MicroRNA-196b is an independent prognostic biomarker in patients with pancreatic cancer

Carcinogenesis, Apr 2017

Kanno, Shinichi, Nosho, Katsuhiko, Ishigami, Keisuke, Yamamoto, Itaru, Koide, Hideyuki, Kurihara, Hiroyoshi, Mitsuhashi, Kei, Shitani, Masahiro, Motoya, Masayo, Sasaki, Shigeru, et al.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


MicroRNA-196b is an independent prognostic biomarker in patients with pancreatic cancer

Carcinogenesis MicroRNA-196b is an independent prognostic biomarker in patients with pancreatic cancer Shinichi Kanno 3 Katsuhiko Nosho 3 Keisuke Ishigami 3 Itaru Yamamoto 3 Hideyuki Koide 3 Hiroyoshi Kurihara 3 Kei Mitsuhashi 3 Masahiro Shitani 3 Masayo Motoya 3 Shigeru Sasaki 3 Tokuma Tanuma 2 3 Hiroyuki Maguchi 2 3 Tadashi Hasegawa 1 3 Yasutoshi Kimura 0 3 Ichiro Takemasa 0 3 Yasuhisa Shinomura 3 4 Hiroshi Nakase 3 0 Department of Surgery, Surgical Oncology and Science, Sapporo Medical University School of Medicine , Sapporo 060-8543 , Japan an d 1 nD,epartment of Surgical Pathology, Sapporo Medical University School of Medicine , Sapporo 060-8543 , Japan 2 Department of Gastroenterology, Teine Keijinkai Hospital , Sapporo, Japa 3 Department of Gastroenterology and Hepatology, Sapporo Medical University School of Medicine , Sapporo 006-0811 , Japan 4 Department of Gastroenterology, Ikeda Municipal Hospital , Ikeda 563-0025 , Japan Pancreatic cancer is a highly aggressive malignancy, with <50% patients surviving beyond 6 months after the diagnosis, and thus, there is an urgent need to explore new diagnostic and therapeutic approaches for this disease. Therefore, we conducted microRNA (miRNA) array analysis to detect miRNA molecules potentially associated with pancreatic cancer malignancy. To assess the identified miRNAs, we performed quantitative reverse transcription-PCR on 248 pancreatic ductal adenocarcinomas (UICC stage II). We also examined miRNA expression [microRNA-21 (miR-21) and microRNA-31 (miR-31)] and epigenetic alterations, including CpG island methylator phenotype (CIMP), potentially associated with the identified miRNAs. For functional analysis, we conducted proliferation and invasion assays using a pancreatic cancer cell line. miRNA array analysis revealed that microRNA-196b (miR-196b) was the most up-regulated miRNA in pancreatic cancer tissues compared with normal pancreatic duct cells. High miR-196b expression was associated with miR-2P1 =( 0.0025) and miR-31 (P = 0.0001) expression. It was also related to poor prognosis in the multivariate analysis using overall survival (hazard ratio: 1.66; 95% confidence interval: 1.09-2.54P; = 0.019). Functional analysis demonstrated that miR-196b inhibitor decreased cell proliferation and that miR-196b mimic promoted cancer cell invasion. In conclusion, a significant association of high miR196b expression with poor prognosis was observed in pancreatic cancer. Our data also revealed that miR-196b played an oncogenic role and that the transfection of the miR-196b inhibitor had an anti-tumour effect in the pancreatic cancer cell line. These results suggest that miR-196b is a promising diagnostic biomarker and therapeutic target in pancreatic cancer. Introduction Pancreatic cancer is a highly aggressive malignancy, with <50% need to elucidate the pathogenesis of pancreatic cancer and patients surviving beyond 6 months after the diagnosis1–(16). explore new possibilities for diagnosing and treating it. Because chemotherapeutic options only marginally prolong MicroRNAs (miRNAs) constitute a class of small non-coding life, the current mortality of patients with pancreatic cancer is RNA molecules (21–25 nucleotides) that function as post-tr-an nearly identical to its incidence. Therefore, there is an urgent scriptional gene regulators. They can function as oncogenes Abbreviations CI CIMP FFPE HR miRNA MSI OS PDAC PFS SD confidence interval CpG island methylator phenotype formalin-fixed paraffin-embedded hazard ratio microRNA microsatellite instability overall survival pancreatic ductal adenocarcinoma progression-free survival standard deviation RNA extraction and microRNA (miRNA) array analysis Total RNA was extracted from formalin-fixed paraffin-embedded (FFPE) tissues using the miRNeasy FFPE Kit (Qiagen, Valencia, CA). The TaqMan® Array Human MicroRNA A+B Cards Set v3.0 (Applied Biosystems, Foster City, CA) was used for simultaneous measurement of the expression of 760 miRNAs on a microfluidic PCR platform. In brief, 1μ g of total RNA was reverse transcribed using the Megaplex Pools Kit (Applied Biosystems), following which miRNAs were amplified and detected by PCR with sp-e cific primers and TaqMan probes. PCR was run in the 7900HT Fast RealTime PCR system (Applied Biosystems) and SDS 2.2.2 software (Applied Biosystems) was used for comparative analysis of the cycle threshold (ΔCT). U6 snRNA (RNU6B; Applied Biosystems) served as an endogenous controlΔ.CT was calculated by subtracting theT vCalues of U6 from theCT values of the gene of interest. Expression of each miRNA in the tumour samples was calculated using the equation−Δ2CT, where ΔCT = (CT miRNA − CT U6). or tumour suppressors (8,17–21). Therefore, they have been increasingly recognised as useful biomarkers for various human cancers. In pancreatic cancer, the overexpression of several miRNAs including microRNA-21 (miR-21), microRNA-31 (miR- Quantitative reverse transcription-PCR (qRT-PCR) of 31), microRNA-155 (miR-155), microRNA-198 (miR-198), micro - miR-21, miR-31 and microRNA-196b RNA-210 (miR-210) and microRNA-222 (miR-222) is higher than that in normal pancreatic tissues and are unfavourable prog n-os miR-21-5p, miR-31-5p and microRNA-196b (miR-196b)-5p expression lev tic factors2( ,3,10,12–14,22,23 ). els were analysed by quantitative RT-PCR using the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems) and TaqMan MicroRNA The term ‘the CpG island methylator phenotype’ (CIMP) Assays (Applied Biosystems), as previously described ( 1 ). In brief, 5  ng has been repeatedly used over the past decade to describe CpG of total RNA were reverse transcribed using specific stem–loop RT pr-im island promoter methylation in various human malignancies ers, following which they were amplified and detected by qRT-PCR with ( 1,24–28 ). CIMP represents a distinct form of epigenetic inst-a specific primers and TaqMan probes. PCR was run in triplicate using the bility in colorectal cance2r0(,24,27,29), causing most sporadic 7500 Fast Real-Time PCR System (Applied Biosystems). SDS v1.4 software colorectal cancers with microsatellite instability (MSI) through(Applied Biosystems) was used for comparative ΔCT analysis. We defined the epigenetic inactivation oMfLH1. We recently reported that high expression level groups of miR-21, and miR-31 as the fourth level (Q4) a CIMP-positive status andMLH1 methylation were detected in a quartile as previously described1)( and miR-196b as the second level in 12% and 3.9% of pancreatic cancers 1(). Our results also in a dichotomous category. showed that CIMP positivity is associated with an unfavo-ur DNA extraction and pyrosequencing of KRAS able prognosis of pancreatic cancer1(). CIMP may be useful analysis not only for molecular characterization but also for the as-sess ment of a response to treatment in various human cancers Genomic DNA was extracted from FFPE tissues using the QIAamp DNA (25); however, miRNAs specific to CIMP status in pancreatic FFPE Tissue Kit (Qiagen) (17). Using the extracted genomic DNA, PCR and targeted pyrosequencing were performed foKrRAS (codon 12/13/61/146) as cancers remain largely unknown. previously described ( 1 ). Against this background, we chose to examine which mole-c ular changes could represent new biomarkers of malignancy in Sodium bisulfite treatment and Real-Time PCR pancreatic cancers. In this study, we conducted miRNA array (MethyLight) to measure promoter methylation of analysis to detect miRNA molecules potentially associated with CACNA1G, CDKN2A (p16), IGF2, MLH1 and RUNX3 malignancy and to examine whether they could be used as diagnostic biomarkers and therapeutic targets using a specimen database of 248 pancreatic ductal adenocarcinomas (PDACs). Bisulfite modification of genomic DNA was performed using the BisulFlashTM DNA Modification Kit (Epigentek, Brooklyn, NY1)).( We quantified DNA methylation in four CIMP-specific promotersC[ACNA1G, CDKN2A (p16), IGF2 and RUNX3] and MLH1 using Real-Time PCR (MethyLight) 1(). CIMP positive status was defined as the presence of three/four or more methylated promotersC[ACNA1G, CDKN2A (p16), IGF2 and RUNX3] and CIMP negative was defined as the presence of zero/four to two/four methylated promoters as previously described1)(. Materials and methods Histopathological evaluations of pancreatic cancer tissue specimens In total, 248 tissue specimens of UICC stage II PDAC from patients who had Pancreatic cancer cell line and miRNA transient undergone surgical treatment at Sapporo Medical University Hospital and transfection Teine Keijinkai Hospital between 1998 and 2013 were collected and a-na lysed. It is challenging to explore new possibilities for treating pancreatic The human pancreatic cancer cell line PK-45P (RCB0241) provided by the cancer at UICC stages III or IV because the survival rate in these cases is RIKEN BRC (Supplementary Figure  1, available atCarcinogenesis Online) very low. Therefore, in this study, we focused on stage I and II cases only. was utilised. Total RNA was extracted from cell pellets using TRIzol In our database, the number of UICC stage I cases was very smalNl (= 11) Reagent (Invitrogen by Life Technologies, Carlsbad, CA). Here, 5  nmol/l compared with the number of stage II cases N(  =  248). For this reason, of miR-196b mimic, 50 nmol/l of miR-196b inhibitor and a negative conwe included stage II cases only. Histological findings related to pancreatic trol (Ambion, Austin, TX) were transfected into the pancreatic cancer cell cancer specimens were evaluated by pathologists who were blinded to line. Cells were transfected using Lipofectamine 2000 (Invitrogen by Life the clinical and molecular information. Overall survival (OS) was defined Technologies), in accordance with the manufacturer’s instructions. as the time from the surgical treatment of pancreatic cancer to death by any cause or the last follow-up. The patients were followed until death Assays for proliferation and invasion or until July 2015, whichever came first. Informed consent was obtained The proliferation of miRNA transfectants was analysed by measuring from all patients before specimen collection. This study was approved by the uptake of tritiated thymidine in the 3-(4,5-dimethylthiazol-2-yl)the institutional review boards of the three participating institutions and 2,5-diphenyltetrazolium bromide assay (MTT assay; Sigma Aldrich, St. complied with the tenets of the Helsinki Declaration. Louis, MO). In brief, transfected cells were seeded into 96-well plates to a density of 5 × 103 cells per well. After incubation for 0, 24 and 48 h, MTT assays were performed using Cell Counting Kit-8 (Dojindo, Tokyo, Japan), in accordance with the manufacturer’s instructions. Cell invasion was performed using the CytoSelect 24-well Cell Invasion Assay kit (Cell Biolabs, Inc., San Diego, CA), following the manufacturer’s directions. This kit included polycarbonate membrane with μ8m- pores that had been precoated with basement membrane matrix. The PK-45P cell suspension containing 1.0×  106 cells/ml of serum-free medium with 5 nmol/l of the miR-196b mimic or the negative control was added to the inside of each insert. Each insert was then transferred to a lower 1 well that was filled with 50μ0l of RPMI containing 10% fetal bovine serum as 2 a chemoattractant. After incubation for 48 h, the inserts were removed; the 3 non-invading cells were then removed using cotton-tipped swabs, and the 4 invading cells were fixed, stained and analysed under a microscope (Olympus, 5 Tokyo, Japan). Cells were counted in four random fields per membrane. 6 7 Statistical analysis 8 No. Relative microRNA expression Fold change (pancreatic cancer/ normal pancreatic duct) Name of microRNA (miR Base ID) microRNA-196b microRNA-221 microRNA-31 microRNA-124a microRNA-142-5p microRNA-184 microRNA-125a-3p microRNA-708 microRNA-449 microRNA-22 microRNA-155 microRNA-185 microRNA-15a microRNA-542-5p microRNA-671-3p The JMP (version 10)  software applications were used for all calculations 9 (SAS Institute, Cary, NC).To assess associations among the clinical, patho-log 10 ical and molecular characteristics, we used either tχh2teest or Fisher’s exact 11 test. To compare mean patient ages, analysis of variance was performed. 12 The Kaplan–Meier method and log-rank test were performed to assess 13 the association between miR-196b expression status of pancreatic cancer 14 and patient mortality. To adjust the results for confounders, we used -mul 15 tivariate Cox proportional hazards regression models to calculate a hazard ratio (HR) according to tumour miR-196b expression status of pancreatic The fold change is expressed as the median of that for pancreatic cancer tissue cancer. divided by that for normal pancreatic duct tissue for each microRNPA-.values The multivariate model initially included tumour location in the -pan were determined by the Mann–WhitneyU test. creas (head, body or tail)K,RAS mutation status (in codon 12/13/61/146; present versus absent), CIMP status (CIMP positive versus CIMP negative) and miR-21 and miR-31 expression levels (high versus low expression lev- levels in each samples. As a result, in all of the cases, miR-196b els). To avoid overfitting, variables in the final model were selected using expression levels in pancreatic cancer cells were significantly backward stepwise elimination with a threshold oPf = 0.10. higher than those of cancer stromaP (=  0.027). In contrast, no significant differenceP( = 0.08) was observed between cancer cell Results and the corresponding pancreatic cancer tissue (Supplementary Figure 2). Detection of high-level miR-196b expression in pancreatic cancer tissue in miRNA array analysis Distribution of miR-196b expression in pancreatic cancer and association of miR-196b with clinicopathological and molecular features Using laser capture microdissection method, we have collected and extracted RNA of the pancreatic cancer tissues from FFPE tissues of pancreatic cancer patients. In contrast, all samples of We utilised 248 stage II pancreatic cancer specimens for which normal pancreatic ductal epithelium were selectively isolated miR-196b expression data were available. miR-196b expresfrom the FFPE tissues of normal pancreatic regions using the MMI sion was quantified in pancreatic cancer specimens and was Cellcut plus microdissection system (MMI, Eching, Germany). calculated using the equation−Δ2CT, where ΔCT  =  (CT miR-196b To examine the miRNA expression signature in pancreatic − CT U6). The distributions of miR-196b expression in the 248 cancer, 14 pancreatic cancer tissues were randomly selected pancreatic cancer specimens were as follows: mean, 0.0014; from the pancreatic cancer specimens for miRNA array anal-y median, 0.00040; standard deviation (SD), 0.0033; range, 0.0– sis. Median expression levels in the pancreatic cancer group 0.035 (Figure 1). Specimens with miR-196b expression were then were compared with those in the normal pancreatic duct group dichotomised for further analysis into a low expression group (N = 5). MiRNA array data revealed differential expression in only (< median) and a high expression group (≥ median). Table  2 miR-196b ( P  <  0.05 by Mann–Whitney U test) between the two shows the clinicopathological and molecular features of pan-cre groups (Table 1). Of the 760 miRNAs, miR-196b was up-regulated atic cancers according to miR-196b expression. High miR-196b the most often (9.78-fold changeP,  = 0.0014). expression was significantly associated with miR-21P( = 0.0025) To examine the miRNA expression signature in CIMP-positive and miR-31 ( P = 0.0001) expression. pancreatic cancer, miRNA array analysis was performed. The results showed that CIMP was positive in four specimens and ne-g High miR-196b expression and patient survival ative in 10; however, the miRNA array data revealed no significant The influence of high miR-196b expression on clinical outcome difference in expression levels in individual miRNAs between was assessed in stage II pancreatic cancer patients. During the CIMP-positive and -negative specimens (data not shown). follow-up of the 248 patients eligible for survival analysis, m- or tality occurred in 100. The median follow-up period for OS was miR-196b expressions in the pancreatic cancer cells 31.1 months. In the Kaplan–Meier analysis, significantly shorter and the corresponding cancer stroma survival was observed in the miR-196b high expression group Using laser capture microdissection method, we have collected [median OS (months): 21.2 versus 40.2; log-rank P  =  0.0053; and extracted RNA of the pancreatic cancer cells and the-cor Figure 2]. responding cancer stroma from 10 FFPE tissues of pancreatic In the univariate Cox regression analysis, compared with the cancer patients, respectively and analysed miR-196b expression miR-196b low expression group, significantly higher mortality rates were observed in the miR-196b high expression group [hazard ratio (HR): 1.74; 95% confidence interval (CI): 1.18–2.60; P = 0.0057; Table 3]. In the multivariate Cox regression analysis Percentages (%) indicate the proportion of patients with a specific clinical, pathological, or molecular feature within a given dichotomous category of tumour microRNA-196b expression. The P-values were calculated using analysis of variance for age aχn2dtest or Fisher’s exact test for all other variables. CIMP, CpG island methylator phenotype; SD, standard deviation after adjusting for T and N stages, a significantly shorter OS was observed in the miR-196b high expression group than in the low expression group (HR: 1.72; 95% CI: 1.15–2.59;P = 0.0074; Supplementary Table 1, available aCtarcinogenesis Online). For the multivariate analysis, we initially included tumour location,KRAS mutation status, CIMP status and miR-21 and miR-31 expression levels in the model. Similarly, compared with the miR-196b low expression group, an independent association with poorer prognosis was observed in the miR-196b highexpression group in the multivariate analysis of OS (HR: 1.66; 95% CI: 1.09–2.54; P = 0.019; Table 3). Functional analysis of miR-196b expression in pancreatic cancer cell line The mimic and inhibitor of miR-196b were transfected into PK-45P. In the proliferation assay, significantly decreased cell proliferation was observed as a result of transfection of the miR196b inhibitor F(igure  3A). The basement membrane invasion assay revealed enhanced invasiveness of the miR-196b mimic (Figure 3B) after transfection into PK-45P. Discussion The results of the miRNA array analysis revealed that miR-196b was the most up-regulated gene in pancreatic cancer tissue co-m pared with normal pancreatic duct tissue. In a database of 248 MicroRNA-196b expression P 0.25 0.86 0.27 0.54 0.11 0.21 0.73 0.26 and glioblastoma, the upregulation of miR-196b is associated with a poor prognosis (46–48). Previous studies have suggested that several microRNAs are associated with a poor prognosis of pancreatic cancer patients; specifically, the increased expression of miR-21, miR-31, miR155, miR-198, microRNA-203 (miR-203), microRNA-205 (miR205), miR-210 and miR-222 was significantly associated with a shorter disease-free survival and OS4,(12–14,22,49). Low microRNA-34a (miR-34a), microRNA-141 (miR-141) and microRNA-375 (miR-375) expression in pancreatic cancer tissue was also si-g nificantly correlated with shorter O1S2,1(5,22). In the current study, using a large database, we revealed that high miR-196b expression was independently associated with a shorter pr-og Figure  2. Kaplan–Meier survival curves for pancreatic cancer according to nosis in a multivariate stage-stratified Cox model. Although the miR-196b expression level. Significantly lower overall survival rates were miR-196b has been reported to be up-regulated in pancreatic observed in patients with high miR-196b expression than in those with low cancer tissues and to be useful as a diagnostic factor of pan-cre expression (log-rank testP: = 0.0053). atic cancer5(,8,9), this is the first study to identify an association of high miR-196b expression with an unfavourable prognosis in pancreatic cancer patients. Tmaobrltea 3l.itTyhoef apassnoccrieaattioicncoafntcuemropautriemnictrsoRNA-196b expression with creOatuirc dcaantacesrhcoewllesd wtehraet smiginR-if1i9c6abntelxyprheigsshieorntlheavneltshionsepai-nn the Overall survival cancer stroma, suggesting that miR-196b is up-regulated in pa-n creatic cancer cells but not in the cancer stroma. The high miRMicroRNA-196b Univariate Multivariate 196b expression level in pancreatic cancer tissues may reflect expression the expression levels in cancer cells. Taken altogether, miR-196b status TotalN HR (95% CI) HR (95% CI) using pancreatic cancer tissue may be a useful biomarker as Low expression 124 1 (referent) 1 (referent) an alternative to laser capture microdissection for examining High expression 124 1.74 (1.18–2.60) 1.66 (1.09–2.54) miR-196b expression in pancreatic cancer samples. We also P 0.0057 0.019 found that a high miR-196b expression level was significantly associated with miR-21 and miR-31 expression in pancreatic The multivariate Cox model included the variables of microRNA-196b expres- cancers. No previous study has reported the relationship of miRsion status stratified by tumour locatioKnR,AS mutation status, CIMP status 196b with miR-21 and miR-31 in pancreatic cancer; therefore, and microRNA-21 and microRNA-31 expression. CI, confidence interval; CIMP, the exact mechanism is still unknown. One possibility is that CpG island methylator phenotype; HR, hazard ratio these oncogenic miRNAs in pancreatic cancer work synergis-ti cally because they influence networks that control innate and patients with stage II pancreatic cancer, high miR-196b expre-s adaptive immunities and apoptosis by regulating signalling sion was associated with miR-21 and miR-31 expression. We pathways. also identified that high miR-196b expression was an unfavou-r With regard to the functional analysis of miR-196b, a recent able prognostic factor in pancreatic cancer patients, indepe-nd study reported that the up-regulation of miR-196b promotes the ent of clinicopathological and molecular features. Functional proliferation and invasiveness of gastric cancer cells by regu-lat analysis demonstrated that an inhibitor of this microRNA and ing the PI3K/AKT/mTOR pathway (32). Moreover, miR-196b has a mimic of it, which resulted in decreased cell proliferation and been reported to promote invasive and migratory phenotypes increased cancer cell invasion, respectively. These results s-ug in oral cancer by targeting NME4, leading to the regulation of gest that miR-196b is an independent prognostic biomarker and downstream molecules, including activating p-JNK, suppres-s therapeutic target in pancreatic cancer. ing TIMP1 and augmenting MMP1/9 (50). The current study has The gene for miR-196b is located in a highly evolutionarily revealed that in pancreatic cancer cells, the miR-196b inhibitor conserved region between theHOXA9 and HOXA10 genes on decreased cell proliferation. Furthermore, the miR-196b mimic chromosome 7 (7p15.2) in humans ( 30). Recently, several inves-ti promoted cell invasion, suggesting that miR-196b functions as gations have reported that target molecules of miR-196b are HOX an oncogenic miRNA in human pancreatic cancer as well as in clusters (such as A5, A9, B6, B7 and C8), FAS, PI3K/AKT/mTOR gastric cancer and oral cancer. signaling pathway, Bcr-Abl expression and c-myc 1(8,31–40). CIMP is associated with both favourable and unfavo-ur Previous studies have reported that miR-196b is diagnostic able prognoses, as well as different clinical characteristics, and prognostic factor in the digestive tract cancers. In eso-pha depending on the tumour type. In pancreatic cancer, we geal cancer tissues, miR-196b was abnormally expressed co m- recently reported that a CIMP-positive status is related to poor pared with that in normal tissues4( 1 ). Lu et  al. (42) reported prognosis ( 1 ). With regard to the association of CIMP status that plasma miR-196b level was substantially up-regulated in with miRNA expression, our current miRNA array data using patients with oral premalignant lesions as well as in oral -can pancreatic cancer samples showed no significant difference cer patients and that miR-196b may be circulating biomarker for in expression in individual miRNAs between CIMP-positive the early detection of oral cancer. In colorectal cancer, a recentand CIMP-negative specimens. Further analysis using largestudy has reported that miR-196b is associated not only with scale samples is expected to clarify the association of miRNA BRAF mutations but also with poor prognosis after adjustment expression with epigenetic alterations, including CIMP status, for MSI status 4( 3 ). In gastric cancer, miR-196b is one of the in pancreatic cancers. microRNAs found to be significantly associated with OS44(,45). Our current study was limited by the cross-sectional design Furthermore, in lung adenocarcinoma, acute myeloid leukemia and the potential for bias (i.e., selection bias) that could have confounded the results. Nevertheless, our multivariate reg-res Foundation (to S.K.), Suhara Memorial Foundation (to K.N.), The sion analysis was adjusted for potential clinical and molecular Japan Society for the Promotion of Science (JSPS) Grant-in-Aid confounders. Although further study is needed to confirm our for Scientific Research (16K07145 to K.N.) and JSPS Grant-in-Aid results, this is an interesting first step toward the improved for Scientific Research (15K08954 to Y.S.). understanding and diagnosis of this high grade malignant d-is ease. The importance of large-scale studies for cancer research Acknowledgements cannot be emphasised enough. Small studies with null results are likely to remain unpublished compared with small studies We deeply thank the pathology departments of Sapporo Medical with significant results; this leads to publication bias. Compared University Hospital and Teine Keijinkai Hospital for providing with previous studies, the present study examined the miR-196b the tissue specimens. The authors would like to thank Enago expression status in a much larger unbiased cohort of panc-re (www.enago.jp) for the English language review. Study co-n atic cancer patients. cept and design: S.K., K.N. and H.N. Data acquisition: S.K., K.I., In conclusion, a significant association of high miR-196b I.Y., H.K., K.M.  and H.K. Data analysis and interpretation: S.K., expression with poor prognosis was observed in pancre- K.N.  and T.H. Drafting of the manuscript: S.K., K.N.  and H.N. atic cancer patients. Our data also revealed that miR-196b Critical revision of the manuscript for important intellectual played an oncogenic role and that the transfection of miR- content: S.K. and K.N. Statistical analysis: S.K. and K.N. Material 196b inhibitor had an anti-tumour effect in a pancreatic support: M.S., M.M., S.S., T.T., H.M., Y.K.  and I.T. Study supervicancer cell line. These results suggest that miR-196b is a sion: K.N., Y.S. and H.N. promising diagnostic biomarker and therapeutic target in Conflict of Interest Statement: None declared. pancreatic cancer. Supplementary material Supplementary data are available aCtarcinogenesis online. Funding This work was supported by Project Mirai Cancer Research Grants (to K.N.), Takeda Science Foundation (to K.N.), Daiwa Securities Health Foundation (to K.N.), Sapporo Jikeikai Tomoiki 4. Wang, J. et al. (2014) Circulating microRNAs in pancreatic juice as ca-n 28. Arai, E. et al. (2012) Single-CpG-resolution methylome analysis iden-ti didate biomarkers of pancreatic cancer. J. Cancer, 5, 696–705. fies clinicopathologically aggressive CpG island methylator phenotype 5. Yu, J. et  al. (2012) MicroRNA alterations of pancreatic intraepithelial clear cell renal cell carcinomas. Carcinogenesis, 33, 1487–1493. neoplasias. Clin. Cancer Res., 18, 981–992. 29. Ogino, S. et al. (2009) CpG island methylator phenotype, microsatellite 6. Hoskins, J.W. et al. (2014) Transcriptome analysis of pancreatic cancer instability, BRAF mutation and clinical outcome in colon cancer. Gut, reveals a tumor suppressor function for HNF1A. Carcinogenesis, 35, 58, 90–96. 2670–2678. 30. Popovic, R. et al. (2009) Regulation of mir-196b by MLL and its overe-x 7. Rizzato, C.et al. (2016) Association of genetic polymorphisms with sur- pression by MLL fusions contributes to immortalization. Blood, 113, vival of pancreatic ductal adenocarcinoma patients. Carcinogenesis, 3314–3322. 37, 957–964. 31. Micevic, G. et al. (2016) DNMT3b modulates melanoma growth by con8. Szafranska, A.E. et  al. (2007) MicroRNA expression alterations are trolling levels of mTORC2 component RICTOR. Cell Rep., 14, 2180–2192. linked to tumorigenesis and non-neoplastic processes in pancreatic 32. Li, N.A. et al. (2016) miR-196b regulates gastric cancer cell proliferation ductal adenocarcinoma. Oncogene, 26, 4442–4452. and invasion via PI3K/AKT/mTOR signaling pathway. Oncol. Lett., 11, 9. Schultz, N.A.et al. (2012) MicroRNA expression profiles associated with 1745–1749. pancreatic adenocarcinoma and ampullary adenocarcinoma. Mod. 33. Taverna, S. et  al. (2015) Curcumin inhibits in vitro and in vivo chronic Pathol., 25, 1609–1622. myelogenous leukemia cells growth: a possible role for exosomal d-is 10. Hernandez, Y.G.et al. (2016) MicroRNA in pancreatic ductal adenocar-ci posal of miR-21. Oncotarget, 6, 21918–21933. noma and its precursor lesions. World J. Gastrointest. Oncol., 8, 18–29. 34. Panda, A.C. et  al. (2014) miR-196b-mediated translation regulation of 11. Antwi, S.O. et  al. (2016) Pancreatic cancer: associations of inflamm-a mouse insulin2 via the 5’UTR. PLoS One, 9, e101084. tory potential of diet, cigarette smoking and long-standing diabetes. 35. Liu, Y. et al. (2013) Low expression of miR-196b enhances the expression Carcinogenesis, 37, 481–490. of BCR-ABL1 and HOXA9 oncogenes in chronic myeloid leukemogen12. Frampton, A.E. et al. (2014) A microRNA meta-signature for pancreatic esis. PLoS One, 8, e68442. ductal adenocarcinoma. Expert Rev. Mol. Diagn., 14, 267–271. 36. Yu, H. et al. (2012) Evolution of coding and non-coding genes in HOX 13. Frampton, A.E. et al. (2015) microRNAs with prognostic significance in clusters of a marsupial. BMC Genomics, 13, 251. pancreatic ductal adenocarcinoma: A meta-analysis. Eur. J. Cancer, 51, 37. Li, Z. et  al. (2012) miR-196b directly targets both HOXA9/MEIS1 onc-o 1389–1404. genes and FAS tumour suppressor in MLL-rearranged leukaemia. Nat. 14. Vychytilova-Faltejskova, Pe.t  al. (2015) MiR-21, miR-34a, miR-198 and Commun., 3, 688. miR-217 as diagnostic and prognostic biomarkers for chronic pancre-a 38. Yekta, S. et  al. (2004) MicroRNA-directed cleavage of HOXB8 mRNA. titis and pancreatic ductal adenocarcinoma. Diagn. Pathol., 10, 38. Science, 304, 594–596. 15. Zhu, Z.M. et al. (2014) Prognostic significance of microRNA-141 expre-s 39. Mo, J.S. et al. (2015) MicroRNA 196B regulates FAS-mediated apoptosis sion and its tumor suppressor function in human pancreatic ductal in colorectal cancer cells. Oncotarget, 6, 2843–2855. adenocarcinoma. Mol. Cell. Biochem., 388, 39–49. 40. Das, R.P. et al. (2015) Elucidation of the molecular interaction between 16. Yamamoto, K. et  al. (2014) Loss of histone demethylase KDM6B miRNAs and the HOXA9 gene, involved in acute myeloid leukemia, enhances aggressiveness of pancreatic cancer through downregu-la by the assistance of argonaute protein through a computational tion of C/EBPα. Carcinogenesis, 35, 2404–2414. approach. Genomics Inform., 13, 45–52. 17. Nosho, K. et al. (2014) Association of microRNA-31 with BRAF mutation, 41. Liu, S.G. et al. (2013) Differential expression of miRNAs in esophageal colorectal cancer survival and serrated pathway. Carcinogenesis, 35, cancer tissue. Oncol. Lett., 5, 1639–1642. 776–783. 42. Lu, Y.C. et al. (2015) Combined determination of circulating miR-196a 18. Hornstein, E. et  al. (2005) The microRNA miR-196 acts upstream of and miR-196b levels produces high sensitivity and specificity for early Hoxb8 and Shh in limb development. Nature, 438, 671–674. detection of oral cancer. Clin. Biochem., 48, 115–121. 19. Kurihara, H.et al. (2016) The relationship between EZH2 expression and 43. Slattery, M.Le.t al. (2016) Colorectal tumor molecular phenotype and microRNA-31 in colorectal cancer and the role in evolution of the- ser miRNA: expression profiles and prognosis. Mod. Pathol., 29, 915–927. rated pathway. Oncotarget, 7, 12704–12717. 44. Li, C.Y. et  al. (2017) Identification and functional characterization of 20. Ito, M.et al. (2015) Association ofFusobacterium nucleatum with clinical microRNAs reveal a potential role in gastric cancer progression. Clin. and molecular features in colorectal serrated pathway. Int. J. Cancer, Transl. Oncol., 19, 162–172. 137, 1258–1268. 45. Lim, J.Y. et al. (2013) Overexpression of miR-196b and HOXA10 charac21. Vychytilova-Faltejskova, Pe.t  al. (2016) Serum-based microRNA sig- terize a poor-prognosis gastric cancer subtype. World J. Gastroenterol., natures in early diagnosis and prognosis prediction of colon cancer. 19, 7078–7088. Carcinogenesis, 37, 941–950. 46. Li, X. et al. (2014) An eight-miRNA signature as a potential biomarker 22. Osayi, S.N. et al. (2014) Biomarkers as predictors of recurrence follo-w for predicting survival in lung adenocarcinoma. J. Transl. Med., 12, 159. ing curative resection for pancreatic ductal adenocarcinoma: a review. 47. Díaz-Beyá, M. et  al.; Cooperative AML group CETLAM (Grupo Biomed Res. Int., 2014, 468959. Cooperativo Para el Estudio y Tratamiento de las Leucemias Agudas 23. Papaconstantinou, I.G.et al. (2013) Expression of microRNAs in patients y Mielodisplasias). (2014) MicroRNA expression at diagnosis adds r-el with pancreatic cancer and its prognostic significance. Pancreas, 42, evant prognostic information to molecular categorization in patients 67–71. with intermediate-risk cytogenetic acute myeloid leukemia. Leukemia, 24. Toyota, M.et al. (1999) CpG island methylator phenotype in colorectal 28, 804–812. cancer. Proc. Natl. Acad. Sci. U. S. A., 96, 8681–8686. 48. Ma, R. et al. (2012) Upregulation of miR-196b confers a poor prognosis 25. Hughes, L.A. et al. (2013) The CpG island methylator phenotype: what’s in glioblastoma patients via inducing a proliferative phenotype. PLoS in a name? Cancer Res., 73, 5858–5868. One, 7, e38096. 26. Kusano, M. et al. (2006) Genetic, epigenetic, and clinicopathologic fe-a 49. Greither, T.et al. (2010) Elevated expression of microRNAs 155, 203, 210 tures of gastric carcinomas with the CpG island methylator phenotype and 222 in pancreatic tumors is associated with poorer survival. Int. and an association with Epstein-Barr virus. Cancer, 106, 1467–1479. J. Cancer, 126, 73–80. 27. Weisenberger, D.J.et al. (2006) CpG island methylator phenotype unde-r 50. Lu, Y.C. et al. (2014) OncomiR-196 promotes an invasive phenotype in lies sporadic microsatellite instability and is tightly associated with oral cancer through the NME4-JNK-TIMP1-MMP signaling pathway. BRAF mutation in colorectal cancer. Nat. Genet., 38, 787–793. Mol. Cancer, 13, 218. 1. Mitsuhashi , K. et  al. ( 2015 ) Association of Fusobacterium species in pancreatic cancer tissues with molecular features and prognosis . Oncotarget , 6 , 7209 - 7220 . 2. Karandish , F. et al. ( 2016 ) Biomarkers and targeted therapy in pancr-e atic cancer . Biomark. Cancer , 8 ( Suppl 1 ), 27 - 35 . 3. Ma , M.Z. et  al. ( 2013 ) Candidate microRNA biomarkers of pancreatic ductal adenocarcinoma: meta-analysis, experimental validation and clinical significance . J. Exp. Clin. Cancer Res. , 32 , 71 .

This is a preview of a remote PDF: https://academic.oup.com/carcin/article-pdf/38/4/425/17680087/bgx013.pdf

Kanno, Shinichi, Nosho, Katsuhiko, Ishigami, Keisuke, Yamamoto, Itaru, Koide, Hideyuki, Kurihara, Hiroyoshi, Mitsuhashi, Kei, Shitani, Masahiro, Motoya, Masayo, Sasaki, Shigeru, Tanuma, Tokuma, Maguchi, Hiroyuki, Hasegawa, Tadashi, Kimura, Yasutoshi, Takemasa, Ichiro, Shinomura, Yasuhisa, Nakase, Hiroshi. MicroRNA-196b is an independent prognostic biomarker in patients with pancreatic cancer, Carcinogenesis, 2017, 425-431, DOI: 10.1093/carcin/bgx013