HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Nagvekar, N. Articles by Willcox, N. Search for Related Content PubMed PubMed Citation Articles by Nagvekar, N. Articles by Willcox, N.

Scanning a DRB3*0101 (DR52a)-Restricted Epitope Cross-Presented by DR3: Overlapping Natural and Artificial Determinants in the Human Acetylcholine Receptor¹

Nita Nagvekar^*, Louise Corlett^*, Leslie W. Jacobson^*, Hidenori Matsuo^2,*, Robert Chalkley, Paul C. Driscoll, Shrikant Deshpande, Edward G. Spack and Nicholas Willcox^3,*

^* Neuroscience Group, Institute of Molecular Medicine, University of Oxford, Oxford, United Kingdom; Ludwig Institute for Cancer Research, and Department of Biochemistry and Molecular Biology, University College, London, United Kingdom; and Anergen Inc., Redwood City, CA 94063.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
A recurring epitope in the human acetylcholine receptor (AChR) subunit (146–160) is presented to specific T cells from myasthenia gravis patients by HLA-DRB3*0101—"DR52a"—or by DR4. Here we first map residues critical for DR52a in this epitope by serial Ala substitution. For two somewhat similar T cells, this confirms the recently deduced importance of hydrophobic "anchor" residues at peptide p1 and p9; also of Asp at p4, which complements this allele’s distinctive Arg⁷⁴ in DRß. Surprisingly, despite the 9 sequence differences in DRß between DR52a and DR3, merely reducing the bulk of the peptide’s p1 anchor residue (Trp¹⁴⁹Phe) allowed maximal cross-presentation to both T cells by DR3 (which has Val⁸⁶ instead of Gly). The shared K⁷¹G⁷³R⁷⁴N⁷⁷ motif in the helices of DR52a and DR3 thus outweighs the five differences in the floor of the peptide-binding groove. A second issue is that T cells selected in vitro with synthetic AChR peptides rarely respond to longer Ag preparations, whereas those raised with recombinant subunits consistently recognize epitopes processed naturally even from whole AChR. Here we compared one T cell of each kind, which both respond to many overlapping 140–160 region peptides (in proliferation assays). Even though both use Vß2 to recognize peptides bound to the same HLA-DR52a in the same register, the peptide-selected line nevertheless proved to depend on a recurring synthetic artifact—a widely underestimated problem. Unlike these contaminant-responsive T cells, those that are truly specific for natural AChR epitopes appear less heterogeneous and therefore more suitable targets for selective immunotherapy.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The primary function of HLA-DR molecules is to present fragments of Ags to specific CD4⁺ T cells. In sharp contrast with DR, the DRB1* ß-chains are highly polymorphic. Some alleles differ only at isolated sites such as position 57 in DRß, where Asp may change to Val, Ser, or Ala. At the opposite end of the helix that forms one wall of the peptide-binding groove, there is a recurring Gly/Val dimorphism at position 86 (1). The Val⁸⁶ side chain severely limits access of bulky "p1 anchor" residues (near the N terminus of the bound peptide) to the major hydrophobic "pocket 1," so reducing their binding and presentation (2, 3, 4, 45). By contrast, other DR alleles differ elsewhere by short stretches of sequence ("cassettes") that recur in different alleles (1). Thus, in the middle of the helix, DR1 and some DR4 subtypes share a Arg⁷¹-Ala⁷³Ala⁷⁴–Thr⁷⁷ motif, whereas DR3 has Lys⁷¹-Gly⁷³Arg⁷⁴–Asn⁷⁷ instead (1) (see Discussion and Fig. 6). While certain peptides can bind "promiscuously" to several alleles (5, 6), others show allele-specific motifs that reflect critical interactions with polymorphic residues in the peptide-binding grooves (5, 6, 7, 8, 9).

View larger version (36K): [in this window] [in a new window]   FIGURE 6. The suggested location in HLA-DR (a) of the 149–159 epitope (b), and sequence differences between DR52a, DR52b, DR3, and DR4 in the floor of the binding groove (c) and in the helix of DRß (d). In c and d, these amino acids are listed in the order in which they might contact the bound peptides. (This figure is adapted from the diagram of the DR1-HA 307-319 complex in Ref 27 .) b, Also shown is AChR 203-215.

In stark contrast, T cell lines initially selected in vitro with synthetic AChR peptides are very heterogeneous and almost never, in our experience, respond to natural epitopes from longer Ags (19, 25). We suggested several explanations (25): 1) a low sensitivity; 2) specificity for cryptic epitopes or 3) specificity for peptides bound in an unnatural register to surface class II molecules, or with extra residues or truncations at either end; 4) new primary responses to contaminants in the original peptides, as we noted in 4/20 cases (25).

The recognition of a natural 146–160 epitope by clone KB-D1 has now enabled us firstly to map a DR52a-restricted sequence and test its cross-presentation by DR3. Secondly, it has permitted detailed comparisons with one of our most sensitive peptide-selected lines (P9[p138]) that has never responded to any longer Ags, even though it recognizes multiple independent peptides that span the same 149–157 core, and thus appeared to be specific for the correct sequence (25).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Antigens

The recombinant polypeptides and synthetic peptides (p) are described in Ref. 25 and were generated in the Oxford laboratory unless stated otherwise. Selected peptides were further analyzed and separated by high resolution HPLC on a reverse-phase C-18 column with an aqueous acetonitrile gradient containing 0.1% trifluoroacetic acid. Fractions corresponding to different peaks were pooled and lyophilized.

Mass spectrometry

Mass spectrometry was performed on a Voyager Elite XL MALDI-TOF mass spectrometer (Perseptive Biosystems, Framingham, MA), equipped with a UV (337 nm) nitrogen laser and delayed extraction. Samples were analyzed in positive ion reflectron mode, and then the major component of each was identified using postsource decay. Reflectron spectra were calibrated externally using a peptide standard calibration mix covering the mass range of interest (Perseptive Biosystems). Spots were prepared by mixing samples 1:2 with matrix, which was a saturated solution of 2,5-dihydroxybenzoic acid (Sigma Aldrich, Poole, U.K.) dissolved in water, before spotting 1 µl onto the sample plate and allowing it to dry.

Peptide binding

The assay is detailed in Ref. 26 ; in brief, HLA-DR molecules were solubilized from the B cell lines shown in Table I and isolated by affinity chromatography with the anti-DR mAb L243. They were incubated overnight at 37°C with a dose range of the test peptide plus a constant 1 µM concentration of biotinylated APO-B 100 peptide 1273–1291 (sequence IPDNLFLKSDGRIKYTLNK) (6, 8). The HLA-DR:peptide complexes were captured (for 2 h at 4°C) onto triplicate microtiter wells precoated with mAb L243, washed, and incubated with europium-Streptavidin for 90 min (Wallac, Turku, Finland), before further washing and analysis on a Wallac 1234 DELFIA Research Fluorometer. The IC₅₀ is the peptide concentration required to halve the binding of the indicator peptide, and was determined using the program SOFTmax Pro.

Antigen-presenting cells

PBL were separated on Ficoll/metrizoate, washed, and irradiated with 30 Gy from a ¹³⁷Cs source (PBLx). The EBV-transformed B cell lines and their HLA-DR alleles are summarized in Table I; before use, they were preincubated with mitomycin C (MMC) at 50 µg/ml for 1 h and washed. In most experiments, the APC were then incubated at 37°C for 5 h at 5 x 10⁶/ml with 10–50 µg/ml of peptide before thorough washing and coculture with the T cells.

T cells and proliferation assays

The T cells and their origins are summarized in Table II. Responses of all three T cells were clearly blocked by the anti-HLA-DR mAb L243 and not by anti-DP or anti-DQ (data not shown). Clone PM-A1 was maintained by restimulation with r130-265 (1 µg/ml) plus DR4⁺ (subtype 0401) PBLx every 13–15 days, and expansion with IL-2 every 3–4 days thereafter (23). Clone KB-D1 and line P9[p138] were restimulated with PHA (1:2000) plus DR3⁺-DR52a⁺ PBLx plus IL-2 on day 0, and subsequent IL-2 as above; for KB-D1, 25 pg/ml of IL-4 was also added on day 0 only (24).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
In this study, we compare the AChR 146–160 epitope recognized by two somewhat similar T cells (Table II). We also evaluate a motif that was recently deduced by aligning 14 DR52a-restricted epitopes; its main features are hydrophobic anchors at p1 and p9, and especially an Asp (or occasionally an Asn or Glu) at p4 (13).

The epitope core for clone KB-D1 and line P9[p138]

The epitope core proved to be very similar for KB-D1 and P9[p138] (Table III). It differs slightly for the DR4-restricted T cell PM-A1 (23), which responds maximally to p149–163, whereas Thr¹⁴⁸ was almost essential for KB-D1, and so apparently was Gly¹⁴⁷ for P9[p138]. All of these T cells clearly required Trp¹⁴⁹. Their near-maximal responses to p147–163 and p143–156 imply no need for long extensions at either end of this 148–156 core. They also show that, in potent peptides with N termini extending as far as position 143, the Ala¹⁵⁷ is not essential (45).

View this table: [in this window] [in a new window]   Table III. The epitope core of three 147–158-specific T cells1

Like clone KB-D1 (24), line P9[p138] proved to be HLA-DR52a- restricted (Figs. 1 and 2). Thus, peptides with the natural sequence (W¹⁴⁹) were presented maximally by APC sharing only DR52a (Fig. 1, a and b, and Fig. 2, open symbols), whereas APC with only DR3 (±DR52b, filled symbols) presented undetectably to KB-D1 and weakly to P9[p138].

View larger version (24K): [in this window] [in a new window]   FIGURE 1. The HLA class II restrictions of clone KB-D1 and line P9[p138]. The T cells were cocultured with samples of the same peptide-pulsed B cells; DR52a = line HHKB, DR1301+-DR52a+. DR3 = line L0081785, DR3+-DR52b+. In a and b, the peptides had Trp149, but they had Phe149 for c and d. Presentation to clone KBD1 by DR3 of peptides with Phe at p1 was much stronger in Figs. 2 and 3.

View larger version (14K): [in this window] [in a new window]   FIGURE 2. Transfected mouse L cells expressing DR52a alone, DR3 alone, or DR52b alone were pulsed with the indicated peptides (20 µg/ml); r3-181 was present continuously at 1 µg/ml, and has Trp149. The results are expressed as the percentage of the maximal presentation by DR3+-DR52a+ PBLx of p145-163 (Phe149); they are means of four experiments with DR52a and two each with DR3 and DR52b.

View larger version (33K): [in this window] [in a new window]   FIGURE 3. Ala scanning of p145–163 for peptide presentation (a and b) and binding (c). Residues of the 145-163 parent sequence with Phe149 that are critical for binding to solubilized DR3 molecules (c) or for presentation by peptide-pulsed DR52a+ (a) or DR3+ B cells (b) to clone KB-D1 (solid lines) or line P9[p138] (dashed lines). We used the same B cell lines as in Fig. 2. Each point is the mean of two to six experiments. Peptides were tested at 20 µg/ml for clone KB-D1 and 50 µg/ml for line P9[p138]. Shown above are the residues in the parent sequence that were changed to Ala, Tyr, or Lys.

To scan the epitope and compare presentation by DR52a and DR3, we synthesized (in a second laboratory) a series of substituents, changing each residue from 146-160 in turn to Ala, except for Ala¹⁵⁷ which was replaced with Lys or Tyr instead. To permit presentation by both DR52a and DR3, we introduced a Phe¹⁴⁹ into the "parent" sequence of these "Alanogues." We first assayed for binding to solubilized DR molecules (Fig. 3c), focusing on DR3 because it is expressed more strongly than DR52a (10). Only two positions proved to be critical. As expected, changing Phe¹⁴⁹ to Ala abolished binding; so also did a Phe¹⁴⁹Tyr substitution (Fig. 3c). Notably, Asp¹⁵² at p4 was clearly essential too, as reported previously (8, 9, 14, 15), whereas none of the other substitutions had significant effects. In general (although a different labeled indicator peptide was used), the IC₅₀ values for most other Alanogues were comparable with those for DR4 0404, which also has Val⁸⁶ in DRß, and were about 5-fold lower for DR4 0401, with Gly⁸⁶ (46).

Residues required for presentation by DR52a

To measure their "functional binding," we next tested the presentation of each Alanogue after prepulsing the DR52a⁺ or DR3⁺ B cell lines (and then washing) before coculture with the T cells. That also prevents the T cells’ own class II molecules from presenting these peptides to each other, as occurs when they are used as a continuous stimulus (23, 24, 25) (which we also tested).

Presentation by the natural DR52a gave very similar profiles for the two T cells (Fig. 3a). Substituting residues at either end (146 and 158–160) with Ala had little effect (Fig. 3a); however, Gly¹⁴⁷ at p -2 and Thr¹⁴⁸ appeared more important, especially for P9[p138], as previously noted in Table III (cf. 8, 15). As expected, the Ala¹⁴⁹ substituent at p1 showed no detectable functional binding. By contrast, the Tyr¹⁴⁹ variant was presented maximally to both T cells (Fig. 3a); it evidently binds well to DR52a, although not to DR3 (Fig. 3, b and c). Notably, all the other residues in the 149-157 core proved to be critical too, with the single exception of Ser¹⁵⁴ at p6, which is evidently less important here than in DR4 (45).

Residues required for cross-presentation by DR3

While presentation to KB-D1 by DR3 again depended greatly on p1 and p4 (rather than p6), there were four major differences from that by DR52a (solid symbols in Fig. 3, a and b). At p1, the Tyr¹⁴⁹ variant was not now presented detectably. However, the Ala¹⁴⁹ substituent stimulated surprisingly well when present continuously (70% of maximum for both T cells; data not shown); it could evidently adopt a more recognizable conformation in DR3 than in DR52a, although not binding firmly enough to resist washing. Notably too, the Ala¹⁵⁷Lys substituent at p9 was presented much better by DR3 than by DR52a, whereas the Ile¹⁵⁸Ala variant at p10 was now much less acceptable to KB-D1. The failure to recognize the variants at p2, p3, p5, p7, and p8, even though they bound well to DR3, shows that these are TCR contact residues (27); it is particularly striking with the conservative Gly¹⁵³Ala (or Ser) substitutions at the primary contact p5, which were not recognized by either DR52a-restricted T cell.

For the P9[p138] line plus DR3 (dotted lines in Fig. 3), we noted several striking contrasts with KB-D1; however, the overall binding register was clearly the same, as shown by the closely similar patterns not only at position 149 (p1), and from 152 (p4) through 157 (p9), but also at both termini (Fig. 3b). There were four main differences from KB-D1: 1) the Thr¹⁴⁸Ala and 2) the Tyr¹⁵¹Ala variants were now presented much better by DR3 than by DR52a (Fig. 3b). 3) With the Ala¹⁵⁷Lys substituent, this difference was less extreme than with KB-D1 (Fig. 3b), whereas 4) the Ile¹⁵⁸Ala variant was now presented equally well by DR3 and DR52a to line P9[p138]. These contrasts with KB-D1 were particularly striking because samples of the same DR3⁺ APC prepulsed with the Ala¹⁵¹ variant were completely unrecognizable by the KB-D1 T cells but stimulated P9[p138] very well. That apparently reflects aberrant recognition by this peptide-selected T cell (see below).

Further differences in recognition of variants at p3

We saw similarly contrasting perceptions of variants at p3 in the shorter p143-156 (Fig. 4); these all had the natural Trp¹⁴⁹ at p1, and were presented significantly only by DR52a, and not by DR3 (not shown). Line P9[p138] responded well regardless of whether Tyr¹⁵¹ was changed to benzyl serine (with a similar bulk) or to naphthyl alanine (similar bulk to Trp) and even better with Ala¹⁵¹ (Fig. 4), whereas none of these variants was recognized detectably by clone KB-D1 (on the same prepulsed APC), even as a continuous stimulus (not shown). This group of peptides was synthesized in a third laboratory (Fig. 4, legend).

View larger version (28K): [in this window] [in a new window]   FIGURE 4. Responses of clone KB-D1 and line P9[p138] to samples of the same DR52a+ B cells prepulsed with variants of Tyr151 in p143-156 (at 50 µg/ml). Because these peptides all have Trp149 at p1, there was no response with DR3+ APC without DR52a (data not shown). These peptides were synthesized by Dr. S. C. Savill and R. Cotton (Zeneca Research, Alderley Park, Macclesfield, U.K.).

Although it recognizes many overlapping peptides synthesized in three laboratories, line P9[p138] eventually proved to be specific for synthetic contaminants. The final proof was its consistent failure to respond detectably (Fig. 5b) to a supposedly identical previous synthesis of the same p145–163 (Phe¹⁴⁹) that stimulated so well in the parent sequence and its many Alanogue variants (Fig. 3). In stark contrast, both versions were recognized very well by KB-D1 and PM-A and even by a DR4-restricted "sister line" selected against the same p138-167 from the same donor (P9; Fig. 5b); for KB-D1, the presentation of this "previous" version by DR52a and DR3 was almost identical to that in Fig. 1c (not shown).

View larger version (21K): [in this window] [in a new window]   FIGURE 5. High resolution HPLC analysis (a) of the 145-163 (Phe149) parent peptide; b, recognition of its separated peaks 1 and 2 by four different T cells; c, sequences of their major components. Peaks 1 and 2 were separated and tested for T cell stimulation (b); the intervening overlap fraction contained both components, and evoked responses similar to those against peak 2 (data not shown). For KB-D1 and P9[p138], the APC were DR52a+ DR3+ PBLx; we used the DR4:0401+ B cell line Boleth for the DR4-restricted "sister" subline that was selected in parallel against the same peptide from the same donor as P9[p138] (25 ) and also uses a different Vß (not shown). For PM-A1, the APC were the autologous B cell line. Shown here are the means of two experiments with the DR52a-restricted T cells and one with those to DR4; in all of them, the peptides were tested both after prepulsing the APC (at 50 µg/ml) and on continuous exposure (at 20 µg/ml). Highlighted in c are the Thr148 that is missing in side peak 2, and its extra Ala between p8 and p10.

Hence, for this line, the more crucial difference from the natural epitope is probably the omission of Thr¹⁴⁸ at p -1, and evidently recurred in several syntheses ( Figs. 1–4). The duplicated Ala¹⁵⁷ is clearly not essential in shorter sequences such as p143–156 (Table III, Fig. 4), although this artifact may also be influential in longer peptides (see Discussion). Previously, we had also excluded the alternative explanations 1) to 3) listed in the introduction, by carefully comparing this line with clone KB-D1. We chose this pair because of their all-or-nothing difference in responsiveness to natural AChR Ags (24, 25) despite their close similarities, not only in peptide preferences and DR52a/DR3 cross-restrictions ( Figs. 1–3), but also in TCR Vß2 usage and even in Th0 phenotype (Table II). That comparison revealed 1) no gross differences in peptide sensitivity (Fig. 1), 2) no sign of specificity for cryptic epitopes in the AChR subunit (e.g., requiring exposure by trypsinization (25), not shown), and 3) no dependence on long extensions or truncations at either end of the epitope core (Table III), or on its binding in an unnatural register (Fig. 3).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
We report first the detailed scanning of a natural epitope, which confirms a recently deduced DR52a-binding motif (13). This epitope is presented to potentially pathogenic T cell clones from myasthenic patients, not only by DR52a as well as DR4, but also by DR3, after one simple modification. This "polygamy" reflects the presence both of generic anchor residues at p1, p6, and p9 and of a distinctive Asp at p4 that adapts it well for binding to these three alleles in particular. Second, our results strongly suggest that specificity for synthetic contaminants is much more common than we previously suspected among T cell lines initially selected in vitro with AChR peptides; even one of our most sensitive T cells that responds to multiple overlapping peptides nevertheless proved to depend on a recurring artifact.

Scanning a polygamous DR52a/DR4-restricted natural epitope cross-presented by DR3

The 146–160 region clearly includes an important natural AChR epitope; exactly the same sequence is recognized by two independent T cell clones—PM-A1 and KB-D1—selected from MG patients with recombinant AChR subunits, from which it can be processed efficiently by various APCs including PBLx, macrophages (24, 30), and dendritic, endothelial (31), and muscle (32) cells. It is also a major pathogenic epitope in whole Torpedo AChR for some mouse strains (33). The DR4-restricted clone PM-A1 is considerably more sensitive than the DR52a-restricted KB-D1 or P9[p138], especially to low Ag doses (not shown), possibly because of differences between the T cells (e.g., in accessory molecules or in TCR affinity), or between DR52a and DR4, e.g., in their affinity for the peptide. Nevertheless, the similarities extend to the TCR contact residues as well as the binding register. For P9[p138], as well as these two clones, Gly¹⁵³ at p5 was completely indispensable; neither the Ala¹⁵³ nor the Ser¹⁵³ variants were recognized detectably, even though they bound well (46). Likewise, Thr¹⁵⁰ and Tyr¹⁵¹ at p2 and p3 were very important too. Secondary TCR contacts include p -2 and p -1, especially for the DR52a-restricted T cells as noted previously (15); perhaps also Val¹⁵⁵ and Val¹⁵⁶ at p7 and p8, although these could influence the peptide binding or orientation instead (2, 16, 34, 35).

Wu et al. have recently deduced a peptide-binding motif for DR52a by aligning a series of 14 DR52a-restricted epitopes (13). We can now add two further very similar sequences (Fig. 6), including a second and potentially immunodominant DR52a-restricted epitope in the AChR subunit sequence (36). Furthermore, our present results strongly support the anchor roles deduced for the hydrophobic residues at p1 and p9, and for the Asp at p4 (13), all of which proved to be crucial for both of our T cells.

Evidently, a bulky anchor at p1 is particularly important for DR52a; it is aromatic in 7 of the 16 epitopes, and only in one is it as small as Ala (13). Despite their similar binding motifs, until recently, DR52a and DR3 were believed to present distinct sets of peptides (9, 16). Remarkably, however, we achieved maximal cross-presentation by DR3 to both T cells merely by reducing the bulk of the peptide’s p1 residue, which further emphasizes the critical importance of this anchoring in pocket 1. Similarly, for anchoring in DR4, p1 is often the only crucial residue, although the Asp¹⁵² at p4 and the Ser¹⁵⁴ at p6 may also be important for presentation to specific T cells (45, 46).

By contrast, for DR52a, an Asp at p4 is just as prevalent (13 of 16 cases) as it is in DR3-restricted epitopes (6, 7, 8, 9, 14, 15, 16). It was also indispensable for binding to DR3 (Fig. 3c), probably because of critical interactions with the Arg⁷⁴ (16, 27, 28, 37) and/or the Tyr²⁶ (16) that these two alleles uniquely share (1). Two findings further support the key role of these shared residues. First, despite its Gly⁸⁶, DR52b completely failed to present any of these peptides/Alanogues to either T cell: it differs crucially by Arg⁷⁴Gln and Tyr²⁶Phe substitutions in DRß, which may either prevent or alter these interactions (37), although it is otherwise somewhat similar to both DR52a and DR3 (1) (Fig. 6). Conversely, in DR3, where these Tyr²⁶ and Arg⁷⁴ residues are shared, only a single change in the p1 anchor is required for efficient cross-presentation, despite the eight other differences from DR52a (Fig. 6). In theory, the cross-presentation to clone KB-D1 might depend on the synthetic error that is required by line P9[p138]. However, that seems very unlikely, since KB-D1 is highly sensitive to the natural sequence in the native Ag (24); also because the correct (previous) peptide that was unrecognizable to P9[138] was cross-presented to KB-D1 exactly as in Fig. 1c. Furthermore, we have recently obtained very similar results with our second DR52a-restricted (AChR ) epitope (36) (Fig. 6); changing the Trp²⁰⁵ to Phe, Leu, or Met again clearly allowed efficient cross-presentation by DR3 to one clone (not shown) although, interestingly, not to two others that may require a different peptide orientation/conformation (16, 34, 35, 38).

At p9, near their C termini, 12 of the 16 DR52a-restricted epitopes also have aliphatic residues (e.g., Leu, Ile), and a Pro here prevented binding in one case (13). With our epitopes, this position was more important for T cell recognition (Fig. 3b). Furthermore, DR3 was more permissive than DR52a at p9, as it presented the Ala¹⁵⁷Lys substituent much better to both T cells (cf. Refs. 9 and 15); indeed a basic residue here may confer specificity for DR3 rather than DR52a (9, 14, 16). Perhaps, in DR52a, there is repulsion between the Lys at p9 and the bulky Phe³⁷ in DRß, or alternatively with the Arg⁷⁶ in DR, which cannot form a salt bridge to the Val⁵⁷ in DR52a as it does to the Asp⁵⁷ in DR3 (27, 28) (Fig. 6). By analogy, in DR4, when this Asp⁵⁷ changes to Ser (in the 0405 subtype), there is a preference for an Asp or Glu at peptide p9 (7), but, surprisingly, that is not evident in DR52a-restricted epitopes (13).

In conclusion, while DR52a and DR3 similarly prefer an Asp at p4, a range of amino acids at other core positions is evidently compatible with both alleles (Figs. 3 and 6). Moreover, since DR52a tolerates larger anchors at p1, whereas DR3 is more permissive at p9, they do not merely duplicate the same specificity; hence, it is not surprising that they are so often in such tight linkage disequilibrium. Indeed, Bontrop et al. (12) proposed that the less polymorphic DR52 alleles have been retained in primate evolution because they present conserved bacterial heat shock protein epitopes so well, and thus broaden the diversity of the responding T cells.

Recognition of artificial epitopes by AChR peptide-selected T cells

In this comparison with the natural epitope of clone KB-D1 (which responds well to native AChR (24)), we expected to find that line P9[p138] required the correct epitope sequence (because of its sensitivity to so many independent overlapping peptides), but presented aberrantly (39). However, it eventually proved to be specific for a peptide contaminant, a major difficulty with peptide selection that we had substantially underestimated previously (25).

A second flaw, inferred by Viner et al. (39), is that short peptides can adopt unnatural conformations/orientations after binding directly to surface class II molecules (at neutral pH), because they bypass the natural processing and loading steps in (acidic) endosomes. Moreover, many T cells recognize these conformational differences specifically (39). While a definitive understanding would require detailed structural analyses, the artifactual absence of Thr¹⁴⁸ at p -1 may likewise affect the conformation/orientation of the present epitope. Together, these suggestions may explain the starkly contrasting perceptions by our two T cells of certain peptides (Figs. 3b and 4), even though they bound in the same register. For example, the Val⁸⁶ of DR3 unexpectedly permitted weak presentation to line P9[p138] of some potent peptides, despite their Trp¹⁴⁹ at p1, although not to KB-D1 (Fig. 1, a and b); this bulky anchor may be more easily accommodated in pocket 1 because of the incorrect Gly¹⁴⁸ at p -1 (instead of the larger Thr). Similarly, the contrasting responses to the Ala variants at p -1 and p3 in the contexts of DR52a and DR3 (Fig. 3, a and b) again suggest distinct conformational influences of the Val⁸⁶Gly interchange in DRß, which is the only difference between these alleles at this end of the groove.

Third, line P9[p138] tolerated a remarkably greater variety of TCR contact residues at p3 than did KB-D1 (Figs. 3b and 4). This line presumably recognizes an alternative orientation that reveals the peptide backbone near p3; its exposure evidently depends partly on a C-terminal truncation after Val¹⁵⁶ at peptide p8 (Fig. 4) and/or on the ThrGly replacement at p -1. Finally, at the other end, the distinctive response by P9[p138] to the Ile¹⁵⁸Ala variant (Fig. 3b) may reflect the presence of the inserted Ala between p8 and p9.

The widespread recurrence of the vital impurity is much more significant than its exact nature; it was evidently replicated several times in three specialist laboratories (Figs. 3 and 4, legends). Therefore, recognition of these multiple overlapping peptides at high sensitivity (e.g., by P9[p138]) is clearly no guarantee of specificity for the correct sequence, which we have only confirmed for 1 of the 20 sublines that we selected with synthetic peptides (25); even that was for a cryptic epitope that had to be unmasked by proteolytic cleavage of one recombinant polypeptide (25).

Minority errors such as mini-deletions, insertions, persisting protecting groups, or D-amino acids can easily escape detection, even on high resolution HPLC. Therefore, it is crucial to confirm T cell specificity with longer recombinant polypeptides and preferably with the natural autoantigen, to which our recombinant AChR-selected T cells are highly sensitive (23, 24, 25, 30). However, that is seldom attempted with peptide-selected T cells, because this widespread limitation is underrecognized. A further weakness is that there may be hindering residues at either end (40), unexpected inserts (41), or posttranslational modifications in some natural Ags that are not reproduced in synthetic peptides (reviewed in Ref. 42), including loss of amide groups (42, 43). The surprising prevalence of these problems among AChR peptide-selected T cells (19) may reflect unnatural peptide conformations after direct binding to surface class II molecules (39), new primary responses in culture at high concentrations by naïve CD45RA⁺ T cells (44), tolerance to the natural sequence/conformation (42), or low relative frequencies of the truly AChR-specific T cells in MG.

In conclusion, our findings strongly suggest that many of the very heterogeneous T cell responses to AChR peptides reported previously (21, 22) may likewise be neither truly autoimmune nor disease-specific. By contrast, the recombinant approach has revealed recurring recognition of two epitopes naturally processed from whole AChR, 146-160 and 201-219; also the frequent use of DR52a, an underrecognized candidate susceptibility gene in MG. Together, these results suggest that the truly pathogenic T cells may be much more restricted, which would have major implications for the prospects for specific immunotherapy.

	Acknowledgments

 
We thank Messrs. G. Plant and R. Cotton, Profs. G. B. Ferrara and N. Groome, and Drs. C. Savill, S. J. Curnow, and S. G. E. Marsh for invaluable help and advice; Prof. A. W. Boylston for mAbs; and many patients and colleagues for PBL.

	Footnotes

 
¹ This work was supported by grants from the Sir Jules Thorn Charitable Trust, the Medical Research Council, the Wellcome Trust, the Myasthenia Gravis Association/Muscular Dystrophy Group of Great Britain and the Royal Society.

² Current address: Dept. of Neurology, Kawatana National Hospital, Nagasaki 859-36, Japan.

³ Address correspondence and reprint requests to Dr. Nick Willcox, Neurosciences Group, Institute of Molecular Medicine, The John Radcliffe Hospital, Headington, Oxford, OX3 9DS, UK. E-mail address:

⁴ Abbreviations used in this paper: MG, myasthenia gravis; AChR, acetylcholine receptor; IC₅₀, peptide concentration required for 50% inhibition of binding of indicator peptide to DR3 molecules; MMC, mitomycin C. PBLx, 30 Gy-irradiated PBL.

Received for publication August 28, 1998. Accepted for publication January 6, 1998.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Marsh, S. G., J. E. Moses, J. G. Bodmer. 1992. HLA class II sequence polymorphism detectable by serology. K. Tsuji, and M. Aizawa, and T. Sasazuki, eds. HLA 1991 610. Oxford University Press, Oxford, U.K.
2. Krieger, J. I., R. W. Karr, H. M. Grey, W. Y. Yu, D. O’Sullivan, L. Batovsky, Z. L. Zheng, S. M. Colon, F. C. A. Gaeta, J. Sidney, et al 1991. Single amino acid changes in DR and antigen define residues critical for peptide-MHC binding and T cell recognition. J. Immunol. 148:2331.[Abstract]
3. Demotz, S., C. Barbey, G. Corradin, A. Amoroso, A. Lanzavecchia. 1993. The set of naturally processed peptides displayed by DR molecules is tuned by polymorphism of residue 86. Eur. J. Immunol. 23:425.[Medline]
4. Newton-Nash, D. K., D. D. Eckels. 1993. Differential effect of polymorphism at HLA-DR1 ß-chain positions 85 and 86 on binding and recognition of DR1-restricted antigenic peptides. J. Immunol. 150:1813.[Abstract]
5. O’Sullivan, D., J. Sidney, E. Appella, L. Walker, L. Phillips, S. M. Colon, C. Miles, R. W. Chesnut, A. Sette. 1990. Characterization of the specificity of peptide binding to four DR haplotypes. J. Immunol. 145:1799.[Abstract]
6. Chicz, R. M., R. G. Urban, J. C. Gorga, D. A. A. Vignali, W. S. Lane, J. L. Strominger. 1993. Specificity and promiscuity among naturally processed peptides bound to HLA-DR alleles. J. Exp. Med. 178:27.[Abstract/Free Full Text]
7. Rammensee, H-G., T. Friede, S. Stevanovic. 1995. MHC ligands and peptide motifs: first listing. Immunogenetics 41:178.[Medline]
8. Malcherek, G., V. Gnau, S. Stevanovic, H-G. Rammensee, G. Jung, A. Melms. 1994. Analysis of allele-specific contact sites of natural HLA-DR17 ligands. J. Immunol. 153:1141.[Abstract]
9. Sidney, J., C. Oseroff, S. Southwood, M. Wall, G. Ishioka, F. Koning, A. Sette. 1992. DRB1*0301 molecules recognize a structural motif distinct from the one recognized by most DRß alleles. 1992. J. Immunol. 149:2634.[Abstract]
10. Berdoz, J., J. Gorski, A-M. Termijtelen, J-M. Drayer, C. Irle, D. Schendel, B. Mach. 1987. Constitutive and induced expression of the individual HLA-DR ß and chain loci in different cell types. J. Immunol. 139:1336.[Abstract]
11. Demotz, S., A. Lanzavecchia, U. Eisel, H. Niemann, C. Widmann, G. Corradin. 1989. Delineation of several DR-restricted tetanus toxin T cell epitopes. J. Immunol. 142:394.[Abstract]
12. Bontrop, R. E., D. G. Elferink, N. Otting, M. Jonker, R. R. P. de Vries. 1990. Major histocompatibility complex class II-restricted antigen presentation across a species barrier: conservation of restriction determinants in evolution. J. Exp. Med. 172:53.[Abstract/Free Full Text]
13. Wu, S., K. Maslanka, J. Gorski. 1997. An integrin polymorphism that defines reactivity with alloantibodies generates an anchor for MHC class II peptide binding: a model for unidirectional alloimmune responses. J. Immunol. 158:3221.[Abstract]
14. Geluk, A., K. E. van Meijgaarden, S. Southwood, C. Oseroff, J. W. Drijfhout, R. R. P. de Vries, T. H. M. Ottenhoff, A. Sette. 1994. HLA-DR3 molecules can bind peptides carrying two alternative specific submotifs. J. Immunol. 152:5742.[Abstract]
15. Geluk, A., K. E. van Meijgaarden, A. A. M. Janson, J. W. Drijfhout, R. H. Meloen, R. P. De Vries, T. H. Ottenhoff. 1992. Functional analysis of DR17(DR3)-restricted mycobacterial T cell epitopes reveals DR17-binding motif and enables the design of allele-specific competitor peptides. J. Immunol. 149:2864.[Abstract]
16. Geluk, A., X-T. Fu, K. E. van Meijgaarden, Y. Y. R. Jansen, R. R. P. De Vries, R. W. Karr, T. H. M. Ottenhoff. 1994. T cell receptor and peptide-contacting residues in the HLA-DR17(3) ß1 chain. Eur. J. Immunol. 24:3241.[Medline]
17. Willcox, N.. 1993. Myasthenia gravis. Curr. Opin. Immunol. 5:910.[Medline]
18. Hohlfeld, R., K. V. Toyka, K. Heininger, H. Gross-Wilde, I. Kalies. 1984. Autoimmune human T lymphocytes specific for acetylcholine receptor. Nature 310:244.[Medline]
19. Hawke, S., H. Matsuo, M. Nicolle, G. Malcherek, A. Melms, N. Willcox. 1996. Autoimmune T cells in myasthenia gravis: heterogeneity and potential for specific immunotargeting. Immunol. Today 17:307.[Medline]
20. Melms, A., G. Malcherek, U. Gern, H. Wietholter, C. A. Muller, R. Schoepfer, J. Lindstrom. 1992. T cells from normal and myasthenic individuals recognize the human acetylcholine receptor: heterogeneity of antigenic sites on the -subunit. Ann. Neurol. 31:311.[Medline]
21. Oshima, M., T. Ashizawa, M. S. Pollack, M. Z. Atassi. 1990. Autoimmune T cell recognition in myasthenia gravis on the extracellular part of the subunit. Eur. J. Immunol. 20:2563.[Medline]
22. Protti, M. P., A. Manfredi, R. M. Horton, M. Bellone, B. M. Conti-Tronconi. 1993. Myasthenia gravis: recognition of a human autoantigen at the molecular level. Immunol. Today 14:363.[Medline]
23. Ong, B., N. Willcox, P. Wordsworth, D. Beeson, A. Vincent, D. Altmann, J. S. S. Lanchbury, G. C. Harcourt, J. I. Bell, J. Newsom-Davis. 1991. Critical role for the Val/Gly⁸⁶ HLA-DRß dimorphism in autoantigen presentation to human T cells. Proc. Natl. Acad. Sci. USA 88:7343.[Abstract/Free Full Text]
24. Nagvekar, N., A-M. Moody, P. Moss, I. Roxanis, J. Curnow, D. Beeson, N. Pantic, J. Newsom-Davis, A. Vincent, N. Willcox. 1998. A pathogenetic rôle for the thymoma in myasthenia gravis; autosensitization of IL4-producing T cell clones recognizing extracellular acetylcholine receptor epitopes presented by minority class II isotypes. J. Clin. Invest. 101:2268.[Medline]
25. Matsuo, H., A-P. Batocchi, S. Hawke, M. Nicolle, L. Jacobson, A. Vincent, J. Newsom-Davis, N. Willcox. 1995. Peptide-selected T cell lines from myasthenia gravis patients and controls recognize epitopes that are not processed from whole acetylcholine receptor. J. Immunol. 155:3683.[Abstract]
26. Tompkins, S. M., P. A. Rota, J. C. Moore, P. Jensen. 1993. A europium fluorimmunoassay for measuring binding of antigen to class II MHC glycoproteins. J. Immunol. Methods 163:209.[Medline]
27. Stern, L. J., J. H. Brown, T. S. Jardetzky, J. C. Gorga, R. G. Urban, J. L. Strominger, D. C. Wiley. 1994. Crystal structure of the human class II MHC protein HLA-DR1 complexed with an influenza virus peptide. Nature 368:215.[Medline]
28. Ghosh, P., M. Amaya, E. Mellins, D. C. Wiley. 1995. The structure of an intermediate in class II MHC maturation: CLIP bound to HLA-DR3. Nature 378:457.[Medline]
29. Johnson, J. P., T. Meo, G. Riethmüller, D. J. Schendel, R. Wank. 1982. Direct demonstration of an HLA-DR allotypic determinant on the low molecular weight (beta) subunit using a mouse monoclonal antibody specific for DR3. J. Exp. Med. 156:104.[Abstract/Free Full Text]
30. Hawke, S., N. Willcox, G. Harcourt, A. Vincent, J. Newsom-Davis. 1992. Stimulation of human T cells by sparse antigens captured on immunomagnetic particles. J. Immunol. Methods 155:41.[Medline]
31. Savage, C. O. S., C. J. Brooks, G. C. Harcourt, J. K. Picard, W. King, D. M. Sansom, N. Willcox. 1995. Human vascular endothelial cells process and present autoantigen to human T cell lines. Int. Immunol. 7:471.[Abstract/Free Full Text]
32. Baggi, F., M. Nicolle, A. Vincent, H. Matsuo, N. Willcox, J. Newsom-Davis. 1993. Presentation of endogenous acetylcholine receptor epitope by an MHC class II-transfected human muscle cell line to a specific CD4⁺ T cell clone from a myasthenia gravis patient. J. Neuroimmunol. 46:57.[Medline]
33. Shenoy, M., M. Oshima, M. Z. Atassi, P. Christadoss. 1993. Suppression of experimental autoimmune myasthenia gravis by epitope-specific neonatal tolerance to synthetic region 146-162 of acetylcholine receptor. Clin. Immunol. Immunopathol. 66:230.[Medline]
34. Reay, P. A., R. M. Kantor, M. M. Davis. 1994. Use of global amino acid replacements to define the requirements for MHC binding and T cell recognition of moth cytochrome c (93–103). J. Immunol. 152:3946.[Abstract]
35. Brett, S. J., D. McKean, J. York-Jolley, J. A. Berzofsky. 1989. Antigen presentation to specific T cells by Ia molecules selectively altered by site-directed mutagenesis. Int. Immunol. 1:130.[Abstract/Free Full Text]
36. Hill, M. E., D. Beeson, P. Moss, L. Jacobson, A. Bond, L. Corlett, A. Vincent, J. Newsom-Davis, N. Willcox. 1999. An immunodominant epitope in the adult-specific acetylcholine receptor subunit. Ann. Neurol. 45:224.[Medline]
37. Mellins, E., B. Arp, D. Singh, B. Carreno, L. Smith, A. H. Johnson, D. Pious. 1990. Point mutations define positions in HLA-DR3 molecules that affect antigen presentation. Proc. Natl. Acad. Sci. USA 87:4785.[Abstract/Free Full Text]
38. Wu, S., J. Gorski, D. D. Eckels, D. K. Newton-Nash. 1996. T cell recognition of MHC class II-associated peptides is independent of peptide affinity for MHC and sodium dodecyl sulfate stability of the peptide/MHC complex. J. Immunol. 156:3815.[Abstract]
39. Viner, N. J., C. A. Nelson, B. Deck, E. R. Unanue. 1996. Complexes generated by the binding of free peptides to class II MHC molecules are antigenically diverse compared with those generated by intracellular processing. J. Immunol. 156:2365.[Abstract]
40. Grewal, L. S., K. D. Moudgil, E. E. Sercarz. 1995. Hindrance of binding to class II major histocompatibility complex molecules by a single amino acid residue contiguous to a determinant leads to crypticity of the determinant as well as lack of response to the protein antigen. Proc. Natl. Acad. Sci. USA 92:1779.[Abstract/Free Full Text]
41. Harcourt, G., A-P. Batocchi, S. Hawke, D. Beeson, N. Pantic, L. Jacobson, N. Willcox, A. Vincent, J. Newsom-Davis. 1993. Detection of -subunit isoforms in human muscle acetylcholine receptor by specific T cells from a myasthenia gravis patient. Proc. R. Soc. London Ser. B. 254:1.[Medline]
42. Chen, W., N. J. Ede, D. C. Jackson, J. McCluskey, A. W. Purcell. 1996. CTL recognition of an altered peptide associated with asparagine bond rearrangement. Implications for immunity and vaccine design. J. Immunol. 157:1000.[Abstract]
43. Molberg, Ø., S. N. Mcadam, R. Körner, H. Quarsten, C. Kristiansen, L. Madsen, L. Fugger, H. Scott, O. Norén, P. Roepstorff, et al 1998. Tissue transglutaminase selectively modifies gliadin peptides that are recognized by gut-derived T cells in celiac disease. Nat. Med. 4:713.[Medline]
44. Elson, C. J., R. N. Barker, S. J. Thompson, N. A. Williams. 1995. Immunologically ignorant autoreactive T cells, epitope spreading and repertoire limitation. Immunol. Today 16:71.[Medline]
45. Hawke, S., H. Matsuo, M. Nicolle, P. Wordsworth, L. Corlett, E. Spack, S. Deshpande, P. Driscoll, and N. Willcox. 1999. Cross-restriction of a T cell clone to HLA-DR alleles associated with rheumatoid arthritis: clues to arthritogenic peptide motifs. Arthritis Rheum. Vol 42 In press.
46. Matsuo, H., Hidenori, M., L. Corlett, S. Hawke, M. Nicolle, P. Driscoll, S. Deshpande, E. Spack, and N. Willcox. 1999. Distant interactions between dimorphisms in HLA-DR4 radically affect recognition of defined peptides by a specific T cell clone. Int. Immunol. In press.

This article has been cited by other articles:

				J. Wahlstrom, A. Eklund, and J. Grunewald Reactivity to Mycobacterial Antigens by Patients with Lofgren's Syndrome as a Model of Phenotypic Susceptibility to Disease and Disease Progression Am. J. Respir. Crit. Care Med., October 1, 2009; 180(7): 685 - 686. [Full Text] [PDF]

				R. Rayment, T. W. Kooij, W. Zhang, C. Siebold, M. F. Murphy, D. Allen, N. Willcox, and D. J. Roberts Evidence for the Specificity for Platelet HPA-1a Alloepitope and the Presenting HLA-DR52a of Diverse Antigen-Specific Helper T Cell Clones from Alloimmunized Mothers J. Immunol., July 1, 2009; 183(1): 677 - 686. [Abstract] [Full Text] [PDF]

				J. Grunewald and A. Eklund State of the Art. Role of CD4+ T Cells in Sarcoidosis Proceedings of the ATS, August 15, 2007; 4(5): 461 - 464. [Abstract] [Full Text] [PDF]

				J. Grunewald, A. Eklund, and O. Olerup Human Leukocyte Antigen Class I Alleles and the Disease Course in Sarcoidosis Patients Am. J. Respir. Crit. Care Med., March 15, 2004; 169(6): 696 - 702. [Abstract] [Full Text] [PDF]

				J Grunewald, J Wahlstrom, M Berlin, H Wigzell, A Eklund, and O Olerup Lung restricted T cell receptor AV2S3+ CD4+ T cell expansions in sarcoidosis patients with a shared HLA-DR{beta} chain conformation Thorax, April 1, 2002; 57(4): 348 - 352. [Abstract] [Full Text] [PDF]

				Z. Chen, N. Dudek, O. Wijburg, R. Strugnell, L. Brown, G. Deliyannis, D. Jackson, F. Koentgen, T. Gordon, and J. McCluskey A 320-Kilobase Artificial Chromosome Encoding the Human HLA DR3-DQ2 MHC Haplotype Confers HLA Restriction in Transgenic Mice J. Immunol., March 15, 2002; 168(6): 3050 - 3056. [Abstract] [Full Text] [PDF]

				M. Yassai, A. Afsari, J. Garlie, and J. Gorski C-Terminal Anchoring of a Peptide to Class II MHC Via the P10 Residue Is Compatible with a Peptide Bulge J. Immunol., February 1, 2002; 168(3): 1281 - 1285. [Abstract] [Full Text] [PDF]

				R. Raju, E. G. Spack, and C. S. David Acetylcholine Receptor Peptide Recognition in HLA DR3-Transgenic Mice: In Vivo Responses Correlate with MHC-Peptide Binding J. Immunol., July 15, 2001; 167(2): 1118 - 1124. [Abstract] [Full Text] [PDF]

				W. Wienhold, G. Malcherek, C. Jung, S. Stevanovic, G. Jung, H. Schild, and A. Melms An example of immunodominance: engagement of synonymous TCR by invariant CDR3{beta} Int. Immunol., June 1, 2000; 12(6): 747 - 756. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Nagvekar, N. Articles by Willcox, N. Search for Related Content PubMed PubMed Citation Articles by Nagvekar, N. Articles by Willcox, N.