Fig 2: Similar genome association profiles of ß-catenin, MafB, and WDR77.(A) Heatmap of ChIP-seq signals of the indicated proteins in tamoxifen- (TAM) or ethanol (ETH)-treated cells. Each row of the heatmap corresponds to a genomic region bound by ß-catenin in transformed cells, with rows sorted by the average ChIP-seq signal across six experiments. TSS-proximal and TSS-distal peaks are shown separately. (B) Percentage of ChIP-seq peaks of the indicated proteins that are TSS proximal. (C) Correlation of TCF and LEF ChIP-seq signals at TSS-proximal and TSS-distal peaks of ß-catenin in transformed cells-TAM peaks. Pearson correlation coefficients are shown.

Fig 3: The direct interaction of β-catenin, MafB, and WDR77.(A) Co-IP of endogenous β-catenin, MafB, and WDR77 proteins in the cytoplasm (Cyt), nucleoplasm (Nuc), and chromatin (Chr) fractions of transformed (T) and non-transformed (E) cells. IgG was used as a co-IP control. (B) Western blots of the indicated proteins in fractions from the cell-free extract prior to the co-IP. Fractions are defined by MEK1/2 (cytoplasm), U1 SnRNP70 (nucleoplasm), and H3 (chromatin). (C) co-IP of recombinant β-catenin, MafB, and WDR77 proteins.

Fig 4: β-catenin, MafB, and WDR77 are important for transformation.(A) Western blots of β-catenin, MafB, and WDR77 in CRISPR knocked-out (CRISPR-KO) and CRISPR control cells (CRISPR-Ctrl). (B, C) Relative growth in low attachment (B) or normal (high-attachment)(C) conditions of transformed (tamoxifen treated; TAM shown in darker colors) and non-transformed (ethanol treated; ETH shown in lighter colors) cells with the indicated knockouts. (D, E) Relative STAT3 (D) or AP-1 (E) reporter (luciferase) activity in non-transformed and transformed cells upon the indicated siRNA knockdowns. Error bars indicated ± SD of 3 replicates.

Fig 5: Prostate cancer‐derived miR‐152‐3p induces osteoclastogenesis through silencing Mafb. (a) Representative TRAP staining images of BMMs cultured with respective transfected sEVs. Bar represents 400 μm. Quantification of multinucleated TRAP+ osteoclasts per well, n = 8. (b) Relative mRNA expression levels of Acp5, Nfatc1, Atp6v0d2 and Ctsk in BMMs cultured with respective transfected sEVs, n = 3. (c) Relative luciferase activity of the wild‐type Mafb 3′ UTR in BMMs was reduced by C4/miR‐152‐3p‐sEVs incubation, while transfection with miR‐152‐3p inhibitors abolished this effect, n = 3. (d) Representative images TRAP staining of BMMs cultured with C4/miR‐152‐3p‐sEVs. sEVs were pre‐incubated with Annexin V (2 μg/ml) or pretrasfected miR‐152‐3p inhibitors, recipient BMMs were restored Mafb expression through plasmid transfection. Bar represents 400 μm. Quantification of multinucleated TRAP+ osteoclasts per well, n = 8. (e) Western blot and densitometric analysis of MAFB, NFATc1, c‐Fos, MITF and β‐actin protein levels in BMMs in indicated groups, n = 3. The data in the figures represent the averages ± SD. Statistically significant differences between the treatment and control groups are indicated as * (P < 0.05) or ** (P < 0.01). N.S. means no significant differences between two groups