microRNA analysis of gastric cancer patients from Saudi Arabian population

Clinical samples

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^-∆CT method [17], wherein ∆C_t = C_t(tumor)-C_t(control) and control is the sample without treatment.

Expression profiles of miRNA in gastric cancer

We used miRNA 4.0 array gene chips, to evaluate miRNA expression profiles between FFPE GC and normal tissues. By adjusting average change >2-fold and value <0.05 as a cut-off level, 1082 expressed genes were differentially expressed between the cancerous and normal control. Among them, one-way ANOVA showed that only 33 miRNAs had significant expression, where 8 (has-mir-193a-5p, has-mir-200c-3p, has-mir-1227-5p, has-mir-1909-3p, has-mir-1378, has-mir-3613-3p, has-mir-5196-5p and has-mir-7704) were apparently expressed in early stage cancer tissues (p < 0.05) and 25 (hsa-miR-22-3p, hsa-miR-28-3p, hsa-miR-29a-3p, hsa-miR-99a-5p, hsa-miR-27b-3p, hsa-miR-143-3p, hsa-miR-145-5p, hsa-miR-1251-5p, hsa-miR-126-3p, hsa-miR-185-5p, hsa-miR-200c-3p, hsa-miR-130b-3p, hsa-miR-378a-3p, hsa-miR-877-5p, hsa-miR-128, hsa-miR-378c, hsa-miR-3613-3p, hsa-miR-4532, hsa-miR-4668-5p, hsa-miR-4800-5p, hsa-miR-6124, hsa-miR-6812-5p, hsa-miR-7150, hsa-miR-7704, hsa-miR-7847-3p) were in advanced stage cancer samples as compare to normal control (p < 0.05) (Table. 2). As we used 2-fold as cut off value so some aberrant miRNAs in the chip analysis may be neglected. Differentially expressed genes were related to signal transduction, metabolism, angiogenesis, cell structure and cycle, and gene protein expression were mostly down-regulated. We found 13 new among 33 aberrant miRNAs that have not been reported before in GC studies. These miRNAs named as hsa-miR-1227-5p, hsa-miR-1909-3p, hsa-miR-3178, hsa-miR-7704 (in early-stage GC) and hsa-miR-1251-5p, hsa-miR-877-5p, hsa-mir-1281, hsa-miR-4532, hsa-miR-6124, hsa-miR-6812-5p, hsa-mir-7150, hsa-mir-7704 and hsa-mir-7847-3p (in late-stage GC) (Table. 2). There are many concerns about the integrity of miRNA from FFPE tissues and their suitability for microarray as RNA might be modified in chemical reaction and become fragmented in FFPE tissues [18] but due to very small size of miRNA (about 19-22 nucleotides) it’s not degraded in FFPE preparation. Many previous studies of miRNA microarray supporting a correlation in miRNA expression results from fresh and FFPE cancer tissues samples [19–21]. We found in all samples good value of RNA integrity and optical density (data not shown). In order to assess miRNA data after filtering we built box plot to visualize the distribution of data. The distributions of log2 ratios are almost same among miRNA chips. The probe-level data for the box plots are distributed from about 2–14 on the log scale (Fig. 1a). To check variation between cancer and normal tissues samples we assess the data using scatter-plot. Around 70–80 % miRNA seems to be same in cancerous and normal tissues used in this study (Fig. 1b).

Transcript ID(Array Design)	Accession	Fold Change (linear) (GC vs. Control)	FDR p-value (GC vs. Control)
Detected in early stage (I and II)
hsa-miR-193a-5p	MIMAT0004614	−2.06796	0.77052133
hsa-miR-200c-3p	MIMAT0000617	−22.552013	0.77052133
hsa-miR-1227-5p	MIMAT0022941	−1.717431	0.77052133
hsa-miR-1909-3p	MIMAT0007883	−1.951977	0.77052133
hsa-miR-3178	MIMAT0015055	−1.826664	0.72999652
hsa-miR-3613-3p	MIMAT0017991	2.981591	0.72999652
hsa-miR-5196-5p	MIMAT0021128	−1.959981	0.77052133
hsa-miR-7704	MIMAT0030019	−1.893422	0.72999652
U79	_	1.738033	0.74726754
Detected in early stage (III and IV)
hsa-miR-22-3p	MIMAT0000077	−1.90434	0.10052633
hsa-miR-28-3p	MIMAT0004502	−1.87632	0.145678142
hsa-miR-29a-3p	MIMAT0000427	−1.96579	0.196109756
hsa-miR-99a-5p	MIMAT0000097	−1.97584	0.21331947
hsa-miR-27b-3p	MIMAT0000419	−2.29128	0.275381924
hsa-miR-143-3p	MIMAT0000435	−2.20679	0.132397944
hsa-miR-145-5p	MIMAT0000437	−2.20593	0.271934938
hsa-miR-1251-5p	MIMAT0028700	−2.70147	0.014814394
hsa-miR-126-3p	MIMAT0005903	−1.9555	0.292936143
hsa-miR-185-5p	MIMAT0001339	−1.55828	0.212836342
hsa-miR-200c-3p	MIMAT0000455	−2.18027	0.113424849
hsa-miR-130b-3p	MIMAT0000691	−2.41458	0.126769698
hsa-miR-378a-3p	MIMAT0000732	−1.78021	0.074619747
hsa-miR-877-5p	MIMAT0004949	1.696113	0.064059796
hsa-miR-1281	MIMAT0005939	−1.97312	0.064954696
hsa-miR-378c	MIMAT0016847	−2.51734	0.016444786
hsa-miR-3613-3p	MIMAT0017991	2.507303	0.212235198
hsa-miR-4532	MIMAT0019071	−2.05911	0.046570249
hsa-miR-4668-5p	MIMAT0019745	2.097354	0.268773791
hsa-miR-4800-5p	MIMAT0019978	1.510005	0.257957349
hsa-miR-6124	MIMAT0024597	2.156251	0.088004364
hsa-miR-6812-5p	MIMAT0027524	1.583428	0.250692071
hsa-miR-7150	MIMAT0028211	1.628161	0.277892706
hsa-miR-7704	MIMAT0030019	−1.64036	0.016444786
hsa-miR-7847-3p	MIMAT0030422	1.59577	0.244713443
U3	_	1.780994	0.060093686
U63	_	−1.69859	0.116804752
U3-2B	_	1.761599	0.070969781
U3-2	_	1.761599	0.070969781
U3-3	_	1.761599	0.070969781
U3-4	_	1.761599	0.070969781

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 early-stage 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 down-regulation 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 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 [39–41].

In late-stage GC tissue samples, 8 miRNAs were up-regulated and 17 showed down-regulation with 10 new miRNAs (has-mir-1251-5p, has-mir-877-5p, has-mir-1281, has-mir-378c, has-mir-4532, has-mir-6124, has-mir-6812-5p, has-mir-7150, has-mir-7704 and has-mir-7847-3p) not reported before (Table. 2). Many miRNAs involve in tumorigenesis and altered in variety of cancers are detected in our study such as tumor-suppressor (has-mir-22-3p, has-mir-28-3p, has-mir-99a-5p, has-mir-130b-3p, has-mir-378a-3p, has-mir-3613-3p, has-mir-4668-5p and has-mir-4800-5p) are reported before in colorectal [42], cervical [22] and breast cancer [43]. Some of the miRNAs (has-mir-29a-3p, has-mir-27b-3p, has-mir-143-3p, has-mir-145-5p, has-mir-126-3p, has-mir-185-5p, has-mir-200c-3p) differential expressed in late-stage GC samples had similar kind of expression in previously published GC studies. The down-regulation of has-miR-29a-3p in our study is consistent with previous study of GC where decreased expression of has-miR-29a-3p promotes cell proliferation by suppressing the expression of cell cycles regulators [44, 45]. The receptor tyrosine kinase like orphan receptor 1 (ROR1) protein is an oncogenic protein. The miR-27b-3p suppress the expression of ROR1 in GC hence function as tumor suppressor [46]. Similar expression of has-mir-27b-3p has been detected in our study. Furthermore, other GC related miRNAs such as has-mir-143-3p, has-mir-145-5p, has-mir-126-3p, has-mir-185-5p have found to be down-regulated in our study have shown similar expression in other studies and reported as tumor-suppressors [47–49]. In addition to miRNAs, we found seven small nucleolar RNA (snoRNAs) differentially expressed in both early and late-stage tumor samples. In our study, only 1 snoRNA (U63) showed down-regulation and others (U79, U3, U3-2b, U3-2, U3-3 and U3-4) were up-regulated and are related to human C/D box group (Table. 2). Several recent studies have highlighted the role of snoRNAs as miRNAs precursors and have similar function like miRNAs [50, 51]. Another study on Epstein-Barr virus genome, suggests their role in infection [52]. What role these snoRNAs may play in GC is interesting to explore.

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 [24–34]. 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.