Background:  Recent studies have shown that PD-1 expression on HIV-1 specific CD8⁺ T cells correlates positively with plasma viral load, and the expression differs between epitope among a single person. This data suggested that there was qualitative difference between HIV-1-specific CD8⁺ T cells. PD-1 expression may be regulated by antigen encounter in vivo. To clear an association between PD-1 expression and antigen stimulation, we studied about PD-1 expression on CD8⁺ T cells specific for an immunodominant HIV-1 epitope, HLA-A*2402-restricted Nef138-10. A stereotypic amino acid substitution (phenylalanine for tyrosine at second position [2F]) in Nef138-10 were seen in HLA-A*2402⁺ patients specifically, suggesting HIV-1 with 2F were selected in vivo.

Methods:  We made 2 HLA-A24 tetramers; 1 presented Nef138-10(Tet [wild type]), the other presented Nef138-10(2F)(Tet [2F]). Peripheral blood mononuclear cells (PBMC) from 17 HLA-A*2402⁺ HIV-1-infected patients were cultured with Nef138-10 peptide for 2 weeks, and stained with these tetramers and anti-PD-1 antibody. Viral sequences of Nef138-10 were 2F in all patients. Staining pattern of these tetramers and PD-1 expression of tetramer⁺ CD8⁺ T cells were analyzed.

Results:  We found the staining pattern of dual-tetramer staining was quit different between patients. We grouped tetramer-positive CD8⁺ T cells into 2 populations, CD8⁺ T cells that could bind to Tet(2F) (Tet [wild type]⁺/Tet[2F]⁺ and Tet[wild type]^–/Tet[2F]⁺) and CD8⁺ T cells that bound to only Tet(wild type) (Tet[wild type]⁺/Tet[2F]^–). PD-1 expression on Tet(2F)⁺ cells were much higher than Tet(^–) and Tet(wild type)⁺/Tet(2F)^– cells and the difference was significant (p <0.0001 and p = 0.0001 respectively, Mann-Whitney U-test). Namely, higher expression of PD-1 was found in the population which could bind to 2F variant which was present antigen, in spite of lower PD-1 expression on Tet(wild type)⁺/Tet(2F)^– cells that bound only to antigen from eradicated HIV-1. Furthermore, we investigated an association between staining pattern of dual-tetramer and clinical status of the patients. CD4 T cell count, but not HIV viral load, was significantly higher in patients that PD-1^low Tet(wild type)⁺/Tet(2F)^– population was dominant compared with patients that PD-1^high Tet(2F)⁺ population was dominant.

Conclusions: These data suggest that PD-1 expression may be regulated by the amount of antigen encountered in vivo and HIV-1-specific PD-1^low CD8⁺ T cells which have proliferative capacity in vitro are maintained in patients with high CD4 count.