Requirement for T-bet in the aberrant differentiation of unhelped memory CD8+ T cells

Abstract

Immunity to intracellular pathogens requires dynamic balance between terminal differentiation of short-lived, cytotoxic effector CD8+ T cells and self-renewal of central-memory CD8+ T cells. We now show that T-bet represses transcription of IL-7Ralpha and drives differentiation of effector and effector-memory CD8+ T cells at the expense of central-memory cells. We also found T-bet to be overexpressed in CD8+ T cells that differentiated in the absence of CD4+ T cell help, a condition that is associated with defective central-memory formation. Finally, deletion of T-bet corrected the abnormal phenotypic and functional properties of "unhelped" memory CD8+ T cells. T-bet, thus, appears to function as a molecular switch between central- and effector-memory cell differentiation. Antagonism of T-bet may, therefore, represent a novel strategy to offset dysfunctional programming of memory CD8+ T cells.

Figure 1.

T-bet–mediated repression of IL-7Rα in effector CD8⁺ T cells. (A) IL-7Rα expression on CD8⁺ T cells from blood of wild-type or Tbx21 ^−/− mice 8 d after LCMV infection. Top row shows CD8⁺ T cells; middle and bottom rows show H-2D^bGP33⁺ or H-2D^bNP396⁺ T cells, respectively; the percentage of events are indicated. (B) Quantitative RT-PCR (Q-PCR) of IL-7Rα mRNA from P14 CD8⁺ (left) or DO11.10 CD4⁺ (right) T cells stimulated with peptide plus APCs and transduced with retrovirus. Cells sorted 5 d after transduction based on GFP. (C) IL-7Rα expression on DO11.10 CD4⁺ T cells in T_H2 conditions stimulated and transduced as in B. Cells stained 5 d after transduction. (D) Q-PCR of P14 CD8⁺ T cells from spleens 8 d after infection sorted for IL-7Rα^hi or IL-7Rα^lo expression. Naive (CD44^lo) CD8⁺ T cells are included in some graphs for reference. (E and F) Q-PCR of T-bet, Eomes, or Blimp-1 mRNA in LCMV-specific CD8⁺ T cells. H-2D^bGP33⁺ plus H-2D^bNP396⁺ CD8⁺ T cells sorted from spleens of wild-type mice at the indicated day after infection. Unfractionated tetramer-positive cells (E) or tetramer-positive cells fractionated by CD62L^hi or CD62L^lo expression (F). Values for Q-PCR represent the mean ± the SEM of triplicate determinations, normalized to HPRT. All results are representative of at least three experiments.

Figure 2.

Enhanced generation of central–memory CD8⁺ T cells in T-bet–deficient mice. (A–C) Phenotype of GP33-specific CD8⁺ T cells from spleens of wild-type or Tbx21 ^−/− mice 8 (A), 30 (B), or 140 d (C) after LCMV infection. Plots display H-2D^bGP33⁺ events; the percentage of events are indicated. For bottom row, splenocytes were stimulated with GP_33-41 peptide; numbers indicate the percentage producing both IFN-γ and IL-2. Results are representative of three experiments with multiple mice per time point. (D and E) Q-PCR of H-2D^bGP33⁺ plus H-2^bNP396⁺ CD8⁺ T cells from spleens of wild-type or Tbx21 ^−/− mice. (D) CCR7 and Blimp-1 mRNA 60 d after infection. Naive represents CD44^loCD8⁺ T cells from uninfected wild-type mice. (E) Eomes mRNA 60 and 100 d after infection. Day 0 represents CD44^loCD8⁺ T cells from uninfected wild-type or Tbx21 ^−/− mice. Results are representative of three similar experiments. (F) Quantification of LCMV-specific CD8⁺ T cells from wild-type or Tbx21 ^−/− mice 8, 30, and 60 d after infection. Left graph shows H-2D^bGP33⁺ cells as the percentage of CD8⁺ in blood. Middle and right graphs show numbers of CD8⁺ T cells from spleen or lymph node (pooled axillary, inguinal, cervical, paraaortic, and mesenteric), respectively, producing IFN-γ in response to GP_33-41. Results represent the mean ± the SEM for at least three mice per data point.

Figure 3.

Enhanced secondary expansion and protection of T-bet–deficient memory cells. Equal numbers (2 × 10⁵) of wild-type or Tbx21 ^−/− GP33-specific CD8⁺ T cells (Thy1.2⁺) isolated 45 d after LCMV infection and transferred to naive wild-type recipients (Thy1.1⁺). 1 d after transfer, recipient mice were infected with 2.5 × 10⁵ CFU of L. monocytogenes. 4 d after challenge, spleens and livers were harvested to assess CD8⁺ T cell expansion and perform quantitative bacterial cultures. (A) Wild-type or Tbx21 ^−/− GP33-specific CD8⁺ T cell expansion. Graphs display numbers of transferred cells (Thy1.2⁺) from spleens of recipient mice producing IFN-γ or IL-2 in response to GP_33-41. (B) Cytokine production by wild-type or Tbx21 ^−/− GP33-specific CD8⁺ T cells from spleens of recipient mice. Plots display CD8⁺ events; numbers indicate the percentage of cells producing both IFN-γ and IL-2 in response to GP_33-41. (C) Bacterial load in mice that received wild-type or Tbx21 ^−/− GP33-specific CD8⁺ T cells. Spleens and livers were homogenized in 1% Triton X-100. Bacteria quantified by limiting dilution culture. Data for A and C represent the mean ± the SEM of six recipients of wild-type and four recipients of Tbx21 ^−/− cells. Results are representative of three similar experiments.

Figure 4.

Enhanced T-bet expression in unhelped memory CD8⁺ T cells. (A) Phenotype of P14 cells from wild-type or Cd4 ^−/− hosts 60 d after infection. For bottom row, splenocytes were stimulated with GP_33-41. Plots show CD8⁺ events; numbers indicate the percentage of P14 cells (Thy1.1⁺) expressing indicated surface marker or cytokine. (B and C) Intranuclear T-bet staining of P14 cells from wild-type or Cd4 ^−/− hosts 75 d after infection. (B) Plots show CD8⁺ events; numbers indicate MFI of T-bet staining in bulk P14 population (Thy1.1⁺). (C) Plots show P14 cells (Thy1.1⁺) from Cd4 ^−/− hosts; numbers represent MFI of T-bet staining in subsets with high or low expression of indicated marker. (D) Q-PCR of T-bet and Eomes mRNA in P14 cells from wild-type or Cd4 ^−/− hosts 60 d after infection. (E) Q-PCR of CCR7 and Blimp1 mRNA in P14 cells from wild-type or Cd4 ^−/− hosts 31, 46, or 60 d after infection. Values for Q-PCR represent the mean ± the SEM of triplicate determinations, normalized to HPRT. All results are representative of at least three experiments.

Figure 5.

T-bet–dependent dysfunction of unhelped memory CD8⁺ T cells. (A and B) Wild-type or Tbx21 ^−/− mice were left untreated (No Tx) or treated with CD4-depleting antibody (0.2 mg GK1.5 i.p.; CD4 depleted) 1 d before and 1 d after LCMV infection. (A) Surface phenotype of GP33-specific CD8⁺ T cells from spleens 90 d after infection. Plots show H-2D^bGP33⁺ events; numbers represent the percentage of tetramer-positive cells expressing indicated surface marker. (B) Cytokine production of GP33-specific CD8⁺ T cells. Plots show CD8⁺ events; numbers indicate the percentage of cells producing both IFN-γ and IL-2 or CD40L in response to GP_33-41. Results are representative of two independent experiments.