Abstract Body

Background: About 1,500 of CTL epitopes including 250 best-defined epitopes and their restricting HLA allele data have accumulated in the Los Alamos database. However, most of these have been derived from subtype B-infected Caucasians, and from subtype C-infected Africans. Studies from south-east Asia, where about 1.5 million patients are living with HIV, and where about 90% of them are infected with subtype CRF01_AE, remain sparse. Although HLA-B*57 is a protective allele amongst Africans and Caucasians, its prevalence amongst Asians is lower than other ethnic groups. Methodology: 1) 557 HIV-1 CRF01_AE-infected Thais were recruited, and their HLA type and viral load were determined to statistically analyze the association of each allele in viral control. 2) In silico molecular dynamics was used to evaluate the effect of HLA structure variants on epitope binding. 3) Recently HLA-B*35-restricted epitope Gag 253-262 NPPIPVGDIY, which D260E escape mutation causing HLA-epitope binding instability and consequent viral load increase, was reported. Sequence mutation at this site was statistically analyzed amongst HLA-B*35:05 positives and HLA-B*35:01 positives. Results: HLA-B*35:05 was identified as the most protective allele (p=0.003, q=0.17), along with HLA-B*57:01 (p=0.044, q=0.31). Structurally HLAB* 35:05 belongs to PY, which has epitope binding motif of Pro (P) at HLA B pocket and Tyr (Y) at F pocket. However, unlike other HLA-B*35 alleles with Arg (R) at HLA residue 97, HLA-B*35:05 has unique sequences at Thr (T) at 94, Leu (L) at 95, and Ser (S) at 97, located at the peptide-binding groove. Analysis of 3D HLA structure and molecular dynamics showed that S97 in HLA-B*35:05 led to 1) less flexibility in the groove, and 2) shorter distances between the α-helixes compared to the other PY, HLA-B*35:01, suggesting fewer peptide variants will be able to bind to the groove, leading to induction of stronger virus-specific immunity by HLA-B*35:05. In the viral sequence at Gag 253-262 NY10, mutations were significantly identified amongst HLAB* 35:01 positives; p=0.0001 by Fisher’s exact test at D260E, p=0.0025 at I261X, and p=0.019 at flanking N252X. However, there was no significant mutations amongst HLA-B*35:05 positives in this site. Conclusions: Our data indicates the existence of a protective effect of HLA-B*57 across ethnic groups and highlights HLA-B*35:05 as an allele uniquely protective in subtype CRF01_AE-infected Thais, due to its divergence from conventional PY structural sequences at the peptide-binding groove. Identification of unique protective alleles in each endemic area provides the opportunity to better define the nature of HLA-mediated immune control and will be valuable for CTL vaccine development.