Journal of Surgical Research 150, 219–226 (2008)

The Effect of Doxorubicin on MEK-ERK Signaling PredictsIts Efficacy in HCC1

Jennifer Choi, M.D.,* Michele Yip-Schneider, Ph.D.,*,2 Faith Albertin, B.S.,* Chad Wiesenauer, M.D.,*Yufang Wang, M.S.,* and C. Max Schmidt, M.D., Ph.D.*,†,‡,§,¶,2

*Department of Surgery, Indiana University School of Medicine, Indianapolis, Indiana; †Department of Biochemistry/Molecular Biology,Indiana University School of Medicine, Indianapolis, Indiana; ‡Walther Oncology Center, Indiana University School of Medicine,

Indianapolis, Indiana; §Indiana University Cancer Center, Indianapolis, Indiana; and ¶Richard L. Roudebush VA Medical Center,Indianapolis, Indiana

Submitted for publication September 5, 2007

doi:10.1016/j.jss.2008.01.029

Background. Hepatocellular cancer (HCC) is a lead-ing cause of cancer-related death worldwide. Histori-cally, doxorubicin (DOX) has been widely used againstunresectable HCC with variable response rates.

Materials and methods. We hypothesized that DOXcombined with mitogen-activated protein kinasekinase-extracellular signal-regulated kinase (MEK-ERK) targeted therapy may provide enhanced anti-cancer effects. Human HCC cell lines (HepG2, Hep3B)were treated with DOX and MEK enzyme inhibitors,U0126 or PD184161, alone or in combination. Growth,apoptosis, and ERK expression/MEK activity were re-spectively determined by proliferation assay, DNAfragmentation enzyme-linked immunoassay or fluoro-chrome inhibitor of caspases, and Western blot.

Results. DOX (0.01–1 �M) decreased cell prolifera-tion in Hep3B cells (IC50 � 0.12 �M) at 48 to 72 h; DOXwas less effective in HepG2 cells (IC50 � 0.25 �M). Atearly time points (30 min) after DOX treatment ofHep3B cells, MEK activity was unchanged at low dosesand decreased at higher doses; after 24 h, phospho-ERK levels increased at higher doses. Contrarily, inHepG2 cells, DOX caused a sustained, dose-dependentincrease in phospho-ERK levels at early and late timepoints. The MEK inhibitor U0126 decreased phospho-ERK in both HCC lines. In contrast to DOX, HepG2cells were more sensitive than Hep3B cells to U0126.The combination of DOX with U0126 (or PD184161)resulted in greater inhibition of proliferation in

1 Jennifer Choi and Michele Yip-Schneider contributed equally tothis work.

2 To whom correspondence and reprint requests should be ad-dressed at Department of Surgery, Indiana University School ofMedicine, 1044 W. Walnut St., Building R4, Rm. 041, Indianapolis,

IN 46202. E-mail: myipschn@iupui.edu.

219

HepG2 but not in Hep3B cells. This effect may be me-diated in part by enhanced apoptosis.

Conclusions. The effect of DOX on early and lateinduction of MEK activity predicts its chemotherapeu-tic response in HCC. Furthermore, this effect may alsodetermine the utility of MEK inhibitor combinationtreatment. © 2008 Elsevier Inc. All rights reserved.

Key Words: doxorubicin (DOX); MEK; MAPK/ERK;U0126; hepatocellular cancer.

INTRODUCTION

Hepatocellular cancer (HCC) remains a leadingcause of cancer-related death worldwide [1, 2]. Withinthe last decade, the incidence of HCC has been rising inthe United States in association with an increase inhepatitis C infection [3, 4]. At present, early detectionand resection provide the best chance for cure; how-ever, most patients are unresectable at presentationdue to the advanced stage of disease.

Doxorubicin (DOX), cisplatin, and gemcitabine arehistorically the most widely used chemotherapeuticsfor the treatment of HCC [5–9]; of these three, DOX(adriamycin hydrochloride, DOX) is the most commonsingle agent used. Reports of its efficacy are not con-sistent, with response rates varying up to 20% and amedian survival of only 4 months. Thus, for the major-ity of patients, no successful treatments are availableto date. Combining DOX with a targeted therapeuticapproach may be warranted to provide much neededtreatment options.

Multiple etiologies of HCC have been identified, in-cluding aflatoxin B, hepatitis B virus, hepatitis C virusand cirrhosis [10]. Regardless of the specific etiology,

most HCC is linked to a common cycle of hepatocyte

0022-4804/08 $34.00© 2008 Elsevier Inc. All rights reserved.

220 JOURNAL OF SURGICAL RESEARCH: VOL. 150, NO. 2, DECEMBER 2008

necrosis followed by rapid proliferation. This processcan ultimately lead to genetic or epigenetic changesthat stimulate growth/survival signaling pathways.The mitogen-activated protein (MAP) kinase kinase(MEK)-extracellular signal-regulated kinase (ERK)pathway is known to play a critical role in regulatingcell growth and may therefore be an important thera-peutic target [11]. In response to an extracellular sig-nal, a kinase cascade is initiated, activating MEK,which leads to the phosphorylation and activation of itsdownstream target, ERK (isoforms 1/2). ERK thentranslocates to the nucleus and initiates the transcrip-tion of specific genes that stimulate cell growth andsurvival. Pharmacological inhibitors of the MEK (i.e.,PD098059 and U0126) have been developed thatspecifically target this pathway both in vitro and invivo [12].

Additional support for targeting the MEK-ERKpathway include reports by our laboratory and othersthat the expression and activity of MEK-ERK pathwayintermediates are increased in human HCC tissues[13, 14]. Furthermore, we have recently demonstratedthat a novel MEK inhibitor PD184161 inhibited cellgrowth and induced apoptosis in a panel of HCC celllines in vitro [15]. In HCC tumor xenografts, PD184161was found to inhibit tumor growth initially but notlong-term possibly due to acquired resistance. Takentogether, these findings suggest that although MEKinhibitors may hold promise for the treatment of HCC,the efficacy of this class of inhibitors may be improvedby combining with other therapeutics drugs to sup-press survival and resistance pathways.

In the present study, we evaluated the combinationof DOX and the MEK inhibitors U0126 and PD184161in HCC. We report that DOX alone inhibits growth andtransiently decreases ERK phosphorylation in Hep3Bcells. In contrast, a sustained increase in ERK phos-phorylation occurs in HepG2 cells, which are moreresistant to growth inhibition by DOX. The MEKinhibitor/DOX combination is associated with de-creased ERK phosphorylation, enhanced growth inhi-bition and induction of apoptosis in HepG2 but notHep3B cells. Taken together, these findings suggestthat the effect of DOX on MEK activity correlates withtherapeutic response. Furthermore, DOX’s effect onthe MEK-ERK pathway may predict whether combina-tion treatment with a MEK inhibitor can enhance itsefficacy.

MATERIALS AND METHODS

Cell Culture

HepG2 and Hep3B cells were obtained from American Type Cul-ture Collection (Bethesda, MD) and maintained in modified Eaglesmedia-alpha containing 10% fetal bovine serum, 100 U/mL penicil-lin, and 100 mg/mL streptomycin. Cells were grown in a monolayer

in 5% CO2 at 37°C. For single agent studies, cells were plated and

drugs were added the following day for the indicated time periods.For combination drug studies, the cells were plated and then thenext day, DOX (Sigma, St. Louis, MO) was administered, followed bythe MEK inhibitors, U0126 (CalBiochem, San Diego, CA) orPD184161 (Pfizer, Ann Arbor, MI) 18 h later. Indicated treatmenttimes were from the time of DOX addition.

Proliferation Assays and Cell Counts

Cell growth was determined using a colorimetric assay, CellTiter96 AQueous One solution cell proliferation assay (Promega, Madison,WI), in which a tetrazolium compound is bioreduced by cells into acolored formazan product. The assay was performed according tomanufacturer’s protocol. Briefly, cells were plated in 96-well platesat a density of 2 � 104/cm2. After 24 h, DOX or vehicle control wasadministered followed by U0126 18 h later. Following treatment forthe indicated time period, cells were incubated with 20 �L of Cell-Titer 96 AQueous One solution reagent. Absorbance was read at 490nm. Proliferation assays were confirmed by trypan blue exclusioncell counts. In both methods, percent cell growth was determined bya ratio of average absorbance (or counts) of treatment wells to that ofthe control wells (100%).

Western Blot Analysis

Cells were lysed in radioimmune precipitation buffer (phosphatebuffer solution, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1%sodium dodecyl sulfate, 1 mM phenylmethylsulfonyl fluoride, 10�g/mL aprotinin, 1 mM sodium vanadate [Na3VO4]). Cell lysateswere centrifuged at 10,000 rpm for 10 min. Supernatants (10 �g totalprotein) were then resolved by sodium dodecyl sulfate polyacryl-amide gel electrophoresis on 4% to 20% gradient gels (Invitrogen,Carlsbad, CA). Separated proteins were transferred to Immobilon Pmembranes (Millipore, Bedford, MA) then incubated for 1 h in block-ing solution (Tris-buffered saline, 0.05% Tween, and 5% nonfat drymilk). Membranes were washed and incubated with primary anti-body according to manufacturer’s recommendations. After washingwith Tris-buffered saline Tween, they were incubated with the ap-propriate secondary antibody (horseradish peroxidase-conjugatedIgG) for 60 min at room temperature. Membranes were washed priorto detection using the enhanced chemiluminescence detection sys-tem (Amersham Pharmacia Biotech, Piscataway, NJ). The primaryantibodies used included a specific phospho-p42/44 MAP kinase(Thr202/Tyr204) antibody (Cell Signaling, Beverly, MA) and totalERK1/2 (K-23) antibody (Santa Cruz Biotech, Santa Cruz, CA). Den-sitometric analysis to determine relative expression was performedusing the Scion Image program (NIH).

Apoptosis

Apoptosis was measured using two techniques. First, enzyme-linked immunoassay (ELISA), the Cell Death Detection ELISA(Roche Diagnostics GmbH, Indianapolis, IN), allowed quantitativedetermination of the amount of cytoplasmic histone-associated DNAfragments induced by cells undergoing apoptosis. Cells were platedin 96-well plates and DOX and/or MEK inhibitors added sequentiallyfor the indicated times. Cell lysates were prepared, placed intostreptavidin-coated microtiter plates, and incubated with anti-histone biotin and anti-DNA peroxidase antibodies. After washingand addition of the substrate, the plates were read at 405 nm.Relative apoptosis was determined by a ratio of the average absor-bance of the treatment wells to that of the control wells (set equalto 1).

Apoptosis was also measured using a fluorochrome inhibitor ofcaspases (FLICA) apoptosis detection kit. The polycaspases FLICAkit FAM-VAD-FMK (Immunochemistry Technologies, Bloomington,MN) permits quantitative detection of caspases by using a caspaseinhibitor conjugated to a fluorescent marker, which is detectable by

flow cytometry. Propidium iodide (PI) staining allows for quantita-

221CHOI ET AL.: EFFICACY OF DOX FOR HCC THERAPY

tion of membrane integrity and cellular necrosis. Cells were plated in6-well plates and the next day, treated with DOX, U0126, or thecombination and grown for 48 h to 70%–80% confluence. Floatingand adherent cells were incubated with the FLICA reagent for 1 h at37°C. Cell suspensions were then incubated with propidium iodideprior to flow cytometric analysis.

Statistics

Statistical analyses, Student’s t-test, and one-way analysis ofvariance with Tukey post hoc, were performed (Prism 3.02 software,Graphpad, San Diego, CA). P � 0.05 was considered significant.

RESULTS

Effect of DOX on HCC Cell Growth

Two human HCC cell lines, HepG2 and Hep3B, weretreated with increasing concentrations of DOX for 48 or72 h. The effect on cell growth was determined byperforming a proliferation assay. At both time points,DOX decreased cell growth in a dose-dependent man-ner (Fig. 1A and B). DOX inhibited cell growth moreeffectively in Hep3B than in HepG2 cells; the IC50, orconcentration that inhibits growth by 50%, for DOX at72 h was approximately 2-fold higher in HepG2 than inHep3B cells (IC50 � 0.25 versus 0.12 �M, respectively).

Effect of DOX on MEK-ERK Activity

To determine the effect of DOX on the MEK-ERKsignaling pathway, cell lysates were prepared fromDOX-treated HepG2 and Hep3B cells. ERK activation,reflected by the level of phosphorylated ERK 1/2 iso-forms, was detected by Western blot analysis. DOXtreatment for 30 min dose-dependently increased thelevel of active, phosphorylated ERK in HepG2 cells; incontrast, the level of phospho-ERK was unchanged atlower doses of DOX and then decreased below basal athigher doses in the more sensitive Hep3B cells (Fig.2A). Total ERK protein levels were not affected. Treat-ment with DOX for 24 h also increased the level of

FIG. 1. Effect of DOX on HCC growth. HepG2 (dotted line) andCell growth was determined by proliferation assay and expressed as

�/– SEM is shown in a representative experiment performed in triplica

phospho-ERK in HepG2 cells (Fig. 2B). In Hep3B cells,phospho-ERK levels were not decreased at 24 h and, infact, increased at higher doses of DOX. Thus, there isan early, sustained increase in ERK phosphorylationin HepG2 cells but in Hep3B cells, an initial decreasefollowed by an increase in phospho-ERK levels. Differ-ential kinetics of ERK phosphorylation appear to cor-relate with differential DOX-mediated growth inhibi-tion.

Inhibition of the MEK-ERK Pathway by theMEK Inhibitor U0126

We have previously shown that U0126 dose-dependently inhibits the growth of HepG2 and Hep3Bcells [16]. HepG2 cells were more sensitive to U0126than Hep3B cells; the IC50 was approximately 10-foldhigher in Hep3B than HepG2 cells (IC50 � 5 versus 0.5�M, respectively). To confirm inhibition of MEK activ-ity by U0126, HepG2 and Hep3B cells were treatedwith increasing concentrations of U0126 for 30 min, 24or 48 h. Cells lysates were prepared and analyzed byWestern blot to measure the level of active, phosphor-ylated ERK. Phospho-ERK was decreased by very lowconcentrations of U0126 (0.1 �M) in HepG2 cells at allthree time points (Fig. 3A). In Hep3B cells, althoughphospho-ERK levels were similarly reduced by 0.1 �MU0126 at 30 min, at the longer time points higherconcentrations (�1 �M) were required to inhibit ERKphosphorylation (Fig. 3B). Thus, greater suppression ofMEK activity by U0126 was associated with greatergrowth inhibition in HepG2 cells and conversely forHep3B cells.

Growth Inhibition by DOX and/or MEK Inhibitors

To determine whether MEK inhibition would en-hance the response to DOX by suppressing the DOX-induced increase in ERK phosphorylation, HepG2 and

3B (solid line) cells were treated with DOX for (A) 48 h or (B) 72 h.rcent cell growth relative to control treated cells (100%). The mean

Heppe

te. *P � 0.05, HepG2 versus Hep3B.

222 JOURNAL OF SURGICAL RESEARCH: VOL. 150, NO. 2, DECEMBER 2008

Hep3B cells were treated with DOX and/or U0126 andgrowth was assessed. Initially, DOX and U0126 wereadded simultaneously, but no enhancement in growthinhibition was noted (unpublished observations). Sub-sequently, since DOX was observed to increase ERKphosphorylation by 24 h in both cell lines, DOX wasadded first followed by U0126 18 h later to prevent thesustained induction of phospho-ERK. The combinationof DOX (0–1 �M) and U0126 (1 or 5 �M) was a moreeffective growth inhibitor than either agent alone inHepG2 cells (Fig. 4A and B). Similar results were ob-tained when HepG2 cells were treated for 48 h withDOX and another MEK inhibitor PD181461 (Fig. 4C).

FIG. 2. Effect of DOX on ERK activation. HepG2 and Hep3B celanalyzed by Western blot to detect phosphorylated ERK and total Eanalysis was performed to determine relative phospho-ERK express

FIG. 3. Effect of U0126 on ERK phosphorylation. (A) HepG2 and

lysates were analyzed by Western blot to detect phosphorylated and to

In contrast, when Hep3B cells were treated with DOXand U0126, there was no significant difference in cellgrowth inhibition by DOX alone relative to the combi-nation (Fig. 5). These results suggest that HepG2 cellsare uniquely sensitive to growth inhibition by theDOX/MEK inhibitor combination.

Effect of DOX/U0126 on ERK Phosphorylation

HepG2 and Hep3B cells were treated with DOXand/or U0126 for 24 h. Cell lysates were prepared andanalyzed by Western blot to detect phosphorylatedERK. ERK phosphorylation was completely sup-

ere treated with DOX for (A) 30 min or (B) 24 h. Cell lysates werelevels as indicated. A representative blot is shown. Densitometric

.

Hep3B cells were treated with U0126 for 30 min, 24 h, or 48 h. Cell

ls wRKion

(B)

tal ERK levels. A representative blot is shown.

in triplicate are shown. *P � 0.05, combination versus single agents.

223CHOI ET AL.: EFFICACY OF DOX FOR HCC THERAPY

pressed to undetectable levels in U0126 as well ascombination-treated HepG2 cells; in contrast, phospho-ERK levels in Hep3B cells treated with the combina-tion were decreased but similar to that of U0126 alone,correlating with less efficacy of the combination inthese cells (Fig. 6).

Induction of Apoptosis by DOX/U0126

HepG2 and Hep3B cells were treated with DOXand/or U0126 for 48 h. Apoptosis was measured by aDNA fragmentation ELISA. Relative apoptosis in HepG2was induced by DOX alone and further increased by theaddition of U0126; although this difference was consis-tently observed, it did not reach significance (Fig. 7A).These results were confirmed by performing a caspase-based apoptosis assay (Fig. 7B). In contrast, apoptosis

FIG. 5. DOX and/or U0126-mediated growth effects in Hep3Bcells. Hep3B cells were treated with DOX (solid black bars) and/orU0126 (hatched bars) at (A) 1 �M or (B) 5 �M for 72 h. Cell growthwas determined by proliferation assay and expressed as percent cellgrowth relative to control treated cells (100%). Data are presented asmean �/– SEM. A representative experiment performed in triplicate

FIG. 4. Combination therapy with DOX/MEK inhibitor inHepG2 cells. HepG2 cells were treated with DOX (solid black bars)and U0126 (hatched bars) at (A) 1 �M or (B) 5 �M for 72 h. Cellgrowth was determined by proliferation assay and expressed aspercent cell growth relative to control treated cells (100%). C) DOXand/or PD181461 treatment of HepG2 cells for 48 h. Results areexpressed as mean �/– SEM. Representative experiments performed

is shown. *P � 0.05, combination versus single agents.

hos

224 JOURNAL OF SURGICAL RESEARCH: VOL. 150, NO. 2, DECEMBER 2008

induced by DOX alone was not enhanced by the additionof U0126 in Hep3B cells (Fig. 7C).

DISCUSSION

For the majority of HCC patients with unresectabletumors, systemic chemotherapy is the main treatment

FIG. 6. ERK phosphorylation levels in combination-treated cells.Cell lysates were prepared and analyzed by Western blot to detect p

FIG. 7. Apoptosis induced by combination therapy. (A) HepG2 cmeasured by a DNA fragmentation ELISA. Relative apoptosis waspresented as the mean �/– SEM from two independent experiments ptreated with DOX alone (0.5 �M) or in combination with U0126 (5measured by ELISA in Hep3B cells after 48 h treatment with DOX

experiment performed in duplicate is shown.

option available. Anthracyclines such as DOX havebeen the most effective single agents investigated inclinical trials to date; however, the response rate hasbeen variable and the median survival time remains at4 months. Combination chemotherapy with otheragents, including targeted agents, may result in higherresponse rates [17].

pG2 and Hep3B cells were treated with DOX and/or U0126 for 24 h.pho- and total ERK protein levels. Representative blots are shown.

were treated with DOX and/or U0126 for 48 h and apoptosis wastermined relative to control treated cells (set equal to 1). Data areormed in duplicate. (B) Apoptosis measured by FLICA in HepG2 cells

) for 48 h. Representative experiments are shown. (C) Apoptosis/or U0126. Data are presented as mean �/– SEM. A representative

He

ellsdeerf

�Mand

225CHOI ET AL.: EFFICACY OF DOX FOR HCC THERAPY

In the present study, we evaluated the combinationof DOX and a MEK inhibitor in vitro using two differ-ent human HCC cell lines. The two cell lines displayeddifferential sensitivity to DOX alone with Hep3B cellsbeing more sensitive than HepG2 cells. Interestingly,the kinetics of DOX-induced ERK activation also dif-fered in these two cell lines. Specifically, there was asustained increase in MEK-ERK activation in HepG2cells but in Hep3B cells there was a delayed increasein MEK activity. The differential growth response toDOX appears to correlate with differential activationof the MEK-ERK pathway.

We hypothesized that if sustained MEK-ERK activa-tion correlated with a poorer response to DOX as inHepG2 cells, then suppression of MEK activity shouldenhance the response to DOX. Treatment of HepG2cells with the DOX/U0126 combination effectively sup-pressed DOX-induced ERK phosphorylation and en-hanced growth inhibition. These effects may be medi-ated in part by enhanced apoptosis. In contrast,although combination-treated Hep3B cells showedsome decrease in phospho-ERK levels, U0126 additiondid not enhance growth inhibition by DOX nor did itincrease DOX-induced apoptosis. Taken together,these results suggest that the combination may bemore effective in HCC cells that demonstrate sustainedERK activation as well as sensitivity to MEK inhibi-tors (i.e., HepG2).

Other groups have previously shown sustained ERKactivation by DOX in various cell types [18–22]. DNAdamaging agents besides DOX, such as etoposide, ion-izing radiation, and ultraviolet irradiation also demon-strated similar effects on ERK activation [18, 22]. Therole of sustained ERK activation in causing cell cyclearrest and apoptosis was confirmed by addingPD98059 or U0126 to DOX and showing that the DOXeffects were attenuated by inhibiting MEK-ERK acti-vation in HepG2 and other cell types [18, 19, 22]. Incontrast, we observed that the U0126/DOX combina-tion was more effective at inhibiting cell growth andinducing apoptosis than the single agents in HepG2cells. In fact, others have also shown that ERK activa-tion by DOX can be antiapoptotic [20, 21]. Differenttiming of drug addition may account for the discrepan-cies in these reports. In our study, the drugs wereadded sequentially, DOX followed by U0126 after 18 h;in the other studies, the MEK inhibitor was eitheradded simultaneously or preincubated prior to DOXaddition. The timing of drug addition may be critical,given the complexity of MEK-ERK signaling and cross-talk with other cellular pathways. Subtle differences inkinetics may be magnified by downstream signals, re-sulting in substantially different outcomes.

The MEK-ERK signaling pathway clearly plays acritical role in HCC. We have shown in the present

study that DOX differentially influences MEK-ERK

signaling in HCC cells and this correlates with theefficacy of DOX as a single agent. Our findings alsosuggest that in HCC cells in which DOX results inearly and sustained ERK activation, combinationtreatment with a MEK inhibitor may be more effective.Efficacy of the combination may be further influencedby sensitivity to MEK inhibitors. Thus, effects on theMEK-ERK pathway may be an important determinantof response to DOX as well as potential efficacy of MEKinhibitor combination therapy.

ACKNOWLEDGMENT

The authors acknowledge funding for this work by Clarian HealthValues Fund.

REFERENCES1. El-Serag HB, Rudolph KL. Hepatocellular carcinoma: Epidemi-

ology and molecular carcinogenesis. Gastroenterology 2007;132:2557.

2. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2006. CACancer J Clin 2006;56:106.

3. Bruno S, Crosignani A, Maisonneuve P, et al. Hepatitis C virusgenotype 1b as a major risk factor associated with hepatocellu-lar carcinoma in patients with cirrhosis: A 17-year prospectivecohort study. Hepatology 2007;46:1350.

4. Davila JA, Weston A, Smalley W, et al. Utilization of Screeningfor hepatocellular carcinoma in the United States. J Clin Gas-troenterol 2007;41:777.

5. Chlebowski RT, Brzechwa-Adjukiewicz A, Cowden A, et al.Doxorubicin (75 mg/m2) for hepatocellular carcinoma: Clinicaland pharmacokinetic results. Cancer Treat Rep 1984;68:487.

6. Friedman MA. Primary hepatocellular cancer—present resultsand future prospects. Int J Radiat Oncol Biol Phys 1983;9:1841.

7. Lai CL, Wu PC, Chan GC, et al. Doxorubicin versus no antitu-mor therapy in inoperable hepatocellular carcinoma. A prospec-tive randomized trial. Cancer 1988;62:479.

8. Lin DY, Lin SM, Liaw YF. Nonsurgical treatment of hepatocel-lular carcinoma. J Gastroenterol Hepatol 1997;12:S319.

9. Yang TS, Lin YC, Chen JS, et al. Phase II study of gemcitabinein patients with advanced hepatocellular carcinoma. Cancer2000;89:750.

10. Blum HE. Hepatocellular carcinoma: Therapy and prevention.World J Gastroenterol 2005;11:7391.

11. Murphy LO, Blenis J. MAPK signal specificity: The right placeat the right time. Trends Biochem Sci 2006;31:268.

12. Roberts PJ, Der CJ. Targeting the Raf-MEK-ERK mitogen-activated protein kinase cascade for the treatment of cancer.Oncogene 2007;26:3291.

13. Ito Y, Sasaki Y, Horimoto M, et al. Activation of mitogen-activated protein kinases/extracellular signal-regulated ki-nases in human hepatocellular carcinoma. Hepatology 1998;27:951.

14. Schmidt CM, McKillop IH, Cahill PA, et al. Increased MAPKexpression and activity in primary human hepatocellular car-cinoma. Biochem Biophys Res Commun 1997;236:54.

15. Klein PJ, Schmidt CM, Wiesenauer CA, et al. The effects of anovel MEK inhibitor PD184161 on MEK-ERK signaling andgrowth in human liver cancer. Neoplasia 2006;8:1.

16. Wiesenauer CA, Yip-Schneider MT, Wang Y, et al. Multipleanticancer effects of blocking MEK-ERK signaling in hepato-

cellular carcinoma. J Am Coll Surg 2004;198:410.

226 JOURNAL OF SURGICAL RESEARCH: VOL. 150, NO. 2, DECEMBER 2008

17. Yeo W, Mok TS, Zee B, et al. A randomized phase III study ofdoxorubicin versus cisplatin/interferon alpha-2b/doxorubicin/fluorouracil (PIAF) combination chemotherapy for unresectablehepatocellular carcinoma. J Natl Cancer Inst 2005;97:1532.

18. Lee ER, Kim JY, Kang YJ, et al. Interplay between PI3K/Aktand MAPK signaling pathways in DNA-damaging drug-induced apoptosis. Biochim Biophys Acta 2006;1763:958.

19. Manov I, Bashenko Y, Eliaz-Wolkowicz A, et al. High-Doseacetaminophen inhibits the lethal effect of doxorubicin inHepG2 cells: The role of P-glycoprotein and mitogen-activated

protein kinase p44/42 pathway. J Pharmacol Exp Ther 2007;322:1013.

20. Niiya M, Niiya K, Shibakura M, et al. Involvement of ERK1/2and p38 MAP kinase in doxorubicin-induced uPA expression inhuman RC-K8 lymphoma and NCI-H69 small cell lung carci-noma cells. Oncology 2004;67:310.

21. Small GW, Somasundaram S, Moore DT, et al. Repression ofmitogen-activated protein kinase (MAPK) phosphatase-1 by an-thracyclines contributes to their antiapoptotic activation of p44/42-MAPK. J Pharmacol Exp Ther 2003;307:861.

22. Tang D, Wu D, Hirao A, et al. ERK activation mediates cellcycle arrest and apoptosis after DNA damage independently of

p53. J Biol Chem 2002;277:12710.