Nasopharyngeal carcinoma (NPC) is one of the most common malignant tumors in South China, and its morbidity and mortality are ranked first globally. The etiology of NPC is multifactorial. Accumulated epidemiological and etiological evidence indicate that NPC develops from a complex interaction between genetic factors, exposure to chemical carcinogens and latent Epstein-Barr virus (EBV) infection (1).

In recent years, studies have indicated that genetic changes in the classical oncogenes and tumor suppressor genes are rare in NPC, and the epigenetic changes in tumor suppressor genes, particlarly DNA methylation, play an important role in its development. NPC exhibits a high frequency of genetic CpG island methylation of the tumor. Moreover, tumor-associated gene methylation changes in NPC cells is involved in many functional processes, including cell cycle regulation, DNA repair, apoptosis, tumor invasion and metastasis (1–3). However, to date, and to the best of our knowledge, the epigenetic changes of the MutS homolog human 3 (MSH3) gene in NPC have not yet been investigated.

Mismatch repair genes maintain the stability of the genome by correction of base mismatch during DNA recombination and replication, and this eliminates the potential for carcinogenic changes by mutant cell growth via induction of cellular apoptosis. Thus, the mismatch repair of genes in the maintenance of genomic integrity, the repair of damage caused by cancer-specific factors and the anticancer process itself all play a very important role (4).

MSH3 is a member of the mismatch repair genes, which is located on chromosome 5q11–13. The MSH3 gene consists of 24 exons and 23 introns and is expressed in a variety of human tissues. MutSβ is a heterodimer of MSH2 and MSH3, and is primarily responsible for the repair of insertion-deletion loops (IDLs) of 1–15 nucleotides, as well as DNA loop and DNA double-strand breaks, which can trigger apoptosis of cells with high levels of DNA damage, and can also regulate the sensitivity of tumor cells to radiotherapy and chemotherapeutic drugs (5–10).

It has been suggested that human MSH3 (hMSH3) deficiency can drive hamartomatous polyposis syndrome tumorigenesis (11). MSH3 deficiency in human colonic epithelial cells results in elevated microsatellite instability at selected tetranucleotide repeats, and in the formation of DNA double-strand breaks, as well as significant proteomic changes, although it lacks oncogenic transformation (12). MSH3 knockout mice develop cancers only in later life (13); however, a double knockout of MSH3 and MSH6 renders mice far more susceptible to cancer than the single knockout of either gene (14). Therefore, MSH3 is considered a tumor suppressor gene.

Previous studies have considered that gene mutation and homozygous deletion are the main mechanisms of tumor suppressor gene transcription inactivation. However, recent studies have confirmed that promoter methylation is the third mechanism of the tumor suppression of gene transcription inactivation, and in some cases the only mechanism of tumor suppressor gene inactivation (15). MSH3 protein expression has been shown to be downregulated in colorectal (16), gastric (17), bladder (18), prostate (19), ovarian (20) and other cancers, although each cancer type presents with its own inactivation mechanism. Mutations have rarely been found in MSH3 (21), and the loss of heterozygosity (LOH) and gene promoter GpG island methylation are the predominant modes of inactivation of MSH3 (16–18,22).

In this study, to screen novel epigenetic inactivation of tumor suppressor genes in NPC, we used a genome-wide screening of genes that were downregulated by promoter hypermethylation. In our previous study, the mRNA expression levels were frequently absent or downregulated in MSH3-methylated NPC primary tumor biopsies as compared to normal nasopharyngeal epithelial (NNE) tissues (unpublished data). However, MSH3-unmethylated cases exhibited upregulated or parallel mRNA expression levels. Thus, MSH3 may be a target gene with its expression suppressed by promoter hypermethylation in NPC.

In this study, we detected the methylation status, and the mRNA and protein expression levels in NPC primary tumor biopsies. We further analyzed the correlation between promoter methylation and mRNA expression. The above-mentioned evidence supports our hypothesis that MSH3 is epigenetically inactivated in NPC by promoter hypermethylation.

Tumor cells frequently exhibit deficiencies in the signaling or repair of DNA damage. These deficiencies probably contribute to the pathogenesis of many diseases; however, a type of genetic stability is required, which can be obtained by overexpressing specific DNA repair genes, in order to produce primary tumor cells that are sufficiently genetically stable to be able to invade and give rise to distant metastases. They also present an opportunity to target the tumor. The dual roles of the DNA repair pathway in cancer highlights how an understanding of DNA repair processes can be used in the development of novel cancer treatments. Thus, the role and expression of the regulatory mechanism of the DNA mismatch repair gene, MSH3, in NPC is of particular importance and is largely unresolved.

In the present study, we demonstrated that the mRNA and protein expression levels of MSH3 in the NPC tissues were downregulated, which indicated that MSH3 played a major role as a tumor suppressor gene in the development of NPC. Our MSP results revealed that MSH3 promoter methylation was detected in 50% of the primary tumors, but was not detected in the NNE tumors. Therefore, MSH3 methylation was a frequent and tumor-specific process in NPC. The MSP and semi-quantitative RT-PCR data exhibited a correlation between the mRNA expression levels and the methylation status in NPC primary tumors. In this context, methylation inactivation appears to be a major mechanism in the loss of MSH3 expression. Thus, MSH3 gene methylation has a similar role in gastric cancer, bladder cancer and NPC, and yet plays differential roles in colorectal cancer, ovarian cancer and oral squamous cell carcinoma, in which LOH was the major regulatory mechanism for the inactivation of MSH3 gene expression (16–18,22).

Vymetalkova et al found the significant overexpression of the MSH3 gene in colon tumors as compared to adjacent mucosal tissues (26). By contrast, it has been reported that the high frequency of LOH, as well as aberrant protein expression, indicate an involvement of impaired MSH3 in the low level of microsatellite instability in colorectal cancer (16). Kawakami et al revealed that decreased hMSH3 protein expression levels in bladder cancer may play a significant role in the progression of bladder tumors. An inverse correlation between hMSH3 protein expression levels with the pathological grade was also found (18). However, in the present study, we demonstrated that MSH3 mRNA and protein expression levels in the NPC tissues were significantly and inversely associated with the variable T stage. This observation suggested that MSH3 mRNA and protein expression could be considered an indicator of local invasion by the primary tumor. Thus, the role of MSH3 as a tumor suppressor gene in NPC is similar to that of colorectal and bladder cancer, although the specific mechanisms of functional gene expression of MSH3 in different cancers vary considerably.

Conde et al discovered that the variant in MSH3, referred to as Ala1045Thr, was associated with a decreased risk of breast cancer (27). The results presented in the studies by Jafary et al and Hirata et al demonstrated that MSH3 codon 222 and MSH3 codon 1036 polymorphisms may represent an increased risk factor for sporadic prostate cancer (28,29). Michiels et al reported that MSH3 single-nucleotide polymorphisms may be associated with an increased risk of lung cancer (30). Koessler et al reported a single nucleotide polymorphism of rs863221, located in MSH3 that was associated with disease-specific survival in patients with colorectal cancer (31). Dong et al discovered that MSH3 single-nucleotide polymorphisms correlated with overall survival in patients with pancreatic cancer (32). However, the majority of the cases were followed-up for >5 years, and thus the 5-year survival rate of the patients was not analyzed. In the future, the role of MSH3 in the pathogenesis, prognosis and risk estimation of NPC needs to be further clarified after the relevant follow-up data of the past 5 years is analyzed statistically.

Previous evidence has suggested that MSH3-deficient as compared to proficient colorectal cancer cells exhibit an increased sensitivity to the irinotecan metabolite, SN-38, and to oxaliplatin, but not to 5-FU (9). Another study reported an attempt to identify an association between MSH2 and MSH3 genetic variants and development of radiosensitivity in breast cancer patients (10). Thus, the sensitivity of nasopharyngeal cancer patients to radiotherapy and chemotherapy is related to MSH3, which need to be confirmed by further experiments.

In addition, methylation-mediated inactivation is potentially a reversible phenomenon (33). As the MSH3 gene product has an anticancer effect, turning this process around and upregulating MSH3 by using demethylating agent treatment may probably prevent or reverse the malignant phenotype, and might therefore translate into a novel demethylating agent and therapeutic target in NPC.

In this study, it is a pity that we have only used tissue samples and have not performed any in vitro experiments on cell lines. We aim to further analyze the mRNA expression of the MSH3 gene in NPC cell lines before and after treatment with the methyltransferase inhibitor, 5-aza-2-deoxycytidine, in future studies. We aim to provide evidence to further verify that MSH3 is epigenetically inactivated in NPC by promoter hypermethylation. The present study revealed that MSH3 is frequently inactivated by its promoter methylation and its mRNA and protein expression correlate with the primary tumor stage in NPC, suggesting that MSH3 mRNA and protein expression can be considered as such, an indicator of primary tumor local invasion, and would have a potential value in clinical applications. MSH3 may probably translate to a novel demethylating agent and therapeutic target in NPC. Whether the MSH3 gene can be used as a predictive indicator of the sensitivity of patients with NPC to radiotherapy and chemotherapy should be confirmed by further experiments.