Progenitor and terminal subsets of CD8+ T cells cooperate to contain chronic viral infection

Abstract

Chronic infections strain the regenerative capacity of antiviral T lymphocyte populations, leading to failure in long-term immunity. The cellular and molecular events controlling this regenerative capacity, however, are unknown. We found that two distinct states of virus-specific CD8(+) T cells exist in chronically infected mice and humans. Differential expression of the T-box transcription factors T-bet and Eomesodermin (Eomes) facilitated the cooperative maintenance of the pool of antiviral CD8(+) T cells during chronic viral infection. T-bet(hi) cells displayed low intrinsic turnover but proliferated in response to persisting antigen, giving rise to Eomes(hi) terminal progeny. Genetic elimination of either subset resulted in failure to control chronic infection, which suggests that an imbalance in differentiation and renewal could underlie the collapse of immunity in humans with chronic infections.

Fig. 1

Eomes expression is up-regulated during chronic viral infection and supports enhanced CD8⁺ T cell exhaustion. (A) Sorted naïve or D^bgp276-specific CD8⁺ T cells from effector [acute; Armstrong day 8 (d8)], memory (acute; Armstrong d30), or exhausted (chronic; clone 13, d30) time points were analyzed for Eomes mRNA expression by real-time quantitative polymerase chain reaction. (B) Flow cytometric analysis of Eomes protein expression in D^bgp33-specific CD8⁺ T cells at indicated numbers of days post-infection (p.i.) (C) Mean fluorescence intensity (MFI) of Eomes expression from cells in (B). Graph displays means ± SEM. (D) Flow cytometric analysis of T-bet versus Eomes expression in D^bgp33-specifc CD8⁺ T cells at indicated days p.i. Gates are based on T-bet–negative and Eomes-negative cells. (E) Longitudinal frequency of GFP⁺YFP⁺ D^bgp33-specific CD8⁺ T cells in Blimp1-YFP/Eomes^gfp/+ dual reporter mice as analyzed by flow cytometry (GFP and YFP, green and yellow fluorescent protein, respectively). (F) Flow cytometric analysis of inhibitory receptor versus Eomes expression in D^bgp33-specific CD8⁺ T cells at d30 p.i. (G) Flow cytometric analysis of T-bet, Eomes, and PD-1 expression of D^bgp33-specific CD8⁺ T cells from wild-type (WT), T-bet KO (TKO), and Eomes cKO (EKO) mice at d22 p.i. (H) Blimp1 relative mRNA in D^bgp276-specific CD8⁺ T cells from WT and EKO mice at d15 p.i. (*P < 0.05; Mann-Whitney). (I) IFN-γ and TNF-α expression in CD8⁺ T cells from indicated mice after peptide pool stimulation as in fig. S1 (**P < 0.01; unpaired t test). In (B) to (G), numbers denote frequency of gated population; in (F) to (I), cells are from clone 13 infection. All data are representative of two to five independent experiments with at least three mice per experimental group.

Fig. 2

Conversion from T-bet^hi precursors to Eomes^hi progeny is critical for the antiviral CD8⁺ T cell response. (A) Flow cytometric analysis of T-bet, Eomes, and PD-1 expression versus BrdU incorporation in D^bgp33-specific CD8⁺ T cells monitored from d15 to d22 p.i. Gates are placed between T-bet^hi and Eomes^hi populations (fig. S4B). (B) Flow cytometric analysis of T-bet, Eomes, and PD-1 expression versus CFSE in D^bgp276-specific donor CD8⁺ T cells monitored from d15 to d22 p.i. (C) Flow cytometric analysis of BrdU incorporation in D^bgp276-specific CD8⁺ T cells from WT, TKO, and EKO mice monitored from d15 to d22. Upper panels are representative stains; lower panels are summarized data (*P < 0.05, **P < 0.01, ***P < 0.001; unpaired t test). In (A) to (C), data are representative of two to five independent experiments with at least three mice per experimental group. Gates were set on the basis of staining controls for BrdU, CFSE, GFP, or T-bet^hi and Eomes^hi subsets. (D) Flow cytometric analysis of PD-1 expression versus CFSE dilution of sorted PD-1^int or PD-1^hi D^bgp276-specific CD8⁺ T cells 7 days after transfer. (E) Eomes^+/+ and Eomes^F/F CD8⁺ T cells were isolated at d15 p.i. and treated in vitro with Tat-Cre. Two weeks after transfer, D^bgp276-specific CD8⁺ T cells were assessed for Eomes and PD-1 expression and CFSE dilution by flow cytometry. (F) T-bet^+/+ and T-bet^F/F CD8⁺ T cells were isolated at d15 p.i. and treated in vitro with Tat-Cre. Two weeks after transfer, D^bgp276-specific CD8⁺ T cells were assessed for T-bet and PD-1 expression and CFSE dilution by flow cytometry. Data in (D) to (F) are representative of two to four independent experiments.

Fig. 3

Deletion of T-bet or Eomes leads to impaired maintenance of the CD8⁺ T cell response and loss of viral control. (A) Longitudinal frequency of D^bgp33-specific CD8⁺ T cells in the blood of infected mice (PBMCs, peripheral blood mononuclear cells). (B) Total D^bgp276-specific CD8⁺ T cells in the spleens of indicated mice at d30 p.i. (**P < 0.01, ***P < 0.001; unpaired t test). (C) Viral load in the serum longitudinally and kidney at d90 p.i. (PFU, plaque-forming units). (D) Viral load in kidneys of bone marrow chimeras transplanted with indicated bone marrow at d90 p.i. (*P < 0.05,**P<0.01;Mann-Whitney). All panels used LCMV clone 13 infection. Data are representative of two to six independent experiments with at least four mice per experimental group.

Fig. 4

Persistent antigen may progressively deplete the T-bet^hi Eomes^lo CD8⁺ T precursor pool. (A) CD8⁺ T cells from infected mice were isolated at d8 p.i. and transferred to mice infected with either WT or V35A clone 13. Flow cytometry plots display T-bet, Eomes, and PD-1 expression in D^bgp33-specific CD8⁺ T cells in the presence or absence of antigen. (B) Left: Flow cytometric analysis of CFSE dilution and recovery of sorted D^bgp33-specific CD8⁺ T cells 2 weeks after transfer into WT or V35A infections. Right: Mean numbers (TSEM) of D^bgp33-specific CD8⁺ T cells from three mice per experimental group (two-way analysis of variance). Numbers followed by x denote relative difference between the indicated groups. (C) Left: Flow cytometric analysis of Eomes and PD-1 expression versus BrdU incorporation in D^bgp276-specific CD8⁺ T cells from WT mice infected with clone 13 with or without CD4⁺ T cell depletion. Right: Summary data for multiple mice (**P < 0.01, ***P < 0.001; unpaired t test). Data in (A) to (C) are representative of two or three independent experiments with at least three mice per experimental group. (D) Total T-bet^hi D^bgp33-specific CD8⁺ T cells were enumerated at the indicated time points with or without CD4⁺ T cell depletion. Data are aggregated across three independent experiments. (E) Flow cytometric analysis of T-bet and Eomes expression in intrahepatic HCV-specific CD8⁺ T cells isolated from patients with resolved or chronic infections. (F) Frequency of Eomes^hi and T-bet^hi HCV-specific CD8⁺ T cells in the liver of patients with resolved or chronic infections (*P < 0.05, ***P < 0.001; unpaired t test). In (E) and (F), a total of 3 and 10 samples were analyzed from resolved and chronic infections, respectively.