A Mir-29B Byproduct Sequence Exhibits Potent Tumor-Suppressive

Published OnlineFirst March 15, 2018; DOI: 10.1158/1535-7163.MCT-17-0850

Small Molecule Therapeutics Molecular Cancer Therapeutics A miR-29b Byproduct Sequence Exhibits Potent Tumor-Suppressive Activities via Inhibition of NF-kB Signaling in KRAS-Mutant Colon Cancer Cells Akira Inoue1, Tsunekazu Mizushima1, Xin Wu2, Daisuke Okuzaki3, Nanami Kambara2, Sho Ishikawa2, Jiaqi Wang2, Yamin Qian2, Haruka Hirose2, Yuhki Yokoyama2, Ryo Ikeshima1, Masayuki Hiraki1, Norikatsu Miyoshi1, Hidekazu Takahashi1, Naotsugu Haraguchi1, Taishi Hata1, Chu Matsuda1, Yuichiro Doki1, Masaki Mori1, and Hirofumi Yamamoto1,2

Abstract

We previously demonstrated that miR-29b-3p is a hopeful effect and named this byproduct miRNA sequence "MIRTX." miRNA-based therapy against colorectal cancer. In this study, we MIRTX directly targeted the 30-UTR of CXCR2 and PIK3R1 mRNA aimed to clarify a value of miR-29b-1-5p as a next-generation and suppressed the NF-kB signaling pathway in KRAS-mutated treatment, especially for KRAS-mutant colorectal cancer. RT-PCR colorectal cancer cells. MIRTX induced apoptosis in DLD1 with assay showed that the expression of miR-29b-3p was high, and its downregulation of antiapoptotic BCL2, BCL-xL, and MCL1 and partner strand, miR-29b-1-5p, level was only negligible in clinical upregulation of cleaved caspase-3 and cleaved PARP. In mouse colorectal cancer samples. Mimic-miR-29b-1-5p significantly xenograft models, systemic administration of MIRTX using a inhibited proliferation of KRAS-mutant colorectal cancer cell lines super carbonate apatite as a delivery vehicle significantly inhibited DLD1 and SW480 and KRAS wild-type HT29 cells. Proliferative tumor growth of DLD1 and HT29 cells without any particular activity was further examined by either miR-29b-1-5p strand or its toxicities. In conclusion, these findings indicate that inhibition opposite complementary sequence because miR-29b-1-5p is a of NF-kB signaling by this novel miRNA-based therapeutic passenger miRNA and may have no physiologic function. We could be a promising treatment against refractory KRAS- found that completely opposite complementary strand to miR- mutant colorectal cancer and KRAS wild-type colorectal cancer. 29b-1-5p, but not miR-29b-1-5p, possessed a potent antitumor Mol Cancer Ther; 17(5); 977–87. 2018 AACR.

Introduction and new therapeutic strategies are urgently needed to improve the prognosis of this disease. Colorectal cancer is the third leading cause of cancer-related Recent advances in our understanding of the speciﬁc signaling death in the United States, with an estimated 134,490 new cases pathways in cancer cells have offered novel targeted therapies for and 49,190 deaths in 2016 (1). Although progress has been made patients with colorectal cancer. Indeed, previous studies have by improving treatment approaches, including surgery, chemo- demonstrated that EGFR mAbs signiﬁcantly improve the prog- therapy, and irradiation, patients with metastatic colorectal cancer nosis of patients with wild-type KRAS metastatic colorectal cancer (mCRC) still have a very poor prognosis (2). Therefore, a better (3–5). However, 40% of colorectal cancers harbor KRAS muta- understanding of the molecular mechanisms of colorectal cancer tions (6), and treating these patients with mutant KRAS mCRC remains one of the biggest challenges in oncology. One possible way to overcome this issue is to develop molecular targeted drugs 1Department of Surgery, Gastroenterological Surgery, Graduate School of that could inhibit KRAS downstream effectors, such as MEK1/2. Medicine, Osaka University, Osaka, Japan. 2Department of Molecular Pathology, For example, the MEK1/2 inhibitors PD-0325901, BAY 86-9766, Division of Health Sciences, Graduate School of Medicine, Osaka University, and AS703026 have been evaluated in clinical trials (7, 8), but 3 Osaka, Japan. Genome Information Research Center, Research Institute for the trials were discontinued, or an effective dose could not be Microbial Diseases, Osaka University, Osaka, Japan. achieved because of drug toxicity. Note: Supplementary data for this article are available at Molecular Cancer Recent evidence indicates that the NF-kB transcription factor is Therapeutics Online (http://mct.aacrjournals.org/). constitutively activated in KRAS-mutant colorectal cancer (9, 10). Corresponding Author: Hirofumi Yamamoto, Department of Molecular Pathol- Therefore, targeting NF-kB signaling pathway represents a prom- ogy, Division of Health Sciences, Graduate School of Medicine, Osaka University, ising strategy against KRAS-mutant colorectal cancer, although Yamadaoka 1-7, Suita City, Osaka 565-0871, Japan. Phone/Fax: 816-6879-2591; this has not been successful so far in colorectal cancer. The NF-kB E-mail: [email protected] signaling pathway is involved in tumor development, such as doi: 10.1158/1535-7163.MCT-17-0850 tumor cell growth, antiapoptosis, metastasis, angiogenesis, and 2018 American Association for Cancer Research. resistance against chemotherapeutics (9, 11). Moreover, NF-kB

www.aacrjournals.org 977

Inoue et al.

seems to play a crucial role in inflammation and oncogenesis samples were stored at 80C in RNAlater until RNA extraction. through regulating multiple molecules (9). Several molecules are Written informed consent was obtained from all patients, in involved in NF-kB signaling pathway. It is reported that PIK3R1, accordance with the guidelines approved by the Institutional the major downstream signaling molecule of AKT, is an intrinsic Research Board. This study was conducted under the supervision activator of NF-kB (12, 13) and that expression levels of IL8 and its of the Ethics Board of Osaka University Hospital. receptor CXCR2 are associated with tumor proliferation and progression through a signaling network involving NF-kB, Transient transfection of miRNA besides PI3K/AKT and MAPK (14–17). Mature MIRTX sense (50-UCUAAACCACCAUAUGAAACCA- Nucleic acid (NA)–based medicine, which includes antisense-, GC-30) and antisense (50-GCUGGUUUCAUAUGGUGGUU- siRNA-, and miRNA-based therapies, shows potential as a next- UAGA-30), mimic-miR-34a sense (50-UGGCAGUGUCUUAG- generation treatment. miRNAs are single-stranded, noncoding CUGGUUGU-30) and antisense (50-ACAACCAGCUAAGACACU- RNAs of 20–22 nucleotides, being involved in various biological GCCA-30), and mimic-miR-negative control sense (50-AUCC- processes and act to posttranscriptionally or transcriptionally GCGCGAUAGUACGUA-30) and antisense (50-UACGUACUA- regulate the expression of multiple target genes by binding to UCGCGCGGAU-30) were obtained from Gene Design Inc. Mod- 0 complementary sequences, mainly in the 3 -untranslated regions ification of the one-strand sequence to inactivation was per- 0 (3 -UTR; ref. 18). formed by Gene Design Inc. or Thermo Fisher Scientific. Cells In the miR-29 family, we and other investigators have shown were transfected with miRNAs at the concentration of 20 to 40 that miR-29b-3p is a tumor suppressor and a prognostic factor in nmol/L by Lipofectamine RNAiMAX (Invitrogen) according to the various human malignancies, including colorectal cancer, lung protocol provided by the manufacturer. cancer, lymphoma, leukemia, and other types of human malignancies, as it regulates cell proliferation, differentiation, apoptosis, RNA isolation – fi migration, and invasion (19 22). These ndings suggest that Total RNA, including miRNA, was isolated from tissue samples miR-29b-3p is a promising next-generation treatment for and cell lines using the miRNeasy Kit (Applied Biosystems) cancers. On the other hand, miR-29b-1-5p, the partner strand of according to the manufacturer's protocol. Total RNA concentra- miR-29b-3p in the primary miRNA duplex was formerly known as tion and purity were reassessed with a NanoDrop ND-1000 miR-29b-1 , and it is believed to be preferentially degraded (23, spectrophotometer (NanoDrop Technologies). 24); however, it was recently recognized that miRNA species may also access Argonaute (Ago) complexes and regulate targets in miRNA expression certain conditions (25, 26). Indeed, it is recently reported that We measured miRNA expression levels in tissues and cells with miR-29b-1-5p exerts a function to suppress the aggressive triple- the TaqMan miRNA Assays (Applied Biosystems). The reverse negative breast cancer cells and bladder cancer cell line (27, 28). transcription reaction was performed with the TaqMan MicroRNA In this study, we examined miR-29b-1-5p expression in com- RT Kit (Applied Biosystems). Quantitative real-time PCR was parison with that of miR-29b-3p in colorectal cancer tissues. We performed with the 7900HT Sequence Detection System (Applied also investigated the antitumor effect in vitro by the nucleic acid Biosystems). Amplification data were normalized to endogenous mimics for miR-29b-1-5p, especially in KRAS-mutant colon can- RNU6B expression. Relative expression was quantified with the cer cells. As a result, we found that miR-29b-1-5p sequence had no DDC 2 method. tumor-inhibitory effects, whereas its byproduct, a complementary t nucleic acid sequence, exhibited a potent tumor-inhibitory activity in vitro and in xenograft mouse models with no drug toxicity. Cell proliferation assay – 4 This artificial miRNA sequence may be a hopeful treatment to the Cells were seeded at a density of 5 7 10 cells per well in patients with KRAS-mutant refractory colorectal cancer. 24-well dishes and cultured for 24 to 72 hours to determine proliferation. Cell counting was performed with an automatic hematology analyzer (Celltac; Nihon Kohden). Cellular prolifer- Materials and Methods ation was also evaluated by WST-8 assay using Cell Counting Kit-8 Cell lines and cell culture (Dojindo Molecular Technologies, Inc.). Human colorectal cancer cell lines HT29 and DLD1 were obtained from the ATCC in 2001. SW480 was obtained from Flow cytometry analysis ATCC in 2010. Stocks were prepared after passage 2 and stored in To examine the apoptosis, flow cytometry analysis of Annexin V liquid nitrogen. All experiments were performed with cells of was performed as described previously (30). Apoptotic cells were passage of <8. These cell lines were authenticated by morphologic stained by using Alexa Fluor 488 Annexin V/Dead Cell Apoptosis inspection, short tandem repeat profiling, and mycoplasma test- Kit (Thermo Fisher Scientific) and counted by flow cytometry ing. Mycoplasma testing was also performed by the authors in using the SH800Z Cell Sorter (Sony Biotechnology Inc.). 2015. Cells were cultured in DMEM containing 10% FBS at 37C fi in a humidi ed incubator with 5% CO2. TUNEL assay Cells were fixed with freshly prepared 10% buffered neutral Clinical tissue samples formalin overnight at room temperature. Apoptotic DNA frag- Colorectal cancer samples were collected from 40 patients who mentation within the cells was analyzed with an in situ DeadEnd had surgery at Osaka University Hospital (Osaka, Japan) and Fluorometric TUNEL System Assay Kit (Promega), according to related hospital between 2012 and 2013. These patients included the manufacturer's instructions. Hoechst was used for nuclear 3 stage I, 15 stage II, 11 stage III, and 11 stage IV patients according staining. The green fluorescence of apoptotic cells was observed to the AJCC TNM classification (7th edition; ref. 29). The tumor with a BZ-X700 fluorescence microscopy (KEYENCE).

978 Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics

A Novel Small RNA against KRAS-Mutant Colon Cancer

Western blot analysis fected with 100 ng of a pmirGLO construct, which contained The protein samples were subject to SDS-PAGE, transferred either the 30-UTR of PIK3R1 or CXCR2,10ngofRenilla luciferase onto PVDF (polyvinylidene difluoride) membranes, and incu- reporter vector (pRL-CMV, Promega), and 5 pmol of either miR- bated with mAbs. Antibodies against MCL1, BCL2, BCL-xL, BAX, NC or MIRTX. Forty-eight hours after cotransfection, cells were PIK3R1, phosphorylated AKT, phosphorylated mTOR, NF-kB assayed for both firefly and Renilla luciferase with the Dual- (p65), and phosphorylated NF-kB (phosphorylated p65) were Luciferase Reporter Assay System (Promega), according to the obtained from Cell Signaling Technology. Antibody against manufacturer's protocol. All transfection experiments were con- CXCR2 (ab14935) was obtained from Abcam. Antibodies against ducted in triplicate. b-actin (ACTB) were obtained from Sigma-Aldrich. Antibody against survivin was obtained from Novus. After incubation with In vivo tumor growth assay secondary antibodies, signals were detected with the ECL Detec- Cells were mixed with Matrigel (BD Biosciences) and medium 6 tion System (GE Healthcare). at a 1:10 ratio (v:v). Cells (2 10 ) in 200 mL of medium/Matrigel solution were injected subcutaneously into both sides of the lower NF-kB luciferase reporter assay back region of female nude mice (NIHON CLEA). Using super All cells were seeded onto 96-well plates and transfected with carbonate apatite (sCA) as a vehicle (31, 32), formulated miRNA NF-kB luciferase reporter vector (pGL4.32[luc2P/NF-kB-RE/ (40% conjugation) was administered intravenously (40 mg/injec- Hygro]; Promega) using Lipofectamine 2000 (Invitrogen) in tion); all injections were delivered via the tail vein, after tumor 3 OptiMEM-reduced serum medium (GIBCO, Invitrogen). The volumes reached 50 mm . Mice were treated eight times with transfection efficiency was evaluated with a Renilla luciferase formulated MIRTX or miR-NC, three times a week. Mature MIRTX reporter vector (pRL-CMV, Promega). NF-kBtranscriptional and miR-negative control for in vivo use grade were obtained activities were determined in dual luciferase reporter assays. from Gene Design Inc. Tumor volumes were determined as Forty-eight hours after cotransfection with both reporter vectors described earlier (33, 34). All animal experiments were per- and the target miRNA, cells were collected with the reporter formed in accordance with currently prescribed guidelines, and lysis buffer from the Luciferase Assay System (Promega). The experimental protocols (ID; DOUI 25-062) were approved by luciferase-mediated luminescence was measured with the Dual- an Institutional Animal Care and Use Committee in Osaka Luciferase Reporter Assay System (Promega) according to the University (Osaka, Japan). manufacturer's protocol. All transfection experiments were conducted in triplicate. IHC Formalin-fixed, paraffin-embedded tissues were prepared and m fi Prediction of MIRTX target genes sectioned for 4 m. These were deparaf nized with xylene, and First, to predict MIRTX-dependent repressed gene sets via then rehydrated in graded alcohols. Immunostaining was per- complementary base-pairing between the 7-mer seed (MIRTX; formed using antibody with VECTASTAIN Elite ABC Kit (Vector 50-CUAAACC-30) and the last exon containing the 30-UTR Laboratories), according to the manufacturer's protocol. Anti- a sequence, the last exons containing the 30-UTRs of human tran- CXCR2 antibody (ab14935) and anti-PI3 kinase p85 antibody scripts were obtained as fasta files from the Ensembl FTP Server [M253] (ab86714), and anti-BCL2 antibody [E17] (ab32124) – k (version hg19). A total of 1,043 ensembl transcripts were found to were obtained from Abcam. Anti NF- B (p65) antibody and k have a perfectly MIRTX-targeted sequence (50-GGTTTAG-30), at antiphosphorylated NF- B (phosphorylated p65) antibody were least in the 30-UTR, by using standard linux tools such as grep. obtained from Cell Signaling Technology. Statistical analysis pmirGLO reporter plasmid construction and luciferase reporter Statistical analyses were performed with the JMP10 program assay (SAS Institute) and GraphPad Prism ver. 6.00 for Windows The 30-UTR of human PIK3R1 mRNA, including the MIRTX- (GraphPad Software). Each experiment was performed at least binding site, was amplified by RT-PCR. The primer sequences three times. Data are expressed as mean SD. Mean values were were forward-GCTCGCTAGCCTCGAAACAGCGCTCACCTTTG- compared using the Student t test or one-way ANOVA analysis. TTT and reverse-ATGCCTGCAGGTCGACTGAACATACCAGCT- Expression levels of miRNAs in colorectal cancer tissue samples TTTGCAT. The amplified product (431 bp) was subcloned and were analyzed with the Wilcoxon signed rank test. Sequential cell ligated into the Xho I and Sal I sites in the pmirGLO Dual- proliferation assays and tumor growth assays were analyzed with Luciferase miRNA Target Expression Vector (Promega) using the two-way ANOVAs for repeated measures. P values <0.05 were In-Fusion HD Cloning Kit (Clontech). The 30-UTR of human considered statistically significant. CXCR2 mRNA, including the MIRTX-binding site, was amplified by RT-PCR. The primer sequences were forward-GCTCGCTAG- CCTCGACCCATTGTGGTCACAGGAA and reverse-ATGCCTG- Results CAGGTCGACACAATCTCGGCTCACTGC. As before, the ampli- Effect of mimic-miR-29b-1-5p on cell proliferation fied product (431 bp) was subcloned and ligated into the Xho I In vitro proliferation assays were performed with the mimic- and Sal I sites in the pmirGLO Dual-Luciferase miRNA Target miR-29b-1-5p, which was synthesized as the double-strand RNA Expression Vector (Promega) using the In-Fusion HD Cloning Kit with miR-29b-1-5p and its complementary byproduct sequence (Clontech). (Fig. 1). The results indicated that mimic-miR-29b-1-5p signifi- Luciferase assays were conducted with 1 104 DLD1 and cantly suppressed cell proliferation of KRAS-mutated DLD1 and SW480 cells per well in 96-well plates. Transfections were per- SW480 cells, and KRAS wild-type HT29 colorectal cancer cell lines formed with Lipofectamine 2000 (Invitrogen) in OptiMEM- compared with parent or negative control miRNA (Fig. 2A; reduced serum medium (GIBCO, Invitrogen). Cells were trans- Supplementary Fig. S1A and S1B). Notably, the growth-inhibitory

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 979

Inoue et al.

Mimic-miR-29b-1-5p

Complementary sequence 3 - CGACCAAAGUAUACCACCAAAUCU -5 (MIRTX) Figure 1. miR-29b-1-5p Sequence of hsa-miR-29b. The hsa- 5-GCUGGUUUCAUAUGGUGGUUUAGA -3 miR-29b precursor is a small RNA molecule in the shape of a stem loop or hairpin. Each arm of the hairpin is processed into mature miR-29b-3p (guide strand) and miR-29b-1-5p Pri-miR-29b hp://www.mirbase.org/ (passenger strand). MIRTX is a novel small RNA sequence that is a completely opposite complementary sequence to miR-29b-1-5p. miR-29b-1-5p 5- GCUGGUUUCAUAUGGUGGUUUAGA -3 miR-29b-3p 3- UUGUGACUAAAGUUUACCACGAU -5

effect by mimic-miR-29b-1-5p was equivalent to or more effective variety of miR-29b-3p expression in clinical colorectal cancer than miRNA-34a (Fig. 2A), which is well known for its antitumor samples, whereas the expression of miR-29b-1-5p was only neg- effects (35). ligible in all samples tested (P < 0.01; Fig. 2B). This finding indicates that miR-29b-1-5p is unstable in colorectal cancer Expression of miR-29b-1-5p and miR-29-3p in clinical tissues, and it is uncertain that the degradable passenger sequence colorectal cancer samples can inhibit cell growth. To examine whether the miR-29b-1-5p Despite the outstanding growth-inhibitory activity, miR-29b-1- strand is definitely functional or not, we used a modified miRNA 5p, formerly described as miR-29b, is generally considered as a mimic from Gene Design Inc. in which activity of the comple- degradable passenger miRNA in many cell types (23, 24). There- mentary sequence is chemically impaired; thus, the function fore, we examined miR-29b-1-5p expression in comparison with derived from single strand of miR-29b-1-5p can be evaluated. that of the guide strand, miR-29-3p. RT-PCR assay showed a We also prepared another miRNA mimic, in which miR-29b-1-5p

Figure 2. Effects of mimic-miR-29b-1-5p on cell proliferation and expression level of hsa-miR-29b in clinical colorectal cancer (CRC) samples. A, Overexpression of mimic- miR-29b-1-5p significantly suppressed cell proliferation in DLD1, SW480, and HT29 cells compared with parent or negative control mimic-miR-NC and was equivalent to or more effective than mimic-miR-34a, which is well known for its antitumor effects. B, Expression level of miR-29b-3p and miR-29b-1-5p in clinical colorectal cancer tumor samples (n ¼ 40). C, Inhibitory effects on proliferation of KRAS-mutant colorectal cancer cell lines. A complementary strand, but not miR-29b-1-5p, significantly inhibited cell proliferation. Modification of the one-strand sequence to inactivation was performed by Gene Design Inc. All data, mean SD. , P < 0.01.

980 Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics

A Novel Small RNA against KRAS-Mutant Colon Cancer

sequence is in turn modified to inactivation. As a result, we that MIRTX suppressed the protein expression of antiapoptotic found that miR-29b-1-5p sequence had no growth-inhibitory proteins MCL1, BCL2, and BCL-xL, whereas MIRTX increased the effect in DLD1 and SW480 and that an opposite complemen- expression of cleaved PARP and cleaved caspase-3 compared with tary strand to miR-29b-1-5p exhibited an obvious growth negative control miRNA treatment (Fig. 3C). inhibition(Fig.2C).AnalternativesystemofmirVanamiRNA mimic from Thermo Fisher Scientificfurtherconfirmed the The target genes of MIRTX result (Supplementary Fig. S1C). Therefore, we hereafter des- The seed sequence (2–8 position from 50-end) is thought to ignate this antitumor byproduct sequence "MIRTX" meaning a play a pivotal role in regulating gene expression by binding the 30- therapeutic miRNA (Fig. 1) and explored the underlying mech- UTR of the target mRNA. In silico whole-genome sequence analysis anism for its potent antitumor effect. demonstrated that 1,043 genes possessed a binding site to the MIRTX seed sequence (CCAAAUC, Supplementary Fig. S2A). MIRTX induces apoptosis cDNA microarray data (GEO accession number; GSE83634) A drastic decrease in KRAS-mutated DLD1 cells with treatment revealed that 74 of the 1,043 genes were downregulated at of MIRTX may be attributable to apoptosis. Terminal deoxynu- <0.66-fold in MIRTX- transfected KRAS-mutated DLD1 cells cleotidyl transferase–mediated dUTP nick end labeling (TUNEL) compared with negative control miRNA-transfected cells (Sup- assay showed that a number of TUNEL-positive fluorescence cells plementary Fig. S2B; Supplementary Table S1). Of them, we appeared with treatment of MIRTX compared with control cul- focused on CXCR2 encoding a G-protein–coupled transmem- tures (Fig. 3A). Flow cytometry analysis of Annexin V showed that brane chemokine receptor, and PIK3R1 [PI3K, regulatory subunit treatment of MIRTX significantly increased apoptotic fraction at 1 (alpha)] as the targets of MIRTX because they regulate apoptosis 24 and 48 hours (P < 0.01, Fig. 3B). Western blot analyses showed and cell survival via NF-kB/survivin and AKT/mTOR signaling

A DLD1 KRASG13D B Parent miR-NC MIRTX

AF-488

20 * 100 * NC NC 15 80 MIRTX MIRTX 60 10 40 5 20 0 0 Annexin-posive cells (%) 241 h Annexin-posive cells (%) 481 h *P < 0.01 C

Cleaved PARP

Cleaved caspase-3

ACTB

Figure 3. Proapoptotic effects of MIRTX in KRAS-mutant colorectal cancer model. A, TUNEL assay showed that the number of apoptotic cells was higher in MIRTX-treated cells than in parent or miR-NC–treated cells. B, Flow cytometry analysis of Annexin V was performed in parental DLD1 cells, miR-NC–treated cells, and MIRTX-treated cells. The number of apoptotic cells was quantiﬁed in triplicate. C, Western blot results showed the protein expression of MCL1, BCL2, BCL-xL, BAX, cleaved PARP, and cleaved caspase-3 in either negative control miR or MIRTX-treated DLD1 cells in 24, 48, and 72 hours. MIRTX signiﬁcantly suppressed the expression of the antiapoptotic proteins MCL1, BCL2, and BCL-xL compared with negative control miRNA. b-Actin (ACTB) was the loading control. All data, mean SD. , P < 0.01.

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 981

Inoue et al.

Figure 4. MIRTX directly binds and targets CXCR2 and PIK3R1 mRNA in KRAS-mutant colorectal cancer cell lines. A and C, Schematic illustration shows that CXCR2 and PIK3R1 possess a putative binding site on the 30-UTR of its mRNA for the seed sequence of MIRTX. B and D, The luciferase activity of a plasmid that included the 30-UTR region of CXCR2 and PIK3R1 in DLD1 and SW480 cells treated with miR-NC or MIRTX. All data, mean SD. , P < 0.01.

pathway, respectively (Supplementary Fig. S3A). MIRTX overex- the luciferase reporter assays showed MIRTX signiﬁcantly suppression inhibited the luciferase reporter activity of a plasmid that pressed NF-kB transcriptional activities in DLD1 and SW480 included the 30-UTR region of CXCR2 (Fig. 4A and B) and PIK3R1 cells, but not in HT29 cells (Fig. 5B). MIRTX suppressed PIK3R1 (Fig. 4C and D) in KRAS-mutated DLD1 and SW480, indicating protein expression; however, the PI3K downstream signaling that MIRTX directly binds to the 30-UTR of CXCR2 mRNA and molecule, phosphorylated AKT, was rather upregulated possi- PIK3R1 mRNA in these KRAS-mutated colorectal cancer cells. blybyacompensatorymechanism and phosphorylated mTOR was not substantially changed (Fig. 5C). In addition, MIRTX MIRTX inhibits CXCR2/NF-kB/survivin pathway but not PI3K/ did not suppress SRE or AP1 transcriptional activities, which AKT/mTOR pathway increase in response to activation of the MAPK/ERK or MAPK/ MIRTX suppressed CXCR2 protein expression, leading to inhi- JNK signaling pathway in colorectal cancer cell lines (Supple- bition of downstream molecules such as phosphorylation of p65 mentary Fig. S4). subunit of NF-kB that is thought to be important for transcriptional activity of NF-kB (16, 36, 37), and antiapoptotic survivin in Systemic administration of formulated MIRTX inhibits tumor KRAS-mutated DLD1 and SW480 (Fig. 5A). On the other hand, growth in vivo decrease in expression of the CXCR2 protein and phosphorylated Because MIRTX is an artiﬁcial byproduct sequence and it is p65 was not evident in KRAS wild-type HT29 cells. Concordantly, not present in the mouse body, we evaluated the toxicity of

982 Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics

A Novel Small RNA against KRAS-Mutant Colon Cancer

Figure 5. MIRTX suppresses NF-kB signaling activity in KRAS-mutant colorectal cancer cell lines. A, Western blot analyses of CXCR2, NF-kB (p65), phosphorylated NF-kB (p-p65), and survivin in DLD1, SW480, and HT29 cells treated with miR-NC or MIRTX. b-Actin (ACTB) was the loading control. B, The luciferase activities of a plasmid that contained the pGL4.32[luc2P/NF-kB- RE/Hygro] in DLD1, SW480, and HT29 cells treated with miR-NC or MIRTX. The data show that MIRTX signiﬁcantly suppressed NF-kB transcriptional activities in DLD1 and SW480 cells. All data, mean SD. , P < 0.01. C, Western blot analyses of PIK3R1, phospho-AKT, and phospho-mTOR in DLD1, SW480, and HT29 cells treated with miR-NC or MIRTX. b-Actin (ACTB) was the loading control.

formulated MIRTX side effects. We observed no differences in Discussion mortality or bodyweight loss between the groups throughout In this study, we identified a novel small RNA sequence the experiment (Fig. 6A). Blood chemistry tests showed no that could significantly inhibit KRAS-mutant colorectal cancer abnormalities except for a modest increase in phosphorus cell growth and designated this potentially therapeutic RNA (Supplementary Fig. S5A), and hematoxylin and eosin sequence "MIRTX." Our results show that MIRTX exhibits a (H&E)–stained sections of each organ tissue (brain, heart, potent growth-inhibitory and proapoptotic effect in vitro and lung, liver, spleen, kidney, colon) revealed no particular his- in vivo. Moreover, we also showed that MIRTX could directly tologic damage in formulated MIRTX-treated mice (Supple- target PIK3R1 and CXCR2, suppressing NF-kB signaling path- mentary Fig. S5B). We then investigated the in vivo tumor- ways. These results suggest a potential therapeutic strategy in inhibitory effect of MIRTX by systemic administration in a KRAS-mutant colorectal cancer. To our knowledge, this report tumor xenograft mouse model. We used a super carbonate is the first to describe the antitumor function of this artificial apatite as the delivery vehicle via tail vein injections. The small RNA sequence. results showed that the systemic administration of this for- It is generally thought that hsa-miR-29b precursor (primary mulated MIRTX significantly inhibited tumor growth of DLD1 miRNA: pri-miRNA) is a small RNA molecule in the shape of a and HT29 in the mice (Fig. 6B, Supplementary Fig. S5C and stem loop or hairpin. After being captured by RISC, each arm of S5D). We then focused on KRAS-mutated DLD1 cells in which the hairpin is processed into the two single RNA strands. One- MIRTX downregulated CXCR2 and PIK3R1, resulting in inhi- strand RNA (guide RNA, mature miRNA) is stable and function- bition of NF-kBcascadein vitro. With RT-PCR analyses, we ally binds to the target mRNA and the other (passenger RNA), confirmed that DLD1 tumor xenografts that were administered initially discriminated with a star mark, is unstable and destined systemic formulated MIRTX exhibited a significantly higher to undergo degradation (23, 24, 38, 39). For example, miR-29b-1- concentration of MIRTX compared with negative control 5p is described as miR-29b on 3D-Gene Human miRNA Oligo miRNA (Fig. 6C). Western blot analyses of DLD1 xenografts chips (Toray Industries), which means it is degraded and has no demonstrated that expression of CXCR2, PIK3R1, phosphor- apparent human physiologic function (38). Indeed, a number of ylated p65, BCL2, and survivin decreased. In contrast, cleaved studies >500 on PubMed database focused on miR-29b-3p, caspase-3 increased with treatment of MIRTX (Fig. 6D). IHC whereas there are only 6 reports on miR-29b-1-5p (27, 28, 40– analyses showed concordant results that expression of CXCR2, 43). We also confirmed that expression levels of miR-29b-1-5p PIK3R1, phosphorylated NF-kB (p-p65), and BCL2 decreased were very low in clinical colorectal cancer tumor samples, which in MIRTX-treated tumors compared with the control tumors, contrasts to high expression of miR-29b-3p using the same while intense expression of the whole NF-kB (p65) protein was colorectal cancer samples. maintained (Fig. 6E).

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 983

Inoue et al.

Figure 6. Systemic administration of formulated sCA-MIRTX inhibited tumor growth in vivo. Tumor xenograft mouse models received intravenous administrations ofMIRTXormiR-NCusingsCAasavehicleondays0,2,4,7,9,12,14,and16via a tail vein injection (arrows, days of injections). Each injection contained 40 mg of formulated oligo with a 40% conjugation rate. A, Body weight loss was evaluated for the parent (n ¼ 8), treated with sCA-control-miR (n ¼ 8), and treated with sCA-MIRTX (n ¼ 8). B, DLD1 and HT29 tumor xenograft mouse models received intravenous administration of MIRTX or control-miRNA using sCA. Tumor volumes were evaluated for the parent (n ¼ 8), treated with miR-NC (n ¼ 8), treated with MIRTX (n ¼ 8) on days 0, 2, 4, 7, 9, 12, 14, and 16. Tumors were resected on day 18. All data, mean SD. (, P < 0.01, two-way ANOVA). C, RT-PCR analyses showing the high expression level of MIRTX in tumor xenografts administered formulated sCA-MIRTX, compared with parent or sCA-control-miR–administered tumors. D, Western blot analyses showing the expression of CXCR2, PIK3R1, NF-kB (p65), phosphorylated NF-kB (p-p65), BCL2, survivin, and cleaved caspase-3 (C-caspase-3) in resected xenograft tumors from parent, control-miRNA, and MIRTX-treated mice. b-Actin (ACTB) was the loading control. All data, mean SD. , P < 0.01. E, The representative images of H&E staining and IHC staining for CXCR2, PI3 kinase p85a,NF-kB (p65), phosphorylated NF-kB (phosphorylated p65), and BCL2. Each scale bar, 50 mm.

984 Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics

A Novel Small RNA against KRAS-Mutant Colon Cancer

As mentioned in the Introduction, a recent report demonstrated when we examined protein expression of PIK3R1 and CXCR2 in a tumor-suppressive effect by miR-29b-1-5p in triple-negative HT29 cells, they were not changed (Fig. 5A–C), suggesting that an breast cancer (28). Therefore, it is of course worth investigating alternative mechanism exists. a role of miR-29b-1-5p in colon cancer; however, the actual The use of siRNA or miRNA as therapeutics or prognostic motivation to start this project emerged from our unpublished biomarkers has shown promise in the treatment of cancer, and data related to a previous study (32), in which mimic-miR-29b-1- several anticancer small RNA–based drugs have entered clinical 5p exhibited a considerably potent growth-inhibitory effect com- trials (19, 47, 48). However, many obstacles that hinder the pared with other miRNAs tested (Supplementary Fig. S6). functional efficacy and safety of miRNAs in the clinic remain to To guarantee the activity of guide miRNA strand alone and be resolved. Effective delivery is the most challenging hurdle in the to exclude off-target effect from the opposite complementary development of miRNAs as a broad therapeutic platform (48). In sequence, companies supply optimally conditioned miRNA mim- this study, we applied an in vivo pH-sensitive delivery system for ic. These include mirVana miRNA Mimics (Thermo Fisher Scien- nucleic acids using sCA nanoparticles. Our previous studies dem- tific), miRCURY LNA microRNA Mimic (Exiqon), and the one onstrated that sCA nanoparticles had high affinity with nucleic from Gene Design Inc. As anticipated, we found that a completely acids and efficiently delivered miRNA and siRNA into various opposite complementary strand to miR-29b-1-5p named MIRTX tumors in mouse models compared with other types of in vivo possessed a potent growth-inhibitory effect, whereas a mature systemic delivery vehicles, such as liposome or atelocollagen (12, miR-29b-1-5p strand did not, using the conditioned miRNA 31, 33, 49). Accumulating evidence provides a strong rationale for mimics obtained from the two different suppliers. therapeutically targeting the NF-kB pathway in many types of Although miR-29b-3p and MIRTX are both the partners of miR- cancers (9, 11, 44). However, despite the pharmaceutical industry's 29b-1-5p, their sequences are rather different. This is because each aggressive effort to develop specific NF-kB inhibitors for indica- arm in the pri-miR-29b is processed with 3 base-shift in RISC (Fig. tions in oncology, no such inhibitor has been clinically approved 1), which caused totally different seed sequences between miR- because of the toxicities associated with the global suppression of 29b-3p and MIRTX, ACCACGA and CCAAAUC, respectively. NF-kB (50). We showed that intravenously administered sCA– Although miR-29b-3p is a putative anti-oncomiRNA (19–21), MIRTX conjugates abundantly accumulated in the xenograft MIRTX showed much stronger antiproliferative activity in DLD1 tumors in nude mice. After treatment, xenograft tumors exhibited (Supplementary Fig. S7). Also as shown in Fig. 2A, MIRTX significant downregulation of PIK3R1, CXCR2, and phosphory- exhibited an equivalent to or more effective than the putative lated p65, and antiapoptotic protein expression, as observed in in anti-onco miR-34a (35) in colorectal cancer cells. Moreover, vitro experiments. Moreover, systemic administration of sCA– MIRTX displayed a very strong inhibition of growth and invasion MIRTX conjugates in colorectal cancer mouse models showed no in a series of KRAS-mutated pancreatic cancer cells (H. Yama- specific toxicities. These results suggest that targeting the NF-kB moto; unpublished observations). pathway by MIRTX could offer a potentially great therapeutic On the basis of in silico computational analysis and cDNA efficacy and less dose-limiting toxicities in colorectal cancer. microarray analysis, we focused on PIK3R1 and CXCR2 because In conclusion, we identified a novel small RNA sequence, these molecules are involved in NF-kB signaling pathway (Sup- MIRTX, that significantly inhibited KRAS-mutant colorectal plementary Fig. S3, refs. 9, 13, 44). As mentioned in Introduction, cancer cell growth in vitro and in vivo. We also demonstrated NF-kB is activated in KRAS-mutated colorectal cancer (9, 10). We that MIRTX could suppress NF-kB signaling pathways via direct indeed demonstrated that MIRTX could directly bind to the 30-UTR inhibition of CXCR2 and PIK3R1. These results suggest a novel region of CXCR2 and PIK3R1 mRNAs and suppress protein expres- RNA-based therapeutic strategy against refractory KRAS-mutant sion of both genes in KRAS-mutated colorectal cancer cells, DLD1 colorectal cancer in addition to KRAS wild-type colorectal cancer. and SW480. Earlier studies demonstrated that IL8 and its receptor CXCR2 are significantly upregulated in the tumor and its micro- Disclosure of Potential Conflicts of Interest environment in colorectal cancer (16, 45). CXCR2 is known as No potential conflicts of interest were disclosed. a cell receptor that directly regulates NF-kB signaling pathway (14, 44). On the other hand, PIK3R1 is reportedly shown to be an Authors' Contributions intrinsic activator of NF-kB (10, 13). As a result, we found that Conception and design: A. Inoue, H. Takahashi, Y. Doki, M. Mori, targeting the multiple molecules by MIRTX led to effective inhi- H. Yamamoto bition of the NF-kB signaling pathway, exerting an antitumor effect Development of methodology: A. Inoue, R. Ikeshima, H. Yamamoto and triggering apoptosis in KRAS-mutated colon cancer cells. Acquisition of data (provided animals, acquired and managed patients, It is of interest that AKT was unexpectedly activated (phos- provided facilities, etc.): A. Inoue, T. Mizushima, X. Wu, N. Kambara, S. Ishikawa, J. Wang, Y. Qian, H. Hirose, M. Hiraki, N. Miyoshi, H. Yamamoto phorylated) or compensated when PIK3R1 was downregulated by Analysis and interpretation of data (e.g., statistical analysis, biostatistics, MIRTX (Fig. 5C). Studies have reported that these signaling path- computational analysis): A. Inoue, D. Okuzaki, H. Takahashi, M. Mori, ways are complex and can result in pathway activation by feed- H. Yamamoto back loops or cross-talk signal activations (46). These pathway Writing, review, and/or revision of the manuscript: A. Inoue, T. Mizushima, activations are also considered to be among the causes of drug Y. Yokoyama, H. Takahashi, N. Haraguchi, H. Yamamoto resistance, which suggests possible limitations from targeting a Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Inoue, T. Mizushima, C. Matsuda, M. Mori single molecule. In this regard, it is assumed that MIRTX would be Study supervision: A. Inoue, T. Mizushima, N. Haraguchi, T. Hata, Y. Doki, more effective if it is used in combination with the AKT inhibitor M. Mori, H. Yamamoto or miR-4689 that inhibits KRAS and AKT1 simultaneously (32). We should emphasize that MIRTX is effective not only to KRAS- Acknowledgments mutated colorectal cancer, but also to KRAS wild-type colorectal We thank the Yamamoto's laboratory members, especially Yui Kubota, cancer (e.g., HT29, Colo201: see Supplementary Fig. S6), but Kazuki Oishi, and Mayu Ito, for experimental assistance in this work. We

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 985

Inoue et al.

express our respect to Emeritus Prof. Toshihiro Akaike (FAIS, Ibaraki, Japan) for The costs of publication of this article were defrayed in part by the payment of developing the original carbonate apatite gene delivery system. This work was page charges. This article must therefore be hereby marked advertisement in supported by Grants-in-Aid for Scientiﬁc Research (KAKENHI, no. 21390360, accordance with 18 U.S.C. Section 1734 solely to indicate this fact. no. 30322184, no. 24390315; to H. Yamamoto). H. Yamamoto has been awarded grant support from Project MEET (Osaka University Graduate School Received September 1, 2017; revised January 6, 2018; accepted February 21, of Medicine) and Mitsubishi Tanabe Pharma Corporation. 2018; published ﬁrst March 15, 2018.

References 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin 20. Wu Y, Crawford M, Mao Y, Lee RJ, Davis IC, Elton TS, et al. Therapeutic 2016;66:7–30. delivery of microRNA–29b by cationic lipoplexes for lung cancer. Mol Ther 2. Colvin H, Mizushima T, Eguchi H, Takiguchi S, Doki Y, Mori M. Gastro- Nucleic Acids 2013;2:e84. enterological surgery in Japan: the past, the present and the future. Ann 21. Garzon R, Heaphy CE, Havelange V, Fabbri M, Volinia S, Tsao T, et al. Gastroenterol Surg 2017;1:5–10. MicroRNA 29b functions in acute myeloid leukemia. Blood 2009;114: 3. Van Cutsem E, Lenz HJ, Kohne CH, Heinemann V, Tejpar S, Melezinek I, 5331–41. et al. Fluorouracil, leucovorin, and irinotecan plus cetuximab treatment 22. Qi Y, Huang Y, Pang L, Gu W, Wang N, Hu J, et al. Prognostic value of the and RAS mutations in colorectal cancer. J Clin Oncol 2015;33:692–700. MicroRNA–29 family in multiple human cancers: a meta-analysis and 4. Tabernero J, Van Cutsem E, Diaz-Rubio E, Cervantes A, Humblet Y, Andre T, systematic review. Clin Exp Pharmacol Physiol 2017;44:441–54. et al. Phase II trial of cetuximab in combination with fluorouracil, leu- 23. Schwarz DS, Hutvagner G, Du T, Xu Z, Aronin N, Zamore PD. Asymmetry in covorin, and oxaliplatin in the first-line treatment of metastatic colorectal the assembly of the RNAi enzyme complex. Cell 2003;115:199–208. cancer. J Clin Oncol 2007;25:5225–32. 24. Khvorova A, Reynolds A, Jayasena SD. Functional siRNAs and miRNAs 5. Saeki H, Emi Y, Kumashiro R, Otsu H, Kawano H, Ando K, et al. Impact of exhibit strand bias. Cell 2003;115:209–16. second-line and later cetuximab-containing therapy and KRAS genotypes 25. Okamura K, Phillips MD, Tyler DM, Duan H, Chou YT, Lai EC. in patients with metastatic colorectal cancer: a multicenter study in Japan. The regulatory activity of microRNA species has substantial influence Surg Today 2014;44:1457–64. on microRNA and 3' UTR evolution. Nat Struct Mol Biol 2008;15: 6. De Roock W, Claes B, Bernasconi D, De Schutter J, Biesmans B, Fountzilas 354–63. G, et al. Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy 26. Okamura K, Chung WJ, Lai EC. The long and short of inverted repeat genes of cetuximab plus chemotherapy in chemotherapy-refractory metastatic in animals: microRNAs, mirtrons and hairpin RNAs. Cell Cycle 2008;7: colorectal cancer: a retrospective consortium analysis. Lancet Oncol 2010; 2840–5. 11:753–62. 27. Xu F, Zhang Q, Cheng W, Zhang Z, Wang J, Ge J. Effect of miR-29b-1 and 7. LoRusso PM, Krishnamurthi SS, Rinehart JJ, Nabell LM, Malburg L, Chap- miR-29c knockdown on cell growth of the bladder cancer cell line T24. J Int man PB, et al. Phase I pharmacokinetic and pharmacodynamic study of the Med Res 2013;41:1803–10. oral MAPK/ERK kinase inhibitor PD-0325901 in patients with advanced 28. Drago-Ferrante R, Pentimalli F, Carlisi D, De Blasio A, Saliba C, Baldac- cancers. Clin Cancer Res 2010;16:1924–37. chino S, et al. Suppressive role exerted by microRNA-29b-1-5p in triple 8. Lim HY, Heo J, Choi HJ, Lin CY, Yoon JH, Hsu C, et al. A phase II study of the negative breast cancer through SPIN1 regulation. Oncotarget 2017;8: efficacy and safety of the combination therapy of the MEK inhibitor 28939–58. refametinib (BAY 86–9766) plus sorafenib for Asian patients with unre- 29. Sobin LH, Gospodarowicz MK, Wittekind C. TNM classification of malig- sectable hepatocellular carcinoma. Clin Cancer Res 2014;20:5976–85. nant tumors. Oxford, United Kingdom: Wiley-Blackwell; 2010. 9. Karin M, Cao Y, Greten FR, Li ZW. NF-kappaB in cancer: from innocent 30. Nakamura M, Nagano H, Sakon M, Yamamoto T, Ota H, Wada H, et al. bystander to major culprit. Nat Rev Cancer 2002;2:301–10. Role of the Fas/FasL pathway in combination therapy with interferon- 10. Jana A, Krett NL, Guzman G, Khalid A, Ozden O, Staudacher JJ, et al. NFkB alpha and fluorouracil against hepatocellular carcinoma in vitro. J Hepatol is essential for activin-induced colorectal cancer migration via upregulation 2007;46:77–88. of PI3K-MDM2 pathway. Oncotarget 2017;8:37377–93. 31. Wu X, Yamamoto H, Nakanishi H, Yamamoto Y, Inoue A, Tei M, et al. 11. Viatour P, Merville MP, Bours V, Chariot A. Phosphorylation of NF-kappaB Innovative delivery of siRNA to solid tumors by super carbonate apatite. and IkappaB proteins: implications in cancer and inflammation. Trends PLoS One 2015;10:e0116022. Biochem Sci 2005;30:43–52. 32. Hiraki M, Nishimura J, Takahashi H, Wu X, Takahashi Y, Miyo M, et al. 12. Han SS, Yun H, Son DJ, Tompkins VS, Peng L, Chung ST, et al. NF-kappaB/ Concurrent targeting of KRAS and AKT by MiR–4689 is a novel treatment STAT3/PI3K signaling crosstalk in iMyc E mu B lymphoma. Mol Cancer against mutant KRAS colorectal cancer. Mol Ther Nucleic Acids 2015;4: 2010;9:97. e231. 13. Hutti JE, Pfefferle AD, Russell SC, Sircar M, Perou CM, Baldwin AS. 33. Takeyama H, Yamamoto H, Yamashita S, Wu X, Takahashi H, Nishi- Oncogenic PI3K mutations lead to NF-kappaB-dependent cytokine expres- mura J, et al. Decreased miR–340expressioninbonemarrowis sion following growth factor deprivation. Cancer Res 2012;72:3260–9. associated with liver metastasis of colorectal cancer. Mol Cancer Ther 14. Ning Y, Labonte MJ, Zhang W, Bohanes PO, Gerger A, Yang D, et al. The 2014;13:976–85. CXCR2 antagonist, SCH–527123, shows antitumor activity and sensitizes 34. Chowdhury EH, Maruyama A, Kano A, Nagaoka M, Kotaka M, Hirose S, cells to oxaliplatin in preclinical colon cancer models. Mol Cancer Ther et al. pH-sensing nano-crystals of carbonate apatite: effects on intracellular 2012;11:1353–64. delivery and release of DNA for efficient expression into mammalian cells. 15. Ning Y, Manegold PC, Hong YK, Zhang W, Pohl A, Lurje G, et al. Inter- Gene 2006;376:87–94. leukin–8 is associated with proliferation, migration, angiogenesis and 35. Li XJ, Ren ZJ, Tang JH. MicroRNA–34a: a potential therapeutic target in chemosensitivity in vitro and in vivo in colon cancer cell line models. Int human cancer. Cell Death Dis 2014;5:e1327. J Cancer 2011;128:2038–49. 36. Pahl HL. Activators and target genes of Rel/NF-kappaB transcription 16. Balkwill F. Cancer and the chemokine network. Nat Rev Cancer 2004;4: factors. Oncogene 1999;18:6853–66. 540–50. 37. Yu LL, Yu HG, Yu JP, Luo HS, Xu XM, Li JH. Nuclear factor-kappaB p65 17. Duckworth C, Zhang L, Carroll SL, Ethier SP, Cheung HW. Overexpression (RelA) transcription factor is constitutively activated in human colorectal of GAB2 in ovarian cancer cells promotes tumor growth and angiogenesis carcinoma tissue. World J Gastroenterol 2004;10:3255–60. by upregulating chemokine expression. Oncogene 2016;35:4036–47. 38. Guo L, Lu Z. The fate of miRNA strand through evolutionary analysis: 18. He L, Hannon GJ. MicroRNAs: small RNAs with a big role in gene implication for degradation as merely carrier strand or potential regulatory regulation. Nat Rev Genet 2004;5:522–31. molecule? PLoS One 2010;5:e11387. 19. Inoue A, Yamamoto H, Uemura M, Nishimura J, Hata T, Takemasa I, et al. 39. Filipowicz W, Bhattacharyya SN, Sonenberg N. Mechanisms of post-tran- MicroRNA–29b is a novel prognostic marker in colorectal cancer. Ann Surg scriptional regulation by microRNAs: are the answers in sight? Nat Rev Oncol 2015;22 Suppl 3:S1410–8. Genet 2008;9:102–14.

986 Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics

A Novel Small RNA against KRAS-Mutant Colon Cancer

40. Yamaguchi N, Osaki M, Onuma K, Yumioka T, Iwamoto H, Sejima T, et al. 46. Mendoza MC, Er EE, Blenis J. The Ras-ERK and PI3K-mTOR path- Identiﬁcation of microRNAs involved in resistance to sunitinib in renal cell ways: cross-talk and compensation. Trends Biochem Sci 2011;36: carcinoma cells. Anticancer Res 2017;37:2985–92. 320–8. 41. Zhou S, Li J, Xu H, Zhang S, Chen X, Chen W, et al. Liposomal curcumin 47. Toiyama Y, Tanaka K, Inoue Y, Mohri Y, Kusunoki M. Circulating cell- alters chemosensitivity of breast cancer cells to Adriamycin via regulating free microRNAs as biomarkers for colorectal cancer. Surg Today 2016; microRNA expression. Gene 2017;622:1–12. 46:13–24. 42. Shi J, Liu H, Wang H, Kong X. MicroRNA expression signature in degen- 48. de Fougerolles A, Vornlocher HP, Maraganore J, Lieberman J. Interfering erative aortic stenosis. Biomed Res Int 2016;2016:4682172. with disease: a progress report on siRNA-based therapeutics. Nat Rev Drug 43. Uematsu Y, Akai S, Tochitani T, Oda S, Yamada T, Yokoi T. MicroRNA- Discov 2007;6:443–53. mediated Th2 bias in methimazole-induced acute liver injury in mice. 49.OgawaH,WuX,KawamotoK,NishidaN,KonnoM,KosekiJ,etal. Toxicol Appl Pharmacol 2016;307:1–9. MicroRNAs induce epigenetic reprogramming and suppress malignant 44. Yang G, Rosen DG, Liu G, Yang F, Guo X, Xiao X, et al. CXCR2 promotes phenotypes of human colon cancer cells. PLoS One 2015;10: ovarian cancer growth through dysregulated cell cycle, diminished apo- e0127119. ptosis, and enhanced angiogenesis. Clin Cancer Res 2010;16:3875–86. 50. Tornatore L, Sandomenico A, Raimondo D, Low C, Rocci A, Tralau- 45. Li A, Varney ML, Singh RK. Expression of interleukin 8 and its receptors in Stewart C, et al. Cancer-selective targeting of the NF-kappaB survival human colon carcinoma cells with different metastatic potentials. Clin pathway with GADD45beta/MKK7 inhibitors. Cancer Cell 2014;26: Cancer Res 2001;7:3298–304. 495–508.

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 987

A miR-29b Byproduct Sequence Exhibits Potent Tumor-Suppressive Activities via Inhibition of NF- κB Signaling in KRAS-Mutant Colon Cancer Cells

Akira Inoue, Tsunekazu Mizushima, Xin Wu, et al.

Mol Cancer Ther 2018;17:977-987. Published OnlineFirst March 15, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-17-0850

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2018/03/15/1535-7163.MCT-17-0850.DC1

Cited articles This article cites 49 articles, 10 of which you can access for free at: http://mct.aacrjournals.org/content/17/5/977.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/17/5/977. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.