Tetramerization of STAT5 promotes autoimmune-mediated neuroinflammation

Abstract

Signal tranducer and activator of transcription 5 (STAT5) plays a critical role in mediating cellular responses following cytokine stimulation. STAT proteins critically signal via the formation of dimers, but additionally, STAT tetramers serve key biological roles, and we previously reported their importance in T and natural killer (NK) cell biology. However, the role of STAT5 tetramerization in autoimmune-mediated neuroinflammation has not been investigated. Using the STAT5 tetramer-deficient Stat5a-Stat5b N-domain double knockin (DKI) mouse strain, we report here that STAT5 tetramers promote the pathogenesis of experimental autoimmune encephalomyelitis (EAE). The mild EAE phenotype observed in DKI mice correlates with the impaired extravasation of pathogenic T-helper 17 (Th17) cells and interactions between Th17 cells and monocyte-derived cells (MDCs) in the meninges. We further demonstrate that granulocyte-macrophage colony-stimulating factor (GM-CSF)-mediated STAT5 tetramerization regulates the production of CCL17 by MDCs. Importantly, CCL17 can partially restore the pathogenicity of DKI Th17 cells, and this is dependent on the activity of the integrin VLA-4. Thus, our study reveals a GM-CSF-STAT5 tetramer-CCL17 pathway in MDCs that promotes autoimmune neuroinflammation.

Keywords: CCL17; STAT5 tetramers; Th17; autoimmunity; monocytes.

Fig. 1.

STAT5 tetramerization in hematopoietic-derived cells is required to promote the pathogenesis of EAE. (A–F) EAE was induced in WT and DKI mice by active immunization. (A) EAE disease scores. n = 15 from three experiments. (B–F) Spinal cords were harvested on day 7 or 14 post-EAE induction. (B) Representative WT and DKI hematoxylin and eosin (H&E) (Top) and Weil’s myelin staining (Bottom) on spinal cord sections harvested on day 14. Black arrows indicate areas of immune cell infiltration. Red arrows indicate areas of demyelination. (C) Pathological evaluation of inflammation and demyelination was performed using H&E-stained and Weil’s myelin-stained spinal cord sections. n = 10 from three experiments. (D) Percent of total infiltrating CD45^hi immune cells, CD4⁺ T cells (CD45^hi TCRβ⁺ CD4⁺), CD8⁺ T cells (CD45^hi TCRβ⁺ CD8⁺), monocytes (CD45^hi Ly6C⁺ CD11b⁺ CD64⁺ Ly6G⁻), and neutrophils (CD45^hi CD11b⁺ Ly6C⁺ CD64⁺ Ly6G⁺) in the spinal cord 7 and 14 d following EAE induction. (E) Percent of monocyte-derived dendritic cells (CD11c⁺ MHCII⁺) and monocyte-derived macrophages (MerTK⁺ CD24⁺) in the spinal cord 14 d following EAE induction. Representative dot plots are shown. (F) Percent of IFNγ, IL-17A, and GM-CSF–expressing CD4⁺ T cells following isolation from spinal cords on day 14. Cells were stimulated with phorbol myristate acetate (PMA) and ionomycin for 4 h. Representative dot plots are shown. Data from D–F are combined results with n = 10 to 11 from three experiments. (G) EAE disease scores from chimeric mice in which WT bone marrow cells were transferred to lethally irradiated WT or DKI recipients, followed by active EAE induction. n = 6 from two experiments. Data from A and C–G are mean $\pm \mathrm{ }$ SEM. Statistical significance is indicated by the P value with NS meaning no statistical significance. (A and G) Area under the curve with unpaired t test. (C–F) Unpaired t test.

Fig. 2.

STAT5 tetramers in both CD4⁺ T cells and antigen-presenting cells are critical for the pathogenesis of EAE. (A) EAE disease scores from WT and DKI recipient mice following transfer of WT or DKI Th17 donor cells. Data are combined results with n = 8 to 9 from three experiments. (B and C) WT or DKI CD4⁺ T cells expanded under Th17 conditions in the presence of WT or DKI APCs were transferred into WT recipient mice. (B) EAE disease scores. (C) Percent of IFNγ, IL-17A, and GM-CSF–expressing WT or DKI Th17 cells following expansion with either WT or DKI APCs. Th17 cells were stimulated with PMA and ionomycin for 4 h. Data from B and C are combined results with n = 6 to 7 from two experiments. (D) Smoothed scatterplot showing differentially expressed genes for the comparison between WT and DKI Th17 cells. Data were generated with two independent experiments; in each experiment, CD4⁺ T cells were isolated from two to three immunized mice and pooled. (E) Visualization of GSEA analysis for genes in Th17 cells based on expression changes (DKI/WT) from high to low (red to blue in the spectrum) aligned with genes up-regulated in pathogenic Th17 cells defined in ref. (vertical bars). NES, normalized enrichment score. (F) DAVID GO enrichment analysis on biological processes (BP3) for genes that were down-regulated in DKI Th17 cells compared to WT Th17 cells. Data from A–C are mean $\pm \mathrm{ }$ SEM. Statistical significance is indicated by the P value with NS meaning no statistical significance. (A and B) Area under the curve with one-way ANOVA and Tukey’s test for multiple comparisons. (C) Unpaired t test.

Fig. 3.

STAT5 tetramers regulate gene expression in MDCs. (A) Numbers of up-regulated and down-regulated genes in the bone marrow monocytes, spinal cord monocytes, and spinal cord MDCs in the comparison of DKI versus WT mice. (B) Venn diagram comparing STAT5 tetramer-dependent genes that were differentially regulated in the bone marrow monocytes, the spinal cord monocytes, and the spinal cord MDCs. (C) Smoothed scatterplot showing differentially expressed genes for the comparison of DKI versus WT spinal cord MDC cells. (D) Gene expression levels of Ccl5, Ccl6, Ccl22, Cxcl3, and Ccl17 in the WT and DKI bone marrow monocytes, spinal cord monocytes, and spinal cord MDCs on day 14 following active EAE induction. (E) DAVID GO enrichment analysis on biological processes (BP3) for genes that were down-regulated in the DKI spinal cord MDCs compared to WT MDCs. A single RNA-Seq dataset per group was generated using spinal cord cells pooled from three independent experiments with n = 18.

Fig. 4.

GM-CSF–mediated STAT5 tetramerization in MDCs regulates CCL17 production. (A) Expression of Ccl17 mRNA in WT and DKI bone marrow–enriched monocytes following GM-CSF stimulation. Shown is expression relative to Actb. Data shown are combined results from six experiments. (B) Level of CCL17 protein present in the culture supernatants of WT and DKI MDCs differentiated for 8 d with GM-CSF. Concentrations were determined using a multiplex bead array from three experiments. (C) Expression of Ccl17 mRNA in WT and DKI splenic DCs following stimulation with GM-CSF. Shown is expression relative to 18S. Data shown are combined results from four experiments. (D) Level of CCL17 protein in the culture supernatants of splenic DCs following GM-CSF stimulation. CCL17 protein concentration was determined by ELISA from three experiments. (E) STAT5B binding to the Ccl17 promoter in splenic DCs following GM-CSF stimulation was determined by ChIP-Seq. Data are from two experiments. (F) Concentration of CCL17 protein in spinal cord homogenates harvested on days 7 and 14 following EAE induction. Shown are combined results with n = 9 to 11 from three experiments. (G) Expression of Ccl17 mRNA in WT and DKI splenic DCs isolated 14 d after priming with MOG_35–55 in complete Freund's adjuvant (CFA). Bone marrow monocytes stimulated with GM-CSF served as the positive control. Expression was relative to Actb. Data shown are combined results from two experiments. (H and I) EAE induced via active immunization in Ccr2^RFP/RFP and B6 mice. (H) EAE disease scores. (I) Concentration of CCL17 protein in Ccr2^RFP/RFP and B6 spinal cord homogenates 14 to 16 d following EAE induction was determined by multiplex bead array. Data from H and I are combined results with n = 5 to 6 from three experiments. (J–M) EAE induced by adoptive transfer of WT Th17 cells into WT recipient mice. Recipient mice were intravenously administered with anti–GM-CSF antibody or an isotype control antibody on days 10 and 12 (black arrows). The spinal cords were harvested on day 14. (J) EAE disease scores. (K) Concentration of CCL17 protein in the spinal cord homogenates was determined by multiplex bead array. (L) Percent of infiltrating CD45^hi immune cells and monocytes/MDCs (CD45^hi CD64⁺ CD11b⁺ Ly6C⁺ Ly6G⁻) in the spinal cords. (M) Percent of differentiated monocytes (F4/80⁺ MHCII⁺). Representative dot plots are shown. In J–M, combined results with n = 8 from three experiments are shown. Data shown are mean $\pm \mathrm{ }$ SEM. Statistical significance is indicated by the P value with NS meaning no statistical significance. (H and J) Area under the curve with unpaired t test. (A–D, I, and K–M) Unpaired t test.

Fig. 5.

CCL17 promotes the pathogenic function of STAT5 tetramer-deficient Th17 cells. (A) Percent of CCR4-expressing lymphoid cells (CD45^hi CD11b⁻), myeloid cells (CD45^hi CD11b⁺), microglia (CD45^int CD11b⁺), and astrocytes (CD45⁻ ACSA-2⁺) in the spinal cords of WT and DKI mice 14 d following active EAE induction assessed by flow cytometric analysis. n = 9 from three experiments. (B) Percent of CCR4-expressing CD4⁺ T cells, CD8⁺ T cells, and B cells in the spinal cords of mice from adoptive transfer experiments detailed in Fig. 4 J–M. Data are combined with n = 8 from three experiments. (C and D) EAE was induced in DKI recipient mice following transfer of DKI Th17 cells expanded with or without exogenous CCL17. (C) EAE disease scores. (D) Percent of IFNγ, IL-17A, and GM-CSF–expressing CD4⁺ T cells in spinal cords on day 14 following transfer. Cells isolated from the spinal cords were stimulated with PMA and ionomycin for 4 h. Data from C and D are combined results from two experiments with n = 6. Representative dot plots are shown. (E) Number of WT Th17 cells that crossed a Transwell insert in response to medium, CCL20, or CCL17. Data are combined results from three experiments. (F) Visualization of GSEA result for genes in Th17 cells sorted based on expression changes (with CCL17/without CCL17) from high to low (red to blue in the spectrum) aligned with genes that were down-regulated in the DKI Th17 cells compared to WT Th17 cells (vertical bars). NES, normalized enrichment score. Data shown are mean $\pm \mathrm{ }$ SEM. Statistical significance is indicated by the P value with NS meaning no statistical significance. (C) Area under the curve with unpaired t test. (A, B, D, and E) Unpaired t test.

Fig. 6.

STAT5 tetramers promote MDC and CD4⁺ T cell interactions and extravasation in the meninges. (A–D) Whole mount spinal cord meninges from WT^reporter and DKI^reporter mice harvested 10 d following Th17 cell transfer. Meninges were stained with anti-CD4 antibody and DAPI. (A) Representative image from the WT meninges showing interactions between CD4⁺ T cells, CCR2⁺ monocytes, CD11c⁺ dendritic cells, and CCR2⁺CD11c⁺ MDCs within cell clusters (Top) and outside of cell clusters (Bottom) from the large stitch tiled image. (B) Representative WT and DKI volume projections without analysis. (C) Representative volume projections showing the number of CCR2⁺ monocytes and the number of CCR2⁺CD11c⁺ MDCs (yellow). Bar graphs show quantification results. (D) Representative volume projections showing the number of CD4⁺ and the number of interactions between the CCR2⁺ monocytes/MDCs and the CD4⁺ T cells (blue). Bar graphs show quantification results. Data from A–D are combined with n = 3 to 4 from two experiments with three z-stack images from each meninge. Quantification was performed using General Analysis 3 (GA3) in the NIS Elements software. (E and F) Whole mount spinal cord meninges from WT recipient mice receiving WT Th17 cells and DKI recipient mice receiving either DKI Th17 cells or DKI Th17 cells expanded with CCL17. Meninges were harvested on day 4 and stained with anti-CD31 (red) and anti-CD4 (white) antibodies. (E) Representative volume projections without analysis. (F) Representative volume projection showing the location of CD4⁺ T cells relative to CD31⁺ vessels. Yellow arrows indicate areas where cells are confined within the vessels. Blue arrows indicate where CD4⁺ T cells have extravasated. Data from E and F are combined with n = 3 WT, 6 DKI and 3 DKI + CCL17 from two experiments. Three z-stack images were taken from each meninge. (G) EAE disease scores from DKI recipient mice receiving DKI T cells, DKI T cells expanded with CCL17 and an isotype control antibody, DKI T cells expanded with CCL17 and anti–LFA-1, and DKI T cells expanded with CCL17 and anti–VLA-4 with n = 2 DKI, 6 DKI + CCL17 + isotype, 4 DKI + CCL17 + anti–LFA-1, and 6 DKI + CCL17 + anti–VLA-4 from three experiments. Data from C, D, and G are mean $\pm \mathrm{ }$ SEM. Statistical significance is indicated by the P value with NS meaning no statistical significance. (G) Area under the curve with one-way ANOVA and Tukey’s test for multiple comparisons. (C and D) Unpaired t test.