Expansion of HIV-1-specific CD28–CD45RA–CD8+ T cells in chronically HIV-1-infected individuals

There is growing evidence that cytotoxic T lymphocytes (CTL) play an important role in protection against HIV-1 infection and the suppression of HIV-1 replication in infected hosts. Analysis of CD8 T cells in HIV-1-infected individuals is therefore expected to contribute to the clarification of AIDS immunopathogenesis. Many phenotype analyses of CD8 T cells from HIV-1-infected individuals have been performed. The number of CD28−CD8+ T cells increases early in primary infection [1] and further during the progression towards AIDS [2–5]. It has been suggested that CD28−CD8+ T cells are involved in the control of virus replication [6–8]. A recent study [9] indirectly showed expansion of the CD28− phenotype in HIV-1-specific CD8 T cells. All the above studies imply that CD28−CD8+ T cells are cytotoxic effector cells. In the present study, we investigated CD28 and CD45RA expression on HIV-1-specific CD8 T cells by using HLA-B*3501 tetrameric complexes (tetramer) for six HIV-1 CTL epitopes. CTL clones specific for the six HLA-B*3501-restricted HIV-1 epitopes [10] were stained with the HLA-B*3501 tetramers and monoclonal antibodies specific for CD28 and CD45RA. These CTL clones expressed neither CD28 nor CD45RA (data not shown). This result is consistent with a recent study [11] showing that CD28−CD45RA−CD8+ and CD28−CD45RA+ CD8+ T cells are effector type CTL. Next, we examined peripheral blood mononuclear cells (PBMC) from the seven HIV-1-seronegative individuals and the six chronically HIV-1-infected individuals who showed a higher number of tetramer+CD8+ T cells. Analysis of CD28 expression on total CD8 T cells showed that the mean percentage of CD28−CD8+ T cells was much higher in chronically HIV-1-infected individuals (76.6%) than in HIV-1-uninfected individuals (40.6%) (Table 1). The CD28CD45RA phenotype of total CD8 T cells from the HIV-1-infected individuals was very different from that of the HIV-1-uninfected individuals. Compared with CD8 T cells from HIV-1-uninfected individuals, CD28−CD45RA− cells were significantly increased whereas CD28+CD45RA+ cells were decreased in chronically HIV-1-infected individuals (Table 1).Table 1: CD28 and CD45RA expression on HLA-B*3501 tetramer binding CD8 T cells from chronically HIV-1-infected individuals carrying HLA-B35. Analysis of the CD28CD45RA phenotype of tetramer+CD8+ T cells in PBMC from six HIV-1-infected individuals showed that 70.5~91.6% of total tetramer+CD8+ T cells were CD28−CD8+ T cells. In five of the six HIV-1-infected individuals, the percentage of CD28−CD45RA−CD8+ T cells was higher in tetramer+CD8+ T cells than in total CD8 T cells. However, in one individual (KI-001), a high number of CD28−CD45RA+CD8+ T cells was observed (Table 1). Therefore, in HIV-1-infected individuals, effector type CD8 T cells (CD28−CD45RA− and CD28− CD45RA+) form the major population of HIV-1-specific CD8 T cells. The percentage of the CD28−CD45RA− phenotype in tetramer+CD8+ T cells from two individuals (KI-019 and KI-038) was markedly reduced after highly active anti-retroviral therapy reduced the viral load (data not shown). These observations indicate that the expansion of HIV-1-specific CD28−CD45RA−CD8+ T cells results from chronic HIV-1 antigen stimulation. Previous studies [3–8] showed that the number of CD28−CD8+ T cells increased in the primary infection of HIV-1 and in chronically HIV-1-infected individuals. This was confirmed by the present study, which showed that the mean percentage of CD28−CD8+ T cells in total CD8 T cells was much higher in chronically HIV-1-infected individuals than in HIV-1-uninfected individuals. Furthermore, a recent study demonstrated that CD28−−CD8+ T cells are dominant in CD8 T cells responding to HIV-1 epitope stimulation [9]. However, this is an indirect demonstration of the expansion of HIV-1-specific CD28−−CD8+ T cells in HIV-1-infected individuals. The use of HLA-B*3501 tetramers in the present study directly demonstrated the expansion of the CD28− phenotype (70.5~91.6%) in HLA-B*3501-restricted, HIV-1-specific CD8 T cells. Effector CTL may be discriminated from naive or memory T cells by the pattern of CD45RA, CD28 and CD27 expression [11,12]. Naive CD8 T cells express CD45RA, CD28 and CD27 and memory CD8 T cells express CD28 and CD27 but lose CD45RA, whereas effector CD8 T cells express CD45RA but lose both CD28 and CD27. The expansion of CD45RA−CD28−CD27−CD8+ T cells is observed after viral infection, and these cells are enriched for perforin-containing T cells [11], indicating that T cells with this phenotype are also effector CD8 T cells. HIV-1-specific CD8 CTL clones used in the present study expressed neither CD28 nor CD45RA. CD28−CD45RA−CD8+ T cells have a high level of perforin in their cytoplasm (unpublished observation). Therefore, it is likely that these HIV-1-specific CD28−CD45RA−CD8+ T cells expanded in the PBMC of HIV-1-infected individuals are effector CTL. Acknowledgements The authors would like to thank Michiyo Tokunaga for technical assistance and Sachiko Sakai for secretarial assistance. Hiroko Tomiyamaa Shinichi Okab Graham S. Oggc Setsuko Idab Andrew J. McMichaelc Masafumi Takiguchia