Modular utilization of distal cis-regulatory elements controls Ifng gene expression in T cells activated by distinct stimuli

Abstract

Distal cis-regulatory elements play essential roles in the T lineage-specific expression of cytokine genes. We have mapped interactions of three trans-acting factors-NF-kappaB, STAT4, and T-bet-with cis elements in the Ifng locus. We find that RelA is critical for optimal Ifng expression and is differentially recruited to multiple elements contingent upon T cell receptor (TCR) or interleukin-12 (IL-12) plus IL-18 signaling. RelA recruitment to at least four elements is dependent on T-bet-dependent remodeling of the Ifng locus and corecruitment of STAT4. STAT4 and NF-kappaB therefore cooperate at multiple cis elements to enable NF-kappaB-dependent enhancement of Ifng expression. RelA recruitment to distal elements was similar in T helper 1 (Th1) and effector CD8(+) T (Tc1) cells, although T-bet was dispensable in CD8 effectors. These results support a model of Ifng regulation in which distal cis-regulatory elements differentially recruit key transcription factors in a modular fashion to initiate gene transcription induced by distinct activation signals.

Figure 1. DNase I hypersensitivity map of extended Ifng locus in naïve, Th1 and Th2 cells

DNase I HS profiles of T helper subsets aligned with a VISTA plot of syntenic regions of the Ifng and IFNG gene loci. DNase–chip analyses were performed on CD4⁺ T cells from OT-II transgenic mice that were naïve, differentiated under Th1 polarizing conditions (5 d), or differentiated under Th2 polarizing conditions (14 d). Prior to processing, cells were either left unstimulated (“U”), or were restimulated with anti-CD3 plus anti-CD28 (“TCR”) or rIL-12 plus rIL-18 (“IL-12+IL-18”). P, Ifng promoter; CTCF, CCCTC-binding factor site.

Figure 2. RelA is essential for acute induction of Ifng gene expression

(A) CD4⁺ and CD8⁺ T cells from Rela ^fl/+.Cd4-cre⁻ or Rela ^fl/fl.Cd4-cre⁺ mice were isolated and labeled with CFSE. CD4⁺ T cells were cultured under Th1 cell conditions for 5 d (upper panel), and CD8⁺ T cells were cultured under Tc1 condition for 3 d (lower panel). Recovered T cells were reactivated as indicated and intracellular IFN-γ was assessed by flow cytometry. Plots were gated on live CD4⁺ or CD8⁺ T cells. Numbers in each plot indicate the percentage of the gated cells in each quadrant. (B) Th1 cells were incubated for 1 hour with indicated concentrations of the RelA inhibitory peptide, (Ser 276), then activated with anti-CD3 plus anti-CD28 for 4 h and analyzed for intracellular IFN-γ. Dot plots are gated on live CD4⁺ T cells, and numbers indicate the percentage of viable CD4⁺ T cells in the indicated subgates. (C) Th1 cells were incubated for 1 hour with 150 μM of the inhibitory RelA peptide (lower panel) or control peptide (upper panel), then activated with the indicated stimuli for 4 h and analyzed for intracellular cytokines. Plots are gated on live CD4⁺ T cells; numbers indicate the percentage of viable CD4⁺ T cells in the indicated subgates. (D) Comparison of mean fluorescence intensities (MFI) of IFN-γ⁺ cells from (C). (E) Th1 cells were activated with IL-12 with or without IL-18, following 1 h pre-incubation with RelA inhibitory peptide or the indicated controls, and nuclear translocation of RelA was assesed by immunoblotting. Cyclophilin B was used as loading control. All data are representative of at least two independent experiments.

Figure 3. Differential recruitment of NF-κB p65 to distinct Ifng CNSs

Th1 cells that were either left unstimulated (“resting”), stimulated for 4 hours with 50ng/ml PMA plus 750 ng/ml Ionomycin or with anti-CD3 plus anti-CD28 (A), or restimulated with 10 ng/ml rIL-12 plus 25 ng/ml rIL-18 (B) were processed for ChIP using an NF-κB p65-specific antibody, and analysis of immunoprecipitated DNA was performed using real-time PCR with primer sets designed to detect DNA at the indicated CNS elements (-55, -34, -22, -5, +18–20, +29, +40, +46 and +55) and the Ifng promoter (Pro). The Il2 promoter (Il2 Pro) and 16S ribosomal protein (16Srp) promoter were used as positive and negative controls, respectively. Values were normalized against input DNA and expressed relative to resting Th1 cells, which was <0.05% of input DNA. Data represent Mean ± SEM from at least three independent experiments. Statistical significance is relative to binding at 16Srp under the same conditions (* p<0.01).

Figure 4. Co-recruitment of STAT4 and RelA at multiple cis elements that drive Ifng gene transcription

Th1 cells were either left unstimulated or stimulated as indicated for 4 h and ChIP was performed for the indicated CNS sites using antibodies specific for NF-κB p65 (A) or STAT4 (B). Relative RelA binding (A) was calculated as described in Fig. 3. Data are representative of at least three independent experiments (* p<0.01 for Th1 samples reactivated with IL-12+IL-18 versus IL-18 alone). (B) Recruitment of STAT4 was calculated as percentage of input DNA. Data represent mean ± SEM from at least three independent experiments. (# p<0.05 versus STAT4 recruitment in resting Th1 samples; ** p<0.01 versus recruitment of STAT4 to 16Srp in resting Th1 cells).

Figure 5. RelA recruitment is dependent on T-bet in Th1 cells but not in Tc1 cells

(A),(C) Cytokine production by Th1 and Tc1 cells generated from WT or Tbx21^−/− mice was evaluated by intracellular cytokine staining. Flow cyometry plots are gated on live CD4⁺ (A) or CD8⁺ (B) T cells. Percentage values shown indicate frequencies of cells positive for the indicated cytokines. (B),(D) ChIP was performed for NF-κB p65 and enrichment was calculated as described in Fig. 3 following restimulation with PMA and ionomycin (top panel) or IL-12 plus IL-18 (bottom panel). Data are representative of at least three independent experiments. T cells from Tbx21^−/− or WT mice were polarized under Th1 (B) or Tc1 (D) conditions and restimulated prior to ChIP. * P < 0.01, # P<0.05 versus Tbx21^−/− samples.

Figure 6. Remodeling of the Ifng locus in Th1 cells is T-bet–dependent

ChIP was performed on the indicated T cell effectors with antibodies specific for mono,di,tri-methylated H3K4 (top panels) or trimethylated H3K27 (bottom panels). CD4⁺ T cells (A) or CD8⁺ T cells (B) from wildtype or Tbx21^−/− mice were activated under Th1- or Tc1 polarizing conditions respectively, then subject to ChIP. * P < 0.01, # P<0.05 versus Tbx21^−/− Th1 samples. Values were normalized against the 16S ribosomal protein promoter, which was assigned a value of 1. Data are representative of at least three independent experiments.

Figure 7. Impaired IL-12 and IL-18 signaling do not account for diminished RelA and STAT4 recruitment to the Ifng locus in T-bet–deficient Th1 cells

(A),(B) mRNA transcripts of the indicated genes were quantitated in Th1- and Tc1-polarized cells generated from WT or Tbx21^−/− mice. Expression was normalized against β2-microglobulin, and relative expression compared to wildtype controls (dashed lines; assigned value of “1.00”). Data represent mean ± SEM from three independent experiments. (C) Th1-polarized cells derived from WT or Tbx21^−/− mice were restimulated with IL-12 + IL-18 for 4 h and nuclear translocation of p65 evaluated by immunoblotting, using cyclophilin B as a loading control. Blots were re-probed for T-bet as a control for T-bet expression. (D) RelA ChIP was performed on indicated T cell populations and relative recruitment at the Il2 promoter was assessed as in Fig 3. Data represent mean ± SEM from at least three independent experiments. (E) Th1-polarized cells from WT, Tbx21^−/− or Stat4^−/− mice were left untreated (resting) or reactivated (IL-12) and phosphorylation of tyrosine residue 693 (pY693) of STAT4 was evaluated by flow cytometry. Histograms are gated on viable CD4⁺ T cells. (F) STAT4 ChIP was performed following IL-12+IL-18 activation of Tbx21^−/− and WT Th1-polarized cells, as in Fig. 4. Data are represented as fractions of input DNA and represent mean ± SEM from at least three independent experiments (*p<0.01, #P<0.05; stimulated WT versus stimulated Tbx21^−/−).