Fig 2: MiDAC negatively regulates the occupancy of the MLL3/4 complexes at regulatory elements genome-wide.(A) MA plots displaying the log2 fold change of UTX (left) and MLL4 (right) occupancy in Dnttip1 KO compared with WT mESCs. A significantly higher number of UTX and MLL4 up-regulated peaks (FC > 2) than down-regulated peaks (FC > 2) is detected in Dnttip1 KO versus WT mESCs. (B, C) Heat maps centered on (B) 1,283 UTX/MLL4 up-regulated peaks (both FC > 2) or (C) 2,396 regions with up-regulated H3K4me2 (FC > 2) in Dnttip1 KO compared with WT mESCs. Occupancy of DNTTIP1, UTX, MLL4, H4K20ac, H3K4me1, H3K4me2, H3K4me3, and H3K27ac in WT and Dnttip1 KO mESCs is displayed. The mean signal of each ChIP-seq peak was determined, and a two-sided Wilcoxon rank sum test was performed between WT and Dnttip1 KO mESCs for each histone mark and protein ChIP sample for which a significant change is claimed. (D) Gene set enrichment analysis showing transcriptional up-regulation of genes associated with regions of increased H4K20ac (left), UTX/MLL4 up-regulated peaks (middle), and regions of increased H3K4me2 (right) in Dnttip1 KO versus WT mESCs.

Fig 3: 2C-ChIP analysis of a differentiation time course identifies a HOXA cluster region resistant to H3K27me3 demethylation. a Time points of RA-induced NT2-D1 differentiation examined in this study. b Steady state transcript levels of HOXA and lncRNA genes as measured by RT-qPCR during RA treatment. RNA levels relative to actin are shown on the y-axis. Dashed lines indicate values below which measurements are unreliable. Oligonucleotide sequences used for quantification are shown in Additional file 1: Table S1. All RNA measurements are from at least three PCRs. c-h 2C-ChIP analysis of H3K4me3 and H3K27me3 levels, and of the chromatin-binding proteins Ash2L, SUZ12, CTCF, and UTX during the RA induction time course. 2C-ChIP tracks are presented as described in Fig. 2f and Additional file 7: Figure S1. The dashed black box identifies a HOXA cluster region slow to lose H3K27me3 signal. 2C-ChIP BED files including surrounding negative controls are in Additional file 6: BED file 7–30

Fig 4: T-cell development in the absence of H3K27 demethylases.(a) Contour plots of CD4 and CD8 expression in thymocytes (left) or splenocytes (right), and expression of TCRß and CD24 on CD4 SP thymocytes (middle) from wild-type (WT), Jmjd3f/fCd4-Cre (Jmjd3 KO), Utxf/fCd4-Cre (Utx KO) or Jmjd3f/f Utxf/fCd4-Cre (dKO) mice. Numbers adjacent to outlined areas indicate percentage T-cells. Numbers of mice for each genotype indicated in (b). The gate (middle column) defines mature (TCRhi CD24lo) CD4 SP thymocytes. Data is representative of 10 (Jmjd3 KO), 9 (Utx KO) and 5 (dKO) separate experiments. (b) Dot plots show absolute cell numbers of mature (TCRhi CD24lo) CD4 SP thymocytes or spleen CD4+ T-cells from mice in (a). The statistical significance was determined by unpaired t-test (two-tailed). *: P<0.05, **: P<0.01, ***: P<0.001, ****: P<0.0001. Each symbol represents one individual mouse. Error bars indicate s.d.

Fig 5: Utx is a tumor suppressor and genetically inactivated in T-ALL Utx whereas is dispensable for physiological T cell developmenta, b, Study of lymphoid development on Utx-/Y compared to Utx+/+ (or UTX+/Y, not shown) background. FACS analyses for CD4+ and CD8+ expression (a) and the relative proportions of CD4+, CD8+ double positive thymocytes across different genotypes (b) are shown. A representative example of three independent samples (biological replicates) is shown. c, Illustration of the transplantations scheme for the in vivo leukemia studies. d, e, T-ALL progresses faster in the male knockout background for Utx (Utx-/y) compared to female wildtype background (Utx+/+), in Notch1-IC/GFP, as it is demonstrated by white blood counts (d) in the peripheral blood as well as GFP+ cells representation upon transplantation of wild type progenitors (e) from female mice (Utx+/+) compared to the corresponding knock-out cells (Utx-/y). f, Survival study of the recipients for male wild-type (Utx+/Y, n=7) and knock-out (Utx-/Y, n=5) mice expressing Notch1-deltaE(?E)/GFP (the allele with the weaker oncogenic action compared to Notch1-IC). g, h, Survival analysis of recipients upon transplantation of wild type progenitors from female mice (Utx+/+) compared to the corresponding knock-out cells (Utx-/y) infected with N1-IC (g) or N1-?E (h). i, qPCR validation of expression levels of one down-regulated (Suz12) and one up-regulated (Il7ra) gene in Utx-/Y (compared to UTX+/Y). The average of three independent samples/studies is presented. j, Targeted Sanger sequencing in pediatric led to identification of three cases with frameshift mutations. The positions of mutations are indicated by dashed lines in the electropherograms. k, Identification of one in-frame deletion (p.A14_A17del, #1, top panel), one splice acceptor site (#2, second panel) and one missense mutation (#3, third panel) in adult T-ALL. #4 is an adult T-ALL case with wild-type UTX (control, bottom panel). Mutations are indicated by red-colored characters. l, Levels of UTX in CUTLL1 T-ALL cells in the absence (-dox) or presence (+dox) of doxycycline. m, n, Apoptosis analysis through measuring annexin-V staining using control LacZ-expressing and UTX-expressing CUTLL1 in the absence (-dox) or presence (+dox) of doxycycline. Representative plots (l, n) as well as average representation (l, m) of three independent experiments are shown.