mirna-491-5p and git1 serve as modulators and biomarkers ... · mirna-491-5p and git1 serve as...

Molecular and Cellular Pathobiology

miRNA-491-5p and GIT1 Serve as Modulators andBiomarkers for Oral Squamous Cell CarcinomaInvasion and Metastasis

Wei-Chieh Huang1,3, Shih-Hsuan Chan1, Te-Hsuan Jang1, Jer-Wei Chang1, Ying-Chin Ko4, Tzu-Chen Yen6,Shang-Lun Chiang4, Wei-Fan Chiang8, Tien-Yu Shieh9, Chun-Ta Liao7, Jyh-Lyh Juang1, Hsueh-Chun Wang2,Ann-Joy Cheng5, Ya-Ching Lu5, and Lu-Hai Wang1,3

AbstractMicroRNAs offer tools to identify and treat invasive cancers. Using highly invasive isogenic oral squamous cell

carcinoma (OSCC) cells, established using in vitro and in vivo selection protocols frompoorly invasive parental cellpopulations, we used microarray expression analysis to identify a relative and specific decrease in miR-491-5p ininvasive cells. Lower expression of miR-491-5p correlated with poor overall survival of patients with OSCCs. miR-491-5p overexpression in invasive OSCC cells suppressed their migratory behavior in vitro and lung metastaticbehavior in vivo. We defined the G-protein—coupled receptor kinase-interacting protein 1 (GIT1)—as a directtarget gene formiR-491-5p control. GIT1 overexpressionwas sufficient to rescuemiR-491-5p–mediated inhibitionof migration/invasion and lung metastasis. Conversely, GIT1 silencing phenocopied the ability of miR-491-5p toinhibit migration/invasion and metastasis of OSCC cells. Mechanistic investigations indicated that miR-491-5poverexpression or GIT1 attenuation reduced focal adhesions, with a concurrent decrease in steady-state levels ofpaxillin, phospho-paxillin, phospho-FAK, EGF/EGFR-mediated extracellular signal–regulated kinase (ERK1/2)activation, andMMP2/9 levels and activities. In clinical specimens of OSCCs, GIT1 levels were elevated relative topaired normal tissues and were correlated with lymph nodemetastasis, with expression levels of miR-491-5p andGIT1 correlated inversely inOSCCs, where they informed tumor grade. Together, ourfindings identify a functionalaxis for OSCC invasion that suggests miR-491-5p and GIT1 as biomarkers for prognosis in this cancer. Cancer Res;74(3); 751–64. �2013 AACR.

IntroductionOral squamous cell carcinoma (OSCC) is the sixth most

common cancer worldwide and accounts for more than 95% ofall head and neck cancer (1). The incidence of oral cancer inTaiwan increased by 30% in the last 5 years, mortality rate grew

by 25% with 30 to 49 years old men having the highest rate.More than 44,000 patients annually require sustained medicaltreatment (2, 3). Unlike other types of solid tumors, themetastasis of OSCCs mostly involves local invasion and oftenis restricted to the head and neck area. Cervical lymph nodemetastasis is a critical prognostic factor for OSCCs, thepatients without such metastasis usually have higher survivalrates (4–6). Like other cancers, oral cancer metastasis requiresan extensive remodeling and degradation of extracellularmatrix (ECM), in part, via increased expression of matrixmetalloproteinases (MMP; ref. 7). However, the molecularmechanism regulating the invasion and metastasis of OSCCsis still largely unclear.

MicroRNAs (miRNA) are an evolutionarily conserved groupof small RNAs of 18 to 24 nucleotides that inhibit or stimulategene expression. microRNAs have been known to play impor-tant roles in various cancer progression and metastasis (8, 9);however, study of their roles in OSCC metastasis is relativelyscarce (10).

GIT1 is a multifunctional scaffold protein found to beassociated with paxillin and capable of stimulating lamellipo-dia formation and spreading of cells (11–13). Although theGIT1/paxillin complex is known to play a role in regulatingfocal adhesion formation and cell migration (14), the preciserole of GIT1 in metastasis of cancer especially in OSCCs is

Authors' Affiliations: 1Institute of Molecular and Genomic Medicine,2Division of Environmental Health and Occupational Medicine, NationalHealth Research Institute, Miaoli; 3Department of Life Sciences, NationalCentral University, Jhongli; 4Environment-Omics-Disease Research Cen-ter, China Medical University Hospital, Taichung; 5Departments of MedicalBiotechnology, Chang Gung University; 6Nuclear Medicine and MolecularImaging Center, Chang Gung Memorial Hospital; 7Head and Neck Oncol-ogy Group, Chang GungMemorial Hospital, Taoyuan; 8Department of Oraland Maxillofacial Section, Chi-Mei Medical Center, Liouying; and 9Depart-ment of Oral Hygiene, Kaohsiung Medical University, Kaohsiung, Taiwan

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

W.-C. Huang and S.-H. Chan contributed equally to this work and are co-first authors.

Corresponding Author: Lu-Hai Wang, National Health Research Instituteand National Central University, No. 35, Keyan Road, Zhunan, MiaoliCounty 35053, Taiwan. Phone: 886-37-246166, ext. 31010; Fax: 886-37-585242; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-1297

�2013 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 751

on May 9, 2020. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 12, 2013; DOI: 10.1158/0008-5472.CAN-13-1297

http://cancerres.aacrjournals.org/

unclear. Focal adhesion kinase (FAK) is a ubiquitouslyexpressed non–receptor protein tyrosine kinase (PTK). Previ-ous studies have indicated that FAK functions as a positiveregulator of tumor invasion and is overexpressed in variouscancers, including breast, head and neck, and ovarian cancers(15, 16). FAK/paxillin interaction controls cell motility in partthrough extracellular signal–regulated kinase (ERK1/2) acti-vation (17). The EGF receptor (EGFR) is a ubiquitous receptortyrosine kinase that is often upregulated in head and neckcancer (18). Activation of EGFR signaling is known to promotecancer cell proliferation and metastasis (19, 20). A previousreport showed that GIT1 linked EGFR-mediated activation ofERK1/2 in 293 cells (21). However, the role of EGFR/GIT1/FAKsignaling in oral cancer is unknown.

In this study, we found that miR-491-5p was substantiallydownregulated in 5 highly invasive OSCC lines and was alsowidely attenuated in OSCCs compared with paired normaltissues. Our data show that miR-491-5p targets GIT1 to inhibitOSCC cell focal adhesion formation, invasion, and metastasisvia regulation of FAK, paxillin, ERK1/2, andMMP2/9. Thus, thisstudy suggests that miR-491-5p is a metastasis suppressor andthat miR-491-5p and GIT1 are significant biomarkers for OSCCmetastasis.

Materials and MethodsTissue samples acquisition

All clinical samples were acquired via protocols approved bythe respective Institutional Review Boards. Thirty-three oralnormal and tumor tissue pairswere obtained from theOral andMaxillofacial Surgery Unit, Chi Mei Medical Center, Tainan,Taiwan, and Department of Public Health and EnvironmentalMedicine, School of Medicine, College of Medicine, KaohsiungMedical University, Taiwan. Tumor and normal oral tissuesamples from surgical resection were snap-frozen in liquidnitrogen or embedded in RNAlater. The clinical characteristicsof these 33 N/T pairs are shown in Supplementary Table S1.Forty-eight oral cancer samples for cDNA microarray analysis(22) and 189 OSCC tissues array slides were obtained fromNuclear Medicine, Chang Gung Memorial Hospital at LinkouTaiwan. Commercial tissue array slides containing 134 OSCCtissues were purchased from US Biomax Inc.

Cell culture, DNA and RNA transfections, and stable cellline generation

Humanoral cancer cell lines CGHNC9, SAS, SCC25, OECM-1,and OC-3 have been described (23). CGHNC9, OECM-1, andSAS cells were grown in Dulbecco's Modified Eagle Medium(DMEM; Invitrogen) with 10% FBS (Biological Industries).SCC25 cells were grown in DMEM/HAM F12 (Invitrogen)supplemented with 10% FBS and hydrocortisone (0.4 mg/mL,Sigma-Aldrich). OC-3 cells were grown in 1:1mixture of DMEMand KSFM (Invitrogen) with 10% FBS. All cells were incubatedat 37�C in 5% CO2. Cells were transfected using the LipofectA-MINE 2000 (Invitrogen) and LipofectAMINE RNAiMAX (Invi-trogen). ThemiR-491-5p cell lines stably expressing pcDNA6.2-EmGFP-miRcon or pcDNA6.2-EmGFP-miR-491-5p were estab-lished by transfection with the respective plasmids followed byselectionwith blasticidin (Sigma-Aldrich). ThemiR-491-5p and

GIT1 cell lines stably expressing pcDNA6.2-GFP-miR-491-5p orpcDNA-flag-GIT1 were established by transfection with therespective plasmids followed by selection with neomycin(Sigma-Aldrich).

Vectors, antibodies, and reagentsThe miR-491-5p was constructed in pcDNA6.2-EmGFP.

GIT1 coding sequence was cloned in pCMV-HA- andpcDNA3-flag. The luciferase-30UTR-wt reporter or luciferase-30UTR-mutant (mt) plasmids were prepared by inserting theGIT1-30-UTR-wt carrying a putativemiR-491-5p binding site oritsmutant sequence into the pGL3-control plasmid (sequencesare shown in Supplementary Table S2). N-terminally (1–676amino acid) truncated FAK (dominant-negative FAK) anddominant-negative MEK (K97A mutant) were constructed inpcDNA3. Antibodies for Western blotting and immunohis-tochemistry (IHC) are mouse anti-GIT1 from BDTransductionLaboratories and Bethyl Laboratories, Inc. The primers (forplasmid construction), antibodies, and reagents used aredescribed in the Supplementary Material and Table S2.

30UTR reporter assayspGL3-GIT1-30UTR-wt or pGL3-GIT1-30UTR-mt was co-

transfected with pre-miR-491-5p or pre-anti-miR-491-5p intoOSCC cells. Luciferase assay was conducted using an assay kit(Promega). Renilla luciferase was cotransfected as a controlfor normalization.

Reverse transcriptase PCR and quantitative reversetranscriptase PCR

Reverse transcriptase (RT)-PCR and quantitative RT (qRT)-PCR were used to detect the miR-491-5p and mRNA expres-sion. We designed a stem-loop RT primer specifically hybrid-izing with miR-491-5p or RNU6B, respectively, the latter wasused for normalization (24). The detailed conditions andprimers used for mRNA expression are listed in the Supple-mentary Material and Table S2.

Cell chemotatic migration and invasion assayChemotatic migration and invasion ability of OSCC cells

were assessed using the Falcon Cell Culture Inserts with orwithout Matrigel (BD Biosciences) coating as described (25).The detailed conditions are described in the SupplementaryMaterial.

In vivo metastasis assaysCB17-SCID mice were used for injection of OSCC cells

orthotopically at oral buccal mucosa or via tail vein andmonitored for 21 to 42 days before sacrifice. Lung tissues wereremoved, fixed, paraffin-embedded, serially sectioned, and sub-jected to hematoxylin and eosin (H&E) staining. The detailedconditions are described in the Supplementary Material.

Focal adhesion assay and immunofluorescencemicroscopy

Immunostaining was conducted to detect the vinculin leveland focal adhesions in C9-lung-IV2 cells. The detailed condi-tions are described in the Supplementary Material.

Huang et al.

Cancer Res; 74(3) February 1, 2014 Cancer Research752




Western blottingDetailed procedure is described in the SupplementaryMate-

rials and Methods.

Paxillin degradation assaysC9-lung-IV2 cells were transfected with negative control

(NC) or GIT1-siRNA-3 and incubated for 48 hours, and MG132at 10 mmol/L was added to the cells for 8 hours. C9-lung-IV2cells were transfected with negative control or GIT1-siRNA,and cycloheximide (Sigma-Aldrich) at 20 mg/mL was added tothe cells for 0, 4, and 12 hours.

IHC and FISHIHCwas conducted to detect GIT1 expression fromparaffin-

embedded oral cancer specimens. The slides were stained withprimary antibody using an automatic slide stainer BenchMarkXT (Ventana Medical Systems). FISH was conducted using thetyramide signal amplification (TSA) technology and the50biotin-labeled miR-491-5p probe or a negative control probe[25 (nmol/L)/L; GeneDireX, Inc.] as previously described (26).The IHC score of GIT1 and FISH score of miR-491-5p for eachspecimen was graded as follows: þ, weak; þþ, moderate;þþþ, strong.

Gelatin zymographyGelatin zymography was used to detect MMP2/9 activity in

C9-lung-IV2 cells using conditions detailed in the Supplemen-tary Material.

Statistical analysisSurvival data was analyzed using the Kaplan–Meiermethod.

Differences between experimental groupswere calculatedwiththe Mann–Whitney U test. Differences with P < 0.05 areconsidered to be statistically significant. The linear correlationcoefficient (Spearman's r) was used to estimate the relation-ship between miR-491-5p and GIT1 levels in the oral tumorspecimens.

ResultsEstablishment of isogenic pairs of high and low invasiveOSCC linesRecent evidence has shown that microRNAs play a role in

regulating cancer metastasis (8). To identify relevant micro-RNAs associated with invasive phenotypes of oral cancer, weestablished isogenic pairs of high and low invasive oral cancerlines through in vitro and in vivo selections. Using Boydenchamber invasion assay (in vitro selection), the oral cancer celllines OC-3, OECM-1, SAS, and SCC25were selected 5 to 8 cyclesto derive the highly invasive sublines OC-3-I5, OECM-1-I8, SAS-I5, and SCC25-I6 in which the number following I denotes thecycles of selection. OC-3 andCGHNC9 lineswere selected 1 to 2cycles by in vivo injection of the respective cells into tail vein ofCB17-SCID mice (in vivo selection), followed by isolation oftumor cells grown from lymph node metastases or lungmetastases to obtain OC-3-I5-lymph-node-IV1, OC-3-I5-lymph-node-IV2, OC-3-lung-IV1, OC-3-I5-lung-IV2, andCGHNC9-lung-IV2 (C9-lung-IV2) where the number following

IV denotes the cycle selection. After invasion selection, wetested those lines for their migratory and invasive ability byconducting Boyden chamber migration/invasion assays. Asshown in Supplementary Fig. S1, the migratory and invasiveability of the selected invasive lineswas dramatically increased.

Downregulation of miR-491-5p was found in the highlyinvasive OSCC cells and was correlated with poorsurvival of the OSCC patient

UsingmicroRNA array, we attempted to identifymicroRNAsassociated with the invasive phenotype of the 5 selected highlyinvasive OSCC lines (SAS-I5, OECM-1-I8, SCC25-I6, OC-3-I5-,and OC-3-I5-lung-IV2) compared with their correspondingparentals. A series of microRNAs with 2-fold differentialexpression between the low and high invasive isogenic lineswere identified. The differentially expressed microRNAs ofeach isogenic pair as shown in each circle and the microRNAcommon among all 5 isogenic pairs are displayed by the Venndiagrams (Fig. 1A). The miR-491-5p has a greater than 2-foldlower expression in all 5 in vitro and in vivo selected invasivelines (Fig. 1A). The reduced expression of miR-491-5p wasverified by RT-PCR, the in vivo selected invasive lines displayeda greater decrease of the microRNA compared with theirparentals (Fig. 1B and C). Furthermore, we examined 33pairs of oral normal and cancer tissues for the miR-491-5pexpression. The miR-491-5p was underexpressed in 29 of 33OSCC samples compared with their matched normal tissues(Fig. 1D).

Kaplan–Meier survival analysis of 189 cases showed thatOSCC samples with lowmiR-491-5p expression correlatedwitha significantly poorer survival than those with highmiR-491-5pexpression (Fig. 1E, top). The miR-491-5p low expression alsohad a correlation with lymph node metastasis (Fig. 1E, bot-tom). The results suggest that miR-491-5p may play a role insuppressing OSCC lymph nodemetastasis and could serve as aprognostic marker for OSCCs.

miR-491-5p inhibits migration, invasion, and lungmetastasis of OSCC cells

Next, we tested whether miR-491-5p was able to suppressmigration and invasion of oral cancer cells by Boyden chamberassay. As expected, miR-491-5p significantly inhibited migra-tion and invasion ofOSCC cells, which could be partially or fullyrescued by anti-miR-491-5p (Fig. 2A). To evaluate whethermiR-491-5p inhibited cancer cell metastasis in vivo, we used2 experimental metastasis models, namely, tail vein injectionand orthotopic implantation of cells in SCID mice. In the tailvein injection model, expression of miR-491-5p resulted ingreatly decreased targeting of the C9-lung-IV2 cells to lung 24hours after injection (Supplementary Fig. S2). As shown in Fig.2B, the percentage of lung tumor area in the miR-491-5ptransiently transfected group was reduced to 45% of thecontrol. Similarly, in the orthotopic model (Fig. 2C), stableoverexpression of miR-491-5p significantly suppressed lungmetastasis of C9-lung-IV2 OSCC cells. miR-491-5p resulted infewer metastatic foci and small nodules in lung. These resultsindicate that miR-491-5p is able to potently inhibit migration,invasion, and metastasis of OSCC cells.

miRNA-491-5p Suppresses OSCC Metastasis

www.aacrjournals.org Cancer Res; 74(3) February 1, 2014 753




GIT1 is a direct target of miR-491-5pTo understand the mechanism of the miR-491-5p regulated

OSCC cell invasion andmetastasis, we attempted to identify itstarget genes by using 2 computational software, TargetScanand DIANA-microT. Those methods identified 17 candidategenes to be potential targets of miR-491-5p (SupplementaryTable S3). We were interested in those involved in regulating

focal adhesion complexes as they had been implicated incancer cell invasion and metastasis (15, 27). Using GeneOntology (GO) analysis, we found a putative target of miR-491-5p, GIT1, which belongs to the focal adhesion related genefamily. To determine whether GIT1 is a direct target of miR-491-5p, we constructed Luc-GIT1-UTR-wild-type (wt) and its30UTR mutant (mt) plasmids. Luciferase reporter assays

SSC25-I6(41)

SAS-I5(122)

miR-491-5p1

A D

B

C

E

33 normal/tumor pairs

**

Normal

P = 0.017

P = 0.001

0 20 40 60 80

Time (months)

(+)

1 95

14 79

(–)

Lymph node metastasis

100 120

miR-491-5p

high expression

miR-491-5p

low expression

miR-491-5p

high expression

miR-491-5p

low expression

OSCC

Ove

rall

surv

ivalmiR-491-5p

U6

miR-491-5p

U6

SCC25

Pare

nta

lP

are

nta

l

Pare

nta

l

Lung-I

V1

I5-L

ung-I

V1

I5-L

ung-I

V2

Lung-I

V2

I5-L

ym

ph-n

ode-I

V1

I5-L

ym

ph-n

ode-I

V2

Pare

nta

l

Pare

nta

l

Pare

nta

l

I5 I5 I6I8

SASOECM-1OC-3

OC-3 CGHNC9

In vitro selected OSCC cell lines

In vivo selected OSCC cell lines

0.25

0.20

0.15

0.10

0.05

0.00

1.0

0.8

0.6

0.4

0.2

0.0

Rela

tive

miR

-491-5

p e

xpre

ssio

n

(miR

-491-5

p/R

NU

6B

)

OECM1-I8(34)

OC3-I5(51)

OC3-I5-IV2(62)

Figure 1. The miR-491-5p expression is downregulated in highly invasive OSCC lines and oral cancers. A, decreased miR-491-5p expression is sharedby 5 highly invasive OSCC lines. Venn diagram shows number of differentially expressed miRNAs of each isogenic pair and the single downregulatedmiRNA common to all 5 pairs. The expression of miRNAs for each highly invasive OSCC line was normalized to the corresponding parental cell line.B andC, expression level ofmiR-491-5p in the in vitro and 2 series of in vivo selected highly invasive cell lineswasmeasured byRT-PCR. TheU6 small nuclearRNAwas used as an internal control. D, qRT-PCR reveals downregulation ofmiR-491-5p in oral tumors comparedwith their matched normal oral tissues. Therelative amount of miR-491-5p to small nuclear RNU6B RNA was calculated using the equation 2�DCt, where DCT ¼ (CT_miR-491-5P � CT_RNU6B RNA).E, top, Kaplan–Meier survival curves of 189 patients with OSCCs divided by miR-491-5p expression. Bottom, association between miR-491-5p expressionlevel and lymph nodemetastasis status of 189 patients withOSCCs as evaluated by Spearman's r. The level ofmiR-491-5pwas determined by FISH. Tissuesexpressing miR-491-5p at levels lower than median level were assigned to the low and those above the median level were assigned to the highexpression group. ��, P < 0.01.

Huang et al.





showed that miR-491-5p suppressed more than 50% transcrip-tional activity of the Luc-GIT1-30UTR-wt reporter comparedwith the control, whereas the GIT1-30UTR-mt where the puta-tivemiR-491-5p binding site hadbeenmutatedwas refractile tothe inhibition (Fig. 3A). We also found that miR-491-5p sig-nificantly inhibited the GIT1 mRNA expression in CGHNC9and SCC25 cells (Fig. 3B), which could be partially reversed byanti-miR-491-5p. As expected, Western blotting showed that

overexpression of miR-491-5p dramatically suppressed GIT1protein expression in CGHNC9 and SCC25 cells and theinhibition was substantially reverted by anti-miR-491-5p (Fig.3C). Consistently, the in vivo selected lines with lowermiR-491-5p expression showed elevated GIT1 mRNA levels (Supple-mentary Fig. S3). Those results suggest that miR-491-5p direct-ly targets GIT1, resulting in downregulation of its mRNA andprotein.

miR-491-5p

miR-491-5p

+

Anti-491-5p

Anti-491-5p

miR-cont

A B

C

miR-491-5p

miR-491-5p+

Anti-491-5p

miR-cont

miR-491-5p

miR-cont

Anti-491-5p

miR-491-5pmiR-491-5p

n = 5

n = 5

miR-cont

miR-cont

miR-491-5pmiR-cont

miR-491-5pmiR-cont

miR-491-5pmiR-cont

T

T

T

T

T

T

TT

Lu

ng

me

tasta

sis

in

dex (

%)

SCC25-I6OC-3 I5 lung IV2CGHNC9

SCC25-I6OC-3 I5 lung IV2CGHNC9

* ***

****

** **

******

***

*

*

2

1.5

1

0.5

0

120

100

80

60

40

20

0

Lung m

eta

sta

sis

index (

%) 120

100

80

60

40

20

0

Re

lative

mig

ratio

n a

bili

ty

1.2

1

0.8

0.6

0.4

0.2

0Rela

tive

inva

sio

n a

bili

ty

Figure 2. The miR-491-5p inhibits migration, invasion, and metastasis of OSCC cells. A, transient transfection of miR-491-5p significantly inhibited migrationand invasion ofCGHNC9,OC-3-I5-lung-IV2, andSCC25-I6 cells, whichwas rescued by anti-miR-491-5p. B, lungmetastasis via tail vein injection of C9-lung-IV2 cells intoSCIDmicewas significantly reduced in the cells transiently transfectedwithmiR-491-5p. The transfection efficiency as judgedby fluorescence ofFAM-labeled miR-491-5p was over 90% (Supplementary Fig. S11). C, lung metastasis via orthotopic implantation of C9-lung-IV2 cells at oral buccalmucosa was dramatically reduced in the cells stably expressing miR-491-5p precursor. The images show H&E staining of the lung metastatic tumors(�200). In tail vein injection model, lung metastasis was detected in all mice in 3 to 4 weeks, whereas in an orthotopic model, on the average fromthree independent experiments, 80% of mice developed lung metastasis in 6 to 8 weeks. Lung metastasis index was calculated as follows: metastatictumor areas/total lung areas. T, metastatic tumor areas. Histograms represent means � SD from 3 independent experiments. �, P < 0.05; ��, P < 0.01.






Suppression of GIT1 by siRNA inhibits OSCC cellmigration, invasion, and metastasis

GIT1 is a multidomain scaffold protein that is usually foundto be associated with focal adhesions and is able to promotecell motility (28). We found that GIT1 was upregulated in thehighly invasive oral cancer lines compared with their parentallines (Supplementary Fig. S4). To assess whether GIT1 levelcorrelated withmetastasis, we analyzed cDNAmicroarray dataof a series of 48 OSCC samples.Wefirst determined themedianGIT1 expression level from the samples and then divided theminto 2 groups, that is, the higher or the lower expression thanthe median level. We found that the higher GIT1 expressionwas significantly associated with the development of lymphnode metastasis (Fig. 4A), as the increased level of GIT1expression appeared to be correlated with cases having highernumber of positive lymph nodes involvement (Fig. 4B). Next,we examined whether GIT1 could affect cell migration, inva-sion, andmetastasis. Three sets of GIT1-siRNAs were tested fortheir inhibitory efficacy byWestern blotting. GIT1-siRNA-3 hadthe highest knockdown effect (Fig. 4C, left) and was used in thesubsequent experiments. Knockdown of GIT1 significantlyinhibited OSCC cell migration and invasion (Fig. 4C, right).GIT1 was able to rescue miR-491-5p-mediated inhibition ofOSCC cell invasion (Fig. 4D). To further assess the role of GIT1

in regulatingmetastasis, we injected GIT1-siRNA expressing orthe control plasmid-transfected C9-lung-IV2 cells into SCIDmice via tail vein. Knockdown of GIT1 resulted in about 5-foldreduction in the lung metastasis (Fig. 4E).

Re-expression of GIT1 significantly reverses miR-491-5p–mediated suppression of invasion and metastasis ofC9-lung-IV2 cells

Ectopic expression of GIT1 could partially rescue cell inva-sion inhibited by miR-491-5p in vitro (Fig. 4D). Using the tailvein injection assay, we also found that ectopic expression ofGIT1 rescued miR-491-5p–mediated inhibition of lung metas-tasis about 3-folds (Fig. 4F). The above results support ourhypothesis that miR-491-5p targets GIT1 to inhibit OSCCinvasion and metastasis.

Inhibition of GIT1 by miR-491-5p enhances degradationof paxillin and impairs focal adhesion signaling in OSCCcells

Previous studies indicated that GIT1 interacted with focaladhesion–related proteins, including paxillin and FAK (29, 30).To investigate the detailedmechanism throughwhich themiR-491-5p targeted GIT1 to inhibit cell migration and invasion, wefirst examined the effect on FAK and paxillin activation upon

miR-491-5p

miR-cont

B C

A

miR-491-5p

Anti-491-5p

miR-cont

miR

-491-5

p+

Anti-4

91-5

p

miR

-491-5

p

miR

-cont

miR

-491-5

p+

Anti-4

91-5

p

miR

-491-5

p

miR

-cont

SCC25

SCC25

CGHNC9

SCC25CGHNC9

GIT1-3¢UTR-wt-luc

GIT1-3¢UTR-mt-luc

miR-491-5p

miR-cont

GIT1-3¢UTR-wt-luc

GIT1-3¢UTR-mt-luc

CGHNC9

GIT1

Actin

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Re

lativ

e lu

cife

rase

act

ivity

Re

lativ

e G

IT1

exp

ress

ion

(GIT

1/G

AP

DH

)

1.2

1

0.8

0.6

0.4

0.2

0 Re

lativ

e lu

cife

rase

act

ivity 1.2

1

0.8

0.6

0.4

0.2

0

miR-491-5p miR-491-5p

GIT1 3¢-UTR-wt GIT1 3¢-UTR-mt5′3′ -5′

3′ 5′3′ -5′

3′Figure 3. Identification of GIT1 asthe direct target of miR-491-5p. A,effect of miR-491-5p on GIT1 30-UTR-wt (GIT1 30UTR-wt-luc) andGIT1 30-UTR-mutant luciferasereporters (GIT1 30UTR-mt-luc).Top, GIT1 30-UTR-wt sequenceand the GIT1 30-UTR-mt in whichthe sequence in bold wasmutagenized to abolish the miR-491-5p binding. Bottom, luciferaseassays showing decreased activityafter cotransfection of GIT130-UTR-wt with miR-491-5p inCGHNC9 (left) and SCC25 (right)OSCC cells. The GIT1 30-UTR-mtreporter was not affected by miR-491-5p. B, expression of GIT1mRNA in CGHNC9 and SCC25cells was measured by qRT-PCRfrom RNAs extracted from cellstransfected with miR-491-5p aloneor together with anti-miR-491-5p(anti-491-5p). Glyceraldehyde-3-phosphate dehydrogenase(GAPDH) was used as an internalcontrol. C, the expression of GIT1protein was detected by Westernblotting of total proteins from cellstransfected with miR-491-5p aloneor together with anti-miR-491-5p.b-Actin was used as an internalloading control. Histogramsrepresent means � SD from 3independent experiments.�, P < 0.05.

Huang et al.





A

B D

E F

C

CGHNC9

CGHNC9

Migration

Migration

Invasion

Invasion

miR-491-5p

pCMV-vector

pCMV-HA-GIT1

miR-cont

miR-491-5p

miR-491-5p miR-491-5p + GIT1

pCMV-vector

pCMV-HA-GIT1

miR-cont

miR-cont

miR-491-5p miR-491-5p + GIT1miR-cont

GIT1 low expression

GIT1 high expression

n = 23

n = 48

n = 5 n = 6

P = 0.0017

r = 0.443

7 7.5 8

GIT1 expression level

(log2 intensity)

8.5 9

n = 25

*

GIT1

Actin

GIT1

NC

-siR

NA

GIT

1 s

iRN

A-1

NC-siRNA

T

T

TT

T

T

T

TT

T

GIT1-siRNA

NC-siRNA GIT1-siRNA

GIT

1 s

iRN

A-2

GIT

1 s

iRN

A-3

NC-siRNAGIT1-siRNA-1

GIT1-siRNA-2GIT1-siRNA-3

Actin

**

***

***

***

* **

120

100

80

60

40

20

0

Re

lative

mig

ratio

n a

nd

inva

sio

n a

bili

ty

1.4

1.2

1

0.8

0.6

0.4

0.2

0

Rela

tive

inva

sio

n a

bili

ty

1.2

1

0.8

0.6

0.4

0.2

0

Lym

ph

no

de

–p

ositiv

e n

um

be

rs

25

20

15

10

5

0

Lym

ph n

ode–positiv

e n

um

bers

Lung m

eta

sta

sis

index (

%)

120

100

80

60

40

20

0Lung m

eta

sta

sis

index (

%)

25

20

15

10

5

0

–5

Figure 4. GIT1 expression correlates with lymph node metastasis and knockdown of GIT1 inhibits OSCC cell migration, invasion, and metastasis.A and B, analysis of cDNA microarray data from 48 OSCC tumor samples revealed an association between high number of lymph nodesinvolvement and increased GIT1 expression, determined from the array analysis. Spearman's r analysis was used to evaluate the relationship of GIT1expression and lymph node metastasis. C, left, Western blotting of GIT1 in CGHNC9 cells transfected with the GIT1-siRNA or negative control (NC).Right, a dramatic decrease in migration and invasion ability was observed in CGHNC9 cells transfected with GIT1-siRNA compared with the NC. D,left, Western blotting of GIT1 from CGHNC9 cells transfected with the GIT1 plasmid or vector control. Right, overexpression of GIT1 significantlyrescued CGHNC9 cell invasiveness inhibited by miR-491-5p. E, lung metastasis following tail vein injection of C9-lung-IV2 cells transiently transfectedwith GIT1-siRNA was dramatically reduced. F, lung metastasis index shows that transient transfection of GIT1 plasmid into C9-lung-IV2 cellsexpressing miR-491-5p significantly rescued their lung metastasis generated by tail vein injection. All images show H&E staining of the lungmetastases (�200). Histograms represent means � SD from 3 independent experiments. �, P < 0.05; ��, P < 0.01.






GIT1 depletion in OSCC cells. We hypothesized that suppres-sion of GIT1 could lead to inhibition of focal adhesion signalingand expression of molecules involved in cancer cell invasion.Indeed, suppression of GIT1 bymiR-491-5p reduced the level ofp-FAK, its interactive protein paxillin and phospho-paxillin(Fig. 5A) whereas the mRNA of paxillin remained unaffected(Supplementary Fig. S5). GIT1 has been shown to bind and

regulate activation of paxillin through FAK (13, 30). As thepaxillin mRNA was not affected by GIT1 knockdown, wechecked whether the decreased paxillin was due to enhancedproteasome degradation. The level of paxillin in GIT1-depletedcells with or without addition of MG132 (a proteasome inhib-itor) was compared. The result showed that treatment ofMG132 reversed the effect of GIT1 depletion on paxillin level

A

B

D

C

C9-lung-IV2

C9-lung-IV2 C9-lung-IV-2

Cyclohex (h) 0 4 12 0 4 12

Vinculin F-actin DAPI Merge

miR-491-5p

Anti miR-491-5p

miR-cont

miR

-491-5

p

miR

-491-5

p+

Fla

g-G

IT1

miR

-491

-5p

+ Fla

g-GIT

1

miR

-co

nt

miR

-491

-5p

miR

-con

t

GIT1

Actin

GIT1

Actin

GIT1

Actin

NC-siRNA

NC-siRNA

GIT1-siRNA

GIT1 siRNA

MG132 (h)

GIT1-siRNA

NC-s

iRNA

GIT

1-siRNA

0 0 8

p-FAK

p-FAK

FAK

FAK

p-Paxillin

Paxillin

Paxillin

Paxillin

p-Paxillin

NC

-siR

NA

GIT

1-s

iRN

A

* **

60

50

40

30

20

10

0

Ave

rage F

A a

rea/μ

m2

Figure 5. GIT1 depletion enhances degradation of paxillin and reduces FAKphosphorylation and decreases focal adhesion formation. A,Western blot analysisshowing expression of miR-491-5p or knockdown of GIT1 decreased phosphorylation of paxillin, FAK, and reduced level of paxillin protein. B, knockdownof GIT1 decreased paxillin protein, which was blocked by MG132, while reduced phosphorylation of FAK was not. Treatment with the MG132 wasat 10mmol/L for 8 hours.C, knockdownofGIT1 significantly decreasedpaxillin stability inOSCCcells. Thepaxillin levels in control andGIT1-depletedC9-lung-IV2 cells treated with cycloheximide (20 mg/mL) for the indicated times show a faster decay in the GIT1 depleted cells. D, C9-lung-IV2 cells expressingthe indicated plasmids were seeded in dishes coated with fibronectin for 120 minutes (top) and analyzed for total focal adhesion (FA) area per cell (bottom).The focal adhesions and stress fibers were revealed with anti-vinculin (green) and rhodamine (red), respectively. Histograms represent means � SD from 3independent experiments. �, P < 0.05.

Huang et al.





(Fig. 5B), suggesting that knockdown ofGIT1 increased paxillindegradation via proteasome pathway in oral cancer cells. Tofurther confirm depletion of GIT1, resulting in decreasedpaxillin stability, we conducted protein degradation experi-ment in cells treated with cycloheximide (Cyclohex). Paxillinwas stable over a course of 12 hours in the control cells. Incontrast, paxillin stability was significantly decreased in theGIT1-depleted cells (Fig. 5C). We further tested whether GIT1depletion could affect OSCC cell focal adhesion and lamelli-podia formation. As shown in Fig. 5D, vinculin level and focaladhesions were significantly decreased in the miR-491-5p–transfected C9-lung-IV2 cells, which was reversed by over-expression of GIT1. Likewise, the focal adhesions were dis-rupted in the si-GIT1–transfected C9-lung-IV2 cells (Fig. 5D).These results suggest that miR-491-5p–mediated decrease ofGIT1 results in reduced FAK activation and increased paxillindegradation, leading to reduced focal adhesion formation ofOSCC cells.

Dominant-negative form of FAK partially inhibitedOSCC cell invasiveness that could be rescued by GIT1We further determined the role of FAK in miR-491-5p–

targeted GIT1 downregulation and inhibition of migration andinvasion. Our data showed that GIT1 promoted cell invasioncould be reversed by dominant-negative FAK (DN-FAK; Sup-plementary Fig. S6B). Western blot analysis confirmed thedownregulation of p-FAK by DN-FAK in C9-lung-IV2 cellsoverexpressing GIT1 (Supplementary Fig. S6A). Those datasuggested that downregulation of GIT1 by miR-491-5p couldsuppress migration, invasion, and metastasis of OSCC cellsthrough degradation of paxillin and reduced focal adhesionsignaling, including FAK and ERK1/2 activation.

miR-491-5p targets GIT1 to regulate expression level andactivity of MMP2/9 via EGFR/ERK1/2 signaling pathwayMMP2/9 have been implicated in tumor cell invasion and

metastasis (31, 32). We observed that overexpression of miR-491-5p or knockdown of GIT1 decreased MMP2/9 mRNA(Fig. 6A) and activity (Fig. 6B) in OSCC cells. Previous studiesindicated that the expression level and activity of MMP2/9were regulated by ERK1/2 in HEK293 and monocytic cells(33, 34). Our data also showed that the expression level ofMMP2/9 was reduced by PD98059 (an MEK inhibitor) inOSCC cells (Supplementary Fig. S7). GIT1 was shown to linkthe EGFR to ERK1/2 activation in HEK293 cells (21). How-ever, the role of GIT1 in the regulation of MMPs in oralcancer cells is unknown. To check whether the miR-491-5p-GIT1 pathway could regulate ERK1/2 signaling, which inturn affected MMP2/9 expression, we examined the effect ofmiR-491-5p and GIT1 on EGF-induced ERK1/2 activation.CGHNC9 cells were transfected with control-siRNA, miR-491-5p, or GIT1-siRNA, serum starved for 24 hours, and thenstimulated with EGF. In the control cells, EGF rapidlystimulated robust activation of ERK1/2. In contrast, miR-491-5p or GIT1-siRNA–treated cells showed significantlyreduced activation of ERK1/2 (Fig. 6C). Thus, reduced GIT1expression is associated with dramatic inhibition of EGF-stimulated ERK1/2 activation. Consistently, overexpression

of miR-491-5p resulted in significantly inhibition of thegrowth beyond 72 hours especially upon EGF stimulation(Supplementary Fig. S8B). In addition, our data showed thatGIT1 increased MMP2/9 mRNA expression could be reducedby dominant-negative MEK or PD98059 (Supplementary Fig.S9B). Western blot analysis confirmed the downregulation ofp-ERK1/2 by DN-MEK and PD98059 in C9-lung-IV2 cellsoverexpressing GIT1 (Supplementary Fig. S9A). These resultssuggest that miR-491-5p targets GIT1 to suppress MMP2/9expression and activity, and this is likely to be due toinhibition of ERK1/2 activation in OSCC cells.

Correlation of miR-491-5p and GIT1 expression withgrades of OSCCs and inverse correlation between miR-491-5p and GIT1

There have been no studies reporting the role of GIT1 inoral cancer so far. We examined GIT1 expression in thematched normal and cancerous oral tissue specimens toassess its clinical connection. We found that in 26 of 33 casesof OSCC samples, the GIT1 level was 1.5- to 23-fold highercompared with their matched normal tissues (Fig. 7A), andthe GIT1 level was inversely correlated with miR-491-5pexpression (Fig. 7B). Furthermore, 189 OSCC samplesshowed that 51 patients with grade 1 had longer survivaltime than 138 patients with grade 2 and 3 (Fig. 7C). We nextdetermined whether tumor grade was correlated with GIT1expression by IHC staining of oral cancer tissue arrays. Foreach section, staining was measured as weak (þ), moderate(þþ), and strong (þþþ). We observed that GIT1 level washigher in the moderately (grade 2) and poorly differentiated(grade 3) than in the well differentiated tumors (grade 1; Fig.7D). About 79% of cases with grade 2 to 3 had moderate-to-strong expression of GIT1, whereas 79% of cases with grade 1had weak expression of GIT1 (Fig. 7D, table). Thus the GIT1level appeared to correlate with the aggressiveness of oralcancer. Furthermore, miR-491-5p signals were relativelyweak in moderately and poorly differentiated tumor samples(Fig. 7D). Seventy-seven percent of cases with grade 2 to 3had weak-to-moderate expression of miR-491-5p, whereas89% of cases with grade 1 showed a moderate-to-strongexpression of miR-491-5p (Fig. 7D, table). The results indi-cate that miR-491-5p and GIT1 expression levels are inverse-ly correlated and lower miR-491-5p together with higherGIT1 correlates with the advanced grades of the disease.

DiscussionOSCC has a high incidence of neck metastasis and often

metastasizes contralaterally (35). It is widely known that thelymph node metastases is an important prognosticator pre-dicting survival of patients with OSCCs (36–38). Despite accu-mulating evidences pointing to the regulatory role of micro-RNAs in cancer progression and invasion, its regulatory role inoral cancer metastasis is still poorly understood. This studyaimed to identify relevant microRNAs and their downstreamtarget genes involved in the regulation of cancer metastasis inOSCCs. For the first time, we establish the miR-491-5p-GIT1-FAK-paxillin and miR-491-5p-GIT1-ERK1/2 pathways in the






regulation of OSCC cell invasion andmetastasis. We also showthat low expression of miR-491-5p and high expression of GIT1correlate with aggressiveness and lymph node metastasis ofOSCC. Furthermore, we showed that the level of miR-491-5p issignificantly correlated with overall survival of OSCCs. Inter-estingly, in a separate study, we found that GIT1 also played animportant role in lymph node metastasis of breast cancer(Chan and colleagues, unpublished).

We found that miR-491-5p level decreased significantly inthe highly invasive OSCC lines compared with their corre-sponding less invasive parental cells and that miR-491-5psuppressed migration, invasion, and metastasis of oral cancercells mainly via targeting GIT1 to affect FAK and ERK1/2

signaling. A recent study reported the inhibitory effect ofmiR-491-5p on expression of MMP9 and invasion of glioblas-toma multiforme cells (39). Our data also show that MMP9mRNA expression level and activity are decreased in the OSCCcells overexpressing miR-491-5p most likely via inhibition ofERK1/2 activation (Fig. 6). In another study, it was shown thatoverexpression of CD44 30UTR interacted with and trappedendogenous miR-491-5p, leading to increased levels of fibro-nectin and collagen, resulting in enhanced cell motility andinvasion in humanbreast cancer cells (40). Ourfinding inOSCCis in agreement with those studies suggesting miR-491-5pfunctions as an invasion suppressor in glioma cells and breastcancer cells. However, we show here that miR-491-5p inhibits

miR

-491

-5p

miR

-491

-5p

+Ant

i-491

-5p

miR

-con

t

A

B C

miR-491-5p

NC-siRNA

GIT1-siRNA

GIT1-siRNA

miR-cont

miR-491-5p miR-491-5pControl

0 5 10 30 0 5 10 30 0 5 10 30

NC-siRNA

GIT1-siRNA

NC

-siR

NA

GIT

1-si

RN

A

GIT1

Actin

GIT1

ERK1/2

p-ERK1/2

ActinMMP9

MMP2

miR-cont

miR-491-5p

NC-siRNA

GIT1-siRNA

EGF, min

miR-cont

miR-491-5p

NC-siRNA

GIT1-siRNA

miR-cont

GCHNC9

CGHNC9

C9-lung-IV2

*

*

**

**

1.2

1

0.8

0.6

0.4

0.2

0Rela

tive

MM

P2 e

xpre

ssio

n

(MM

P2/G

AP

DH

)

1.2

1

0.8

0.6

0.4

0.2

0Rela

tive

MM

P9 e

xpre

ssio

n

(MM

P9/G

AP

DH

)

Figure 6. Effect of GIT1 on activation and expression of MMP2/9 and on EGF-induced ERK1/2 phosphorylation. A, relative MMP2/9 mRNA levels in CGHNC9cells transfected with the indicated plasmids. B, top, Western blotting of GIT1 expression in CGHNC9 cells transfected with miR-491-5p, GIT1-siRNA,or negative control (NC). Bottom, CGHNC9 cells were preincubated in serum-free medium for 24 hours, and conditioned media were extracted andanalyzed by gelatin zymography. C, CGHNC9 cells were transfected with the indicated plasmids for 8 hours, serum-starved for 24 hours, and then treatedwith 10 ng/mL EGF for the indicated times. Histograms represent means � SD from 3 independent experiments. ��, P < 0.01.

Huang et al.





OSCC invasion and metastasis via targeting GIT1, which isdistinct from those 2 studies. As shown in Supplementary Fig.S8A, miR-491-5p although had no significant effect on prolif-eration of OSCC cells before 72-hour time point did affect thegrowth beyond that time point especially upon EGF stimula-

tion likely due to inhibition of ERK1/2 activation (Supplemen-tary Fig. S8B). Our data also showed that expression of miR-491-5p in C9-lung-IV2 cells resulted in a significant decrease oflung metastasis compared with the control in SCID mice at24-hour time point (Supplementary Fig. S2). It appears that

A B

C

E

D

0 20 40 60 80

Time (months)

EGF

EGFR

Transcription/

activation

Proteasomal

degradationIntegrin

signaling

Invasion/metastasis

Fibronectin

Integrin

α β

Y397FAK

ERK1/2

MMP2/9

P P

P

P

P

FAKPAX

PAXSHD PBD

100 120

GIT1 and miR-491-5p correlate with the aggressiveness of human oral cancer

GIT1 expression

miR-491-5p expression

Oral specimen No. specimens Weak (+) Moderate (++) Strong (+++) P*

Grade 1 86 68 (79%) 10 (12%) 8 (9%) –

Grades 2 and 3 48 10 (21%) 15 (31%) 23 (48%) <0.0001

Grade 1 86 10 (11%) 21 (24%) 55 (65%) –

Grades 2 and 3 48 25 (52%) 12 (25%) 11 (23%) <0.0001

miR-491-5p

GIT1

33 normal/tumor (N/T) pairs

OSCC

Grade 1

(n = 51)

Grade 2

(n = 114)

Grade 3

(n = 24)

P < 0.01

P < 0.01

Ove

rall

surv

ival

Normal

N = 33

P = 0.0344

r = –0.371

0 0.01 0.02

Grade 1 Grade 2G

IT1

miR

-491-5

pGrade 3

0.03 0.04

0.04

0.03

0.02

0.01

0.00

0.05

0.04

0.03

0.02

0.01

0.00

1.0

0.8

0.6

0.4

0.2

0.0

Rela

tive

GIT

1 e

xpre

ssio

n

(GIT

1/G

AP

DH

)

Relative GIT1 expression

(GIT1/GAPDH)

Rela

tive

miR

-491-5

p e

xpre

ssio

n

(miR

-491-5

p/R

NU

6B

)

**

Figure 7. miR-491-5p and GIT1 expression levels exhibits an inverse relationship in OSCC specimens. A, qRT-PCR shows upregulation of GIT1 mRNA inOSCC tumors compared with matched normal oral tissues. B, a negative correlation between miR-491-5p and GIT1 RNA levels in 33 OSCC tumorspecimens. Spearman's r analysis was used. C, overall survival timewas significantly associatedwith low grade (grade 1) comparedwith high grades (grades2 and 3) in 189 OSCC samples. D, a typical example of the expression levels of miR-491-5p and GIT1 protein as determined by FISH (for miR-491-5p)and IHC (for GIT1) of oral cancer tissue prepared from 134 paraffin-fixed carcinoma (grades 1–3) specimens. E, a proposed scheme of the negativeregulation of the GIT1/FAK/paxillin and EGF signaling pathways by miR-491-5p in OSCC cells. ��, P < 0.01.






miR-491-5p targets and regulates distinct pathways to mani-fest its biologic function in different types of cancer cells. Ourstudy identified miR-491-5p as a novel metastasis suppressorin oral cancer.

The differential expression level of miR-491-5p and differ-ence in migration/invasion ability is difficult to equate quan-titatively. We think at least there are 3 possible explanations;first, a given miRNA could have multiple targets involved inregulating migration/invasion; second, a 2-fold difference inmiRNA could lead to a greater decrease in the target, resultinginmore than 2-fold difference inmigration/invasion; and third,the same effector gene could be targeted bymiRNAs other thanmiRNA-491-5p.

As a scaffold protein associated with integrin, GIT1 playsa functional role in cell migration (11, 12), focal adhesionformation (28), and angiogenesis (41). For example, GIT1promotes migration in A431 cells by interacting with theMYO18A-PAK2-bpix complex (11). GIT1 is also an impor-tant regulator for VEGF-induced endothelial cell podosomeformation (41). Moreover, several previous studies showedthat GIT1 was associated with p-ERK1/2 in focal adhesionsafter EGF stimulation in HeLa cells (42, 43) and affectedfocal adhesion formation in osteoblasts following platelet-derived growth factor stimulation (28). Although GIT1 hasbeen implicated in the regulation of cell motility, its role ininvasion and metastasis of OSCC cells has not beenreported before. For the first time, our study indicates thatGIT1 plays an important role in invasion and metastasis oforal cancer cells and it is targeted and downregulated bymiR-491-5p in the highly invasive and poorly differentiatedtumor cells.

Regulation of focal adhesion complexes involving FAK andpaxillin are known to be important in cell invasion (44). Ourdata show that knockdown of GIT1 led to decreased FAKactivity (Fig. 5A) and increase of proteasome-mediateddegradation of paxillin (Fig. 5B and C). Together, thesefindings suggest that miR-491-5p inhibits invasion andmetastasis of OSCC cells through depletion of GIT1, leadingto disruption of focal adhesion complexes. Although GIT1 isoften implicated in the regulation of cell focal adhesion,information about its role and regulatory mechanism for theinvasion and metastasis in oral cancer has not been exploredpreviously. Interaction between FAK and paxillin plays animportant role in focal adhesion signaling and regulation ofcell motility (45), in part, via ERK1/2 activation (17). Fur-thermore, a previous study indicated that paxillin�/� MEFcells showed several phenotypical changes, includingreduced cell migration (46). In addition, paxillin�/� MEFstill have focal adhesion (with abnormal shapes). Our dataalso showed that GIT1-depleted OSCC cells had decreasedlevel and activity of paxillin protein and reduced cell migra-tion. Depletion of FAK inhibited the phosphorylation ofpaxillin as expected as paxillin is a substrate of FAK andknockdown of paxillin also inhibited phosphorylation of FAK(Supplementary Fig. S10). These results suggested thatdepletion of GIT1 reduced paxillin stability, leading todecreased FAK phosphorylation in agreement with thereport of Hagel and colleagues (46). Therefore, the loss of

focal adhesions upon miR-491-5p overexpression or knock-down of GIT1 may not be attributed to the reduction ofpaxillin, but mainly to the reduction of GIT1. Among thepotential upstream activation of ERK1/2 is EGFR, which hasbeen previously implicated in oral cancer development andprogression (47, 48). Our data show that depletion of GIT1significantly reduces EGF-induced ERK1/2 activation inOSCC cells (Fig. 6C). This is in agreement with a previousstudy indicating that GIT1 links the EGFR to ERK1/2 acti-vation by associating with MEK1 (21). Thus, we propose aregulatory signaling pathway of OSCC metastasis consistingof miR-491-5p and its downstream target, GIT1, whichfunctions as a scaffold for focal adhesion signaling andEGF-induced ERK1/2 activation.

We showed that transient transfection with miR-491-5p orknockdown of GIT1 dramatically inhibited MMP2/9 mRNAexpression and activity (Fig. 6A and B). We also showed thattreatment of dominant-negative MEK or PD98059 (Supple-mentary Fig. S7) reduced the MMP2/9 expression level. It wasshown that ERK1/2 pathway enhanced the invasion of lungcancer cells by increasingMMP2/9 expression and activity (33).Thus, the inhibition ofMMP2/9 bymiR-491-5p and knockdownof GIT1 is likely, in part, due to decreased ERK1/2 activation inOSCC cells.

Our analysis of clinical samples revealed a low miR-491-5pand high GIT1 expression pattern in the advanced OSCC andthere appeared to exist an inverse relationship between miR-491-5p and GIT1 expression. ThemiR-491-5p seems to have noeffect onOSCC cell growth. Thus,miR-491-5p expression is lesslikely to be involved in tumor initiation or growth; rather, it ismore likely to be involved in the invasive progression when itneeds to be downregulated to allow overexpression of GIT1 forcancer cell motility and MMP production/activation. It wouldbe interesting to explore the upstream regulator for miR-491-5p downregulation during the metastatic progression ofOSCCs.

In conclusion, our study has identifiedmiR-491-5p as a novelmetastatic suppressor through regulation of its downstreamtarget, GIT1, resulting in perturbation of FAK/paxillin andERK1/2 signaling as well as MMP expression and activity inoral cancer cells (Fig. 7E). The miR-491-5p and GIT1 couldpotentially serve as metastatic prognosis biomarkers andtargets for intervention of OSCC metastasis.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: L.-H. Wang, W.-C. Huang, S.-H. Chan, T.-Y. Shieh,A.-J. ChengDevelopment of methodology: L.-H. Wang, W.-C. Huang, S.-H. Chan,T.-H. Jang, J.-W. Chang, W.-F. Chiang, A.-J. ChengAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): L.-H.Wang,W.-C. Huang, S.-H. Chan, S.-L. Chiang,W.-F. Chiang, C.-T. Liao, J.-L. Juang, H.-C. Wang, Y.-C. LuAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): L.-H. Wang, W.-C. Huang, S.-H. Chan, Y.-C. Ko,W.-F. ChiangWriting, review, and/or revision of the manuscript: L.-H. Wang, W.-C.Huang, S.-H. Chan, C.-T. Liao

Huang et al.





Administrative, technical, or material support (i.e., reporting ororganizing data, constructing databases): L.-H. Wang, W.-C. Huang,S.-H. ChanY.-C. Ko, T.-C. Yen, C.-T. Liao, J.-L. JuangStudy supervision: L.-H. Wang, C.-T. LiaoSecured grant support for the study and acquisition of clinical speci-mens: L.-H. Wang

AcknowledgmentsThe authors thank the Pathology and Microarray Core Laboratories of the

National Health Research Institutes for H&E and IHC staining and microRNAarray analysis, respectively. They also thank Hsin-Lei Yao for discussion.

Grant SupportThis work was supported by grants from National Science Council, Taiwan

(01D2-MMNSC16), and the Department of Health, Taiwan (DOH100-TD-111-004).

The costs of publication of this article were defrayed in part by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received May 8, 2013; revised October 26, 2013; accepted November 13, 2013;published OnlineFirst December 12, 2013.

References1. SamanW.Global epidemiology of oral and oropharyngeal cancer. Oral

Oncol 2009;45:309–16.2. Ho P-S, Ko Y-C, Yang Y-HC, Shieh T-Y, Tsai C-C. The incidence of

oropharyngeal cancer in Taiwan: an endemic betel quid chewing area.J Oral Pathol Med 2002;31:213–9.

3. Chen Y-J, Chang JT-C, Liao C-T, Wang H-M, Yen T-C, Chiu C-C,et al. Head and neck cancer in the betel quid chewing area:recent advances in molecular carcinogenesis. Cancer Sci 2008;99:1507–14.

4. Ferlito A, Rinaldo A, Devaney KO,MacLennan K,Myers JN, PetruzzelliGJ, et al. Prognostic significance of microscopic and macroscopicextracapsular spread from metastatic tumor in the cervical lymphnodes. Oral Oncol 2002;38:747–51.

5. Mamelle G, Pampurik J, Luboinski B, Lancar R, Lusinchi A, Bosq J.Lymph node prognostic factors in head and neck squamous cellcarcinomas. Am J Surg 1994;168:494–8.

6. Liao C-T, Hsueh C, Lee L-Y, Lin C-Y, Fan K-H, Wang H-M, et al.Neck dissection field and lymph node density predict prognosis inpatients with oral cavity cancer and pathological node meta-stases treated with adjuvant therapy. Oral Oncol 2012;48:329–36.

7. Kurahara S-i, Shinohara M, Ikebe T, Nakamura S, Beppu M, Hiraki A,et al. Expression of MMPS, MT-MMP, and TIMPs in squamous cellcarcinoma of the oral cavity: correlations with tumor invasion andmetastasis. Head Neck 1999;21:627–38.

8. Nicoloso MS, Spizzo R, Shimizu M, Rossi S, Calin GA. MicroRNAs[mdash] the micro steering wheel of tumour metastases. Nat RevCancer 2009;9:293–302.

9. Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D, et al.MicroRNA expression profiles classify human cancers. Nature 2005;435:834–8.

10. P�erez-Say�ans M, Pilar G-D, Barros-Angueira F, Su�arez-Pe~narandaJM, Fern�andez AC, G�andara-Rey JM, et al. Current trends in miRNAsand their relationship with oral squamous cell carcinoma. J Oral PatholMed 2012;41:433–43.

11. Hsu R-M, Tsai M-H, Hsieh Y-J, Lyu P-C, Yu J-S. Identification ofMYO18A as a novel interacting partner of the PAK2/bPIX/GIT1 com-plex and its potential function in modulating epithelial cell migration.Mol Biol Cell 2010;21:287–301.

12. Manabe R-i, Kovalenko M, Webb DJ, Horwitz AR. GIT1 functionsin a motile, multi-molecular signaling complex that regulatesprotrusive activity and cell migration. J Cell Sci 2002;115:1497–510.

13. Hajdo-Milasinovic A, van der Kammen RA, Moneva Z, Collard JG.Rac3 inhibits adhesion and differentiation of neuronal cells bymodifying GIT1 downstream signaling. J Cell Sci 2009;122:2127–36.

14. Schmalzigaug R, Garron M-L, Roseman JT, Xing Y, Davidson CE,Arold ST, et al. GIT1 utilizes a focal adhesion targeting-homologydomain to bind paxillin. Cell Signal 2007;19:1733–44.

15. Zhao J, Guan J-L. Signal transduction by focal adhesion kinase incancer. Cancer Metastasis Rev 2009;28:35–49.

16. Brunton VG, Frame MC. Src and focal adhesion kinase as therapeutictargets in cancer. Curr Opin Pharmacol 2008;8:427–32.

17. Subauste MC, Pertz O, Adamson ED, Turner CE, Junger S, HahnKM. Vinculin modulation of paxillin–FAK interactions regulates

ERK to control survival and motility. J Cell Biol 2004;165:371–81.

18. Nicholson RI, Gee JMW, Harper ME. EGFR and cancer prognosis. EurJ Cancer 2001;37 Suppl 4:9–15.

19. Ciardiello F, Tortora G. A novel approach in the treatment of cancer:targeting the epidermal growth factor receptor. Clin Cancer Res 2001;7:2958–70.

20. Tsai P-C, Hsieh C-Y, Chiu C-C, Wang C-K, Chang L-S, Lin S-R.Cardiotoxin III suppresses MDA-MB-231 cell metastasis through theinhibition of EGF/EGFR-mediated signaling pathway. Toxicon 2012;60:734—43.

21. Yin G, Haendeler J, Yan C, Berk BC. GIT1 functions as a scaffold forMEK1-extracellular signal-regulated kinase 1 and 2 activation byangiotensin II and epidermal growth factor. Mol Cell Biol 2004;24:875–85.

22. Peng C-H, Liao C-T, Peng S-C, Chen Y-J, Cheng A-J, Juang J-L, et al.A novel molecular signature identified by systems genetics approachpredicts prognosis in oral squamous cell carcinoma. PLoS One2011;6:e23452.

23. Lu Y-C, Chen Y-J, Wang H-M, Tsai C-Y, Chen W-H, Huang Y-C, et al.Oncogenic function and early detection potential of miRNA-10b in oralcancer as identified by microRNA profiling. Cancer Prev Res 2012;5:665–74.

24. Wu R-M, Wood M, Thrush A, Walton EF, Varkonyi-Gasic E. Real-timePCR quantification of plant miRNAs using universal ProbeLibrarytechnology. Biochemica 2007;2:12–5.

25. Yeh Y-M, Chuang C-M, Chao K-C, Wang L-H. MicroRNA-138 sup-presses ovarian cancer cell invasion andmetastasis by targeting SOX4and HIF-1a. Int J Cancer 2013;133:867–78.

26. Silahtaroglu AN, Nolting D, Dyrskjot L, Berezikov E, Moller M,Tommerup N, et al. Detection of microRNAs in frozen tissue sec-tions by fluorescence in situ hybridization using locked nucleic acidprobes and tyramide signal amplification. Nat Protoc 2007;2:2520–8.

27. SakamotoS,McCannRO, Dhir R, KyprianouN. Talin1 promotes tumorinvasion and metastasis via focal adhesion signaling and anoikisresistance. Cancer Res 2010;70:1885–95.

28. RenY,YuL, Fan J,Rui Z,HuaZ,ZhangZ, et al. PhosphorylationofGIT1tyrosine 321 is required for association with FAK at focal adhesionsand for PDGF-activated migration of osteoblasts. Mol Cell Biochem2012;265:109–18.

29. Zhang ZM, Simmerman JA, Guibao CD, Zheng JJ. GIT1 paxillin-binding domain is a four-helix bundle, and it binds to both paxillinLD2 and LD4 motifs. J Biol Chem 2008;283:18685–93.

30. Zhao Z-s, Manser E, Loo T-H, Lim L. Coupling of PAK-interactingexchange factor PIX to GIT1 promotes focal complex disassembly.Mol Cell Biol 2000;20:6354–63.

31. Mendes O, Kim H-T, Lungu G, Stoica G. MMP2 role in breast cancerbrain metastasis development and its regulation by TIMP2 andERK1/2. Clin Exp Metastasis 2007;24:341–51.

32. Liu Z, Li L, Yang Z, Luo W, Li X, Yang H, et al. Increased expression ofMMP9 is correlated with poor prognosis of nasopharyngeal carcino-ma. BMC Cancer 2010;10:270.

33. Luo Y, Liang F, Zhang Z-Y. PRL1 promotes cell migration and invasionby increasing MMP2 and MMP9 expression through Src and ERK1/2pathways†. Biochemistry 2009;48:1838–46.






34. Shim Y-J, Kang B-H, Jeon H-S, Park I-S, Lee K-U, Lee I-K, et al.Clusterin induces matrix metalloproteinase-9 expression via ERK1/2and PI3K/Akt/NF-kB pathways in monocytes/macrophages. J LeukocBiol 2011;90:761–9.

35. Liao C-T, Huang S-F, Chen IH, Chang J-C, Wang H-M, Ng S-H, et al.Risk stratification of patients with oral cavity squamous cell carcinomaand contralateral neck recurrence following radical surgery. Ann SurgOncol 2009;16:159–70.

36. Capote A, Escorial V, Mu~noz-Guerra MF, Rodríguez-Campo FJ,Gamallo C, Naval L. Elective neck dissection in early-stage oralsquamous cell carcinoma—does it influence recurrence and survival?Head Neck 2007;29:3–11.

37. Greenberg JS, El Naggar AK, Mo V, Roberts D, Myers JN. Disparity inpathologic and clinical lymph node staging in oral tongue carcinoma.Cancer 2003;98:508–15.

38. Capote-Moreno A, Naval L, Mu~noz-Guerra MF, Sastre J, Rodríguez-Campo FJ. Prognostic factors influencing contralateral neck lymphnode metastases in oral and oropharyngeal carcinoma. J Oral Max-illofac Surg 2010;68:268–75.

39. Yan W, Zhang W, Sun L, Liu Y, You G, Wang Y, et al. Identification ofMMP-9 specific microRNA expression profile as potential targets ofanti-invasion therapy in glioblastoma multiforme. Brain Res 2011;1411:108–15.

40. Rutnam ZJ, Yang BB. The non-coding 30 UTR of CD44 inducesmetastasis by regulating extracellular matrix functions. J Cell Sci2012;125:2075–85.

41. Wang J, Taba Y, Pang J, Yin G, Yan C, Berk BC. GIT1mediates VEGF-induced podosome formation in endothelial cells. Arterioscler ThrombVasc Biol 2009;29:202–8.

42. Yin G, Zheng Q, Yan C, Berk BC. GIT1 is a scaffold for ERK1/2activation in focal adhesions. J Biol Chem 2005;280:27705–12.

43. Zhang N, Cai W, Yin G, Nagel DJ, Berk BC. GIT1 is a novel MEK1-ERK1/2 scaffold that localizes to focal adhesions. Cell Biol Int 2010;34:41–7.

44. Kornberg LJ. Focal adhesion kinase and its potential involvement intumor invasion and metastasis. Head Neck 1998;20:745–52.

45. Turner CE. Paxillin and focal adhesion signalling. Nat Cell Biol 2000;2:E231–6.

46. Hagel M, George EL, Kim A, Tamimi R, Opitz SL, Turner CE, et al. Theadaptor protein paxillin is essential for normal development in themouse and is a critical transducer of fibronectin signaling.Mol Cell Biol2002;22:901–15.

47. Shuyi Y, Feng W, Jing T, Hongzhang H, Haiyan W, Pingping M,et al. Human beta-defensin-3 (hBD-3) upregulated by LPS viaepidermal growth factor receptor (EGFR) signaling pathwaysto enhance lymphatic invasion of oral squamous cell carcinoma.Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2011;112:616–25.

48. Sheu JJ-C, Hua C-H, Wan L, Lin Y-J, Lai M-T, Tseng H-C, et al.Functional genomic analysis identified epidermal growth factor recep-tor activation as themost common genetic event in oral squamous cellcarcinoma. Cancer Res 2009;69:2568–76.

Huang et al.





2014;74:751-764. Published OnlineFirst December 12, 2013.Cancer Res Wei-Chieh Huang, Shih-Hsuan Chan, Te-Hsuan Jang, et al. Oral Squamous Cell Carcinoma Invasion and MetastasismiRNA-491-5p and GIT1 Serve as Modulators and Biomarkers for

Updated version

10.1158/0008-5472.CAN-13-1297doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/12/12/0008-5472.CAN-13-1297.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/74/3/751.full#ref-list-1

This article cites 48 articles, 15 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/74/3/751.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/74/3/751To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-13-1297

http://cancerres.aacrjournals.org/content/suppl/2013/12/12/0008-5472.CAN-13-1297.DC1

http://cancerres.aacrjournals.org/content/74/3/751.full#ref-list-1

http://cancerres.aacrjournals.org/content/74/3/751.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/74/3/751


mirna-491-5p and git1 serve as modulators and biomarkers ... · mirna-491-5p and git1 serve as...

Documents