Control of memory CD4 T cell recall by the CD28/B7 costimulatory pathway.

The CD28/B7 costimulatory pathway is generally considered dispensable for memory T cell responses, largely based on in vitro studies demonstrating memory T cell activation in the absence of CD28 engagement by B7 ligands. However, the susceptibility of memory CD4 T cells, including central (CD62L(high)) and effector memory (T(EM); CD62L(low)) subsets, to inhibition of CD28-derived costimulation has not been closely examined. In this study, we demonstrate that inhibition of CD28/B7 costimulation with the B7-binding fusion molecule CTLA4Ig has profound and specific effects on secondary responses mediated by memory CD4 T cells generated by priming with Ag or infection with influenza virus. In vitro, CTLA4Ig substantially inhibits IL-2, but not IFN-gamma production from heterogeneous memory CD4 T cells specific for influenza hemagglutinin or OVA in response to peptide challenge. Moreover, IL-2 production from polyclonal influenza-specific memory CD4 T cells in response to virus challenge was completely abrogated by CTLA4Ig with IFN-gamma production partially inhibited. When administered in vivo, CTLA4Ig significantly blocks Ag-driven memory CD4 T cell proliferation and expansion, without affecting early recall and activation. Importantly, CTLA4Ig treatment in vivo induced a striking shift in the phenotype of the responding population from predominantly T(EM) in control-treated mice to predominantly central memory T cells in CTLA4Ig-treated mice, suggesting biased effects of CTLA4Ig on T(EM) responses. Our results identify a novel role for CD28/B7 as a regulator of memory T cell responses, and have important clinical implications for using CTLA4Ig to abrogate the pathologic consequences of T(EM) cells in autoimmunity and chronic disease.

Divergent generation of heterogeneous memory CD4 T cells.

Mechanisms for the generation of memory CD4 T cells and their delineation into diverse subsets remain largely unknown. In this study, we demonstrate in two Ag systems, divergent generation of heterogeneous memory CD4 T cells from activated precursors in distinct differentiation stages. Specifically, we show that influenza hemagglutinin- and OVA-specific CD4 T cells activated for 1, 2, and 3 days, respectively, exhibit gradations of differentiation by cell surface phenotype, IFN-gamma production, and proliferation, yet all serve as direct precursors for functional memory CD4 T cells when transferred in vivo into Ag-free mouse hosts. Using a conversion assay to track the immediate fate of activated precursors in vivo, we show that day 1- to 3-activated cells all rapidly convert from an activated phenotype (CD25(high)IL-7R(low)CD44(high)) to a resting memory phenotype (IL-7R(high)CD25(low)CD44(high)) 1 day after antigenic withdrawal. Paradoxically, stable memory subset delineation from undifferentiated (day 1- to 2-activated) precursors was predominantly an effector memory (CD62L(low)) profile, with an increased proportion of central memory (CD62L(high)) T cells arising from more differentiated (day 3-activated) precursors. Our findings support a divergent model for generation of memory CD4 T cells directly from activated precursors in multiple differentiation states, with subset heterogeneity maximized by increased activation and differentiation during priming.

Novel phenotypes and migratory properties distinguish memory CD4 T cell subsets in lymphoid and lung tissue.

Memory T cells are heterogeneous in expression of lymph node homing receptors, delineating "central-memory" (TCM, CD62Lhi/CCR7+) and "effector-memory" (TEM, CD62Llo/CCR7-) subsets that migrate to lymphoid and non-lymphoid tissues, respectively. It is not known how these subsets arise or how homing receptor expression and tissue origin determine their functional and migratory properties. Here, we investigated the role of CD62L expression in the generation, function, distribution and migration of heterogeneous memory CD4 T cells specific for influenza hemagglutinin (HA). We found that CD62Lhi and CD62Llo memory subsets are generated independent of CD62L expression by the activated precursor, and both subsets distribute into spleen and lung. Functionally, spleen- and lung-derived CD62L memory subsets produce effector cytokines at similar kinetics but differ strikingly in cell surface phenotype and migration: the CD62Llo memory subset expresses a classic memory phenotype (CD45RBlo/CD44hi/CD11a(hi)), while the CD62Lhi subset expresses an unconventional phenotype (CD45RBhi/CD44int/CD11a(int)), defining a new polyclonal memory subset. The CD62Lhi subset also trafficked more efficiently than CD62Llo cells into lymph nodes; however, only lung but not spleen CD62Llo memory T cells homed to lung. Our results reveal novel phenotypic heterogeneity of memory CD4 T cells co-segregating with CD62L expression and tissue-specific tropism of non-lymphoid memory CD4 T cells.

Anti-CD3 priming generates heterogeneous antigen-specific memory CD4 T cells.

Anti-CD3 activation of peripheral T cells is used in adoptive immunotherapy for cancer and HIV infection, but the long-term fate of anti-CD3-primed T cells in vivo is not known. In this study, we demonstrate that anti-CD3-mediated activation of influenza hemagglutinin (HA)-specific TCR-transgenic CD4 T cells results in generation of a long-lived HA-specific memory CD4 T cell population when transferred into lymphocyte-deficient and intact mouse hosts. This anti-CD3-primed memory population is indistinguishable from HA peptide-primed memory CD4 T cells in terms of phenotype, rapid recall function, and enhanced proliferative capacity. Moreover, anti-CD3 priming generates phenotypically heterogeneous memory subsets in lymphoid and non-lymphoid sites. Our results suggest that anti-CD3 has potential efficacy in generating memory responses in adoptive immunotherapies and vaccines and that the tissue distribution and maintenance of heterogeneous lymphoid and non-lymphoid memory T cell subsets are a stochastic process that can occur independent of antigen or TCR specificity.

Modulation of memory CD4 T cell function and survival potential by altering the strength of the recall stimulus.

Optimization of long term immunity depends on the functional persistence of memory T cells; however, there are no defined strategies for promoting memory T cell function and survival. In this study, we hypothesized that varying the strength of the recall stimulus could modulate the function and survival potential of memory CD4 T cells. We tested the ability of peptide variants of influenza hemagglutinin (HA) exhibiting strong and weak avidity for an HA-specific TCR, to modulate HA-specific memory CD4 T cells in vitro and in vivo. In vitro stimulation with a weak avidity peptide (L115) uncoupled memory CD4 T proliferation from effector cytokine production with low apoptosis, whereas stimulation with a strong avidity peptide (Y117) fully recalled memory T cell functions but triggered increased apoptosis. To determine how differential recall would affect memory T cells in vivo, we boosted BALB/c hosts of transferred, CFSE-labeled HA-specific memory CD4 T cells with native HA, Y117, and L115 variant peptides and found differences in early Ag-driven memory T cell proliferation and IL-7R expression, with subsequent changes in memory T cell yield. High avidity boosting resulted in rapid proliferation, extensive IL-7R down-regulation, and the lowest yield of HA-specific memory cells, whereas low avidity boosting triggered low in vivo proliferation, maintenance of IL-7R expression, and the highest memory T cell yield. Our results indicate that memory CD4 T cell function and survival can be modulated at the recall level, and can be optimized by low level stimulation that minimizes apoptosis and enhances responses to survival factors.

Association of Gag-specific T lymphocyte responses during the early phase of human immunodeficiency virus type 1 infection and lower virus load set point.

T lymphocyte responses to human immunodeficiency virus (HIV) type 1 Gag were measured in 9 patients by interferon-gamma enzyme-linked immunospot assay at 3 time points within 12 months of infection. Patients with early recognition of HIV-1 Gag had lower subsequent HIV-1 load set points, as measured during the first 2 years of infection, compared with those of patients with undetectable Gag-specific responses (median, 4.27 vs. 5.05 log(10) HIV-1 RNA copies/mL, respectively; P=.028). An inverse correlation existed between the magnitude of the Gag-specific responses and the HIV-1 load set point (r=-0.733; P=.025). Early sustained T lymphocyte responses to HIV-1 Gag may be important for the establishment of virus load set point.