| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Medicine (C.S.H., L.P.H., L.B., Å.L.), University of Washington, Seattle, Washington 98195; Department of Woman and Child Health (E.Ö., B.P.) and Department of Molecular Medicine, Clinical Genetics (I.K.), Karolinska Institute, 171 76 Stockholm, Sweden; Department of Public Health and Clinical Medicine (O.R.), Umeå University, 901 87 Umeå, Sweden; and Department of Medicine (M.L.-O., C.T.), University Hospital, 221 00 Lund, Sweden
Address correspondence and requests for reprints to: Christiane S. Hampe, Department of Medicine, Box 357710, University of Washington, Seattle, Washington 98195. E-mail:champe{at}u.washington.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a previous study (25), we demonstrated that GAD65Ab in newly diagnosed Type 1 diabetes patients can distinguish between variants of GAD65 (human, mouse, and rat) which differ mainly at the N-terminal part of the molecule. The difference in species-reactivity was more distinct in newly diagnosed teenagers and young adults whose autoantibodies preferred human GAD65 as opposed to mouse or rat GAD65, whereas Type 2 diabetes patients and GAD65Ab-positive healthy individuals showed a broader reactivity. The aim of the present study was to compare the GAD65Ab epitope patterns between three groups of GAD65Ab-positive phenotypes: newly diagnosed Type 1 diabetes patients (group I), healthy adults (group II), and first-degree relatives to Type 1 diabetes patients (group III).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Three groups (I, II, and III) of GAD65Ab-positive individuals
and one group of GAD65Ab-negative individuals were employed in this
study (Table 1
). Group I consisted of
GAD65Ab-positive newly diagnosed Type 1 diabetes patients (n =
243). One subset of this group (n = 125) consisted of 0- to
18-yr-old patients with newly diagnosed Type 1 diabetes. An additional
blood sample was obtained for some patients (n = 67) 5 yr after
onset of the disease. All of these patients were part of a study
conducted at the St. Görans Children Hospital (Stockholm, Sweden)
and represented 80% of all children diagnosed in Stockholm during
1993–1995. The frequency of GAD65Ab-positive sera in the original
group was 76%. The second subset contained 118 15- to 35-yr-old
randomly selected newly diagnosed Swedish Type 1 diabetes patients.
These patients were registered in 1992–1993 in the Diabetes Incidence
Study in Sweden (DISS) and were previously reported to be positive for
GAD65Ab (26). The frequency of GAD65Ab-positive samples in
the original patient group was 74%. The serum samples of all diabetes
patients in this study were obtained at the clinical diagnosis of
diabetes.
|
Group III consisted of 41 healthy GAD65Ab-positive first-degree relatives of Type 1 diabetes patients. These subjects were identified by screening 0- to 93-yr-old first-degree relatives in families with at least two siblings with diabetes to a total of 1170 probands with Type 1 diabetes. The families were identified in the DISS and the Swedish Childhood Diabetes registry. The GAD65Ab-positive first-degree relatives were 7–74 yr of age, and none of them were known to have developed diabetes (in a follow-up time of minimal 2 yr and maximal 12 yr).
We also analyzed sera from 131 randomly selected 15- to 35-yr-old healthy Swedish individuals. These individuals represented the control group in the DISS 92/93 study (25). Small aliquots of serum samples were kept frozen at -80 C.
Construction of fusion molecules
The GAD65, GAD67, and chimeric GAD complementary DNA (cDNA)
molecules used in the present study are summarized in Table 2. The constructions of full-length
murine GAD65 cDNA clones are described elsewhere (25).
Full-length rat GAD67 cDNA (28) was inserted into pGEM4
(Promega Corp., Madison, WI) and coded pEx12
(29). Fusion proteins were constructed by substituting
GAD65 sequences with corresponding regions of GAD67. The N-fusion
molecule consisting of the amino terminal amino acid residues of
human
GAD65(1–243)GAD67(249–593)
was constructed by introducing a NarI site at position 747 of the GAD67
cDNA clone by PCR. The resulting GAD67 fragment was cloned into the
GAD65 cDNA of human using the native NarI site at position 727. To
create the N+M fusion molecule
GAD65(1–423)GAD67(426–593)
, a SphI site was introduced at
nucleotide position 1279 of the GAD67 cDNA by PCR, and the resulting
GAD67 fragment was exchanged with the corresponding fragment of the
GAD65 cDNA of human using the native SphI site at position
1266. To create the M+C fusion molecule
GAD67(1–249)GAD65(243–585),
a NarI site at position 747 of the GAD67 cDNA was introduced by PCR.
The corresponding part of the molecule of GAD65 was replaced using
a native NarI site in its cDNA. The fusion molecule M was created
by introducing a SphI site at nucleotide position
1279 of the GAD67 cDNA by PCR and exchanging the corresponding
sequence of the construct
GAD67(1–249)GAD65243–585) (fusion
protein M+C) using a native SphI site at position
1266. The construct GAD65(1–243)
GAD67(249–426)GAD65(423–585)
(fusion protein N+C) was created by introducing a SphI site
at position 1279 of the GAD67 cDNA. The resulting GAD67 fragment was
exchanged with the corresponding cDNA fragments of construct
GAD65(1–243)GAD67(249–593)
(N-fusion molecule) using the native SphI site. The fusion
protein C
GAD67(1–426)GAD65(423–585)
was prepared in our laboratory by Dr. Dorota B. Schranz. Each construct
was verified by DNA sequence analysis before use. The in
vitro translated fusion proteins were assessed by SDS-PAGE and
were of the expected molecular weights.
|
GAD65Abs were detected by a previously described RIA (29, 30) with recent modifications (25). Recombinant 35S-GAD65, IA-2, and fusion proteins were produced by in vitro-coupled transcription/translation with SP6 RNA polymerase and nuclease-treated rabbit reticulocyte lysate (Promega Corp.), as described previously (30). The in vitro-translated 35S-GAD65 and fusion proteins were kept at -80 C and used within 2 weeks of preparation. Equal amounts of labeled antigens were used in the RIA as verified by densitometric analysis by SDS-PAGE.
IA-2Abs were detected as described previously (31). The
upper limit of the normal range was established as the 99th percentile
of the levels of 131 healthy control subjects (Table 1
). Autoantibodies
to 125I-insulin (Amersham Pharmacia Biotech, Buckinghamshire, UK) were measured as described
previously in a protein A Sepharose-based immunoprecipiation assay
(32).
Statistical analysis
Results are shown as antibody levels expressed as percent
binding of human GAD65, which was set at 100%. Antibody levels were
expressed as a relative index to correct for interassay variation using
one positive and one negative standard serum described previously
(29, 30). All samples were analyzed in duplicate
determinations, and the intra-assay average coefficient of variation
was 6.1%. The upper limit of the normal range was established for each
GAD65, GAD67, and fusion molecule as the 99th percentile of the levels
of the 131 healthy control subjects (Table 1). The Juvenile Diabetes
Foundation islet cell autoantibodies standard, which is also GAD65Ab
positive (33), was used as the GAD65Ab-positive
standard. A randomly selected control serum from a healthy volunteer
was used as negative standard.
The correlation between antibody indices was calculated using the Spearman rank correlation test. The significance of differences between antibody levels was tested with the nonparametric Mann-Whitney U test. A P value of less than 0.05 was considered significant (1, P < 0.05–0.01; **, P < 0.001–0.01; and ***, P < 0.001).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most fusion proteins and GAD isoforms were bound equally well by
sera from all three GAD65Ab-positive groups (data not shown). However,
there were significant differences in the binding to GAD67 and the N
terminus of GAD65. GAD67 was bound significantly better by sera from
both healthy individuals (group II) (P = 0.0015) and
first-degree relatives (group III) (P = 0.003) compared
with Type 1 diabetes patients (group I) (Fig. 1). We observed that GAD65Ab and GAD67Ab
indices do not correlate in any of the study groups (Fig. 2). The GAD65/67 reactive sera were
tested by immunoblotting to determine whether these sera react with
denatured GAD65. They failed to recognize GAD65 under these conditions,
indicating that the shared epitope is conformational, not linear (data
not shown).
|
|
The level of binding to the N terminus was higher in healthy
individuals (group II) and first-degree relatives (group III) compared
with Type 1 diabetes patients (Fig. 3).
This binding pattern was observed both in the GAD65-specific group
and in the GAD65/67-reactive group.
|
In all three GAD65Ab-positive study groups, we observed that
GAD65/67-reactive sera reacted better with any of the fusion proteins
compared with GAD65-specific sera (Fig. 4, A–C). These differences were
especially significant in the Type 1 diabetes patients (Fig. 4A). In
sera of group I every tested protein was bound significantly better by
GAD65/67-reactive sera than by GAD65-specific sera. This binding
pattern was less obvious in healthy individuals (Fig. 4B) and
first-degree relatives (Fig. 4C). In these two groups, significantly
preferred binding by the GAD65/67-reactive sera was observed only for
the N-, C-, and M-fusion proteins.
|
GAD65-specific sera. Compared with GAD65, binding to all
fusion proteins was significantly reduced in all three GAD65Ab-positive
sera groups (Fig. 5, A–C), with the
exceptions of the fusion protein N+M and mouse GAD65. Whereas
GAD65Ab-specific sera of Type 1 diabetes patients (Fig. 5A) bound human
GAD65 significantly better than mouse (P = 0.019), sera
of healthy individuals (Fig. 5B) and first-degree relatives (Fig. 5C)
did not differentiate between human and mouse GAD65. Sera of Type 1
diabetes patients showed a significant (P = 0.0001)
reduction in binding to the N+M-fusion protein compared with GAD65.
However, in healthy individuals and first-degree relatives no
significant difference was observed in the binding to the N+M fusion
protein and GAD65.
|
), or between GAD65 and the
fusion proteins N+M, M+C, and C (Fig. 6A). Only the binding of
the fusion proteins M (P = 0.011), N+C
(P = 0.0002), and N (P = 0.0001) was
significantly reduced compared with GAD65. The binding pattern was
similar in group II (Fig. 6B), where no significant differences in
binding were observed, except for the N (P = 0.019) and
M-fusion proteins (P = 0.009). Group III sera only
differentiated between GAD65 the N+C fusion protein and the N-fusion
protein (P = 0.0085), while binding to GAD65 and the
other fusion proteins did not differ (Fig. 6C).
|
Blood samples of Type 1 diabetes patients (0- to 18-yr-old subgroup) were obtained at onset of the disease and 5 yr later. When tested for antibody binding to human and mouse GAD65, GAD67, and the N-fusion protein, no significant differences in the binding pattern compared with the onset samples were observed (data not shown).
Serum samples from first-degree relatives who are positive for more than one autoantibody show restricted epitope specificity
Serum samples from the 41 first-degree relatives were also tested for their reactivity to IA-2 and insulin. It was found that 27% (11 of 41) of the samples were positive for GAD65Ab and IA-2Ab and 36% (4 of 11) of these double positive samples tested positive also for insulin autoantibodies. When analyzed for binding to GAD67 and the N-fusion protein, we observed that samples with more than one autoantibody reactivity show restricted epitope specificity. Whereas 13 of 30 (43%) GAD65Ab-positive sera also bound GAD67, only 3 of 11 (27%) sera positive for two autoantibodies were GAD67Ab positive (P = 0.04). Significant differences were also observed in the reactivity to N-fusion protein when comparing double with single autoantibody positive samples. None of double positive samples (0 of 11) reacted with the N-fusion protein, compared with 13 of 30 (43%) of the single positive samples (P < 0.0001).
The N-terminal part increases the antigenicity of the middle epitope
Next, we analyzed the influence of the first 240 amino acids on
the binding of both the middle and carboxyterminal region of GAD65. The
N+M fusion protein showed a considerable better binding compared with
both N and M fusion proteins (Figs. 5 and 6). This binding pattern was
observed in all sera groups but was particularly obvious for
GAD65-specific Type 1 diabetes patients (Figs. 5a and 6a) because here
the binding to the N-fusion protein was only 1% compared with binding
to GAD65. Binding to the M-fusion protein was reduced to 21%
(P = 0.0001) compared with GAD65. In contrast, fusion
protein N+M was only reduced to 65% (P = 0.0001)
compared with GAD65 and was, hence, bound significantly better than the
N-fusion protein (P = 0.0001) or the M-fusion protein
(P = 0.0001).
The effect of the N terminus on binding to the C-epitope was less obvious. A significantly better binding of the N+C fusion protein compared with the N terminus alone was observed only for Type 1 diabetes patients [both GAD65 specific (P = 0.0001) and GAD65/67 reactive (P = 0.0014)] and in GAD65-specific first-degree relatives (P = 0.0002). There were no significant differences in the binding to the C-fusion protein compared with the N+C fusion protein.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
None of the observed findings was age or gender related. No significant differences in the epitope recognition between sera from female or male individuals in the same sera group were observed (data not shown). Because the mean age varies (16–47 yr) between the sera groups, we tested the possibility that the differences in epitope recognition were age related. However, no significant differences in the binding pattern were observed when different age-classes of the same group were compared (data not shown), suggesting that the observed findings are not age related.
GAD65/67-reactive sera of all study groups showed a higher reactivity to all fusion proteins compared with GAD65-specific sera. We, therefore, conclude that antibodies that react with both GAD65 and GAD67 have a broader epitope spectrum compared with GAD65-specific antibodies.
GAD65-specific sera in Type 1 diabetes patients are dependent on the intact molecule of GAD65 as antibodies in group I showed significantly better binding to GAD65 compared with any of the fusion proteins. Furthermore, antibodies in group I are also sensitive to minor amino acid substitutions in the human GAD65 sequence—such as demonstrated by the use of mouse GAD65 and, in part also, GAD67. It should be noted that the differences in amino acid sequences between these molecules are located mainly at the N-terminal end of the molecule. We, therefore, propose that Type 1 diabetes often develops in association with conformational dependent GAD65-specific antibodies directed to a restricted epitope, which is dependent on the amino acid sequence at the N-terminal end of human GAD65. The GAD65-specific sera identified in the healthy individuals and the first-degree relatives, on the other hand, showed broader epitope specificity. The latter sera recognized fusion proteins and could not differentiate between human and mouse GAD65. We showed that the binding to GAD67 and to the N-terminal epitope of GAD65 was significantly lower in Type 1 diabetes patients than in healthy individuals and in first-degree relatives. Furthermore, we compared binding to human and mouse GAD65, GAD67, and the N-fusion protein between serum samples taken at onset of Type 1 diabetes and 5 yr later. Because no differences in the binding pattern were observed, we conclude that the observed epitope specificities tend to remain constant over time. Our findings are in concordance with previous studies that failed to identify epitopes for Type 1 diabetes patients’ sera located at the N terminus of GAD65 (9, 10, 34). Significantly higher C-terminal GAD65Ab indices in Type 1 diabetes patients than in healthy control children were reported previously (34). We did not observe such a binding pattern in the present study. However, the sera of the two studies differ considerably. Whereas the 28 healthy individuals in our study were adults (age range, 40–60 yr), Falorni et al. (34) studied sera from nine children with a maximal age of 14 yr. It is possible that the reactivity to the C terminus increases with time, although we did not observe any age-related change in the binding pattern to the C terminus. In a recent study of 29 GAD65Ab-positive children to parents with Type 1 diabetes the middle epitope of GAD65 was proposed as the initiating region as sera of 28 of 29 of the children reacted with this region (35). Their findings in healthy GAD65Ab-positive children that 36% of their GAD65Ab-specific sera reacted with the N-terminal part and 27% of the sera were GAD65/67 cross-reacting support our data in GAD65Ab-positive healthy individuals and first-degree relatives.
Although the N-terminal part does not seem to carry any epitope for GAD65-specific Type 1 diabetes patients, this part of the molecule significantly enhances the binding to the M-fusion protein in this group. Similar trends were observed in the other sera groups, emphasizing the importance of the N terminus. These results cannot be explained by a possible concealment of the middle region in this fusion protein because polyclonal antibodies raised to this region showed exceptionally good binding to this molecule (data not shown).
Our data confirm reports (11, 36) that Type 1 diabetes patients recognize highly conformation-dependent epitopes. We also show that the GAD65Ab epitope reactivity is different in Type 1 diabetes than in nondiabetic GAD65Ab-positive individuals (healthy individuals and first-degree relatives to Type 1 diabetes patients). Our laboratory has demonstrated already (25) that antibodies in Type 1 diabetes patients recognize more specific epitopes because they clearly differentiate between human and rodent GAD65, whereas this is not the case in healthy individuals or first-degree relatives (in this study). Of note, samples of GAD65Ab-positive first-degree relatives that bind also to IA-2 showed a restricted epitope specificity compared with samples that bound to GAD65 only. Long-term follow-up of these individuals will be necessary to fully delineate the relationship of this antibody profile to diabetes risk. Furthermore, it will be of importance to investigate whether progression to Type 1 diabetes is accompanied by a reduction of recognized epitopes. These results might have major significance in the study of GAD65Ab in the evolution of Type 1 diabetes and in the prediction of the disease.
|
Acknowledgments |
|---|
We thank S. Blaylock for her assistance in preparing this manuscript.
|
Footnotes |
|---|
Received March 27, 2000.
Revised August 2, 2000.
Accepted September 1, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Nalini, L. K. Gaur, M. Maldonado, C. S. Hampe, L. Rodriguez, G. Garza, A. Lernmark, and A. Balasubramanyam HLA Class II Alleles Specify Phenotypes of Ketosis-Prone Diabetes Diabetes Care, June 1, 2008; 31(6): 1195 - 1200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Oak, L. K. Gilliam, M. Landin-Olsson, C. Torn, I. Kockum, C. R. Pennington, M. J. Rowley, M. R. Christie, J. P. Banga, and C. S. Hampe The lack of anti-idiotypic antibodies, not the presence of the corresponding autoantibodies to glutamate decarboxylase, defines type 1 diabetes PNAS, April 8, 2008; 105(14): 5471 - 5476. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S Ronkainen, S. Hoppu, S. Korhonen, S. Simell, R. Veijola, J. Ilonen, O. Simell, and M. Knip Early epitope- and isotype-specific humoral immune responses to GAD65 in young children with genetic susceptibility to type 1 diabetes. Eur. J. Endocrinol., October 1, 2006; 155(4): 633 - 642. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 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]
|
|
|
|
 |
|
 M. Maioli, E. Alejandro, G. Tonolo, L. K. Gilliam, L. Bekris, C. S. Hampe, D. A. Obinu, A. Manconi, L. Puddu, K. Lynch, et al. Epitope-Restricted 65-Kilodalton Glutamic Acid Decarboxylase Autoantibodies among New-Onset Sardinian Type 2 Diabetes Patients Define Phenotypes of Autoimmune Diabetes J. Clin. Endocrinol. Metab., November 1, 2004; 89(11): 5675 - 5682. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Maldonado, C. S. Hampe, L. K. Gaur, S. D'Amico, D. Iyer, L. P. Hammerle, D. Bolgiano, L. Rodriguez, A. Rajan, A. Lernmark, et al. Ketosis-Prone Diabetes: Dissection of a Heterogeneous Syndrome Using an Immunogenetic and {beta}-Cell Functional Classification, Prospective Analysis, and Clinical Outcomes J. Clin. Endocrinol. Metab., November 1, 2003; 88(11): 5090 - 5098. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Padoa, J. P. Banga, A.-M. Madec, M. Ziegler, M. Schlosser, E. Ortqvist, I. Kockum, J. Palmer, O. Rolandsson, K. A. Binder, et al. Recombinant Fabs of Human Monoclonal Antibodies Specific to the Middle Epitope of GAD65 Inhibit Type 1 Diabetes-Specific GAD65Abs Diabetes, November 1, 2003; 52(11): 2689 - 2695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Kobayashi, S. Tanaka, M. Okubo, K. Nakanishi, T. Murase, and A. Lernmark Unique Epitopes of Glutamic Acid Decarboxylase Autoantibodies in Slowly Progressive Type 1 Diabetes J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4768 - 4775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. E. Winter, N. Harris, and D. Schatz Immunological Markers in the Diagnosis and Prediction of Autoimmune Type 1a Diabetes Clin. Diabetes, October 1, 2002; 20(4): 183 - 191. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. S. Hampe, I. Kockum, M. Landin-Olsson, C. Torn, E. Ortqvist, B. Persson, O. Rolandsson, J. Palmer, and A. Lernmark GAD65 Antibody Epitope Patterns of Type 1.5 Diabetic Patients Are Consistent With Slow-Onset Autoimmune Diabetes Diabetes Care, August 1, 2002; 25(8): 1481 - 1482. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), healthy individuals (x) and first-degree
relatives (
) were compared. The correlation factor and
P value calculated by the Sperman rank correlation test
are indicated.
