Cell division cycle associated 2 (CDCA2) upregulation promotes the progression of hepatocellular carcinoma in a p53-dependant manner

Ethics approval

This study was approved by the Ethics Committee of Yantai Yuhuangding Hospital (IRB no.: YHD 2018-151). Written informed consent was obtained from all individual participants included in the study prior to study commencement for enrollment and for the use of patient clinical information for publication.

Molecular cloning

The CDCA2 overexpression plasmid was constructed by inserting the 3,072-bp coding sequence of CDCA2 (GenBank accession number: NM_152562.4) into the p3xFlag-CMV−7.1 vector. The CYPJ overexpression plasmid was constructed by inserting the 1178-bp coding sequence of PPIL3 (GenBank accession number: NM_032472.4) into the pCMV-myc vector. A GV493 vector containing three CDCA2-specific shRNA sequences was constructed to knock down CDCA2 (shCDCA2-1: gcAAACCTTTCAGAGGAGAAA, shCDCA2-2: gcCGTTCTCAGTTCTCCTAAT, shCDCA2-3: ccACAGTAACCGTAGAGCAAT).

RNA extraction and real-time quantitative PCR (qRT-PCR)

Pathologically confirmed specimens were collected during surgery from patients treated at Yantai Yuhuangding Hospital (Shandong, China), total RNA was extracted, and qRT-PCR analysis was performed on an FTC-3000P Real-Time Quantitative Thermal Cycler (Funlyn, Canada) as previously described (Wang et al., 2021). All the experiments were performed at least in triplicate. The primer sequences used for qRT-PCR were: CDCA2-F: 5′-AAAGGAGGAACACCTGTTTGTA-3′, CDCA2-R: 5′-CCTGAAAGGGTCTCAGAGATTG-3′, β-Tubulin-F: 5′-GGACCGCATCTCTGTGTACTA-3′, β-Tubulin-R: 5′-CTTTGGCCCAGTTGTTACCT- 3′, p21-F: 5′-GTGGACCTGTCACTG TCTTG- 3′, p21-R: 5′-TTAGGGCTTCCTCTTGGAGA- 3′, TP53-F: 5′-GGAAGAGAAT CTCCGCAAGAAA- 3′, TP53-R: 5′-CTCATTCAGCTCTCGGAACATC- 3′, CYPJ-F: 5′-GGAAGAGGAGGCAACAGTATTT- 3′, CYPJ-R: 5′-GGCAACTTCTCCAACTCATCTA-3′.

Western blotting

Total proteins were extracted from cells lysed with RIPA lysis buffer (Beyotime, China) and subjected to Western blotting analysis using rabbit/mouse primary antibodies from Affinity Biosciences, USA [anti-CDCA2 (DF3543), anti-Cyclin E1 (AF6235), anti-GAPDH (T0004), anti-β-Actin (T0022), anti-β-Tubulin (T0023)], or Cell Signaling Technology, USA [Cell Cycle Regulation Antibody Sampler Kit (9932T), Stress and Apoptosis Antibody Sampler Kit (8357T), Epithelial-Mesenchymal Transition (EMT) Antibody Sampler Kit (9782T)], and HRP-conjugated goat anti-rabbit/mouse secondary antibodies (ZSGB-BIO, China). The immunoblots were developed with ECL reagent (Thermo Scientific, USA). Densitometric analysis was performed to determine the grayscale intensity of the target bands using ImageJ software (imagej.nih.gov/ij).

Immunohistochemistry (IHC)

IHC was performed as previously described (Lin et al., 2017) except that HRP-conjugated secondary antibodies (ZSGB-BIO, China) were used for staining and a DAB Kit (ZSGB-BIO, Beijing, China) was used for color development.

Cell culture and transfection

HCC cell lines were obtained from iCell Bioscience (China). The cells were routinely cultured in MEM (iCell Bioscience, Shanghai, China) for HepG2 and SK-Hep1 cells or in DMEM (Biological Industries, Beit-Haemek, Israel) containing nonessential amino acids (Invitrogen, Waltham, MA, USA) for Huh7 cells, and the media were supplemented with 10% fetal bovine serum (Biological Industries, Beit-Haemek, Israel). The cells were cultured at 37 °C with 5% CO₂ humidified air according to standard procedures.

Transient plasmid transfection was performed with XtremeGENE HP DNA Transfection Reagent (Roche, Switzerland) following the manufacturer’s instructions.

Lentivirus-mediated CDCA2 knockdown

Huh7 and SK-Hep1 cells were both infected with lentivirus carrying the GV493 vector with CDCA2-specific shRNA (shCDCA2, target sequence: gcAAACCTTTCAGAGGAGAAA) or a nontargeting control sequence (shNC, sequence: TTCTCCGAACGTGTCACGT) at MOIs of 5 and 30, respectively, and uninfected cells were used as the control. After three days of growth, the cells were observed with a fluorescence microscope to confirm the expression of EGFP in >90% of cells. Puromycin dihydrochloride (MedChemExpress, Monmouth Junction, NJ, USA) was added to the complete growth media at a final concentration of 8 µg/mL for Huh7 cells and 6 µg/mL for SK-Hep1 cells, and this medium was refreshed every 3 days until all the uninfected cells had died and no more death occurred in the infected cells. The cells were then digested and seeded in 96-well plates at a concentration of 1 cell/well and continued to grow until stable strains were obtained.

CCK-8 assay

The proliferation of cells was measured using Cell Counting Kit-8 (Dojindo Laboratories, Japan) as previously described (Wang et al., 2021).

Cell viability assays were performed by treating the cells (5 × 10³/well) with cisplatin (CDDP) in DMSO for 48 hrs. An equal volume of DMSO was added to the control wells. Each sample was analyzed in triplicate. The OD₄₅₀ was measured after 2 hr of incubation at 37 °C with 100 µL serum-free MEM containing 10 µL CCK-8 solution. Cell viability = OD_sample/OD_control × 100%.

Transwell migration assay

Transwell migration assays were performed as previously described (Yang et al., 2020) with 4 × 10⁴ cells seeded per 8-µm Transwell chamber (Corning, Corning, NY, USA).

Colony formation assay

Colony formation assays were performed as previously described (Yang et al., 2020).

Tunel

TUNEL assay was performed as previously described (Wang et al., 2021).

Flow cytometry

Flow cytometry analysis of cell cycle and apoptosis was performed on a Moflo XDP flow cytometer (Beckman Coulter, Brea, CA, USA) using the Cycletest™ Plus DNA Reagent Kit (BD Biosciences, USA) and FITC Annexin V Apoptosis Detection Kit I (BD Biosciences, San Jose, CA, USA).

Bioinformatics analysis

Clinical data and RNA-seq v2 data from The Cancer Genome Atlas (TCGA) database (http://cancergenome.nih.gov/) were retrieved from cBioPortal (http://www.cbioportal.org/) and analyzed. The CDCA2 transcription level and promoter methylation level in HCC tissues were compared using the UALCAN database (http://ualcan.path.uab.edu/) (Chandrashekar et al., 2017). Gene function enrichment analysis was performed using the Gene Ontology Consortium (GO) (http://geneontology.org/).

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8.0 software. The log-rank (Mantel-Cox) test was used to compare Kaplan–Meier survival curves. Numerical data with a normal distribution were compared by Student’s t-test for two samples or one-way ANOVA for three or more samples. Categorical data were compared with Fisher’s exact test. Ratios were compared by Kruskal–Wallis test for two samples or Mann–Whitney test for three or more samples.

Elevated CDCA2 expression in HCC was associated with malignant features and worse prognosis

CDCA2 expression in HCC tissues and patient survival were first analyzed using RNA-seq data from UALCAN. The results showed a significantly higher transcription level of CDCA2 in HCC tissues (n = 371) than in normal tissues (n = 50) (P < 0.0001, Fig. 1A), and CDCA2 expression increased with clinical stage (Fig. 1B). Interestingly, CDCA2 expression was significantly higher in patients expressing mutant TP53 than in patients expressing wild-type TP53 (P < 0.0001) (Fig. 1C). The survival rate of the high CDCA2 group was worse than that of the low CDCA2 group in the complete dataset, which included all ethnic groups (P < 0.0001), and in Asian patients (P < 0.0001), but this trend was not observed in White Caucasian patients (P = 0.3846) (Figs. 1D and 1E).

The correlations between CDCA2 mRNA expression and the clinicopathological characteristics of HCC patients were analyzed using the TCGA LIHC_pan_caner_atlas_2018 dataset. The results suggested correlations between high CDCA2 expression and lower age (P = 0.0213), advanced clinical stage (P = 0.0001), and higher tumor grade (P = 0.0011). In addition, CDCA2 expression in Asian patients was significantly higher than that in non-Asian patients (P = 0.0405), and CDCA2 was more frequently mutated in the high expression group (P = 0.0081) (Table 1).

We confirmed the mRNA expression level of CDCA2 in 40 pairs of HCC tissues and normal liver tissues by qRT-PCR, and the results showed higher expression in HCC tissues than in normal tissues (P < 0.001) (Fig. 1F). CDCA2 expression was also studied in 48 paraffin-embedded HCC tissues by IHC (Fig. 1G). The correlations between CDCA2 protein expression and the clinicopathological characteristics of the patients were analyzed, and the results showed an association of high CDCA2 protein expression with high tumor grade (P = 0.0405) and Glypican-3 positivity (P = 0.049) (Table 2). Although not statistically significant, there were higher percentages of high-CDCA2 samples in groups with advanced clinical stage and larger tumor size.

CDCA2 knockdown and overexpression in HCC cell lines

The levels of CDCA2 in the L02 liver cell line and three HCC cell lines, namely, Huh7, SK-Hep1 and HepG2 cells, were detected by qRT-PCR and Western blotting (Figs. 1H and 1I). All three HCC cell lines had significantly higher CDCA2 expression than the L02 cell line. CDCA2 expression was highest in Huh7 cells and lowest in HepG2 cells. Therefore, CDCA2 was transiently overexpressed in HepG2 cells (HepG2-pCDCA2), and cells transfected with an empty vector served as a control (HepG2-p3xFlag). Three CDCA2-specific shRNAs were transfected into Huh7 cells to achieve transient knockdown of CDCA2. shCDCA2-1 and shCDCA2-2 appeared to have similar knockdown efficiency, both better than shCDCA2-3 (Fig. 1I), and shCDCA2-1was packaged into lentivirus to achieve the stable knockdown of CDCA2 in the Huh7 and SK-Hep1 cell lines (Lv-shCDCA2) for subsequent functional and mechanistic studies, and cells transduced with nontargeting sequences served as a control (Lv-shNC). The alteration of CDCA2 expression was confirmed by qRT-PCR and Western blotting (Figs. 1H and 1J).

CDCA2 promoted HCC cell proliferation and migration in vitro

A CCK-8 assay was used to investigate the role of CDCA2 in the proliferation of HCC cells. The results showed that knockdown of CDCA2 significantly attenuated the growth of Huh7 and SK-Hep1 cells, while overexpression of CDCA2 promoted the proliferation of HepG2 cells (Fig. 2A). Cell clonogenic ability was also studied, and the results suggested that CDCA2 overexpression led to an increased number of colonies formed by HepG2 cells (P = 0.0136) (Fig. 2B). These results confirmed the pro-proliferative role of CDCA2 in HCC cells.

We then investigated the role of CDCA2 in HCC cell migration by transwell assay and found reduced migration capacity after CDCA2 was knocked down in Huh7 cells (Fig. 2C). However, cell migration was not significantly affected in SK-Hep1-Lv-shCDCA2 or HepG2-pCDCA2 cells compared to their controls.

We then investigated the expression of key EMT markers in HCC cells using Western blotting. The results showed increased levels of E-cadherin and decreased levels of Vimentin and the E-cadherin suppressor ZEB-1 in CDCA2-knockdown Huh7 cells compared to control cells, suggesting inhibition of the EMT pathway, and increased levels of N-cadherin in CDCA2-overexpressing HepG2 cells, indicating increased mesenchymal features. The levels of EMT markers were unchanged, or not expressed, in CDCA2-knockdown SK-Hep1 cells (Fig. 2D).

CDCA2 promoted the G1/S transition of HCC cells possibly by the upregulation and activation of CCND1/CDK4/6 and CCNE1/CDK2

The cell cycle progression of the three HCC cell lines was studied by flow cytometry. After CDCA2 knockdown, an increase in the proportion of Huh7 cells in the G1 phase was observed, whereas CDCA2 overexpression caused a decrease in the proportion of HepG2 cells in the G1 phase and an increase in the S phase (Fig. 2E). CDCA2 knockdown did not significantly affect the cell cycle progression of SK-Hep1 cells. These data suggested that CDCA2 could promote the transition from the G1 phase to the S phase in the cell cycle.

We examined the expression of key components of cell cycle regulation, cyclins (CCNs) and cyclin-dependent kinases (CDKs), by Western blotting. The results showed reduced levels of CCND1/CCNE1/CDK2/CDK4 in CDCA2-knockdown Huh7 cells, reduced levels of CCND1/CCNE1 in CDCA2-knockdown SK-Hep1 cells and increased levels of CCND1/CCND3/CCNE1/CDK2/CDK6 in CDCA2-overexpressing HepG2 cells (Fig. 2F). We also explored the mRNA expression of cyclins and CDKs in the TCGA LIHC_pan_caner_atlas_2018 dataset, which includes 366 patients, and found that CDCA2 expression was strongly or moderately correlated with the expression of CCNE1 (r = 0.635, P < 0.0001), CDK2 (r = 0.552, P < 0.0001) and CDK4 (r = 0.547, P < 0.0001) but was not correlated with the expression of CCND1, CCND3 and CDK6. Taken together, our results suggested that CDCA2 might promote the G1/S phase transition of HCC cells in part by upregulating CCND1/CDK4/6 and CCNE1/CDK2 expression.

To further understand the mechanism by which CDCA2 promotes the growth of HCC, the expression of the CDK inhibitors p21, p27, and p18 in HCC cells was detected by qRT-PCR and/or Western blotting. We observed upregulation of p18 and p27 expression and a lack of p21 expression in Huh7-Lv-shCDCA2 cells and the upregulation of p21 expression and downregulation of p18 and p27 expression in HepG2 cells (Figs. 2F and 2G).

CDCA2 played an antiapoptotic role in HCC cells by inactivating the p38 MAPK pathway and activating the JNK/c-Jun pathway

The apoptosis of CDCA2-overexpressing HepG2 cells was studied by flow cytometry. Apoptosis was induced by treatment with 5 mg/L CDDP for 24 hrs. In both the CDDP-treated and untreated groups, CDCA2 overexpression protected HepG2 cells from apoptosis, especially early apoptosis (Fig. 3A). The apoptosis of CDCA2-knockdown Huh7 and SK-Hep1 cells was investigated by TUNEL staining. Apoptosis was induced by treatment with 10 mg/L and 20 mg/L CDDP for 24 hrs. The results showed that Huh7 and SK-Hep1 cells were not as sensitive to CDDP-induced cell death as HepG2 cells. In particular, the apoptosis of Huh7 cells was not induced by CDDP at concentrations up to 20 mg/L, but this resistance was abolished by CDCA2 knockdown (P = 0.0418) (Fig. 3B). Treatment with 20 mg/L CDDP induced the apoptosis of SK-Hep1 cells (P = 0.0006 for the control group, P = 0.0049 for the CDCA2-knockdown group), and CDCA2 knockdown enhanced the chemosensitivity of SK-Hep1 cells to CDDP (P = 0.0435) (Fig. 3C).

Western blotting analysis of stress and apoptosis markers showed no change in the Caspase 3/cleaved PARP and p38 MAPK/MAPKAPK2 pathways and inactivation of the JNK/c-Jun and p53 pathways in Huh7-Lv-shCDCA2 cells compared to the control cells. In SK-Hep1-Lv-shCDCA2 cells, there were increased levels of Caspase 3/cleaved PARP, activation of the p38 MAPK/MAPKAPK2 pathway, and inhibition of the JNK/c-Jun and p53 pathways. In CDCA2-overexpressing HepG2 cells, we found activation of p53 and JNK/c-Jun and inactivation of p38 MAPK (Fig. 3D). The data described above suggested that CDCA2 plays antiapoptotic and proliferative roles, possibly through the inactivation of the p38 MAPK pathway and the activation of the JNK/c-Jun pathway in HCC cells.

GO analysis of genes correlated with CDCA2 in HCC

Genes correlated with the expression of CDCA2 were pooled from cBioPortal, and of these genes, 386 genes with a Pearson’s correlation coefficient of ≥0.6 were analyzed by GO for functional enrichment analysis. The enriched molecular functions among these proteins included DNA binding, methylated histone binding, protein kinase activator activity, microtubule binding, cyclin-dependent protein Ser/Thr kinase activity, etc. The pathways positively enriched among these genes were DNA replication (19.37-fold enrichment, FDR = 2.36E−08) and p53 pathways (6.27-fold enrichment, FDR = 8.11E−04). The enriched reactome pathways included mitotic metaphase/anaphase transition, G2/M DNA replication checkpoint, mitotic G1 phase and G1/S transition, cyclin A/B1/B2-associated events during G2/M transition, CHK1/CHK2 (Cds1)-mediated inactivation of the cyclin B:CDK1 complex, TP53 regulation of cell cycle gene transcription, regulation of TP53 activity through phosphorylation, etc.

CDCA2 expression was regulated by CYPJ

Our research group previously discovered that the peptidyl-prolyl cis/trans-isomerase (PPIase) CYPJ could promote the growth of HCC by preventing G1 phase arrest by activating the cyclin D1 promotor (Chen et al., 2015). We therefore investigated the possible relationship between CDCA2 and CYPJ in HCC cells. We measured the mRNA and protein levels of CDCA2 and CYPJ in three HCC cell lines with altered CDCA2 or CYPJ expression (Figs. 4A and 4B). Our results showed no change in the CYPJ protein levels in CDCA2-knockdown or CDCA2-overexpressing HCC cells. However, overexpressing CYPJ significantly upregulated the expression of CDCA2. These results suggested that CYPJ might be a regulator of CDCA2.

We then compared SK-Hep1 and HepG2 cells overexpressing CDCA2 with control cells treated with different concentrations of cyclosporine A (CsA), a PPIase inhibitor of most cyclophilins. The results showed that CDCA2 could alleviate the inhibition of cell growth caused by reduced cyclophilin activity (Figs. 4C and 4D).