Gastric cancer is one of the leading causes of cancer mortality in the world. Aberrant expression of microRNAs (miRNAs) is the hallmark of this disease. MiRNAs are endogenous non-coding RNAs that are involved in many biological processes (e.g., cell proliferation, differentiation, apoptosis, invasion and development) through gene repression. Deregulation of miRNA expression in gastric tumors and cancer cell lines have been documented to contribute in tumorigenesis, and the expression signature may correlate with different cancer types and clinicopathological features. Here, we summarized the updated gastric cancer-associated miRNAs and the downstream targets in the process of tumorigenesis. Recently, many researchers make use of the miRNA microarray platform to profile miRNA expression in gastric cancer and correlated with different clinical parameters. Its application on cancer diagnosis, prognosis and predicting treatment response rate are still underway and needs further investigation. Emerging roles of miRNAs with oncogenic or tumor suppressive properties in gastric tumorigenesis were discussed. Epigenetic silencing of miRNA by hypermethylation of promoter CpG island was also observed in gastric cancer. However, detailed mechanisms of how miRNAs regulate gene expression in gastric cancer has not been well studied. In this review, we highlight the up-to-date findings on the deregulated miRNAs in gastric cancer, and the potential use of miRNA in the clinical settings, such as diagnostic/prognostic markers and chemotherapeutic tools.

Gastric cancer is the second leading cause of cancer mortality in the world, and has a particularly high incidence in Asian countries including China and Japan. Despite the declining incidence of gastric cancer, there are still over 1 million cases newly diagnosed and 850000 deaths globally each year[1]. The high mortality rate is mainly due to late presentation, since early stage of gastric cancer is either asymptomatic or presents with non-specific symptoms. The survival rate depends on the stage of gastric cancer at the time of diagnosis[2]. In Western countries, the 5-year survival ranged from 5%-20%, whereas in Japan, the survival rate was about 50% due to early diagnosis[3]. Today, surgery remains the mainstay of potentially curative treatment for gastric cancer. Nevertheless, over 50% of the patients may still develop recurrence after curative resection. A good screening method for early detection is the best way to reduce gastric cancer mortality. Due to the associated side effects with endoscopy, including perforation, aspiration pneumonia or bleeding, thus making endoscopic screening not a common practice in the community. Hence there is an urge for the discovery of biomarkers for non-invasive early detection in gastric cancer patients.

MicroRNAs (miRNAs) are endogenous noncoding regulatory RNAs with 17-25 nucleotides, which play important roles in post-transcriptional gene regulation. The ability to bind complementary sequences in 3’-untranslated regions (3’-UTR) of various target mRNAs leading to direct mRNA degradation or translational repression. MiRNAs regulate gene expression and contribute to development, differentiation, inflammation, and carcinogenesis. At present, over 24500 entries have been listed in the miR Registry Database (release 20, http://www.mirbase.org/), implicating more than 30000 mature miRNA products, and the number of miRNAs is expected to increase exponentially in the future. Studies have shown that over 30% of human genes are regulated by miRNAs, in which a single miRNA controls over hundreds of RNA.

MiRNA plays pivotal role in biological processes including cell proliferation, metastasis, differentiation, development and apoptosis[4,5]. Accumulating evidence showed that the miRNA profiles were differentially expressed in cancerous tissues and normal counterparts[6,7]. The fact that miRNAs are very specific for different types of tissues and even for types of cells within those tissues, many studies profiled the miRNA patterns in various cancer types, which put forward the diagnostic and prognostic values of miRNAs in clinical applications. To date, there are many different platforms to study the expressions of miRNA, for example, northern blots, real-time PCR, primer extension and microarrays. Microarray has been the most widely used for miRNA research, not only it is a more user-friendly platform, its high throughput property makes possible to profile the whole genome of miRNA.

Deregulation of miRNA in human cancers was a result of impaired miRNA-biogenesis, genomic or epigenetic alterations, leading to the proto-oncogenic or tumor suppressive role of miRNA in tumorigenesis. For oncogenic miRNA, inhibitor of miRNA blocks the function of miRNA; while tumor suppressive miRNA, reconstitute miRNA precursor produce an anti-tumor effect. Hence targeting specific miRNA could be a possible alternative for treating gastric cancer patients.

The biogenesis of miRNA involves the transcription of miRNA by RNA polymerase II to form large double-stranded precursor (Pri-miRNA) in the nucleus. It appeared in a hairpin structure which was then cleaved by RNase III endonuclease Drosha to produce a 60-70 nucleotides hairpin precursor (Pre-miRNA) and exported to the cytoplasm by a nuclear membrane export receptor Expotin-5. Pre-miRNA is then cleaved by another RNase III enzyme Dicer to form a mature double-stranded miRNA (about 22 nucleotides) which comprises a mature miRNA guide strand and passenger strand. The guide strand (mature miRNA) is incorporated into RNA-induced silencing complex (RISC) to target mRNA via 2 mechanisms: (1) cleavage of mRNA at the miRNA binding site; and (2) repression of translation. Due to the fact that miRNAs are present in the genomic regions that involved in cancers, miRNAs become increasingly recognized as potential marker for diagnosis and prognosis.

Microarray is a useful and convenient platform to profile the miRNA expression in human cancers. By comparing the expression in gastric cancerous tissue with non-tumor tissue, the distinct miRNA signatures are associated with progression and perhaps prognosis of gastric cancer[8]. The miRNA signatures have a higher accuracy and reproducibility than mRNA expression profiles, over 80% of the paired gastric samples were classified correctly by the miRNA signatures. Prediction and validation of downstream targets become increasingly important in miRNA research, here; we listed the targets of upregulated and downregulated miRNAs in gastric cancer (Table 1).

Over 92% of the gastric cancer patients demonstrated an upregulation of miR-21 in solid human tumors, including gastric cancer[9]. Currently, it has been reported that miR-21 was upregulated not only in cancerous tissues but also in Helicobacter pylori (H. pylori)-infected gastric mucosa[10]. Gastric cancer is a result of multistep and long-term interactions between genetic and environmental factors which process from chronic gastritis, atrophic gastritis, intestinal metaplasia, glandular atrophy and dysplasia[11]. The miRNA that was associated with H. pylori-induced inflammation was miR-218, overexpression of this miR abrogated nuclear factor-kappa B (NF-κB) activation[12]. It was hypothesized that miR-21 might augment the progression of infected normal mucosa to chronic gastritis with unknown mechanisms. The signal transducer and activator of transcription 3 (STAT3) activated the induction of miR-21[13]. Activation of NF-κB and interleukin (IL)-6 stimulated STAT3 signaling, which may explain the H. pylori-mediated upregulation of miR-21. Interestingly, there is a transcription binding site of NF-κB located in miR-21 transcriptional elements, suggesting that miR-21 upregulation is a result of NF-κB activation in gastric cancer[14].

Numerous evidence revealed that miR-21 attributed to gastric cancer through enhanced cell proliferation and inhibited apoptosis. On top of that, miR-21 also has the ability to incite cell invasion and migration. It has been reported that RECK, a tumor suppressor gene, is the target of miR-21. It involves in the process of metastasis and angiogenesis through regulating metalloproteases (MMPs)[10]. MiR-125b, miR-199a and miR-100 have been shown to be the most important progression-related miRNAs in gastric cancer and pancreatic adenocarcinoma[12,15], implicating that miRNA may have different functions depends upon the tumor site. There are some miRNAs (miR-32, miR-182 and miR-143) that are found to be associated with intestinal-type gastric cancer, this study implicated the usefulness of miRNA expression profiles may serve as diagnostic biomarkers for different subtypes of gastric cancers[16]. Expressions of miR-143, miR-145, miR-9, miR-443, miR-31, and miR-34 have been reported to downregulate in gastric cancer[17-19]. The roles of miR-143 and miR-145 on cell proliferation have also been demonstrated in other gastrointestinal cancers[20]. Ectopic expression of miR-143 and miR-145 showed significant growth retardation and sensitized to 5-fluorouracil treatment in gastric cancer cells[17].

There is a strong correlation of elevated expression of high mobility group AT-hook 2 (HMGA2), a nonhistone chromosomal protein that modulate translation, with tumor invasiveness in gastric cancer. HMGA2 was inversely regulated by the let-7 miRNA family. High expression of HMGA2 correlated with tumor invasion and was an independent prognostic factor in gastric cancer[21]. In addition, miR-214 was reported to modulate hedgehog signaling, where activation of hedgehog contributes to gastric cancer[8]. High expression of miR-214 was identified to correlate with unfavorable outcome in gastric cancer. In ovarian cancer, miR-214 regulated the downstream target PTEN to induce cell survival[22]. Furthermore, miR-196b expression has been shown to be significantly higher in gastric cancer tissues than normal counterparts[23]. Silence of ETS2 (a transcriptional modulator) enhanced miR-196b expression which promoted gastric cancer cell migration and invasion by increased vimentin, MMP-2 and MMP-9 expressions and suppressed E-cadherin expression.

Recently, many researchers are looking into the potential use of miRNAs as prognostic tools other than diagnostic application, however, this specific marker needs to correlate closely with clinical outcomes and metastatic potential. Profiling of miRNAs seems to be advantageous over mRNAs in terms of cancer phenotypes differentiation. A current report suggested a prognostic signature for gastric cancer which consist of four risk miRNAs (miR-10b, miR-21, miR-223 and miR-338; with hazard ratio > 1) and three protective miRNAs (miR-30a-5p, miR-126 and let-7a; hazard ratio < 1), and was associated with clinical outcomes[39]. Another study showed that low expressions of miR-21 and miR-181b are associated with overall survival in patients treated with S-1 and doxifluridine[40]. Low expression of miR-125a-3p is correlated with tumor size, invasion, metastasis and advanced clinical stage, and an independent prognostic marker for gastric cancer[41]. Ectopic expression of miR-125a-3p showed remarkable retardation of gastric cancer growth in vitro, illustrating the tumor suppressive property and potential clinical use of miR-125a-3p. On the other hand, the other strand of miR-125a, miR-125-5p also showed similar tumor suppressive effect in gastric cancer, and low expression was associated with poor prognosis[42]. Other study showed that HMGA2 expression was directly correlated with tumor invasiveness and prognosis, which is modulated by let-7 family[21]. A similar observation was also seen in pituitary adenomas, in which low expression of let-7 induced HGMA2 and related to tumor proliferation and invasion[43].

Though surgery is a promising treatment for gastric cancer, however, over 50% of patients will have recurrence or systemic metastasis. Chemotherapy has been widely used in the management of patients with advanced disease to prolong survival. However, multidrug resistance (MDR) often hinders successful chemotherapeutic treatment. And there are the two mechanisms exist in tumor cells towards cytostatic drugs: intrinsic or acquired drug resistance.

MiR-138 was found to be highly expressed in MDR variant when compared with its parental leukemia cells[44]. Increased accumulation of adriamycin inside the cells resulted in apoptosis through the repression of P-gylcoprotein (P-gp) and Bcl2 expressions. Multidrug resistance-associated protein (MRP-1) is another transporter, other than P-gp, that expels cytostatic drugs out of the cell. Transfection with miR-326 reduced expression of MRP-1 and sensitized the cells to VP-16 and doxorubicin in resistance cells[45]. Similar approach by using miR-27a and miR-451 antagomirs decreased P-gp and multidrug resistance-1 (MDR1) mRNA expression which increased the intracellular concentration of cytotoxic drugs in the cells[46]. When the miRNA expression profiles in MDR variant were compared with its parental gastric cell line, a small subset of miRs has been excavated. Downregulation of miR-15b and miR-16 in MDR cells restrained the cells from apoptosis by modulation of Bcl2[47]. Ectopic expression of miR-15b and miR-16 sensitized MDR cells to chemotherapeutic agent, which lead to cell apoptosis[48]. These findings are of great impact on the functions of miRNAs, and notably improve/sensitize the efficacy of chemotherapeutic agent in gastric cancer.

The important roles of miRNAs in cancer biology have been extensively studied in recent years. Ideally, miRNA could be a better therapeutic tool over single gene therapy, due to its ability to target multiple genes. However, several problems were encountered when translated into clinical applications. Firstly, as agreed by most of the researchers, identification of the downstream targets of miRNA is intricating. Despite several target prediction algorithms (e.g., TargetScan, miRTar and miRecords etc.) currently available to envisage miRNA targets, however, the complication is that various miRNA could target the same mRNA, hence, justification from results algorithms are inevitable. Microarray and real-time RT-PCR are the analyzing techniques that are commonly used for screening and validation. With the use of locked nucleic acid oligos for in situ hybridization, knockdown or PCR studies, these enhance the sensitivity and specificity for the analysis of miRNAs. Detailed studies on the mechanism of miRNA biogenesis and interaction with RISC in gastric carcinogenesis would favor the manipulation of miRNA expression which might help in designing a more specific chemotherapeutic agent.

There are increasing demand on miRNA profiling with the expansion of its application into differentiating subtypes and predict treatment response in gastric cancers. The stability of miRNAs in the circulation makes it suitable as diagnostic and prognostic markers in various cancers. Also, it is likely to acquire a miRNA signature, instead of a single miRNA, to provide useful information for the clinicians to make decision on personalized management of the disease. With a better understanding of these miRNAs and their target genes, it would open up new perspectives for more sophisticated and effective therapeutic agents for treating gastric cancer.