word "mel" for honey and 'sweet'. In fact,
one of the carliest diagnostic tests used for diabetes mellitus was tasting of urine for sweetness.
Diabetes mellitus is a complex, polygenic disease descrïbed by a heterogeneous group of
syndromes rather than a single disease entity. Acute symptoms of insulin deficiency are
characterized by hyperglycemia, altered glucose and lipid metabolism, plyuria, thirst and weight
loss (3). There are two comrnon foms of diabetes: type 1 diabetes and type II diabetes mellitus.
Type 1 or insulin- dependent diabetes @DM), the topic of my research, accounts for about
15% of al1 diabetes cases, although as many as thirty percent of al1 diabetes patients have clear
signs of autoirnmunity, but absolute insulin deficiency develops only late. Such patients represent
a heterogeneous group and are difficult to categorize (4). The daily clinical routine tends to treat
adult onset diabetes without strict insulin requirement as Type II diabetes. IDDM is by far the
most common chronic disease afflicting children and young adults (5). Over the years after
disease onset, IDDM leads to a long list of life threatening complications that reflect widespread
microvascular disease. In the disease categones classified as loss of kidney function, blindness,
heart disease, stroke and loss of limbs, IDDM is the leading underlying cause (6) . These
complications reflect abnormal supply and demand kinetics of insulin, with chronic
h ypergl ycemia: glucose is duectl y mi togenic for rnicrovasc ular endothelium.
1.11. Clinical Manifestations of IDDM
Hyperglycernia, which is the central biochemical feature of the disease reflects impaired
glucose uptake into muscle and adipose tissue and the release of glucose into the circulation by the
liver (3). The body counteracts excess giucose in the blood by excreting it in the urine which
results in polyuria and severe dehydration. The increase in plasma osmolarity stimulates the thirst
center in the hypothalamus which produces the typicai symptoms of diabetes mellitus, polyuria,
polydypsia, and polyphagia. As hyperglycernia and glucosuria intensifies, ketonemia develops.
Patients with D D M depend on exogenous insulin for sunival (3).
Although the discovery of insulin by Banting and Best (University of Toronto) has
prevented the early dernise of patients due to rapid ketoacidosis, insulin treatment does not impede
the chronic complications of the disease such as nephropathy, retinopathy, and neuropathy and
other micro- and eventually macro-vascular disease manifestations (7). For example, diabetes is the
leading cause of blindness, end-stage rend disease, and diabetic patients are 2-4 time more likely
to develop heart disease (6). The overall life-expectancy of patients with IDDM is 36 years after
diagnosis, which usudly occurs around 8- 10 years of age (6). While the prevalence of IDDM is
about 0.5% in Caucasians, it varies drarnatically between countries and different ethnic groups (5,
8, 9) . IDDM can develop at any stage in life but is most frequent before the age of twenty (8).
Thus, many patients are faced with Iife-long chronic complications. In addition, the high cost
associated with chronic care and the nsing incidence of type 1 diabetes world-wide has placed an
enonnous econornic burden on Our societies, with one in seven health care dollars going to
diabetes and its complications (10). Furthemore, there is a worldwide, dramatic rise in diabetes
incidence. which cannot be explained in terms of genetic diabetes susceptibiiity. As a result, there
is an ever increasing need for research aimed at cunng this disorder.
1.111. Animal models for type 1 disbetes
Some of the major advances in the study of IDDM are owed to two spontaneous IDDM
animai models, the BioBreeding (BB) rat, discovered in Ottawa, and the non-obese diabetic (NOD)
mouse. Disease in these animals develops naturally but can also be induced by varioiis
interventions which can serve as specialized tools and models for aspects of the disease (1 1, 12).
T lt e B B r a t was developed in Canada in the 1970's (13). Diabetes in these animals
develops in about 50-80% of animals between 60-120 days of age (14). The metabolic and
pathologie charactenstics of diabetes in BB rats is highly similar to diabetic patients. For instance:
BB rats develop rapid ketoacidosis following hyperglycemia; there is no gender-bias with respect
to disease incidence, but infiltration of the islets occurs shortly before disease onset (14, 15). BB
rats are severely lymphopenic (16) and lack the RT6-positive T ce11 population which is important
for disease susceptibility (17). Mordes and colleagues have shown that transfusion of the RT6
positive cells from normal strains prevents diabetes in these rats (18). Although there are clear
benefits to this model, the severe lymphopenia and the rapid progress in transgenic technology in
mice have made the NOD mouse the major research tool in many laboratories (6).
T h e N O D m O u s e was originally derived from a cataract prone mouse strain in
Japan by Makino and colleagues in the 1970's (19). Diabetes develops spontaneously in NOD
mice which begins with a peri-islet ce11 infiltration by mononuclear cells at about 5-7 weeks of age
followed by more and more invasive insulitis and overt diabetes within 4-6 months (19). Quite
sirnilar to the human disease, development of insulitis is a prerequisite of IDDM, however, only a
subset of animals with insulitis progress to diabetes. Both the age of onset and the spontaneous
incidence varies from colony to colony, pexhaps in part due to genetic drift (20) but in large part
due to different environmenta1 factors such as temperature, diet, infections and stress (21). On the
other hand, the NOD mouse displays some features not common in human IDDM. For example,
NOD diabetes develops slowly, ketoacidosis develops late and female NOD mice develop diabetes
more frequently (60-80%) than males (10-40%) - intriguingly, a similar fom of IDDM is
common among Japanese patients with the disease (21). NOD mice also exhibit a more
widespread autoimmune disposition, most notable is the lymphocytic infiltration of the salivary
glands (sialitis) and thyroid. NOD lymphocytes have also been reported to be resistant to
apoptosis-inducing signals such as cyclophospharnide and gamma irradiation (22). Since
apoptosis plays a key role in maintaining homeostasis of the immune system, abnormaiities, even
if subtle, in this pathway are attractive candidates for contributing roles in the polygenic trait that
generates diabetes risk (see below).
A d d i t i o n a 1 rn O d e i s of Z D D M : Autoimmune diabetes cm be rapidly induced
by adoptive transfer of spleen cells or purified T cells from diabetic animals into irradiated non-
diabetic recipients (23) or NOD scidscid mice (24). Both CD4+ and CDS+ T cells are required to
transfer disease (1 1, 12, 25). A majority of recipients become diabetic within 4 weeks following
transfer (23).
Alternatively, diabetes can be accelerated with large doses of an alkylating agent,
cyclophosphamide (CY) (26). Usuaily a majority of animals become diabetic within 2-3 weeks
following either two injections of 200mgkg or a single injection of 350 mgkg (26, 27). CY
treatrnent, however, does not induce diabetes in normal mice or NOD-related but diabetes resistant
(NOR) rnice (28). Disease onset is associated with a shift from a TH2 to TH1 response (29,30)
and an increase in islet-infiltrating CD8+ T cells (31). Although the underlying mechanism of
action of CY in this model is not fully understood, it has been suggested that CY treatment
selectively eliminates unidentified regulatory element of the immune systern which may inhibit
effector cells that mediate 6-ce11 destruction (26,32). Holmberg and colleagues have shown that
CY induces apoptosis in lymphocytes and suggested that diabetes ensues because regulatory T
cells are more sensitive to apoptosis mediated by CY (22,33).
Despite considerable differences between NOD mouse diabetes and human IDDM, the
NOD mouse remains a representative and convenient model for human IDDM: it is perhaps an
attractive model for a human autoimmune disease in general (19, 34). Remarkably, many key
features of the human disease are reflected in NOD mice. IDDM is a polygenic disease, and as in
humans, the pnmary susceptibility allele maps to the MHC locus (35,36). Furthemore, many of
the sarne autoantigens that are targeted by diabetic autoimmunity in human IDDM are also
targeted in NOD mice, and the propensity for autoimmunity in other tissues is shared by IDDM
patients (1,6, 37). The remarkable similarity in the pathogenesis of diabetes between humans and
NOD rnice with respect to genetics, environmental factors. targeted autoantigens, and disease
pathology indicates that these rodents provide robust models to test antigen-specific immuno-
intervention strategies (1,37-42).
IJV. Etiology of type 1 diabetes
The etiology of type 1 diabetes is stilI unclear. NonetheIess, it is clear that both genetic and
environmental factors play important roles in orchestrating the disease. IDDM is a classical
exampte of polygenic diseases, where multiple environmental factors are believed to contribute to
disease progression (6).
1.IV.I. Genetic predisposition
For many years, the high rate of familial transmission of IDDM has indicated that it is a
hereditary disease (1). For instance, the risk of developing type 1 àiabetes is estimated to be
approxirnately 7% for siblings and 6% for offspring (l), with children to diabetic fathers nearly
three times as diabetes-prone than children of diabetic rnothers, likely reflective of a protective
materna1 effect rather than genetic constitution (43). Pioneering genetic studies based on
populations and multiplex families suggested a clear disease association with the major
histocompatibility complex (MHC) (8,44) - a discovery first made at McMaster University (45) . The finding that HLA identical siblings cany an IDDM risk of about 20-30% (44) compared to a
risk of 30 percent in monozygotic twins (46, 47) indicated that much of the disease risk is
conferred by the MHC. Michaelsen and Thorsby proposed that susceptibility to type 1 diabetes in
humans is associated most strongly with the HLA-DQ @Q8) and HLA-DR @R3,4) alleles (loci
homologous to the 1-A and 1-E in the mouse) (48-50). In fact, ninety percent of diabetic
Caucasians possess the DR3 and/or the DR4 allele (8). The MHC conferred risk association was
initialIy made by serotyping (45, 51) and has since k e n confirmed by direct genomic typing (50,
52). However, i t should be noted that risk alleles found in patients are also found in many normal
conuols and, thus, disease susceptibility is not caused by mutant class II MHC alleles found
exclusively in diabetics (8). In order to identify which DR or DQ molecule was associated with
diabetic autoirnmunity, Todd and colleagues performed extensive sequence cornparison of DQB
dleles found cornmonly in diabetic patients with those that were rare. They found that DQf3-chains
most commonly found in patients had either serine, alanine, or valine at position 57 on both
chromosomes while DQB alleles found rarely in diabetics had aspartic acid at this position (50,
53).
IDDM development in the NOD mouse is also linked closely to its MHC. Remarkably, the
IAB-chain in NOD rnice, the murine homolog of the DQB gene, also has serine at residue 57 rather
than aspartic acid found in most non-diabetic mouse strains (36). Because LA (equivaIent to
human DQ) is the only major class II MHC molecule expressed by NOD mice, Todd et al,.
postulated that the unique LAM7 molecule was responsible for diabetic autoimmunity in NOD
mice (8). It was no surprise that NOD mice carrying a transgene encoding an IAB-chain coding
for aspartic acid at position 57 showed abnorrnal disease incidence (54). Moreover, N3D mice
congenic for MHC from B IO (H-2b) (55) or NON (non-obese normal, H-2nbl) mice (56) do not
develop insulitis or diabetes. Sarvetnick's lab demonstrated that normal mice carrying NOD MHC
c m be pushed towards diabetogenesis by an ILI0 transgene under control of the insuiin promoter,
indicative of the paramount devance of MHC in disease susceptibility (57).
Mapping of diabetes risk to non-ASP residues is particularly interesting considering the
critical location of this residue within the class II DQ molecule. Crystal structure data from HLA-
DR1 suggest that the aspartic acid at position 57 may mediate MHC-peptide interaction (58).
Collectively, data from human IDDM and rodent models provide strong evidence for a critical role
for the class II MHC region in detennining susceptibility to IDDM, and hence contributes
si p i fican tl y to Iddl (the pnmary IDDM suscepti bility locus).
The low disease concordance among identical twins (-35%) and HLA-identical siblings
(>20%) indicates that although the MHC is a major component of disease susceptibility, it is not
the only genetic locus involved (1). Furthemore, although the NOD MHC class II (IA~~) is
required for the developrnent of diabetes (28), by itself it is not sufficient to cause diabetes (7).
When MHC of NOD is expressed on B6 or B 10 genetic background, these mice do not develop
insulitis or diabetes spontaneously (59). Indeed, several studies have confirmed that the
development of autoimmune diabetes is controlled by multiple genes (36, 56,60,61). Linkage
analysis and microsatellite mapping in IDDM families and NOD rnice have identified several
susceptibility regions which so far include al least 19 Idd loci in humans and 18 in NOD mice.
These studies are complex, as they use diabetes as readout, despite the fact that disease
developrnent has critical environmentai triggers, accelerators and decellerators (6). The strongest
evidence so fa. for a non-MHC risk gene in humans is, Idd2, which maps within 596 bp of the
insulin gene (62, 63) and contributes approximately 10% of the familial clustenng of type I
diabetes. Most susceptibility genes still remain to be identified (60).
I.IV.11. Role of environmental factors
Despite a large body of evidence supporting the genetic basis for IDDM, only a fraction of
those genetically at nsk develop diabetes. The concordance rate for IDDM between identical twins
is less than 50% (46,47) and both inbred strains of rodents, the NOD mouse and BB rat, are not
100% concordant for disease. This emphasizes a critical role for environmental factors in the
etiology of type 1 diabetes (5). Only to some extent rnight these data be explained by post-
conceptional genetic discordance (e.g. T ce11 or B ce11 repertoire generatïon).
Evidence from epiderniological studies in humans and experimentation with animal models
strongly suggest an important role for the environment as a nsk factor, trigger or course modifier
of autoimmunity in type 1 diabetes. The world-wide incidence of diabetes varies drarnatically kom
country to country and stri king1 y di fferen t diabetes incidence has been reported among geneticall y
similar populations in different countries (64). Migration studies comparing the incidence of
diabetes in Asian migrants to the UK. and Japanese migrants to Hawaii aiso support this notion
(65, 66). More convincing evidence for the importance of environmental factors in influencing
diabetes development is the increase in diabetes incidence documented over the last 40 years in
several countries which cannot be explained by changes in the gene pool (7).
The search for environmental factors that influence the development of diabetes has
identified several putative candidates including: diet, viral infections, toxins, duration of breast
feeding, temperature, climate, latitude, psycho social factors such as stress (5, 67). Among the
various infectious agents, Coxsackie B4 virus, enteroviruses in general, rubella, and recently, a
retroviral superantigen have been proposed as possible triggers of type 1 diabetes (5,68,69).
C O w rn i l k p r O t e i n One candidate environmental trigger is bovine serum albumin
(BSA), a protein found in cow's milk (70, 71). Dietary exposure to cow milk early in childhood
has been long proposed (7 1-74) as a critical diabetes risk factor (75-77). The finding that children
newly diagnosed with IDDM had elevated Ievels of anti-BSA IgG antibodies implicated BSA as a
critical molecule possibly responsible for the diabetogenic effect of cow's milk (74,78-80).
C O n t r O v e r s y This has become a widely debated and sornewhat polarized theme in the
diabetes literature, with over 30 publications in the last thirty months alone (reviewed in (8 1)). Our
laboratory continues to play a central role in this debate. Evidence in support of the BSA
hypothesis developed by Our [ab, included: 1. ) Early induction of tolerance to BSA prevents
diabetes in BB rats (71). 2. ) Diabetes-prone rodents (82) tend to have an elevated serum
concentration of anti-BSA antibodies. 3. ) Exposure of infants with diabetes risk-associated MHC
to cow milk-based formula raises the iDDM risk 13.2fold (83):. This study confirmed numerous
predecessors that found elevated disease nsk associated with early cow milk exposure. As
expected, risk levels were smaller when comparing diabetic children with disease with those from
the general population without attention to genetic risk elements (reviewed in (81)).
There are two separate issues here - the possible role of BSA in diabetes development, and
the diabetogenic effect of cow milk based formula given to geneticall y susceptible infants before
the age of 3-4 months. These issues may not be synonymous. Indeed, abnormal irnmunity to cow
milk proteins is highly prevalent in children with newly diagnosed IDDM, and BSA may only be
one element of concem in a cow milk-based weaning diet (81). A prospective, blinded, placebo-
controlled trial is under way in Europe to test the hypothesis that delay of weaning to dairy
products will reduce the incidence of IDDM (84-88). Our laboratory participates in this trial.
T h e T O r O n t O m O d e 1 Dosch et al. in Toronto proposed a molecular mimicry
hypothesis for IDDM which holds that early exposure to the mimicry antigen BSA would expand
self-reactive T cells against pancreatic 8-cells (71,79). The key finding that anti-BSA antibodies
precipitated a 69 kD protein, islet ce11 antigen 69 (ICA69), in lysates of islet cells crystallized their
hypothesis into a model (70). According to this model, expansion of ICA69 cross-reactive T cells
would engender autoreactive T ce11 pools to overcome penpheral tolerance and initiate andfor
sustain the disease. Once autoreactive T cells were generated, they could be maintained by ICA69
expressed in i3-cells, since the removal of B-cells results in the loss of diabetes-associated T cells
(89)-
Considenng the pivotal role of T cells in diabetes pathogenesis. direct evidence in support of
BSA or ICA69 specific T cells did not corne until fairly recently, as the ICA69 locus had to be
found, the gene cloned and analyzed (90-95). In 1995, Cheung and CO-workers demonstrated that
newly-diabetic children had T cells that were sensitized to BSA and they identified the target
epitope as the ABBOS peptide (aa 152-169) (96). Once we cloned ICA69 and produced the
recombinant molecule, it was demonstrated that T cells sensitized to BSA in patients were also
cross-reactive with recombinant ICA69 (93). The cross-reactive T ce11 epitope in ICA69 was
mapped to the Tep69 peptide (aa 36-47) of ICA69 (93).
M y r o l e i n t h e p r O c e s s In favor of a disease-associated role of dietary BSA,
we have shown that NOD mice reared on diet free of intact BSA (hydrolyzed casein diet :
Nutramigen) are protected from diabetes (97). Nutramigen is the test diet in our European diabetes
prevention trial (98). A search for an underlying mechanism revealed that mice protected from
diabetes failed to generate ICA69 autoreactive T ce11 pools, thus maintaining tolerance to ICA69.
However, when such mice were first irnmunized with BSA, ICA69 reactive T cells werc readily
recruited. This suggested that ICA69 autoreactive T cells are part of the normal NOD mouse T ce11
repertoire but were non-responsive unless they were previously recruited and selected by BSA.
However, these are complex issues not fully understood (99). For example, immunization with
ABBOS protected from IDDM and generated ABBOS-specific T ce11 repertoires that did not
cross-react with Tep69, the major homologous T ce11 epitope in ICA69 (42). To better understand
this cross reactivity and modes of disease modification by these peptides was central to my project
(see below).
Spleen cells from diet-protected mice failed to transfer diabetes when grafled alone and when
CO-transferred with diabetic spleen cells from diabetic mice they inhibited adoptive diabetes
transfer. This suggested that a regulatory population of T cells in Nutrarnigen-fed mice mediated
protection from disease. These data are consistent with Our BSA mimicry model, where T cells
encountering lower affinity ABBOS peptide can escape the anergy inducing effects of high-
affinity self-peptide Tep69 (93, 100). However, addition of oral BSA to Nutramigen-fed rodents
does not break the protective effect of the diet, but in a recent BB rat study such animals
nevertheless developed high levels of anti-BSA antibodies (101). We are collaborating with this
group and our prelirninary data indicate that these apparently BSA-specific antibodies may indeed
be anti-ICA69 antibodies (unpublished observations).
I.V. Pathogenesis of type 1 diabetes
In Our laboratory we consider diabetes and its scientific approaches concurrently for human
and rodent disease. Sirnilarities in observations strengthen their impact. One striking example is
the identification of an identical T ceIl epitope sequence in ICA69 targeted by diabetic
autoimmunity in humans and mice (42),. This situation is not common. For exarnple, the
Coxsackie B4 epitope with a homologous sequence in GAD65 is not a natural epitope targeted by
human autoreactive T cells (102, 103)
H u m a n E v i d e n c e The finding that diabetes was induced following allogeneic bone
marrow transplantation from a diabetic donor to a non-diabetic individual (104, 105) and that
treatment with immunosuppressive agents, notably cyclosporin, deIays disease course in both
rodents (106), and humans (107, 108) characterizes type 1 diabetes as an autoimmune disease (1).
Progressive IDDM in patients is characterized by an infiammatory mononuclear ce11
infiltration of the pancreatic islet cells (insulitis ), which is followed by 6-ce11 destruction and
hyperglycernia (109). Microscopic examination of the insulitis lesion in pancreas biopsies from
recently diagnosed diabetic patients reveal that it consists pnmarily of T cells, CD4 and CD8 T
cells (mosrly CD8 T cells), and some B cells and macrophages (1 10, 11 1). hterestingly, few islets
devoid of B-cells, containing only a and delta cells, remain free of insulitis suggesting that the
autoimmunity is 0-cell specific. However, extra-islet autoimmunity is commonly observed in
diabetic patients ( 1 12- 114). By 2-4 months after diagnosis, almost 90% of a patient's B-cell mass
is lost ( I 11). For many years, 0-cell destruction was thought to occur gradually in a linear fashion
over a long period, but now it is believed to take place late in the disease process (1 15).
Onset of type I diabetes in humans is frequently associated with autoantibodies to a wide
array of cytosolic constituents of B-cells including insulin (1 16), proinsulin (1 17), glutarnic acid
decarboxylase (GAD) (1 18), I-A2 (1 19) and ICA69 (120, 121). As immune response markers,
autoantibodies have been proven to be quite useful in accurately predicting disease risk in relatives
of diabetic patients since only about 5% of first-degree relatives of patients with type I diabetes
also develop diabetes (122, 123). Nonetheless, no conclusive experimental evidence points to a
pathogenic role for these autoantibodies (I) , although B cells are critical part of the disease
process, most likely because of peculiar antigen presenting functions ( 124).
In humans a large body of necessarily indirect evidence suggests that IDDM is a T cell
mediated disease (1). This was first suspected because a majority of mononuclear cells in islet-
lesion were T cells (1 10, 125), and because diabetic patients had increased frequency of activated T
cells (as indicated by IL-2 receptor up regulation) in their circulation and islet lesions (126-128).
The most persuasive argument for a critical role of T cells in autoimmune diabetes is obviously the
critical role played by disease-associated class XI MHC (see above). In addition, several groups
showed that peripheral blood T cells in patients with pre-clinical and clinical IDDM were
sensitized to human islet antigens (94, 129-133). More recently, several groups have demonstrated
T cell reactivity to GAD (134), heat shock protein 65 (hsp 65) (135), IA-2 (136, 137) and ICA69
(93). T cell reactivity to B-cell antigens in tissue extracts is well documented (1, 136, 138).
Interestingly, T cell autoreactivities are not species restricted, with NOD mouse T cells recognizing
human islet antigens and human patient T cells recognizing rodent islets (136, 139). Collectively,
these observations suggest, but in no way prove a disease prerequisite role of T cells in diabetic
autoirnrnuni ty .
E x p e r i rn e n ta I A n i m a I s Studies of the BB rat and NOD mouse have more
mechanistically demonstrated key roles for T cells in the pathogenesis of type I diabetes (1 1, 12,
140, 141). As in patients, the majority of cells in insulitic lesions of NOD mice are T cells (142,
143). Definitive evidence for a critical role of T cells has come from the following studies:
prevention or suppression of diabetes in NOD mice treated with reagents that block T cell function
(144-147); mice congenitally athyrnic or neonatally thymectomized fail to develop disease (148-
150); diabetes can be transferred to non-diabetic NOD mice (NOD-SCID or in-adiated NOD) and
BB rats by adoptive transfer of T cells or islet reactive T ce11 clones from diabetic animals (25,
15 1-156). Thus far, consensus is solid in reports of animal data which show that both CD4+ and
CD8+ T cells are necessary for diabetes development (1 1, 12, 24, 25, 143, 145, 151, 157-159).
However, rnuch remains to be leamed about the individual contributions of each class of T cell to
the disease pathogenesis.
Considering the central role of T cells in IDDM, T cell sensitization to B-ce11 autoantigens is
considered a key event in the initiation and/or progression of disease (1,6, 89). Thus far, several
candidate autoantigens have been identified either by detection with autoantibodies andlor
autoreactive T cells (6). None of these antigens has unequivocally been shown to be the triggering
antigen(s), but some can definitely affect disease development in experimental settings:
GAD6Y67, HSP65, Prdinsulin and ICA69 (40-42, 160-164).
1.W. ICA69: An autoantigen in type 1 diabetes meMitus
Research in Our laboratory is focused on understanding the pathoetiological role of ICA69
in IDDM. Several lines of evidence suggest that ICA69 is a target autoantigen in type 1 diabetes:
(1) detection of ICA69 autoantibodies as well as autoreactive T cells to ICA69 in diabetic patients
(93,96, 120); (2) presence of ICA69 autoantibodies in newly diagnosed diabetic and pre-diabetic
patients (165-169). More recentiy, a Dutch group concluded that over 80% of patients with recent-
onset diabetes had either ICA69 autoreactive T cells or autoantibodies to the molecule (170). Thus,
ICA69 is one of the most frequently targeted autoantigens in IDDM.
M O 1 e c u 1 a r p r O p e r t i e s In Our laboratory, ICA69 was identified and
subsequently cloned based on its cross-reactivity to a cow rnilk protein, BSA (94). Eisenbarth's
group independently cloned the gene through screening of a pancreatic cDNA library with patient
sera (120). In retrospect, the molecule was first detected as a 69 kDa band of unknown identity
precipitated by anti-BSA antibodies in endogenously radiolabelled islet ce11 extracts (171).
In terestingly , this group has become one of the most vocal advocates against a role of BSAIICA69
immunity in diabetes (172).
The protein was fulIy cloned and characterized extensively in our laboratory. Human, mouse,
and rat ICA69 sequences have been characterized: the human gene has been mapped to
chromosome 7p22 (95) and the munne locus to chromosome 6 (120), aithough the initial
chromosomal fine mapping and synteny provided by the Boston group turned out to be in error
(91). The amino acid sequence of ICA69 is highly conserved between human and mouse and
shares 85% homology overall, with large regions of identity and three clusten of diversification
(91).
F u n c t i O n a L s t u d i e s Tissue expression foiiows a neuroendocrine pattern with high
expression in 8-cells, brain. and testes (92). The biological function of ICA69 is still unknown and
is being addressed in our laboratory through knock-out in mice and C. elegans, the latter carrying
a highl y conserved homologous gene: in transgenic C. elegans carrying GFP under the ICA69-
hornolog promoter, every single neuron are painted green (Pillon, M. et al., unpublished
observation). Tep69 knock out (KO) (targetted dismption of the Tep69 region of ICA69 gene)
mice are viable and k ing bred ont0 NOD background using a speed congenic approach (124),
development of full ICA69- KO animals is at the stage of biastocyst aggregation as this thesis is
witten.
Collectively, our data suggest that there is immunological cross-reactivity between ABBOS
and Tep69 (both naturally generated T ce11 epitopes in patients and NOD rnice (42, 93)),
sequences which are fully conserved between rodents and humans. Consequentiy, Our laboratory
proposed a molecuiar rnimicry model for IDDM, where exposure to dietary BSA would recruit T-
cells cross-reactive with ICA69 and thus generate and sustain autoimrnunity significant levels: it
was Ohashi et al. who first proposed that p l sizes of autoreactive T cells, rather than presence or
absence, determine the autoimmune potential (173).
Since IDDM is a T ce11 mediated disease, evidence for T ce11 sensitization to BSA and
ICA69 in recent-onset diabetic patients added credibility to the mimicry model (93,96), the most
recent data corne from a large, collaborative patient study that was only recently unblinded and is
being analyzed now (166-169, 174). The specific target epitopes consistently mapped to the
ABBOS peptide (aa 152-169 of BSA) and the Tep69 peptide (aa 43-47 of ICA69). More recently,
antigenic-rnirnicry between these homologous T-ce11 epitopes was directly confirmed in vivo in
NOD mice. NOD rnice when immunized with BSA (in CFA) can readily generate T ceil responses
to ICA69 (42). However, much remains to be leamed about a specific role for ICA69 in the
pathogenesis of IDDM and about its potential as a immunotherapeutic agent. In this, our approach
to study human and rodent IDDM with specific, focused questions has been of mutual benefit.
Chapter II. Purpose of the present studies There is a great deai of indirect evidence for a significant role for ICA69 in IDDM. Direct
evidence, however, has k e n limited to a single study demonstrating diabetes prevention foilowing
neonatal tolerance induction to Tep69 in NOD mice. Furthermore, evidence for molecular rnimicry
between BSNABBOS and ICA69lTep69 has thus far been indirect. In order to provide formal
proof for molecular mirnicry we have here generated and characterized a T ce11 hybridoma which
demonstrates the proposed mirnicry, and through studies of fine specificity we provide evidence
that such T cells are common in NOD rnice. We have also utilized two independent animal models
which mimic late stage pre-diabetes to decipher possible roles of ICA69 in progressive diabetic
autoirnmunity in vivo.
11.1. In vitro studies
Previously we proposed a molecular mimicry model for the pathoetiology of type 1 diabetes.
According to our model, exposure to the dietary antigen, BSAIABBOS, would expand ultimately
self reactive T cells against the B-ce11 antigen, ICA69nep69 (6,70). In agreement with Our model,
we have demonstrated that T ce11 antigenic mirnicry between the dietary antigen BSA and ICA69
exists in newly diabetic patients (93) and NOD rnice (42). In NOD mice, immunization with BSA
or ABBOS recruits T cells reactive to recombinant ICA69 (recICA69) or Tep69, and vice versa
(42). Sirnilarly, cross-tolerance c m also be induced between BSA and ICA69 (42). These data,
however, do not rule out the possibility that multiple or overlapping T ce11 populations may be
responsible for the apparent mimicry response. To provide conclusive evidence for the existence of
single T ce11 clones, in NOD mice, which can recognize both rnirnicry epitopes, we generated T ce11
hybridomas frorn NOD mice immunized with the ABBOS peptide. An alanine mapping (ALA
mapping) strategy was also undertaken to identiQ critical amino acid within the rnimicry peptides
which mny be involved in MHC andor TCR interaction. Finally, to c o n f m the clonal origin of the
hybridoma and to identify the TCR variable genes involved, 1 am in the process of cloning the
TCR with a view to generate a TCR transgenic NOD mouse.
1I.II. In vivo studies
ICA69 is a very common target of diabetic autoirnrnunity in patients (121). Furthermore,
neonatal tolerization of NOD mice with the main, identified T ce11 epitope of ICA69 (Tep69) leads
to a dramatic reduction in the incidence of diabetes. These findings suggest some role for ICA69
as a target autoantigen in IDDM. The primary objective of the present studies is to determine the
role of Tep69 specific T cells in the induction and/or progression of diabetes in the NOD mouse
model of IDDM. Another major goal in Our lab is to test the immunotherapeutic potential of
ABBOS or Tep69. Autoimmune Ne11 destruction in IDDM is viewed as a failure to promote or
maintain self-tolerance (6, 99). Since it was previously shown that NOD mice exhibit a loss of
tolerance to ICA69 (42) we argued that the induction of specific tolerance to ICA69 (Tep69) could
therefore represent a means of preventing disease (175). Prevention of disease is contingent on a
critical role for ICA69 reactive T cells in the pathogenesis of IDDM. Several therapeutic strategies
based on the administration of diabetes-associated autoantigens have successfully been used to
prevent IDDM in NOD mice. For example, intrathymic (39) or intravenous injection of glutamic
acid decarboxylase (40) or irnrnunization with heat shock protein 60 (176) provide significant
disease protection (177). Similarly, imrnunization with GAD (178, 179) or insulin B chain (180)
can abolish or delay disease development in this animal model. There is a distinct need for models
to explain the effects of single agent irnmunotherapy for which we coined the term 'antigenic
synergy' (42).
In the present study, we used the adoptive transfer and CY models for studying the role of
ICA69 and used soluble Tep69 and ABBOS peptides to manipulate disease course in vivo. These
models are probably close to pre-IDDM in humans. The two models provide rapid and highly
synchronized diabetes development with reduced roles for environmental factors that may interfere
with disease development (42). Furthermore, results from o u lab demonstrate that the adoptively
transferred disease model recapinilates core aspects of IDDM development at an accelerated Pace
(42). Autoreactive T-cells are often difficult to detect in spontaneously diabetic mice, but become
readily detectable following adoptive disease transfer (42). However, on the whole, we encountered
considerable difficulty in detecting relevant T ce11 repertoires consistently, and, in the CY model, we
failed in this respect because of the dramatic resistance of NOD T cells to apoptosis. In culture,
CY treated spleen cells display elevated background proliferation which interferes with Our assay
system.
Given that Tep69-specific T ce11 are spontaneously generated during the course of adoptive
disease transfer, if T cell sensitization to an autoantigen plays an important role in the initiation
andor progression of type 1 diabetes, manipulation of the corresponding reactive T cells should
modify the incidence andor progression of diabetes (1). The ability to alter the outcome of disease
by either prevention or precipitation could give important insights about the function of the
autoantigen in diabetic autoimmunity. The early finding that neonatal tolenzation with Tep69
prevents disease was encouraging in this respect. Consequently, we ernployed a similar tolerance
induction strategy and tested the ability of soluble Tep69 and ABBOS peptide to modify the
course of IDDM in the adoptive transfer as well as the CY models of diabetes development.
Chapter III. Specific Aims: 1. Demonstrate that a single T cell clone from NOD mice can recognize both
BSA/ABBOS and ICA69mep69
2. Identify the structurai prerequisites of T ce11 mimicry between ABBOS and Tep69
3. Establish adoptive transfer and CY systern for rapid IDDM development
4. Modify disease course in both systems with antigen-specific strategies
5. Study mechanisms of antigen-specific disease modification
Chapter IV. Materials and Methods
1V.I. Peptide and protein antigens:
Tep69, ABBOS, BSA193 and OVA152 peptides were synthesized a it the Alberta Peptide
Institute at the University of Alberta in Edmonton, Alberta Al1 peptides were pdf ied by high-
performance liquid chromatography (HPLC) (purity 299%) and c o n f m e d by sequencing. The
peptides used are shown in table 1 (numeric codes denote NHZ-terminal amino acid position):
-- --
Table 1: Test antieens used in this study
peptide 524 lsynthetic 1>99% pure
j3-isoform IE. coli 1 truncated C-terminus t 1
Sigma Chem. Inc
BSA-148 'ABBOS"
synthetic 1~99% pure
AFIKATGKKEDE purified Fraction V
synthetic ~ 9 9 % pure
FKADEKKFWGKYL synthetic 1 homolog ICA69-350
EDKGACLLPKIE I
Sigma Chem. Inc lpurified 1 Fraction V 1 r
OVA152 lsynthetic 1299% pure
As a source of recombinant ICA69 protein, the human ICA69-B isoform was
used (92). This naturally occurring protein has 97% sequence identity with the
murine molecule (91). Of the three amino acids that differ between human and
mouse sequence, two amino acid changes are conservative, and the Tep69 region is
identicd. BSA (fraction V) and Ovalburnin (OVA fraction V) were purchased from
Sigma (St . Louis, MO). Alanine replacement peptides were s ynthesized and puri fied
by Dr. J. Gorga (Pittsburgh, PA) within the terms of a collaboration (see table 2,
below).
Table 2. List of peptides & ALA replacement peptides BSA 147-1 64 [ABBOS) BSA 147-1 64,K155A BSA 147-164,K1600 8SA 147-164,C147A BSA 147-1 64,0148A BSA 147-164,€149A BSA 147-1 64,F150A BSA 147-1 64,KlSl A BSA 147-164,0153A BSA 147-1 64,El54A BSA 147-164,KlSSA BSA 147-164,K156A BSA 147-1 64,F157A BSA 147-1 64,W158A BSA 147-1 64,Gl59A BSA 147-164,K160A BSA 147-1 64,Yl6lA BSA 147-1 64, L I 62A BSA 147-1 64,Y 163A BSA 147-164,E164A ICA69 33-50 (Te~69) ICA69 33-50,K39A ICA69 33-50, K43A ICA69 33-50, K44A ICA69 33-50,H48A ICA69 33-50,H48D
DEFKADEKKFWGKYLYE CDEFKADEAKFWGKYLYE CDEFKADEKKFWGQYLYE ADEFKADEKKFWGKYLYE - CAEFKADEKKFWGKYLYE CDAFKADEKKFWGKYLYEF CDEAKADEKKFWGKYLYE CDEFAADEKKFWGKYLYE COEFKAAEKKFWGKYLYE CDEFKADAKKFWGKYLYE CDEFKADEAKFWGKYLYE CDEFKADEKAFWGKYLYE CDEFKADEKKAWGKYLYE CDEFKADEKKFAGKYLYE CDEFKADEKKFWAKYLYE CDEFKADEKKFWGAYLYE CDEFKADEKKFWGKALYE CDEFKADEKKFWGKYAY E CDEFKADEKKFWGKYLAE CDEFKADEKKFWGKYLYA
TKQAFIKATGKKEDEHVV TKQAFIAATGKKEDEHVV TKQAFIKATGAKEDEHVV TKQAFIKATGKAEDEHVV TKQAFIKATGKKEDEAVV TKQAFIKATGKKEDEQVV
IV.11. Animals:
NODLt mice (Jackson Laboratones, Bar Harbor, ME) were purchased andor bred
in our rodent facility. NOD males. 6 to 8 weeks old. were used as recipients of adoptive transfer
spleen cell grafts from diabetic females or for CY experiments. Diagnosis of diabetes required
glucosuna (TesTape, Lilly, Toronto) and persistent (non-fasting) hyperglycemia detected by
electronic monitor (SureStepm, Life technologies Inc., Bumaby, BC). Positive animals with
blood-glucose levels 813.8 mmoVL on two consecutive days were considered diabetic. Al1 animals
were kept in the same housing unit in our facility under approved animal-care protocols.
IV-III. Isolation of spleen cens for adoptive transfer
Recently diabetic donor mice were killed by cervical dislocation, and spleen single-ce11
suspensions were made from freshly isolated spleens using a ce11 stomacher (Seward, Cincinnati,
OH). Cells were pelleted at 1200 RPM for 10 min. and red blood cells were lysed with Tns-
ammonium chloride for 3-5 min. Cells were irnmediately washed twice in PBS, counted (Coulter
Counter ZM, Hialeah, FL) and used for adoptive transfer within 2 hours. Prior to counting and
transferring, spleen ce11 suspensions were passaged through 7 0 p nylon ce11 strainer to remove
ce11 clurnps (Falcon, Franklin Lakes, NI). More than 90% of transferred cells were viable as
determined by dye exclusion.
IV.IV. Adoptive transfer of diabetes
Randomized recipients were whole-body irradiated (650 rad) 1 day pnor to grafting.
Splenocytes were pooled fiom groups of 4-6 recently diabetic NODLt females and 5-20x10~ (as
required) splenocytes were injected intravenously (i.v.) into inadiated recipients. NOD males (6-8
weeks old) were used as recipients of adoptive transfers of diabetic spleen ce11 grafts. In some
experiments. diabetic spleen ce11 donors were treated with Tep69 or PBS i.v. 1 day before transfer.
Glucosuria was screened 2-4 times a week and at the end of the expriment, followed by
measurements of blood glucose when indicated.
1V.V. Induction of diabetes by CY
Based on pilot experiments, a single injection of CY (Sigma, St. Louis, MO) was
administered i.p. to NOD male mice (6-8 weeks old) at a dose of 250 mgkg (30mg/mL in saline).
Animals were monitored 2-4 times a week for diabetes after CY injection, followed by
measurements of blood glucose when indicated.
IV.VI. Peptide injections
In transfer experiments, newly diabetic, female spleen ce11 donors received a single i.v.
injection (100 or 400 pg, see below) of Tep69, ABBOS, GAD65, peptide 524, table 1, or control
BSA-193 peptide (0.5 mg/mL) (42) dissolved in PBS 1 day pnor to removal of spleens for
adoptive transfer. Peptide treatrnent (100pg or 400pg, Lv.) of spleen ce11 recipients was done 1 day
after adoptive transfer. PBS served as vehicle control as indicated. In CY experirnents, 100 pg of
Tep69 or control OVA152 peptide (0.5 mg/mL,Table 1) was injected i.v. 5 days before or 5 days
after CY injection.
IV-VIL Tissue culture
Al1 assays with E12.3 cells (see below) were carried out in HS medium (RPMI-1640
medium supplemented with 2mM L-glutamine, lûûU/mL penicillin, 100pgIrnL streptomycin,
50pM O-2-mercaptoethanol and 10% equine sem (Hyclone, Logan, UT). Equine albumin does
not contain sequences homologous to Tep69 while bovine semm contains large amounts of bovine
serum albumin and thus ABBOS sequence. Humidified incubaton were set at 37°C and 5% COZ
RPMJ 1640 supplemented with L-glutamine, &ME, and 10% bovine serum was used to establish
hybndomas.
IV.VILI. Generation of ABBOS-speeific T ceil hybridomas
T ce11 h ybridomas were establis hed by pol yeth ylene gl ycol-induced fusion of an tigen-
specific spleen ceH blasts with the BW5147 (ATCC) thymoma line which does not express TCR
a or B chains on its surface (181, 182). Spleen cells were obtained from NOD mice imunized
with ABBOS peptide emulsified in complete Freund's adjuvant (CFA) 10 days prior to fusion.
Spleen cells were stimulated in vitro for 3 days with 50Cig/rnl of ABBOS. Blasts were cultured for
an additional 3 days in the presence of 1OUIrnl recombinant IL-2. Stimulated cells were mixed
with B W5 147 cells at a 10: 1 ratio. BW5147 is hypoxanthine-guanine phosphoribosyl transferase
(HGPRT) negative. The cells were washed and resuspended in 50% (vlv) polyethylene glycol
according to standard protocols (183). The fused ce11 mixture was washed and cultured ovemight
in bulk in RPMI-1640 medium supplemented with 10% calf serum, 50pM ZME, and 2mM L-
glutamine. The following day, cells were plated into 96-well plates and incubated at 37°C over
night in hypoxanthine-aminopterin-ttiymidine supplemented medium (HAT, 5rnM hypoxanthine,
3Oj.M aminopterin, 0.8m.M thymidine, Sigma, St. Louis, MO); fresh HAT solution was re-added
every second day. After 6 days, wells with growing clones were transferred to 24-well plates for
expansion. Expanded hybridomas were then transferred to culture flasks and passaged in HS
medium. Hybridomas of interest were sub-cloned twice by limiting dilution at 0.33 cells/well.
While al1 subclones showed reactivity to ABBOS and Tep69 peptides, clone El 2.3 was used in the
expenments described here.
IV.IX. E12.3 hybridoma proliferation assay
Replicate cultures of 2x104 E12.3 hybndoma cells were plated in the presence or absence of
synthetic peptide or protein antigens with either 4x1@ irradiated (5000 rad, cesium source)
syngeneic NOD mouse spleen cells, 2x104 irradiated C3G7 B-lyrnphoma cells transfected with
NOD H-2g7 6-chah (A generous gift from Dr. Unanue), or no APC in % well flat bottom micro-
well plates for 2 to 4 days (184). The concentration of synthetic peptides or protein antigens was
0.5- 1 .O pg/well unless indicated othenvise. Each well was pulsed with 1pCi of 3H-Thymidine (6.7
Ci/mmol) 6-18 hours before harvesting, and DNA incorporation was measured in a liquid
scintillation counter. Al1 experiments were repeated three times on independent days.
1V.X. E12.3 ce11 death assay
Ce11 death of stimulated E12.3 cells was measured by particle concentrated fluorescence
irnmunoassay as described previously (185). Bnefly, replicate cultures of lx104 E12.3 hybridoma
cells were plated in the presence or absence of synthetic peptide or protein antigens and with or
without l x 104 irradiated C3G7 B lymphoma cells or 4x10~ irradiated NOD spleen cells (5000
rad, cesium source) as relevant antigen presenting cells into 96 well flat bottom micro well plates
for 2-4 days . The concentration of the synthetic peptide or protein was 0.5-1.0 pglwell unless
indicated otherwise. After incubation, 1OpL of 8pg/mL 2,7-bis-2-carboxyethyl-5-6-
carboxyfluorescein @ C E 0 dye (Molecular Probes, Eugene, OR) was added directly to the cells
and incubated again for 1 hour (37°C and 5% CO2). The relative fluorescence u n i s of the
BCECF dye uptake by viable cells was then measured by Pandex Screen Machine (Mundelein,
ME). A standard curve was used to determine the incorporation of BCECF in RFUkeIl as
described (1 86- 188).
IV-XI. NOD T ce11 proliferation assay ;
NOD rnice 4-6 weeks of age received a single injection (50pg s.c.) of BSA in complete
Freund's adjuvant (CFA) and in vitro recall responses of pooled regional lymph node (LN) cells
were measured 9 days later. Replicate culture of 2 x 1 6 LN cells were incubated for 5 days in
serum-free HGM-af medium (PromoCelL, Heidelberg, Germany) in the presence of 0.5 pg of
ABBOS, Tep69, or alanine replacement peptides. 3~thymidine (6.7 Ci/mmol, IpCieIwell) was
added ovemight, pnor to harvesting and scintillation counting.
IV.Xm. Flow cytometric analysis
E12.3 cells or NOD spleen cells (lxl06/assay) were washed in staining buffer (PBS with
0.1% BSA) and then incubated with a pre-tested arnount (10pg) of 2.462 (rat anti-mouse
CD16/CD32 (Fcy IlIl i l receptor) antibody (Ehrmingen, Mississauga, ON) for 15 min. on ice.
Cells were then incubated with 2pg of fluorescein isothiocyanate- W C ) conjugated hamster anti-
mouse CD3& (Pharmingen) of IgGl isotype or 2pg FITC-conjugated isotype control
(Phanningen) antibody for 30 min. on ice. After two washes with staining buffer, cells were
analyzed with a FACSCAN andyzer (Becton Dickinson, San Jose, CA).
IV.XI1. Molecular biology
Total RNA was isolated from E12.3 hybridoma cells using ID Pure RNA II Plus Kit (ID
Labs, London, ON). After counting, 5 x 106 cells were re-pelleted and lysed in 500& of Lysis
Buffer. After lysis, 20 pL of 'Adsorbin beads' were added to the suspension and vortexed. After
incubation for 5 min. on ice, the rnixhm was centrifuged spun and supernatant was transferred to a
new 1.5 mL microcentrifuge tube. Next, 500p.L of Tris-buffered phenol (Gibco BRL,
Mississauga, ON), lOOpL of chloroforrn (Gibco BRL) and 50 pL of Buffer A was added,
vortexed, and cenuifuged at 14,000 rpm for 10 min.. The RNA was isopropanol precipitated,
washed in cold 70% ethanol and resuspended in 20-50 pL of RNAse-free DEPC-treated distilled
water.
IV.XIV. Cloning of TCR a and B variable region genes
This work is stili in progress and probably will not be compieted by in time for this thesis.
Below 1 describe parameters of the efforts to date to isolate the E12.3 TCR. The obstacle in this is
the high prevalence of endogenous BW (fusion partner) TcR transcripts in and out of frarne.
Obtaining this TcR would have been a considerable bonus for my thesis as well as my involvement
in subsequent work and the remaining weeks in the lab will be dedicated to this effort (and to
writing two publications of my work).
An anchored polymerase chah reaction (PCR) strategy (189) was used to amplify
rearranged TCR a and B regions from total RNA isolated from E12.3 cells. Primers specific for
the constant region of TCR cx (C-a) and TCR 0-chahs (C-0) were used to synthesize cDNA.
Subsequently, a poly-Guanidine (poly-G) homopolymer taii was added at the cDNA 5' end, using
calf thymus terminai deoxynucleotide tramferase (TdT). cDNA was then amplified by PCR using
sequence-specific primers for the poly-G tail and the constant-region of TCR a (C-a) or 0 chah
(C-l3) respectively. The cloning strategy is summarized in figure 1. Total RNA prepared from
E12.3 cells was reverse transcribed. Bnefly, 5 pg of total RNA (9pL in DEPC-treated water) was
denatured at 90°C for 5 min., chilled on ice and added to the remaining reaction components in a
~ ~ ~ C C C C C C C C C
dC Primer
Clone cDNA into pGEhl-T
E u y v.ctor
final volume of 40 pL (20 U RNA Guard
(Pharmacia Biotech, Montreal, Que.), 1X fmt
strand buffer (5OmM Tris-HCI, pH 8.3),
40mM KCl, and 6rnM MgCls), 200U
Moloney murine leukemia virus reverse
transcriptase (M-MLV-RT, Gibco BRL),
1rnM dNTP's (Pharmacia), 1 mM DTT, and
10 pmol of a synthetic antisense
oligonucleotide annealing to C-a (Z743: 5'-
AGATTCCATGGTITTCGGCAC-3') or C-
6 ( 2 8 3 1 : 5 ' -
GACCCCACTGTGGACCTCCITGCC-3').
To synthesize cDNA, the reaction was
incubated at room temperature for 10 min.,
then at 37°C for one hour.
cDNA was purified using the
QIAquick PCR purification kit according to the manufacturer's instructions (Qiagen, Mississauga,
ON). A poly-G tail was added to the cDNA with 17 U TdT (Gibco BRL), 1X TdT buffer (0.2 M
potassium cacodylate, pH 7.2, 2 mM COClz, and 0.2mM D I T ) (Gibco BRL), and 1m.M dGTP
(Pharmacia) in a final reaction volume of 30 pL. The reaction was canied out at 37'C/10 min. and
the enzyme was inactivated (68"C/15 min.). The modified cDNA's containing TCR a and TCR O
sequences, respectively, were purified using the QIAquick Nucleotide Removal Kit (Qiagen),
resuspended in 20 pL DEPC-treated, distilled water and amplified using the Hot-Start XL PCR kit
(Perkin Elmer, Mississauga, ON). The 30 S upper layer contained lx-XL Buffer II, 2U of rTth
DNA polymerase, 5 pL cDNA, and DEPC-treated water to 30 pL.
The lower layer contained: lx-XL Buffer II, 1.1 mM MgCOAC1~,0.2m.M dNTI"s, 200 ng
of a nested TCR constant region-specific pnmer containing a restriction site for SstI (C-a: 2745:
5'-AGACCAGGAGCTCTTAACTGGTACAC-3' or C-B: 2742: 5 ' -
TGCTTTTGAGAGCTCAAACAAGGAGACCTTGGG-3'), 35 ng of the anchor/dC pnmer (5'-
GGTCGACGCGGCCGCGGAGGCCCCCCCCCCCCCC-3'), 200 ng of an anchor primer
(JB4: 5'-GGTCGACGCGGCCGCGGAGGAGGCC-3') and DEPC-treated water to 20 pL. The
RT-PCR was performed using a thermal cycler (Stratagene, La Jolla, CA) with the foliowing
cycling parameters: denaturation at 9S°C for 1 min., annealing at 72°C for 1 min., and extension at
55°C for 2 min. for 30 cycles.
PCR products were purified with the QIAquick PCR purification kit and ligated into the
pGEM-T Easy vector using the pGEM-T vector system according to manufacturer's instructions
(Promega, Madison, WI). This system uses the single adenosine and thymidine overhangs for
l i gation to linear vector. DHS-a competent cells (Life Technologies) were transformed with the
ligation products by heat shock and plated on LB-agar containing 3.5 mg ampicillin, 100 pL of
100 mM IPTG, and 20 pL of 4% X-gal in DMF. The plates were incubated at 37°C overnight.
Several colonies were picked from both TCR-a and TCR-8 plates, growing in 5 mL liquid LB + 5Omg/mL of ampicillin at 37°C with shaking at 225 rpm overnight.
Plasmids were isolated using QIAprep Spin Plasmid Kit and 1.5 mL of the overnight
cultures, according to manufacturer's instructions (Qiagen). The pGEM-T Easy vector contains
two EcoRi sites on either side of the cloning site. To confirm the presence of inserts, 2jiL of the
plasmid preparations were digested with 10U of EcoRI (Pharmacia) in 2X OnePhor-Al1 Buffer
Plus (Pharmacia) for 1 hr at 37°C.
Positive clones were sequenced on an automated instrument using T7 and T3 primers and
sequences, 400-600 base pairs in length were submitted electronically to GenBank for analysis
with the BLASTn program.
IV.XV. Statistics.
Where applicable, Fisher's exact test and two-tailed, Mann-Whitney tests were used to
analyze proliferation data with significance set at 5%. Fisher as well as paired tests were employed
to analyze in vivo results.
Chapter V. Results The core element of my work was a focus on the major T ce11 epitope in ICA69, Tep69, a
naturally generated, immunodorninant region of the molecule, and its mimicry counterpart, the
ABBOS peptide from BSA (42, 93). Below, 1 will first describe in vitro studies with a T ce11
hybridoma, work that for a first time formally estabtished antigenic mimicry between Tep69 and
ABBOS, and that established the structural prerequisites for this mimicry. 1 will then move to
describe in vivo studies that had the aim to detennine if these peptides, in particular the
endogenous Tep69, was able to modify diabetes development in NOD mice. My findings were to
a certain extent surprising as they delineated a profound abnormality in NOD rnice exposed to
high doses of systemic peptide. This tolerogenic regimen in normal animals is immunogenic in
NOD rnice.
V.I. In vitro studies
The work below represents the characterization of a T ce11 hybridoma, E12.3, that was
initially generated by R.K. Cheung, another member of our laboratory. 1 took these ce11 over,
formally cloned the line and conducted the experiments described here. A special project student,
S. Wiener, helped with some of the proliferation assays. He has now joined the laboratory as an
h u n o l o g y graduate student and will take over my work as 1 leave for Medical School.
1
100 Io00
CPM ([WJTdR) F f l v r r ~ g n ~ ~ I n n i L 1 I L n d E l f J œ i h R e p i ï c a ~ ~ d c u ~ ~ o f &IO. El= hybridonu cclls WCTC culauul wilb 4.01101 i d a l c d NOD spleen cdb u APCI. Afca 4 r h y s in BSA-frct HS mdium in Ibr pmrncc a rbvœc d m i n a synlbrUc pcp& uit igau (05 pghvefl). 1 pCï JH-ïhymidinr (6.7 Cilmmol) vu akkd CO cuh 4 l . and incorpontion inio DNA w u Iirurrrcd by liquid-rcinulIation counting.Three indcpcnduic cxpcrimnli w a c prformcd R u u l i s var ccunbùrd ami expuscd as mran cpm + SD of üiplicarc cultrrrrs
V.I.I. Antigenic mimicry between
ABBOS and Tep69
In order to determine if the NOD
mouse T ce11 repertoire cm generate T cells
clones which c m recognize both BSA and
ICA69 we generated T ce11 hybridoma lines
from NOD mice immunized with the
immunodorninant epitope of BSA, i.e. the
ABBOS peptide (97). Several hybridomas
were generated as described in chapter II
and data from clone E12.3 are presented.
Once the El 3.3 hybridoma was successfully
sub-cloned, two questions remained to be
answered; (i) is the E12.3 hybridoma specific for its imrnunizing antigen ABBOS and if so, (ii)
does this T ce11 clone cross-react with its mimicry antigen, ICA69 and its epitope Tep69?
To address these questions we performed standard T ce11 proliferation as well as a ce11 deah
assay previously established in our laboratory (Figure 2) (186, 188, 190). A dramatic decrease in
proliferation was observed when E12.3 cells were cultured with BSA or ABBOS peptide provided
the presence of NOD-relevant antigen presenting cells (APC, e.g. irradiated syngeneic NOD
spleen cells). Control antigens Ovalbumin or peptide OVA152 had no such effects (p < 0.0001).
The BW hybridoma fusion partner showed no responses in any combination.
These data indicated an antigen-specific response of E12.3 to BSA and ABBOS. A ciramatic
growth inhibition was also observed when cells were stimulated with the recombinant BSA
rnimicry Ag, recICA69, or its immunodominant T ce11 epitope, Tep69 peptide (p < 0.0001).
Antigen-specific growth inhibition is a common feature of T ce11 hybridomas, which much like
immature thymocytes, undergo growth arrest and activation-induced ce11 death (AICD) upon
cognate stimulation through their T ce11 Ag receptor (191). In contrast to normal T cells which
undergo prohferation, antigen-mediated activation often induces an irreveaible ce11 cycle block in
T ce11 hybndomas (192, 193).
ABBBOS !no APC
Ovd bumin
RN W. Antigtm-sycllïc .cciratk. - idud œü dath of El23 ctllr Ccll vinhiiiry w u aucssd by particle caocmtrad innnunoflwrcurna ~ s a y . E l l 3 cclls wcr. culturd for 4 h y s wiih urtigcn (OJglrrcll) and imdiiiicri APC( c l.Ox 104 1- As' oslufcclcd C3G7 ccllr) Ïn HS medi- I O i J of BCECF dyc solurH>d l*) w u abdçd d k d y w d WU. Afvr ia~ubaau# f a Ihr r 37'0 5% C G . upwltc of dyc by virblc alb un rncnnnai by f'aadcx Sersm Mrhmc and is uptd os &vc il- imiU (m.
To determine if ce11 death was indeed
associated with the observed growth inhibition in
responding E12.3 cells, we used an established ce11
death assay (187) in Ag-stimulated E12.3 cells.
This assay is based on the BCECF vital dye which
is metabolized in the cytosol of living cells into
fluorescent esters (194). The data presented in
figure 3 indicate a dramatic loss of ce11 viability in
E12.3 cells cultured with the mimicry antigens
BSA/ABBOS and recICA69/Tep69, in the
presence of irradiated APCs (p < 0.0001). No
significant loss of viability was seen in E12.3 cells
exposed to the control antigens Ovalbumin or
OVA152 in the presence of APCs. It should be noted that the antigen specific efiect observed was
not caused by non-specific toxicity of the peptide or protein Ags since no growth inhibition was
observed when Et2.3 cells were cultured with the mimicry antigens in the absence of APCs. These
data demonstrate that antigen specific, APC-rnediated presentation of the relevant proteins and
peptides leads to activation of the E12.3 hybridoma and consequent growth inhibition and ce11
death, characteristic of AICD, presumably involving Fas-FasL interactions (191). These
observations formally establish the presence of antigenic mimicry between BSA/ABBOS and
ICA69TTep69 at the level of a single T ce11 clone.
V.I.11. MHC Restriction of Mirnicry Response
Figure 4 I
OVA 152
Ovdbumin
BSA i I 1 1' I CPM ( [ 3 ~ ~ d ~ )
FJ. Antigcn-specific growth inhibillon of El23 cdls. 2.0~101 El2 hybndomri cclls were culnrred w i h ZOxlû' imdiaced syngcncic imtigco prcxnting cclis (LAC' w f e c t e d C3G7 4 1 s ) . Cclls werc plaird in 96-weI1 fht boiiom micro-wcll plates and inabatcd (37'C. 5% CO2) for 4 days io BSA-frce HS medium in the prcxoce or absence of protein or synht ic peptide mtigens (0.5 pdwcll). 1 )ici JH-Thymidine (6.7 Cimmoi) was added to cjch well ovcrnighr. and incorporation of d c w t i v i t y into DNA was m c d by tiquid- scintillation counting. Tbrre indepcndcnt expcriwnts wcrc pcifomed. RcniIts wcrc cornbincd md expresscd as mean cpm t SD o f triplicak culturcs
To characterize the MHC-specificity of the
observed mimicry response we used APCs
from B6 (EEb) and Balbk (H2d), neither of
which allowed effective antigen presentation,
while NOD APC's perforrned as before (not
shown). To determine if, as expected,
presentation involved the NOD I A ~
heterodimers, we used the C3G7 ce11 line, a B
lymphoma ce11 line (M12) stably transfected
with the NOD H-2g7 B-chain (184). It is
important to note that the parent Ml2 ce11 line
only expresses endogenous 1 - ~ d a chain and
no other MHC molecules, M12-I-~d a is
identical to the NOD I - ~ g 7 a chain. Results in
figure 4 clearly indicate that C3G7 cells
bearing only NOD 1 - ~ g 7 class il MHC are fully able to substitute for NOD spleen cells as APCs.
These results demonstrate that the activation of EI2.3 T ceIl hybndoma is antigen-specific and
MHCNOD-dependen t.
V.1.m. Mimicry Response is Antigen Dose Dependent
The dose dependence of E12.3'~ antigenic mimicry response is shown in figure 5.
Responses are expressed as the % inhibition of E12.3 growth (3H-TdR incorporation) when
cultured with specific rnimicry antigens @SA, ABBOS, recICA69, Tep69) in the presence of
APCs (C3G7 cells) relative to the growth of E12.3 when culhued under the sarne conditions but in
the presence of only control antigen (Ovalbumin). The presence of a clear dose dependent
response in E12.3 reflects optimal loading of the relevant peptides into class II heterodimers and
these data indicate that E12.3 cells follow normal T ce11 physiology.
Finure 5. Dose-dependence of mimicry respoasc in E U 3 cells. 1.0 X IO4 E12.3 cells w e e cultured with 2.0 X 10' imdiated [-Ag7 transfccted C3G7 cells. After 4 days in HS medium in the presence or absence of protein or synthetic peptide antigens (0.1-10.0 pglwell). 1 pCi 3H-Thymidine (6.7 Ci/mrnol) was added and the incorporation into DNA w;ls mcasured by Liquid-scintillation counting- Four independent sxperimtnts were performed. Results were combined and exptessed as mean cpm 2 SD of inpliute cultures.
V.I.N. Structural prerequisites for mimicry
Once antigenic-rnirnicry was established at a clona1
level, we investigated the structural prerequisites of
molecular mimicry between ABBOS and Tep69.
Previously, ABBOS and Tep69 were identified as
the minimal, immunodorninant, epitopes in BS A and
reclCA69, respectively (42, 93, 96). To identify
critical amino acid residues involved in TCR and/or
MHC interaction within the mimicry epitopes, we
used ALA mapping, i.e. a series of synthetic peptides
containing single arnino acid substitutions. We
obtained a set of alanine replacement peptides for
ABBOS and Tep69 (a kind gift from Dr. J. Gorga,
Pittsburgh, PA). Table 2 provides a listing of the peptides and their corresponding alanine
substitutions.
We then performed T ce11 proliferation assays as before, culturing E11.3 cells in the
presence of native, unsubstituted or an alanine substituted peptide in the presence of imadiated
C3G7 cells as APC. If substitution
with an aianine residue alters either
MHC binding or TcR contacts critical
for T ce11 activation, this should be
reflected in a lessening or loss of
E12.3 activation (195). Figure 6
surnmarizes results obtained. Single
alanine substitution at any of three
lysine residues (aa 15 1,155, or 156 of
BSA) and residues D, E, and F (aa
153, 154, and 157 of BSA) within the
ABBOS peptide completely abrogated
the peptide's ability to trigger cognate
ce11 death in E12.3. Interestingly, and
in keeping with the mimicry between
- -
Figure 6 C - Q O - ABBOS A
F Tep69
q , E I
50.000 100.000 50.~000 100.000 Cell Survivai (CPM. [ 3 ~ d R )
Replicement of critiaû uniw acid d d u a wilhin ABBOS and Tep69 riîb iluiint abrogmtcs tbeir a p c i t y to inducc growtâ inhibition io E I U cdls Replicrite cultures of 2Ox lWE12.3 cclls were culturrd uithZ-OxlW irradiated 1-Ar' uansfeclcd C3G7cells as APCs in HS medium in the presence or absence of ALA rcplacmnt peptides for ABBOS (left) or Tep69 (rigfit) (0.5 pglwell). Aftcr 4 days. 1 pCi 3H-Thymidine (6.7 Cimmol) was addrd to u c h wcll. and incorporation into DNA was measmeci by liquid-scintillation counting. Two independent experiments were performed. Results were combined and expresscd as rncan cprn SD of mplicate cultures. ALA peptide which were not taled duc to iinsolubility am noted. (* replacerrunt A wiih D. sec ABBOS-D rephccmcnt ucptidc rcfer to figure 7).
ABBOS and Tep69, it is these same three residues that constitute the (entire) homology between
BSA and ICA69 in this region (94). These observations document the exquisite specificity of the
E l 2 3 clone.
Arnino acids Threonine and Glycine (aa 41 and 42 in ICA69) correspond to Aspartate and
Glutamate (aa 153 and 154 of BSA). Ala replacements were insoluble, but we previously observed
that exchanging these residues between ABBOS and Tep69 (DE<->TG) had no effect on T ce11
responses to either peptide, while replacement of alanine (aa 152) with ASP rendered this ABBOS
homolog ineffective ('ABBOS-D', see below, figure 7 and ref. (93)), presumably interfering with
TcR contact (195). It was sornewhat surpnsing that substitution of Glutamate with the small
glycine residue did not affect peptide binding andor recognition. Since Our present results
highlight the importance of these amino acids in ABBOS (KADEKK), it would be justifiable to
assume that the T and G residues in between the fiat two lysines within Tep69 (KATGKK) are
also important for TCR and/or MHC interaction.
Al1 other ALA substituted ABBOS peptides did not significantly affect stimulation of E12.3
cells. However, replacement of any of the three homologous, identically spaced lysine residues in
Tep69 also abrogated the ability of this peptide to trigger its cognate response. Unfominately, we
were unable to test each and every replacement peptide in Tep69 due to solubility probiems.
Nevertheless, the Tep69 data available thus fully confirm conclusions obtained with ABBOS
replacement peptides and they define the structural prerequisites for the antigenic rnimicry between
the two epitopes. Four identically spaced residues are sufficient to mediate exquisite specificity of
these rnirnicry epitopes with little contribution of neighboring side chains. These four critical
residues include a lysine in position 10, which is part of the anchor motif for human and mouse
DQA-A binding peptides (195).
Taken together, Our data suggest that the KAxyKK motif is a rather absolute requirement for
mimicry, while there is more flexibility in the choice of side chains for residues flanking this motif.
Structural prerequisites for mimicry and peptide recognition are thus not identical.
V.I.V. Immunization generates E12.3-1ike T cells
We next detennined if the fine specificity profile delineated in E12.3 cells could be detected
following irnmunization with BSA. NOD males were immunized once with BSA in complete
Freund's adjuvant and regional lymph node cells were stimulated in microcultures as described
(97). As shown in figure 7, the single round irnmunization had generated T cells that recognized
Tep69 and ABBOS, but these responses could not be recalled with single amino acid replacement
peptides containing alanine instead of the lysines in the core KAxyKK motif, or replacing the
alanine with Asp. Replacing a distant, C-terminal K or H had no measurable effect on recall
responses. Thus, a single BSA imrnunization generates T ce11 pools with the same fine specificity
patterns as E12.3 (which itself was denved from an ABBOS-immunized mouse). These
observations suggest that E12.3-like cells represent a common rather than rare NOD T ce11
response to BSA.
Figure 7 T ceil Responses (CPM, means+ SD [ 3 ~ ~ d ~ ) lm 1OOOO 1OOOOO
Tep69 (TKQAF IKATGKKEDEHW)
(TKQAFIAATGKKEDEHVV)
(TKQAFIKATGAKEDEHVV)
(TKQAFIKATGKAEDEHVV)
(TKQAFIKATGKAEDEAVV) 1 '
4BBOS-D (CDEFKDDEKKFWGKYLYE)
(CDEFKADEAKFWGKYLYE)
ABBOS (CDEFKADEKKWGKYLYE)
(CDEFKADEKKFWG AYLYE) 11
t
m r e 7. T ce11 responses to alanine replacement peptides in NOD mice. Regional LN cells were obtained from 3 NOD males immunized with 50 pg BSA s-c. 9 days prior to sacrifice. Triplicate cultures of 2x105 LN cells were pooled and cultured with 0.5 pg of ABBOS, Tep69, or the corresponding alanine substituted peptides. 6-18 hr before harvesting, 1pCi of [3mthymide (6.7 Ci/mmol) was added to each well and isotope incorporation was analyzed by liquid scintillation counting. Results from two experiments were combined and expressed as mean cpm * SD.
(CDEFAADEKKFWGKYLYE) 1 4
V.I.W. E12.3 TcR
The E12.3 clone was analyzed phenotypically. Given the drarnatic and very robust,
reproducible and sequence-specific peptide response data, it was not swprising to find rather high
level expression of TcR as measured by flow cytometry after staining for CD3 (Figure 8). RT-
PCR for the detection of TcR message indicated a high abundance of CD3, TcR-a and -f3-
transcnpts (not shown). Cloning attempts of the specific T ce11 receptor involved in the rnimicry
response is ongoing. To date, the eight TcR a and -&clones sequenced al1 were derived from the
BW fusion partner, one 0-chain was
productively rearranged (VBUDB2.l/JB2.5 and
Val. UJocS.5) and another is not functional,
with several stops in the Vi3 5.2/DB2.1/JB2.5
region. This confirms published data (196).
The sequenced alpha transcnpt include a
previously reported non-functional
pseudogene rearrangement (V-a 16.1) and a
productively reamnged but not translated V-a
l.l/J-a 2.5 chain (181, 197). Thus, the
relevant NOD mouse-derived TCR alpha and
beta transcnpts are likely very much 'diluted'
by endogenous, non-functional B W (fusion
partner) TcR transcripts that compete for the
same promoter activity . 1 am planning to
isolate and sequence additional TcR clones,
and will test an antisense-RNAse H strategy to
selectively destroy B W-derÏved transcripts.
V.11. In vivo studies
Although the demonstrated antigenic
Log,, fluorescence intensity M8. Expression d C D ~ E on BWS147, NOD spkcn .ad El= dk. BW5 147 celis (A & 8) . NO0 spleen celis (C & D) or El 23 (E & F) wcrc mina i vidi anti-mouse CD3& Ab (B. O. & O) while conml samples (A. C & E) wcre staincd with igGI iswpt conuol Ab. Thc exprrssioci of CD3E vrrs anaiyscd by fiow qomctry. Gghc -nt OC BWSI47 ctlls. fifty paccnt OC s p I a a l i s and one hundrd paunc for El23 cclis naincd positive.
mimicry between ABBOS and Tep69 epitopes was unequivocal, we had previously noted
considerable functional differences between these peptides. Thus, neonatal tolerization of NOD
mice with Tep69 was far more effective and long-lasting than that following neonatal ABBOS
exposure (42). Consequently, only Tep69 tolerized mice were significantly protected h m diabetes
development. In contrast, immunization of young adult mice prior onset of insulitis, with ABBOS,
but not Tep69 significantly reduced diabetes incidence (42). While it is tempting to suspect that
affinity/avidity differences may contribute to these functional differences among the two mimiay
peptides, we have been unable to derive relevant data - perhaps due to the low level TcR expression
on E12.3 cells. We therefore extended Our previous data to search for further differences between
the two peptides. In this we focused on experimental strategies that resemble more pre-diabetic
autoimmunity than the natural course of disease development. If irnrnunotherapy of IDDM would
32
reach human application, it would likely involve subjects with definitive signs of pre-diabetes
(198).
V.1I.I. Modification of adoptively transferred diabetes
If, as suggested by our previous observations in the rather special case of neonatal tolerance
induction, Tep69/ICA69 is involved in the disease process leading to IDDM, it should be possible
Figure 9
Days post transfer Fin 9. Cumulative incidence of adoptively transferred diabetes is dependent on the the number of dïabetic spleen c e k transferred. Either O-. 5-, IO-, 15-, or 20-x106spleen cells pooled from 3-4 recently diabetic female mîce were transferred to irradiated male NOD male ( 6 8 wk old) recipients. Recipients were screened frequently for diabetes and were considered diabetic when blood-glucose 2 260 mg/kg on two consecutive
to manipulate disease development by
antigen-specific immune intervention in
the course of this process (1). Our
previous studies coutd not determine
whether Tep69-specific T cells actually
acted as effectors of 0-ce11 destruction or
indirectly influenced 0-ce11 demise in
some fashion (199). In order to better
charactenze the functional significance of
Tep69 reactive T cells during disease
progression and to investigate the
immunotherapeutic potential of Tep69
peptide treatment, we employed a
tolerance induction strategy using soluble
peptide (37) to modify disease course in
the adoptive transfer as well as
cyclophosphamide (CY) induced rnodels
of diabetes development. These two
models served as a consistent and rapid
means of producing disease, less susceptible to environmental influences that shape diabetes
development during the slow, natural development of the disease in NOD rnice (21).
V.II.11. Calibration of adoptive diabetes transfer
Pilot experiments with i.v. Tep69 peptide treatment in adoptive uansfer recipients of diabetic
spleen cells unexpectedly failed to show any protection from the disease but instead pointed to
peptide-dependent acceleration of disease development. To allow observation of both, disease
precipitation as well as disease protection, it was necessary to generate conditions for an
intermediate incidence of diabetes (-50%) following adoptive transfer of diabetic spleen cells into
irradiated recipients. Even subtle departure from
identity between related peptides can
drarnatically alter their agonist activity (200-
205).
We therefore performed a spleen ce11
dose titration with 0-20x 106 pooled diabetic
spleen cells, transferred to groups of 10-
sublethally irradiated NOD males per dose
(Figure 9). Adoptive transfer of 0, 5-, IO-, 15-
or 20x10~ diabetic spleen cells generated a
cumulative diabetes incidence of O%, 20%,
50%, 75%, and 100% respectively, establishing
a clear-cut ce11 dose relationship with ensuing DDM as expec ted (23). Subsequent experiments
used a ce11 input of 107, which generated the intermediate rate of diabetes development desired
V.II.111. Tep69 precipitates adoptively transferred diabetes
The following experimental plan was executed (Figure 10): 400pg of Tep69, ABBOS?
GAD65 peptide p34 (40). or BSA- 193 (42) control peptide were injected i.v. 1 day after adoptive
transfer of 10' diabetic spleen cells into five groups of 8-10 mice (Figure 11). After 5 weeks. the
C I A m 9
BSA193, i ) PBS. n=8 4 - O - 3 50
2 ABBOS, n=9
25
O
Doys P m TMSIu Hguu~U. Eucet or various LI. pcpüdc Lmtmcn. oii dopclvcly l n a s î d diabetes JOOpg Tep69. ABBOS or GAD65 peplidc w u injecicd i.v. in PBS ( 1 0 q i I ) 1 &y posi a iufcr . PBS pmvidcd a vchiclc con ml and BSAI93 w a peptide conltol. ,Mice werc moniiorcd frqucntly for glycoiuria. and were considercd di&c if blood nlucosc kvcb wat >13.8 mmoM. alucau.
diabetes incidence was lower in ABBOS treated
rnice (44%, p<0.0 1) than in BSA-193, GAD65,
or PBS injected animals (66%, 55%, and 62%
respectively). In conuast, an increased diabetes
incidence was observed afier injection of Tep69
(89%, p<0.01). Since treatment with BSA-193
or PBS did not a significant effect on disease,
the precipitation of diabetes with Tep69 was
antigen-spi fic.
The difference between ABBOS and
Tep69 peptides was highly significant
(p<0.001), suggesting that despite the clear cut
mimicry between the two peptides, they are not
at al1 functionally equivalent. Small changes in
primary systemic exposure to a high dose of antigen or peptides following intravenous injection is
a recognized and well characterized tolerogenic regimen (198,206,207). Our observations in i.v.
Tep69 injected animals implied that NOD mice may be unable to normally execute tolerance
induction through this pathway. Proposed explanations of failures in NOD mouse self-tolerance
have previously been associated with peculiar, low-avidity I-A~' binding characteristics (184). To
better characterize the Tep69 effect,
we asked if pre-treatment of diabetic spleen cell donors prior to transfer could enhance its diabetogenic effect, as may be expected for a proper immunization response. As well, we reduced
the peptide doses by 75% because the high doses used in the first series of experiments may
merely have overwhelmed an immune system inherently prone to tolerize ineffectively.
The experimental design used is described in figure 12, results are illustrated in figure 13.
Treatment of adoptive transfer recipients with Tep69 enhanced IDDM development at a similar
rate as before and raised disease incidence from 56% to 77% (p<0.01). Intravenous Tep69
injection of spleen cell donors (100pg, i.v.) 1 day prior removal of their spleen cells was without
significant effect (incidence, 508, p 0.34). However, i.v. injection of Tep69 into donors as well as
t Tep69 i.v. PBS i-v.
Dav -1
1 PBS i.v.
Tep69-wTep69 Tep69->PBS Group croup
1 PBS i.v.
PBS->PBS Croup
Monitor Diabetes by Glycosuria and Blood Glucose I Fieui-e 12. Pre-transfer and post-transfer i.v. Tep69 peptide injection
recipients produced rapid and complete diabetes development (incidence: 100%, p4.002).
This suggested, that the effect of i.v. Tep69 injection into donor and recipient mice showed
synergy, through prior treatment of the adoptive spleen cell donors. While injection of only the
Days post transfer m u . ENect OC i.v Tep69 peptide on diabetcs incidence followinl adoptive transfer of l07spleea ctlls from àiabetic fuiinles. Subsets of micl received 1 0 p g of Tep69 either before or afier masfer, or both before and aftei transfer. PBS (lcQ.11) provided a vehicle control.
donor had no major effect by itself.
The fact that pre-treating or
'boosting' with systemic peptide
leads to rapid and complete diabetes
development supports Our above
conclusion that NOD mice fail to
normally undergo tolerization
following systemic peptide
exposure. These data also indicate
that Tep69 responsive T cells are
principally able to support
progressive beta ce11 autoirnrnunity,
and they imply that immunotherapy
in a diabetes-prone host may face
challenges of disease exacerbation
rather than mitigation, difficulties not
expected in normal strains of mice.
The natural history of IDD in humans and mice involves a protracted penod of pre-IDD.
Many factors may impact on disease development b208)- We chose to focus on accelerated models
of IDD to circumvent the long prodromal disease stages (32, 209) Peptide therapy appears to
hasten andor enhance the development of diabetogenic T cells already present in the graft. To
determine if this possibility is correct, we employed a second model of accelerated IDD: the
cyclophosphamide model. There, the cytotoxic dmg CY alkylates majority of T cells, and IDD
develops during a rapid wave of regeneration in T ce11 and monocyte compaiûnents.(22).
Thirty four 7-week-old rnice were divided into 3 groups of 10, 10, and 14 mice respectively
(Figure 14). The first group (n=IO) received a single i.v. injection of 100pg of Tep69 peptide
while the two other groups received vehicle only (PBS). 5 days after peptidelPBS treatment, al1
three groups received a single injection of CY i.p to induce diabetes. The CY dose had been
calibrated in pilot studies to generate an intermediate diabetes incidence of about 50%. Five days
after the CY injection, the pre-treated and half of the untreated rnice received an i.v. injection of
100pg of Tep69 or a control peptide, OVA152 (Figure 14). Four weeks later, dl anirnals that
wouid develop IDDM had converted to diabetes, no additional cases were observed over a 12 week
period. Very sirnilar to the Tep69 effect in adoptively transferred mice, a significant increase in rate
and incidence of diabetes was observed in mice that received a single Tep69 injection five days
after CY treatment (60% vs. 43%, p 4-02). The diabetogenic effect of Tep69 was significantly
enhanced in rnice receiving two
Tep69 peptide injections prior
to and after CY treatment (9096,
p ~0.02).
On face value these
observations are consistent with
the foregoing conclusions, in
that i.v. administration of Tep69
promotes diabetogenesis in an
antigen-specific, boostable
fashion. These data are
consistent with our previous
studies in other expenmental
rnodels (42): colIectively they
assign an important role to
ICA69 and its major T ce11
epitope, Tep69, in early,
intermediate and late NOD
mouse diabetes development.
'peptide ' m e 14. Effect of i.v. Tep69 treatment on CY-induced diabetes. NOD males were injected with 100pg Tep69 i-v., either 5 days before (Re-CY), or 5 days after (PostCY) CY injection. The control group received OVA152 convol peptide- The difference in incidence between the control p u p and the experimental group is statistically significant at P < 0.02 by Mann-Whimey or Wilcoxon non-pararnetrïc paired test.
V.1I.V. Conventional immunization with Tep69
Conventional imrnunization with diabetes antigens has been reported to generate diabetes
protection, although, as in Our case, exact mechanisms remain elusive (41, 175, 177, 179, 180).
NOD male mice, 8 weeks old, were immunized with Tep69 or OVA152 peptide (50 ug i.p.)
ernulsified in incomplete Freund's adjuvant (IFA). Diabetes was accelerated two weeks later by
injection of 250mgkg CY. The cumulative incidence of diabetes in mice immunized with the
control peptide, OVAl S2 (n=lO) was 60% by four weeks (Figure 15). However, mice immunized
with Tep69 i.p. (n=10) exhibited a significantIy lower in diabetes incidence than these control mice
(p 0.02).
Mechanisms for this protection remain unresoived. For example, we were unable to detect
THU2 biases in Tep69-specific T cells (data not shown). However, it is uncertain if T ce11
functional measurements after CY treatment are meaningful in face of the massive T ce11
regeneration at the same time. Our data concur for our test antigen and its BSA mimic (42), as well
Figure 15
bePtiae Days after CY iqjection -rn uccd diabetes in wrel5. Cumuiativc incidence of CY ' d
mice immunized i.p. with Tep69 and incornpletc Freund's adjuvant (IFA) or OVAlS2 and IFA. Two groups of 10 male NOD mice (7-8 weeks of age) were immunized wirh a single injection of Tep69 or conuol peptide (P). Two weeks later diabetes was induced wirh a single i-p. injection of CY (250 mgkg). The cumulative diabetes incidence was statistically significanr (p c 0.02) as deiermined by Mann-Whitney or Wikoxon. two-taiIed, non-pyynetnc test
as for other diabetes antigens tested (e.g.
(41)), that traditionai modes of rodent
immunization reduce or bloc k diabetes
development. In this we c m conclude that the
T ce11 repertoire targeting ICA69 is focused
on the Tep69 epitope. and that modification
of this T ce11 pool in any way impacts directly
on progression of diabetic autoimrnunity to
diabetes.
The data generated here, demonstrate a
defect in NOD mouse peripheral tolerance
induction. It is intuitively obvious that having
identified a T ce11 repertoire (Tep69-specific)
which is involved in diabetogenesis, that any
modulation of its activity will influence
disease outcorne. We conclude, that al1
models of NOD mouse diabetes are
susceptible to specific modification by
Tep69, and that this modification can entail both, promotion and arrest of progressive diabetic
autoimmunity .
Chapter VI. Discussion IDDM is a chronic autoimmune d i s e s caused by the T cell-mediated destruction of insulin
producing 0-cells (1, 6). Both, environmental and genetic factors contribute to diabetes
pathoetiology (7). The initiai triggenng event(s) of IDDM are unknown. However, a number of
candidate autoantigens have been shown to play a d e in the initiation andor progression of type 1
diabetes. Insulidpro-insulin, GAD65/67, hsp65l70, IA-2, and ICA69 were identified because they
are frequently targeted by DDM-associated autoantibodies andor autoreac tive Tcells (6, 2 10).
For example. over 80% of patients with recent onset IDDM have humoral andor T-ce11
autoreactivity to ICA69, making ICA69 one of the most frequent autoimmune targets in this group
of autoantigens (170). In NOD mice the same antigens play central roles in beta ce11
autoirnrnunity, since immunomodulation with these antigens consistently prevents disease.
Antigenic mimicry has been suggested for GAD65 (with a Coxsackie B epitope) (21 1). and for
ICA69 (with BSA) (70). Human data suggest that Coxsackie-like epitope in GAD65 is not
nomally targeted in diabetic hosts (102, 103). but there is considerable, indirect evidence that the
Tep69/ABBOS epitopes are mimicry antigens in diabetic patients (93).
Here we have analyzed the T ce11 mirnicry between Tep69 and ABBOS in detail, establish a
fine specificity map for the E12.3 T ce11 hybndoma that recognizes both peptides, define the
structural prerequisites for this mimicry and demonstrate that T cells with the sarne fine specificity
profile are comrnon in BSA imrnunized NOD rnice. We then analyze the effects of these peptides
in vivo. Depending on experimental details, Tep69 will either precipitate or prevent diabetes in
accelerated models of pre-diabetes. Despite the established rnimicry, Tep69 and ABBOS are not
functionally equivalent, with some protective but not disease precipitating activity. These
experiments delineated a novel immune abnormality in NOD mice, which fail to develop penpheral
tolerance following systemic (i.v.) antigen exposure. Overall, Our observations confinn the role of
ICA69JTep69 self-reactivity in NOD mouse IDDM, they formally establish a link of this
autoimrnunity to the mimicry antigen, BSA, and dernonstrate that mirnicry epitopes can differ in
their function within the frameworks of diabetic autoimmunity. The observation of disease
precipitation in models of pre-diabetes cautions against similar outcornes of immunotherapy trials
in subjects wi th de finit ive signs of pre-diabetic autoirnrnunity .
VI.1. Molecular mimicry at the Clona1 Level
Bovine semm albumin is considered a possible environmental trigger for the development of
type 1 diabetes (8 1, 212). In 1992 Our laboratory proposed that antigenic rnirnicry between BSA
and a, then unidentified, 69kDa islet protein (ICA69), played some role in the pathoetiology of
type 1 diabetes (70). The molecular mimicry hypothesis of autoimmunity refers to a breach of
self-tolerance through imrnunological cross-reactivity between naturally occurring, non-self,
proteins and a self-protein that share sufficient structural homology (213). The structural
prerequisites for such mimicry are not well defined, but often may not involve linear sequence
identity (214,215).
In Our model, early exposwe to the rnimicry antigen BSA would recruit self-reactive T cells
targeting the pancreatic 6-ce11 antigen ICA69. Expansion of cross-reactive T ce11 pools beyond a
numerical threshold (173) would then contribute to progressive autoimmune disease (6). Once
ABBOS-reactive T cells have been generated, they would be sustained or amplified in the course
of pre-diabetes progression by 6-ce11 ICA69lTep69. This possibility would be consistent with
findings by Larger et al. (89). These authors chemicaily removed 6-cells at weaning. Serial
transfer of spleen cells failed to generate IDD in recipients, suggesting that the presence of B-cell
autoantigen is a prerequisite for sustaining diabetogenic T ce11 pools. In this study, 0-cells were
removed at a very young age, before insulitis, before detectable islet imrnunity, and before
exposure to dietary cow milk/BSA. This experimental design fails to differentiate between the role
of mirnicry and B-ce11 antigen in the evolution of diabetogenic T cells and their maintenance for a
long period of time. It is well established that diabetogenesis requires an interplay of
environmental and endogenoudgenetic factors. The Larger study confims that self-antigen is
required to maintain antigen-reactive T cells in vivo and should provide an excellent tool to analyze
the potentiai role for rnimicry in potentiating autoimmune diabetes (216).
Antigenic mirnicry between BSA/ABBOS and ICA69Kep69 was first observed in T ce11
populations of newly diabetic patient (93) and then in NOD mice (42). Intriguingly, and perhaps
encouraging for an eventual transfer of insights gained in such NOD mouse experiments, the
same, identical epitopes are targeted by T cells from diabetic patients, with strong, albeit indirect
evidence for similar antigenic rnirnicry (93). Immunization of NOD mice with recICA69 or its
irnrnunodominant epitope, the Tep69 peptide, leads to rapid recniitrnent of T cells specific for BSA
and its imrnunodominant epitope, ABBOS as well as the cognate antigen itself (42). Vice versa,
irnrnunization with BSA or ABBOS recruits T cells specific for ICA69. Similarly, cross-tolerance
cm be induced following neonatal injection with either one of the two mimicry antigens (42).
In Our previous studies, we could not formally rule out the possibility that multiple or
overlapping T cell populations were giving rise to the apparent rnimicry phenornenon. This
concem was reinforced by observations of several differences in the NOD rnouse immune
responses to Tep69 and ABBOS. For example. in responses to neonatal tolenzation, the ABBOS-
induced (cross-)tolerance is short Iived and without effect on diabetes development, while (cross-)
tolerance to Tep69 is long lasting and diabetes protective (42).
ABBOSmep69-crossreactive T ce11 pools are found following irnrnunization in complete
Freund's adjuvant (CFA) or, spontaneously, following adoptive transfer of diabetes in non-
immunized mice (97). However, immunization with ABBOS peptide cm in some instances recruit
T ce11 pools that are unable to recognize Tep69 (42). This failure to display mimicry was observed
in mice protected from diabetes following i.p. immunization with ABBOS emulsified in FA.
When these protected mice were re-immunized with ABBOS (in CFA), we could recall T ce11
proliferative responses to ABBOS but not Tep69, unless animals were first immunized with BSA
(42). This implied that the NOD mouse can generate mirnicry as well as non-mimicry T ce11
repertoires specific for ABBOS, the latter associated with escape from diabetes. In the present
context we propose that the presence of a mirnicry peptide does not necessarily enforce a mimicry
response. How these alternative repertoires are chosen or selected is uncertain, but studies of
affinity/avidity are heading our iist for future experiments.
Collecti vely , these observations clouded Our rnimicry mode1 where ABBOSlTep69 cross-
reactive T cells are core elements. The development of the E12.3 T ce11 hybndoma was thus a
critical experiment, formally demonstrating that a single T ce11 clone shows the proposed antigenic
rnimicry in an unequivocal fashion. E12.3 cells allowed us to test and charactenze critical aspects
of the ABBOSmep69 mirnicry. Cell proliferation and death assays indicated that E12.3 responds
exclusively to mimicry antigens BSNABBOS and ICA69/Tep69. but not to control antigens
Ovalbumin or OVA152 nor a series of closely related Ala replacement peptides. The antigen-
specific response of E12.3 was dose dependent and resulted in activation induced ce11 death
( AICD).
The structural prerequisites for antigenic mirnicry are poorly understood (214). The
mimicry between ABBOS and Tep69 is remarkable as the homology between the two molecules is
Iirnited to four equally spaced residues (KAxyKK, where xy can be DE or TG). As discussed
below, every one of the four shared residues is critical. There are 2 additional protein sequence
regions that show more extensive, linear homology ôetween BSA and ICA69, but no antigenic
mimicry is observed between peptides from these regions which appear not to be immunogenic,
and at least one of these peptides is not naturally generated andor presented (42.94).
Resuits of alanine mapping indicated that a stringent structural requirement for mimicry is
the presence of the KAxyKK motif within the mirnicry peptides. The fact that few arnino acids are
needed to generate mimicry is not unprecedented. Studies in experimental autoimmune
encephalomyelitis, the murine mode1 of multiple sclerosis, demonstrate that synthetic peptides
containinp four native myelin basic protein (MBP) residues c m recall T ce11 responses against
native MBP (217). The same group has now demonstrated that a mimicry peptide from
herpesvirus sairniri with homology of 5 amino acids can stimulate MBP-specific T cells and
induce experimental allergic encephalomyelitis (EAE) in vivo (218). We conclude that the
prerequisites for mirnicry do not depend on contiguous, linear homology, but on specific amino
acid structural motifs.
As expected, the E12.3 response strictly required NOD I-A~' restriction as demonstrated in
studies with I-A~' transfected B lymphoma cells (184). We have not previously tested whether the
rnirnicry peptides can be presented andor recognized by T cells in the context of MHC class 1.
Preliminary experiments with class 1 and with ciass II deficient NOD mice showed weak or absent
responses in both. We are also curious as to whether ABBOS and Tep69 bind within the classical
peptide binding groove of the MHC- Although the location of antigenic binding to MHC has yet
to be determined in NOD mice, it has k e n suggested in human studies that both the ABBOS and
Tep69 peptide contain MHC contact sites outside of the classical peptide binding groove (219),
near the contact regions for superantigen binding, a property suggested for another diabetes target
antigen, insulin (220). My preliminary experiments found that OVA152 fails to compete
ABBOS/Tep69 in a functional cornpetition assay (data not shown). OVAlS2 is a naturally
processed and presented immunogen in NOD mice (42). However, more ngorous expenments are
underway to confirm and extend these fîndings.
1 have started to examine the TCR genes expressed by the E12.3 hybridoma. The data
available to date explain the low level TcR expression through the abundance of endogenous BW
transcripts. Further cloning will be required to identify the functional TcR transcripts involved in
the E 12.3 antigen response.
VI.11. Tep69 Peptide precipitates diabetes
Karges et al. previously demonstrated that diabetic autoimmunity in NOD mice targets
ICA69 with antigenic mimicry between the same T ce11 epitopes, ABBOS and Tep69, as observed
in diabetic patients (42). The finding that neonatal tolerance induction to Tep69 in NOD prevented
diabetes suggested a possible role for Tep69 in the disease process in NOD mice. Taken together
with evidence that the NOD mouse model reflects many of the key features of D D M including
genetic predisposition and susceptibility to environmental influence of disease (1, 19, 28, 221,
222), these data implied that the NOD mouse provides a model for studying ICA69 and its role in
the pathogenesis of IDDM (42). ICA69 is the only diabetes target antigen with a bown, common
environmental mimicry antigen, and exposure of genetically susceptible infants less than 3-4
months of age to infant formula, that contains this antigen is associated with a more than 10 fold
higher risk to develop the disease (83). The largest and most recent study from Finland showed
that abnormall y enhanced immuni ty to BSA is an independent marker of diabetes risk (85).
On this background it seemed important to extend our studies of mimicry between the two
relevant epitopes, as well as to assess their role in diabetic autoimmunity in greater detail. Our
expenmental models, adoptive transfer and CY induced IDDM, avoid long term regulatory
influences that characterize the slow, spontaneous diabetes development in both man and NOD
mice, and they more closely model pre-diabetes, i.e. the phase of disease development just
preceding overt diabetes. Since current thinking in the field targets pre-diabetes for immune
intervention, our experiments would also test the potential of imrnunotherapy with the two rnimicry
peptides. An earlier finding that ABBOS- and Tep69- specific T ce11 pools are spontaneously
generated following adoptive transfer of diabetic spleen cells, suggesting a link of ICA69-specific
T celis to disease progression (42). Our studies extend these observations.
Our data show, for the fmt time, that autoimmune diabetes cm be acceleratedprecipitated by
i-v. injection of a soluble peptide autoantigen. Intravenous injection of Tep69 peptide significantly
increased the incidence of adoptively transferred diabetes. In contrast, i.v. injection of ABBOS
peptide had a marginally protective effect on disease. Tep69 acted in strictly antigen specific
fashion, since control peptides and vehicle (OVAlSuBSA193) were without effect. The effect of
Tep69 could be boosted by a second i.v. injection, thus following classical immune response
patterns. Consistent with these results, i.v. injection of Tep69 also precipitated CY-induced
diabetes. The ability to modulate diabetes in the two IDDM models reinforces the notion that
Tep69 plays a critical role in the disease process. Our data imply a more direct role for Tep69-
specific T cells in the progression of diabetic autoirnmunity and B-ce11 destruction. However, we
have not, so far, obtained evidence for direct effector function of Tep69-specific T cells associated
with beta ce11 death. Perhaps due to the need for serum- (i.e. BSA-) free culture conditions, we
have been unable to generate relevant T ce11 lines.
Our experiments on one hand deiineated striking sirnilarities between ABBOS and Tep69 as
analyzed with E12.3 hybridoma cells. On the other hand, we found considerable differences in
their functional role within the immune system. ABBOS failed to precipitate disease and had a
small protective effect, reminiscent of the disease protection observed after conventional
imrnunization with ABBOS (42). Mimicry does thus not necessarily include identical
functionality, indeed, Tep69 and ABBOS peptides had essentially opposing effecü in our data set.
Together with the earlier evidence that ABBOS is not a good (neonatal) tolerogen (42). potentially
different roles emerge for each of the mimicry peptide pair: ABBOS may initiate early T ceIl poots,
and Tep69 may play a more driving role, consistent with the 'driving' role of beta cells in disease
progression (89).
Disease protection by various regirnen correlates with the absence of T ce11 reactivity to
Tep69/recICA69 and progressive autoirnmunity with the ability to activate and expand Tep69-
specific T cells: (1) diabetes protection with systemic immunization with ABBOS (in F A ) was
associated with a loss of reactivity to Tep69 (ICA69) (42); (2) rnice protected from diabetes by
feeding a BSA-free diet do not show ICA69 responses even after immunization, unless primed
with BSA/ABBOS, while mice that progressed to overt diabetes often show spontaneous T ce11
responses to ICA69 (97).
The disease precipitating effect of Tep69 was unexpected. Intravenous antigedpeptide
injection is a well described strategy for peripheral tolerance induction (198,207,223,224). The
observation that NOD mice develop boostable responses following i.v. peptide treatment delineates
a new abnormaiity in the NOD immune response system. Further experiments are in progress to
better characterize and understand this abnormality. The defect is not absolute, since neonatal
tolerization with Tep69 is effective (42), as is intrathymic antigen delivery (39) although not in al1
studies (225) and it may involve only certain antigens and not or less so others (such as ABBOS
in this study and GAD65 in another (40)). Tolerance induction by i.v. antigen exposure results in
apoptotic cell death of many reactive T cells. It is tempting to speculate that the failure of NOD
mice to undergo tolerization following i.v. Tep69 injection is related to the documented apoptosis
resistance of NOD T cells (22, 33, 226, 227). Consistent with our findings, two groups recently
reported diabetes acceleration following intrathymic injection of GAD65 peptide (225) and i.v.
injection of a mixture of GAD6YGAD67 (228).
Efforts to measure Tep69-specific T cells induced by i.v. injection were unsuccessful. In
particular, T ce11 proliferation assays with spleen or LN cells derived from CY treated rnice show
highly elevated background proliferation. CY treatment induces a drarnatic decline in lymphocyte
numbers followed by a massive wave of repopulation with lymphoproliferation, splenomegaly and
general IL-2 receptor upregulation (26,229).
In concordance with Wicker's findings, results of Our diabetic spleen ce11 titration
experiment indicate that the incidence of adoptively transferred diabetes is proportional to the
number of diabetogenic effector T cells transferred (23). Since Tep69 nearly doubled the diabetes
incidence even though the sarne number of diabetic spleen cells were transferred to al1 groups it
would be reasonable to assume that peptide exposure resulted in expansion of diabetogenic T ce11
pools. We have also demonstrated that the diabetogenic effect of i.v. Tep69 is antigen- and peptide
specific as well as boostable. We propose that a similar mechanism underlies the Tep69 effects in
CY treated mice. CY produces a quick, dramatic reduction of peripheral lymphocytes (229). This
would leave fewer Tep69-reactive T cells (memory cells) and explain why Tep69 peptide treatment
5 days before CY injection had a more pronounced effect on diabetes incidence than when Tep69
was administered 5 Qys after CY treatment. During pilot expenments we noted that administration
of higher CY doses abolishes the diabetogenic effect of CY, presumably, because al1 antigen-
reactive T cells are eliminated
Although these explanations are rational, formal proof for this model is still lacking and
would require demonstration of clonal expansion of Tep69-specific T cells, for instance, by
limiting dilution analysis. The generation of a T ce11 clone which can transfer diabetes would
further argue in favor of a diabetogenic effect for Tep69-specific T cells. Ultimately, a Tep69-
specific, TCR transgenic mouse would enable us to follow the fate of such T cells in vivo and
aliow us to test o w hypothesis in depth (230).
An alternative to Our proposed clona1 expansion hypothesis would be that high affinity
Tep69 administered i.v. induces development of TH1 type cells while lower affinity ABBOS
induces TH2 type cells (231). A pro-inflammatory TH1 biased response is associated with
destructive autoimmunity while a TH2 biased response is known to mediate protection from
diabetes (29,232,233). This model could also explain the observed difierence between i.v. Tep69
(precipitation of diabetes) and ABBOS (suppression of diabetes) treatments. A recent study by
Tisch et al. have suggested a similar mechanism to explain suppression of diabetes following i.v.
administration of GAD65 protein (234).
Extensive previous studies showed that injection of Tep69, i.v. or i.p. with incomplete
adjuvant did not precipitate spontaneous diabetes development (Karges e t al. unpublished). NOD
mouse diabetes development proceeds a slow Pace and is probably regulated inherently with a
balance of progression and retardation elernents only slightly biased towards the fonner (235).
TCR transgenic mice carrying rearranged T ce11 receptor genes of a diabetogenic T ce11 clone
BDC2.5 develop diabetes equally slowly, although cloned T cells from the same mice transfer
IDDM rapidly in NOD scid mice (230). While these mechanisms remain unclear, we submit that
they interfere with rapid IDDM development following Tep69 treatrnents, whereas disease
development in adoptive transfer and CY models does not appear subject to such regulatory
elements (23,26,236).
It has been suggested that the decision between T ce11 tolerance and T ce11 activation reflects
diverse factors such as : (1) Timing of peptide treatment; (2) route of administration;(3) dose of
peptide ; (4) and the nature of the immunogen (175,231,237,238). The present results emphasize
the importance of each of these factors on the outcome of various peptide treatment strategies.
Since systemic injection of Tep69 in neonatal mice prevented diabetes while a similar approach in
lare-stage pre-diabetic environment led to a different outcome (42). the stage in the disease process
when the immune intervention (strategy) is administered rnay also represent a very important
deterniinant.
On the whole, Our unexpected results raise a note of caution about antigen-specific
interventions to treat diabetes. Irnmunotherapy of diabetes is an exciting prospect. However, our
results caution against oversimplification of such efforts. Blanas et al.. showed that oral
administration of self-antigens can induce CDS+ cytotoxic T-cells capable of inducing
autoimmune diabetes (239). Genain et al., showed that an immunotherapeutic shift of immunity
from TH1 to a less aggressive TH2 response can protect animals in the short term, but may
exacerbate autoimrnunity later (240). These findings argue that mechanisms of autoimmunity are
complex and their understanding and long term outcornes of current strategies are important for
the development of diabetes prevention regimen.
Chapter VII. Future Directions Our first goal is to decipher the exact mechanism underlying Tep69-specific diabetes
precipitation. Identification of exact TCR a and 6 chahs involved in Tep69lABBOS mirnicry by
cloning and transfection experiments in TCR-1- hybridomas will enable us to proceed towards
generation of transgenic mice canying this TCR. This system would not only enable us to test
core aspects of the molecular mimicry model and our clonal expansion hypothesis, but also would
help further our understanding of autoimmune disease. Eventually, we plan to apply the successful
strategies from the adoptive transfer model to the spontaneous disease model to see if the same or
different mechanisms are at play. Our ultimate goal is to develop a rationai immunotherapy
strategy that c m be transferred to the human disease without undue fears of adverse complications.
Our laboratory is very interactive, and 1 becarne involved more peripherally in several other
diabetes projects. These will not be detailed to here. However, working together with a summer-
and a then a project student, 1 generated an ICA69 DNA-based expression plasmid that has just
been demonstrated to work well when ttansfected into cos-7 cells in vitro (not shown). And it has
shown promise as DNA vaccine: when adrninistered to NOD rnice intradermally we found
considerable anti-ICA69 antibody responses (not shown). This route of administration is usually
associated with a TH2 bias and we have hopes that this regirnen may delay or prevent IDDM, as is
has shown such promise in EAE (241). Since my data clearly demonstrate the possible dangers of
high-dose peptide immunotherapy which may precipitate rather than deviate autoimmune disease,
the DNA vaccination approach has gained interest in Our view. Unfortunately 1 will not be able to
continue this work and determine if the long-term expression typicaI for DNA vaccines (242)
leads to modification of disease course. When this effort is completed, 1 will be a CO-author on
resulting publications, since 1 constnicted the vector and was principally involved in its initial
c haractenzation.
References 1. Bach, J.F. 1994. Insulin-dependent diabetes mellitus as an autoimmune disease. Endocrine Rev 15516-542.
2. Sodeman, W.A., and T.A. Sodeman, editors. 1985. Sodeman 's Pathologic Physiology: rnechanisms of disease. 7 ed. Sodeman's Pathologic Physiology: mechanisms of disease. Edited by W.A. Sodeman. W. B. Saunders, Philadelphia.
3. Becker, K.L., editor. 1990. Principles and Pructice of Endocrinology and Metabolim. Edited by R.C. G.S.A.K. Eisenbarth. Lippincott Company, Philadelphia.
4. Todd, J.A. 1996. Transcribing diabetes. Nature 384:407-408.
5. Akerblom, H.K., and M. Knip. 1998. Putative Environmental Factors in type 1 Diabetes. Diabetes Metab. Rev. 143 1-67.
6. Karges, W., J. Ilonen, B.H. Robinson, and H.-M. Dosch. 1995. Self and Non-Self Antigen in Diabetic Autoimmunity: Molecules and Mechanisms. Molec Aspects Med 16:79-2 13.
7. Castano, L., and G.S. Eisenbanh. 1990. Type-1 Diabetes: A chronic autoimmune disease of human, mouse, and rat. Ann. Rev. Immunol. 8647-679.
8. Todd, I.A., J.I. Bell, and H.O. McDevitt. 1988. A molecular basis for genetic susceptibility to insulin-dependent diabetes mellitus. meview] [3 1 refs]. Trenuk Genet 4, no. 5: 129-34.
9. Lipton, R.B., M. Kocova, R.E. LaPorte, J.S. Dorman, T.J. Orchard, W.J. Riley, A.L. Drash, D.J. Becker, and M. Trucco. 1992. Autoirnmunity and genetics contribute to the risk of insulin-dependent diabetes rnellitus in families: islet ce11 antibodies and HLA DQ heterodimers. Amer J Epidemiol 136, no. 5:5O3- 12.
IO. American Diabetes Association. 1998. Cost of Diabetes on the rise. Professional Secrion Quarterly Winter 19985
1 1. Metroz-Dayer, M.D., A. Mouland, C. Brideau, D. Duhamel, and P. Poussier. 1990. Adoptive transfer of diabetes in BB rats induced by CD4 T lymphocytes. Diabetes 39, no. 8:928-32.
12. Edouard, P., J.C. Hiserodt, C. Plamondon, and P. Poussier. 1993. CDS+ Tsells are required for adoptive transfer of the BB rat diabetic syndrome. Diabetes 42, no. 3:390-7.
13. Nakhooda, A.F., A.A. Like, C.I. Chappel, F.T. Murray, and E.B. Marliss. 1977. The spontaneously diabetic Wistar rat. Metabolic and morphologie studies. Diabetes 26. no. 2: 100-12.
14. Guberski, D.L., L. Butler, S.M. Manzi, M. Stubbs, and A.A. Like. 1993. The BBZ(Wor rat: clinical characteristics of the diabetic syndrome. Diabe~ologia 36, no. 10:912-9.
15. Bach, J.F. 1988. Mechanisms of autoimrnunity in insulin-dependent diabetes mellitus. [Review] [77 refs]. Clin Exp Immunol 72, no. 1: 1-8.
16. Poussier, P., A.F. Nakhooda, J.A. Falk, C. Lee, and E.B. Marliss. 1982. Lymphopenia and abnormal lymphocyte subsets in the "BB" rat: relationship to the diabetic syndrome. Endocrinol 1 10: 1825- 1827.
17. Greiner, D.L., E.S. Handler, K. Nakano, J.F. Mordes, and A.A. Rossini. 1986. Absence of the RT-6 T ce11 subset in diabetes-prone BB/W rats. J Immun01 136, no. 1: 148-5 1.
18. Mordes, J.P., D.L. Gailina, E.S. Handler, D.L. Greiner, N. Nakamura, A. Pelltier, and A.A. Rossini. 1987. Transfusions enriched for W3125 helperlinducer T lymphocytes prevents spontaneous diabetes in the BB/W rat. Diabetologia 30: 22-29.
19. Kikutani, H., and S. Makino. 1992. The murine autoimmune diabetes model: NOD and related strains. meview] [174 refs]. Adv Immunol 5 1 :285-322.
20. Baxter, A.G., M. Koulmanda, and T.E. Mandel. 1991. High and low diabetes incidence nonobese diabetic (NOD) mice: origins and characterisation. Autoimmwti~ 9, no. l:6 1-7.
2 1. Pozzilli, P., A. Signore, A.J. Williams, and P.E. Beales. 1993. NOD mouse colonies around the world: recent facts and figures. Immunol Today 14: 193-196.
22. Colucci, F., C.M. Cilio, K. Lejon, C.P. Gonçalves, M.L. Bergman, and D. Holmberg. 1996. Programrned cell &atb in the pathogenesis of murine IDDM: resistance to apoptosis induced in lymphocytes by c yclophosphamide. J Autoimmun 9, no. 2:27 1-6.
23. Wicker, L.S., B.J. Miller, and Y. Mullen. 1986. Transfer of autoimmune diabetes mellitus with spienoc ytes from nonobese diabetic (NOD) miçe. Diabetes 35, no. 8: 855-60.
24. Christianson, S.W., L.D. Shultz, and E.H. Leiter. 1993. Adoptive transfer of diabetes into irnmunodeficient NOD-scidscid rnice. Relative contributions of CD4+ and CD8+ T-cells from diabetic versus prediabetic NOD.NON-Thy- la donors. Diabetes 42, no. 144-55.
25. Bendelac, A., C. Carnaud, C. Boitard, and J.F. Bach. 1987. Syngeneic transfer of autoimmune diabetes from diabetic NOD mice to healthy neonates. J Erp Med 166:823-832.
26. Harada, M., and S. Makino. 1984. Promotion of spontaneous diabetes in non-obese diabetic- prone rnice by c yclophospharnide. Diaberolog. 27:604-606.
27. Charleton, B., A. Bacelj, R.M. Slattery, and T.E. Mandel. 1989. Cyclophosphamide-induced diabetes in NODMrehi mice: evidence for suppression in spontaneous autoimmune diabetes mellitus. Diabetes 38:441-447.
28. Wicker, L.S., B.J. Miller, L.Z. Coker, S.E. McNally, S. Scott, Y. Mullen, and MC. Appel. 1987. Genetic control of diabetes and insulitis in the nonobese diabetic (NOD) mouse. J. Exp. Med. 165: 1639- 16%.
29. Rothe, H., A. Faust, U. Schade, R. Kleemann, G. Bosse, T. Hibino, S. Martin, and HI Kolb. 1994. Cyclophosphamide treatment of femaIe non-obese diabetic mice causes enhanced expression of inducible nitric oxide synthase and interferon-gamma, but not of interleukin-4. Diabetologia 37, no. 1 1 : 1 154-8.
30. Rothe, H., V. Burkart, A. Faust, and H. Kolb. 1996. Interleukin-12 gene expression is associated with rapid development of diabetes mellitus in non-obese diabetic mice. Diabetologia 39, no. 1 : 1 19-22.
31. Kay, T.W., I.L. Campbell, and L.C. Harrison. 1991. Characterization of pancreatic T lymphocytes associated with beta ce11 destruction in the non-obese diabetic (NOD) mouse. Journal of Autoirnmunity 4, no. 2:263-76.
32. Yasunarni, R., and J.-F. Bach. 1988. Anti-suppressor effect of cyclophosphamide on the development of spontaneous diabetes in NOD mice. Eur. J. Immunol. 18:481-484.
33. Colucci, F., M.L. Bergman, G.C. Penha, C.M. Cilio, and D. Holmberg. 1997. Apoptosis resistance of nonobese diabetic peripheral lymphocytes linked to the IddS diabetes suscepti bili ty region. Proc Natl Acad Sci U S A 94, no. 16: 8670-4.
34. Trembleau, S., G. Penna, S. Gregori, M.K. Gately, and L. Adonni. 1997. Deviation of pancreas-infiltrating cells to Th2 by interleukin-12 antagonist administration inhibits autoimmune diabetes. Eur J Immunol27, no. 9:2330-9.
35. Davies, J., Y. Kawagushi, S.T. Bennett, J.B. Copeman, H.J. Cordell, L.E. Ritchard, P.W. Reed, S.C.L. Gough, S.C. Jenkins, S.M. Palmer, K.M. Balfour, B.R. Rowe, M. Farrall, A.H. Barnett, S.C. Bain, and J.A. Todd. 1994. A genome-wide search for human type 1 diabetes susceptibility genes. Nature 37 1: 130-136.
36. Wicker, L.S., LA. Todd, and L.B. Peterson. 1995. Genetic control of autoimmune diabetes in the NOD mouse. Annu Rat lmmunol 13, no. 179: 179-200.
37. Parïsh, N.M., P.R. Hutchings, L. O'Reilly, R. Quartey-Papafïo, D. Healey, P. Ozegbe, and A. Cokke. 1995. Tolerance induction as a therapeutic strategy for the control of autoimmune endocrine disease in mouse models. Immuno l Rev 144269-300.
38. Bowman, M.A., E.H. Leiter, and M.A. Atkinson. 1994. Prevention of diabetes in the NOD mouse: implications for therapeutic intervention in human disease. lmmunol Today 15, no. 3: 1 15-20.
39. Tisch, R., X.D. Yang, S.M. Singer, R.S. Liblau, L. Fugger, and H.O. McDevitt. 1993. Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. N m r e 366:72-75.
40. Kaufman, D.L., M. Clare-Salzler, J. Tian, T. Fortshuber, G.S.P. Ting, P. Robinson, M.A. Atkinson, E. Sercarz, A.J. Tobin, and P.V. Lehman. 1993. Spontaneous loss of T cell tolerance to glutamic acid decarboxylase in munne insulin dependent diabetes. Nature 366:69-72.
41. Elliott, J.F., H.Y. Qin, S. Bhatti, R. Singh, D.K. Smith, J. Lauzon, and B. Singh. 1994. Immunization with the large isoform of mouse glutamic acid decarboxylase (GAD67) prevents autoimmune diabetes in NOD rnice. Diabetes 43: l494- 1499.
42. Karges, W., D. Harnmond-McKibben, R. Gaedigk, N. Shibuya, R. Cheung, and H.-M. Dosch. 1997. Loss of self-tolerance to ICA69 in non-obese diabetic mice. Diabetes 46: 1548- 1556.
43. Dosch, H.-M., J.M. Martin, B.M. Robinson, H. Akerblom, and J. Karjalainen. 1993. An Irnrnunological Basis for Disproportionate Diabetes Risks in Children with a Type 1 Diabetic Mother or Father. Diabetes Care 16:949-95 1.
44. Rotter, J.L, and E.M. Landaw. 1984. Measuring the genetic contribution of a single locus to a multilocus disease. Clin Genet 26, no. 6529-42.
45. Singal, D.P., and M.A. Blajchman. 1973. Histocompatibility (HL-A) antigens, lymphocytotoxic antibodies and tissue antibodies in patients with diabetes mellitus. Diabetes 22, no. 6:429-32.
46. Gottlieb, M.S., and H.F. Root. 1968. Diabetes melitus in twins. Diabetes 17:693-704.
47. Barnett, A.H., C. Eff, and R.D.G. Leslie. 1981. Diabetes in identical twins. Diabetologia 20: 87-93.
48. Michelsen, B., R. Wassmuth, J. Ludvigsson, A. Lemmark, G.T. Nepom, and L. Fisher. 1990. HLA heterozygosity in insuiin-dependent diabetes is rnost frequent at the DQ locus. Scand J Immun013 1, no. 4:405-13.
49. Thonby, E., and K.S. Ronningen. 1993. Particular HLA-DQ molecules play a dominant role in determining susceptibility or resistance to type 1 (insulin-dependent) diabetes mellitus. Diabetologia 36, no. 537 1-377-
50. Todd, I.A., J.I. Bell, and H.O. McDevitt. 1987. HLA-DQ$ gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Namre 329599-604.
5 1. Platz, P., B.K. Jakobsen, N. Morling, L.P. Ryder, A. Svejgaard, M. Thomsen, M. Christy, H. Krornann, J. Benn, J. Nerup, A. Green, and M. Hauge. 1981. HLA-D and -DR antigens in genetic analysis of insulin dependent diabetes mellitus. Diabetologia 21, no. 2: 108- L 5.
52. Caillat-Zucman, S., H.J. Garchon, J. Timsit, R. Assan, C. Boitard, S.I. Djilali, P. Bougneres, and J.F. Bach. 1992. Age-dependent HLA genetic heterogeneity of type 1 insulin-dependent diabetes mellitus. Journal of Clinical Investigation 90, no. 6:2242-50.
53. Morel, P.A., J.S. Doman, J.A. Todd, H.O. McDevitt, and M. Tnicco. 1988. Aspartic acid at position 57 of the HLA-DQ beta chain protects against type 1 diabetes: a family study [published erratum appears in Proc Nat1 Acad Sci U S A 1989 Feb;86(4):1317]. Proc Nat1 Acad Sci OS A 85, no. 21:8lll-S.
54. Lund, T., L.O. OTReilly, P. Hutchings, O. Kanagawa, E. Simpson, R. Gravely, P. Chandler, J. Dyson, J.K. Picard, A. Edwards, D. Kioussis, and A. Cooke. 1990. Revention of insulin- dependent diabetes mellitus in non-obese diabetic mice by transgenes encoding modified I- AP chah or normal I-E a c h a h Namre 345:727-729.
55. Wicker, L.S., M.C. Appel, F. Dotta, A. Pressey, B.I. Miller, N.H. DeLarato, P.A. Fischer, R.J. Boltz, and L.B. Peterson. 1992. Autoimmune syndromes in major histocompatibility complex (MHC) congenic strains of nonobese diabetic (NOD) rnice. The NOD MHC is dominant for insulitis and cyclophosphamide-induced diabetes. Journal of Experimental Medicine 176, no. 157-77.
56. Prochazka, M., D.V. Serreze, S.M. Worthern, and E.H. Leiter. 1989. Genetic control of diabetogenesis in NODILt mice. Diaberes 38: 1446-1455.
57. Lee, M.-S., R. Mueller, L.S. Wicker, L.B. Peterson, and N. Sarvetnick. 1996. IL-10 is necessary and sufficient for autoimmune diabetes in conjunction with NOD MHC homozygosity. J. Exp. Med. 183:2663-2668.
58. Brown, J.H., T.S. Jardetzky, J.C. Gorga, L.J. Stem, R.G. Urban, J.L. Strorninger, and D.C. Wiley. 1993. Three-dimensional structure of the human class II histocompatibility antigen HLA-DR 1 [see comments]. Nature 364, no. 6432:33-9.
59. Wicker, L.S., J.A. Todd, and L.B. Peterson. 1995. Genetic control of autoimmune diabetes in the NOD mouse. Ann Rev Immun01 13: 179-200.
60. Tisch, R., and H.O. McDevitt. 19%. Insulindependent diabetes mellitus. Cell85:291-297.
61. McAleer, M.A., P. Reifsnyder, S.M. Palmer, M. Prochazka, J.M. Love, J.B. Copeman, E.E. Powell, N.R. Rodrigues, J.B. Pnns, D.V. Serreze, N.H. DeLarato, L.S. Wicker, L.B. Peterson, N.J. Schorl, J.A. Todd, and E.H. Leiter. 1995. Crosses of NOD rnice with the related NON strain - A polygenic mode1 for IDDM. Diabetes 44: 1186-1 195.
62. Bell, G.I., S. Horita, and J.H. Kararn. 1984. A polymorphic locus near the human insulin gene is associated with insulin-dependent diabetes mellitus. Diabetes 33: 1504- l5O9.
63. Lucassen, A.M., C. Julier, J.P. Beressi, C. Boitard, P. Froguel, M. Lathrop, and J.I. Bell. 1993. Susceptibility to insulin dependent diabetes mektus maps to a 4.1 kb segment of DNA spanning the insulin gene and associated VNTR. Nahtre Genetics 4, no. 3:305-10.
64. Drash, A.L. 1990. What do epiderniologic observations tell us about the etiology of insulin dependent diabetes mellitus? Schweiz Med Wochenschr 120, no. 3:3945.
65. Tull, E., N. Tajima, T. Podar, J. Dorman, C. Moy, and R. Morte . 1992. Epidemics, migrants and the death of pancreas. Zn Epidemiology and Etiology of Insulin-Dependent Diabetes Mellitus in the Young, vol. 21. C. Levy-Marcha1 and P. Czemichow, editors. Karger, Basel. 56-65.
66. Staines, A., S. Hanif, S. Ahmed, P.A. McKinney, S. Shera, and H.J. Bodansky. 1997. Incidence of insulin dependent diabetes mellitus in Karachi, Pakistan. Arch Dis Clzild 76, no. 2121-3.
67. Todd, J.A., T.J. Aitrnan, R.J. Comall, S. Ghosh, J.R. Hall, C.M. Herne, A.M. Knight, J.M. Love, M.A. McAleer, and J.B. Prins. 1991. Genetic analysis of autoimune type 1 diabetes mellitus in mice. Nature 35 1, no. 6327542-7.
68. Conrad, B., E. Weidrnann, G. Trucco, W.A. Rudert, R. Behboo, C. Ricordi, H. Rodriquez- Rilo, D. Finegold, and M. Tmcco. 1994. Evidence for superantigen involvement in IDDM aetiology. Nature 37 1:35 1-355.
69. Conrad, B., R.N. Weissmahr, J. Boni, R. Arcari, J. Schupbach, and B. Mach. 1997. A human endogenous retroviral superantigen as candidate autoimmune gene in type 1 diabetes. Ce11 90, no. 2303-13.
70. Karjalainen. J., J.M. Martin, M. Knip, J. Ilonen, B.H. Robinson, E. Savilahti, H.K. Akerblom, and H.-M. Dosch. 1992. A bovine albumin peptide as a possible trigger of insulin dependent diabetes meIlitus. N Engl J Med 327, no. 5302-307.
71. Dosch, &M., J. Karjalainen, J. Morkowski, LM. Martin, and B.H. Robinson. 1992. Nutritional triggers of IDDM. In Epidemiology and etiology of Insulin-dependent diabetes in the Young, vol. 21. C. Lévy-Marchai and P. Czemichow, editors. Karger, Basel. 202-2 17.
72. Elliott, R.B., and J.M. Martin. 1984. Dietary protein: a trigger of insulin-dependent diabetes in the BB rat? Diabetologia 26, no. 297:297-299.
73. Daneman, D., L. Fishman, C. Clarson, and J.M. Martin. 1987. Dietary triggers of insulin- dependent diabetes in the BB rat. Diabetes Res 5, no. 93:93-97.
74. Martin, J.M., B. Trink, D. Daneman, H.-M. Dosch, and B.H. Robinson. 1992. Milk proteins in the etiology of insulin-dependent diabetes mellitus (IDDM). Ann Med 23447-452.
75. Dahl-Jgrgensen, K., G. Joner, and K.F. Hanssen. 1991. Relationship between cow milk consumption and incidence of IDDM in childhood. Diabetes Cure 14, no. 11: 1081-1083.
76. Scott, F.W. 1990. Cow rnilk and insulin-dependent diabetes mellitus: is there a relationship. Am. J. Nutrition 5 1 :489-49 1.
77. Gerstein, H. 1994. Cow's milk exposure and type 1 diabetes mellitus. Diabetes Cure 17:13- 19.
78. Glemm, M., B.H. Robinson, and LM, Martin. 1989. Could bovine serum alburnin be the initiating antigen ultimately responsible for the development of insulin dependent diabetes mellitus? Diaberes Res 10: 103- 107.
79. Karjalainen, J., J. Martin, J. KNp, J. Ilonen, B. Robinson, E. Savilahti, H. Akerblom, and H. Dosch. 1992. A bovine aibumin peptide as a possible tngger of insulin-dependent diabetes mellitus. N Engl J Med 327:302-307.
80. Saukkonen, T., E. Savilahti, O. Vaarala, E.T. Virtale, J. Tuomilehto, H.K. Akerblom, and and the Childhood Diabetes in Finland Study Group. 1994. Children with newly diagnosed insulin-dependent diabetes mellitus have increased levels of antibodies to bovine serum albumin but not to ovalbumin. Diabetes Care 17:970-976.
81. Harnmond-McKibben, D., and H--M. Dosch. 1997. Cow rnilk, BSA and IDDM: can we senle the controversies? Diab Cure 20:897-901.
82. Beppu, H., WE. Winter, M.A. Atkinson, N.K. Maclaren, K. Fujita, and H. Takahashi. 1987. Bovine alburnin antibodies in NOD rnice. Diaberes Res 657-69.
83. Perez-Bravo, F.. E. Carrasco, MD. Gutierrez-Lopez, M.T. Martinez, G. Lopez, and M. Garcia de los Rios. 1996. Genetic predisposition and environmental factors leading to the development of insulin-dependent diabetes mellitus in Chilean children. J Mol Med 74: 105- 109.
84. Vaarala, O., T. Saukkonen, E. Savilahti, T. Klemola, and H.K. Akerblom. 1995. Development of immune response to cow's milk proteins in infants receiving cow's milk or hydrolyzed formula. J AZZerg Clin Immun01 96:9 17-923.
85. Saukkonen, T., S.M. Virtanen, M. Karppinen, H. Reijonen, J. Ilonen, L. Rasanen, H.K. Akerblom, E. Savilahti, and a.t.C.D.i.F.S. Group. 1998. Significance of cow's Mlk protein antibodies as nsk factor for childhood IDDM: interactions with dietary cow's milk intake and HLA-DQB i genotype. DiabetoZogia 4 1 :72-78.
86. Akerblom, H.K., H.-M. Dosch, B.H. Robinson, M. Knip, J. Karjalainen, E. Savilahti, J. Ilonen, S.M. Virtanen, A. Reunanen, T. Saukkonen, and J.M. Martin. 1993. 1s dietary intervention for the prevention of insulin-dependent diabetes mellitus feasible? In Pediatr Adolesc Endocnn. vol. 23.2. Laron, editor. 23 vols. Karger, Basel. 97-104.
87. Akerblom, H. 1993. Aspects of intervention trials for the prevention of Type 1 (insulin- dependent) diabetes mellitus. Diabetologia 365 1-53.
88. Akerblom, H.K., E. Savilahti, T.T. Saulckonen, A. Paganus, S.M. Virtanen, K. Teramo. M. Knip, J. Ilonen, H. Reijonen, J. Karjalainen, O. Vaarala, and A. Reunanen. 1993. The case for elirnination of cow's milk in early infancy in the prevention of Type 1 diabetes: the Finnish experience. Diabetes Metab Rev 9:269-278.
89. Larger, E., C. Becourt, J.F. Bach, and C. Boitard. 1995. Pancreatic islet beta cells drive T cell- immune responses in the nonobese diabetic mouse model. J Exp Med 18 1, no. 5: 1635-42.
90. Karges, W., R. Gaedigk, M.F. Hui, R.K. Cheung, and H.-M. Dosch. 1997. Molecular Cloning of Murine ICA69: Diabetes-prone Mice Recognize the Human Autoirnmune- Epitope, Tep69, Conserved in Splice Variants from both Species. Biochim Biophys Acta 1360, no. 2:97-101.
9 1. Gaedigk, R., W. Karges, M.F. Hui, and H.-M. Dosch. 1996. Genomic organization of human and mouse ICAp69, a target antigen in in diabetic autoimmunity. Genomics 38:382-391.
92. Karges, W., M. Pietropaolo, C. Ackerley, and HM. Dosch. 1996. Gene expression of islet ce11 antigen p69 (ICAp69) in man, mouse and rat. Diabetes 455 13-521.
93. Miyazaki, I., R.K. Cheung, R. Gaedigk, M.F. Hui, I. Van der Meulen, R.V. Rajotte, and H.- M. Dosch. 1995. T ce11 activation and anergy to islet ce11 antigen in type 1 diabetes. J Immurzol 1%: 1461-1469.
94. Miyazaki, I., R. Gaedigk, M.F. Hui, R.K. Cheung, J. Morkowski, R-V. Rajotte, and H.-M. Dosch. 1994. Cloning of human and rat p69, a candidate autoimmune target in Type I diabetes. Biochim Biophys Acta 1227: 101-104.
95. Gaedigk, R., A,M.V. Duncan, 1. Miyazaki, B.H. Robinson, and H.-M. Dosch. 1994. ICA1 encoding p69, a protein linked to the development of type 1 diabetes mellitus, maps to chromosome 7p22. Cytogen Ce11 Genet 66:274-276.
96. Cheung, R.K., J. Karjalainen, J. VanderMeulen, D. Singal, and H.-M. Dosch. 1994. T cells of children with insulin dependent diabetes are sensitized to bovine serum albumin. Scand. J. Immunol. 40623-628.
97. Karges, W., D. Hammond-McKibben, R.K. Cheung, M. Visconti, N. Shibuya, D. Kemp, and H.M. Dosch. 1997. lrnmunological Aspects of Nutritional Diabetes Prevention in NOD Mice. A Pilot Study for the Cow's Milk-Based IDDM Prevention Trial. Diabetes 46557- 564.
98. Akerblom, H.K., E. Savilahti, M. Knip, J. Vandermeulen, and H.-M. Dosch. 1996. Early nutrition and type 1 diabetes. In Long-term consequences of early feeding, vol. 36. J. Boulton, editoï. 36 vols. Lippincott-Raven, Philadelphia. 207-224.
99. Ohashi, P.S., S. Oehen, K. Buerki, H. Pircher, C.T. Ohashi, B. Odermatt, B. Malissen, R.M. Zinkemagel, and H. Hengartner. 1991. Ablation of "tolerance" and induction of diabetes by virus infection in viral antigen transgenic mice. Cell65, no. 2:305-17.
100. Kawai, K., and P.S. Ohashi. 1995. LmmunologicaI function of a defined T cell popdation tolerized to low-affinity self antigen. Naîure 374:68-69.
10 1. Malkani, S., D. Nompleggi, J.W. Hansen, DL. Greiner, J.P. Mordes, and A.A. Rossini. 1997. Dietary cow's milk protein does not alter the frequency of diabetes in the BB rat. Diabetes 46, no. 7: 1 133-40.
102. Schloot, N.C., B.O. Roep, D.R. Wegmann, L. Yu, T.B. Wang, and G.S. Eisenbarth. 1997. T- ce11 reactivity to GAD65 peptide sequences shared with coxsackie virus protein in recent- onset IDDM, post-onset IDDM patients and control subjects. Diabetologia 40, no. 3:332-8.
103. Endl, J., H. Otto, G. Jung, B. Dreisbusch, F. Donie, P. Stahl, R. Elbracht, G. Schrnitz, E. Meinl, M. Hummel, A.G. Ziegler, R. Wank, and D.J. Schendel. 1997. Identification of naturally processed T ce11 epitopes from glutamic acid decarboxylase presented in the context of HLA-DR alleles by T lymphocytes of recent onset IDDM patients. J Clin Invest 99, no. 10:2405-15.
104. Lampeter, E.F., M. Homberg, K. Quabeck, U.W. Schaefer, P. Wernet, J. Bertrams, W.H. Grosse, F.A. Gries, and H. Kolb. 1993. Transfer of insulin-dependent diabetes between
HLA-identical siblings by bone marrow transplantation [see comments]. Loncet 341, no. 8855: 1343-4.
105. Vialettes, B., and D. Maraninchi. 1993. Transfer of insulindependent diabetes between HLA- identical siblings by bone marrow transplantation. Lancer 342, no. 8864.
106. Mon, Y., M. Suko, H. Okudaira, 1. Matsuba, A. Tsuruoka, A. Sasaki, H. Yokoyarna, T. Tanase, T. Shida, M. Nishimura, and et ai. 1986. Preventive effects of cyclosporin on diabetes in NOD mice. Diabetologia 29, no. 4:244-7.
107. Feutren, G., L. Papoz, R. Assan, B. Vialettes, G. Karsenty, P. Vexiau, H. Du Rostu. M. Rodier, J. Sirmai, A. Lallemand, and et al. 1986. Cyclosporin increases the rate and length of remissions in insulin-dependent diabetes of recent onset. Results of a multicentre double- blind trial. h c e t 2, no. 84991 19-24.
1 08. Bac hl J.F. 1993. S trategies in immunotherapy of insulin-dependent diabetes mellitus. Annals of the New York Academy of Sciences 696, no. 364:364-76.
109. Gepts, W., and P.M. Lecompte. 1981. The pancreatic islets in diabetes. Am J Med 70:105- 115.
110. Hanninen, A., S. Jalkanen, M. Saimi, S. Toikkanen, G. Nikolakaros, and 0. Simell. 1991. Macrophages, T ce11 Rceptor usage, and endotheliai ceil activation in the pancreas at the onset of insulin-dependent diabetes mellitus. Journal of Clinical investigation 90, no. 5: 1901-10.
1 1 1. Hanafusa, T., A. Miyazaki, J. Miyagawa, S. Tarnura, M. Inada, K. Yarnada, Y. Shinji, H. Katsura, K. Yamagata, N. Itoh, and et ai. 1990. Examination of islets in the pancreas biopsy specimens from newly diagnosed type 1 (insulin-dependent) diabetic patients. Diabetologia 33, no. 2: 105-1 1.
112. Landin-Olsson, M., F.A. Karlsson, A. Lemmark, and G. Sundkvist. 1992. Islet ce11 and thyrogastric antibodies in 633 consecutive 15- to 34-yr-old patients in the diabetes incidence study in Sweden. Diabetes 41, no. 8: 1022-7.
113. Bianchi, G., P. Montanari, A. Fabbti, A. Gamberini, M. Zoli, and G. Marchesini. 1995. Thyroid volume in type 1 diabetes patients without overt thyroid disease. Acta Diabetol 32:49-52.
114. Rensch, M.J., J.A. Memitch, M. Lieberman, B.D. Lang, D.R. Davis, and P.R. Magnelly. 1996. Gluten sensitive enteropathy in patients with insulin-dependent diabetes mellitus. Ann Int Med l24:564-567.
115. Shimada. A., B. Charlton, E.C. Taylor, and C.G. Fathman. 1996. Beta-cell destruction may be a late consequence of the autoimmune process in nonobese diabetic mice. Diabetes 45, no. 8: 1063-7.
1 16. Palmer, J.P., C.M. Asplin, P. Clemons, K. Lyen, O. Tatpati, P.K. Raghu, and T.L. Paquette. 1983. Insulin antibodies in insulin-dependent diabetics before insulin ueatment. Science 222: 1337-1339.
117. Kuglin. B., 1. Rjasanowski, J. Bertrams. F.A. Gries, H. Kolb, and D. Michaelis. 1990. Antibodies to proinsulin and insulin as predictive markers of type 1 diabetes. Diabetic Medicine 7, no. 4:3 104.
118. Baekkeskov, S., H.J. Aanstoot, S. Christgau, A. Reee, M. Solimena, M. Cascalho, F. Folli, H. Richter-Olesen, and P. De Camilli. 1990. Identification of the 64K autoantigen in IDDM as the GAB A-synthesizing enzyme glutamic acid decarboxylase. Nature 347: 15 1 - 156.
1 19. Rabin, DU., S.M. Pleasic, LA. Shapiro, H. Yoo-Warren, J. Oles, J.M. Hicks, D.E. Goldstein, and P.M. Rae. 1994. Islet ce11 antigen 512 1s a diabetes-specific islet autoantigen related to protein tyrosine phosphatases. J Immun01 1523 183-3 188.
120. Pietropaolo, M., L. Castano, S. Babu, R. Buelow, Y.-L. Kuo, S., S. Martin, A. Martin, A.C. Powers, M. Prochazka, J. Naggen, E.H. Leiter, and G.S. Eisenbarth. 1993. Islet ce11 autoantigen 69 k D (ICA69). Molecular cloning and characterization of a novel diabetes- associated autoantigen. J Clin Invest 92:359-37 1 .
121. Roep, B.O., G. Duinkerken, G.M. Schreuder, H. Kolb, V.R. de, and S. Martin, 1996. HLA- associated inverse correlation between T ceIl and antibody responsiveness to islet autoantigen in recent-onset insulin-dependent diabetes mellitus. Eur J Immun01 26, no. 6: 1285-9.
122. Landin, O.M., J.P. Palmer, A. Lemmark, L. Blom, G. Sundkvist, L. Nystrom, and G. Dahlquist. 1992. Predictive value of islet ce11 and insulin autoantibodies for type 1 (insulin- dependent) diabetes mellitus in a population-based study of newly-diagnosed diabetic and matched control children. Diabetologia 35, no. 1 1 : 1068-73.
123. Atkinson, M.A., and N.K. Maclaren. 1993. Islet ce11 autoantigens in insulin-dependent diabetes. J Clin Invest 92, no. 4:1608-16.
124. Serreze, DY., H.D. Chapman, D.S. Vamum, M.S. Hanson, P.C. Reifsnyder, S.D. Richard, S.A. Fleming, E.H. Leiter, and L.D. Shultz. 1996. B lymphocytes are essential for the initiation of T cell-mediated autoimmune diabetes: analysis of a new "speed congenic" stock of NOD.Ig mu nul1 mice. J Exp Med 184, no. 52049-53.
135. Bottazzo, G.F., B.M. Dean, J.M. McNally, E.H. MacKay, P.G. Swift, and D.R. Garnble. 1985. In situ characterization of autoimmune phenornena and expression of HLA molecules in the pancreas in diabetic insulitis. N Engl J Med 313, no. 6:353-60.
126. Chatenoud, L., G. Feutren, D.L. Nelson, C. Boitard, and J.F. Bach. 1989. Effect of cyclosporin on interleukin Zreiated T-lymphocyte parameters in IDDM patients. Diabetes 38, no. 2:249-56.
127. Karlsson-Parra, A., M. Kobbah, U. Ewald, T. Tuvemo, U. Forsum, and L. Klareskog. 1988. Circulating class II transplantation antigen-expressing T lymphocytes in children with insulin-dependent diabetes mellitus at diagnosis. Acta Paediatr Scand 77, no. 4:554-8.
138. Jackson, R.A., M. A. Morris, B.F. Haynes, and G.S. Eisenbarth. 1982. Increased circulating Ia-antigen-bearing T cells in type 1 diabetes mellitus. N Engl J Med 306, no. 13:785-8.
129. Atkinson, M.A., D.L. Kaufman, L. Campbell, K.A. Gibbs, S.C. Shah, D.F. Bu, M.G. Erlanger, A.J. Tobin, and N.K. Maclaren. 1992. Response of peripheral-blood mononuclear cells to glutamate decarboxylase in insulindependent diabetes. h c e t i:458459.
130. Atkinson, M.A., M.A. Bowman, L. Campbell, B.L. D m w , D.L. Kaufrnan, and N.K. MacLaren. 1994. Cellular immunity to a determinant common to glutamate decarboxylase and coxsackie vms in insulin-dependent diabetes. J. Clin. Invest, 94:2 125-2 129.
131. Harrison, L.C., H. De Aizpurua, T. Loudovaris, I.L. Campbell, J.S. Cebon, B.D. Tait, and P.G. Colman. 1991. Reactivity to human islets and fetal pig proislets by peripheral blood
mononuclear cells from subjects with preclinicai and clinical insulin-dependent diabetes. Diabetes 40, no. 9: 1128-33.
132. Harrison, L.C., S.X. Chu, H.J. DeAizpuma, M. Graham, MC. Honeyman, and P.G. Colman. 1992. Islet-reactive T cells are a marker of preclinicai insulin-dependent diabetes. J Clin Invest 89, no. 4: 1 161-5,
133. Lohmann, T., R.D. Leslie, M. Hawa, M. Geysen, S. Rodda, and M. Londei. 1994. hunodominant epitopes of glutamic acid decarboxylase 65 and 67 in insulin-dependent diabetes mellitus. h c e t 343, no. 8913: 1607-8.
134. Hummel, M., B.I. Durinovic, and A.G. Ziegler. 1996. Relation between cellular and humord immunity to islet ce11 antigens in type 1 diabetes. J Autoimmun 9, no. 3:427-30.
135. Roep, B.O., A.A. Kallan, G. Duinkerken, S.D. Arden, J.C. Hutton, G.J. Bniining, and R.R. de Vries. 1995. T-ceII reactivity to beta-ceIl membrane antigens associated with beta-ce11 destruction in IDDM. Diabetes 44, no. 3:278-83.
136. Durinovic, B.I., M. Hummel, and A.G. Ziegler. 1996. Cellular immune response to diverse islet ce11 antigens in DDM. Diabetes 45, no. 6:795-800.
137. Kawasaki, E., G.S. Eisenbarth, C. Wasmeier, and J-C. Hutton. 1996. Autoantibodies to protein tyrosine phosphatase-like proteins in type 1 diabetes. Overlapping specificities to phogrin and ICA5 l m - 2 . Diabetes 45, no. 10: 1344-9.
138. Brooks-Worrell, B.M., G.A. Starkebaum, C. Greenbaum, and J.P. Palmer. 1996. Peripheral blood mononuclear cells of insulin-dependent diabetic patients respond to multiple islet ce11 proteins. J Immun01 157, no. 125668-74.
139. Gelber, C., L. Paborsky, S. Singer, D. McAteer, R. Tisch, C. Jolicoeur, R. Buelow, H. McDevitt, and C.G. Fathman. 1994. Isolation of nonobese diabetic mouse T-cells that recognize novel autoantigens involved in the early events of diabetes. Diabetes 43, no. 1:33- 9.
140. Thivolet, C., A. Bendetac, P. Bedossa, J.-F. Bach, and C. Carnaud. 1991. CD8+ T ce11 homing to the pancreas in the nonobese diabetic mouse is CD4+ T cell-dependent. J. Irnmunol. 146%-88.
141. Hutchings, P.R., E. Simpson, L.A. O'Reilly, T. Lund, H. Waldmann, and A. Cooke. 1990. The involvement of Ly2+ T cells in beta ce11 destruction. J Autoimmun 1: 101-9.
143. Miyazaki, A., T. Hanafusa, K. Yarnada, J. Miyagawa, H. Fujino-Kurihara, H. Nakajima, K. Nonaka, and S. Tarui. 1985. Predominance of T lymphocytes in pancreatic islets and spleen of pre-diabetic non-obese diabetic (NOD) mice: a longitudinal study- Clin Exp Immunol60, no. 3 :622-30.
143. Koike, T., Y. Itoh, T. Ishii, 1. Ito, K. Takabayashi, N. Maruyarna, H. Tomioka, and a. Yoshida. 1987. Preventive effect of monoclonal anti-L3T4 antibody on development of diabetes in NOD rnice. Diabetes 36535-538.
144. Harada, M., and S. Makino. 1986. Suppression of overt diabetes in NOD rnice by anti- thymocyte serum or anti-Thy I ,2 antibody . Jikken Dobutsu 35, no. 4:SO 1-4.
145. Shizuru, J.A., C. Taylor-Edwards, B.A. Banks, A.K. Gregory, and C.G. Fathman. 1988. Irnmunotherapy of the nonobese diabetic mouse: treatment with an antibody to T-helper lymphocytes. Science 240:659-662.
146. Hayward, A.R., M. Shriber, -4. Cooke, and H. Waldmann. 1993. Revention of diabetes but not insulitis in NOD rnice injected with anabody to CD4. JAutoimrnun 6, no. 3:301-10.
147. Sempe, P., P. Bedossa, M.F. Richard, M.C. Villa, J.F. Bach, and C. Boitard. 1991. Anti- a l p h a t a T ce11 receptor monoclonal antibody provides an efficient therapy for autoimmune diabetes in nonobese diabetic (NOD) rnice. European Journal of Zmmunology 21, no. 5: 1163-9.
148. Makino, S., M. Harada, Y. Eshimoto, and Y. Hayashi. 1986. Absence of insulitis and overt diabetes in athymic nude mice with NOD genetic background. Jikken Dobutsu 35, no. 4:495- 8.
149. Ogawa, M., T. Maruyama, T. Hasegawa, T. Kanaya, and F. Kobayashi. 1985. The inhibitory effect of neonatal thymectomy on the incidence of insulitis in NOD rnice. Biomed Res 6: 103- 10s.
150. Dardenne, M., F. Lepault, A. Bendelac, and J.F. Bach. 1989. Acceleration of the onset of diabetes in NOD mice by thymectorny at weaning. Eur / Immunol 19, no. 5:889-95.
15 1. Miller, B.J., MC. Appel, J.J. O'Neil, and L.S. Wicker. 1988. Both the Lyt-2 and L3T4 T ce11 subsets are required for the transfer of diabetes in nonobese diabetic mice. J. Immunol. 14052-58.
152. Haskins, K., M. Portas, K. Bergman, K. Lafferty, and B. Bradley. 1989. Pancreatic islet specific T ce11 clones from nonobese diabetic mice. Proc. Natl. Acad Sei. USA 86:8000.
153. Nakano, N., H. Kikutani, H. Nishimoto, and T. Kishimoto. 199 1. T celt receptor V gene usage of islet beta cell-reactive T cells is not restricted in non-obese diabetic mice. J Erp Med 173, no. 5: 1091-7.
154. Haskins, K., and M. McDuffie. 1990. Acceleration of diabetes in young NOD rnice with a CD4+ islet-specific T ce11 clone. Science 249, no. 4975: 1433-6.
155. Yagi, H., M. Matsumoto, M. Kunimoto, J. Kawaguchi, S. Makino, and M. Harada. 1992. Analysis of the roles of CD4+ and CDS+ T cells in autoimmune diabetes of NOD rnice using transfer to NOD athymic nude mice. Eur. J. Immunol. 22:2387-2393.
156. Bradley, B., K. Haskins, F.G. LaRosa, and K.J. Lafferty. 1992. CD8 T cells are not required for islet destruction induced by a CD4+ islet-specific Tcell clone. Diabetes 41: 1603-1608.
157. Wang, B., A. Gonzalez, C. Benoist, and D. Mathis. 1996. The role of CD8+ T cells in the initiation of insulin-dependent diabetes meIlitus. Eur J Immunol 26, no. 8: 1762-9.
158. Wicker, L.S., E.H. Leiter, J.A. Todd, R.J. Renjilian, E. Peterson, P.A. Fischer, P.L. Podolin, M. Zijlstra, R. Jaenisch, and L.B. Peterson. 1994. Beta 2-microglobulin-deficient NOD mice do not develop insulitis or diabetes. Diabetes 43, no. 3:5004.
159. Serreze, D.V., E.H. Leiter, G.J. Christianson, D. Greiner, and D.C. Roopenian. 1994. Major histocompatibility complex class 1-deficient NOD-BZmnull mice are diabetes and insulitis resistant. Diabetes 43, no. 3505-9.
160. French, M.B., J. Allison, D.S. Cram, H.E. Thomas, C.M. Dempsey, A. Silva, H.M. Georgiou, T.W. Kay, L.C. Hanison, and A.M. Lew. 1997. Transgenic expression of mouse proinsulin II prevents diabetes in nonobese diabetic mice. Diubetes 46, no. 1:34-9.
161. Griffin, A C , W. Zhao, K.W. Wegmann, and W.F. Hickley. 1995. Experimental autoimmune insulitis. Induction by T lymphocytes specific for a peptide of proinsulin. Am J Parhol 147, no. 3:845-57.
162. Daniel, D., and D.R. Wegmann. 1996. Intranasal administration of insulin peptide B: 9-23 protects NOD mice from diabetes. ANI N Y AC& Sci 778, no. 37 1 :37 1-2.
163. Crawford, M., D. Daniel, D. Wegmann, H. Yang, and R.G. Gill. 1997. Autoimmune islet darnage mediated by insulin-specific T cells. Transplant Proc 29, no. 1-2:758-9.
164. Elias, D., D. Markovits, T. Reshef, R. van der Zee, and 1.R. Cohen. 1990. Induction and therapy of autoirnmune diabetes in the non-obese diabetic (NODLt) mouse by a 65-kDa heat shock protein. Proc Natl Acad Sci USA 87, no. 4: 1576-80.
165. Martin, S., J. Kardorf, B. Schulte, E.F. Lampeter, F.A. Gries, 1. Melchers, R. Wagner, J. Bertrams, B.O. Roep, A. Pfutzner, M. Pietropaolo, and H. Kolb. 1995. Autoantibodies to the islet antigen ICA69 occur in IDDM and in rheumatoid arthritis. Diabetologia 38, no. 3:35 1-
166. Dosch, &-M., R.K. Cheung, J.C. Gorga, M. Trucco, and DJ. Becker. 1997. Abnorxnal T ce11 autoreachvity in IDDM. Diabetes 46 (Suppl 1):58A.
167. Dosch, H.-M., J.C. Gorga, R.K. Cheung, J. Gay, and D.J. Becker. 1997. Structural prerequisites of antigenic mirnicry between ICA69 (Tep69) and BSA (ABBOS). 16th Int. Diabetes Fed. Congr., Helsinki, Finland.
168. Dosch, H.-M., R.K. Cheung. and DJ. Becker. 1998. Are Islet Self-Antigens distinguished by their ability to induce T ce11 Anergy or to recruit Z A P 7 O ? Diabetes 47 Suppl 1 :A2O8.
169. Dosch, H.-M., R.K. Cheung, M. Pietropaolo, and D.J. Becker. 1998. Do Roinsulin- & IA-2 specific T cells mark Progression of Diabetic Autoimrnunity? Diabetes 47 Suppl 1:A208.
170. Roep, B.O., G. Duinkerken, H. Kolb, R.R.P. De Vries, and S. Martin. 1995. HLA associated inverse correlation between T ce11 proliferation and antibody production to islet ce11 autoantigen ICA69 in IDDM. Autoimmunity 2126.
17 1. Colman, P.G., I.L. Campbell, T.W.H. Kay, and L.C. Harrison. 1987.64 000-Mr auto-antigen in type 1 diabetes: evidence against its surface location on hurnan islets. Diabetes 36: 1432- 40.
172. Harrison, L.C. 1994. BSA immunity in IDDM- just a load of bull? Int Diabetes Monit 6, no. 57 .
173. Ohashi, P.S., S. Oehen, P. Aichele, H. Pircher, B. Odermatt, P. Herrera, Y. Higuchi, K. Buerki, H. Hengartner, and R.M. Zinkemagel. 1993. Induction of diabetes is influenced by the infectious virus and local expression of MHC class 1 and tumor necrosis factor-alpha. J Zmmunol 150, no. 1 l :S 85-94.
174. Dosch, H.-M., R.K. Cheung, and D.J. Becker. 1996. T ce11 responses in shipped blood samples of newly diabetic children. 15th Immun01 Diabetes Workshop, Canberra, AU.
175. Aichele, P., D. Kyburz, P.S. Ohashi, B. Odermatt, R.M. Zinkemagel, H. Hengartner, and H. Pircher. 1994. Peptide-induced T-ce11 tolerance to prevent autoirnmune diabetes in a transgenic mouse model. Proc Natl Acod Sci U S A 91, no. 2:444-8.
176. Elias, D., and I.R. Cohen. 1994. Peptide therapy for diabetes in NOD mice. Loncet 343, no. 8 899: 704-6.
177, Elias, D., A. Meilin, V. Ablamunits, O.S. Birk, P. Carmi, W.S. Konen, and I.R. Cohen. 1997. Hsp6O peptide therapy of NOD mouse diabetes induces a Th2 cytokine burst and downregulates autoimmunity to various beta-ce11 antigens. Diabetes 46, no. 5:758-64.
178. Elliot, J., H. Qin, S. Bhatti, D. Smith, R. Singh, T. Dillon, J. Lauzon, and B. Singh. 1994. Immunization with the larger fom of mouse glutamic acid decarboxylase (GAD67) prevents autoirnrnune diabetes in NOD mice. Diaberes 43: 1494-1499.
179. Tian, J., M. Clare-Salzler, A. Hershenfeld, B. Middleton, D. Newman, R. Mueller, S. Arita, C. Evans, M.A. Atkinson, Y. Mullen, N. Sarvetnick, A.J. Tobin, P.V. Lehmann, and D.L. Kaufman. 1996. Modulating autoimmune responses to GAD inhibits disease progression and prolongs islet graft survival in diabetes-prone mice. Nature Med 2: 1348-1353.
180. Ramiya, V.K., X.Z. Shang, P.G. Pharis, C.H. Wasserfail, T.V. Stabler, A.B. Muir, D.A. Schatz, and N.K. Maclaren. 1996. Antigen based therapies to prevent diabetes in NOD mice. J Autoimmun 9, no. 3:349-56.
181. White, J., M. Blackman, J. Bill, J. Kappler, P. Marrack, D.P. Gold, and W. Born. 1989. Two better ce11 lines for making hybridomas expressing T ce11 receptors. J. Immunol. 143: 18% 1825.
182. Kappler, J.W., B. Skidmore, J. White, and P. Marrack. 198 1. Antigen-indicible, H-2 restricted, interleukin-2 producing T ce11 hybridomas. Lack of independent antigen and H-2 recognition. J. Exp. Med. 153:1198-1206.
183. Rock, K.L., and B. Benacerraf. 1983. MHC-restricted T ceIl activation: analysis with T ce11 hybridomas. meview] [65 refs] . Immunol Rev 7629-57.
184. Carrasco-Marin, E., 1. Shimizu, O. Kanagawa, and E.R. Unanue. 1996. The Class II MHC I- Ag7 Molecules from Non-Obese Diabetic Mice Are Poor Peptide Binders. J Immrcnol l56:4SO - 458.
185. Leeder, J.S., H.M. Dosch, P.A. Harper, P. Lam, and S.P. Spielberg. 1989. Fluorescence- based viability assay for studies of reactive dmg intermediates. Anal Biochem 177, no. 2:364- 72.
186. Cheung, R.K., and H.-M. Dosch. 1991. The Tyrosine Kinase lck is Critically Involved in the Growth Transformation of Human B Lymphocytes. J Biol Chem 266, no. 14:8667-8670.
187. Miyazaki, I., R.K. Cheung, and H.-M. Dosch. 1993. Viral IL10 is critical for B ceIl Growth transformation by EBV. J Exp Med 178:43947.
188. Cheung, R.K., 1. Miazaki, and H.-M. Dosch. 1993. Unexpected patterns of EBV gene expression during early stages of B ce11 transformation. Int. J. Immunol. 5707-716.
189. Santamaria, P., T. Utsugi, B.J. Park, N. Avenll, S. Kawazu, and J.W. Yoon. 1995. Beta-cell- cytotoxic CDS+ T cells from nonobese diabetic mice use highly homologous T ce11 receptor alpha-chain CDR3 sequences. J Zmmunol 154, no. 5:%94-503.
190. Miyazaki, T., R.K. Cheung, and H.-M. Dosch. 1994. The role of VIL-10 and other EBV latency genes early during B ce11 growth transformation. In Serninars in Virology, vol. 5. M.E. Oldstone, editor. Acad. Press, London (UK). 405-4 14.
191. Shi, Y.F., M.G. Szalay, L. Paskar, B.M. Sahai, M. Boyer, B. Singh, andD-R. Green. 1990. Activation-induced cell kath in T ceil hybridomas is due to apoptosis. Morphologic aspects and DNA fragmentation [published erratum appears in J Immunol 1990 Dec 1; l45(ll):394S]. J Imrnunol144, no. 9:3326-33.
192. Ashwell, J.D., R.E. Cunningham, P.D. Noguchi, and D. Hernandez. 1987. Ce11 growth cycle block of T ce11 hybridomas upon activation with antigen. J E p Med 165, no. 1: 173-94.
193. Sussrnan, J.J., T. Saito, E.M. Shevach, R.N. Germain, and J.D. Ashwell. 1988. Thy-l- and Ly-6-mediated lymphokine production and growth inhibition of a T ce11 hybridoma require co-expression of the T ce11 antigen receptor complex. J Immunol 140, no. 8:2520-6.
194. Dosch, H.M., P. Lam, D. Guerin, and J.S. Leeder. 1988. Charactenstics of Particle Concentration Fluorescence Lmmunoassay (PCFIA): Novel alternatives to ELISA and RIA. In: P.-L. Gembala (Ed), Baxter Healthcare Corporation: Chigago, IL. pp 5-22 -
195. Alexander, J., J. Sidney, S. Southwood, J. Ruppert, C. Oseroff, A. Maewal, K. Snoke, H.M. Serra, R.T. Kubo, A. Sette, and H.A. Grey. 1994. Development of high potency universal DR-restricted helper epitopes by modification of high affinity DR-blocking peptides. Immuniiy l:75 1-761.
196. Lee, N.E., and M.M. Davis. 1988. T ce11 receptor beta-chain genes in BW5147 and other AKR tumors. Deletion order of murine V beta gene segments and possible 5' regulatory regions. J Irnmunol 140, no. 5: 1665-75.
197. Letourneur, F., and B. Malissen. 1989. Derivation of a T ceIl hybridoma variant deprived of functional T ce11 receptor alpha and beta chain transcripts reveals a nonfunctional alpha- rnRNA of BW5147 origin. Eur J Immunoll9, no. 122269-74.
198. Liblau, R., R. Tisch, N. Bercovici, and H.O. McDevitt. 1997. Systemic antigen in the treatrnent of T-cell-mediated autoimmune diseases. [Review] [50 refs]. immun01 Today 18, no. lî:~99-6W.
199. von Herrath, M., and A. Holz. 1997. Pathological changes in the islet milieu precede infiltration of islets and destruction of betacells by autoreactive lymphocytes in a transgenic mode1 of virus-induced IDDM. J Autoimmun 10, no. 3:Z 1-8.
200. Sloan-Lancaster, J., A.S. Shaw, J.B. Rothbard, and P.M. Ailen. 1994. Partial T ce11 signaling: altered phospho-x and lack of zapïO recruitment in APL-induced T ce11 anergy. Cell79:913- 922.
201. Sloan-Lancaster, J., and P.M. Allen. 1995. Signalling events in the anergy induction of T helper 1 cells. [Review] [33 refs]. Ciba Fou& Symp 195, no. 189:189-96.
202. Madrenas, J., L.A. Chau, J. Smith, J.A. Bluestone, and R.N. Germain. 1997. The efficiency of CD4 recniitment to ligand-engaged TCR controls the agonist/partial agonist properties of peptide-MHC molecule ligands. J Erp Med 185, no. 2:219-29.
203. Reis e Sousa, C., E.H. Levine, and R.N. Germain. 1996. Partial signaling by CD8+ T cells in response to antagonist ligands. J Exp Med 184, no. 1: 149-57.
204. Madrenas, J., R.L. Wange, J.L. Wang, N. Isakov. L.E. Samelson, and R.N. Germain. 1995. Zeta phosphorylation without ZAP-70 activation induced by TCR antagonists or partial agonists [see comrnents]. Science 267, no. 5 1975 15-8.
205. Allen, P.M. 1994. Peptides in Positive and Negative Selection: A Delicate Balance. Cell 76593-596.
206. Parish, N.M., and A. Cooke. 1995. Animal models of autoimrnune endocrine disease and their uses in developing new methods of intervention. meview] [IO8 refs]. Baillieres Clin Endocrinol Metab 9, no. 1 : 175-98.
207. Liblau, R.S., R. Tisch, K. Shokat, X. Yang, N. Dumont, C.C. Goodnow, and H.O. McDevitt. 1996. Intravenous injection of soluble antigen induces thymic and peripheral T-cells apoptosis. Proc Natl Acad Sei U SA 93, no. MO3 1-6.
208. Atkinson, M.A., and E.H. Leiter. 1999. The NOD mouse mode1 of type 1 diabetes: As good as it gets? Nat Med 5, no. 6:601-4.
209. Andre, I., A. Gonzalez, B. Wang, J. Katz, C. Benoist, and D. Mathis. 1996. Checkpoints in the progression of autoimrnune disease: lessons from diabetes models. meview]. Proc Natl Acad Sci U S A 93, no. 6:2260-3.
210. Roep, B.O. 1996. T-ce11 responses to autoantigens in LDDM. The search for the Holy Grail. meview]. Diabetes 45, no. 9: 1147-56.
21 1. Tian, J., P.V. Lehmann, and D.L. Kaufman. 1994. T ce11 cross-reacticity between coxsackievirus and glutamate decarboxylase is associated with a murine diabetes susceptibility allele. J Ekp Med 180: 1979- 1984.
212. Akerblom, H.K., and 0. Vaarala. 1997. Cow milk proteins, autoimmunity and type 1 diabetes [comment]. Exp Clin Endocrinol Diaberes 105, no. 2:83-5.
2 13. Oldstone, M.B.A. 1987. Molecular mimicry and autoimrnune disease. Cell50:8 19-820.
214. Wucherpfennig, K.W., and J.L. Strorninger. 1995. Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate hurnan T ce11 clones specific for myelin basic protein. Ce11 80, no. 5:695-705.
215. Wucherpfennig, K.W., and J.L. Strorninger. 1995. Selective binding of self peptides to disease-associated major histocompatibility complex (MHC) molecules: a mechanism for MHC-linked susceptibility to human autoimrnune diseases. J fip Med 18 1, no. 5: 1597- 1601.
216. Rao, V.P., A.E. Kajon, K.R. Spindler, and G. Carayanniotis. 1999. Involvement of epitope mimicry in potentiation but not initiation of autoimmune disease. J lmmunol 162, no. 10: 5888-93.
2 17. Gautam, A., C.B. Lock, D.E. Srnilek, C.I. Pearson, L. Steinman, and HO. McDevitt. 1994. Minimum structural requirements for peptide presentation by major histocompatibility complex class II molecules: Implications in induction of autoimmunity. Pmc. Natl. Acad. Sci. USA 9 1:767-77 1.
218. Gautam, A.M., R. Liblau, G. Chelvanagarn, L. Steinman, and T. Boston. 1998. A viral peptide with limited homology to a self peptide can induce clinicai signs of experimental autoimmune encephalomyelitis. J lmmunol 176, no. 7:60-4.
219. Gorga, J-C., D.F. Graziano, B.O. Boehm, H.-M. Dosch, and M. Trucco. 1996. GAD and BSA peptide binding to HLA-DQ molecules have permissive or proteetive associations with Type 1 diabetes. 15th Immun01 of Diabetes Workshop, Canberra, Ausirdia,
220. Tornpkins, S.M., J-C. Moore, and P.E. Jensen. 1996. An insulin peptide that binds an alternative site in class II major histwompatibility cornplex. J Fxp Med 183, no. 3567-66.
221. Leiter, E.H., D.V. Serreze, and M. Prochazka. 1990. The genetics and epidemiology of diabetes in NOD mice. Zmmunology Today 1 1, no. 5: 147-9.
222. Serreze, D.V., and E.H. Leiter. 1994. Genetic and pathogenic basis of autoimmune diabetes in NOD mice. [Review] [62 refs]. Curr Opin Zmmunol6, no. 6900-6.
223. Webb, S., C. Moms, and J. Sprent. 1990. Extrathymic tolerance of mature T cells: Clona1 elimination as a consequence of immunity. Ce11 63:1249-1256.
224. Gerling, K., DY. Serreze, S.W. Christianson, and E.H. Leiter. 1992. Intrathymic islet ce11 transplantation reduces beta-ce11 autoimmunity and prevents diabetes in NOD/Lt mice. Diabetes 4 1, no. 1 2: 1672-6.
225. Cetkovic-Cvrlje, M., I.C. Gerling, A. Muir, M.A. Atkinson, J.F. Elliot, and E.H. Leiter. 1997. Retardation or acceleration of diabetes in NOD/Lt mice mediated by intrathyrnic administration of candidate beta-cell antigens. Diabetes 46, no. 12: 1975-82.
226. Larnhamedi-Cherradi, S.E., J.J. Luan, L. Eloy, G. Fluteau, J.F. Bach, ar,d H.J. Garchon. 1998. Resistance of T-cells to apoprosis in autoimmune diabetic (NOD) mice is increased early in life and is associated with dysregulated expression of Bcl-x. Diabetologia 41, no. 2: 178-84.
227. Penha-Goncalves, C., K. Leijon, L. Persson, and D. Holmberg. 1995. Type 1 diabetes and the control of dexamethazone-induced apoptosis in mice maps to the same region on chromosome 6. Genomics 28, no. 3:39844.
228. Zekzer, D., F.S. Wong, O. Ayalon, 1. Millet, M. Altieri, S. Shintani, M. Solimena, and R.S. Sherwin. 1998. GAD-reactive CD4+ Th 1 cells induce diabetes in NOD/SCID mice. J Clin invest 101, no. 1:68-73.
229. Zhang, Z.L., H.M. Georgiou, and T.E. Mandel. 1993. The effect of cyclophosphamide treatment on lymphocyte subsets in the nonobese diabetic mouse: a cornparison of vanous lymphoid organs. Autoimmunity 15, no. 1 : 1- 10.
230. Katz, J.D., B. Wang, K. Haskins, C. Benoist, and D. Mathis. 1993. Foilowing a diabetogenic T ce11 from genesis through pathogenesis. Ce11 74, no. 6: 1089-100.
231. Heiffer, C., J. Stein, S. Southwood, H. Ketelaar, A. Sette, and K. Bonornly. 1995. Altered peptide ligands can control CD4 T lymphocyte differentiation in vivo. J Eip Med 181, no. 4: 1569-74.
232. Fox, C.J., and J.S. Danska. 1997. IL-4 expression ac the onset of islet inflammation predicts nondestructive insulitis in nonobese diabetic mice. J lmmunol 158, no. 5% 14-24.
233. Liblau, R.S., S.M. Singer, and H.O. McDevitt. 1995. Th1 and Th2 CD4+ T cells in the pathogenesis of organ-specific autoimmune diseases. lmmunol TG- 16, no. 1:34-8.
234. Tisch, R., R.S. Liblau, X.D. Yang, P. Liblau, and H.O. McDevitt. 1998. Induction of GAD65 specific regulatory T-cells inhibits ongoing autoimmune diabetes in nonobese diabetic mice. Diabetes 47, no. 6:894-9.
235. Rohane, P.W., A. Shimada, D.T. Kim, C.T. Edwards, B. Charlton, L.D. Shultz, and C.G. Fathman. 1995. Islet-infiltrating lymphocytes from prediabetic NOD mice rapidly transfer diabetes to NOD-scidkid rnice. Diabetes 44, no. 5550-4.
236. Charlton, B., A. Bacelj, R.M. Slattery, and T.E. Mandel. 1989. Cyclophosphamide induced diabetes in NOD/Wehi mice: evidence for suppression in spontaneous autoimmune diabetes me Ili tus. Diabetes 38:44 1 -449.
237. Kearny, E.R., K.A. Pape, D.Y. Loh, and M.K. Jenkins. 1994. Visualization of peptide- specific T ce11 immunity and peripheral toterance induction in vivo. h w i i r y 1:327-339.
238. Vaysburd, M., C. Lock, and H. McDevitt. 1995. Prevention of insulin-dependent diabetes mellitus in nonobese diabetic mice by imrnunogenic but not by tolerated peptides. J Erp Med 182, no. 32397-902.
239. Blanas, E., F.R. Carbone, J. Allison, J.F.A.P. Miller, and W. Heath. 1996. Induction of autoimmune diabetes by oral administration of autoantigen. Science 274: 1707-1709.
240. Genain, C.P., K. Abel, N. Belmar, F. Villanger, D.P. Rosenberg, C. Linington, F.S. Raine, and S.L. Hauser. 1996. Late complications of immune deviation therapy in a non-human primate. Science 274: 2054-2057.
241. Waisman, A., P.J. Ruiz, D.L. Hirschberg, A. Gelman, J.R. Oksenberg, S. Brocke, F. Mor, I.R. Cohen, and L. Steinman. Suppressive vaccination with DNA encoding a variable region gene of the T-ce11 receptor prevents autoimmune encephalomyelitis and activates Th2 immunity [see cornments]. .
242. Xiang, Z.Q., S.L. Spitalnik, J. Cheng, J. Erikson, B. Wojczyk, and H.C. Ertl. 1995. Immune responses to nucleic acid vaccines to rabies virus. Virology 209, no. 2569-79.