A molecular switch in immunodominant HIV-1-specific CD8 T-cell epitopes shapes differential HLA-restricted escape

Presentation of identical HIV-1 peptides by closely related Human Leukocyte Antigen class I (HLAI) molecules can select distinct patterns of escape mutation that have a significant impact on viral fitness and disease progression. The molecular mechanisms by which HLAI micropolymorphisms can induce differential HIV-1 escape patterns within identical peptide epitopes remain unknown. Here, we undertook genetic and structural analyses of two immunodominant HIV-1 peptides, Gag180–188 (TPQDLNTML, TL9-p24) and Nef71–79 (RPQVPLRPM, RM9-Nef) that are among the most highly targeted epitopes in the global HIV-1 epidemic. We show that single polymorphisms between different alleles of the HLA-B7 superfamily can induce a conformational switch in peptide conformation that is associated with differential HLAI-specific escape mutation and immune control. A dominant R71K mutation in the Nef71-79 occurred in those with HLA-B*07:02 but not B*42:01/02 or B*81:01. No structural difference in the HLA-epitope complexes was detected to explain this observation. These data suggest that identical peptides presented through very similar HLAI landscapes are recognized as distinct epitopes and provide a novel structural mechanism for previously observed differential HIV-1 escape and disease progression.


Background
The human leukocyte antigen (HLA) locus on chromosome 6 is the most polymorphic region of the human genome. The extreme diversity of HLA class I (HLAI) loci allows optimal binding of peptides derived from the vast array of environmental pathogens [1]. The HLAI residues that are polymorphic are mainly those forming the peptide-binding groove, which contains six binding pockets (A to F) that define the size and chemical characteristics of the specific peptide repertoire that can be accommodated by each HLAI molecule. Interaction between peptide and HLAI is usually governed by the compatibility of residues at the N-and C-terminus of the peptide (peptide anchor residues) within the highly polymorphic binding pockets. Cognate T-cell receptors (TCRs) expressed on CD8 + T-cells detect pathogenderived peptides presented by HLAI molecules on the surface of infected cells [2].
Minor differences between the HLAI molecules expressed ('HLAI micropolymorphisms'), even of a single amino acid, can have a profound impact on both T-cell immunity and disease outcome during a range of infections, including HIV-1 [3][4][5][6][7][8]. This disparity has been partly explained by the selection of escape mutations that have different consequences for viral fitness [8][9][10]. For example, the dominant HIV-1-specific epitope restricted by two closely-related HLAI molecules within the B7 superfamily, HLA-B*42:01 and HLA-B*81:01, is the same peptide, Gag 180-188 TPQDLNTML (TL9-p24) [11,12]. The escape mutation selected in each case differs, with variation most commonly arising at Gag-182 (position-3 in the epitope) in HLA-B*42:01-positive subjects, and at Gag-186 (position-7) in HLA-B*81:01positive subjects. In vitro studies indicate that the escape mutants selected at Gag-186 have a dramatic negative impact on viral replication capacity, in contrast to the minimal effect of mutation at Gag-182 [13]. HLA-B*81:01, together with HLA-B*57 and HLA-B*58:01, is one of the group of HLAI molecules most strongly associated with immune control of HIV-1 [5,14,15] and the viral replicative capacity in HLA-B*81:01 + subjects has been reported to be lower than in subjects expressing any other allele [13]. The mutation principally responsible for this protective effect is the above-mentioned HLA-B*81:01-driven T186S variant that fails to yield replicating virus stocks in studies of C clade virus [13]. This is in line with the lower viral loads and higher CD4 counts observed in HLA-B*81:01 positive subjects in studies of exclusively C-clade infection [9,14,16]. Notably, in a study of cohorts in Zambia, Rwanda and Kenya involving several clades of virus, including clade A which carries a different consensus residue at Gag-186, stable CD4 counts but not lower viraemia was observed in association with HLA-B*81:01 [15].
In this study we focus on two HIV-1-specific epitopes and their presentation by 4 different members of the HLA-B7 superfamily, HLA-B*07:02, HLA-B*42:01, HLA-B*42:02 and HLA-B*81:01. These alleles are highly prevalent in Sub-Saharan Africa, the region worst afflicted by the HIV-1 pandemic, one or more being expressed in 35-40% of people comprising these populations [14]. The identical TL9-p24 and Nef 71-79, RPQVPLRPM (RM9-Nef) epitopes are presented by these alleles, and represent two of the dominant HIV-1-specific responses in Southern African study cohorts [14], being recognised by approximately 70% and 40%, respectively, of subjects expressing HLA B7-supertype alleles, and targeted by approximately 25% and 20%, respectively of all subjects in Southern African study cohorts, irrespective of HLAI type [14,17].
In the case of both epitopes, TL9-p24 and RM9-Nef, distinct patterns of escape mutations are induced by different HLAI molecules within the B7 superfamily [17]. We here have undertaken a genetic and structural approach to better understand the molecular mechanisms by which HLAI micropolymorphisms can influence the precise nature of T-cell escape and immune control through identical HIV-1 epitopes. These data demonstrate that even a single HLAI polymorphism can switch the conformation of a peptide in the HLAI binding groove, substantially altering both the peptide residues positioned to contact incoming TCRs, and consequently of the impact of variation at different residues within the same epitope.
Thus, HLAI micropolymorphisms among these 4 HLA-B alleles result in distinct patterns of immunodominance at a population level. However, even when the frequency of recognition of epitopes was similar, such as TL9-p24 targeting in HLA-B*42:01-positive or HLA-B*81:01-positive subjects; and RM9-Nef targeting in subjects expressing any of the 4 closely-related B7 superfamily HLAI molecules, selection pressure on HIV-1 was HLAI-specific.
A conformational switch induces altered presentation of TL9-p24 by HLA-B*81:01 compared to HLA-B*07:02 and HLA-B*42:01 We hypothesised that differential HLAI-specific selection pressure operating on the same epitope may result from structural differences in the HLA-peptide complex. To explore this notion, we first solved the atomic structures of HLA-B*07:02, HLA-B*42:01 and HLA-B*81:01 with TL9-p24. In addition, we solved the atomic structures of HLA-B*07:02, HLA-B*42:01, HLA-B*42:02 and HLA-B*81:01 with RM9-Nef (see below). All structures were determined to extremely high resolutions, between 1.18 Å and 2.09 Å, with crystallographic R work / R free ratios within accepted limits as shown in the theoretically expected distribution [19] (Additional file 1: Tables S1 and S2). The electron density around the HLAI binding groove and the peptide was unambiguous in all of the structures.
The total number of contacts, buried surface area and surface complementarities were comparable in TL9-p24 structures in complex with HLA-B*07:02, HLA-B*42:01 and HLA-B*81:01 ( Table 2). The overall conformation of  Figure 3G). Overall, therefore,  the altered network of contacts between the TL9-p24 peptide and the polymorphic residues in HLA-B*81:01, compared to those of HLA-B*07:02 and HLA-B*42:01, resulted in the upward display of different residues within the solvent exposed, central peptide bulge ( Figure 3H).
HLAI micropolymorphisms alter direct interactions with the RM9-Nef peptide, explaining differences in pHLAI stability and overall epitope presentation Met9 acted as primary anchor residues and Arg1, Val4 and Arg7 pointed away from the groove for potential TCR interactions ( Figure 4A). The total number of contacts, buried surface area and surface complementarities were comparable in all four structures (Table 3). Although the overall conformation of the RM9-Nef peptide backbone was similar for all 4 HLAI molecules, we observed important differences in the side chain orientations that could potentially impact T-cell recognition and viral escape, as detailed below. Despite distinct selection pressure at position 1 of RM9-Nef by HLA-B*07:02 (R71K) ( Figure 2F) (Table 1), peptide residue Arg1 was in a similar conformation in all 4 HLAIs ( Figure 4A). However, the altered conformation of Leu6 in the RM9-Nef-HLA-B*81:01 complex, compared with the other three complexes has similarities with the conformational switch described showing differences around peptide residues Arg1, Leu6 and Arg7 (black arrows pointing up indicate that the corresponding residue is solvent exposed and available for TCR contact, black arrows pointing down indicate that the corresponding residue is buried in the HLA groove; no arrow indicates a position between solvent exposed and buried). The position of the circled residue (Leu6) may be important to explain differential escape when presented by different HLAIs. The other two major differences in RM9-Nef peptide presentation were located at the junction between HLAI residue 147 and peptide residue Pro8, and between HLAI residue Tyr9 and peptide residue Pro2. The smaller side chain of Leu147 in HLA-B*81:01 formed no contacts with Pro8, whereas Trp147 in the 3 other HLAIs made multiple interactions with Pro8 ( Figure 4D). Similarly, in all but HLA-B*42:02, in which the side chain of His9 was too short, Tyr9 could contact Pro2 ( Figure 4E). These differences could explain the slightly reduced stability of RM9 bound to HLA-B*81:01 and HLA-B*42:02 ( Figure 2B,D).

Discussion
Taken together, these data demonstrate that, from the perspective of the TCR, TL9-p24 'looks' completely different in the context of HLA-B*81:01 than it does in the context of HLA-B*07:02 or HLA-B*42:01. This difference in potential TCR contact residues is consistent with the differential viral escape patterns observed between these closely related HLAI molecules. Indeed, 4 of these differences are only selected by HLA-B*81:01, of which two impact on viral replicative capacity [13].
We were unable to generate enough soluble HLA-B*42:02-TL9 protein to generate a crystal structure, consistent with the low stability and lack of recognition of this pHLAI [7]. HLA-B*42:02 has a unique polymorphism within the B-pocket and contains His at residue 9 as opposed to Tyr9 in HLA-B*07:02, HLA-B*42:01 and HLA-B*81:01 molecules. Tyr9 in HLA-B*07:02, HLA-B*42:01 made direct contacts with the primary N-terminal anchor residue (Pro2) in the TL9-p24 peptide. Thus, His at peptide position 9 could alter this interaction and destabilise the HLA-B*42:02-TL9 complex. This notion is consistent with the lack of detectable responses observed to TL9-p24 in HLA-B*42:02 individuals [7], and is most likely a result of the low HLA-B*42:02-TL9 stability demonstrated here by both peptide-HLAI offrates and by temperature dependent circular dichroism experiments.
The four closely-related RM9-Nef peptide-HLAI structures show that similar conformations do not preclude HLAI-specific selection pressures, such as for R71K, that is only observed in the context of HLA-B*07:02. Our structural analysis demonstrated that R71 (position 1 in the peptide) pointed up, away from the HLAI groove and could therefore act as a putative TCR contact. Indeed, our previous work has shown that N-terminal peptide residue 1 can serve as an important TCR contact [20] in some systems. However, direct binding by a TCR to this residue would depend on the overall orientation of the TCR as most contacts are usually made with the central bulge of the peptide (normally residues 4-6 in the canonical 9-mer peptide). Thus, sensitivity to changes at peptide N-terminal position 1 would likely be highly dependent on the TCR sequence deployed by T-cells recognising RM9-Nef in the context of divergent HLAIs. Because the selection of self-ligands during thymic education are likely different depending on the HLA type, we speculate that HLA-B*07:02 expressing individuals may select a TCR, sensitive to the R71K mutation, that is not selected in HLA-B*42:01/42:02/81:01 positive individuals. TCR structures obtained from HLA-B*07:02 restricted RM9 specific T-cell clones sensitive to the R71K mutation are needed to investigate this notion further. These findings further highlight that HLA-allele specific HIV sequence changes at a population level are a highly sensitive measure of HLA-allele specific selective immune pressure.
Finally, these data demonstrate that a substantial alteration of the conformation of peptide residue P6 in the HLA-B*81:01-RM9 complex is, as with the TL9-p24 Gag-HLA-B*81:01 structure, associated with the differential selection of escape mutations observed in vivo. These observations in relation to both the TL9-p24 and RM9-Nef epitopes support the notion that different restriction elements presenting the same viral epitope in a structurally distinct conformation have an impact on the patterns of viral escape and, thereby, potentially also on immune control.

Conclusions
These data suggest that identical peptides presented through subtly different HLAI alleles can be recognized as distinct epitopes and provide a novel structural mechanism for previously observed differential HLA allele specific patterns of HIV-1 escape and disease progression.

Study subjects and HIV-1 sequence analysis
We studied 1,327 adults with chronic antiretroviral therapy (ART)-naïve C-clade HIV-1 infection recruited from Durban, South Africa [8,9] and from the Thames Valley Cohort, United Kingdom [21]. Informed consent was obtained from all participating individuals, and institutional review boards at the University of KwaZulu-Natal, Massachusetts General Hospital, and the University of Oxford approved the study. HIV-1 sequences from Gag and Nef proteins were generated [8] and analyzed [9] as previously described.

pHLAI stability assays
The measurement of pHLAI stability was determined with a dissociation assay based on radiolabeled β2m and biotinylated HLAI, as recently described [23]. Briefly, biotinylated HLAI heavy chain, 125 I-labeled β2m, and peptide were allowed to fold into pHLAI complexes in streptavidincoated scintillation microplates (Flashplate PLUS, Perkin Elmer, Boston, MA) for 24 h at 18°C. Excess of unlabeled β2m was added and dissociation was initiated by placing the microplate in a scintillation reader (TopCount NXT, Perkin Elmer, Boston, MA) operating at 37°C. The scintillation signal was monitored by continuous reading of the microplate for 24 h. Half-lives were calculated from dissociation curves using the exponential decay equation in Prism v.5.0a (GraphPad, San Diego, CA). Assays were performed in duplicate; the mean value of two experiments is reported.
Additionally, the thermal stability of HLA-B complexes was assessed by circular dichroism (CD) spectroscopy monitoring the change in ellipticities at 218 nm. Data were collected on an Aviv Model 215 spectropolarimeter (Aviv Biomedical Inc., Lakewood, NJ) using an 0.1-cm quartz cell. Proteins were dissolved in PBS at concentrations of 3 μM. Melting curves were recorded in 0.5°C intervals from 4°C up to a maximum temperature when protein aggregation was observed. Melting curves were analyzed assuming a two-state trimer-to-monomer transition from the native (N) to unfolded (U) conformation N 3 ↔ 3U with an equilibrium constant K = [U] 3 /[N 3 ] = F/[3c 2 (1-F) 3 ] where F and c are the degree of folding and protein concentration, respectively. Data were fitted as described [24]. Fitted parameters were the melting temperature T m , van't Hoff's enthalpy ΔH vH , and the slope and intercept of the native baseline. As all protein complexes aggregated to various degrees upon unfolding, the ellipticity of the unfolded state was set as a constant of −1.36 M −1 cm −1 [25].

Construct design
The HLAI heavy chains and β2m chain were generated by PCR mutagenesis (Stratagene) and PCR cloning. All sequences were confirmed by automated DNA sequencing (Lark Technologies). The HLAI heavy chains (residues 1-248) (α1, α2 and α3 domains), and β2m (residues 1 -100) were also cloned and used to make the pHLAI complexes. The HLAI α chains and β2m sequences were inserted into separate pGMT7 expression plasmids under the control of the T7 promoter [2].

Protein expression, refolding and purification
Competent Rosetta DE3 E.coli cells were used to produce the HLAI heavy chains and β2m in the form of