ANT2 suppression by shRNA restores miR-636 expression, thereby downregulating Ras and inhibiting tumorigenesis of hepatocellular carcinoma.

0.64 MB·English

by  Jang, Ji-Young

Experimental&MolecularMedicine(2013)45,e3;doi:10.1038/emm.2013.1
&2013KSBMB. Allrightsreserved2092-6413/13
www.nature.com/emm
ORIGINAL ARTICLE
ANT2 suppression by shRNA restores miR-636
expression, thereby downregulating Ras and inhibiting
tumorigenesis of hepatocellular carcinoma
Ji-Young Jang1, Young-Sin Lee1, Yoon-Kyung Jeon1,2, Kyoungbun Lee2, Ja-June Jang2 and Chul-Woo Kim1,2
MicroRNAs (miRNAs) participate in diverse biological functions and carcinogenesis by inhibiting specific gene expression.
We previously reported that suppression of adenine nucleotide translocase 2 (ANT2) by using the short hairpin RNA (shRNA)
approach has an antitumor effect in several cancer cells. We here examined the influence of ANT2 on expression of miRNAs
in hepatocellular carcinoma (HCC) to further elucidate the tumor-suppressive mechanism of ANT2 shRNA. We first carried
out screening for miRNAs, whose expression is regulated by ANT2 suppression in the Hep3B HCC cell line using miRNA
microarrays. Validation of candidate miRNAs was done by incorporating clinical samples, and their effects on the tumorigenesis
of HCC were studied in vitro and in vivo. miR-636 was one of the miRNAs whose expression was highly upregulated by
ANT2 suppression in miRNA microarray analysis, as confirmed by real-time reverse transcription-polymerase chain reaction.
Notably, miR-636 was markedly downregulated in HCC tissues compared with matched non-neoplastic liver in clinical samples.
Restoration of miR-636 in Hep3B cells led to significant reduction of cell proliferation and colony formation. miR-636
restoration resulted in a decreased level of Ras, one of the putative targets of miR-636, and inactivation of its signaling
pathway. Moreover, tumorigenesis was efficiently suppressed by miR-636 in an in vivo tumor xenograft model of HCC. The data
suggest that miR-636 might function as a tumor suppressor miRNA affecting HCC tumorigenesis via downregulation of Ras,
and that ANT2 suppression by shRNA could exert an anticancer effect by restoring miR-636 expression in HCC.
Experimental& Molecular Medicine (2013) 45, e3;doi:10.1038/emm.2013.1;publishedonline 10January2013
Keywords: adenine nucleotide translocase2;hepatocellular carcinoma; miR-636; Ras signaling pathway;RNA interference
INTRODUCTION resultshavebeencontentious,thesemiRNAswerefoundtobe
Cancerisaffectedbymanymolecularpathwaysinvolvingboth consistentlydownregulatedinseveralhumancancers.8Among
canonical protein-coding genes and recently discovered non- the downregulated miRNAs in HCC is the hepato-specific
coding genes. MicroRNAs (miRNAs), a class of small RNAs miR-122; downregulation of miR-122 may have a more direct
17–27 nucleotides in length, are representative molecules role in tumorigenesis through the activation of cyclin G1.9
transcribed from noncoding genes, and control gene expres- The Ras signaling pathway controls cell proliferation, diff-
sion at the level of mRNA translation.1 An aberrant miRNA erentiation and survival in all multicellular organisms.10–13
expressionsignatureisahallmarkofseveraldiseases,including Consequently, proper regulation of Ras signaling is critically
cancer,2 and miRNAs can function as oncogenes or tumor important for normal development, and dysregu-
suppressors.3 Moreover, these tumor-associated miRNAs can lation of the Ras pathway leads to pathologic diseases.14–16
serve as biomarkers for tumor diagnosis and prognosis or as Indeed, constitutively active Ras promotes cell proliferation,
therapeutic targets.4 Several recent studies have reported that neoplastictransformationandtumorigenesis.14,17,18Oncogenic
miRNA expression is deregulated in human hepatocellular Ras proteins are commonly detected in human cancers,
carcinoma (HCC) in comparison with matched non- including B90% of pancreatic cancers, B70% of malignant
neoplastic tissue.5,6 The downregulated miRNAs in HCC neoplasias and B30% of all human cancers.19 Despite the
include several members of the large let-7 family and the central importance of Ras signaling in human cancer, direct
miR-143/145.7 Although for members of the let-7 family the inhibition of hyperactivated Ras has not achieved sufficient
1TumorImmunityMedicalResearchCenter,CancerResearchInstitute,SeoulNationalUniversityCollegeofMedicine,Seoul,Koreaand2Departmentof
Pathology,SeoulNationalUniversityCollegeofMedicine,Seoul,Korea
Correspondence:DrC-WKim,DepartmentofPathology,SeoulNationalUniversityCollegeofMedicine,28Yongon-dong,Jongno-gu,Seoul110-799,Korea.
E-mail:[email protected]
Received8August2012;accepted18September2012
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
2
clinicalefficacy.15,18,20WereasonedthatdownregulationofRas and cultured in Dulbecco’s Modified Eagle Medium supplemented
bythespecificmiRNAmightprovideamoreeffectivestrategy with10%fetalbovineserum,100Uml(cid:2)1penicillin,and100mgml(cid:2)1
for inhibiting oncogenic Ras signaling and Ras-mediated streptomycin (Gibco, Grand Island, NY, USA) in a humidified 5%
tumorigenesis in human cancer. CO2/95% airatmosphere at371C.
Adeninenucleotidetranslocase(ANT) isa nuclear-encoded
Antibodies and reagents
protein abundantly located in the inner mitochondrial mem-
Anti-phospho-Akt and anti-Akt antibodies were purchased from
brane,andtheroleofthisproteinistocatalyzetheexchangeof
CellSignalingTechnology (Beverly, MA, USA).
mitochondrial adenosine triphosphate with cytosolic adeno-
sine diphosphate. ANT therefore plays an important role in
Construction of ANT2 shRNA expression vector
cellular energy metabolism by influencing mitochondrial
The ANT2 shRNA expression vector used to achieve specific down-
oxidative phosphorylation. In addition, ANT is the major
regulation of ANT2 was described previously.30 ANT2 small-
component ofthe mitochondrial permeability–transition pore
interfering RNA (50-GCAGAUCACUGCAGAUAAGTT-30) that was
complex that interacts with Bcl2 family proteins, thereby
designed to be complementary to exon 2 of ANT2 (GenBank
contributing to mitochondria-mediated apoptosis.21,22 ANT-
accession number NM001152) was synthesized, and DNA vectors
deficient mice are able to form mitochondrial permeability– expressing the shRNA forms of the small-interfering RNAs were
transition pore complex, which has prompted discussion generatedusingpSilencer3.1-H1puroplasmidswithaTTCAAGAGA
regarding the roles of ANT in mitochondrial permeability– linker sequence that forms looped structures (Ambion, Austin, TX,
transition pore complex.23 Human ANT has four isoforms USA).ThevectorexpressingANT2shRNAwasusedthroughoutthis
(ANT1,ANT2,ANT3andANT4)andtherelativeexpressions study.AscrambleshRNA(Ambion)withnosignificanthomologyto
of these isoforms are dependent on developmental stages, human gene sequences was used as a control to detect nonspecific
proliferation status, as well as tissue types or cell types.14,15 effects.
ANT2 is specifically expressed in undifferentiated cells or
Transfection
tissues that are able to proliferate and regenerate; for
Fortransfection,cellswereplatedoneithersix-wellplates(2(cid:3)105cells
example, lymphocytes, kidney and liver. The expression
perwell)or100-mmdishes(2(cid:3)106cells),andallowedtoadherefor
of ANT2 was recently found to be upregulated in several
24h. Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was used
hormone-dependent cancers, and the induction of ANT2
forthetransfection.CellsweretransfectedwitheitherpSilencer3.1-H1
expression in cancer cells was shown to be directly associated
puro ANT2 shRNA (Ambion) or pSilencer 3.1-H1 puro scramble
with glycolytic metabolisms, raising a question regarding the shRNA vectors (Ambion). Transfected cells were cultured for 6h,
role of ANT2 during carcinogenesis.24–27 Therefore, ANT2 whereuponthemediumwasreplacedwithfreshmediumsupplemen-
repression effectively leads to cell growth arrest and increases tedwith10%FBS.Thecellswereharvested24–48haftertransfection.
mitochondrial membrane potential from human cells as well Negative mimics, miR-636 mimics, negative inhibitor or miR-636
aschemosensitizedcancercells,implyingthatANT2actsasan inhibitor was transfected intothecells using the same method.
antiapoptotic oncoprotein.17,28,29 Thus, ANT2 could be a
promising candidate for cancer therapy based on specific miRNAs microarray statistical analysis
miRNA regulation. Agilent miRNA assays (Agilent Technologies Inc, Santa Clara, CA,
USA) integrate eight individual microarrays on a single glass slide.
We hypothesized that upregulation of a specifc miRNA by
Each microarray includes approximately 15000 features containing
ANT2 short hairpin RNA (shRNA) may suppress HCC
probes sourced from the miRBASE public database. The probes are
tumorigenesis. Presently, miR-636 miRNA expression was
60-meroligonucleotidesthataredirectlysynthesizedonthearray.In
highlyupregulatedbyANT2suppression.miR-636restoration
this study we used the human microRNA microarray v2.0, which
resultedinadecreaseinlevelofRas,oneoftheputativetargets
contains723humanand76humanviralmiRNAs,eachreplicated16
ofmiR-636,andinactivationofitssignalingpathway.Thedata times. Three hundred sixty-two miRNAs are interrogated by two
suggest that miR-636 might function as a tumor suppressor different oligonucleotides, 45 miRNAs by three and 390 miRNAs by
miRNA affecting HCC tumorigenesis via downregulation of four.Only twomiRNAsareinterrogated byasingle oligonucleotide.
Ras, and that ANT2 suppression by shRNA could exert an Thearrayalsocontainsasetofpositiveandnegativecontrolsthatare
anticancer effect by restoring miR-636 expression in HCC. replicated a variety of times. Some of the positive control probes
Therefore, the purpose of our study was to determine target non-miRNA human RNAs. Each of these targets was inter-
rogatedwithfourdifferentprobes,withfivereplications.Thesignals
whether or not ANT2 shRNA-mediated miR-636 restoration
from these positive controls can be bright ordim depending on the
suppressed HCC cell proliferation, and to investigate the
sample and, according to Agilent, they do not behave consistently
potential of miR-636 for future use as a therapeutic agent
enough tobeused fornormalization.
targeting HCC cells.
Quantitative real-time RT–PCR (RT–qPCR)
MATERIALS AND METHODS ForRT–qPCR,totalRNAsfromthesamesamplesusedinthemiRNA
Cell lines and culture microarrayswere tested by using aniCylerIQreal-time PCR system
The human hepatoma cell line, Hep3B, HepG2 and SNU449 were (Bio-Rad, Hercules, CA, USA). Results are shown as fold-change.
purchased from the American Type Culture Collection (ATCC, RT–qPCRwascarriedoutwithMir-XmiRNAFirst-StrandSynthesis
Manassas,VA,USA)orKoreanCellLineBank(KCLB,Seoul,Korea) and SYBR Green Real time PCR Master Mix (TaKaRa, Otsu, Japan)
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
3
accordingtothemanufacturer’sinstructions.Thehousekeepinggene in100mlphosphate-bufferedsalineandtheninjectedsubcutaneously
U6 was used for standardization of the initial miRNA content of a into the right or left side of the posterior flank of six BALB/c nude
sample. Relative changes of gene expression were calculated by the mice,respectively.Tumorgrowthwasexamineddailyandthetumor
following formula, and the data were represented as fold upregula- volumes were calculated every week using the formula for hemi-
tion/downregulation: fold change¼2(cid:2)DDCt, where DDCt¼(Ct of ellipsoids: V¼length (cm)(cid:3)width (cm)(cid:3)height (cm)(cid:3)0.5236.
gene of interest, treated(cid:2)C of HK gene, treated)(cid:2)(Ct of gene After 5 weeks, the mice were killed and the tumors were dissected
t
ofinterest,control(cid:2)C ofHKgene,control),C isthethresholdcycle and photographed.
t t
number and HKisthe housekeeping gene.
Statistical analysis
Western blotting GraphPad Prism software (Brothersoft, Beijing, China) was used to
For western blot analysis, cells were harvested 24 or 48h after conductstatisticalanalysis.Resultsareexpressedasthemean±s.e.m.
transfection and lysed with lysis buffer consisting of 5mMl(cid:2)1 For P-value calculation, unpaired Student’s t-tests (two tailed) were
ethylenediamine tetraacetic acid, 300mMl(cid:2)1 NaCl, 0.1% NP-40, utilized for all comparisons. P-values o0.05 were considered as
0.5mMl(cid:2)1NaF,0.5mMl(cid:2)1Na3VO4,0.5mMl(cid:2)1phenylmethylsulfonyl significant.
fluoride, and 10mgml(cid:2)1 each of aprotinin, pepstatin and leupeptin
(Sigma-Aldrich, St Louis, MO, USA). Following centrifugation at RESULTS
15000g for 30min, the concentrations of supernatant proteins were ANT2 shRNA induces upregulation of miR-636 in Hep3B
analyzedusingtheBradfordreagent(Bio-Rad).Foranalysisofprotein HCC cells
contents,50mgtotalproteinwassubjectedto10%SDS-PAGEandthe
miRNA microarray screenings were done to identify ANT2
resolved proteins were transferred to polyvinylidene difluoride
shRNA-regulated miRNAs in Hep3B HCC cells (Figure 1a).
membranes (Millipore, Bedford, MA, USA), which were incubated
Amongthe887humanmiRNAsonthemicroarray,8miRNAs
with the antibodies indicated above. Immunoblots were visualized
wereupregulated,and14miRNAsweredownregulated,witha
using an enhanced chemiluminescence detection system (Amersham
greater than twofold difference between sc shRNA and ANT2
Pharmacia Biotech, Uppsala, Sweden).
shRNA transfection in Hep3B. As a result, we obtained
Cell proliferation assay significant candidate miRNAs and reaffirmed by selecting a
Cell proliferationwas measured using MTT (3-(4,5-dimethylthiazol- specific miRNA, miR-636, among the miRNAs. As shown in
2-yl)-2,5-diphenyltetrazolium bromide) (Sigma-Aldrich). The reduc- Figure 1b, miR-636 was upregulated by ANT2 shRNA.
tion of tetrazolium salts is now widely accepted as a reliable way to Furthermore, we confirmed this result in the other HCC cell
examinecellproliferation.TheyellowtetrazoliumMTTisreducedby lines, HepG2and SNU449(Supplementary Data 1).However,
metabolically active cells, in part by the action of dehydrogenase the functional attributes of miR-636 associated with liver
enzymes, to generate reducing equivalents such as nicotinamide tumor progression have not been experimentally established.
adenine dinucleotide and nicotinamide adenine dinucleotide phos-
Moreover, among candidate miRNAs, only miR-636 expres-
phate.The resulting intracellular purpleformazancan besolubilized
sion in HCC tissues was significantly lower than that in non-
and quantified spectrophotometrically. Results of cell proliferation
neoplastic liver tissue.
assaysarepresentedasthemeanvaluesofthreereplicateexperiments
performedintriplicate.
Downregulation of miR-636 in HCC compared with non-
neoplastic liver tissue
Luciferase assay
To see if miR-636 might have some biological relevance in
Hep3BcellsweretransfectedwithmiR-636oranegativecontrol(miR
HCC,RT–qPCRof20HCCtumortissuepairswasconducted.
control),eachofconcentration100nM,for6hinreducedserumand
antibiotics-freeOpti-MEMwithLipofectamine2000.After24h,cells Expression of miR-636 was markedly lower in HCC tissue
were transfected with Ras 30 untranslated region (UTR)-luciferase compared with matched non-neoplastic liver tissue
expressionvector.Fireflyluciferasewasmeasuredincelllysatesusing (Figure 1c), which implied that the downregulation of miR-
a Signosis Biosignal Capture Luc-Screen System on a Fusion plate 636 might be associated with tumorigenesis. We performed
reader(Perkin-Elmer,Waltham,MA,USA). Firefly luciferaseactivity Mann–Whitney test and w2 test for a two-sample test of
wasusedfornormalizationandasaninternalcontrolfortransfection miR-636 expression levels. For multiple-testing adjustment,
efficiency.
therewasasignificantdifferenceofmeanmiRexpressionlevel
between the two groups only if the P-value was o0.005. As
Soft agar colony formation assay
ANT2 is highly expressed in proliferating cells, and ANT2
Cells transfected with miR-636 or a negative control (miR control)
inductionincancercellsisdirectlyassociatedwithcarcinogen-
wereplatedintriplicateintheappropriatemediumcontaining0.3%
esis,25–27 ANT2 was enhanced in HCC tissue compared with
Bacto-agar on top of 0.6% bottom agar and cultured for 2 weeks.
non-neoplastic liver tissue (data not shown).
Colonies were visualized by staining with 0.1% p-iodonitro tetra-
zoliumviolet(Sigma-Aldrich).Monolayercolonieswerestainedwith
crystal violetdye aftermethanolfixation. RestorationofmiR-636byANT2shRNAdownregulatesRas
gene transcripts, in which Ras is a critical downstream
Tumorigenicity assays in nude mice target of miR-636
For in vivo tumorigenicity assay, negative mimics (negative control) K-Ras or H-Ras mRNAs were significantly decreased in cells
ormiR-636mimic-transfectedHep3Bcells(2(cid:3)105)weresuspended with overexpression of miR-636 or cells in which ANT2
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
4
[RT-qPCR]
[miRNA microarray] 4
+4 3.5 p<0.05
+3
++---210123 5 miR1-6336 15 Relative miR-636expression (2-ddCt) 12..15253
-4 0.5
sc shRNA vs ANT2 shRNA 0
sc shRNA ANT2 shRNA
[RT-qPCR]
1.4
p<0.005
1.2
36dCt) 1
6d
miR-n (2- 0.8
ve sio 0.6
aties
Relxpr 0.4
e

0.2
0
Normal liver tissue HCC tissue
(n=20) (n=20)
Figure 1 Knockdown of adenine nucleotide translocase 2 (ANT2) by short hairpin RNA (shRNA) induces specific restoration of miR-636
in Hep3B and expression of miR-636 was remarkably lower in hepatocellular carcinoma (HCC) tissue. (a) RNA extraction was performed
24haftertransfectionofHep3BcellswithANT2shRNAformicroRNA(miRNA)microarrayanalysis.Thearrayswereprogressedusingan
Agilent scanner and microarray data extracted with Agilent Feature Extraction software. (b) Quantitative real-time RT–PCR (RT-qPCR)
analysis of the effect of ANT2 shRNA-induced restoration on miR-636 expression. Cells were transfected with scramble or ANT2 shRNA.
Total RNA was extracted 24h after transfection and subjected to RT-qPCR using specific primers for human miR-636 or U6 (internal
control). (c) RT–qPCR of 20 HCC tumor tissue pairs was conducted. RNAs extraction from the paraffin block was used by the Ambion
mirVanamiRNAisolationkitandsubjectedtoRT–qPCRusingspecificprimersforhumanmiR-636orU6(internalcontrol).
expressionwas knockdown (Figure 2a). Bioinformatic predic- levels(Figure2c).Furthermore,weconfirmedthisresultinthe
tion recognized 1–7 miR-636 binding sites in the 30UTRs of other HCC cell lines, HepG2 and SNU449 (Supplementary
theRasgene.Totestwhetherthepredictedbindingsitesinthe Data 3). Moreover, we confirmed that phosphorylated Akt
target genes could mediate repression of translation by levels were decreased by knockdown of ANT2 by shRNA and
miR-636(Figure2b).Ras30UTRsweresubcloneddownstream slightly recovered by miR-636 inhibitor in Figure 2d. Taken
of the luciferase gene. The reporter with the 30UTR of Ras together,thesefindingssuggestedthatANT2shRNAtreatment
showed a significantly lower luciferase activity in overexpres- led to restoration of miR-636 expression in Hep3B cells,
sion of miR-636 or knockdown of ANT2 cells (Figure 2b). accompanied by suppression of Ras/PI3K/Akt signaling. It is
TheresultwasconfirmedintheHepG2andSNU449HCCcell well known that Akt plays important roles as a mediator of
lines (Supplementary Data 2). Taken together, the evidence Ras-mediated cell survival and proliferation; as a result,
demonstrates that Ras is a critical downstream target of inhibition of Akt by Akt-specific inhibitor also inhibits cell
miR-636. proliferation of Hep3B (Supplementary Data 4).
Upregulation of miR-636 by ANT2 shRNA inactivates the Upregulation of miR-636 by ANT2 shRNA inhibits cell
Ras/phosphoinositide 3-kinase (PI3K)/Akt signaling proliferation and suppresses colony formation
pathway With regard to the biological functions of miR-636, we
Restoration of miR-636 by ANT2 shRNA downregulates Ras observed that restoration of miR-636 expression suppresses
transcripts and inactivates the Ras/PI3K/Akt signaling HCC cell proliferation. The knockdown of ANT2 and inhibi-
pathway. Ras protein is among the best-characterized proto- tion of miR-636 inhibited cell proliferation of Hep3B
oncogenes and Ras signaling regulates diverse cellular func- (Figure 3a). Overexpression of miR-636 by miR-636 mimics
tions, including cell growth, survival and migration. With (Supplementary Data 5) or ANT2 knockdown by ANT2
regard to the regulation of Ras/Mek/Erk or Ras/PI3K/Akt shRNA also inhibited cell proliferation of Hep3B.30
activity, we found that phosphorylated Akt levels were Overexpression of miR-636 also impaired anchorage-
decreased by miR-636 mimics but not phosphorylated Erk independent growth in soft agar, significantly reducing the
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
5
K-Ras
H-Ras [RT-qPCR]
Relative expression (2-ddCt) 000011......02468124 p<p0<.000.0505 p<p0<.00.5005 Relative Ras-UTRLuciferase activity (Fold) 00001.....246812 p<0.[0R0e5porter gene assayp]<0.005
Negative miR-636 sc ANT2 0
Negative miR-636 sc ANT2
shRNA shRNA
shRNA shRNA
[Western-blot] [Western-blot]
Negative mimics + - sc shRNA + - + -
miR-636 mimics - + ANT2 shRNA - + - +
Negative mimics - - + -
P-Akt miR-636 mimics - - - +
P-Akt
Akt
Akt
P-Erk
P-Erk
Erk Erk
Figure2 miR-636regulatesRasexpressionatthepost-transcriptionallevelandrestorationofmiR-636byadeninenucleotidetranslocase
2(ANT2)shorthairpinRNA (shRNA) suppressesRas/phosphoinositide3-kinase/Aktsignaling.(a) After 24h oftransfectionwithnegative/
miR-636 mimics or scramble/ANT2 shRNA, total RNA was extracted and subjected to RT-qPCR using specific primers for human K-Ras,
H-Ras or glyceraldehyde 3-phosphate dehydrogenase as the internal control. (b) Human embryonic kidney 293 cells were transiently
co-transfected with Ras–30 untranslated region-luciferase reporter gene vectors and then stimulated with negative/miR-636 mimics or
scramble/ANT2 shRNA for 18h. Luciferase activities in cell extracts were analyzed by dual-luciferase reporter assay system and
normalizedusingpRL-TK-luciferaseactivity(Renillaluciferaseactivity)ineachsample.Datapresentthemean(foldincreaseinluciferase
activitycomparedwithcells stimulatedwithnegativemimicsorscrambledshRNA)±s.d.oftriplicates.(c)After24h oftransfectionwith
negativeormiR-636mimics,cellextractswerepreparedforwesternblottingwithantiphospho-Akt/anti-Aktantibodiesorantiphospho-Erk/
anti-Erk antibodies. (d) Hep3B cells were transfected with scrambled/ANT2 shRNA and/or negative/miR-636 inhibitor, and cell extracts
werepreparedforwesternblottingwithantiphospho-Akt/anti-Aktantibodyorantiphospho-Erk/anti-Erkantibody.
number and size of the colonies. Overexpression of miR-636 DISCUSSION
also reduced the proliferation of Hep3B cells (Figure 3b). Microarrays have been used in several studies of miRNA
Moreover, there was significant reduction in cell proliferation expression profiling in HCC. One of the earliest works to
as well as lower anchorage-independent growth in soft agar detect the relationship between miRNAs and HCC31 reported
(Figure 3c). that three miRNAs were expressed more exuberantly in HCC
samples than in nontumorous tissues, while five miRNAs had
lower expression in HCC tissue. The prediction accuracy of
Overexpression of miR-636 suppresses in situ tumor classifyingHCC and nontumorous tissues with eightmiRNAs
formation was 97.8%. Another study reported that 18 miRNAs were
We determined whether miR-636 and its target gene are overexpressed in HCC and 6 miRNAs were overexpressed in
associated with tumorgenesis. miR-636 mimics transfected nontumorous samples.32 Aberrant expression of miR-21
Hep3Bcellsimplantedsubcutaneously intotheflanksofnude contributes to progression and metastasis of HCC, and
mice. Overexpression of miR-636 significantly suppressed miR-21 can participate in the process of HCC development
overall tumor growth, as assessed by tumor volume by targeting phosphatase and tensin homolog (PTEN33).
(Figure4).Thus,miR-636wouldseemtoregulatetumorigen- Another study used a supervised algorithm to predict
esis, at least inpart through its target Ras.Ras expressionwas metastasis characterization in the examination of miRNA
significantlydecreased in miR-636-overexpressed tumor tissue expression profiles of 482 cancerous and noncancerous
(data not shown). Consistent with the in vitro results, over- specimens from radical resection of 241 patients with
expression of miR-636 Hep3B cells significantly suppressed HCC.34 In comparative miRNA expression studies, several
overalltumorgrowthinvivo.Theseinvitroandinvivoresults miRNAs that are important in HCC initiation, development
imply that the gradual loss of miR-636 over a long period of andmetastasis,suchasmiR-21,havebeenidentified.Aberrant
time is likely to commit hepatocytes to tumorigenesis. expression of miR-21 can contribute to HCC growth and
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
6
[MTT assay] [MTT assay]
1.2
y 1 p<0.05 p<0.05 1.4 K-Ras p<0.05 p<0.005
abilit 0.8 bility 1.12 H-Ras
cell vi 0.6 ell via 0.8
Relative 00..24 Relative c 000...246
0 0
sc shRNA + - + - 24 48 72
ANT2 shRNA - + - + After transfection (h)
Negative mimics - - + -
miR-636 mimics - - - +
[Soft agar colony formation assay]
60
well 50
ers/ 40
b
m 30
nu 42% reduction
y 20
n
o
ol 10
C
0
Negative miR-636-
mimics mimics
Negative miR-636-
mimics mimics
Figure 3 Restoration of miR-636 by adenine nucleotide translocase 2 (ANT2) short hairpin RNA (shRNA) inhibits cell proliferation and
upregulation of miR-636 by ANT2 shRNA suppresses colony formation. (a) Hep3B cells were transfected with scramble, ANT2 shRNA,
negative or miR-636 inhibitor. Cell viability (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)) assays were performed
24h after transfection. Three independent experiments were performed. Data were analyzed using the Student’s t-test. (b) Hep3B cells
were transfected with negative or miR-636 mimics. Cell viability (MTT) assays were performed 24, 48 and 72h after transfection. Three
independent experiments wereperformed.Datawereanalyzedusing the Student’s t-test.(c) Colonyformation after transfectionofHep3B
cellswithnegativeormiR-636mimics.
spreadby modulatingPTENexpressionandPTEN-dependent
pathways involved in mediating the phenotypiccharacteristics
of cancer cells, such as cell growth, migration and invasion.33
ThefunctionofmiR-636hasyetnotbeenstudiedincancer
cells. The only reports on miR-636 describe miR-636 as a
potential candidate responsible for the post-transcriptional
silencing of GR in GC-resistant cells.35 For the first time, our
study suggests that miR-636 is a tumor suppressor miRNA
affecting HCC tumorigenesis via suppression of the Ras
signaling pathway. Restoration of miR-636 efficiently
suppressed Hep3B cell proliferation in vitro and
tumorigenesis in vivo at a level comparable to ANT2 shRNA
treatment.
Physiological and oncogenic activation of Ras stimulates a
wide range of downstream signaling pathways. The first Ras
effector pathway to be identified was the Raf–Mek–Erk Figure 4 Overexpression of miR-636 suppresses in situ tumor
pathway.36–39 This pathway is an essential shared element formation.Balb/cnudemicewerechallengedwithnegativeormiR-
636 mimic-transfected 2(cid:3)105 Hep3B cells by subcutaneous
of mitogenic signaling involving tyrosine kinase receptors,
injection into the left or right flanks. Tumor sizes were measured
leading to a wide range of cellular responses, including
by a caliper every week and volumes calculated using the formula:
growth, differentiation, inflammation and apoptosis.40 The m12(cid:3)m2(cid:3)0.5236, where m1 represents the short tumor axis and
second best-characterized Ras effector family is PI3Ks, which m thelongaxis,untilday73followingtumorchallenge.
2
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
7
play important roles as mediators of Ras-mediated cell 14Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100:
survival and proliferation.41,42 PI3Ks are important regu- 57–70.
15Downward J. Targeting RAS signalling pathways in cancer therapy.
lators of cellular growth, transformation, adhesion, apoptosis,
NatRevCancer2003;3:11–22.
survival and motility.43–46 Ras-dependent PI3K activation is 16Schubbert S, Shannon K, Bollag G. Hyperactive Ras in developmental
linked to prosurvival signaling due to the activation of Akt. disordersandcancer.NatRevCancer2007;7:295–308.
17BosJL.Rasoncogenesinhumancancer:areview.CancerRes1989;49:
Our findings that inactivated PI3K or Akt induces apoptosis
4682–4689.
and that dominant-negative Akt has the ability to enhance 18DownwardJ.Cancerbiology:signaturesguidedrugchoice.Nature2006;
apoptosis highlight the importance of PI3K/Akt signaling in 439:274–275.
promoting survival.41,47–49 Moreover, our results demonstrate 19Roberts PJ, Der CJ. Targeting the Raf-MEK-ERK mitogen-activated
proteinkinasecascadeforthetreatmentofcancer.Oncogene2007;26:
the importance of Ras–PI3K interaction in Ras-driventumor- 3291–3310.
igenesis. Although multiple Ras effectors are essential to 20Sebolt-Leopold JS, Herrera R. Targeting the mitogenactivated protein
kinasecascadetotreatcancer.NatRevCancer2004;4:937–947.
initiate tumor formation, only signaling through the PI3K/
21MarzoI,BrennerC,ZamzamiN,SusinSA,BeutnerG,BrdiczkaDetal.
Aktpathwayseemstobenecessarytomaintaintumorgrowth. Thepermeabilitytransitionporecomplex:atarforapoptosisregulationby
The reduction of Ras oncogene by miR-636 induces inactiva- caspaseandbcl-2-relatedproteins.JExpMed1998;187:1261–1271.
22Marzo I, Brenner C, Zamzami N, JurgensmeierJM, Susin SA,VieiraHL
tion of the PI3K/Akt pathway and inhibits tumor growth.
etal.Baxandadeninenucleotidetranslocatorcooperateinthemitochon-
Tumor maintenance in Ras-driven tumor growth is required drialcontrolofapoptosis.Science1998;281:2027–2031.
for PI3K signaling. 23KokoszkaJE,WaymireKG,LevyS,SlighJE,CaiJ,JonesDPetal.The
ADP/ATP translocator is not essential for the mitochondrial permeability
In summary, the deduced expression profiles in HCC cell
transitionpore.Nature2004;427:461–465.
lines show that miR-636 can act as a tumor suppressor 24Faure VH, Heddi A, Giraud S, Chautard D, Stepien G. Expression of
through the direct control of Ras expression. The restoration oxidativephosphorylationgenensinrenaltumorsandtumoralcelllines.
MolCarcinog1996;16:165–172.
of miR-636 by ANT2 shRNA contributes to suppression of
25ChevrollierA,LoiseauD,ChabiB,RenierG,DonayO,MalthieryYetal.
tumorformation.ANT2suppressionbyshRNAmightbeable ANT2 isoform required for cancer cell glycolysis. J Bioenerg Biomembr
to exert an anticancer effect in HCC by regulation of tumor 2005;37:307–316.
26Chevrollier A, Loiseau D, Gautier F, Malthlery Y, Stepien G. ANT2
suppressor miRNAs.
expressionunderhypoxicconditionsproducesoppositecell-cyclebehavior
in143BandHepG2cancercells.MolCarcinog2005;42:1–8.
27ChevrollierA,LoiseauD,StepienG.WhatisthespecificroleofANT2in
ACKNOWLEDGEMENTS cancercells?MedSci(Paris)2005;21:156–161.
Thiswork wassupported by the Global Core ResearchCenter 28LeBrasM,Borgne-SanchezA,TouatZ,ElDeinOS,DeniaudA,MaillierE
etal.Chemosensitizationbyknockdownofadeninenucleotidetranslocase-
(GCRC) GrantNo. 2012-0001190fromtheNational Research
2.CancerRes2006;66:9143–9152.
Foundation (NRF),MinistryofEducation, Science and Technology 29Jang JY, Kim MK, Jeon YK, Joung YK, Park KD, Kim CW. Adenovirus
(MEST), RepublicofKorea, also by GrantNo SS100003fromthe adenine nucleotide translocator-2 shRNA effectively induces apoptosis
Seoul R&BD Program. andenhanceschemosensitivitybythedown-regulationofABCG2inbreast
cancerstem-likecells.ExpMolMed2012;44:251–259.
30Jang JY, Choi Y, Jeon YK, Kim CW. Suppression of adenine nucleotide
translocase-2byvector-basedsiRNAinhumanbreastcancercellsinduces
apoptosisandinhibitstumorgrowthinvitroandinvivo.BreastCancerRes
1 Rosa V, Carlo CM. MiRNAs and cancer. Am J Pathol 2009; 174: 2008;10:1–14.
1131–1138. 31Murakami Y, Yasuda T, Saigo K, Urashima T, Toyoda H, Okanoue T
2 IorioMV,CroceCM.MicroRNAinvolvementinhumancancer.Carcinogen- et al. Comprehensive analysis of microRNA expression patterns in
esis2012;33:1126–1133. hepatocellular carcinoma and non-tumorous tissues. Oncogene 2006;
3 BaohongZ,XiaopingP,GeorgePC,ToddAA.MicroRNAsasoncogenesand 25:2537–2545.
tumorsuppressors.DevBiol2007;302:1–12. 32Huang YS, Dai Y, Yu XF, Bao SY, Yin YB, Tang M et al. Microarray
4 Scott AW, Andr T. MicroRNA signatures as diagnostic and therapeutic analysis of microRNA expression in hepatocellular carcinoma and non-
targets.ClinChem2008;54:943–944. tumorous tissues without viral hepatitis. J Gastroenterol Hepatol 2008;
5 Laura G, Francesca F, Elisa C, Silvia S, Giovanni L, Carlo MC et al. 23:87–94.
MicroRNAinvolvementinhepatocellularcarcinoma.CellMolMed2008; 33Meng F, Henson R, Wehbe-Janek H, Ghoshal K, Jacob ST, Patel T.
12:2189–2204. MicroRNA-21regulatesexpressionofthePTENtumorsuppressorgenein
6 ManuelaF,SylwiaEW,RamiroG,HansjuergA,RamanaD,Chang-GongL humanhepatocellularcancer.Gastroenterology2007;133:647–658.
et al. A microRNA signature of Hypoxia. Mol Cell Biol 2007; 27: 34Budhu A, Jia HL, Forgues M, Liu CG, Goldstein D, Lam A et al.
1859–1867. Identification of metastasis-related microRNAs in hepatocellular carci-
7 PanagiotaC,KaterinaV,IoannisS,ChristosP,EvangelosB.Alterationsof noma.Hepatology2008;47:897–907.
microRNAsinsolidcancersandtheirprognosticvalue.Cancers2010;2: 35Michael AT, Ashley LB, Nancy LK, Steven TR, Preethi HG. Role for
1328–1353. MicroRNAsinregulatingglucocorticoidresponseandresistanceinmulti-
8 BenjaminB,ParkSM,AnnikaH,AndreaEM,MarcusEP,EndocrR.The plemyeloma.HormCancer2011;2:182–189.
roleoflet-7incelldifferentiationandcancer.Cancer2010;17:F19–F36. 36MoodieSA,WillumsenBM,WeberMJ,WolfmanA.ComplexesofRas.GTP
9 JunX,YaxingX,JunliH,SaifengW,ChangfeiL,SongdongM.MiR-122in with Raf-1 and mitogen-activated protein kinase kinase. Science 1993;
hepaticfunctionandliverdiseases.ProteinCell2012;3:364–371. 260:1658–1661.
10Bollag G, McCormick F. Regulators and effectors of ras proteins. 37Warne PH, Viciana PR, Downward J. Direct interaction of Ras and the
AnnuRevCellBiol1991;7:601–632. amino-terminalregionofRaf-1invitro.Nature1993;364:352–355.
11CampbellSL,Khosravi-FarR,RossmanKL,ClarkGJ,DerCJ.Increasing 38VojtekAB,HollenbergSM,CooperJA.MammalianRasinteractsdirectly
complexityofRassignaling.Oncogene1998;17:1395–1413. withtheserine/threoninekinaseRaf.Cell1993;74:205–214.
12MarshallC.HowdosmallGTPasesignaltransductionpathwaysregulate 39Zhang ZF, Settleman J, Kyriakis JM, Takeuchi-Suzuki E, Elledge SJ,
cellcycleentry?CurrOpinCellBiol1999;11:732–736. Marshall MS et al. Normal and oncogenic p21ras proteins bind to
13ShieldsJM,PruittK,McFallA,ShaubA,DerCJ.UnderstandingRas:‘it the amino-terminal regulatory domain of c-Raf-1. Nature 1993; 364:
ain’tover‘tilit’sover’.TrendsCellBiol2000;10:147–154. 308–313.
Experimental&MolecularMedicine
AnticancereffectbyrestoringmiR-636expression
J-YJangetal
8
40RouxPP,BlenisJ.ERKandp38MAPKactivatedproteinkinases:afamily 46CourtneyKD,CorcoranRB,EngelmanJA.ThePI3Kpathwayasdrugtarget
ofproteinkinaseswithdiversebiologicalfunctions.MicrobiolMolBiolRev inhumancancer.JClinOncol2010;28:1075–1083.
2004;68:320–344. 47CoxAD,DerCJ.ThedarksideofRas:regulationofapoptosis.Oncogene
41VivancoI,SawyersCL.Thephosphatidylinositol3-kinaseAKTpathwayin 2003;22:8999–9006.
humancancer.NatRevCancer2002;2:489–501. 48DownwardJ.Rassignallingandapoptosis.CurrOpinGenetDev1998;8:
42CastellanoE,DownwardJ.RoleofRASintheregulationofPI3-kinase. 49–54.
CurrTopMicrobiolImmunol2011;346:143–169. 49DownwardJ.PI3-kinase,Aktandcellsurvival.SeminCellDevBiol2004;
43CantleyLC.Thephosphoinositide3-kinasepathway.Science2002;296: 15:177–182.
1655–1657.
44VoliniaS,HilesI,OrmondroydE,NizeticD,AntonacciR,RocchiMetal.
ThisworkislicensedunderaCreativeCommons
Molecularcloning,cDNAsequence,andchromosomallocalizationofthe
human phosphatidylinositol 3-kinase p110 alpha (PIK3CA) gene. Geno- Attribution-NonCommercial-NoDerivs 3.0 Un-
mics1994;24:472–477. ported License. To view a copy of this license, visit http://
45FrumanDA,MeyersRE,CantleyLC.Phosphoinositidekinases.AnnuRev
creativecommons.org/licenses/by-nc-nd/3.0/
Biochem1998;67:481–507.
SupplementaryInformationaccompaniesthepaperontheExperimental&MolecularMedicinewebsite(http://www.nature.com/emm)
Experimental&MolecularMedicine