| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
* Medicine 1 and
Medicine Laboratory, Istituto Scientifico San Raffaele, Milan, Italy;
Institut fur Diabetesforshung, Munich, Germany;
Division of Medicine, University of Bristol, Bristol, United Kingdom; and
¶ Department of Medicine, King’s College School of Medicine, London, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
are autoantigens of type 1 diabetes in humans. Autoantibodies
are predominantly against IA-2, and IA-2-specific epitopes are major
autoantibody targets. We used the close homology of IA-2 and IA-2 to
design chimeras and mutants to identify humoral IA-2-specific epitopes.
Two major IA-2 epitopes that are absent from the related autoantigens
IA-2 and IA-2
13 splice variant ICA512.bdc were found contiguous
to each other within IA-2 juxtamembrane amino acids 611–620 (epitope
JM1) and 621–630 (epitope JM2). JM1 and JM2 are recognized by sera
from 67% of patients with IA-2 Abs, and relatives of patients with
type 1 diabetes having Abs to either JM epitope had a >50% risk for
developing type 1 diabetes within 6 years, even in the absence of
diabetes-associated HLA genotypes. Remarkably, the presence of Abs to
one of these two epitopes was mutually exclusive of the other; JM2 Abs
and not JM1 Abs were found in relatives with HLA DR3/4, DR4/13, or
DR1/4 genotypes; and the binding of autoantibodies to the JM2 epitope,
but not the JM1 epitope, markedly affected proteolysis of IA-2. This is
a unique demonstration of HLA-associated B cell responses to epitopes
within a single autoantigen in humans and is consistent with
modification of Ag processing by specific Ab-influencing peptide
presentation by HLA molecules. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
are novel proteins of the protein tyrosine phosphatase (PTP) family
(1, 2, 3, 4). Both are transmembrane proteins expressed within
secretory granules in brain and neuroendocrine cells including the
pancreatic islet cells (4, 5). The amino-terminal
intralumenal domains show only modest homology to each other (20%),
whereas their cytoplasmic domains are closely related (80%), being
characterized by a single PTP-like domain of 
300 amino acids and a
short juxtamembrane (JM) portion of 100 amino acids which links the PTP
domain to the transmembrane region (1, 2, 3, 4). These PTP-like
proteins have no or only weak phosphatase activity against common PTP
substrates (2, 6) and are postulated to have a regulatory
role in secretory granule exocytosis (7, 8, 9).
IA-2 and IA-2 are also major autoantigens of type I diabetes
(1, 6, 10, 11, 12). Autoantibodies to these proteins can be
detected in sera of 60–80% of patients with recent onset type 1
diabetes and their detection in sera of nondiabetic individuals
together with other islet autoantibodies such as GADA and IAA is highly
predictive for a future development of the disease (13, 14). Humoral autoimmunity is directed to the cytoplasmic portion
of the two proteins (15). Autoantibodies specific to the
IA-2 JM region and PTP-like domain, specific to the IA-2 PTP-like
domain, and cross-reactive between the IA-2 and IA-2 PTP-like
domains have been identified (15, 16, 17, 18, 19, 20). Those directed
against IA-2-specific epitopes are frequently detected in the early
humoral response, while IA-2-specific Abs are relatively infrequent,
suggesting that humoral autoimmunity to IA-2 is secondary to that of
IA-2 and arises via epitope spreading (21). Therefore,
characterization of the IA-2-specific epitopes may be an important step
toward the understanding of the maturation of the autoimmune response
against these PTPs.
We previously reported the presence of IA-2-specific epitopes within
the first 82 JM IA-2 intracellular amino acids (15). Part
of this region is encoded by exon 13 of the IA-2 gene
and is deleted in an alternatively spliced form of IA-2,
ICA512.bdc (22). In this study we have used
chimeric IA-2/IA-2 constructs and IA-2 peptides to identify minimal
epitopes involved in Ab binding and show that most of the IA-2 JM Ab
reactivity is directed against two potentially linear contiguous
epitopes with distinct HLA-DR allele associations that are absent in
the alternatively spliced ICA512.bdc.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
First, to define the minimal epitopes of the IA-2 JM Abs, sera
from 19 patients with IA-2 JM Abs were used. These were selected from a
cohort of 217 patients with new onset type 1 diabetes previously
assayed for IA-2 and IA-2 Ab epitope reactivity (16).
Next, to determine the prevalence of Abs to individual IA-2 JM
epitopes, sera from 106 IA-2 Ab-positive first-degree relatives of
patients with type 1 diabetes were used. These included 20 children
participating in the German BABYDIAB study (23), 42
relatives from the Munich family study (24), and 43
relatives from the Barts Oxford family study (25). The 106
IA-2 Ab-positive relatives were from 105 unrelated families and
included one sibling pair. Broad Ab reactivity to the IA-2 JM
and PTP regions was previously reported in a portion of subjects from
these cohorts (16, 21). The relatives had a median age of
11.8 years at Ab testing (interquartile range: 5.2–19.2 years), and 54
developed type 1 diabetes during follow-up (median time to diabetes or
last contact if no diabetes: 3.3 years; range: 0.1–14.5 years). All
relatives also had autoantibodies to insulin and/or glutamic acid
decarboxylase.
IA-2/IA-2 constructs
Constructs used in this study are shown in Fig. 1
. The IA-2604–979
(IA-2IC) and IA-2741–1033 (IA-2IC)
constructs (16) were used to construct the chimeras. All
the chimeric constructs were obtained by in-frame joining of different
portions of IA-2 and IA-2 using the Seamless (Stratagene, La Jolla,
CA) method with appropriate oligonucleotide primers. The correct in
frame sequence of joining boundaries was confirmed by DNA
sequencing.
|
Abs to each construct were measured by radio binding assay to
[35S]methionine-labeled, in vitro transcribed
and translated proteins as previously described (16).
Purified plasmid DNA of the constructs were obtained by Quantum Prep
(Bio-Rad, Hercules, CA) spin columns and in vitro transcribed and
translated using the TnT SP6-coupled rabbit reticulocyte lysate system
(Promega, Madison, WI) in the presence of 40 µCi of Redivue
L-[35S]methionine (Amersham
Pharmacia Biotech, Uppsala, Sweden). Unincorporated
[35S]methionine was removed by gel
chromatography on a NAP5 column (Amersham Pharmacia Biotech). In all
cases the expected size of the product was confirmed on SDS-PAGE. A
total of 20,000 cpm of product in TBST (25 µl of 50 mM Tris, 150 mM
NaCl, 1% Tween 20 (pH 7.4)) was incubated overnight at 4°C with 2
µl of sera and an excess of bacterially expressed intracellular
IA-2 to compete out binding to the non-IA-2-specific epitopes of the
chimeras and mutants as previously described (16). The
quantity necessary for complete inhibition of binding to the IA-2
was determined by titration against sera with high-titer
IA-2-binding Abs. Protein A-Sepharose (Amersham Pharmacia Biotech)
was subsequently added for 1 h, washed five times with 800 µl
ice-cold TBST, and counted in a TopPlate beta counter (Canberra
Packard, Groningen, The Netherlands). To control for binding efficiency
a serum with Abs to PTP epitopes shared between IA-2 and IA-2 and no
reactivity to IA-2- or IA-2-specific epitopes was included in each
assay. This serum should not be influenced by mutations from IA-2 to
IA-2 because binding is equivalent to both proteins. Sera from IA-2
Ab-negative normal control subjects were also included in each assay.
Positivity against each protein was defined by the upper limit of
normal control sera.
Further definition of epitopes within the IA-2 JM region was performed by competition with 20-mer peptides corresponding to IA-2 601–620, IA-2 611–631, IA-2 621–641, and IA-2 631–651. Serum (2 µl) was preincubated with 1 µg of peptide for 1 h at 4°C before testing in the radiobinding assay.
HLA class II typing
HLA DR and DQ typing was performed on DNA using sequence-specific oligonucleotides in 80 of the IA-2 Ab-positive first-degree relatives and in the 19 IA-2 JM Ab-positive patients with type 1 diabetes.
Statistical analysis
Analyses of the associations of JM Ab specificities with HLA class II alleles and genotypes were performed using Fisher’s exact test. Kaplan-Meier survival analysis was used to determine the risk for developing type 1 diabetes. For all statistical methods the Statistical Package for Social Sciences (SPSS, Chicago, IL) was used.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To identify the IA-2-specific epitopes within the JM region,
IA-2/IA-2 chimeras were made joining different portions of the IA-2
JM with the remaining part of the molecule belonging to IA-2 (Fig. 1
). Sera from the 19 patients with IA-2 JM Abs were then tested against
these chimeras in the presence of an excess of unlabeled IA-2IC to
inhibit binding to non-IA-2 regions of the chimeras (Table I). Eight sera (nos. 1–8) recognized
JM601–620, JM601–642,
JM601–682, and JM601–777
chimeras, suggesting the presence of at least one epitope within IA-2
residues 601–620. In all eight sera binding could also be obtained
against the IA21–620 and to IA-2IC, indicating
that an IA-2 JM epitope recognized by these sera should be contained
within IA-2 residues 604–620 (IA-2 JM1 epitope). In seven of these
eight sera, binding to the JM601–682 chimera
could be totally inhibited by competition with unlabeled
JM601–620 chimera, suggesting that the IA-2 JM
reactivity was solely against the JM601–620
epitope, whereas in one serum (no. 8) binding to the
JM601–682 chimera was only partially inhibited,
indicating the presence of Abs to additional IA-2 JM epitopes.
|
Two sera (nos. 15 and 16) bound the JM601–777 chimera, only suggesting the presence of epitope(s) within or extending to the IA-2 PTP region. Three sera (nos. 17–19), despite showing binding to the original IA-2389–779 deletion construct, did not bind any of the JM601–620, JM601–642, JM601–682, and JM601–777 chimeras. One of these sera (no. 17) bound a truncated IA-21–620 protein, suggesting that there is an epitope that includes residues upstream from the intracellular region. For this serum, binding to the original IA-2389–779 construct was not inhibited by competition with an IA-2IC construct (data not shown), further suggesting that the nonintracellular portion of IA-2 was essential for binding to this epitope.
Competition with peptides identifies small contiguous IA-2 JM epitopes
Further definition of JM1 and JM2 epitopes was performed using
20-mer peptides to compete the binding to the
JM601–682 chimera (Fig. 2). Sera with single JM1 or JM2
specificities in the absence of Ab binding to IA-2 PTP or IA-2
epitopes were used in these competition studies. Binding in sera with
JM1 Abs was completely inhibited by competition with IA-2 601–620 and
IA-2 611–630 peptides but was unaffected by competition with IA-2
591–610, 621–640, and 631–650 peptides, indicating that the JM1
epitope is contained with IA-2 residues 611–620. Binding in sera with
JM2 Abs was completely inhibited by competition with IA-2 611–630 and
IA-2 621–640 peptides but was unaffected by competition with IA-2
591–610, 601–620, and 631–651 peptides, indicating that the JM2
epitope is contained within IA-2 residues 621–630.
|
We determined Ab reactivity to the distinct JM1 and JM2
epitopes in 106 IA-2 Ab-positive first-degree relatives of patients
with type 1 diabetes (Fig. 3). JM1 Abs
were identified by binding to JM601–620
and JM601–682 chimeras and JM2 Abs by
binding to the JM601–642 and
JM601–682 but not the
JM601–620 chimera. The additional presence of
JM2 Abs in sera with JM1 Abs was ascertained when binding to the
JM601–682 chimera was only partially inhibited
by competition with excess unlabeled JM601–620
chimera. Sera from 72 (68%) of these 104 relatives had Abs to JM1
and/or JM2 epitopes. IA-2 JM-1 Abs were found in 34 relatives and IA-2
JM-2 Abs were found in 40 relatives. Abs to IA-2 JM-1 and IA-2 JM-2
epitopes were almost mutually exclusive, with only two relatives having
Abs to both epitopes (p < 0.0001). Nineteen
relatives had Abs to either JM1 or JM2 epitopes in the absence of Abs
to the IA-2 PTP domain (data not shown).
|
We previously reported HLA associations in the type and broadness
of IA-2 Ab reactivity that appeared in early childhood
(21). Children with the high diabetes risk HLA DR3/4 or
DR4/4 genotypes developed broad IA-2 Ab reactivity to multiple IA-2
epitopes, whereas those without these genotypes often developed Abs to
the IA-2 JM region before PTP Abs. HLA DR typing was available in 83 of
the 106 IA-2 Ab-positive relatives in this study, including 28 with no
IA-2 JM Abs, 23 with JM1 Abs only, 31 with JM2 Abs only, and 1 with
both JM1 and JM2 Abs (Table II). HLA
genotypes in the 23 relatives with JM1 Abs only differed markedly from
those of the relatives with JM2 Abs only or no JM Abs. Fewer than half
(11 of 23) of the JM1 Ab-positive relatives had DR4-containing
genotypes, whereas 30 of 31 JM2 Ab-positive and 24/28 IA-2 JM
Ab-negative relatives had HLA DR4 (p < 0.0001,
JM1 Ab positive vs JM2 Ab positive; p < 0.01, JM1 Ab
positive vs JM Ab negative). Moreover, HLA DR genotypes found in JM2
Ab-positive or JM-negative relatives were relatively restricted when
compared with those in relatives with JM1 Abs. The high diabetes risk
DR3/4 genotype was present in 12 (39%) JM2 Ab-positive relatives and
11 (39%) JM Ab-negative relatives compared with only one (4%) of the
JM1 Ab-positive relatives (p < 0.005 vs JM2 Ab
positive and p < 0.01 vs JM Ab negative). Six of 32
JM2 Ab-positive relatives had the less common DR4/13 genotype and five
had the DR1/4 genotype, whereas only one of the JM1 Ab-positive
(p < 0.01) and one of the JM Ab-negative
(p < 0.005) relatives had these genotypes. The
one relative with both JM1 and JM2 Abs was DR4/12. Consistent with our
previous observation (21), none of the 14 HLA-typed
relatives with IA-2 JM Abs in the absence of IA-2 PTP Abs had the HLA
DR3/4 genotype.
|
Footprinting demonstrates differences in protection of Ag proteolysis between IA-2 JM1 and JM2 Abs
To demonstrate whether binding of IA-2 to JM1 and JM2 Abs may
alter Ag proteolysis, footprinting studies using mild trypsin treatment
of IA-2 Ag bound to JM-specific Abs were undertaken (Fig. 4). Mild trypsin digestion of IA-2IC
resulted in minor bands at 40, 30, 27, and 24 kDa and a major band of
small molecular mass fragments using 0.1 mg/ml trypsin, with a
progressive loss of the minor bands when 1 mg/ml trypsin was used (Fig. 4A). Trypsin treatment of IA-2IC bound to Abs of the JM1
specificity yielded a similar pattern (Fig. 4B), whereas
IA-2IC bound to Abs of the JM2 specificity was markedly more resistant
to trypsin digestion with a major 40-kDa fragment and a relatively
small amount of low molecular mass fragments remaining even when 1
mg/ml trypsin was used (Fig. 4C). The footprinting pattern
of IA-2IC complexed to IA-2-specific PTP domain Abs gave a similar
pattern to that seen with JM2 Abs (Fig. 4B).
|
We previously reported an increased risk for developing type 1
diabetes in young relatives who had Abs to the JM region of IA-2
(21). Therefore, we examined diabetes development in the
104 IA-2 Ab-positive relatives according to IA-2 JM Ab status (Fig. 5). Of the 54 relatives who developed
type 1 diabetes on follow-up, 36 were IA-2 JM Ab positive. The 6-year
risk after first IA-2 Ab detection was 63% (95% confidence
interval (CI): 40–86%), 49% (95% CI: 33–65%), and 70% (95% CI:
51–89%) in relatives who had JM1 Abs only, JM2 Abs only, and no JM
Abs, respectively. Despite having low-risk HLA genotypes, the 12
non-DR4 IA-2 JM1 Ab-positive relatives had a similar risk to develop
diabetes within 6 years (67%) as the DR4 IA-2 Ab-positive relatives
(data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
are autoantigens of
type 1 diabetes in humans. Autoimmunity is predominantly against IA-2,
and IA-2-specific epitopes are major targets of type 1
diabetes-associated autoantibodies. In this study we have identified
two major contiguous epitopes of the short IA-2 JM region. The two
major JM epitopes are found within IA-2 residues 611–620 (epitope JM1)
and 621–630 (epitope JM2), account for most of the Ab reactivity
against the JM region, and are recognized by sera from the majority of
IA-2 Ab-positive relatives. Remarkable is the finding that
autoantibodies to these epitopes are almost mutually exclusive of each
other and associate with distinct HLA class II genotypes. To our
knowledge this is the first example of HLA-associated B cell responses
to epitopes within a single autoantigen in humans.
This mapping of IA-2 JM1 and JM2 humoral epitopes suggested that they
are contained within short linear adjacent sequences. The complete JM1
epitope appeared to be contained within residues 611–620 and contact
residues for Ab binding outside 611–620 seem unlikely. This was
deduced by the observation that 1) binding could be obtained
against IA-21–620, the
JM601–620 chimera, and to
IA-2604–979, indicating that residues 604–620
should be sufficient for Ab binding and 2) complete inhibition of
binding was obtained with the overlapping peptides 601–620 and
611–630, whereas no inhibition was obtained with IA-2 peptide
591–610. Abs recognizing the JM2 epitopes bound a chimera in which
IA-2 residues 601–620 were deleted and in which IA-2 residues 643–979
were replaced by the corresponding residues of IA-2. Together with
previous studies which showed that JM Abs bind the JM region when IA-2
PTP 683–979 is deleted (15), the binding data suggest
that necessary residues are contained within IA-2 aa 621–642, although
we cannot exclude that residues common between IA-2 and IA-2 in
region 643–682 are also involved. That inhibition of binding was
achieved by peptides 611–630 and 621–640, but not 631–650, suggested
that the essential residues involved in Ab binding are within 621–630.
This is also supported by a recent report identifying an epitope within
IA-2 aa 623–631 (19).
The location of these two epitopes is intriguing from the viewpoint of
IA-2 function. It was recently reported that IA-2 contains a PDZ domain
that includes residues 663–700 within the JM region, and that this is
involved in the binding of IA-2 to 2-syntrophin which in turn links
IA-2 and secretory granules to the cytoskeleton (7, 8).
Release of the secretory granule is postulated to result through
cleavage by µ-calpain and involving residues 643–659
(8). Cleavage would result in a truncated protein
containing the two epitopes and would dissociate these JM epitopes from
those contained within the PTP domain of IA-2. In our experience, IA-2
autoantibodies rarely if ever require both the JM and the PTP domain
for Ag recognition and usually do not bind epitopes that span across
the JM-PTP junction. Therefore, it could be postulated that
autoreactivity is generated through the distinct intracellular portions
that result from cleavage rather than the intact molecule. Remarkably,
both these functional sites lie outside the binding sites of IA-2
autoantibodies. Similarly, autoantibodies to GAD and to insulin, the
other major diabetes-associated autoantigens, rarely bind epitopes that
are close to the functional domains of the proteins (26, 27). These observations imply that autoantibodies in type 1
diabetes do not directly interfere with protein function, and that
functional regions of proteins may normally be protected from Ab
binding.
The JM region of IA-2 appears to be highly immunogenic and to act as a
distinct autoantigen from the PTP-like domains of IA-2 and IA-2
(15). It is unique to IA-2, having only weak homology to
the corresponding region of IA-2, and whereas autoantibodies binding
to the IA-2 PTP-like domain often cross-react with IA-2,
autoantibodies binding to the JM region of IA-2 have not been
identified (16, 17). Moreover, an alternatively spliced
form of IA-2 (ICA512.bdc) that lacks exon 13 is also recognized by IA-2
PTP Abs (22), but the deletion of residues 557–629 as a
result of the exon 13 splicing would remove both JM1 and JM2 epitopes.
Because these epitopes are important autoantibody targets, the use of
ICA512.bdc IA-2 alone for autoantibody measurements is not advocated.
Indeed, 19 of the 106 IA-2-positive relative sera, including 6 of the
54 who developed type 1 diabetes, had IA-2 Abs recognizing only the JM1
or JM2 epitopes (data not shown) and would not be expected to bind the
alternatively spliced IA-2.
The uniqueness of the immunogenic JM region to IA-2 and not the related
IA-2 and ICA512.bdc proteins has potentially interesting
implications for tolerance loss. Autoantigens have been detected in
human thymus during fetal life (28). Remarkable is that
transcription of the alternatively spliced ICA512.bdc form (which lacks
the JM epitopes), but not IA-2, could be detected in human thymus
(29). Therefore, the IA-2 JM epitopes and not those
present on the remainder of the molecule or on IA-2 could be
"absent" during thymic education, and IA-2 JM autoreactive T cells
could escape deletion or anergy, leading to an increased susceptibility
to autoimmunity. So far, only isolated reports of IA-2 T cell epitopes
have appeared (30, 31), and only a limited number of
potential epitopes have been suggested, none of which reside in the
IA-2 exon 13 region (30, 31, 32). Nevertheless, Ab responses
suggest that the JM and PTP regions of IA-2 represent quite distinct
entities for autoantigen recognition (15); therefore, T
cell epitopes within the JM region are likely. An important role of the
JM region of IA-2 early in tolerance loss to IA-2 is further suggested
by our previous report that the IA-2 JM epitopes are the first IA-2 Ab
epitopes that are recognized in almost 50% of subjects
developing type 1 diabetes-associated autoantibodies (21).
Moreover, in examining the age of first autoantibody appearance in the
BABYDIAB cohort (21), we found that children who develop
IA-2 JM Abs did so earlier and progressed to clinical diabetes more
rapidly than children who developed islet autoantibodies without IA-2
JM specificities (median age of first IA-2 Ab appearance was 1.9 years
in IA-2 JM-positive children vs 3.1 years in IA-2-positive JM-negative
children; p = 0.02, Mann-Whitney U
test).
Striking in our study was the observation that, despite the high prevalence of Abs to JM1 and JM2 epitopes, the simultaneous presence of Abs to both epitopes was rare in individual patients, suggesting that sustained expansion of both JM1- and JM2-responsive B cells was inhibited. An important factor in determining which of the B cells are expanded appeared to be related to HLA class II genes. HLA DR4-containing haplotypes were almost always found in relatives and patients with IA-2 JM2 Abs, but were present in only half of the relatives with IA-2 JM1 Abs. Moreover, none of the genotypes found in JM2 Ab-positive relatives were found in JM1 Ab-positive relatives. DR3/4, DR13/4, and DR1/4 were the predominant genotype (74%) of IA-2 JM2 Ab-positive relatives, whereas only 9% of the IA-2 JM1 Ab-positive relatives had these genotypes. HLA DR3 and DR13 almost always share the same allele at the DRB3 locus, potentially explaining why both were associated with IA-2 JM2 Ab positivity. The mechanism leading to HLA-associated Ab reactivity to distinct epitopes is not entirely clear. Recent analysis of HLA class II genotype structure and diabetes risk strongly indicated that risk was related to autoantigen peptide binding to HLA class II molecules (33). Peptides are generated by proteolytic enzymes within vesicles of the endocytic pathway, and it has been postulated that selective responses may arise through modified processing of Ag due to its binding to specific B cell receptors (34). It is known that the dissociation rate of monovalent Ag from high-affinity Ab is slower than the time taken for Ag capture, endocytosis, and processing by professional APCs. Thus, when high-affinity Abs drive Ag uptake, either directly via B cell membrane Ig or indirectly via FcRs, the substrate for processing may frequently be an Ag/Ab complex. Ab-bound Ag can both prevent the generation of peptides and generate new "cryptic" epitopes and thereby alter the response of Ag-specific T cell clones (34, 35). In this study we find that proteolysis of IA-2 Ag bound by JM2 Abs in vitro differs from that of IA-2 bound by JM1 Abs or unbound Ag. Binding of IA-2 JM2 Abs protected against proteolysis of IA-2, a finding consistent with the relatively restricted HLA class II genotypes found in subjects with these Abs as compared with those with IA-2 JM1 Abs. It is conceivable that B cell receptor binding to the JM2 autoantibody epitopes preferentially generates T cell epitopes that can bind to HLA class II molecules expressed in DR3/4, DR13/4, or DR1/4 APCs and result in T cell help for JM2-specific B cells, whereas B cell receptors binding to JM1 autoantibody epitopes either do not generate these peptides or generate peptides that are preferentially bound by other class II molecules. Although the mechanism by which a HLA-associated epitope-specific Ab response occurs is speculative, the observation is consistent with processing and presentation of Ab-bound IA-2 Ag having an important role in the determination of the survival of specific B cell clones.
In conclusion, this study has demonstrated how chimeric construct
screening together with point mutation can be used to identify specific
autoantibody epitopes. This approach has identified two major
diabetes-related adjacent potentially linear epitopes within the IA-2
JM region that are absent from the related autoantigens IA-2 and
ICA512.bdc. Relatives with Abs to either JM epitope have a high risk
for developing type 1 diabetes, even in the absence of
diabetes-associated HLA genotypes. However, the presence of Abs to one
of these two epitopes was mutually exclusive of the other; the epitope
specificity of Abs was strongly linked ot HLA class II genotype;
and the binding of Abs to one of the epitopes, but not the other,
markedly affected proteolysis of IA-2. These findings are consistent
with processing of specific Ab-bound Ag being a determinant for peptide
presentation by HLA class II. This approach, together with the
generation and characterization of mAbs (36) and their
screening with peptide libraries (37), should lead to the
definition of minimal epitopes that could be potentially be used in the
modeling of molecular mimicry.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Ezio Bonifacio, Department of Medicine 1, Istituto Scientifico San Raffaele, Via Olgettina 60, 20132 Milan, Italy. E-mail address: bonifacio.ezio{at}hsr.it
3 Abbreviations used in this paper: IA, islet Ag; JM, juxtamembrane; PTP, protein tyrosine phosphatase; CI, confidence interval.
Received for publication January 17, 2001. Accepted for publication February 13, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-syntrophin and nNOS in pancreatic -cells. Eur J. Cell Biol. 79:621.[Medline]
2-syntrophin and Ca2+/µ-calpain-mediated cleavage of ICA512 upon stimulation of insulin secretion. EMBO J. 20:4013.[Medline]
, as an autoantigen in insulin-dependent diabetes mellitus: precursor of the 37-kDa tryptic fragment. Proc. Natl. Acad. Sci. USA 19:2307.
in binding of autoantibodies in IDDM. Diabetologia 40:1327.[Medline]
and with six other members of the protein tyrosine phosphatase family: recognition of antigenic determinants by IDDM sera. J. Autoimmun. 10:245.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C. Pihoker, L. K. Gilliam, C. S. Hampe, and A. Lernmark Autoantibodies in Diabetes Diabetes, December 1, 2005; 54(suppl_2): S52 - S61. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Dromey, S. M. Weenink, G. H. Peters, J. Endl, P. J. Tighe, I. Todd, and M. R. Christie Mapping of Epitopes for Autoantibodies to the Type 1 Diabetes Autoantigen IA-2 by Peptide Phage Display and Molecular Modeling: Overlap of Antibody and T Cell Determinants J. Immunol., April 1, 2004; 172(7): 4084 - 4090. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Starwalt, E. L. Masteller, J. A. Bluestone, and D. M. Kranz Directed evolution of a single-chain class II MHC product by yeast display Protein Eng. Des. Sel., February 1, 2003; 16(2): 147 - 156. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
