Concurrent Targeting of KRAS and AKT by MiR-4689 Is a Novel Treatment Against Mutant KRAS Colorectal Cancer

KRAS mutations are a major cause of drug resistance to molecular-targeted therapies. Aberrant epidermal growth factor receptor (EGFR) signaling may cause dysregulation of microRNA (miRNA) and gene regulatory networks, which leads to cancer initiation and progression. To address the functional relevance of miRNAs in mutant KRAS cancers, we transfected exogenous KRASG12V into human embryonic kidney 293 and MRC5 cells with wild-type KRAS and BRAF genes, and we comprehensively profiled the dysregulated miRNAs. The result showed that mature miRNA oligonucleotide (miR)-4689, one of the significantly down-regulated miRNAs in KRASG12V overexpressed cells, was found to exhibit a potent growth-inhibitory and proapoptotic effect both in vitro and in vivo. miR-4689 expression was significantly down-regulated in cancer tissues compared to normal mucosa, and it was particularly decreased in mutant KRAS CRC tissues. miR-4689 directly targets v-ki-ras2 kirsten rat sarcoma viral oncogene homolog (KRAS) and v-akt murine thymoma viral oncogene homolog 1(AKT1), key components of two major branches in EGFR pathway, suggesting KRAS overdrives this signaling pathway through inhibition of miR-4689. Overall, this study provided additional evidence that mutant KRAS functions as a broad regulator of the EGFR signaling cascade by inhibiting miR-4689, which negatively regulates both RAS/mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K)/AKT pathways. These activities indicated that miR-4689 may be a promising therapeutic agent in mutant KRAS CRC.


Introduction
Recent advances in our comprehension of the specific signaling pathways in cancer cells have offered novel targeted therapies for patients with metastatic colorectal cancer (mCRC). Studies have shown that monoclonal antibodies that target epidermal growth factor receptors (EGFRs) provide a survival benefit in mCRCs that harbor wild-type KRAS. 1,2 A retrospective analysis of randomized, phase 2 OPUS and phase 3 CRYSTAL clinical trials revealed that cetuximab combined with a first-line chemotherapy (FOLFOX4 or FOL-FIRI) significantly improved the prognosis of patients with wild-type KRAS mCRC compared to chemotherapy alone. However, these effects were not observed in patients with mutant KRAS mCRC. 2,3 Because KRAS mutations occur in ~40% of patients with mCRC, 4 novel therapeutic strategies are urgently needed for treating this population.
Ligand binding to the extracellular domain of EGFR results in phosphorylation of the tyrosine kinase located on an intracellular domain. This activation of the receptor then induces the activation of intracellular effectors via intracellular signaling pathways, including one where rapidly accelerated fibrosarcoma (RAF) activates the dual kinase, mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) kinase (MEK), which then activates the ERK pathway. In addition, EGFR signals via the stress-activated protein kinase (SAPK)/c-Jun N-terminal kinase (JNK) pathway, the phosphoinositide 3-kinase (PI3K)/AKT pathway, the signal transducers and activators of transcription (STAT) 3 pathway, and the phospholipase Cγ pathway. [5][6][7] Cetuximab binds to EGFR with a high specificity and blocks ligand-induced phosphorylation of the receptor. This blocking leads to the inhibition of EGFR downstream signaling pathways, including the RAF/MEK/ERK pathway, which is mainly involved in cell proliferation, and the PI3K/AKT pathway, which is mainly involved in cell survival and tumor invasion. 8,9 However, once point mutations in the KRAS oncogene occur, the downstream pathways become constitutively active (RAF/ MEK/ERK, SAPK/JNK, and possibly PI3K/AKT pathways). 10 This constitutive activity is considered the major reason that mutant KRAS CRCs are resistant to anti-EGFR therapy. 1,7 Combination therapies with multiple inhibitors are currently in development to overcome resistance to anti-EGFR therapy. However, a clinical trial for testing the MEK1/2 inhibitor, PD0325901,was discontinued in phase 2 because of severe toxicity, which included blurred vision and acute neurotoxicity. 11,12 Another MEK1/2 inhibitor, RDEA119, combined with sorafenib also led to severe adverse events, including liver failure, sepsis/hepatic encephalopathy, and tumor lysis syndrome for patients with hepatocellular carcinoma in a phase 2 study. 13 A multicenter phase 1/2 study to evaluate another MEK1/2 inhibitor, AS703026, plus 5-fluorouracil, Figure 1 Screening for candidate miRs that could suppress SRE or AP1 transcription activity in HEK293-KRAS. Mutant KRAS cDNA was introduced into HEK293 or MRC5 cells. MiRNA expression profiles were compared between these cells with a miRNA microarray analysis (see Supplementary Figure S1). (a) Schematic diagram of the results of microarray analysis. The protocol identified 6 miRNAs common to both cell lines that showed >8.0-fold decreased expression in HEK293-KRAS and MRC5-KRAS compared to control HEK293 and MRC5 cells. In addition to those six miRNAs, we identified miR-4685-3p with a Target Scan search, which indicated that it was likely to bind to the 3′UTR region of MEK2, and it functioned downstream of the KRAS signaling pathway (see Supplementary Table S1 and S2). (b) Activated KRAS signaling through (left) RAF-MEK-ERK and (right) MEKK1-SEK1-JNK transduction pathways. Addition of EGF or introduction of a mutant KRAS gene into HEK293 enhanced (left) SRE and (right) AP1 transcription of the luciferase reporter gene. (c) Screening candidate miRs to identify those that inhibit the two signal transduction pathways. Several miRNAs significantly suppressed SRE or AP1 transcription of luciferase compared with miR-NC in HEK293-KRAS cells. We selected miR-4689, miR4685-3p, miR-296-5p, and miR-4442 for further evaluations in mutant KRAS CRC cells. All data represent the mean ± SEM; * P < 0.05, ** P < 0.01. 12   leucovorin, and irinotecan (FOLFIRI) for mutated KRAS mCRC was unable to proceed to phase 2, because effective doses of AS703026 could not be achieved due to its toxicity. 14 Therefore, novel effective therapeutic alternatives are critically needed. MicroRNAs (miRNAs) are endogenous, single-stranded, small, noncoding RNAs of 20-22 nucleotides. miRNAs bind to complementary sequences, mainly in the 3′-untranslated regions (3′-UTR), of multiple target mRNAs, and they can either cause translational repression or facilitate cleavage of their target mRNAs. 15,16 miRNAs are involved in crucial biological processes, including development, differentiation, apoptosis, and proliferation. 17,18 To date, a few studies have shown, through searches with Target Scan and/or sequence binding prediction programs, that miR-30b, miR-18a, and miR-143 specifically inhibit KRAS expression in colon cancer cells. [19][20][21] This study aimed to discover a powerful, therapeutic miRNA that could target mutant KRAS in CRC. For this purpose, we performed a microarray analysis in KRAS-transfected, human embryonic kidney 293 (HEK293) and human lung fibroblasts (MRC5). We screened the ability of candidate miRNAs to suppress signal transduction in the RAF/ MEK/ERK and MEKK/SEK/JNK pathways. Because miRNAs can bind to multiple target regions through imperfect complementarity, 22 we also investigated whether candidate miRNAs could concurrently inhibit other transduction pathways, such as the PI3K/AKT and STAT3 pathways.

KRAS G12V overexpression downregulates a subset of miRNAs
We first examined whether mutated KRAS could induce dysregulation of miRNA expression. With a miRNA microarray, we compared the profile of miRNAs expressed in HEK293 and MRC5 cells that overexpressed KRAS G12V to those expressed in parental cells that harbored wild-type KRAS. The results indicated that a subset of miRNAs, including miR-4270, miR-4689, miR-296-5p, miR-3619-3p, miR-4731-3p, and miR-4442, was significantly down-regulated in KRAS G12V transfected cells. Most of these miRNAs were commonly down-regulated in both cell types (see Supplementary Figure S1 and Supplementary Tables S1 and S2). We also included miR-4685-3p for comparison in further analyses, because miR-4685-3p potentially targeted MEK2, a downstream target of KRAS, based on an in silico prediction obtained with Target Scan 6.2 software 23 (Figure 1a).

miR-4689 inhibits KRAS signaling activity
To determine the functional relevance of these miRNAs in KRAS signaling, we used serum response element (SRE) and activator protein-1 (AP1) reporter assays. The pGL4.33[luc2P/SRE/Hygro] vector contained SRE, which drives the transcription of the luciferase reporter gene, luc2P, in response to activation of the MAPK/ERK signaling pathway. 24 The pGL4.44[luc2P/AP1 RE/Hygro] vector contained an AP1 response element that drives the transcription of   luc2P, in response to activation of the MAPK/JNK signaling pathway. We confirmed that the addition of EGF or the introduction of a mutant KRAS significantly enhanced both SRE and AP1 transcription activities in HEK293 cells (Figure 1b). In HEK293 cells that overexpressed KRAS G12V (HEK293-KRAS), most of the tested miRNAs inhibited SRE and AP1 transcriptional activities, except miR-3619-3p (in SRE-reporter cells) and miR-4731-3p (in AP1-reporter cells) (Figure 1c). Further analyses in mutant KRAS CRC cell lines, including the SW480 (KRAS G12V ), DLD1 (KRAS G13D ), and HCT116 (KRAS G13D ) lines, revealed that only miR-4689 potently suppressed both SRE and AP1 transcriptional activities in all tested mutated KRAS cell lines (Figure 2a). Furthermore, miR-4689, but not miR-4685-3p, significantly inhibited cell proliferation in these cell lines (Figure 2b).

miR-4689 is downregulated in mutant KRAS CRC
Next, we compared the expression of miR-4689 in nine CRC cell lines of which seven cell lines were mutated KRAS or BRAF cells. Wild-type KRAS or BRAF cell lines had significantly higher miR-4689 levels than did mutated KRAS or BRAF CRC cell lines (Figure 3a, P = 0.02). miR-4689 was downregulated in mutant KRAS CRC clinical samples when compared with wild-type KRAS samples (Figure 3b, P = 0.007). Consistent with its tumor suppressive properties, the expression of miR-4689 was significantly downregulated in CRC tissues compared with its expression in normal mucosa (Figure 3c, P = 0.0002).
In vitro mechanistic studies confirmed an inverse correlation between KRAS and miR-4689 expression levels. Thus, forced expression of KRAS gene in nontumor KEK293 and MRC5 significantly decreased miR4689 expression (see Supplementary Figure S2a,b, P = 0.018 and 0.03, respectively). In contrast, knockdown of KRAS gene expression with the shKRAS vector increased miR-4689 expression in KRAS mutated DLD1 cells (P = 0.002, see Supplementary Figure S3a).

miR-4689 directly targets KRAS mRNA
To gain further insight into how miR-4689 regulated the KRAS signaling pathway, we searched for potential targets

AKT1 is also a target of miR-4689
Although miR-4689 and siRNA-mediated silencing of MEK1 and MEK2 had similar effects on ERK phosphorylation (Figure 5a), miR-4689 inhibited proliferation more potently than the siRNA treatment in mutated KRAS DLD1 cells (Figure 5b). This raised the possibility that miR-4689 might target additional proliferation-associated genes outside of the MAPK/ERK pathway. We first searched for a miR-4689 binding site on the 3′UTRs of key molecules in other pathways related to EGFR signaling; however, no putative sequence was found. Recent evidence showed that, in addition to 3′UTRs, miRNAs can target open reading frames (ORFs) and 5′UTRs of mRNAs to regulate gene expression. [25][26][27] Further investigation of related molecules showed that the AKT1 gene, a crucial component of the PI3K pathway, carried a highly conserved miR-4689 binding element on its ORF sequence (nucleotides 69-90; Figure 5c). Luciferase activity of a reporter plasmid that harbored the AKT1 ORF sequence was significantly inhibited with overexpression of miR-4689; this result suggested that miR-4689 might directly bind the coding sequence of AKT1 mRNA (Figure 5d). Inhibition of the AKT1 gene by miR-4689 was further confirmed in Western blot (Figure 5e) and real-time quantitative reverse-transcription PCR (qRT-PCR) (Figure 5f) assays.

miR-4689 induces apoptosis in mutant KRAS cells
Because miR-4689 could regulate both MAPK/ERK and PI3K/AKT pathways, we next investigated whether miR-4689 had a proapoptotic effect in mutated KRAS DLD1 cells. Transfection of miR-4689 reduced the levels of phosphorylated ERK and phosphorylated AKT (Figure 6a), and increased the number of Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick-End Labeling (TUNEL)-positive apoptotic cells (Figure 6b) in vitro. Overexpression of miR-4689 induced the expression of proapoptotic genes, BAX, BAD, BAK, and Cyt C, and suppressed the expression of antiapoptotic genes, BCL2 and BCLXL (Figure 6b). In contrast, knockdown of KRAS or AKT1 alone led to only limited effects on altered expression of the apoptosis-related genes (see Supplementary Figure S4). These results imply that We confirmed that the expression of KRAS, AKT1, and other crucial components of the MAPK/ERK and PI3K/AKT signaling pathways decreased in mouse xenografts (Day 9, after three single systemic administrations of miRNAs; Figure 6d). Xenografts, which were exposed to the systemic miRNAs, exhibited ~38.8-fold higher concentrations of miR-4689 compared to the negative control; this result indicated efficient drug delivery to the focal site (Figure 6d, P = 0.002).

In vivo safety test of miR-4689
To examine in vivo safety, we investigated the toxicity of repeat administration of miR-4689 (40 g per injection × 7 times) at day 14. We observed no mortalities or body weight loss (Figure 7a). Blood chemistry tests revealed no physiologically significant differences between miRNA negative control oligonucleotide (miR-NC) and miR-4689 treated groups (Figure 7b), except for the slight increase in BUN (P = 0.001). Hematoxylin and eosin-stained sections of each organ tissue (brain, heart, lung, liver, kidney, spleen, small intestine, and colon) showed that miR-4689 caused no particular histological damage (Figure 7c).

Discussion
In this study, we focused on dysregulated miRNAs in oncogenic KRAS-driven CRC. Earlier studies have identified miRNAs that targeted components of the EGFR signaling To identify mutant KRAS-related miRNAs, we introduced KRAS G12V into noncancerous HEK293 and MRC5 cell lines, rather than wild-type KRAS CRC cells. This strategy allowed us to exclude numerous potential background disorders of miRNAs, caused by various genetic and epigenetic alterations observed in cancer, which facilitated the identification of miRNAs induced solely by KRAS gene alterations.
We selected for therapeutic miRNAs, mainly by evaluating the RAF/MEK/ERK pathway, and measuring SRE activity after treatment with specific miRNAs. Inhibition of the MAPK/ ERK pathway is a minimum requirement for suppression of mutant KRAS tumor cells. Indeed, Migliardi et al. 31 showed that cotreatment with a MEK inhibitor and a PI3K/mTOR inhibitor suppressed tumor growth in patient-derived xenografts of mutant KRAS colorectal carcinomas. Misale et al. 32 showed that the cetuximab-resistant DiFi and Lim1215 cells recovered sensitivity in response to combinatorial targeting of MEK and EGFR. Lee et al. 33 reported that combination treatment with simvastatin, which inhibits the MAPK/ERK pathway, and cetuximab showed synergistic antitumor effects in xenograft tumors that originated from mutant KRAS CRC cells. Those findings suggested that combined blockade of multiple pathways is critical for targeting tumors that harbor KRAS mutations.
After screening for SRE/AP1 activity, we identified miR-4689 and miR-4685-3p as potent inhibitors in the RAF/MEK/ ERK or MEKK/SEK/JNK pathway. It is of interest that miR-4689, but not miR-4685-3p, inhibited proliferation of mutant KRAS SW480 cells, and that only miR-4689 could correct altered RAF/MEK/ERK activity (Figure 2). Based on these findings, an investigation is currently underway to pursue the possibility of developing a tailor-made medicine based on reporter assays performed with surgically resected tumor cells. The in vitro proliferation assay showed that miR-4689 potently inhibited cell growth in all tested CRC cell lines with KRAS mutations. To our knowledge, this was the first report to describe miR-4689 function in cancer development.
Our results showed that miR-4689 could directly bind to the 3′UTR region of KRAS and to the ORF of AKT1 mRNAs. It is known that miRNAs preferably bind to 3′UTR regions to regulate gene expression; however, recent evidence had suggested that ORFs and 5′ UTRs also included numerous miRNA binding elements, and the binding of miRNAs to these regions could lead to activation or inhibition of target mRNAs. [25][26][27] A few well-understood examples include miR143, which binds to the ORF of BRAF mRNA, and miR145, which binds to the ORF of DNA Fragmentation Factor-45 (DFF45) mRNA, to regulate gene expression. 20,34 In our experiments, miR-4689 inhibited the luminescence intensity of a luciferase reporter that carried a miR-4689 binding element from the AKT1 coding region; this result indicated that miR-4689 directly interacted with the AKT1 ORF (Figure 5d). We found that miR-4689 could significantly inhibit AKT1 expression, at both the mRNA and protein levels (Figure 5e,f).
Earlier studies have shown that targeting either the RAF/MEK/ERK or the PI3K/AKT pathway alone was not sufficient to regulate the entire EGFR signaling pathway, because one signal transduction branch could compensate for diminished signaling in another branch. For example, a selective MEK inhibitor was shown to induce activation of AKT in basal-like breast cancer models and in pancreatic cancer cell lines. [35][36][37] In addition, PI3K/AKT pathway activation caused resistance to MEK inhibitors in mutant KRAS cells. 38 Indeed, we validated that knockdown of KRAS alone could not down-regulate the ERK phosphorylation in DLD1 and that it led to only limited growth inhibition, which is consistent with an earlier finding reported by Singh et al. 39 (see Supplementary Figure S3b). In this regard, the ability of miR-4689 to target both RAF/MEK/ERK and PI3K/AKT pathways through direct inhibition of KRAS and AKT1 represents a promising strategy against mutant KRAS tumors. We further demonstrated that cotransfection of miR-4689 together with a KRAS-and/or AKT-expression vector in DLD1 restored cell proliferation (see Supplementary Figure S5). This result indicates that the targeting of both KRAS and AKT1 is essentially important to inhibit proliferation of KRAS mutated DLD1. Our results showed that miR-4689 effectively reduced phosphorylation of MEK/ERK and dramatically induced proapoptotic factors, including BAX, BAD, BAK, and Cyt C (Figures 5 and 6), which almost completely blocked cell proliferation in vitro (Figure 2b) and in vivo (Figure 6c). The growth inhibitory effect of miR-4689 was much stronger than that of a siRNA-mediated knock-down of either MEK1 or MEK2; even though the ERK phosphorylation levels were similar. This result suggested that miR-4689 could affect multiple targets simultaneously (Figure 5b). We found that the expression level of miR-4689 was significantly reduced in mutant KRAS and mutant BRAF cell lines compared with either wild-type cell line although we had only two KRAS or BRAF wild CRC cells examined. In clinical specimens, we found that miR-4689 was significantly down-regulated in human CRC samples compared with those of normal mucosa samples; moreover, miR-4689 expression was much lower in cancer tissues that harbored mutant KRAS compared with those that harbored wild-type KRAS. Moreover, mechanistic studies showed that overexpression of KRAS gene decreased miR-4689 expression in HEK293 and MRC5 cells (see Supplementary Figure S2b had no serious adverse effects on each organ tissue (brain, heart, lung, liver, kidney, spleen, small intestine, and colon). Data represent the mean ± SEM; n = 3 for each group. 24 Figure S3a). These findings suggested that mutant KRAS/BRAF CRC cells may have a mechanism to maintain low levels of the potentially toxic miR-4689. Given that the restoration of tumor suppressive miRNAs is a relatively safe, effective strategy for cancer treatment, 40,41 systemic or focal administration of miR-4689 represents a novel therapeutic strategy for treating mutated KRAS cancers. Indeed, we did not detect any serious adverse effects in miR-4689 treated mice (Figure 7).
Several siRNA-and miRNA-based therapeutic methodologies have been advanced for delivering these drugs to liver and kidney, and these are undergoing phase 1 or 2 trials for treating hypercholesterolemia, acute kidney injury, liver cancer, transthyretin-mediated amyloidosis, and hepatitis C virus. 42,43 However, systemic delivery technology of siRNAs and miRNAs for treating solid tumors has been impeded by many limitations. 44 In this study, we applied a potential in vivo, pH sensitive delivery system for miRNAs based on super carbonate apatite nanoparticles. 29,30 We showed that a systemic administration of miR-4689 mixed with the nanoparticles in nude mice inhibited the growth of pre-established DLD1 (KRAS G13D ) xenografts compared to a comparable miR-NC administration. After treatment, xenograft samples exhibited a downregulation of KRAS and AKT1 on Western blots. The proapoptosis and antiapoptosis changes observed in vitro were also observed in vivo.
In conclusion, we demonstrated that miR-4689 potently suppressed oncogenic KRAS-driven EGFR signaling pathways through direct inhibition of KRAS and AKT1 (Figure 8).
These results may open up new possibilities for a miRNA targeted therapy against intractable mutated KRAS CRC.

Materials and methods
Cell line. HEK293, MRC5, and human CRC cell lines, including DLD1, HCT116, SW480, LoVo, RKO, HT29, COLO201, and CaCO-2 were obtained from the American Type Culture Collection (Rockville, MD). CaR-1 with wild-type KRAS or BRAF 45 was obtained from JCRB Cell Bank (Japanese Collection of Research Bioresources Cell Bank). All cells were cultured in Dulbecco's modified Eagle's medium, which contained 10% fetal bovine serum, 100 U/ml penicillin, and 100 mg/ml streptomycin. Cultured were incubated at 37°C in a humidified incubator of 95% air and 5% CO 2 .
Transient miRNA/siRNA/plasmid transfection. Mature miRNA oligonucleotides (miRs) and miR-NCs were obtained from Gene Design (Osaka, Japan). Small interfering (si) RNA oligonucleotides that specifically targeted MEK1 and MEK2 mRNAs (siMEK1, siMEK2) were obtained from Cell Signaling Technology (Beverly, MA). siRNA oligonucleotides that specifically targeted AKT1 mRNA (siAKT1) and scrambled negative control siRNA were purchased from OriGene Technologies (Rockville, MD). We used pCMV6-Entry plasmid inserted the KRAS G12V and AKT1 sequence with SgfI-MluI restriction sites and pCMV6-Entry plasmid as the control (OriGene, Rockville, MD). Cells were transiently transfected with Lipofectamine RNAiMax or Lipofectamine 2000 (Invitrogen, Carlsbad, CA) according to the protocol provided by the manufacturer.
shKRAS. The double-stranded short hairpin RNA clone specifically targeting of KRAS gene was purchased from Broad Institute (Cambridge, MA). pLKO.1 vector nontarget short hairpin RNA was used as a control vector. KRAS short hairpin RNA was introduced by lentiviral delivery and selected with puromycin. After stringent washes, fluorescent signals were scanned with the 3D-Gene Scanner (Toray Industries) and analyzed with 3D-Gene Extraction software (Toray Industries). The raw data of each spot was normalized to the mean intensity of the background signal, which was determined from blank spot signal intensities and 95% confidence intervals. Measurements of duplicate spots with signal intensities >2 SD above the background signal intensity were considered valid. The relative expression level of a given miRNA was calculated by comparing the signal intensities of the averaged valid spots to their mean value throughout the microarray experiments after normalization by their median values adjusted equivalently. miRNA microarray data has been approved and assigned GEO accession numbers (GEO accession no. GSE60370).
Real-time, quantitative PCR analysis of messenger RNA expression. Complementary DNA was synthesized from 1.0 µg total RNA with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA), according to the protocol provided by the manufacturer. qRT-PCR was performed with specifically designed oligonucleotide primers and the LightCycler 480 Real-Time PCR system (Roche Diagnostics, Mannheim, Germany). The amplification products were detected with the LightCycler-DNA master SYBR green I (Roche Diagnostics), according to standard protocols, and the level of target gene expression was calculated. The expression of the target gene was normalized relative to the mRNA expression levels of actin-beta (ACTB), which was used as an endogenous control. The specifically designed PCR primers were, as follows: ACTB forward primer 5′-GATG AGATTGGCATGGCTTT-3′, ACTB reverse primer 5′-CACC TTCACCGTTCCAGTTT-3′; KRAS forward primer 5′-ATTCC TTTTATTGAAACATCAGCA-3′, KRAS reverse primer 5′-TCG-GATCTCCCTCACCAAT-3′; and AKT1 forward primer 5′-AA CACCATGGACAGGGAGAG-3′, AKT1 reverse primer 5′-CATC TTGGTCAGGTGGTGTG-3′.
Real-time qRT-PCR for miRNA expression. The reverse transcription reaction was performed with the TaqMan MicroRNA RT Kit (Applied Biosystems, Foster City, CA). Real-time quantitative PCR was performed with TaqMan MicroRNA Assays (Applied Biosystems) and the 7500HT Sequence Detection System (Applied Biosystems). The target miRNA signal was normalized relative to that of the endogenous control, RNU48. Data were analyzed according to the comparative C t method. 46 Serum response element/AP1 luciferase reporter assay. All cells were seeded onto 96-well plates and transfected with the SRE luciferase reporter vector (pGL4. 33 The coding sequence of AKT1 mRNA, including the miR-4689 binding site, was amplified by RT-PCR. The primer sequences were, as follows: Forward-GCTCGCTAGCCTCG AATGAGCGACGTGGCTATTG; Reverse-ATGCCTGCAGG TCGACGCCACAGAGAAGTTGTTG. The amplified product (175 bp) was subcloned and ligated into the Sal I and Xho I sites in the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega) with the In-Fusion HD Cloning Kit (Clontech). The entire sequence (insert and vector) was confirmed by DNA sequencing.
pmirGLO luciferase reporter assay. Luciferase assays were conducted with 1 × 10 4 DLD1 (KRAS G13D ) cells per well in 96-well plates. Transfections were performed with Lipofectamine 2000 (Invitrogen) in OptiMEM reduced serum media (GIBCO, Invitrogen). Cells were transfected with 100 ng of a pmirGLO construct, which contained either the 3′-UTR of KRAS or the coding sequence of AKT1, and 5 pmol of either miR-NC or miR-4689. Thirty-six hours after transfection, cells were assayed for both firefly and renilla luciferase with the Dual-Luciferase Reporter Assay System (Promega), according to the manufacturer's instructions. All transfection experiments were conducted in triplicate.
TUNEL assay. Apoptotic DNA fragmentation within the cell was analyzed with an in situ DeadEnd Fluorometric TUNEL System Assay Kit (Promega), according to the manufacturer's instructions. In brief, 2 × 10 5 cells/well were plated on Lab-Tek II CC2 Chamber Slides and treated with 50 nM miR-4689 for 48 hours. Cells were fixed in 4% paraformaldehyde at 4°C for 30 minutes, permeabilized in 0.1% Triton X-100, and labeled with fluorescein-12-dUTP with terminal deoxynucleotidyl transferase. The localized green fluorescence of apoptotic cells (fluorescein-12-dUTP) was evaluated by measuring fluorescence intensity with BZ-9000software (KEY-ENCE, Japan).
Statistical analysis. Each experiment was repeated at least three times. Data are expressed as mean ± SEM. Mean values were compared with the Student's t-test. Expression of miRs in wild-type and mutant KRAS CRC tissue samples were analyzed with the Wilcoxon rank sum test. Expression levels of miRs in normal colonic mucosa and CRC tissue samples were analyzed with the Wilcoxon signed-rank test. Sequential cell proliferation and tumor growth were analyzed with two-way ANOVAs for repeated measures. Discrete variables were assessed with Fisher's exact test. P values ≤0.05 were considered statistically significant. Statistical analyses were performed with JMP software version 9.0 (SAS Institute, Cary, NC). Figure S1. Comparison of miRNA microarray results between KRAS G12V overexpressed and negative control transfected cells. Figure S2. Overexpression of KRAS gene decreased miR-4689 in HEK293 and MRC5. Figure S3. The knockdown effects of KRAS to the expression of miR-4689, the ERK phosphorylation, and cell proliferation in KRAS mutated DLD1 cell line. Figure S4. Immunoblots for the expression change of apoptosis-related molecules under knockdown of KRAS or AKT1. Figure S5. The effects of KRAS-or/and AKT1-overexpression on the inhibition of cell proliferation by miR-4689. Table S1. miRNAs down-regulated in HEK293-KRAS compared to HEK293-Ctrl. Table S2. miRNAs down-regulated in MRC5-KRAS compared to MRC5-Ctrl. Table S3. Relationship between KRAS mutations and clinicopathological factors.