MiR-92a Promotes Cell Metastasis of Colorectal Cancer Through

ORIGINAL ARTICLE–COLORECTAL CANCERMiR-92a Promotes Cell Metastasis of Colorectal Cancer Through PTEN-Mediated PI3K/AKT PathwayTao-Wei Ke,MD1,2,Po-Li Wei,MD,PhD4,5,6,7,Ken-Tu Yeh,MD3,William Tzu-Liang Chen,MD2,and Ya-Wen Cheng,PhD71Institute of Medicine,Chung-Shan Medical University,Taichung,Taiwan;2Division of Colorectal Surgery,Department of Surgery,China Medical University Hospital,Taichung,Taiwan;3Department of Pathology,Changhua Christian Hospital,Changhua,Taiwan;4Department of Surgery,College of Medicine,Taipei Medical University,Taipei,Taiwan; 5Division of General Surgery,Department of Surgery,Taipei Medical University Hospital,Taipei Medical University, Taipei,Taiwan;6Cancer Center,Taipei Medical University Hospital,Taipei Medical University,Taipei,Taiwan;7Graduate Institute of Cancer Biology and Drug Discovery,Taipei Medical University,Taipei,TaiwanABSTRACTBackground.MicroRNAs regulate gene expression at the posttranscriptional level and play important roles in tumor development,progression,and metastasis.The aim of this study was to investigate the role of microRNA-92a(miR-92a)in metastasis of colorectal cancer(CRC). Methods.One hundredfifty-eight CRC patients were enrolled.The expression of miR-92a,PTEN,and E-cad-herin was analyzed by real-time PCR.Univariate(Kaplan–Meier)analysis was used to analyze primary outcomes included5-year overall survival and tumor recurrence. CRC cell model studies were used to analyze the miR-92a-involved CRC metastasis.Results.The expression of miR-92a in tumor tissues was significantly positively correlated with lymph node metastasis in CRC patients(p=0.012).After adjusting for age,sex,and disease differentiation,this correlation remained significant(p=0.01).In addition,there was a negative correlation between levels of miR-92a and the PTEN gene(p\0.0001).No any association of miR-92a and E-cadherin was found(p=0.128).Patients with high miR-92a/low PTEN had poorer overall survival and dis-ease-free survival rates than those with high miR-92a/high PTEN,low miR-92a/high PTEN,and low miR-92a/low PTEN.The association of levels of miR-92a and PTEN with tumor cell migration in CRC was also confirmed in CRC cell models.Conclusions.We suggest that miR-92a is involved in lymph node metastasis of CRC patients through PTEN-regulated PI3K/AKT signaling pathway.Colorectal cancer(CRC)is one of the most common malignancies and a leading cause of cancer-related mor-bidity and mortality in the United States and Western Europe.1The incidence rate has grown in Taiwan over the past two decades,and it is currently the second most common cancer in Taiwan(Department of Health,ROC, 1990–2011).Epidemiologic studies have suggested that Western diets and a change in lifestyle might play a role in the increased incidence rate.2The classic adenoma–carci-noma sequence is the bedrock of our understanding of tumorigenesis in CRC.2miRNAs are an evolutionarily conserved class of small noncoding RNAs,which regulate gene expression by matching the sequence of target genes,either completely or partially,at the30-untranslated region,causing the sup-pression of protein translation or mRNA degradation.3,4 More than50%of these miRNA genes are located in cancer-associated genomic regions or in fragile sites,sug-gesting that miRNAs are deeply involved in the pathogenesis of cancers,including cancer metastasis.5,6 Tao-Wei Ke and Po-Li Wei contributed equally to this study.Electronic supplementary material The online version of thisarticle(doi:10.1245/s10434-014-4305-2)contains supplementarymaterial,which is available to authorized users.ÓSociety of Surgical Oncology2014First Received:22July2014Y.-W.Cheng,PhDe-mail:ywc@.twAnn Surg OncolDOI10.1245/s10434-014-4305-2Recently,studies have shown that miR-92a is deregulated in several types of cancers7–9and that up-regulation of miR-92a is correlated with a worse prognosis.9The involvement of miR-92a in the tumorigenesis and metas-tasis of lung cancers,hepatocellular carcinoma,and esophageal squamous cell carcinoma has also been repor-ted.7–10Moreover,miR-92a has been shown to promote metastasis and regulate epithelial-mesenchymal transition (EMT)via suppression of the expression of E-cadherin and PTEN.9,11We hypothesized that miR-92a promotes metastasis and regulates the EMT in CRC via suppression of E-cadherin and PTEN.Thus,we analyzed the correlations between miR-92a,E-cadherin,and PTEN gene expression and CRC clinical outcomes.The role of miR-92a in CRC tumor metastasis was also investigated in this study. PATIENTS AND METHODSPatients and SpecimensCRC tumor tissues were collected from158nonse-lected patients who underwent surgical resection for CRC at the Department of Surgery,China Medical University Hospital,between December2008and December2010. Informed written consent was obtained from all the sub-jects and/or guardians before use of their resected specimens.The acquisition of samples and their sub-sequent examination were approved by the institutional review board of China Medical University.None of the participants had a history of cancer.The clinical stages and pathologic features of primary tumors were defined according to the criteria of the American Joint Commis-sion on Cancer.Patients with distant metastasis were excluded,and those with nodal involvement received postoperative chemotherapy.Postoperative follow-up vis-its were scheduled for1and2months after surgery,then every3months during thefirst2years and every 6months thereafter,or more frequently if needed.The median duration of follow-up after a curative resection was4.8years.Real-Time PCR-Based Detection of miR-92a,E-cadherin,and PTEN GenesTotal RNA was extracted from the tumor tissues of the CRC patients using TRIzol(Invitrogen,Carlsbad,CA).For E-cadherin and PTEN gene expression analysis,real-time quantitative PCR was performed in afinal volume of25l L containing1l L of each cDNA template,10pmol of E-cadherin and PTEN gene-specific primer,and12.5l L of a SYBR Green master mix.Quantification was performed with the comparative threshold cycle method,with water as the negative control.An arbitrary threshold was chosen based on the variability of the baseline.The average threshold cycle values for the target gene were normalized to an endogenous housekeeping gene encoding18S rRNA. Expression of mature miRNA was detected by a TaqMan miRNA assay(Applied Biosystems,Foster City,CA)and normalized using the2-DD CT method relative to U6B.All TaqMan PCRs were performed in triplicate.The definition of high and low expression of E-cadherin,PTEN,and miR-92a was dependent on the mean value of genes expression of all patients’normal tissues.Expression levels higher than the mean were defined as high expression.Expression levels lower than the mean were defined as low expression.MiR-92a Precursor and Inhibitor TransfectionCells were grown to confluence in6-well plates.miR-92a mimics or miR-92a inhibitor(Life Science)and neg-ative control(Life Science)cells were transfected using Lipofectamine2000transfection reagent(Invitrogen) according to the manufacturer’s protocol.MiR-92a inhib-itor was used to knock down the miR-92a expression in the SW480cell line.The mimic miR-92a was used to over-express miR-92a in the HCT-15cell line.Transfection efficiency was evaluated by real-time PCR.Cell Invasion and Migration AssayInvasion and migration assay were performed using 24-well insert(8l m pore size;Millipore,Billerica,MA). Cells(19105)in0.2mL of serum-free culture medium were seeded in the upper well,and10%fetal bovine serum–containing culture medium was added to the lower well as a chemoattractant.For invasion assay,the upper side offilter was coated by5mg/mL Matrigel(BD Bio-sciences).After migration for16h and invasion for24h, cells on the upper side offilter were removed,and cells adherent to the underside of the membrane werefixed by 4%paraformaldehyde at37°C for15min,permeabilized by0.5%Triton X-100,and stained with hematoxylin.The number of migrating or invading cells was quantified by counting infive random microscopicfields(9100)per filter.Statistical AnalysisAll data were analyzed by SPSS software,version13.0 (IBM,Armonk,NY).A Chi square test was used to com-pare miR-92a,PTEN,and E-cadherin gene expression with clinicopathologic parameters.A probability of less than 0.05was considered statistically significant.Kaplan–Meier survival curves were constructed for overall survival(OS),T.-W.Ke et al.and the log-rank test was used to evaluate the survival curve difference between miR-92a and PTEN expression. In this study,survival was defined as the time from the date of surgical intervention to December31,2013.RESULTSmiR-92a Expression was Correlated with Lymph Node Metastasis and Tumor Stage of CRCTo understand whether the expression of miR-92a was correlated with clinical parameters of CRC,158tumor tissues were collected from CRC patients.As shown in Table1,the level of expression of miR-92a in patients with nodal metastasis was significantly higher than in those with no metastasis(86.1vs.67.4%,p=0.006).In addition, patients with advanced-stage disease had higher miR-92a expression than patients with early-stage disease(85.1vs.67.9%,p=0.011).No significant association was found between miR-92a expression and other clinical parameters (age,sex,T factor,M factor,and differentiation).PTEN but not E-cadherin Gene Expressionwas Negatively Correlated with miR-92amiR-92a has been shown to promote metastasis and regulate EMT via suppression of E-cadherin and PTEN gene expression.9,11We further analyzed the correlation of E-cadherin,PTEN,and miR-92a gene expression in the tumor tissue of the CRC patients.As shown in Table2,no significant correlation was found between E-cadherin and miR-92a expression in the tumor tissues of the CRC patients(p=0.128).However,the expression of PTEN was significantly negatively correlated with levels of miR-92a(p\0.0001).Using linear correlation analysis,we also found a negative correlation between miR-92a and PTEN but not between miR-92a and E-cadherin(Supple-mental Fig.1;r=-0.245,p=0.002for miR-92a and PTEN;r=-0.060,p=0.455for miR-92a and E-cad-herin).In addition,PTEN gene expression was negatively correlated with lymph node metastasis(Supplemental Fig.2;p\0.01).Patients with lymph node metastasis had lower PTEN gene expression compared to patients without metastasis.Expression Levels of miR-92a and High PTEN Gene Expression in Tumor Tissues of CRC Patients were Correlated with OSWe expected that the expression of miR-92a and PTEN would be associated with clinical outcomes of CRC.As shown by the results of the Kaplan–Meier analysis,the survival rate of patients with high levels of miR-92a was lower than that of patients with low levels of miR-92a (p=0.029;Fig.1a).We also analyzed the association between PTEN expression and clinical outcomes.As shown in Fig.1b,low PTEN expression was associated with poor survival rates in CRC patients(p=0.014).TABLE1Association of miR-92a expression and clinical parame-ters in tumor tissues of colorectal cancer patientsParameters MiR-92a levels pTotal Low,n(%)High,n(%)(n=158)(n=38)(n=120)Age0.452 B65year6313(20.6)50(79.4)[65year9525(26.3)70(73.7)Gender0.348 Female6418(28.1)46(71.9)Male9420(21.3)74(78.7)T factor0.097 172(28.5)5(71.5)2255(20.0)20(80.0)39829(29.6)69(70.4)4282(7.1)26(92.9)N factor0.00608628(32.6)58(67.4)1?27210(13.9)62(86.1)M factor0.079 014237(26.1)105(73.9)1161(6.3)15(93.7)Stage0.011 I?II8427(32.1)57(67.9)III?IV7411(14.9)63(85.1) Differentiation0.738 Well82(25.0)6(75.0)Moderate13233(25.0)99(75.0)Poor183(16.7)15(83.3)TABLE2Association of miR-92a,E-cadherin,and PTEN expres-sion in tumor tissues of colorectal cancer patientsParameters MiR-92a level pTotal Low(%)High(%)(n=158)(n=38)(n=120)E-cadherin0.128 Low10930(27.5)79(72.5)High498(16.3)41(83.7)PTEN\0.0001 Low837(8.4)76(91.6)High7531(41.3)44(58.7)miR-92a,and PTEN Expression in Colorectal Cancer MetastasisLevels of E-cadherin gene were not correlated with clinical outcome in patients(p=0.297,Fig.1c).The prognostic significance of the combination of miR-92a and PTEN was also indicated by Kaplan–Meier analysis.Patients with miR-92a low/PTEN high had longer OS than did patients with miR-92a low/PTEN low tumors,miR-92a high/PTEN low tumors,and miR-92a high/PTEN high tumors (p=0.047,Fig.1d).Thus,we suggest that both PTEN and miR-92a expression could be used as prognostic factors for CRC patients.Expression Levels of miR-92a and PTEN in Tumor Tissues of CRC Patients were Correlated with DFS We expected that miR-92a and PTEN expression would be associated with DFS in CRC patients.The results of the Kaplan–Meier analysis demonstrated that levels of mir-92a were correlated with DFS in patients(p=0.025,Supple-mental Fig.3a).In addition,low PTEN expression was associated with poor DFS in patients(p=0.007,Supplemental Fig.3b).Levels of E-cadherin gene were not correlated with tumor recurrence of CRC patients (p=0.071,Supplemental Fig.3c).Patients with miR-92a low/PTEN high had longer DFS than did patients with miR-92a low/PTEN low tumors,miR-92a high/PTEN low tumors,and miR-92a high/PTEN high tumors(p=0.027, Supplemental Fig.3d).Thus,we suggest that both PTEN and miR-92a expression could be used as biomarkers for tumor recurrence of CRC patients.Migration and Invasion Ability of CRC Cell Lines were Correlated with miR-92a-Mediated PTEN Expression To investigate the mechanism by which miR-92a pro-motes lymph node metastasis,we examined the roles of miR-92a and PTEN in CRC cell migration.We used two CRC cell lines:SW480and HCT-15.As shown in Fig.2a, the expression of miR-92a in the SW480cell line was high, whereas that of PTEN was low.In contrast,the HCT-15 CRC cells had low miR-92a and high PTENgene T.-W.Ke et al.expression(Fig.2b).Knockdown of miR-92a in the SW480CRC cell line resulted in a significant increase in PTEN gene expression levels compared to those of the parental cells(p\0.001;Fig.2a).The reverse correlation was found in pre-miR-92a-transfected HCT-15cancer cell lines and their parental controls(p\0.001;Fig.2b).In addition,cell migration and invasion ability decreased significantly after transfection with mimic miR-92a com-pared to the negative control in the SW480cell line (Fig.2c).The reverse correlation was found in pre-miR-92a-transfected HCT-15cancer cell lines and their parental controls(Fig.2d).These observations suggest that miR-92a plays an important role in promoting the migration and invasion ability of CRC cells.miR-92a Promotes Cell Migration via Activationof PTEN/PI3K/AKT Signaling PathwayTo understand whether miR-92a promotes the migration of CRC cells via PTEN inhibition,the downstream genes of the PTEN signaling pathway,PTEN/PI3K(phosphatidyli-nositide3-kinases)/AKT,were analyzed in mimic miR-92a and pre-miR-92a-transfected SW480and HCT-15CRC cell lines.As shown in Fig.3,after knockdown of miR-92a in the SW480cell line,PTEN gene expression levels increased significantly,and downstream expression of PI3K and phosphate-AKT(pAkt)was reduced.The reverse correlation was found in pre-miR-92a-transfected HCT-15 cancer cell lines(Fig.3).DISCUSSIONUp-regulation of miR-92a is associated with several types of human malignant solid tumors,such as CRC,lung cancers,hepatocellular carcinoma,and esophageal squa-mous cell carcinoma.7–11Zhang et al.found that the expression of miR-92a was significantly up-regulated in thePTEN PI3K AKT P-AKT -actinmiR92aN N miR92asignaling pathway inactivation wasexpression.a The downstream genes PI3Kwere decreased in mimic miR-92alines compared to parental controlPI3K and phosp-AKT protein expressionmiR-92a-transfected HCT-15cancercontrol cellsmiR-92a,and PTEN Expression in Colorectal Cancer Metastasistissues of CRC patients with lymph node metastasis.11The present study also revealed that miR-92a expression was correlated with lymph node metastasis and the tumor stage of CRC patients,with those with lymph node metastasis having higher miR-92a expression(Table1).According to previous studies,miR-92a seems to be an oncogene that promotes tumor cell migration and invasion by regulation of E-cadherin,PTEN,RECK,and C13orf125.7,9–11In the present study,we found that miR-92a and lymph node metastasis were associated with inactivation of PTEN but not with inactivation of E-cadherin(Table2and Supple-mental Fig.1).Previous reports showed promoter hypermethylation of the E-cadherin gene in40%of CRC patients.12,13Another study found that low E-cadherin gene expression was caused by low miR-200c expression.14 Thus,we suggest that PTEN but not E-cadherin is an important target gene of miR-92a in CRC metastasis.PTEN is a tumor suppressor gene.Deficient PTEN expression leads to activation of the PI3K/AKT signaling pathway,promoting antiapoptosis and cell prolifera-tion.15–18A previous study suggested that the PTEN/PI3K/ pAkt pathway may play an important role in sporadic colon carcinogenesis and that reduced PTEN expression may predict relapse in CRC patients.19Studies observed a significant correlation between PTEN expression and poor prognosis in primary prostate,breast cancer,and non-small cell lung cancers.20–22The present study also found that patients with low PTEN had poor OS and DFS compared to patients with higher PTEN expression(Fig.2 and Supplemental Fig.3).In addition,we also found that patients high miR-92a/low PTEN had poor OS and DFS compared to those with high miR-92a/high PTEN,low miR-92a/high PTEN,and low miR-92a/low PTEN(Fig.2d and Supplemental Fig.3d).These results were also con-firmed in cell model studies.In the SW480colon cancer cell line,knockdown of miR-92a expression restored the expression of PTEN.The invasion and migration ability were also reduced(Fig.3c).Thus,we propose that down-regulation of the PTEN gene by miR-92a may play an important role in tumor recurrence and patient’s clinical outcome in CRC.In conclusion,we demonstrated that miR-92a expression levels in the tumor tissues of CRC patients were positively correlated with nodal metastasis.miR-92a promoted metastasis via suppression of PTEN gene expression and activation of the PI3K/AKT pathway.Moreover,the expression levels of miR-92a and PTEN predicted clinical outcomes and tumor recurrence of CRC.Therefore,we suggest that miR-92a might be used as a biomarker to monitor tumor recurrence in CRC. ACKNOWLEDGMENT This work was supported by grants from the National Science Council(NSC-100-B-040-012)and Ministry of Health and Welfare(MOHW103-TD-B-111-01and MOHW103-TDU-B-212-113001)of Taiwan.DISCLOSURE The authors declare no conflict of interest.REFERENCES1.Jemal A,Siegel R,Xu J,Ward E.Cancer statistics.CA Cancer JClin.2010;60:277–300.2.Leslie A,Carey FA,Pratt NR,Steele RJ.The colorectal ade-noma–carcinoma sequence.Br J Surg.2002;89:845–60.3.Bartel DP.MicroRNAs:genomics,biogenesis,mechanism,andfunction.Cell.2004;116:281–97.4.Wang H,Wu J,Meng X,et al.MicroRNA-342inhibits colorectalcancer cell proliferation and invasion by directly targeting DNA methyltransferase1.Carcinogenesis.2011;32:1033–42.5.Calin GA,Sevignani C,Dumitru CD,et al.Human micro-RNAgenes are frequently located at fragile sites and genomic regions involved in cancers.Proc Natl Acad Sci U S A.2004;101:2999–3004.6.Shatseva T,Lee DY,Deng Z,Yang BB.MicroRNA miR-199a-3p regulates cell proliferation and survival by targeting caveolin-2.J Cell Sci.2011;124:2826–36.7.Hayashita Y,Osada H,Tatematsu Y,et al.A polycistronicmicroRNA cluster,Mir-17-92,is overexpressed in human lung cancers and enhances cell proliferation.Cancer Res.2005;65:9628–32.8.Shigoka M,Tsuchida A,Matsudo T,et al.Deregulation of MiR-92a expression is implicated in hepatocellular carcinoma devel-opment.Pathol Int.2010;60:351–7.9.Chen ZL,Zhao XH,Wang JW,et al.MicroRNA-92a promoteslymph node metastasis of human esophageal squamous cell car-cinoma via E-cadherin.J Biol Chem.2011;286:10725–34.10.Lin HY,Chiang CH,Hung WC.STAT3upregulates MiR-92a toinhibit RECK expression and to promote invasiveness of lung cancer cells.Br J Cancer.2013;109:731–8.11.Zhang G,Zhou H,Xiao H,Liu Z,Tian H,Zhou T.MicroRNA-92a functions as an oncogene in colorectal cancer by targeting PTEN.Dig Dis Sci.2014;59:98–107.12.Miranda E,Destro A,Malesci A,et al.Genetic and epigeneticchanges in primary metastatic and nonmetastatic colorectal can-cer.Br J Cancer.2006;95:1101–7.13.Lind GE,Thorstensen L,Løvig T,et al.A Cpg island hyper-methylation profile of primary colorectal carcinomas and colon cancer cell lines.Mol Cancer.2004;3:28.14.Hur K,Cejas P,Feliu J,et al.Hypomethylation of long inter-spersed nuclear element-1(LINE-1)leads to activation of proto-oncogenes in human colorectal cancer metastasis.Gut.2014;63:635–46.15.Cantley LC,Neel BG.New insights into tumor suppression;PTEN suppresses tumor formation by restraining phosphoinosi-tide3-kinase/Akt pathway.Proc Natl Acad Sci U S A.1999;96:4240–5.16.Khaleghpour K,Li Y,Banville D,et al.Involvement of PI-3kinase signaling pathway in progression of colon adenocarci-noma.Carcinogenesis.2004;25:241–8.17.Stiles B,Gilman V,Khanzenzon N,et al.Essential role of Akt-1/protein kinase B in PTEN-controlled tumorigenesis.Moll Cell Biol.2002;22:3842–51.18.Osaki M,Oshimura M,Ito H,et al.PI3K-Akt pathway:its functionsand alterations in human cancer.Apoptosis.2004;9:667–76. 19.Colakoglu T,Yildirim S,Kayaselcuk F,et al.Clinicopathologicalsignificance of PTEN loss and the phosphoinositide3-kinase/AktT.-W.Ke et al.pathway in sporadic colorectal neoplasms:is PTEN loss predictor of local recurrence?Am J Surg.2008;195:719–25.20.Tang JM,He QY,Guo RX,et al.Phosphorylated Akt overex-pression and loss of PTEN expression in non-small lung cancer confers poor prognosis.Lung Cancer.2006;51:181–91.21.McMenamin ME,Soung P,Perera S,et al.Loss of PTENexpression in paraffin-embedded primary prostate cancercorrelates with high Gleason score and advanced stage.Cancer Res.1999;59:4291–6.22.Saal LH,Holm K,Maurer M,et al.PIK3CA mutations correlatewith hormone receptors,node metastasis,and ERBB2,and are mutually exclusive with PTEN loss in human breast carcinoma.Cancer Res.2005;65:2554–9.miR-92a,and PTEN Expression in Colorectal Cancer Metastasis。