CXXC finger protein 1 is critical for T-cell intrathymic development through regulating H3K4 trimethylation

T-cell development in the thymus is largely controlled by an epigenetic program, involving in both DNA methylation and histone modifications. Previous studies have identified Cxxc1 as a regulator of both cytosine methylation and histone 3 lysine 4 trimethylation (H3K4me3). However, it is unknown whether Cxxc1 plays a role in thymocyte development. Here we show that T-cell development in the thymus is severely impaired in Cxxc1-deficient mice. Furthermore, we identify genome-wide Cxxc1-binding sites and H3K4me3 modification sites in wild-type and Cxxc1-deficient thymocytes. Our results demonstrate that Cxxc1 directly controls the expression of key genes important for thymocyte survival such as RORγt and for T-cell receptor signalling including Zap70 and CD8, through maintaining the appropriate H3K4me3 on their promoters. Importantly, we show that RORγt, a direct target of Cxxc1, can rescue the survival defects in Cxxc1-deficient thymocytes. Our data strongly support a critical role of Cxxc1 in thymocyte development.


Supplementary Figure 3
Intrathymic T cell development was severely impaired in Cxxc1 deficient mice (a) Quantification of DN, DP, CD4+ or CD8+ SP thymocytes subpopulations. The P value was calculated by unpaired t-test (two-tailed). NS, not significant. ***P < 0.001. Error bars present s.d. (b) Thymocytes from Cxxc1 fl/fl or Cxxc1 fl/fl Lck Cre mice were analyzed. (c) Lineage negative cells were analyzed for the profile of DN1-DN4 subsets. (d) Quantitation of the absolute cell numbers of CD4 + and CD8 + T cell subsets, as the product of total splenocytes multiplied by the percentage of cells found in that population. The statistical significance was calculated by unpaired t-test (two-tailed). NS, not significant. ***P < 0.001. (e)Total splenocytes from indicated mice were analyzed for CD4 and CD8 expression by flow cytometry. Data is from six mice analysis (a-e). (f, g) CD4 + SP or CD8 + SP cells sorted from indicated mice were used for genome DNA PCR analysis using indicated primers (f); or were used for Real-time Quantitative PCR analysis (g). Data comes from independently sorted samples sets (each from 2 mice). Error bars indicate s.d.(h-k) DN3 cells sorted from Cxxc1 fl/fl CreERT2 + mice were co-cultured with OP9-DL1 at presence or absence of 4-hydroxytamoxifen (4-HT) for 5 days. PCR analysis of Cxxc1 deletion and the Loxp-flanked allele (f); detection of Cxxc1 expression by Real-time Quantitative PCR assays (g); the number of thymocytes (h); and flow cytometry analysis of thymocytes (i) were shown. Data is from four mice experiments. The statistical significance was calculated by unpaired t-test (two-tailed). ***P < 0.001. Error bars indicate s.d.

Supplementary Figure 4
The Setd1 interacting domain of Cxxc1 is required for intrathymic T cell development (a) DP thymocytes were sorted from wild-type mice, then lysed and immunoprecipitated with anti-Cxxc1, anti-Setd1 or rabbit immunoglobulin G (IgG; control antibody) and detected with anti-Cxxc1 or anti-Setd1 (top), immunoblot analysis of lysates (without immunoprecipitation) with anti-Cxxc1 or anti-Setd1 (bottom). Data is from three independently sorted DP cell (each from 1 mouse). (b) Schematic view of CXXC1 protein and strategies of generation of mutated and truncated CXXC1 proteins. (c) Wild-type DN3 cells were transfected with indicated retrovirus and were co-cultured with OP9-DL1 cell line; GFP positive cell were sorted after 5 days and subjected to Western Blot analysis. Data comes from two separately sorted DN3 cell (each from 2 mice). (d) DN3 cells from wild-type mice were sorted and transfected with indicated retrovirus, GFP + thymocytes were analyzed by FACS after co-culture with OP9-DL1 for 5 days. Data comes from two separately sorted DN3 cell (each from 1 mouse). (e) Detection of DNMT1 protein expression in DP thymocytes. Thymocytes from WT or Cxxc1-deficient mice were permeabilized with Foxp3 staining buffer, followed by staining with anti-DNMT1 primary antibody and anti-mouse IgG fluorescein-conjugated secondary antibody. Data is from independent three mice experiments (c) Hierarchical clustering of 387 Cxxc1 target genes with at least 2 fold expression difference between WT and Cxxc1-deficient DP thymocytes.

Supplementary Figure 6
Epigenetic regulation of Cxxc1 target loci (a, b) H3K27me3 (a) and H3K9me2 (b) modifications on Cd8a-Cd8b, Zap70, Lck and Rorc loci in wild-type and Cxxc1-deficient DP thymocytes, detected by ChIP-PCR assays. Data is from three independently sorted DP cell sets (each from 2 mice). (c) Cxxc1 binding profile and H3K4me3 on the locus of Bdnf gene in wild-type and Cxxc1-deficient DP thymocytes. (d-f) H3K4me3 (d), H3K9ac (e), and H3K27ac (f) modifications on Bdnf locus in wild-type and Cxxc1-deficient DP thymocytes. Experiments are same as Fig. 4 (d-f). Error bars indicate s.d. (g) Detection of Lck protein level in indicated DP thymocytes by Western Blot. Data comes from three independently sorted DP cell sets (each from 2 mice).

Supplementary Figure 7
Profile of expression, DNA methylation and binding of Cfp1, Setd1, H3K4me3 during T cell development.
(a) Indicated populations sorted from wild-type mice were subjected to Real-time Quantitative PCR analysis. Data is from independently sorted samples (each from 2 mice). (b) Detection of DNA methylation on promoter regions of indicated genes from different populations of wild-type mice were analyzed by MeDIP assays. DN1/2: DN1 mixed with DN2 cell. Data comes from independently sorted samples (each from 3 mice). (c-e) Analysis of H3K4me3 modification(c), binding of Cxxc1, (d) and Setd1 (e) by ChIP-qPCR on Cd8a, Rorc, Zap70 and Lck loci in indicated populations. DN1/2: DN1 mixed with DN2 cell. Data comes from three independently sorted samples (for H3K4me3, each from 2 mice; for Setd1 and Cxxc1, each from 4 mice). Errors bars indicate s.d.

Supplementary Figure 8
Bcl2 transgene could restore the survival of Cxxc1-deficient DP thymocytes but not TCR signaling (a) DP thymocytes sorted from wild-type or Cxxc1-deficient mice were stimulated with anti-CD3 plus anti-CD28 antibodies for indicated time. The cell lysate were subjected to Western Blot using indicated antibodies. Data is from three independently sorted DP cell sets (each from 5 mice). (b) Quantitation of thymocyte subpopulations in indicated mice. The statistical significance were determined by unpaired t-test (two-tailed). ***P <0.001. (c) Flow cytometry analysis of thymocytes from wild-type or Cxxc1-deficient mice on hBcl2 transgenic background. (d) Percentage of DP thymocytes in indicated mice was shown. The P value were calculated by unpaired t-test (twoailed). NS, not significant. ***P<0.001. Data comes from six mice analysis (b-d). (e) DP thymocytes sorted from wild-type, Cxxc1deficient, hBcl2-TG or hBcl2-TG-Cxxc1-deficeint mice were cultured and harvested at indicated time. Viability was measured using flow cytometry by staining of Annexin V. Data comes from three independently sorted DP cell sets (each from 3 mice). Results are presented as mean ± s.d. (f) Annexin V staining of thymocytes from indicated mice following staining of anti-CD4 and anti-CD8 antibodies, see (c).

Supplementary Figure 9
Full-size Western Blots.