Transcriptional repressor ZEB2 promotes terminal differentiation of CD8+ effector and memory T cell populations during infection

Abstract

ZEB2 is a multi-zinc-finger transcription factor known to play a significant role in early neurogenesis and in epithelial-mesenchymal transition-dependent tumor metastasis. Although the function of ZEB2 in T lymphocytes is unknown, activity of the closely related family member ZEB1 has been implicated in lymphocyte development. Here, we find that ZEB2 expression is up-regulated by activated T cells, specifically in the KLRG1(hi) effector CD8(+) T cell subset. Loss of ZEB2 expression results in a significant loss of antigen-specific CD8(+) T cells after primary and secondary infection with a severe impairment in the generation of the KLRG1(hi) effector memory cell population. We show that ZEB2, which can bind DNA at tandem, consensus E-box sites, regulates gene expression of several E-protein targets and may directly repress Il7r and Il2 in CD8(+) T cells responding to infection. Furthermore, we find that T-bet binds to highly conserved T-box sites in the Zeb2 gene and that T-bet and ZEB2 regulate similar gene expression programs in effector T cells, suggesting that T-bet acts upstream and through regulation of ZEB2. Collectively, we place ZEB2 in a larger transcriptional network that is responsible for the balance between terminal differentiation and formation of memory CD8(+) T cells.

Figure 1.

ZEB2 deficiency results in a reduced CD8⁺ T cell response. (A) Microarray analysis using Gene 1.0 ST array was performed on OT-I T cells sorted on the indicated days after Lm-OVA infection. (B) Total P14 cells on the indicated days (left) or KLRG1^hi and KLRG1^lo P14 populations on day 14 after LCMV infection (right) were sorted, and ZEB2 expression levels were determined by qPCR. Expression is normalized to day 5 after infection or the KLRG1^lo population, respectively. Zeb2^+/+ and Zeb2^−/− T cell kinetics were monitored after LCMV infection. (C–E) The frequency of endogenous CD8⁺ tetramer⁺ cells (C) or congenically marked P14 cells from single (D) or mixed transfers (E) on the indicated days of infection is shown. In mixed transfer recipients, the percentage of P14 cells of total CD8⁺ T cells in the spleen, lymph node (LN), liver, and lung was analyzed on day 8 of infection. (F) For mixed transfers, proliferation of P14 cells in the spleen was assessed by BrdU incorporation on days 5 and 6 of infection and Annexin V staining in 7AAD⁻ P14 cells on day 5 of infection. Unstained BrdU control is displayed in gray. Numbers in histograms indicate the percentage of Annexin V⁺7AAD⁻ P14 cells of a representative mouse. Data are from three (A) or two to five (B–D) independent experiments with n = 2–6. Mean ± SEM is shown. Unpaired, two-tailed Student’s t test was performed to determine significance. *, P ≤ 0.05; ***, P < 0.001.

Figure 2.

Lack of ZEB2 leads to a selective loss of terminally differentiated shorter-lived effector CD8⁺ T cells. (A) Co-transferred Zeb2^+/+ and Zeb2^−/− P14 T cells were analyzed on the indicated days after LCMV infection for KLRG1 and CD127 expression. Numbers indicate the percentage of the population in total P14 cells from a representative mouse. (B) Quantification of the frequency and total numbers of P14 populations represented in A on day 8 after LCMV infection. (C) Kinetics of the total KLRG1^loCD127^hiCD8⁺ T cell population over the LCMV infection. (D) Expression of surface markers on Zeb2^+/+ and Zeb2^−/− P14 cells on the indicated days after LCMV infection. (E) Intracellular cytokine staining after gp33 peptide restimulation on day 10 of LCMV infection. Numbers indicate the frequency for each quadrant from a representative mouse. (F) Quantification of IFNγ⁺ and IFNγ⁺IL2⁺ P14 cells. Data are representative of two to three independent experiments with n = 2–3. Mean ± SEM is shown. Unpaired, two-tailed Student’s t test was performed to determine significance. **, P < 0.01; ***, P < 0.001.

Figure 3.

ZEB2 is necessary for the down-regulation of CD127 during a secondary response to pathogen. Endogenous CD8⁺tetramer⁺ cells in the PBL were analyzed in mice primarily infected with VSV-OVA, followed 30 d later with Lm-OVA (left). Co-transferred Zeb2^+/+ and Zeb2^−/− P14 T cells in the PBL were assessed in mice primarily infected with Lm-gp33, followed 30 d later with LCMV (right). (A) Kinetic analysis of antigen-specific response to secondary infection. (B) KLRG1 and CD127 expression at the peak of infection after rechallenge. Numbers are the percentage of the population in total antigen-specific CD8⁺ T cells from a representative mouse. (C) Quantification of the frequency of antigen-specific CD8⁺ T cell populations represented in B. Data are representative of two independent experiments with n = 2–4. Mean ± SEM is shown. Unpaired, two-tailed Student’s t test was performed to determine significance. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 4.

Microarray analysis of ZEB2-deficient CD8⁺ T cells. (A) Volcano plot comparing differential gene expression between ZEB2-deficient and WT P14 cells. (B) Overlay of differential gene expression with ImmGen clusters II (orange), III (yellow), IV (green), and VII (blue; Best et al., 2013). (C) ChIP analysis performed with a ZEB2 antibody on OT-I cells on day 8 of Lm-OVA infection at E-boxes in the promoter and first intron of Il7r or in the proximal promoter and 2 kb upstream of Il2. Numbers in the bottom corners indicate the number of genes in that region. Data are from two independent experiments with three (A and B) or five (C) mice per group.

Figure 5.

T-bet is upstream of ZEB2 in terminally differentiating CD8⁺ T cells. (A) T-bet expression in ZEB2-deficient and -sufficient P14 cells at day 8 of LCMV infection (top) and in CD44^hiCD8⁺ T cells with (Zeb2^Tg/+) or without (Zeb2^+/+) constitutive ZEB2 expression (bottom). (B) T-bet and ZEB2 expression in Tbx21^+/+ and Tbx21^+/− OT-I T cells isolated at day 8 after Lm-OVA infection, as assessed by qPCR. (C) ChIP analysis was performed with a T-bet antibody on bulk or KLRG1^hiCD8⁺ T cells isolated on day 8 of LCMV infection at the indicated sites 5′ to the Zeb2 coding sequence and in the 3′ UTR. Tbx21^−/− splenocytes were used as a negative control. (D) Comparison of gene expression in Tbx21^−/− versus Tbx21^+/+ CD8⁺ OT-I T cells at day 6 after Lm-OVA infection, plotted against that for Zeb2^−/− versus WT CD8⁺ P14 T cells at day 6 after LCMV infection. Gray lines indicate 1.5-fold cut-off. (E) Volcano plots comparing differential gene expression between ZEB2-deficient and WT P14 cells overlaid with those genes highly up-regulated (>1.5 fold) in Tbx21^−/− (top left) or Tbx21^+/+ (top right) OT-I cells or comparing differential gene expression between Tbx21^−/− and Tbx21^+/+ OT-I cells overlaid with those highly up-regulated (>1.5 fold) in Zeb2^−/− (bottom left) or Zeb2^+/+ (bottom right). Data are from two (A and C) or four (B) independent experiments, with n = 2–5. Microarray analysis represents two to three independent replicates with n = 3.

Figure 6.

Heterozygotic mutation of ZEB2 in MWS patients does not affect T cell populations of the blood. (A) Zeb2^+/+, Zeb2^+/−, and Zeb2^−/− T cell kinetics were monitored after VSV-OVA infection. The frequency of endogenous CD8⁺tetramer⁺ cells in the CD8⁺ PBL population is shown. (B) Quantification of the frequency of antigen-specific KLRG1 and CD127 CD8⁺ T cell populations on day 8 at the peak of infection. Data are representative of five independent replicates with n = 3–6. (C) CD4⁺ T cell, CD8⁺ T cell, and CD19⁺ B cell populations in the blood of MWS patients and WT controls. (D) Quantification of the frequency of populations represented in C. (E) KLRG1 and CD127 expression on CD8⁺ T cell populations of MWS patients and controls. (F) Analysis of naive (CCR7⁺CD45RA⁺), central memory (CCR7⁺CD45RA⁻), effector memory (CCR7⁻CD45RA⁻), and CD45RA⁺ effector memory (CCR7⁻CD45RA⁺) CD8⁺ T cells in the blood of MWS patients and controls. Numbers in contour plots are the percentage of the indicated population of a representative sample. Data are pooled from two independent experiments with n = 5–6. Mean ± SEM is shown. Unpaired, two-tailed Student’s t test was performed to determine significance. *, P < 0.05; **, P < 0.01; ***, P < 0.001.