PrognosticSigniﬁcanceofAMPKActivationandTherapeutic ...clincancerres.aacrjournals.org/content/clincanres/early/...log-rank tests, and Cox regression analysis were used for survival

Human Cancer Biology

Prognostic Significance of AMPKActivation and TherapeuticEffects of Metformin in Hepatocellular Carcinoma

Longyi Zheng1,2,3, Wen Yang1,3, Fuquan Wu1,3, Chao Wang1,3, Lexing Yu1,3, Liang Tang1,3, Bijun Qiu3,4,Yuqiong Li1,3, Linna Guo1,3, Mengchao Wu1,3, Gensheng Feng1,3, Dajin Zou2, and Hongyang Wang1,3,5

AbstractPurpose: The AMP-activated protein kinase (AMPK) serves as an energy sensor in eukaryotic cells and

occupies a central role in linking metabolism and cancer development. However, the phosphorylation

status of AMPK and its therapeutic value in human hepatocellular carcinoma (HCC) remain unclear.

Experimental Design: The phosphorylation status of AMPK (Thr172) was determined by immuno-

blotting and immunostaining in specimens from 273 patients with HCC (including 253 patients with

hepatitis B virus -related HCC). Kaplan–Meier survival analysis was used to determine the correlation

with prognosis. The effects of therapeutic metformin/AMPK activation were assessed in cultured human

HCC cell lines and primary HCC cells in vitro and in xenograft tumors model in vivo. To define the

mechanisms of anticancer effects of metformin, we examined its influence on AMPK activation and

NF-kB pathway.

Results: AMPK is dysfunctional in patients with HCC, and low p-AMPK staining is correlated with

aggressive clinicopathologic features and poor prognosis. Activation of AMPK by metformin not only

inhibited HCC cells growth in vitro and in vivo, but also augmented cisplatin-induced growth inhibition in

HCC cells. Knockdown of AMPKa expression can greatly decrease the inhibitory effect of metformin,

indicating that AMPK activation is required for the anticancer action of metformin. Mechanistically,

metformin/AMPK activation inhibited NF-kB signaling through upregulation of IkBa. Activation of NF-

kB signaling by ectopic expression of P65 or overexpression of an undegradable mutant form of IkBaattenuated the anticancer effects of metformin.

Conclusions: These results present novel insight into a critical role of AMPK in HCC progression.

Anticancer effects of therapeuticmetformin/AMPK activation unravelmetformin’s potential in treatment of

HCC. Clin Cancer Res; 1–9. �2013 AACR.

IntroductionHepatocellular carcinoma (HCC) is the third leading

cause of cancer-related mortality worldwide (1). Although

the short-term prognosis of HCC has been improved, thelong-term prognosis of HCC remains dismal even afterradical excision. Therefore, it is necessary to identify noveleffective therapeutic strategies for the management ofHCC (2).

The AMP-activated protein kinase (AMPK) is a highlyconserved heterotrimeric serine/threonine kinase. It servesas an energy sensor in all eukaryotic cells (3) and alsooccupies a central role in linking metabolism and cancerdevelopment (4). AMPK is activated typically by an increasein the cellular AMP/ATP ratio under conditions such asglucose deprivation, hypoxia, ischemia, and heat shock (5).In addition, it is also activated by several hormones andcytokines such as adiponectin and leptin and by the anti-diabetic drug metformin (6). Once activated, AMPK pro-motes energy production and limits energy usage to ensurecellular survival (5).

Various molecules and signaling pathways have beenidentified to be regulated by AMPK (7). Recent studies alsoindicate that AMPK activation strongly suppresses cell pro-liferation in nonmalignant cells, as well as in tumor cells.AMPKmediates these effects throughmultiple mechanisms

Authors' Affiliations: 1International Cooperation Laboratory on SignalTransduction, Eastern Hepatobiliary Surgery Institute; 2Department ofEndocrinology, Changhai Hospital, Second Military Medical University;3National Center for Liver Cancer Research; 4Liver Transplant Centers andHepatic Surgery; and 5State Key Laboratory of Oncogenes and relatedGenes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao TongUniversity School of Medicine, Shanghai, People's Republic of China

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

L. Zheng, W. Yang, and F. Wu contributed equally to this work.

Corresponding Authors: Hongyang Wang, International CooperationLaboratory on Signal Transduction, Eastern Hepatobiliary Surgery Insti-tute/Hospital, 225Changhai Road, Shanghai 200438, People's Republic ofChina. Phone: 86-21-81875361; Fax: 86-21-65566851; E-mail:[email protected]; and Dajin Zou, Department of Endocrinology,Changhai Hospital, 168 Changhai Road, Shanghai 200433, People'sRepublic of China. Phone: 86-21-81873279; Fax: 86-21-53961632; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-13-0203

�2013 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org OF1

Cancer Research. on January 18, 2020. © 2013 American Association forclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 13, 2013; DOI: 10.1158/1078-0432.CCR-13-0203

http://clincancerres.aacrjournals.org/

including regulationof cell cycle, apoptosis, autophagy, andinhibition of protein synthesis, de novo fatty acid synthesis(8). Altered levels of AMPK have been linked with manyhuman diseases including cancer, and the modulation ofAMPK has emerged as an important target for the treatmentof obesity, diabetes, and cancer.

Despite the growing evidence to a link between AMPKand cancer, little is knownabout the phosphorylation statusof AMPK and its significance in human HCC. The objectiveof the present study was to evaluate the state of AMPKphosphorylation in HCC and determine its prognosticvalue in patients with HCC. In particular, we explored theeffects of therapeutic metformin/AMPK activation on HCCcell growth in vitro and in vivo. We also identified thesignaling pathway by which AMPK activation inhibits HCCcell growth.

Materials and MethodsCollection of human tissue specimen

Ten human normal liver tissues were obtained from thedistal normal liver tissue (at least 2 cm away from thehemangioma front) of patients with liver hemangioma.Formalin-fixed and paraffin-embedded HCC tissues from273 consecutive patients and fresh-frozenHCC tissues from19 patients who underwent radical resection in EasternHepatobiliary Surgery Hospital (Shanghai, China) fromSeptember 2005 to July 2009 were retrieved for immuno-histochemical or Western blot analysis. All human samplecollection procedures were approved by the Ethical ReviewCommittee of theHospital. Informed consentwas obtainedin all cases before surgery. The diagnosis of HCC wasconfirmed by pathologic results. All patients were followedupuntilDecember 2011,with amedianobservation timeof32months.Overall survival (OS)was defined as the intervalbetween the dates of surgery and death. Time to recurrence(TTR) was defined as the interval between the dates of

surgery and first recurrence of disease; if recurrence was notdiagnosed, patients were censored on the date of death orthe last follow-up. First recurrence was classified as recur-rence, distant recurrence, or a combination of both (9).

Statistical analysisValues presented are expressed as mean � SD. After

acquiring all data for histologic parameters and in vitroassays, Fisher exact test and Student t test were applied todetermine statistical significance. Kaplan–Meier analysis,log-rank tests, and Cox regression analysis were used forsurvival analysis. Data analysis was conducted by the SPSSsoftware (version 16; SPSS).

Chemicals and reagents, histopathologic and immuno-histochemical evaluation, cell culture, RNA extraction andreal-time PCR (RT-PCR), Western blot assay, dual-lucifer-ase reporter system, ELISA, in vivo tumorigenicity experi-ments, siRNA transfections, and isolation of primary HCCcells have been described in Supplementary Materials andMethods.

ResultsExpression of p-AMPK in HCC tissues

Wefirst evaluated the status of AMPK phosphorylation innormal human liver and paired tumorous and nontumor-ous HCC samples. As shown in Fig. 1A and B, normal livertissues showed strong p-AMPK (Thr172) expression. In 18of the 19 specimens, significant decreased level of p-AMPK(normalized by corresponding total AMPK expression lev-el) was observed in tumors relative to paired nontumoroustissues. Furthermore, immunohistochemical analysis alsoshowed that p-AMPK level was down-regulated in 61.8%(81/131) of the patients with HCC (Fig. 1C). These datarevealed that p-AMPK expression was downregulated in avast majority of HCC tissues.

Low p-AMPK expression correlated with poorprognosis of HCC

Furthermore, based on the immunohistochemical anal-ysis of HCC tissues, all 273 patients with HCCwere dividedinto two groups: high p-AMPK expression group (n ¼ 76)and low p-AMPK expression group (n¼ 197). Intriguingly,as shown in Table 1, patients in low-expression group weresignificantly associated with aggressive clinicopathologicfeatures [high-serum a-fetoprotein (AFP) level, incompletetumor encapsulation, late tumor–node–metastasis (TNM)stage, portal venous invasion, and distant metastasis]. Tofurther evaluate its prognostic value, p-AMPK expressionwas examined by immunohistochemistry (IHC) in sectionsfrom 252 HCC specimens with TTR and OS time. Kaplan–Meier survival analysis revealed that the prognosis of thepatientswithHCCwith lower p-AMPKexpression (n¼183)in tumor tissues was worse than those with higher p-AMPKexpression (n ¼ 69; Fig. 1D). Then multivariate survivalanalysis was conducted to identify the prognostic factors forrecurrence. As shown in Table 2, low p-AMPK expressionwas found to be an independent poor prognostic factor for

Translational RelevanceHepatocellular carcinoma (HCC) is a major health

problem with 700,000 new cases per year worldwide.The AMP-activated protein kinase (AMPK) occupies acentral role in linking metabolism and cancer develop-ment. In this study, through large-scale analysis of speci-mens, we showed for the first time that AMPK is dys-functional in patients with HCC, and low p-AMPKstaining is correlated with aggressive clinicopathologicfeatures and poor prognosis. Moreover, in vitro andin vivo study also revealed that activation of AMPK bymetformin inhibited NF-kB and STAT3 signaling activ-ity, and thus inhibited HCC cells growth. Therefore, lowp-AMPK expression could serve as a valuable predictingfactor for recurrence and poor survival of patients withHCC, and therapeutic metformin/AMPK activationcould be useful for the treatment of HCC.

Zheng et al.

Clin Cancer Res; 2013 Clinical Cancer ResearchOF2




recurrence in the patient. These data suggested that lowp-AMPK expression correlated with recurrence and poorsurvival of patients with HCC.

Therapeutic metformin/AMPK activation inhibits HCCcells growth in vitroTo address whether modulation of p-AMPK level can

affect HCC cells growth, we aimed to activate AMPK in HCCcells. Metformin, a widely used drug for treatment of type IIdiabetes, is known to activate AMPK in various tissues. Asshown in Fig. 2A and Supplementary Fig. S2, treatment ofHCC cells with metformin led to significant activation ofAMPK and inhibited the growth of HCC cells at a dose andtime-dependent manner. Next, we examined whether met-formin’s inhibitory effect depended on activation of AMPK.For this, we first used AICAR (5-Aminoimidazole-4-carbox-amide 1-b-D-ribofuranoside), an AMPK-specific activator. Asshown in Supplementary Fig. S3, treatment with AICAR alsoshowed inhibitory effect on HCC cells proliferation. Then,knockdown of AMPK isoforms was conducted by transient

transfection of siRNA oligos against two catalytic subunitsof AMPKa1 and a2 (Supplementary Fig. S4). We observedthat knockdown AMPKa1/2 expression can greatly decreasethe inhibitory effect of metformin (Supplementary Fig. S5).These data indicated that metformin inhibited HCC cellgrowth mainly through activation of AMPK.

Therapeutic metformin/AMPK activation blocks cellcycle, induces cell apoptosis, and sensitizes HCC cellstoward chemotherapy

We further investigate the effect of metformin/AMPKactivation on other biologic characteristics of HCC cells.Metformin treatment caused cell-cycle block in G0–1 andS phase, induced apoptosis, and decreased clone forma-tion ability of HCC cells (Fig. 2B and Supplementary Fig.S6). Chemoresistance is a major obstacle to the efficacy ofchemotherapeutic treatment of HCC. We next studied thepotential effect of metformin/AMPK activation on che-mosensitivity. As shown in Fig. 2C and SupplementaryFig. S7, cisplatin exerted a stronger inhibitory effect on

Figure 1. p-AMPKdownregulation is a commonevent inHCC tissues, and lowp-AMPKexpression correlateswith poor prognosis ofHCC.A–C, representativeWestern blot analysis showing (A and B) and immunohistochemical staining (C) of the expression of p-AMPK in normal liver tissues (N), tumor tissues (T),and paired peritumor tissues (L). B, p-AMPK/AMPK ratios of tumor tissues were also expressed as fold change of paired peritumor tissues. D, the OStime and TTR of 252 patients with HCC were compared between the low and high p-AMPK groups.

Anticancer Effect of Therapeutic Metformin/AMPK Activation on HCC

www.aacrjournals.org Clin Cancer Res; 2013 OF3




cells treated with metformin. Because PTEN (deleted onchromosome ten)/Abcg2 has been implicated in HCCresistance toward chemotherapy (10), we next investigat-ed whether this suppressive growth effect of metfomin wasmediated through modulation of the PTEN pathway.Western blot analysis revealed that metformin/AMPKactivation increased PTEN protein activity, which wasaccompanied by decreased expression levels of Abcg2,wherever knockdown AMPKa1/2 expression abolished thiseffect (Fig. 2C and D).

Therapeutic metformin/AMPK activation inhibitedNF- kB and IL-6/STAT3 signaling activity

Because NF-kB is critically involved in the regulation ofdivergent physiologic and pathologic processes (11), wefurther investigated whether NF-kB is also involved in theinhibitory effects of metformin/AMPK activation in HCCcells. Luciferase assay showed that NF-kB activities wererepressed by metformin treatment (Fig. 3A). RT-PCR alsoshowed that metformin treatment repressed NF-kB–regu-lated gene transcription (Fig. 3A). Interleukin (IL)-6, one ofthemost important NF-kB–dependent cytokines, is amajorSTAT3 activator. IL-6/STAT3 signaling pathway has alsobeen closely linked with HCC development (12). As shownin Supplementary Fig. S8A, ELISA assays revealed thatmetformin treatment decreased IL-6 synthesis and secre-tion. Western blot analysis also showed that addition ofmetformin decreased p-STAT3 level both in the rest stateand after IL-6 stimulation (Supplementary Fig. S8B).

Because NF-kB activation requires nuclear translocationof P65 subunit of NF-kB, we examined the subcellularlocalization of TNF-a–activated P65 in the presence ofmetformin using immunofluorescent confocalmicroscopy.After a 1-hour treatment with 100 ng/mL TNFa, there wasextensive nuclear staining for the P65 protein (Fig. 3B).However, pretreatment with 5 mmol/L metformin blockedTNFa-stimulated translocation of P65 from cytoplasm tonuclear. We also observed an inverse relationship betweenp-AMPK expression and p65 nuclear staining using immu-nohistochemistry (IHC) in 82 human HCC tissues (Sup-plementary Fig. S9).

The activity of NF-kB is tightly regulated by interactionwith inhibitory IkB proteins. To determine whether met-formin attenuated IkBa degradation, wemeasured the levelof total IkBa byWestern blot analysis. As shown in Fig. 3C,metformin treatment increased the expression of IkBa inboth HepG2 and SMMC7721 cells. Consistent with met-formin, activation of AMPK by AICAR had similar inhibi-tory effect on NF-kB activities (Supplementary Fig. S10).Conversely, knockdown of AMPKa1/2 expression not onlyresulted in increased NF-kB activities, but also inhibited theupregulation of IkBa expression by metformin treatment(Fig. 3C). These data further verified that metformin inhi-bits NF-kB activity via activation of AMPK.

Furthermore, to investigate whether metformin/AMPKactivation inhibited HCC cell growth through inhibitionof NF-kB signaling, SMMC7721 cells were transfectedwith P65 to activate NF-kB signaling (Supplementary

Table 1. Correlation between expression ofp-AMPK (Thr172) and clinicopathologicalcharacteristics in 273 HCCs

Highp-AMPK

Lowp-AMPK

(Thr172;n ¼ 76)

(Thr172;n ¼ 197) Pa

Age<50 43 109 0.892�50 33 88

GenderMale 67 165 0.451Female 9 32

HBV InfectionYes 69 184 0.446No 7 13

AFP (ng/mL)�200 24 109 0.000448<200 52 88

Liver cirrhosisYes 56 147 0.878No 20 50

Maximal tumor size (cm)<5 24 67 0.775�5 52 130

Tumor multiplicitySingle 66 153 0.0932Multiple 10 44

Tumor encapsulationIncomplete 35 131 0.00235Complete 41 66

Edmondson gradeI/II 23 28 0.00324III/IV 53 169

Pathologic TNM stageEarly stage (I-II) 56 95 0.00014Late stage (III) 20 102

Portal vein thrombosisAbsence 70 150 0.00207Gross 6 47

Distant metastasisAbsent 73 163 0.00281Present 3 34

Child-Pugh stageb

A 72 183 0.78677B 4 14

MELDc Score<10 71 188 0.99999�10 3 9

aFisher exact test.bNo patients with Child-Pugh C were included.cMELD (the model for end-stage liver disease). Significantresults (P < 0.05) are given in bold.

Zheng et al.





Fig. S11). As shown in Fig. 3D, forced expression ofP65 attenuated the effect of metformin. Because met-formin inhibited NF-kB signaling activity through upre-gulation of IkBa, we next investigated whether the inhib-itory effect of metformin can also be rendered on HCCcells overexpressing an undegradable mutant form of IkB[IkBa superrepressor (IkBaSR)]. As expected, transfectionof IkBaSR also attenuated the growth inhibiting effects ofmetformin (Fig. 3D). In addition, transfection of IkBaSRalso prevented metformin-induced upregulation of PTENexpression (Supplementary Fig. S12), suggesting thatmetformin increased PTEN expression through inhibitionof NF-kB activation. Together, these data suggested thatmetformin/AMPK activation inhibited IkBa degradationand thus decreased NF-kB and IL-6/STAT3 signalingactivity. The anticancer effects of metformin are mediated,at least in part, by inhibiting NF-kB and IL-6/STAT3signaling activity.

Therapeutic metformin/AMPK activation inhibits HCCcells growth in vivoTo evaluate the antitumor effects ofmetformin in vivo, we

generated HCC tumor xenografts by subcutaneous inocu-lation of SMMC7721 or HCC-LM3 cells in nudemice. Micewere daily treated with vehicle or metformin for 7 to 8weeks. During the treatment, metformin did not signifi-cantly affect body weight, average blood glucose level, andliver function of the mice (Supplementary Fig. S13). Asshown in Fig. 4A, metformin administration significantlyinhibited the growth of SMMC7721 or HCC-LM3 cell-derived tumors over the course of the experiment. More-over, metformin-treated tumors exhibited enhanced AMPKphosphorylation and reduced Stat3 phosphorylation,which was accompanied by a decrease of Ki67 staining andan increase of IkBa (Fig. 4B).

Therapeutic metformin/AMPK activation inhibitsprimary HCC cells growth

To explore the effect of therapeutic metformin/AMPKactivation on the proliferation of primaryHCC cells, freshlyisolated primary HCC cells were treated with metformin.As shown in Fig. 4C and D, metformin exerted inhibitoryeffects at a dose- and time-dependent manner and induc-ed apoptosis on the primary HCC cells. Western blotanalysis further confirmed thatmetformin enhanced AMPKphosphorylation, increased PTEN phosphorylation, andreduced STAT3 phosphorylation, which was accompaniedby an increase of IkBa expression level (SupplementaryFig. S14). These data further verified that therapeutic met-formin/AMPK activation elicits anticancer effects on prim-ary HCC cells.

DiscussionIn mammals, AMPK has been described as a sensor of

cellular and whole-body energy homeostasis (13). Thus,AMPK is a widely accepted pharmacologic target for thetreatment of metabolic syndrome and type II diabetes.During the last 10 years, there is a considerable amount ofevidence showing that AMPK is implicated in cancer cellgrowth and metabolism (14). By regulating a variety oftissue- and cell-specific downstream targets, AMPK controlsintracellular energy homeostasis to maintain the propergrowth rates (15). Likewise, under conditions of metabolicstress, AMPK activation can regulate various processes,including development, cell-cycle progression, apoptosis,and autophagy. Inactivation of AMPK has been implicatedin tumorigensis andmalignant behaviors in several cancers,including prostate, lung, and breast.

Despite the fact that hepatitis B virus (HBV) or hepatitis Cvirus infection is themajor risk factor forHCCdevelopment

Table 2. Univariate and multivariate Cox regression analysis of risk factors for recurrence

Univariate analysis Multivariable analysis

Variable HR (95% CI) P HR (95% CI) P

Age (�50 vs. <50) 1.036 (0.764–1.405) 0.819Sex (male vs. female) 0.836 (0.546–1.280) 0.41HBV Infection (absent vs. present) 0.968 (0.56–1.674) 0.907AFP (ng/mL) (<400 vs. �400) 1.815 (1.336–2.466) 0.00014 1.598 (1.163–2.196) 0.004Liver cirrhosis (absent vs. present) 1.062 (0.754–1.496) 0.731Tumor multiplicity (single vs. multiple) 1.866 (1.321–2.635) 0.00040 1.811 (1.258–2.608) 0.001Maximal tumor size (cm; <5 vs. �5) 1.337 (0.960–1.862) 0.086Tumor encapsulation (present vs. absent) 1.667 (1.209–2.299) 0.0018 1.322 (0.947–1.850) 0.101Edmondson grade (I/II vs. III/IV) 1.428 (0.985–2.070) 0.060Pathologic TNM stage (I–II vs. III) 1.807 (1.335–2.447) 0.00013 1.092 (0.772–1.545) 0.617Portal vein thrombosis (absence vs. gross) 1.864 (1.300–2.672) 0.00071 1.621 (1.104–2.381) 0.014p-AMPK level (low vs. high) 0.512 (0.352–0.745) 0.00047 0.609 (0.409–0.906) 0.014Child-Pugh stage (A vs. B) 1.190 (0.574–2.467) 0.648MELDa Score (<10 vs. �10) 1.354 (0.653–2.807) 0.434

aMELD (the model for end-stage liver disease). Significant results (P < 0.05) are given in bold.






Figure 2. Therapeutic metformin/AMPK activation inhibits HCC cellsgrowth and sensitizes HCC cellstoward chemotherapy. A,metformin inhibits human HCCcells growth. The results areexpressed as the percentage ofviable cells over cells in controlgroup. B, metformin decreasesclone formation ability of HCCcells. Treatments were conductedin triplicate; representative wellsare shown. (�,P <0.05; ��,P <0.01).C, therapeutic metformin/AMPKactivation augmented cisplatin-induced growth inhibition in HCCcells. Left, HepG2 and HCC-LM3cells were treated with cisplatin,metformin, or their combination for48 hours. Cell viability wasgenerated by CCK-8 assay. Theresults are expressed as mean �SD values from three independentexperiments (��, P < 0.01). Right,expression of PARP, p-AMPK, p-PTEN, and Abcg2 was determinedvia Western blot analysis aftertreatment with cisplatin,metformin, or their combination for48 hours. Representative resultsfrom three experiments wereshown.D, knockdownof AMPKa1/2 abolished themetformin-inducedincrease in PTEN expression.

Zheng et al.





worldwide, emerging evidence has also indicated that alco-hol consumption, diabetes, obesity, or other importantlifestyle factors contributed to HCC tumorigenesis (16). Soit is interesting to evaluate the expression and function ofthe fuel-sensing enzyme, AMPK, in HCC. In this study, weinvestigated the status of AMPK phosphorylation in HCCfor the first time. On the basis of human HCC clinicalspecimens, we showed that the p-AMPK (Thr172) wasdownregulated in majority of the patients with HCC, andlow p-AMPK staining correlated with poor prognosis ofHCC. Although our work was being completed, Lee andcolleagues also reported that AMPK-a2 is a tumor suppres-sor in HCC, and inactivation of AMPK-a2 promotes hepa-tocarcinogenesis by destabilizing p53 in a SIRT1-dependentmanner (17). Taken together, these findings imply thatAMPK may serve as a negative regulator in liver, and theloss of inhibitory effect of AMPK contributes to progressionand invasion of HCC.AMPK is also a promising target for cancer therapy. Many

recent studies have shown that exercise or pharmacologicactivators of AMPK, such as metformin, phenformin,

AICAR, A769662, cannabinoids, and aspirin, caused AMPKactivation and inhibited or delayed the onset of tumors indifferent animal cancer models (8, 18, 19). Populationstudy also showed that in type II diabetes mellitus patientswith HCC, metformin therapy is associated with a reducedHCC risk and seems to have a protective effect on HCCdevelopment (20, 21). In this study, we mainly selectedmetformin to modulate AMPK activity and showed theinhibitory effect of metformin on cultured HCC cells orprimary HCC cells growth. We further showed that metfor-min treatment resulted in cell-cycle block and inducedapoptosis. Interestingly, metformin also increased chemo-sensitivity of HCC cells, consistent with observation inclinical study and mouse xenografts models (22, 23).

Metformin exerts its effect through both AMPK-depen-dent and -independent mechanisms. Recent studies alsoshowed that downregulation of AMPK did not affectmetformin action on prostate cancer cell growth andmTOR inhibition, suggesting the existence of an alter-native AMPK-independent pathway (24). In this study,AMPKa1/2 knockdown greatly decreased, but not

Figure 3. Therapeutic metformin/AMPK activation inhibited NF-kB and IL-6/STAT3 signaling activity. A, metformin treatment repressed NF-kB reporter geneactivity and expression of NF-kB–regulated gene transcription. Left, SMMC7721 or HepG2 cells with or without metformin treatment were transfected withreporter gene. Right, metformin treatment repressed NF-kB–regulated gene transcription. Data represent the mean� SD of three independent experiments(�, P < 0.05; ��, P < 0.01). B, effect of metformin on TNFa-stimulated nuclear translocation of the P65. Cells were fixed, stained, and analyzed using confocalmicroscopy. Scale bar ¼ 50 mm. C, metformin inhibits NF-kB signaling activity via activation of AMPK. HepG2 cells were transfected with the AMPKa1/2siRNA or control siRNA, and then incubated with metformin. NF-kB signaling activity was assayed by luciferase assay (top) and IkBa expression wasdetermined by Western blot analysis (bottom). Representative results from three experiments were shown. �, P < 0.05; ��, P < 0.01. D, ectopic expression ofIkBa superrepressor (IkBaSR) or P65 rescues metformin mediated growth inhibition in SMMC7721 cells. The growth inhibition rate was detected by CCK-8.The results are expressed as mean � SD values from three independent experiments (��, P < 0.01).






completely reversed the inhibitory effect of metformin,further indicating that the effect of metformin on HCCcells is mainly dependent on activation of AMPK. How-ever, whether metformin exerts some of the antitumoraleffects on HCC cells through AMPK-independent mech-anism remains to be elucidated.

It is estimated that about 15% of human cancers, includ-ing colon cancer, gastric cancer, and HCC are associatedwith chronic infections and inflammation (25). Multiplesignaling pathways are involved in inflammation-mediatedtumorigenesis and also in human HCC development,among which NF-kB and STAT3 are likely the centralsignaling hubs. NF-kB and STAT3, each have a central rolein regulating the expression of a large number of down-stream genes that control cell proliferation, apoptosis, stressresponses, and immune functions. Most importantly, a rolefor NF-kB–regulated expression of the STAT3-activatingcytokine IL-6 has recently emerged both in viral hepatitisand in hepatosteatosis (26). Expression of IL-6 is elevated in

cirrhosis and HCC, and high-serum IL-6 level is also anindependent risk factor for progression from chronicviral hepatitis to HCC (27). Therefore, both the pathwaysthat control IL-6 expression and those that control its abilityto activate STAT3 offer interesting opportunities to thera-peutic intervention, as well as prevention. In fact, severalIKK/NF-kB or STAT3 inhibitors have been developed andprovided promising results in preclinical models (11, 28).We showed that activation of AMPK by metformin red-uced IkBa degradation, resulting in inhibition of NF-kBsignaling, decreased IL-6 expression, and STAT3 signal-ing. These findings are supported by the observation thatinhibitory effect of metformin on proliferation is sign-ificantly attenuated in cells transfected with P65 orIkBaSR, which activated NF-kB signaling or inhibitedIkBa degradation, respectively. Therefore, these preclin-ical results provide further evidence that metforminmight be considered as a valuable modulator for HCCprevention and therapy.

Figure4. Therapeuticmetformin/AMPKactivation inhibitsHCCcells growth in vivo andprimaryHCCcells in vitro. A andB,SMMC7721orHCC-LM3cellswereinjected subcutaneously in the right flank of male nude mice. A, tumor growth curve after administration of vehicle (squares) or metfomin (circles). Resultsrepresent the mean � SD of 8 mice in each group. Representative images of the dissected tumors after treatment are shown. �, P < 0.05; ��, P < 0.01.B, Hematoxylin and eosin (H&E) and immunostaining of p-AMPK, Ki67, and active caspase-3 expression in tumor tissues were shown (top). Western blotanalyses of representative tumors for each group are shown (bottom). C and D, freshly isolated HCC cells were treated with various concentrations ofmetformin for 24 or 48 hours. C, a cell viability assay (CCK-8) was conducted. The results are expressed as the percentage of viable cells over cells in controlgroup. D, metformin induces apoptosis in primary HCC cells. Primary HCC cells were exposed to metformin (5 mmol/L) for 48 hours and then labeled withAnnexin V. The results are expressed as mean � SD values from three independent experiments. �, P < 0.05; ��, P < 0.01.

Zheng et al.





In summary, we have revealed that activation of AMPKby metformin inhibited NF-kB and STAT3 signaling, andthus inhibitedHCCcells growth in vitro and in vivo. Althoughneeding further clinical trials to evaluate its safety andefficacy, our results indicated that therapeutic AMPK activa-tion should be an attractive target for HCC treatment.Considering that aberrant NF-kB and STAT3 signaling activ-ity is one of the most important characteristics of inflam-mation-mediated tumorigenesis, our study also unravelsmetformin’s potential in the treatment of human tumors.

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

Authors' ContributionsConception anddesign:H.Wang, L.-Y. Zheng,W.Yang, F.-Q.Wu,C.Wang,M-C. Wu, G.-S. Feng, D. ZouDevelopment of methodology: H. Wang, L.-Y. Zheng, W. Yang, L. TangAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): L.-Y. Zheng, W. Yang, F.-Q. Wu, C. Wang, L. Yu,L. Tang, B. Qiu, L. Guo

Analysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): L.-Y. Zheng, W. Yang, F.-Q. Wu, L. Tang,M-C. Wu, G.-S. Feng, D. ZouWriting, review, and/or revision of the manuscript: H. Wang, L.-Y.Zheng, W. Yang, M-C. Wu, G.-S. Feng, D. ZouAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): H. Wang, L. Yu, Y.-Q. LiStudy supervision: H. Wang, M-C. Wu, G.-S. Feng, D. Zou

AcknowledgmentsThe authors thank Dong-Ping Hu, Dan Cao, Shan-Hua Tang, Dan-Dan

Huang, and Shan-Na Huang for their technical assistances.

Grant SupportThis work was supported by National Natural Science Foundation of

China (81221061, 31201026, 30900770) and Chinese National Key Project(2012ZX10002-009, 2013ZX10002-011).

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

Received January 22, 2013; revised July 23, 2013; accepted August 2, 2013;published OnlineFirst August 13, 2013.

References1. Bosch FX, Ribes J, Diaz M, Cleries R. Primary liver cancer: worldwide

incidence and trends. Gastroenterology 2004;127:S5–S16.2. Villanueva A, Hernandez-Gea V, Llovet JM. Medical therapies for

hepatocellular carcinoma: a critical view of the evidence. Nat RevGastroenterol Hepatol 2013;10:34–42.

3. Hardie DG, Carling D, Carlson M. The AMP-activated/SNF1 proteinkinase subfamily: metabolic sensors of the eukaryotic cell? Annu RevBiochem 1998;67:821–55.

4. Jones RG, Thompson CB. Tumor suppressors and cell metabolism: arecipe for cancer growth. Genes Dev 2009;23:537–48.

5. Mihaylova MM, Shaw RJ. The AMPK signalling pathway coordinatescell growth, autophagy and metabolism. Nat Cell Biol 2011;13:1016–23.

6. ShackelfordDB,ShawRJ. TheLKB1-AMPKpathway:metabolismandgrowthcontrol in tumour suppression.NatRevCancer 2009;9:563–75.

7. Hardie DG. The AMP-activated protein kinase pathway–new playersupstream and downstream. J Cell Sci 2004;117:5479–87.

8. Motoshima H, Goldstein BJ, Igata M, Araki E. AMPK and cell prolif-eration–AMPK as a therapeutic target for atherosclerosis and cancer.J Physiol 2006;574:63–71.

9. Yang W, Wang C, Lin Y, Liu Q, Yu LX, Tang L, et al. OV6(þ) tumor-initiating cells contribute to tumor progression and invasion in humanhepatocellular carcinoma. J Hepatol 2012;57:613–20.

10. Lee TK, Castilho A, Cheung VC, Tang KH, Ma S, Ng IO. Lupeol targetsliver tumor-initiating cells through phosphatase and tensin homologmodulation. Hepatology 2011;53:160–70.

11. Baud V, Karin M. Is NF-kappaB a good target for cancer therapy?Hopes and pitfalls. Nat Rev Drug Discov 2009;8:33–40.

12. Toffanin S, Friedman SL, Llovet JM. Obesity, inflammatory signaling,and hepatocellular carcinoma-an enlarging link. Cancer Cell 2010;17:115–7.

13. Winder WW, Hardie DG. AMP-activated protein kinase, a metabolicmaster switch: possible roles in type 2 diabetes. Am J Physiol 1999;277:E1–10.

14. Luo Z, Saha AK, Xiang X, Ruderman NB. AMPK, the metabolicsyndrome and cancer. Trends Pharmacol Sci 2005;26:69–76.

15. Foretz M, Viollet B. Regulation of hepatic metabolism by AMPK.J Hepatol 2011;54:827–9.

16. Baffy G, Brunt EM, Caldwell SH. Hepatocellular carcinoma in non-alcoholic fatty liver disease: an emerging menace. J Hepatol 2012;56:1384–91.

17. Lee CW, Wong LL, Tse EY, Liu HF, Leong VY, Lee JM, et al. AMPKpromotes p53 acetylation via phosphorylation and inactivation ofSIRT1 in liver cancer cells. Cancer Res 2012;72:4394–404.

18. Vara D, Salazar M, Olea-Herrero N, Guzman M, Velasco G, Diaz-Laviada I. Anti-tumoral action of cannabinoids on hepatocellular car-cinoma: role of AMPK-dependent activation of autophagy. Cell DeathDiffer 2011;18:1099–111.

19. Hawley SA, Fullerton MD, Ross FA, Schertzer JD, Chevtzoff C, WalkerKJ, et al. The ancient drug salicylate directly activates AMP-activatedprotein kinase. Science 2012;336:918–22.

20. Donadon V, Balbi M, Mas MD, Casarin P, Zanette G. Metformin andreduced risk of hepatocellular carcinoma in diabetic patients withchronic liver disease. Liver Int 2010;30:750–8.

21. Chen HP, Shieh JJ, Chang CC, Chen TT, Lin JT, Wu MS, et al.Metformin decreases hepatocellular carcinoma risk in a dose-depen-dent manner: population-based and in vitro studies. Gut 2013;62:606–15.

22. Jiralerspong S, Palla SL, Giordano SH, Meric-Bernstam F, Liedtke C,Barnett CM, et al. Metformin and pathologic complete responses toneoadjuvant chemotherapy in diabetic patients with breast cancer.J Clin Oncol 2009;27:3297–302.

23. Rocha GZ, Dias MM, Ropelle ER, Osorio-Costa F, Rossato FA,Vercesi AE, et al. Metformin amplifies chemotherapy-inducedAMPK activation and antitumoral growth. Clin Cancer Res 2011;17:3993–4005.

24. Ben Sahra I, Laurent K, Loubat A, Giorgetti-Peraldi S, Colosetti P,Auberger P, et al. The antidiabetic drug metformin exerts an antitu-moral effect in vitro and in vivo through a decrease of cyclin D1 level.Oncogene 2008;27:3576–86.

25. Coussens LM, Werb Z. Inflammation and cancer. Nature 2002;420:860–7.

26. Park EJ, Lee JH, Yu GY, He G, Ali SR, Holzer RG, et al. Dietaryand genetic obesity promote liver inflammation and tumorigen-esis by enhancing IL-6 and TNF expression. Cell 2010;140:197–208.

27. Wong VW, Yu J, Cheng AS, Wong GL, Chan HY, Chu ES, et al. Highserum interleukin-6 level predicts future hepatocellular carcinomadevelopment in patients with chronic hepatitis B. Int J Cancer 2009;124:2766–70.

28. Groner B, Lucks P, Borghouts C. The function of Stat3 in tumor cellsand their microenvironment. Semin Cell Dev Biol 2008;19:341–50.






Published OnlineFirst August 13, 2013.Clin Cancer Res Longyi Zheng, Wen Yang, Fuquan Wu, et al. Effects of Metformin in Hepatocellular CarcinomaPrognostic Significance of AMPK Activation and Therapeutic

Updated version

10.1158/1078-0432.CCR-13-0203doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2013/08/14/1078-0432.CCR-13-0203.DC1Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/early/2013/09/11/1078-0432.CCR-13-0203To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-13-0203

http://clincancerres.aacrjournals.org/content/suppl/2013/08/14/1078-0432.CCR-13-0203.DC1

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/early/2013/09/11/1078-0432.CCR-13-0203


Documents

PrognosticSigniﬁcanceofAMPKActivationandTherapeutic ...clincancerres.aacrjournals.org/content/clincanres/early/...log-rank tests, and Cox regression analysis were used for survival