Fig 2: LDB2 overexpression inhibits tumor cell proliferation and migration(A) Overexpression of LDB2 was confirmed by WB in HCC samples. GAPDH was chosen for loading control. (B and C) Cell proliferation ability was analyzed by MTT assays and colony formation assays. (D) LDB2 overexpression arrested cell cycle. (E) Overexpression of LDB2 inhibited cell migration. (F) LDB2 overexpression promotes cell apoptosis as shown by WB. (G) LDB2 overexpression inhibited tumor formation in vivo. 2×106 WT or LDB2-overexpressing sample cells were injected into nude mice. 5 weeks post injection, tumor weight was measured. *p<0.05, **p<0.01 and ***p<0.001 by two-tailed Student’s t test. All data presented are shown as means ± SD collected from three independent experiments.

Fig 3: LDB2 deletion promotes liver cancer cell proliferation and migration(A) LDB2 was deleted in HCC samples using two individual sgRNAs through CRISPR/Cas9 technology. The deletion efficiency was validated by Western blot. (B and C) Deletion of LDB2 enhanced proliferation of HCC sample cells as examined by MTT assays and colony formation assays. (D) Percentages of cell-cycle distribution of WT or LDB2-deleted HCC sample cells were checked by FACS. (E) Percentages of cell apoptosis in LDB2-depleted HCC samples were measured by WB. (F and G) LDB2-deleted cells showed increased migration ability as measured by migration assays and WB. (H) The weights of tumors were measured 5 weeks post injection. 2×106 WT or LDB2-deleted sample cells were injected into nude mice. *p<0.05 and **p<0.01 by two-tailed Student’s t test. All data presented are shown as means ± SD collected from three independent experiments.

Fig 4: LDB2 recruits BRD7 to HEY1 promoter(A) BRD7 was identified as a potential interactive protein of LDB2 by MS. (B) LDB2 directly associated with BRD7 as shown by co-IP assays. 1×107 sample cells were lyzed and supernatant was used for incubation with antibody. (C) BRD7 enriched on HEY1 promoter. (D) LDB2 deletion impaired the enrichment of BRD7 on HEY1 promoter. (E) BRD7 deletion promoted HEY1 expression. Total RNA was extracted from WT or BRD7-deleted sample cells and RT-qPCR was conducted. Fold changes were normalized to endogenous Actb. (F) Overexpression of LDB2 or BRD7 inhibited HEY1 expression. Fold changes were normalized to endogenous Actb. (G) BRD7 expression was negatively correlated with HEY1 expression according to Park’s cohort. *p<0.05, **p<0.01 and ***p<0.001 by two-tailed Student’s t test. All data presented are shown as means ± SD collected from three independent experiments.

Fig 5: LDB2 is weakly expressed in HCC tumors(A) Expressionanalysis ofLDB2 in normal (n=215) and tumor tissues (n=224) according to Wang’s cohort and Zhang’s cohort. (B) Total RNAs were extracted from human HCC samples. Then expression levels ofLDB2 were analyzed in 20 pairs of peritumor and HCC samples by RT-qPCR. Fold changes were normalized to endogenous Actb. T, tumor; P, peritumor. (C) Expression levels ofLDB2 in peritumor and HCC samples were examined by Western blot. GAPDH was chosen for sample loading control. (D) Protein levels of LDB2 were analyzed by IHC in peritumor and tumor samples. Scale bars, 100µm. (E) Expression level of LDB2 was negatively correlated with clinical stages according to Wang’s cohort. (F) Expression level of LDB2 was negatively correlated with poor prognosis by Kaplan–Meier survival analysis according to Wang’s cohort. The mean expression levels of LDB2 served as cut-off values. *p<0.05, **p<0.01 and ***p<0.001 by two-tailed Student’s t test. All data presented are shown as means ± SD collected from three independent experiments.