VSV Selectively Destroys Tumors

7/28/2019 VSV Selectively Destroys Tumors

1/12

Neurobiology of Disease

Systemic Vesicular Stomatitis Virus Selectively Destroys

Multifocal Glioma and Metastatic Carcinoma in Brain

Koray Ozduman, Guido Wollmann, Joseph M. Piepmeier, and Anthony N. van den PolDepartment of Neurosurgery, Yale University School of Medicine, New Haven, Connecticut 06520

Metastatic tumorsand malignantgliomas make up themajorityof cancers in thebrain. They areinvariably fatal andthereis currentlyno

cure. From in vitro comparisons of a number of viruses, we selected one that appeared the best in selectively killing glioblastoma cells.This replication-competent virus, the glioma-adapted vesicular stomatis virus strain VSVrp30a, was used for in vivo tests with theunderlying view that infection of tumor cells will lead to an increase in the number of viruses subsequently released to kill additional

tumor cells. Intravenous injection of VSVrp30a expressing a greenfluorescent protein reporter, rapidly targeted and destroyed multipletypes of human and mouse tumors implanted in the mouse brain, including glioblastoma and mammary tumors. When tumors wereimplanted both in the brain and peripherally, emulating systemic cancer metastasis, tumors inside and outside the brain were simulta-

neously infected. Intranasal inoculation, leading to olfactory nerve transport of the virus into the brain, selectively infected and killedolfactory bulb tumors. Neither control cortical wounds nor transplanted normal mouse or human cells were targeted, indicating viral

tumor selectivity. Control viruses, including pseudorabies, adeno-associated, or replication-deficient VSV, did not infect the braintumor.Confocal laser time-lapse imaging through a cranial window showed that intravenous VSVinfects the tumor at multiple sites and

kills migrating tumor cells. Disrupted tumor vasculature, suggested by dye leakage, may be the port of entry for intravenously deliveredVSV. Quantitative PCR analysis of how VSVrp30a selectively infected tumor cells suggested multiple mechanisms, including cell surface

binding and internalization.

Key words: brain cancer; glioma; metastasis; vesicular stomatitis virus; oncolytic virus; fluorescence microscopy

IntroductionCancer in the brain is one of the deadliest diseases. A commonbrain cancer is glioblastoma, which is a primary cancer of thebrain; tumor cells can migrate from the body of a tumor intomultiple brain systems. Metastases, which migrate into the brainfrom cancer originating outside the CNS, can colonize multiplesites within the brain. Each year 200,000 people in the UnitedStates are diagnosed with a brain tumor (Central Brain TumorRegistry of the United States, 2005). The malignant disease usu-ally startswith subtlechanges in cerebral function,becomes man-ifest with seizures or raised intracranial pressure, progresses toloss of motor and mental function, and finally to death. There is

no cure or effective treatment, and the median survival is mea-sured in months despite modern therapy (Posner and Chernik,1978; DeAngelis, 2001; Ohgaki and Kleihues, 2005; Stupp et al.,2005). Failure of conventional treatment modalities such as sur-gery, radiation and chemotherapy results from multifocality ofdisease, infiltrative growth within otherwise normal brain tissue,impermeability of the CNS to therapeutic agents, substantial tu-

mor genotypic variability, persistence of cancer-initiating cells,

and rapid genetic evolution of resistant clones. Cancer is a sys-temic disease, and multiple metastatic lesions or large infiltrative

gliomas require an efficient delivery method to target dispersed

tumor cells.A unique approach to combating brain cancer is the use of

viruses (Aghi and Rabkin, 2005). Nonreplicating viruses are used

to deliver toxic or therapeutic genes.Such viral vectors have so farbeen only partially effective because only a small percentage of

tumor cells are infected (Puumalainen et al., 1998; Rainov andRen, 2003). An alternate strategy is the use of replicating viruses

that target, infect, produce progeny virus to infect more tumor

cells, and kill infected cells. Some viruses that have been engi-neered to kill dividing tumor cells, such as recombinant herpes

simplex or adenovirus, have shown promise, but these viruses

have not been able to disseminate to all regions of the tumor(Khuri et al., 2000; Pulkkanen and Yla-Herttuala, 2005). Other

viruses, including reovirus, measles, or pox viruses, show meritfor infecting tumor cells (Parato et al., 2005). In a previous series

of in vitro experiments on nine different viruses with reported

intrinsic oncolytic potential against gliomas, a rhabdovirus, ve-sicular stomatis virus VSVrp30a, outperformed other candidates

(Wollmann et al., 2005). It was also superior in infecting gliomaxenografts in a subcutaneous glioma model. VSV also has potent

oncolytic activity against a broad spectrum of systemic cancercells tested in vitro (Stojdl et al., 2003). To clearly analyze viralinfection and spread in brain tumors, we used red fluorescent

Received June 14, 2007; revised Jan. 2, 2008; accepted Jan. 3, 2008.

This work was supported by National Institutes of Health Grants A1/NS48854 and C A124737 (A.N.v.d.P.). K.O.

wassupported byan AnnaFullerMolecularOncologyFellowship.We thankDr. P.K. Ghoshfor expertisein generat-

ingfluorescent tumor cells, Dr.D. D. Spencerfor supplyinghumanbraintissue,Y. Yangand V.Rogulinfor excellent

technical facilitation, and Dr. J. Rose for helpful suggestions.

Correspondence should be addressed to Anthony N. van den Pol, Department of Neurosurgery, Yale University,

School of Medicine, 333 Cedar Street, New Haven, CT 06520. E-mail: [email protected]:10.1523/JNEUROSCI.4905-07.2008

Copyright 2008 Society for Neuroscience 0270-6474/08/281882-12$15.00/0

1882 The Journal of Neuroscience, February 20, 2008 28(8):18821893


2/12

protein (RFP)-expressing cancer cells and a recombinant VSVthat expresses green fluorescent protein (GFP). A battery of ge-neticallydistinct human cancercell lines andseveral mouse xeno-graft models allowed us to study different aspects of VSV oncol-

ysis. A single intravenous dose of VSV simultaneously andselectively targeted multifocal brain tumors and subcutaneousflank tumors. Time-lapse confocal laser imaging of VSV oncoly-

sis in the live mouse brain suggested that infection spread withinthe tumor, probably enhanced by a local leaky bloodbrain bar-rier (BBB).Onco-selectivity of VSVcan be attributed in part to itshigh sensitivity to interferon antiviral defense, which allows nor-mal cells to attenuate infection but appears to be defective in themajority of tumors tested (Stojdl et al., 2000, 2003). In addition,we show with quantitative reverse transcription (RT)-PCR ex-periments that multiple additional mechanisms, including moreefficient virus binding and internalization, enhance preferentialtargeting of tumor cells by VSVrp30a.

Materials and MethodsCell culture. U87MG human high-grade glioma cells and 4T1 mouse

mammary carcinoma cells were obtained from American Type CultureCollection (Rockville, MD). A549 and Calu-1 human lung carcinomasand T-47D, MCF-7, and BT474 human breast carcinoma cells were pro-vided by the Yale Cancer Center (Yale University, New Haven, CT).U118MG glioma cells were kindly provided by Dr. R. Matthews (YaleUniversity). Primary human glioblastoma cultures wereestablished fromtissue derived from two patients undergoing resective surgery. Nontu-mor primary human astrocytes were prepared as explant cultures fromhuman temporal lobectomy material, removed for intractable epilepsy,solely for the benefit of the patient, and the use was approved by the YaleUniversityHuman Investigation Committee. Astrocyte identity was con-firmed by positive immunofluorescence for human glial fibrillary acidicprotein. Cells were maintained in Minimal Essential Medium (Invitro-gen, Carlsbad, CA) supplemented with 10% fetal bovine serum, 1% so-dium pyruvate, 100 M nonessential amino acids, and 25 mM HEPES

buffer. For VSV susceptibilityexperiments, 2 105

cells wereculturedinsix-well plates and infected with VSVrp30a at a multiplicity of infection(MOI) of 1. The total number of viable and infected cells was counted at12 h intervals in 15 high-magnification microscopic fields.

Transfection. For stable transfection, RFP [tetrameric phosphorylatedDiscosoma red (pDsRed1)N1 and monomeric pDsRed-monomerC1], cyan fluorescent protein [phosphorylated cytomegalovirus (pCM-V)CFP], and blue fluorescent protein (pCMVBFP) plasmids wereused. pCMVCFP and pCMVBFP were generated by exchanging theRFP expression cassette from pDsRed monomerC1 with CFP or BFPcassettes from pRSET/CFP (Invitrogen). Cells were transfected usingLipofectamine 2000reagent (Invitrogen), enriched,and maintained withG418 (Sigma, St. Louis, MO). rU87 and rU118 cells were sorted forbrightest fluorescence on a FACSvantage SE fluorescence-activated cellsorter (BD Biosciences, San Jose, CA). Growth characteristics were as-sessed using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bro-mide (MTT) (Invitrogen)assayaccording to theinstructions of theman-ufacturer. Optical density was read at 570 nm using a Dynatech(Alexandria, VA) MR500 ELISA plate reader and corrected from back-ground control.

Mouse procedures. Animal experiments and postoperative care wereperformed in accordance with institutional guidelines of the Yale Uni-versity Animal Careand Use Committee. Immunodeficient homozygousCB17-SCID (CB17SC-M) and NCr-Nude mice (NCr-Foxn1nu) at anage of 45 weeks were obtained from Taconic Farms (Germantown,NY). Animals were stereotactically grafted with 3 105 tumor cellsbilaterally into the striatum (2 mm lateral, 0.4 mm rostral to bregma at 3mmdepth) orcortexor with1 105 cells unilaterally in the leftolfactorybulb (0.75 mm lateral, 4.5 mm rostral to bregma at 1.5 mm depth). For

cranial window experiments, a teardrop-shaped skin flap (with an an-teroposterior axis along the parasagittal plane crossing the midpupillaryline) was removed, and a 5-mm-wide round craniectomy was drilled

under an operating microscope (Zeiss, Oberkochen, Germany). rU87glioma cells, 5 104, were injected along a trajectory parallel to thecortex immediately below the pia mater. The craniotomy was coveredwith a round glass coverslip of 8 mm diameter and glued to the bonemargins usingcyanoacrylate. Animals were followed with daily measure-ments of weight, food and water consumption, and overall health andactivity, and the cranial window animals were given parenteral antibiot-ics. During in vivo brain imaging, mice were lightly anesthetized to re-

duce movement, and the head was maintained in a horizontal positionduring image acquisition.

Animals were killed with a pentobarbital overdose and perfused tran-scardially with4% paraformaldehyde. For determination of blood brainbarrier integrity, 2% Evans Blue (Sigma) in 100 l of sterile saline wasinjected intravenously 1 h before the animals were killed.

Viruses. Ten days after tumor grafting, 100 l of medium containingvirus (107 plaque forming units (PFU) of VSVrp30a, VSVGFP, VSVGGFP, or recombinant pseudorabies virus (PRV)GFP or 10 12 viri-ons adeno-associated virusserotype2 (AAV)GFP]was injected throughthe tail vein. VSVGFP and VSVGGFP were generously provided byDr. J. K. Rose (Yale University). PRVGFP strain 152 was kindly pro-vided by Dr. L. W. Enquist (Princeton University, Princeton, NJ) (Smithet al., 2000) andAAVGFP waskindly provided by Dr.K. R. Clark (OhioState University, Columbus, OH). Details on the generation and charac-teristics of the tumor-adapted VSVGFP isolate VSVrp30a were de-scribed previously (Wollmann et al., 2005). VSV, PRV, andAAV were allshown to infect glioma cells previously (Wollmann et al., 2005).

Immunocytochemistry. Cryosections were mounted and counter-stained with 4,6-diamidino-2-phenylindole (DAPI) (Invitrogen). Forcleaved caspase 3 immunocytochemistry, 20m sections were incubatedwith a 1:150 dilution of polyclonal rabbit cleaved caspase 3 antiserum(Cell Signaling Technology, Danvers, MA), followed by detection with a1:200 dilution of biotinylated anti-rabbit antibody (Invitrogen) and ABCreagents detected with diaminobenzidine and hydrogen peroxide (Vec-tor Laboratories, Burlingame, CA). For von Willebrand factor staining, a1:500 dilution of polyclonal rabbit antibody (Abcam, Cambridge, MA)was used following the same procedure outlined above.

Imaging. Histologicalsections were studied eitheron an Olympus Op-

tical (Tokyo, Japan) IX71 fluorescent microscope fitted with a SPOT-RTcamera (Diagnostic Instruments, Sterling Heights, MI) or on an Olym-pus Optical Fluo-300 laser confocal microscope using Fluoview 300 soft-ware version 4.3. Corresponding phase-contrast, green, red, and bluefluorescence photomicrographs were fused and corrected for color andcontrast using Adobe Photoshop 7 software (Adobe Systems, San Jose,CA). Several high-resolution pictures were used for montage of wholecoronal sections.

Tumor area calculations and cell and nuclear counting were doneusing NIH ImageJ software. Animals with cranial windows were imageddaily or more frequently under an Olympus Optical SZXZB12 stereomicroscope and Olympus Optical Fluo-300 laser confocal microscope.

Virus binding and entry experiments. Binding, entry, and uncoating ofVSVrp30a were compared in rU87 glioma cells and normal human as-trocytes using quantitative RT-PCR. To assess membrane-bound frac-tion, cells were incubated at 4C with VSVrp30a for 20 min and washedfive times before RNA extraction. To quantify intracellular virus, cellswere trypsinized for 10 min and washed five times in PBS. VSV waspooled in the endosomal compartment by incubating for 90 min in thepresence of 5 mM ammonium chloride (Sigma). Ammonium chlorideblocks acidificationof the endosome and thereforeblocks viraluncoatinginto the cytoplasm. These experiments were done in triplicate using anMOI of 10.

Real-Time PCR for viral GFP. RNA was extracted from cell lysatesusing RNeasy kit (Qiagen, Valencia, CA), and 1 g of total RNA wasreverse transcribed by random hexamer priming using the Super ScriptIII reverse transcriptase kit (Invitrogen). This was followed by quantita-tive PCR using TaqMan gene expression assays (Applied Biosystems,Foster City, CA) for GFP and human glyceraldehyde-3-phosphate dehy-

drogenase (GAPDH). Primer sequences for GFP were as follows: sense,5-GAGCGCACCATCTTCTTCAAG-3; antisense, 5-TGTCGCC-CTCGAACTTCAG-3.

Ozduman et al. VSV Targets Brain Cancer J. Neurosci., February 20, 2008 28(8):18821893 1883


3/12

ResultsTumor cells stably transfected with redfluorescent protein gene reliablyform tumorsTo choose the optimum nontoxic fluores-cent protein that would allow simulta-neous imaging with GFP encoding virus,

we tested several different fluorescent pro-teins, including cyan (pCMVCFP), blue(pCMVBFP), and red (tetrameric pD-sRed1N1 and monomeric pDsRed-monomerC1). BFP was subject to quickphotobleaching, and CFP was not practicalfor use simultaneously with GFP-expressingVSV because of overlap in their fluorescentspectrum. Transfection of U87 high-gradeglioma cells with tetrameric RFP resulted intoxicity in vitro. Stable transfection ofU87MGor U118MG glioma cells witha plas-midencoding themonomeric DsRed protein

resulted in bright red fluorescence and didnot alter growth characteristics as demon-strated by MTT assay. Over a 24 h period,absorbance measurements increased by89.8 33.2% in rU87 cells and by 103.8 32.5% in U87MG cells (statistically nonsig-nificant p 0.75, n 6, Students t test),indicating that the red glioma cells were rep-licating at their usual rate. After selection inG418 and fluorescence-activated cell sorting,all tumor cells expressed red fluorescent pro-tein (Fig. 1A).

Bilateral brain injections of either rU87or rU118 cells in SCID or nude mice

yielded tumors in 100% (90 of 90) of injec-tions. Ten days after injection, noninfectedrU87 xenografts in SCID mice had a meancalculated volume of 7.13 5.76 mm3

(n 4; calculated as ellipsoid volume: 4/3x y z). Horizontal and vertical radii xand

yrepresent half of the horizontal and verti-cal diameters in the coronal section thatharbored the largest tumor cross-sectionarea. The anteroposterior radius z repre-sents half the tumor craniocaudal length,calculated by multiplying section thicknesswith the number of sections where the tu-

mor is present. Tumor xenografts grewpredominantly by expansion (Fig. 1C), al-though in a number of cases we found tu-mor cells separated from the main tumorbody and invading the surrounding paren-chyma. Tumors grew with no spontaneousregression or hemorrhage until theyreached a lethal size (Fig. 1C). For humanereasons, mice were killedat the onset of symptoms before tumor-induced death.

Intravenously injected VSVrp30a targets multifocal braintumors with high selectivity

In a previous study, we showed that, in animals bearing twosubcutaneous tumors, intratumoral VSV injection into one tu-morresultedin selectiveinfection also in thesecond tumor on the

contralateral side of the body. This strongly suggested vascularviral dissemination and led us to the hypothesis that VSV couldbe delivered to brain tumors through the intravascular route.However, entry into brain tumors is more complex than else-where in thebody because theblood brain barrier canbe a major

obstacle, even for relatively small-sized chemical or biologicalagents (Pardridge, 2002). We tested whether intravenously ad-ministered VSV could target intracranial tumor xenografts. In

Figure 1. Intravenously injected VSVrp30a infects intracranial xenografts at multiple sites and spreads within the tumor toinfect thetumor massand infiltrating tumorcells. Stably RFP-transfectedand fluorescence-activatedcell-sorted U87MGhumanhigh-gradegliomacellsgrew in vitro withoutanysignsofcytotoxicity(A) andformedbraintumorswithno signsof spontaneousregression (B), eventually resulting in the animals demise (C). When tumor-bearing animals were given a single intravenousinjectionof VSVrp30a,multipleindependent foci(arrows inD)ofVSVinfectionwerefoundinsidethetumormass.Tumorcellsininfected areas demonstrated marked cytopathy, and most were lysed (small arrow in E, F). Viral progeny released from these

initiallyinfectedcells spreadto neighboringtumor cells(large arrowin E, F) creatingdistinctareasof cellsin differentphasesofVSV infection. In addition to infecting the whole tumor, VSVrp30a also infected individual infiltrative tumor cells close to thetumor mass (arrows in H) as well as in isolated clusters remote from the main tumor bulk (IK). Despite widespread infectionwithin the tumor mass (green fluorescent area in Gmarked with arrows), no infection was noted in the surrounding brainparenchyma. LNschematize the general outline of VSV oncolysis in the SCID mouse brain tumor model: After intravenousinjection(L),VSVrp30ainfectionstarts atfew scatteredfoci(M) aroundbraintumor vasculatureand spreads toinfect thewholetumor (N). Scale bars:A, B, 50m; C, 500m; D, 300m; E, F, HK, 20m; G, 100m.

1884 J. Neurosci., February 20, 2008 28(8):18821893 Ozduman et al. VSV Targets Brain Cancer


4/12

mice that were implanted with bilateral brain rU87 xenografts(n 9 animals, 18 tumors), intravenous injection of 10 7 PFU ofVSVrp30a through the tail vein consistently resulted in infectionof the intracranial tumors. A single injection was sufficient toinfectgliomas.After an injectionof 107 PFU, VSVrp30a infectionwas found in different regions inside the tumor, suggesting mul-tiple sites of virus seeding (Fig. 1 D). Infection inside the tumor

started with a time lag after intravenous delivery but was found inall tumors analyzed after 24 h.

VSVrp30a replicates and spreads within the tumor massVSV is a replication-competent virus and spreads within the tu-mor. Early in thecourse of infectionin thenon-infected areas,redtumor cells with intact cellular outlines were seen. Within theVSV-infected area, infected and dead tumor cells and debris (Fig.1 E, F, small arrows) were surrounded by VSV-infected cells withmembranous GFP staining (Fig. 1E, F, large arrows). The pres-ence of newly infected cells at the boundary between killed andnon-infected cells suggests that VSV spreads within the tumor.Analyzing animals at different time points within the course ofVSV infection and using serial histological sections through thetumor, we found that the virus had spread from initially infectedfoci to encompass the whole tumor (Fig. 1G). In addition to thewidespread infection and tumor cell killing inside the main tu-mor mass, tumor cells that were migrating into the normal brainat the tumor margin were also infected, with remarkable sparingof neighboring normal brain cells in immediate contact with thetumor (Fig. 1H). Although most tumor xenografts grew by ex-pansion, some isolated cell clusters of tumor cells remote fromthe main tumor bulk were observed. Of note, these distant cellconglomerates were also targeted by VSV (Fig. 1IK). Thecourseof VSV oncolysis in the brain is schematized in Figure 1 LN.

To document viral spread within the tumor mass, we killedanimals 24, 48, and 72 h post-inoculation (HPI) and demon-

strated a continuously increasing area of infection in the tumormass, resulting in tumor cell death, as indicated by the loss ofcellular integrity and the appearance of small granular debrisafter viral infection (Fig. 2). Six tumors were used for analysis ofVSV and four for VSVrp30a infection at each time point. Inanimals killed early in the course of infection, we documentedareas of infected and killed tumor cells surrounded by other cellsthat appeared live at the time of death but with early signs ofinfection and expression of viral reporter gene (Fig. 2). Intactcellular outlines and homogeneous intracellular DsRed signalwere used to distinguish healthy cells from dead or dying cells. Inaddition, DAPI counterstain showed nuclear breakdown com-monly found in dead cells, and areas of dead cells showed small

granular debris in the place of intact cells (see additional detailsbelow and Fig. 4). The mean infected area in coronal brain sec-tions with thelargest tumor diameter increasedfrom50to60%at 1 d post-inoculation (DPI) to what appeared to be infection ofalmost the entire tumor by 3 DPI (Fig. 2). As shown in the frac-tionated bar graph in Figure 2A, the proportion of viable tumorcells showing normal cellular and nuclear morphology decreasedsimultaneously. Finally, the spread of infection was documentedin real time using the cranial window animal model as describedbelow.

In vitro, VSVrp30a was more effective at selective destructionof gliomas than the parental VSVGFP. To determine whetherthis difference persisted in vivo, we compared the relative abilities

of the two VSVs to infect gliomas in the mouse brain using thesame concentration of virus inoculum. Both VSVrp30a (n 9animals, 18 tumors) and VSVGFP (n 9 animals, 18 tumors)

successfully targeted the brain tumors after tail vein inoculationVSVGFP. A high tumor selectivity of VSV infection was ob-served during the time period of our study. In animals implanted

with multiple tumors, all brain tumors were infected simulta-neously (Fig. 3). VSV infection and resultant oncolysis were ob-served in all tumor masses regardless of the size and location.

Figure 2. The whole tumor mass is infected within 72 h. The spread of VSVrp30a infectionencompassesthewholetumormasswithinafewdays. A showsthetimeprogressionoftheVSVoncolysis. n6forVSV; n4forVSVrp30aateachtimepoint.Theinfectedareaincreasesandthe number of viable cells decreases. Black indicates the area of tumor lysis with little GFPdetection because of GFP diffusion out of dead cells. VSVrp30a has a slightly higher rate ofspread withinthe tumor than VSVGFP (not statistically significant)at 2 and3 d afterinocula-tion.BGshowthreerepresentativerU87xenograftsanalyzed24(B, C),48(D,E),and72(F,G)h after infection respectively. Scale bars: B, 100m; C, E, G, 50m;D, F, 300m.



5/12

Small andlargetumors were infected at dif-ferent localizations in the brain whetherstriatal, cortical, subventricular, or withinthe olfactorybulb.Tumors in thebrain andperiphery were also simultaneously in-fected (see below). In animals killed withinthe first 72 HPI, when the entire tumor

mass was infected by VSVrp30a, we didnotobserve GFP expression in thebrain paren-chyma surrounding the tumor, choroidplexus, or in the leptomeninges. After thisperiod of high tumor-specific infection(7296 HPI in our model), when the wholetumor bulk showed marked cytopathy, wedid observe select periventricular groupsofinfected neurons and ependymal cells inimmunodeficient mice.

VSVrp30a infection causes tumor cell deathVSV infection results in death of tumorcells in vitro. Similarly, VSVrp30a induced

widespread tumor cell death in the brain invivo. Such a fast and efficienteliminationofthe tumor is important because remainingresistant cells have the potential of repopu-lating the tumor mass, which is commonlyobserved in gliomas after conventionaltreatment. Widespread cell death was de-tected within the tumor mass. Cell deathwas indicated by loss of cellular outlines onconfocal and light microscopy. Differentialinterference contrast (DIC) microscopyshowed loss of cellular outlines and diffu-sion of intracellular tumor cell DsRed with

fluorescent imaging and nuclear fragmen-tation assessed with DAPI staining (Fig. 4).Infected tumors stained positively for thecellular apoptosis execution protein acti-vated caspase 3 (Fig. 4 H). There was a sta-tistically significant decrease in the numberof cells with intact cellular ( p 0.0001,ANOVA test) and nuclear morphology( p 0.0001, ANOVA test) and a statisti-cally significant increase in the number oftumorcells immunostained positive for ac-tivated caspase 3 ( p 0.0001, ANOVAtest) after infection (Fig. 4 IK). Infectedtumor cells underwent lysis several hoursafter expression of viral proteins. An im-portant observation was that the mouseneovasculature and its cellular constitu-ents, including endothelial cells, were notinfected, as shown by the intact vessel ar-chitecture in DIC images (Fig. 4M)andthepositive staining for the endothelial marker von Willebrand fac-tor of these non-infected cells (Fig. 4N, O). After widespread tu-mor cell lysis, both viral GFP and tumor DsRed intensity de-creased in the tumor mass, which can be attributed to loss ofcellular integrity and resultant diffusion of intracellular content.

Oncolytic potential of VSVrp30a generalizes to human glioma and

mammary carcinomaVSVrp30a was adapted to human gliomas by passaging it onU87MG cells for manygenerations.Therefore, to testthe hypothesis

that the in vivo oncolytic potential of VSVrp30a was not restricted to

the cell line on which it was developed, we also used two other types

of tumor implanted into the mouse brain. One of the cell types was

U118MG, a cell line derived from high-grade human glioma but

with different genetic anomalies than found in U87MG. The two

glioma types have been shown to differ from each other in their p53

status, c-myc, and epidermal growth factor receptor and PDGF(platelet-derived growth factor) receptor overexpression; they share

p14, p16 deletion, and low expression of multidrug resistance gene.

Figure3. VSVrp30atargetsmultipleseparatexenograftsinthebrain.AsingleintravenousinjectionofVSVrp30awasgiventoanSCIDanimalbearingbilateralstriatalredfluorescentrU87graftsandathirdincidentaltumormassthathasseededthroughthe

CSF to the lateral ventricle. A is a montage of fluorescent and phase-contrast images from a representative coronal brain slice

prepared48HPI.A

showscolocalizationof viral-encoded GFPand tumorcell-expressedRFP, simultaneouslyin all tumormasses.BDweretakenfromtheventricularborderoftherightxenograftasmarkedon A. EGarefromtheincidentaltumormassattheventricular wall (arrow). HJare from brain parenchyma, distant from the tumor graft. Hshows phase-contrast image, and I

showsthe absence of viralfluorescence.Jshowshealthynucleiof cellsin thebrain parenchyma(BP), arachnoid(Ar), andcorticalblood-vessels (CV). Similarly, no infection was observed outside the tumor in the slice preparation at 72 HPI. Ctx, Cortex; LV,lateral ventricle; Tm, TM, tumor. Scale bars: A, 500m; BJ, 20m.



6/12

U118 cells were stably transfected and sorted using the same proto-col we used for U87MG cells and will be called rU118 herein. SCIDmice bearing rU118 human glioblastoma xenografts (n 12 tu-mors) were killed 48 and 72 h (six tumors each) after a single intra-venous injection of VSVrp30a; widespread selective infection, andapoptotic cell death wasnoted(Fig.5AC). There was no significantstatistical difference in VSV induction of apoptosis between rU118tumors and rU87 tumors 72 HPI (p 0.1855, Students t test),suggesting that the virus targeted both types of glioma cells.

We also tested a mouse mammary carcinoma, the 4T1 cell.4T1 cells were transiently transfected with the monomeric DsRed

gene and injected into the brain and periphery 48 h after trans-fection. We used 4T1 mouse mammary carcinoma cells to createan animal model with simultaneous peripheral and multiple

brain tumors to mimic the setting of a dis-seminated systemic cancer. VSVrp30a in-fected all cranial (Fig. 5 D, E) and subcuta-neous flank grafts (Fig. 5F) in SCID mice(n 2 mice) after a single intravenousinjection.

VSVrp30a kills a wide variety of humancancer cells that can cause brain tumorsTo test the tumor cell tropism, we testedVSVrp30a on five additional human carci-noma lines. Our choice was guided by thefact that lung and breast cancers are themost common sources for brain metasta-sis. All human cancer lines tested, includ-ing A549, Calu-1 lung carcinoma, andT-47D, MCF7, and BT-474 breast carci-noma, were infected and killed byVSVrp30a at an MOI of 1. All cells exhib-ited viral green fluorescence by 60 h afterinfection, and all cells, dividing and nondi-

viding, in the culture dish were killed by4872 h after infection (Fig. 5I,J). In addi-tion, multiple cultures were establishedfrom tissue derived from resected glioblas-toma surgery. These primary high-gradegliomas were grown in culture and infectedwith VSVrp30a at an MOI of 1. Completeinfection and presence of cytopathic effectswere observed within 36 h after virus appli-cation (Fig. 5G,H), indicating that the on-colytic capacity of VSV seen on establishedglioma and other cancer cell lines couldtranslate to tumor cells directly derived

from human cancer patients.IV delivery of VSV efficiently targets andlyses brain xenografts in nude miceSCID mice are deficient in both humoraland cellular immune defense. Nude mice,conversely, have an immune defect pri-marilyaffecting T-cells. To test whether theless compromised immune status wouldinterfere with tumor targeting, we testedstriatal rU87 glioma xenografts in NCr-Nude mice (n 6 tumors). Tumor xeno-grafts were infected and destroyed withsimilar efficiency and kinetics in the nude

mouse as earlier in the SCID mouse (Fig.5KL). At 72 h after virus inoculation,there was no significant difference in dying

cells revealed by activated caspase 3 immunostaining betweenrU87 tumors in SCID animals, rU118 tumors in SCID animals,and rU87 tumors in nude mice ( p 0.085, ANOVA test). Allbrain tumors were similarly targeted by the virus.

Intranasal VSV targets olfactory bulb tumorsAfter intranasal inoculation, VSV can enter the CNS. Nasal mu-cosal infection results in infection of olfactory neurons, and VSVcan gain access to the CNS through the olfactory nerve, whichprojects to the glomeruli in the olfactory bulb (van den Pol et al.,

2002). We tested the hypothesis that VSV could enter the brainalong the olfactory nerve and target brain tumors in the olfactorybulb. rU87 cells were stereotactically grafted into one of the two

Figure 4. VSV infectionof braintumor xenograftsresults in widespreadtumor celldeath. Slicepreparations fromuninfected

(AD) and infected (EH) rU87 xenografts in SCID mice. Uninfected tumor xenografts have normal cellular morphology asoutlined by cytoplasmic RFP transgene expression (A), have no viral green fluorescence (B), have normal appearing nuclei (C),anddonot stainwiththeapoptosismarkercleavedcaspase3 (D).VSV-infectedtumorcells,incontrast,expressviralGFP( F),losecellular integrity (E), undergo chromatolysis (G), and stain positively for the apoptotic marker (H) at 48 HPI. Graphs (IK)indicate the increase in viral tumor cell killing during the first 3 d after virus injection. Laser confocal DIC imaging showsuninfected(L)andinfected(M) gliomaxenografts. Notethat, despite widespread cellular debrisinside the VSV-infected tumorxenograft,vessel architectureis sparedand endothelialcells appearmorphologicallyhealthy(arrows).Noninfected cellsconsti-tuting the vessel wall stained positive for endothelial marker von Willebrand factor (arrows in N,O). Scale bar, 50 m.



7/12

olfactory bulbs in SCID mice. Four daysafter tumor implantation, animals were in-oculated in each nostril with 25l contain-ing 2.5 107 PFU of VSVrp30a. VSV wastransported to the periphery of both olfac-tory bulbs after intranasal inoculation andinfected the olfactory nerve and the tumor

but did not show much infection in deeperparts of the control contralateral bulb lack-ing tumor (Fig. 6F). SCID mice with VSVin the olfactory bulb survived VSV infec-tion for 2 weeks. A long survival in this tu-mor model allowed us to demonstrate thatfew tumor cells could be found in the bulbafter the viral challenge. In animals killed 3DPI (four tumors) or 5 DPI (six tumors),VSV infection was located within the tu-mor but not in the surrounding paren-chyma or in the intact contralateral olfac-tory bulb (Fig. 6). In nine additional micekilled 7, 10, 15, 16, or 17 d after virus inoc-ulation, we found cellular debris indicativeof a viral infection but did not find surviv-ing tumor cells, suggesting that the viruswas successful in eliminating the tumor.

Although we found evidence of infectedolfactory nerve fibers that could serve asentry to theCNS, there is still thepossibilitythat VSV entered the blood circulationthrough intranasal application and mightreach the olfactory bulb via the circulatorysystem. To address this question, we placedperipheral sentinel tumors into the flank ofthree animals simultaneously with olfac-

tory tumor grafting. Seven days later, micereceived 2.5 107 PFU VSVrp30a intoeach nostril. At 3 d after inoculation, botholfactory bulb and peripheral tumors wereanalyzed. Interestingly, no viral infectionwas found at the remote sentinel tumor(Fig. 6GI), whereas strong GFP expres-sion was evident at the nasal mucosa, olfac-tory nerve and bulb, and at the olfactorytumor graft, suggesting a direct path ofVSVrp30a through the olfactory nerveroute.

Real-time microscopy of VSVrp30aoncolysis in cortical glioma xenograftsTo study the kinetics of VSV oncolysis, wealso used a time-lapse in situ human braintumor imaging model. rU87 cells wereimplanted subpially intothe SCID mousecortex (n 3) under a glass coverslippermanently mounted on a parasagittalcraniectomy (Fig. 7A, B). By imaging tu-mor cells in the living animal brain using fluorescent and laserconfocal microscopy, we could detect VSV infection as early as16 HPI after tail vein injection of 10 7 PFU of VSVrp30a (Fig.7C) and follow viral infection over several days in the live

brain. Tumor growth and infection were observed every 4 8 hover a course of 4 d. VSV infection spread through the wholetumor bulk within 96 h (Fig. 7C); although multiple points of

infection were seen, the majority of tumor cells appeared to beinfected by a locally spreading VSV infection. Laser confocalmicroscopy gave us a cellular or even subcellular resolution,allowing us to see not only single cells but also their processes.

Using this technique, we found that VSV infected single infil-trating red tumor cells but not the surrounding normal braintissue (Fig. 7 D, E). Vascular patency and circulation appeared

Figure5. Brainandmammarycancersweretargetedindifferentxenograftmodelsofthebrainandtheperiphery.ToexcludethepossibilitythatVSVrp30atumortargeting mightbe specific tothe rU87glioma cell line in theSCIDmousemodel, wetestedthevirusforitsoncolyticactivityondifferentcancercelllines invivo and invitro.XenograftsofrU118high-gradegliomacellsweretargeted with VSVrp30a,and thevirusspread throughout thewholetumormassand induced apoptosiswithin 72 h (AC).4T1mouse mammary cancer cells in the brain (D, E) and in the flank (F) were also effectively and selectively targeted and killed inSCID mice after a single intravenous injection of VSVrp30a (arrows indicate tumornormal tissue border). Tissue from patientsundergoing glioblastoma surgery was cultured and completely infected by VSVrp30a within 36 h (G, H). Three other humanbreastcarcinoma(I)andtwo humanlungcarcinomacelllines(J)weretested invitro,andallwerekilledwithin4872haftertheadditionof VSVrp30a ata multiplicityof infectionof 1. T-cell-deficient nude mice received thesamerU87 gliomaxenograftandweresubjectedtotheidenticalvirusinjectionprotocolasintheT-andB-cell-deficientSCIDmousemodel.ComparablewithSCIDmice, glioma xenografts in the nude mouse brain were efficiently targeted and killed with comparable kinetics (KM). At 3 d

aftervirus injection,virus-induced apoptosis wasobserved in themajorityof tumorcells. Scalebars:A,D,300m;B, 50m; C,20m; E, 30m; F, 100m;G, H, 30m; K, L, 20m.



8/12

normal in tumor vessels within the infected area even at ad-

vanced stages of tumor cell lysis; in live real-time imaging,

erythrocytes could be seen moving through the uninfected

walls of the blood vessels within the tumor even when all

surrounding tumor cells were infected and dying.

Mechanism of VSV tumor targeting

Theblood brain barrier maybe leaky in some types of tumors. In

normal circumstances, VSV does not enter the CNS through the

hematogenous route. Therefore, to explore what makes tumorxenografts permeable to VSV, we tested bloodbrain barrier in-

tegrity in the xenograft neovasculature. Strong Evans Blue stain-

ing was detected within bilateral striatal tumors in SCID (n 4

tumors) and nude mice (n 2 tumors) 1 h after intravenous

injection of the dye. On all sections, the Evans Blue stained area

completely overlapped with the red tumor mass, indicating a

uniform permeability of the tumor to dye (Fig. 8A, B). Control

experiments were performedto address theeffect of needleinjury

alone on Evans Blue leakage. Evans Blue leakage was observed

only immediately after the injury. Ten days after injury, the same

time given for tumor formation, no dye leakage could be ob-

served (Fig. 8C). All brains showed dye leakage in circumven-tricular areas in which the blood brain barrier is normally per-

missive for dye infiltration.

Sterile needle injury or human astrocyte transplants do notget infectedTo test the hypothesis that tumor targeting of VSV was not a conse-

quence of nonspecific parenchymal brain damage at the injection

site, bilateral sterile stab wounds were done using the same (Hamil-

ton point style 2) needle at the same striatal coordinates (n 4

wounds). When the same amount of intravenous virus was injected

10 d after this needle injury, we did not detect any viral GFP within

thebrain (Fig. 8D). In parallel, we addressed thequestion of whether

the virus targets human cells, or transplanted cells in general, rather

than tumor cells specifically. When human (n 4 injections) or

mouse (n 4 injections) normal astrocytes were transplanted into

the same coordinates used for tumor cells, no viral GFP expression

was detected in the brain 72 h after intravenous virus injection (Fig.

8E). These data support the view that VSVrp30a does not enter the

brain because of the injury of cell transplantation, nor does it target

normal humanor mouse cells transplanted into the brain. Infectionof glioma xenografts appears specific for replication-competent

VSV. To test whether replication was essential for effective tumor

targeting, we tested a recombinant replication deficient VSV (VSV

GGFP)for itscapacityto infect intracranial xenografts. Thisvirus

can infect cells but, because of the deletion of the VSV-G gene, can-

not produce infectious progeny, resulting in only a single cycle ofinfection. Intravenous injection of 10 7 PFUs of VSVGGFP to

animals with bilateral striataltumors (n 2) did notresult in tumor

Figure 6. Gliomas within the olfactory bulb are infected after intranasal inoculation. Intranasally given VSV can enter the brain through the olfactory nerve. When rU87 glioma cells wereunilaterally implanted into the olfactory bulbs of SCID mice, tumor xenografts grew similar to the striatal xenografts (A, C). Four days after implantation, animals were challenged with intranasal

VSVrp30a andtheviruswas documented toinfect theolfactoryxenografts (B,D, F).VSV replicationwas documented inaxonswithin theolfactorynerve(E,andarrowinB)andinsomebutnotallof theolfactory glomeruli, which indicatesthe port of VSVentry (B, E).Viruscan enter both tumor-containingand control bulbs throughbotholfactorynerves,but only tumor-bearing side showsinfection in the central part (F). Simultaneously grown subcutaneous tumor grafts (G) did not show infection after intranasal VSVrp30a application (I), indicating that virus travels through theolfactoryroute rather thansystemically.Stereomicroscopicpicture of an intact subcutaneous flanktumor(G)andsectionofthesametumor(H).L-OB,Leftolfactorybulb;R-OB,rightolfactorybulb;

EPL, external plexiform layer; GL, glomeruli; Mi, mitral cell layer; ON, olfactory nerve. Scale bars: A, B,G, 300m; E, F, 200m; C, D, H, I, 50m.



9/12

infection (Fig. 8G). As a positive control,GFP-expressing scattered cells were detectedin the liver after intravenous injection ofVSVGGFP (Fig. 8H). To exclude thepossibility that targeting of intracranial tu-mor xenografts by VSV was not the result ofnonspecific viral entry, we used two other vi-

ruses that were previously shown to infect orlyse glioma cells in vitro. Intravenous injec-tion of a replication-incompetent adeno-associated virus type 2 (AAVGFP; n 2; 3DPI) or a replication-competent pseudora-bies virus (PRVGFP; n 2; 2 DPI) did notresult in infection of intracranial rU87 xeno-grafts (Fig. 8F,I). All three of these viruseswere shown previously to infect tumor cellsin vitro, and this was reconfirmed in vitro forthe stocks used in this study.

Mechanism of VSV tumor selectivityOur data are consistent with the view that

VSV selectively targets glioma in the brain,suggesting that some mechanism underliesthe relative selectivity of the virus to thetumor. A correlation exists between VSVinfection and various defects in the inter-feron system in tumor cells (Stojdl et al.,2000; Balachandran et al., 2001). We previ-ously compared a series of five differentbrain tumor cell lines and normal humancontrol astrocytes and demonstrated aninterferon-dependent selectivity of VSV forglial tumor cells over their nontumor coun-terpart (Wollmann et al., 2007). However,

antiviral defense may not exclusivelyaccountfor the selectivity, and several additionalmechanisms have been proposed (Wagnerand Rose, 1996; Balachandran and Barber,2004; Barber, 2005; Lyles and Rupprecht,2007). No specific receptor has been identi-fied for VSV; after binding, the virion under-goes clathrin-mediated endocytosis, and, af-ter acidification in the endosome, theuncoated virus enters the cytoplasm. To testthe hypothesis that VSVrp30a binds and en-ters tumor cells with a higher efficiency, wecompared viral binding and entry in rU87glioma cells and primary normal adult hu-man astrocytes in vitro using quantitative PCR (Fig. 9). Viral quan-

tity was normalized against cellular GAPDH. In separate quantita-tive PCR experiments, we confirmed that GAPDH was expressed atcomparable quantities in normal astrocytes and glioma cells (resultsnotshown).To test forthe quantityof VSVthat bindsto cell surface,

we incubated the cells with virus at 4C. After 20 min of incubation,we washed the cells five times with buffered medium before RNAextraction. Compared with human astrocytes, the quantity of cell-bound VSVrp30a at 20 min after infection appeared to be slightly(1.8times) greater for rU87glioma cells (statistically significant;p

0.005, n 6, Students ttest)(Fig.9A). Whenwe quantified surface-bound plus intracellular VSVrp30a after a 30 min incubation at an

MOI of10, weagainfound a somewhat greater viralquantityin rU87cells (p 0.05, n 6, Students ttest)(Fig. 9B). We then comparedhow efficientlythe cells internalizedVSV.For this purpose, we incu-

bated the cells with VSVrp30a at an MOI of 10 for 30 min and

incubated them for another 1 h after changing the medium three

times to eliminate unbound virus. At 90 min after infection, cells

were washed five times with PBS and trypsinized, and intracellular

VSV was quantified. Finally, to measure only the endosomal frac-

tion, the same experiment was duplicated in the presence of ammo-

nium chloride, which inhibits viral uncoating from the endosomes.

At 90 min after infection, we found 2.6 times more virus in the

endosomes and 3.6 times more virus intracellularly in rU87 cells

compared with normal astrocytes (both statistically significant;p

0.001, n6,andp0.003, n6, respectively,Studentsttest) (Fig.

9C,D). These data suggest that VSV binding to the plasma mem-

brane andvirus internalizationappear to be greateringlioma thaninnormal astrocytes. These factors may contribute to the greater level

of VSV infection found in glioma.

Figure7. Time-lapseconfocalimagingofVSVtargetingofgliomainlivemousebrain.Kineticsofviraloncolysiswereanalyzedusing rU87 cells implanted subpially in SCID mouse cortex under a cranial glass coverslip.A shows the localization of the tumorcellsin thelivinganimal brain.Fluorescence microscopy during96 h afteran intravenousinjectionof VSVrp30a showedthattheviral GFP expression became apparent 16 h after inoculation (B). The infection spread within the tumor and progressed toencompassthewholetumormass( C). In situ laserconfocalmicroscopy revealed thatsingleinfiltrative tumorcells weretargetedaftersuchanintravenousviraldelivery.Notethatsingletumorcellsinfiltratingthecortex(arrow)expressviralmembranousGFP(D, E). Scale bars: D, 50m; E, 10m.



10/12

DiscussionMetastatic tumors and glioblastomas make up the majority ofcancers in the brain and generally result in death within months;there is currently no cure. This therapeutic failure is mostly attribut-able to the incomplete elimination of all cancer cells. We show thatintravenous inoculation of VSVrp30a resulted in selective infectionand destruction of human glioblastoma and mammary cancer cellstransplanted into the mouse brain. Systemic delivery of a limiteddose of the replication-competent VSVrp30a resulted in selectiveinfection of multifocal tumors and in production of viral progenythat spread throughout the tumor within 72 h. Such a local self-

amplification is not found with other conventional cytotoxic treat-ments such as chemotherapy or radiation. Different types of cancercells were targeted, with sparing of normal brain cells, underlining

VSV as a potential therapeutic against multi-ple types of cancer in brain.

Tumor targeting by systemic deliveryGliomas and brain metastases are differentin their tumor biology, and growth patternand kinetics. The mode of action of viral

therapeutics differs fundamentally fromconventional methods such as radiation orchemotherapy. The unique property ofVSV that separatesit from many other con-ventional treatment technologies is thatVSV shows oncolytic activity against abroad spectrum of cancer types, includingboth systemic cancers and primary cancersof the brain, despite substantial genotypicvariability. We tested VSV as a potentialoncolytic candidate that can be intrave-nously administered to treat multiple typesof brain cancer regardless of the kind oftumor or the status of systemic disease.

Our study shows that a singleintravenousinjection of VSVrp30a effectively targetsmultifocal or infiltrative tumors simulta-neously in the brain and periphery. By 3 dafter inoculation, tumors generally werecompletely or almost completely infected,and tumor cells were dying or dead. In con-trast to VSV, several other oncolytic viruses(adenovirus, herpes simplex) have been ap-plied by systemic injections and were able toinfect only a small part of the tumor (RainovandKramm, 2001; Lyonset al., 2006; Liu andKirn, 2007). In our experiments, after intra-

venous injection, selective viral infection wasfound in all tumors and tumor types, and inboth SCID and nude mice on different ge-netic backgrounds, indicating a high effi-ciency of VSV-tumor targeting. A notablefinding was the demonstration of simulta-neous oncolytic infection in bothsubcutane-ous flank and brain tumors in the same ani-mal after a single systemic virus injection. Inaddition to demonstrating the efficiency of asystemic VSV-based therapy, this findingemphasized the potential of VSV to infectwidely metastatic cancers. Given the onco-

lytic activity of VSV against a range of cancercells with diverse genetic defects (Stojdl et al.,

2003; Lichty et al., 2004; Barber, 2005), an intravenous injection ofVSVcan potentially be used to target simultaneously both brain andperipheral metastases of a systemic cancer.

Direct intratumoral injection may be useful for some tumors.To date, such intratumoral injections of viruses have resulted in avery limited viral spreadaround theinjection site (Pulkkanen andYla-Herttuala, 2005). The ability of VSV to infect multiple inde-pendent central and peripheral sites and spread in them is sub-stantive,making it possible to treat disseminateddisease andeventarget minute tumor colonies that are otherwise difficult to de-tect. Xenograft models are limited in that human gliomas often

grow expansively rather than infiltratively in mice. Nonetheless,infiltrating cells were observed within the host brain, and thesewere targeted by VSV.

Figure 8. VSV selectively enters brain tumors through a leaky bloodbrain barrier but does not infect control wounds ornormalbrain cellimplants.After intravenousinjection, EvansBlue leakedinto tumorxenograftsin whichit colocalized(arrows)with tumoral RFP fluorescence, indicating an incompetent blood brain barrier within the entire tumor (A, B). No Evans Blueleakage was observed in areas with needle injury only (see arrows) (C). Nonspecific brain damage inflicted by a sterile needleinjury(D)orhumanastrocytetransplants(E)in thesame brainregionusedfor tumorgraftingwasnotenough toallowVSVrp30aentryafterintravenousinjectionofthesamedoseofVSVrp30aasinexperimentalanimals.Tumortargetinginthebrainrequiredreplication-competentVSV,and replication-deficientVSV(VSVdeltaG-GFP)didnot infect rU87xenografts inthe mousebrain,butnumerousinfected cellswere documented in theliver (G, H).No infectionin thebraintumor,liver,or spleenwas detected afterintravenousinjectionof pseudorabiesvirus (PRVGFP)or adeno-associatedvirus (AAVGFP),both of whichare knownto infect

glioma cells (F, I). LV, Lateral ventricle. Scale bars:A, 300m; D, H, 100m; E, G, 50m; F, I, 20m.



11/12

Olfactory nerve routeWe also tested a novel approach of target-ing brain tumors through neural circuits.VSV can enter the brain through the olfac-tory nerve (Reiss et al., 1998; van den Pol etal., 2002). Inoculation of the olfactory mu-cosa led to complete infection of the tumor

via the olfactory nerve; after infection, tu-mor cells were absent from the olfactorybulb. Relatively little infection was foundoutside the tumor in the tumor- bearingbulb or in the contralateral normal bulb.This finding indicates that neural routes ofVSV dissemination in brain can be ex-ploited for targeting tumors in specific re-gions and facilitating of distant tumors orinfiltrative tumor cells with minimal infec-tion in normal brain parenchyma.

Detection of recombinant virus and tumor

To localize the virus, we used a green fluorescent reporter genecoupled to the VSV-G gene in the position of the fifth VSV gene.This would shift the viral L-gene to the sixth position, resulting inattenuated L-protein synthesis and a slight reduction in replica-tion (Dalton and Rose, 2001), an advantage when consideringtreatment of the brain. Live microscopic imaging of the brainallowed us for the first time to follow the time course with single-cell resolution, from before inoculation to a point when VSV hadspread throughout the tumor. In experiments in which we im-aged VSVoncolysis through a glass windowabove thebrainusingtime-lapse confocal laser microscopy, green viral infection of redcortical tumors spread throughout the tumor, mostly by localexpansion of the area of infected cells. Blood vessels appeared

mostly undamaged even late in tumor infection. Furthermore,vessel cells that stained positive for the endothelial marker vonWillebrand factor appeared to be spared from infection, favoringthe hypothesis that virus entry is facilitated by vessel leakage.With live recording, blood cells could be seen moving throughpatent vessels throughout the tumor. In control experiments,VSV did not enter the normal brain through the vasculature, nordid it enter the brain away from the tumor in mice bearing tu-mors, raising the question as to how VSV entered the brain. Wetested the hypothesis that a local disruption of the BBB existedusing intravenous dye injections and found dye entered the brainonly in areas of tumors but not at 10-d-old control needleinjuries, indicating an incompetent BBB. This leakage was selec-

tive for the tumor vasculature because additional experimentsshowed that simple focal brain damage or implantation of non-malignant cells into the brain did not result in VSVrp30a infec-tion in the CNS. The ability to cross the BBB does not generalizeto all viruses because we found no evidence that pseudorabiesvirus or AAV infected the tumor. Even a leaky BBB preventedadenovirus and herpes from entering brain tumors (Rainov andKramm, 2001). Furthermore, a replication-deficient VSV (Milleret al., 2004;Okuma et al., 2006) that can replicate only in the firstcell infected without subsequent release of infectious virionsshowed no tumor invasion; that the replication-incompetent vi-rus can be detected was shown by the presence of a few solitaryinfected, GFP-expressing liver cellsand byin vitro experiments in

which the virus showed robust infection and reporter gene ex-pression in cultured glioma. This suggests that ongoing viral rep-lication is needed to initiate tumor infection. VSV entering the

tumor maytherefore arise,at least in part, from infectious daugh-ter virions released by infected peripheral cells.

Infiltrating tumor cells that migrate away from the main tu-mor body into normal brain escape local treatments and are animportant cause of tumor recurrence. Confocal imaging in bothlive and fixedsections demonstrated that thesecells are selectivelytargeted by VSVrp30a, most likely a result of virus progeny dif-fusing from the main tumor bulk. Despite the heavy infection ofthe tumor and infiltrative tumor cells, we did not observe infec-tion in neighboring normal astrocytes or neurons in immediatecontact with the tumor, indicating the high tumor specificity ofthe infection. In this study, VSV rapidly infected and killed twotypes of glioma and one type of mammary cancer in vivo. It alsokilled five systemic carcinoma cell lines (lung and breast cancer)and cells from two primary glioblastoma in vitro, indicating thepotential for VSV to kill a diverse array of cancers that may arise

or metastasize in the brain. In addition to its diverse tropism,VSV, unlike some other viruses, is not dependent on specificgenetic tumor defects such as the p53 status, RAS or myc onco-gene activity, cell cycle status, or hypoxic environment of thetumor cell (Balachandran et al., 2001; Connor et al., 2004; Woll-mann et al., 2005).

Mechanisms of viral selectivity for cancer cellsWe found a remarkably high degree of selective infection of thebrain tumors with relatively little infection of the surroundingnormal brain; the ratio of infected tumor cells to infected normalbrain cells was 10,000:1, or better. A large fraction of varioustumors display defects in their interferon system (Stojdl et al.,

2003). VSV is highly sensitive to interferon-mediated antiviraldefense mechanisms with which normal cells effectively controlVSV infection and spread (Trottier et al., 2005). We and othersdemonstrated previously VSV mutants and recombinants thatare enhanced in their susceptibility to interferon have shownpromise in targeting cancers (Stojdl et al., 2003; Ahmed et al.,2004; Ebert et al., 2005). VSVdeltaM51, which has an enhancedinterferon susceptibility, has also been shown to target gliomassystemically and prolong survival (Lun et al., 2006). We took analternative strategy and isolated VSVrp30a by raising VSV formany generations on gliomas under selective pressure (Woll-mann et al., 2005). VSVrp30a showed enhanced binding andinternalizationin glioma cells, mechanisms that could contribute

to tumor selectivity.Because human cancers show more complex genotypic alter-ations compared with rodent tumors, we used human glioblas-

Figure9. VSVbindingandentryismoreefficientinrU87gliomacells.QuantificationofVSVbyRT-PCRatdifferentintervalsindistinctcellularcompartmentsindicatedthatVSVrp30abindsandenterstherU87humangliomacelllinemoreefficientlythanits

normal cellular counterpart. Cell-surface-bound VSV after 20 min of incubation with the virus was significantly greater in rU87cells compared with normal human astrocytes (A). When surface-bound plus internalized virus was quantified at 30 min, therewasa trendtoward more virus inrU87but thedifferencewasnotsignificant(B).At90minafterinfection,therewassignificantlymoreVSVintheendosomalcompartmentinrU87cells(C)whencomparedwithastrocytes.Alsothetotalamountofintracellularvirus was significantly greater in rU87 cells (D). All experiments were done in triplicate using VSVrp30a at an MOI of 10.



12/12

toma models in immunodeficient mice (Hahn and Weinberg,2002). A limiting factor in this study is that such mice are moresusceptible to viral infections of any kind. Wild-type mice andhumans survive peripheral VSV infections, but immunodeficientmice do not, making it difficultto study long-term results of VSVtargeting of human cancers in animals. Despite the vulnerabilityof immunodeficient mice to VSV, we nevertheless demonstrated

a therapeutic time window of complete tumor oncolysis withminimal brain infection outside the tumor. With longer survivaltimes than used in the present study,VSV could potentially infectnormal brain cells. Attenuated recombinant VSV strains mayprovide an additional margin of safety when used in the brain ofnonhuman primates (Johnson et al., 2007) andin mice (Flanaganet al., 2003; Clarke et al., 2007) after intracerebral injections orintranasal administration.

ReferencesAghi M, Rabkin S (2005) Viral vectors as therapeutic agents for glioblas-

toma. Curr Opin Mol Ther 7:419430.Ahmed M, Cramer SD, Lyles DS (2004) Sensitivity of prostate tumors to

wild type and M protein mutant vesicular stomatitis viruses. Virology330:3449.

Balachandran S, Barber GN (2004) Defective translational control facilitatesvesicular stomatitis virus oncolysis. Cancer Cell 5:5165.

Balachandran S, Porosnicu M, Barber GN (2001) Oncolytic activity of ve-sicular stomatitis virus is effective against tumorsexhibitingaberrant p53,

Ras, or myc function and involves the induction of apoptosis. J Virol75:34743479.

Barber GN (2005) VSV-tumor selective replication and protein translation.Oncogene 24:77107719.

Central Brain Tumor Registry of the United States (2005) Statistical report:primary brain tumors in the United States, 19982002. Chicago: CentralBrain Tumor Registry of the United States.

ClarkeDK, Nasar F, LeeM, Johnson JE,WrightK, Calderon P, GuoM, Natuk

R, Cooper D, Hendry RM, Udem SA (2007) Synergistic attenuation ofvesicular stomatitis virus by combination of specific G gene truncationsand N gene translocations. J Virol 81:20562064.

Connor JH, Naczki C, Koumenis C, Lyles DS (2004) Replication and cyto-pathic effect of oncolytic vesicular stomatitis virus in hypoxic tumor cellsin vitro and in vivo. J Virol 78:8960 8970.

Dalton KP, Rose JK (2001) Vesicular stomatitis virus glycoprotein contain-ing the entire green fluorescent protein on its cytoplasmic domain is

incorporated efficiently into virus particles. Virology 279:414421.DeAngelis, L. M (2001) Brain tumors. N Engl J Med 344:114123.Ebert O, Harbaran S, Shinozaki K, Woo SL (2005) Systemic therapy of ex-

perimental breast cancer metastases by mutant vesicular stomatitis virusin immune-competent mice. Cancer Gene Ther 12:350358.

Flanagan EB, Schoeb TR, Wertz GW (2003) Vesicular stomatitis viruseswith rearranged genomes have altered invasiveness and neuropathogen-esis in mice. J Virol 77:57405748.

Hahn WC, Weinberg RA (2002) Modelling the molecular circuitry of can-

cer. Nat Rev Cancer 2:331341.

Johnson JE, Nasar F, Coleman JW, Price RE, Javadian A, Draper K, Lee M,Reilly PA, Clarke DK, Hendry RM, Udem SA (2007) Neurovirulenceproperties of recombinant vesicular stomatitis virus vectors in non-

human primates. Virology 360:36 49.Khuri FR, Nemunaitis J, Ganly I, Arseneau J, Tannock IF, Romel L, Gore M,

Ironside J, MacDougall RH, Heise C, et al. (2000) a controlled trial ofintratumoral ONYX-015, a selectivelyreplicating adenovirus, in combi-

nation with cisplatin and 5-fluorouracil in patients with recurrent headand neck cancer. Nat Med 6:879 885.

Lichty BD, Power AT, Stojdl DF, Bell JC (2004) Vesicular stomatitis virus:re-inventing the bullet. Trends Mol Med 10:210216.

Liu TC, Kirn D (2007) Systemic efficacy with oncolytic virus therapeutics:clinical proof-of-concept and future directions. Cancer Res 67:429 432.

Lun X, Senger DL, Alain T, Oprea A, Parato K, Stojdl D, Lichty B, Power A,Johnston RN,HamiltonM, etal. (2006) Effects of intravenously admin-

istered recombinant vesicular stomatitis virus (VSV(deltaM51)) on mul-

tifocal and invasive gliomas. J Natl Cancer Inst 98:15461557.

Lyles DS, Rupprecht CE (2007) Rhabdoviridae. In Fields Virology, D. M.

Knipe, and P. M. Howley, eds. (Lippincott Williams and Wilkins), pp.

13631408.

Lyons M, Onion D, Green NK, Aslan K, Rajaratnam R, Bazan-Peregrino M,

Phipps S, Hale S, Mautner V, Seymour LW, Fisher KD (2006) Adenovi-

rustype 5 interactionswith human blood cells maycompromise systemic

delivery. Mol Ther 14:118128.MillerMA, LavineCL, Klas SD,Pfeffer LM,Whitt MA (2004) Recombinant

replication-restricted VSV as an expression vector for murine cytokines.

Protein Expr Purif 33:92103.

Ohgaki H, Kleihues P (2005) Population-based studies on incidence, sur-

vival rates, and genetic alterations in astrocytic and oligodendroglial gli-

omas. J Neuropathol Exp Neurol 64:479 489.

Okuma K,BoritzE, WalkerJ, SarkarA, AlexanderL, Rose JK (2006) Recom-

binant vesicular stomatitis viruses encoding simian immunodeficiency

virus receptors target infected cells and control infection. Virology

346:8697.

Parato KA, Senger D, Forsyth PA, Bell JC (2005) Recent progress in the

battle between oncolytic viruses and tumours. Nat Rev Cancer 5:965976.

Pardridge WM (2002) Drug and gene delivery to the brain: the vascular

route. Neuron 36:555558.

Posner JB, Chernik NL (1978) Intracranialmetastasesfrom systemic cancer.Adv Neurol 19:579592.

Pulkkanen KJ, Yla-Herttuala S (2005) Gene therapy for malignant glioma:

current clinical status. Mol Ther 12:585598.

Puumalainen AM, Vapalahti M, Yla-Herttuala S (1998) Gene therapy for

malignant glioma patients. Adv Exp Med Biol 451:505509.

Rainov NG, Kramm CM (2001) Vector delivery methods and targeting

strategies for gene therapy of brain tumors. Curr Gene Ther 1:367383.

Rainov NG, Ren H (2003) Oncolytic viruses for treatment of malignant

brain tumours. Acta Neurochir Suppl 88:113123.

ReissCS, Plakhov IV,Komatsu T (1998) Viralreplicationin olfactoryrecep-

torneurons andentryinto theolfactory bulb andbrain. AnnN Y Acad Sci

855:751761.

Smith BN, Banfield BW, Smeraski CA, Wilcox CL, Dudek FE, Enquist LW,

Pickard GE (2000) Pseudorabies virus expressing enhanced green fluo-

rescent protein: A tool for in vitro electrophysiological analysis of trans-

synaptically labeled neurons in identified central nervous system circuits.

Proc Natl Acad Sci USA 97:92649269.

Stojdl DF, Lichty B, Knowles S, Marius R, Atkins H, Sonenberg N, Bell JC

(2000) Exploitingtumor-specificdefects in the interferonpathway witha

previously unknown oncolytic virus. Nat Med 6:821825.

Stojdl DF, Lichty BD, tenOever BR, Paterson JM, Power AT, Knowles S,

MariusR, Reynard J, Poliquin L, AtkinsH, etal. (2003) VSVstrains with

defects in their ability to shutdown innate immunity are potent systemic

anti-cancer agents. Cancer Cell 4:263275.

Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ,

Belanger K, Brandes AA, Marosi C, Bogdahn U, et al. (2005) Radiother-

apy plus concomitant and adjuvant-temozolomide for glioblastoma.

N Engl J Med 352:987996.

Trottier MD, Palian BM, Reiss CS (2005) VSV replication in neurons is

inhibited by type I IFN at multiple stages of infection. Virology

333:215225.

van den Pol AN, Dalton KP, Rose JK (2002) Relative neurotropism of a

recombinant rhabdovirus expressing a green fluorescent envelope glyco-

protein. J Virol 76:13091327.

Wagner RR, Rose JK (1996) Rhabdoviridae: the viruses and their replica-

tion. In Fields virology, B. N. Fields, and D. M. Knipe, eds. (Philadelphia,

Raven Press).

Wollmann G, Tattersall P, van den Pol AN (2005) Targeting human glio-

blastoma cells: comparison of nine viruses with oncolytic potential. J Vi-

rol 79:6005 6022.

Wollmann G, Robek MD, van den Pol AN (2007) Variable deficiencies in

the interferonresponseenhance susceptibility to vesicular stomatitis virus

oncolyticactionsin glioblastomacells butnot in normalhumanglial cells.

J Virol 81:14791491.


Documents

VSV Selectively Destroys Tumors