Our recent published studies have suggested that impaired immune reconstitution after allogeneic stem cell transplantation is associated with a greater proportion of circulating late memory T cells, defined phenotypically. To characterize the relationship between immunophenotypic markers of T cell maturation and functional attributes of T cells, we optimized an 8-color, 10-parameter cytokine flow cytometry (CFC) approach and studied T cells from healthy donors. T cells were exposed to stimuli that both bypass (PMA:Ionomycin, P:I) and signal through the T cell receptor (Staph enterotoxin B, SEB; and CMV pp65 peptide pools) and stained with CD45RA and CD27 to demarcate naïve (N, CD45RA+CD27+), and three progressively mature memory subsets: M1 (CD45RA−CD27+), M2 (CD45RA−CD27−), M3 (CD45RA+CD27−) CD4+ and CD8+ T cells. We assessed the 15 possible combinations of cells producing IL-2, IFNγ, TNFα, and MIP1β alone or in combination within maturation subsets. When we initially studied the production of individual cytokines, we found that the bulk of IL-2 production was produced by activated N and M1 cells in both CD4 and CD8 lineages. In contrast, IFNγ and MIP1β were produced by later maturation stages (M2 and M3) of CD4+ and CD8+ T cells. In contrast to the polarized production of individual cytokines at the extremes of the maturation spectrum, early and middle memory cells (M1 and M2) cells produced heterogeneous combinations of cytokines (e.g, IL-2+IFNγ+ and TNFα+MIP-1β+ cells). We also found that IL-2/IFNγ co-producing cells, shown to be particularly important for the control of chronic viral pathogens, exist mainly in the M1 and M2 stages, and not the M3 stage. The above results were consistent with both P:I and SEB stimulation, and across several healthy subjects tested. Our cross-sectional results were confirmed by in vitro differentiation experiments, wherein we sorted naive (CD45RA+CD27+) CD4+ and CD8+ T cells and demonstrated that their function evolved as expected following stimulation with PHA and IL-2, which resulted in differentiation into M1 and M2 cells in culture. Finally, we stimulated PBMC from healthy CMV-seropositive donors with a CMV pp65 peptide mixtures and examined maturation and cytokine production. Consistent with prior observations, most CMV-specific T cells were M2 and M3 cells. Surprisingly, the most abundant functional subsets consisted of cells producing either MIP1β alone or MIP1β and other cytokines. Consistent with our results following polyclonal stimulation, we found that IL-2/IFNγ co-producing CMV specific T cells existed in M1 and M2, but not in the M3 stage. These results demonstrate that:

  1. Functional cytokine signature is strongly associated with T cell maturation stage;

  2. Nearly all IL-2 production occurs in N, M1 and M2 cells;

  3. M3 cells produce little IL-2, but substantial amounts of MIP1β;

  4. IL-2/IFNγ co-production is rare in M3 cells, but exist in M1 and M2 cells, perhaps suggesting why late stage skewing of memory T cells may lead to functional T cell impairment in vivo; and

  5. that MIP1β is the most abundant cytokine produced by CMV-specific T cells.

Overall, our results demonstrate that phenotypically defined maturation stages in both CD4+ and CD8+ T cell lineages are strongly associated with functional signatures irrespective of stimulus type, and that multidimensional analyses of human T cells may be beneficial when assessing human T cells in the clinical setting.

Author notes

Disclosure: No relevant conflicts of interest to declare.