GATA3 induces the pathogenicity of Th17 cells via regulating GM-CSF expression

Abstract

T-bet-expressing Th17 (T-bet⁺RORγt⁺) cells are associated with the induction of pathology during experimental autoimmune encephalomyelitis (EAE) and the encephalitic nature of these Th17 cells can be explained by their ability to produce GM-CSF. However, the upstream regulatory mechanisms that control Csf2 (gene encoding GM-CSF) expression are still unclear. In this study, we found that Th17 cells dynamically expressed GATA3, the master transcription factor for Th2 cell differentiation, during their differentiation both in vitro and in vivo. Early deletion of Gata3 in three complimentary conditional knockout models by Cre-ERT2, hCd2 ^Cre and Tbx21 ^Cre, respectively, limited the pathogenicity of Th17 cells during EAE, which was correlated with a defect in generating pathogenic T-bet-expressing Th17 cells. These results indicate that early GATA3-dependent gene regulation is critically required to generate a de novo encephalitogenic Th17 response. Furthermore, a late deletion of Gata3 via Cre-ERT2 in the adoptive transfer EAE model resulted in a cell intrinsic failure to induce EAE symptoms which was correlated with a substantial reduction in GM-CSF production without affecting the generation and/or maintenance of T-bet-expressing Th17 cells. RNA-Seq analysis of Gata3-sufficient and Gata3-deficient CNS-infiltrating CD4⁺ effector T cells from mixed congenic co-transfer recipient mice revealed an important, cell-intrinsic, function of GATA3 in regulating the expression of Egr2, Bhlhe40, and Csf2. Thus, our data highlights a novel role for GATA3 in promoting and maintaining the pathogenicity of T-bet-expressing Th17 cells in EAE, via putative regulation of Egr2, Bhlhe40, and GM-CSF expression.

Keywords: Bhlhe40; GATA3; GM-CSF; Th17; experimental autoimmune encephalomyelitis; pathogenicity.

Figure 1

GATA3 is transiently induced during Th17 cell differentiation. (A, B) Naïve C57BL/6 CD4 T cells cultured under Th17 polarization conditions and monitored for GATA3 expression at 0, 24, 48, and 72 hr. (A) Representative FACS plots at the indicated timepoints depicting GATA3 and RORγt staining amongst CD4⁺CD44⁺Foxp3^- T cells and representative GATA3 histograms within CD4⁺CD44⁺Foxp3^-RORγt⁺ T cells. (B) Mean % of GATA3-expressing (i.e., GATA3⁺RORγt⁺ Th17) cells among CD4⁺CD44⁺Foxp3^-RORγt⁺ populations in (A) from three independent experiments. (C, D) Representative RORγt and GATA3 reporter expression by Rorc ^E2-Crimson Gata3 ^ZsGreen Foxp3 ^RFP CD4 T cells cultured under Th17 polarization conditions as in (A). (D) Mean % of GATA3-expressing (i.e., GATA3⁺RORγt⁺ Th17) cells among CD4⁺CD44⁺Foxp3^-RORγt⁺ populations in (C) from two independent experiments. (E–H) The in vivo kinetics of GATA3 expression within draining lymph node (dLN) Th17 cells in response to MOG_35-55/CFA immunization. (E, F) Representative GATA3 and RORγt staining (E) within CD4⁺CD44⁺Foxp3^- dLN T cells and the mean % of GATA3-expressing cells among CD4⁺CD44⁺Foxp3^-RORγt⁺ T cells (F) from immunized C57BL/6 mice (n=5-8 mice/timepoint from three independent experiments). (G, H) Representative GATA3 and RORγt staining (G) within donor 2D2 CD4⁺CD44⁺Foxp3^- dLN T cells and the mean % of GATA3-expressing cells among 2D2 CD4⁺CD44⁺Foxp3^-RORγt⁺ T cells (H) from immunized 2D2 naïve T cell recipient Cd45.1/Cd45.2 mice (n=4 mice/timepoint from two independent experiments). UI – unimmunized, PI – post immunization. For statistical comparisons, a one-way ANOVA was conducted with Tukey Post-Hoc testing for group comparisons. Significance levels are denoted as follows: **p <0.01; ***p <0.001.

Figure 2

Early GATA3 expression is required to generate a de novo T-bet⁺ Th17 cell response and EAE symptoms. (A) Mean EAE clinical scores from Tcra^-/- hosts that received 4x10⁶ CD4⁺CD44⁺ T cells harvested from the draining lymph nodes (dLNs) of Cre-ERT2-Gata3 ^fl/fl mice 6 days after MOG/CFA immunization with vehicle or tamoxifen treatment (n=12 mice/group from three independent experiments). On the day of the cell transfer procedure, the Tcra^-/- host mice were boosted with MOG_35-55/IFA and Pertussis Toxin as described in the methods. (B, C) Representative T-bet and RORγt staining amongst d0 Vehicle or Tamoxifen-treated Cre-ERT2-Gata3 ^fl/fl d6 dLN CD3⁺CD4⁺CD44⁺Foxp3^- T cells (B), and the mean percentages of T-bet and RORγt expressing T cell subsets (C). (D, E) Representative IFNγ and IL-17A staining amongst Vehicle or Tamoxifen treated Cre-ERT2-Gata3 ^fl/fl d6 dLN CD3⁺CD4⁺CD44⁺Foxp3^- T cells (D), and the mean percentages of IFNγ and IL-17A expressing T cell subsets (E, n =5 mice/group from two independent experiments). (F) Mean EAE clinical scores from MOG_35-55/CFA immunized Gata3 ^fl/fl and hCd2 ^Cre Gata3 ^fl/fl mice. n=12 mice/group from three independent experiments. (G, H) Representative T-bet and RORγt staining amongst Gata3 ^fl/fl and hCd2 ^Cre Gata3 ^fl/fl d6 post-immunization dLN CD3⁺CD4⁺CD44⁺Foxp3^- T cells (G), and the mean percentages of T-bet and RORγt expressing T cell subsets (H). (I, J) Representative IFNγ and IL-17A staining amongst Gata3 ^fl/fl and hCd2 ^Cre Gata3 ^fl/fl d6 dLN CD3⁺CD4⁺CD44⁺Foxp3^- T cells (I), and the mean percentages of IFNγ and IL-17A expressing T cell subsets (J). n=6 mice/group from two independent experiments. For statistical comparisons, unpaired student’s T tests were used. Significance levels are denoted as follows: *p <0.05; **p <0.01; ***p <0.001.

Figure 3

GATA3 expression is required for the development and/or maintenance of early T-bet⁺ Th17 cells. (A) Mean EAE clinical scores from MOG_35-55/CFA immunized Gata3 ^fl/fl and Tbx21 ^Cre Gata3 ^fl/fl mice. n=10 mice/group from three independent experiments. (B) Mean cell transfer EAE clinical scores from Tcra^-/- mice that received d6 dLN Gata3 ^fl/fl or Tbx21 ^Cre Gata3 ^fl/fl donor CD4⁺CD44^hi T cells and immunized with MOG_35-55/IFA. n=10 mice/group from three independent experiments. (C, D) Representative T-bet and RORγt staining amongst Gata3 ^fl/fl and Tbx21 ^Cre Gata3 ^fl/fl d6 post-immunization dLN CD3⁺CD4⁺CD44^hiFoxp3^- T cells (C). The mean percentages of T-bet- and RORγt-expressing CD4 T cell subsets within the d6 dLN (D). (E, F) Representative IFNγ and IL-17A staining amongst Gata3 ^fl/fl and Tbx21 ^Cre Gata3 ^fl/fl d6 post-immunization dLN CD3⁺CD4⁺CD44^hiFoxp3^- T cells (E). The mean percentages of IFNγ and IL-17A positive dLN CD3⁺CD4⁺CD44^hiFoxp3^- T cell subsets. n=6 mice/group from two independent experiments. For statistical comparisons, unpaired student’s T tests were used. Significance levels are denoted as follows: *p <0.05; **p <0.01; ***p <0.001.

Figure 4

GATA3 is required for GM-CSF production and CD4 T cell-mediated encephalomyelitis. (A, B) Characterization of the pre-transfer d6 dLN CD3⁺CD4⁺CD44^hi Cd45.1/Cd45.2 and Cre-ERT2-Gata3 ^fl/fl populations. (A) Representative IFNγ vs IL-17A staining and (B) the mean percentages of IFNγ and IL-17A CD4⁺CD44^hiFoxp3^- T cell subpopulations in the pre-transfer isolates. (C) Mean transfer EAE clinical scores from Tcra^-/- hosts that received 4x10⁶ Cre-ERT2-Gata3 ^fl/fl d6 dLN CD4⁺CD44^hi T cells (A, B) and a vehicle or tamoxifen treatment. n=15 mice/group from four independent experiments. (D-G; H-K) Characterizations of the Cre-ERT2-Gata3 ^fl/fl donor CD4 T cells within vehicle or tamoxifen treated Tcra^-/- hosts post-transfer. Representative IFNγ and IL-17A staining of vehicle or tamoxifen treated Cre-ERT2-Gata3 ^fl/fl donor CD4 T cells in the Tcra ^-/- host CNS (D) and the spleen (F). The mean percentages of IFNγ and IL-17A subpopulations within the CNS (E) and spleen (G). Representative GM-CSF and IL-17A staining of vehicle or tamoxifen treated Cre-ERT2-Gata3 ^fl/fl donor CD4 T cells within the Tcra ^-/- host CNS (H) and the spleen (J). The mean percentages of IL-17A and GM-CSF positive donor CD4 T effector cells within the CNS (I) and the spleen (K). n=12 recipient mice/group from four independent experiments. For statistical comparisons, unpaired student’s T tests were used. Significance levels are denoted as follows: *p <0.05; **p <0.01; ***p <0.001.

Figure 5

The post-differentiation effects of GATA3 on GM-CSF production is cell intrinsic. Mixed co-transfers of CD45.1⁺CD45.2⁺ C57BL/6 and CD45.2⁺ Cre-ERT2-Gata3 ^fl/fl d6 dLN CD4⁺ T effector cells were conducted. (A, B) The starting ratios of FACS-sorted donor Cd45.1/Cd45.2 C57BL/6 and Cd45.2 Cre-ERT2-Gata3 ^fl/fl d6 dLN CD4 T effector cells shown as a representative FACS plot (A) and population means (B). (C-E) The post-transfer ratios of donor CD45.1⁺CD45.2⁺ C57BL/6 and CD45.2⁺ Cre-ERT2-Gata3 ^fl/fl cells within host Tcra ^-/- mice treated with corn oil (vehicle) or tamoxifen. (C) Representative CD45.2 and CD45.1 staining within the CNS. The mean percentages of corn oil or tamoxifen treated CD45.1⁺CD45.2⁺ C57BL/6 and CD45.2⁺ Cre-ERT2-Gata3 ^fl/fl CD4 T cells within the CNS (D) and spleen (E). (F, G) Representative post-transfer corn oil or tamoxifen treated CD45.1/CD45.2 and Cre-ERT2-Gata3 ^fl/fl donor CD4 T effector IFNγ and IL-17A staining (F) and subpopulation means (G) within the CNS of Tcra ^-/- recipient mice. (H, I) Representative post-transfer donor Cd45.1/Cd45.2 and Cre-ERT2-Gata3 ^fl/fl donor CD4 T effector GM-CSF and IL-17A staining (H) and subpopulation means (I) within the CNS of Tcra ^-/- recipient mice. n=12 mice/group from three independent experiments. For statistical comparisons, unpaired student’s T tests were used. Significance levels are denoted as follows: *p <0.05; **p <0.01; ***p <0.001.

Figure 6

GATA3 regulates pro-inflammatory gene expression in EAE. RNA-Seq analysis was performed using CNS-infiltrating CD4⁺CD44^hiCD25^-CD45.1⁺CD45.2⁺ C57BL/6 and CD45.1^-CD45.2⁺ Cre-ERT2-Gata3 ^fl/fl donor T effector cells from tamoxifen-treated Tcra^-/- co-transfer EAE mice at the peak. Differentially expressed genes were identified using Partek Flow Genomic Suite (Partek) and curated based on an FDR threshold of <0.05. Differentially expressed genes were clustered and displayed as a heatmap. The results are representative of biological duplicates.

Figure 7

GATA3 is required for normal Bhlhe40 and Egr2 expression at all stages. (A–D) The expression of Egr2 and Bhlhe40 within Gata3-sufficient (Cd45.1/Cd45.2) and Gata3-deficient (Cre-ERT2-Gata3 ^fl/fl) CNS-infiltrating CD3⁺CD4⁺CD44^hiFoxp3^- T effector cells from co-transfer EAE Tcra^-/- recipient mice treated with tamoxifen (cell transfer d0). (A) Representative Egr2 and RORγt staining and summary statistics (B) amongst tamoxifen treated CNS-infiltrating CD4⁺CD44^hiFoxp3^- Cd45.1/Cd45.2 and Cre-ERT2-Gata3 ^fl/fl cells. (C) Representative Bhlhe40 and RORγt staining and summary statistics (D) amongst tamoxifen treated CD4⁺CD44^hiFoxp3^- Cd45.1/Cd45.2 and Cre-ERT2-Gata3 ^fl/fl cells. n=4 mice/condition from two independent experiments. (E–H) The expression of Egr2 and Bhlhe40 within Gata3-sufficient (Vehicle, d0) or Gata3-deficient (Tamoxifen, d0) day 6 dLN CD3⁺CD4⁺CD44^hiFoxp3^- T effector cells from MOG_35-55/CFA-immunized Cre-ERT2-Gata3 ^fl/fl mice. (E) Representative Egr2 and RORγt staining and the corresponding summary statistics (F) from vehicle control or tamoxifen treated d6 dLN Cre-ERT2-Gata3 ^fl/fl CD4 T effector cells. (G) Representative d6 dLN CD4 T effector Bhlhe40 and RORγt staining and the corresponding summary statistics (H) from immunized and vehicle or tamoxifen treated Cre-ERT2-Gata3 ^fl/fl mice. n=4 mice/condition from two independent experiments. For statistical comparisons, unpaired student’s T tests were used. Significance levels are denoted as follows: *p <0.05; **p <0.01; ***p <0.001.

Figure 8

A model of regulatory network involving GATA3, Egr2, Bhlhe40 and Csf2. The top differentially expressed genes between CD4⁺CD44^hiCD25^-CD45.1⁺CD45.2⁺ C57BL/6 and CD45.1^-CD45.2⁺ Cre-ERT2-Gata3 ^fl/fl donor T effector cells in Figure 6 were used in an ingenuity pathway analysis to visualize regulatory connections between GATA3, Egr2, Bhlhe40 and Csf2. Downregulated genes from the dataset were overlaid (green).