hur suppresses fas expression and correlates with patient ... · hur suppresses fas expression and...

Cell Death and Survival

HuR Suppresses Fas Expression and Correlateswith Patient Outcome in Liver CancerHaifeng Zhu1, Zuzana Berkova1, Rohit Mathur1, Lalit Sehgal1, Tamer Khashab1,Rong-Hua Tao1, Xue Ao1, Lei Feng2, Anita L. Sabichi3, Boris Blechacz4, Asif Rashid5, andFelipe Samaniego1

Abstract

Hepatocellular carcinomas (HCC) show resistance to che-motherapy and have blunt response to apoptotic stimuli. HCCcell lines express low levels of the Fas death receptor and areresistant to FasL stimulation, whereas immortalized hepato-cytes are sensitive. The variable Fas transcript levels andconsistently low Fas protein in HCC cells suggest posttranscrip-tional regulation of Fas expression. The 30-untranslated region(UTR) of Fas mRNA was found to interact with the ribonu-cleoprotein Human Antigen R (HuR) to block mRNA transla-tion. Silencing of HuR in HCC cells increased the levels of cellsurface Fas and sensitized HCC cells to FasL. Two AU-richdomains within the 30-UTR of Fas mRNA were identified asputative HuR-binding sites and were found to mediate thetranslational regulation in reporter assay. Hydrodynamic trans-fection of HuR plasmid into mice induced downregulation of

Fas expression in livers and established functional resistance tothe killing effects of Fas agonist. Human HCC tumor tissuesshowed significantly higher overall and cytoplasmic HuR stain-ing compared with normal liver tissues, and the high HuRstaining score correlated with worse survival of patients withearly-stage HCC. Combined, the protumorigenic ribonucleo-protein HuR blocks the translation of Fas mRNA and effectivelyprevents Fas-mediated apoptosis in HCC, suggesting that tar-geting HuR would sensitize cells to apoptotic stimuli andreverse tumorigenic properties.

Implications:Demonstrating how death receptor signaling path-ways are altered during progression of HCC will enable thedevelopment of better methods to restore this potent apoptosismechanism. Mol Cancer Res; 13(5); 809–18. �2015 AACR.

IntroductionHepatocellular carcinoma (HCC) is the ninth leading cause

of cancer-related deaths with a 5-year survival rate of 11%(1–3). The immune surveillance system can identify cancerouscells and then eliminate them through Fas death receptoractivation. However, tumor cells resist recognition by immunecells and apoptosis (4, 5). Healthy liver tissue expresses abun-dant levels of Fas that is typically silenced during HCC trans-formation, but HCC rarely acquires mutations that disable Fassignaling (4, 6).

HuR modulates posttranscriptional processing of target pre-mRNAs or mRNA stabilization and translation through interac-tion with AU-rich elements (ARE) within 30-untranslated regions

(UTR) of the target mRNAs (7, 8). Loss of HuR leads to cell deathand pathologic overexpression ofHuR is almost uniformly linkedto transformation (9).HuRhas beenpreviously shown tomediateskipping of Fas exon 6, leading to the production of a solubledecoy Fas (7, 10, 11). However, the relationship between HuRexpression andmodulation of Fas signaling in HCC has not beenpreviously described.

In the current study, we revealed that elevated levels of HuRinterfere with the translation of Fas mRNA, leading to decreasedFas expression and subsequent resistance to Fas-mediated apo-ptosis in HCC-derived cell lines and clinically aggressive HCC.

Materials and MethodsCell culture

Terminally differentiated human hepatic cells HepaRG(EMD Millipore) were cultured in Williams E Medium withculture medium supplement (EMD Millipore). Immortalizednormal hepatocytes (CRL4020) and HCC-derived cell lines(HepG2, Hep3B, SNU-182, SNU-398, SNU-449, and SNU-475;ATCC) were cultured in DMEM. Myeloma cell line AMBL6(a gift from Dr. Robert Orlowski; The University of Texas MDAnderson Cancer Center, Houston, TX) was maintained inRPMI-1640 medium. Both media were supplemented with10% FBS (Sigma-Aldrich). All cell lines were regularly testedfor mycoplasma (Lonza).

Transient knockdown of HuRTransient knockdown of HuR was achieved by electroporation

of specific HuR-targeting SMARTpool-designed siRNAs (Sir-HuR)

1Department of Lymphoma and Myeloma,The University of Texas MDAnderson Cancer Center, Houston, Texas. 2Department of Statistics,The University of Texas MDAnderson Cancer Center, Houston, Texas.3Baylor College of Medicine, Houston, Texas. 4Department of Gastro-enterology, Hepatology, and Nutrition, The University of Texas MDAnderson Cancer Center, Houston, Texas. 5Department of Pathology,The University of Texas MDAnderson Cancer Center, Houston, Texas.

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

Corresponding Author: Felipe Samaniego, Department of Lymphoma andMyeloma, The University of Texas MD Anderson Cancer Center, SCR2.2016,7455 Fannin Street, Houston, TX 77054. Phone: 713-563-1509; Fax: 713-563-7314; E-mail: [email protected]

doi: 10.1158/1541-7786.MCR-14-0241

�2015 American Association for Cancer Research.

MolecularCancerResearch

www.aacrjournals.org 809

on June 21, 2020. © 2015 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst February 12, 2015; DOI: 10.1158/1541-7786.MCR-14-0241

http://mcr.aacrjournals.org/

and the siCONTROL nontargeting siRNA (Sir-CON; both fromDharmacon/Thermo Fisher Scientific) into cells using the NeonTransfection System according to the manufacturer's instructions(Life Technologies; ref. 12).

Cell survival assayCell survival was evaluated by the MTT assay according to the

manufacturer's protocol (Promega; refs. 12, 13).

Cell death assayTreated cells were harvested, and incubated with propidium

iodide (BD Pharmingen) before flow cytometry to measure thenumber of dead cells.

Immunoblot analysisCells and mouse livers were analyzed as described previously

(12–14) using primary and secondary antibodies listed in theSupplementary Materials.

Confocal microscopyFixed cells were incubatedwith anti-Fas (Abcam) and anti-HuR

(EMD Millipore) primary antibodies followed by Alexa Fluor647-conjugated goat anti-mouse or Alexa Fluor 488-conjugatedgoat anti-rabbit secondary antibodies (both from Life Technolo-gies). The stained cells were mounted in ProLong Gold antifadereagent with DAPI; Life Technologies). Images of cells wereacquired and analyzed as described previously (12).

qRT-PCRTotal RNA was isolated using the TRIzol reagent (Life Tech-

nologies) according to the manufacturer's instructions. Comple-mentary DNA was synthesized using a reverse transcription-PCRarchive kit (Life Technologies) according to the manufacturer'sprotocol. All qRT-PCR tests were performed using the ABI Ste-pOnePlus Real-Time PCR-System (Applied Biosystems/Life Tech-nologies) according to the manufacturer's instructions. Primersequences and PCR amplification profile are listed in the Supple-mentary Materials.

CloningTwo evolutionarily conserved AREs in the Fas 30-UTR were

cloned into a 30-UTR-luciferase vector (OriGene) to make theconstructs Luc-Seq1 and Luc-Seq2. The full-length Fas 30-UTRluciferase vector was used as a template to amplify Seq1 and Seq2using the primer pairs introducing AsiSI and XhoI restrictionendonucleases sites (Listed in the SupplementaryMaterials). PCRproducts and vectors were digested with AsiSI and XhoI enzymesand ligated using T4 DNA ligase (all from New England Biolabs).

Luciferase reporter assayLipofectamine 2000 (Life Technologies) was used to transfect

HepG2 cells with plasmids: Luc-Fas 30-UTR, Luc-Seq1 orLuc-Seq2, Sir-HuR or Sir-CON, Luc-Null, and Renilla luciferasereporter pRL-TK vector (both from Promega) as a transfectionnormalization control. Twenty-four hours after transfection, cellswere harvested and analyzed using dual-luciferase reporter assaykit (Promega). Mean luciferase activity levels were derived fromthree independent experiments, following normalization to theRenilla luciferase activity.

Ribonucleoprotein immunoprecipitationThe ribonucleoprotein immunoprecipitation was performed

using the Magna RIP RNA-Binding Protein ImmunoprecipitationKit for HuR (EMD Millipore) following the manufacturer'sinstructions. qRT-PCR was conducted to determine the presenceand quantities of Fas mRNAs.

mRNA stability assayCells were treated with actinomycin D (Sigma-Aldrich) for 8 or

24hours. The total RNAwas isolated andqRT-PCRwas conductedas described above to determine the amount of Fas mRNA.

Analysis of newly synthesized proteinsCells were incubated in medium devoid of methionine (Pro-

mega) overnight before incubation with methionine mimic,Click-iT (L-azidohomoalanine) for nascent protein synthesis (LifeTechnologies), for 4 hours. Cells were then lysed for immunoblotanalysis.

Surface Fas and FasL-binding assayCells were collected and incubated with mouse IgG blocking

reagent (Life Technologies) before incubation with the phycoer-ythrin-conjugated anti-Fas antibody UB2 (Beckman Coulter) orthe corresponding isotype control IgG1-phycoerythrin (BDBiosciences).

To detect FasL binding, cells were incubated with FasL-FLAG(Enzo Life Sciences) followed by PhycoLink anti-FLAG-RPE anti-body (ProZyme). Phycolink anti-FLAG-RPE staining withoutFasL-FLAG was used as negative control. Surface Fas and boundFasL were measured by flow cytometry and analyzed as describedpreviously (12).

Animal experimentsAll animal experiments were performed in accordance with

the guidelines of MD Anderson Cancer Center's InstitutionalAnimal Care and Use Committee. Five- to 6-week-old C57BL/6mice (Harlan Laboratories) were hydrodynamically transfectedwith pcDNA3.1-HuR (a gift from Dr. Terry Dixon; MedicalUniversity of South Carolina, Charleston, SC ) or pcDNA3.1(14). Mice were challenged by an intraperitoneal injection of alethal dose of the agonistic anti-Fas antibody Jo2 (BD Bios-ciences) or buffer 24 hours after the transfection. Mouse sur-vival was monitored for up to 12 hours after the challenge.Livers were excised at 12 hours or at the time of death forfurther analysis.

IHCTissue microarray slides of human HCC and normal liver

tissues (IMH-360; IMGENEX/Novus Biologicals) were subjectedto deparaffinization, rehydration, and antigen retrieval beforeblocking using the BLOXALL kit (Vector Laboratories; ref. 15).Slides were then incubated with primary anti-HuR antibody(EMD Millipore) followed by biotinylated secondary antibody.The signal was enhanced with a VECTASTAIN Elite ABC Kit, anddeveloped with a 3, 30-diaminobenzidine (DAB) peroxidase kit(both from Vector Laboratories). Slides were counterstainedusing hematoxylin and mounted with Permount. HuR stainingintensity and a percentage of positive cells were scored indepen-dently by two investigators using the following grading system:staining intensity (0, undetectable; 1, low; 2, moderate; and 3,

Zhu et al.

Mol Cancer Res; 13(5) May 2015 Molecular Cancer Research810




high); percentage of positive cells (0, 0%–24%; 1, 25%–49%; 2,50%–74%; and 3, 75%–100%) and cytoplasmic staining(0, absent; and 1, present). A HuR staining score was obtainedfrom the sum of staining intensity and percentage of positive cellgrades (0, 0–1; 1, 2–3; 2, 4; and 3, 5–6).

Statistical analysisData are reported as the mean � SD of three samples or the

mean � SEM from three independent experiments unless indi-cated otherwise. Differences between groups were comparedusing the two-tailed Student t test. Summary statistics, includingmean, standard deviation, median and range for continuousvariable age, frequency counts, and percentages for categoricalvariables (such as stage and score variables), are reported. TheFisher's exact test was used to evaluate the association betweenscore variables and stage. The Wilcoxon rank-sum test or theKruskal–Wallis test was used to evaluate the difference in agebetween/among patient groups. Kaplan–Meier method was usedto estimate overall survival (OS).MedianOS time inmonths with95% confidence intervals (CI) was calculated. The log-rank testwas used to evaluate the difference in OS between patient groups.Statistical software SAS 9.1.3 (SAS) and S-Plus 8.0 (TIBCO Soft-ware Inc.) were used for all the analyses. A P-value < 0.05 wasconsidered statistically significant.

ResultsHCC cell lines are resistant to FasL

Neither FasL nor sFasL suppressed the viability of HCC celllines at the doses tested, whereas the highest doses significantlydecreased the survival rate of HepaRG and CRL4020 cells(Fig. 1A). To explore the mechanism underlying the resistanceof HCC cells to FasL, we evaluated the expression of Fas protein,Fas mRNA, and Fas signaling-related molecules (FADD, pro-caspase-8, and BID). Fas protein levels were low to undetect-

able in all six HCC cell lines, but robust in HepaRG andCRL4020 cells (Fig. 1B), whereas the levels of Fas mRNA variedgreatly among the HCC cell lines, but Fas mRNA was absent inHep3B (Fig. 1C). The expression levels of Fas signaling com-ponents were similar in all HCC cell lines to those found inCRL4020 cells with the exception of Hep3B cells that hadundetectable levels of FADD, procaspase-8, and BID (data notshown). These results suggested that reduced levels of Fas inHCC cell lines are associated with their resistance to Fas-mediated apoptosis.

Expression of HuR inversely correlates with expression ofFas in HCC cell lines

Regulation of Fas isoform expression was shown to originatefrom HuR-promoted skipping of exon 6, leading to expressionof soluble Fas instead of signaling-capable membrane-boundFas receptor (7, 10). To determine whether Fas levels correlatewith levels of HuR, we performed immunoblot analysis ofHCC cell lines and control CRL4020 cells. All HCC cell linesshowed higher levels of HuR compared with CRL4020 cells(Fig. 2A). Confocal microscopy confirmed very low to unde-tectable levels Fas in HCC cell lines while showing high HuRstaining that extended into cytoplasm (Fig. 2B). The observedcorrelation between elevated expression of HuR and lack ofFas expression suggested that HuR might regulate Fas expres-sion in HCC cell lines.

To assess HuR-promoted exon 6 skipping in our cell model, Faspre-mRNA splicing was evaluated in HCC cell lines and CRL4020cells by qRT-PCR using primers flanking exon 6 of Fas. Thesignificant amounts of a shorter band, corresponding to solubleFasmRNAs, were generated from the positive control AMBL6 cellsand Hep3B cells, whereas amounts of sFAS mRNA were compa-rable among the remaining HCC cell lines and control CRL4020cells (Supplementary Fig. S1A). These results suggested that

A

B C

100

50

00 100 200 0 50 100

Su

rviv

al (

% n

on

trea

ted

)

sFasL (ng/mL)

HepaRG

HepG2Hep3BSNU398SNU449SNU182SNU475

CRL4020

FasL (ng/mL)

1.5

1.0

0.5

0

Fas

mR

NA

rel

ativ

e to

Hep

aRG

**

**

****

HepaR

G

HepG2

Hep3B

SNU398

SNU449

SNU182

SNU475

CRL4020

HepaR

G

HepG2

Hep3B

SNU398

SNU449

SNU182

SNU475

CRL4020

FAS

b-Actin

Figure 1.Fas death receptor protein and mRNAlevels and responsiveness of HCC celllines to Fas activation. A, terminallydifferentiated human hepatic cells(HepaRG), human TERT immortalizedhepatocytes (CRL4020), and six HCCcell lines were incubated with FasL orsuperFasL for 24 hours and cellsurvival was determined using MTTassay. �� , P � 0.01. B, cell extractswere analyzed by immunoblotanalysis for basal expression of Fasrelative to b-actin. C, total mRNA wasisolated from the indicated cell linesand the content of Fas mRNA wasevaluated relative to GAPDH mRNAby RT-PCR. Data in A and C arepresented as the mean� SEM of threeindependent experiments.

HuR in Liver Cancer

www.aacrjournals.org Mol Cancer Res; 13(5) May 2015 811




alternative splicing of Fas pre-mRNA is not the primary mecha-nism of Fas downregulation in HCC cell lines.

HuR binds to Fas mRNAThe presence ofHuR in the cytoplasmofHCC cell lines (Fig. 2B)

suggested possible HuR involvement in regulation of Fas byexerting its effects on FasmRNA. To confirm that cytoplasmic HuRoperates normally in HCC cell lines, we evaluated a previouslyreported effect of HuR on FasL mRNA (16). As expected, HuRknockdown in HepG2 cells significantly lowered levels of FasLmRNA compared with control cells (Supplementary Fig. S1B).

We thus shifted our focus to 30-UTR of Fas mRNA that containssequences with the potential to bind numerous regulatory proteins(http://rbpdb.ccbr.utoronto.ca/). Computational analysis of FasmRNA using the UCSC Genome Browser (http://genome.ucsc.edu/) revealed two putative HuR-binding sites (Seq1 and Seq2)in its 30-UTR. A two-dimensional structure prediction algorithm(RNAfold, http://rna.tbi.univie.ac.at/cgi-bin/RNAfold.cgi) furthersupported a high probability of HuR binding to these two con-served sites (Fig. 2C). Using a ribonucleoprotein immunoprecip-itation followed by qRT-PCR, we probed for a possible associationbetween HuR protein and Fas mRNA in HepG2 cells. Fas mRNAwas amplified from HuR-specific but not control IgG precipitates(Fig. 2D, left) that originated from samples with equivalent con-centrations of input RNA containing comparable Fas mRNA levels(Fig. 2D, right). This newly identified association strongly impli-cated HuR in regulating Fas expression.

HuR regulates Fas mRNA translation in HCC cellsHuR typically stabilizes the transcripts bearing AREs in

their 30-UTRs or alters rates of translation of its mRNA targets(8, 9, 17). To reveal how HuR affects the Fas expression at themRNA level, we selected readily transfectable HepG2 cells for

further studies. We first knocked down HuR in HepG2 cellsusing HuR-specific si-RNA (Sir-HuR) or control si-RNA (Sir-CON; Fig. 3A). qRT-PCR showed no significant difference inthe levels of Fas mRNA between HuR knockdown and controlHepG2 cells (Fig. 3B). To check the stability of Fas mRNA, weblocked transcription in HuR knockdown and control HepG2cells and analyzed the levels of Fas mRNA over 24 hours.HuR knockdown and control cells showed no significantdifferences in the rate of decay of Fas mRNA (Fig. 3C). Wenext examined the effect of HuR on Fas translation by usinglabeled amino acid incorporation. As shown in Fig. 3D,robust levels of Fas protein were produced in HuR knock-down HepG2 cells, whereas little Fas was produced in controlcells (Fig. 3D).

We next cloned the two putative HuR-binding sites (Fig. 2C)downstream of the luciferase reporter gene. HepG2 cells weretransfected with Sir-HuR or Sir-CON combined with plasmidsencoding luciferase reporter with the entire Fas mRNA 30-UTR,Seq1 or Seq2, illustrated in Fig. 3E. The reporter expression wascompletely blocked with Fas 30-UTR, but only partially with Seq1and Seq2 alone (Fig. 3E). HuR knockdown partially restoredexpression of the reporter with the entire Fas 30-UTR, and fullyrestored expressionof reporterwith Seq1 andSeq2 (Fig. 3E). Thus,Seq1 and Seq2 blocked reporter expression in the presence ofHuR. Taken together, these results suggested that HuR negativelyaffects translation of Fas rather than the stability of FasmRNA andthat this effect is at least partially mediated through HuR inter-action with Seq1 and Seq2 in the 30-UTR of Fas mRNA.

Knockdown of HuR increases Fas expression, binding of FasL,and FasL-induced apoptosis

To evaluate a more generalized effect of HuR on Fas proteinexpression, HuR was knocked down in three HCC lines

A B

CD

DAPI

FAS

HuR

Anti-HuRAnti-Fas

CR

L40

20H

epG

2S

NU

398

SN

U44

9

CRL4020

HepG2

Hep3B

SNU398

SNU449

SNU182

SNU475

b-Actin

Seq1 Seq2

60

40

20

0

1.5

1

0.5

0F

as m

RN

A (

% G

AP

DH

)

Fas

mR

NA

fo

ld e

nri

chm

ent

Seq1 Seq2 IgG Anti-HuR IgG Anti-HuR

Figure 2.Expression of HuR in HCC celllines and presence of putativeHuR-binding elements withinthe 30-UTR of Fas mRNA. A,immunoblot analysis of Fas andHuR expression in HCC cell lines.Panels showingFas andb-actin arethe same as shown in Fig. 1B. B,HepG2 cells stained byimmunofluorescent anti-Fas(green) and anti-HuR (red)antibodies, and nuclear DAPI stain(blue) were observed by confocalmicroscopy. C, location andstructure of two possible HuR-binding sites within the Fas mRNA30-UTR. Dashed rectangles showlocation of conserved putativeHuR-binding sites Seq1 and Seq2.D, analysis of the HuR bindingto Fas mRNA in HepG2cells by ribonucleoproteinimmunoprecipitation followed byqRT-PCR (left); analysis of inputRNA (right).

Zhu et al.





(HepG2, SNU449, and SNU398) showing a wide range of FasmRNA levels (Fig. 1C). Silencing of HuR dramatically increasedtotal Fas protein expression in all three cell lines (Fig. 4A)suggesting a generalized regulation of Fas expression by HuR inHCC cell lines.

Increased total cellular Fas protein expression in HuR knock-down HepG2 cells was accompanied by elevated surface Faslevels and subsequently also increased FasL binding, bothdetected by flow cytometry (Fig. 4B and C). Finally, sFasLinduced 70.5 � 1.7% cell death in HuR knockdown HepG2cells, but only 22.8 � 0.7% cell death in control HepG2 cells(Fig. 4D and Supplementary Fig. S2A). Immunoblot analysisof cells stimulated by sFasL showed greater cleavage/activationof caspases-8 and -3 and lower BID levels in HuR knockdownthan in control HepG2 cells (Supplementary Fig. S2B), con-firming that HuR-dependent regulation of Fas expressionrestores early Fas signaling and leads to an improved cell deathresponse.

Overexpression of HuR in livers lowers Fas levels and protectsthe mice from a lethal challenge with Fas agonistic antibody

To assess the in vivo effects of HuR on Fas-mediated liver cellkilling, we overexpressed HuR in mouse livers by hydrody-namic transfection and 24 hours later treated the mice with alethal dose of Fas agonistic antibody Jo2 (12). HuR-transfectedmice demonstrated significantly improved survival than did

control vector-transfected mice (10 mice per group; P ¼0.02; Fig. 5A). Gross examination of livers exposed to Jo2revealed blackened liver tissue typical of extensive hemorrhag-ing in vector-transfected livers, but only limited blackening inHuR-transfected livers (Fig. 5B). Immunoblot analysis of livertissues confirmed the overexpression of HuR, that was associ-ated with lower levels of Fas and the apparently reducedprocessing of Fas signaling molecules (procaspases-3, -8, andBID, and PARP) in HuR-compared with control vector-trans-fected mice after a challenge with Jo2 agonistic antibody(Fig. 5C). These in vivo data confirmed the cell culture dynam-ics showing that HuR blocks Fas-mediated apoptosis by sup-pressing Fas protein expression and consequently decreasingthe amplitude of Fas signaling that translates to lower rates ofapoptosis.

HuR is overexpressed in HCC tissuesConsistent HuR overexpression in HCC cell lines compelled us

to examine whether humanHCC tumors also show elevated HuRexpression. A liver cancer tissuemicroarray (TMA of 59 liver tissuecores from 44 patients; available characteristics are summarizedin Table 1, and were stained with anti-HuR antibody (Fig. 6A).Staining intensity, percentage of positive cells, and cytoplasmicstaining were evaluated independently by two researchers. HuRstaining score was obtained by combining scores for the intensityand percentage of positive cells.

Fas

β-Actin

Sir

-CO

N

Sir

-Hu

R

A B C

D E

1

0.5

0

3

2

1

0

Fas

mR

NA

(%

GA

PD

H) 1.5

1

0.5

0

Fas

mR

NA

(%

GA

PD

H)

1.5

1

0.5

0R

elat

ive

luci

fear

ase

acti

vity

Sir-CON Sir-HuR Sir-CON Sir-HuR

Sir-CON Sir-HuR

Fas 3¢UTR

Fas 3¢UTRLuc-Null Seq1 Seq2

0 8 24Time (h)

Sir-CON

Sir-HuRH

uR

mR

NA

(%

GA

PD

H)

Figure 3.The effect of HuR knockdown on the FasmRNA levels, stability, and translation in HepG2 cells. A, levels of HuRmRNA in HepG2 cells transfected with HuR-targetingsi-RNA (Sir-HuR) or control si-RNA (Sir-CON) were evaluated by qRT-PCR. B, the levels of Fas mRNA in Sir-HuR and Sir-CON transfected cells were analyzedby qRT-PCR. C, transcription was inhibited by incubation of cells transfected with Sir-HuR and Sir-CON with actinomycin D. Cells were harvested at indicatedtimes post transfection and analyzed for expression of FasmRNAbyqRT-PCR. D, click-it protein synthesis assaywas used tometabolically label Sir-HuR andSir-CONtransfected cells to evaluate translation of Fas mRNA by immunoblot analysis. E, schematic representation of luciferase reporter constructs with Fas 30-UTRand identified HuR-binding sequences (Seq1 and Seq2), which were used to evaluate their effects on reporter translation in Sir-HuR or Sir-CON–transfected cells byusing dual-luciferase reporter assay. �� represent significant differences (P < 0.001) between adjacent Sir-CON and Sir-HuR bars. Data in A–C are presented as themean � SEM of three independent experiments. Data in E represent mean � SD of six separate wells. Experiment was repeated twice and the same trendwas observed.

HuR in Liver Cancer





Our first objective was to evaluate the association betweenHuRstaining score and disease stage for 55 tissue samples (4 cholan-giocarcinoma tissues were excluded from the analysis). Theassociation between HuR staining score (0–1 or 2–3) and diseasestage [0, early (I/II), or late (III/IV)] was statistically significant(P ¼ 0.0013) and showed a significant increasing trend (c2 for

trend P ¼ 0.0007; Fig. 6B). Associations of scores for stainingintensity, percentage of positive cells, and cytoplasmic stainingwith disease stage were also significant (P ¼ 0.002, 0.002, and0.02, respectively).

The second objective was to look at a possible association ofdisease stage and HuR staining score with the OS for 42 patients

HepG2 SNU449

Fas

HuR

β-Actin

SNU398

SiRNA: CON CONHuR HuR CON HuR

A

B C D

Co

un

t/su

rfac

e F

as

Co

un

t/su

rfac

e F

asL

sFasL(100 ng/mL)

Nontreatment

Cel

l dea

th (

% t

ota

l)

IgG

0 0

0 102 103 104 105

PE-A0 102 103 104 105

PE-A

Anti-Fas/Sir-CONAnti-Fas/Sir-HuR

IgMAnti-FLAG/Sir-CONAnti-FLAG/Sir-HuR Sir-CON

Sir-HuR

80

60

40

20

0

Figure 4.Effects of HuR knockdown on Fasprotein expression, binding of FasL andapoptosis. A, the expression of HuR andFas detected by immunoblot analysis inindicatedHCCcell lines transfectedwithSir-HuR or Sir-CON. B, surface Faswas detected by flow cytometry ofPE-conjugated Fas-antibody UB2-labeled Sir-HuR or Sir-CON transfectedHepG2 cells. C, the binding of FLAG-tagged Fas ligand to the surface ofSir-HuR or Sir-CON–transfected HepG2cells was evaluated by flow cytometryafter staining with fluorescently labeledanti-FLAG antibody. D, apoptosis ofSir-HuR or Sir-CON–transfected HepG2cells in response to FasL was evaluatedby PI exclusion and flow cytometry.�� , P � 0.001. Graphical representationof the flow cytometry data in a bargraph is shown. Corresponding flowcytometry plots are provided inSupplementary Fig. S2A.

A B

C

Wit

ho

ut

Jo2

pcDNA3.1 pcDNA3.1-HuR

HuR

Fas

Pro–caspase-8

Pro–caspase-3

Cleavedcaspase-8

Cleavedcaspase-3

Cleaved PARP

b-Actin

BID

pcDNA3.1120

100

80

60

40

20

00 1 2 3 4 5 6 7 8 9

P = 0.02

Time (h)10 11 12

Su

rviv

al (

%)

pcDNA3.1-HuR

pcDNA3.1

pcDNA3.1-HuR

Jo2 antibody

Wit

h J

o2

Figure 5.Effect of HuR overexpression on micesurvival and liver cell apoptosis afterchallenge with agonistic anti-Fasantibody. Mice were transfected withpcDNA3.1 or pcDNA3.1-HuR byhydrodynamic transfection method and24 hours later they were challengedwith a lethal dose of agonistic anti-Fasantibody Jo2. A, Kaplan–Meiersurvival curves for pcDNA3.1 andpcDNA3.1-HuR–transfected mice.B, the gross appearance of livers fromchallenged and control mice.C, immunoblot analysis of Fas apoptoticsignaling proteins in livers of transfectedmice.

Zhu et al.





with HCC. We confirmed that patients with early-stage (I/II)disease had better survival than late-stage (III/IV) patientsalthough differences showed marginal significance; P ¼ 0.0579(Fig. 6C). There were no statistically significant differencesin survival by gender (P¼ 0.1303) although, association betweenstage of the disease (I/II and III/IV) and gender was statisticallysignificant (P¼ 0.0471); 95% of stage III/IV disease patients weremale compared with 68.2% of patients with stage I/II disease.Patients with the highest HuR staining score (score 3) hadworse survival than patients with undetectable to moderate HuRstaining score (scores 0–2), although the observed differenceswerenot statistically significant (Fig. 6D). Interestingly, analysis ofsurvival of 22 HCC patients with the early-stage (I/II) diseasesuggested that HuR staining score of 3 is associated with worseprognosis (Fig. 6E). However, the difference in OS between theundetectable to moderate (0–2) and high score groups (3) wasnot significant (P ¼ 0.086). The 5-year OS rates for these groupswere 0.78 (95% CI, 0.55–1.00) and 0.38 (0.19–0.76),respectively.

DiscussionIn this report, we show that HuR is consistently overexpressed

in HCC-derived cell lines and HCC tumor tissues. We show aninverse correlation between the expression of HuR and the deathreceptor Fas in HCC-derived cell lines. HuR interacts with the 30-UTR of Fas mRNA via two AREs to suppress its translation thatexplains the differential Fas expression in normal hepatocytescomparedwithHCC.We confirmedHuR-mediated Fas regulationin three HCC cell lines, normal liver cells, and inmouse liver. In acohort of unselected patients, the HuR levels correlated signifi-cantly with the advanced clinical HCC stage and high HuRstaining was associated with shorter survival of patients with theearly-stage (stage I or II) HCC. The significance of the observedcorrelations is likely to be improved if larger cohorts of HCCpatients were studied.

The properties ofHuR protein extending beyond Fas regulationhave generated wide interest. HuR controls the expression of apanel of proteins that are critical for transformationandaggressive

tumor growth (9). Abundant expression of HuR correlates withaggressive growth of renal cell carcinoma, breast cancer, gastriccancer and a suggested correlationmay hold true for HCC, as well(9, 18).

Under unstimulated conditions in nontransformed cells,HuR is expressed constitutively at low levels and primarilyresides in the nucleus. Transformed cells or cells under stressshow enhanced expression of HuR protein and enhancedexport of HuR mRNA to the cytoplasm. Several recent reportselucidated some of the mechanisms of HuR upregulation inHCC; hepatitis B encoded X protein (HBx) induces expressionof HuR while murine double minute 2 (Mdm2) stabilizes HuRprotein by NEDDylation and regulates its nucleocytoplasmicshuttling (19, 20). The cytoplasmic translocation is a prereq-uisite for HuR's ability to regulate mRNAs stability or transla-tion through binding to their AREs (15). HuR positively reg-ulates the stability of over 80 mRNAs that encode proteinsinvolved in apoptosis inhibition, inflammation, and cell pro-liferation (9), including HAUSP, a regulator of p53 stability,during progression of nonalcoholic steatohepatitis to HCC(21). In contrast, the expression of a few HuR targets (c-myc,wnt5a, and p27) has been shown to be suppressed by HuRthrough blocking their translation (9, 17, 22, 23). In the case ofthe Fas receptor, HuR has been previously shown to mediatesplicing out Fas exon 6, leading to the production of a solubledecoy Fas (7, 10, 11). Our data indicate that in HCC, HuRsuppresses expression of the Fas receptor by blocking FasmRNA translation without significantly affecting Fas mRNAstability or splicing. To our knowledge, this is the first report ofHuR-mediated inhibition of Fas translation and confirms thatconstitutive overexpression of HuR in HCC may be responsiblefor the low Fas-expressing phenotype.

HuR blocks apoptosis and promotes survival by regulatingmRNAs that encode proteins such as p21, Mcl-1, and Bcl-2(24, 25). Our data revealed that in addition to the abovementioned apoptosis regulating proteins, HuR also blocksFas-mediated apoptosis by interfering with the expression ofFas receptor. Using flow cytometry, we confirmed increasedcell surface expression of Fas receptor and subsequentlyincreased binding of FasL and apoptosis in HuR knockdowncells. Consistent with the cell and tissue analysis, the over-expression of HuR in mouse livers protected liver tissue fromthe injury induced by a Fas agonistic antibody, as evidencedby decreased liver hemorrhage, and attenuated apoptoticsignaling in HuR-transfected mice compared with controlvector-transfected mice.

The downregulation of Fas, gain of FasL, and production ofsoluble Fas are potential mechanisms contributing to HCCdevelopment (26, 27) and were implicated to occur as acoordinated event in HCCs. The existence of a commonupstream regulator of Fas and FasL expression was proposed(28). Our findings indicate that HuR downregulates Fas andpromotes expression of FasL in HCC suggesting that HuR is thissought common regulator.

HuR has been shown to suppress immunity through modula-tion of cytokines levels via regulation of their mRNAs. For exam-ple, HuR binds the mRNAs of MKP-1 and TGFb and promotestheir expression and coordinately suppresses tumor targetingimmune responses (29, 30). HuR has a wider role in circumvent-ing inflammation overall by posttranscriptional suppression ofinflammatory cytokines production (17).

Table 1. TMA characteristics

Patient characteristic n (%)

Number of patients 44 (100)Age, yearsMedian 57Range 20–77

GenderMale 35 (79.5)Female 9 (20.5)

DiagnosisCholangiocarcinoma 2 (4.5)HCC þ cholangiocarc 1 (2.3)HCC 41 (93.2)

StageI 14 (31.8)II 9 (20.5)III 14 (31.8)IV 7 (15.9)

TMA coresPrimary tumor tissue 40 (67.8)Metastatic tissue 4 (6.8)Matched metast. tissue 6 (10.2)Matched normal tissue 9 (15.2)

HuR in Liver Cancer





Alteration of the Fas/FasL system is regarded as a key compo-nent of immune evasion in HCC. One of the mechanisms ofimmune evasion is HuR-mediated stabilization of FasL mRNA(16), which we confirmed also occurs in HCC cell lines assilencing of HuR decreased FasL mRNA levels. Our finding ofHuR-mediated regulation of Fas expression fits well with the HuRfunctions in apparent evasion of immune surveillance (5). Theseevents coupled with the wide ranging effects of HuR on apoptosiscan shield tumors from the immune system.

Treatment of early-stage HCC is by surgical resection. Use oftargeted therapy with sorafenib and other chemotherapies showsincremental clinical improvements. However, most patients withnonresectable, late-stage HCC die within 3 to 6months, and their5-year survival remains a dismal 11% (1).Most effector responsesto chemotherapy depend on an intact and inducible Fas deathreceptor signaling, which is typically blocked in HCC. Our anal-ysis showed that interventions aimed to block HuR expressionallow replenishment of Fas levels and confer cells with thecapacity to respond to apoptotic stimulation with FasL. Restora-tion of Fas expression inHCCwould enable effective apoptosis inresponse to chemotherapy treatments inducing cell killing. Inother cancer settings, the direct modulation of Fas-associatedproteins (PMLRAR, nucleolin) determines the fate of cell apo-ptotic stimulus (12, 14).

Expression patterns of Fas and FasL have been used todistinguish cohorts of HCC patients with short (11 months)and long (52 months) disease-free intervals (27). Havingrecognized in our study that HuR regulates the expression ofboth Fas receptor and FasL, we anticipated that HuR expressionlevels would integrate the prognostic significance of Fas andFasL on survival of patients with HCC. Indeed, high expressionof HuR was associated with shorter survival of patients withearly-stage HCC.

Having demonstrated a role of HuR in HCC apoptosis regu-lation, we queried whether HuR can be targeted pharmacologi-cally. In order for HuR to exert its effects, it must dimerize prior tobinding its targets. Low-molecular-weight inhibitors of HuRdimerization affected levels of HuR-targeted ARE-containingmRNAs and inhibited cell proliferation (31). Also, targeting ofHuR–mRNA complexes by competing AREs has successfully dis-placed mRNAs leading to suppression of proliferation (32).Nuclear export of HuR is an almost exclusive property of stressedor transformed cells and thus a promising target for therapy (33).Thus, targeting of HuR may offer us the ability to modulate abroad range of HuR-mediated protumorigenic effects, includingcancer cell proliferation, survival, angiogenesis, invasion, andmetastases, by interfering with the actions of a single target(24, 25, 34–36).

Normal Stage StageStageStageliver I II III IV

Survival of stage I/II patientsby HuR staining score

Survival by stage

Survival by HuR staining score

22.2% 77.3% 87.5%

77.8% 22.7% 12.5%

968472604836241200.0

Normal liver Stage I/II Stage III/IV

0.2

0.4

0.6

0.8

1.03

2

1

0P = 0.0579

1: I/II (E / N = 11 / 22)2: III/IV (E / N = 14 / 20)

Time (months)

Pro

bab

ility

Hu

R s

tain

ing

sco

re

968472604836241200.0

0.2

0.4

0.6

0.8

1.0

P = 0.2638

1:0–1:2 (E / N = 7 / 15)2:3 (E / N = 18 / 27)

Time (months)

Pro

bab

ility

8472604836241200.0

0.2

0.4

0.6

0.8

1.0

P = 0.086

1:0–1:2 (E / N = 2 / 9)2:3 (E / N = 9 / 13)

Time (months)

Pro

bab

ility

A

B C

D E

Figure 6.HuR expression and correlation withthe stage of HCC and survival. Livercancer TMAwas stained with anti-HuRantibody and scored for stainingintensity and percentage of positivecells independently by twoinvestigators. An HuR staining scorewas obtained from the sum of scoresfor staining intensity and percentageof positive cells. A, examples of HuRstaining in samples representingdifferent stages of HCC. B, distributionof HuR staining score in stages of HCC.Increasing trend of HuR stainingscores 2–3 with increasing HCCstage was significant; c2 for trendP ¼ 0.0007. C, Kaplan–Meier survivalcurves of 42 patients with HCC withearly (I/II) and late (III/IV) stage HCC.(P ¼ 0.058). D, Kaplan–Meier survivalcurves of 42 patients with HCC withundetectable tomoderate (0–1–2) andhigh (3) HuR staining scoresdisregarding the stage of HCC.(P ¼ 0.264). E, Kaplan–Meier survivalcurves of 22 patients with early (I/II)stage HCC according to low (0–1–2)versus high (3) HuR staining score(P ¼ 0.086).

Zhu et al.





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

Authors' ContributionsConception and design: H. Zhu, R.-H. Tao, A.L. Sabichi, B. Blechacz,F. SamaniegoDevelopment of methodology: H. Zhu, L. Sehgal, R.-H. Tao, B. Blechacz,F. SamaniegoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H. Zhu, R. Mathur, L. Sehgal, R.-H. Tao, A. Rashid,F. SamaniegoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): H. Zhu, Z. Berkova, L. Sehgal, L. Feng, B. Blechacz,A. Rashid, F. SamaniegoWriting, review, and/or revision of the manuscript: H. Zhu, Z. Berkova,R. Mathur, L. Sehgal, T. Khashab, R.-H. Tao, X. Ao, L. Feng, A.L. Sabichi,B. Blechacz, A. Rashid, F. SamaniegoAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): H. Zhu, R. Mathur, X. Ao, F. Samaniego

Study supervision: F. SamaniegoOther (developed the animal methodology, designed and performed allthe animal experiments, collected the data, and reviewed the manuscript):R.-H. Tao.

Grant SupportThis work was supported by grants from the National Cancer Institute

(CA158692), the National Institute of Diabetes and Digestive and KidneyDiseases (DK091490), and the American Cancer Society (MSRG-10-052-01-LIB). The MD Anderson Flow Cytometry and Cellular Imaging Facility wasfunded by a Cancer Center Support Grant from the National Cancer Institute(P30CA16672).

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

Received May 5, 2014; revised January 2, 2015; accepted January 3, 2015;published OnlineFirst February 12, 2015.

References1. Blechacz B, Mishra L. Hepatocellular carcinoma biology. Recent Results

Cancer Res 2013;190:1–20.2. Altekruse SF, McGlynn KA, Reichman ME. Hepatocellular carcinoma

incidence, mortality, and survival trends in the United States from 1975to 2005. J Clin Oncol 2009;27:1485–91.

3. Parkin DM. Global cancer statistics in the year 2000. Lancet Oncol2001;2:533–43.

4. Nagao M, Nakajima Y, Hisanaga M, Kayagaki N, Kanehiro H, Aomatsu Y,et al. The alteration of Fas receptor and ligand system in hepatocellularcarcinomas: how do hepatoma cells escape from the host immune sur-veillance in vivo? Hepatology 1999;30:413–21.

5. Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoe-diting: from immunosurveillance to tumor escape. Nat Immunol 2002;3:991–8.

6. Kubo K, Matsuzaki Y, Okazaki M, Kato A, Kobayashi N, Okita K. The Fassystem is not significantly involved in apoptosis in human hepatocellularcarcinoma. Liver 1998;18:117–23.

7. Izquierdo JM. Cell-specific regulation of Fas exon 6 splicing medi-ated by Hu antigen R. Biochem Biophys Res Commun 2010;402:324–8.

8. Brennan CM, Steitz JA. HuR and mRNA stability. Cell Mol Life Sci2001;58:266–77.

9. Abdelmohsen K, GorospeM. Posttranscriptional regulation of cancer traitsby HuR. Wiley Interdiscip Rev RNA 2010;1:214–29.

10. Izquierdo JM. Hu antigen R (HuR) functions as an alternative pre-mRNAsplicing regulator of Fas apoptosis-promoting receptor on exon definition.J Biol Chem 2008;283:19077–84.

11. Alvarez E, Castello A, Carrasco L, Izquierdo JM. Poliovirus 2A pro-tease triggers a selective nucleo-cytoplasmic redistribution of splicingfactors to regulate alternative pre-mRNA splicing. PloS ONE 2013;8:e73723.

12. Wise JF, Berkova Z, Mathur R, Zhu H, Braun FK, Tao RH, et al. Nucleolininhibits Fas ligand binding and suppresses Fas-mediated apoptosis in vivovia a surface nucleolin-Fas complex. Blood 2013;121:4729–39.

13. Berkova Z,Wang S,Wise JF, Maeng H, Ji Y, Samaniego F. Mechanism of Fassignaling regulation by humanherpesvirus 8K1oncoprotein. JNatl CancerInst 2009;101:399–411.

14. Tao RH, Berkova Z, Wise JF, Rezaeian AH, Daniluk U, Ao X, et al.PMLRARalpha binds to Fas and suppresses Fas-mediated apoptosisthrough recruiting c-FLIP in vivo. Blood 2011;118:3107–18.

15. Vazquez-ChantadaM, Fernandez-RamosD, EmbadeN,Martinez-LopezN,Varela-Rey M, Woodhoo A, et al. HuR/methyl-HuR and AUF1 regulate theMAT expressed during liver proliferation, differentiation, and carcinogen-esis. Gastroenterology 2010;138:1943–53.

16. Drury GL, Di Marco S, Dormoy-Raclet V, Desbarats J, Gallouzi IE. FasLexpression in activated T lymphocytes involves HuR-mediated stabiliza-tion. J Biol Chem 2010;285:31130–8.

17. Katsanou V, Papadaki O, Milatos S, Blackshear PJ, Anderson P, Kollias G,et al. HuR as a negative posttranscriptional modulator in inflammation.Molecular cell 2005;19:777–89.

18. Ronkainen H, Vaarala MH, Hirvikoski P, Ristimaki A. HuR expression is amarker of poor prognosis in renal cell carcinoma. Tumour Biol2011;32:481–7.

19. Hung CM, Huang WC, Pan HL, Chien PH, Lin CW, Chen LC, et al.Hepatitis B virus X upregulates HuR protein level to stabilize HER2expression in hepatocellular carcinoma cells. Biomed Res Int 2014;2014:827415.

20. EmbadeN, Fernandez-RamosD, Varela-ReyM, BerazaN, SiniM, Gutierrezde Juan V, et al. Murine doubleminute 2 regulates Hu antigen R stability inhuman liver and colon cancer through NEDDylation. Hepatology2012;55:1237–48.

21. Martinez-Lopez N, Varela-Rey M, Fernandez-Ramos D, Woodhoo A, Vaz-quez-Chantada M, Embade N, et al. Activation of LKB1-Akt pathwayindependent of phosphoinositide 3-kinase plays a critical role in theproliferation of hepatocellular carcinoma from nonalcoholic steatohepa-titis. Hepatology 2010;52:1621–31.

22. Nabors LB, Gillespie GY,Harkins L, King PH.HuR, a RNA stability factor, isexpressed in malignant brain tumors and binds to adenine- and uridine-rich elementswithin the 30 untranslated regions of cytokine and angiogenicfactor mRNAs. Cancer Res 2001;61:2154–61.

23. Leandersson K, Riesbeck K, Andersson T. Wnt-5a mRNA translation issuppressed by the Elav-like protein HuR in human breast epithelial cells.Nucleic Acids Res 2006;34:3988–3999.

24. Abdelmohsen K, Lal A, Kim HH, Gorospe M. Posttranscriptionalorchestration of an anti-apoptotic program by HuR. Cell Cycle 2007;6:1288–92.

25. Wang W, Furneaux H, Cheng H, Caldwell MC, Hutter D, Liu Y, et al. HuRregulates p21 mRNA stabilization by UV light. Mol Cell Biol 2000;20:760–9.

26. Hammam O, Mahmoud O, Zahran M, Aly S, Hosny K, Helmy A,et al. The role of fas/fas ligand system in the pathogenesis of livercirrhosis and hepatocellular carcinoma. Hep Monthly 2012;12:e6132..

27. Ito Y, Monden M, Takeda T, Eguchi H, Umeshita K, Nagano H, et al. Thestatus of Fas and Fas ligand expression can predict recurrence of hepato-cellular carcinoma. Br J Cancer 2000;82:1211–7.

28. Roskams T, Libbrecht L, Van Damme B, Desmet V. Fas and Fas ligand:strong co-expression in human hepatocytes surrounding hepatocellularcarcinoma; can cancer induce suicide in peritumoural cells? J Pathol2000;191:150–3.

29. Wang X, Liu Y. Regulation of innate immune response by MAP kinasephosphatase-1. Cell Signal 2007;19:1372–82.

30. Beck C, Schreiber H, Rowley D. Role of TGF-beta in immune-evasion ofcancer. Microsc Res Tech 2001;52:387–95.

HuR in Liver Cancer





31. Meisner NC, Hintersteiner M, Mueller K, Bauer R, Seifert JM,Naegeli HU, et al. Identification and mechanistic characterizationof low-molecular-weight inhibitors for HuR. Nat Chem Biol 2007;3:508–15.

32. SunDQ,WangY, LiuDG.Cancer cell growth suppression by a62ntAU-richRNA from C/EBPbeta 30UTR through competitive binding with HuR.Biochem Biophys Res Commun 2012;426:122–8..

33. Siddiqui N, Borden KL. mRNA export and cancer. Wiley Interdiscip RevRNA 2012;3:13–25.

34. WangW,CaldwellMC, Lin S, FurneauxH,GorospeM.HuR regulates cyclinA and cyclin B1 mRNA stability during cell proliferation. EMBO J2000;19:2340–50.

35. Levy NS, Chung S, Furneaux H, Levy AP. Hypoxic stabilization of vascularendothelial growth factor mRNA by the RNA-binding protein HuR. J BiolChem 1998;273:6417–23.

36. Sheflin LG, Zou AP, Spaulding SW. Androgens regulate the binding ofendogenous HuR to the AU-rich 30UTRs of HIF-1alpha and EGF mRNA.Biochem Biophys Res Commun 2004;322:644–51.


Zhu et al.




2015;13:809-818. Published OnlineFirst February 12, 2015.Mol Cancer Res Haifeng Zhu, Zuzana Berkova, Rohit Mathur, et al. Outcome in Liver CancerHuR Suppresses Fas Expression and Correlates with Patient

Updated version

10.1158/1541-7786.MCR-14-0241doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2015/02/11/1541-7786.MCR-14-0241.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/13/5/809.full#ref-list-1

This article cites 36 articles, 9 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/13/5/809.full#related-urls

This article has been cited by 3 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://mcr.aacrjournals.org/content/13/5/809To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-14-0241

http://mcr.aacrjournals.org/content/suppl/2015/02/11/1541-7786.MCR-14-0241.DC1

http://mcr.aacrjournals.org/content/13/5/809.full#ref-list-1

http://mcr.aacrjournals.org/content/13/5/809.full#related-urls

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

mailto:[email protected]

http://mcr.aacrjournals.org/content/13/5/809


hur suppresses fas expression and correlates with patient ... · hur suppresses fas expression and...

Documents