ZNF143-Mediated H3K9 Trimethylation Upregulates CDC6 by Activating MDIG in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the most common cause of death worldwide, and the molecular mechanism of uncontrolled HCC progression remains unclear (1). Accumulating evidence has suggested that dysregulation of the cell cycle is one of the most important hallmarks of HCC (2). A selective CDK4/6 inhibitor showed encouraging results in preclinical models of HCC and might represent a novel therapeutic strategy for HCC treatment (3). Elucidation of the mechanisms underlying the pathogenesis and molecular biology of dysregulated cell cycle in HCC is fundamental for the development of effective therapeutic treatments.

Zinc finger protein 143 (ZNF143) can be activated by IGF1 and is a transcription factor of the C2H2 family. This protein contains a transcription activation domain, 7 Kruppel-like C2H2 zinc finger motif domain, and a C-terminal domain (4, 5). Some studies have shown that ZNF143 possesses potent oncogenic properties and can promote growth (6, 7). In prostate cancer, ZNF143 is involved in cisplatin resistance by regulating the transcription of DNA repair genes (8). In addition, ZNF143 is closely related to tumor malignancy via regulating cell motility and exerting tumor-suppressive effects in breast cancer (9). Intriguingly, Haibara and colleagues found that two small molecules inhibit ZNF143 activity, suggesting an opportunity for cancer therapeutics (10). To date, the exact biological role of ZNF143 in HCC has not been investigated.

An increasing body of evidence suggests that epigenetic changes such as DNA methylation and other posttranslational modifications of chromatin play critical roles in HCC tumorigenesis (11, 12). Of these epigenetic alterations, dynamic histone modifications play a pivotal role in the regulation of gene transcription (13). Histone methylation can be regulated by histone methyltransferase and demethylases. Mineral dust-induced gene (MDIG), also known as myc-induced nuclear antigen with a molecular weight of 53 kDa (MINA53), was first identified in alveolar macrophages obtained from coal miners (14), contains one JmjC domain and can demethylate H3K9me3 (15, 16). Numerous studies have demonstrated that MDIG is overexpressed in malignancies and can promote cell migration, cell-cycle transition, and proliferation (14, 17).

Various studies have shown that functionally related genes involved in epigenetic reprogramming can be controlled by specific transcription factors (18, 19). Elucidation of the epigenetic alterations induced by transcription factors will provide more comprehensive and precise insight into the regulatory mechanisms of transcription factors. In the current study, we identified cell division cycle 6 (CDC6) as a downstream target gene of ZNF143 during the regulation of cell cycle in HCC. ZNF143 promoted CDC6 expression by attenuating the enrichment of H3K9me3 in CDC6 promoter region through activation of MDIG. Our findings reveal the fundamental roles of ZNF143 in cell-cycle control and identify the ZNF143–MDIG–CDC6 axis as a potential target for anticancer therapy.

Human primary HCC/matched adjacent noncancerous liver tissue specimens were obtained with informed consent from the Guangxi Cancer Institute (Nanning, China), Zhejiang University (Hangzhou, China), and the Qidong Liver Cancer Institute (Qidong, China). None of the patients received preoperative radiation or chemotherapy. This study was approved by the Research Ethics Committee of Renji Hospital, Shanghai Jiao Tong University School of Medicine and in accordance with Declaration of Helsinki. Written informed consent was received from participants prior to inclusion in the study.

The mRNA expression data were downloaded from The Cancer Genome Atlas (TCGA) database (https://tcga-data.nci.nih.gov/tcga/). HCCDB data were directly downloaded from the HCCDB dataset (http://lifeome.net/database/hccdb/home.html; ref. 20).

Tissue microarrays (TMA) were constructed, and the diagnosis was confirmed by three pathologists. IHC was performed as described previously (21). Briefly, the sections were incubated overnight with ZNF143, MDIG, and CDC6 at 4°C and then evaluated by two independent observers in a blinded manner. The antibodies used in this study were anti-ZNF143 (HPA003263, 1:10, Sigma-Aldrich), anti-MDIG/MINA53 (PA5-31300, 1:25, Invitrogen), anti-CDC6 (HPA050114, 1:10, Sigma-Aldrich), and anti-Ki67 (GT209429, 1:50, GeneTech). Scores of staining intensity were: 0, negative; 1, weak; 2, moderate; 3, strong. Scores of positively stained cell proportion were: 0, no positive; 1, <10%; 2, 10%–35%; 3, 35%–75%; 4, >75%.The results were scored 0 to 4 by two independent investigators. A score of 0–2 was considered to represent low expression and a score of 3–4 was considered to represent high expression.

The human HCC cell lines MHCC-97L, MHCC-LM3, and MHCC-97H were kindly provided by the Liver Cancer Institute of Zhongshan Hospital, Fudan University (Shanghai, China). Huh7 was obtained from the Riken Cell Bank (Tokyo, Japan). Li-7 cells were purchased from the Cell Bank of the Institute of Biochemistry and Cell Biology, China Academy of Sciences (Shanghai, China). The Hep3B and HEK-293T cell lines were obtained from the ATCC. Cells were all cultured in DMEM (Sigma-Aldrich) containing 10% FBS (Gibco) and maintained at 37°C with 5% CO₂. All cell lines were authenticated by STR profiles in the past 6 months and were used within 10 passages after reviving from the frozen stocks. Cells were free of Mycoplasma contamination was determined by PCR assay.

The human ZNF143 cDNA (NM_003442.6) and CDC6 cDNA (NM_001254.4) was subcloned into the GV492 vector (Genechem). Short hairpin RNA (shRNA) targeting ZNF143 (shZNF143#1/shZNF143#2/shZNF143#3), CDC6 (shCDC6#1/shCDC6#2), and their negative control (shNC) were obtained from Genechem. Target sequences are listed in Supplementary Table S1. The cDNA sequences of human MDIG and shMDIG target sequences were described in our previous study (21). The process of lentivirus production and cell transfection was performed as described previously (21). siRNAs targeting CBX3 (siCBX3#1/siCBX3#2) and CBX5 (siCBX5#1/siCBX5#2) and their negative control (siNC) were obtained from GenePharma. Target sequences are listed in Supplementary Table S1.

Total RNA was isolated with TRIzol reagent (Invitrogen) and reverse-transcribed into cDNA by using the PrimeScript RT Reagent Kit (RR037A, TaKaRa Bio). Quantitative real-time PCR assays were carried out by using TB Green Premix Ex Taq II (RR820A, TaKaRa Bio) on an Applied Biosystems 7500 Software version 2.0.5 real-time PCR system (Thermo Fisher Scientific) according to the manufacturer's instructions. GAPDH was used as an internal loading control. The primer sequences are listed in Supplementary Table S2.

Western blot analysis was performed as described previously (22). Antibodies used in this study were anti-ZNF143 (sc-100983, 1:200, Santa Cruz Biotechnology), anti-MDIG (ab155335, 1:200, Abcam), anti-CDC6 (sc-9964, 1:50, Santa Cruz Biotechnology), anti-H3K9me3 (ab8898, 1:50, Abcam), anti-β-actin (A1978, 1:10,000, Sigma-Aldrich), anti-rabbit IgG (A0545, 1:5,000, Sigma-Aldrich), and anti-mouse IgG (A4416, 1:5,000, Sigma-Aldrich), anti-MDIG (39-7300, 1:200, Thermo Fisher Scientific), anti-CBX3(11650-2-AP, 1:200, ProteinTech), anti-CBX5 (11831-1-AP, 1:200, ProteinTech), and anti-Flag (F1804, 1:200, Sigma-Aldrich).

The transfections of the plasmid DNAs and siRNAs were carried out with Lipofectamine 2000 transfection reagent (Invitrogen) following the manufacturer's instructions.

A total of 1,000 cells were seeded in 96-well plates in three replicates. Next, 10 μL MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] reagent (5 mg/mL, Sigma-Aldrich) was added to each well and incubated for 4 hours at 37°C. Then, the media were removed and 100 μL DMSO (Sigma-Aldrich) was added to the well. The OD value was measured at 570 nm every 24 hours.

Colony formation assay was performed according to previous methods (21). A total of 1,000 cells were seeded into 6-well culture plates and cultured for approximately 15 days. Then, the cells were washed twice with PBS, fixed in neutral buffered formalin for 30 minutes, and stained with Giemsa (Sigma-Aldrich) for another 30 minutes. Three independent experiments were performed.

Flow cytometry analysis was performed as described previously (22). In brief, 1 × 10⁶ cells were plated into 6-well plates. After adhering to the well, the cells were treated with 2 mmol/L thymidine (Sigma-Aldrich) for 24 hours and harvested by trypsin after releasing for 0 and 24 hours. Then, the cells were washed with PBS twice and fixed with 70% ethanol at −20°C overnight. Before analysis by flow cytometry, the cells were washed twice with 1× PBS and resuspended in 200 μL PI solution [50 μg/mL PI (Sigma-Aldrich), 100 μg/mL RNase (Sigma-Aldrich), 0.2% Triton X-100] at 4°C for 30 minutes. The results were analyzed by Modfit 3.2 software. Three independent experiments were performed.

For the tumor xenograft assays, 2 × 10⁶ Li-7 cells infected with ZNF143/Vector, 2 × 10⁶ MHCC-97H cells transfected with shNC/shZNF143#2/shZNF143#3, 2 × 10⁶ Li-7 cells infected with Vector + shNC/ZNF143 + shNC/ZNF143 + shMDIG#2/ZNF143 + shMDIG#3, or 2 × 10⁶ MHCC-97H cells transfected with shNC/shCDC6#1/shCDC6#2 were resuspended in 200 μL of serum-free DMEM and inoculated subcutaneously into one flank of each nude mouse (nu/nu, male, 4 weeks, n = 9 per group). After 4 weeks, all mice were sacrificed, and the tumor weights were measured. The animal experimental protocols were approved by the Shanghai Medical Experimental Animal Care Commission and performed in accordance with the institutional ethical guidelines for animal experiments.

A dual luciferase assay was performed as described previously (21). The normal and mutant promoter region of MDIG was subcloned into the pGL3 vector (Promega). The primers for cloning are provided in Supplementary Table S3 (21, 23). MHCC-97L and MHCC-LM3 cells were transiently transfected with the corresponding pGL3 reporter constructs and the PRL-TK reporter plasmid together with ZNF143 or Vector using Lipofectamine 2000 (Invitrogen). Cells were lysed 48 hours after transfection. The dual luciferase assay reporter system was used according to the manufacturer's instructions (Promega).

MHCC-97H/Hep 3B/Huh7 cells, and MHCC-97H/Huh7 cells transfected with shNC/shMDIG#3 or siNC/siCBX3#2/siCBX5#1, and Li-7/MHCC-LM3 cells overexpressing ZNF143/Vector, and MHCC-97H/Hep 3B cells transfected with shNC/shZNF143#2 were fixed and immunoprecipitated using the EZ-Magna ChIP assay kit as recommended by the manufacturer (Millipore). Antibodies used to immunoprecipitate purified chromatin were anti-H3K4me3 (ab213224, Abcam), anti-H3K9me3 (ab8898, Abcam), anti-H3K27me3 (ab6002, Abcam), and anti-ZNF143 (16618-1-AP, ProteinTech), anti-MDIG (39-7300, Thermo Fisher Scientific), anti-CBX3 (11650-2-AP, ProteinTech), anti-CBX5 (11831-1-AP, ProteinTech). Primers used to amplify the promoter regions of MDIG, CDC6, or other S-phase genes are shown in Supplementary Table S2.

Coimmunoprecipitation (co-IP) assays were performed using MHCC-97H/Huh7 cells, and MHCC-97L/MHCC-LM3 cells overexpressing ZNF143. The cells were harvested in RIPA (Upstate Biotechnology) lysis buffer for 40 minutes on ice and centrifuged at 12,000 × g for 10 minutes. The protein A/G agarose beads were incubated with antibody overnight at 4°C while rotating. After washing, the complexes were subjected to Western blotting analysis. Antibodies used were anti-MDIG (12214-1-AP, ProteinTech), anti-CBX3 (11650-2-AP, ProteinTech), anti-CBX5 (11831-1-AP, ProteinTech), and anti-Flag (F1804, Sigma-Aldrich).

For mRNA profiling analysis, MHCC-97H and Huh7 cells were stably transfected with shMDIG#3 or shNC and then subjected to a human gene expression microarray (Affymetrix) assay (GeneTech, Shanghai, China). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (genes with P < 0.05, fold change > 1.5) was carried out by DAVID software (https://david.ncifcrf.gov).

Statistical analysis was carried out using SPSS 19.0 software and GraphPad Prism 5 and ImageJ software. Two-tailed and unpaired Student t tests were used for two group comparisons. The differences in ZNF143, MDIG, and CDC6 expression levels between the paired HCC tissues and adjacent nontumorous liver tissues were compared by paired t tests. Pearson correlation analysis was performed to analyze the correlation of two molecules. Survival curves were estimated using the Kaplan–Meier method and compared using the log-rank test. Data are shown as the mean ± SD. P < 0.05 was considered statistically significant.

To investigate the role of ZNF143 in HCC, we first analyzed ZNF143 mRNA expression in TCGA database. As shown in Fig. 1A, ZNF143 expression in HCC samples was significantly higher than that in their matched benign counterpart. We then examined the mRNA and protein expression of ZNF143 in HCC and matched noncancerous tissues from our lab by qPCR and Western blot analyses. Consistent with the results of TCGA dataset, ZNF143 was upregulated in HCC tissue compared with control tissue (Fig. 1B–D). The HCCDB dataset showed that HCC is one of the tumors displays obvious upregulation of ZNF143 (Supplementary Fig. S1A; ref. 20). These results indicated that ZNF143 expression is substantially increased in HCC. Notably, patients with ZNF143 overexpression exhibited worse overall survival in TCGA dataset (Fig. 1E). Moreover, ZNF143 expression was positively correlated with histologic grades in TCGA dataset (Fig. 1F). These results suggested that upregulation of ZNF143 contribute to poor clinical outcomes of patients with HCC.

To evaluate the biological effects of ZNF143 in HCC, we first examined ZNF143 expression in immortalized normal hepatocyte L-02 cell and a series of HCC cell lines by Western blot analyses (Supplementary Fig. 2A). We stably established ZNF143-overexpressing cell lines in MHCC-97L, MHCC-LM3, Li-7 cells (Flag-tagged ZNF143/Vector), and ZNF143 knockdown cell lines in Hep 3B, MHCC-97H cells (Mock/shNC/shZNF143#1/shZNF143#2/shZNF143#3) with lentiviral transfection. The protein level of ZNF143 in ZNF143-overexpressing or knockdown HCC cells were verified by Western blot analysis (Supplementary Fig. 2B). MTT assays indicated that ectopic expression of ZNF143 promoted HCC cell proliferation, whereas knockdown of ZNF143 led to a marked reduction in proliferation (Fig. 2A and B). Plate colony formation assays demonstrated that ZNF143 enhanced the clonogenicity of HCC cells (Fig. 2C and D). To extend the in vitro results, we explored the role of ZNF143 in tumor growth using a xenograft model. As shown in Fig. 2E and F, in nude mice, when ZNF143 was overexpressed in Li-7 cells, the tumors showed enhanced growth. In contrast, tumors with shZNF143#2 and shZNF143#3 exhibited smaller volumes and lower weights than tumors with shNC in MHCC-97H cells. Western blot analysis showed that ZNF143 expression remained high in xenografts from ZNF143-overexpressing cells and low in xenografts from ZNF143 knockdown cells (Supplementary Fig. 2C). Moreover, the Ki67 percentage score of tumor cells was relatively increased in ZNF143-overexpressing group and decreased in shZNF143 group when compared with their negative control (Fig. 2G). The expression of ZNF143 was positively correlated with Ki67 and PCNA in TCGA dataset (Fig. 2H). Altogether, these findings suggested that ZNF143 is a positive regulator of HCC proliferation.

To further clarify the effect of ZNF143 on HCC cell proliferation, we evaluated the cell-cycle distribution by flow cytometry after upregulating and knocking down ZNF143 expression in HCC cell lines. As shown in Supplementary Fig. S2D and S2E, ZNF143 knockdown in MHCC-97H cells led to G₁–S arrest, together with a substantial decline in the cell population in S-phase, and ZNF143 overexpression significantly facilitated the cell-cycle transition by increasing the population of S-phase cells and decreasing the population of G₁ phase cells among MHCC-LM3 cells. Then, we further added 2 mmol/L thymidine to synchronize cells at the G₁–S phase border. After release for 24 hours, flow cytometry analysis showed that the percentage of cells at G₁ phase was significantly lower in ZNF143-overexpressing MHCC-LM3 cells and was higher in ZNF143 knockdown MHCC-97H cells than their corresponding control cells (Fig. 3A and B). These results suggested that ZNF143 is crucial for HCC cell-cycle progression in the G₁–S transition.

We next explored the mechanism by which ZNF143 regulates cell proliferation and cell-cycle progression. Given the role of ZNF143 in the G₁–S transition, we examined the mRNA expression of S-phase genes in HCC cells after ZNF143 overexpression and knockdown by qPCR analyses (Fig. 3C; ref. 24). Among these genes, CDC6 attracted our attention because of its marked change, and its regulatory mechanism in HCC has not been fully elucidated (25, 26). Western blot results further showed that ZNF143 could activate CDC6 at the protein level (Fig. 3D). Then, we evaluated biological effects of CDC6 in HCC cells. We established stable overexpression of CDC6 (CDC6/Vector) in MHCC-97L and MHCC-LM3 cells and stable knockdown of CDC6 (Mock/shNC/shCDC6#1/shCDC6#2) in MHCC-97H and Huh7 cells. The protein expression levels of CDC6 in HCC cells were verified by Western blot analysis (Supplementary Fig. S3A). The plate colony formation and MTT assays revealed that CDC6 enhanced HCC cells proliferation and clonogenicity (Supplementary Fig. S3B–S3E). In xenograft model, the tumors from MHCC-97H-shCDC6 cells showed less active proliferative ability at the implantation site than control (Supplementary Fig. S3F and S3G). Flow cytometry further suggested that CDC6 promotes G₁–S transition in HCC cells (Supplementary Fig. S3H and S3I). In addition, the plate colony formation and MTT assays revealed that CDC6 knockdown inhibited cell proliferation induced by ZNF143 overexpression in HCC cells, and CDC6 reexpression in the ZNF143-knockdown cells rescued the proliferation ability (Fig. 3E–J; Supplementary Fig. S4A–S4E).

Global changes in the epigenetic landscape are a hallmark of cancer (27). A set of functionally related genes involved in epigenetic reprogramming can be controlled by specific transcription factors (28–30). Here, we investigated whether ZNF143 modulates epigenetic changes during activation of CDC6. In contrast to other histone modifications, histone methylation has been highlighted because of its highly specific dynamics with respect to gene regulation (31). Here, we focused on the distribution of the active marker H3K4me3 and the repressive markers H3K27me3 and H3K9me3 on the promoter region of CDC6. Six pairs of primers (Supplementary Table S2) were used to detect possibly altered sites in CDC6 promoter region (Supplementary Fig. S4F). As shown in Fig. 4A and B, ChIP-qPCR assays indicated that the enrichment of H3K9me3 was obviously decreased at −0.2 kb and +0.2 kb promoter regions of CDC6 after ZNF143 overexpression in both MHCC-LM3 and Li-7 cells, and ZNF143 knockdown increased H3K9me3 at CDC6 promoter in MHCC-97H and Hep3B cells. The enrichment of H3K4me3 and H3K27me3 on CDC6 promoter showed different results in ZNF143-overexpressing HCC cells (Supplementary Fig. 4G-J). Taken together, our data suggest that alterations in histone methylation, especially H3K9me3, might contribute to the activation of CDC6 by ZNF143.

We further examined the mechanism underlying ZNF143 regulates H3K9me3 in HCC. Histone methylation is regulated by two classes of enzymes with opposing activities: histone methyltransferases and histone lysine demethylases (31–33). Given role in the regulation of cancer growth, we focused on histone lysine demethylases of H3K9me3(31). On the basis of previous studies, we selected KDM4A, KDM4B, KDM4C, KDM4D, and MDIG (MINA53) for further analysis (32, 34). qPCR showed that ZNF143 upregulated MDIG expression (Fig. 4C). MDIG is overexpressed in a variety of human cancers and plays a key role in cell proliferation (16, 21, 35). We then examined the protein expression of MDIG in HCC cells with altered ZNF143 expression. As shown in Fig. 4D, the MDIG protein level was upregulated in ZNF143-overexpressing HCC cells and significantly inhibited in ZNF143 knockdown HCC cells.

To further elucidate the mechanism by which ZNF143 activates MDIG expression in HCC, we analyzed the MDIG promoter region for possible ZNF143 binding sites. JASPAR software indicated that two potential ZNF143 DNA-binding motifs exist in the promoter region of MDIG (Fig. 4E and F). Luciferase assays showed that relative luciferase activity of the MDIG promoter was significantly induced by ZNF143 overexpression, and mutation of binding site 1, but not other sites, abolished ZNF143-mediated induction of the MDIG promoter reporter activity (Fig. 4G). ChIP assay also indicated that endogenous ZNF143 was recruited to binding site 1 of the MDIG promoter (Fig. 4H). These data suggested that ZNF143 promotes MDIG transcription by binding to its promoter.

A series of studies have shown that MDIG plays an important role in tumorigenesis (14, 16, 36). Our previous study showed that MDIG overexpression promoted HCC cell proliferation, migration and spreading (21). Gene expression profiling of MDIG-knockdown (shMDIG) versus negative control (shNC) revealed that 66 genes, including CDC6, were downregulated and 32 genes were upregulated after MDIG was knocked down in both MHCC-97H and Huh7 cells (Fig. 5A; Supplementary Table S4). The enriched KEGG pathways of 66 overlapping genes included the cell cycle (Fig. 5A). Therefore, we explored the role of MDIG in cell-cycle progression. The results showed that MDIG knockdown in Huh7 cells led to G₁–S arrest. Overexpression of MDIG in MHCC-LM3 cells contributed to G₁–S phase transition (Supplementary Fig. S5A–S5G). These results suggested that MDIG is crucial for HCC cell-cycle progression.

We next examined whether CDC6 expression can be regulated by MDIG. As shown in Fig. 5B and Supplementary Fig. S5H and S5I, Western blot and qPCR analyses indicated that MDIG promoted the expression of CDC6. Our previous study showed that MDIG could suppress the formation of H3K9me3 (21). We hypothesized that alterations of epigenetic modifications might contribute to the increase of CDC6, and the enrichment of H3K9me3 in CDC6 promoters was analyzed. Six pairs of primers were used to detect possibly altered sites in CDC6 promoter (Supplementary Fig. 4F). As shown in Fig. 5C, the enrichment of H3K9me3 was specifically increased at the promoter regions of CDC6 in MHCC-97H and Huh7 cells after MDIG knockdown. In addition, H3K9me3 enrichment in the promoter region of genes regulating S-phase entry were analyzed by ChIP-qPCR. After knocking down MDIG, the enrichment of H3K9me3 was increased at the promoter regions of CCNA2, MCM3, CDC6 in MHCC-97H cells and CCNA2, MCM3, CDC6, TK1 in Huh7 cells (Fig. 5D). ChIP assay also showed the enrichment of MDIG was found at the promoters of these genes (Fig. 5E; Supplementary Fig. S5J). Then, we next try to explore the mechanism mediating the recruitment MDIG to CDC6 promoter. ZNF143 is a transcription factor, we first speculated whether ZNF143 mediates this recruitment. Co-IP showed that endogenous MDIG and Flag-tagged ZNF143 cannot interact with each other (Supplementary Fig. S6A and S6B). CBX3 and CBX5 encoding HP1 proteins can recognize and bind to H3K9me3 (37). Previous study showed a protein interaction between MDIG and CBX3, CBX5 (38). We next determine whether CBX3 or CBX5 mediates this recruitment. Co-IP of endogenous proteins confirmed the interaction between MDIG and CBX3 or CBX5 (Supplementary Fig. S6A, S6C, and S6D). ChIP assay further demonstrated the enrichment of CBX3, CBX5 on CDC6 promoter region (Supplementary Fig. S6E). We next investigated the effect of HP1 proteins depletion on the recruitment of MDIG on CDC6 promoter region. We silenced the expression of CBX3 or CBX5 by siRNA (Supplementary Fig. S6F and S6G). After knocking down CBX3, MDIG was reduced at CDC6 promoter region both in MHCC-97H and Huh7 cells. However, after knocking down CBX5, the enrichment of MDIG at CDC6 promoter was decreased only in MHCC-97H cells (Supplementary Fig. S6H). The mRNA expression of CDC6 was decreased mainly after decreasing CBX3 (Supplementary Fig. S6I and S6J). The result showed that CBX3 plays a major role in this recruitment in HCC cells. Interestingly, ZNF143 knockdown also decreased the enrichment of CBX3 in CDC6 promoter region (Supplementary Fig. S6K). In addition, MTT and colony formation assays showed that knocking down CDC6 inhibited the proliferation induced by MDIG overexpression (Fig. 5F–H; Supplementary Fig. S6L).

To further examine whether ZNF143 promotes CDC6 expression by activating MDIG, we first examined the expression of MDIG, CDC6, and H3K9me3 after ZNF143 overexpression and knockdown by Western blot analyses. As shown in Fig. 6A and B, MDIG and CDC6 were upregulated and H3K9me3 was decreased in ZNF143-overexpressing HCC cells. After ZNF143 knockdown, the opposite effects were observed (Fig. 6C and D). In addition, similar results were observed in ZNF143-overexpressing mouse tumor xenografts (Supplementary Fig. S6M). ZNF143 upregulated the expression of MDIG and CDC6 in xenograft tumor tissue derived from Li-7 cells with ZNF143 overexpression and MHCC-97H cells with ZNF143 knockdown were also demonstrated by IHC analysis (Supplementary Fig. 7A). To determine the role of MDIG in the upregulation of CDC6 induced by ZNF143, we knocked down MDIG under conditions of ZNF143 overexpression. The results showed that the upregulation of CDC6 induced by ZNF143 overexpression was inhibited and H3K9me3 was increased by MDIG knockdown (Fig. 6E and F; Supplementary Fig. S7B and S7C). Meanwhile, the enhancement of tumor growth, cell proliferation, and colony formation induced by ZNF143 overexpression were inhibited by MDIG knockdown in HCC cells (Fig. 6G–I; Supplementary Fig. S7D–S7F). The protein levels of ZNF143 and MDIG in mouse tumor tissues were examined by Western blot analyses (Supplementary Fig. S7G). In addition, the decreased proliferation and clonogenicity induced by shZNF143 could be rescued by MDIG overexpression in ZNF143-knockdown cells (Supplementary Fig. S7H–S7K).

Given that ZNF143 increased CDC6 expression by upregulating MDIG, we first analyzed the expression of MDIG and CDC6 in HCC and matched adjacent nontumor liver tissues in TCGA dataset and our HCC patient tissues. As shown in Supplementary Fig. S8A and S8B, the relative mRNA expression of MDIG and CDC6 was higher in most HCC tissues than their matched nontumorous liver tissues. Kaplan–Meier survival analysis indicated that patients with high expression of MDIG or CDC6 showed significantly worse overall survival in TCGA dataset (Supplementary Fig. S8C). The mRNA level of ZNF143 was positively correlated with MDIG and CDC6 expression in HCC tissues (Fig. 7A and B). The positive correlations among ZNF143, MDIG and CDC6 at the protein level were also demonstrated by IHC analysis of 211 patients with HCC (Fig. 7C and D; Supplementary Fig. S8D). Intriguingly, the ZNF143 and MDIG protein abundance fluctuated conformably during the cell-cycle progression in HEK-293T cells (Supplementary Fig. 8E). These data confirmed that ZNF143 expression was correlated with MDIG and CDC6 expression and was clinically relevant in HCC. Taken together, our data support a role for the ZNF143/MDIG/CDC6 axis in promoting HCC progression and suggest that ZNF143 might serve as a biomarker and potential target for HCC diagnosis and therapy (Fig. 7E).

Dysregulation of the cell cycle leading to hyperactive cell division is often found in cancers, and sustained proliferative signaling is a well-recognized hallmark of malignancy (39, 40). Pharmacologic inhibitors of cyclin-dependent kinases have shown promising effects in patients with breast and other cancers (41). In this study, our data showed that ZNF143 promotes HCC cell proliferation, colony formation, tumor growth, and cell-cycle transition.

Epigenetic alterations associated with cell-cycle disorders are common phenomena in various cancers, including HCC (32, 33). A series of studies have shown that transcription factors work in concert with histone modification enzymes to exert their function as gene regulators (18, 42). ZNF143 is a novel chromatin-looping factor that contributes to the architectural foundation of the genome by providing sequence specificity at promoters (43). The CDC6 protein is essential for initiation of DNA replication and is overexpressed in various cancers (25, 44). In this study, we found that CDC6 plays a role in cell proliferation of HCC. Intriguingly, our results showed that ZNF143 regulates the expression of CDC6 through epigenetic alteration of H3K9me3 modification by activating MDIG. MDIG is a cell growth regulating gene that contains a JmjC domain and suppresses the formation of H3k9me3 (14, 45). We identified MDIG as a direct target of ZNF143. ZNF143 overexpression significantly promoted MDIG expression and decreased H3K9me3 expression in HCC. Our previous study showed that MDIG overexpression promoted HCC cell proliferation, cell migration and invasion (21). In this study, we further demonstrated that MDIG enhances HCC cell cycle progression. Our result showed that beyond CDC6, further genes regulating S-phase entry can be regulated by MDIG and H3K9me3, we believed that MDIG-dependent histone demethylation of promoters of S-phase genes might a general mechanism to control gene expression in the S phase. In addition, we demonstrated that mainly CBX3 mediate recruiting MDIG to CDC6 promoter. Heterochromatin protein 1 (HP1) proteins interact with other molecules were originally identified as critical components in heterochromatin-mediated gene silencing (37). Interestingly, our result showed HP1 proteins did not exert transcriptional repression on CDC6 expression. After knocking down HP1, proteins, especially CBX3, slightly decreased CDC6 expression. In addition, whether other molecular medicated the recruitment MDIG to the CDC6 promoter will be conducted further research in the future. All these results further elucidate the role of ZNF143.

Our study also showed that ZNF143 is overexpressed in HCC tissue compared with adjacent normal tissue. Furthermore, by analyzing the clinical features, we identified ZNF143 as a potential biomarker of reduced survival in HCC. Paek and colleagues demonstrated that IGF-1 induces the expression of ZNF143 in colon cancer cells through PI3K and reactive oxygen species (4). Whether wortmannin, an inhibitor of PI3K, diphenyleneiodonium (DPI), an NADPH oxidase inhibitor, and monodansylcardavarine (MDC), a receptor internalization inhibitor, have therapeutic effects in HCC with high ZNF143 expression needs to be studied in the future. More importantly, Haibara and colleagues discovered two novel small molecules that inhibit ZNF143 activity (10). All these discoveries provided the opportunity that ZNF143 might be a therapeutic target during HCC progression. In addition, future studies should also aim at characterizing the protein complexes engaged by ZNF143 in order to provide other rational for novel drugs that can inhibit the function of ZNF143 in HCC cells.

To the best of our knowledge, our report is the first attempt to validate the oncogenic role of ZNF143 in HCC. Our data suggested that ZNF143 is a prognostic biomarker and is overexpressed in HCC tissues. ZNF143 promotes proliferation by activating cell-cycle progression in HCC cells. Importantly, we proposed a novel model for the ZNF143–MDIG–CDC6 regulatory axis, which might provide insight into the function of ZNF143 in HCC development.

No potential conflicts of interest were disclosed.

Conception and design: L. Zhang, H. Li, J. Li

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Ge, F. Zhao, Q. Zhou, X. Chen, T. Chen, H. Xie, Y. Cui, M. Yao

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Zhang, Q. Huo, H. Tian, H. Li, J. Li

Writing, review, and/or revision of the manuscript: L. Zhang, Q. Huo, H. Li, J. Li

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L. Zhang, Q. Huo

Study supervision: J. Li

This work was supported, in part, by grants from the National Key Program for Basic Research of China (973; 2015CB553905), National Natural Science Foundation of China (81972580, 81773152, 81672832), the National Key Sci-Tech Special Project of China (2018ZX10723204-006), and Key Discipline and Specialty Foundation of Shanghai Municipal Commission of Health and Family Planning (2018BR20).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.