Fig 2: Schematic illustration of the mechanisms underlying the regulation of PHF8 and KDM3A by AR, MYC and miR-22DuringNED (left), androgen deprivation reduces AR activity and MYC expression, miR-22 is derepressed from the downregulated MYC and consequently, PHF8 and KDM3A are downregulated by the elevated miR-22. Short IL-6 treatment downregulates MYC, although, it increases AR activity. The upregulation of miR-22 can be caused by the elevated AR activity and the derepression from the downregulated MYC. In CRPC cells (right), miR-22 is repressed by MYC, despite of elevated AR activity. The basal expression level of miR-22 partially contributes to the restored expression of PHF8 and KDM3A in CRPC cells.

Fig 3: MYC sustains the expression of PHF8 and KDM3A by repressing miR-22A. Expression of the indicated proteins, as assessed by western blotting, in LNCaP, LNCaP-Abl and LNCaP-IL-6 cells treated with 20 nM scrambled (Ctrl) or MYC siRNA for 60 Hours. B. Expression of miR-22, as assessed by RT-qPCR, in cells treated as in A. C. Expression of the indicated proteins, as assessed by western blotting, in LNCaP cells over-expressing a constitutively active MYC construct and a tamoxifen-inducible mock construct (mock-ER) or MYC-ER-HA. 1 μM tamoxifen was added to the medium for 48 hours. D. Expression of miR-22 in LNCaP cells overexpressing mock-ER or MYC-ER-HA as in C. E. Expression of the indicated proteins, as assessed by western blotting, in LNCaP cells over-expressing a doxycycline-inducible MYC-HA construct and treated with medium containing vehicle (Reg), 20 ng/ml IL-6 (IL-6) or CS-FBS (CS-FBS) for 6 days. Doxycycline (dox) was used at 0.5 μg/ml for the indicated amount of time. F. Expression of miR-22 in LNCaP cells treated as in E as assessed by RT-qPCR. The expression of miR-22 was normalized to the vehicle treatment (no doxycycline). All experiments were performed at least three times independently, and standard deviation is indicated by bars. *: p<0.05; **: p<0.01.

Fig 4: Knockdown of KDM3A results in epigenetic control and transcriptional activation of the androgen responseA. Map of transcriptional regulation of KDM3A on androgen signaling. Red indicates positive response of KDM3A on gene (down with shRNA knockdown of KDM3A); blue indicates negative response of KDM3A on gene (up with shRNA knockdown of KDM3A). B. Gene set enrichment analysis of ranked transcriptomic data upon shRNA knockdown of KDM3A indicated significant enrichment of hallmark gene set of androgen response with p-value and false discovery rate q-value below 1.0E-20.

Fig 5: siRNA screening of histone demethylases required for Wnt/β-catenin-mediated transcription.(a) Analysis of luciferase activity of 293 T/Top cells treated with the indicated siRNA relative to those transfected with scramble siRNA. The relative luciferase activity was normalized against the protein concentration of each cell lysate sample. Values mean±s.d. from three independent experiments. (b) Real-time RT-PCR analysis of the relative expression of histone demethylases in human of 293 T/Top cells transfected with targeted siRNA relative to their expression in cells with scramble siRNA. Values are mean±s.d. of three independent experiments. (c) KDM3A, KDM3B and JMJD1C were knocked down by siRNA in 293 T cells. (d,e) Knockdown of endogenous KDM3A, KDM3B and JMJD1C inhibited AXIN2 and DKK1 expression induced by LiCl (d) or Wnt3a (e) in 293 T cells. Values are mean±s.d. of triplicate samples from a representative experiment. AB, siRNA for KDM3A/B; A+B+C, siRNA for KDM3A, KDM3B and JMJD1C. (f) KDM3A and KDM3B enhanced Topflash reporter activities induced by β-catenin in 293 T cells. Values are mean±s.d. for triplicate samples from a representative experiment. β-Cat; the mutant form of β-catenin. (g) Expression of KDM3A rescued Wnt3a-induced Topflash activity in 293 T cells depleted of KDM3A/B. Values are mean±s.d. of triplicate samples from a representative experiment. (h) Expression of KDM3A rescued AXIN2 and DKK1 expression in 293 T cells depleted of KDM3A/B. Values are mean±s.d. of triplicate samples from a representative experiment. (i,j) Wild-type KDM3A, but not KDM3Amut (KDM3A-H1120A), enhanced Topflash reporter activities induced by β-catenin (i) or Wnt3a in 293 T cells. Values are mean and±s.d. for triplicate samples from a representative experiment. *P<0.05; **P<0.01, unpaired two-tailed Student' s t-Test. (k) Western blot showed the knockdown of KDM3A/B by shRNA in HCT116 cells. (l) The knockdown of KDM3A/B inhibited the expression of Wnt target genes by Real-time RT-PCR. (m) Quantitative comparison of genes down-regulated by knockdown of KDM3A/B and Wnt/β-catenin target gene signatures in adenoma (left panel) and carcinoma (right panel) using GSEA. P<0.001, false discovery rate q<0.001.