identiﬁcation of transmembrane protein 98 as a novel ...xiao bing liu1,2,3, chang chun ling1, yuen...

Cancer Biology and Signal Transduction

Identification of Transmembrane Protein 98 as a NovelChemoresistance-Conferring Gene in HepatocellularCarcinoma

Kevin Tak-Pan Ng1,2,3, Chung Mau Lo1,3, Dong Yong Guo1, Xiang Qi1,3, Chang Xian Li1, Wei Geng1,2,3,Xiao Bing Liu1,2,3, Chang Chun Ling1, Yuen Yuen Ma1,3, Wai Ho Yeung1,2,3, Yan Shao1,3,Ronnie Tung-Ping Poon1,2,3, Sheung Tat Fan1,2,3, and Kwan Man1,2,3

AbstractChemoresistance is one of the major obstacles in systemic chemotherapy and targeted therapy for

patients with advanced hepatocellular carcinoma. To identify novel chemoresistance-associated targets in

hepatocellular carcinoma, chemoresistant hepatocellular carcinoma cell lines were established. By com-

paring the global gene expression profiles between chemoresistant and chemosensitive cell lines, eight

novel chemoresistance-associated genes were identified to be significantly associated with the commonly

augmented chemoresistance of hepatocellular carcinoma cells. One upregulated candidate named trans-

membrane protein 98 (TMEM98) was found to be overexpressed in 80 of 118 (67.80%) of patients with

hepatocellular carcinoma. TMEM98 mRNA in tumor tissues was significantly higher than nontumor

tissues of patients with hepatocellular carcinoma (P < 0.0001). Upregulation of TMEM98 was significantlycorrelated with advanced tumor stage (P ¼ 0.048), high incidence of early tumor recurrence (P ¼ 0.005),poor overall survival (P ¼ 0.029), and poor disease-free survival (P ¼ 0.011) of patients with hepatocellularcarcinoma after hepatectomy. Importantly, upregulation of TMEM98 mRNA in patients with hepatocel-

lular carcinoma who received transarterial chemoembolization (TACE) treatment was significantly higher

than in patients without TACE treatment (P ¼ 0.046). Moreover, patients with poor response to TACEtreatment had higher degree of TMEM98 upregulation than the responsive patients. In vitro and in vivo

studies showed that suppression of TMEM98 in chemoresistant hepatocellular carcinoma cells restored

their chemosensitivity, while forced overexpression of TMEM98 enhanced their chemoresistance. The

mechanism of TMEM98 in conferring chemoresistance of hepatocellular carcinoma might be possibly

through activation of the AKT pathway and deactivation of p53. In conclusion, we identified a panel of

novel common chemoresistance-associated genes and demonstrated that TMEM98 is a chemoresistance-

conferring gene in hepatocellular carcinoma. Mol Cancer Ther; 13(5); 1285–97. �2014 AACR.

IntroductionLiver cancer is the sixth most common cancer and the

thirdmost common cause of cancer-related deathsworld-wide (1).Hepatocellular carcinoma represents 70% to 85%of the total liver cancer (1). Up to 70% of patients withhepatocellular carcinoma have been suffering from lim-ited treatment options because of late diagnosis and/or

advanced stage of the disease when, however, surgicaltreatments including liver transplantation and hepatecto-my as well as regional therapy are not feasible. Currently,there is no proven effective conventional systemic che-motherapy for patients with advanced hepatocellularcarcinoma because of the inherent chemoresistant natureof hepatocellular carcinoma and with intolerable cytotox-icity, resulting in thedismalprognosis of thesepatients (2–4).Hepatocellular carcinoma is a heterogeneousdisease interms of etiology, molecular, and carcinogenic mechan-isms as well as biologic behaviors, which can collectivelycontribute to diverse mechanisms of chemoresistanceamong patients with hepatocellular carcinoma (5). Recentsuccess of clinical trial of single-agent sorafenib in treatingadvanced patients with hepatocellular carcinoma hasshed lights for future development of targeted therapyfor patients with advanced hepatocellular carcinoma (6).However, most of the targeted agents have demonstratedavery lowresponse rate, including sorafenib (3, 7), leavingthe problem of chemoresistance to be solved. Therefore,

Authors' Affiliations: 1Department of Surgery; 2Center for CancerResearch; and 3State Key Laboratory for Liver Research, the Universityof Hong Kong, Pokfulam, Hong Kong SAR, China

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Corresponding Author: Kwan Man, Department of Surgery, Centre forCancer Research and State Key Laboratory for Liver Research, TheUniversity of Hong Kong, Room L9-55, Li Ka Shing Faculty of MedicineBuilding, 21 Sassoon Road, Pokfulam, Hong Kong, China. Phone: 86-852-28199646; Fax: 86-852-28199634; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-13-0806

�2014 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 1285

on July 5, 2021. © 2014 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst March 7, 2014; DOI: 10.1158/1535-7163.MCT-13-0806

http://mct.aacrjournals.org/

identification of novel molecular targets must be veryimportant for understanding the underlying mechanismsof chemoresistance of hepatocellular carcinoma and even-tually fordevelopingmore effective therapeutic regimens.

The development of chemoresistance in hepatocellularcarcinoma could be either intrinsic and/or acquired.There have been severalmolecular targets associatedwithchemoresistance of hepatocellular carcinoma. Chemore-sistance of hepatocellular carcinoma can be achieved byupregulation of the drug transporter family known as theadenosine triphosphate-binding cassette (ABC) transpor-ters such as ABCB1, ABCC1, ABCC2, and ABCC3 (8–11),leading to increment of drugs efflux system to removedrugs out of cells. High levels ofMRP2,MRP3,MRP4, andMRP5 genes have been found in an acquired cisplatin-resistant hepatocellular carcinoma cell line (12). More-over, recent findingshave suggested that liver cancer stemcells contribute to chemoresistance of hepatocellular car-cinoma. For instance, upregulation of Octamer 4 gene, atranscriptional factor of pluripotent cells, can significantlyaugment chemoresistance of hepatocellular carcinomacells via the Oct4–AKT–ABCG2 pathway (13). CD133þ

hepatocellular carcinoma cancer stem cells can also con-tribute to chemoresistance through activation of the Akt/PKB pathway (14). Overexpression of granulin epithelinprecursor andABCB5 in liver cancer stem cells can lead tosignificant increment of chemoresistance (15). Besides, astudy from acquired doxorubicin-resistant hepatocellu-lar carcinoma cell lines has identified a panel of differ-entially overexpressed genes and subsequently charac-terized that upregulation of TOP2A gene is one of thecontributors of acquired doxorubicin-resistance inhepatocellular carcinoma (16). Yet, the molecular infor-mation governing chemoresistance in hepatocellularcarcinoma so far is far away from achieving effectivetherapeutic regimens.

To search for novel targets,we started fromestablishingchemoresistant sublines from a human metastatic hepa-tocellular carcinomacell line and further identifiedapanelof common differential genes linking to the developmentof acquired chemoresistance of hepatocellular carcinomacells. Moreover, we investigated one of the upregulatedcandidates, transmembrane protein 98 (TMEM98), tounveil its clinical significance and roles in chemoresis-tance of hepatocellular carcinoma.

Materials and MethodsPatients

Onehundred and eighteen patientswith hepatocellularcarcinoma who underwent liver resection betweenDecember 1999 and May 2009 were recruited fromDepartment of Surgery, Queen Marry Hospital, the Uni-versity of Hong Kong. Twelve normal liver tissues wererecruited from living donors at the same hospital. The lastfollow-up date for the patients with hepatocellular carci-noma was on October 2010. Twenty-five clinical sampleswere obtained from patients with hepatocellular carcino-ma who received TACE treatment and they were subse-

quently subject for other treatments, including hepatec-tomy and liver transplantation. The technical details ofTACE protocol have been described in previous paper(17). The periods of TACE treatment for these patientsmay vary, from 2 to 105 months. The responsivenessafter TACE treatment were defined by our pathologistbased on the pathologic report from the last CT scanwhich were defined into 5 categories by our medicalspecialists: 1, no abnormality; 2, appearance of newtumor; 3, tumor same size; 4, decrease in tumor load;and 5, increase in tumor load. We defined the category 1and 4 as "responsive group," while categories 2, 3, and 5as "poorly responsive group." All clinical samples wereobtained from patients with signed consent. The studywas approved by the Ethics Committee of the Univer-sity of Hong Kong.

Cell linesHuman hepatocellular carcinoma cell lines, MHCC97L

(metastatic), PLC (nonmetastatic), and Hep3B (nonmeta-static), were cultured at Dulbecco’s modified Eagle medi-um (DMEM) medium with 10% FBS (Invitrogen). Toestablish chemoresistant cells, MHCC97L/CisR, andMHCC97L/DoxR, MHCC97L was chronically incubatedwith increased concentrations of cisplatin (PharmachemieBV) or doxorubicin (Sigma-Aldrich) for 12 months,starting from concentration of 100 ng/mL of cisplatinor 20 ng/mL of doxorubicin. On average, the concen-tration of cisplatin or doxorubicin was increased forevery 2 to 3 weeks. Proliferation rates of the chemore-sistant sublines were examined for each month. BeforecDNA microarray analysis, the chemoresistant sublineswere cultured in DMEM medium without drug for1 month. MHCC97L/CisR2 and MHCC97L/DoxR2were sublines under exposure of cisplatin and doxoru-bicin, respectively, for additional of 2 months.

Cloning and transfectionFull length of human TMEM98 cDNA was cloned into

pcDNA3.1(þ) vector (Invitrogen). The primers for cloningincluded forward primer: 50-GTACCAGGATCCAGCATGGAGACTGTGGTGATTGTT-30; reverse primer: 50-CTCGAGTCTAGA CTAAATTGCAGACTGCTCCTGCA-30. Transfection of plasmids to cells was performed byLipofectamine-2000 (Invitrogen). Stable transfectants wereselected fromG418-containingmedium for 2weeks. siRNAof targeting human TMEM98 mRNA and negative controlsiRNA were purchased from Invitrogen. siRNAs weretransfected to cells by using Lipofectamine RNAiMAX(Invitrogen) according to manufacturer’s instruction.

Proliferation assaysMTT (Invitrogen) assay and colony formation assay

were performed as previously described (18). Cells weretreatedwithdrugs for 72hours and2weeks forMTTassayand colony formation assay, respectively. Each experi-ment consisted of 3 replications and at least 3 individualexperiments were carried out.

Ng et al.

Mol Cancer Ther; 13(5) May 2014 Molecular Cancer Therapeutics1286




cDNA microarray analysiscDNA microarray profiles of MHCC97L, MHCC97L/

CisR, andMHCC97L/DoxRwere conducted by gene chipsystem Human U133 Plus 2.0 (Affymetrix Inc.), byGenome Research Centre, The University of Hong Kong(18). Microarray data (GEO accession number: GSE54175)were analyzed by GeneSpring Version 10 (Agilent Tech-nologies). A 3-fold difference was used to select differen-tial genes.

Quantitative reverse transcription PCRTotal RNA from cells and liver tissue samples were

purified by TRizol regent (Invitrogen). Method of quan-titative reverse transcription PCR (qRT-PCR) analysiswasdescribed as in previous study (19). The expression of 18Sribosomal RNA was used as internal control. Primersused in this study were listed in Supplementary Table S1.

Western blot analysisMethod of Western blot analysis was described in

previous study (18). TMEM98 antibody was purchasedfrom Sigma-Aldrich Corporation. Antibodies, includingAKT, phospho-AKT(Ser473), BCL-XL, phospho-GSK-3b(Ser9), p21, p53, phospho-p53(Ser6), phospho-p53(Ser9),phospho-p53(Ser15), phospho-p53(Ser20), and phospho-p53(Ser392), were purchased from Cell Signaling Tech-nology. b-Actin antibodywas purchased from Santa CruzBiotechnology.

Apoptosis assayCells (3 � 105) were seeded onto 6-well plate for 24

hours. The cells were harvested and stained withAnnexin-V Fluor Staining Kit (Roche) according to man-ufactory’s instruction and analysis by flow cytometer.Early apoptosis was defined as Annexin V–positive andpropidium iodide (PI)-negative cells. Late apoptosis wasdefined as Annexin V–positive and PI-positive cells. Thenumber of apoptotic cells included early and late apopto-sis. Each experimentwas analyzed in triplicate and at least3 independent experiments were performed.

Animal modelXenograft ectopic liver tumor model in nude mice was

adopted (18). For drug treatment, single dose of cisplatin(5mg/kg) or doxorubicin (5mg/kg)was administrated tonude mice at day 5 after subcutaneous injection of cells (5� 105 cells/100 mL). Tumor volume was calculated as thefollowing equation: tumor volume (cm3)¼ length�width� thickness. At least 6 mice were performed for eachexperimental group. Animal study was specificallyapproved byAnimal (Control of Experiments) OrdinanceChapter 340, the Department of Health, Hong Kong Spe-cial Administrative Region [Ref.: (12–63) in DH/HA&P/8/2/3 Pt. 37].

Statistical analysisTMEM98 mRNA of clinical samples was analyzed

by Prism Version 5.01 (Graphpad). The difference of

TMEM98 mRNA between tumor and nontumor tissuesof each patient with hepatocellular carcinoma wasdetermined as: DDDCt(TMEM98) ¼ DDCt(tumor) �DDCt(nontumor). Statistical analysis of clinical para-meters was carried out using SPSS 16 for Windows(SPSS Inc.). Receiver operating characteristic (ROC)curve was generated to analyze the sensitivity and 1-specificity of DDDCt(TMEM98) value to predict overallsurvival of patient with hepatocellular carcinoma afterhepatectomy. Youden index was used to determine highderegulation (High group) and low deregulation (Lowgroup) of patient with hepatocellular carcinoma. Theassociation of TMEM98 deregulation and clinicopatho-logic parameters was analyzed by a x2 test. The prog-nostic value of TMEM98 mRNA for predicting overalland disease-free survivals of patient with hepatocellularcarcinoma after hepatic resection was calculated byKaplan–Meier analysis with the log-rank test. For dis-ease-free survival analysis, patient with hepatocellularcarcinoma under the category of hospital mortality wereexcluded. Cox proportional hazard regression modelwas performed to test factors that were significantlyassociated with the overall survival or disease-freesurvival of the patient with hepatocellular carcinoma.P value < 0.05 was considered to be statisticallysignificant.

ResultsEstablishment of chemoresistant hepatocellularcarcinoma sublines

After continuous incubation of increasing concentra-tions of cisplatin or doxorubicin to a human metastatichepatocellular carcinoma cell line named MHCC97L for12 months, 2 chemoresistant sublines, MHCC97L/CisRand MHCC97L/DoxR, were established. The morpholo-gy of MHCC97L/CisR and MHCC97L/DoxR sublineswas similar to MHCC97L. In a nondrug culture medium,these sublines grew slightly faster than MHCC97L (Sup-plementary Fig. S1). The in vivo growth rate of thesesublines without drug was similar to MHCC97L (Supple-mentary Fig. S2).

The in vitro chemoresistance of MHCC97L/CisR andMHCC97L/DoxR sublines was significantly increasedcompared with MHCC97L. MTT assay showed that che-moresistance of MHCC97L/CisR to cisplatin (IC50 ¼20 mg/mL) was nearly 10-folds higher than MHCC97L(IC50¼ 2.2mg/mL).MHCC97L/DoxRexhibitedmore than25-fold increase of resistance to doxorubicin comparedwith MHCC97L (IC50 of doxorubicin: 40 mg/mL vs.1.5 mg/mL; Fig. 1A). The colony-forming ability ofMHCC97/CisR and MHCC97L/DoxR retained similarcolony-forming ability at 1,000 ng/mL of cisplatin(Fig. 1B) and 200 ng/mL of doxorubicin (Fig. 1C), respec-tively, whereas MHCC97L was dramatically suppressedin >500 ng/mL of cisplatin or >50 ng/mL of doxorubicin(Fig. 1B and C). The IC50 of MHCC97L/CisR2and MHCC97L/DoxR2 was approximately 1.3- and1.2-fold of the MHCC97L/CisR and MHCC97L/DoxR,

The Role of TMEM98 in Chemoresistance of Hepatocellular Carcinoma

www.aacrjournals.org Mol Cancer Ther; 13(5) May 2014 1287




MTT AssayA

B

C

D E

MHCC97L

Cisplatin (ng/mL)

Cisplatin concentration (ng/mL)

0

0 50 100 200 300 500

0 50

Time (day)0 5 10 15 20 25 30

Time (day)0 5 10 15 20 25 30

100 200 300 500

500 800 1,000 2,000 3,000

Cisplatin (ng/mL)

Colony formation assay

Colony formation assay

Doxorubicin (ng/mL)

Doxorubicin concentration (ng/mL)

In vivo xenograftmodel under cisplatin treatment

In vivo xenograftmodel under doxorubicin treatment

Doxorubicin (ng/mL)

0 5,000 10,000 15,000 20,000 20,0000 40,000 60,000 80,00025,000 30,000

MHCC97L/CisR

MHCC97LMHCC97L/CisR

MHCC97LMHCC97L/CisR

MH

CC

97L

MH

CC

97L/

Cis

R

0 500 800 1,000 2,000 3,000

Num

ber

of c

olon

ies

250

200

150

100

50

0

Num

ber

of c

olon

ies

250

200

150

100

50

0

MH

CC

97L

MH

CC

97L/

Cis

R

MHCC97L

MHCC97L/DoxR

MHCC97LMHCC97L/DoxR

MHCC97LMHCC97L/Dox

MH

CC

97L

MH

CC

97L/

Dox

R

MH

CC

97L

MH

CC

97L/

Dox

RP

erce

ntag

e of

via

ble

cells

(%

)1009080706050403020100

Tum

or v

olum

e (m

m3 )

1009080706050403020100 Tu

mor

vol

ume

(mm

3 )

1009080706050403020100

Per

cent

age

of v

iabl

e ce

lls (

%)

1009080706050403020100

MTT Assay

Ng et al.





respectively (data not shown). The in vivo chemoresistanceof MHCC97/CisR and MHCC97L/DoxR was increasedcompared with MHCC97L under cisplatin and doxorubi-cin treatment, respectively (Fig. 1D and E).

Identification of common chemoresistance-associated genesThe experimental procedures to identify common che-

moresistant candidates were illustrated in Supplementa-ry Fig. S3. To avoid overestimation of differential genesfrom cDNA microarray analysis, 3-fold difference wasused as filter condition rather than 2-fold difference.Comparing to cDNA microarray profile of MHCC97L,537 and 678 differential genes were identified inMHCC97L/CisR and MHCC97L/DoxR respectively, inwhich 235 genes were commonly differentially expressedin both sublines. Among them, 164 genes were commonlyupregulated (Supplementary Table S2) and 71 geneswerecommonly downregulated (Supplementary Table S3). Byapplying 10-fold difference as secondary selection crite-rion, 36 candidate genes were identified to be commonlydifferentially expressed. After validation by qRT-PCRanalysis, 19 common differential genes, 7 upregulatedand 12 downregulated, were identified (SupplementaryTable S4).To examine whether the degree of differential expres-

sion is correlated to the degree of chemoresistance ofhepatocellular carcinoma cells, the expression of 19 dif-ferential genes were further verified in MHCC97L,MHCC97L/CisR, MHCC97L/CisR2, MHCC97L/DoxR,and MHCC97L/DoxR2 cells. Among the upregulatedgenes, the expression levels of APOB, RUNDC3B,SYK, TMEM47, and TMEM98 genes were significantlyincreased along with the increased degrees of chemore-sistance of cells, reaching the highest level in bothMHCC97L/CisR2 and MHCC97L/DoxR2 comparedwith MHCC97L/CisR andMHCC97L/DoxR, respective-ly (Supplementary Fig. S4A). Among the downregulatedgenes, the expression levels of FGB, ZNF284, and BTB11were significantly lowered along with increased degreesof chemoresistance of cells, reaching the lowest level inboth MHCC97L/CisR2 and MHCC97L/DoxR2 com-pared with MHCC97L/CisR and MHCC97L/DoxR,respectively (Supplementary Fig. S4B). Totally, 8 che-moresistance-associated genes were identified to be sig-nificantly associated with common chemoresistance ofhepatocellular carcinoma cells.To identify hepatocellular carcinoma–associated

candidate genes, mRNA expression levels of APOB,RUNDC3B, SYK, TMEM47, and TMEM98 genes werepreliminarily investigated in 37 pairs of tumor andadjacent nontumor liver tissues of patients with hepa-tocellular carcinoma. One of them named TMEM98 was

found to be overexpressed in approximately 65% of thepatients with hepatocellular carcinoma (SupplementaryTable S5). Owing to chemoresistance- and hepatocellu-lar carcinoma–associated features of TMEM98 gene, itwas selected for further characterization.

Deregulation of TMEM98 predicts poor prognosis ofpatients with hepatocellular carcinoma

Among 118 patients with hepatocellular carcinoma, 80patients (67.80%)were found to differentially overexpressTMEM98 mRNA (DDDCt(TMEM98) � 1). The averagerelative DDCt(TMEM98) value among tumor liver tissues,nontumor liver tissues, and healthy donor liver tissueswere 7.00, 5.34, and 4.29, respectively (Fig. 2A). Theexpression level of TMEM98mRNA in tumor liver tissuewas significantly higher than in nontumor liver tissues(unpaired 2-tailed t test, P < 0.0001; paired 2-tailed t test,P < 0.0001) and healthy donor liver tissues (unpaired 2-tailed t test, P¼ 0.0002). Agreedwith the result from qRT-PCR analysis, Western blot analysis showed thatTMEM98 protein was overexpressed in tumor tissue ofpatients with hepatocellular carcinoma compared withnontumor tissues and healthy donor tissues (Fig. 2B).

Fifty-one (43.22%) and 67 (56.78%) patients with hepa-tocellular carcinoma were defined as TMEM98 highderegulation group (High group) and low deregulationgroup (Low group) respectively according to Youdenindex analysis. High deregulation of TMEM98 mRNAwas significantly correlated with the presence ofadvanced New AJCC stage (P ¼ 0.001) and recurrenceof tumor within first year (P ¼ 0.005; Table 1). Moreover,low deregulation of TMEM98 mRNA was significantlycorrelated with early pathologic tumor–node–metastasis(pTNM) stage (P ¼ 0.048).

Kaplan–Meier analysis illustrated that patients withhigh deregulation of TMEM98 mRNA were significantlyassociated with poor overall survival (log-rank ¼ 4.741,P ¼ 0.029) and poor disease-free survival (log-rank ¼6.543, P ¼ 0.011; Fig. 2C). The mean periods of overalland disease-free survival for the High group were 66.6and 38.8 months, whereas for the Low group were 85.8and 62.8 months.

Cox proportional hazard regression analysis was usedto find out the independent predictors for predictingoverall survival and disease-free survival of patients withhepatocellular carcinoma after hepatectomy among select-ed 5 factors, including TMEM98 mRNA, pTNM stage,venous infiltration, New AJCC stage, and serum a-feto-protein (AFP) level, which were significantly associatedwith the overall survival of patients with hepatocellularcarcinoma by Kaplan–Meier analysis (SupplementaryTable S6). Univariable Cox proportional hazard regressionanalysis showed that TMEM98 mRNA was a significant

Figure 1. Establishment of chemoresistant hepatocellular carcinoma sublines. A, MTT assay under cisplatin and doxorubicin treatments. Colony formationassay under cisplatin treatment (B) and doxorubicin treatment (C), D, xenograft ectopic nudemousemodel of MHCC97L andMHCC97L/CisR under cisplatintreatment. E, xenograft ectopic nude mouse model of MHCC97L and MHCC97L/DoxR under doxorubicin treatment. ��, P < 0.01.






A B

C

E

G

F

D

15

10

5

0

1.0

0.8

0.6

0.4

0.2

0.0

HCC-T

HC

C #

1

HC

C #

2

HC

C #

3

HC

C #

4

HC

C #

5

HC

C #

6

HC

C #

7

HCC-NT

Liver tissues

0.00 25.00

Overall survival (month) Disease-free survival (month)

Patients with TACE treatment

Poorly responsive Responsive

n = 18

n = 3

50.00 75.00 100.00 125.00 0.00 25.00 50.00

P = 0.029* P = 0.011*

Low, n = 67

Low, n = 65

High, n = 51

High, n = 51

75.00 100.00 125.00

Cum

sur

viva

l

Healthy donor

Hea

lthy

dono

rs

TMEM98

10

5

0

–5

–10

5

4

3

2

1

0

1 2 T NT NT NT NTT T T NTT NTT NTT

HCC HCC HCC HCC

TACE

With TACE

876543210

0 500 0 50 100 200 500

Rel

ativ

e ra

tio (

fold

) 6

5

4

3

2

1

0Rel

ativ

e ra

tio (

fold

)

1,000

Cisplatin concentration (ng/mL) Doxorubicin concentration (ng/mL)

2,000 3,000

Expression level of TMEM98 mRNA respondingto cisplatin treatment

Expression level of TMEM98 mRNA respondingto doxorubicin treatment

Without TACE

MHCC97L

PLC

Hep3B

MHCC97L

PLC

Hep3B

Non-TACE

NTT NT NT NTT T T

β-Actin

TMEM98

β-Actin

Rel

ativ

e ex

pres

sion

leve

lto

TM

EM

98 m

RN

A (

log 2

)

Rel

ativ

e T

/NT

rat

ioof

TM

EM

98 m

RN

A (

log 2

)

Rel

ativ

e T

/NT

rat

ioof

TM

EM

98 m

RN

A (

log 2

)

1.0

0.8

0.6

0.4

0.2

0.0

Cum

sur

viva

l

Figure 2. Clinical significance of TMEM98 in hepatocellular carcinoma. A, TMEM98 mRNA in liver tissues of patients with hepatocellular carcinoma andhealthy donors. B, expression of TMEM98 protein in 5 pairs of tumor and nontumor tissues of patients with hepatocellular carcinoma. C, Kaplan–Meieranalysis of overall and disease-free survival of patients with hepatocellular carcinoma. D, Western blot analysis of TMEM98 protein betweennon-TACE and TACE treatment groups. E, comparison of tumor to nontumor ratio (T/NT) of TMEM98 mRNA between patients with and withoutTACE treatment. F, comparison of tumor to nontumor ratio (T/NT) of TMEM98 mRNA between patients with hepatocellular carcinoma with poorlyresponsive and responsive effects in the TACE treatment group. G, TMEM98mRNA in different hepatocellular carcinoma cells in responding to cisplatinor doxorubicin treatment. �, P < 0.05; ��, P < 0.01.

Ng et al.





predictor for both of overall survival (HR ¼ 1.818; 95%confidence interval (CI), 1.05–3.14; P¼ 0.032) and disease-free survival (HR ¼ 1.816; 95% CI, 1.14–2.89; P ¼ 0.012)of patients with hepatocellular carcinoma after hepatecto-my (Supplementary Table S6). Multivariable Cox propor-tional hazard regression analysis revealed that TMEM98was not an independent factor for overall or disease-freesurvival of patients with hepatocellular carcinoma.To determine whether the deregulation of TMEM98

is correlated to increased chemoresistance of patientswith hepatocellular carcinoma, the expression levels ofTMEM98 mRNA and protein patients with hepatocellularcarcinoma received TACE treatment (TACE group) was

examined. The expression level of TMEM98 protein in theTACE groupwas found to be higher than in the non-TACEgroup (Fig. 2D). Eighty percent of patients with hepatocel-lular carcinoma were found to overexpress TMEM98mRNAin tumor tissues afterTACE treatment. The averagetumor to nontumor (T/NT) ratio of TMEM98 mRNA inthe TACE group was significantly higher than patientswithout TACE treatment (P ¼ 0.046; Fig. 2E). Moreover,in the TACE group, patients with poor response to TACEtreatment showed a higher degree of T/NT ratio ofTMEM98 mRNA [DDDCt(TMEM98) ¼ 2.92] comparedwith patients responded to TACE treatment [DDDCt(TMEM98) ¼ 1.42; Fig. 2F].

Table 1. Correlation analysis of TMEM98 mRNA and clinicopathologic features of patients withhepatocellular carcinoma

TMEM98 mRNA (n)

Clinicopathologic features Number (n) Low High P

SexMale 94 54 40 0.772Female 24 13 11

Age� 55 years 64 33 31 0.213> 55 years 54 34 20

pTNM stagea

Early stage (I–II) 34 24 10 0.048c

Advanced stage (III–IV) 83 42 41New AJCCa

Stage I to II 78 53 25 0.001c

Stage III to IV 39 13 26Venous infiltrationAbsent 51 34 17 0.059Present 67 33 34

Cirrhosisa

No 42 20 22 0.169Yes 74 45 29

Encapulationa

Absent 23 14 9 0.668Present 22 12 10

Tumor sizea

< 5 cm 33 20 13 0.598� 5 cm 69 38 31

AFP levela

� 1000 ng/ml 91 55 36 0.141> 1000 ng/ml 27 12 15

Hepatitis B surface antigena

Negative 17 10 7 0.854Positive 101 57 44

Recurrence within first yearb

No 68 46 22 0.005c

Yes 46 19 27

aTotal number less than 118 because of missing data.bFour patients were excluded because of their death within the first year without recurrence.c, P < 0.05.






MTT assay of MHCC97L/CisR subline MTT assay of MHCC97L/DoxR sublineA

B

C

D

E

F


Cisplatin concentration (μg/mL)


Cis

plat

intr

eatm

ent

Dox

orub

icin

trea

tmen

t

0 5,000 10,000 15,000 20,000 0 5,000 10,000 15,000 20,000

Time (day)0 5 10 15 20 25 30

Time (day)0 5 10 15 20 25 30

Cisplatin (μg/mL)

0 1 3 5

0 1 3 5

0 1 3 5

0 1 3 5

si-Nilsi-Negsi-TMEM98

Si-TMEM98

Si-Ctr Si-TMEM98Si-Ctr

Si-TMEM98Si-Ctr

si-TMEM98

si-Ctr

si-Nilsi-Negsi-TMEM98


Doxorubicin concentration (μg/mL)


In vivo xenograftmodel under cisplatin treatment

In vivo xenograftmodel under doxorubicin treatment

Doxorubicin (μg/mL)

0 5,000 10,000 15,000 20,000 25,000 30,000 0 5,000 10,000 15,000 20,000 25,000 30,000

MH

CC

97L-

3.1

MH

CC

97L-

TM

EM

98

MH

CC

97L-

3.1

MH

CC

97L-

TM

EM

98

MHCC97L-3.1MHCC97L-TMEM98

MHCC97L-3.1MHCC97L-TMEM98

16

14

12

10

8

6

4

2

0

30

25

20

15

10

5

0

Perc

enta

ge o

f via

ble

cells

(%)

Perc

enta

ge o

f via

ble

cells

(%)

Perc

enta

ge o

f apo

ptot

ic c

ells

(%)

Apoptosis assay on MHCC97L/CisR

MTT assay of MHCC97L-derivedtransfectants in cisplatin

pcDNA3.1TMEM98-1TMEM98-2

pcDNA3.1TMEM98-1TMEM98-2

MTT assay of MHCC97L-derivedtransfectants in doxorubicin

Apoptosis assay on MHCC97L/DoxR

Perc

enta

ge o

f apo

ptot

ic c

ells

(%)

1009080706050403020100

Perc

enta

ge o

f via

ble

cells

(%)

1009080706050403020100 Pe

rcen

tage

of v

iabl

e ce

lls (%

)

1009080706050403020100

1009080706050403020100

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

1009080706050403020100

1009080706050403020100

Ng et al.





TMEM98 is a novel chemoresistance-conferringtarget in hepatocellular carcinomaTreatment with low dosages of cisplatin or doxorubicin

in short period could upregulate TMEM98mRNA in bothMHCC97L and PLC cells in a dose-dependent manner,whileHep3Bgained upregulation ofTMEM98 in cisplatintreatment but not in doxorubicin treatment (Fig. 2G).Upregulation of TMEM98 mRNA after cisplatin or doxo-rubicin treatment could be persisted after removal of drug(Supplementary Fig. S5).TMEM98 siRNA could significantly suppress the

expression of TMEM98mRNA to 96 hours after transfec-tion in MHCC97L/CisR and MHCC97L/DoxR (Supple-mentary Fig. S6). Suppression of TMEM98 expressionresulted in significant decrement of the chemoresistanceofMHCC97L/CisR to cisplatin (si-Neg: IC50¼ 15 mg/mL;si-TMEM98: IC50 ¼ 5 mg/mL; Fig. 3A) and MHCC97L/DoxR to doxorubicin (si-Neg: IC50 > 30 mg/mL; si-TMEM98: IC50 ¼ 11 mg/mL; Fig. 3A). Suppression ofTMEM98 expression in MHCC97L/CisR or MHCC97L/DoxR cells by siRNA could significantly enhance theirapoptosis under cisplatin (Fig. 3B) or doxorubicin(Fig. 3C) treatment compared with control cells. More-over, overexpression of TMEM98 in MHCC97L cellscould significantly increase their chemoresistance to bothcisplatin (MHCC97L-pcDNA3.1: IC50 ¼ 4.0 mg/mL;MHCC97L-TMEM98-1: IC50 ¼ 5.5 mg/mL; Fig. 3D) anddoxorubicin (MHCC97L-pcDNA3.1: IC50 ¼ 4.2 mg/mL;MHCC97L-TMEM98-1: IC50 ¼ 10 mg/mL; Fig. 3D). Over-expression of TMEM98 in PLC cell line also increased itschemoresistance (Supplementary Fig. S7). Most important-ly, overexpression of TMEM98 in MHCC97L cell line sig-nificantly enhanced its in vivo resistance to both cisplatin(Fig. 3E) and doxorubicin (Fig. 3F). However, overexpres-sion of TMEM98 in MHCC97L cell line could not increasesorafenib chemoresistance (Supplementary Fig. S8).

TMEM98 modulates chemoresistance ofhepatocellular carcinoma through AKT and p53pathwaysTo understand the molecular mechanisms of TMEM98

on chemoresistance of hepatocellular carcinoma, 2 impor-tant signaling pathways, including the AKT and p53pathways, were investigated. Activation of the AKT[i.e., phospho-AKT(Ser473)] pathway was found inMHCC97L/CisR and MHCC97L/DoxR compared withMHCC97L (Fig. 4A). Suppression of TMEM98 by siRNAresulted in repression of activation of AKT (Fig. 4B).Moreover, suppression of TMEM98 in MHCC97L/CisRand MHCC97L/DoxR could repress EGF- and IGF2-induced activation of AKT and its downstream targets,

including BCL-XL and phospho-GSK-3b(Ser9) (Fig. 4C).Furthermore, suppression of TMEM98 inMHCC97L/CisRand MHCC97L/DoxR could inhibit the activation of theAKT pathway under drug environment in a dose-depen-dent manner (Fig. 4D). In addition, forced overexpressionof TMEM98 in MHCC97L led to elevation of AKT activa-tion under cisplatin or doxorubicin treatment (Fig. 4E).

Deactivation of p53, illustrated by deactivations ofphospho-p53(Ser15) and phospho-p53(Ser392), wasfound in MHCC97L/CisR and MHCC97L/DoxR, whileother forms of activated p53, including phospho-p53(Ser20), phospho-p53(Ser6), and phospho-p53(Ser9),could not be detected (Fig. 5A). Suppression of TMEM98in chemoresistant sublines restored the activation of phos-pho-p53(Ser15) and phospho-p53(Ser392) (Fig. 5B). Thedegree of activation of the p53 signaling pathway inMHCC97L/CisR and MHCC97L/DoxR was elevatedunder drug environment after suppression of TMEM98(Fig. 5C). Meanwhile, forced-overexpression of TMEM98inMHCC97L could repress p53 activation under cisplatinor doxorubicin treatment (Fig. 5D).

DiscussionSystemic chemotherapy has been adopted to treat

advanced patients with hepatocellular carcinoma formore than 30 years, but the survival outcome of thesepatients remained unsatisfactory (3, 4). The emergence oftargeted therapy, encouraged by improved survival ben-efits from sorafenib, will become a major strategy fortreatment of advanced patients with hepatocellular car-cinoma. However, inadequate knowledge on the molec-ular mechanisms of chemoresistance of hepatocellularcarcinoma from intrinsic and acquired pathways hindersthe development of effective targeted therapy on eradi-cating cancer cells (7, 20). Therefore, identificationofnoveltargets becomes an important task not only to understandmolecular mechanism of chemoresistance of hepatocellu-lar carcinoma but to eventually develop new effectivetherapeutic strategies for advanced patients with hepato-cellular carcinoma (21).

We applied a step-by-step approach to identify novelchemoresistance-associated genes. The in vitro and in vivogrowth rates of the chemoresistant sublines without drugtreatment were similar to parental MHCC97L cell line,indicating that the selection process did not alter theproliferation and tumorigenesis of the sublines. The invitro and in vivo chemoresistance of the chemoresistantsublines significantly higher than MHCC97L indicatedthat the acquired molecular changes of these sublinesunder drug selection, therefore, may be mainly prone todevelop their chemoresistance. There were 235 commonly

Figure 3. Functional role of TMEM98 in chemoresistance of hepatocellular carcinoma. A, MTT assay of MHCC97L/CisR and MHCC97L/DoxR sublines incisplatin or doxorubicin treatment after suppression of TMEM98 expression. B, apoptosis assay ofMHCC97L/CisR in cisplatin treatment after suppression ofTMEM98 expression. C, apoptosis assay of MHCC97L/DoxR in doxorubicin treatment after suppression of TMEM98 expression. D, MTT assay of TMEM98-overexpressing MHCC97L cells in cisplatin or doxorubicin treatment. Cisplatin treatment (E), or doxorubicin treatment (F) on in vivo ectopic model ofMHCC97L-3.1 and MHCC97L-TMEM98. �, P < 0.05; ��, P < 0.01.






differential genes, resting more than 300 and 400 geneswere distinctly differential in MHCC97L/CisR andMHCC97L/DoxR, respectively. These data indicated thatthe acquired chemoresistance of hepatocellular carcinomais contributed by acquired genetic or epigenetic changes ofa variety of genes, which were either or both drug-specificand nonspecific. Hepatocellular carcinoma can resist todifferent chemotherapeutic and targeting agents (2, 20).Identification of common differential genes is indispens-able for understanding the molecular mechanism of che-moresistance of hepatocellular carcinoma on differentdrugs.

The first question we would like to ask was whetherchange in expression of these genes was correlated to thechanges of chemoresistance in hepatocellular carcinoma.Among 19 highly differential genes, 5 genes were foundto have significantly increasing upregulation and 3 geneswere found to have significantly increasing downregula-tion in both MHCC97L/CisR2 and MHCC97L/DoxR2comparing to MHCC97L/CisR and MHCC97L/DoxR,respectively, indicating that the increasing differentialexpressions of these genes are significantly correlated tothe acquired chemoresistance in hepatocellular carcino-ma. So far, the roles of these genes in chemoresistance of

Figure 4. The roles of TMEM98 inthe AKT pathway. A, Western blotanalysis of TMEM98, AKT, andphospho-AKT(Ser473) protein inMHCC97L and chemoresistantsublines. B, Western blot analysisof TMEM98, AKT, and phospho-AKT(Ser473) upon suppression ofTMEM98 in chemoresistantsublines. C, suppression ofTMEM98 inhibits EGF- and IGF2-mediated AKT activation inMHCC97L/CisR and MHCC97L/DoxR sublines. D, suppression ofTMEM98 in chemoresistantsublines suppressedAKT signalingunder cisplatin or doxorubicintreatment. E, overexpression ofTMEM98 in MHCC97L cell lineenhanced AKT activation undercisplatin or doxorubicin treatment.�, si-Ctr; þ, si-TMEM98.

Ng et al.





hepatocellular carcinoma is not yet clear, suggestingthat they may be novel chemoresistance-associatedcandidates.Genes involved in tumorigenesis have been also found

to contribute to chemoresistance of hepatocellular carci-noma (13, 15, 22, 23). Identification of hepatocellularcarcinoma–associated targets is thus critical not only forunderstanding the molecular mechanisms of both intrin-sic and acquired chemoresistance but also for developinghepatocellular carcinoma–associated targeted therapy.From preliminary examination of clinical samples, oneof the differential genes, TMEM98, was found to be over-

expressed in most of the hepatocellular carcinoma tissue.Human TMEM98 protein, composed of 226 amino acidresidues, is a potential single-pass transmembrane pro-tein located in endoplasmic reticulum. TMEM98 has beenfound to be overexpressed in adenocarcinoma subtype ofadenosquamous carcinoma whose patients have relativepoor prognosis than other subtypes (24).TMEM98mRNAhas been included into adenocarcinoma-like expressionsignature, which is associated with an epithelial mesen-chymal transition and activated b-catenin pathway (24).So far the function of TMEM98 is unclear. In our study,TMEM98 mRNA was found to be overexpressed in early

Figure 5. The roles of TMEM98 inthe p53 pathway. A, Western blotanalysis of p53 and differentphosphorylated p53 proteins inMHCC97L and chemoresistantsublines. B, Western blot analysisof p53 and differentphosphorylated p53 proteins inchemoresistant sublines aftersuppression of TMEM98. C,suppression of TMEM98 enhancedp53 activation in chemoresistantsublines under drug treatments. D,overexpression of TMEM98 inMHCC97Lcell line suppressedp53activation under drug treatments.�, si-Ctr; þ, si-TMEM98.






70% of tumor tissues of patients with hepatocellularcarcinoma, indicating that this gene is hepatocellularcarcinoma–associated and its deregulation in hepatocel-lular carcinoma may also possibly contribute to intrinsicchemoresistance of hepatocellular carcinoma. Further-more, high deregulation of TMEM98 in hepatocellularcarcinoma was found to be significantly associated withadvanced New AJCC stage. Low deregulation ofTMEM98 in hepatocellular carcinoma was significantlyassociatedwith early pTNMstage andnearly significantlyassociated with absence of venous infiltration. Thesedemonstrated a positive correlation between TMEM98deregulation and progressive phenotype of hepatocellu-lar carcinoma. Moreover, high deregulation of TMEM98in hepatocellular carcinoma was found to be significantlycorrelated with higher incidence of early tumor recur-rence and poor overall and disease-free survivals ofpatients with hepatocellular carcinoma after hepatecto-my, suggesting that TMEM98 mRNA may be a potentialprognostic marker for patients with hepatocellular carci-noma. The diagnostic value of TMEM98 protein inpatients with hepatocellular carcinoma is also valuablefor further characterization.

TACE has been widely used to improve the survivalof patients with unresectable hepatocellular carcinoma(25). Although a recent study demonstrated improvedsurvival rates and low mortality rate using TACE withlipiodol in 8,510 patients with unresectable hepatocellu-lar carcinoma (26), the survival rate of advanced hepato-cellular carcinoma from TACE treatment thus far hasbeen unsatisfactory (25, 27) suggesting a pressing needto identify molecular mechanism linking to acquiredchemoresistance during TACE treatment. Our resultshowed that a significantly increasing upregulation ofTMEM98 mRNA in tumor was found in patients withhepatocellular carcinoma received TACE treatment com-pared with patients with hepatocellular carcinomawithout TACE treatment, indicating that hepatocellularcarcinoma tumor cells acquired higher changes ofTMEM98 expression than nontumor cells after chemo-therapy. In addition, the degree of deregulation ofTMEM98 in patients who were poorly responsive toTACE treatment was higher than that in TACE-respon-sive patients, implying that acquired overexpression ofTMEM98 in hepatocellular carcinoma may be linkingto the acquired chemoresistance of patients with hepato-cellular carcinoma after chemotherapy. The in vitrostudy indicated that TMEM98 is responsive to chemo-therapeutic agents.Altogether, the abovedata suggestedapossible association of acquired TMEM98 expressionin development of chemoresistance in hepatocellular car-cinoma. However, the sample size of patients receivedTACE treatment in this study was insufficient to reachsolid conclusion because of the lack of patients withhepatocellular carcinoma who were subjected for TACEas the first treatment option before surgical treatment.

In our functional study, suppression of TMEM98expression could trim down chemoresistance of chemore-

sistant sublines, while forced overexpression of TMEM98could increase chemoresistance of different hepatocellu-lar carcinoma cells. These data indicated that TMEM98may plays important roles in the development and main-tenance of chemoresistance in hepatocellular carcinomaand targeting suppression of TMEM98 in hepatocellularcarcinoma may be a potential strategy to overcome che-moresistance of hepatocellular carcinoma.

Several lines of evidence illustrated that activation ofAKT and deactivation of p53 play important roles onchemoresistance of cancers and hepatocellular carcinomaleading to important targets for treating advancedpatientswith hepatocellular carcinoma (28–32). In our study, thechemoresistant sublines exhibitedhigher level of activatedAKT and lower levels of activated forms of p53 thanparental MHCC97L, indicating that AKT and p53 may beinvolved in the development of chemoresistance.Ourdatademonstrated that TMEM98 could modulate chemoresis-tance of hepatocellular carcinoma cells throughmodifyingthe status of the AKT and p53 pathways.

In conclusion, our clinical and experimental evidencessuggested that TMEM98 is not only a prognostic markerfor patients with hepatocellular carcinoma but also anovel molecular target associated with intrinsic andacquired chemoresistance of hepatocellular carcinoma.Furthermore, TMEM98 may confer chemoresistance ofhepatocellular carcinoma by activation of the AKT sig-naling pathway and deactivation of p53. These findingsshould provide important information for developingeffective strategy in the future to overcome chemoresis-tance of hepatocellular carcinoma.

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

Authors' ContributionsConception and design: K.T.-P. Ng, C. Mau Lo, K. ManDevelopment of methodology: K.T.-P. Ng, K. ManAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):K.T.-P.Ng,C.MauLo,D.Y.Guo,C.X. Li, X.B. Liu,Y.Y. Ma, R.T.-P. PoonAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): K.T.-P. Ng, C.X. Li, X.B. Liu, Y.Y. Ma,K. ManWriting, review, and/or revision of the manuscript: K.T.-P. Ng, C. MauLo, S.T. Fan, K. ManAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): K.T.-P. Ng, X. Qi, W. Geng,C.C. Ling, W.H. Yeung, Y. Shao, K. ManStudy supervision: C. Mau Lo, R.T.-P. Poon, S.T. Fan, K. Man

Grant SupportThis study was supported by the Collaborative Research Funds

(HKU5/CRF/08 & HKU3/CRF/11R by C.M. Lo and K. Man) of theResearch Grant Council of Hong Kong, and the Small Project Funding(201109176183 by K.T.-P. Ng and K. Man) and the Seed Funding Pro-gramme for BasicResearch (201306159004 byK.T.-P.NgandK.Man) of theUniversity of Hong Kong.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received September 25, 2013; revised February 6, 2014; accepted Feb-ruary 17, 2014; published OnlineFirst March 7, 2014.


Ng et al.




References1. JemalA,BrayF,CenterMM,Ferlay J,WardE, FormanD.Global cancer

statistics. CA Cancer J Clin 2011;61:69–90.2. Zhu AX. Systemic therapy of advanced hepatocellular carcinoma: how

hopeful should we be? Oncologist 2006;11:790–800.3. Worns MA, Weinmann A, Schuchmann M, Galle PR. Systemic ther-

apies in hepatocellular carcinoma. Dig Dis 2009;27:175–88.4. Yau T, Chan P, Epstein R, Poon RT. Evolution of systemic therapy of

advanced hepatocellular carcinoma. World J Gastroenterol 2008;14:6437–41.

5. Wong CM, Ng IO. Molecular pathogenesis of hepatocellular carcino-ma. Liver Int 2008;28:160–74.

6. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al.Sorafenib in advanced hepatocellular carcinoma. N Engl J Med2008;359:378–90.

7. Yau T, Chan P, Epstein R, Poon RT. Management of advancedhepatocellular carcinoma in the era of targeted therapy. Liver Int2009;29:10–7.

8. Nies AT, Konig J, Pfannschmidt M, Klar E, Hofmann WJ, Keppler D.Expression of the multidrug resistance proteins MRP2 and MRP3 inhuman hepatocellular carcinoma. Int J Cancer Journal international ducancer 2001;94:492–9.

9. Silverman JA, Thorgeirsson SS. Regulation and function of the mul-tidrug resistance genes in liver. Prog Liver Dis 1995;13:101–23.

10. Chenivesse X, FrancoD, Brechot C.MDR1 (multidrug resistance) geneexpression in human primary liver cancer and cirrhosis. J Hepatol1993;18:168–72.

11. WarmannS,GohringG, TeichmannB, Geerlings H, Pietsch T, Fuchs J.P-glycoproteinmodulation improves in vitro chemosensitivity inmalig-nant pediatric liver tumors. Anticancer Res 2003;23:4607–11.

12. Wakamatsu T, Nakahashi Y, Hachimine D, Seki T, Okazaki K. Thecombination of glycyrrhizin and lamivudine can reverse the cisplatinresistance in hepatocellular carcinoma cells through inhibition ofmultidrug resistance-associated proteins. Int J Oncol 2007;31:1465–72.

13. Wang XQ, Ongkeko WM, Chen L, Yang ZF, Lu P, Chen KK, et al.Octamer 4 (Oct4) mediates chemotherapeutic drug resistance in livercancer cells through a potential Oct4-AKT-ATP-binding cassette G2pathway. Hepatology 2010;52:528–39.

14. Ma S, Lee TK, Zheng BJ, Chan KW, Guan XY. CD133þ HCC cancerstem cells confer chemoresistance by preferential expression of theAkt/PKB survival pathway. Oncogene 2008;27:1749–58.

15. Cheung ST, Cheung PF, Cheng CK, Wong NC, Fan ST. Granulin-epithelin precursor and ATP-dependent binding cassette (ABC)B5regulate liver cancer cell chemoresistance. Gastroenterology 2011;140:344–55.

16. Pang E, Hu Y, Chan KY, Lai PB, Squire JA, Macgregor PF, et al.Karyotypic imbalances and differential gene expressions in theacquired doxorubicin resistance of hepatocellular carcinoma cells.Lab Invest 2005;85:664–74.

17. Poon RT, Ngan H, Lo CM, Liu CL, Fan ST, Wong J. Transarterialchemoembolization for inoperable hepatocellular carcinomaandpost-resection intrahepatic recurrence. J Surg Oncol 2000;73:109–14.

18. NgKT, LeeTK,ChengQ,WoJY,SunCK,GuoDY, et al. Suppressionoftumorigenesis and metastasis of hepatocellular carcinoma by shRNAinterference targeting on homeoprotein Six1. Int J Cancer 2010;127:859–72.

19. Ng KT, Qi X, Kong KL, Cheung BY, Lo CM, Poon RT, et al. Over-expression of matrix metalloproteinase-12 (MMP-12) correlates withpoor prognosis of hepatocellular carcinoma. Eur J Cancer 2011;47:2299–305.

20. Wysocki PJ. Targeted therapy of hepatocellular cancer. Expert OpinInvest Drugs 2010;19:265–74.

21. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet2012;379:1245–55.

22. Yoo BK, Gredler R, Vozhilla N, Su ZZ, Chen D, Forcier T, et al.Identification of genes conferring resistance to 5-fluorouracil. ProcNatl Acad Sci U S A 2009;106:12938–43.

23. Geng W, Ng KT, Sun CK, Yau WL, Liu XB, Cheng Q, et al. The role ofproline rich tyrosine kinase 2 (pyk2) on cisplatin resistance in hepato-cellular carcinoma. PLoS One 2011;6:e27362.

24. Imadome K, IwakawaM, Nakawatari M, Fujita H, Kato S, Ohno T, et al.Subtypes of cervical adenosquamous carcinomas classified byEpCAM expression related to radiosensitivity. Cancer Biol Therapy2010;10:1019–26.

25. Llovet JM, Bruix J. Systematic review of randomized trials for unre-sectable hepatocellular carcinoma: chemoembolization improves sur-vival. Hepatology 2003;37:429–42.

26. Takayasu K, Arii S, Ikai I, OmataM, Okita K, Ichida T, et al. Prospectivecohort study of transarterial chemoembolization for unresectablehepatocellular carcinoma in 8510 patients. Gastroenterology 2006;131:461–9.

27. Llovet JM,RealMI,Montana X, PlanasR,Coll S, Aponte J, et al. Arterialembolisation or chemoembolisation versus symptomatic treatment inpatients with unresectable hepatocellular carcinoma: a randomisedcontrolled trial. Lancet 2002;359:1734–9.

28. Fabregat I, Roncero C, Fernandez M. Survival and apoptosis: adysregulated balance in liver cancer. Liver Int 2007;27:155–62.

29. Galluzzi L, Senovilla L, Vitale I, Michels J, Martins I, Kepp O, et al.Molecular mechanisms of cisplatin resistance. Oncogene 2012;31:1869–83.

30. Seitz SJ, Schleithoff ES, Koch A, Schuster A, Teufel A, Staib F, et al.Chemotherapy-induced apoptosis in hepatocellular carcinomainvolves the p53 family and is mediated via the extrinsic and theintrinsic pathway. Int J Cancer 2010;126:2049–66.

31. Guan YS, La Z, Yang L, He Q, Li P. p53 gene in treatment ofhepatic carcinoma: status quo.World J Gastroenterol 2007;13:985–92.

32. Fabregat I. Dysregulation of apoptosis in hepatocellular carcinomacells. World J Gastroenterol 2009;15:513–20.






2014;13:1285-1297. Published OnlineFirst March 7, 2014.Mol Cancer Ther Kevin Tak-Pan Ng, Chung Mau Lo, Dong Yong Guo, et al. Chemoresistance-Conferring Gene in Hepatocellular CarcinomaIdentification of Transmembrane Protein 98 as a Novel

Updated version

10.1158/1535-7163.MCT-13-0806doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2014/03/06/1535-7163.MCT-13-0806.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/13/5/1285.full#ref-list-1

This article cites 32 articles, 2 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/13/5/1285.full#related-urls

This article has been cited by 7 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://mct.aacrjournals.org/content/13/5/1285To request permission to re-use all or part of this article, use this link


http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-13-0806http://mct.aacrjournals.org/content/suppl/2014/03/06/1535-7163.MCT-13-0806.DC1http://mct.aacrjournals.org/content/13/5/1285.full#ref-list-1http://mct.aacrjournals.org/content/13/5/1285.full#related-urlshttp://mct.aacrjournals.org/cgi/alertsmailto:[email protected]://mct.aacrjournals.org/content/13/5/1285http://mct.aacrjournals.org/

identiﬁcation of transmembrane protein 98 as a novel ...xiao bing liu1,2,3, chang chun ling1, yuen...

Documents