help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism , doi:10.1210/jc.2005-2537
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow View responses
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Inaba, H.
Right arrow Articles by De Groot, L. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Inaba, H.
Right arrow Articles by De Groot, L. J.
Related Collections
Right arrow Thyroid
Right arrow Autoimmunity
The Journal of Clinical Endocrinology & Metabolism Vol. 91, No. 6 2286-2294
Copyright © 2006 by The Endocrine Society

Thyrotropin Receptor Epitopes and Their Relation to Histocompatibility Leukocyte Antigen-DR Molecules in Graves’ Disease

Hidefumi Inaba, William Martin, Anne S. De Groot, Shuwen Qin and Leslie J. De Groot

Endocrinology Division (H.I., S.Q., L.J.D.G.), Department of Medicine, and TB/HIV Research Lab (A.S.D.G.), Brown University, Providence, Rhode Island 02903; and EpiVax, Inc. (W.M., A.S.D.G.), Providence, Rhode Island 02903

Address all correspondence and requests for reprints to: Leslie J. De Groot, M.D., Brown University/Medicine/Endocrinology, Box G, Room E-308, 70 Ship Street, Providence, Rhode Island 02903. E-mail: leslie_degroot{at}brown.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Context: Graves’ disease (GD) is characterized by autoimmunity to the TSH receptor (TSHR).

Objective: We sought to identify T cell epitopes in TSHR that initiate this immune response and their interaction with human histocompatibility leukocyte antigen (HLA) molecules predisposing to GD.

Design: We examined the affinity of 31 overlapping peptides spanning the TSHR extracellular domain for binding in vitro to five purified HLA-DR molecules; DRB1*0101 (DR1), DRB1*1501 (DR2), DRB1*0301 (DR3), DRB1*1101 (DR5), and DRB1*0701 (DR7). We scanned the TSHR extracellular domain using a T cell epitope-mapping algorithm, EpiMatrix. We compared these results with clinical studies of GD patients measuring in vitro T cell responses to the peptides.

Setting: The study was conducted at a university laboratory.

Patients: Patients included 200 serial adult clinic patients with GD.

Intervention: There were no interventions.

Main Outcome Measurements: Binding affinity of epitopes, predicted affinity, and reported T cell stimulation data were measured.

Results: Most peptides bound with intermediate or high affinity to one or more HLA-DR molecule. Peptides binding to HLA-DR3 and HLA-DR5, which predispose to GD, exhibited moderate binding affinities overall, whereas most peptides binding to GD-protective HLA-DR7 bound with high affinity. These differences may relate to T cell selection in the thymus. Binding affinity of peptides correlated strongly with EpiMatrix-predicted affinity for HLA-DRB1*0101, DRB1*1501, DR3, and DRB1*0701 but not HLA-DR5. Average IC50 values correlated significantly with clinical T cell stimulation data.

Conclusions: Three different methods for identifying immunogenic peptides did not provide a uniform picture of important TSHR epitopes. However, peptide 132–150 (GIFNTGLKMFPDLTKVYST) was identified by three methods as an important epitope in GD; the possible importance of peptides 145–163, 158–176, 207–222, 248–263, 272–291, and 343–362 was also identified.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
GRAVES’ DISEASE (GD) is an organ-specific autoimmune disease. The pathogenesis of this disease remains uncertain. Environmental and endogenous factors are believed to contribute to GD, in conjunction with certain genetic factors. As in other organ-specific autoimmune diseases, autoimmune reactions are the known cause of clinical disease.

Patients with GD present with hyperthyroidism due to uncontrolled stimulation of the thyroid by antibodies that bind to the TSH receptor (TSHR) (1). CD4+ T cells recognizing TSHR peptides presented by human histocompatibility leukocyte antigen (HLA) class II molecules play a central role in the development of this autoimmunity. Attention has focused on the extracellular domain (ECD) of the TSHR because it is exposed on the surface of the cell and because a large portion of the ECD appears to be shed into surrounding tissue fluid during receptor processing (2). The ECD of the TSHR is composed of 418 amino acids linked to the transmembrane domain of 346 residues. The transmembrane domain contains seven transmembrane segments. The location of CD4+ T cell epitopes within the TSHR-ECD has been the subject of a number of studies.

Genetic factors clearly play an important role in GD. An association between certain HLA class II alleles and GD has been documented. Recently the CTLA-4 gene region has also been identified as a second locus conferring susceptibility to GD (3). This gene encodes an important negative regulator of the immune system.

Certain HLA class II genes such as HLA-DR3 and HLA-DQA1*0501 predispose to GD among Caucasian patients (4, 5). In contrast, HLA-DR7 or HLA-DQA1*0201 may provide protection against GD (6, 7). The role of HLA class II molecules is to selectively bind and present peptides, derived from processed protein captured by endocytosis, in the HLA class II epitope presenting cleft (8). Immunogenic proteins such as TSHR-ECD are endocytosed by antigen-presenting cell (APC), processed in an endosomal pathway in which they are trimmed to 10–20 amino acid peptides, displace the common CLIP peptide from its association with HLA class II molecules, and are then transported to the cell surface in which they are displayed to T cell receptors.

As has been documented for other autoimmune diseases, it is not unusual to find multiple reactive epitopes within the same autoantigen. This has been observed in the case of glutamic acid decarboxylase peptides in type 1 diabetes mellitus with HLA-DR4 (9) and for type IV collagen peptides in Goodpasture’s disease with HLA-DR15 (10) and acetylcholine receptor peptides in myasthenia gravis with HLA-DR3 (11). Many studies have partially defined the epitopes important in development of GD. We have previously reported sequences containing amino acid residues 145–163, 158–176, 207–222, 248–263, 272–291, and 343–362 considered to be important epitopes recognized by T cells from patients with GD (12, 13, 14, 15, 16, 17). Martin et al. (18) found TSHR peptides 52–71, 142–161, 202–221, and 247–266 to be frequently recognized by CD4+ T cells from patients with GD. Tandon et al. (19) found that TSHR 146–165, 160–179, and 202–221 were also possibly relevant.

One explanation of the association between HLA and GD is that high-risk HLA alleles present autoantigenic epitopes more efficiently than low-risk HLA alleles (20). Alternatively, both protective and predisposing alleles may be capable of presenting TSHR-derived epitopes but differ with respect to the functional characteristics of the CD4+ T cells that respond to the epitopes. Competition between high- and low-risk alleles for binding of autoantigenic fragments derived from TSHR could also affect the development of GD. Due to a higher affinity for specific fragments, protective alleles might prevent binding and presentation of crucial epitopes by high-risk alleles. High-affinity binding during thymic T cell selection might lead to deletion of cells reactive to specific epitopes. On the other hand, CD4+CD8–CD25+ regulatory T cells generated in thymus may play a protective role in GD. And finally, some HLA alleles may be less likely to present key epitopes that promote TSHR antibodies.

To elucidate the mechanism underlying the role of HLA class II molecules, we purified HLA-DR molecules from Epstein-Barr virus-transformed human B-lymphoblastoid cell lines (BLCL) and evaluated peptide binding to the HLA-DR molecules in vitro. We also scanned the TSHR-ECD using the EpiMatix T cell epitope-mapping algorithm. Then we compared these two methods with clinical T cell stimulation data.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Antibodies

L243 (anti-HLA-DR monoclonal antibody) hybridoma cells, HB-55, were purchased from American Type Culture Collection (Manassas, VA) and expanded in serum-free medium (Life Technologies, Inc., Grand Island, NY). This antibody is reported to bind any DR and produce minimal contamination with DQ protein (21). Antibody secreted by these cells was purified by using protein A-Sepharose chromatography. An L243-Sepharose column was prepared by coupling 10 mg of purified L243 with 2 ml of cyanogen bromide-activated Sepharose 4B (Pharmacia, Uppsala, Sweden).

Cell lines

The following homozygous Epstein-Barr virus-transformed BLCLs were kindly provided by Gerard Nepom and Susan Masewicz (Virginia Mason Research Center, Seattle, WA): LG2 (HLA-DRB1*0101, DR1), MGAR (HLA-DRB1*1501, DR2), QBL (HLA-DRB1*0301, DR3), Sweig (HLA-DRB1*1101, DR5), and PLH (HLA-DRB1*0701, DR7). These transformed cells were expanded in RPMI 1640 with 2 mM L-glutamine, 10 mM HEPES buffer, 50 µg/ml streptomycin, 50 U/ml penicillin, and 10% heat-inactivated fetal bovine serum (Life Technologies).

Purification of HLA-DR molecules

HLA-DR molecules were purified from 108 BLCLs expressing a known DR protein in the homozygous state, as previously described (12). All of the cell manipulations were performed at 4 C. Cells were lysed in 20 ml of lysis buffer [10 mM Tris, 150 mM NaCl, 0.5% Triton X-100, 1 mM EDTA, 1 mM Pefabloc (pH 7.4)]. The lysates were cleared by centrifugation at 100,000 x g for 60 min at 4 C. Lysates were passed through a protein A column and then through an L243 (anti-HLA-DR)-Sepharose affinity column. The column was washed extensively with PBS and 0.1% Triton X-100 (pH 7.4) and eluted with 10 ml of 50 mM glycine-NaOH, 0.1% Triton X-100 (pH 11.5). Eluates were immediately neutralized with 2 M glycine-HCl (pH 3.0), dialyzed against 50-fold volume of 10 mM Tris, 137 mM NaCl, 0.1% Triton X-100 (pH8), and concentrated with centricon (10 kDa; Amicon, Beverly, MA). The HLA-DR molecules were then evaluated on 10% SDS-PAGE for purity. The concentration of HLA-DR molecules was determined using BCA protein concentration assay kit (Pierce, Rockford, IL).

Peptides

Thirty-one individual TSHR-ECD peptides 16–21 amino acids in length were prepared by solid-phase peptide synthesis, each one overlapping the previous sequence by five to six amino acids (22). Peptides were purified by reverse-phase HPLC.

Mycobacterium tuberculosis 65-kDa heat shock protein peptide 3–13 (hsp3–13) (restricted by HLA-DR3) and influenza hemagglutinin peptide 307–319 (HA307–319) (restricted by HLA-DR1, HLA-DR2, HLA-DR5, and HLA-DR7) were used as controls, respectively (23, 24).

Both the hsp3–13 and HA307–319 peptides were biotinylated and used as competitor peptides.

All synthetic peptides were checked for purity by amino acid analysis and mass spectrometry. Purity was greater than 90% by these methods.

Binding and competition assays of HLA-DR and TSHR-ECD peptides

Serial dilutions (0.01–100 µM) of nonbiotinylated TSHR-ECD peptides and hsp3–13 (or HA307–319) were incubated in a 96-well plate with a given purified HLA-DR molecule (100 ng) for 45 min, followed by the addition of a single concentration (0.2 µM) of biotinylated hsp3–13 or HA307–317 peptide in 150 mM citrate phosphate buffer containing 0.75% n-octyl-ß-D-glucopyranoside and 1 mM Pefabloc (pH 5.4). The plate was incubated at 37 C overnight. Each well of a 96-well plate (Coaster, Cambridge, MA) was coated with 100 µl of anti-HLA-DR capture antibody (L243, 10 µg/ml) in 0.2M borate buffer (pH 8.1) and incubated overnight at 4 C. The antibody plate was washed five times with PBS, containing 0.05% Tween 20, blocked with PBS blocking solution [including 5% fetal bovine serum/fetal calf serum (1:1)] at room temperature for 3 h, and washed. Tris buffer (50 µl) [50 mM (pH 8.0)] containing 0.75% n-octyl-ß-D-glucopyranoside was added to each well. Bound major histocompatibility complex (MHC)-peptide complexes were transferred from the first incubation plate to this plate, which was incubated at 4 C, overnight. After washing, europium-labeled streptavidin (Wallac, Gaithersburg, MD) was added to each well and incubated at room temperature for 60 min, followed by enhancement buffer (Wallac) for 20 min at room temperature. Fluorescence was measured with a Delfia 1232 fluorometer (Wallac). Each assay of each peptide was done in triplicate, and at least three complete studies were done of binding of all peptides to each HLA-DR molecule. The results of the three separate assays were in close agreement and are presented as an average of the results. IC50 of greater than 100 were not determined and are reported as greater than 100.

Sum of T cells score

To determine T cell proliferation in response to TSHR peptide, peripheral blood mononuclear cells (PBMCs) (3 x 105/well) were cultured with or without antigen and pulsed with [3H]thymidine (1 µCi/well; ICN Radiochemicals, Irvine, CA) for 16 h as previously described (13, 15). [3H]thymidine incorporation was measured by liquid scintillation counting. Results were expressed as a stimulation index (SI), the ratio of [3H]thymidine uptake in the presence of antigen to [3H]thymidine uptake in culture medium alone.

A summary of clinically derived T cell stimulation data were defined as follows and is reported below. For each of our previous studies of T cell stimulation by TSHR epitopes, we scored statistically significant positive responses for each epitope as 1 and amalgamated the results. Epitopes positive in stimulation of T cell lines or T cell clones were also given a score of 1. We included in the same manner the data by Martin et al. (18) and Tandon et al. (19) for epitopes recognized in their studies. These scores thus represented an arbitrary but useful compilation of available reported studies evaluating T call reactivity to TSHR peptides. HLA typing data were available on only one small subset of 13 patients in our studies

EpiMatrix Z-score

EpiMatrix is a T cell epitope-mapping algorithm that is used to identify putative HLA ligands/T cell epitopes contained within protein sequences. Computation is performed by comparing peptides sequences with a set of HLA allele-specific coefficient matrices. Each matrix contains a set of 180 coefficients, one for each of nine positions, or pockets, contained in the floor of the MHC binding groove and for each of the 20 amino acids. To complete an analysis, target protein sequences are parsed into overlapping 9-mer frames in which each frame overlaps the last by eight amino acids. For any given frame, each amino acid is assigned a coefficient based on its type and position within the HLA binding groove. Coefficients are summed to produce a raw score. Raw scores are then normalized with respect to a distribution derived from a large set of randomly generated peptide sequences. The resulting Z-scores from this distribution are directly comparable across prediction for different alleles. For this study the sequences of each of the synthesized peptides were screened against the same alleles used in the binding assays. For each peptide we selected the highest scoring 9-mer peptide, within the synthetic peptide sequence, to represent the score for the sequence. EpiMatrix has been used in a large number of studies (25, 26) to predict T cell epitopes. The algorithm has been shown to be an accurate predictor of class I and class II epitopes (25, 26)

Statistical analysis

Pearson product moment correlation coefficients (r values) were calculated for IC50 values vs. Z-scores, IC50 values vs. sum of T cells, and Z-scores vs. sum of T cells. Pearson correlation coefficients were also calculated for IC50 values vs. SI of DR3+ patients and Z-scores vs. SI of DR3+ patients. Probabilities for the correlations of less than 0.05 were considered statistically significant.

For comparison between in vitro peptide binding assays and EpiMatrix predictions, IC50 cut-off values for positive and negative binders were set to 50 µM, and Z-score cut-off values were set to 1.44 to compare sensitivity and specificity.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
A summary of the in vitro binding affinities of the 31 TSHR peptides to five HLA-DR molecules (DR1, DR2, DR3, DR5, and DR7) is shown in Table 1Go. These HLA-DR alleles were evaluated in this study because of their known associations with GD. HLA-DR3 and DR5 (less so) are considered to predispose to GD, HLA-DR7 to be protective, and HLA-DR1 and HLA-DR2 to be neutral, according to most published studies (5, 6, 7, 27). Peptides demonstrating binding affinities of less than 10 µM were considered to be high-affinity binders (shown in bold). Those peptides demonstrating binding affinity between 10 and 50 µM were considered to be intermediate binders, and those peptides demonstrating binding affinity greater than 50 µM were considered low-affinity binders. This categorization is similar to that used in other studies (28).


View this table:
[in this window]
[in a new window]
  		TABLE 1. HLA-DR binding affinities of TSHR-ECD-derived peptides (IC50 in micromoles)

 
All but two peptides (20-35, 30-49) bound with intermediate or high affinity to at least one HLA-DR molecule. Peptide 57–75 bound with IC50 of less than 2 to three HLA-DR alleles (DR2, DR5, and DR7). Peptide 132–150 bound with IC50 greater than 10 for HLA-DR1 and HLA-DR7 and IC50 of less than 2 for HLA-DR2, DR3, and DR5. Surprisingly, most of the peptides binding HLA-DR7 bound with high affinity, compared with the number of peptides that bound with high affinity to HLA-DR3 and DR5. Many more peptides bound with moderate affinity to HLA-DR3 and DR5. Most of the peptides binding to HLA-DR1 and DR2 bound with low affinity. No peptide bound with high affinity to all HLA-DR molecules. Only two peptides, 109–124 and 132–150, bound with high affinity to HLA-DR3. No peptide bound with high affinity to HLA-DR3 and DR5 and low affinity to HLA-DR7.

An EpiMatrix analysis of predicted epitopes within the TSHR-ECD amino acid sequence was performed for each HLA-DR allele using progressive 9-mer sequences (overlapping by eight amino acids). The 9-mer core epitope sequence, within the specific 14–20 amino acid synthetic peptide, predicted by EpiMatrix to have the highest Z-score for a DR molecule, is reported in the columns labeled "P" in Table 2Go. Each DR molecule was separately analyzed. The predicted binding affinity for the 9-mer sequence is given as a Z-score in the columns in Table 2Go headed by the specific DR allele. Peptides that receive Z-scores greater than 1.64 are generally considered to be in the top 5% of all binding scores, and Z-scores greater than 1.28 are in the top 10% of all binding scores. Peptides receiving Z-scores above 1.64 are considered to be moderately likely to bind, whereas peptides scoring 1.28–1.64 are considered only weakly likely to bind. Lower scoring peptides generally do not bind. The 9-mers with the highest Z-score within each of the sequences labeled 1–31 in Table 1Go were chosen to represent the predicted binding affinity of each HLA-DR allele for each of the 31 peptides in Table 1Go.


View this table:
[in this window]
[in a new window]
  		TABLE 2. EpiMatrix Z-scores for TSHR peptides and HLA-DR molecules

 
Correlations between the observed binding affinities of the peptides and EpiMatrix predicted binding affinities are shown in Table 3AGo. Because a higher binding affinity is reflected by a smaller IC50 value and the EpiMatrix Z-score is higher for stronger predicted binding, a strong relationship between prediction and binding affinity is reflected by a negative number in the correlation table. Significant correlations were noted between EpiMatrix Z-scores and calculated IC50 values for alleles HLA-DR1, DR2, DR3, and DR7. No correlation was evident between EpiMatrix predictions for HLA-DR5 and binding to the HLA-DR5 molecule produced by the Sweig cell line. These results indicate that the in vitro binding affinity and predicted epitope binding to specific HLA-DR molecules were in fair to high agreement.


View this table:
[in this window]
[in a new window]
  		TABLE 3. Correlations among peptide binding assay, EpiMatrix Z-scores, and clinical T cell stimulation assay

 
Another analysis for comparing of the IC50 values and EpiMatrix Z-scores for selecting MHC ligands is shown in Table 4Go. In this comparison, IC50 cut-off values were set to 50 µM (>50 µM was considered nonbinding; lower was considered binding). Z-score cut-off values were set based on a retrospective review of corresponding EpiMatrix Z-scores and observed binding affinities (IC50). Z-score cut-off scores were set to maximize and balance of predictive sensitivity (the percentage of true positives identified) and specificity (the percentage of true negatives rejected). Thus, Z-score cut-off value is 1.44 for all HLA-DR molecules. In this table the IC50 cut-off values for the competitive binding assay included strong and moderate affinity binders. This analysis indicates strong and significant agreement on specificity and sensitivity between the two methods for four of the five DR molecules studied.


View this table:
[in this window]
[in a new window]
  		TABLE 4. Correspondence between EpiMatrix Z-scores and observed IC50 values, assuming assignment by binding assay is correct

 
We also evaluated associations between peptide binding affinities and observed ability of the same epitopes to stimulate T cells from GD patients in vitro. As described above, we combined data related to the responses of T cells from GD patients to these synthetic peptides from a number of clinical studies. Our summary included selected published data by Martin et al. (18) and Tandon et al. (19). The resulting compilation is presented in Table 5Go and the score is also presented in Table 1Go. There was a weak (appropriately negative) but not significant correlation between observed IC50 values for each HLA-DR allele and the combined data (sum of T cell score) (Table 3BGo). Because the T cell stimulation data are derived from patient populations with a mixture of HLA-DR molecules, this lack of correlation is not surprising. In contrast, we found that the average IC50 for the five HLA-DR molecules (Table 1Go) was correlated with the sum of T cells (Table 3BGo, last column). This method of comparison might better reflect the relationship between binding to a range of HLA alleles and the immunogenicity of the peptides in T cell assays derived from a diverse population when HLA typing data were not available. Although certain HLA alleles have been associated with GD, no data describing the association between HLA-DR and T cell responses to peptides derived from TSHR-ECD have been published.


View this table:
[in this window]
[in a new window]
  		TABLE 5. Summary of clinical T cell stimulation data from GD patients

 
We were able to retrospectively identify one small group of 13 HLA-DR3+ patients from one of the studies published by our group (17) (Table 3CGo). All of these subjects had one copy of the DRB1*0301/DQA1*0501/DQB2*0201 haplotype and carried DR1, 2, 4, 6, 10, or 11 as the second DR allele. There was a weakly positive correlation between responses by these patients in vitro to the peptides and the results of the HLA-DR3 binding assays that did not reach significance (the correlation was –0.29), suggesting that binding affinity is related to T cell stimulation in this group of patients (Table 3CGo). There was no significant correlation between the EpiMatrix Z-score and the sum of T cell score for the clinical studies, as expected, because the patients were not categorized by HLA-DR type (Table 3Go, D and E).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The goal of our study was to identify peptide epitopes in the TSHR-ECD that may be important in the development of GD. GD is often referred to as a B cell-mediated disease because of the importance of antibodies causing hyperthyroidism. However, CD4+ T cells play a critical role in the development of antibody responses, affinity maturation, and isotype switching, and they also play an important role in the long-term cytopathic response that is observed in all forms of autoimmune thyroid disease. Certain immunodominant T cell epitopes appear to be critically important to the development of other autoimmune diseases such as diabetes mellitus type 1, Goodpasture’s disease, and myasthenia gravis (9, 10, 11). Peptides that act as MHC class II T cell epitopes are usually 10–20 amino acids in length because the binding groove of the MHC class II molecules is open at both ends. The peptide binds in the HLA groove because the side chains of certain amino acids (anchor residues often at positions 1, 4, 6, and 9) comprising the core 9-mer sequence fit into the pockets present in the floor of the binding groove. The longer length includes flanking regions, which are considered to stabilize the peptide in the binding groove. Longer extensions are tolerated and sometimes preferred. MHC class II T cell epitopes are so defined by their ability to selectively stimulate T cells obtained from patients who have a specific autoimmune disease (11) or in some instances by recovering naturally processed peptides from APCs (9, 10). Because the ability of peptides to stimulate T cells is absolutely dependent on their binding to HLA molecules, HLA binding assays have been used to screen peptides for possible immunogenicity. Likewise, T cell epitope-mapping algorithms use information about anchor residues binding pocket affinities to predict peptide binding to HLA molecules, and these algorithms have proven to be a useful tool for the prospective selection of T cell epitopes for research studies (25, 26, 29). Binding affinity is also important in the process of thymic T cell selection, during which T cells having high-affinity binding for peptides present in the fetal thymus are deleted (30).

Different interpretations of the relation of binding affinity to importance in disease process have been offered. A common theme is that high-affinity binding to HLA-DR correlates with both efficiency in antigen presentation and T cell deletion in the thymic T cell selection process (9, 10, 11). Alternatively, some reports indicate that peptides with high-affinity binding to HLA-DR molecules lead to thymic deletion of the cognate T cells and that peptides that exhibit only moderate binding affinities are more apt to enter in the circulation and participate in autoimmune disease (28, 31, 32).

It must also be noted that the interaction of epitope, APC, and T cell involves numerous other factors, including variations in the specific type of APC involved and the nature of the responding T cell. As a result, development of tolerance occurs at both central and peripheral locations (33, 34). Because of immunological tolerance, severe autoimmune reactions are generally suppressed. But surprisingly, 10–30% of females and 5–15% of males develop autoantibodies against thyroid-related antigens, such as TSHR, thyroglobulin, or thyroid peroxidase (35, 36, 37, 38).

The current studies do not evaluate the affinity of binding of the T cell receptor complex to the epitope-HLA complex, which must in fact be a crucial interaction. We do have data on the T cell responses to the epitopes, presented herein as sum of T cell scores. It is known that amino acids in the third, fifth, and eighth positions of the core nonamer tend to point up, out of the binding cleft, and make contact with the T cell receptor (39). This aspect of epitope function can be more thoroughly analyzed when epitope-specific T cell lines or clones are available. T cell receptor contact modeling algorithms are not currently available. B cell epitope mapping algorithms have been described, but none have been shown to be accurate predictors of B cell immunogenicity. Finally, whereas some T cell epitopes overlap B cell epitopes (40), in most cases there is no good correlation between B cell epitopes and T cell epitopes. Also, and perhaps more importantly in this study, the relatively arbitrary start and stop points chosen for synthetic peptides may, or may not, interrupt the linear sequence of a naturally processed epitope.

In this study, we determined binding affinities of a set of overlapping 16–21 mer synthetic peptides covering the TSHR-ECD to five kinds of purified HLA-DR molecules. This approach is valuable because it defines affinity of peptide binding to a specific HLA allele. However, it does not exactly mimic the complexities of antigen presentation in patients, who would generally be heterozygous for DR genes and would also carry functional DQ alleles. We also determined the probable epitopes by using a T cell epitope-mapping algorithm. Nearly all peptides bound to at least one of the HLA-DR molecules, which is expected and in fact represents the presumed function of having different HLA-DR molecules in the population by providing protection against pathogens to at least some subset of individuals. One interesting observation that we made was that many of the peptides (11 of 31) that we selected for this study bound with high affinity to HLA-DR7, which has been associated with protection from GD. Only two of 31 bound with high affinity to DR3. One explanation for this observation may be that high-affinity binding contributed to the removal of disease-causing T cells in the process of thymic deletion in subjects bearing DR. There is good evidence that TSHR is present in the thymus (41). In addition, Schmidt et al. (42) have shown that part of the mechanism for MHC-linked resistance to autoimmunity is negative selection of pathogenic autoreactive T cells, implying high affinity of some immunodominant peptides for disease-protective MHC molecules.

In contrast, only two peptides (109-124, 132-150) bound with high affinity to HLA-DR3, an allele that has been associated with thyroid disease (Table 1Go). One of the two (peptide 132–150) also binds strongly to GD-related HLA-DR5. These two peptides are not among those showing high affinity for HLA-DR7. Four peptides also displayed high affinity binding to HLA-DR5, another GD-associated allele. The implication of this observation is not entirely clear because high-affinity binding has been associated with thymic deletion. Eighteen of 31 peptides bound with moderate affinity to DR3. The lack of many high-affinity binders and the predominance of moderate affinity binders for this allele might lead to retention of autoreactive TSHR-ECD-specific T cells in HLA-DR3+ patients

In a previous study, we examined binding of 13 TSHR-ECD peptides to HLA-DR3 using a different assay (12). The peptides in that study were classified as immunodominant or not. Peptides classified as nonimmunodominant had similar IC50 values in the two studies. Almost all nonimmunodominant had IC50 values greater than 40 in both studies. The peptides designated as immunodominant in the prior study had IC50 values of approximately 2–20, whereas in the present study, all but one had IC50 values of 20–40. The reasons for this small shift are uncertain but may relate to the different techniques used. Peptide 109–124 demonstrated the same very high affinity in both studies.

We previously performed several T cell immunogenicity studies using blood samples from patients with GD and reported several sequences that appeared to be immunodominant epitopes (peptides 145–163, 158–176, 207–222, 248–263, 272–291, and 343–362). In this study each of these peptides bound with moderate affinity (IC50 14–50) to DR3.

We evaluated each of the peptides by comparing the results of our binding affinity studies with those of a T cell epitope-mapping algorithm and found a significant correlation between the predictions and HLA-DR binding for four of the five HLA-DR molecules evaluated (Table 3AGo). There was no correlation between HLA-DR5 predictions and in vitro binding studies. These observations support the use of T cell epitope mapping using the EpiMatrix algorithm for the prediction of HLA ligands, except for HLA-DR5. The reason for the discrepancy between HLA-DR5 prediction and binding is under evaluation. The peptides that were synthesized often truncated the predicted epitope, which could have led to reductions in the correlation between predictions and immunogenicity. Furthermore, flanking amino acids act to stabilize the MHC molecule outside the binding groove. Indeed, amino terminal truncation of the flanking residues (Table 2Go) seemed to reduce the likelihood that peptides predicted to bind would bind in vitro, whereas COOH terminus truncation did not affect the accuracy of the binding predictions. This is due to the relative importance of the amino terminal amino acids in MHC/peptide interactions. Shorter amino terminal flanks could reduce the ability of the important pocket-binding side chain from the key amino acid to interact with the P1 binding pocket in the HLA binding groove.

We also compared the IC50 values and T cell epitope predictions with results from T cell stimulation assays using GD patients’ blood samples and found a poor correlation with the T cell stimulation assays. Although this is disappointing, perhaps it should be anticipated. First, the sequences we chose may not correspond to those produced in vivo in APCs. Also, eptitope binding assays and predictions relate to specific DR structures, but there is effectively no information about the HLA alleles in the patient populations studied. Furthermore, whereas binding of epitope to HLA is required for T cell response, it is not sufficient. The T cell receptor must recognize the epitope in the context of the HLA molecule, and amino acids in positions 3, 5, and 8 in the core sequence are important in this regard. Epitopes can bind effectively but not stimulate a T cell, depending on single amino acid differences in the structure. However, the correlation between IC50 values and T cell immunogenicity was significant when values for all HLA-DRs were averaged, in a manner that might be more representative of a population with diverse DRs.

Evaluating these TSHR epitopes by three distinct methods does not provide a uniform picture of their importance in development of GD. The results are summarized in Table 6Go by criteria of high in vitro binding affinity to DR molecules, high EpiMatrix Z-score, and clinical T cell stimulation data. Only peptide 132–150 satisfies all criteria, which confirms its possible immunogenic importance. Peptide 158–176, which is most active in clinical assays, fails by other criteria. Peptides 11 (145–163), 16 (207–222), 20 (248–263), 22 (272–291), and 27 (343–362) are identified again as possibly important epitopes, based on observed activity in clinical studies and moderate binding affinity.


View this table:
[in this window]
[in a new window]
  		TABLE 6. Comparison of TSHR-ECD IC50, EpiMatrix score, and T cell stimulation

 
Confirmation of importance of specific TSHR epitopes has clinical significance because it might offer therapeutic possibilities through use of selective T cell removal. In this study some important epitopes may have been missed because the method used to select peptide sequences was not based on epitope prediction. This sort of algorithm should be used in future studies to select peptides, and amino terminal flanking regions should be included when peptides are synthesized. Furthermore, the poor correlation between binding affinity and T cell immunogenicity once again reinforces the need for HLA typing in the context of clinical studies on epitope selection. In future studies, recovery of naturally processed epitopes from HLA-DR3 molecules can also be performed as a mean of epitope identification.


   Acknowledgments
 
We thank Paul Knopf for reviewing the manuscript and Daniel Rivera and Elizabeth Bishop for many helpful suggestions in our research.


   Footnotes
 
This work was supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Grant RO1 DK027384.

First Published Online April 4, 2006

Abbreviations: APC, Antigen-presenting cell; BLCL, B-lymphoblastoid cell line; ECD, extracellular domain; GD, Graves’ disease; HLA, histocompatibility leukocyte antigen; hsp3–13, heat shock protein peptide 3–13; MHC, major histocompatibility complex; PBMC, peripheral blood mononuclear cell; SI, stimulation index; TSHR, TSH receptor.

Received November 22, 2005.

Accepted March 29, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Rees Smith B, McLachlan SM, Furmaniak J 1988 Autoantibodies to the thyrotropin receptor. Endocr Rev 9:106–121[Abstract/Free Full Text]
  2. Couet J, Sar S, Jolivet A, Vu Hai MT, Miligrom E, Misrahi M 1996 Shedding of human thyrotropin receptor ectodomain. J Biol Chem 271:4545–4552[Abstract/Free Full Text]
  3. Yanagawa T, Hidaka Y, Guimaraes V, Soliman M, DeGroot LJ 1995 CTLA-4 gene polymorphism associated with Graves’ disease in a Caucasian population. J Clin Endocrinol Metab 80:41–45[Abstract]
  4. Yanagawa T, Mangklabruks A, Chang Y-B, Okamoto Y, Fisfalen M-E, Curren PG, DeGroot LJ 1993 Human histocompatibility leukocyte antigen DQA1*0501 allele associated with genetic susceptibility to Graves’ disease in Caucasian population. J Clin Endocrinol Metab 76:1569–1574[Abstract]
  5. Uno H, Sasazuki H, Tamia H, Matsumoto H 1981 Two major genes, linked to HLA and Gm, control susceptibility to Graves’ disease. Nature 292:768–770[CrossRef][Medline]
  6. Lavard L, Svejgaard A 1997 HLA-class 2 associations in juvenile GD; indication of a strong protective role of the DRB1*0701, DQA1*0201. Tissue Antigens 50:639–641[Medline]
  7. Chen Q-Y, Huang W, She J-X, Baxter F, Volpe R, Maclaren NK 1999 HLA-DRB1*08, DRB1*03/DRB3*0101, and DRB3*0202 are susceptibility genes for Graves’ disease in North American Caucasians, whereas DRB1*07 is protective. J Clin Endocrinol Metab 84:3182–3186[Abstract/Free Full Text]
  8. Ghosh P, Amaya M, Mellins E, Wiley DC 1995 The structure of an intermediate in class II MHC maturation: CLIP bound to HLA-DR3. Nature 378:457–462[CrossRef][Medline]
  9. Nepom GT, Lippolis JD, White FM, Masewicz S, Marto JA, Herman A, Luckey CJ, Falk B, Shabanowitz J, Hunt DF, Engelhard VH, Nepom BS 2001 Identification and modulation of a naturally processed T-cell epitope from the diabetes-associated autoantigen human glutamic acid decarboxylase 65 (hGAD65). Proc Natl Acad Sci USA 98:1763–1768[Abstract/Free Full Text]
  10. Phelps RG, Turner AN, Rees AJ 1996 Direct Identification of naturally processed autoantigen-derived peptides bound to HLA-DR15. J Biol Chem 271:18549–18553[Abstract/Free Full Text]
  11. Raju R, Spack EG, David CS 2001 Acetylcholine receptor peptide recognition in HLA DR3-transgenic mice: in vivo responses correlate with MHC-peptide binding. J Immunol 167:1118–1124[Abstract/Free Full Text]
  12. Sawai Y, DeGroot LJ 2000 Binding of human thyrotropin receptor peptides to a Graves’ disease-predisposing human leukocyte antigen class II molecule. J Clin Endocrinol Metab 85:1176–1179[Abstract/Free Full Text]
  13. Soliman M, Kaplan E, Abdel-Latif A, Scherberg N, DeGroot LJ 1995 Does thyroidectomy, radioactive iodine therapy, or antithyroid drug treatment alter reactivity of patients’ T cells to epitopes of thyrotropin receptor in autoimmune thyroid diseases? J Clin Endocrinol Metab 80:2312–2321[Abstract]
  14. Fisfalen ME, Soltani K, Kaplan E, Palmer EM, van Seventer GA, Straus FH, Diaz M, Ober C, DeGroot LJ 1997 Evaluating the role of Th0 and Th1 clones in autoimmune thyroid disease by use of Hu-SCID chimeras. Clin Immunol Immunopathol 85:253–264[CrossRef][Medline]
  15. Soliman M, Kaplan E, Guimaraes V, Yanagawa T, DeGroot LJ 1996 T cell recognition of residue 158–176 in thyrotropin receptor confers risk for development of thyroid autoimmunity in siblings in a family with Graves’ disease. Thyroid 6:545–551[Medline]
  16. Fisfalen ME, Palmer EM, Van Seventer GA, Soltani K, Sawai Y, Kaplan E, Hidaka Y, Ober C, DeGroot LJ 1997 Thyrotropin-receptor and thyroid peroxidase-specific T-cell clones and their cytokine profile in autoimmune thyroid disease. J Clin Endocrinol Metab 82:3655–3663[Abstract/Free Full Text]
  17. Soliman M, Kaplan E, Yanagawa T, Hidaka Y, Fisfalen ME, DeGroot LJ 1995 T cells recognize multiple epitopes in the human thyrotropin receptor extra-cellular domain. J Clin Endocrinol Metab 80:905–914[Abstract]
  18. Martin A, Nakashima M, Zhou A, Aroson D, Werner AJ, Davies TF 1997 Detection of major T-cell epitopes on human thyroid stimulating hormone receptor by overriding immune heterogeneity in patients with Graves’ disease. J Clin Endocrinol Metab 82:3361–3366[Abstract/Free Full Text]
  19. Tandon N, Freeman MA, Weetman AP 1992 T-cell responses to synthetic TSH receptor peptides in Graves’ disease. Clin Exp Immunol 89:468–473[Medline]
  20. Sinha AA, Lopez MT, McDevitt HO 1990 Autoimmune diseases: the failure of self tolerance. Science 248:1380–1388[Abstract/Free Full Text]
  21. Gorga JC, Horejsi V, Johnson DR, Raghupathy R, Strominger JL 1987 Purification and characterization of class II histocompatibility antigens from a homozygous human B cell line. J Biol Chem 262:16087–16094[Abstract/Free Full Text]
  22. Houghten RA 1985 General method for the rapid solid-phase synthesis of large numbers of peptides: specificity of antigen-antibody interaction at the level of individual amino acids. Proc Natl Acad Sci USA 82:5131–5135[Abstract/Free Full Text]
  23. Sidney J, Oseroff C, Southwood S, Wall M, Ishioka G, Koning F, Sette A 1992 DRB1*0301 molecules recognize a structural motif distinct from the one recognized by most DR1 alleles. J Immunol 149:2634–2640[Abstract]
  24. O’Sullivan D, Arrhenius T, Sidney J, Del Guercio MF, Albertson M, Wall M, Oseroff C, Southwood S, Colon SM, Gaeta FC, Sette A 1991 On the interaction of promiscuous antigenic peptides with different DR alleles. Identification of common structural motifs. J Immunol 147:2663–2669[Abstract/Free Full Text]
  25. De Groot AS, Jesdale B, Martin W, Saint Aubin C, Sbai H, Bosma A, Lieberman J, Skowron G, Mansourati F, Mayer KH 2003 Mapping cross-clade HIV-1 vaccine epitopes using a bioinformatics approach. Vaccine 21:4486–4504[CrossRef][Medline]
  26. De Groot AS, Bishop EA, Khan B, Lally M, Marcon L, Franco J, Mayer KH, Carpenter CC, Martin W 2004 Engineering immunogenic consensus T helper epitopes for a cross-clade HIV vaccine. Methods 34:476–487[CrossRef][Medline]
  27. Tomer Y, Davies TF 2003 Searching for the autoimmune thyroid disease susceptibility genes: from gene mapping to gene function. Endocr Rev 24:694–717[Abstract/Free Full Text]
  28. Geluk A, van Meijgaarden KE, Schloot NC, Drijfhout JW, Ottenhoff TH, Roep BO 1998 HLA-DR binding analysis of peptides from islet antigens in IDDM. Diabetes 47:1594–1601[Abstract]
  29. Meister GE, Roberts CGP, Berzofsky JA, De Groot AS 1995 Two novel T cell epitope prediction algorithms based on MHC-binding motifs; comparison of predicted and published epitopes from Mycobacterium tuberculosis and HIV protein sequences. Vaccine 13:581–591[CrossRef][Medline]
  30. Kappler JW, Roehm N, Marrack P 1987 T-cell tolerance by clonal elimination in the thymus. Cell 49:273–280[CrossRef][Medline]
  31. Mason K, Denney Jr DW, McConnell HM 1995 Myelin basic protein peptide complexes with the class molecules I-Au and I-Ak form and dissociate rapidly at neutral pH. J Immunol 154:5216–5227[Abstract]
  32. Carrasco-Marin E, Shimizu J, Kanagawa O, Unanue ER 1996 The class II MHC I-Ag7 molecules from non-obese diabetic mice are poor peptide binders. J Immunol 156:450–458[Abstract]
  33. Mondino A, Khoruts A, Jenkins MK 1996 The anatomy of T-cell activation and tolerance. Proc Natl Acad Sci USA 93:2245–2252[Abstract/Free Full Text]
  34. Miller JF, Morahan G 1992 Peripheral T-cell tolerance. Annu Rev Immunol 10:51–69[CrossRef][Medline]
  35. DeGroot LJ, Quintans J 1989 The causes of autoimmune thyroid disease. Endocr Rev 10:537–562[Abstract/Free Full Text]
  36. Weetman AP 2005 Autoimmune thyroid disease. In: DeGroot LJ, Jameson JL, eds. Endocrinology. 5th ed. Chap 99. Philadelphia: WB Saunders; 1979–1994
  37. Fisfalen M-E, DeGroot LJ, Quintans J, Franklin WA, Soltani K 1988 Microsomal antigen reactive lymphocyte lines and clones deriver from thyroid tissue of patients with Graves’ disease. J Clin Endocrinol Metab 66:776–784[Abstract/Free Full Text]
  38. Costagliola S, Morgenthaler NG, Hoermann R, Badenhoop K, Struck J, Freitag D, Poertl S, Weglohner W, Hollidt JM, Quadbeck B, Dumont JE, Schumm-Draeger PM, Bergmann A, Mann K, Vassart G, Usadel KH 1999 Second generation assay for thyrotropin receptor antibodies has superior diagnostic sensitivity for Graves’ disease. J Clin Endocrinol Metab 84:90–97[Abstract/Free Full Text]
  39. De Oliveira DB, Harfouch-Hammoud E, Otto H, Papandreou NA, Stern LJ, Cohen H, Boehm BO, Bach J, Caillat-Zucman S, Walk T, Jung G, Eliopoulos E, Papadopoulos GK, van Endert PM 2000 Structural analysis of two HLA-DR-presented autoantigenic epitopes: crucial role of peripheral but not central peptide residues for T-cell receptor recognition. Mol Immunol 37:813–825[CrossRef][Medline]
  40. Brett SJ, Cease KB, Ouyang CS, Berzofsky JA 1989 Fine specificity of T cell recognition of the same peptide in association with different I-A molecules. J Immunol 143:771–779[Abstract]
  41. Spitzweg C, Joba W, Heufelder AE 1999 Expression of thyroid-related genes in human thymus. Thyroid 9:133–141[Medline]
  42. Schmidt D, Verdaguer J, Averill N, Santamaria P 1997 A mechanism for the major histocompatibility complex-linked resistance to autoimmunity. J Exp Med 186:1059–1075[Abstract/Free Full Text]



eLetters:

Read all eLetters

IC50 comparisons of binding to different class II using different reference peptides are flawed
Richard G Phelps
JCEM Online, 9 Nov 2006 [Full text]
Response to letter from Prof RG Phelps -
Leslie J De Groot, et al.
JCEM Online, 18 Jan 2007 [Full text]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow View responses
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Inaba, H.
Right arrow Articles by De Groot, L. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Inaba, H.
Right arrow Articles by De Groot, L. J.
Related Collections
Right arrow Thyroid
Right arrow Autoimmunity

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