After more than 30 years of active research on HIV/AIDS, many challenging questions remain unanswered. For example, some individuals progress to AIDS within a year after HIV acquisition whilst others never develop the disease. Host genetic factors play a role in disease progression but the precise pathways are still unknown. Studies aimed at deciphering this are hampered by inter-individual variability and HIV’s high propensity to mutate: as a consequence the host-virus interactions can vary significantly between different infected people. How innate and adaptive immune mechanisms slow HIV replication is still ill defined and differs substantially amongst different ethnic groups. As most studies to date have focused on patients of European ancestry, it has become a global health priority to determine how HIV control is achieved in non-Caucasians (27
). In the present study, we assessed the protective effect of -35
SNP in Han Chinese subjects who were infected in a short time-frame by a very narrow range of HIV-1 strains. In addition to being the first report on the effect of -35
SNP in an Asian population, this study confirms the protective effect of the variant in a setting where confounders such as viral strain diversity and time after HIV acquisition are largely controlled. We analysed six substitutions in HIV’s proviral DNA which were associated with HLA-C (C-mut) in order to determine whether immune pressure on the virus was stronger in -35CC
than in -35TT
subjects. C-mut were identified using sequencing data from PBMC proviral DNA rather than plasma RNA, as mutations that had reached fixation in the population were more relevant to the purpose of the analysis than quasispecies from an actively replicating population (28
). The key finding of this study is that subjects with the protective CC
genotype at -35 show a higher C-mut frequency than their TT
counterparts, suggesting that the virus is exposed to a stronger immune pressure in presence of the CC
genotype. We suggest that HLA-C-restricted CTLs play a role in the protective effect of the variant, without excluding the possibility that NK-cells or some HLA alleles also contribute to viral control.
Of all SNPs associated with delayed HIV disease progression, -35
SNP is consistently prominent, clearly implicating HLA-C as a key player in HIV control. The mechanisms underlying this have been partially elucidated with the identification of the 3′UTR variant but remain subject to considerable debate. On the other hand, there are associations between HLA-C and several autoimmune diseases (29
). Notably, -35
SNP is also associated with increased susceptibility to psoriasis where disease is thought to be mediated by HLA-Cw*06 restricted CTLs specific for self peptides, a response that could potentially be elicited as a result of enhanced HLA-C expression (30
). HLA-C is involved in other autoimmune disorders suggesting that it is somehow implicated in T-cell tolerance. HLA-C binds a restricted range of self peptides as a result of limited polymorphism in its α1 domain and prolonged association with chaperones in the endoplasmic reticulum. This could result in HLA-C molecules being more available to bind pathogen-derived peptides, which could result in specific targeting of key viral molecules, thus countering pathogen replication. Supporting this, the molecule H-2Ld
corresponding to HLA-C in the mouse, is poorly expressed and defective in binding endogenous self peptides but is the dominant CTL-restriction element for recognition of Vesicular Stomatitis and Lymphocytic Choriomeningitis viruses (32
). Therefore, the contribution of HLA-C to pathogen recognition in peripheral and lymphoid organs could be distinct but complementary to that of HLA-A and –B.
HLA-C molecules could also play a role during thymic T-cell selection (33
). The efficacy of thymic negative selection is proportional to the expression level of MHC complexes in the thymic medulla, which means that fewer clones with a high avidity for self MHC-peptide complexes are deleted when MHC expression is low. Assuming that HLA-C expression in medullary epithelial and dendritic cells is lower than that of HLA-A and –B (as it is in other tissues), HLA-C-mediated negative selection might be less stringent. Thymic negative selection imposes specificity on the T-cell repertoire, consequently a less efficient negative selection leads to the generation of a highly avid T-cell repertoire (34
). Therefore, it is tempting to speculate that the T-cell repertoire restricted by HLA-C comprises several high-avidity clones able to effectively recognise viral peptides in the periphery. Selection of a high avidity T-cell repertoire could be a double-edged sword: it might lead to better pathogen recognition but also result in the generation of auto-immune pathologies. The dual association of -35
SNP linking HLA-C to better HIV control and susceptibility to psoriasis seems to corroborate this hypothesis. Moreover differential HLA-C expression could also influence thymic positive selection by allowing a more diverse range of T-cell clones to be selected in subjects with the CC
genotype where HLA-C expression is thought to be higher than in TT
individuals. This hypothesis remains to be tested.
We report here that increased expression of HLA-C in the -35
SNP leads to the generation of escape mutations in HIV. This is the first function-based evidence linking -35
SNP to HIV control in vivo
. Whether HLA-C expression levels are directly responsible for the protective effect of -35
SNP (or 263del) or strong LD between the variants and other protective genes in the HLA locus exert viral control is still unclear. This is extremely difficult to disentangle, therefore we cannot exclude the possibility that some protective HLA-alleles have an effect on delayed progression to AIDS independently or in conjunction with -35
SNP. On the other hand, as LD between HLA alleles in the Han Chinese substantially differs from that of the Caucasians, we were able to partially separate some associations with -35
SNP to conclude that the protective effect of the variant can be dissociated from the effect of the main protective HLAs (i.e. B*57, B*27). Although polymorphisms in the binding site of miRNA-148a (HLA-C 3′UTR) provide a mechanism explaining how HLA-C expression can be modulated, we cannot exclude that HLA-C alleles in LD with -35CC
play a role in protection. However a growing body of evidence support the concept that increased HLA-C expression results in a more effective immune response through mechanisms such as NK-cell licensing and activation, antigen cross presentation to CTLs and increased survival of APCs (35
). Further studies are required, nevertheless our data support the concept that HIV-specific CTLs are key components of HIV control in vivo
and highlight the relevance of exploring HLA-C-restricted CTLs in the design of a potential HIV vaccine.