HIV-1–specific CD8+ T cells can mature to memory in patients receiving ART. Part 4
Although the basis for this generalized failure of HIV-1–specific T cells to mature into terminally differentiated effectors is still not clear, a complete blockade in T cell maturation would predict that the HIV-1–specific CD8+ T cell responses could not mature to memory in patients receiving ART and controlling virus well. It is interesting to note that HIV-1–specific T cells can mature to memory given a sufficient level of control over the virus. Of 11 patients with PVLs <50 HIV RNA copies/mL, as measured by the Ultrasensitive assay, 6 maintained responses that were primarily of an acute phenotype, and 5 had responses with a significant proportion of memory cells. It is important to note that the range of CD8+ T cells within the acute and memory phenotype was 21%–77% and 19%–77%, respectively, and that a difference of >3% was significant in our assay (see Subjects, Materials, and Methods). Longitudinal phenotype measures for some of the patients showed that the HIV-1–specific T cell responses shift toward memory as PVLs decrease and revert to the acute expansion phenotype as PVL increase. For the 11 patients taken together, the ability to shift to memory does not seem to relate to starting PVLs or CD4+ T cell counts but rather to the duration of a high degree of control. Indeed, on the basis of the routine Amplicor and Ultrasensitive PVL measurements for 10 of these patients, we were able to demonstrate that the fraction of HIV-1–specific CD8+ T cells in acute expansion correlated inversely with duration of virus control in patients receiving ART.
Therefore, we find that combined CD45RA/CD27 and CD45RA/CD28 staining can be used effectively to distinguish the acute, effector, and memory subsets in both resolved and persistent infections. This is in contrast to the recent conclusion by Appay et al., who suggested that memory and effector subsets are not accurately described by combined CD27/CD28 staining. This conclusion is based on the observation that, at selected timepoints very early in acute HIV-1 infection (before seroconversion) and in acute EBV and HCV infection, the CD8+ T cells can be mixed for CD28 expression. We believe that this reflects an initial priming period, during which the cells lose CD28 expression as they are being activated. This was clearly shown by Trimble et al., whose detailed study of acute versus convalescent EBV infection indicated that CD28 expression was sequentially lost on expanding CD8+ T cells during the first 10 days of infection. There was a parallel increase in both CD38 and HLA-DR expression, so that the peak of CD38 and HLA-DR expression coincides with the nadir of CD28 expression. A detailed longitudinal analysis of CD28 expression on tetramer-positive cells in HIV-1 infection also shows that, within the acute period of infection, the cells do become primarily CD28−, but this happens over a period of 2–4 weeks. Therefore, in the very early stages of acute viral infection, it appears that the loss of CD28 expression will span the period that it takes to fully activate the cells and that, during this period, combined CD27/CD28 staining may not fully depict all of the cells entering the acute phase. However, once the infection is established and the cells are fully primed, the differentiation states, as defined by these markers, do appear to correspond to acute, memory, and effector CD8+ T cell subsets. There is no doubt that in resolved FLU infection, when antigen is clearly absent or low, the tetramer-positive cells are memory. In well controlled HIV-1 infection, when the PVLs are highly suppressed for an extended duration, the cells approach the same maturation state as that for FLU, and this progression toward maturation correlated with PVL. In the longitudinal analyses of patients commencing or discontinuing ART, the cells are seen to shift toward the memory or acute subsets as the PVL decreases or increases, respectively. These data, coupled with the differences in the dominant maturation state for EBV and CMV tetramer-positive cells, support the view that antigen burden drives the maturation of CD8+ T cells toward the acute, memory, or terminally differentiated effector stages and that the markers accurately reflect these maturation states