Virus-Associated RNA I-Deleted Adenovirus, a Potential Oncolytic Agent Targeting EBV-Associated Tumors

Virus-Associated RNA I–Deleted Adenovirus, a Potential Oncolytic

Agent Targeting EBV-Associated Tumors

Yaohe Wang,1Shao-An Xue,

2Gunnel Hallden,

1Jennelle Francis,

1Ming Yuan,

1

Beverly E. Griffin,2,3and Nick R. Lemoine

1

1Cancer Research UK Molecular Oncology Unit, Institute of Cancer, Barts and the London School of Medicine and Dentistry,Queen Mary University of London; 2Division of Medicine, Hammersmith Hospital, Imperial College London; and3Imperial College London at St. Mary’s, London, United Kingdom

Abstract

Given the growing number of tumor types recognizablyassociated with EBV infection, it is critically important thattherapeutic strategies are developed to treat such tumors.Replication-selective oncolytic adenoviruses representa promising new platform for anticancer therapy. Virus-associated I (VAI) RNAs of adenoviruses are required forefficient translation of viral mRNAs. When the VAI gene isdeleted, adenovirus replication is impeded in most cells(including HEK 293 cells). EBV-encoded small RNA1 isuniformly expressed in most EBV-associated human tumorsand can functionally substitute for the VAI RNAs ofadenovirus. It enables replication to proceed throughcomplementation of VAI-deletion mutants. We hypothesizedthat VAI-deleted adenovirus would selectively replicate inEBV-positive tumor cells due to the presence of EBV-encodedsmall RNA1 with no (or poor) replication in normal or EBV-negative tumor cells. In this report, we show that high levelsof replication occurred in the VAI-deleted mutant in the EBV-positive tumor cells compared with low (or negligible) levelsin EBV-negative and normal human primary cells. Corre-spondingly, high toxicity levels were observed in EBV-positivetumor cells but not in EBV-negative tumor or normal humanprimary cells. In vivo , VAI-deleted adenovirus showedsuperior antitumoral efficacy to wild-type adenovirus inEBV-positive tumor xenografts, with lower hepatotoxicitythan wild-type adenovirus. Our data suggest that VAI-deletedadenovirus is a promising replication-selective oncolyticvirus with targeting specificity for EBV-associated tumors.(Cancer Res 2005; 65(4): 1523-31)

Introduction

EBV, a ubiquitous human herpesvirus found latently expressedin 90% of the human population, is the causative agent ofinfectious mononucleosis and posttransplant immunoprolifera-tive disorders in immunocompromised patients. It is closelyassociated with the development of a variety of malignantdiseases, including endemic Burkitt’s lymphoma, B-cell lympho-ma of immunocompromised patients, nasopharyngeal carcinoma,Hodgkin’s disease, T-cell lymphoma, natural killer cell leukemia/lymphoma, smooth-muscle tumors, and gastric cancer (1, 2).

More recent reports have linked EBV with conventional epithelialcancers of other sites, including breast (3–5), lung (6–9), prostate(7), liver (10), colon (7), and also with lymphoepithelioma-likecarcinoma of the esophagus (11). Given the ever-growingnumber of tumor types associated with EBV infection, thera-peutic strategies to treat EBV-associated tumors may beconsidered a high priority in oncology.Replication-selective, oncolytic viruses provide a new platform

to treat cancer. Promising clinical trial data with mutantadenoviruses have shown both their antitumor potency andsafety (12, 13). Two main approaches are currently being used toengineer adenoviruses with tumor-selective replication. The firststrategy limits the expression of the E1A gene to tumor tissuesthrough the use of tumor- and/or tissue-specific promoters;consequently, E1A-induced S phase entry and activation of viraland cellular genes will only occur in tumor tissues. A secondstrategy optimizes tumor selectivity by deletion of viral genesthat are critical for efficient replication in normal cells butexpendable in tumor cells (14). The most commonly describedoncolytic viruses explored over the past 10 years have beenmutant adenoviruses where at least 25 different ones have beendescribed for use in cancer treatments to date (15).Several studies have shown similarities in the function of

adenovirus and EBV genes (16–20). Based on the similaritiesdescribed, we believe that a window of opportunity exists in thedesign of replication-selective adenovirus for EBV-associatedtumors. The adenovirus genome encodes two RNA polymeraseIII-directed, f160-nucleotide-long RNAs, the so-called virus-associated RNA I and RNA II. These accumulate to high levelsduring the late stages of viral infection (21–23). VAI RNA isobligatory for efficient translation of viral and cellular mRNAs(24, 25). It binds to and blocks the activation of cellular double-stranded RNA-dependent serine/threonine protein kinase (PKR),which phosphorylates protein translation initiation factor eIF-2a,enabling protein synthesis to proceed at normal levels (26–29). Amutant that fails to produce VAI RNA (dl331) grows more poorlythan its parent in human 293 cells (25).EBV also expresses two low-molecular-weight RNAs, designat-

ed EBV-encoded RNAs (EBER) 1 and 2, transcribed by RNApolymerase III (30). Although there is no striking nucleotidesequence homology between virus-associated RNAs and EBERs,the RNAs are similar in size, genomic organization, and degreeof secondary structure; they have similar functions (19, 20, 31).In particular, the two small RNAs encoded by EBV can efficientlycomplement the VAI RNA–mediated translational defect inadenovirus-infected cells, as shown by constructing an Ad5substitution mutant in which the two virus-associated RNAgenes have been deleted and replaced by an EBV DNA segmentencoding the two EBERs (32, 33). Given the uniform expression

Requests for reprints: Nick R. Lemoine, Cancer Research UK MolecularOncology Unit, Institute of Cancer, Barts and the London School of Medicine andDentistry, Queen Mary University of London, John Vane Science Building,Charterhouse Square, London EC1 6BQ, United Kingdom. Phone: 44-20-7014-0420;E-mail: [email protected].

I2005 American Association for Cancer Research.

www.aacrjournals.org 1523 Cancer Res 2005; 65: (4). February 15, 2005

Research Article

Research. on November 7, 2015. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

of EBERs in human EBV-associated tumors, generally at highlevels, and the similar functions between VAI RNA and EBER1,we hypothesized that VAI RNA–defective adenovirus could becomplemented by EBER1 in EBV-positive tumors, thus restrictingvirus-induced cell killing and replication to EBV-positive tumors.In the present report, we tested this hypothesis. We show thatVAI-deleted adenovirus is a potential therapeutic agent that canspecifically target EBV-associated tumors.

Materials and Methods

Cell Lines and Cell Culture. Human Burkitt’s lymphoma (Raji),

epidermoid carcinoma of larynx (Hep-2), lung cancer (A549), andembryonic kidney (HEK; a subclone named JH-293) cell lines were all

obtained from the Cancer Research UK Central Cell Service (Clare Hall,

Middlesex, United Kingdom). The Jijoye cell line was obtained from

American Type Culture Collection (Manassas, VA) and the humannasopharyngeal carcinoma line C666-1 from Prof. Dorly P. Huang (The

Chinese University of Hong Kong, Hong Kong, China). The presence of

EBV has been consistently shown in C666-1 (34). Raji, Jijoye, Hep-2, and

C666 cells were maintained in RPMI containing 10% fetal bovine serum,2% L-glutamine, and antibiotics. JH-293 and A549 cells were maintained

in DMEM supplemented with 10% fetal bovine serum and antibiotics.

The human gastric cancer cell lines AGS and AGS-EBV were a generousgift from Prof. Lindsey Hutt-Fletcher (Louisiana State University,

Shreveport, LA). The AGS-EBV cell line was obtained by G418 selection

of AGS cells that were infected with a recombinant Akata virus in which

a neomycin resistance cassette had been inserted into the nonessentialBDLF3 open reading frame. Both the AGS and AGS-EBV cell lines were

maintained in Ham’s F12 medium, supplemented with 10% fetal bovine

serum, antibiotics, and G418 at 500 Ag/mL (AGS-EBV cells only). Normal

human bronchial epithelial cells (NHBE) and normal prostate epithelialcells (PrEC) were obtained from Clonetics Corp. (San Diego, CA) and

were propagated as recommended by the manufacturer. All cells were

maintained at 37jC in a humidified atmosphere containing 10% CO2.

Viruses. VAI-deleted adenovirus dl331 (25), wild-type Ad5 and E3B-deleted adenoviral mutant dl309 (35) were gifts from Prof. Bayar

Thimmappaya (Northwestern University, Chicago, IL) and Dr. Thomas

Shenk (Princeton University, Princeton, NJ). The replication-defective Ad5mutant dl312 was inactivated by psoralen-UV, as previously described, to

serve as a nonreplicating particle control for in vitro cell survival and in

vivo efficacy studies. Ad5, dl309, and dl312 adenoviruses were amplified

in the human embryonic kidney cell line HEK 293; dl331 was generatedin A549 cells. All adenoviruses were purified and titered as previously

described (36).

Viral Infection and Replication Assay. Cells were seeded in six-well

plates (3 � 105-1 � 106 cells/well, according to their growth properties) andinfected with viruses at 100 particles per cell when 70% to 80% confluence

had been reached. For all human tumor cells, viral infections were done in

standard growth medium without serum. Primary human cells wereinfected in predefined growth medium, with no alterations to growth factor

components. Infections were done in a low volume of medium for 2 hours

at 37jC/10% CO2. Medium was then replaced either with fresh medium

containing 2% fetal bovine serum for tumor cells or with growth factors forprimary human cells. Experiments were repeated under each condition in

triplicate. Cells and media were collected at 48 and 72 hours postinfection.

The collected samples were freeze-thawed thrice, and replicating virus was

determined by the limiting dilution assay as described (36) with the use ofA549 and 293-T cells (293 cells transformed with SV40 T antigen) as

indicator cells. A549 supports VAI-deleted adenovirus replication to an

equivalent level to wild-type adenovirus (37), and SV40 T antigen (in 293-T)

could rescue the translational defect resulting from infection with VAI-deleted adenovirus (38).

In vitro Cell Survival Assay. For cell survival assays, 1 � 104 cells (1 �103 cells of Hep-2) were seeded in each well of 96-well plates in 100 ALmedium and infected 24 hours later with serial dilutions of different

viruses. Cell survival was determined by 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfony)-2H-tetrazolium assay (MTS, Prom-

ega, Madison, WI ) 6 days postinfection, and the EC50 values of different

adenoviruses obtained as previously described (39). For NHBE and PrEC

cells, 100 AL fresh medium was added to each well 3 days postinfection.All assays were done at least thrice with each concentration of virus

in sextuplicate.

Detection of EBER1 Expression by Reverse Transcription-PCR. TotalRNA was extracted from different cells, using an RNA extraction kit(Qiagen, Crawley, United Kingdom) according to the manufacturer’s

instructions. To synthesize cDNA, 1 Ag of total RNA was treated with

10 units of DNase I (Roche, Lewes, United Kingdom) at 37jC for 30

minutes, as described (40). After purification, DNase I-treated RNA wasreverse transcribed by Superscript II reverse transcriptase (Invitrogen,

Renfrew, United Kingdom) with 50 pmol pd(N)6 random hexamer primers

(Amersham Biosciences, Chalfont St. Giles, United Kingdom) in 20 AL. ForPCR amplification, 1 AL cDNA was added to 49 AL PCR reaction mixture

[75 mM Tris-HCl (pH 8.8), 20 mM (NH4)2SO4, 0.01% (vol/vol) Tween 20,

1.5 mM MgCl2, 0.16 mM (each) deoxynucleoside triphosphate, and 30

pmol of each primer] containing 1.5 units of TaqDNA polymerase(ABgene, Epsom, United Kingdom). To analyze expression of EBER1, the

primer pairs used were 5V-AGGACCTACGCTGCCCTAGA-3V ( forward) and5V-CCCTAGTGGTTTCGGACACAC-3V (reverse). After 30 cycles of amplifi-

cation, PCR products were separated by electrophoresis on 1% agarosegels containing ethidium bromide. The human b-actin gene was used as

an internal control.

Western Blotting. Control and infected cells were lysed at indicatedtimes with lysis buffer [20 mmol/L Tris-HCl (pH 7.6), 400 mmol/L NaCl, 50

mmol/L KCl, 5 mmol/L h-mercaptoethanol, 1% Triton X-100, 20% glycerol,

100 Ag/mL phenylmethylsulfonyl fluoride, 1 Ag/mL leupeptin, and 10 Ag/mL

aprotinin]. Protein lysates were quantitated by the bicinchoninic acid assay(Sigma, Poole, United Kingdom). Proteins were separated by reducing SDS-

10% polyacrylamide gel electrophoresis, electrotransferred onto nitrocellu-

lose membranes and probed with a mouse anti-total PKR antibody (Santa

Cruz Biotechnology, Santa Cruz, CA), or a rabbit anti-phospho-PKR(Biosource, Camarillo, CA), or rabbit anti-eIF-2a[pS52] phosphospecific

antibody (Biosource), or mouse anti-human proliferating cell nuclear

antigen antibody (Research Monoclonal Antibody Service, Cancer ResearchUK, London, United Kingdom). Immunocomplexes were detected by

incubation with a horseradish peroxidase-conjugated secondary antibody

(DAKO, Glostrup, Denmark) and visualized by enhanced chemiluminescence

reagent (Amersham Biosciences).In vivo Efficacy, Replication, and Gene Expression. Two approaches

were used to evaluate the antitumor efficacy of VAI-deleted adenovirus and

wild-type Ad5 in vivo as described below. All animal experiments were

approved by the Cancer Research UK Animal and Safety Committee and theUnited Kingdom Home Office.

In the first protocol, cultured C666-1 cells were infected with vehicle buffer,

or psoralen-UV-inactivated dl312, wild-type Ad5, and dl331, for 2 hours at 1,000particles per cell. Fourteen hours postinfection, 10% preinfected C666-1 cells

mixed with 90% uninfected cells, and 1.5 � 107 mixed cells, in 200 AL of PBS,

were injected s.c. into the flanks of 6 BALB/c nu/nu mice (4-6 weeks old) for

each group. Tumor growth, as measured by volume, was monitored twiceweekly.

In the second protocol, 2 � 107 C666-1 cells were suspended in 200 ALnormal saline and injected s.c. into the flanks of athymic (nu/nu) mice (4-6

weeks old) and allowed to grow. Volume was estimated twice weekly using

the formula: volume = (length � width2)(3.142/6). Once tumors reached the

size range of 33.5 to 117 AL, mice were stratified by tumor size and thenrandomly allocated to treatment groups; treatment groups were balanced by

tumor size at the time of treatment initiation in all cases (P > 0.8, t test for

tumor volumes). To treat the established tumors, 100 AL of PBS without or

with virus (Ad5, dl331, or replication-incompetent control adenovirus,PUVdl312) were directly injected into them. Injections were introduced

through a single central tumor puncture site and three to four needle tracts

made radially from the center. For efficacy experiments, tumors were injected

(1 � 1010 particles/d) on 5 consecutive days, and for biological end point

Cancer Research

Cancer Res 2005; 65: (4). February 15, 2005 1524 www.aacrjournals.org



studies, tumors were injected once only, on day 1. Tumors were harvested at

stated time points after injection for analysis of viral gene expression,

replication, and general histopathology as described (15). Tumor size andanimal survival were monitored. Animals were sacrificed when tumor

reached the size limited by Home Office regulations or after 3 months.

Survival analysis was done according to the method of Kaplan-Meier.

Viral Hepatoxicity. Murine lung adenocarcinoma CMT-64 (5� 106) cells,were injected s.c. into both flanks of immunocompetent C57/BL6 mice

(obtained from Cancer Research UK Biological Resources). The animals were

randomized into four groups (n = 9 per group) and treated with PBS,PUVdl312, wild-type Ad5, and dl331. I.v. injection of viruses (1 � 1010

particles in a volume of 100 AL) via the tail vein was initiated once the tumor

size reached 40 to 80 mm3. At 24, 48, and 72 hours postinjection, mice were

sacrificed (three mice per time point), liver samples were collected andprocessed for general histopathology, and viral protein expression was

analyzed as described (15).

Statistical Analysis. Unpaired t test or one-way ANOVA were used for

all in vitro studies and in vivo toxicity studies. Kaplan-Meier curves were

compared using the log-rank test.

ResultsEBER1 Expression in Normal Human Primary and Trans-

formed Cells. To determine the selectivity of replication of VAI-deleted adenovirus targeting EBV-associated tumors and cell killing,a panel of human cell lines including normal primary and EBV-negative and EBV-positive tumor cell lines was used, as listed inTable 1. Given that EBV resides in 90% of the human population, theexpression pattern of EBER1 in all cells was predetermined bysemiquantitative reverse transcription-PCR before testing ourcomplementation hypothesis. Raji, Jijoye, AGS-EBV, and C666-1expressed varying levels of EBER1 RNA whereas other cells werenegative (Fig. 1; Table 1). Among the four EBV-positive cell lines, thehighest EBER1 RNA expression levels were observed in the C666-1and Raji cells, whereas slightly lower levels were detected in theAGS-EBV cells previously stably transfected with the EBV genome.Selectivity of EBER1-Positive and EBER1-Negative Cells to

Oncolysis by VAI-Deleted Adenovirus. EC50 values were deter-mined in each cell to compare the cell killing potencies of VAI-deleted and wild-type adenovirus in normal and tumor cells withdifferent EBER1 status (Table 1). The ratios of EC50 value for dl331and wild-type Ad were compared and correlated to EBER1 status ineach cell line (Fig. 2A). In the panel of human cells examined, thedl331 mutant showed a cell killing potency that was equivalent orsuperior to wild-type adenovirus in the EBV-positive AGS-EBV, Raji,Jijoye, and C666-1 cells. In contrast, the dl331 mutant was markedlyattenuated in the EBV-negative human cells tested both in tumor celllines and normal primary human cells. Surprisingly, Panc-1 cells, ahuman ras mutant pancreatic cancer cell line previously reported tobe susceptible to dl331 (41), showed a very low level of cytopathiceffect to dl331 in the present study. These results show that the VAI-deleted adenovirus could selectively kill EBER1-positive cell lines,whereas they have little effect on EBV-negative and normal cells. Inaddition, the VAI-deleted adenovirus was as potent as wild-typeadenovirus in the EBER1-positive tumor cell lines.Viral Replication of VAI-Deleted Adenovirus in Normal

Primary Human, EBV-Negative and EBV-Positive Tumor Cells.To determine whether dl331-induced selective cell killing wasindicative of replication induction in EBV-positive cells, replica-tion of mutant and wild-type adenovirus was compared inindividual cell lines. The EBV-positive (AGS-EBV, Jijoye, Raji, andC666-1), EBV-negative (JH-293, Panc-1, Hep-2, and AGS-ATCC)and normal human primary (NHBE and PrEC) cells were infectedwith VAI-deleted and wild-type adenoviruses, and cells wereharvested at 48 or 72 hours. The VAI-deleted mutant replicated tolevels of 20% to 73% that observed for wild-type Ad5 in all four

Table 1. Human cells evaluated and EBER1 expression

Cell name Derived tissues EBER1 status

NHBE Normal humanbronchial epithelial cells

�

PrEC Normal human

prostate epithelial cells

�

HEK Human embryonalkidney cells

�

A549 Human lung

cancer cells

�

Panc-1 Human pancreatic

cancer cells

�

Hep-2 Human epidermoid

carcinoma of larynx

�

AGS Human stomach

cancer cells

�

AGS-EBV AGS-transfected

with EBV

+

Jiyoye African Burkitt’s

lymphoma cells

+

Raji African Burkitt’slymphoma cells

+

C666-1 Human nasopharyngeal

carcinoma cells

+

Figure 1. EBER1 RNA expression in a panel of humancells by reverse transcription-PCR. Total RNA extraction,reverse transcription of RNA, and PCR amplification weredescribed in Materials and Methods. PCR products wereseparated on 1% agarose gels and visualized byethidium bromide staining. Lane M, 100 bp ladder; lane 1,DNase I–treated EBV-C15 NPC DNA; lane 2, NHBE;lane 3, PrEC; lane 4, JH-293; lane 5, A549; lane 6,Panc-1; lane 7, HeLa; lane 8, Hep-2; lane 9, AGS-ATCC;lane 10, AGS-EBV; lane 11, Jijoye; lane 12, Raji,lane 13, C666-1 late passage; lane 14, C666-1 earlypassage; lane 15, primers only.

Oncolytic VAI-Deleted Mutant Adenovirus




EBV-positive tumor cell lines (Fig. 2B). The highest level ofreplication of dl331 was observed in C666-1 cells, which alsoexpress the highest level of EBER1 RNA among these cells. Viralproduction by VAI-deleted adenovirus in EBV-negative cells waslower than in EBV-positive cells. To our surprise, viral productionby VAI-deleted adenovirus in normal primary cells was f4,000times less than wild-type adenovirus in NHBE, and 500 to 1,000times less in PrEC cells, respectively. Of note, among EBV-negative human tumor cell lines, only Panc-1 showed a relativelevel of replication, that is f20% that of wild-type Ad5 at72 hours.

Because in each cell line, viral infection, replication, lysis, andspread are dependent on their intrinsic properties, it may not bereasonable to compare two tumor cell lines derived from differentsources. Therefore, data from matched cell lines, such as AGS andAGS-EBV, could be more meaningful. The viral replication ratio ofdl331/Ad5 in AGS-EBV at 48 hours proved to be 16 times (P < 0.01)higher than in AGS cells, and after 72 hours increased to 33 times(P < 0.001; Fig. 2B).These results show that VAI-deleted adenovirus can selectively

replicate in EBV-positive tumor cells with no significant replicationbeing observed in normal primary cells.

Figure 2. Selective cell killing and replication of VAI-deleted adenovirus in EBV-positive and EBV-negative cells. A, EC50 value comparison for wild-type Ad5 andVAI deletion mutant. The EC50 values of Ad5 and mutant dl331 (VAI-deleted) in different human transformed cell lines were obtained by a cell death assay(MTS assay) as described in Materials and Methods. The EC50 values of the mutant virus were compared with the wild-type control and an EC50 ratio wascalculated as Ad5/dl331. A ratio >1.0 shows that the potency of mutant virus is greater than that of wild-type virus. The data represent triplicate experiments.B, comparison of viral production by Ad5 and dl331 adenovirus in normal human primary cells and EBV-positive and EBV-negative tumor cells. Measurementswere done by tissue-culture infectious dose-50% assay on A549 cells as described in Materials and Methods. Viral production by wild-type Ad5 and dl331 indifferent cells was calculated as plaque-forming unit per cell (n = 3), compared and expressed as the ratio of dl331/Ad5. A ratio of 1 would indicate that the mutantreplicates as well as wild-type virus.

Cancer Research




Changes in PKR and eIF-2A Activation in Response to ViralInfection. VAI RNA regulates mRNA translation and controlsviral replication through the PKR/eIF-2a pathway; EBER1 has asimilar function (19, 42, 43). To test whether the selectivity ofthe VAI mutant for EBV-positive cells occurs through a blockedPKR-eIF-2a pathway, total PKR, phosphorylated-PKR, andphosphorylated-eIF-2a were analyzed by immunoblotting inEBV-positive cells (C666-1) and EBV-negative cells (Hep-2 andPanc-1) at 24 hours postinfection with wild-type Ad5, dl309, anddl331 (Fig. 3). Viral infection resulted in increased expressionlevels of total PKR in all these cell lines and there was nodifference in the level of total PKR between viruses (Fig. 3, 1). Innormal cells, PKR is active through a mechanism involvingautophosphorylation in response to viral infection. Changes inphosphorylated PKR levels were therefore determined in thethree cells following infection with different adenoviruses(Fig. 3, 2). As expected, dl331 infection of EBV-negative cellsshowed a significant elevation of phosphorylated PKR comparedwith wild-type adenovirus infection. However, EBV-positiveC666-1 cells still showed elevated levels of phosphorylatedPKR following dl331 treatment, whereas infection with wild-type Ad5 and dl309 mutant blocked PKR phosphorylation (Fig.3, 2). We further investigated phosphorylation of the substrateof PKR, eIF-2a, using a rabbit anti-eIF-2a[pS52] phosphospe-cific antibody. As observed in Fig. 3, 3 , infection with dl331resulted in a minor increase in the level of phosphorylated eIF-2a in comparison with wild-type adenovirus infection in EBV-positive cells; however, dl331 produced a significantly elevatedlevel of phosphorylated eIF-2a compared with wild-typeadenovirus in EBV-negative cells. These results indicate thatEBER 1 may complement the regulation of eIF-2a by a routenot exclusively dependent on PKR. In addition, no differencesin phosphorylated PKR and phosphorylated eIF-2a levels wereobserved in normal human primary cells following wild-type orVAI-deleted adenovirus infection (data not shown) although thecells showed a significant difference in viral replication withlow or no replication occurring with VAI mutant.Antitumoral Efficacy of the VAI-Deleted Adenovirus In vivo

Is Equivalent or Superior to that of Wild-type Adenovirus inEBV-Positive Tumors. To evaluate whether dl331 could inhibit

tumor growth in vivo , preinfected C666-1 cells were implanted

s.c. in BALB/C nu/nu mice as described in Materials and

Methods. Tumor formation and growth were monitored over

time. Tumors formed in all PBS-treated control animals and in

83.3% (5 of 6) of animals that received inactivated, nonreplicating

control adenovirus at 27 days postimplantation. No tumor

formed in any of the animals that received 10% of wild-type

Ad5- or dl331-preinfected cells up to 205 days postimplantation

with the mixed tumor cells. Statistical analysis confirmed

significant differences between both of wild-type Ad5- and

dl331-treated groups and each control group (P < 0.01).The antitumor efficacy of dl331 was also confirmed in preestab-

lished s.c. C666-1 tumors treated by intratumoral virus administrationas described in Materials and Methods. Tumor growth and survivalrate over time were monitored (Fig. 4A and B). The dl331 mutantsignificantly inhibited tumor growth compared with the two controlgroups, and was superior to wild-type Ad5 (Fig. 4A). Significantlyprolonged survival was also observed in the dl331-treated animalscompared with Ad5-treated animals (Fig. 4B). The dl331 and Ad5treated groups had 75% and 50%, respectively, animals alive after75 days compared with controls with no animals survivingafter 50 days.To confirm cell death and active viral replication in the tumor

mass, H&E and immunohistochemical staining with antiviralprotein E1A and hexon antibodies were done on cross sectionstaken on day 5 after single viral injections (Fig. 4C). H&Estaining confirmed the killing of tumor cells from live virus-treated tumors, with dl331 resulting in more severe cytopathiceffects than wild-type adenovirus (Fig. 4C , 1). Equivalent levels ofviral proteins E1A and hexon were observed in EBV-positive tumorsafter single treatmentswith dl331 andwild-type adenovirus (Fig. 4C , 2and 3), demonstrating that viral replication had occurred within thetumor cells.VAI-Deleted Mutant Results in Reduced Hepatotoxicity

In vivo . Intratumoral treatment with replication-selective onco-lytic adenovirus is well tolerated in both mice and humans (44).However, systemic administration might have higher toxicityrisks in organs such as liver. The systemic delivery ofreplication-competent adenoviruses, even when modified fortumor-selective replication, has previously been shown to causedose-limiting toxicity in mice, associated with liver necrosisresulting from direct exposure to adenoviral coat proteins, viralearly gene expression, viral replication, and/or de novo late geneexpression (45). The host immune response may also play a rolein mediating this dose-limiting toxicity. To explore this topic, theacute hepatotoxicity of dl331 and wild-type adenoviruses in therecently developed CMT-64 immunocompetent mouse tumormodel (36) was investigated, as described in Materials andMethods. Histopathologic observations showed that cytopathiceffects of hepatocytes following dl331 treatment decreased withtime, whereas those of wild-type adenovirus were consistentlyhigher at three time points. Viral late gene (hexon) expressionand cytopathic effects in the liver after dl331 treatment weresignificantly lower than that found with wild-type adenovirus

Figure 3. PKR and eIF-2-a alterations inEBV-positive and EBV-negative tumor celllines after infection with dl331 andwild-type adenovirus. Cells were infectedwith Ad5, dl309, or dl331 at 100 particlesper cell as described in Materials andMethods. Twenty-four hours later,proteins were extracted, separated bygel-electrophoresis, and immunoblottedfor total PKR, phosphorylated PKR(P-PKR), phosphorylated eIF2-a(P-eIF2-a), and proliferating cell nuclearantigen (PCNA). Lane 1, uninfected-cells;lane 2, wild-type Ad5-infected cells;lane 3, dl309-infected cells; lane 4,dl331-infected cells. Data are fromduplicate experiments.





Figure 4. Antitumoral efficacy of VAI deletion mutant (dl331) in established C666-1 (EBV-positive) tumors in vivo . C666-1 human nasopharyngeal carcinomaxenografts implanted s.c. in C57BL/6 nu/nu mice as described in Materials and Methods. PBS or viruses (1 � 1010 particles) were injected intratumorallyon 5 consecutive days. Single injections were done for the biological time-points study. A, tumor growth curves; animals treated with dl331 (n = 7) andwild-type Ad5 (n = 6) showed significant delays in tumor growth compared with the control groups injected with PBS (n = 7) or the inactivated particles controlgroup (n = 7; P < 0.01). B, animal survival curves. dl331 and wild-type Ad5 treatment produced significantly higher survival rates than the control group(P < 0.001); the survival rate of dl331-treated mice was superior to Ad5. C, representative histologic changes and viral protein expression in C666-1 xenograftson day 5 after treatment with different adenoviruses, as indicated. H&E staining (original magnification, �200) shows that Ad5 and dl331 producesignificant cytopathic effects in tumor cells (i.e., cell death, eosinophilic cytoplasm, enlarged nuclei, and inclusion bodies). Control tumors show nosignificant change. Immunohistochemical analysis of such sections (original magnification, �200) shows that the remaining tumor cells stain positively forviral proteins E1A and hexon (brown ), whereas the PBS-treated tumors show no immunoreactivity; the PUVdl312-treated tumors show only weak nuclear stainingfor the hexon coat protein and no E1A staining.

Cancer Research




after 48 hours: There was no difference of E1A expressionbetween wild-type and VAI-deleted adenoviruses infection. Theviral gene hexon expression was consistent with the cytopathiceffect alteration. Figure 5 represents examples of histopathologyand adenovirus hexon protein staining in liver after viraltreatment.

Discussion

The presence and expression of the EBV genome undoubtedlycontributes to the development of many, if not all, tumorsassociated with this virus, opening up an unique opportunity todevelop novel anticancer treatments directed against viral, ratherthan cellular, targets. A number of therapies have been proposedto target EBV-associated tumors based on expression of viralgenes, as recently reviewed (46). Some examples are thoseinhibiting the expression of EBV-encoded oncogenes or inductionof EBV episome loss. Others involve the use of EBV-dependentpromoters and replication origins to express toxic genes,preferentially within EBV-positive tumors, the deliberate induc-tion of the lytic EBV replication in tumor cells, or variousimmune therapies targeting EBV viral proteins. In the presentstudy, we exploited the fact that EBERs are uniformly expressedin most human EBV-associated tumors, engineering a comple-mentary mutation in a replication-competent adenovirus toproduce a selective oncolytic viral therapy for these tumors.Indeed, a VAI-deleted mutant, dl331, showed selectivity for EBV-positive cells with the therapeutic index being especiallyadvantageous for these cells as opposed to normal primaryhuman cells. The data presented here show that this mutantcould be used to treat EBV-associated human tumors.VAI RNA–PKR-eIF-2a has been recognized as a major pathway

in host defense against viral infection and replication. Based oncommon functions exhibited by adenovirus VAI RNA and Ras for

inactivating PKR, it was recently reported that a VAI mutantadenovirus could be used selectively to treat tumors with anactivated ras oncogene (41). In the present study, VAI-deletedadenovirus showed selective cell killing and replication in EBV-positive tumor cells but not significant in a tumor cell line(Panc-1) with activated Ras pathway. Our results show that inthe EBV-positive tumor-derived cell line C666-1, dl331 induced ahigher level of phosphorylated-PKR but a minor increase level ofphosphorylated-eIF-2a when compared with wild-type adenovi-rus. These findings suggest that EBER may act through anotherpathway independent of PKR to affect host protein translation.It is known that the phosphorylation of eIF2a is mediated by

at least three different protein kinases: the IFN-inducible dsRNA-activated kinase PKR, the erythroid cell-specific, heme-regulatedkinase HCR (HRI) and the nutrient-regulated protein kinase GCN2(47). A recent report (48) showed that EBER-1 could enhanceprotein synthesis by a PKR-independent mechanism, stronglysupporting our data. In addition, SV40 antigen can rescue thetranslational defect in monkey and human cells that result frominfection of VAI-deleted adenovirus. However, the large T antigenof SV40 does not prevent the activation of PKR and thecomplementation appears to occur at a step downstream of PKRactivation, maybe by dephosphorylation of eIF-2a (38), and EBER1may act through a similar pathway. In the EBV-negative tumor celllines, Hep-2 and Panc-1, dl331 induced higher levels of phosphor-ylated-PKR and phosphorylated-eIF2a than wild-type adenovirusalthough Panc-1 has a mutant ras oncogene that could be expectedto inactivate PKR. These results are consistent with the poor cellkilling and viral replication observed in response to dl331 in thesecells. However, 72 hours after infection with dl331, viral replication inPanc-1 cells approached 20% of that of wild-type adenovirus,comparable with the observation in some EBV-positive cells. Thissuggests that partial complementation occurred in the cells but therewas very low cytotoxicity with dl331 in Panc-1 cells compared with

Figure 5. Histopathologic changes and viral protein expression in liver after adenovirus treatment. Livers were harvested at the same time points after treatmentwith different viruses and histopathologic changes and viral protein expression was examined as described in Materials and Methods. Top , representative of H&Estaining (original magnification, �200), 48 hours after i.v. injection of different viruses. Black arrows , cytopathic effects, showing cells with enlarged nuclei andinclusion bodies as well as eosinophilic cytoplasm. Bottom , representative of viral hexon protein staining (original magnification, �200) at 48 hours after virusinjection, with strong staining in Ad5-treated group.





the EBV-positive cells. The mechanism is not clear at this time.Of note, in the normal human primary cells tested, dl331 showedgood selectivity of viral replication compared with wild-typeadenovirus; however, no differences were observed in the level ofphosphorylated PKR and phosphorylated eIF2a after dl331 and Ad5treatments in these cells (data not shown). This suggests that VAIRNA apparently has other functions that control viral replication inthese cells. Indeed, VAI RNA of adenovirus, like EBER-1, can activatethe 2-5 (A) synthetases, which also affect viral replication (20, 31). Leiet al. (49) reported that adenovirus VAI RNA may affect viral andcellular gene expression by modulation of RNA editing byantagonizing the RNA-editing activity of the ADAP adenosinedeaminase. Further investigation will be required to explore themolecular mechanism for this.In the present study, one most interesting result is that dl331

showed superior cell killing in vitro and antitumoral efficacyin vivo compared with wild-type adenovirus, in EBV-positivecells, although the absolute level of viral replication of dl331 waslower than wild-type adenovirus. Our data indicate that EBV-positive C666-1 cells still have a relatively high level ofphosphorylated PKR after dl331 infection. PKR has beenreported to be indispensable for cell apoptosis by apoptoticstimulators such as E2F-1 and tumor necrosis factor-a (50, 51),and activation of PKR can induce the expression of Fas andtrigger apoptosis through the Fas-associated death domainprotein/caspase-8 death signaling pathway (52, 53). Thus, it isreasonable to postulate that higher PKR levels in dl331-infectedcells may promote adenovirus-induced cell death and this wouldexplain the lower level of replication in these cells. In addition,an increasing production of IFN-h in the absence of VAI RNAmay also explain why dl331 showed superior cell killing andantitumoral efficacy compared with wild-type adenovirus. VAIRNA also has similar functions to E3L of vaccinia virus. Deletionof the E3L has been shown to activate IFN regulatory factor 3(IRF3) and increase IFN-h production in cells infected byvaccinia virus lacking this gene (54). Of note, dl331 was derivedfrom dl327, which has a large deletion in the adenoviral E3region (25). This deletion may be another factor that affects theantitumoral efficacy because we have found that E3B deletion ofadenovirus can enhance potency compared with wild-typeadenovirus in vitro and in vivo in nude mice but not inimmunocompetent mouse tumor models (15).Although dl331 has superior antitumoral efficacy compared

with wild-type adenoviruses in vivo , the proportion of cases(2 of 7) showing complete regression of tumor after viral

treatment alone was still low. It would therefore be desirable todevelop other strategies that maintain selectivity but enhancethe efficacy of the adenovirus. Based on our findings (15, 55),deletion of the E1B 19K gene but retention of E3B seemspromising in this regard. In addition, it is conceivable thatcombination of adenovirus and conventional approach, such aschemotherapy, radiotherapy, and immunotherapy, will improveantitumor efficacy.Replication-selective, oncolytic adenoviruses have been evalu-

ated extensively in epithelial tumors but relatively little inlymphoid malignancies mainly owing to the perceived resistanceof the latter to adenovirus infection. However, in the presentstudy, we found that this resistance is not absolute and that,whereas most lines are refractory, adenovirus can infect someB-cell lymphoma lines, such as the Burkitt’s lymphoma derivedlines Raji and Jijoye, resulting in cytotoxicity to these cells. This isconsistent with observations described elsewhere using primarychronic lymphocytic leukemia and B-cell lymphoma lines (56).Taken together, our findings suggest that oncolytic adenovirus isalso a potential agent to treat at least a proportion oflymphomas. We found that the infectivity of Burkitt’s lymphomalines is associated with the expression of adenovirus receptors,such as coxsackievirus adenovirus receptor and heparan sulfateglycosaminoglycans, but that genetic alterations also play animportant role in the cytotoxicity of adenoviruses in lymphomacells (57); the mechanism behind the latter finding is notunderstood at present.In conclusion, our data suggest that a VAI-deleted adenovirus can

induce replication and cell lysis selectively in EBV-positive tumorcells. The replication-selective, oncolytic adenovirus showedefficient tumor killing in vitro and in vivo for EBV-positive tumorcells, and lower hepatotoxicity than wild-type adenovirus. Itrepresents a potential therapeutic agent for targeting human EBV-associated tumors, an increasingly important group of malignantdiseases.

Acknowledgments

Received 8/27/2004; revised 11/10/2004; accepted 11/21/2004.Grant support: Cancer Research United Kingdom.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Profs. Bayar Thimmappaya, Lindsey Hutt-Fletcher, and Dorly P. Huangfor the generous gifts of valuable materials; Gary Martin (Biological Resources Unitand the Central Cell Service of Cancer Research UK) for valuable support; andDrs. Iain McNeish and Georges Vassaux for helpful insights and reading of themanuscript.

References1. Kawa K. Epstein-Barr virus-associated diseases inhumans. Int J Hematol 2000;71:108–17.

2. Rickinson AB, Kieff E. Epstein Barr Virus. In: KnipeDM and Howley PM. Fields Virology, Vol. 2, Phila-delphia, PA: Lippincott Williams & Wilkins; 2001.pp. 2597–612.

3. Labrecque L, Barnes D, Fentiman I, Griffin BE.Epstein-Barr virus in epithelial cell tumors: a breastcancer study. Cancer Res 1995;55:39–45.

4. Bonnet M, Guinebretiere J, Kremmer E, et al.Detection of Epstein-Barr virus in invasive breastcancers. J Natl Cancer Inst 1999;91:1376–81.

5. Fina F, Romain S, Ouafik L, et al. Frequency andgenome load of Epstein-Barr virus in 509 breastcancers from different geographical areas. Br J Cancer2001;84:783–90.

6. Wong M, Chung L, Yuen S, et al. In situ detection ofEpstein-Barr virus in non-small cell lung carcinomas.J Pathol 1995;177:233–40.

7. Grinstein S, Preciado M, Gattuso P, et al. Demon-stration of Epstein-Barr virus in carcinomas of varioussites. Cancer Res 2002;62:4876–8.

8. Morbini P, Riboni R, Tomaselli S, Rossi A, Magrini U.EBER- and LMP-1-expressing pulmonary lymphoepi-thelioma-like carcinoma in a Caucasian patient. HumPathol 2003;34:623–5.

9. Huber M, Pavlova B, Muhlberger H, Hollaus P,Lintner F. Detection of the Epstein-Barr virus inprimary adenocarcinoma of the lung with Signet-ringcells. Virchows Arch 2002;44:25–30.

10. Sugawara Y, Makuuchi M, Takada K. Detection ofEpstein-Barr virus DNA in hepatocellular carcinomatissues from hepatitis C-positive patients. ScandJ Gastroenterol 2000;35:981–4.

11. Chen P, Pan C, Hsu W, Ka H, Yang A. Epstein-Barrvirus-associated lymphoepithelioma-like carcinoma ofthe esophagus. Hum Pathol 2003;34:407–11.

12. Hawkins L, Lemoine N, Kirn D. Oncolytic biothe-rapy: a novel therapeutic platform. Lancet Oncol2002;3: 17–26.

Cancer Research




13. Kirn D. Clinical research results with dl1520 (Onyx-015), a replication-selective adenovirus for the treat-ment of cancer: what have we learned? Gene Ther2001;8:89–98.

14. Kirn D. Replication-selective oncolytic adenovi-ruses: virotherapy aimed at genetic targets in cancer.Oncogene 2000;19:6660–9.

15. Wang Y, Hallden G, Hill R, et al. E3 genemanipulations affect oncolytic adenovirus activity inimmunocompetent tumor models. Nat Biotechnol2003;21:1328–35.

16. Earl P, Jones E, Moss B. Homology between DNApolymerases of poxviruses, herpesviruses, and adeno-viruses: nucleotide sequence of the vaccinia virus DNApolymerase gene. Proc Natl Acad Sci U S A 1986;83:3659–63.

17. Parker G, Crook T, Bain M, Sara E, Farrell P, AlldayM. Epstein-Barr virus nuclear antigen (EBNA)3C is animmortalizing oncoprotein with similar properties toadenovirus E1A and papillomavirus E7. Oncogene1996;13:2541–9.

18. Horvath J, Faxing C, Weber J. Complementation ofadenovirus early region 1a and 2a mutants by Epstein-Barr virus immortalized lymphoblastoid cell lines.Virology 1991;184:141–8.

19. Sharp T, Schwemmle M, Jeffrey I, et al. Comparativeanalysis of the regulation of the interferon-inducibleprotein kinase PKR by Epstein-Barr virus RNAs EBER-1 and EBER-2 and adenovirus VAI RNA. Nucleic AcidsRes 1993;21:4483–90.

20. Sharp TV, Raine DA, Gewert DR, Joshi B, Jagus R,Clemens M. Activation of the interferon-inducible (2V-5V) oligoadenylate synthetase by the Epstein-Barr virusRNA, EBER-1. Virology 1999;257:303–13.

21. Rechel PA, Merrick WC, Siekierka J, Mathews M.Regulation of a protein synthesis initiation factor byadenovirus virus-associated RNA1. Nature 1985;313:196–200.

22. Soderlund H, Pettersson U, Vennstrom B,Philipson L, Mathews M. A new species of virus codedlow molecular weight RNA from cells infected withadenovirus type2. Cell 1976;7:585–93.

23. Mathews M. Genes for VA-RNA in adenovirus 2. Cell1975;6:223–9.

24. Schneider RJ, Weinberger C, Shenk T. AdenovirusVAI RNA facilitates the initiation of translation invirus-infected cells. Cell 1984;37:291–8.

25. Thimmappaya B, Weinberger C, Schneider RJ,Shenk T. Adenovirus VAI RNA is required for efficienttranslation of viral mRNAs at late times after infection.Cell 1982;31:543–51.

26. Katze M, Decarato D, Sfer B, Galabru J,Hovanessian G. Adenovirus VAI RNA complexes with68,000 Mr protein kinase to regulate its phosphoryla-tion and activity. EMBO J 1987;6:689–98.

27. Kitajewski JR, Schneider RJ, Safer B, et al.Adenovirus VAI RNA antagonizes the antiviralaction of interferon by preventing activation ofthe interferon-induced eIF-2a kinase. Cell 1986;45:195–200.

28. Schneider RJ, Safer B, Munemitsu SM, Samuel CE,Shenk T. Adenovirus VAI RNA prevents phosphoryla-tion of the eukaryotic initiation factor 2a subunitsubsequent to infection. Proc Natl Acad Sci U S A 1985;82:4321–4.

29. Siekierka J, Mariano T, Reichel PA, Mathews M.Translational control by adenovirus: lack of virusassociated RNAI during adenovirus infection results inphosphorylation of initiation factor eIF-2 and inhibi-tion of protein synthesis. Proc Natl Acad Sci U S A1985;82:1959–63.

30. Rosa M, Gottlieb E, Lerner MR, Steitz J. Strikingsimilarities are exhibited by two small Epstein-Barrvirus-encoded ribonucleic acids and the adenovirus-associated ribonucleic acids VAI and VAII. Mol CellBiol 1981;1:785–96.

31. Desai SY, Patel RC, Sen GC, Malhotra P, Ghadge GD,Thimmapaya B. Activation of interferon-inducible 2V-5V oligoadenylate synthetase by adenoviral VAI RNA.J Biol Chem 1995;270:3454–61.

32. Bhat RA, Thimmappaya B. Two small RNAsencoded by Epstein-Barr virus can functionallysubstitute for the virus-associated RNAs in the lyticgrowth of adenovirus 5. Proc Natl Acad Sci U S A1983;80:4789–93.

33. Bhat RA, Thimmappaya B. Construction andanalysis of additional adenovirus substitution mutantsconfirm the complementation of VAI RNA function bytwo small RNAs encoded by Epstein-Barr virus. J Virol1985;56:750–6.

34. Cheung S, Huang D, Hui A, et al. Nasopharyngealcarcinoma cell line (C666-1) consistently harbouringEpstein-Barr virus. Int J Cancer 1999;83:121–6.

35. Pilder S, Logan J, Shenk T. Deletion of the geneencoding the adenovirus 5 early region 1b 21,000-molecular-weight polypeptide leads to degradation ofviral and host cell DNA. J Virol 1984;52:664–71.

36. Hallden G, Hill R, Wang Y, et al. Novel immuno-competent murine tumor models for the assessmentof replication-competent oncolytic adenovirus effica-cy. Mol Ther 2003;8:412–24.

37. Kitajewski J, Schneider RJ, Safer B, Shenk T. Anadenovirus mutant unable to express VAI RNAdisplays different growth responses and sensitivity tointerferon in various host cell lines. Mol Cell Biol1986;6:4493–8.

38. Rajan P, Swaminathan S, Zhu J, et al. A noveltranslational regulation function for the simian virus40 large-T antigen gene. J Virol 1995;69:785–95.

39. Jakubczak JL, Ryan P, Gorziglia M, et al. An oncolyticadenovirus selective for retinoblastoma tumor sup-pressor protein pathway-defective tumors: depen-dence on E1A, the E2F-1 promoter, and viralreplication for selectivity and efficacy. Cancer Res2003;63:1490–9.

40. Xue S, Lu Q, Poulsom R, Karran L, Jones M,Griffin BE. Expression of two related viral early genesin Epstein-Barr virus-associated tumors. J Virol2000;74:2793–803.

41. Cascallo M, Capella G, Mazo A, Alemany R. Ras-dependent oncolysis with an adenovirus VAI mutant.Cancer Res 2003;63:5544–50.

42. Meurs E, Chong K, Glabru J, et al. Molecular cloningand characterization of the human double-strandedRNA-activated protein kinase induced by interferon.Cell 1990;62:379–90.

43. Clark P, Schwemmle M, Schickinger J, Hilse K,Clemens MJ. Binding of Epstein-Barr virus small RNAEBER-1 to the double-stranded RNA-activated proteinkinase DAI. Nucleic Acids Res 1991;19:243–8.

44. Vile R, Ando D, Kirn D. The oncolytic virotherapytreatment platform for cancer: unique biological andbiosafety points to consider. Cancer Gene Ther2002;9:1062–7.

45. Johnson L, Shen A, Boyle L, et al. Selectivelyreplicating adenoviruses targeting deregulated E2Factivity are potent, systemic antitumor agents. CancerCell 2002;1:325–37.

46. Israel B, Kenney S. Virally targeted therapies forEBV-associated malignancies. Oncogene 2003;22:5122–30.

47. Clemens MJ. Protein kinase that phosphorylatedeIF-2 and eIF2B, and their role in eukaryotic celltranslational control. In: Hershe J, Mathews M,Sonenber N, editors. Translational control. Cold SpringHarbor (NY): Cold Spring Harbor Laboratory Press;1996. p. 139–72.

48. Laing K, Elia A, Jefre I, et al. In vivo effects of theEpstein-Barr virus small RNA EBER-1 on proteinsynthesis and cell growth regulation. Virology2002;297:253–69.

49. Lei M, Liu Y, Samuel C. Adenovirus VAI RNAantagonizes the RNA-editing activity of theADAP adenosine deaminase. Virology 1998;245:188–96.

50. Vorburger SA, Pataer A, Yoshida K, et al. Role forthe double-stranded RNA activated protein kinasePKR in E2F-1-induced apoptosis. Oncogene 2002;21:6278–88.

51. Jeffrey IW, Bushell M, Tilleray V, Morley S,Clemens M. Inhibition of protein synthesis inapoptosis: differential requirements by the tumornecrosis factor-a family and a DNA-damagingagent for caspases and the double-stranded RNA-dependent protein kinase. Cancer Res 2002;62:2272–80.

52. Balachandran S, Kim CN, Yeh WC, Mak TW,Bhalla KN, Barber GN. Activation of the dsRNA-dependent protein kinase, PKR, induces apoptosisthrough FADD-mediated death signaling. EMBO J1998;17:6888–902.

53. Donze O, Dostie J, Sonenberg N. Regulatableexpression of the interferon-induced double-strandedRNA dependent protein kinase PKR induces apo-ptosis and Fas receptor expression. Virology 1999;256:322–9.

54. Xiang Y, Condit R, Vijaysri S, Jacobs B,Williams B, Silverman R. Blockade of interferoninduction and action by the E3L double-strandedRNA binding proteins of vaccinia virus. J Virol 2002;76:5251–9.

55. Liu T, Hallden G, Wang Y, et al. An E1B-19 kDagene deletion mutant adenovirus demonstratestumor necrosis factor-enhanced cancer selectivityand enhanced oncolytic potency. Mol Ther 2004;9:786–803.

56. Medina D, Sheay W, Goodell L, et al. Adenovirus-mediated cytotoxicity of chronic lymphocytic leuke-mia cells. Blood 1999;94:3499–508.

57. Wang Y, Tan P, Francis J, et al. Anti-CD22immunolipoplexes enhance transduction of adenovi-rus in B lymphoma cells independent of adenovirusreceptor but do not enhance the cytotoxicity ofoncolytic adenovirus. Mol Ther 2004;9:S171.





2005;65:1523-1531. Cancer Res Yaohe Wang, Shao-An Xue, Gunnel Hallden, et al. Oncolytic Agent Targeting EBV-Associated Tumors

Deleted Adenovirus, a Potential−Virus-Associated RNA I

Updated version

http://cancerres.aacrjournals.org/content/65/4/1523

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/65/4/1523.full.html#ref-list-1

This article cites 54 articles, 22 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/65/4/1523.full.html#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] at

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

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications


http://cancerres.aacrjournals.org/content/65/4/1523

http://cancerres.aacrjournals.org/content/65/4/1523.full.html#ref-list-1

http://cancerres.aacrjournals.org/content/65/4/1523.full.html#related-urls

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

mailto:[email protected]

mailto:[email protected]


Documents

Virus-Associated RNA I-Deleted Adenovirus, a Potential Oncolytic Agent Targeting EBV-Associated Tumors