Increased Expression of SETD7 Promotes Cell Proliferation by Regulating Cell Cycle and Indicates Poor Prognosis in Hepatocellular Carcinoma

Purpose To investigate the role of SET domain containing 7 (SETD7) in hepatocellular carcinoma (HCC) and determine whether SETD7 can be used as a predictor of overall survival in HCC patients. Methods mRNAs and proteins of SETD7 and related genes in HCC tumor samples and paired adjacent non-tumorous liver tissues (ANLTs) (n = 20) or culture cells were determined by quantitative real-time PCR and Western blot. Cell proliferation and apoptosis with SETD7 knockdown SMMC-7721 cells or SETD7 overexpressed HepG2 cells were analyzed by CCK8 assay or flow cytometry. Gene expression alterations in SETD7 knockdown of SMMC-7721 cells were determined by digital gene expression (DGE) profiling. Defined data on patients (n = 225) with HCC were retrieved for the further study. Tissue microarrays (TMAs) were performed using paraffin tissues with tumor and ANLTs. SETD7 and related proteins were determined by TMAs immunohistochemistry. Statistical analyses were conducted to associate SETD7 expression with tumor features and patient outcomes, as well as related proteins expression. Results SETD7 expression was significantly higher in HCC tumor tissues than in ANLTs. SETD7 overexpression in vitro can promote HepG2 cell proliferation, whereas SETD7 knockdown can inhibit SMMC-7721 cell proliferation by regulating the cell cycle. SETD7 expression was significantly correlated with five genes expression. Increased SETD7 is associated with metastasis, recurrence, large tumor size, and poor tumor differentiation, and indicates poor prognosis in HCC patients. Conclusions SETD7 plays a critical role in HCC, and its immunohistochemistry signature provides potential clinical significance for personalized prediction of HCC prognosis.


Introduction
Changhai Hospital between 2009 and 2013. The patients had been diagnosed with HCC according to the WHO Classification of Tumor of the Digestive System. Clinical data, including patient characteristics, clinical presentation, tumor differentiation, sites of lesion, laboratory findings, objective response, and survival were collected from the hospital information system. The summary of clinicopatholgic characteristics are in S1 Table. All patients provided informed written consent for sample collection and permission to use for research purposes. The protocol for all experiments was approved by Ethics Committee of the Second Military Medical University.

RNA extraction, reverse transcription and qRT-PCR
The total RNA was isolated from tissues or cells using the TRIzol reagent (Invitrogen). The cDNA was synthesized following the manufacturer's instructions (ThermoFisher). qRT-PCR of SETD7, zinc finger and BTB domain containing 20 (ZBTB20), cyclin-dependengt kinase inhibitor 2D (CDKN2D) were performed using FastStart Universal SYBR Green Master (Rox) (Roche). qRT-PCR assay were performed using an ABI quantitative PCR model 7300 (Applied Biosystems). Primers for SETD7, ZBTB20, CDKN2D, GAPDH were designed as follows S2 Table. GAPDH was used as internal reference. The qRT-PCR was performed in triplicate and included no-template controls. Relative expression was calculated with the comparative CT method.
SETD7 silencing sequences and overexpression plasmid 3 interfering RNAs (siRNA) targeting human SETD7 and the scrambled siRNA were designed and synthesized by GenePharma (Shanghai, China) (S4 Table). The SETD7 overexpression or control plasmid was constructed by Genechem as S1 Fig.

Cell proliferation, cell cycle and apoptosis analysis
For cell proliferation, SMMC-7721 cells were plated at a concentration of 5×10 3 cells each well in 96-well plates that media containing 10% FBS, after incubation for 24 h, the cells were transfected with si-SETD7-2 or the scrambled RNA at a final concentration of 100 nM using FuGENE HD Transfection Reagent (Roche, Basel, Switzerland), according to the manufacturer's protocol. Meanwhile, HepG2 cells were transfected with pGV141-SETD7 or pGV141 plasmid in similar way (0.2 μg plasmid/ per well). Cell proliferation assay was performed by Cell Counting Kit-8 (CCK-8) solution (Obio technology, Shanghai, China) according to the manufacturer's protocol. For cell cycle and apoptosis analysis, cells were plated at a concentration of 5×10 5 cells each well in 6-well plates, and transfected in the same way. After incubation for 36 h, cells were trypsinized, collected after washing twice with phosphate-buffered saline, fixed in 75% cold ethanol, then follows cell cycle and apoptosis analysis protocol (BD) and performed using a FACS cytometer (FACS, CA, USA).These experiments were performed in triplicate.

Digital gene expression profiling (DGE) and bioinformatics
After SMMC-7721 cells was transfected with si-SETD7-2 or the scrambled RNA for 24 h, the cells were harvested in TRIzol reagent (SETD7 knockdown efficiency were confirmed) then sent to DGE analysis. Sequencing was performed at RiboBio Co., Guangzhou through the illumineHi-Seq2500. Data was aligned to the Ensembl transcript annotations using bowtie and RSEM as described previously [19]. All other bioinformatics analysis was performed using glbase [20].

Statistical analysis
Results were analyzed with SPSS 21.0 statistical software. Student's t test was used to evaluate the quantitative variables. Pearson test or correction for continuity chi square test was used to evaluate the correlation between SETD7 expression and clinicopathologic parameters. Kaplan-Meier analysis was used to estimate the survival probability. Log-rank test was used to estimate the comparison of survival curves between groups. Univariate and multivariate Cox regression analysis was used to determine contribution of SETD7 expression to the survivals. Significance was defined as p< 0.05.

Results
Expression of SETD7 in HCC tumor tissues, paired ANLTs and cell lines SETD7 exhibited a markedly higher expression in HCC tumor tissues than in ANLTs, as determined by qRT-PCR ( Fig 1A). This finding was verified by Western blot analysis ( Fig 1B); the gray value of the Western blot result was calculated ( Fig 1C). SMMC-7721, QGY-7703, Bel-7404, HCC-0010, HepG2, and HL-7702 cells were detected by Western blot analysis to determine the SETD7 expression in different liver cancer cell lines and the normal liver cell line. The results exhibit as following (in descending order of their SETD7 levels): SMMC-7721, Bel-7404, QGY-7703, and HCC-0010. Almost negative results were obtained in HepG2 and HL-7702 ( Fig 1D). IHC assay revealed that SETD7 is significantly higher in HCC tumor tissues than in ANLTs ( Fig 1E). The IHC scores of SETD7 in HCC tumor tissues and ANLTs are indicated in the box plots ( Fig 1F).

SETD7 expression changes liver cancer cell proliferation by regulating the cell cycle
Considering the higher expression of SETD7 in HCC tumor tissues than in ANLTs, we determined whether SETD7 would affect cell function. Three interference sequences S1, S2, and S3 were designed to knock down SETD7 in SMMC-7721. SETD7 overexpression plasmid was constructed and transfected into HepG2 cells. Interfering efficiency and overexpression capacity were detected by Western blot analysis after transfection for 24 h; specifically, three interference sequences obtained almost same knockdown efficiency (around 40-50%) (Fig 2A and 2B). CCK8 assay showed that SETD7 knockdown inhibited SMMC-7721 cell growth, whereas SETD7 overexpression increased HepG2 cell growth, compared with the control group ( Fig 2C and 2D).
To explore the approach of SETD7 in regulating cell proliferation, apoptosis was detected in SMMC-7721 and HepG2. Apoptosis rates were not altered significantly in SETD7 knockdown SMMC-7721 cells or in SETD7 overexpressed HepG2 cells, compared with the control group (data not shown). Cell-cycle assay revealed that cells in the G1 phase were significantly decreased in SETD7 knockdown SMMC-7721 cells, compared with the control (Fig 2E); this result showed an opposite effect on HepG2 cells with SETD7 overexpressed (Fig 2F).

DGE analysis of SETD7 knockdown in SMMC-7721 cells
To elucidate the mechanism underlying SETD7 regulation of cell proliferation, DGE analysis was performed in SETD7 knockdown SMMC-7721 cells by transfecting the S2 interference sequence. The results showed that ZBTB20, zinc finger and SCAN domain-containing protein 26 (ZSCAN26), and CDKN2D were significantly downregulated, whereas actin, gamma 2, smooth muscle, enteric (ACTG2) and carboxypeptidase A5 (CPA5) were significantly upregulated after SETD7 knockdown (Fig 3A). For ZBTB20 and CDKN2D correlate with cell cycle, the mRNA and protein levels of ZBTB20 and CDKN2D were subsequently verified by qRT-PCR and Western blot analysis in SETD7 knockdown SMMC-7721 cells and in SETD7 overexpressed HepG2 cells (Fig 3B-3E). SETD7 correlated with H3K4me2, ZBTB20, and CDKN2D via TMA IHC analysis To verify whether SETD7 is more highly expressed in HCC tumor tissues than in ANLTs and that SETD7 correlates with ZBTB20 and CDKN2D in vitro, TMA IHC staining for SETD7, H3K4me2, ZBTB20, and CDKN2D was performed on the 225 pairs of HCC tumor tissues and ANLTs (Fig 4). SETD7, CDKN2D and ZBTB20 are widely expressed in the nuclei and cytoplasms of HCC specimens. H3K4me2 immunoreactivity was detected in the nuclei. As expected, SETD7, H3K4me2, CDKN2D, and ZBTB20 are significantly higher in HCC tumor tissues than in ANLTs (Fig 4). The expression of SETD7 is closely related to the expression of ZBTB20 (P = 0.002), H3K4me2 (P = 0.04), and CDKN2D (P = 0.004) in HCC (Table 1).

Relationship between SETD7 expression and clinicopathologic parameters
To further investigate the clinical significance of SETD7 expression in the development and progression of HCC, primary HCC samples were divided into two groups, low expression of SETD7 (IHC score: 0-2, n = 74) and high expression of SETD7 (IHC score: 3-6, n = 151). Analysis indicated high expression of SETD7 in HCC tumor tissues is significantly correlated with several crucial clinicopathologic parameters, such as metastasis (P = 0.009), recurrence (P = 0.012), poor tumor differentiation (P = 0.047), and large tumor size (P = 0.001), which are associated with the increased risk of early recurrence and reduced survival in HCC patients ( Table 2) [21].

Prognostic values of the IHC signature
Kaplan-Meier analyses were performed to investigate the relationship between SETD7 expression and clinical prognosis. Patients with high SETD7 expression exhibited a significantly shorter overall survival (OS) than those with low SETD7 expression (Fig 5) (Table 3).

Discussion
In this study, SETD7 is more highly expressed in HCC tumor tissues than in ANLTs. To identify the biological significance of SETD7, SETD7 was knockdown in SMMC-7721 cells by small interfering RNA, for SMMC-7721 cell line is basically high expression of SETD7. On the other hand, SETD7 was overexpressed in HepG2 cells through transfecting overexpression plasmid, for HepG2 cell line is basically low expression of SETD7. Furthermore, SETD7 is significantly correlated with the expression of ZBTB20, CDKN2D, ZSCAN26, ACTG2, and CPA5. ZBTB20 (also known as DPZF, HOF, and ZNF288), which belongs to the BTB/POZ zinc finger family of proteins [22], represses forkhead box O1 (FoxO1) expression in lung cancer cells and promotes cell growth. Overexpression of ZBTB20 exhibits a significantly increases percentage of cells in the S phase and reduced the percentage of cells in the G1/G0 phase [23]. CDKN2D (also named p19 INK4d ) which belongs to the INK4 family of cyclin-dependent kinase inhibitors, can bind to CDK4/6; overexpression of CDKN2D leads to cell-cycle arrest in both G1 and G2  phases [24]. By inference, SETD7 promoting cell proliferation is a complex process of multiple factors involved in. As the sequence-specific transcriptional repressor of alpha fetoprotein (AFP) [25], ZBTB20 knockout mice show some genes differentially involve in growth, metabolism, and detoxification in the liver [26]. Overexpression of ZBTB20 in HCC is strongly associated with recurrence, metastasis, and vein invasion and is also an independent prognostic factor for HCC [27]. In addition, overexpression of CDKN2D reverses the effect of miR-451, which contributes to esophageal carcinoma malignancy [28]. These results support that SETD7 regulating ZBTB20 and CDKN2D plays an important role in HCC.
ZSCAN26 also known as SREZBP, ZNF187, SRE-ZBP, is member of the C 2 H 2 zinc finger family of proteins exemplified by transcription factor IIIA and the Drosophila Kriippel protein, shows specific SRE-binding activity, and its molecular function conclude metal ion binding, nucleic acid binding and transcription factor activity [29,30]. ACTG2 links with E-cadherin via β-catenin. Recent studies identified several heterozygous missense variants in ACTG2 in megacystis-microcolon-intestinal hypoperistalsis patients had occurred de novo or were inherited [31][32][33][34]. The research of CPA5 is focus on mast cell diseases in zebrafish [35][36][37].
SETD7 regulation of the ZBTB20 and CDKN2D expression may involve complex mechanisms, such as methylation of both histone and non-histone proteins. Although the mechanism underlying such regulation remains unknown in the present study, which is our research limitation, the biological function and clinical significance of SETD7 in HCC still provide useful information.
Univariate and multivariate analysis revealed that high SETD7 expression, large tumor size and metastasis are associated with reduced OS. Previous studies showed that the size of tumors, recurrence and metastasis are important determinants of prognosis in HCC, small tumor diameter, without metastasis associated with relatively improved survival and prognosis [21].
Aberrations in H3K4 methylation occur in various cancers. Upregulation of MLL1 and SMYD3 increases H3K4 methylation, as well as induces breast and colorectal cancer, fibrosarcoma, and HCC [3]. However, there are some reports to the contrary, for example, high level of H3K4me2 is associated with increased survival for patients with large-cell or squamous-cell lung carcinoma [38], and low levels of H3K4me2 indicate poor survival in resectable pancreatic adenocarcinoma and metachronous liver metastasis of colorectal cancer [39]. We assume that H3K4me2 level in different tissue cells contribute differently to cancer progression. In the current study, H3K4me2 is more highly expressed in HCC tumor tissues than in ANLTs, and high H3K4me2 level in HCC tumor tissues is correlated with poor survival. Hence, H3K4me2 may be a dependent factor associated with SETD7, which is correlated with HCC development and progression.

Conclusions
This study determined that SETD7 plays a critical role in HCC by regulating the cell cycle, and SETD7 can be an independent prognostic factor for the OS of patients with HCC. These findings can provide new insights into the biological and clinical significance of SETD7 in HCC.