Elevated expression of minichromosome maintenance 3 indicates poor outcomes and promotes G1/S cell cycle progression, proliferation, migration and invasion in colorectal cancer

Abstract Background: The minichromosome maintenance (MCM) family, a core component of DNA replication, is involved in cell cycle process. Abnormal proliferation has been identified as a crucial process in the evolution of colorectal cancer (CRC). However, the roles of the MCM family in CRC remain largely unknown. Methods: Here, the expression, prognostic significance and functions of the MCM family in CRC were systematically analyzed through a series of online databases including CCLE, Oncomine, HPA, cBioPortal and cancerSEA. Results: We found all MCM family members were highly expressed in CRC, but only elevation of MCM3 expression was associated with poor prognosis of patients with CRC. Further in vitro and in vivo experiments were performed to examine the role of MCM3 in CRC. Analysis of CCLE database and qRT-PCR assay confirmed that MCM3 was overexpressed in CRC cell lines. Moreover, knockdown of MCM3 significantly suppressed transition of G1 to S phase in CRC cells. Furthermore, down-regulation of MCM3 inhibited CRC cell proliferation, migration, invasion and promoted apoptosis. Conclusion: These findings reveal that MCM3 may function as an oncogene and a potential prognosis biomarker. Thus, the association between abnormal expression of MCM3 and the initiation of CRC deserves further exploration.


Introduction
Colorectal cancer (CRC) is the third most prevalent malignant cancer globally. Besides fast progression, CRC is associated with high mortality [1]. In 2019, approximately 1,762,450 and 606,880 new cases and deaths reported in the United States of America, respectively [2]. Although the advent of more accurate early screening methods and personalized treatments has stabilized the number of CRC cases in developed countries, the incidence and mortality in developing countries, including China, continue to rise significantly [3,4]. Currently, the TNM staging method is based on the American Joint Committee on Cancer (AJCC) and the Union for International Cancer Control (UICC). It is used to assess the prognosis of patients and guide the choice of treatment. However, given that treatment outcomes of individual patients are unpredictable [5], in-depth understanding of the underlying molecular mechanisms and identification of CRC markers are required to develop effective therapeutic strategies.
Uncontrolled self-replication of tumor cells promotes cancer progression, and refers not only to abnormal proliferation at the primary site but also to rapid growth in the target organs for metastasis [6].
Cell proliferation is regulated by multiple mechanisms, with DNA replication being a crucial regulator that ensures precise control of the cell cycle [7]. The minichromosome maintenance family (MCMs) was first identified in Saccharomyces cerevisiae, where its mutation was shown to weaken the stability of minichromosome in proliferating cells, thereby stalling continuity of the cell cycle [8,9]. So far, nine homologues of MCM proteins have been identified in eukaryotic cells including MCM2-7, MCM8, MCM9 and MCM10 [10][11][12][13]. These proteins are involved in the initiation of DNA replication [14], transcription [15], cycle checkpoint [16] and RNA splicing [17]. Among them, MCM2-7 recruits ORC, CDC6 and CDT1 to form heteromultimers, which function as helicases driving the transition from M to G1 phase [18]. Although similar to MCM2-7 helicase in terms of the replication process, the proteins involved in loading MCM8 and MCM9 are different [19]. Recent studies have indicated that during homologous recombination, MCM8 and MCM9 can stabilize each other and form a complex that promotes DNA synthesis thereby reactivating a broken replication fork [20]. MCM10 is highly conserved in eukaryotic cells, and promotes continuous replication of the cell cycle by recruiting DCAF1-Cul4-E3 [21].
Accumulating evidence indicates that DNA replication-related proteins can be novel therapeutic targets for tumors [22]. Due to their close association with the cell replication process, the MCM family proteins are considered ideal biomarkers for malignant cell proliferation. Besides, abnormal expression of MCMs has been detected in a variety of tumors, including breast cancer [23], hepatocellular carcinomas [24], renal cancer [25], colorectal cancer [26] and gliomas [27]. Furthermore, evidence has emerged that associates abnormal expression of MCMs with cancer prognosis. For instance, MCM2 is a promising prognostic marker in lung squamous cell carcinoma, where its overexpression indicates a shorter survival period [28]. High expression level of MCM5 correlates with tumor invasiveness and unfavorable prognosis of cervical cancer [29]. MCM4 influences the cell cycle in colorectal cancer by regulating (Skp2)-P27 axis and Ohmyungsamycin A expression [30]. High expression of MCM7 promotes cell proliferation by activating the MAPK signaling pathway, leading to worse prognosis in hepatocellular carcinoma [31]. The depletion of MCM9 can lead to an early susceptibility to colorectal cancer [32].
In the MCM2-7 complex, MCM3 functions as a flexible linker through its C-terminal [33]. Unlike cell cycle biomarkers ki-67 and proliferating cell nuclear antigen (PCNA), MCM3 possesses higher stability and is less susceptible to interference from external factors [34]. Although several lines of evidence indicate that MCMs play a key role in tumor progression, but their prognostic value in CRC remain unknown. In the present study, we aimed to screen for valuable prognostic and diagnostic markers to imporve the treatment of CRC by conducting an integrated analysis of several databases based on the mRNA expression profile of the MCM family.

Oncomine database analysis
Oncomine (http://www.oncomine.org) is currently the world's largest oncogene chips database and integrated data-mining platform for cancer genetic information. It covers 65 microarray data set, 4700 chips, and 480 million gene expression data. Oncomine is used for analysis of gene expression differences, identification of outliers, predicting co-expressed genes, and facilitating the discovery of new tumor biomarkers, or therapeutic targets [35]. Further, it integrates TCGA and GEO data as well as some bioinformatics analysis tools for standardized analysis of tumor transcriptome data. In the present study, we used the database to analyze the transcription levels of distinct MCMs in tumor and normal tissues for different types of cancers. The results were filtered using the threshold: fold change ≥ 2; P-value ≤ 1 × 10 −4 ; gene rank ≥ top 10%.

The human protein atlas database
The Human Protein Atlas (HPA) is a large database containing tissue and cell distribution information for 24,000 proteins in humans. These include the immunohistochemical (IHC) test results showing the distribution and expression of each protein in 48 human normal tissues, 20 tumor tissues, 47 cell lines, and 12 blood cells, which have been validated and indexed by experts [36]. It also includes the possible roles of specific proteins in the development of different tumors, making it an platform for tumor research. Therefore, IHC images of cancer and normal tissues in patients with CRC were obtained from the HPA database, and the protein expression of MCMs was visualized.

Analysis of cBioPortal database
With the widespread application of chip and high-throughput sequencing technology, massive genomic data have been generated in the field of oncology, and a variety of databases have been designed to meet different research needs. The cBioPortal (http://www.cbioportal.org/) database integrates data from 126 tumor genomic studies, including large-scale tumor research projects such as TCGA and Oncomine, and contains data from 28,000 specimens [37]. In  addition, some samples also include information such as clinical prognosis. This database provides researchers with high-quality molecular profiling, and clinical prognostic correlation of large-scale cancer genomics projects, within a short time [38]. In our study, we analyzed the relationship between clinical prognosis and MCMs in patients with CRC using the database.

CancerSEA database analysis
CancerSEA (http://biocc.hrbmu.edu.cn/CancerSEA/) is the first single-cell sequencing database designed to comprehensively explore the functional states of cancer cells at single-cell level. The single-cell sequencing results in the CancerSEA database are derived from 72 data sets in the SRA, GEO and ArrayExpress websites. In total, it comprises 41,900 cancer cells from 25 types of cancer, and functional analysis results from data sets such as HCMDB, Cyclebase and StemMapper, which have redefined 14 functional states [39]. Therefore, we used the CancerSEA database to analyze the functional correlation of the MCM family with CRC.

Cancer cell line encyclopedia database analysis
The Cancer cell line encyclopedia (CCLE) (https://portals.broadinstitute.org/ccle) contains sequencing information of 947 human cancer cell lines from more than 30 tissue sources, and includes data visualization [40]. To deepen our understanding of DNA mutations, gene expression and chromosome copy number information of specific genes, we used CCLE to analyze the mRNA expression of MCM family members in cell lines derived from different tumor types. The raw microarray data of MCMs in CRC cell lines were downloaded from CCLE and then converted to a single value for each probe set using the RMA algorithm and quantile normalization.

Cell culture
The normal human colon cell line NCM460 and human CRC cell line COLO205 were obtained from the Chinese Academy of Sciences Cell Bank. All the cells were cultured in roswell park memorial institute (RPMI)-1640 medium containing 10% fetal bovine serum (PAN-Biotech, Adenbach, Bavaria) in a 37 • C constant temperature incubator with 5% CO 2 .

Lentivirus transfection and stable cell line selection
The lentivirus-mediated GV248 vector (Genechem, China) was used to express short hairpin RNA (shRNA) targeted MCM3. The sequences of shRNAs targeting MCM3 were as follows: 5 -CCA GAG GTA GAC GTA TTT ATT-3 (sense) and 5 -GCC TGG CTA GGG TTA GGA CTT-3 (antisense). Lentiviral particles were transfected into 293T cells with GV248-shMCM3 constructs. The stable cell lines were selected by puromycin (4 μg/ml) 72 h after transfection. The medium containing puromycin was replaced every 3 days for 2 weeks.

RNA isolation and quantitative real-time PCR
TRIzol reagent (Invitrogen, U.S.A.) was used to extract total RNA from cells. The extracted RNA was reverse-transcribed using the first strand cDNA synthesis kit (Thermo, #K1642). Then, qRT-PCR was carried out on a Bio-Rad CFX96 system (Bio-Rad, Hercules, CA). Primer sequences in the present study were: 5 -CGA GAC CTA GAA AAT GGC AGC C-3 (forward) and 5 -GCA GTG CAA AGC ACA TAC CGC A-3 (reverse) for MCM3; 5 -GTC TCC TCT GAC TTC AAC AGC G-3 (forward) and 5 -ACC ACC CTG TTG CTG TAG CCA A-3 (reverse) for GAPDH. GAPDH was used as a normalizing control and the relative mRNA expression was calculated using the CT method.

Cell apoptosis and cycle analysis
Cell apoptosis and cycle were detected by flow cytometry (FCM). Briefly, adherent cells were separated into single cells using 0.25% trypsin; the cell suspensions were mixed with PBS and centrifuged for 5 min at 1000 r/min. The supernatant was discarded and 1 × 10 6 cells were resuspended in 500 μl of PBS and incubated in AnnexinV-APC and DAPI in the dark for 20 min. A cell suspension containing 1 × 10 6 cells was stained with propidium iodide (PI) and analyzed by FCM for cell cycle distribution.

Cell colony formation assay
Colony formation assay was conducted to evaluate the effect of MCM3 knockdown on cell proliferation. Transfected cells (shMCM3 or shCtrl) were seeded in six-well plates at a density of 500 per well; three replicate wells were set for each group. About 2 ml of RPMI-1640 medium containing 10% FBS was added into each well and cultured for 2 weeks. The formed clones were washed twice with PBS, fixed with methanol for 20 min and subsequently stained with 0.2% Crystal Violet for 10 min. A cell colony with more than 50 cells was considered a positive colony and the number of colonies were counted under a microscope.

Cell migration and invasion assay
The migration and invasion ability of cells was determined using wound healing and transwell assays. For the wound healing assay, 5 × 10 5 cells were seeded in a 6-well plate and allowed to grow until a confluence of 95%. An artificial wound was created with a 10 μl sterile tip. Cells were incubated a serum-free medium for 24 h, after which the wound was imaged under a microscope. Cell migration was assessed by measuring the spacing between the two boundaries of the cell area. For cell transwell assay, 100 μl of serum-free medium containing 1 × 10 5 cells was plated in the upper chamber whereas a medium containing 10% FBS was added to the lower chamber and cultured for 24 h under standard conditions. The cells that had migrated into the lower chamber were fixed with 4% polymethanol for 20 min and stained with 0.2% Crystal Violet for 5 min. Invaded cells were observed under an inverted microscope and counted in five random fields of view.

Xenograft in vivo analysis
Ten BALB/c nude mice were purchased from the Shanghai Laboratory Animal Center (Shanghai, China). All animal studies were conducted in full compliance with the principles and protocols approved by the Ethics Committee on Animal Center of North Sichuan Medical College. After the mice grew to 4 weeks old, they were administered 2% cocaine hydrochloride anesthesia on the skin surface. To evaluate the tumorigenicity of stable knockdown of MCM3 cells in vivo, CRC cells were transfected with lentivirus (COLO205-shMCM3) to knockdown the expression of MCM3 or with COLO205-shCtrl. About 100 μl of the cell suspension (1 × 10 7 cells/ml) was injected into the lateral abdomen of nude mice (n = 5 per group). The survival status of mice was observed and their tumor size recorded every 3 days. Four weeks after the injection, mice were killed by cervical dislocation. After their heart and breathing had completely stopped, the formed neoplasms were excised and weighed for immunohistochemical assay.

Immunohistochemistry staining
Excised tumor tissues were continuously sliced into 5-μm sections. The slides were dewaxed in xylene and dehydrated with a series of ethanol concentrations. Antigen retrieval was performed by treating the tissue sections with citrate buffer and heating them a microwave. The tissues were incubated with primary mouse monoclonal antibody against PCNA (dilution, 1:500; cat. no., 60097-1-Ig; Proteintech) overnight at 4 • C. Next, they were incubated with HRP-labeled secondary antibody (Santa cruz, sc-516102) for 30 min and stained with the DAB immunohistochemistry color development kit (Sangon Biotech, China). Finally, the dyed tissue sections were examined under a microscope (Olympus, Japan).

Statistical analysis
The GraphPad Prism (Version 8.0 GraphPad Software, CA) was used for statistical analyses. The significance of differences between groups was evaluated using the Student's t-test. Statistical significance of MCM family expression between colorectal cancer and normal tissues from Oncomine was provided by the program. Survival data of MCM family mRNA expression were obtained from cBioPortal. Survival curves were plotted via the Kaplan-Meier method and compared with the log-rank test. P values less than 0.05 were considered statistically significant.

Results mRNA expression profile of MCM family members in CRC
CCLE analysis revealed that the expression of MCM3 in colorectal cancer ranked 15th among all types of cancer ( Figure 1A). Analysis performed on the Oncomine database showed that the mRNA expression of all MCM family members was overexpressed in CRC compared with normal tissues across diversified datasets ( Figure 1B). MCM mRNA expression in CRC and normal tissues were summarized in Table 1. In a TCGA dataset with the largest sample size (n=237) from the Oncomine database, MCM3 transcripts in the colon and rectal adenocarcinoma tissues were 1.668-and 1.660-fold higher than relevant normal tissues respectively ( Figure 2B). Moreover, MCM2 was 2.083-and 2.199-fold higher in colon and rectal adenocarcinoma samples respectively compared with normal tissues (Figure 2A). The MCM4 was 1.690-and 2.030-fold as high as the normal tissues in colon and rectal adenocarcinoma samples, respectively ( Figure 2C). MCM5 was higher 1.240-fold in colon and 1.213-fold in rectal adenocarcinoma  samples than in normal tissues ( Figure 2D). The MCM6 was 2.053-and 2.142-fold higher in colon and rectal adenocarcinoma samples respectively compared with the normal tissues ( Figure 2E). MCM7 was 2.380-and 2.176-fold higher in the colon and rectal adenocarcinoma samples in that order compared with contrastingly relevant normal tissues ( Figure 2F). Furthermore, MCM8 was found to be 1.980-and 1.937-fold higher in colon and rectal adenocarcinoma samples respectively as compared with relevant normal tissues ( Figure 2G). MCM9 was 1.926-fold in colon and 1.999-fold in rectal adenocarcinoma samples lower than relevant normal tissues ( Figure 2H). MCM10 was also 3.484-fold in colon and 3.309-fold in rectal adenocarcinoma samples lower than relevant normal tissues ( Figure 2I).

Protein expression of MCMs in CRC tissues
The expression of MCMs in CRC tissues was analyzed using MCMs immunohistochemical results obtained from Human Protein Atlas database ( Figure 3). We found that protein levels of all MCM members were higher in CRC tissues than in normal tissues. These results were consistent with those seen in mRNA expression.

The impact of MCM family expression on prognosis of CRC patients
Further analysis of the prognostic implication of mRNA levels of MCM family in CRC performed in cBioPortal database revealed that patients with high MCM3 expression had worse DFS/PFS outcome compared with those with low MCM3 expression (P=0.009339, Figure 5B). However, mRNA levels of the remaining MCM family members were not significantly correlated with OS or DFS/PFS (P>0.05, Figures 4 and 5A,C-I).

The functions of MCM family members in a single CRC cell
Heterogeneity associated with different functional phenotypes of tumor cells has been major obstacle to effective treatment of cancer. Recent advances in single-cell sequencing (scRNA-seq) technology have provided a tool for understanding the functional status of tumor cells at the cellular level. Functional correlation analysis of cancerSEA showed that the functional phenotypes of MCM family in CRC cells were positively correlated with cell cycle, DNA damage, DNA repair, proliferation and metastasis, but negatively correlated with quiescence and stemness ( Figure  6B). In addition, MCM3 played a role in cell cycle, invasion and proliferation in CRC cells ( Figure 6A).

The mRNA expression of MCM3 in CRC cell lines
Having shown that MCM3 expression in COLO205 cell was higher than in other CRC cell lines based on CCLE analysis ( Figure 7A), we used COLO205 cells in the subsequent analysis. Thus, qRT-PCR assay was performed to determine the mRNA levels of MCM3 in CRC cell lines and normal cell lines. As shown in Figure 7B, the mRNA expression of MCM3 was higher in the COLO205 cell line compared with the normal cell line NCM460.

Effect of MCM3 knockdown on cell cycle and apoptosis
Accumulating evidence has confirmed that members of the MCM family play a key role in DNA replication. Our previous analysis uncovered that MCMs are interlinked with cell cycle-related genes. Therefore, we investigated the effect of MCM3 knockdown on the cell cycle of CRC cells. Results shown in Figure 7C indicate that the protein levels of MCM3 were significantly reduced after MCM3 knockdown. The proportion of cells in the G1 phase was higher, while the proportion of cells in the S phase was significantly lower in sh-MCM3 group than in the sh-Ctrl group ( Figure 8A,B). On the other hand, apoptosis was significantly higher in cells expressing sh-MCM3 compared with cells of sh-Ctrl group ( Figure 8C). Mechanistically, MCM3 knockdown significantly inhibited the expression of proteins associated with cell cycle initiation such as CDK2, CDK4, CyclinD1 and CyclinE1 ( Figure 8D). These results suggested that MCM3 interferes with cell cycle progression causing abnormal proliferation of CRC cells.

MCM3 knockdown inhibits CRC cells proliferation, migration, invasion and tumor growth in vitro and in vivo
To determine the biological functions of MCM3 in CRC, MCM3 was knocked down in CRC cells. A colony formation assay performed on the cells indicated that MCM3 silencing decreased the number of colonies in formed by COLO205 cells (P<0.05, Figure 9A). Furthermore, wound healing and transwell assays revealed a significantly lower cell migration and invasion ability in the sh-MCM3 group than in sh-Ctrl group (P<0.05, Figure 9B,C). To determine the effect of MCM3 knockdown on tumorigenesis, COLO205 cells transfected with sh-MCM3 and negative control cells were injected into the lateral abdomen of nude mice. The tumor growth (volume and weight) was dramatically inhibited in the sh-MCM3 group unlike in the control group ( Figure 9D). IHC staining using PCNA as an indicator of tumor cell proliferation revealed the PCNA expression was substantially lower in the sh-MCM3 group compared with the control group ( Figure 9E). Collectively, these results demonstrated that MCM3 regulated the progression of CRC cells.

Discussion
Strict control of DNA replication is a prerequisite for normal cell division [41]. Therefore, deregulation of DNA replication leads to abnormal gene phenotypes that drive malignant transformation of normal cells [42,43]. Currently, chemotherapy and small-molecule inhibitors targeting key enzymes of the DNA replication process are used to treat tumors [44]. Given that these targeted drugs are non-specific, they affect the entire replication process of normal and tumor cells leading to concomitant death [45]. Prevailing evidence indicates that MCM family proteins act as helicases in the initial stage of DNA replication, and are key regulators of cell cycle checkpoints [46]. For instance, MCM2-7 complexes contribute to the initiation of cell cycle, and are therefore potential markers for tumor screening and treatment.
To the best of our knowledge, comprehensive bioinformatics analysis of the MCM family in CRC has not been performed. Our study is, therefore, the first to analyze mRNA expression profile of MCM family in CRC by integrating information from several large databases. We examined mRNA expression profiles of the MCM family in 20 common human tumors using the Oncomine database. The results showed that all MCM family members were up-regulated in CRC and these results were corroborated by immunohistochemistry analysis of colorectal cancer and normal tissues [47]. Immunostaining results from the HPA database demonstrated that the expression of the MCM family proteins in CRC was consistent with that of mRNA expression. Further, Kaplan-Meier analysis revealed that high expression of MCM3 was associated with poor DFS/PFS. No such association was found for other members of MCM family.
MCM3 is the catalytic core of the MCM2-7 complex which besides recognizing the origin of replication, generates ATP-dependent enzyme activity by interacting with proteins involved in the cell division cycle [48]. High expression of MCM3 has been reported in other types of neoplasia including papillary thyroid carcinoma [49], cervical squamous cell carcinoma [50], ameloblastoma [51] and malignant melanoma [52]. Consistently, our results reveal that MCM3 is highly expressed in CRC cell lines based on CCLE analysis and qPCR assay. These results show the potential of MCM3 to serve as a marker of tumor cell proliferation and a prognostic predictor.
Interestingly, Western blot analysis revealed that MCM3 was positively related to classic cyclins such as CDK2 and CDK4. This partially confirmed that MCM3 promotes DNA synthesis [53]. A study by Branzei et al. indicated that the MCM2-7 complex controls progression from G1 to S phase during replication by regulating DNA replication checkpoints [54]. Elsewhere, it was reported that deregulation of cell cycle promoted genomic instability and apoptosis leading tumorigenesis [55]. Consistent with these findings, analysis of the CancerSEA showed that MCM3 participated in the cell cycle, proliferation and metastasis. These functions were verified experimentally through in vitro and in vivo tests which showed that MCM3 deletion significantly suppressed the G1 phase and promoted apoptosis. Moreover, MCM3 knockdown dramatically inhibited COLO205 cell proliferation, migration and invasion both in vitro and in vivo. We therefore propose that MCM3 could be a promising biomarker of CRC.