MCM6 promotes metastasis of hepatocellular carcinoma via MEK/ERK pathway and serves as a novel serum biomarker for early recurrence

Tumor tissue specimens, including tumor and matched adjacent non-tumor tissues, were obtained from 70 patients who had undergone curative liver resection. Detailed clinical pathological parameters were listed in Table 1. Thirty normal hepatic tissues were obtained from patients suffering from benign hepatic lesions who underwent resection. The serum samples, including 34 healthy people, 32 patients with cirrhosis, and 31 patients with HCC (obtained by collecting venous blood at the time of the primary diagnosis before operation and 8 days after surgical resection) were used in this study for measuring the serological level of MCM6 by ELISA. Serum samples were frozen and stored at −80 °C until measurement. All the selected HCC cases for serological examination had received curative hepatectomy and been provided with a uniform follow-up to the cohort. Curative hepatectomy was defined to involve (1) complete removal of all nodules with the resection margin greater than 10 mm, (2) the absence of invasion of the main trunk and first-order branches of the portal vein, common hepatic duct and its first-order branches or main trunk of the hepatic vein and inferior vena cava (3) the absence of intra- or extra-hepatic metastasis, and (4) the absence of residual tumor or portal tumor thromboses on postoperative imaging. In this study, all specimens were obtained and used under protocols approved by the Integrated Hospital of Traditional Chinese Medicine of Southern Medical University Office for Protection of Human Subjects.

MCM6 expression data based on RNA-Seq were extracted from The Cancer Genome Atlas (TCGA) database for 291 patients with HCC. Patients were divided into low and high expression groups, and the prognostic values of MCM6 for overall survival, cumulative recurrence, and early-recurrence were assessed.

HCC cell lines (HCC-LM3, SMMC7721, Huh7 and PLC/PRF/5) were obtained from the Cancer Research Institute of Southern Medical University in Guangzhou, China. HepG2 cells were purchased from the Chinese Academy of Sciences Cell Bank (Shanghai, China). All of the cells were routinely maintained in high-glucose DMEM (Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Life Technologies) at 37 °C with 5% CO₂.

Lentiviral particles expressing shRNA against MCM6 or control sequence were constructed by Cyagen Bioscience Inc. (Guangzhou, China). LM3 and SMMC-7721 cells were transfected with lentiviral vectors and selected with 750 mg/mL G418 (Gibco). MCM6 expression was confirmed by qPCR, and the levels of MCM6 protein were measured using western blotting.

Total RNA was isolated from cells using the E.Z.N.A. Total RNA Kit I (Cat. No 66834–02 Omega). Reverse transcription was performed using the PrimeScript RT Reagent Kit (TaKaRa Biotech, Dalian, China). qRT-PCR reactions were performed using a Bio-Rad system (Bio-Rad Labs, Hercules, CA) with SYBR Green PCR Master Mix (TaKaRa Biotech). The cycling conditions were denaturation at 95 °C for 15 s, with annealing at 60 °C for 15 s and fluorescence collection at 72 °C for 10 s. The 2 − ∆∆Ct method was used for relative quantification and statistical analysis. Independent experiments were done in triplicate. The primers used in this study are listed in the Additional file 1: Table S1.

Immunohistochemical staining of tissue was performed according to the manufacturer’s instructions (DAKO, Glostrup, Denmark) on formalin-fixed, paraffin-embedded tissue sections that had been cut to 4-μm thickness. The specimens were dried at 62 °C for 2 h and then dewaxed in xylene and rehydrated using graded alcohols, followed by incubation in 3% hydrogen peroxide for 15 min to exhaust the endogenous peroxidase activity. The antigens were then retrieved in Tris-EDTA (pH 9.0)/ 0.01 M sodium citrate buffer (pH 6.0) using a microwave oven for 25 min, followed by blocking with 10% goat serum for 30 min to prevent non-specific staining. Primary antibodies were incubated overnight at 4 °C in a humidified chamber, followed by HRP-labeled anti-mouse/rabbit secondary antibody (DAKO) incubation for 1 h at room temperature. Antibody binding was detected by DAB, and the reaction was stopped by immersion in distilled water once the brown color appeared. Finally, specimens were counter-stained with hematoxylin for 3 min. Two different pathologists who specialized in liver cancer evaluated the IHC results without knowledge of the clinical data. Both the extent and intensity of immunostaining were taken into consideration when analyzing the data. The intensity of staining was scored from 0 to 3, and the extent of staining was scored from 0% to 100%. Final quantitation of staining was obtained by multiplying the two scores. The protein expression was classified as high expression if the score was higher than 1.5 and low expression if the score was 1.5 or less. The antibodies are listed below: MCM6 (1:4000; Abcam, Cambridge, UK), Ki-67 (1:200, ZSGB-BIO, China), Vimentin (1:100, ZSGB-BIO, China) and E-cadherin (1:100, ZSGB-BIO, China).

Cells grown on coverslips were rinsed with phosphate-buffered saline (PBS) and fixed with cold 4% paraformaldehyde for 5 min at RT. Subsequently, the cells were blocked with Triton X-100 at a concentration of 0.2% for 30 min. Cells were then blocked for 1 h with 5% BSA and washed for 30 min, followed by incubation with primary monoclonal antibodies against MCM6 (1:500; Abcam, Cambridge, UK). Vimentin (1:100, ZSGB-BIO, China) and E-cadherin (1:50, ZSGB-BIO, China) overnight at 4 °C. The next day, the coverslips were incubated for 1 h in a dark room with Alexa Fluor 564 goat anti-rabbit IgG and Alexa Fluor 488 goat anti-mouse IgG (1:100 dilution; Bioworld Technology, Inc). Furthermore, the coverslips were stained with DAPI for 5 min at 4 °C. Finally, an LSM80 fluorescent microscope (ZESS, Germany) was used to observe the expression in cells.

Total proteins were extracted with RIPA lysis buffer supplemented with protease inhibitors, separated by SDS-PAGE and transferred to a PVDF membrane (Millipore, Billerica, MA). The membrane was blocked with 5% non-fat milk at room temperature for 1 h and incubated with the appropriate antibody. Protein bands were visualized by the chemiluminescent HRP detection system (Millipore, Billerica, MA). The antibodies are listed below: MCM6 (1:2000, Proteintech, Chicago, IL), E-cadherin (1:1000, CST, Boston, MA), Vimentin (1:1000, CST, Boston, MA), GAPDH (1:5000, Ray antibody, China), ZO1 (1:1000, CST, Boston, MA), Fibronectin (1:1000, Wanleibio, China), p-ERK1/2 (1:300, Wanleibio, China), ERK1/2(1:500, Wanleibio, China), p-MEK1/2 (1:500, ABclonal, China), and MEK1/2 (1:1000, Wanleibio, China).

Cell viability was analyzed using the Cell Counting Kit-8 (CCK-8) (Dojindo Molecular Technologies Inc. Shanghai, China). Cells at a density of 2 × 10³/well were seeded into 96-well plates and cultured in 100 μL of DMEM containing 10% FBS for 4 days. Ten microliters of CCK8 solution was added to each plate, and the cells were incubated for 3 h at 37 °C. The absorbance value (OD) of each well was measured at 450 nm.

For the EdU incorporation assay, proliferating cells were examined using the Cell-Light EdU Apollo 567 in vitro Imaging Kit (RiboBio, Guangzhou, China) according to the manufacturer’s protocol. Cells at a density of 5 × 10³/well were seeded into 96-well plates and incubated with 10 mM EdU for 2 h, followed by fixation with 4% paraformaldehyde, permeabilization in 0.2% Triton X-100 and staining with Apollo fluorescent dyes. Next, 50 μL/well DAPI was used to stain the cell nuclei for 10 min. The number of EdU-positive cells was counted under a fluorescent microscope in five random fields.

For the wound-healing assay, cells were grown to confluence in a six-well plate. Artificial wound tracks were created by scraping the confluent cell monolayers with a pipette tip. The cells were fed with serum-free medium. The ability of the cells to migrate into the wound area was assessed every 24 h after scratching.

Cells were plated in 6-well culture plates at 100 cells/well. After incubation for 2 weeks at 37 °C, the cells were washed twice with PBS and stained with 0.1% crystal violet solution. The number of colonies containing ≥50 cells was counted under a microscope. The colony formation efficiency was calculated as (number of colonies/number of cells inoculated) × 100%.

Cell motility was assessed by cell migration and invasion assays using transwell chambers with or without Matrigel (BD, Biosciences, CA). Approximately 6 × 10⁴ cells in medium without FBS were seeded on transwell chambers with or without Matrigel and incubated at 37 °C for 15 h. Medium containing 10% FBS was put in the lower chamber. The invasive cells attached to the lower surface of the membrane insert were fixed, stained using Giemsa (Jiancheng, Jiangsu, China) and quantified.

The serologic level of MCM6 was detected using the human MCM6 ELISA kit (Cusabio, Wuhan, CN) according to the manufacturer’s instructions. One hundred microliters of serum samples and standard proteins was added to the appropriate wells and incubated at 37 °C for 2 h. After being washed three times with washing buffer, HRP-conjugated secondary antibody (100 μL/well) was added to the cells, which were incubated at 37 °C for 1 h. The plate was washed three times, and the cells were then incubated with DAB solution for 15–30 min at 37 °C. Then, 50 μL of Stop Solution was added to each well, the optical density at 450 nm was detected, and the concentration of MCM6 was calculated from the standard curve.

Both subcutaneous and orthotopic models were used for animal studies. A total of 2 × 10⁶ SMMC7721 cells transfected with MCM6 shRNA, lentiviral vectors and negative control (scramble) vector in 0.1 ml PBS medium was used in both the subcutaneous model and orthotopic model. The cells were injected into the left leg of 4–6-week-old male BALB/c nude mice. The subcutaneous tumor size was calculated and recorded every 3 days with a Vernier caliper using the following equation: tumor volume (mm³) = (L × W²)/2, where L = tumor long axis and W = short axis. The measurements were repeated three times. The mice were maintained in a barrier facility on HEPA-filtered racks and fed an autoclaved rodent diet. After 27 days, the mice were killed and the tumor tissues were surgically excised, weighed and stained with hematoxylin and eosin (H&E). For the orthotopic model, the cell suspension was injected into the left hepatic lobe of 4-week-old nude mice with a micro-syringe. Transplanted cells were allowed to grow for up to 6 weeks, when the mice were sacrificed. The livers and lungs were dissected, fixed, and paraffin-embedded. All animal handling and procedures were approved by the Animal Care and Use Committee of Southern Medical University.

All statistical analyses were performed with SPSS statistical software (version 21.0; SPSS, IBM, Armonk, NY). Survival curves were constructed using the Kaplan–Meier method and analyzed by the log-rank test. Significant prognostic factors found by univariate analysis were entered into a multivariate analysis using the Cox proportional hazard regression model. The Pearson’s chi-square test was used to analyze the association of MCM6 expression with various clinicopathologic characteristics. A paired Student’s t test or χ2 tests was used to compare the values between subgroups. Receiver operating characteristic (ROC) curve analysis was used to determine the predictive value of the parameters, and the differences in the area under the curve (AUC) were detected. R project for Statistical Computing (R version 3.4.3) was used to calculate Harrell C-index. Data were expressed as the mean ± SD of at least three sample replicates. A value of P < 0.05 was statistically significant, and single, double, and triple asterisks indicate statistical significance of * P < 0.05, ** P < 0.01 and *** P < 0.001.

To evaluate the expression of MCM6 in HCC tissues, we first conducted immunohistochemistry assay in 70 HCC specimens (tumor and matched adjacent non-tumorous tissues) and 30 normal hepatic tissue specimens. We observed that MCM6 was primarily localized in the nucleus. The results revealed that MCM6 expression in HCC tissues was markedly higher than the adjacent non-tumor liver tissue and normal liver tissue (Fig. 1a). We detected high expression of MCM6 in 50/70 (71.4%) HCC tissues, compared with only 8/70 (11.4%) adjacent non-tumor tissues and no staining in normal tissues (P < 0.01; Fig. 1b).

Second, to verify the functions of MCM6, we tested the correlation of MCM6 expression status in 70 HCC specimens with 12 widely recognized clinicopathological parameters. Statistical analysis indicated that high MCM6 expression was associated with hepatitis B infection, liver cirrhosis, multiple tumor nodules, high serum AFP and early recurrence (Table 1). Furthermore, univariate analysis revealed that tumor size, GGT level, vascular invasion, early recurrence and MCM6 expression were unfavorable predictors for TTR and OS (Table 2). To assess the prognostic significance of MCM6, we do the Kaplan–Meier survival analysis. The 1-, 3- and 5-year recurrence rates in the MCM6-positive group were significantly higher than in the MCM6-negative group (58.0%, 80.1% and 81.5% versus 19.2%, 55.1% and 61.9%, respectively; Fig. 1c). Similarly, the 1-, 3- and 5-year OS rates in the MCM6-positive group were significantly lower than in the MCM6-negative group (61.2%, 38.3% and 32.4% versus 95.1%, 70.7% and 45.1%, respectively; Fig. 1d). Further multivariate analysis demonstrated that MCM6 was an independent prognostic factor for both the OS (P < 0.05) and TTR (P < 0.05) of HCC patients (Table 3).

In addition, to confirmed the prognostic value of MCM6, we applied MCM6 to 291 HCC samples with RNA-sequencing data in the Cancer Genome Atlas project (TCGA). Based on the relative expression of MCM6, 124 and 167 patients were classified into high- and low- expression groups, respectively. We performed systematic Kaplan-Meier survival and clinicopathological features analysis of the whole mRNA transcriptomics and found that high MCM6 mRNA expression was significantly correlated with early recurrence (recurrence rates 80.2% vs. 53.2%) (Fig. 1e) and predicted poor survival and high recurrence risk (Fig. 1f-g).

In vitro experiments suggested that MCM6 expression in high-metastasis potential cell lines, such as HCC-LM3 and SMMC7721, was higher than in the low-metastasis potential cell lines PLC/PRF5 and HepG2 (Fig. 2a). These data indicated that MCM6 might regulate the invasion and metastasis of HCC cells. This speculation was confirmed by cytological experiments. To understand the function of MCM6 in HCC cells, we designed three short hairpin RNAs (shRNA1, shRNA2, and shRNA3) to silence MCM6 expression. More than fourfold decrease in MCM6 expression was observed in SMMC7721 and HCC-LM3 cells treated with shRNA compared with the control group by qRT-PCR (Student’s t-test, with P < 0.05 for both) and western blotting; shRNA3 was the most effective shRNA and was chosen for further study (Fig. 2b-c). The wound-healing and Matrigel-coated (for invasion) or –uncoated (for migration) Transwell assays showed that HCC invasion and metastasis were effectively inhibited by MCM6 knockdown. The results showed that SMMC7721-scramble and HCC-LM3-scramble cells had a faster wound closure rate and more invasive properties than shMCM6 cells (Fig. 2d-e). These results uncovered the role of MCM6 in promoting HCC progression in vitro.

Subsequently, we examined the effect of MCM6 expression on HCC cells growth in vitro. Using CCK-8 assay (Fig. 2f), colony formation (Fig. 2g), and Edu incorporation assays (Fig. 2h), we found that knockdown MCM6 markedly suppressed cell proliferation in SMMC7721 and HCC-LM3 cells.

Invasion and metastasis are responsible for the unsatisfactory long-term survival. Emerging evidence suggested that EMT endows epithelial cells with their migratory and invasive capacity and is involved in many cancer metastasis processes, including HCC [17, 18]. To explore the relationship between MCM6 and the epithelial–mesenchymal transition (EMT), we performed real-time PCR to assess EMT markers in MCM6 knockdown cells. We found that the suppression of MCM6 resulted in higher E-cadherin and ZO-1 levels, accompanied by decreased levels of the mesenchymal marker vimentin, but did not significantly influence the other EMT markers (Fig. 3a). Western blot and immunofluorescence staining also confirmed that MCM6 expression inversely correlated with E-cadherin but positively correlated with vimentin expression (Fig. 3b and c).

Next, we sought to determine the signaling mechanisms involved in MCM6-mediated EMT. The extracellular signal-regulated kinase 1/2 (ERK, also known as p42/44 mitogen-activated protein kinase MAPK) signaling pathway is one of the most intensively studied signaling mechanisms that regulates migration, invasion, and EMT in cancer [19‐21]. We found that MCM6 affected activation of ERK signaling. We analyzed the activation status of ERK and the upstream gene MEK1/2 in cells with stable MCM6 knockdown. The data showed that MCM6 increased the levels of phosphorylated MEK1/2 (p-MEK1/2) and ERK1/2 (p-ERK1/2), but the total MEK and ERK had similar expression levels in all experimental conditions (Fig. 3d). Meanwhile, we used immunofluorescence assays to detect the localization and expression of ERK protein. It is widely known that ERK shuttling between the nucleus and cytoplasm provides a switch-like transition effect in the activation of the ERK signal, and active nuclear ERK1/2 levels were assessed by immunofluorescence. Consistent with the western blot data, the confocal photos revealed that the activation of ERK1/2 was strongly suppressed when MCM6 was knocked down (Fig. 3e). Taken together, these data suggest that MCM6 promotes EMT, at least in part, through the upregulation of MEK/ERK signaling pathways.

To further investigate the effects of MCM6 in vivo, we established subcutaneous xenograft tumor and orthotopic xenograft tumor models in nude mice. After 4 weeks, SMMC7721-Scramble cell-derived tumors at the subcutaneous implantation sites were larger and grew more rapidly than SMMC7721-shMCM6 cell-derived tumors. (Fig. 4a). Consistently, the liver orthotopic xenograft tumor also showed that MCM6 knockdown had imposed significant regulatory effects on HCC metastasis. The intrahepatic and pulmonary metastasis rates in mice with tumors derived from SMMC7721-shMCM6 cells were significantly lower than the corresponding control group (Fig. 4b and c). Moreover, the effect of MCM6 on E-cadherin and vimentin was also observed by IHC in the xenograft specimens (Fig. 4d). We found that MCM6 expression inversely correlated with E-cadherin but positively correlated with vimentin and Ki-67 expression. Taking these results together, our studies demonstrated that MCM6 could promote HCC growth and metastasis in vivo.

Previous studies demonstrated that the MCM6 levels in plasma had diagnostic value for HCC [16]. Additionally, our earlier results revealed that the elevated expression of MCM6 in tissues and serum is correlated with early recurrence (Table 1) (Additional file 1: Table S2). Therefore, to further explore its potential value in early recurrence non-invasive serologic detection, we examined MCM6 expression in patient serum. Firstly, when discerning HCC patients from cirrhosis and healthy controls, serum MCM6 had significant sensitivity (P < 0.0001, Fig. 5a). Moreover, the data obviously revealed that the serum MCM6 concentration from HCC patients who received curative hepatectomy dropped sharply compared with the pre-treatment level (2039.17 ± 206.02 pg/mL preoperative versus 1077.51 ± 234.31 pg/mL postoperative, P < 0.05). In addition, in the non-early recurrence subgroup, plasma MCM6 levels declined more dramatically after operation and some were even reduced to the normal range over a certain time (Fig. 5b).

To analyze the value of serum MCM6 in predicting the prognosis of HCC, we prepared the receiver operating characteristic (ROC) curve for serum MCM6 expression according to early recurrence. The cut off value was determined using receiver operating characteristic (ROC) curves and the Youden’s Index. (Youden’s index = sensitivity + specificity-1). According to Youden’s index, optimal sensitivity (78.6%) and specificity (76.5%) were seen at a cut-off of ≥941.635 pg/ml to distinguish the non-early recurrence group from the early recurrence group. To estimate the value of the dynamic changes in serum MCM6 in the early recurrence of HCC after liver resection, we analyzed the differences in the recurrence rate of the two groups. Early-recurrence was observed in 8 of 10 (80%) patients with high postoperative MCM6 levels (≥ 941.635 pg/ml), whereas only 6 of 21 (28%) patients with low postoperative MCM6 levels (< 941.635 pg/ml) had early-recurrence (Fig. 5c). Additionally, a high postoperative MCM6 level correlated significantly with cumulative recurrence after curative hepatectomy (Fig. 5d).

Next, the predictive performance was compared between serum preoperative MCM6 and postoperative MCM6 (Fig. 5e). Our data showed that after curative hepatectomy, the area under the ROC curve (AUROC) for MCM6 was 0.773 (95% confidence interval [CI], 0.594–0.953). The postoperative serum MCM6 was more informative than the preoperative level in predicting early recurrence in HCC. Furthermore, the ROC analysis also revealed that postoperative MCM6 is superior to postoperative AFP in differentiating early recurrence. (Fig. 5f). The AUC of postoperative MCM6 was 0.773 (95% confidence interval [CI], 0.594–0.953), which was larger than that of AFP at 0.681 (95% confidence interval [CI], 0.474–0.887). Additionally, the sensitivities of the postoperative MCM6 and AFP were 78.6% and 71.4%, respectively. Notably, the sensitivity of the combination with AFP and MCM6 was increased to 80.1% (Table 4). Harrell C-index could also proved that postoperative serum MCM6 not only significantly correlated with tumor early recurrence but also displayed improved indices of prognostic performance. Harrell C-index = 0.7249 95%CI (0.6143–0.8356). This result showed that the prediction model is accurate and the credibility is high (a C-index of 1 indicates 100% predictive accuracy). To summarize, in patients with resectable HCC, high postoperative serum MCM6 individuals had a higher tendency of early recurrence. In particular, patients with dual positivity of postoperative serum MCM6 and AFP should be supplied with more intensive care and close follow-up after they undergo tumor resection.