Over Expression of Minichromosome Maintenance Genes is Clinically Correlated to Cervical Carcinogenesis

Minichromosome Maintenance (MCM) proteins play important roles in cell cycle progression by mediating DNA replication initiation and elongation. Among 10 MCM homologues MCM 2–7 form a hexamer and assemble to the pre-replication complex acting as replication licensing factors. Binding and function of MCM2-7 to pre-replication complex is regulated by MCM10 mediated binding of RECQL4 with MCM2-7. The purpose of this study is to explore the role of MCMs in cervical cancer and their correlation with the clinical parameters of cervical cancer. We have investigated sixty primary cervical cancer tissue samples, eight cervical cancer cell lines and thirty hysterectomised normal cervical tissue. The expression profiling of MCMs was done using semi-quantitative RT-PCR, immunoblotting and immunohistochemistry. MCM2, 4, 5, 6, 7, 10 and RECQL4 are significantly over-expressed in cervical cancer. Among these, MCM4, 6 and 10 show increased frequency of over expression along with advancement of tumor stages. MCM4, 5 and 6 also show differential expression in different types of lesion, while MCM2 and MCM10 are over expressed in cervical cancer irrespective of clinico-pathological parameters. Our data indicates the role of MCM4, MCM5, MCM6, MCM10 and RECQL4 in the progression of cervical cancer.


Introduction
DNA replication is the event of common interest in the study of initiation and progression of cancer. A normal cell maintains its entry and exit into cell cycle by several checkpoints and ''licensing'' its DNA replication only once per cell cycle. This licensing mechanism includes the formation of pre-replication complexes (pre-RCs) in late M and early G1 phases and their subsequent activation at the G1-S boundary. The pre-RCs mark the replication origins and control bidirectional DNA synthesis from these origins when S phase is initiated. Pre-RC assembly involves sequential recruitment of several proteins on replication origin. The reaction starts by the initial binding of origin recognition complex (ORC). Subsequent binding of CDC6 and CDT1 provide a landing pad for the further recruitment of putative DNA helicases as Minichromosome Maintenance (MCM) 2-7 complex [1]. Other important members of pre-RC are MCM10 and RECQL4 [2]. At the G1-S transition, the activity of two kinases, CDC7 and cyclins E/A-CDK2, recruit additional factors to pre-RCs, resulting in the formation of pre-initiation complexes (pre-ICs) [3]. Additionally, CDC7 and CDK2 activate the MCM2-7 helicases, which together with formation of pre-IC result in recruitment of DNA polymerases and initiation of DNA replication. Paradoxically, during late S and M phases, high activity of cyclin-dependent kinase (CDK) results in dissolution of the pre-RCs and destruction of selective pre-RC components, thereby preventing DNA re-replication [4].
MCM proteins were first recognized in the yeast Saccharomyces cerevisiae as mutants defective in the maintenance of mini chromosomes, suggesting a role in plasmid replication and cell cycle [1]. At least 10 homologues, MCM1-10, have been characterized in humans. Among these, MCM2-7 and MCM10 are involved in DNA replication [5]. Expression profiling of isolated MCM genes in multiple malignancies has been reported [6]. Deregulation of MCMs by reducing or increasing the levels of a single MCM leads to disruptions in genome stability in yeast [7,8,9]. Since MCM activity is essential for DNA replication in dividing cells and is lost in quiescence [10], MCMs are obvious markers for proliferation. Molecular studies suggest that increased levels of MCMs mark not only proliferative malignant cells [11,12], but also precancerous cells and the potential for recurrence [13,14]. Experimental evidence has identified RECQL4 and MCM10 as most important components of pre-RC [15]. During DNA replication, MCM10 mediates RECQL4 association with MCM2-7 complex on the origin [2]. RECQL4 is up regulated in actively proliferating virus transformed human B cells, fibroblasts and umbilical endothelial cells [16]. However, the expression profile of MCM10 with respect to RECQL4 and other MCMs is poorly understood in cancers. Since the expression level of MCM proteins in several dysplasias and neoplasias is up-regulated manifold, these proteins can be useful as potential diagnostic and prognostic marker for human malignancies [17].
Here, we show the expression profiles of MCM2, MCM3, MCM4, MCM5, MCM6, MCM7, MCM10 and RECQL4 in cervical cancer (CC) cell lines and primary tumors. We have also shown correlation of the expression levels to clinical parameters.

Primary Tumor Biopsies, Controls and Cell Lines
Sixty primary tumor biopsy samples, 30 hysterectomised controls and 8 cervical cancer cell lines have been used for this study. Tumor biopsies were collected in RNA later from the patients attending the Department of Radiotherapy and Radiation and normal cervical tissue were taken from subjects without any history of cervical cancer and who underwent hysterectomy for other reasons, in the Department of Obstetrics and Gynaecology, Institute of Medical Sciences, Banaras Hindu University, Varanasi. All biopsies were obtained after appropriate informed written consent of the subjects and approval of the institutional review board/ethical committee of the Institute of Medical Sciences, Banaras Hindu University. The biopsy samples were sub categorized according to the clinical information (Table 1). Eight cervical cancer cell lines (HeLa, SiHa, SW756, C-4I, C-33A, CaSki, MS751, and ME-180) [18] were kind gift from Dr. VVS Murty, Columbia University, New York, USA. HPV typing of primary tumor samples was done by PCR amplification using primers of HPV-16, HPV-18 and L1 consensus sequences [19].

PCR
Total RNA was isolated from tissue samples preserved in RNA later (Ambion, USA), using TriZol (Invitrogen, USA) and from cell lines harvested from semi confluent culture flasks. Good quality RNA (as confirmed by integrity of 28S and 18S rRNA on agarose gel and A 260 /A 280 ratio) was reverse transcribed by cDNA kit (Applied Biosystems, USA) according to manufacturer's protocol. The cDNA was used for PCR amplification by gene specific primers ( Table 2). Primers were amplified in 2X PCR Master Mix (Thermo Scientific, USA) in a condition optimized through temperature and MgCl 2 concentration gradient. All the PCR experiments were accompanied with no-template control and bactin as internal control. After agarose gel electrophoresis, band intensities were calculated by SpotDenso densitometry software (AlphaImager, USA) and normalized with b-actin. Thus a normalized expression level indicated the ratio between band intensity of the target gene and b-actin for each sample. A gene was considered to be up-regulated when it had a normalized band intensity value higher than mean 626 standard deviation (SD) of normal cervix. Fold change for each sample is calculated by the expression value of a particular sample divided by average of the normal for a particular gene.

Immunoblotting
Whole protein fraction was isolated from cell lines (n = 8) primary tumors (n = 6) and control samples (n = 6) using lysis buffer (50 mM Tris-Cl P H7.4, 1 m EDTA, 2 mM EGTA, 1 mM DTT, 100 mM PMSF, 100 mM NaCl, 1% NP-40, 0.1% SDS and protease inhibitor cocktail). Protein concentration was determined by Bradford assay. 100 mg of protein per sample was run on SDS-PAGE and transferred on PVDF membrane (Millipore, USA). Mouse monoclonal anti-MCM4 antibody (Santa Cruz Biotechnology, USA) was used and the blots were detected with NBT-BCIP substrate system against ALP-tagged secondary antibody.

Immunohistochemistry
Immunohistochemistry was done on paraffin embedded tissue sections. 5 mm thick sections of primary tumor (n = 2) and control (n = 2) were cut and a microarray of two subsequent sections of each sample were mounted on poly L-lysine coated slides. Tissue sections were deparaffinized and treated for 10 minutes in citrate buffer for antigen retrieval in water bath at 95uC. Rabbit polyclonal anti-MCM4 primary antibody (Abcam, UK) and rabbit polyclonal anti-Ki67 antibody (Pierce Biotechnology, USA) were used in 1:250 and 1:1000 dilutions respectively. Peroxidase conjugated secondary antibody was used against the primary antibody. For chromogenic detection, 3, 39-diaminobenzidine tetrahydrochloride (DAB) (Sigma, USA) was used as the substrate for peroxidase. Slides were counterstained with hematoxylin. Cells with brown nuclei were considered as positively stained for MCM4. Negative control experiment was performed without using primary antibody.

Statistical Analysis
Statistical analysis was done by SPSS version 16.0 for windows. Mann-Whitney and Kruskal-Wallis tests were used to compare two or more than two groups respectively. Where a Kruskal-Wallis test gave significant result, two individual groups of that particular parameter were compared by Mann-Whitney test. Expression trends among different stages were analyzed by Spearman correlation as required. Data is represented in the text as (mean of expression level 6 SD) for expression level. P values ,0.05 were considered as significant.  Figure 1C). The average expression level of MCM4 gradually increases with the advancement of the disease ( Figure 2B). Frequency of MCM4 over-expression gradually increases with the advancement of stage of disease. It is over expressed in 16% cases of stage I (n = 06), 22% cases of stage II (n = 22) and 31% cases of stage III (n = 32) of the disease ( Figure 3B). There is a trend of increasing MCM4 expression with the advancement of the disease, however, correlation is not significant (Spearman's rho = 0.079; p = 0.546). Kruskal-Wallis test comparing expression levels of normal and different types of lesion shows significance (p = 0.026). Interestingly not all the groups, but ulceroinfiltrative (p = 0.014), proliferative (p = 0.021) and ulcerative (0.012) show significant difference compared to normal ( Figure 4A). In primary tumors and cell lines, average fold changes in MCM4 expression is 2.08 (61.6) and 2.61 (61.04) respectively. Immunoblotting data is concordant with the RT-PCR data for MCM4 showing up-regulation in tumors ( Figure 5). Immunohistochemical analysis shows that MCM4 expression is undetectable in the normal cervical epithelium ( Figure 6A) whereas the MCM4 positive cells are dispersed in the cancerous cervical epithelium ( Figure 6B). Proliferation in the cancerous cervical epithelium is confirmed by the expression of Ki67 in comparison to normal ( Figure 6C and Figure 6D).

Expression Profiling of MCMs and RECQL4 in CC Cell Lines and Primary Tumors
MCM5 shows significant up-regulation in primary tumor (0.20660.152, median = 0.27, p = 0.018) and cell lines (0.30560.05, median = 0.32, p = 0.0001) compared to normal (0.12860.098, median = 0.12) ( Figure 1D). Average expression level of MCM5 has been shown in figure 2C. 20% cases of stage I, 22% cases of stage II and only 9.3% cases of stage III show up regulation of MCM5 ( Figure 3C). No positive correlation is found between expression levels of MCM5 with increasing tumor stages (Spearman's rho = 20.308, p = 0.078). Kruskal-Wallis test shows significant difference between the expression level of MCM5 in normal and different types of lesions (p = 0.007). There is no difference found among the groups but individually some of the groups show differential expression compared to normal. MCM5 shows significant up regulation in infiltrative (p = 0.048), proliferative (p = 0.006) and ulcerative (p = 0.032) types of lesions as compared to normal ( Figure 4B).
MCM6, which we found to be up-regulated in tumor  Figure 2E). 66% cases of stage I, 77% cases of stage II and 100% cases of stage III show MCM10 over expression ( Figure 3E). Although, there is an increasing trend, no significant correlation (Spearman's rho = 0.108, p = .410) is found between expression level and tumor stages. Also, MCM10 is significantly up-regulated in all lesion types compared to normal ( Figure 4D).

HPV Status
All the primary tumors were HPV positive, among which more than 71% were positive for HPV 16. Of eight cell lines C-4I, SiHa, were HPV 16 positive, HeLa, Me-180, MS751 and SW756 were HPV 18 positive and CaSki was both HPV 16 and 18 positive, while one cell line (C33A) was HPV negative. There was no significant correlation of the expression of any of the MCMs with HPV status or with HPV types.

Discussion
Replication licensing can be positively correlated with the proliferative potential of eukaryotic cells [20]. Perpetually growing tumor cells require continuous licensing. Many tumors such as osteosarcoma [21], ductal breast carcinoma [22], medulloblastoma [23], prostate carcinoma [24], oral squamous cell carcinoma [25] and many others show over-expression of minichromosome maintenance genes. There are isolated reports of the deregulated expression of individual MCMs in cervical cancer. A recent immunohistochemical study has shown that MCM2 is differentially expressed in normal epithelium compared to high grade cervical intraepithelial neoplasia (CIN) and invasive cancer [26]. MCM3 and MCM4 have been shown to be over-expressed in CC cell lines [27]. Unlike the previous study, MCM3 does not show any significant change in our analysis, while MCM4 is significantly up-regulated. Western blot analysis and immunohistochemistry results are concordant with RT-PCR expression profiles. Since the expression of this helicase is relatively more frequent in higher grade tumors, it may serve as potential stage-specific marker for CC.
Dosage alterations of these replication associated genes have vivid cytogenetic background. MCM2 which is over-expressed in cervical cancer irrespective of any clinical parameter is located at 3q21. 3q21 shows high level of amplification in seven CC cell lines [28]. Overall, 3q shows frequent copy number gains by comparative genomic hybridization in cervical cancer [29]. Comprehensive cytogenetic approaches marked 8q as a region of high chromosomal gain in CC cell lines [30]. Two of the replisome associated genes, MCM4 (8q11.2) and RECQL4 (8q24.3) are included in this region. MCM4 has been detected as osteosarcoma driver gene as found to be over-represented in both copy number and expression profiles [21].
Not all but some of the genes as MCM4, MCM6, MCM10 and RECQL4 show a positive trend with increasing tumor stages. The discrepancy of correlation of MCM2, MCM5 and MCM7 may be due to a very small sample size in tumor stage I (N = 06). However, the differential over-expression of MCM4 and MCM6 in our study indicates their critical role in CC progression. Frequent over expression of MCM10, which partners with RECQL4 and binds to MCM2-7 complex activating its helicase activity, indicates that MCM10 is critical in the tumor progression.
In conclusion our study provides a comprehensive report of the expression profile of all the major MCM genes involved in human DNA replication and RECQL4, an important replisome associ-   ated factor in cervical cancer. Studies with larger sample size specifically of lower tumor stages can show significant correlation between expression levels of these genes and progressing tumor stages. This may give a better idea about the potentiality of these genes as stage specific markers. Further studies with precancerous lesions may provide clues as to whether these MCMs and RECQL4 can be therapeutic targets in cervical cancer.