miR-137 targets Cdc42 expression, induces cell cycle G1 arrest and inhibits invasion in colorectal...

miR-137 targets Cdc42 expression, induces cell cycle G1 arrestand inhibits invasion in colorectal cancer cells

Ming Liu1, Nan Lang1, Meng Qiu1, Feng Xu1, Qiu Li1, Qiulin Tang1, Ji Chen1, Xi Chen1, Siyuan Zhang1,

Zhen Liu1, Jitao Zhou1, Yajie Zhu1, Yu Deng1, Yi Zheng2 and Feng Bi1

1 Laboratory of Signal Transduction & Molecular Targeted Therapy, State Key Laboratory of Biotherapy/Department of Medical Oncology,

West China Hospital, Sichuan University, Chengdu, Sichuan Province, 610041, China2 Division of Experimental Hematology and Cancer Biology, Children’s Hospital Research Foundation, University of Cincinnati, Cincinnati, OH

miRNAs have emerged as post-transcriptional regulators that are critically involved in the pathogenesis of a number of human

cancers. Cdc42, one of the best characterized members of the Rho GTPase family, is found to be up-regulated in several types

of human tumors and has been implicated in cancer initiation and progression. In the present study, we have identified miR-

137 as a potential regulator of Cdc42 expression. A bioinformatics search revealed a putative target-site for miR-137 within

the Cdc42 30 UTR at nt 792–798, which is highly conserved across different species. Expression of miR-137 in colorectal

cancer cell lines was found inversely correlated with Cdc42 expression. miR-137 could significantly suppress Cdc42 30 UTRluciferase-reporter activity, and this effect was not detectable when the putative 30 UTR target-site was mutated. Consistent

with the results of the reporter assay, ectopic expression of miR-137 reduced both mRNA and protein expression levels of

Cdc42 and mimicked the effect of Cdc42 knockdown in inhibiting proliferation, inducing G1 cell cycle arrest, and blocking

invasion of the colorectal cancer cells, whereas anti-miR-137 expression led to the opposite effect. Furthermore, expression of

miR-137 suppressed the immediate downstream effector of Cdc42, PAK signaling. Our results suggest that miR-137 may have

a tumor suppressor function by directly targeting Cdc42 to inhibit the proliferation and invasion activities of colorectal cancer

cells. They raise an interesting possibility that Cdc42 activity and function can be controlled by miRNAs in addition to the

classic regulators such as guanine nucleotide exchange factors and GTPase-activating proteins.

IntroductionmicroRNAs (miRNAs) are a class of naturally occurringsmall, noncoding RNAs measuring 18–25 nucleotides inlength. More than 700 miRNAs have been annotated in thehuman genome thus far.1 miRNAs in conjunction with theeffects of Argonaute (Ago) result in a guide molecule in post-transcriptional gene silencing, which can either suppress

translation or degrade miR-bound mRNA by partially com-plementing with the 3 untranslated region (30 UTR) of thecognate mRNAs.2–4 Any given miRNA is likely to have mul-tiple gene targets. In fact, it is estimated that as high as 30%of protein-coding genes could be targets of miRNAs. BecausemiRNAs can target multiple transcripts, they are involved indiverse cell biological processes including cell differentiation,cell proliferation, apoptosis, metabolism, protein secretionand viral infection. Recent studies have also implicated miR-NAs in tumorigenesis and cancer progression.4–6 Increasingnumbers of miRNAs including miR-21,7 miR-200,8 miR-10,9,10 miR-1511,12 and miR-1611,12 have been reported to dis-play aberrant expressions in cancers and may be involved inmodulating cancer cell behaviors.

Rho GTPases form a subfamily of the Ras superfamily of20–30 kDa GTP-binding proteins that have been shown tomediate a plethora of cellular effects such as cytoskeletalreorganization, membrane trafficking, proliferation, apopto-sis/survival, cell polarity, cell adhesion, cell cycle and genetranscription.13,14 Cdc42 (cell division cycle 42), one of thebest-characterized members of the Rho GTPase family, hasbeen shown to be up-regulated in several types of humancancers, including testicular cancer,15 colorectal cancer,16

breast cancer,7 head and neck carcinoma18 and melanoma.19

The up-regulated Cdc42 activity may impair c-Cbl-mediated

Key words: miR-137, colorectal cancer, Cdc42, small G protein, cell

cycle

Abbreviations: Cdc42: cell division cycle 42; ERK1/2: extracellular

signal-regulated kinase 1/2; PAK1: p21 protein (Cdc42/Rac)-

activated kinase 1; miRNA: microRNA; MLC: myosin light chain;

Si-Cdc42: siRNA targeting cdc42 coding sequences

Grant sponsor: National Natural Science Foundation of China;

Grant number: 30672378

DOI: 10.1002/ijc.25452

History: Received 5 Dec 2009; Revised 14 Apr 2010; Accepted 29

Apr 2010; Online 12 May 2010

Correspondence to: Feng Bi, Laboratory of Signal Transduction &

Molecular Targeted Therapy/Department of Medical Oncology, West

China Hospital, Sichuan University, Chengdu, Sichuan Province,

610041, China, Tel: 86-28-85164044, Fax: 86-28-85164046, E-mail:

[email protected]

Can

cerCellBiology

Int. J. Cancer: 128, 1269–1279 (2011) VC 2010 UICC

International Journal of Cancer

IJC

EGFR degradation, contribute to EGFR hyperactivity andinduce proteasomal degradation of p21CIP1 leading to anincrease in cell proliferation and migration.20,21 In addition,downregulation of Cdc42 signals can inhibit anchorage-inde-pendent growth and induce apoptosis via the PI(3)K-Akt andErk signaling cascades and the p53 tumor suppressor.22 Ourprevious studies in primary gene targeted mammalian cellshave shown that Cdc42 is critically involved in actin filopodiaformation, cell motility, directional migration and cellgrowth. These functional outcomes may be through regula-tion of PAK1, GSK3b, MLC, ERK1/2, JNK and/or NF-jBpathways.23 Given these diverse functions and signalingevents dependent on Cdc42, Cdc42 may contribute to multi-ple stages of tumorigenesis and tumor progression.

To date, biochemical studies have identified a tightly regu-lated GTP-binding/GTP-hydrolysis cycle that is essential forthe regulation of Cdc42 activity. Interestingly, in humantumors, no mutations in Cdc42 coding sequences have beenfound, suggesting that either the regulation of GTP-bindingactivity of Cdc42 by the classic regulators such as guaninenucleotide exchange factors, GTPase-activating proteins andGDP-dissociation inhibitors is sufficient in modulating its ac-tivity, or additional mode of regulation in the expression levelmay be involved in controlling its activity and function. Thecurrent study is aimed at elucidating the regulatory mecha-nism of Cdc42 by miRNAs. Our data show that miR-137 cannegatively regulate Cdc42 expression and suppress the growthand invasion of colorectal cancer cells.

Material and MethodsCell culture

All cell lines were maintained in RPMI 1640 medium (Invi-trogen, Gaithersburg, MD, USA) supplemented with 10%heat-inactivated fetal bovine serum (FBS)(Gibco, Gaithers-burg, USA), 100 U/ml of penicillin G sodium, and 100 lg/mlstreptomycin sulfate (Sigma, Saint Louis, MO, USA) in ahumidified atmosphere containing 5% CO2 at 37 �C.

Transfection

Control oligo, hsa-miR-137, control anti-miR, anti-miR-137and Si-Cdc42 (siRNA targeting cdc42 coding sequences,50-CCUCUACUAUUGAGAAACU dTdT-30, 30-dTdT GGAGAUGAUAACUCUUUGA-50) were chemically synthesized byRiboBio (Guangzhou, China). RNA oligonucleotides weretransfected into cells using Lipofectamine 2000 (Invitrogen)according to the manufacturer’s protocol.

Plasmid construct

The 30 UTR of the human Cdc42 gene (NM_001791) wasPCR amplified from human genomic DNA using primers50-GCTCTAGAGCCCTTTCTGCACAGC-30 and 50-GCTCTAGAGCTAACCACTGGTTAATG-30, and cloned into theXba1-site of pGL3-control vector (Promega, Madison, WI,USA), which is designated pGL3-Cdc42-wt after sequencing.

Site-directed mutagenesis of the miR-137 target-site in theCdc42 30 UTR was carried out using site-directed mutagenesiskit (TaKaRa, Dalian, China), with pGL3-Cdc42-wt as a tem-plate, and named pGL3-Cdc42-mut (primers: FW, 50-GCTACTAGTTTAAATCAAACTAAAGAT-30, RV, 50-TTAGTATGATGCCGACACCAG-30).

Luciferase assay

Cells were seeded in 24-well plates 1 day before transfection.For reporter assays, the cells were transiently cotransfectedwith 0.3 lg wt or mutant reporter plasmid and 60 nM con-trol miRNA or miR-137 using lipofectamine 2000 (Invitro-gen). Firefly and Renilla luciferase activities were measuredconsecutively by using Dual Luciferase Assay (Promega)according to the manufacturer’s instructions, and normalizedfor transfection efficiency by the control vector containingRenilla luciferase, pRL-TK (promega). Three independentexperiments were performed in triplicate.

Cell growth assay

SW1116, Lovo and Colo320 colorectal cancer cells wereseeded in 96-well plates 1 day before transfection. After trans-fecting with miR-137, Si-Cdc42, control oligo, anti-miR-137 orcontrol anti-miR, the Cell Counting Kit-8 (Dojindo, Kuma-moto, Japan) was used to determine relative cell growth.

Cell cycle assay

Cells were harvested by trypsinization 72 hr after transfec-tion, washed three times with ice-cold PBS and fixed with70% ethanol overnight at 4�C. The fixed cells were rehy-drated in PBS and subjected to PI/RNase staining followedby fluorescence activated cell sorter scan (FACS) analysis(Becton Dickinson, Mountain View, CA, USA). The percent-age of cells in each phase of the cell cycle was estimatedusing ELITE software.

Matrigel invasion assay

Twenty-four hours after transfection, 4 �104 cells were sus-pended in 0.25 ml of culture medium with 1% FBS andplated in the top chamber with matrigel-coated membrane(24-well insert; pore size, 8 mm; Becton Dickinson). The cellswere incubated for 48 hr, after which the cells that did notinvade through the pores were removed by a cotton swab.Cells on the lower surface of the membrane were stainedwith Hematoxylin and Eosine for visualization, and counted.

Quantitative PCR analysis

Total RNA was extracted using Trizol (Invitrogen), treatedwith DNase I (Takara) to eliminate contaminating genomicDNA, and reverse-transcribed into cDNA with the ReverseTranscriptase M-MLV (TaKaRa). Real-time PCR was per-formed using a SYBR Premix Ex TaqTM kit (TaKaRa) on theiQ5 Real-Time PCR Detection System (Bio-Rad, Hercules, CA,USA). PCR primers used were as follows: Cdc42 FW, 50-gatggtgctgttggtaaa-30 and RV, 50-taactcagcggtcgtaat-30; b-actin FW,

Can

cerCellBiology

1270 miR-137 regulates Cdc42 expression


50-tgacgtggacatccgcaaag-30 and RV, 50-ctggaaggtggacagcgagg-30.For analysis of miRNA expression by qRT-PCR, reverse tran-scription and PCR were carried out using Bulge-LoopTM

miRNA qPCR Primer Set for hsa-miR-137 (RiboBio, MQP-0101), and U6 snRNA (RiboBio, MQP-0201) according tothe manufacturer’s instructions. Expression of Cdc42, relativeto b-actin and miR-137, relative to U6, was determined usingthe 2�DDCT method.

Western blot

Total-cell lysates were prepared using RIPA buffer (150 mMNaCl, 1% NP-40, 50 mM Tris-HCl (pH 7.4), 1 mM phenyl-methylsulfonyl fluoride, 1 lg/ml leupeptin, 1 mM deoxy-cholic acid and 1 mM EDTA) containing a cocktail of prote-ase inhibitors and phosphatase inhibitors (Calbiochem,Darmstadt, Germany). Equal amounts of protein sample (40–60 lg) was separated by 12% SDS-PAGE and transferred toPVDF membrane (Millipore, Bedford, MA, USA) using theBio-Rad semidry transfer system. The following antibodieswere used for Western blotting: Anti-Cdc42(Becton Dickin-son), Anti-b-actin (Santa Cruz, California, USA), anti-PAK1,anti-p44/42, anti-MLC, anti-cyclin D1, anti-Phospho-PAK1(Thr-423)/PAK2(Thr-402), anti-phospho-p44/42 MAPK

(Erk1/2) (Thr202/Tyr204), anti-phospho-MLC(Ser-473) (CellSignaling Technology, MA, USA). Blotted proteins weredetected and quantified using the ODYSSEY Infrared Imag-ing System (LI-COR Biosciences, Lincoln, NE, USA).

Statistical analysis

Assays for characterizing phenotypes of cells were analyzedby Student’s test or one-way ANOVA and correlationsbetween groups were calculated with Pearson. P values of<0.05 were deemed statistically significant. Data analysis wasachieved using SPSS for windows version 14.0 (SPSS, Chi-cago, USA).

ResultsThe Cdc42 30 UTR contains a putative target site for

miR-137 that is highly conserved across species

To associate miRNAs with the regulation of Cdc42 expres-sion, we performed a bioinformatic search for putative miR-NAs that are predicted to target cdc42 gene using PicTar.24

We found that the 30 UTR of human cdc42 harbors severalputative miRNA target sites, including those for hsa-miR-1,hsa-miR-15, hsa-miR-16, hsa-miR-18, hsa-miR-27, hsa-miR-29, hsa-miR-32, hsa-miR-33, hsa-miR-92, hsa-miR-101, hsa-

Figure 1. 30 UTR of cdc42 contains a putative target site for miR-137 that is highly conserved across species. (a) Schematic representation

of the Cdc42 mRNA putative sites targeted by miR-137. (b) Comparison of nucleotide sequences between the miR-137 seed sequences and

its target in different species. Sequence analysis indicated that miR-137 target sequence at nt 792–798 of the cdc42 30 UTR is highly

conserved across different species. The target-site for miR-137 within the cdc42 30 UTR at nt 792–798 (AAGCAAU) is underlined.

Can

cerCellBiology

Liu et al. 1271


miR-103, hsa-miR-107, hsa-miR-128, hsa-miR-133, hsa-miR-137, hsa-miR-141, hsa-miR-144, hsa-miR-185, hsa-miR-186,hsa-miR-195, hsa-miR-199, hsa-miR-204, hsa-miR-206, hsa-miR-211, hsa-miR-320 and hsa-miR-342. We initially focusedour studies on hsa-miR-101, hsa-miR-133 and hsa-miR-137,since these three miRNAs have been reported to be down-regulated in several types of cancer cell lines and tumors.However, in subsequent tests, we soon found that the level ofhsa-miR-101 or hsa-miR-133 has no correlation with Cdc42protein expression, and that hsa-miR-101 and hsa-miR-133have no effect on Cdc42 30 UTR luciferase-reporter activity.Additionally, both the Cdc42 protein and mRNA were notaffected by the ectopic expression of hsa-miR-101 and hsa-miR-133 (data not shown). Therefore, we focused our studieson hsa-miR-137 exclusively. A sequence analysis revealed

that the targeting sequence for miR-137 is located at nt 792–798 of the cdc42 30 UTR, and this region is highly conservedacross different species (Fig. 1).

Figure 3. The putative miR-137 binding site in cdc42 30 UTR is

functional in a luciferase assay. pGL3-Cdc42-wt or pGL3-Cdc42-mut

luciferase constructs containing a wild-type or a mutated Cdc42 30

UTR were cotransfected into SW1116, Lovo and Hela cells with

control miR or miR-137, respectively. Luciferase activity was

determined 48 hr after transfection. The ratio of normalized sensor

to control luciferase activity is shown. Data are shown as the mean

6 SD and were obtained from three independent experiments

performed in triplicate. * indicates a significant difference from

control miR-transfected cells (p < 0.05).

Figure 2. miR-137 and Cdc42 expressions are inversely correlated

in colorectal cancer cell lines. (a) Western-blot analysis for Cdc42

in 7 colorectal cancer cell lines. (b) The band intensities in a were

quantified with the ODYSSEY Infrared Imaging System (LI-COR

Biosciences). Data represent the mean 6 SD. from three

independent experiments. (c) qRT-PCR analysis for miR-137

expression in 7 colorectal cancer cell lines. The miR-137

expression was normalized to U6 expression using the 2�DDCT

method. Cdc42 protein expressions are inversely correlated with

miR-137 expressions (p ¼ 0.013). Data are representative of three

independent experiments performed in triplicate.

Can

cerCellBiology



Figure 4. miR-137 regulates the expression of Cdc42. (a) Western-blot analysis of SW1116 and Lovo cells 72 hr after transfected with miR-

137, Si-Cdc42 or control oligo for Cdc42 protein expression. (b) The band intensities in a were quantified with the ODYSSEY Infrared

Imaging System (LI-COR Biosciences). Data represent the mean 6 SD from three independent experiments. (c) qRT-PCR analysis of Cdc42

mRNA expression in SW1116 and Lovo cells 72 hr after transfected with miR-137, Si-Cdc42 or control oligo. Cdc42 mRNA expression was

normalized to b-actin mRNA expression, and data are shown as a ratio of miR-137- and Si-Cdc42-transfected cells to control oligo-

transfected cells using the 2�DDCT method. (d) qRT-PCR analysis of miR-137 in SW1116 and Lovo cells transfected with miR-137, Si-Cdc42

or control oligo. The miR-137 expression was normalized to U6 expression, and data are shown as a ratio of miR-137-and Si-Cdc42-transfected

cells to control oligo-transfected cells using the 2�DDCT method. Data are representative of three independent experiments performed in

triplicate. (e) Western-blot analysis of Colo320 cells 72 hr after transfected with anti-miR-137 or control anti-miR for Cdc42 protein expression.

(f) The intensities of the blots in e were quantified with the ODYSSEY Infrared Imaging System (LI-COR Biosciences). Data represent the

mean 6 SD from three independent experiments. * indicates a significant difference from control oligo-transfected cells (p < 0.05).

Can

cerCellBiology

Liu et al. 1273


The level of miR-137 is inversely correlated with Cdc42

protein expression in colorectal cancer cells

To examine whether miR-137 may functionally affect Cdc42expression, we first determined expression levels of miR-137and Cdc42 protein in 7 different colorectal cancer cell lines.In cell lines with low endogenous miR-137 expression, asmeasured by miRNA real-time PCR (i.e., Lovo, HT-29,COLO205, SW620, SW1116), a relatively high level of Cdc42protein was observed. Conversely, cell lines with relativelyhigh miR-137 expression (i.e., Colo320 and SW480) showedsignificantly lower amounts of Cdc42 protein when comparedwith those containing lower miR-137 (Fig. 2). These resultsfrom the available cell lines suggest a reciprocal relationshipbetween levels of miR-137 and Cdc42 protein expression inhuman colorectal cancer cells (p ¼ 0.013), justifying a furtherexamination of the role of miR-137. However, because of thelimited number of colorectal cancer cell lines available, a de-finitive conclusion on this issue needs to wait until a more

systemic study is done using larger numbers of primary colo-rectal cancer samples.

The putative 30 UTR target site of cdc42 is directly

regulated by miR-137

From the earlier observations, we hypothesize that cdc42 is atarget of miR-137. The 30 UTR of human cdc42 gene wascloned into the Xba1-site of pGL3-luciferase reporter vector(pGL3-control) (Promega) to test whether it might serve as adirect functional target of miR-137. This construct wasnamed pGL3-Cdc42-wt. In parallel, another luciferase re-porter construct in which the putative miR-137 targetingregion AAGCAAU, located within nt 792–798, was specifi-cally mutated and predicted to abolish miR-137 binding, wasdesignated pGL3-Cdc42-mut. Transient transfection ofSW1116 (Fig. 3a), Lovo (Fig. 3b) and HeLa (Fig. 3c) cellswith pGL3-Cdc42-wt and miR-137 led to a significantdecrease of luciferase activity when compared with the

Figure 5. miR-137 influences the proliferation potential of coloretal cancer cells. Proliferation potential of SW1116 (a) and Lovo (b) 24 hr

after transfected with miR-137, Si-Cdc42 or Control oligo, and Colo320 cells (c) 24 hr after transfection with anti-miR-137 or Control anti-

miR, was determined by the CCK-8 Assay. The data represent means 6 SD from 3 independent experiments performed in triplicate.

Can

cerCellBiology



control. The activity of the mutant reporter construct, how-ever, was unaffected by the cotransfection with miR-137.

Next, to examine whether miR-137 could affect Cdc42expression in colorectal cell lines, we transfected colorectalcell lines SW1116 and Lovo that show a relatively highCdc42 expression, with miR-137, a control oligo or Si-Cdc42.The protein and mRNA expression levels of Cdc42 were ana-lyzed by Western blotting and real-time PCR, respectively.As shown in Figure 4, compared with the negative control,transient expression of miR-137 led to a significant decreasein Cdc42 protein (Figs. 4a and 4b) and mRNA expressions(Fig. 4c), similar to that caused by transfection of Si-Cdc42in both cell lines. Furthermore, to determine whether trans-fection of cell lines with anti-miR-137 affects Cdc42 proteinexpression, Colo320 cells with a relatively high miR-137expression and low Cdc42 expression, were transiently trans-fected with anti-miR-137 or control-anti-miR. Down-regula-tion of endogenous miR-137 with anti-miR-137 led to a sig-nificant increase in Cdc42 protein expression in Colo320 cells(Figs. 4e and 4f). Taken together, these data suggest that the30 UTR of Cdc42 is a functional target site for miR-137 inthe colorectal cancer cells.

miR-137 regulates proliferation, invasion and G0/G1

progression of coloretal cancer cells

Cdc42 is known to play a key role in the regulation of cellcycle progression at the G1/S transition. We tested whetherthe cell growth potential of colorectal cancer cells transfectedwith miR-137 was inhibited as a consequence of Cdc42expression suppression. Figure 5 shows the results of CCK-8assays where the ectopic expression of miR-137 or Si-Cdc42markedly inhibits the proliferation potentials of SW1116 cells(Fig. 5a) and Lovo cells (Fig. 5b), and anti-miR-137 enhancesthe proliferation potentials of Colo320 cells (Fig. 5c). Thus,reducing Cdc42 expression in SW1116 and Lovo cells, eitherby miR-137 or Si-Cdc42, results in similar growth pheno-types of a marked inhibition of cell proliferation, whileincreasing Cdc42 expression by anti-miR-137 has an oppositeeffect on Colo320 cells.

To further understand the mechanisms by which cell pro-liferation is affected, flow cytometry was performed to ana-lyze the cell cycle phase distribution of these cells. The cellcycle progression of SW1116 cells (Fig. 6a) and Lovo cells(Fig. 6b) transfected with miR-137 or Si-Cdc42 was stalled atthe G1 phase with a significant decrease in S and G2/Mphases compared with cells transfected with control oligo.Furthermore, while the percentage of G0/G1 phase decreasedin Colo320 cells after transfection with anti-miR-137, theother phases increased accordingly (Fig. 6c). These resultsindicate that miR-137 can reduce cell proliferative potentialby mimicking Cdc42 reduction and support the role of thismiRNAs as a negative regulator of Cdc42 regulated cellgrowth.

In addition to proliferation regulation, Cdc42 is involvedin the invasive phenotype of tumor cells that may contribute

to the morbidity and mortality of patients with cancer. Todetermine whether miR-137 may regulate invasion, SW1116and Lovo cells were transfected with miR-137, Si-Cdc42 orcontrol oligo, and Colo320 cells were transfected with anti-miR-137 or a control anti-miR. In a matrigel invasion assay,we found a significant decrease in invasive activity of miR-137-transfected cells, mimicking that of Si-Cdc42-transfectedcells (Figs. 7a and 7b). In parallel, we observed a significantincrease in invasive activity of Colo320 cell upon inhibition

Figure 6. miR-137 induces cell cycle G1 arrest in coloretal cancer

cells. Cell cycle progression of SW1116 (a) and Lovo (b) 72 hr after

transfected with miR-137, Si-Cdc42 or Control oligo, and Colo320

(c) 72 hr after transfected with anti-miR-137 or Control anti-miR,

was determined by FACS analysis. Data are representative of 3

independent experiments. * indicates a significant difference from

control oligo-transfected control cells (p < 0.05).

Can

cerCellBiology

Liu et al. 1275


of miR-137 by anti-miRNA (Fig. 7c). These data suggest thatmiR-137 negatively regulates invasion of the colorectal cancercells, and this may, at least in part, be due to its targetingeffect on Cdc42.

Ectopic expression of miR-137 suppresses

Cdc42/PAK signaling

To determine whether the downstream signals of Cdc42 areaffected by ectopic expression of miR-137, SW1116 and Lovocells were transfected with miR-137, Si-Cdc42 or controloligo, and the phosphorylation and/or the expression levelsof PAK1 (p21 protein (Cdc42/Rac)-activated kinase 1), myo-sin light chain (MLC), extracellular signal-regulated kinase1/2 (ERK1/2) and cyclin D1 were examined by Western blot-ting. As shown in Figure 8, the phosphorylation levels ofPAK1, MLC, ERK1/2 and the expression of Cyclin D1, theknown downstream signals of Cdc42, decreased significantlyby miR-137 expression. These signaling effects again mim-icked those of Cdc42 siRNA knockdown, suggesting that thefunction of miR-137 in the cancer cells is likely, at least inpart, through the Cdc42 regulated PAK signaling pathway.This is consistent with the results of our previous study per-

formed in Cdc42-/- and Cdc42GAP-/- primary mouse em-bryonic fibroblasts (MEFs) produced by gene targeting as theCdc42 loss- or gain-of-activity cell model.23

DiscussionIncreasing studies in the past few years have shown thatmiRNAs could serve functionally as ‘‘oncogenes’’ or ‘‘tumorsuppressor genes’’ and regulate multiple cellular processes rel-evant to tumorigenesis and cancer progression. Despite inten-sive efforts in this area, only a handful of the cases studieshave identified the gene targets of specific miRNAs involvedin human tumor formation. Since it is important to definethe function and mechanism of miRNAs in oncogenesis ofdifferent types of human tumors, in this study, we focusedon the role of miR-137, which has been found to be down-regulated in colorectal cancer,25 glioblastoma multiforme26

and oral cancer.27

As a key member of the Rho GTPase family, Cdc42 isknown to be involved in proliferation, migration, invasion,apoptosis and angiogenesis, in several types of cancers.13,14

Our previous studies suggest that Cdc42 may contribute tothe hypoxia-mediated angiogenesis28 and is critically involved

Figure 7. miR-137 influences coloretal cancer cell invasion. The invasive activities of SW1116 (a) and Lovo (b) 24 hr after transfected with

miR-137, Si-Cdc42 or control oligo, and Colo320 (c) 24 hr after transfected with anti-miR-137 or Control anti-miR, were assayed in a

matrigel-coated transwell, and the cells that successfully invaded into the matrigel were quantified 48 hr after plating. Data are

representative of 3 independent experiments. * indicates a significant difference from control oligo-transfected cells (p < 0.05).

Can

cerCellBiology



Figure 8. Ectopic expression of miR-137 reduces PAK signaling similarly to that by Cdc42 knockdown. (a) Western-blot analysis of SW1116

and Lovo 72 hr after transfected with miR-137, Si-Cdc42 or control oligo for PAK1, MLC, ERK1/2, cyclin D1, phospho-PAK1, phospho-MLC,

phospho-ERK1/2. (b) The band intensities in a were quantified with the ODYSSEY Infrared Imaging System (LI-COR Biosciences). Data

represents the mean 6 SD from three independent experiments. * indicates a significant difference from control oligo-transfected control

cells (p < 0.05).

Can

cerCellBiology

Liu et al. 1277


in actin filopodia formation, cell motility, directional migra-tion and cell growth. Cdc42 is also required for the regula-tion of multiple signaling pathways, including PAK1, GSK3b,MLC, ERK1/2, JNK and NF-jB, in Cdc42�/� andCdc42GAP�/� primary mouse embryonic fibroblasts.23

Unlike Ras, that is found to be constitutively activated bypoint mutations in its coding sequence in human tumors, todate, no coding region sequence mutation has been detectedfor the cdc42 gene in human cancers.29 This suggests thatCdc42 is likely to be regulated by different mechanisms suchas overexpression, deregulation of the GTP-binding cycleand/or deregulation of its stability. To understand the contri-bution of Cdc42 to cancer progression, it is essential to betterdefine the regulatory mechanism of Cdc42.

Since we observed that the 30 UTR of human Cdc42 har-bors one putative site recognized by miR-137, and moreimportantly, miR-137 and Cdc42 expressions seem to beinversely correlated in several colorectal cancer cells, wehypothesized that Cdc42 may be directly regulated by miR-137 at the posttranscriptional level. By a luciferase-based re-porter assay, we show that one predicted miR-137 targetingsite in the Cdc42 30 UTR is functional, and that expression ofmiR-137 causes a significant decrease in both mRNA andprotein of Cdc42. Consistently, the proliferation and invasionactivities of the colorectal cancer cells were inhibited by miR-137 similar to that of Cdc42 siRNA, but were enhanced byanti-miR-137 that elevated Cdc42 expression. An examina-tion of the effects by miR-137 expression on downstream sig-naling of Cdc42 further confirm that miR-137 mimics Cdc42downregulation in several intracellular signaling events. To-gether, these functional and signaling findings strongly sug-gest that miR-137 can acts by downregulating Cdc42 expres-sion to impact on cancer cell behavior.

Cell cycle progression involves the sequential activation ofdifferent cyclin-dependent kinases (CDKs). The expression ofcyclins and other regulators of CDKs vary during cell cycle

progression and are often deregulated in cancers. Cdc42 hasbeen shown to contribute to the regulation of cell cycle pro-gression in G1/S-phase transition, but the signal transductionpathways involved are still unclear.30,31 Previous studies haveindicated miR-137 inhibits proliferation of glioblastoma mul-tiforme cells by an unknown mechanism.26 Our results mayprovide a clue for the further study of the mechanisminvolved. It is tempting to speculate that miR-137 mayachieve proliferative inhibition and cell cycle G1 arrest induc-tion via negatively regulating Cdc42 expression, which inturn affects expression and activities of downstream mole-cules of Cdc42, such as ERK1/2 and cyclin D1.

In addition to effects on the cell cycle, Cdc42 regulatescell cytoskeleton and adhesion, both of which are functionsthat are important for cell migration and invasion in severaltypes of cancer.13,16,17 To this end, the protein level of Cdc42is found to be elevated with high incidence (60%) in colo-rectal cancer.16 Cdc42 activation promotes lamellopodiumformation and concomitantly enhanced cell invasion inhuman colon cancer cells.32 In our study, ectopic expressionof miR-137 inhibits invasion in colorectal cancer cells, simi-lar to the effects of siRNA downregulation of Cdc42,whereas anti-miR-137 stimulates invasion in Colo320 cells.The phospho-PAK1 and -MLC are also suppressed by miR-137 and Cdc42 siRNA, suggesting that miR-137 may regu-late the cancer cell invasion through the Cdc42-PAK1-MLCpathway.

In conclusion, our studies show that miR-137 expressioncan markedly suppress Cdc42 expression and effectively in-hibit Pak-MLC or ERK signaling and the proliferation/inva-sion of colorectal cancer cells, mimicking the effects of Cdc42knockdown. Anti-miR-137 shows an opposite effect onCdc42 expression and cancer cell behaviors. The results sug-gest that miR-137 is a regulator of Cdc42 activity and mayserve as a useful tool in exploring the potential therapeuticbenefits of Cdc42 targeted therapy in colorectal cancer.

References

1. Griffiths-Jones S, Saini HK, van Dongen S,Enright AJ. miRBase: tools for microRNAgenomics. Nucleic Acids Res 2008;36:D154–D158.

2. Chi SW, Zang JB, Mele A, Darnell RB.Argonaute HITS-CLIP decodes microRNA-mRNA interaction maps. Nature 2009;460:479–86.

3. Bartel DP. MicroRNAs: genomics,biogenesis, mechanism, and function. Cell2004;116:281–97.

4. Lai EC. MicroRNAs are complementary to30 UTR sequence motifs that mediatenegative post-transcriptional regulation.Nat Genet 2002;30:363–4.

5. Wu W, Sun M, Zou GM, Chen J.MicroRNA and cancer: Current status andprospective. Int J Cancer 2007;120:953–60.

6. Calin GA, Croce CM. MicroRNAsignatures in human cancers. Nat RevCancer 2006;6:857–66.

7. Asangani IA, Rasheed SA, Nikolova DA,Leupold JH, Colburn NH, Post S, AllgayerH. MicroRNA-21 (miR-21) post-transcriptionally downregulates tumorsuppressor Pdcd4 and stimulates invasion,intravasation and metastasis in colorectalcancer. Oncogene 2008;27:2128–36.

8. Korpal M, Lee ES, Hu G, Kang Y. ThemiR-200 family inhibits epithelial-mesenchymal transition and cancer cellmigration by direct targeting of E-cadherintranscriptional repressors ZEB1 and ZEB2.J Biol Chem 2008;283:14910–4.

9. Sasayama T, Nishihara M, Kondoh T,Hosoda K, Kohmura E. MicroRNA-10b isoverexpressed in malignant glioma and

associated with tumor invasive factors,uPAR and RhoC. Int J Cancer 2009;125:1407–13.

10. Li Ma, Julie Teruya-Feldstein, Robert A.Weinberg. Tumour invasion and metastasisinitiated by microRNA-10b in breastcancer. Nature 2007;449:682–8.

11. Cimmino A, Calin GA, Fabbri M, IorioMV, Ferracin M, Shimizu M, Wojcik SE,Aqeilan RI, Zupo S, Dono M, Rassenti L,Alder H, Volinia S, Liu CG, Kipps TJ,Negrini M, Croce CM. miR-15 and miR-16induce apoptosis by targeting BCL2. ProcNatl Acad Sci 2005;102:13944–9.

12. Xia L, Zhang D, Du R, Pan Y, Zhao L, SunS, Hong L, Liu J, Fan D. miR-15b andmiR-16 modulate multidrug resistance bytargeting BCL2 in human gastric cancercells. Int J Cancer 2008;123:372–9.

Can

cerCellBiology



13. Etienne-Manneville S, Hall A. RhoGTPases in cell biology. Nature 2002;420:629–35.

14. Sahai E, Marshall CJ. RHO-GTPases andcancer. Nat Rev Cancer 2002;2:133–42.

15. Kamai T, Yamanishi T, Shirataki H, TakagiK, Asami H, Ito Y, Yoshida K.Overexpression of RhoA, Rac1, and Cdc42GTPases is associated with progression intesticular cancer. Clin Cancer Res 2004;10:4799–805.

16. Gomez Del Pulgar T, Valdes-Mora F,Bandres E, Perez-Palacios R, Espina C,Cejas P, Garcıa-Cabezas MA, Nistal M,Casado E, Gonzalez-Baron M, Garcıa-Foncillas J, Lacal JC. Cdc42 is highlyexpressed in colorectal adenocarcinomaand downregulates ID4 through anepigenetic mechanism. Int J Oncol 2008;33:185–93.

17. Fritz G, Just I, Kaina B. Rho GTPases areover-expressed in human tumors. Int JCancer 1999;81:682–7.

18. Abraham MT, Kuriakose MA, Sacks PG,Yee H, Chiriboga L, Bearer EL, DelacureMD. Motility-related proteins as markersfor head and neck squamous cell cancer.Laryngoscope 2001;111:1285–9.

19. Tucci MG, Lucarini G, Brancorsini D,Zizzi A, Pugnaloni A, Giacchetti A, RicottiG, Biagini G. Involvement of E-cadherin,beta-catenin, Cdc42 and CXCR4 in theprogression and prognosis of cutaneousmelanoma. Br J Dermatol 2007;157:1212–6.

20. Hirsch DS, Shen Y, Wu WJ. Growth andmotility inhibition of breast cancer cells byepidermal growth factor receptordegradation is correlated with inactivationof Cdc42. Cancer Res 2006;66:3523–30.

21. Bao W, Thullberg M, Zhang H,Onischenko A, Stromblad S. Cellattachment to the extracellular matrixinduces proteasomal degradation ofp21(CIP1) via Cdc42/Rac1 signaling. MolCell Biol 2002;22:4587–97.

22. Zugasti O, Rul W, Roux P, Peyssonnaux C,Eychene A, Franke TF, Fort P, Hibner U.Raf-MEK-Erk cascade in anoikis iscontrolled by Rac1 and Cdc42 via Akt. MolCell Biol 2001;21:6706–17.

23. Yang L, Wang L, Zheng Y. Gene targetingof Cdc42 and Cdc42GAP affirms thecritical involvement of Cdc42 in filopodiainduction, directed migration, andproliferation in primary mouse embryonicfibroblasts. Mol Biol Cell 2006;17:4675–85.

24. Krek A, Grun D, Poy MN, Wolf R,Rosenberg L, Epstein EJ, MacMenamin P,da Piedade I, Gunsalus KC, Stoffel M,Rajewsky N. Combinatorial microRNAtarget predictions. Nat Genet 2005;37:495–500.

25. Bandres E, Agirre X, Bitarte N, Ramirez N,Zarate R, Roman-Gomez J, Prosper F,Garcia-Foncillas J. Epigenetic regulation ofmicroRNA expression in colorectal cancer.Int J Cancer 2009;125:2723–43.

26. Silber J, Lim DA, Petritsch C, Persson AI,Maunakea AK, Yu M, Vandenberg SR,

Ginzinger DG, James CD, Costello JF,Bergers G, Weiss WA, Alvarez-Buylla A,Hodgson JG. miR-124 and miR-137 inhibitproliferation of glioblastoma multiformecells and induce differentiation of braintumor stem cells. BMC Med 2008;6:14.

27. Kozaki K, Imoto I, Mogi S, Omura K,Inazawa J. Exploration of tumor-suppressive microRNAs silenced by DNAhypermethylation in oral cancer. CancerRes 2008;68:2094–105.

28. Xue Y, Bi F, Zhang X, Zhang S, Pan Y, LiuN, Shi Y, Yao X, Zheng Y, Fan D. Role ofRac1 and Cdc42 in hypoxia induced p53and von Hippel-Lindau suppression andHIF1alpha activation. Int J Cancer 2006;118:2965–72.

29. Rihet S, Vielh P, Camonis J, Goud B,Chevillard S, de Gunzburg J. Mutationstatus of genes encoding RhoA, Rac1, andCdc42 GTPases in a panel of invasivehuman colorectal and breast tumors. JCancer Res Clin Oncol 2001;127:733–8.

30. Catherine F. Welsh. Rho GTPases as keytransducers of proliferative signals in G1cell cycle regulation. Breast Cancer ResTreat 2004;84:33–42.

31. Jaffe AB, Hall A. Rho GTPases:biochemistry and biology. Annu Rev CellDev Biol 2005;21:247–69.

32. Jaffe T, Schwartz B. Leptin promotesmotility and invasiveness in human coloncancer cells by activating multiple signal-transduction pathways. Int J Cancer 2008;123:2543–56.

Can

cerCellBiology

Liu et al. 1279


miR-137 targets Cdc42 expression, induces cell cycle G1 arrest and inhibits invasion in colorectal...

Documents

Transcript of miR-137 targets Cdc42 expression, induces cell cycle G1 arrest and inhibits invasion in colorectal...