The role of HIV-1-specific CD4+ T-cell responses in controlling HIV-1 infection remains unclear. Previous work has suggested that such cells are eliminated in the early stages of infection in most subjects, and thus cannot substantially contribute to host defense against HIV-1. Here, using flow cytometric detection of antigen-induced intracellular cytokines, we show that significant frequencies of gag specific, T-helper-1 CD4+ memory T cells are detectable in most subjects with active/progressive HIV-1 infection (median frequency, 0.12% of memory subset; range, 0-0.66%). Median frequencies of these cells were considerably higher in nonprogressive HIV-1 disease (0.40%), but there was substantial overlap between the two groups (range of nonprogressors, 0.10-1.7%). Continuous HIV-1 suppression with anti-retroviral therapy was associated with a time-dependent reduction in median frequencies of gag-specific CD4+ memory T cells: 0.08% in subjects treated for 4-24 weeks, and 0.03% in subjects treated for 47-112 weeks. Thus, functional HIV-1-specific CD4+ T cells are commonly available for support of anti-HIV-1 effector responses in active disease, but their decline with anti-retroviral therapy indicates that immunologic participation in long-term HIV-1 control will probably require effective vaccination strategies.

The rhesus macaque (RM) is a critical animal model for studies of viral pathogenesis and immunity, yet fundamental aspects of their cellular immune response remain poorly defined. One such deficiency is the lack of validated phenotypic signatures for their naive and memory T cell subsets, and the resultant unavailability of accurate information on their memory T cell development, homeostasis, and function. In this study, we report a phenotypic paradigm allowing definitive characterization of these subsets and their comprehensive functional analysis. Naive T cells are optimally delineated by their homogeneous CD95lowCD28highβ7 integrinint (CD4+) or CD95lowCD28intCD11alow (CD8+) phenotypes. This subset 1) was present in blood and secondary lymph tissues, but not effector sites; 2) vastly predominated in the fetal/neonatal immune system, but rapidly diminished with postnatal age; 3) lacked IFN-γ production capability, and specific responses to RM CMV; and 4) demonstrated low in vivo proliferative activity. CD4+ and CD8+ memory subsets were CD95high, but otherwise phenotypically heterogeneous and included all IFN-γ production, RM CMV-specific responses, effector site T cells, and demonstrated high in vivo proliferative activity (∼10 times the naive subset). These analyses also revealed the RM “effector memory” subset within the overall memory population. This population, best defined by lack of CD28 expression, contained the majority of RM CMV-specific cells, was highly enriched in extralymphoid effector sites, and comprised an increasing proportion of total memory cells with age. The effector memory subset demonstrated similar in vivo proliferative activity and survival as CD28+ “central memory” T cells, consistent with independent homeostatic regulation.

The highly regulated secretion of effector cytokines by CD4 ϩ T cells plays a critical role in immune protection against pathogens such as cytomegalovirus. Here, we directly compare the frequency and functional characteristics of cytomegalovirus-specific CD4 ϩ memory/effector T cells in normal and HIV ϩ subjects using a novel, highly efficient multiparameter flow cytometric assay that detects the rapid intracellular accumulation of cytokine(s) after short-term (6 h) in vitro antigen stimulation. Responses in this assay correlate precisely with independent measures of sensitization history (e.g., seroreactivity), and allow the simultaneous assessment of multiple cytokines in single effector T cells. Healthy HIV Ϫ individuals manifested an average of 0.71, 0.72, 0.38, and 0.06% CD4 ϩ T cells responding to cytomegalovirus with ␥ -IFN, TNF-␣ , IL-2, and IL-4 production, respectively, with the simultaneous production of ␥ -IFN, TNF-␣ , and IL-2 being the most common effector phenotype. Significantly, overall cytomegalovirus-specific CD4 ϩ effector frequencies were markedly higher among 40% of HIV ϩ subjects (2.7-8.0%), and demonstrated a predominately polarized ␥ -IFN ϩ /TNF-␣ϩ /IL-2 Ϫ /IL-4 Ϫ phenotype. In contrast, CD4 ϩ effector frequencies for heterologous, nonubiquitous viruses such as the mumps virus were low or absent in the HIV ϩ group. These data suggest the existence of homeostatic mechanisms in HIV disease that selectively preserve memory T cell populations reactive with ubiquitous pathogens such as cytomegalovirus-likely at the expense of T cell memory to more sporadically encountered infectious agents. ( J. Clin. Invest. 1997. 99:1739-1750.)

No abstract