Identifying HIV-1 peptides that HIV-specific CTLs recognize (epitope) is crucial for studies of elimination of HIV-infected cells and CTL-escape mechanism of HIV-1.
Since these epitopes are presented by multiple HLA class I molecules, the ones restricted by each HLA class I antigen need to be revealed.
Also, diversity of HLA class I antigens differs among ethnic backgrounds and there is a huge difference among Caucasians, Asians and Blacks. Therefore, already- identified epitopes presented by HLA class-I antigen among Caucasians cannot always be useful for Asians. Moreover, because HIV-1 is highly variable among subtypes, epitopes need to be analyzed according to each subtype.

Under such circumstances, we have previously demonstrated that HIV-1 epitopes can be identified by Reverse Immunogenetics. By this method, we have succeeded in unveiling numerous HLA class 1 binding HIV-1 peptides, as well as identifying 33 types of epitopes of HIV subtype B presented by 5 types of HLA antigens, HLA-A*1101, - A*2402, -A*3303, -B*3501, -B*5101, which are frequently seen among Asian people. These epitopes are stored in HIV Immunology database at the Los Alamos National Laboratory, and many researchers in and outside Japan have used them for their studies.

Subtype B is prevalent in Japan and the Western countries, while in Africa, which is worst-affected, and in Asia where the infection is spreading rapidly, subtypes A, C, and E are predominant. Hence, from the global standpoint, epitope analysis of these subtypes is of great importance. Particularly, epitope analysis of subtype E is urgent, since subtypes E and C are prevailing in Southeast Asia and China, and also, about half of the sexually transmitted infections in Japan is subtype E. We have established a method of identifying the epitopes specific to subtype E, by using known epitopes of subtype B. In this method, we first examined the subtype E amino acid sequence corresponding to 8 types of epitopes of HLA-A*1101. This analysis enabled identification of 3 types of epitopes commonly recognized by subtypes B and E, and the sequences of 3 epitopes specific to subtypes B and E. This method is believed to be effective in identifying unknown epitopes of a subtype using already identified epitopes of another subtype.