Fig 2: The effect of H2A.Z and/or SFRP1 silencing on the proliferation and invasion of ICC cells. HuccT-1 and RBE cells were transduced with LV-shRNA-NC (negative control), LV-shR-H2A.Z and/or siR-SFRP1 to generate cells with silencing. Proliferation, invasion and wound healing were evaluated. A The dynamic proliferation of ICC cells. B The result of nude mouse tumorigenicity assay, and the expression of PCNA in mouse tissues. C The invasion of ICC cells (200 × magnification). D The wound healing ability of ICC cells. The data shown are representative images or expressed as individual mean values or the means ± SDs of each group in three separate experiments. *p < 0.05, **p < 0.01, ***p < 0.001

Fig 3: Upregulation of H2A.Z is associated with downregulation of SFRP1 and poor survival. A-B The levels of H2A.Z and SFRP1 in 25 primary ICC and paired adjacent nontumor tissues were examined by IHC. Scale bar, 100 μm. B The expression levels of H2A.Z (upper panel) and SFRP1 (lower panel) were semiquantitatively analysed. C The mRNA transcript levels of H2A.Z and SFRP1 were measured by quantitative RT–PCR in 20 pairs of ICC and paracancerous tissues, and Pearson correlation coefficients were calculated (r = − 0.6776, **p = 0.001). D The mRNA transcript levels of H2A.Z and SFRP1 were quoted from the GSE76297, and Pearson correlation coefficients were calculated (r = − 0.04623, ****p = 0.001). E Kaplan–Meier analysis of overall survival in 50 ICC patients with higher H2A.Z (n = 25) and lower SFRP1 (n = 25) expression vs. patients with lower H2A.Z (n = 25) and higher SFRP1 (n = 25) expression. The data shown are representative images, individual mean values or the mean ± SD of each group in three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001

Fig 4: Schematic illustration of the role of H2A.Z in regulating SFRP1 promoter activity in ICC. In the nucleus, H2A.Z recruits KDM1A, H3K4me1 and H3K4me2 to form a complex to promote demethylation upstream of the SFRP1 promoter and coordinates with HP1α, which is located at downstream CpG islands and recruits SUV39H1/H3K9me3/DNMTs to repress SFRP1 transcription

Fig 5: H2A.Z binds to the upstream region of the SFRP1 promoter and reduces SFRP1 expression. A Graphical representation of the CpG sites in the SFRP1 promoter. B ChIP assay of KDM1A, H3K4me1 and H3K4me2 enrichment in the P3 region upstream of the SFRP1 promoter. IgG was used as the negative control. Relative fold enrichment = [%(ChIP/Input)]/[%(IgG/Input)]. C BSP analysis of the CpG site methylation rate. Each dot represents one CpG site. White represents the absence of methylation, and pale green to deep yellow represent increasing methylation rates. D Quantitative RT–PCR analysis of the SFRP1 mRNA transcript level relative to the GAPDH mRNA transcript level in the blank control, LV-shRNA-NC, LV-shR-H2A.Z and LV-NC-H2A.Z + 5-aza-dC groups. The data are expressed as the individual mean values or the mean ± SD of each group in three independent experiments. **p < 0.01, ***p < 0.001