MHC class I (MHC I)–restricted virus-specific CTLs are implicated as critical components in the control of this naturally occurring lentivirus and in the protective immune response to the successfully applied attenuated equine infectious anemia virus vaccine in the horse. Nevertheless, the structural basis for how the equine MHC I presents epitope peptides remains unknown. In this study, we investigated the binding of several equine infectious anemia virus–derived epitope peptides by the ability to refold recombinant molecules and by thermal stability, and then by determining the x-ray structure of five peptide–MHC I complexes: equine MHC class I allele (Eqca)-N*00602/Env-RW12, Eqca-N*00602/Gag-GW12, Eqca-N*00602/Rev-QW11, Eqca-N*00602/Gag-CF9, and Eqca-N*00601/Gag-GW12. Although Eqca-N*00601 and Eqca-N*00602 differ by a single amino acid, Eqca-N*00601 exhibited a drastically different peptide presentation when binding a similar CTL epitope, Gag-GW12; the result makes the previously reported function clear to be non–cross-recognition between these two alleles. The structures plus Eqca-N*00602 complexed with a 9-mer peptide are particularly noteworthy in that we illuminated differences in apparent flexibility in the center of the epitope peptides for the complexes with Gag-GW12 as compared with Env-RW12, and a strict selection of epitope peptides with normal length. The featured preferences and unconventional presentations of long peptides by equine MHC I molecules provide structural bases to explain the exceptional anti-lentivirus immunity in the horse. We think that the beneficial reference points could serve as an initial platform for other human or animal lentiviruses.

Equine infectious anemia virus (EIAV), the infectious etiology of EIA, was the first virus identified as causing animal infectious diseases in 1904 and is also a lentivirus that shares genetic and structural similarities with HIV-1 (13). However, unlike the progressive degenerative pathologies of other human or animal lentiviruses, horse-infected EIAV only causes dynamic recurring disease cycles and associated waves of viremia in the first year of infection (4). In most equids, this chronic disease stage is followed by a long-term asymptomatic state that can last for the lifespan of the infected horse, despite carrying highly virulent EIAV (2, 47). This distinguished disease progress makes horse infection by EIAV a pivotal animal model for investigating the mechanism of successful persistence of nonprogression and for investigating the vaccine development for lentiviruses, including HIV-1 (3, 8).

CTL responses contribute to the resolution of early episodes of viremia and clinical disease in the EIAV-infected horse (913). In the first year of a rapid onset of consecutive disease cycles, the peak bases of the viremia waves with irregular intervals are associated with the appearance of robust CTL responses (12, 14). During this process, virus-specific CTLs recognize epitope peptides that are 8–11 aa in length, which, in most cases, are presented by equine MHC class I (MHC I) molecules (15). Thus far, >20 CD8+ T cell (also called CTL) epitopes derived from the EIAV proteome have been identified (12, 1618). The CTL epitopes Env-RW12 (RVEDVTNTAEYW) and Gag-GW12 (GSQKLTTGNCNW) derived from the EIAV Env and Gag proteins, respectively, are two of the earliest defined CD8+ T cell epitopes (12) that are commonly used in EIAV-specific T cell immunity studies of equines (9, 12, 19, 20). Env-RW12 and Gag-GW12 tetramer+ CD8+ T cells have been identified in vivo as early as 14 d after EIAV inoculation, and the cell frequencies remain at a detectable level during the acute disease episodes (14). Env-RW12 has been defined as an immunodominant CTL epitope and can trigger a robust level of CTL immunity compared with Gag-GW12 (11, 13, 14, 20). However, escape viral variants under the selection pressure of Env-RW12–specific T cell responses have also been identified in most infected horses, although it has been reported that Env-RW12–specific T cells can recognize the rare mutants at a reduced level (9, 11, 13, 14, 16). Moreover, despite the immunodominance of Env-RW12, no correlation between Env-RW12–specific CD8+ T cells and the control of viremia has been elucidated. In contrast, CTL epitope Gag-GW12 variants rarely arose during the EIAV infection, and the persistence of viral control in horses correlates with Gag-GW12–specific CD8+ T cell responses (14, 16).

The most extraordinary feature of the epitope peptides Env-RW12 and Gag-GW12 is their different recognition in different horses. It has been determined that the virus-specific CTL responses of different individuals are predominantly restricted by the MHC I alleles. Immunogenic epitope peptides recognized by virus-specific CTLs are presented by MHC (equine leukocyte Ag [ELA] in the horse and HLA in humans) molecules. The subtle changes in the peptide-binding region of MHC molecules that leads to amino acid differences have a significant impact on the immune responses to a similar peptide (21, 22). It has been reported that a single amino acid polymorphism within the α2 domain of ELA can alter or even abolish recognition of Env-RW12 and another EIAV-derived CTL epitope peptide, Rev-QW11 (QAEVLQERLEW) (20). Equine MHC class I allele (Eqca)-N*00602 (formerly designated as the equine MHC I 7-6 allele) is associated with the serologically defined ELA-A1 alleles (19, 2325), which are prevalent in the Arabian horse and mixed-breed pony. Another ELA allele, Eqca-N*00601, formerly designated as the equine MHC I 114 allele, which differs from Eqca-N*00602 allele by a single amino acid (E152V) within the peptide-binding groove (20, 25), resulted in the abrogation of epitope peptide Gag-GW12–specific CTL recognition even though Gag-GW12 can still bind to Eqca-N*00601 (20). This phenomenon highlights the immunologic importance of MHC I polymorphism for antiviral immunity. However, the structural mechanism has not been explored.

The unique disease pathogenesis of EIAV infection in horses, the distinguishing characteristics of the anti-EIAV CTL immunity in an equine MHC I–restricted manner (26, 27), the anti-EIAV vaccine, which has been used in China since the 1970s, and the successful clearing of the EIA disease (EIAD) by the 1980s (28) all make it worthwhile to determine an in-depth structure of the preferences of equine MHC I molecule. In this study, we chose a structural approach to explore the mechanism of the successful anti-lentivirus immunity. A series of structures highlighting the interactions between the EIAV-derived epitope peptides and the correlated equine MHC I alleles Eqca-N*00602 and Eqca-N*00601, together with biochemical analyses, was performed. The structural results led to a better understanding of the preference of equine MHC I molecules with unconventional epitope presentation as an efficacious avenue for anti-lentivirus CTL immunity in the horse and may also be of practical use for the vaccine development of other lentiviruses.

ELA-A1–restricted EIAV CTL epitope peptides named Env-RW12 (RVEDVTNTAEYW), Gag-GW12 (GSQKLTTGNCNW), and Rev-QW11 (QAEVLQERLEW) (12) were selected (Table I). Additionally, to screen potential shorter peptides binding to the ELA-A1 haplotype Eqca-N*00602, the NetMHCpan 2.3 server (http://www.cbs.dtu.dk/services/NetMHCpan-2.3/) was used to predict the candidate peptides (29, 30). A total of 15 peptides (Table I) with higher scores were chosen. The peptides were synthesized and then purified through reverse-phase HPLC (SciLight Biotechnology) with a purity of >90%. The peptides were stored as freeze-dried powders at −80°C and were dissolved in DMSO before use.

The cDNA coding the extracellular domain (residues 1–274) of equine Eqca-N*00602 (GenBank accession no. AY225155) and Eqca-N*00601 (GenBank accession no. AY374512; http://www.ncbi.nlm.nih.gov/genbank) genes was synthesized (Beijing Sunbiotech) and ligated into the cloning vector pMD18-T (Takara). After cutting with the restriction enzymes NdeI and XhoI, the cDNA fragments were ligated into the prokaryotic expression vector pET21a(+) (Novagen). The expression plasmids were transformed into Escherichia coli strain BL21 (DE3). The expression vector pET21a(+) for the mouse β2-microglobulin (β2m) was previously constructed in the laboratory (31). All proteins were expressed as inclusion bodies and purified as previously described (32). The purified inclusion bodies of the three proteins were solubilized in 8 M urea buffer with a concentration of 30 mg/ml.

The preparation of complexes with different peptides was performed using the gradual dilution method as described previously (33, 34). In brief, the ELA-A1, β2m, and peptide in a 1:1:3 molar ratio were gradually diluted into the refolding buffer (100 mM Tris [pH 8.0], 400 mM l-arginine, 2 mM EDTA, 5 mM reduced glutathione, and 0.5 mM glutathione disulfide) and incubated for 48 h at 4°C. Subsequently, the remaining soluble portion within the refolding system was concentrated and purified by chromatography on a Superdex 200 16/60 column followed by Resource Q anion-exchange chromatography (GE Healthcare) (35).

Crystals of equine MHC I molecules were screened using the hanging drop vapor diffusion method with the Crystal Screen kit I/II and Index kit (Hampton Research) at 291 K. The crystals of Eqca-N*00601 and Eqca-N*00602 complexed to different peptides were obtained in solution of 0.02 M magnesium chloride hexahydrate, 0.1 M HEPES (pH 7.5), 22% (w/v) polyacrylic acid sodium salt 5100 at a final concentration of 10–20 mg/ml. The pH of the solution and the concentration of the polyacrylic acid sodium salt 5100 were optimized slightly for different complexes. For cryoprotection, the crystals were transferred to reservoir solutions containing 20% glycerol. X-ray diffraction data were collected at 100 K at Shanghai Synchrotron Radiation Facility Beamline BL17U (Shanghai, China) and subsequently processed and scaled using the Denzo program and HKL-2000 software package (HKL Research). The data collection statistics are shown in Table II.

Table II.
X-ray diffraction data processing and refinement statistics
Eqca-N*00602/Env-RW12Eqca-N*00602/Gag-GW12Eqca-N*00601/Gag-GW12Eqca-N*00602/Rev-QW11Eqca-N*00602/Gag-CF9
Data processing 
 Space group P21 P4121P4121P4121P4121
 Cell parameters 
  a (Å) 82.5 69.8 69.6 69.0 69.5 
  b (Å) 71.4 69.8 69.6 69.0 69.5 
  c (Å) 99.9 207.1 206.6 205.5 206.5 
  α (°) 90.0 90.0 90.0 90.0 90.0 
  β (°) 102.9 90.0 90.0 90.0 90.0 
  γ (°) 90.0 90.0 90.0 90.0 90.0 
 Resolution range (Å)a 50.0–2.3 (2.38–2.3) 50.0–2.5 (2.59–2.5) 50.0–2.6 (2.69–2.6) 50.0–2.6 (2.69–2.6) 50.0–2.2 (2.28–2.2) 
 Total reflections 268,632 190,454 184,244 78,513 233,358 
 Unique reflections 50,685 18,138 17,060 7,697 26,518 
 Completeness (%) 99.5 (100.0) 98.9 (99.8) 99.2 (99.6) 94.9 (98.6) 98.7 (99.5) 
 Rmerge (%)b 12.0 (46.4) 8.4 (58.7) 5.8 (40.8) 7.5 (50.9) 5.8 (53.8) 
 I/σ 13.0 (3.9) 26.1 (5.3) 38.4 (7.0) 34.2 (5.1) 32.7 (4.9) 
Refinement 
 Rwork (%)c 20.0 22.1 21.1 21.2 22.8 
 Rfree (%) 23.4 26.5 27.6 26.8 26.6 
 RMSD 
  Bond lengths (Å) 0.006 0.005 0.008 0.004 0.008 
  Bond angles (°) 0.870 0.759 0.831 0.736 0.853 
 Ramachandran plot quality (%) 
  Most favored 89.4 86.5 86.8 86.5 88.3 
  Disallowed 
Eqca-N*00602/Env-RW12Eqca-N*00602/Gag-GW12Eqca-N*00601/Gag-GW12Eqca-N*00602/Rev-QW11Eqca-N*00602/Gag-CF9
Data processing 
 Space group P21 P4121P4121P4121P4121
 Cell parameters 
  a (Å) 82.5 69.8 69.6 69.0 69.5 
  b (Å) 71.4 69.8 69.6 69.0 69.5 
  c (Å) 99.9 207.1 206.6 205.5 206.5 
  α (°) 90.0 90.0 90.0 90.0 90.0 
  β (°) 102.9 90.0 90.0 90.0 90.0 
  γ (°) 90.0 90.0 90.0 90.0 90.0 
 Resolution range (Å)a 50.0–2.3 (2.38–2.3) 50.0–2.5 (2.59–2.5) 50.0–2.6 (2.69–2.6) 50.0–2.6 (2.69–2.6) 50.0–2.2 (2.28–2.2) 
 Total reflections 268,632 190,454 184,244 78,513 233,358 
 Unique reflections 50,685 18,138 17,060 7,697 26,518 
 Completeness (%) 99.5 (100.0) 98.9 (99.8) 99.2 (99.6) 94.9 (98.6) 98.7 (99.5) 
 Rmerge (%)b 12.0 (46.4) 8.4 (58.7) 5.8 (40.8) 7.5 (50.9) 5.8 (53.8) 
 I/σ 13.0 (3.9) 26.1 (5.3) 38.4 (7.0) 34.2 (5.1) 32.7 (4.9) 
Refinement 
 Rwork (%)c 20.0 22.1 21.1 21.2 22.8 
 Rfree (%) 23.4 26.5 27.6 26.8 26.6 
 RMSD 
  Bond lengths (Å) 0.006 0.005 0.008 0.004 0.008 
  Bond angles (°) 0.870 0.759 0.831 0.736 0.853 
 Ramachandran plot quality (%) 
  Most favored 89.4 86.5 86.8 86.5 88.3 
  Disallowed 

RMSD, root mean square deviation.

a

Values in parentheses are given for the highest resolution shell.

b

Rmerge = ∑hkli |Ii −〈I〉|∑hkliIi.

c

R = ∑hklFobs|−k|Fcal ||/ ∑hkl |Fobs|.

The structures of Eqca-N*00602/Env-RW12 were determined using molecular replacement with the program MOLREP. The search model was Protein Data Bank entry 1Q94 (36) with the water coordinates omitted. Extensive model building was performed manually using COOT (37) and restrained refinement using REFMAC5. Subsequent rounds of refinement were performed using the Phoenix refine program implemented in the PHOENIX package with isotropic ADP refinement and bulk solvent modeling (38). The stereochemical quality of the final model was assessed with the program PROCHECK (39). The structures of other equine MHC I molecules, that is, Eqca-N*00602/Gag-GW12, Eqca-N*00602/Rev-QW11, Eqca-N*00602/Gag-CF9, and Eqca-N*00601/Gag-GW12, were determined using molecular replacement with Eqca-N*00602/Env-RW12 as the search model.

The thermostabilities of Eqca-N*00602 and Eqca-N*00601 complexed with epitope peptides Env-RW12, Gag-GW12, and Rev-QW11 were assessed by circular dichroism (CD) spectroscopy. CD spectra were measured at 20°C on a Jasco J-810 spectropolarimeter equipped with a water-circulating cell holder. Briefly, all the purified protein samples were concentrated to 6 μM in 50 mM NaCl and 20 mM Tris-HCl (pH 8.0). The CD value was real-time monitored by using a 1-mm optical path length cell at 218 nm as the temperature was raised from 25 to 75°C at a rate of 1°C/min. The actual temperature of the protein solution was measured with a thermistor inserted into the cuvette. The proportion of denatured protein was calculated from the mean residue ellipticity (θ) using the standard method: fraction unfolded (%) = (θ − θN)/(θU − θN), where θN and θU are the mean residue ellipticity values in the fully folded and fully unfolded states, respectively. The midpoint transition temperature (Tm) was calculated using the data of denaturation curves in the program Origin 8.0 (OriginLab) (40).

The five structures of the equine MHC molecules were deposited in the Protein Data Bank (http://www.rcsb.org/pdb/home/home.do). The entries are as follows: Eqca-N*00602/Env-RW12: 4ZUV (http://www.rcsb.org/pdb/search/structidSearch.do?structureId=4ZUV); Eqca-N*00601/Gag-GW12: 4ZUW (http://www.rcsb.org/pdb/search/structidSearch.do?structureId=4ZUW); Eqca-N*00602/Gag-GW12: 4ZUT (http://www.rcsb.org/pdb/search/structidSearch.do?structureId=4ZUT); Eqca-N*00602/Rev-QW11: 4ZUS (http://www.rcsb.org/pdb/search/structidSearch.do?structureId=4ZUS); and Eqca-N*00602/Gag-CF9: 4ZUU (http://www.rcsb.org/pdb/search/structidSearch.do?structureId=4ZUU).

EIAV-derived peptides Env-RW12, Gag-GW12, and Rev-QW11 were identified as ELA-A1–restricted CTL epitopes with distinct immunogenicities (Table I) (14, 20). However, the biochemical characteristics of these peptides as they bind to equine MHC I molecules, which may impact the immune features of the epitope peptides, have still not been explored. By utilizing a CD spectroscopy–based assay, the thermostabilities of the three CTL epitopes Env-RW12, Gag-GW12, and Rev-QW11 complexed to two ELA-A1–related equine MHC I molecules Eqca-N*00601 and Eqca-N*00602, which share a similar sequence except for a single amino acid difference encoded at position 152 (V→E), were tested (Table II). All complexes produced similar spectra at 20°C; however, the midpoint thermal denaturation revealed compelling differences in the complex stability (Fig. 1). The Tm values were determined from the melting curves as described previously (4143). With Tm values of 62.41°C and 59.25°C when binding to Eqca-N*00601 and Eqca-N*00602, respectively, Env-RW12 had the highest binding affinity to these two molecules compared with the other peptides (see Table III). This result is in agreement with the immunodominance of CTL recognition of Env-RW12 (11, 13, 14, 20). When comparing the stabilities of the similar peptide binding to Eqca-N*00602 and Eqca-N*00601, we found that both Env-RW12 and Rev-QW11 had higher Tm values when complexed to Eqca-N*00601 compared with Eqca-N*00602 (62.41°C for Eqca-N*00601/Env-RW12 versus 59.25°C for Eqca-N*00602/Env-RW12 and 59.95°C for Eqca-N*00601/Rev-QW11 versus 47.25°C for Eqca-N*00602/Rev-QW11). However, Gag-GW12 was more stable when forming a complex with Eqca-N*00602 (51.25°C for Eqca-N*00601/Gag-GW12 versus 53.85°C for Eqca-N*00602/Gag-GW12). The results indicated a different binding mode of Gag-GW12 to the two closely related equine MHC I molecules compared with the peptides Env-RW12 and Rev-QW11. Taken together, the differences in peptide–MHC I stabilities were not sufficient to explain the differential CTL recognition of these peptides by Eqca-N*00601 and Eqca-N*00602.

Table I.
ELA binding peptides used in this study
TypeNameDerived ProteinPositionSequenceScoreaRefoldingb
Defined epitopes Env-RW12 Envc 195–206 RVEDVTNTAEYWd /e 
Gag-GW12 Gag 21–32 GSQKLTTGNCNW 
Rev-QW11 Rev 57–67 QAEVLQERLEW 0.037 
Predicted peptides Gag-CF9 Gag 172–180 CTSEEMNAF 0.513 
Gag-LW9 Gag 139–147 LTPRGYTTW 0.273 
Pol-EW9 Pol 1063–1071 ESMGGQTPW 0.199 
Pol-HF8 Pol 989–996 HTDNGTNF 0.411 
Gag-VL9 Gag 17–25 VTVQGSQKL 0.331 — 
Gag-LF9 Gag 67–75 LSGQEREAF 0.122 — 
Gag-ML9 Gag 226–234 MTARFIRGL 0.496 — 
Gag-DY8 Gag 333–340 DTLEEKMY 0.139 — 
Env-FM9 Env 136–144 FSYETNRSM 0.566 — 
Env-GF9 Env 124–132 GSFPGCRPF 0.500 — 
Env-TY9 Env 85–93 TTGGILWWY 0.377 — 
Env-NF9 Env 595–603 NTPDSIAQF 0.373 — 
Env-IY9 Env 461–469 IAASATMSY 0.505 — 
Rev-EW9 Rev 89–97 EVLQERLEW 0.049 — 
TypeNameDerived ProteinPositionSequenceScoreaRefoldingb
Defined epitopes Env-RW12 Envc 195–206 RVEDVTNTAEYWd /e 
Gag-GW12 Gag 21–32 GSQKLTTGNCNW 
Rev-QW11 Rev 57–67 QAEVLQERLEW 0.037 
Predicted peptides Gag-CF9 Gag 172–180 CTSEEMNAF 0.513 
Gag-LW9 Gag 139–147 LTPRGYTTW 0.273 
Pol-EW9 Pol 1063–1071 ESMGGQTPW 0.199 
Pol-HF8 Pol 989–996 HTDNGTNF 0.411 
Gag-VL9 Gag 17–25 VTVQGSQKL 0.331 — 
Gag-LF9 Gag 67–75 LSGQEREAF 0.122 — 
Gag-ML9 Gag 226–234 MTARFIRGL 0.496 — 
Gag-DY8 Gag 333–340 DTLEEKMY 0.139 — 
Env-FM9 Env 136–144 FSYETNRSM 0.566 — 
Env-GF9 Env 124–132 GSFPGCRPF 0.500 — 
Env-TY9 Env 85–93 TTGGILWWY 0.377 — 
Env-NF9 Env 595–603 NTPDSIAQF 0.373 — 
Env-IY9 Env 461–469 IAASATMSY 0.505 — 
Rev-EW9 Rev 89–97 EVLQERLEW 0.049 — 
b

Peptides that can help the Eqca-N*00602 H chain renature with horse β2m are marked as +, otherwise −.

c

The virus stain: EIAV-WSU5 (GenBank no. AF247394).

d

The residues within positions of the primary anchors of Eqca-N*00602 are denoted with underlined bold letters.

e

Scoring for peptides with residues >10 is not available.

FIGURE 1.

Thermostability of equine MHC I Eqca-N*00602/Eqca-N*00601 complexes. CD spectropolarimetry was used to access the thermostability of purified equine MHC I Eqca-N*00602 (A) and Eqca-N*00601 (B) complexed with previously identified EIAV-derived CTL epitopes Env-RW12, Gag-GW12, and Rev-QW11 (12). Denaturation in complex structure was monitored at 218 nm as the temperature was ramped up from 25 to 75°C, and the fraction of unfolded material is expressed as a function of the temperature.

FIGURE 1.

Thermostability of equine MHC I Eqca-N*00602/Eqca-N*00601 complexes. CD spectropolarimetry was used to access the thermostability of purified equine MHC I Eqca-N*00602 (A) and Eqca-N*00601 (B) complexed with previously identified EIAV-derived CTL epitopes Env-RW12, Gag-GW12, and Rev-QW11 (12). Denaturation in complex structure was monitored at 218 nm as the temperature was ramped up from 25 to 75°C, and the fraction of unfolded material is expressed as a function of the temperature.

Close modal
Table III.
Tms of Eqca-N*00602 and peptide complexes
PeptidesEqca-N*00602Eqca-N*00601
Env-RW12 59.25 62.41 
Gag-GW12 53.85 51.25 
Rev-QW11 47.25 59.95 
PeptidesEqca-N*00602Eqca-N*00601
Env-RW12 59.25 62.41 
Gag-GW12 53.85 51.25 
Rev-QW11 47.25 59.95 

Tms (°C) were determined by CD.

The crystal structures of the three EIAV-derived long epitopes Env-RW12 (12-mer), Gag-GW12 (12-mer), and Rev-QW11 (11-mer) complexed to Eqca-N*00602 were determined to the resolutions of 2.3, 2.5, and 2.6 Å, respectively (Table II). To reveal the characteristics of the pELA/peptide–MHC I to present peptides with common lengths, we also screened an Eqca-N*00602–binding 9-mer peptide Gag-CF9 (CTSEEMNAF) derived from the EIAV Gag protein (the Eqca-N*00602–binding peptide screening is described in detail below) and determined the structure of Eqca-N*00602/Gag-CF9 at a resolution of 2.2 Å. The pEqca-N*00602 complexed with peptide Env-RW12 was crystallized in the P212121 orthorhombic space group with two molecules in one asymmetric unit, whereas the other three peptides complexed to pEqca-N*00602, that is, Gag-GW12, Rev-QW11, and Gag-CF9, were crystallized in the P41212 orthorhombic space group with only one molecule per asymmetric unit. The pEqca-N*00602 complexes exhibited the known typical architecture of classical MHC I molecules in vertebrates, with binding groove–forming anti-parallel α1 and α2 helices atop a β-sheet bed. The two “hooks” contributed from both α1 and α2 domains that are made of the two loops connecting to β2m (44). The α3 domain and β2m occupied the standard positions below the bed (Fig. 2A–C). The overall structures of pEqca-N*00602 are similar to the previously determined structure of cattle MHC I molecule N*01801 (Fig. 2D) (31). The superposition of Eqca-N*00602/Env-RW12 onto N*01801/IPA generated a root mean square deviation of 0.831 Å, which indicated that the overall structures of equine MHC I molecules conserved the common overall conformation of MHC I molecules from other mammals.

FIGURE 2.

Structural overview of equine MHC I Eqca-N*00602 complexes. (A) Overview of the structure of the Eqca-N*00602/Env-RW12. (B) The peptide Env-RW12 (green sticks and spheres) was presented in the peptide-binding groove. Two molecules in one asymmetric unit of Eqca-N*00602/Env-RW12 are shown. (C) The superposition of different Eqca-N*00602 complexes was formed by four different peptides: Eqca-N*00602/Env-RW12 (green), Eqca-N*00602/Gag-GW12 (cyan), Eqca-N*00602/Rev-QW11 (purple), and Eqca-N*00602/Gag-CF9 (yellow). The peptides were omitted. (D) Structural alignment of MHC H chains and β2m exhibited a similar overview conformation of Eqca-N*00602 (green) and cattle MHC I N*01801 (orange, Protein Data Bank code 3PWU).

FIGURE 2.

Structural overview of equine MHC I Eqca-N*00602 complexes. (A) Overview of the structure of the Eqca-N*00602/Env-RW12. (B) The peptide Env-RW12 (green sticks and spheres) was presented in the peptide-binding groove. Two molecules in one asymmetric unit of Eqca-N*00602/Env-RW12 are shown. (C) The superposition of different Eqca-N*00602 complexes was formed by four different peptides: Eqca-N*00602/Env-RW12 (green), Eqca-N*00602/Gag-GW12 (cyan), Eqca-N*00602/Rev-QW11 (purple), and Eqca-N*00602/Gag-CF9 (yellow). The peptides were omitted. (D) Structural alignment of MHC H chains and β2m exhibited a similar overview conformation of Eqca-N*00602 (green) and cattle MHC I N*01801 (orange, Protein Data Bank code 3PWU).

Close modal

The conformations of the peptides presented by MHC molecules have been confirmed to play a pivotal role in peptide immunogenicity. In the structures of Eqca-N*00602 complexed to the four peptides derived from EIAV determined in the present study, the electron densities of these four peptides clearly exhibited the presenting conformations (Fig. 3A–D). All four peptides were arched with tight contacts at their N and C termini into the peptide-binding groove of pEqca-N*00602. The overlapping P2 and Pc residues of these four peptides clearly exhibited the traditional anchors used by the Eqca-N*00602 peptides presented in the present study (Fig. 3E). Additionally, the residues in the P3 position of the peptides also had a conserved conformation, which indicated a potentially important role for the P3 anchoring for the Eqca-N*00602–restricted peptides. This unconventional P3 anchor revealed by our structures is in accordance with the previous peptide screening data for horse MHC I, in which positions 2 and 3 were identified as possible primary anchor residues (19). These residues located in the central but close to the C terminus of the Eqca-N*00602–binding peptides were directed into the solvent and exhibited diverse presenting conformations (Fig. 3E). The space of the groove of pEqca-N*00602 was completely occupied by the superposed peptides, indicating the different exposed profiles of these Eqca-N*00602–presented peptides.

FIGURE 3.

Diverse presentation of peptides by equine MHC I Eqca-N*00602. (AD) Peptides Env-RW12, Gag-GW12, Rev-QW11, and Gag-CF9 binding to Eqca-N*00602 were presented with electron density at the 1.5-σ contour level; the peptide Env-RW12 from molecule 1 in the asymmetric units of the structure Eqca-N*00602/Env-RW12 is shown as a representative. (E) Diverse presentation of peptides Env-RW12 (green), Gag-GW12 (cyan), Rev-QW11 (purple), and Gag-CF9 (yellow) by Eqca-N*00602. The peptides were aligned according to the superposition of the α1 and α2 domains of the four structures of Eqca-N*00602. The views of the peptide alignments showed that most conformational distinctions were located in the central region of the peptides. The major space of the peptide-binding groove (in red rectangle) of Eqca-N*00602 was occupied by a combination of the different peptides. The conserved conformation of residues in P2, P3, and Pc positions are shown as white sticks in purple ellipses. (F) The superposition of 12-mer peptides Env-RW12 and Gag-GW12 presented by Eqca-N*00602 exhibited different conformation and flexibilities. The coloration of the two peptides was performed according to the isotropic B factors.

FIGURE 3.

Diverse presentation of peptides by equine MHC I Eqca-N*00602. (AD) Peptides Env-RW12, Gag-GW12, Rev-QW11, and Gag-CF9 binding to Eqca-N*00602 were presented with electron density at the 1.5-σ contour level; the peptide Env-RW12 from molecule 1 in the asymmetric units of the structure Eqca-N*00602/Env-RW12 is shown as a representative. (E) Diverse presentation of peptides Env-RW12 (green), Gag-GW12 (cyan), Rev-QW11 (purple), and Gag-CF9 (yellow) by Eqca-N*00602. The peptides were aligned according to the superposition of the α1 and α2 domains of the four structures of Eqca-N*00602. The views of the peptide alignments showed that most conformational distinctions were located in the central region of the peptides. The major space of the peptide-binding groove (in red rectangle) of Eqca-N*00602 was occupied by a combination of the different peptides. The conserved conformation of residues in P2, P3, and Pc positions are shown as white sticks in purple ellipses. (F) The superposition of 12-mer peptides Env-RW12 and Gag-GW12 presented by Eqca-N*00602 exhibited different conformation and flexibilities. The coloration of the two peptides was performed according to the isotropic B factors.

Close modal

When compared with the two 12-mer peptides Env-RW12 and Gag-GW12 presented by Eqca-N*00602, we found that although they are both 12-mers, the main chain conformations of the two peptides were completely different from each other (Fig. 3F). Notably, it was revealed by a higher B factor that the entire length, and especially the residues in the middle region of Gag-GW12, was more flexible than Env-RW12, which may indicate a more diverse TCR docking mode for Gag-GW12 compared with Env-RW12.

Eqca-N*00602 has only one variant residue in position 152 compared with the sequence of another equine MHC I allele, Eqca-N*00601 (Supplemental Fig. 1). However, horses carrying these two alleles have the divergent T cell responses to the similar EIAV-derived peptides. In this study, we elucidated the structural basis of this phenomenon by determining the related structures of the peptide Gag-GW12 complexed to Eqca-N*00601 and Eqca-N*00602.

Through a one-site mutation strategy, we generated the soluble protein of Eqca-N*00601, which has a Val in position 152 corresponding to the Glu152 of Eqca-N*00602 and successfully determined the crystal structure of Eqca-N*00601 with peptide Gag-GW12. The overall structure of Eqca-N*00601 was quite similar to Eqca-N*00602 with a root mean square deviation of 0.171 Å (Supplemental Fig. 2A). The only different portion of the two H chains was located in the electron densities carried by Val152 of Eqca-N*00601 and Glu152 of Eqca-N*00602 (Supplemental Fig. 2B, 2C). When we superposed the Gag-GW12 peptides presented by the two related horse MHC I molecules, two different conformations of peptide Gag-GW12 were observed when presented by Eqca-N*00601 and Eqca-N*00602 (Fig. 4A, 4B). Whereas the main chain of the N terminus (P1–P4) and C terminus (Pc) retained a similar conformation and binding to the peptide-binding grooves, major differences were observed in the middle parts of the peptides (Fig. 4C, Supplemental Fig. 3). In Eqca-N*00602–presented Gag-GW12, an incomplete α helix was observed in the middle portion of the peptide (Thr7 to Asn9). Side chains of residues Thr6, Thr7, Gly8, (no side chain), and Asp11 protruded outside of the peptide-binding groove (Fig. 4D). In contrast, Gag-GW12 in Eqca-N*00601 exhibited a C-terminal budged conformation with residues Leu5, Thr7, Gly8, Asn9, and Cys10 exposed out of the peptide groove (Fig. 4E). The tremendously different exposed landscape of the peptide Gag-GW12 presented by Eqca-N*00601 and Eqca-N*00602 may indicate the different TCR recognitions of this peptide by the horses with the two MHC I alleles.

FIGURE 4.

Divergent presentation of the peptide Gag-GW12 by the closely related equine MHC I alleles Eqca-N*00601 and Eqca-N*00602. (A and B) Distinct binding of the peptide Gag-GW12 to Eqca-N*00601 and Eqca-N*00602. The main chain of Gag-GW12 was colored salmon in the structure of Eqca-N*00601/Gag-GW12 and cyan in the structure of Eqca-N*00602/Gag-GW12. To conveniently observe the shift of similar residues in the two different conformations of Gag-GW12, the side chains of residues from Gag-GW12 are presented in different colors (Lys4 in red, Leu5 in orange, Thr6 in yellow, Thr7 in green, Asn9 in blue, Cys10 in purple, and Asn11 in brown). The two structures were aligned according to the superposition of the α1 and α2 domains of the MHC molecules. Both the side (A) and top (B) views of the peptide alignments demonstrated that most conformational distinctions were located in the central region of the peptides. (C) General side chain orientation for the peptide Gag-GW12 in the structures Eqca-N*00601/Gag-GW12 and Eqca-N*00602/Gag-GW12 as viewed in profile from the peptide N terminus toward the C terminus (up is toward the TCR, down is toward the floor of the peptide binding groove, left is toward the α1 helix domain, and right is toward the α2 helix domain). The pockets within the peptide-binding groove that accommodate the anchors of the peptides are listed under the corresponding anchors. (D and E) The different exposed areas of the peptide Gag-GW12 presented by Eqca-N*00602 (D) and Eqca-N*00601 (E) are shown in vacuum electrostatic surface potential. The exposed area of IPA-M1 in molecule 1 was largely covered by residues Lys4, Thr6, Thr7, Gly8, and Asp11. Residue Glu152 (cyan) of the Eqca-N*00602 H chain protrudes under the middle of the peptide and has a tight contact with the peptide. (D) For Eqca-N*00601/Gag-GW12, Gag-GW12 exposes Lys4, Leu5, Thr7, Gly8, Asp9, and Cys10 for potential TCR docking. The residue Val152 Eqca-N*00601 is shown as sticks and spheres with a salmon color.

FIGURE 4.

Divergent presentation of the peptide Gag-GW12 by the closely related equine MHC I alleles Eqca-N*00601 and Eqca-N*00602. (A and B) Distinct binding of the peptide Gag-GW12 to Eqca-N*00601 and Eqca-N*00602. The main chain of Gag-GW12 was colored salmon in the structure of Eqca-N*00601/Gag-GW12 and cyan in the structure of Eqca-N*00602/Gag-GW12. To conveniently observe the shift of similar residues in the two different conformations of Gag-GW12, the side chains of residues from Gag-GW12 are presented in different colors (Lys4 in red, Leu5 in orange, Thr6 in yellow, Thr7 in green, Asn9 in blue, Cys10 in purple, and Asn11 in brown). The two structures were aligned according to the superposition of the α1 and α2 domains of the MHC molecules. Both the side (A) and top (B) views of the peptide alignments demonstrated that most conformational distinctions were located in the central region of the peptides. (C) General side chain orientation for the peptide Gag-GW12 in the structures Eqca-N*00601/Gag-GW12 and Eqca-N*00602/Gag-GW12 as viewed in profile from the peptide N terminus toward the C terminus (up is toward the TCR, down is toward the floor of the peptide binding groove, left is toward the α1 helix domain, and right is toward the α2 helix domain). The pockets within the peptide-binding groove that accommodate the anchors of the peptides are listed under the corresponding anchors. (D and E) The different exposed areas of the peptide Gag-GW12 presented by Eqca-N*00602 (D) and Eqca-N*00601 (E) are shown in vacuum electrostatic surface potential. The exposed area of IPA-M1 in molecule 1 was largely covered by residues Lys4, Thr6, Thr7, Gly8, and Asp11. Residue Glu152 (cyan) of the Eqca-N*00602 H chain protrudes under the middle of the peptide and has a tight contact with the peptide. (D) For Eqca-N*00601/Gag-GW12, Gag-GW12 exposes Lys4, Leu5, Thr7, Gly8, Asp9, and Cys10 for potential TCR docking. The residue Val152 Eqca-N*00601 is shown as sticks and spheres with a salmon color.

Close modal

Because there is only a one-residue difference between Eqca-N*00601-152V and Eqca-N*00602-152G, the divergent presentation of Gag-GW12 by these two MHC I alleles would be substantially influenced by the E152V mutation. In the pEqca-N*00602/Gag-GW12 structure, the side chain of Glu152 forms two hydrogen bonds with residue Asn11 of peptide Gag-GW12, which brings the main chain of the peptide close to the these two MHC I alleles. In contrast, no direct contact was observed between the side chain of Val152 in the MHC I allele Eqca-N*00601 with peptide Gag-GW12. This lack of direct contact loosened the middle region of Gag-GW12, which formed a bulged conformation out of the groove of pEqca-N*00601 (Fig. 5). Residue Gln at position 155 of Eqca-N*00601 formed hydrogen bonds with Thr6 and Thr7 in pEqca-N*00601/Gag-GW12, whereas it formed different hydrogen bonds with Thr6 in pEqca-N*00602/Gag-GW12. Therefore, Gln155 also contributed to the different binding of Gag-GW12 to Eqca-N*00601 and Eqca-N*00602 in the coordination with residues in position 152. The conformation of Gag-GW12 in the pEqca-N*00601/Gag-GW12 complex was more similar to the conformations of Env-RW12 and Rev-QW11 presented by Eqca-N*00602 (Fig. 6), in which no hydrogen bonds were found between Glu152 and the peptides. Compared with the hydrophilic residue Asp in the position Pc-1 of Gag-GW12, the corresponding residues in the Pc-1 positions of Env-RW12 and Rev-QW11 are Tyr and Glu, respectively, which do not form hydrogen bonds with the Glu152 of Eqca-N*00602 in the structures. Therefore, the tight interaction between residue 152 and the peptides destructed the C terminus–bulged conformation of pEqca-N*00601/pEqca-N*00602–presented peptides (as in pEqca-N*00601/Gag-GW12, pEqca-N*00602/Env-RW12, and Eqca-N*00602/Rev-QW11).

FIGURE 5.

E152V contributes to the conformational divergence of Gag-GW12 presented by Eqca-N*00601 and Eqca-N*00602. (A) Eqca-N*00602–presented Gag-GW12 (cyan) forms tight hydrogen bonds with Glu152 through the residue Asn at the P11 position of the peptide. Direct contact between Glu155 on the H chain and Thr6 on the peptide was also observed. The main chain conformation of the peptide Gag-GW12 was restricted by the two pairs of interactions (in green rectangles) and thus tends to form a partial α helix. (B) Residues Thr6 and Thr7 of Eqca-N*00601–presented Gag-GW12 (salmon) form hydrogen bonds with Glu155 of Eqca-N*00601. No direct contact between Val152 on the H chain and peptide was observed.

FIGURE 5.

E152V contributes to the conformational divergence of Gag-GW12 presented by Eqca-N*00601 and Eqca-N*00602. (A) Eqca-N*00602–presented Gag-GW12 (cyan) forms tight hydrogen bonds with Glu152 through the residue Asn at the P11 position of the peptide. Direct contact between Glu155 on the H chain and Thr6 on the peptide was also observed. The main chain conformation of the peptide Gag-GW12 was restricted by the two pairs of interactions (in green rectangles) and thus tends to form a partial α helix. (B) Residues Thr6 and Thr7 of Eqca-N*00601–presented Gag-GW12 (salmon) form hydrogen bonds with Glu155 of Eqca-N*00601. No direct contact between Val152 on the H chain and peptide was observed.

Close modal
FIGURE 6.

Eqca-N*00601/Eqca-N*00602 present long peptides in unconventional conformations compared with MHCs from other species. Env-RW12 [(A) green] presented by Eqca-N*00602, Gag-GW12 presented by Eqca-N*00602 [(B) cyan] and Eqca-N*00601 (B, salmon), and Rev-QW11 [(C) purple] presented by Eqca-N*00602 are represented by thick sticks. All other peptides, denoted as thin sticks, are 11- to 13-mer peptides presented by other mammalian MHC alleles. The bulged conformations of peptides presented by Eqca-N*00601/Eqca-N*00602 were closer to the C terminus of the peptides, denoted with red arrows. The structures used are HLA-B*3508 (Protein Data Bank codes 2FZ3, 3BW9, and 1ZHL), HLA-B*5703 (2BVQ), cattle N*01301 (2XFX), H-2Db (1JPF), and chicken B21 (3BEV).

FIGURE 6.

Eqca-N*00601/Eqca-N*00602 present long peptides in unconventional conformations compared with MHCs from other species. Env-RW12 [(A) green] presented by Eqca-N*00602, Gag-GW12 presented by Eqca-N*00602 [(B) cyan] and Eqca-N*00601 (B, salmon), and Rev-QW11 [(C) purple] presented by Eqca-N*00602 are represented by thick sticks. All other peptides, denoted as thin sticks, are 11- to 13-mer peptides presented by other mammalian MHC alleles. The bulged conformations of peptides presented by Eqca-N*00601/Eqca-N*00602 were closer to the C terminus of the peptides, denoted with red arrows. The structures used are HLA-B*3508 (Protein Data Bank codes 2FZ3, 3BW9, and 1ZHL), HLA-B*5703 (2BVQ), cattle N*01301 (2XFX), H-2Db (1JPF), and chicken B21 (3BEV).

Close modal

The superposition of these peptides presented by Eqca-N*00601/Eqca-N*00602 with the available crystal structures of longer peptides (residues ≥ available crystal in the MHC I molecules from other species) revealed a comparison of the key features of horse MHC I. Alignment of the features of horse MHC I demonstrated that the main chains of the N-terminal portion (P1–P3) of the peptides were similar. The peptide conformations begin to differ at the P4 residues. Furthermore, from the P5 to Pc-2 residues, the peptides were accommodated in the groove by adopting tremendous divergent conformations, either zigzagging within the groove or bulging out of the groove. The main chain conformations of these long peptides converged again at the Pc-1 position and finally overlapped with each other at the Pc positions. The relatively conserved conformations of the N and C termini of longer peptides presented by MHC I of horses and other species revealed the importance of N- and C-terminal anchoring. Additionally, the divergent conformations in the middle portions of the longer peptides presented by the MHC I molecules reflected the variant TCR binding modes of these peptides. None of the longer peptides presented by MHC I molecules from other vertebrates had a conformation similar to horse Eqca-N*00601/Eqca-N*00602 in the medium regions. The three peptides, that is, Env-RW12 and Rev-QW11 presented by Eqca-N*00602 and Gag-GW12 presented by Eqca-N*00601, had a distinctive C-terminal–bulged conformation at the main chains (P5–Pc-2). The bulged portions of these peptides were located closer to the C-terminal end of the binding groove than the bulged-type conformations of other MHC I–presented peptides. The bulged portions of the peptides (except Gag-GW12 presented by Eqca-N*00602) presented by these two equine MHC I molecules, Eqca-N*00601/Eqca-N*00602, were even closer to the C terminus of the peptides than a peptide presented by cattle MHC I molecule N*01301, which has been demonstrated to have a C-terminal–bulged conformation (45).

The distances between the C-terminal peptides are greater than those between peptides presented by cattle MHC I molecule N*01301 (15). The consistent height of P1–P3 and the Pc residues revealed the structural conservation of the N- and C-terminal anchoring of MHC I in different species. However, the middle portions of the peptides adopted diverse distances to the groove floor formed by the β-sheet. The distances of the Pc-1 and Pc-2 residues to the groove bottom of Env-RW12 and Rev-QW11 presented by Eqca-N*00602 and Gag-GW12 presented by Eqca-N*00601 are the longest distances within the MHC I structures available. This result quantitatively indicated that these peptides are one of the C-terminal–bulged peptides in the MHC I structures reported to date.

The unconventional conformations of Eqca-N*00601/Eqca-N*00602–restricted peptides may be associated with the uncommon groove of Eqca-N*00602, including the positively charged C pocket (Fig. 7A, described in detail below) under the bulged portion of the peptides. The peptides would be extruded out of the groove to avoid the repulsion of the positively charged C pocket, unless it is pulled by the residues (e.g., Glu152 in pEqca-N*00602) on the α1 or α2 helices of the MHC I molecules.

FIGURE 7.

Uncommon peptide binding groove of Eqca-N*00602 revealed by the complex with 9-mer peptide Gag-CF9. (A) The peptide Gag-CF9 (yellow) was accommodated between the two α helices (gray straps) in the groove of Eqca-N*00602. The bottom of the groove represented with vacuum electrostatic surface potential showed the pockets B, C, and F in the groove. The deep blue color of the C pocket indicated a positive charge. (B) The C pocket of Eqca-N*00602. The Glu5 and Asn7 of Gag-CF9 insert into the pocket C of Eqca-N*00602, whereas the middle portion of peptide Env-RW12 bulges out of the groove without any contact with the positively charged C pocket. (C) The Eqca-N*00602 B pocket is shown with the P2 Thr of Gag-CF9 (yellow stick and sphere) and P2 Val of Env-RW12 (green stick and sphere). The major residues composing the B pocket are shown as sticks (yellow for Eqca-N*00602/Gag-CF9 and green for Eqca-N*00602/Env-RW12). The main chains of Env-RW12 and Gag-CF9 are denoted as thin ribbons in green and yellow, respectively. (D) Eqca-N*00602 F pocket with the Pc Phe of Gag-CF9 and P2 Trp of Env-RW12 in the pocket. An addition hydrogen bond was observed between Trp12 of Env-RW12 and Asn73 of Eqca-N*00602. The hydrogen bonds are represented as purple dashed lines.

FIGURE 7.

Uncommon peptide binding groove of Eqca-N*00602 revealed by the complex with 9-mer peptide Gag-CF9. (A) The peptide Gag-CF9 (yellow) was accommodated between the two α helices (gray straps) in the groove of Eqca-N*00602. The bottom of the groove represented with vacuum electrostatic surface potential showed the pockets B, C, and F in the groove. The deep blue color of the C pocket indicated a positive charge. (B) The C pocket of Eqca-N*00602. The Glu5 and Asn7 of Gag-CF9 insert into the pocket C of Eqca-N*00602, whereas the middle portion of peptide Env-RW12 bulges out of the groove without any contact with the positively charged C pocket. (C) The Eqca-N*00602 B pocket is shown with the P2 Thr of Gag-CF9 (yellow stick and sphere) and P2 Val of Env-RW12 (green stick and sphere). The major residues composing the B pocket are shown as sticks (yellow for Eqca-N*00602/Gag-CF9 and green for Eqca-N*00602/Env-RW12). The main chains of Env-RW12 and Gag-CF9 are denoted as thin ribbons in green and yellow, respectively. (D) Eqca-N*00602 F pocket with the Pc Phe of Gag-CF9 and P2 Trp of Env-RW12 in the pocket. An addition hydrogen bond was observed between Trp12 of Env-RW12 and Asn73 of Eqca-N*00602. The hydrogen bonds are represented as purple dashed lines.

Close modal

Thus far, most of the identified Eqca-N*00602–restricted CTL epitopes are long peptides with >11 residues. In this study, we screened the potential Eqca-N*00602–binding peptides with a common length (8–10 residues) among the EIAV proteome utilizing an online software-based prediction (29, 30). One nonamer peptide, Gag-CF9, was shown to successfully assist in the renaturing of pEqca-N*00602 in vitro (Supplemental Fig. 4), and, subsequently, the structure of Gag-CF9 complexed to Eqca-N*00602 was determined. The binding of peptide Gag-CF9 to Eqca-N*00602 was clearly distinguished from the longer peptides Env-RW12, Gag-GW12, and Rev-QW11. In the pEqca-N*00602 groove, the vacuum electrostatic surface potential showed a positively charged C pocket (Fig. 7A), to which the residues Arg97 and Arg114 of Eqca-N*00602 contributed substantially. For the longer peptides Env-RW12, Gag-GW12, and Rev-QW11, bulged conformations exposed the medium region of the peptides out of the peptide-binding groove, whereas no direct contact occurred between the middle portion of the peptides and the groove bottom of Eqca-N*00602. Although Gag-GW12 formed a tight interaction with Glu152 in the α2 helix of Eqca-N*00602, no side chains of Gag-GW12 were inserted into or interacted with the groove bottom of Eqca-N*00602. In contrast, the peptide Gag-CF9 had a tight interaction with the C pocket of Eqca-N*00602. Two salt bridges were formed between the negatively charged residue Glu5 of Gag-CF9 and Arg114 of Eqca-N*00602 and also between the hydrophilic Asn7 of Gag-CF9 and Arg97 of Eqca-N*00602 (Fig. 7B).

The pivotal roles of Glu5 and Asn7 as the secondary anchor residues of short peptide Gag-CF9 may reveal the strict selection of Eqca-N*00602 for the shorter peptides (8–10 residues). Eqca-N*00602 prefers the short peptides with a negative charge or hydrophilic residues in the middle portion (e.g., P5 and P7 residues) of the peptides, which could act as secondary anchoring residues to form tight hydrogen bonds to Arg97 and Arg114 in the C pocket of Eqca-N*00602. Peptides with glycine in the middle portion (such as the P5 and P7 residues) may also survive in the Eqca-N*00602 binding. Longer peptides, such as Env-RW12, Gag-GW12, and Rev-QW11, do not need to follow the rule because of the bulged middle portion of the peptides without any contact with the groove bottom. The low amount of the Eqca-N*00602–binding peptides with common length (8–10 residues) identified to date may also reflect the strict selection of Eqca-N*00602 for these relatively short peptides. To assess the rules for the short peptide binding of Eqca-N*00602, we analyzed the Eqca-N*00602–binding capabilities of a series of peptides with high binding scores in the software-based prediction. We found that the peptides following this motif with a negative charge or a glycine in P5 and a hydrophilic residue in P7 can assist in the renaturing of pEqca-N*00602 in vitro (Table III).

The primary anchors of MHC I molecule–restricted peptides play a pivotal role in the binding between peptides and MHC I. In the present study, the structures of four peptides complexed to horse MHC I provide the possibility of investigating the binding features of the B and F pockets of Eqca-N*00601/Eqca-N*00602 to accommodate the different primary anchors at P2 and Pc of the peptides at the atomic level. We compared the structures of the 12-mer Env-RW12 and 9-mer Gag-CF9 peptide complexed to Eqca-N*00602 as representative structures. Val2 of Env-RW12 and Thr2 of the Gag-CF9 peptide were inserted into the B pocket of Eqca-N*00602, which was formed by Tyr7, Tyr9, Glu63, Asn66, and Met67. The major difference between the two residues (Val2 of Env-RW12 and Thr2 of Gag-CF9) as the P2 anchor is that the Thr2 of Gag-CF9 formed two additional hydrogen bonds with Glu63 and Asn66 (Fig. 7C). Similar hydrogen bonds were observed between the Ser2 of peptide Gag-GW12 and the corresponding residues of the B pocket of Eqca-N*00602 (data not shown). The additional hydrogen bonds between the P2 residues of the Gag-GW12 (Ser2) and Gag-CF9 (Thr2) peptides and the B pocket revealed a preference of Eqca-N*00602 for the hydrophilic residues Ser and Thr with medium size for the P2 anchoring. Compared with the small space of the B pocket, the large F pocket of Eqca-N*00602 accommodated Trp (Env-RW12, Gag-GW12, and Rev-QW11) or Phe (Gag-CF9) with aromatic rings. However, when we compared the binding of Env-RW12’s Trp12 and Gag-CF9’s Phe9 with the F pocket of Eqca-N*00602, an additional hydrogen bond between the Trp12 of Env-RW12 and the Asn73 of Eqca-N*00602 was observed (Fig. 7D). Moreover, Pc-Trp is larger than Pc-Phe and has a more complementary shape for the F pocket, which endowed Pc-Trp with more van der Waals contact with pocket F than for Pc-Phe. Together with the additional hydrogen bond, Trp acts as the preferable Pc anchor for Eqca-N*00602 compared with Phe.

When the peptide-binding motif of horse MHC I Eqca-N*00602 was compared with other mammalian MHC I molecules, the hydrophilic residues Ser and Thr with medium size were found to be two common residues for the P2 anchors of mammalian MHC I molecules, such as human MHC HLA-A*1101, HLA*3301, HLA*6801, rhesus macaque MHC I Mamu-A*01, and Mamu-A*02. With regard to the Pc anchor Trp, although Trp was not a common residue in the protein sequences of pathogens, many mammalian MHC I molecule–restricted peptides use Trp as their Pc anchor, including HLA-A*01, HLA-B*35, HLA-B*57, and Mamu-B*17 (46). Interestingly, swine MHC I SLA-1*0401 had a similar preference as Eqca-N*00602 for both the P2 and Pc anchors (Fig. 8A). In the structure of SLA-1*0401 complexed to peptide S-OIVNW9 (47), the P2 anchor Ser formed two hydrogen bonds with Glu63 and Asn66, which is similar to Eqca-N*00602/Gag-GW12 (Fig. 8B). In the F pocket, the SLA-1*0401–restricted peptide S-OIVNW9 also used Trp as the Pc anchor. However, the only difference between the two tryptophans of Gag-GW12 (Eqca-N*00602 restricted) and S-OIVNW9 (SLA-1*0401 restricted) was the reverse conformation of the two tryptophans (Fig. 8C). The Trp12 of Gag-GW12 formed a hydrogen bond with Asn73 on the α1 helix of Eqca-N*00602, whereas the Trp9 of S-OIVNW9 formed a hydrogen bond with Asp116 on the α1 helix of the SLA-1*0401 groove. The preferences of Eqca-N*00602 for the peptide with primary anchors Ser/Thr for P2 and Trp for Pc, as in many other mammals, reveals an advantage of these residues as MHC I–binding peptide anchors after a long evolutionary selection in the mammalian immune system.

FIGURE 8.

Comparison of primary anchors of peptides presented by horse MHC I Eqca-N*00602 and swine MHC I SLA-1*0401. (A) Superposition of Gag-GW12 (cyan) presented by Eqca-N*00602 and peptide S-OIVNW9 (violet) presented by SLA-1*0401 (Protein Data Bank code 3QQ3). The two peptides have similar primary anchors Ser in P2 and Trp in Pc positions. (B) Similar binding of Ser in P2 positions of peptide Gag-GW12 and S-OIVNW9 with the residues in the F pockets of Eqca-N*00602 and swine MHC I SLA-1*0401. (C) Reverse conformations of Trp in the Pc positions of the peptide Gag-GW12 and S-OIVNW9. Distinct interaction of Trp with the residues of the F pockets of Eqca-N*00602 and swine MHC I SLA-1*0401 are observed. The hydrogen bonds are represented as cyan (for Eqca-N*00602/Gag-GW12) or violet (for SLA-1*0401/S-OIVNW9) dashed lines.

FIGURE 8.

Comparison of primary anchors of peptides presented by horse MHC I Eqca-N*00602 and swine MHC I SLA-1*0401. (A) Superposition of Gag-GW12 (cyan) presented by Eqca-N*00602 and peptide S-OIVNW9 (violet) presented by SLA-1*0401 (Protein Data Bank code 3QQ3). The two peptides have similar primary anchors Ser in P2 and Trp in Pc positions. (B) Similar binding of Ser in P2 positions of peptide Gag-GW12 and S-OIVNW9 with the residues in the F pockets of Eqca-N*00602 and swine MHC I SLA-1*0401. (C) Reverse conformations of Trp in the Pc positions of the peptide Gag-GW12 and S-OIVNW9. Distinct interaction of Trp with the residues of the F pockets of Eqca-N*00602 and swine MHC I SLA-1*0401 are observed. The hydrogen bonds are represented as cyan (for Eqca-N*00602/Gag-GW12) or violet (for SLA-1*0401/S-OIVNW9) dashed lines.

Close modal

EIAV is an important member of the lentivirus family and has been successfully controlled by an attenuated vaccine in horses. This situation might provide a reference model for the development of lentivirus vaccines (3). Therefore, a horse infected with EIAV may act as a decent animal model for other lentivirus studies (8). The successful prevention and control of EIA in China in the 1970s and 1980s, which is credited to the large-scale application of a donkey leukocyte–attenuated EIAV vaccine, may encourage the development of vaccines for human lentiviruses (28). Recently, the systematic genomic analysis of the Chinese EIAV vaccine strain and its parental virulent strains revealed that, although the vaccine strain is comprised of a complex combination of different mutations, the public and private specificities of CTL epitopes may be found in the EIAV strains (48, 49). Thus, the major strategy for EIAV vaccine development was still the attenuated virus because it induces strong CTL and Ab immunities.

In EIAV-infected horses, considerable evidence has demonstrated the dominant role of the CTL epitope Env-RW12 in the EIAV-specific CTL immunities among Eqca-N*00602 horses. However, it is strongly suggested that the immune escape mutants of Env-RW12 could occur in the early stage of EIAV infection, whereas no mutants derived from another Eqca-N*00602 epitope Gag-GW12 have been identified to date (11, 12). It has also been reported that there is a close association between the Gag-GW12–induced CTL immunity and efficient virus control (14, 16). A broader TCR repertoire can be provoked by epitope peptides with flexible, exposed residues, limiting the appearance of escape variants and maintaining host resistance to the disease (5052). The structural differences of the presentation of the two peptides may lead to antigenic variance. Of the structures determined in the present study, Gag-GW12 was revealed to be more flexible than Env-RW12, especially in the central region. This conformational flexibility might lead to a diverse TCR repertoire for Gag-GW12, which restricts the appearance of Gag-GW12–derived mutants. Although the Rev-QW11–specific TCR repertoire is the only peptide of the four used for this study’s structure determination to be quantified thus far (53), we can rationally predict that the TCR repertoire specific for Gag-GW12 is relatively diverse based on our structural analysis.

Furthermore, considering that the immunodominant CTL epitope of Env-RW12 and the relatively subdominant epitope Gag-GW12 correlated with viral control, we cannot distinguish the relative significance of the two peptides. Both peptides acted in their own way in the anti-EIAV CTL responses in different infection stages (14). Considering the divergent presentation of the two peptides by Eqca-N*00602, the different antigenicities of the two peptides may be associated with the inherent characteristics of the horse MHC I molecule. Therefore, the intrinsic features of horse MHC I molecules, especially the distinct CTL epitopes presentation, have been contributed to EIAV control during horse infection and the development of EIAV molecular vaccine.

MHC I genes are the most polymorphic genes among jawed vertebrates. With the development of bioinformatics and genetics, MHC I polymorphisms have been identified to correlate with the susceptibility of different diseases, including the infection of pathogens. It has also been reported that the progression of the chronic virus infection is associated with different MHC I alleles. For example, the HLA-B*5701–carried population has been observed to be long-term nonprogressors in human HIV infection (54, 55), whereas Mamu-B*08 and Mamu-B*17 are two rhesus macaque MHC I alleles associated with a slow progress after the SIV infection (56, 57). Additionally, the structures of the murine MHC I molecules Kbm1 and Kbm8 revealed that subtle changes in the peptide environment impact thermostability and alloreactivity (58). In the horse MHC I molecules, the structures determined in the present study demonstrated that a single amino acid polymorphism could lead to a drastically different interaction of the similar peptide to the two closely related ELA-A1 alleles, that is, Eqca-N*00601 and Eqca-N*00602 molecules (Supplemental Fig. 2). The divergent presentation of the epitope peptide Gag-GW12 by the two molecules might indicate a structural basis for the non–cross-recognition of Gag-GW12 between horses carrying the two alleles. It has been demonstrated that the substitution of the residue in position 152 does not affect the cross-recognition of Env-RW12 and even results in a more efficient recognition of the Eqca-N*00602–presented Rev-QW11 by the Eqca-N*00601+ horse (20). Within the structures of pEqca-N*00602/Env-RW12 and pEqca-N*00602/Rev-QW11 determined in this study, there was no direct interaction between the polymorphic site and peptides, which differs from the tight contacts between Glu152 and Gag-GW12 in the pEqca-N*00602/Gag-GW12 structure. Therefore, we suggest that Env-RW12 and Rev-QW11 may perform in similar conformations when binding to Eqca-N*00601, which has a Val in position 152. This behavior may act as the structural basis for the cross-recognition of Env-RW12 and Rev-QW11 between the horse Eqca-N*00601 and Eqca-N*00602 alleles (20). The concordance between CTL-specific immunogenicities and the conclusions of our structural analyses complement our findings on the epitopes or peptide presentation of the horse MHC I Eqca-N*00601/Eqca-N*00602. Although, to our best of knowledge, there has been no report regarding the comparison of disease progression after EIAV infection between the two Eqcas, we suggest that the distinct peptide presentation of Eqca-N*00601 and Eqca-N*00602 may lead to variant anti-virus CTL immunities.

In the present study, we determined the structures of five horse MHC I molecules, that is, Eqca-N*00602 and Eqca-N*00601 complexed to four EIAV-derived epitopes and one 9-mer peptide with diverse lengths (9-mer to 12-mer). A series of comparative structural evidence elucidated the inherent characteristics of horse MHC I molecules. The horse Eqca-N*00602 molecule binds epitopes in diverse conformations, for example, the distinct presentations of longer epitopes Env-RW12 and Gag-GW12 with different antigenicities. One polymorphic residue between Eqca-N*00602 and Eqca-N*00601 was able to lead to a distinct conformation of GW12, indicating this diverse peptide presentation manner, as the peptide conformation can be changed so easily. Martin and colleagues found drastically different interactions between a self-peptide pVIPR with two HLA subtypes HLA-B*2705 and HLA-B*2709, which only differ with respect to residue 116 (Asp versus His) within the peptide-binding groove (21). The recently determined HLA-A*0301 complexed with an HIV-derived peptide RT313 also reflects the important role of residue 152 in the minor tuning of peptide side chains presented by HLA-A*0301 (Glu152) and HLA-A*1101 (Ala152) (59). Nevertheless, to our knowledge, the governing of the main chain conformation of peptides by the residue in position 152 was only observed in the structures of the Eqcas. Additionally, the structure of Eqca-N*00602 with peptide CF9 showed the structural basis of why the horse Eqca-N*00602 prefers to present longer epitope peptides but has a strict selective binding to a 9-mer. Longer CTL epitope peptides are more flexible and tend to be presented in conformations with diverse manners.

In summary, these unconventional characteristics of epitope–peptide presentation by preferred horse MHC I molecules act as a critical component in the elite CTL immunity among horses and may explain why the attenuated vaccine could successfully protect horses against EIAV infection. These data also provide a beneficial reference for vaccine design against other lentiviruses.

We acknowledge the assistance of the staff at the Shanghai Synchrotron Radiation Facility of China and the KEK Synchrotron Facility in Japan.

This work was supported by the State Key Program of the National Natural Science Foundation of China (Grant 31230074) and the 973 Project of the China Ministry of Science and Technology (Grant 2013CB835302).

The online version of this article contains supplemental material.

Abbreviations used in this article:
CD

circular dichroism

EIA

equine infectious anemia

EIAD

EIA disease

EIAV

equine infectious anemia virus

ELA

equine leukocyte Ag

Env

envelope glycoprotein

Eqca

equine MHC class I allele

Gag

Gag protein

β2m

β2-microglobulin

MHC I

MHC class I

P

protein

Rev

reverse transcriptase

Tm

midpoint transition temperature.

1
Steinbrook
R.
2007
.
One step forward, two steps back—will there ever be an AIDS vaccine?
N. Engl. J. Med.
357
:
2653
2655
.
2
Narayan
O.
1989-1990
.
Immunopathology of lentiviral infections in ungulate animals.
Curr. Opin. Immunol.
2
:
399
402
.
3
Leroux
C.
,
Cadoré
J. L.
,
Montelaro
R. C.
.
2004
.
Equine infectious anemia virus (EIAV): what has HIV’s country cousin got to tell us?
Vet. Res.
35
:
485
512
.
4
Cheevers
W. P.
,
McGuire
T. C.
.
1985
.
Equine infectious anemia virus: immunopathogenesis and persistence.
Rev. Infect. Dis.
7
:
83
88
.
5
McGuire
T. C.
,
O’Rourke
K. I.
,
Perryman
L. E.
.
1990
.
Immunopathogenesis of equine infectious anemia lentivirus disease.
Dev. Biol. Stand.
72
:
31
37
.
6
Sellon
D. C.
,
Fuller
F. J.
,
McGuire
T. C.
.
1994
.
The immunopathogenesis of equine infectious anemia virus.
Virus Res.
32
:
111
138
.
7
Henson
J. B.
,
McGuire
T. C.
.
1974
.
Equine infectious anemia.
Prog. Med. Virol.
18
:
143
159
.
8
Craigo
J. K.
,
Montelaro
R. C.
.
2011
.
Equine infectious anemia virus infection and immunity: lessons for AIDS vaccine development.
Future Virol.
6
:
139
142
.
9
Chung
C.
,
Mealey
R. H.
,
McGuire
T. C.
.
2004
.
CTL from EIAV carrier horses with diverse MHC class I alleles recognize epitope clusters in Gag matrix and capsid proteins.
Virology
327
:
144
154
.
10
McGuire
T. C.
,
Fraser
D. G.
,
Mealey
R. H.
.
2002
.
Cytotoxic T lymphocytes and neutralizing antibody in the control of equine infectious anemia virus.
Viral Immunol.
15
:
521
531
.
11
Mealey
R. H.
,
Leib
S. R.
,
Pownder
S. L.
,
McGuire
T. C.
.
2004
.
Adaptive immunity is the primary force driving selection of equine infectious anemia virus envelope SU variants during acute infection.
J. Virol.
78
:
9295
9305
.
12
Mealey
R. H.
,
Zhang
B.
,
Leib
S. R.
,
Littke
M. H.
,
McGuire
T. C.
.
2003
.
Epitope specificity is critical for high and moderate avidity cytotoxic T lymphocytes associated with control of viral load and clinical disease in horses with equine infectious anemia virus.
Virology
313
:
537
552
.
13
Ridgely
S. L.
,
Zhang
B.
,
McGuire
T. C.
.
2003
.
Response of ELA-A1 horses immunized with lipopeptide containing an equine infectious anemia virus ELA-A1-restricted CTL epitope to virus challenge.
Vaccine
21
:
491
506
.
14
Mealey
R. H.
,
Sharif
A.
,
Ellis
S. A.
,
Littke
M. H.
,
Leib
S. R.
,
McGuire
T. C.
.
2005
.
Early detection of dominant Env-specific and subdominant Gag-specific CD8+ lymphocytes in equine infectious anemia virus-infected horses using major histocompatibility complex class I/peptide tetrameric complexes.
Virology
339
:
110
126
.
15
Liu
J.
,
Zhang
S.
,
Tan
S.
,
Zheng
B.
,
Gao
G. F.
.
2011
.
Revival of the identification of cytotoxic T-lymphocyte epitopes for immunological diagnosis, therapy and vaccine development.
Exp. Biol. Med. (Maywood)
236
:
253
267
.
16
Chung
C.
,
Mealey
R. H.
,
McGuire
T. C.
.
2005
.
Evaluation of high functional avidity CTL to Gag epitope clusters in EIAV carrier horses.
Virology
342
:
228
239
.
17
Lonning
S. M.
,
Zhang
W.
,
McGuire
T. C.
.
1999
.
Gag protein epitopes recognized by CD4+ T-helper lymphocytes from equine infectious anemia virus-infected carrier horses.
J. Virol.
73
:
4257
4265
.
18
Tagmyer
T. L.
,
Craigo
J. K.
,
Cook
S. J.
,
Issel
C. J.
,
Montelaro
R. C.
.
2007
.
Envelope-specific T-helper and cytotoxic T-lymphocyte responses associated with protective immunity to equine infectious anemia virus.
J. Gen. Virol.
88
:
1324
1336
.
19
McGuire
T. C.
,
Leib
S. R.
,
Mealey
R. H.
,
Fraser
D. G.
,
Prieur
D. J.
.
2003
.
Presentation and binding affinity of equine infectious anemia virus CTL envelope and matrix protein epitopes by an expressed equine classical MHC class I molecule.
J. Immunol.
171
:
1984
1993
.
20
Mealey
R. H.
,
Lee
J. H.
,
Leib
S. R.
,
Littke
M. H.
,
McGuire
T. C.
.
2006
.
A single amino acid difference within the α-2 domain of two naturally occurring equine MHC class I molecules alters the recognition of Gag and Rev epitopes by equine infectious anemia virus-specific CTL.
J. Immunol.
177
:
7377
7390
.
21
Huang
J.
,
Goedert
J. J.
,
Sundberg
E. J.
,
Cung
T. D.
,
Burke
P. S.
,
Martin
M. P.
,
Preiss
L.
,
Lifson
J.
,
Lichterfeld
M.
,
Carrington
M.
,
Yu
X. G.
.
2009
.
HLA-B*35-Px-mediated acceleration of HIV-1 infection by increased inhibitory immunoregulatory impulses.
J. Exp. Med.
206
:
2959
2966
.
22
Hülsmeyer
M.
,
Fiorillo
M. T.
,
Bettosini
F.
,
Sorrentino
R.
,
Saenger
W.
,
Ziegler
A.
,
Uchanska-Ziegler
B.
.
2004
.
Dual, HLA-B27 subtype-dependent conformation of a self-peptide.
J. Exp. Med.
199
:
271
281
.
23
Chung
C.
,
Leib
S. R.
,
Fraser
D. G.
,
Ellis
S. A.
,
McGuire
T. C.
.
2003
.
Novel classical MHC class I alleles identified in horses by sequencing clones of reverse transcription-PCR products.
Eur. J. Immunogenet.
30
:
387
396
.
24
Ramsay
J. D.
,
Leib
S. R.
,
Orfe
L.
,
Call
D. R.
,
Tallmadge
R. L.
,
Fraser
D. G.
,
Mealey
R. H.
.
2010
.
Development of a DNA microarray for detection of expressed equine classical MHC class I sequences in a defined population.
Immunogenetics
62
:
633
639
.
25
Tallmadge
R. L.
,
Campbell
J. A.
,
Miller
D. C.
,
Antczak
D. F.
.
2010
.
Analysis of MHC class I genes across horse MHC haplotypes.
Immunogenetics
62
:
159
172
.
26
Holmes
E. C.
,
Ellis
S. A.
.
1999
.
Evolutionary history of MHC class I genes in the mammalian order Perissodactyla.
J. Mol. Evol.
49
:
316
324
.
27
Tallmadge
R. L.
,
Lear
T. L.
,
Antczak
D. F.
.
2005
.
Genomic characterization of MHC class I genes of the horse.
Immunogenetics
57
:
763
774
.
28
Shen
R.-X.
,
Xu
Z.
,
He
X.
,
Zhang
S.
.
1979
.
Study on immunological methods of equine infectious anemia.
China Agric. Sci.
4
:
1
15
.
29
Hoof
I.
,
Peters
B.
,
Sidney
J.
,
Pedersen
L. E.
,
Sette
A.
,
Lund
O.
,
Buus
S.
,
Nielsen
M.
.
2009
.
NetMHCpan, a method for MHC class I binding prediction beyond humans.
Immunogenetics
61
:
1
13
.
30
Nielsen
M.
,
Lundegaard
C.
,
Blicher
T.
,
Lamberth
K.
,
Harndahl
M.
,
Justesen
S.
,
Røder
G.
,
Peters
B.
,
Sette
A.
,
Lund
O.
,
Buus
S.
.
2007
.
NetMHCpan, a method for quantitative predictions of peptide binding to any HLA-A and -B locus protein of known sequence.
PLoS One
2
:
e796
.
31
Li
X.
,
Liu
J.
,
Qi
J.
,
Gao
F.
,
Li
Q.
,
Li
X.
,
Zhang
N.
,
Xia
C.
,
Gao
G. F.
.
2011
.
Two distinct conformations of a rinderpest virus epitope presented by bovine major histocompatibility complex class I N*01801: a host strategy to present featured peptides.
J. Virol.
85
:
6038
6048
.
32
Chen
W.
,
Gao
F.
,
Chu
F.
,
Zhang
J.
,
Gao
G. F.
,
Xia
C.
.
2010
.
Crystal structure of a bony fish β2-microglobulin: insights into the evolutionary origin of immunoglobulin superfamily constant molecules.
J. Biol. Chem.
285
:
22505
22512
.
33
Liu
J.
,
Sun
Y.
,
Qi
J.
,
Chu
F.
,
Wu
H.
,
Gao
F.
,
Li
T.
,
Yan
J.
,
Gao
G. F.
.
2010
.
The membrane protein of severe acute respiratory syndrome coronavirus acts as a dominant immunogen revealed by a clustering region of novel functionally and structurally defined cytotoxic T-lymphocyte epitopes.
J. Infect. Dis.
202
:
1171
1180
.
34
Zhou
M.
,
Xu
Y.
,
Lou
Z.
,
Cole
D. K.
,
Li
X.
,
Liu
Y.
,
Tien
P.
,
Rao
Z.
,
Gao
G. F.
.
2004
.
Complex assembly, crystallization and preliminary X-ray crystallographic studies of MHC H-2Kd complexed with an HBV-core nonapeptide.
Acta Crystallogr. D Biol. Crystallogr.
60
:
1473
1475
.
35
Chu
F.
,
Lou
Z.
,
Gao
B.
,
Bell
J. I.
,
Rao
Z.
,
Gao
G. F.
.
2005
.
Complex assembly, crystallization and preliminary x-ray crystallographic studies of rhesus macaque MHC Mamu-A*01 complexed with an immunodominant SIV-Gag nonapeptide.
Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun.
61
:
614
616
.
36
Li
L.
,
Bouvier
M.
.
2004
.
Structures of HLA-A*1101 complexed with immunodominant nonamer and decamer HIV-1 epitopes clearly reveal the presence of a middle, secondary anchor residue.
J. Immunol.
172
:
6175
6184
.
37
Emsley
P.
,
Cowtan
K.
.
2004
.
Coot: model-building tools for molecular graphics.
Acta Crystallogr. D Biol. Crystallogr.
60
:
2126
2132
.
38
Adams
P. D.
,
Grosse-Kunstleve
R. W.
,
Hung
L. W.
,
Ioerger
T. R.
,
McCoy
A. J.
,
Moriarty
N. W.
,
Read
R. J.
,
Sacchettini
J. C.
,
Sauter
N. K.
,
Terwilliger
T. C.
.
2002
.
PHENIX: building new software for automated crystallographic structure determination.
Acta Crystallogr. D Biol. Crystallogr.
58
:
1948
1954
.
39
Laskowski
R. A.
,
Macarthur
M. W.
,
Moss
D. S.
,
Thornton
J. M.
.
1993
.
Procheck—a program to check the stereochemical quality of protein structures.
J. Appl. Cryst.
26
:
283
291
.
40
Tobita
T.
,
Oda
M.
,
Morii
H.
,
Kuroda
M.
,
Yoshino
A.
,
Azuma
T.
,
Kozono
H.
.
2003
.
A role for the P1 anchor residue in the thermal stability of MHC class II molecule I-Ab.
Immunol. Lett.
85
:
47
52
.
41
Li
L.
,
Bouvier
M.
.
2005
.
Biochemical and structural impact of natural polymorphism in the HLA-A3 superfamily.
Mol. Immunol.
42
:
1331
1344
.
42
Sharma
A. K.
,
Kuhns
J. J.
,
Yan
S.
,
Friedline
R. H.
,
Long
B.
,
Tisch
R.
,
Collins
E. J.
.
2001
.
Class I major histocompatibility complex anchor substitutions alter the conformation of T cell receptor contacts.
J. Biol. Chem.
276
:
21443
21449
.
43
Webb
A. I.
,
Dunstone
M. A.
,
Chen
W.
,
Aguilar
M. I.
,
Chen
Q.
,
Jackson
H.
,
Chang
L.
,
Kjer-Nielsen
L.
,
Beddoe
T.
,
McCluskey
J.
, et al
.
2004
.
Functional and structural characteristics of NY-ESO-1-related HLA A2-restricted epitopes and the design of a novel immunogenic analogue.
J. Biol. Chem.
279
:
23438
23446
.
44
Saper
M. A.
,
Bjorkman
P. J.
,
Wiley
D. C.
.
1991
.
Refined structure of the human histocompatibility antigen HLA-A2 at 2.6 A resolution.
J. Mol. Biol.
219
:
277
319
.
45
Macdonald
I. K.
,
Harkiolaki
M.
,
Hunt
L.
,
Connelley
T.
,
Carroll
A. V.
,
MacHugh
N. D.
,
Graham
S. P.
,
Jones
E. Y.
,
Morrison
W. I.
,
Flower
D. R.
,
Ellis
S. A.
.
2010
.
MHC class I bound to an immunodominant Theileria parva epitope demonstrates unconventional presentation to T cell receptors.
PLoS Pathog.
6
:
e1001149
.
46
Wu
Y.
,
Gao
F.
,
Liu
J.
,
Qi
J.
,
Gostick
E.
,
Price
D. A.
,
Gao
G. F.
.
2011
.
Structural basis of diverse peptide accommodation by the rhesus macaque MHC class I molecule Mamu-B*17: insights into immune protection from simian immunodeficiency virus.
J. Immunol.
187
:
6382
6392
.
47
Zhang
N.
,
Qi
J.
,
Feng
S.
,
Gao
F.
,
Liu
J.
,
Pan
X.
,
Chen
R.
,
Li
Q.
,
Chen
Z.
,
Li
X.
, et al
.
2011
.
Crystal structure of swine major histocompatibility complex class I SLA-1*0401 and identification of 2009 pandemic swine-origin influenza A H1N1 virus cytotoxic T lymphocyte epitope peptides.
J. Virol.
85
:
11709
11724
.
48
Shen
T.
,
Liang
H.
,
Tong
X.
,
Fan
X.
,
He
X.
,
Ma
Y.
,
Xiang
W.
,
Shen
R.
,
Zhang
X.
,
Shao
Y.
.
2006
.
Amino acid mutations of the infectious clone from Chinese EIAV attenuated vaccine resulted in reversion of virulence.
Vaccine
24
:
738
749
.
49
Wang
X.
,
Wang
S.
,
Lin
Y.
,
Jiang
C.
,
Ma
J.
,
Zhao
L.
,
Lv
X.
,
Wang
F.
,
Shen
R.
,
Kong
X.
,
Zhou
J.
.
2011
.
Genomic comparison between attenuated Chinese equine infectious anemia virus vaccine strains and their parental virulent strains.
Arch. Virol.
156
:
353
357
.
50
Messaoudi
I.
,
Guevara Patiño
J. A.
,
Dyall
R.
,
LeMaoult
J.
,
Nikolich-Zugich
J.
.
2002
.
Direct link between mhc polymorphism, T cell avidity, and diversity in immune defense.
Science
298
:
1797
1800
.
51
Price
D. A.
,
West
S. M.
,
Betts
M. R.
,
Ruff
L. E.
,
Brenchley
J. M.
,
Ambrozak
D. R.
,
Edghill-Smith
Y.
,
Kuroda
M. J.
,
Bogdan
D.
,
Kunstman
K.
, et al
.
2004
.
T cell receptor recognition motifs govern immune escape patterns in acute SIV infection.
Immunity
21
:
793
803
.
52
Slifka
M. K.
,
Whitton
J. L.
.
2001
.
Functional avidity maturation of CD8+ T cells without selection of higher affinity TCR.
Nat. Immunol.
2
:
711
717
.
53
Mealey
R. H.
,
Littke
M. H.
,
Leib
S. R.
,
Davis
W. C.
,
McGuire
T. C.
.
2007
.
Cloning and large-scale expansion of epitope-specific equine cytotoxic T lymphocytes using an anti-equine CD3 monoclonal antibody and human recombinant IL-2.
Vet. Immunol. Immunopathol.
118
:
121
128
.
54
Migueles
S. A.
,
Sabbaghian
M. S.
,
Shupert
W. L.
,
Bettinotti
M. P.
,
Marincola
F. M.
,
Martino
L.
,
Hallahan
C. W.
,
Selig
S. M.
,
Schwartz
D.
,
Sullivan
J.
,
Connors
M.
.
2000
.
HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors.
Proc. Natl. Acad. Sci. USA
97
:
2709
2714
.
55
Pereyra
F.
,
Jia
X.
,
McLaren
P. J.
,
Telenti
A.
,
de Bakker
P. I.
,
Walker
B. D.
,
Ripke
S.
,
Brumme
C. J.
,
Pulit
S. L.
,
Carrington
M.
, et al
International HIV Controllers Study
.
2010
.
The major genetic determinants of HIV-1 control affect HLA class I peptide presentation.
Science
330
:
1551
1557
.
56
Loffredo
J. T.
,
Maxwell
J.
,
Qi
Y.
,
Glidden
C. E.
,
Borchardt
G. J.
,
Soma
T.
,
Bean
A. T.
,
Beal
D. R.
,
Wilson
N. A.
,
Rehrauer
W. M.
, et al
.
2007
.
Mamu-B*08-positive macaques control simian immunodeficiency virus replication.
J. Virol.
81
:
8827
8832
.
57
Yant
L. J.
,
Friedrich
T. C.
,
Johnson
R. C.
,
May
G. E.
,
Maness
N. J.
,
Enz
A. M.
,
Lifson
J. D.
,
O’Connor
D. H.
,
Carrington
M.
,
Watkins
D. I.
.
2006
.
The high-frequency major histocompatibility complex class I allele Mamu-B*17 is associated with control of simian immunodeficiency virus SIVmac239 replication.
J. Virol.
80
:
5074
5077
.
58
Rudolph
M. G.
,
Speir
J. A.
,
Brunmark
A.
,
Mattsson
N.
,
Jackson
M. R.
,
Peterson
P. A.
,
Teyton
L.
,
Wilson
I. A.
.
2001
.
The crystal structures of Kbm1 and Kbm8 reveal that subtle changes in the peptide environment impact thermostability and alloreactivity.
Immunity
14
:
231
242
.
59
Zhang
S.
,
Liu
J.
,
Cheng
H.
,
Tan
S.
,
Qi
J.
,
Yan
J.
,
Gao
G. F.
.
2011
.
Structural basis of cross-allele presentation by HLA-A*0301 and HLA-A*1101 revealed by two HIV-derived peptide complexes.
Mol. Immunol.
49
:
395
401
.

The authors have no financial conflicts of interest.

Supplementary data