miR-3140 suppresses tumor cell growth by targeting BRD4 via its coding sequence and downregulates the BRD4-NUT fusion oncoprotein

miR-3140 suppresses tumor cell growth by targeting BRD4 via its coding sequence and downregulates... www.nature.com/scientificreports Correction: Publisher Correction OPEN miR-3140 suppresses tumor cell growth by targeting BRD4 via its coding sequence and Received: 8 November 2017 downregulates the BRD4-NUT Accepted: 27 February 2018 Published online: 14 March 2018 fusion oncoprotein 1,2 1 1 1 Erina Tonouchi , Yasuyuki Gen , Tomoki Muramatsu , Hidekazu Hiramoto , Kousuke 3 1 1,4 Tanimoto , Jun Inoue & Johji Inazawa Bromodomain Containing 4 (BRD4) mediates transcriptional elongation of the oncogene MYC by binding to acetylated histones. BRD4 has been shown to play a critical role in tumorigenesis in several cancers, and the BRD4-NUT fusion gene is a driver of NUT midline carcinoma (NMC), a rare but highly lethal cancer. microRNAs (miRNAs) are endogenous small non-coding RNAs that suppress target gene expression by binding to complementary mRNA sequences. Here, we show that miR-3140, which was identified as a novel tumor suppressive miRNA by function-based screening of a library containing 1090 miRNA mimics, directly suppressed BRD4 by binding to its coding sequence (CDS). miR-3140 concurrently downregulated BRD3 by bind to its CDS as well as CDK2 and EGFR by binding to their 3’ untranslated regions. miR-3140 inhibited tumor cell growth in vitro in various cancer cell lines, including EGFR tyrosine kinase inhibitor-resistant cells. Interestingly, we found that miR-3140 downregulated the BRD4-NUT fusion protein and suppressed in vitro tumor cell growth in a NMC cell line, Ty-82 cells. Furthermore, administration of miR-3140 suppressed in vivo tumor growth in a xenograft mouse model. Our results suggest that miR-3140 is a candidate for the development of miRNA-based cancer therapeutics. e b Th romodomain and extra-terminal domain (BET) family proteins, including BRD2, BRD3, BRD4, and BRDT, contain two conserved bromodomains that are associated with acetylated lysine in histones, facilitating transcrip- 1,2 tional activation as epigenetic readers . Among the BET family proteins, BRD4 has been shown to play a critical 3,4 5 role in promoting tumor growth in several cancers, including acute myeloid leukemia , multiple myeloma , 6 7 8 9 MLL-fusion leukemia , diffuse large B cell lymphoma , triple negative breast cancer , and pancreatic cancer . BRD4 is enriched at super-enhancers of several oncogenes, such as MYC, CCND2, and BCL-XL, in cancer cells, upregulating the transcription of these genes . BRD4 increases the transcription of the other oncogenes BCL2, CDK6, and MYC by binding to the chromatin locus of these genes and recruiting positive transcriptional elonga- 11 6 tion factor complex (P-TEFb) to the promoter . Thus, BRD4 is thought to be a rational target for cancer therapy . NUT midline carcinoma (NMC) is a poorly differentiated carcinoma that arises in the midline of the upper 12,13 aerodigestive tract or the mediastinum . NMC is rare, refractory to conventional treatments, and highly lethal, 12,13 with a median survival period of 6.7–9.5 months . The pathogenesis of NMCs involves the BRD4-NUT fusion gene, which is caused by a unique chromosome translocation t(15; 19)(q13; p13.1) in the majority of cases, although BRD3-NUT fusion by a t(9; 15)(q34; q14), NSD3-NUT fusion by a t(8; 15)(p12; q15), and ZNF532-NUT 13–15 fusion by a t(15; 18)(q14; q23) occur in the remaining few cases . The translocation breakpoints occur within intron 10 of the BRD4 gene (19p13.1) and intron 2 of NUT (15q14), such that the BRD4-NUT protein contains Department of Molecular Cytogenetics, Medical Research Institute, Tokyo Medical and Dental University, Tokyo, Japan. Department of Maxillofacial Surgery, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan. 3 4 Genome Laboratory, Medical Research Institute, TMDU, Tokyo, Japan. Bioresource Research Center, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo, Japan. Correspondence and requests for materials should be addressed to J.I. (email: johinaz.cgen@mri.tmd.ac.jp) SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 1 www.nature.com/scientificreports/ both acetyl-histone binding bromodomains and the extraterminal domain of BRD4 (i.e., the full functional domain of BRD4) . The BRD4-NUT oncoprotein promotes tumor cell growth through the function of BRD4 as 16–18 well as that of NUT . e fir Th st-generation BET bromodomain inhibitor JQ1 binds to the acetyl-lysine binding pocket of BRD4, and thus, JQ1 depletes not only BRD4 but also BRD4-NUT from chromatin by preventing the binding of BRD4 to 17,19 8,20,21 chromatin . As a result, JQ1 inhibits BRD4-mediated transcription of oncogenes, such as MYC . Several studies have shown that BET bromodomain inhibitors are highly effective against various intractable cancers, 8 9 22 including triple negative breast cancer , pancreatic cancer , and NMC , although resistance to BET bromodo- 8,23 main inhibition can be acquired through various mechanisms . Several clinical trials using BET bromodomain 16,24–26 inhibitors have been started . microRNAs (miRNAs) are endogenous small non-coding RNAs of 20–25 nucleotides that decompose or 27,28 inhibit protein translation of mRNA by binding to complementary mRNA sequences . An individual miRNA has multiple target genes, and an individual gene can be targeted by multiple miRNAs . Although many studies have revealed that miRNAs repress gene expression by binding to 3′ untranslated regions (3′UTR) of mRNAs, an increasing body of evidence supports that miRNAs also bind to the coding sequences (CDS) of target mRNAs 29–31 and that miRNA binding to the CDS of mRNAs can effectively suppress translation . Our previous study also revealed that miR-432–3p directly targets KEAP1 via its CDS . Because a single miRNA can simultaneously repress numerous target genes, miRNA mimics targeting several oncogenes may be useful as therapeutic agents for cancer therapy . In the present study, to investigate the novel candidate miRNAs for the development of miRNA-based cancer therapeutics, we conducted function-based screening using a miRNA library containing 1090 miRNAs. We revealed that miR-3140, identified as a novel tumor suppressive miRNA (TS-miRNA), repressed BRD4 directly by binding to its CDS. Furthermore, we showed that miR-3140 also downregulated the BRD4-NUT oncoprotein in NMC cells. miR-3140 suppresses other tumor promoting genes, such as EGFR and CDK2, via 3′UTR. Finally, the effects of miR-3140 on tumor growth were tested in vivo. Results miR-3140 was identified as a novel TS-miRNA by function-based miRNA library screening. To extract novel TS-miRNAs, we performed function-based miRNA library screening from a library containing 1090 miRNA mimics in Panc1 cells. The strategy and brief results of this study are shown in Fig.  1a. In this study, the relative cell growth ratio was determined after transfection of each miRNA in two Panc1-derived clones, PEcadZsG-Panc1 #1 and #2 cells, which were established in our previous study . Figure 1b shows the results of this screening in Panc1 #1 (left) and #2 (right) 72 hours aer t ft ransfection with each miRNA. We set the criteria for extracting TS-miRNAs (cell growth ratio < 0.6), and then extracted 29 miRNAs from Panc1 #1 and 65 miRNAs from Panc1 #2 cells, respectively (Fig. 1c). Twelve miRNAs that inhibited cell growth in both screening assays 35–38 were identified as candidate TS-miRNAs (Table  1) . Then, 4 miRNAs were extracted according to annotation confidence, which means the certainty of the actual existence of a particular miRNA, determined using miRBase (http://www.mirbase.org/) . Among these 4 miRNAs, little is known about the tumor-suppressive function of miR-3140. Thus, we focused on a detailed analysis of miR-3140. miR-3140 inhibited in vitro tumor cell growth in various cancer cell lines. To confirm the growth suppressive effect of miR-3140 observed in the function-based screening, we evaluated the cell proliferation in vitro aer t ft ransfection with miR-3140 or miR-NC in two pancreatic cancer cell lines, Panc1 and MIAPaCa2 cells, respectively. Consistent with the screening results, overexpression of miR-3140 significantly inhibited tumor cell growth in both cell lines (Fig. 1d,e). We next investigated whether miR-3140 suppresses in vitro tumor cell growth in various cell lines including triple negative breast cancer (MDA-MB-231), esophageal cancer (KYSE150), liver cancer (Sk-Hep1), and non-small-cell lung carcinoma (HUT29 and A549). We also tested the effects of miR-3140 in KYSE150 CDDP-R cells, which are resistant to cisplatin . As shown in Fig. 1e, overexpression of miR-3140 markedly reduced tumor cell growth in all the cancer cell lines we tested. Taken together, miR-3140 inhibited in vitro tumor cell growth. miR-3140 repressed CDK2 and EGFR through 3’UTR interactions. To investigate genes downreg- ulated by miR-3140, gene expression array analysis was performed in three cancer cell lines, Panc1, MIAPaCa2, and MDA-MB-231 cells. As shown in a Venn diagram (Fig. 2a, left), the expression levels of 228 genes decreased by more than 2-fold in miR-3140-transfected cells compared with miR-NC-transfected cells. Among 228 genes, 99 genes were predicted to be direct targets of miR-3140 according to the TargetScan program (http://www.tar- getscan.org) because candidate binding-sites for seed sequences of miR-3140 exist in the 3′UTRs (Fig. 2a, right). Among the overlapping genes, we extracted CDK2, CDK6, and EGFR, which are all known to promote tumor 40–42 cell growth . Western blotting showed that the protein levels of CDK2, CDK6, and EGFR were markedly reduced in miR-3140-transfectants of Panc1 and MIAPaCa2 cells, compared with miR-NC-transfectants (Fig. 2b). To determine whether miR-3140 can directly bind to the 3′UTR of CDK2, CDK6 and EGFR, we performed lucif- erase assays using reporter plasmid vectors containing the wild type (Wt) or mutant (Mt) 3′UTR of these genes in Panc1 cells. The luciferase activity of the Wt vectors, except for the 3′UTR of CDK6 (Supplementary Fig. S1), was decreased compared with the empty vector (EV) in miR-3140-transfected cells and completely restored by the Mt vector for CDK2 and EGFR (Fig. 2c). These results indicate that CDK2 and EGFR are direct target genes of miR- 3140 through binding the 3′UTR region, although CDK6 was downregulated indirectly by miR-3140. To investigate whether CDK2 and EGFR participate in the growth suppressive effects of miR-3140, we exam- ined the cell proliferation in vitro aer t ft ransfection with small interfering RNA (siRNA) targeting CDK2 , EGFR, or negative control siRNA. As shown in Fig. 2d, si-CDK2 significantly reduced tumor cell growth in MIAPaCa2 SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 2 www.nature.com/scientificreports/ Figure 1. miR-3140 was identified as a novel tumor suppressive miRNA (TS-miR) by function-based miRNA library screening. (a) The strategy used to identify novel TS-miRs in this study. (b) Results of function-based miRNA library screening in two Panc1-derived clones (PEcadZsG-Panc1 #1 and PEcadZsG-Panc1 #2 cells), using the Pre-miR miRNA Precursor Library-Human V15 consisting of 1,090 mature human miRNA mimics. e ce Th ll growth ratio was assessed with crystal violet staining using a relative ratio normalized based on the cell survival rate of cells transfected with negative control miRNA (miR-NC). A value of 0.6 was used to determine the cut-off value for growth inhibition (dotted line). (c) The Venn diagram showing the overlap of 12 miRNAs between Panc1 #1 (3000 cells/well) and #2 (5000 cells/well). (d) Phase contrast images of Panc1 and MIAPaCa2 cells transfected with 10 nmol/L of miR-NC or miR-3140. Images were obtained 3 days aer t ft ransfection. (e) Cell growth assay in various types of cancer cells. Pancreatic cell lines (Panc1 and MIAPaCa2), a triple negative breast cancer cell line (MDA-MB-231), esophageal squamous cell carcinoma cell lines (KYSE150 and the cisplatin resistant cell line KYSE150 CDDP-R), a liver cancer cell line (Sk-Hep1), and non-small-cell lung cancer cell lines (HUT29 and A549) were transfected with 10 nmol/L of miR-NC or miR-3140. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 3 www.nature.com/scientificreports/ Panc1 Panc1 Annotation #1 #2 Confidence † † ‡ miRNA Precursor Mature Sequence Ratio Ratio Location Host gene (miRBase) Ref hsa-miR-342-5p AGGGGUGCUAUCUGUGAUUGA 0.432 0.323 14q32.2 EVL high has-miR-449b* CAGCCACAACUACCCUGCCACU 0.357 0.446 5q11.2 CDC20 — hsa-miR-608 AGGGGUGGUGUUGGGACAGCUCCGU 0.487 0.503 10q24.31 SEMA4G hsa-miR-634 AACCAGCACCCCAACUUUGGAC 0.443 0.579 17q24.2 PRKCA hsa-miR-671-3p UCCGGUUCUCAGGGCUCCACC 0.563 0.467 7q36.1 CHPF2 high hsa-miR-876-3p UGGUGGUUUACAAAGUAAUUCA 0.407 0.439 9p21.1 LINGO2 high 20q11.22 GDF5 — hsa-miR-1289 UGGAGUCCAGGAAUCUGCAUUUU 0.576 0.338 5q31.1 FSTL4 hsa-miR-1293 UGGGUGGUCUGGAGAUUUGUGC 0.580 0.425 12q13.12 LIMA1 — hsa-miR-3126-3p CAUCUGGCAUCCGUCACACAGA 0.582 0.533 2p13.3 ANTXR1 — hsa-miR-3140 AGCUUUUGGGAAUUCAGGUAGU 0.508 0.459 4q31.3 FBXW7 high — hsa-miR-3165 AGGUGGAUGCAAUGUGACCUCA 0.533 0.370 11q13.4 NUMA1 — hsa-miR-3173 0.551 0.372 14q32.13 DICER1 — Table 1. Summary of 12 miRNAs selected as candidates for novel tumor suppressive miRNAs in functional based screening with Pre-miR miRNA Precursor Library v15. Ratio means the cell growth ratio of viable cells as assessed by crystal violet staining 72 hours aer t ft ransfection with miRNAs, relative to the control transfectants. Reference is TS-miRNA. cells, but not in Panc1 cells. Similarly, si-EGFR suppressed cell proliferation in Panc1 cells (Fig. 2e), but not in MIAPaCa2 cells (Supplementary Fig. S2). These results suggested that CDK2 and EGFR contribute to tumor cell growth in a cell-dependent manner. Thus, downregulation of CDK2 and EGFR may partially contribute to miR-3140-mediated suppression of tumor cell growth. miR-3140 suppressed tumor cell growth in EGFR tyrosine kinase inhibitor (EGFR-TKI)-resistant lung cancer cells. Since miR-3140 directly targeted EGFR, we next examined whether miR-3140 over- came the resistance to EGFR-TKIs. EGFR is one of the most frequently mutated “driver” genes in non-small cell lung carcinoma (NSCLC) and EGFR-TKIs are used for the treatment of EGFR-mutated-NSCLC. Therefore, we examined the ee ff cts of miR-3140 in EGFR-mutated-NSCLC cells. In agreement with past reports, HUT29 cells, which harbor the EGFR L858R mutation, were sensitive to the EGFR-TKIs gefitinib and erlotinib, whereas NCI-H1975 cells, which harbor the EGFR L858R/T790M double mutation, were resistant to gefitinib and erlo- tinib (Supplementary Fig. S3a,b) . Both miR-3140 and si-EGFR reduced the expression of EGFR in NCI-H1975 cells, resulting in the suppression of in vitro tumor cell growth in NCI-H1975 cells (Fig. 2e,f ). es Th e results sug- gested that miR-3140 may overcome the acquired resistance to EGFR-TKIs at least in part by suppressing mutant EGFR in NSCLC cells. Suppression of BET family genes inhibited tumor cell growth. Because the effects of miR-3140 on in vitro tumor cell growth were much greater than those of si-CDK2 and si-EGFR (Figs 1e and 2d–f ), concurrent downregulation of additional target genes may be necessary for the induction of miR-3140-mediated inhibition of tumor cell growth. A growing body of evidence suggested that coding sequence (CDS) of mRNA could be 29–31 directly targeted by a miRNA-containing RISC complex . u Th s, we explored additional target genes of miR- 3140 using the STarMirDB database (http://sfold.wadsworth.org) , which predicts miRNA target genes based on CDS binding. Among the 228 genes downregulated by miR-3140 in 3 cell lines (Fig. 2a), candidate binding-sites for the seed sequence of miR-3140 exist in the CDS of 103 genes (Fig. 3a, Supplementary Table S1). From these 103 overlapping genes, we extracted BRD4 as a candidate target gene of miR-3140, because BRD4 has been shown to play a critical role in promoting tumor growth in several cancers through upregulates the transcription of oncogenes, including MYC and CCND2. Our gene expression array analysis showed that miR-3140 downregu- lated the expression of MYC and CCND2 as well as BRD4 in Panc1 cells, suggesting that BRD4 plays a critical role for tumor cell growth in Panc1 cells. As shown in Fig. 3b, miR-3140 markedly reduced BRD4 protein expression. Furthermore, we found that other BET family genes, BRD2 and BRD3, together with genes for the downstream 10,20 signaling pathways of BRD4, MYC, phosphorylated STAT3 and Cyclin D2 , were significantly reduced by miR- 3140 (Fig. 3b). Transfection of siRNA targeting each BET family gene, especially si-BRD4, markedly inhibited tumor cell growth (Fig. 3c, Supplementary Fig. S4). Conversely, overexpression of BRD4 promoted cell prolifera- tion compared to transfection with an empty vector (Fig. 3d). Taken together, downregulation of BRD4 contrib- uted to miR-3140-mediated suppression of tumor cell growth. miR-3140 repressed BRD4 and BRD3 by directly binding to the CDS region. To examine whether miR-3140 can directly bind to the CDS of BRD4, we first performed luciferase assays using reporter plasmid vec- tors containing Wt or Mt seed sequences in the CDS. Because two candidate binding-sites for the seed sequence of miR-3140 exist within the CDS of BRD4, three mutant constructs (Mt1, Mt2 and Mt1 + 2) were established (Fig. 4a). e Th luciferase activity of the Wt vector was decreased compared with EV in miR-3140 transfected cells and partially recovered with each Mt1 or Mt2 vector and completely restored with the Mt1 + 2 vector (Fig. 4a). We next examined whether miR-3140 suppressed the expression of BRD4 by binding to these positions within the SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 4 www.nature.com/scientificreports/ Figure 2. miR-3140 targeted CDK2 and EGFR by binding their 3′UTR regions. (a) Left, identification of downregulated genes aer ft miR-3140 transfection by a gene expression array. The Venn diagram shows that 228 genes were commonly downregulated (fold change >2) by transfection of miR-3140 in Panc1, MIAPaCa2, and MDA-MB-231 cells. Right, prediction of candidate target genes regulated by miR-3140 via their 3′UTR. The Venn diagram shows that 99 genes were predicted as candidate 3′UTR-targets of miR-3140 by the TargetScan program. (b) Western blot analysis of CDK2, CDK6, and EGFR in Panc1 and MIAPaCa2 cells 72 hours aer ft transfection with 10 nmol/L of miR-NC or miR-3140. (c) Luciferase reporter assays. Panc1 cells were transfected with the pmirGLO Dual Luciferase vectors containing wild type (Wt) CDK2 and EGFR or mutant (Mt) 3’UTR target sites of these genes, and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding site of miR-3140 within the 3′UTR of each gene and mutant sequences. Bottom, results of the luciferase assay; *P < 0.05. (d,e) Evaluation of the effect of si-CDK2 or si-EGFR. Western blot analysis (top) and cell growth assay (bottom) in indicated cell lines aer t ft ransfection with 20 nmol/L of negative control siRNA (si- NC) or siRNA targeting each gene. (f) Evaluation of the effect of miR-3140 in EGFR-TKI-resistant lung cancer cells. Western blot analysis (top) and cell growth assay (bottom) in the indicated cell line aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. Cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. CDS. Synonymous mutations were generated at each site of the Halotag-BRD4 protein (Supplementary Fig. S5a). Each of those Wt and Mt vectors were transfected into MIAPaCa2 cells followed by transfection of miR-NC or miR-3140, respectively. Whereas the expression level of exogenously expressed BRD4-Wt was reduced in the miR- 3140-transfected cells, the miR-3140-induced reduction of exogenous BRD4 was restored by all three patterns of each mutation at binding-sites for the seed sequence of miR-3140 (Fig. 4b). These combined data suggested that miR-3140 can downregulate BRD4 expression by directly targeting its CDS. Similarly, we examined whether BRD3 was also a candidate target gene of miR-3140. Candidate binding-sites for the seed sequence of miR-3140 exist within both the CDS and the 3′UTR of the BRD3 gene. The luciferase activity of the 3′UTR-Wt vector did not decrease, but that of the CDS-Wt vector was decreased compared with EV in miR-3140 transfectant, and this decrease was completely recovered by the CDS-Mt vector (Fig. 4c,d). Synonymous mutations were generated at the CDS site of the Halotag-BRD3 protein (Supplementary Fig. S5b). While the expression level of exogenously expressed BRD3-Wt was decreased in the miR-3140-transfected cells, this miR-3140-induced decrease in exogenous BRD3 was restored by the mutations at a binding-site for the seed sequence of miR-3140 (Fig. 4e), suggesting that miR-3140 can downregulate BRD3 expression by targeting its CDS as well as BRD4. Taken together, miR-3140 targets BRD4 and BRD3 through their binding CDS regions. miR-3140 downregulated the BRD4-NUT fusion protein. Based on the results that miR-3140 directly targets the CDS of BRD4, we investigated whether miR-3140 suppresses the BRD4-NUT fusion gene. The BRD4-NUT fusion is caused by t(15:19) translocation and drives NUT midline carcinoma (NMC), a rare, SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 5 www.nature.com/scientificreports/ Figure 3. miR-3140 downregulated BET family genes. (a) Prediction of candidate targets regulated by miR- 3140 via their coding sequences (CDS). The Venn diagram shows that 103 genes were predicted as candidate CDS-targets of miR-3140 by the STarMirDB program. (b) Western blot analysis of BET family proteins (BRD2, BRD3, and BRD4) and downstream targets of BRD4 (Cyclin D2, MYC, and p-STAT3) in Panc1 and MIAPaCa2 cells 72 hours aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. (c,d) Effects of knockdown of BET family proteins (c), or overexpression of BRD4 (d) on cell growth in Panc1 cells. Cells were transfected with 20 nmol/L of si-NC or siRNA targeting BET family proteins (c), and empty vector or BRD4-HaloTag expression vector (d). Left, western blot analysis of indicated proteins in Panc1 cells 72 hours aer t ft ransfection. Right, results of the cell growth assay. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. 12 21,45 highly lethal cancer (Fig. 5a) . Consistent with past reports , JQ1 reduced MYC and significantly suppressed in vitro tumor growth of Ty-82 cells, a NMC cell line, which harbor t(15;19) bearing the BRD4-NUT fusion gene (Fig. 5b) . As shown in Fig. 5c, miR-3140 repressed the expression of the BRD4-NUT fusion protein and its downstream target MYC in Ty-82 cells. As a result, miR-3140 effectively suppressed in vitro tumor cell growth of Ty-82 cells (Fig. 5c). Next, to test whether miR-3140 can suppress tumor cell growth of JQ1-resistant cells, JQ1-resistant cells (Ty- 82 JQ1-R) were generated from Ty-82 cells (IC50: 0.43 µ M in Ty-82 cells, 1.37 µ M in Ty-82 JQ1-R cells; Fig. 5d). To determine whether Ty-82 JQ1-R cells have developed resistance by acquiring BRD4-NUT-dependence or not, we examined the effects of JQ1 in Ty-82 JQ1-R cells. The expression of MYC, a downstream effector of BRD4-NUT, was not suppressed sufficiently by JQ1 treatment in Ty-82 JQ1-R cells, whereas the expression of MYC was suppressed by JQ1 treatment in a dose-dependent manner in Ty-82 cells (Supplementary Fig. S6a). Knockdown of BRD4-NUT using RNA interference downregulated the expression of MYC and suppressed cell growth both in Ty-82 JQ1-R cells (Supplementary Fig. S6b). These results suggested that JQ1 could not block the effects of BRD4-NUT sufficiently in Ty-82 JQ1-R cells, although the mechanism is unknown. miR-3140 downreg- ulated the expression of BRD4-NUT fusion protein and MYC, and suppressed tumor cell growth in Ty-82 JQ1-R cells as well as Ty-82 cells (Fig. 5c,e). These combined data suggested that the resistance of JQ1 is, at least in part, dependence of BRD4-NUT/MYC pathway. Thus, our results suggested that miR-3140 inhibited cell growth in Ty-82 JQ1-R cells by targeting BRD4-NUT. miR-3140 suppressed tumor growth in a xenograft mouse model. We next examined the therapeu- tic effect of miR-3140 through the local administration of dsRNA mimicking miRNA around MIAPaCa2-derived subcutaneous tumors in nude mice. miR-NC (left) or miR-3140 (right) were administered into the subcutaneous space around tumors 5 times (7, 11, 15, 18, 21 days aer t ft he injection of MIAPaCa2 cells; Fig.  6a). Administration of miRNA did not produce any adverse consequences, such as body weight loss or local damage. As a result, tum- ors treated with miR-3140 at 23 days aer t ft he injection of MIAPaCa2 cells were significantly smaller than tumors treated with miR-NC (Fig. 6b,c, Supplementary Fig. S7a,b). We confirmed that the expression of miR-3140 was significantly high in the miR-3140-treated tumors compared to miR-NC -treated tumors by qRT-PCR (Fig. 6d). Furthermore, immunohistochemical staining showed that the expression of BRD4, BRD3, CDK2, and EGFR, the SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 6 www.nature.com/scientificreports/ Figure 4. miR-3140 directly targeted BRD4 and BRD3 by binding their coding regions. (a) Luciferase reporter assay. MIAPaCa2 cells were transfected with the pmirGLO Dual Luciferase vectors containing Wt or Mt BRD4, or empty vector (EV), and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding sequence of miR-3140 within the CDS of BRD4 and mutant sequences are indicated. Bottom, the results of luciferase assay; *P < 0.05. (b) Western blot analysis of BRD4 in MIAPaCa2 cells. Cells were co-transfected with the Wt or Mt BRD4 expression vector, and aer 24 ft hours, either 10 nmol/L of miR-NC or miR-3140 was additionally transfected. (c,d) Luciferase reporter assays. Panc1 cells were transfected with a reporter plasmid (Wt of BRD3 3′UTR or EV (c), or Wt, Mt of BRD3 CDS or EV (d)) and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding sequences of miR-3140 within the BRD3 3′UTR (c) and the BRD3 CDS (d). Bottom, results of the luciferase assay; *P < 0.05. (e) Western blot analysis of BRD3 in Panc1 cells. Cells were co-transfected with the Wt or Mt BRD3 expression vector, and aer 24 ft hours, either 10 nmol/L of miR-NC or miR-3140 was additionally transfected. SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 7 www.nature.com/scientificreports/ Figure 5. miR-3140 downregulates the BRD4-NUT fusion protein. (a) Schema of the relationship between mRNA of the BRD4-NUT fusion gene in NUT midline carcinoma and the location of miR-3140 target sites. (b) Effects of the BET bromodomain inhibitor JQ1 in Ty-82 cells. Left, the results of the cell growth assay. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. Right, western blot analysis of MYC in Ty-82 cells 3 hours aer t ft ransfection. (c,e) Left, cell growth assay in a NUT midline carcinoma cell line, Ty-82 (c) and its JQ1-resistant cells (Ty-82 JQ1-R) cells (e) which were transfected with 10 nmol/L of miR-NC or miR-3140. Cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. Right, western blot analysis of the BRD4-NUT fusion protein and MYC in Ty-82 (c) and Ty-82 JQ1-R cells (e) 72 hours aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. (d) Dose response curve of JQ1 for Ty-82 (upper) and Ty-82 JQ1-R cells (lower) at 50 hours following treatment with JQ1. Cell indexes were normalized with the last time point before JQ1 treatment. targets of miR-3140, were reduced in the resected tumors treated with miR-3140 (Fig. 6e, Supplementary Fig. S8). Although other targets of miR-3140 may also participate in the tumor suppressive activity, miR-3140 inhibited tumor growth in vivo at least in part by suppressing BRD4, BRD3, CDK2 and EGFR. Discussion Here, we identified miR-3140 as a novel TS-miRNA by function-based miRNA library screening. As summarized in Fig. 6f, we showed that miR-3140 inhibited tumor cell growth in various cancer cells both in vitro and in vivo at least in part by directly targeting BRD4, BRD3, CDK2, and EGFR. Furthermore, we revealed that miR-3140 suppressed the BRD4-NUT oncoprotein in NMC cells and that miR-3140 inhibited in vitro tumor cell growth in NMC cells. We revealed that miR-3140 directly suppressed EGFR and CDK2 via 3′UTR interaction. Both EGFR and CDK2 play a role in cancer progression. EGFR is activated by gain-of-function mutations or amplification in several cancers including lung, head and neck, ovary, colon, and esophagus . Small molecule inhibitors, such as gefitinib and erlotinib, are used in lung cancer patients who have mutations on EGFR . We showed that miR- 3140 suppressed in vitro tumor cell growth by directly reducing EGFR expression in NCI-H1975 cells, which 43,47,48 are EGFR-TKI resistant due to EGFR L858R/T790M double mutations . Although in vivo experiments are needed to confirm the effects of miR-3140 on EGFR-TKI-resistant cells, our results suggest that miR-3140 may overcome the acquired resistance to EGFR-TKIs in lung cancer. CDK2 is another target of miR-3140. The cat- alytic activity of CDK2-CyclinE complexes is hyperactivated in several cancers by Cyclin E amplification, or a loss-of-function mutation of FBXW7, a ubiquitin ligase for Cyclin E degradation, although CDK2 mutations are rare in human cancers . Thus, suppression of CDK2 by miR-3140 may contribute to the inhibition of tumor growth in some cancers. Second, we identified that miR-3140 suppressed BRD4 and BRD3 through binding to their CDS. Because BET bromodomain inhibitors suppressed BET family-mediated transcription of the oncogene MYC, they were 5,22 reported to be promising agents for many cancers . Several BET inhibitors have entered into clinical trials in SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 8 www.nature.com/scientificreports/ Figure 6. era Th peutic effects of miR-3140 for tumor growth in vivo . (a) The experimental schedule for miR- 3140 treatment in nude mice which were subcutaneously inoculated with MIAPaCa2 cells. (b) The representative image of tumor-bearing nude mice at 23 days aer t ft he inoculation of MIAPaCa2 cells. Tumors are denoted by arrowheads. (c) Tumor growth curves of xenograft mouse models treated with miR-NC or miR- 3140 (n = 5, each). Tumor volume was calculated using the following formula: (shortest diameter) × (longest diameter) × 0.5. Bar, SD for 5 mice; *P < 0.05. (d) Expression analysis of miR-3140 in resected tumors. The expression level of miR-3140 was measured by qRT-PCR using the relative ratio normalized based on the expression of RNU6B. Each experiment was performed in duplicate. Bar, SD. (e) Representative images of immunohistochemical staining for BRD4, BRD3, CDK2 and EGFR in resected tumors. Scale bar, 50 μm (f) Schematic models for the mechanism by which miR-3140 suppresses tumor growth. 16,24,25 some cancers, including NMC . However, development of resistance to BET inhibition has been reported, in the same way as occurs in the other molecular targeted drugs. For example, BET inhibitor resistance is associated with increased BRD4 phosphorylation mediated by casein kinase 2 (CK2) in triple negative breast cancer cells . SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 9 www.nature.com/scientificreports/ We showed that miR-3140 suppressed the expression of BRD4-NUT in NMC cells. In the BRD4-NUT fusion gene, the translocation breakpoints of BRD4 is within intron 10 , and the binding sequence of miR-3140 exists within exon 9 of BRD4. Thus, miR-3140 can directly target the BRD4-NUT fusion gene and consequently, miR- 3140 suppressed in vitro tumor cell growth of Ty-82 cells as well as JQ1. Unfortunately, Ty-82 cells hardly formed subcutaneous tumors in our xenograft model. Further in vivo studies of the effect of miR-3140 for NMC cells are needed for future work. Our data suggested that miR-3140 could suppress in vitro tumor cell growth of JQ1-resistant Ty-82 cells. Direct suppression of BRD4-NUT and concurrent downregulation of other tumor promoting genes such as EGFR and CDK2 by miR-3140 may potentially overcome resistance to BET inhibitors in NMC cells, although further studies are needed to clarify the acquired resistance of BET inhibitiors. On the other hand, miR-3140 might not suppress BRD3-NUT, since the translocation breakpoint of BRD3 in BRD3-NUT fusion gene is within intron 9 , and the binding sequence of miR-3140 exists within exon 10 of BRD3. It was reported that the expression of BRD4 were increased by gemcitabine treatment in Panc1 and MIAPaCa2 cells, and the combination of BRD4 silencing and gemcitabine treatment had a synergistic effect on the chemo- therapeutic efficacy . A combined treatment of gemcitabine with miR-3140 also may increase gemcitabine sen- sitivity in pancreatic cancer. In primary samples, we did not observe a correlation between the expression levels of miR-3140 and over- all survival in pancreatic cancer, breast cancer, or acute myeloid leukemia according to the TCGA database (Supplementary Fig. S9a–c). This is because the miR-3140 expression levels are very low in the majority of pri- mary tumor samples, although miR-3140 is predicted to exist in human tissue according to the miRBase database. Further investigation of the expression pattern of miR-3140 in non-tumor tissues is required. Because of recent advances in clinical tumor sequencing and developments in small molecule inhibitors, molecular targeted therapy is one of the key therapeutic strategies along with conventional chemotherapy, radi- ation therapy, and immunotherapy. However, acquired resistance to molecular targeted drugs is a major prob- lem for cancer treatment . Reduction of the molecular target itself and other tumor promoting targets by a miRNA-based therapy may contribute to overcome the drug tolerance that develops against molecularly targeted drugs. In this study, our findings suggest that miR-3140 suppresses tumor cell growth not only in various can- cer cells, but also in EGFR-TKIs-resistant cells and JQ1-resistant cells. Although further in vivo studies of drug delivery systems and possible off-target effects are needed, miR-3140 may be a candidate for the development of miRNA-based cancer therapeutics. Materials and Methods Cell culture. Panc1 and MIAPaCa2 cells were cultured in Dulbecco’s Modified Eagle Medium containing 10% Fetal Bovine Serum (FBS). MDA-MB-231 cells from the American Type Culture Collection (Manassas, VA) were maintained in L-15 medium containing 10% FBS. KYSE150 cells, a gift from Dr. Shimada Y (Toyama University) , NCI-H1650, NCI-H1975, A549, HUT29 from ATCC, the NUT midline carcinoma cell line Ty-82 from JCRB, and 11–18 were maintained in RPMI1640 medium containing 10% FBS. The KYSE150 CDDP- resistant cell line (KYSE150 CDDP-R), which is resistant to cisplatin, was generated previously . Sk-Hep1 cells were maintained in Eagle’s Minimum Essential Medium containing 10% FBS. All cell lines were maintained at 37 °C with 5% CO2. The all experiments were carried out in accordance with the approved guidelines and regu- lations (2010-5-4, 2016-011C2). Function-based miRNA screening. Two Panc1-derived clones (PEcadZsG-Panc1 #1 and #2 cells) were seeded on 96-well plates. Aer 24 ft hours, each clone was transfected in duplicate with each of the 1090 dsRNAs from the Pre-miR miRNA Precursor Library-Human V15 (Thermo Fisher Scientific, CA) or a negative control miRNA (miR-NC) using an RNA concentration of 10 nmol/L. Aer 72 ft hours, viable cell number was assessed by the crystal violet staining assay. The results were normalized to the cell numbers of cells transfected with miR-NC. Transfection of miRNAs and siRNAs. The dsRNA mimicking mature human miR-3140-3p (MC17496) and nonspecific control miRNA (negative control #1) were purchased from Thermo Fisher Scientific. The SMARTpool siRNA for BRD2 (M-004935-02), BRD3 (M-004936-01), BRD4 (M-004937-02), CDK2 (M-003236- 04), EGFR (L-003114-00), nonspecific control siRNAs, and a set of 4 siRNA for BRD4 (MQ-004937-02) were from GE Healthcare (Buckinghamshire, UK). Each SMARTpool siRNA consists of 4 siRNA duplexes designed to target different regions of the same gene. mRNA sequence. miRNAs and siRNAs were transfected individually into cells at the indicated concentrations using Lipofectamine RNAiMAX (Thermo Fisher Scientific) according to the manufacturer’s instructions. In vitro cell growth assay. The in vitro cell growth assay was carried out as described previously . Briey fl , cell viability was measured using the WST-8 assay (Cell Counting Kit-8; Dojindo, Kumamoto, Japan) at the indi- cated number of days aer p ft lating, and the results were normalized to day 1 values. Each assay was carried out in triplicate. Treatment with EGFR-TKI. Gefitinib and Erlotinib were purchased from Cayman Chemicals (Michigan, USA) and were resuspended in dimethyl sulfoxide (DMSO) to a final concentration of 100 mM and 500 mM for long-term storage, respectively. Cells were treated with the medium containing Gefitinib (0.001, 0.01, 0.1, 1, 5, 10, 25 µ M), Erlotinib (0.001, 0.01, 0.1, 1, 5, 10, 25 µ M), or DMSO for 72 hours. Cell viability was assessed by WST-8 assay as described above. Gene expression array analysis. Gene expression array analysis was carried out as previously described . Briefly, Panc1, MIAPaCa2, and MDA-MB-231 cells were transfected with 10 nmol/L of miRNA (miR-NC, or SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 10 www.nature.com/scientificreports/ miR-3140). RNA was extracted 72 hours after transfection. The data were analyzed by GeneSpring software (Agilent Technologies, Japan). Western blotting. Western blotting was performed according to previously reported methods . Primary antibodies for Western blotting were used as follows: antibodies for BRD4 (#13440), BRD2 (#5848), MYC (#9402), NUT (#3625), Cyclin D2 (#3741 S), p-STAT3 (#9145 S), STAT3 (#9139 S), and CDK6 (#3136) were pur- chased from Cell Signaling Technology; antibodies for CDK2 (sc-163) and EGFR (sc-03-G) from Santa Cruz Biotechnology; antibodies for BRD3 (A302-368A) from Bethyl Laboratories (Montgomery, TX); β-actin (A5441) and Vinculin (V9131) from Sigma-Aldrich (Tokyo, Japan); and HaloTag (G921A) from Promega (Tokyo, Japan). Luciferase activity assay. Luciferase reporter plasmids were constructed by inserting the 3′UTR of CDK2, EGFR and CDS of BRD4 and BRD3 downstream of the luciferase gene within the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, Madison, WI, USA). All site-specific mutations used the KOD -Plus- Mutagenesis Kit (TOYOBO, Osaka, Japan). Luciferase reporter plasmids or control plasmid (pmirGLO) were transfected into Panc1 cells using Lipofectamine 2000 (Thermo Fisher Scientific), and 10 nmol/L of miRNA (miR-NC or miR-3140) was also transfected 6 hours later. After 2 days, Firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega), and relative luciferase activity was calculated by normalizing the Firey fl luciferase reading with its corresponding internal Renilla luciferase control. Plasmid construction and transfection. BRD4 and BRD3 cDNA were purchased from Kazusa DNA Research Institute (Chiba, Japan) and were subcloned into the pFN28A HaloTag CMV-neo Flexi Vector (Promega, Madison, WI) to generate a mammalian expression vector. All site-specific mutations were gener - ated using the KOD-Plus-Mutagenesis Kit (TOYOBO). The BRD3 or BRD4 expression vector or the empty vec- tor were transfected into Panc1 and MIAPaCa2 cells using the Lipofectamine 2000 (Thermo Fisher Scientific), according to the manufacturer’s instructions. Establishment of JQ1-resistant cells. JQ1-resistant cells derived from Ty-82 were established by long-term incubation with gradually increasing JQ1 concentrations. JQ1 was purchased from APExBIO (Houston, TX). JQ1 was resuspended in DMSO to a final concentration of 10 mM for long-term storage. The cells were initially exposed to JQ1 at 0.1 µ mol/L in RPMI medium, then cultured in JQ1-free medium to confluence, and then exposed to JQ1 at a higher concentration. This cycle was repeated several times, until cells that were able to survive in RPMI medium including 2.5 μmol/L JQ1 were defined as Ty-82 JQ1-R cells. Real-time xCELLigence impedance analysis of cytotoxicity. Ty-82 cells and Ty-82 JQ1-R cells (3 × 10 ) were seeded in wells of the E-Plate 16 (ACEA Biosciences, San Diego, CA, USA). Approximately 48 hours later, these cells were treated with JQ1 (0.01, 0.05, 0.1, 0.5, 1, 5, 10 µ mol/L) (APExBIO) and DMSO. Cell-electrode impedance was monitored using the xCELLigence RTCA DP system (ACEA Biosciences) to pro- duce time-dependent cell response dynamic curves. Data were collected every 5 min aer t ft reatment with JQ1 for the first four hours, every 15 min for the next 20 hours, and then every 1 hour for an additional 48 hours. Dose response curves of JQ1 for Ty-82 cells and Ty-82 JQ1-R cells at 50 hours following treatment with JQ1 were con- structed. Cell indexes were normalized with the last time point before treatment with JQ1. In vivo tumor growth and miRNA administration. In vivo miRNA administration of miRNA was per- formed as previously described . Six-week-old female BALB/c nude mice were purchased from Oriental Bio Service, Japan. Briefly, a total of 5.0 × 10 cells in 100 μL of PBS were subcutaneously injected into the dorsal side of the mice. Aer t ft umor formation at day 7, a mixture of 1 nmol dsRNA (miR-NC or miR-3140) and 100 μL AteloGene (KOKEN, Tokyo, Japan) was administered around the tumor (miR-NC to the left dorsal side and miR-3140 to the right dorsal side of mice). miRNAs were administered on days 7, 11, 15, 18, and 21, and at 23 days ae ft r cell injection, mice were sacrificed and tumors were resected. Tumor volume was calculated using the following formula: (shortest diameter) × (longest diameter) × 0.5. All experimental protocols conducted on the mice were approved by the Tokyo Medical and Dental University Animal Care and Use Committee. Immunohistochemistry. Immunohistochemistry was performed as previously described . The resected tumors from xenograft mouse model were fixed in 10% formaldehyde in PBS for 24 h and stored in 70% ethanol and then embedded in paran. ffi e Th following primary antibodies were used for immunohistochemistry: an anti- body for BRD4 (HPA061646, 1:500) was purchased from Atlas Antibodies (Stockholm, Sweden), BRD3 (A302- 368A, 1:500) and CDK2 (IHC-00374, 1:500) antibodies were from Bethyl Laboratories, and EGFR (sc-03-G, 1:200) antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). BRD4, BRD3, EGFR, and CDK2 staining were scored semiquantitatively using histo-score (H-score) based on staining intensity and percentage of positive cells. Staining intensity was scored as follows: 0 = none, 1 = weak, 2 = moderate, or 3 = strong. H-score was cal- culated by multiplying the intensity of staining with percentage of cells stained in randomly chosen 3 fields from each specimen . Quantitative RT-PCR (qRT-PCR). Total RNA was extracted using TRIsure reagent (BIOLINE, London, UK) according to the manufacturer’s instructions. For miRNA, total RNA was reverse transcribed using the Taqman Reverse Transcription Kit followed by qRT-PCR performed using Custom Taqman miRNA Assays (Applied Biosystems, Foster City, CA). The miRNA expression was normalized to the internal control RNU6B. e f Th ollowing primers were used for the Taqman assay (Thermo Fisher Scientific): human miR-3140-3p (244524), RNU6B (001093). SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 11 www.nature.com/scientificreports/ Public datasets. To explore the generality of the miRNA expression and clinical features among pancreatic cancer, breast cancer and acute myeloid leukemia (AML), we examined the public datasets from TCGA (http:// cancergenome.nih.gov) retrieved on 24th July 2017. We used TCGA datasets on primary cases for pancreatic cancer, breast cancer and AML, and examined correlation of prognosis and expression of miR-3140. The datasets we used included mRNA-expression data on 174 samples of pancreatic cancer, 972 samples of breast cancer, 89 samples of AML and 1881 miRNAs, respectively. Statistical analysis. Student’s t-test, Paired t-test and log-rank test were performed using R software. P val- ues of < 0.05 were considered significant. Ethical Approval. All experimental protocols conducted on the mice were approved by the Tokyo Medical and Dental University Animal Care and Use Committee (0170036C). References 1. Belkina, A. C. & Denis, G. V. BET domain co-regulators in obesity, inflammation and cancer. Nat. Rev. Cancer 12, 465–477 (2012). 2. Fujisawa, T. & Filippakopoulos, P. Functions of bromodomain-containing proteins and their roles in homeostasis and cancer. Nat. Rev. Mol. Cell Biol. 18, 246–262 (2017). 3. Zuber, J. et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 478, 524–528 (2011). 4. Blobel, G. A., Kalota, A., Sanchez, P. V. & Carroll, M. Short hairpin RNA screen reveals bromodomain proteins as novel targets in acute myeloid leukemia. Cancer Cell 20, 287–288 (2011). 5. Delmore, J. E. et al. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 146, 904–917 (2011). 6. Dawson, M. et al. Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529–533 (2011). 7. Chapuy, B. et al. Discovery and Characterization of Super-Enhancer-Associated Dependencies in Diffuse Large B Cell Lymphoma. Cancer Cell 24, 777–790 (2013). 8. Shu, S. et al. Response and resistance to BET bromodomain inhibitors in triple-negative breast cancer. Nature 529, 413–417 (2016). 9. Sahai, V. et al. BET Bromodomain Inhibitors Block Growth of Pancreatic Cancer Cells in r Th ee-Dimensional Collagen. Mol. Cancer e Th r. 13, 1907–1917 (2014). 10. Lovén, J. et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell 153, 320–334 (2013). 11. Sun, B. et al. Synergistic activity of BET protein antagonist-based combinations in mantle cell lymphoma cells sensitive or resistant to ibrutinib. Blood 126, 1565–1574 (2015). 12. French, C. A. Demystified molecular pathology of NUT midline carcinomas. J. Clin. Pathol. 63, 492–6 (2010). 13. French, C. NUT midline carcinoma. Nat. Rev. Cancer 14, 149–50 (2014). 14. French, C. A. et al. NSD3-NUT fusion oncoprotein in NUT midline carcinoma: Implications for a novel oncogenic mechanism. Cancer Discov. 4, 929–941 (2014). 15. Alekseyenko, A. A. et al. Ectopic protein interactions within BRD4–chromatin complexes drive oncogenic megadomain formation in NUT midline carcinoma. Proc. Natl. Acad. Sci. 114, E4184–E4192 (2017). 16. Stathis, A. et al. Clinical response of carcinomas harboring the BRD4–NUT oncoprotein to the targeted bromodomain inhibitor OTX015/MK-8628. Cancer Discov. 6, 492–500 (2016). 17. Wang, R. & You, J. Mechanistic analysis of the role of bromodomain-containing protein 4 (BRD4) in BRD4-NUT oncoprotein- induced transcriptional activation. J. Biol. Chem. 290, 2744–2758 (2015). 18. Reynoird, N. et al. Oncogenesis by sequestration of CBP/p300 in transcriptionally inactive hyperacetylated chromatin domains. EMBO J. 29, 2943–2952 (2010). 19. Wang, R. et al. Activation of SOX2 expression by BRD4-NUT oncogenic fusion drives neoplastic transformation in NUT midline carcinoma. Cancer Res. 74, 3332–3343 (2014). 20. Mazur, P. K. et al. Combined inhibition of BET family proteins and histone deacetylases as a potential epigenetics-based therapy for pancreatic ductal adenocarcinoma. Nat. Med. 21, 1163–71 (2015). 21. Ott, C. J. et al. BET bromodomain inhibition targets both c-Myc and IL7R in high-risk acute lymphoblastic leukemia. Blood 120, 2843–2852 (2012). 22. Filippakopoulos, P. et al. Selective inhibition of BET bromodomains. Nature 468, 1067–73 (2010). 23. Rathert, P. et al. Transcriptional plasticity promotes primary and acquired resistance to BET inhibition. Nature 525, 543–547 (2015). 24. Berthon, C. et al. Bromodomain inhibitor OTX015 in patients with acute leukaemia: A dose-escalation, phase 1 study. Lancet Haematol. 3, e186–e195 (2016). 25. Amorim, S. et al. Bromodomain inhibitor OTX015 in patients with lymphoma or multiple myeloma: A dose-escalation, open-label, pharmacokinetic, phase 1 study. Lancet Haematol. 3, e196–e204 (2016). 26. Wang, C. Y. & Filippakopoulos, P. Beating the odds: BETs in disease. Trends Biochem. Sci. 40, 468–479 (2015). 27. Lin, S. & Gregory, R. I. MicroRNA biogenesis pathways in cancer. Nat. Rev. Cancer 15, 321–333 (2015). 28. Iorio, M. V. & Croce, C. M. MicroRNA dysregulation in cancer: Diagnostics, monitoring and therapeutics. A comprehensive review. EMBO Mol. Med. 4, 143–159 (2012). 29. Forman, J. J., Legesse-Miller, A. & Coller, H. A. A search for conserved sequences in coding regions reveals that the let-7 microRNA targets Dicer within its coding sequence. Proc. Natl. Acad. Sci. USA 105, 14879–84 (2008). 30. Hausser, J., Syed, A. P., Bilen, B. & Zavolan, M. Analysis of CDS-located miRNA target sites suggests that they can effectively inhibit translation. Genome Res. 23, 604–615 (2013). 31. Ito, Y. et al. Identification of targets of tumor suppressor microRNA-34a using a reporter library system. Proc. Natl. Acad. Sci. 114, 3927–3932 (2017). 32. Akdemir, B., Nakajima, Y., Inazawa, J. & Inoue, J. miR-432 Induces NRF2 Stabilization by Directly Targeting KEAP1. Mol. Cancer Res. molcanres. 0232, 2017, https://doi.org/10.1158/1541-7786.MCR-17-0232 (2017). 33. Kong, Y. W., Ferland-McCollough, D., Jackson, T. J. & Bushell, M. MicroRNAs in cancer management. Lancet Oncol. 13, e249–e258 (2012). 34. Hiramoto, H. et al. miR-509-5p and miR-1243 increase the sensitivity to gemcitabine by inhibiting epithelial-mesenchymal transition in pancreatic cancer. Sci. Rep. 7, 4002 (2017). 35. Yang, H. et al. microRNA-342-5p and miR-608 inhibit colon cancer tumorigenesis by targeting NAA10. Oncotarget 7, 2709–2720 (2016). 36. Fujiwara, N. et al. miR-634 activates the mitochondrial apoptosis pathway and enhances chemotherapy-induced cytotoxicity. Cancer Res. 75, 3890–3901 (2015). 37. Warnecke-Eberz, U., Chon, S. H., Hölscher, A. H., Drebber, U. & Bollschweiler, E. Exosomal onco-miRs from serum of patients with adenocarcinoma of the esophagus: comparison of miRNA profiles of exosomes and matching tumor. Tumor Biol. 36, 4643–4653 (2015). SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 12 www.nature.com/scientificreports/ 38. Tréhoux, S. et al. Micro-RNAs miR-29a and miR-330-5p function as tumor suppressors by targeting the MUC1 mucin in pancreatic cancer cells. Biochim. Biophys. Acta - Mol. Cell Res. 1853, 2392–2403 (2015). 39. Kozomara, A. & Griffiths-Jones, S. MiRBase: Annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 42, 68–73 (2014). 40. Du, J. et al. Critical role of CDK2 for melanoma growth linked to its melanocyte-specific transcriptional regulation by MITF. Cancer Cell 6, 565–576 (2004). 41. Asghar, U., Witkiewicz, A. K., Turner, N. C. & Knudsen, E. S. The history and future of targeting cyclin-dependent kinases in cancer therapy. Nat. Rev. Drug Discov. 14, 130–146 (2015). 42. Gazdar, A. F. & Minna, J. D. Deregulated EGFR signaling during lung cancer progression: Mutations, amplicons, and autocrine loops. Cancer Prev. Res. 1, 156–160 (2008). 43. Sharma, S. V., Bell, D. W., Settleman, J. & Haber, D. A. Epidermal growth factor receptor mutations in lung cancer. Nat. Rev. Cancer 7, 169–181 (2007). 44. Rennie, W. et al. STarMirDB: A database of microRNA binding sites. RNA Biol. 13, 554–60 (2016). 45. Grayson, aR. et al. MYC, a downstream target of BRD-NUT, is necessary and sufficient for the blockade of differentiation in NUT midline carcinoma. Oncogene 33, 1–7 (2013). 46. Kuzume, T. et al. Establishment and characterization of a thymic carcinoma cell line (Ty-82) carrying t(15;19)(q15; p13) chromosome abnormality. Int. J. cancer J. Int. du cancer 50, 259–264 (1992). 47. Ke, E. E. & Wu, Y. L. EGFR as a Pharmacological Target in EGFR-Mutant Non-Small-Cell Lung Cancer: Where Do We Stand Now? Trends Pharmacol. Sci. 37, 887–903 (2016). 48. Politi, K., Ayeni, D. & Lynch, T. The Next Wave of EGFR Tyrosine Kinase Inhibitors Enter the Clinic. Cancer Cell 27, 751–753 (2015). 49. Otto, T. & Sicinski, P. Cell cycle proteins as promising targets in cancer therapy. Nat. Rev. Cancer 17, 93–115 (2017). 50. French, C. A. et al. BRD4-NUT Fusion Oncogene: A Novel Mechanism in Aggressive Carcinoma Advances in Brief BRD4-NUT Fusion Oncogene: A Novel Mechanism in Aggressive Carcinoma 1. Cancer Res. 63, 304–307 (2003). 51. French, C. A. et al. BRD-NUT oncoproteins: a family of closely related nuclear proteins that block epithelial differentiation and maintain the growth of carcinoma cells. Oncogene 27, 2237–2242 (2008). 52. Wang, Y. H. et al. BRD4 promotes pancreatic ductal adenocarcinoma cell proliferation and enhances gemcitabine resistance. Oncol. Rep. 33, 1699–1706 (2015). 53. Shimada, Y., Imamura, M., Wagata, T., Yamaguchi, N. & Tobe, T. Characterization of 21 newly established esophageal cancer cell lines. Cancer 69, 277–284 (1992). 54. Endo, H. et al. Potential of tumor-suppressive miR-596 targeting LGALS3BP as a therapeutic agent in oral cancer. Carcinogenesis 34, 560–569 (2013). 55. Lessey, B. A. et al. Immunohistochemical analysis of human uterine estrogen and progesterone receptors throughout the menstrual cycle. J. Clin. Endocrinol. Metab. 67, 334–40 (1988). Acknowledgements We thank Ayako Takahashi and Rumi Mori for technical assistance. This work was supported by KAKENHI (15H05908, 16K14630, 15K18401, 15K19040) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), and partially supported by the Project for Cancer Research And Therapeutic Evolution (P-CREATE) from Japan Agency for Medical Research and development, AMED. This study was also partly supported by Nanken-Kyoten, TMDU. Author Contributions E. Tonouchi, and Y. Gen were involved in research design, performed the experiments, analyzed data and wrote the manuscript. T. Muramatsu and H. Hiramoto were involved in research design, performed the expriments, and analyzed data. K. Tanimoto contributed to TCGA data analysis. J. Inoue contributed materials. J.Inazawa was involved in research design, wrote the manuscript, and study supervision. Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-018-22767-y. Competing Interests: The authors declare no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2018 SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 13 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Scientific Reports Springer Journals

miR-3140 suppresses tumor cell growth by targeting BRD4 via its coding sequence and downregulates the BRD4-NUT fusion oncoprotein

Free
13 pages

Loading next page...
 
/lp/springer_journal/mir-3140-suppresses-tumor-cell-growth-by-targeting-brd4-via-its-coding-FuJBdfBJwv
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN
2045-2322
D.O.I.
10.1038/s41598-018-22767-y
Publisher site
See Article on Publisher Site

Abstract

www.nature.com/scientificreports Correction: Publisher Correction OPEN miR-3140 suppresses tumor cell growth by targeting BRD4 via its coding sequence and Received: 8 November 2017 downregulates the BRD4-NUT Accepted: 27 February 2018 Published online: 14 March 2018 fusion oncoprotein 1,2 1 1 1 Erina Tonouchi , Yasuyuki Gen , Tomoki Muramatsu , Hidekazu Hiramoto , Kousuke 3 1 1,4 Tanimoto , Jun Inoue & Johji Inazawa Bromodomain Containing 4 (BRD4) mediates transcriptional elongation of the oncogene MYC by binding to acetylated histones. BRD4 has been shown to play a critical role in tumorigenesis in several cancers, and the BRD4-NUT fusion gene is a driver of NUT midline carcinoma (NMC), a rare but highly lethal cancer. microRNAs (miRNAs) are endogenous small non-coding RNAs that suppress target gene expression by binding to complementary mRNA sequences. Here, we show that miR-3140, which was identified as a novel tumor suppressive miRNA by function-based screening of a library containing 1090 miRNA mimics, directly suppressed BRD4 by binding to its coding sequence (CDS). miR-3140 concurrently downregulated BRD3 by bind to its CDS as well as CDK2 and EGFR by binding to their 3’ untranslated regions. miR-3140 inhibited tumor cell growth in vitro in various cancer cell lines, including EGFR tyrosine kinase inhibitor-resistant cells. Interestingly, we found that miR-3140 downregulated the BRD4-NUT fusion protein and suppressed in vitro tumor cell growth in a NMC cell line, Ty-82 cells. Furthermore, administration of miR-3140 suppressed in vivo tumor growth in a xenograft mouse model. Our results suggest that miR-3140 is a candidate for the development of miRNA-based cancer therapeutics. e b Th romodomain and extra-terminal domain (BET) family proteins, including BRD2, BRD3, BRD4, and BRDT, contain two conserved bromodomains that are associated with acetylated lysine in histones, facilitating transcrip- 1,2 tional activation as epigenetic readers . Among the BET family proteins, BRD4 has been shown to play a critical 3,4 5 role in promoting tumor growth in several cancers, including acute myeloid leukemia , multiple myeloma , 6 7 8 9 MLL-fusion leukemia , diffuse large B cell lymphoma , triple negative breast cancer , and pancreatic cancer . BRD4 is enriched at super-enhancers of several oncogenes, such as MYC, CCND2, and BCL-XL, in cancer cells, upregulating the transcription of these genes . BRD4 increases the transcription of the other oncogenes BCL2, CDK6, and MYC by binding to the chromatin locus of these genes and recruiting positive transcriptional elonga- 11 6 tion factor complex (P-TEFb) to the promoter . Thus, BRD4 is thought to be a rational target for cancer therapy . NUT midline carcinoma (NMC) is a poorly differentiated carcinoma that arises in the midline of the upper 12,13 aerodigestive tract or the mediastinum . NMC is rare, refractory to conventional treatments, and highly lethal, 12,13 with a median survival period of 6.7–9.5 months . The pathogenesis of NMCs involves the BRD4-NUT fusion gene, which is caused by a unique chromosome translocation t(15; 19)(q13; p13.1) in the majority of cases, although BRD3-NUT fusion by a t(9; 15)(q34; q14), NSD3-NUT fusion by a t(8; 15)(p12; q15), and ZNF532-NUT 13–15 fusion by a t(15; 18)(q14; q23) occur in the remaining few cases . The translocation breakpoints occur within intron 10 of the BRD4 gene (19p13.1) and intron 2 of NUT (15q14), such that the BRD4-NUT protein contains Department of Molecular Cytogenetics, Medical Research Institute, Tokyo Medical and Dental University, Tokyo, Japan. Department of Maxillofacial Surgery, Graduate School, Tokyo Medical and Dental University, Tokyo, Japan. 3 4 Genome Laboratory, Medical Research Institute, TMDU, Tokyo, Japan. Bioresource Research Center, Tokyo Medical and Dental University, Bunkyo-ku, Tokyo, Japan. Correspondence and requests for materials should be addressed to J.I. (email: johinaz.cgen@mri.tmd.ac.jp) SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 1 www.nature.com/scientificreports/ both acetyl-histone binding bromodomains and the extraterminal domain of BRD4 (i.e., the full functional domain of BRD4) . The BRD4-NUT oncoprotein promotes tumor cell growth through the function of BRD4 as 16–18 well as that of NUT . e fir Th st-generation BET bromodomain inhibitor JQ1 binds to the acetyl-lysine binding pocket of BRD4, and thus, JQ1 depletes not only BRD4 but also BRD4-NUT from chromatin by preventing the binding of BRD4 to 17,19 8,20,21 chromatin . As a result, JQ1 inhibits BRD4-mediated transcription of oncogenes, such as MYC . Several studies have shown that BET bromodomain inhibitors are highly effective against various intractable cancers, 8 9 22 including triple negative breast cancer , pancreatic cancer , and NMC , although resistance to BET bromodo- 8,23 main inhibition can be acquired through various mechanisms . Several clinical trials using BET bromodomain 16,24–26 inhibitors have been started . microRNAs (miRNAs) are endogenous small non-coding RNAs of 20–25 nucleotides that decompose or 27,28 inhibit protein translation of mRNA by binding to complementary mRNA sequences . An individual miRNA has multiple target genes, and an individual gene can be targeted by multiple miRNAs . Although many studies have revealed that miRNAs repress gene expression by binding to 3′ untranslated regions (3′UTR) of mRNAs, an increasing body of evidence supports that miRNAs also bind to the coding sequences (CDS) of target mRNAs 29–31 and that miRNA binding to the CDS of mRNAs can effectively suppress translation . Our previous study also revealed that miR-432–3p directly targets KEAP1 via its CDS . Because a single miRNA can simultaneously repress numerous target genes, miRNA mimics targeting several oncogenes may be useful as therapeutic agents for cancer therapy . In the present study, to investigate the novel candidate miRNAs for the development of miRNA-based cancer therapeutics, we conducted function-based screening using a miRNA library containing 1090 miRNAs. We revealed that miR-3140, identified as a novel tumor suppressive miRNA (TS-miRNA), repressed BRD4 directly by binding to its CDS. Furthermore, we showed that miR-3140 also downregulated the BRD4-NUT oncoprotein in NMC cells. miR-3140 suppresses other tumor promoting genes, such as EGFR and CDK2, via 3′UTR. Finally, the effects of miR-3140 on tumor growth were tested in vivo. Results miR-3140 was identified as a novel TS-miRNA by function-based miRNA library screening. To extract novel TS-miRNAs, we performed function-based miRNA library screening from a library containing 1090 miRNA mimics in Panc1 cells. The strategy and brief results of this study are shown in Fig.  1a. In this study, the relative cell growth ratio was determined after transfection of each miRNA in two Panc1-derived clones, PEcadZsG-Panc1 #1 and #2 cells, which were established in our previous study . Figure 1b shows the results of this screening in Panc1 #1 (left) and #2 (right) 72 hours aer t ft ransfection with each miRNA. We set the criteria for extracting TS-miRNAs (cell growth ratio < 0.6), and then extracted 29 miRNAs from Panc1 #1 and 65 miRNAs from Panc1 #2 cells, respectively (Fig. 1c). Twelve miRNAs that inhibited cell growth in both screening assays 35–38 were identified as candidate TS-miRNAs (Table  1) . Then, 4 miRNAs were extracted according to annotation confidence, which means the certainty of the actual existence of a particular miRNA, determined using miRBase (http://www.mirbase.org/) . Among these 4 miRNAs, little is known about the tumor-suppressive function of miR-3140. Thus, we focused on a detailed analysis of miR-3140. miR-3140 inhibited in vitro tumor cell growth in various cancer cell lines. To confirm the growth suppressive effect of miR-3140 observed in the function-based screening, we evaluated the cell proliferation in vitro aer t ft ransfection with miR-3140 or miR-NC in two pancreatic cancer cell lines, Panc1 and MIAPaCa2 cells, respectively. Consistent with the screening results, overexpression of miR-3140 significantly inhibited tumor cell growth in both cell lines (Fig. 1d,e). We next investigated whether miR-3140 suppresses in vitro tumor cell growth in various cell lines including triple negative breast cancer (MDA-MB-231), esophageal cancer (KYSE150), liver cancer (Sk-Hep1), and non-small-cell lung carcinoma (HUT29 and A549). We also tested the effects of miR-3140 in KYSE150 CDDP-R cells, which are resistant to cisplatin . As shown in Fig. 1e, overexpression of miR-3140 markedly reduced tumor cell growth in all the cancer cell lines we tested. Taken together, miR-3140 inhibited in vitro tumor cell growth. miR-3140 repressed CDK2 and EGFR through 3’UTR interactions. To investigate genes downreg- ulated by miR-3140, gene expression array analysis was performed in three cancer cell lines, Panc1, MIAPaCa2, and MDA-MB-231 cells. As shown in a Venn diagram (Fig. 2a, left), the expression levels of 228 genes decreased by more than 2-fold in miR-3140-transfected cells compared with miR-NC-transfected cells. Among 228 genes, 99 genes were predicted to be direct targets of miR-3140 according to the TargetScan program (http://www.tar- getscan.org) because candidate binding-sites for seed sequences of miR-3140 exist in the 3′UTRs (Fig. 2a, right). Among the overlapping genes, we extracted CDK2, CDK6, and EGFR, which are all known to promote tumor 40–42 cell growth . Western blotting showed that the protein levels of CDK2, CDK6, and EGFR were markedly reduced in miR-3140-transfectants of Panc1 and MIAPaCa2 cells, compared with miR-NC-transfectants (Fig. 2b). To determine whether miR-3140 can directly bind to the 3′UTR of CDK2, CDK6 and EGFR, we performed lucif- erase assays using reporter plasmid vectors containing the wild type (Wt) or mutant (Mt) 3′UTR of these genes in Panc1 cells. The luciferase activity of the Wt vectors, except for the 3′UTR of CDK6 (Supplementary Fig. S1), was decreased compared with the empty vector (EV) in miR-3140-transfected cells and completely restored by the Mt vector for CDK2 and EGFR (Fig. 2c). These results indicate that CDK2 and EGFR are direct target genes of miR- 3140 through binding the 3′UTR region, although CDK6 was downregulated indirectly by miR-3140. To investigate whether CDK2 and EGFR participate in the growth suppressive effects of miR-3140, we exam- ined the cell proliferation in vitro aer t ft ransfection with small interfering RNA (siRNA) targeting CDK2 , EGFR, or negative control siRNA. As shown in Fig. 2d, si-CDK2 significantly reduced tumor cell growth in MIAPaCa2 SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 2 www.nature.com/scientificreports/ Figure 1. miR-3140 was identified as a novel tumor suppressive miRNA (TS-miR) by function-based miRNA library screening. (a) The strategy used to identify novel TS-miRs in this study. (b) Results of function-based miRNA library screening in two Panc1-derived clones (PEcadZsG-Panc1 #1 and PEcadZsG-Panc1 #2 cells), using the Pre-miR miRNA Precursor Library-Human V15 consisting of 1,090 mature human miRNA mimics. e ce Th ll growth ratio was assessed with crystal violet staining using a relative ratio normalized based on the cell survival rate of cells transfected with negative control miRNA (miR-NC). A value of 0.6 was used to determine the cut-off value for growth inhibition (dotted line). (c) The Venn diagram showing the overlap of 12 miRNAs between Panc1 #1 (3000 cells/well) and #2 (5000 cells/well). (d) Phase contrast images of Panc1 and MIAPaCa2 cells transfected with 10 nmol/L of miR-NC or miR-3140. Images were obtained 3 days aer t ft ransfection. (e) Cell growth assay in various types of cancer cells. Pancreatic cell lines (Panc1 and MIAPaCa2), a triple negative breast cancer cell line (MDA-MB-231), esophageal squamous cell carcinoma cell lines (KYSE150 and the cisplatin resistant cell line KYSE150 CDDP-R), a liver cancer cell line (Sk-Hep1), and non-small-cell lung cancer cell lines (HUT29 and A549) were transfected with 10 nmol/L of miR-NC or miR-3140. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 3 www.nature.com/scientificreports/ Panc1 Panc1 Annotation #1 #2 Confidence † † ‡ miRNA Precursor Mature Sequence Ratio Ratio Location Host gene (miRBase) Ref hsa-miR-342-5p AGGGGUGCUAUCUGUGAUUGA 0.432 0.323 14q32.2 EVL high has-miR-449b* CAGCCACAACUACCCUGCCACU 0.357 0.446 5q11.2 CDC20 — hsa-miR-608 AGGGGUGGUGUUGGGACAGCUCCGU 0.487 0.503 10q24.31 SEMA4G hsa-miR-634 AACCAGCACCCCAACUUUGGAC 0.443 0.579 17q24.2 PRKCA hsa-miR-671-3p UCCGGUUCUCAGGGCUCCACC 0.563 0.467 7q36.1 CHPF2 high hsa-miR-876-3p UGGUGGUUUACAAAGUAAUUCA 0.407 0.439 9p21.1 LINGO2 high 20q11.22 GDF5 — hsa-miR-1289 UGGAGUCCAGGAAUCUGCAUUUU 0.576 0.338 5q31.1 FSTL4 hsa-miR-1293 UGGGUGGUCUGGAGAUUUGUGC 0.580 0.425 12q13.12 LIMA1 — hsa-miR-3126-3p CAUCUGGCAUCCGUCACACAGA 0.582 0.533 2p13.3 ANTXR1 — hsa-miR-3140 AGCUUUUGGGAAUUCAGGUAGU 0.508 0.459 4q31.3 FBXW7 high — hsa-miR-3165 AGGUGGAUGCAAUGUGACCUCA 0.533 0.370 11q13.4 NUMA1 — hsa-miR-3173 0.551 0.372 14q32.13 DICER1 — Table 1. Summary of 12 miRNAs selected as candidates for novel tumor suppressive miRNAs in functional based screening with Pre-miR miRNA Precursor Library v15. Ratio means the cell growth ratio of viable cells as assessed by crystal violet staining 72 hours aer t ft ransfection with miRNAs, relative to the control transfectants. Reference is TS-miRNA. cells, but not in Panc1 cells. Similarly, si-EGFR suppressed cell proliferation in Panc1 cells (Fig. 2e), but not in MIAPaCa2 cells (Supplementary Fig. S2). These results suggested that CDK2 and EGFR contribute to tumor cell growth in a cell-dependent manner. Thus, downregulation of CDK2 and EGFR may partially contribute to miR-3140-mediated suppression of tumor cell growth. miR-3140 suppressed tumor cell growth in EGFR tyrosine kinase inhibitor (EGFR-TKI)-resistant lung cancer cells. Since miR-3140 directly targeted EGFR, we next examined whether miR-3140 over- came the resistance to EGFR-TKIs. EGFR is one of the most frequently mutated “driver” genes in non-small cell lung carcinoma (NSCLC) and EGFR-TKIs are used for the treatment of EGFR-mutated-NSCLC. Therefore, we examined the ee ff cts of miR-3140 in EGFR-mutated-NSCLC cells. In agreement with past reports, HUT29 cells, which harbor the EGFR L858R mutation, were sensitive to the EGFR-TKIs gefitinib and erlotinib, whereas NCI-H1975 cells, which harbor the EGFR L858R/T790M double mutation, were resistant to gefitinib and erlo- tinib (Supplementary Fig. S3a,b) . Both miR-3140 and si-EGFR reduced the expression of EGFR in NCI-H1975 cells, resulting in the suppression of in vitro tumor cell growth in NCI-H1975 cells (Fig. 2e,f ). es Th e results sug- gested that miR-3140 may overcome the acquired resistance to EGFR-TKIs at least in part by suppressing mutant EGFR in NSCLC cells. Suppression of BET family genes inhibited tumor cell growth. Because the effects of miR-3140 on in vitro tumor cell growth were much greater than those of si-CDK2 and si-EGFR (Figs 1e and 2d–f ), concurrent downregulation of additional target genes may be necessary for the induction of miR-3140-mediated inhibition of tumor cell growth. A growing body of evidence suggested that coding sequence (CDS) of mRNA could be 29–31 directly targeted by a miRNA-containing RISC complex . u Th s, we explored additional target genes of miR- 3140 using the STarMirDB database (http://sfold.wadsworth.org) , which predicts miRNA target genes based on CDS binding. Among the 228 genes downregulated by miR-3140 in 3 cell lines (Fig. 2a), candidate binding-sites for the seed sequence of miR-3140 exist in the CDS of 103 genes (Fig. 3a, Supplementary Table S1). From these 103 overlapping genes, we extracted BRD4 as a candidate target gene of miR-3140, because BRD4 has been shown to play a critical role in promoting tumor growth in several cancers through upregulates the transcription of oncogenes, including MYC and CCND2. Our gene expression array analysis showed that miR-3140 downregu- lated the expression of MYC and CCND2 as well as BRD4 in Panc1 cells, suggesting that BRD4 plays a critical role for tumor cell growth in Panc1 cells. As shown in Fig. 3b, miR-3140 markedly reduced BRD4 protein expression. Furthermore, we found that other BET family genes, BRD2 and BRD3, together with genes for the downstream 10,20 signaling pathways of BRD4, MYC, phosphorylated STAT3 and Cyclin D2 , were significantly reduced by miR- 3140 (Fig. 3b). Transfection of siRNA targeting each BET family gene, especially si-BRD4, markedly inhibited tumor cell growth (Fig. 3c, Supplementary Fig. S4). Conversely, overexpression of BRD4 promoted cell prolifera- tion compared to transfection with an empty vector (Fig. 3d). Taken together, downregulation of BRD4 contrib- uted to miR-3140-mediated suppression of tumor cell growth. miR-3140 repressed BRD4 and BRD3 by directly binding to the CDS region. To examine whether miR-3140 can directly bind to the CDS of BRD4, we first performed luciferase assays using reporter plasmid vec- tors containing Wt or Mt seed sequences in the CDS. Because two candidate binding-sites for the seed sequence of miR-3140 exist within the CDS of BRD4, three mutant constructs (Mt1, Mt2 and Mt1 + 2) were established (Fig. 4a). e Th luciferase activity of the Wt vector was decreased compared with EV in miR-3140 transfected cells and partially recovered with each Mt1 or Mt2 vector and completely restored with the Mt1 + 2 vector (Fig. 4a). We next examined whether miR-3140 suppressed the expression of BRD4 by binding to these positions within the SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 4 www.nature.com/scientificreports/ Figure 2. miR-3140 targeted CDK2 and EGFR by binding their 3′UTR regions. (a) Left, identification of downregulated genes aer ft miR-3140 transfection by a gene expression array. The Venn diagram shows that 228 genes were commonly downregulated (fold change >2) by transfection of miR-3140 in Panc1, MIAPaCa2, and MDA-MB-231 cells. Right, prediction of candidate target genes regulated by miR-3140 via their 3′UTR. The Venn diagram shows that 99 genes were predicted as candidate 3′UTR-targets of miR-3140 by the TargetScan program. (b) Western blot analysis of CDK2, CDK6, and EGFR in Panc1 and MIAPaCa2 cells 72 hours aer ft transfection with 10 nmol/L of miR-NC or miR-3140. (c) Luciferase reporter assays. Panc1 cells were transfected with the pmirGLO Dual Luciferase vectors containing wild type (Wt) CDK2 and EGFR or mutant (Mt) 3’UTR target sites of these genes, and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding site of miR-3140 within the 3′UTR of each gene and mutant sequences. Bottom, results of the luciferase assay; *P < 0.05. (d,e) Evaluation of the effect of si-CDK2 or si-EGFR. Western blot analysis (top) and cell growth assay (bottom) in indicated cell lines aer t ft ransfection with 20 nmol/L of negative control siRNA (si- NC) or siRNA targeting each gene. (f) Evaluation of the effect of miR-3140 in EGFR-TKI-resistant lung cancer cells. Western blot analysis (top) and cell growth assay (bottom) in the indicated cell line aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. Cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. CDS. Synonymous mutations were generated at each site of the Halotag-BRD4 protein (Supplementary Fig. S5a). Each of those Wt and Mt vectors were transfected into MIAPaCa2 cells followed by transfection of miR-NC or miR-3140, respectively. Whereas the expression level of exogenously expressed BRD4-Wt was reduced in the miR- 3140-transfected cells, the miR-3140-induced reduction of exogenous BRD4 was restored by all three patterns of each mutation at binding-sites for the seed sequence of miR-3140 (Fig. 4b). These combined data suggested that miR-3140 can downregulate BRD4 expression by directly targeting its CDS. Similarly, we examined whether BRD3 was also a candidate target gene of miR-3140. Candidate binding-sites for the seed sequence of miR-3140 exist within both the CDS and the 3′UTR of the BRD3 gene. The luciferase activity of the 3′UTR-Wt vector did not decrease, but that of the CDS-Wt vector was decreased compared with EV in miR-3140 transfectant, and this decrease was completely recovered by the CDS-Mt vector (Fig. 4c,d). Synonymous mutations were generated at the CDS site of the Halotag-BRD3 protein (Supplementary Fig. S5b). While the expression level of exogenously expressed BRD3-Wt was decreased in the miR-3140-transfected cells, this miR-3140-induced decrease in exogenous BRD3 was restored by the mutations at a binding-site for the seed sequence of miR-3140 (Fig. 4e), suggesting that miR-3140 can downregulate BRD3 expression by targeting its CDS as well as BRD4. Taken together, miR-3140 targets BRD4 and BRD3 through their binding CDS regions. miR-3140 downregulated the BRD4-NUT fusion protein. Based on the results that miR-3140 directly targets the CDS of BRD4, we investigated whether miR-3140 suppresses the BRD4-NUT fusion gene. The BRD4-NUT fusion is caused by t(15:19) translocation and drives NUT midline carcinoma (NMC), a rare, SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 5 www.nature.com/scientificreports/ Figure 3. miR-3140 downregulated BET family genes. (a) Prediction of candidate targets regulated by miR- 3140 via their coding sequences (CDS). The Venn diagram shows that 103 genes were predicted as candidate CDS-targets of miR-3140 by the STarMirDB program. (b) Western blot analysis of BET family proteins (BRD2, BRD3, and BRD4) and downstream targets of BRD4 (Cyclin D2, MYC, and p-STAT3) in Panc1 and MIAPaCa2 cells 72 hours aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. (c,d) Effects of knockdown of BET family proteins (c), or overexpression of BRD4 (d) on cell growth in Panc1 cells. Cells were transfected with 20 nmol/L of si-NC or siRNA targeting BET family proteins (c), and empty vector or BRD4-HaloTag expression vector (d). Left, western blot analysis of indicated proteins in Panc1 cells 72 hours aer t ft ransfection. Right, results of the cell growth assay. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. 12 21,45 highly lethal cancer (Fig. 5a) . Consistent with past reports , JQ1 reduced MYC and significantly suppressed in vitro tumor growth of Ty-82 cells, a NMC cell line, which harbor t(15;19) bearing the BRD4-NUT fusion gene (Fig. 5b) . As shown in Fig. 5c, miR-3140 repressed the expression of the BRD4-NUT fusion protein and its downstream target MYC in Ty-82 cells. As a result, miR-3140 effectively suppressed in vitro tumor cell growth of Ty-82 cells (Fig. 5c). Next, to test whether miR-3140 can suppress tumor cell growth of JQ1-resistant cells, JQ1-resistant cells (Ty- 82 JQ1-R) were generated from Ty-82 cells (IC50: 0.43 µ M in Ty-82 cells, 1.37 µ M in Ty-82 JQ1-R cells; Fig. 5d). To determine whether Ty-82 JQ1-R cells have developed resistance by acquiring BRD4-NUT-dependence or not, we examined the effects of JQ1 in Ty-82 JQ1-R cells. The expression of MYC, a downstream effector of BRD4-NUT, was not suppressed sufficiently by JQ1 treatment in Ty-82 JQ1-R cells, whereas the expression of MYC was suppressed by JQ1 treatment in a dose-dependent manner in Ty-82 cells (Supplementary Fig. S6a). Knockdown of BRD4-NUT using RNA interference downregulated the expression of MYC and suppressed cell growth both in Ty-82 JQ1-R cells (Supplementary Fig. S6b). These results suggested that JQ1 could not block the effects of BRD4-NUT sufficiently in Ty-82 JQ1-R cells, although the mechanism is unknown. miR-3140 downreg- ulated the expression of BRD4-NUT fusion protein and MYC, and suppressed tumor cell growth in Ty-82 JQ1-R cells as well as Ty-82 cells (Fig. 5c,e). These combined data suggested that the resistance of JQ1 is, at least in part, dependence of BRD4-NUT/MYC pathway. Thus, our results suggested that miR-3140 inhibited cell growth in Ty-82 JQ1-R cells by targeting BRD4-NUT. miR-3140 suppressed tumor growth in a xenograft mouse model. We next examined the therapeu- tic effect of miR-3140 through the local administration of dsRNA mimicking miRNA around MIAPaCa2-derived subcutaneous tumors in nude mice. miR-NC (left) or miR-3140 (right) were administered into the subcutaneous space around tumors 5 times (7, 11, 15, 18, 21 days aer t ft he injection of MIAPaCa2 cells; Fig.  6a). Administration of miRNA did not produce any adverse consequences, such as body weight loss or local damage. As a result, tum- ors treated with miR-3140 at 23 days aer t ft he injection of MIAPaCa2 cells were significantly smaller than tumors treated with miR-NC (Fig. 6b,c, Supplementary Fig. S7a,b). We confirmed that the expression of miR-3140 was significantly high in the miR-3140-treated tumors compared to miR-NC -treated tumors by qRT-PCR (Fig. 6d). Furthermore, immunohistochemical staining showed that the expression of BRD4, BRD3, CDK2, and EGFR, the SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 6 www.nature.com/scientificreports/ Figure 4. miR-3140 directly targeted BRD4 and BRD3 by binding their coding regions. (a) Luciferase reporter assay. MIAPaCa2 cells were transfected with the pmirGLO Dual Luciferase vectors containing Wt or Mt BRD4, or empty vector (EV), and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding sequence of miR-3140 within the CDS of BRD4 and mutant sequences are indicated. Bottom, the results of luciferase assay; *P < 0.05. (b) Western blot analysis of BRD4 in MIAPaCa2 cells. Cells were co-transfected with the Wt or Mt BRD4 expression vector, and aer 24 ft hours, either 10 nmol/L of miR-NC or miR-3140 was additionally transfected. (c,d) Luciferase reporter assays. Panc1 cells were transfected with a reporter plasmid (Wt of BRD3 3′UTR or EV (c), or Wt, Mt of BRD3 CDS or EV (d)) and aer 6 ft hours, either miR-NC or miR-3140 was additionally transfected. Top, putative binding sequences of miR-3140 within the BRD3 3′UTR (c) and the BRD3 CDS (d). Bottom, results of the luciferase assay; *P < 0.05. (e) Western blot analysis of BRD3 in Panc1 cells. Cells were co-transfected with the Wt or Mt BRD3 expression vector, and aer 24 ft hours, either 10 nmol/L of miR-NC or miR-3140 was additionally transfected. SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 7 www.nature.com/scientificreports/ Figure 5. miR-3140 downregulates the BRD4-NUT fusion protein. (a) Schema of the relationship between mRNA of the BRD4-NUT fusion gene in NUT midline carcinoma and the location of miR-3140 target sites. (b) Effects of the BET bromodomain inhibitor JQ1 in Ty-82 cells. Left, the results of the cell growth assay. The cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. Right, western blot analysis of MYC in Ty-82 cells 3 hours aer t ft ransfection. (c,e) Left, cell growth assay in a NUT midline carcinoma cell line, Ty-82 (c) and its JQ1-resistant cells (Ty-82 JQ1-R) cells (e) which were transfected with 10 nmol/L of miR-NC or miR-3140. Cell growth ratio was assessed with the WST-8 assay based on the relative ratio compared with day 1. Bar, SD for triplicate experiments; *P < 0.05. Right, western blot analysis of the BRD4-NUT fusion protein and MYC in Ty-82 (c) and Ty-82 JQ1-R cells (e) 72 hours aer t ft ransfection with 10 nmol/L of miR-NC or miR-3140. (d) Dose response curve of JQ1 for Ty-82 (upper) and Ty-82 JQ1-R cells (lower) at 50 hours following treatment with JQ1. Cell indexes were normalized with the last time point before JQ1 treatment. targets of miR-3140, were reduced in the resected tumors treated with miR-3140 (Fig. 6e, Supplementary Fig. S8). Although other targets of miR-3140 may also participate in the tumor suppressive activity, miR-3140 inhibited tumor growth in vivo at least in part by suppressing BRD4, BRD3, CDK2 and EGFR. Discussion Here, we identified miR-3140 as a novel TS-miRNA by function-based miRNA library screening. As summarized in Fig. 6f, we showed that miR-3140 inhibited tumor cell growth in various cancer cells both in vitro and in vivo at least in part by directly targeting BRD4, BRD3, CDK2, and EGFR. Furthermore, we revealed that miR-3140 suppressed the BRD4-NUT oncoprotein in NMC cells and that miR-3140 inhibited in vitro tumor cell growth in NMC cells. We revealed that miR-3140 directly suppressed EGFR and CDK2 via 3′UTR interaction. Both EGFR and CDK2 play a role in cancer progression. EGFR is activated by gain-of-function mutations or amplification in several cancers including lung, head and neck, ovary, colon, and esophagus . Small molecule inhibitors, such as gefitinib and erlotinib, are used in lung cancer patients who have mutations on EGFR . We showed that miR- 3140 suppressed in vitro tumor cell growth by directly reducing EGFR expression in NCI-H1975 cells, which 43,47,48 are EGFR-TKI resistant due to EGFR L858R/T790M double mutations . Although in vivo experiments are needed to confirm the effects of miR-3140 on EGFR-TKI-resistant cells, our results suggest that miR-3140 may overcome the acquired resistance to EGFR-TKIs in lung cancer. CDK2 is another target of miR-3140. The cat- alytic activity of CDK2-CyclinE complexes is hyperactivated in several cancers by Cyclin E amplification, or a loss-of-function mutation of FBXW7, a ubiquitin ligase for Cyclin E degradation, although CDK2 mutations are rare in human cancers . Thus, suppression of CDK2 by miR-3140 may contribute to the inhibition of tumor growth in some cancers. Second, we identified that miR-3140 suppressed BRD4 and BRD3 through binding to their CDS. Because BET bromodomain inhibitors suppressed BET family-mediated transcription of the oncogene MYC, they were 5,22 reported to be promising agents for many cancers . Several BET inhibitors have entered into clinical trials in SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 8 www.nature.com/scientificreports/ Figure 6. era Th peutic effects of miR-3140 for tumor growth in vivo . (a) The experimental schedule for miR- 3140 treatment in nude mice which were subcutaneously inoculated with MIAPaCa2 cells. (b) The representative image of tumor-bearing nude mice at 23 days aer t ft he inoculation of MIAPaCa2 cells. Tumors are denoted by arrowheads. (c) Tumor growth curves of xenograft mouse models treated with miR-NC or miR- 3140 (n = 5, each). Tumor volume was calculated using the following formula: (shortest diameter) × (longest diameter) × 0.5. Bar, SD for 5 mice; *P < 0.05. (d) Expression analysis of miR-3140 in resected tumors. The expression level of miR-3140 was measured by qRT-PCR using the relative ratio normalized based on the expression of RNU6B. Each experiment was performed in duplicate. Bar, SD. (e) Representative images of immunohistochemical staining for BRD4, BRD3, CDK2 and EGFR in resected tumors. Scale bar, 50 μm (f) Schematic models for the mechanism by which miR-3140 suppresses tumor growth. 16,24,25 some cancers, including NMC . However, development of resistance to BET inhibition has been reported, in the same way as occurs in the other molecular targeted drugs. For example, BET inhibitor resistance is associated with increased BRD4 phosphorylation mediated by casein kinase 2 (CK2) in triple negative breast cancer cells . SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 9 www.nature.com/scientificreports/ We showed that miR-3140 suppressed the expression of BRD4-NUT in NMC cells. In the BRD4-NUT fusion gene, the translocation breakpoints of BRD4 is within intron 10 , and the binding sequence of miR-3140 exists within exon 9 of BRD4. Thus, miR-3140 can directly target the BRD4-NUT fusion gene and consequently, miR- 3140 suppressed in vitro tumor cell growth of Ty-82 cells as well as JQ1. Unfortunately, Ty-82 cells hardly formed subcutaneous tumors in our xenograft model. Further in vivo studies of the effect of miR-3140 for NMC cells are needed for future work. Our data suggested that miR-3140 could suppress in vitro tumor cell growth of JQ1-resistant Ty-82 cells. Direct suppression of BRD4-NUT and concurrent downregulation of other tumor promoting genes such as EGFR and CDK2 by miR-3140 may potentially overcome resistance to BET inhibitors in NMC cells, although further studies are needed to clarify the acquired resistance of BET inhibitiors. On the other hand, miR-3140 might not suppress BRD3-NUT, since the translocation breakpoint of BRD3 in BRD3-NUT fusion gene is within intron 9 , and the binding sequence of miR-3140 exists within exon 10 of BRD3. It was reported that the expression of BRD4 were increased by gemcitabine treatment in Panc1 and MIAPaCa2 cells, and the combination of BRD4 silencing and gemcitabine treatment had a synergistic effect on the chemo- therapeutic efficacy . A combined treatment of gemcitabine with miR-3140 also may increase gemcitabine sen- sitivity in pancreatic cancer. In primary samples, we did not observe a correlation between the expression levels of miR-3140 and over- all survival in pancreatic cancer, breast cancer, or acute myeloid leukemia according to the TCGA database (Supplementary Fig. S9a–c). This is because the miR-3140 expression levels are very low in the majority of pri- mary tumor samples, although miR-3140 is predicted to exist in human tissue according to the miRBase database. Further investigation of the expression pattern of miR-3140 in non-tumor tissues is required. Because of recent advances in clinical tumor sequencing and developments in small molecule inhibitors, molecular targeted therapy is one of the key therapeutic strategies along with conventional chemotherapy, radi- ation therapy, and immunotherapy. However, acquired resistance to molecular targeted drugs is a major prob- lem for cancer treatment . Reduction of the molecular target itself and other tumor promoting targets by a miRNA-based therapy may contribute to overcome the drug tolerance that develops against molecularly targeted drugs. In this study, our findings suggest that miR-3140 suppresses tumor cell growth not only in various can- cer cells, but also in EGFR-TKIs-resistant cells and JQ1-resistant cells. Although further in vivo studies of drug delivery systems and possible off-target effects are needed, miR-3140 may be a candidate for the development of miRNA-based cancer therapeutics. Materials and Methods Cell culture. Panc1 and MIAPaCa2 cells were cultured in Dulbecco’s Modified Eagle Medium containing 10% Fetal Bovine Serum (FBS). MDA-MB-231 cells from the American Type Culture Collection (Manassas, VA) were maintained in L-15 medium containing 10% FBS. KYSE150 cells, a gift from Dr. Shimada Y (Toyama University) , NCI-H1650, NCI-H1975, A549, HUT29 from ATCC, the NUT midline carcinoma cell line Ty-82 from JCRB, and 11–18 were maintained in RPMI1640 medium containing 10% FBS. The KYSE150 CDDP- resistant cell line (KYSE150 CDDP-R), which is resistant to cisplatin, was generated previously . Sk-Hep1 cells were maintained in Eagle’s Minimum Essential Medium containing 10% FBS. All cell lines were maintained at 37 °C with 5% CO2. The all experiments were carried out in accordance with the approved guidelines and regu- lations (2010-5-4, 2016-011C2). Function-based miRNA screening. Two Panc1-derived clones (PEcadZsG-Panc1 #1 and #2 cells) were seeded on 96-well plates. Aer 24 ft hours, each clone was transfected in duplicate with each of the 1090 dsRNAs from the Pre-miR miRNA Precursor Library-Human V15 (Thermo Fisher Scientific, CA) or a negative control miRNA (miR-NC) using an RNA concentration of 10 nmol/L. Aer 72 ft hours, viable cell number was assessed by the crystal violet staining assay. The results were normalized to the cell numbers of cells transfected with miR-NC. Transfection of miRNAs and siRNAs. The dsRNA mimicking mature human miR-3140-3p (MC17496) and nonspecific control miRNA (negative control #1) were purchased from Thermo Fisher Scientific. The SMARTpool siRNA for BRD2 (M-004935-02), BRD3 (M-004936-01), BRD4 (M-004937-02), CDK2 (M-003236- 04), EGFR (L-003114-00), nonspecific control siRNAs, and a set of 4 siRNA for BRD4 (MQ-004937-02) were from GE Healthcare (Buckinghamshire, UK). Each SMARTpool siRNA consists of 4 siRNA duplexes designed to target different regions of the same gene. mRNA sequence. miRNAs and siRNAs were transfected individually into cells at the indicated concentrations using Lipofectamine RNAiMAX (Thermo Fisher Scientific) according to the manufacturer’s instructions. In vitro cell growth assay. The in vitro cell growth assay was carried out as described previously . Briey fl , cell viability was measured using the WST-8 assay (Cell Counting Kit-8; Dojindo, Kumamoto, Japan) at the indi- cated number of days aer p ft lating, and the results were normalized to day 1 values. Each assay was carried out in triplicate. Treatment with EGFR-TKI. Gefitinib and Erlotinib were purchased from Cayman Chemicals (Michigan, USA) and were resuspended in dimethyl sulfoxide (DMSO) to a final concentration of 100 mM and 500 mM for long-term storage, respectively. Cells were treated with the medium containing Gefitinib (0.001, 0.01, 0.1, 1, 5, 10, 25 µ M), Erlotinib (0.001, 0.01, 0.1, 1, 5, 10, 25 µ M), or DMSO for 72 hours. Cell viability was assessed by WST-8 assay as described above. Gene expression array analysis. Gene expression array analysis was carried out as previously described . Briefly, Panc1, MIAPaCa2, and MDA-MB-231 cells were transfected with 10 nmol/L of miRNA (miR-NC, or SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 10 www.nature.com/scientificreports/ miR-3140). RNA was extracted 72 hours after transfection. The data were analyzed by GeneSpring software (Agilent Technologies, Japan). Western blotting. Western blotting was performed according to previously reported methods . Primary antibodies for Western blotting were used as follows: antibodies for BRD4 (#13440), BRD2 (#5848), MYC (#9402), NUT (#3625), Cyclin D2 (#3741 S), p-STAT3 (#9145 S), STAT3 (#9139 S), and CDK6 (#3136) were pur- chased from Cell Signaling Technology; antibodies for CDK2 (sc-163) and EGFR (sc-03-G) from Santa Cruz Biotechnology; antibodies for BRD3 (A302-368A) from Bethyl Laboratories (Montgomery, TX); β-actin (A5441) and Vinculin (V9131) from Sigma-Aldrich (Tokyo, Japan); and HaloTag (G921A) from Promega (Tokyo, Japan). Luciferase activity assay. Luciferase reporter plasmids were constructed by inserting the 3′UTR of CDK2, EGFR and CDS of BRD4 and BRD3 downstream of the luciferase gene within the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega, Madison, WI, USA). All site-specific mutations used the KOD -Plus- Mutagenesis Kit (TOYOBO, Osaka, Japan). Luciferase reporter plasmids or control plasmid (pmirGLO) were transfected into Panc1 cells using Lipofectamine 2000 (Thermo Fisher Scientific), and 10 nmol/L of miRNA (miR-NC or miR-3140) was also transfected 6 hours later. After 2 days, Firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega), and relative luciferase activity was calculated by normalizing the Firey fl luciferase reading with its corresponding internal Renilla luciferase control. Plasmid construction and transfection. BRD4 and BRD3 cDNA were purchased from Kazusa DNA Research Institute (Chiba, Japan) and were subcloned into the pFN28A HaloTag CMV-neo Flexi Vector (Promega, Madison, WI) to generate a mammalian expression vector. All site-specific mutations were gener - ated using the KOD-Plus-Mutagenesis Kit (TOYOBO). The BRD3 or BRD4 expression vector or the empty vec- tor were transfected into Panc1 and MIAPaCa2 cells using the Lipofectamine 2000 (Thermo Fisher Scientific), according to the manufacturer’s instructions. Establishment of JQ1-resistant cells. JQ1-resistant cells derived from Ty-82 were established by long-term incubation with gradually increasing JQ1 concentrations. JQ1 was purchased from APExBIO (Houston, TX). JQ1 was resuspended in DMSO to a final concentration of 10 mM for long-term storage. The cells were initially exposed to JQ1 at 0.1 µ mol/L in RPMI medium, then cultured in JQ1-free medium to confluence, and then exposed to JQ1 at a higher concentration. This cycle was repeated several times, until cells that were able to survive in RPMI medium including 2.5 μmol/L JQ1 were defined as Ty-82 JQ1-R cells. Real-time xCELLigence impedance analysis of cytotoxicity. Ty-82 cells and Ty-82 JQ1-R cells (3 × 10 ) were seeded in wells of the E-Plate 16 (ACEA Biosciences, San Diego, CA, USA). Approximately 48 hours later, these cells were treated with JQ1 (0.01, 0.05, 0.1, 0.5, 1, 5, 10 µ mol/L) (APExBIO) and DMSO. Cell-electrode impedance was monitored using the xCELLigence RTCA DP system (ACEA Biosciences) to pro- duce time-dependent cell response dynamic curves. Data were collected every 5 min aer t ft reatment with JQ1 for the first four hours, every 15 min for the next 20 hours, and then every 1 hour for an additional 48 hours. Dose response curves of JQ1 for Ty-82 cells and Ty-82 JQ1-R cells at 50 hours following treatment with JQ1 were con- structed. Cell indexes were normalized with the last time point before treatment with JQ1. In vivo tumor growth and miRNA administration. In vivo miRNA administration of miRNA was per- formed as previously described . Six-week-old female BALB/c nude mice were purchased from Oriental Bio Service, Japan. Briefly, a total of 5.0 × 10 cells in 100 μL of PBS were subcutaneously injected into the dorsal side of the mice. Aer t ft umor formation at day 7, a mixture of 1 nmol dsRNA (miR-NC or miR-3140) and 100 μL AteloGene (KOKEN, Tokyo, Japan) was administered around the tumor (miR-NC to the left dorsal side and miR-3140 to the right dorsal side of mice). miRNAs were administered on days 7, 11, 15, 18, and 21, and at 23 days ae ft r cell injection, mice were sacrificed and tumors were resected. Tumor volume was calculated using the following formula: (shortest diameter) × (longest diameter) × 0.5. All experimental protocols conducted on the mice were approved by the Tokyo Medical and Dental University Animal Care and Use Committee. Immunohistochemistry. Immunohistochemistry was performed as previously described . The resected tumors from xenograft mouse model were fixed in 10% formaldehyde in PBS for 24 h and stored in 70% ethanol and then embedded in paran. ffi e Th following primary antibodies were used for immunohistochemistry: an anti- body for BRD4 (HPA061646, 1:500) was purchased from Atlas Antibodies (Stockholm, Sweden), BRD3 (A302- 368A, 1:500) and CDK2 (IHC-00374, 1:500) antibodies were from Bethyl Laboratories, and EGFR (sc-03-G, 1:200) antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). BRD4, BRD3, EGFR, and CDK2 staining were scored semiquantitatively using histo-score (H-score) based on staining intensity and percentage of positive cells. Staining intensity was scored as follows: 0 = none, 1 = weak, 2 = moderate, or 3 = strong. H-score was cal- culated by multiplying the intensity of staining with percentage of cells stained in randomly chosen 3 fields from each specimen . Quantitative RT-PCR (qRT-PCR). Total RNA was extracted using TRIsure reagent (BIOLINE, London, UK) according to the manufacturer’s instructions. For miRNA, total RNA was reverse transcribed using the Taqman Reverse Transcription Kit followed by qRT-PCR performed using Custom Taqman miRNA Assays (Applied Biosystems, Foster City, CA). The miRNA expression was normalized to the internal control RNU6B. e f Th ollowing primers were used for the Taqman assay (Thermo Fisher Scientific): human miR-3140-3p (244524), RNU6B (001093). SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 11 www.nature.com/scientificreports/ Public datasets. To explore the generality of the miRNA expression and clinical features among pancreatic cancer, breast cancer and acute myeloid leukemia (AML), we examined the public datasets from TCGA (http:// cancergenome.nih.gov) retrieved on 24th July 2017. We used TCGA datasets on primary cases for pancreatic cancer, breast cancer and AML, and examined correlation of prognosis and expression of miR-3140. The datasets we used included mRNA-expression data on 174 samples of pancreatic cancer, 972 samples of breast cancer, 89 samples of AML and 1881 miRNAs, respectively. Statistical analysis. Student’s t-test, Paired t-test and log-rank test were performed using R software. P val- ues of < 0.05 were considered significant. Ethical Approval. All experimental protocols conducted on the mice were approved by the Tokyo Medical and Dental University Animal Care and Use Committee (0170036C). References 1. Belkina, A. C. & Denis, G. V. BET domain co-regulators in obesity, inflammation and cancer. Nat. Rev. Cancer 12, 465–477 (2012). 2. Fujisawa, T. & Filippakopoulos, P. Functions of bromodomain-containing proteins and their roles in homeostasis and cancer. Nat. Rev. Mol. Cell Biol. 18, 246–262 (2017). 3. Zuber, J. et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 478, 524–528 (2011). 4. Blobel, G. A., Kalota, A., Sanchez, P. V. & Carroll, M. Short hairpin RNA screen reveals bromodomain proteins as novel targets in acute myeloid leukemia. Cancer Cell 20, 287–288 (2011). 5. Delmore, J. E. et al. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 146, 904–917 (2011). 6. Dawson, M. et al. Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529–533 (2011). 7. Chapuy, B. et al. Discovery and Characterization of Super-Enhancer-Associated Dependencies in Diffuse Large B Cell Lymphoma. Cancer Cell 24, 777–790 (2013). 8. Shu, S. et al. Response and resistance to BET bromodomain inhibitors in triple-negative breast cancer. Nature 529, 413–417 (2016). 9. Sahai, V. et al. BET Bromodomain Inhibitors Block Growth of Pancreatic Cancer Cells in r Th ee-Dimensional Collagen. Mol. Cancer e Th r. 13, 1907–1917 (2014). 10. Lovén, J. et al. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell 153, 320–334 (2013). 11. Sun, B. et al. Synergistic activity of BET protein antagonist-based combinations in mantle cell lymphoma cells sensitive or resistant to ibrutinib. Blood 126, 1565–1574 (2015). 12. French, C. A. Demystified molecular pathology of NUT midline carcinomas. J. Clin. Pathol. 63, 492–6 (2010). 13. French, C. NUT midline carcinoma. Nat. Rev. Cancer 14, 149–50 (2014). 14. French, C. A. et al. NSD3-NUT fusion oncoprotein in NUT midline carcinoma: Implications for a novel oncogenic mechanism. Cancer Discov. 4, 929–941 (2014). 15. Alekseyenko, A. A. et al. Ectopic protein interactions within BRD4–chromatin complexes drive oncogenic megadomain formation in NUT midline carcinoma. Proc. Natl. Acad. Sci. 114, E4184–E4192 (2017). 16. Stathis, A. et al. Clinical response of carcinomas harboring the BRD4–NUT oncoprotein to the targeted bromodomain inhibitor OTX015/MK-8628. Cancer Discov. 6, 492–500 (2016). 17. Wang, R. & You, J. Mechanistic analysis of the role of bromodomain-containing protein 4 (BRD4) in BRD4-NUT oncoprotein- induced transcriptional activation. J. Biol. Chem. 290, 2744–2758 (2015). 18. Reynoird, N. et al. Oncogenesis by sequestration of CBP/p300 in transcriptionally inactive hyperacetylated chromatin domains. EMBO J. 29, 2943–2952 (2010). 19. Wang, R. et al. Activation of SOX2 expression by BRD4-NUT oncogenic fusion drives neoplastic transformation in NUT midline carcinoma. Cancer Res. 74, 3332–3343 (2014). 20. Mazur, P. K. et al. Combined inhibition of BET family proteins and histone deacetylases as a potential epigenetics-based therapy for pancreatic ductal adenocarcinoma. Nat. Med. 21, 1163–71 (2015). 21. Ott, C. J. et al. BET bromodomain inhibition targets both c-Myc and IL7R in high-risk acute lymphoblastic leukemia. Blood 120, 2843–2852 (2012). 22. Filippakopoulos, P. et al. Selective inhibition of BET bromodomains. Nature 468, 1067–73 (2010). 23. Rathert, P. et al. Transcriptional plasticity promotes primary and acquired resistance to BET inhibition. Nature 525, 543–547 (2015). 24. Berthon, C. et al. Bromodomain inhibitor OTX015 in patients with acute leukaemia: A dose-escalation, phase 1 study. Lancet Haematol. 3, e186–e195 (2016). 25. Amorim, S. et al. Bromodomain inhibitor OTX015 in patients with lymphoma or multiple myeloma: A dose-escalation, open-label, pharmacokinetic, phase 1 study. Lancet Haematol. 3, e196–e204 (2016). 26. Wang, C. Y. & Filippakopoulos, P. Beating the odds: BETs in disease. Trends Biochem. Sci. 40, 468–479 (2015). 27. Lin, S. & Gregory, R. I. MicroRNA biogenesis pathways in cancer. Nat. Rev. Cancer 15, 321–333 (2015). 28. Iorio, M. V. & Croce, C. M. MicroRNA dysregulation in cancer: Diagnostics, monitoring and therapeutics. A comprehensive review. EMBO Mol. Med. 4, 143–159 (2012). 29. Forman, J. J., Legesse-Miller, A. & Coller, H. A. A search for conserved sequences in coding regions reveals that the let-7 microRNA targets Dicer within its coding sequence. Proc. Natl. Acad. Sci. USA 105, 14879–84 (2008). 30. Hausser, J., Syed, A. P., Bilen, B. & Zavolan, M. Analysis of CDS-located miRNA target sites suggests that they can effectively inhibit translation. Genome Res. 23, 604–615 (2013). 31. Ito, Y. et al. Identification of targets of tumor suppressor microRNA-34a using a reporter library system. Proc. Natl. Acad. Sci. 114, 3927–3932 (2017). 32. Akdemir, B., Nakajima, Y., Inazawa, J. & Inoue, J. miR-432 Induces NRF2 Stabilization by Directly Targeting KEAP1. Mol. Cancer Res. molcanres. 0232, 2017, https://doi.org/10.1158/1541-7786.MCR-17-0232 (2017). 33. Kong, Y. W., Ferland-McCollough, D., Jackson, T. J. & Bushell, M. MicroRNAs in cancer management. Lancet Oncol. 13, e249–e258 (2012). 34. Hiramoto, H. et al. miR-509-5p and miR-1243 increase the sensitivity to gemcitabine by inhibiting epithelial-mesenchymal transition in pancreatic cancer. Sci. Rep. 7, 4002 (2017). 35. Yang, H. et al. microRNA-342-5p and miR-608 inhibit colon cancer tumorigenesis by targeting NAA10. Oncotarget 7, 2709–2720 (2016). 36. Fujiwara, N. et al. miR-634 activates the mitochondrial apoptosis pathway and enhances chemotherapy-induced cytotoxicity. Cancer Res. 75, 3890–3901 (2015). 37. Warnecke-Eberz, U., Chon, S. H., Hölscher, A. H., Drebber, U. & Bollschweiler, E. Exosomal onco-miRs from serum of patients with adenocarcinoma of the esophagus: comparison of miRNA profiles of exosomes and matching tumor. Tumor Biol. 36, 4643–4653 (2015). SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 12 www.nature.com/scientificreports/ 38. Tréhoux, S. et al. Micro-RNAs miR-29a and miR-330-5p function as tumor suppressors by targeting the MUC1 mucin in pancreatic cancer cells. Biochim. Biophys. Acta - Mol. Cell Res. 1853, 2392–2403 (2015). 39. Kozomara, A. & Griffiths-Jones, S. MiRBase: Annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 42, 68–73 (2014). 40. Du, J. et al. Critical role of CDK2 for melanoma growth linked to its melanocyte-specific transcriptional regulation by MITF. Cancer Cell 6, 565–576 (2004). 41. Asghar, U., Witkiewicz, A. K., Turner, N. C. & Knudsen, E. S. The history and future of targeting cyclin-dependent kinases in cancer therapy. Nat. Rev. Drug Discov. 14, 130–146 (2015). 42. Gazdar, A. F. & Minna, J. D. Deregulated EGFR signaling during lung cancer progression: Mutations, amplicons, and autocrine loops. Cancer Prev. Res. 1, 156–160 (2008). 43. Sharma, S. V., Bell, D. W., Settleman, J. & Haber, D. A. Epidermal growth factor receptor mutations in lung cancer. Nat. Rev. Cancer 7, 169–181 (2007). 44. Rennie, W. et al. STarMirDB: A database of microRNA binding sites. RNA Biol. 13, 554–60 (2016). 45. Grayson, aR. et al. MYC, a downstream target of BRD-NUT, is necessary and sufficient for the blockade of differentiation in NUT midline carcinoma. Oncogene 33, 1–7 (2013). 46. Kuzume, T. et al. Establishment and characterization of a thymic carcinoma cell line (Ty-82) carrying t(15;19)(q15; p13) chromosome abnormality. Int. J. cancer J. Int. du cancer 50, 259–264 (1992). 47. Ke, E. E. & Wu, Y. L. EGFR as a Pharmacological Target in EGFR-Mutant Non-Small-Cell Lung Cancer: Where Do We Stand Now? Trends Pharmacol. Sci. 37, 887–903 (2016). 48. Politi, K., Ayeni, D. & Lynch, T. The Next Wave of EGFR Tyrosine Kinase Inhibitors Enter the Clinic. Cancer Cell 27, 751–753 (2015). 49. Otto, T. & Sicinski, P. Cell cycle proteins as promising targets in cancer therapy. Nat. Rev. Cancer 17, 93–115 (2017). 50. French, C. A. et al. BRD4-NUT Fusion Oncogene: A Novel Mechanism in Aggressive Carcinoma Advances in Brief BRD4-NUT Fusion Oncogene: A Novel Mechanism in Aggressive Carcinoma 1. Cancer Res. 63, 304–307 (2003). 51. French, C. A. et al. BRD-NUT oncoproteins: a family of closely related nuclear proteins that block epithelial differentiation and maintain the growth of carcinoma cells. Oncogene 27, 2237–2242 (2008). 52. Wang, Y. H. et al. BRD4 promotes pancreatic ductal adenocarcinoma cell proliferation and enhances gemcitabine resistance. Oncol. Rep. 33, 1699–1706 (2015). 53. Shimada, Y., Imamura, M., Wagata, T., Yamaguchi, N. & Tobe, T. Characterization of 21 newly established esophageal cancer cell lines. Cancer 69, 277–284 (1992). 54. Endo, H. et al. Potential of tumor-suppressive miR-596 targeting LGALS3BP as a therapeutic agent in oral cancer. Carcinogenesis 34, 560–569 (2013). 55. Lessey, B. A. et al. Immunohistochemical analysis of human uterine estrogen and progesterone receptors throughout the menstrual cycle. J. Clin. Endocrinol. Metab. 67, 334–40 (1988). Acknowledgements We thank Ayako Takahashi and Rumi Mori for technical assistance. This work was supported by KAKENHI (15H05908, 16K14630, 15K18401, 15K19040) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), and partially supported by the Project for Cancer Research And Therapeutic Evolution (P-CREATE) from Japan Agency for Medical Research and development, AMED. This study was also partly supported by Nanken-Kyoten, TMDU. Author Contributions E. Tonouchi, and Y. Gen were involved in research design, performed the experiments, analyzed data and wrote the manuscript. T. Muramatsu and H. Hiramoto were involved in research design, performed the expriments, and analyzed data. K. Tanimoto contributed to TCGA data analysis. J. Inoue contributed materials. J.Inazawa was involved in research design, wrote the manuscript, and study supervision. Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-018-22767-y. Competing Interests: The authors declare no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2018 SCIenTIfIC REPO R ts | (2018) 8:4482 | DOI:10.1038/s41598-018-22767-y 13

Journal

Scientific ReportsSpringer Journals

Published: Mar 14, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off