Int J Clin Exp Med 2018;11(2):664-673 www.ijcem.com /ISSN:1940-5901/IJCEM0054962 Original Article MicroRNA-27a-3p promotes the metastasis and epithelial-mesenchymal transition of gastric cancer via targeting RUNX1 Xiao-Dong Wang1, Qing-He Han2, Bin Song3, Xiao Yang4, Bo Kan5, Ying Song1*, Qing-Hai Yuan2* 1Digestive Endoscopic Department, The Second Hospital of Jilin University, Changchun, Jilin, China; Departments of 2Radiology, 3Neurosurgery, 4Urinary Surgery, The Second Hospital of Jilin University, Changchun, Jilin, China; 5Clinical Laboratory, The Second Hospital of Jilin University, Changchun, Jilin, China. *Equal contributors. Received April 9, 2017; Accepted August 6, 2017; Epub February 15, 2018; Published February 28, 2018 Abstract: Increasing evidence shows that microRNAs (miRNAs) are a novel class of gene regulators, and play a vital role in tumor development and progression. MicroRNA-27a (miR-27a) was previously reported dysregulated expres- sion in human carcinoma, including gastric cancer (GC). However, till now, the mechanism that miR-27a functions as an oncogene is still not well known. As we all known, there are two isoforms of mature miR-27a, miR-27a-5p and miR-27a-3p. In this study, we figure out the link between miR-27a-3p and RUNX1 and the overexpression of miR-27a-3p markedly promote gastric cancer cell proliferation, invasion, and migration in vitro. We also describe that miR-27a-3p have effect on the EMT process. Furthermore, miR-27a-3p repressed RUNX1 expression by directly binding to 3-untranslated region (UTR) of RUNX1 in gastric cancer, and the inverse correlation was observed be- tween the expressions of miR-27a-3p and RUNX1 mRNA in primary GC tissues. In turn, the restoration of RUNX1 led to suppressed proliferation, invasion, and migration of GC cells. In vivo, miR-27a-3p promotes tumor growth of GC. Taken together, these results suggested that the miR-27a-3p/RUNX1 axis promotes GC progression by directly downregulating RUNX1 expression and may be employed as a novel prognostic marker and therapeutic target in the management of gastric cancer. Keywords: miR-27a-3p, gastric cancer, RUNX1, epithelial-mesenchymal transition (EMT), metastasis Introduction in length that suppress gene expression by directly binding to the complementary sequenc- Gastric cancer (GC) is among the most common es in 3’-untranslated region (UTR) of their target malignancies and the second leading cause of gene mRNA to induce mRNA degradation and cancer-related death worldwide [1]. It is pro- suppression of translation [4, 5]. Rapidly grow- posed that the incidence of the GC has signifi- ing evidence suggests that miRNAs were invol- cant region differences, particularly in Eastern ved in diverse physiological and pathological Asia, Eastern Europe, and South America [2]. processes including cell proliferation, differen- Particular clinical methods are currently under- tiation, motility, apoptosis, angiogenesis and way to improve diagnosis and treatment for the metastasis [6]. An increasing number of miR- GC patients, including advanced surgical resec- NAs were shown to be involved in metastasis tion, anti-Her-2 molecular-targeted therapies, and invasion of GC, including miR-21 [7], miR- radiation and chemotherapeutic therapy; nev- 25 [8], miR-29 [9, 10], miR-30b [11], miR-141 ertheless, the overall 5-year survival rates [12], and miR-223 [13]. Although compelling remains as poor as 10% to 40% [3]. Gastric evidences indicated that miR-27a-3p is associ- cancer is a complex genetic disease accompa- ated with gastric carcinogenesis, where miR- nying various genetic and epigenetic alteration. 27a-3p as an oncogene and a promising diag- However, the common molecular mechanism nostic biomarker [14], the role of miR-27a-3p in underlying the proliferation, migration and inva- GC metastasis and invasion remains to be sion of GC remain to be elucidated. demonstrated. MicroRNAs (miRNAs) are a kind of diverse, RUNX1 belongs to the RUNX family with the short, non-coding RNA with 20-22 nucleotides other two members, RUNX2 and RUNX3, each Metastasis of gastric cancer having distinct tissue-specific express patterns humidified environment with 5% CO2 at 37°C, and cell-context dependent functions through respectively. interactions with their common partner, core binding factor beta CBFβ [15]. As a transcrip- RNA extraction and quantitative RT-PCR tion factor, RUNX1 is involved in multiple signal- ing pathways, including the RAS, ERK, TGF-β Total RNA from frozen tissues or cells were and WNT pathways, in biological process and extracted by TRIzol (Invitrogen) according to cancer [16-19]. Early evidences have described the instructions of manufacturer. Isolated total that RUNX1 is deeply associated with leukemia RNA was using a reverse transcription kit (Cat. [20] and breast cancer [21, 22]. However, the No. RR037a; Takara, Tokyo, Japan) to obtain regulatory activity of RUNX1 is not confined to the miRNAs or mRNAs SYBR® Premix Ex Taq™ hematopoietic lineage and breast tissue. Accu- (Cat. No. RR420A, Takara) were employed for mulating evidence suggests significant contri- quantitative RT-PCR of miRNA and mRNA on an butions by RUNX1 as a tumor suppressor in ABI 7500 PCR system (Applied Biosystems). development of gastric cancer [23]. The primers of used in this study are listed as follows: RUNX1-forward: 5’-CTGCCCATCGCTTT- In this present study, we examined the expres- CAAG--GT-3’; RUNX1-reverse: 5’-GCCGAGTAGT- sion patterns of miR-27a-3p and RUNX1 in gas- TTTC ATCATTGCC-3’; GAPDH-forward: 5’-AGAA- tric cancer tissues and cell lines, and validated GGCTGGGGCTCATTTG-3’; GAPDH-reverse: 5’- the negative relationship between miR-27a-3p AGGGGCCATCCACAGTCTTC-3’. Primers for U6 and RUNX1. We found that RUNX1 is a direct and miR27a-3p were purchased from Gene- target of miR-27a-3p, and RUNX1 overexpres- Copoeia (RiboBio). U6 RNA and GAPDH were sion could partially attenuate the effect of miR- used as miRNA and mRNA internal control 27-3p in GC. Furthermore, our data showed respectively. Each sample was run in triplicate. that miR-27a-3p directly downregulated RUNX1 The relative expression was calculated using expression through binding to RUNX1-3’-UTR, the relative quantification equation (RQ) = -ΔΔCt2 . and suggest an important regulatory role in the EMT process in GC. Western blot assay Materials and methods Total protein was extracted from cultured cells by lysing in RIPA buffer, and the lysates were Patients and cancer specimens analyzed using the standard Western blotting analyses. GAPDH proteins serving as a loading Fresh frozen human gastric GC samples and control. Antibodies used: RUNX1 from Abcam; their corresponding normal gastric tissue sam- Fibronectin, Vimentin from Sigma; E-cadherin, ples were obtained from the Tumor Bank α-catenin, γ-catenin, N-cadherin from BD Bio- Facility of Tianjin Medical University Cancer science. The detection was achieved by incuba- Institute and Hospital and National Foundation tion with HRP-conjugated secondary Abs (Santa of Cancer Research (TBF of TMUCIH & NFCR). Cruz Biotechnology). Bound proteins were visu- The types of all the tumors were confirmed by alized with ECL Plus Western Blot Detection the pathologic analysis. The use of all the Reagents (GE Healthcare). human materials were in accordance with the ethical guidelines of the Declaration of Helsinki Plasmid construction and transfection and were approved by the Institutional Review Board. The full length cDNA of RUNX1 was inserted into the empty pCMV-Tag2B vector in the Cell culture restriction digestion sites EcoRI and XhoI. For the construction of human RUNX1-3’UTR, the Human GC cell lines of HGC-27, MGC-803 and potential miR-27a-3p-binding sequences con- BGC823 were purchased from Cell Bank of taining wild-type or mutant seed region were Chinese Academy of Sciences (Shanghai, Chi- synthesized and cloned into the SpeI and na). Cells were cultured in DMEM/high glucose HindIII restriction enzyme digestion sites of (HGC-27, BGC823 or HGC-27) medium (Hy- pMIR-REPORT luciferase vector (Applied Bio- Clone) supplemented with 10% FBS penicillin systems). The following primers were used to (100 U/mL) and streptomycin (100 ug/mL) in a generate specific fragments: RUNX1-forward:5’- 665 Int J Clin Exp Med 2018;11(2):664-673 Metastasis of gastric cancer Figure 1. miR-27a-3p was increased in GC tissues and correlated with lymph node metastatic capacity in GC tis- sues. A. Expression of miR-27a-3p was upregulated in GC tissues compared to normal gastric tissues based on the data downloaded from the GEO database (GSE93415). B. qPCR data of miR-27a-3p in primary gastric cancer (with or without lymph node metastasis). C. Measurement of miR-27a-3p expression in 21 GC specimens and adjacent normal gastric tissues. CGGAATTCATGGCTTCAGACAGCATA--TT-3’, RUN- plot the growth curve, the seeding cells were X1-reverse: 5’-CCCTCGAGTCAGTAGGGCCTCCA- trypsinized and manually countedbyusing Neu- CACGG-3’,UTR-wildtype-forward: 5’-GGACTAGT- bauer chamber under the microscope. The CAGGATCTCGCTGTAGGT-3’, UTR-wildtype-rever- growth curve was then plotted using the cell se: 5’-CCAAGCTTATGGTCAAAGCAAGAAAGAA-3’. counting data. The mutant fragment were also constructed as a negative control (mut1: ACUGUGAA to AAGG- Cell invasion assay GGCA, mut2: CUGUGAA to CGAUACA). Lipofec- tamine 2000 reagent (Invitrogen) was used for Transwell chamber filters (Millipore) were coat- the transfection of plasmids, miRNA mimics ed with Matrigel. After transfection with miR- and miRNA inhibitor. NC, miR-27a-3p mimics, empty vector, and RUNX1 vector respectively, BGC823 cells were Luciferase reporter plasmids and assays suspended in serum-free DMEM media at a concentration of 1×105/mL, and then the count MGC-803 cells were seeded in 24-well plates of 50,000 cells (500 μL) per well were seed- at a density of 5×104 cells per well and tran- ed into the upper chamber of the transwell. siently transfected with mutant or wild-type Following, the chamber was transferred to a luciferase reporter plasmids and Renilla lucifer- well including 500 μL of DMEM media contain- ase control vector (pRL-TK), and then the cells ing 10% fetal bovine serum.