Figures and data
H3K4me3 is required for FOXP3-dependent gene activation in Treg cells.
(A)Venn diagram showing overlap of H3K4me3 enriched regions (this study) and FOXP3 binding sites (Konopacki et al.2019) in sorted CD4+YFP+ Treg cells (left). Genomic distribution of overlapped peaks(right). Note that the overlapped peaks are predominantly enriched at promoters.
(B)Venn diagram showing overlap of H3K27me3 enriched regions (Wei et al.2009) and FOXP3 binding sites in Treg cells (left). Genomic distribution of overlapped peaks (right). Note that the overlapped peaks are predominantly enriched at promoters.
(C)Heatmap showing enrichment of H3K27me3, H3K4me3, and FOXP3 surrounding the TSS. Unsupervised K-means clustering was conducted on H3K27me3 and H3K4me3 signals.
(D)Heatmap showing gene expression levels in Treg cells (RNA-seq data was obtained from Oh et al.2017) The clusters were consistent as in Fig. 1C.
(E)Gene Ontology (GO) pathway analysis of different clusters.
CXXC1 interacts with FOXP3 in Treg cell.
(A)Average CXXC1 CUT&Tag signals around promoters in Treg cells. IgG was used as the control.
(B)Venn diagrams showing the overlap of FOXP3 binding genes, CXXC1 binding genes, and H3K4me3 enriched genes in Treg cells. Gene promoters covered by FOXP3 binding sites, CXXC1 binding sites, or exhibited high H3K4me3 levels were defined as FOXP3-bound genes, CXXC1-bound genes, or H3K4me3-enriched genes.
(C)Heatmaps showing FOXP3 ChIP-seq and CXXC1 CUT&Tag signals at indicated regions.
(D)Interaction between FOXP3 and CXXC1 was assessed by co-IP (forward and reverse) using Treg cell lysates.
(E)Immunofluorescence for FOXP3 and CXXC1 colocalization in Treg cells. Scale bars, 2 μm.
Foxp3creCxxc1fl/fl mice spontaneously develop a fatal early-onset inflammatory disorder.
(A)Survival curves of Foxp3cre (black line) and Foxp3creCxxc1fl/fl (red line) mice (n = 20).
(B)Gross body weight of Foxp3cre and Foxp3creCxxc1fl/fl mice (n = 6).
(C)A representative image of Foxp3cre and Foxp3creCxxc1fl/fl mice.
(D)Representative images showing the spleen and peripheral lymph nodes from Foxp3cre and Foxp3creCxxc1fl/fl mice.
(E)Hematoxylin and eosin staining of the skin, lung, liver, and colon from Foxp3cre and Foxp3creCxxc1fl/fl mice (scale bar, 100 μm).
(F)Flow cytometry analysis of CD62L and CD44 expression on peripheral lymph node CD4+ and CD8+T cells from Foxp3cre and Foxp3creCxxc1fl/fl mice (left). Right, frequency of CD44lowCD62Lhi and CD44hiCD62Llow population in CD4+ or CD8+ T cells (n = 9).
(G)Lymph node cells from Foxp3cre and Foxp3creCxxc1fl/fl mice were stimulated ex vivo with PMA + ionomycin for 4h and analyzed for IFN-γ expressing in CD4+ or CD8+ T cells using flow cytometry (left). Right, percentages of IFN-γ+CD4+ or IFN-γ+CD8+ T cells in the lymph nodes of Foxp3cre and Foxp3creCxxc1fl/fl mice (n = 7).
All mice analyzed were 18-20 days old unless otherwise specified. Error bars show mean ± SD. The log-rank survival curve was used for survival analysis in A, and multiple unpaired t-test or two-tailed Student’s t-test were used for statistical analyses in B, F, and G (****P<0.0001). The flow cytometry results are representative of three independent experiments.
CXXC1 is essential for Treg cells to suppress T cell-mediated EAE and colitis.
(A)Mean clinical scores for EAE in Rag1−/− recipients of 2D2 CD4+ T cells, either alone or in combination with Foxp3cre or Foxp3creCxxc1fl/fl mice after immunization with MOG35–55, complete Freund’s adjuvant (CFA), and pertussis toxin (n=5).
(B and C)Representative histology of the spinal cord of Rag1−/− mice after EAE induction. Hematoxylin and eosin(H&E) staining (upper), Luxol fast blue (F&B) staining (lower). Scale bars, 50 μm (400×) and 200 μm (100×).
(D)Representative flow cytometry plots and quantification of the the percentages of IFNγ+ or IL-17A+ CD4+Vβ11+ T cells (n=5).
(E)Statistical analysis of the percentage CD4+ FOXP3+ Treg cell in CNS tissues 14 days after EAE induction.
(F)Changes in body weight of Rag1−/− mice after colitis induction(n=6).
(G)Haematoxylin and eosin (H&E) staining of colons from T cell-induced colitis mice 6 weeks after T cell transfer. Scale bars, 50 μm (400×) and 200 μm (100×).
(H)Statistical analysis of the percentage CD4+ FOXP3+ Treg cell in the spleen, mesenteric lymph nodes, and colon 6 weeks after colitis induction.
Error bars show mean ± SD. P values are determined by a two-tailed Student’s t-test or two-way ANOVA and Holm–Sidak post hoc test (A, D, and E). (*P<0.05, **P<0.01, ***P < 0.001, ****P<0.0001).
Single-cell transcriptomics reveals distinct Treg cell populations.
(A)UMAP plot showing clusters identified based on variable gene expression of sorted CD4+YFP+ Treg cells. Each dot represents a cell, and each color corresponds to a different population of cell types. Clustering analysis revealed 8 distinct Treg cell populations.
(B)Mean fold changes in cluster abundance between Foxp3cre and Foxp3creCxxc1fl/fl mice.
(C)Pseudotime trajectories of Treg cells based on Slingshot, color-coded by Treg cell subpopulations.
(D)Visualization of density and clonotype richness across Treg clusters from Foxp3cre and Foxp3creCxxc1fl/fl mice.
(E)TCR sharing of expanded clonotypes across all possible combinations of Treg cells from Foxp3cre and Foxp3creCxxc1fl/fl mice.
(F)Heatmap showing Z scores for the average expression of Treg-specific genes in each cluster between Foxp3cre and Foxp3creCxxc1fl/fl mice.
(G-H)Representative flow cytometry plots and quantification of (G) expression of CD25, CD69, ICOS, T-bet, and (H) IFN-γ in CD4+YFP+ Treg cells from Foxp3cre and Foxp3creCxxc1fl/fl mice (n=5 CD25, n=10 CD69, n=5 ICOS, n=6 T-bet, n=5 IFN-γ).
(I)Ki-67 expression (left) and frequency (right) in CD4+FOXP3+ Treg cells from Foxp3cre a-nd Foxp3creCxxc1fl/fl mice (n=6).
Error bars show mean ± SD. P values are determined by a two-tailed Student’s t-test (G-H). (**P<0.01, ***P < 0.001, ****P<0.0001). The flow cytometry results are representative of three independent experiments.
Cxxc1-deficient Treg cells in het-KO female mice exhibit reduced suppression markers under steady-state conditions.
(A)Schematic representation of wild-type and Cre-positive Treg cells in female Foxp3Cre/+mice.
(B)Flow cytometry analysis of the YFP-FOXP3+ (WT) and YFP+FOXP3+ (KO) Treg cells in Foxp3Cre/+ Cxxc1fl/+ (het-WT) and Foxp3Cre/+ Cxxc1fl/fl (het-KO) female mice (left), along with the frequency and absolute numbers of YFP+ cells within the total Treg population (right)(n=5).
(C)Flow cytometry analysis of Ki-67expression (left) and MFI (right) in YFP− and YFP+ cells within the CD4+FOXP3+ Treg cells from 6- to 8-week-old het-KO female mice (n=4).
(D)Representative flow cytometry plots and quantification of ICOS, CD25, CTLA4, and GITR expression in YFP− and YFP+ cells within CD4+FOXP3+ Treg cells from het-KO female mice (n=5).
Error bars show mean ± SD. P values are determined by a two-tailed Student’s t-test (G-H). (**P<0.01, ***P < 0.001, ****P<0.0001). The flow cytometry results are representative of three independent experiments.
CXXC1 regulates Foxp3-dependent molecule H3K4 trimethylation in Treg cells.
(A-B)Heatmaps showing H3K4me3(A)and CXXC1(B) signals centered on narrow, medium, and broad domains. The top panel shows the average CUT&Tag signals around indicated domains.
(C)Venn diagrams showing the overlap of FOXP3 binding genes, CXXC1 binding genes, and H3K4me3-BD-associated genes with decreased H3K4me3 levels after CXXC1 depletion in Treg cells.
(D)Gene Ontology (GO) pathway analysis of the overlap peaks.
(E)Representative genome browser view showing the enrichments of FOXP3, CXXC1, and H3K4me3 in Treg cells.
