microRNA analysis of gastric cancer patients from Saudi Arabian population

Background The role of small non-coding microRNAs (miRNAs) in several types of cancer has been evident. However, its expression studies have never been performed in gastric cancer (GC) patients from Saudi population. First time this study was conducted to identify miRNAs that are differentially expressed in GC patients compared with normal controls. Methods We investigated the role of miRNAs in GC patients using formalin-fixed paraffin-embedded (FFPE) tissues of 34 samples from GC patients (early stage = 7 and late-stage = 26) and 15 from normal control. We have used miRNA microarray analysis and validated the results by Real-time quantitative PCR (RT-qPCR). Results We obtained data of 1082 expressed genes, from cancer tissues and noncancerous tissues (49 samples in total). Where 129 genes were up-regulated (P > 0.05) and 953 genes (P > 0.05) were down-regulated in 49 FFPE tissue samples. Only 33 miRNAs had significant expression in early and late-stage cancer tissues. After candidate miRNAs were selected, RT-qPCR further confirmed that four miRNAs (hsa-miR-200c-3p, hsa-miR-3613, hsa-miR-27b-3p, hsa-miR-4668-5p) were significantly aberrant in GC tissues compared to the normal gastric tissues. Conclusions In this study we provide miRNAs profile of GC where many miRNAs showed aberrant expression from normal tissues, suggesting their involvement in the development and progression of gastric cancer. In early and late-stage miR-200c-3p showed significant down regulation as compare to control samples. Many of miRNAs reported in our study showing up-regulation are new and not reported before may be due to population difference. In conclusion, our results suggest that miR-200c-3p had potential to use as diagnostic biomarker for distinguishing GC patients from normal individuals and can be used for diagnosis of cancer at early stage.


Background
Gastric Cancer (GC) is the second major cause for cancer-related deaths worldwide mostly prevalent in Asian countries, including Korea, Japan and China [1]. In Saudi Arabia, For all cancer types about 2.9 % is accounted for GC and it is ranked 11 th among both male and female population [2]. Multiple factors contribute to the progression of gastric tumors, including diet rich in salted and nitrated food, alcohol consumptions, low consumption of fruits and vegetables, use of tobacco and, especially, Helicobacter pylori infection [3]. Like other malignancies, both genetic and epigenetic factors are involved in the pathogenesis of gastric cancer [4]. Mortality of the disease is potentially reduces if it is diagnosed at early stages in gastric cancer [5].
Tumor markers always remain helpful at early stages in screening of high risks groups. But, the current cancer tumor markers have limited performance in detection of gastric cancer due to low sensitivity and specificity [6]. Therefore, it is required to discover some novel diagnostic biomarkers for the early detection of this malignancy. MicroRNAs (miRNAs) are small group of RNA molecules that regulate expression of different genes by binding to mRNAs. They play an important role in different cellular processes that are necessary to maintain a normal physiological condition. In cancer, particular miRNAs act as tumor supressor or oncogenic hence considered as biomarkers for early diagnosis and accurate prognosis of cancer [7]. They are considered to control a variety of functions in tumor cell including cell proliferation, migration, invasion, differentiation, apoptosis and metabolic processes [8]. Many previous studies suggest that dysregulated expression of miRNAs associated with cancers, and function as tumor inducer and suppressor and their expression may play an important role in cancer progression. The expression of miRNAs in different oncogenic pathways suggests their importance during carcinogenesis [9].
Microarray-based gene expression profiling is a potential technique to study the expression of miRNAs in GC patients [10].
Several previous studies have reported that miRNAs may be used as diagnostic biomarkers in different cancer types [11]. Fang et al. [12] has suggested some oncogenic miRNAs (miR-10b, miR-21, miR-223 and miR-338) and tumor suppressive miRNAs (miR-30a-5p, miR-126 and let-7a) as prognostic signatures in GC patients. Jiang et al. [13] has reported higher expression of miR-421 in early stage GC patients hence suggesting its role as diagnostic biomarker in GC. Using miRNA array in GC patients, abnormal expression of miRNA profiles with up and down-regulated miRNAs has been reported [14].
Although the importance of miRNAs as important prognostic factors in patients with GC is confirmed, but data on the miRNA signature of GC in the Saudi population is missing. Therefore, the present study was undertaken to detect the miRNA expression profile of GC patients and normal gastric FFPE tissue using miRNA 4.0 microarrays. In both early and late GC tumors, 33 miRNAs were found to be differentially expressed and significantly aberrants were validated using RT-qPCR.

Clinical samples
In this study, 34 gastric cancer tissues samples were collected from King Abdulaziz University (KAU) hospital Jeddah. Informed consent was obtained from patients undergoing a surgical procedure to remove a portion of gastric cancer. This study was approved by medical ethical committee of KAU, Jeddah Saudi Arabia (Refer-ence#174-15). Using the tumor-node-metastasis (TNM) staging of the International Union Against Cancer (1997) [15] all the tumor samples were staged and graded according to the World Health Organization criteria [16]. From these specimens, we have collected 34 FFPE biopsy samples from gastric cancer patients and 15 FFPE gastric biopsy samples from healthy individuals along with detailed clinical history from KAU hospital (Table. 1).

RNA extraction and quality analysis
RNeasy FFPE kit was used for extraction of RNA from FFPE tissues according to manufacturer's instructions. RNA was further purified using DNase I treatment (Ambion, Austin, TX) to eliminate any contaminating DNA. RNA concentrations were calculated using a Nanodrop ND-1000 spectrophotometer (NanoDrop, Wilmington, USA). RNA integrity was evaluated by running electropherograms and RNA integrity number, RIN (a correlative measure that indicates intactness of mRNA) was determined using the RNA 6000 PicoAssay for the Bioanalyzer 2100 (Agilent Technologies, Palo Alto, USA).

Affymetrix miRNA arrays methods
The Affymetrix Genechip miRNA 4.0 array process was carried out according to the manufacturer's instructions. 1000 ng RNA samples were labeled with the FlashTag™ Biotin using RNA Labeling Kit (Genisphere, Hatfield, PA, USA). The labeled RNA was further quantified, fractionated and hybridized to the miRNA microarray according to the standard protocol. The labeled RNA was then heated at 99°C and then at 45°C (for 5 mins at both temp). An Affymetrix® 450 Fluidics Station was used for RNA-array hybridization with agitation at 60 rotations per minute at 48°C for 16 h. The chips were washed and stained using a Genechip Fluidics Station 450 (Affymetrix, Santa Clara, California, United States). The chips were then scanned with an Affymetrix GCS 3000 canner (Affymetrix, Santa Clara, California, United States). Using the Affymetrix® GeneChip™ Command Console software (AGCC) signal values were computed.

Raw data preparation and statistic analysis
Raw data were extracted automatically in Affymetrix data extraction protocol using the software provided by AGCC. The CEL files import, miRNA level RMA + DABG-All analysis and result export using Affymetrix® Expression Console™ Software. Array data were filtered by probes annotated species. The comparative analysis between GC samples and control samples was carried out using independent T-test and fold-change where the null hypothesis was that no difference exists among two groups. False discovery rate (FDR) was further controlled by the adjustment of p value using Benjamini-Hochberg algorithm. All Statistical tests and visualization of differentially expressed genes were done using R statistical language v. 3.1.2.

MiRNA quantification by real-time quantitative PCR (RT-qPCR)
For miRNA quantification SYBR Green RT-qPCR assay was used. Using miScript Reverse Transcription kit (Qiagen) one microgram of RNA was reverse-transcribed to cDNA. Further using miScript SYBR Green PCR kit (Qiagen) along with Universal primer and the miRNA-specific primers qPCR is performed in ABI 7900 Real-time PCR system (Applied Biosystems). The miRNA-specific primer sequences were designed using miRNA sequences obtained from the miRBase database. Each reaction mixture contained 10 μl of 2× Fast SYBR Green Master Mix (Life Technologies, Grand Island, NY), 0.5 μl of dye (ROX) II (50x), 1 μl of forward primer, 1 μl of reverse primer, 10 μL of distilled water, and 1 μl of cDNA template. Each sample was run in duplicates and U6 miRNA was used as control to normalize the expression levels of miRNAs. Mean expression values of each miRNAs (dublicate) relative to U6 RNA were calculated using the 2 -ΔCT method [17], wherein ΔC t = C t(tumor) -C t(control) and control is the sample without treatment.

Identification of differentially expressed miRNAs in early vs late-stage GC
We analyzed the expression of miRNAs significantly expressed from cancerous (early and last stage) and normal tissues using Volcano plot filtering. For up and down-regulated miRNAs genes we used ≥ 2.0 fold change and value P value (<0.05). In this plot, differentially expressed genes are statistically significant and are shown in red and green points (Fig. 2). We identified 1082 differently expressed miRNA and only 33 had significant expression, where 9 genes showed up-regulation and 24 genes showed down-regulation of expression in GC (early and late-stage GC) as compare to normal gastric tissues. Further hierarchical clustering analyses  (Fig. 3) were performed on the basis of differentially expressed miRNAs from cancerous versus normal tissues. Based on this hierarchical clustering, we have divided miRNAs expression in three groups: early-stage, late-stage and normal tissues. These results showed change in cancerous and control tissue samples. We have identified many new miRNAs from this study from Saudi population that have never been reported in studies related to cancer as well as in GC. Among 33 aberrant miRNAs (early and late-stage GC) 13 were new and rest of the miRNAs were reported before in cancer studies. Similarly, from remaining 20 miRNAs, only 9 were reported in GC studies before. Most of the miRNAs were down-regulated only 9 were up-regulated. In earlystage GC tissue samples, 8 miRNAs had aberrant expression as compare to normal control. Among them 7 was up-regulated and only 1 showed down-regulation (Table. 2). Some of them were not reported before in GC except has-mir-200c-3p. Some miRNAs identified in early-stage cancer tissues samples, including hsa-miR-193a-5p, hsa-miR-3613 and hsa-miR-5196-5p were reported already in cervical cancer and acute lymphoblastic leukemia [22,23]. In our study, significant downregulation of has-mir-200c-3p has been seen in early and late-stage GC tissue samples. The miR-200 family consists of five members and is encoded by two genes on chromosomes 12 and function as tumor regulator [24,25]. It plays an important role in different types of cancer by inhibition, invasion, migration, proliferation and drug resistance. Previous studies have shown down-regulation of miR-200c in GC that is consistent with our study [26,27]. The miR-200c has Fig. 3 Hierarchical clustering for differentially expressed miRNAs in cancer versus normal pass volcano plot. Red indicates high relative expression and green indicates low relative expression been reported in various types of cancers, including colorectal [28], breast [29], cholangiocarcinoma [30], colorectal [31], hepatocellular [32], lung adenocarcinoma [33], ovarian [34] and renal cell carcinoma [35]. The miR-200 family inhibits the epithelial-mesenchymal transition and metastasis by down-regulating ZEB1 and ZEB2 (Zinc-finger E-box Binding homeobox 1 and 2). The miR-200 inhibits angiogenesis by regulating interleukin-8 and CXCL1 secreted by cancer cells. Marked reduction of angiogenesis has been observed after delivering miR-200 members into the tumour endothelium [36]. Another study reported that extracellular matrix proteins and peptidases are targeted by miR-200 and alters the tumor microenvironment to inhibit metastasis [37]. H. pylori induced infection is another risk factor for progression of GC. Matsushima et al. [38] have characterized that decreased expression of tumor suppressor family miR-200 has been seen in H.pylori positive GC patients. Few other studies reported down-regulation of miR-200 in H. pylori-infected GC and increased expression of anti-apoptotic proteins, Bcl-2 and XIAP hence inhibiting apoptosis [39][40][41].
The present study has limitations due to small number of tumor samples (n = 34), especially samples of early stages. Therefore, results of the present study need further validation using larger group of GC patients in future. Our results showed that significant down-regulation of has-mir-200c-3p is markedly observed in early-stage GC as compare to late-stage GC samples.

Validation of dysregulated miRNAs by quantitative PCR analysis
Further using real-time quantitative PCR analysis, results from expression array were validated. We selected four significantly aberrant miRNAs with 2-fold change including hsa-miR-200c-3p, hsa-miR-3613, hsa-miR-27b-3p, hsa-miR-3613 from both stages to test in cancerous versus normal tissues and miRNA U6 was used as control. The miRNA expression analysis of 2 oncogenic hsa-miR-200c and hsa-miR-27b-3p was significantly down-regulated while hsa-miR-3613 and hsa-miR-3613 showed up-regulation in comparison with normal control (Fig. 4a and b) that is consistent with miRNAs array results. Significantly down-regulation of hsa-miR-200c has been reported previously in many studies of cancer indicating the role of hsa-miR-200c as tumor suppressor [24][25][26][27][28][29][30][31][32][33][34]. hsa-miR-200c had multiple roles in regulating tumor cell growth by inhibiting the metastasis and invasion of hepatocellular carcinoma and gastric carcinoma [35,36]. At last, accordant down-regulation of hsa-miR-200c in GC tissues was detected, in comparison with the normal tissues. We also found aberrant expression of some new miRNAs in GC patients, which have no report of aberrant expression in any other studies.

Conclusions
In conclusion, we explored the miRNAs expression of FFPE gastric tissues from GC patients and normal control. Many miRNAs showed aberrant expression in cancerous versus normal control. As it is evident from many previous studies and also the current results strongly suggest that hsa-miR-200c acts as a tumor suppressor miRNA that plays a potential role in the oncogenesis in humans. Hence highlighting its functions as a tumor-suppressive miRNA and prognostic marker in GC patients in Saudi population. The significance and role of aberrant miRNAs expression of GC patients in Saudi population will be better understood as more miRNAs will be identified. Our study may be helpful in future to identify potential prognostic biomarkers for GC.