Immunological Control of HIV-1 Disease Progression by Rare Protective HLA Allele

ABSTRACT Rare HLA alleles such as HLA-B57 are associated with slow progression to AIDS. However, the evidence for the advantage of rare protective alleles is limited, and the mechanism is still unclear. Although the prevalence of HLA-B*67:01 is only 1.2% in Japan, HLA-B*67:01-positive (HLA-B*67:01+) individuals had the lowest plasma viral load (pVL) and highest CD4 count in HIV-1 clade B-infected Japanese individuals. We investigated the mechanism of immunological control of HIV-1 by a rare protective allele, HLA-B*67:01. We identified six novel HLA-B*67:01-restricted epitopes and found that T cells specific for four epitopes were significantly associated with good clinical outcomes, pVL and/or CD4 count. The wild type or cross-reactive sequences of three protective and immunodominant Pol and Gag epitopes were found in around 95% of the circulating HIV-1, indicating that T cells specific for three conserved or cross-reactive epitopes contributed to good clinical outcomes. One escape mutation (Nef71K) in the Nef protective epitope, which was selected by T cells restricted by either HLA allele in the HLA-B*67:01-C*07:02 haplotype, affected the HLA-B*67:01-restricted RY11-specific T-cell recognition. These results imply that the further accumulation of the Nef71K mutation in the population will negatively affect the control of HIV-1 replication by RY11-specific CD8+ T cells in HIV-1-infected HLA-B*67:01+ individuals. The present study demonstrated that conserved or cross-reactive epitope-specific T cells mainly contribute to control of HIV-1 by a rare protective allele, HLA-B*67:01. IMPORTANCE HLA-B57 is a relatively rare allele around world and the strongest protective HLA allele in Caucasians and African black individuals infected with HIV-1. Previous studies suggested that the advantage of this allele in HIV-1 disease progression is due to a strong functional ability of HLA-B57-restricted Gag-specific T cells and lower fitness of mutant viruses selected by the T cells. HLA-B*57 is a very rare allele and has not been reported as a protective allele in Asian countries, whereas a rare allele, HLA-B*67:01, was shown to be a protective allele in Japan. Therefore, the analysis of HLA-B*67:01-restricted T cells is important to clarify the mechanism of immunological control of HIV-1 by a rare protective HLA allele in Asia. We found that HLA-B*67:01-restricted T cells specific for three conserved or cross-reactive Gag and Pol epitopes are associated with good clinical outcomes in HLA-B*67:01+ individuals. It is expected that T cells specific for conserved or cross-reactive epitopes contribute to a curing treatment.

protective effect against the progression of the disease (1,3,6,(10)(11)(12)(13)(14)(15)(16). Genome-wide association studies have further demonstrated strong genetic associations between single-nucleotide polymorphisms within HLA-B*57 and HLA-B*52 and their protective effect against HIV-1 disease progression (7,14,17,18). These studies implied that HIV-1-specific T cells restricted by these protective HLA alleles contribute to suppression of HIV-1 replication in HIV-1-infected individuals. Indeed, previous studies have demonstrated that the induction of HLA-B*57-and HLA-B*27-restricted CD8 1 T cells targeting HIV-1 Gag epitopes was associated with lower plasma viral load (pVL) in HLA-B*57-positive (HLA-B*57 1 ) and HLA-B*27 1 individuals, respectively (19)(20)(21)(22)(23). These findings suggest that Gag epitope-specific T cells have an important role in suppression of HIV-1 replication in individuals with the protective HLA alleles. HLA-B*57 is a very rare allele in Asian countries, including Japan, Thailand, and northern China (http://www .allelefrequencies.net) and was not reported as a protective allele in Asian countries, suggesting that HIV-1 replication may be controlled by other protective HLA alleles in these countries. A previous study showed that three HLA alleles, HLA-B*67:01, HLA-B*52:01, and HLA-C*12:02, were significantly associated with both low pVL and a high CD4 count in HIV-1 subtype B-infected Japanese individuals, indicating that these three HLA alleles have a protective effect on HIV-1-infected Japanese individuals (8).
Furthermore, it has been demonstrated that the epitope-specific CD8 1 T cells restricted by these three protective HLA alleles effectively suppressed HIV-1 replication in Japanese individuals (24)(25)(26).
HLA-B*67:01 is a rare allele in the world. It is absent or very rare (,0.1%) in Caucasians and Africans, whereas its prevalence in some Asian populations is 0.9%-1.2% (http://www .allelefrequencies.net). Although its prevalence is only 1.2% in Japan (http://hla.or.jp), HLA-B*67:01 1 individuals had the lowest pVL and highest CD4 count in HIV-1 clade B-infected Japanese individuals (8). The protective effect of this allele was also reported in HIV-1infected Chinese people in northern China (27). A previous study on 15 HIV-1-infected HLA-B*67:01 1 Japanese individuals identified only two Gag epitopes restricted by HLA-B*67:01 and suggested that these Gag-specific T cells reduced the pVL (24). HLA-B*67:01restricted T cells specific for other proteins have not been identified; hence, the role of HLA-B*67:01-restricted HIV-1-specific T cells in the slow progression of the disease remains only partially clarified.
A previous study on Caucasian and African-American populations showed an advantage of rare HLA alleles such as the HLA-B*58 supertype, which includes HLA-B*57, on HIV-1 disease progression (5). The advantage of HLA-B*57 may be explained by a strong ability of Gag-specific HLA-B*57-restricted T cells to suppress HIV-1 replication (11,(28)(29)(30) and lower the fitness of mutant viruses selected by these T cells (30)(31)(32). It was speculated that the effect of preadapted HIV-1 transmission was minimal in individuals harboring rare HLA alleles (33). To further understand the mechanism of the immunological control of HIV-1 by rare protective HLA alleles, it is valuable to investigate the role of HIV-1-specific T cells restricted by other HLA alleles which are less prevalent than HLA-B*57. The effect of preadapted HIV-1 transmission on such very rare alleles may be minimal compared with HLA-B*57. HLA-B*67:01 is a good target allele for these studies because the frequency of this allele is approximately 5 times lower than that of HLA-B*57.
In the present study, we investigated the HLA-B*67:01-mediated control of HIV-1 disease progression. We collected samples from 24 treatment-naive HIV-1 subtype B-infected HLA-B*67:01 1 individuals and then sought to identify novel HLA-B*67:01-restricted HIV-1 epitopes. We further analyzed the role of T cells specific for these HLA-B*67:01-restricted epitopes in HIV-1 suppression and accumulation of mutations in these epitopes. Here, we clarified the role of the very rare allele HLA-B*67:01 in HIV-1 infection.
Identification of novel HLA-B*67:01-restrcited HIV-1 T-cell epitopes. Our previous study showed that GagEM11 (EGATPQDLNTM)-induced T cells did not recognize the truncated peptides, whereas GagAN11 (ATPQDLNTMLN)-induced T cells recognized GagTL9 (TPQDLNTML) to a greater extent than the AL10 and AN11 peptides (24). From these results, we previously concluded that EM11 and TL9 are HLA-B*67:01 epitopes. We summarize the previous data related to the recognition of EM11-induced and AN11-induced T cells for the truncated peptides of GagEM11 and Gag AN11 ( Fig. 2A). However, because GagEM11 (EGATPQDLNTM) and GagTL9 (TPQDLNTML) overlap, we speculated that these peptides are cross-recognized or that another peptide is the epitope. We therefore investigated the recognition of EM11-induced T-cells established from KI-474 for the EM11, TL9, and TM8 peptides at different concentrations by performing an intracellular cytokine staining (ICS) assay (Fig. 2B). These T cells recognized EM11 and TL9 peptides at a concentration of 100 nM, whereas they recognized only TL9 peptide at 0.1 to 10 nM. These results suggested that EM11 is not an HLA-B*67:01-restricted epitope. We analyzed T-cell responses to EM11 and TL9 peptides further in 15 HIV-1infected HLA-B*67:01 1 individuals by performing an ex vivo ELISpot assay and compared the T-cell responses to these peptides. All responders to TL9 showed weaker or no responses to EM11 at a concentration of 100 nM (Fig. 2C). These results suggest that the responses of EM11-induced T cells to the EM11 peptide may be related to the cross-recognition of the overlapping part of the EM11 peptide by the T-cell receptor (TCR) of TL9specific T cells. These findings indicate that TL9 is an immunodominant epitope and EM11 is not an epitope.
Next, we analyzed three Nef (Nef4, 6, and 7) and three Pol (Pol4, 8, and 34) peptide cocktails to identify novel HLA-B*67:01-restricted T-cell epitopes. To elucidate HLA-B*67:01-restriction of T-cell responses to these Nef and Pol cocktails, we selected seven responders to these cocktails in the ELISpot assay (Table 1). We stimulated PBMCs from these responders with each peptide cocktail and cultured them for approximately 2 weeks. HLA-B*67:01-restricted T-cell responses to each peptide cocktail in the cultured cells were then analyzed by performing the ICS assay. HLA-B*67:01-restricted T-cell responses to four cocktail peptides were found in these responders as follows: T-cell responses to Nef4 in individuals KI-699 and KI-553, to Nef7 in individual KI-699, to Pol4 in individual KI-475, and to Pol34 in individual KI-475 (Fig. 3A).
These findings indicate that the T cells specific for these four HLA-B*67:01-restricted epitopes together contribute to suppression of the disease progression.
Effective recognition of circulating HIV-1 by T cells specific for Gag and Pol protective epitopes. To clarify the variations in the four protective epitopes among circulating HIV-1, we identified the sequences corresponding to these epitopes in 321 to 351 HIV-1-infected Japanese individuals ( Table 3). The wild-type (WT) sequences of GagTL9 and GagNL11 were highly conserved (.96%) among these individuals, while that of PolAL10 was detected in 83% of 348 individuals. The WT sequence of NefRY11 was found in 59.5% of 321 individuals. These results indicate that the Gag and Pol protective epitopes were relatively conserved among circulating viruses. Mutations in these two Gag epitopes were not found in HLA-B*67:01 1 individuals. PolAL10-10I was Next, we analyzed T-cell responses to PolAL10 and PolAL10-10I peptides using the ELISpot assay. We found that 92.3% of the responders to the PolAL10 peptide had positive responses to the PolAL10-10I peptide, and the responses to the PolAL10-10I peptide positively correlated with the responses to the WT peptide (Fig. 7A). These findings indicate that the PolAL10-10I mutant epitope was cross-recognized by PolAL10-specific T cells. Thus, HLA-B*67:01-restricted T cells specific for two Gag and one Pol protective epitopes can recognize the majority of circulating viruses.
Next, we investigated T-cell responses to NefRY11-1K by performing the ex vivo ELISpot assay in 24 HLA-B*67:01 1 individuals and found that T-cell responses to NefRY11-1K were not detected in any of the individuals tested, although NefRY11-specific T cells were found in 6 individuals (Fig. 8F). Taken together, these results suggest that NefRY11-1K is an escape mutant epitope from RY11-specific T cells. Because all six responders to NefRY11 had both HLA-B*67:01 and HLA-C*07:02, it is not clear whether the response to RY11 results from an HLA-B*67:01-restricted or HLA-C*07:02-restricted T-cell response. To elucidate the HLA-restriction of these T-cell responses, we analyzed the HLArestriction of NefRY11 peptide-induced bulk T cells from five individuals whose PBMCs were available for this analysis. All these bulk T cells recognized 721.221-B*67:01 cells a The statistical analyses of differences in pVL or CD4 count between responders to each epitope and nonresponses were conducted by using the two-tailed Mann-Whitney test. Multiple tests were performed using the q value, a measure of significance in terms of the false-discovery rate. Significant differences (P , 0.05, q , 0.1) are indicated in bold.

DISCUSSION
Although previous studies have suggested an advantage of rare HLA alleles against HIV-1 disease progression (5,36), the effects of rare HLA alleles on the disease progression and its mechanism are still unclear. A previous study on HIV-1-infected Japanese individuals showed that HLA-B*67:01 1 individuals exhibited the lowest pVL and highest CD4 counts among HIV-1 clade B-infected Japanese individuals (8). The prevalence of HLA-B*67:01 in Japan is approximately 1.2%, whereas that of HLA-B*57 in Caucasian and Black African populations is 2.8% to 8% (37)(38)(39)(40)(41)(42)(43). Thus, the rare HLA advantage in HIV-1 infection was found in the Japanese individuals harboring HLA-B*67:01. Because a previous study identified only Gag epitopes presented by HLA-B*67:01 (24), the mechanism of the protective effect of HLA-B*67:01 remains unclear. In the present study, we identified six novel HLA-B*67:01-restricted epitopes and demonstrated that four novel epitopes, NefRY11, NefRM9, NefRL9, and PolAL10, were immunodominant in addition to two previously identified Gag epitopes. Further analysis of T-cells specific for these epitopes showed that responders to NefRY11, PolAL10, GagTL9, or GagNL11 had higher CD4 counts and/or lower pVL than nonresponders. The breadths and total magnitude of T-cell responses to NefRY11, PolAL10, GagTL9, and GagNL11 showed a strong positive correlation with the CD4 count and an inverse correlation with the pVL. Thus, the present study demonstrated that these HLA-B*67:01-restricted T-cells had a strong ability to suppress HIV-1 replication in HIV-1-infected individuals.
NefRY11 and PolAL10 are less conserved than the two Gag epitopes among the four HLA-B*67:01 protective epitopes. Isoleucine at the C terminus of PolAL10 (PolL672I) was most frequently detected among these substitutions. However, PolL672I was not an HLA-B*67:01-associated mutation, suggesting that this is not an escape mutation from   ( (Table 3), and the T-cells response to the NefRY11 epitope was found in only 3 of 14 individuals infected with the NefR71K mutant (data not shown). These findings support the idea that selection and accumulation of NefR71K by HLA-C*07:02-restricted RY11-specific T cells may negatively affect the induction of HLA-B*67:01-restricted RY11-specific T cells if this mutant is transmitted to the HLA-B*67:01 1 individuals. It is assumed that NefRY11-specific T cells failed to suppress the replication of the Nef71K mutant virus in HLA-B*67:01 1 individuals infected with this mutant virus because NefRY11-specific T cells failed to recognize the RY11-1K mutant epitope. We found that three out of five responders to NefRY11 were infected with the Nef71K mutant virus, whereas the other two responders were infected with the WT virus. Responders harboring the mutant virus showed a trend for higher pVL and lower CD4 counts than those harboring the WT virus (median pVL, 61,000 versus 4,850; median CD4 count, 496 versus 899). Although a small number of HLA-B*67:01 1 responders were analyzed, this result implied that NefRY11-specific T cells failed to suppress the replication of the Nef71K mutant virus in HLA-B*67:01 1 individuals infected with the mutant virus. This finding together with the result that RY11-1K mutant-specific T cells were not induced in individuals infected with the Nef71K mutant virus (Fig. 8F)  Previous studies have demonstrated that polyfunctional and broad CD8 1 T-cell responses were induced by an HIV-1 conserved vaccine and that CD8 1 T cells specific for the conserved region contributed to suppression of HIV-1 replication in HIV-1 infection (34,(44)(45)(46)(47)(48). Here, we identified that the CD8 1 T cells specific for four HLA-B*67:01restricted epitopes were significantly associated with good clinical outcomes and that the majority of these protective epitopes were conserved and had a strong effect on suppression of HIV-1 replication. The present study provided additional evidence supporting the concept that CD8 1 T cells specific for the conserved region have a strong ability to suppress HIV-1 replication. Thus, CD8 1 T cells targeting the conserved regions are preferable candidates of effector T cells for an AIDS vaccine (49)(50)(51)(52)(53)(54) and the HIV-1 cure treatment (55).
GagTL9 is known to be an HLA-B*81:01-restricted epitope in HIV-1 subtype C infection (56,57). HLA-B*81:01 is associated with slow progression to AIDS in African populations infected with HIV-1 subtype C strains and the second strongest protective allele in southern Africa (6,30,(58)(59)(60). Previous studies have shown that the magnitude of Tcell responses to the GagTL9 epitope was associated with good clinical outcomes in HLA-B*81:01 1 black African individuals (56,61). Thus, GagTL9 is a protective T-cell epitope presented by both HLA-B*67:01 and HLA-B*81:01. Furthermore, previous studies have shown HLA-B*81:01-associated escape mutations (GagQ182S and GagT186S) within the GagTL9 epitope in chronically HIV-1 subtype C-infected individuals (62,63) and have shown that these mutations were found in 32% of HIV-1 subtype C-infected HLA-B*81:01 1 individuals (62). The emergence of these escape mutants at a late phase of the infection was associated with increased viremia (59). In contrast, the current study in Japan showed that GagTL9 is highly conserved among the circulating viruses and that the GagQ182S/A and GagT186L mutations were found in only 2.9% of the circulating viruses, but they were not detected in the HLA-B*67:01 1 individuals. Thus, the accumulation of these escape mutant viruses is only detected in black African individuals, even though protective GagTL9-specific T cells are elicited in both Japan and Africa. Because these mutations are associated with HLA-B*81:01 in Africa, we assume that GagTL9-specific HLA-B*81:01-restricted T cells can select these mutant viruses. A previous study demonstrated that HLA-B*67:01-restricted T cells cannot recognize GagTL9-3S-7L/TL9-3A-7L mutant peptides (64), implying that these T cells can select these mutant viruses in Japan. However, the reason for the accumulation of these mutations in Africa but not in Japan remains unknown.
HLA-B*67:01 has been reported as one of the HLA alleles associated with autoimmune diseases such as Takayasu arteritis and relapsing polychondritis (65)(66)(67)(68). HLA-B57 and HLA-B*27 are associated with drug hypersensitivity (69,70) and autoimmune diseases such as ankylosing spondylitis (71,72), respectively, while HLA-B*52:01, which Journal of Virology is a protective allele in HIV-1-infected Japanese individuals, is also associated with Takayasu arteritis (65,66,68,(73)(74)(75) and ulcerative colitis (65,68,76). These findings suggest that strong immune responses elicited by these HLA molecules contribute to not only the control of HIV-1 but also the onset of autoimmune diseases and hypersensitivity. It is interesting that these HLA alleles have different effects on infectious diseases and autoimmune diseases.
In the present study, we identified four protective HLA-B*67:01-restricted epitopes and demonstrated that three of them were conserved among circulating HIV-1 viruses. CD8 1 T-cells specific for these epitopes were effectively and stably elicited in the majority of HIV-1-infected HLA-B*67:01 1 individuals. Thus, we demonstrated that HLA-B*67:01 has a strong effect on HIV-1 suppression via immune control by HLA-B*67:01-restricited T cells specific for conserved epitopes and that preadapted mutation affected the recognition of Nef-specific T cells. In the current study, we elucidated part of the mechanism of the rare HLA allele advantage against HIV-1 infection.

MATERIALS AND METHODS
Subjects. All treatment-naive Japanese individuals chronically infected with HIV-1 subtype B were recruited between 2007 and 2019 from the National Center for Global Health and Medicine, Japan. This study was approved by the ethics committees of Kumamoto University (RINRI-1340 and GENOME-342) and the National Center for Global Health and Medicine (NCGM-A-000172-01). Informed consent was obtained from all individuals in accordance with the Declaration of Helsinki. Peripheral blood mononuclear cells (PBMCs) were separated from whole blood. The HLA type of HIV-infected individuals was determined by standard sequence-based genotyping. The pVLs of the individuals were measured using the Cobas TaqMan HIV-1 real-time PCR version 2.0 assay (Roche Diagnostics, NJ, USA).
Peptides. We previously designed overlapping peptides consisting of 11-mer amino acids spanning consensus sequences of Gag, Pol, and Nef of HIV-1 clade B (77). Each 11-mer peptide was overlapped by 9 amino acids. These 11-mer peptides and truncated peptides were synthesized by an automated multiple peptide synthesizer and purified by high-performance liquid chromatography (HPLC). The purity was examined by HPLC and mass spectrometry. Peptides with more than 90% purity were used.
IFN-c ELISpot assay. An ex vivo IFN-g ELISpot assay was performed as previously described (24,48). The number of spots of a T-cell response to each peptide was finally calculated by subtracting the number of spots in the wells without peptides from that in the wells with peptides. The mean value plus 3 standard deviations of spot number for the peptides among 13 HIV-1-naive individuals was 162 spots/ 10 6 CD8 1 T cells (24,48). Therefore, we defined .200 spots/10 6 CD8 1 T cells as a positive response.
Induction of HIV-1-specific T cells from HIV-1-infected individuals. PBMCs from individuals KI-699, KI-553, and KI-475 were incubated with a 1-mM 11-mer peptide cocktail of Nef4 or Nef7, Nef4, and Pol4 or Pol37, respectively, and cultured for 14 days to induce peptide-specific T cells. To establish HIV-1 epitope-specific bulk T cells, PBMCs were stimulated with a specific epitope peptide at 100 nM. After 2 to 3 weeks in culture, epitope-specific bulk T cells were analyzed by performing the intracellular cytokine staining (ICS) assay. The NefRY11 T-cell clone was generated from the RY11-specific bulk T cells by the limiting dilution assay in a 96-U plate, using 200 mL of cloning mixture of 2 Â 10 5 irradiated allogeneic PBMCs from healthy donors, 2 Â 10 4 irradiated 721.221-B*67:01 cells, and 100 nM RY11 peptide in RPMI 1640 containing FCS, 200 U/mL rIL-2, and 2.5% phytohemagglutinin (PHA). After 2 to 3 weeks in culture, the NefRY11-specific CD8 1 T-cell clone was used in the ICS assay.
Preparation of HIV-1 clones. The full-length HIV-1 pNL43 derivative in which the Nef gene was completely replaced with SF2 NEF (pNL43 SF2Nef ) was previously generated (78,79). 293T cells were transfected with the construct, and the infectious HIV-1 virions released into the medium were collected 72 h later.
Intracellular cytokine staining (ICS) assay. 721.221 cells prepulsed with HIV-1 epitope peptide or infected with HIV-1 virus, or CD4 1 T cells infected with HIV-1 virus were cocultured with HIV-1-specific bulkcultured T cells or a T-cell clone in a 96-well plate for 2 h at 37°C. Brefeldin A (10 mg/mL) was then added, and the cells were incubated for another 4 h at 37°C. Next, the cells were fixed with 4% paraformaldehyde and incubated in permeabilization buffer (0.1% saponin, 10% FBS-phosphate-buffered saline [PBS]), after which they were stained with allophycocyanin (APC)-conjugated anti-CD8 monoclonal antibody (MAb; Dako, Denmark), followed by fluorescein isothiocyanate (FITC)-conjugated anti-IFN-g MAb (BD Biosciences). The percentage of IFN-g-producing cells among the CD8 1 T-cell population was determined by FACS Canto II.
Identification of the HIV-1 epitope sequence. Determination of the epitope sequence was performed as previously described (35). Briefly, viral RNA was extracted from plasma samples using a QIAamp MinElute virus spin kit (Qiagen). cDNA was synthesized from the RNA using the SuperScript III first-strand synthesis system for reverse transcription-PCR (RT-PCR) and random hexamers (Invitrogen). The Nef, Gag, and Pol regions were amplified by nested PCR using Taq DNA polymerase (Promega). All sequences were determined using the BigDye Terminator version 3.1 cycle sequencing kit (Applied Biosystems) and ABI 3500 genetic analyzer (Applied Biosystems). The epitope sequence data of NefRY11 from 321 chronically HIV-1 subtype B-infected, treatment-naive Japanese individuals and the epitope sequence data of PolAL10 from 348 individuals were analyzed after excluding individuals with a mixture of amino acid sequences from previously analyzed data (35) and adding new data from nine HLA-B*67:01 individuals. The GagTL9 and GagNL11 sequence data were previously identified from 351 chronically HIV-1 subtype B-infected individuals (24).
Statistical analysis. The frequency of individuals harboring the mutants between HLA 1 and HLAones was statistically analyzed using Fisher's exact test. Groups were compared by performing the unpaired t test or two-tailed Mann-Whitney U tests. A P value of ,0.05 was considered significant. Multiple tests were performed using the q value, a measure of significance in terms of the false-discovery rate. A significance threshold for q of ,0.1 was used.