Five HLA-DP Molecules Frequently Expressed in the Worldwide Human Population Share a Common HLA Supertypic Binding Specificity

T cells recognize a complex between a specific MHC type and a particular Ag-derived epitope. Thus, a given epitope elicits a response only in individuals who express an MHC molecule capable of binding that particular epitope (1–4). MHC molecules are extremely polymorphic, with hundreds of different variants known in humans (5, 6). Much of the observed polymorphism is concentrated in the residues hypothesized to line the peptide-binding groove and form the specific pockets that engage the amino acid side chains of the peptide ligand. Therefore, addressing multiple MHC-binding specificities is required to allow coverage of the human patient populations targeted by an epitope-specific diagnostic or therapeutic intervention.

The issue of population coverage in relation to MHC polymorphism is further complicated by the fact that different MHC types are expressed at dramatically different frequencies in different ethnicities. Thus, without careful consideration, an epitope set associated with an ethnically biased population could result in decreased applicability and usefulness for any diagnostic immuno-prophylactic or -therapeutic applications (7, 8). One means of circumventing the problem of MHC polymorphism relies on the selection of epitopes restricted by multiple MHC types. In the case of HLA class I, the identification of promiscuous epitopes is facilitated by the use of MHC supertypes, which defines sets of A and B class I molecules associated with largely overlapping peptide-binding repertoires (9–12).

In the context of HLA class II, four loci must be considered: DRB1, DRB3/4/5, DQ, and DP. Like HLA class I, each locus is polymorphic, with specific alleles occurring with often widely variable frequencies in different ethnicities. Several studies suggested the existence of class II supertypes encompassing several DR and DQ specificities that, like class I supertypes, describe sets of molecules sharing largely overlapping peptide-binding repertoires (13–26).

Unlike the situation with HLA-DR and -DQ, for which hundreds of T cell epitopes have been identified and reported in the literature, relatively few HLA-DP epitopes have been described [e.g., see the listings compiled by the Immune Epitope Database: www.iedb.org (27)]. Similarly, knowledge of the binding repertoires and specificities of HLA-DP alleles is, by comparison, somewhat limited. This is due, in part, to the common assumption that DP molecules are less important in the immune response than DR or DQ molecules, because the cell surface expression level of DP molecules seems to be ∼10-fold lower than for DR or DQ (28–30). This low expression level also makes DP more difficult to isolate, purify, and characterize, which may have limited the number of studies conducted with these molecules. However, a growing body of literature indicates that DP encodes fully functional molecules, restricting epitope responses in the context of cancer, allergy, and infectious disease (31–44).

A putative motif for DPB1*0201, present in ∼20% of the general population, has been defined using peptide-elution methodology (45, 46). The most frequent DPB alleles are DPB1*0401 and DPB1*0402, both of which are present in about one third of the individuals in the general population. Using peptide-binding studies, Maillere and coworkers defined motifs for both of these DPB1 subtypes (47, 48). Interestingly, these investigators showed that DPB1*0401 and DPB1*0402 share very similar peptide-binding specificity, with nearly identical main anchor motifs, leading them to hypothesize the existence of an HLA-DP supertype. The same investigators also noted that DPB1*0402 and DPB1*0201 differ by only a single residue in the P4 secondary pocket and that the two molecules share identical specificity at the main P1 and P6 anchor positions. Accordingly, they hypothesize that DP4 and DP2 molecules will also share a common repertoire of peptides. In contrast, Dong et al. (34) described a motif for DP9 that differs considerably from that recognized by DP4 and DP2 molecules.

In the current study, using comprehensive panels of single substitution analogs and large peptide libraries, we derived novel detailed binding motifs for two additional HLA-DP molecules: DPB1*0101 and DPB1*0501. We also provide more detailed motifs for DPB1*0201, DPB1*0401, and DPB1*0402, previously characterized on the basis of sets of eluted ligands and/or limited sets of substituted peptides. Unexpectedly, all five DP molecules, originally selected only on the basis of their frequency in human populations, were found to share largely overlapping peptide motifs. This led to the hypothesis that these molecules might also share largely overlapping peptide-binding repertoires. To test this hypothesis, we used a large set of binding data, encompassing several hundred nonredundant peptides, to examine the cross-reactivity between these five common DP molecules. Our findings demonstrate that the DP supertype, originally hypothesized by Malliere and coworkers, extends far beyond what was originally envisioned and includes at least three additional very common DP specificities. These DP supertype molecules are found in >90% of the human population; thus, these findings have important implications for epitope-identification studies and the monitoring of human class II-restricted immune responses.

Peptides for screening studies were purchased from Mimotopes (Clayton, Victoria, Australia) and/or A & A Labs (San Diego, CA) as crude material on a small (1-mg) scale. Peptides used as radiolabeled ligands were synthesized on a larger scale by A & A Labs and purified (>95%) by reversed-phase HPLC.

MHC molecules were purified from EBV-transformed homozygous cell lines by mAb-based affinity chromatography, essentially as described in detail elsewhere (49). Briefly, cells were maintained in vitro by culture in RPMI 1640 medium (Flow Laboratories, McLean, VA), supplemented with 2 mM l-glutamine (Life Technologies, Grand Island, NY), 100 U (100 mg/ml) penicillin-streptomycin solution (Life Technologies), and 10% heat-inactivated FCS (Hazelton Biologics, Lenexa, KS). Large-scale cultures were maintained in roller bottles. HLA molecules were purified from cells lysed at a concentration of 10⁸ cells/ml in 50 mM Tris-HCl (pH 8.5), containing 1% (v/v) Nonidet P-40 (NP-40), 150 mM NaCl, 5 mM EDTA, and 2 mM PMSF. Lysates were passed through 0.45-μm filters and cleared of nuclei and debris by centrifugation at 10,000 × g for 20 min.

For affinity purification, columns of inactivated Sepharose CL4B and Protein A Sepharose were used as precolumns. HLA-DR, -DQ, and -DP molecules were captured by repeated passage of lysates over LB3.1 (anti–HLA-DR), SPV-L3 (anti–HLA-DQ), and B7/21 (anti–HLA-DP) columns. After two to four passages of the lysate, Ab columns were washed with 10-column volumes of 10 mM Tris-HCl (pH 8) with 1% (v/v) NP-40, 2-column volumes of PBS, and 2-column volumes of PBS containing 0.4% (w/v) n-octylglucoside. MHC molecules were then eluted with 50 mM diethylamine in 0.15 M NaCl containing 0.4% (w/v) n-octylglucoside (pH 11.5). A 1/26 volume of 2.0 M Tris (pH 6.8) was added to the eluate to reduce the pH to 8. Eluates were concentrated by centrifugation in Centriprep 30 microconcentrators at 2000 rpm (Millipore, Bedford, MA). Protein purity, concentration, and the effectiveness of depletion steps were monitored by SDS-PAGE and bicinchoninic acid assay.

Assays to quantitatively measure peptide binding to class II MHC molecules are based on the inhibition of binding of a high-affinity radiolabeled peptide to purified MHC molecules. This assay system was described in detail elsewhere (49), but it is briefly described in general terms. The optimal assay conditions for each DP molecule tested in the current study are summarized in the 6Results. To measure the capacity of peptide ligands to bind MHC molecules, 0.1–1 nM radiolabeled peptide was coincubated at room temperature or 37°C with 1 μM to 1 nM purified MHC in the presence of a mixture of protease inhibitors (EDTA, pepstatin A, N-ethylmaleimide, 1,10-phenanthroline, PMSF, and Nα-p-tosyl-l-lysine chloromethyl ketone). Following a 2–4 d incubation, the percentage of MHC-bound radioactivity was determined by capturing MHC/peptide complexes on LB3.1 (DR), L243 (DR), HB180 (DR/DQ/DP), SPV-L3 (DQ), or B7/21 (DP) Ab-coated Optiplates (Packard Instrument, Meriden, CT) and determining bound cpm using the TopCount (Packard Instrument) microscintillation counter. For competitive assays, the concentration of peptide yielding 50% inhibition of the binding of a radiolabeled standard probe peptide are calculated. Under the conditions used, where [label] < [MHC] and IC₅₀ ≥ [MHC], the measured IC₅₀ values are reasonable approximations of the true K_d values (50, 51). Each competitor peptide was tested at six concentrations covering a 100,000-fold dose range in three or more independent experiments. As a positive control, the unlabeled version of the radiolabeled probe was also tested in each experiment.

The peptide-binding repertoire (R) of each MHC molecule was defined as the set of peptides that bind that molecule with an affinity ≥1000 nM. The relationship between two molecules is measured by determining cross-reactivity and repertoire overlap. Cross-reactivity is defined as the fraction of peptides that bind one MHC molecule (i) that also bind a second (ii): (R_i and R_z)/(R_i) × 100%. Repertoire overlap is defined as the fraction of peptides binding either molecule that bind both: (R_i and R_z)/(R_i or R_z) × 100%.

Population coverage was calculated as previously described (52). Gene frequencies (gf) for each HLA allele were calculated from population frequencies obtained from the MHC database [dbMHC; National Center for Biotechnology Information (NCBI), (53)]. Phenotypic frequencies (pf) were calculated using the binomial distribution formula: pf = 1− (1 − ∑gf)². To obtain total potential population coverage, no linkage disequilibrium was assumed.

Analysis of the single amino acid substitution (SAAS) binding data was performed essentially as described previously for analysis of positional scanning combinatorial library data (40). Briefly, the IC₅₀ nanomolar value of each substituted peptide was standardized relative to the geometric mean IC₅₀ nanomolar value of the wild-type peptide. Relative binding values were then represented in a 20 × 15-aa position matrix. Next, an average (geometric) relative binding affinity (ARB) was calculated for each position, encompassing all 20 possible residues. Finally, for each position, the ratio of ARB for the entire library/position-specific ARB was derived. We denominated this ratio, which describes the factor by which the geometric average binding affinity associated with all 20 residues at a specified position differs from that of the average binding affinity of the entire library, as the specificity factor (SF). Positions with the greatest specificity have the greatest SF value. Primary anchor positions were defined as those with SF ≥ 2.4. This criterion identifies positions where the majority of residues are associated with significant decreases in binding capacity.

HLA class II polymorphism is very extensive, and different MHC types are expressed at different frequencies in different ethnicities. Although DP α- and β-chains are polymorphic, polymorphism of the α-chain is generally limited to a few residues; for the most part, based on HLA-DR crystal structures, these residues are in regions not likely to affect binding specificity (54–56). Furthermore, DPA is overrepresented by just a few alleles, where DPA*0103 represents ∼60% of all DPA genes, and four alleles account for ∼90%. For these reasons, most studies focus on HLA-DP β-chains as the main determinant of HLA-DP peptide-binding specificity. The frequency of DPB1 allelic variants expressed in various ethnicities [dbMHC NCBI, (53)] was compiled with the goal of selecting a panel of DP molecules affording coverage of the majority of the human population, irrespective of ethnicity (Table I).

Accordingly, we selected a panel of five HLA-DPB1 molecules (DPB1*0101, DPB1*0201, DPB1*0401, DPB1*0402, and DPB1*0501) representing the most common allelic variants in major populations worldwide. Each DPB1 allele selected is encountered with an average phenotypic frequency >15% across the seven main populations for which frequency data are available (Australia, Europe, North America, Oceania, South America, Southeast Asia, and sub-Saharan Africa). In aggregate, these molecules cover ∼92% of the average population at the DPB1 locus. Importantly, they afford equally high and balanced population coverage, irrespective of the ethnicity considered (range 75–99%, depending on the specific ethnicity considered). Of the targeted specificities, quantitative peptide-binding assays were reported by Maillere and coworkers for DPB1*0401 and DPB1*0402 (47, 48). To establish a complete panel covering all of the common specificities, we undertook development of binding assays for DPB1*0101, DPB1*0201, and DPB1*0501.

To establish assays for the DPB1*0101, DPB1*0201, and DPB1*0501 molecules, we first assembled a panel of potential binders, including peptides known to be HLA-DP epitopes or endogenous ligands presented by these specificities or known to bind other HLA-DP or class II molecules with high affinity. Candidate peptides were radiolabeled and tested in direct-binding assays for their capacity to bind DP molecules purified as described in 1Materials and Methods.

We found that a two-residue C-terminal truncation of a previously described high-affinity DP4 ligand [human aminopeptidase 285–299, sequence EKKYFAATQFEPLAA (47)] was the optimal radiolabeled ligand for DPB1*0101 purified molecules, whereas a two-residue N-terminal truncation of the same peptide (i.e., residues 287–301, sequence KYFAATQFEPLAARL) was optimal for DPB1*0201. The optimal radiolabeled ligand for DPB1*0501 was an M₁₁ to Y, C₇ to T, and R₁₅ to A analog of a DP5 epitope [(57); sequence IGRIAETILGYNPSA]. The specificity of each assay was verified by assessing capture with various HLA class II locus-specific Abs (Supplemental Fig. 1).

Various assay conditions were examined to assess the effects of pH, temperature, duration of incubation, and MHC capture time. A representative analysis is shown in Supplemental Fig. 1. These experiments allowed us to identify optimal conditions for each specific molecular assay (Table II). In general, optimal assay conditions were similar to those used for HLA-DR assays (15, 58, 59). However, signals in DP-binding assays were substantially amplified by using an extended MHC-Ab capture step of up to 20–24 h. These assay conditions also proved optimal for the DPB1*0401 and DPB1*0402 molecules and were used in subsequent competitive-inhibition assays.

Following the biochemical validation of the assays described above, we sought to further validate them biologically by examining the correlation between assay specificity and the binding capacity of known HLA-restricted epitopes. We queried the Immune Epitope Database [www.iedb.org (27)] and the published literature (26, 60, 61) and selected a panel of 27 epitopes and endogenous ligands presented by one or more of the five common DP specificities selected in the current study (Table III). To control for locus specificity, we also selected a panel of 24 HLA-DQ–restricted epitopes.

Each of the peptides was tested for its capacity to bind DPB1*0101, DPB1*0201, DPB1*0401, DPB1*0402, and DPB1*0501. The resulting IC₅₀ nanomolar values are shown in Table III. In 90% of the cases, a DP-restricted epitope bound its corresponding restricting allele with a binding affinity ≥1000 nM; >50% of the cases were associated with affinities ≥100 nM (Fig. 1).

A striking degree of cross-reactivity was also noted between the different DP allelic variants in the context of the panel of HLA-DP–restricted epitopes and endogenous ligands, with the majority (71.4%) of peptides binding four or more of the five molecules with an affinity ≥1000 nM (Fig. 2). This might have been predicted based on the fact that a similar radiolabeled ligand was found to be suitable for four of the DP assays studied herein, as well as previous observations by Maillere and coworkers regarding the existence of a potential HLA-DP supertypic specificity (47). In contrast, only occasional DP binding and cross-reactivities were detected with the panel of DQ-restricted epitopes and ligands. In conclusion, the data presented in this section support the biological relevance of the assays developed.

Next, we sought to use the newly developed assays to define motifs for DPB1*0101, DPB1*0201, and DPB1*0501. To accomplish this task, we took advantage of the observation that the human aminopeptidase 285–299 peptide (sequence EKKYFAATQFEPLAA) used as the ligand for the DP*0101 and DP*0402 assays, fortuitously bound, albeit with varying affinities (0.66–8.3 nM), to the each of the various DP molecules in consideration. A full set of SAAS analogs was synthesized; all 19 possible amino acid substitutions were represented for each position (1–15) of the wild-type peptide.

The entire panel of analogs was then tested for binding in each of the novel DP assays. Binding data were evaluated as described in 1Materials and Methods and summarized as ARBs in allele-specific 20 × 15-aa position matrices. The role of each position was determined by calculating an SF for each position, which describes the factor by which the geometric average binding affinity associated with all 20 residues at a specified position differs from that of the average affinity of the entire library (62, 63). Positions with the greatest specificity have the greatest SF value. Primary anchor positions were defined as those with SF ≥ 2.4.

In a set of preliminary experiments, we tested the panel of SAASs in the DPB1*0401 and DPB1*0402 assays, for which Maillere and coworkers defined a motif based on two hydrophobic/aromatic main anchors at positions 1 and 6 of a prototypical class II nine-residue core (47). Our results confirmed these observations. For DPB1*0401, positions F5 and F10, corresponding to the P1 and P6 anchors, had the greatest influence on binding capacity. These positions were associated with SFs of 15 and 23, respectively; 13 substitutions at position F5 and 14 substitutions at F10 affected binding by ≥10-fold (Supplemental Table IA). No other positions were associated with an SF > 2.0 or had more than two substitutions with a ≥10-fold effect on binding. For DPB1*0402, positions F5, F10, and L13, corresponding to the P1, P6, and P9 anchors, had SFs of 7.7, 8.0, and 3.4, respectively. At positions F5 and F10, 11 substitutions affected binding by ≥10-fold, as did 7 substitutions at L13 (Supplemental Table IB). For both alleles, the most preferred residues at the P1 and P6 anchors were found to be aromatic or hydrophobic. Taken together, these results confirm the observations of Maillere’s group and further validate the assays established.

When the same panel of SAASs was tested for DPB1*0101 binding, positions F5 and F10 had SFs of 3.7 and 2.4, respectively, and nine and seven substitutions at these positions affected binding by ≥10-fold (Table IV). These main anchor positions are identical to the P1 and P6 anchor positions similarly defined for the DPB1*0401 and 0402 molecules. In addition, a very prominent role was demonstrated for position N-2, corresponding to residue K3, for which an SF of 9.4 was calculated, and 13 of the substitutions affected binding to DPB1*0101. No other position was associated with SF > 1.8 or more than three substitutions with >10-fold effects on DPB1*0101 binding.

The SAAS panel was next tested for DPB1*0201 binding. At residues F5 and F10, again corresponding to the P1 and P6 anchors, SFs of 9.1 and 13 were calculated, respectively, and 12 and 14 substitutions affected binding by ≥10-fold, respectively (Table V). In addition, a very prominent role was once again demonstrated for position N-2, corresponding to residue K3; an SF of 3.1 was calculated, and eight of the substitutions affected DPB1*0201 binding by >10-fold. No other position was associated with an SF > 2.1 or more than five substitutions with >10-fold effects on DPB1*0201 binding.

Finally, the panel of SAASs was tested for DPB1*0501 binding. Again, the most significant influences on binding were detected at residues F5 and F10, corresponding to the P1 and P6 anchors, for which SFs of 9.0 and 9.6 were found, respectively; 13 substitutions for each site affected binding by ≥10-fold (Table VI). Position N-4, corresponding to residue E1, had an SF of 2.3, just below the threshold we used to define main anchors, and seven of the substitutions tested affected binding by >10-fold, suggesting that this position has at least some secondary influence on DPB1*0501 binding capacity.

Taken together, the data in this section defined comprehensive quantitative binding motifs for each of the five most common HLA-DP specificities in the general population. In addition to identifying specific main anchor positions and preferences, these motifs provide information about specificity at other positions and the effect on binding of each of the 20 naturally occurring amino acids.

A summary of the motif associated with each of the five common DP allelic variants is presented in Table VII. All five molecules are associated with a P1–P6 spacing of two main anchor residues, for which a prominent specificity for hydrophobic and aromatic residues is generally noted.

However, each allele is also associated with it own unique binding specificity; in two cases, the position 2 residues to the N-terminal side of this core also seemed to significantly affect overall binding capacity. These results, indicating significant similarity in the mode and specificity of binding, are in agreement with the testing of the known epitopes described above, which showed a high degree of similarity in the binding pattern of different DP molecules and clearly demonstrated that each allele is associated with its own binding specificity.

In conjunction with the observations of Maillere and coworkers, these results define a broad DP supertype that includes several DP molecules with specificity for aromatic or hydrophobic residues at positions 1 and 6. This proposed shared specificity (the DP supermotif) is shown in Table VII.

The data presented above demonstrated that the various allelic variants considered have a common motif, but they did not address to what extent corresponding overlaps in binding repertoire exist. To address this point, we tested a panel of 425 nonredundant peptides for its capacity to bind each DP molecule. The set of peptides consisted of 15-mers, overlapping by 10 residues, spanning the entire sequences of the Phleum pratense 1, 2, 3, 4, 5, 6, 7, 11, 12, and 13 pollen Ags, which are implicated in allergic reaction to timothy grass.

We found that 250 (58.8%) of the P. pratense peptides bound none of the DP molecules, and 60 (14.1%) bound only one (Fig. 2). Thirty-five (8.2%) other peptides bound two of the different DP molecules tested. Thirty-one (7.3%) peptides bound three DP molecules, 30 (7.1%) bound four DP molecules, and 19 (4.5%) bound all five DP molecules. Altogether, 80 (18.7%) peptides were promiscuous DP-binding peptides, binding three or more DP molecules. This incidence of promiscuity is appreciably higher than would be expected.

The shared specificities of these DP molecules was also demonstrated by measuring the degree of allele-to-allele cross-reactivity, defined as the fraction of peptides binding a specific DP allele that also bind a second DP allele (Table VIII). Overall, 35–87% (mean, 57.6%) of the peptides that bind one DP molecule also bind any given second DP molecule. The greatest cross-reactivity was seen between DPB1*0401 and DPB1*0201; for which 86.7% of the DPB1*0401-binding peptides also bound DPB1*0201. In general, the lowest cross-reactivities were found for DPB1*0501; an average of only 40% of the peptides binding one DP type also bound DPB1*0501.

Repertoire overlap, defined as the fraction of peptides that bind either of two molecules in a pair that bind both molecules, is a second method for assessing similarities in binding specificity (Table IX). Overall, on average, ∼40% of the repertoires of any pair of DP alleles studied overlap.

Table X presents a list of the peptides binding four or more of the DP molecules with an affinity ≥1000 nM. The inferred main anchors, matching the DP supermotif from Table VII in nature and spacing, are highlighted in bold type. Overall, 34 of 49 (69.4%) of the promiscuous peptides carry the proposed supermotif (Fig. 3). Indeed, of 159 peptides carrying the DP supermotif, 53 (33.3%) bound more than two DP molecules; by contrast, only 27 of 266 (10.2%) peptides that did not carry the supermotif bound three or more DP molecules.