Introduction

Oncolytic viruses (OVs) have been engineered or selected to exploit genetic defects in tumor cells.^1,2,3,4 Unlike conventional gene therapy which involves using viral vectors for targeted gene delivery to diseased tissues, OVs are fully replication-competent and are expected to actively spread through cancerous tissue, killing tumor cells in their path by direct lysis. Indeed, in vitro models have clearly demonstrated that this occurs, and that therapeutic windows of greater than 1,000-fold between normal cells and tumor cells are common.^5,6 In vivo, the situation is far more complex, because there are many barriers to virus delivery and replication that can compromise therapeutic efficacy. Often, much of a systemically administered viral therapeutic is scavenged by normal cells of the liver,⁷ or neutralized in the blood by antibodies⁸ or complement^9,10 or through non-productive binding to white blood cells.¹¹ The effective dose of an OV that reaches a tumor is significantly less than the administered dose and multiplicities of infection in vivo are much lower than commonly used in vitro.¹² Furthermore, the replication and spreading of the virus within tumors is likely to be compromised by the restrictive properties of solid tumor architecture and micro-environment (e.g., extracellular matrix, hypoxia, high interstitial tumor pressure, and low pH).^13,14 The first few days of OV infection and amplification within the tumor are of critical importance, considering that both innate and adaptive immune systems eventually curtail the spreading of the virus. Indeed, despite the excellent in vitro activity of a variety of OVs, systemic delivery in early phase I human trials has resulted in very limited anti-tumor activity.¹⁵ We set out to investigate some of the factors that could limit or enhance OV therapy in vivo, with an eye to identifying key areas of possible intervention for improving therapeutic outcomes. Using two different OVs (vesicular stomatitis virus (VSV) and vaccinia virus) and a combination of imaging approaches, we found that only a small amount of systemically delivered virus reaches tumor sites. Despite poor virus delivery to the tumors, we found that a significant portion of cells within OV-treated tumors were triggered to undergo apoptosis. Importantly, virus infection of a small proportion of tumor cells resulted in indirect killing of uninfected tumor cells. We discovered that viral infection triggered a loss of blood flow to the interior of the tumor, causing induction of massive cellular apoptosis in tumor cells while leaving normal tissues unaffected, and that the absence of vascular perfusion within infected tumors was induced by the recruitment of neutrophils to the tumor bed. The depletion of neutrophils from animals prior to OV administration eliminated apoptosis of uninfected tumor cells and permitted a more extensive replication and spreading of the virus throughout the tumor. Our results reveal a previously unappreciated and unexpected interplay between OVs, tumors, and inflammatory response.

Results

Extensive replication of VSV in subcutaneous tumors despite poor virus delivery

In order to gain a better understanding of the mechanisms of action of the therapeutic viral agents, we used a combination of imaging strategies to assess viral replication in vivo. Initially, BALB/c mice with subcutaneous CT-26 colon cancer tumors were intravenously infected with a recombinant version of VSV that expresses green fluorescent protein (GFP). Fluorescence microscopy of tumors excised 24 hours after infection showed limited VSV infection of CT-26 cells (Figure 1a), thereby suggesting that only a small amount of the virus had leaked out of the tortuous network of blood vessels at the rim of the tumor.

Figure 1.

Systemically administered vesicular stomatitis virus (VSV) is poorly delivered to subcutaneous tumor but is rapidly, extensively, and selectively amplified in tumor cells surrounding tumor neovasculature. (a) BALB/c mice with CT-26 subcutaneous tumors were treated intravenously with VSV expressing green fluorescent protein (GFP) and euthanized 24 hours later. A dissecting fluorescence microscope was used for performing in situ imaging of tumor. Overlay of bright field and green fluorescence. (b) BALB/c mice with CT-26 subcutaneous tumors were intravenously infected with VSV. Pairs of mice were sacrificed and perfused with phosphate-buffered saline at time points between 5 minutes and 144 hours. Organs were frozen and virus titers were subsequently determined by plaque assay. Data are represented as means standard error. (c) BALB/c mice with CT-26 tumors were treated with VSV expressing GFP intravenously and euthanized 48 hours later. Sections were stained for apoptosis by TUNEL assay and VSV localization is detected by immunofluorescence. Hoechst stain and overlay are also shown (20). PFU, plaque forming units.

Full figure and legend (59K)

Given that large amounts of systemically administered OV resulted in only limited infection of tumors, we investigated to what extent systemically administered oncolytic virus reached the tumor bed, and also studied the kinetics of virus replication in normal and tumor tissues. Mice with subcutaneous CT-26 tumors were treated intravenously with 10⁹ plaque forming units of VSV (AV1) and pairs of mice were sacrificed at time points between 5 minutes and 6 days. Homogenized tissues were titrated to quantify virus delivery and replication within various organs. Immediately following virus administration, the highest virus titer was found in the liver (7.2 0.5 log₁₀ plaque forming units/g), followed by the spleen (5.5 0.13), whereas virus titer in the tumor was much lower (3.0 1.1). Strikingly, at 24 hours after injection, titers in the tumors had increased by over 4 logs, while liver titers had decreased by the same amount (Figure 1b). This suggests that the attenuated form of VSV is able to productively infect only the tumor, while much of the injected virus is rapidly cleared by the liver (likely due to mechanical deposition as well as uptake in Kupffer cells^16,17). This observation is consistent with the results of our imaging study that showed that only a small amount of virus is delivered to tumors, and that it replicates at the tumor rim.

Replication of VSV in the tumor rim induces extensive apoptosis in the uninfected tumor core

The finding relating to poor OV delivery and replication within tumors was in sharp contrast to our original expectations, given that CT-26 cells are exquisitely vulnerable to VSV in vitro⁵ and that VSV is very effective therapeutically against localized and disseminated CT-26 tumors.⁵ We therefore re-confirmed the virus distribution as seen by GFP imaging, by carrying out immunofluorescence analysis of thin sections of tumors, using antibodies directed against VSV proteins. In addition, we detected terminal deoxynucleotidyl transferase-mediated dUTP nick end label (TUNEL) staining, a marker of apoptosis, as we would expect to see significant cell killing in CT-26 tumors that respond to OV therapy. As suggested by the experiments described earlier, infection was, in general, sparse and confined to the outer edge of the tumor (Figure 1c). In contrast, we found extensive apoptosis throughout much of the tumor, even in areas not infected with virus (Figure 1c). This observation was confirmed by immunohistochemical detection of VSV and active caspase 3, and the results demonstrated poor virus distribution with extensive apoptosis of uninfected cells at both 24 and 48 hours after infusion (Figure 2a). These results suggest that a major portion of the in vivo tumor killing activity of VSV is not by direct viral infection but, rather, by an apoptotic signal associated with neighboring virus infection. Importantly, active replicating virus was found to be a pre-requisite for the induction of apoptosis in non-infected tumor cells, as verified by the fact that this phenomenon was not observed following the administration of ultraviolet (UV)-inactivated virus (Figure 2b). In addition, this indirect apoptotic effect is confined to tumor cells and was not detected in adjacent normal tissue (Figure 2a).

Figure 2.

Vesicular stomatitis virus (VSV) infection triggers the killing of uninfected tumor cells and leads to loss of tumor perfusion in subcutaneous CT-26 tumors. (a) BALB/c mice with CT-26 tumors were treated intravenously with VSV expressing green fluorescent protein (GFP) and sacrificed at the indicated time points. Representative scans of serial sections of tumors treated as indicated and stained for immunohistochemical analysis to detect VSV and active caspase 3, so as to visualize areas of virus infection and apoptosis, respectively. Prior to being sacrificed, mice were perfused with fluorescent microspheres for 5 minutes. Tumor sections were prepared and analyzed for fluorescence using a microarray scanner. Tumors excised from VSV-treated mice show no fluorescence in the tumor core (large light areas), indicating that tumor tissue in VSV-treated mice is not perfused (). In contrast, adjacent normal tissue remains well perfused (+). Ten tumors were examined and representative tumors are shown. (b) BALB/c mice with CT-26 tumors were treated intravenously with ultraviolet (UV)-inactivated VSV particles and sacrificed 24 hours later. Fluorescent microspheres are distributed evenly throughout the tumor as evidenced by uniform fluorescence (black dots). H&E, hematoxylin and eosin.

Full figure and legend (237K)

OV infection triggers a loss of tumor perfusion in different models of tumor and virus

We reasoned that the massive cell death resulting from a very restricted viral infection could be caused by an acute loss of blood perfusion within the tumor. In order to visualize perfusion, 100 nm fluorescent beads (microspheres), were injected intravenously into VSV-treated mice with CT-26 tumors. The mice were then sacrificed after 5 minutes. Microspheres remain in the vascular system and are distributed throughout perfused regions of the normal organs and the tumor¹⁸ and can be visualized microscopically. We confirmed the presence of microspheres within blood vessels by co-injecting Hoechst dye #33242.^19,20 Indeed, Hoechst-stained endothelial cells are found surrounding the luminally contained fluorescent beads (Figure 3a). In tumors treated with phosphate-buffered saline (PBS), a relatively uniform distribution of fluorescent beads and Hoechst staining is observed (Figure 3b). In contrast, 24 hours after infection with VSV tumor perfusion is markedly affected, as measured by both microsphere distribution and Hoechst staining (Figure 3b). The fact that apoptotic staining occurred simultaneously with the loss of tumor perfusion is quite striking (Figure 2a). As predicted, in adjacent normal tissues in which we observed no secondary apoptosis, microspheres were found uniformly distributed (Figure 2a).

Figure 3.

Absence of intra-tumor blood flow is confirmed with in vivo Hoechst staining of endothelial cells. (a) Mice with CT-26 tumors were perfused with Hoechst #33342 stain and fluorescent microspheres for 5 minutes prior to being sacrificed. Sections were visualized with a fluorescence microscope. Pictures captured at 40. (b) Mice with CT-26 tumors were treated intravenously with vesicular stomatitis virus (VSV) or phosphate-buffered saline (PBS) and were perfused with Hoechst #33342 stain and fluorescent microspheres for 5 minutes, 24 hours after infection. Sections were visualized under a fluorescence microscope and representative pictures are shown (20).

Full figure and legend (101K)

The killing of uninfected tumor cells and the loss of tumor perfusion was not unique to the murine CT-26 colon cancer model.This phenomenon was observed also when nude mice implanted with the human SW620 tumor cell line were systemically treated with VSV (Figure 4 and Supplementary Figure S1). Furthermore, upon administration of an oncolytic version of vaccinia virus²¹ to mice with CT-26 tumors, we observed a similar loss of blood flow in the core of the tumor (Figure 4 and Supplementary Figure S2). However, in the case of vaccinia (in contrast to the results with VSV), killing of uninfected cells was not initiated until approximately 120 hours after infection, consistent with the slower replication cycle of this virus.

Figure 4.

Killing of uninfected tumor cells occurs with the use of vaccinia virus and in a xenograft model of human cancer. BALB/c mice with CT-26 tumors were treated with ultraviolet (UV)-inactivated vesicular stomatitis virus (VSV) particles, perfused with fluorescent microspheres 24 hours after injection, and sacrificed 5 minutes later. Fluorescent microspheres are distributed evenly though out the tumor, evidenced by uniform fluorescence (black dots). BALB/c mice with CT-26 tumors were treated intravenously with VSV expressing green fluorescent protein (GFP), and the tumors were analyzed for perfusion 24 hours later. BALB/c mice with CT-26 tumors were treated intra-peritoneally with vaccinia virus (vvDD) and perfused with fluorescent microspheres 120 hours after infection. Three tumors were tested. CD1 nude mice with subcutaneous SW620 human colon carcinoma tumors were treated intravenously with VSV expressing GFP and perfused with microspheres 24 hours after infection. Five tumors were tested, and data from representative tumors are shown. vvDD, double-deleted vaccinia virus.

Full figure and legend (89K)

Pro-inflammatory molecular signature in VSV treated tumors

In order to elucidate the molecular events that cause the acute loss of blood flow within the tumor, we carried out transcript profiling of tumors treated with either VSV or VSV inactivated by UV (Table 1). Most genes that are up-regulated in response to VSV therapy are nuclear factor-B-responsive, as would be expected from a natural infection.²² Using a quantitative real-time polymerase chain reaction approach (on 5 additional tumors in each group) we confirmed that a number of the genes are reproducibly up-regulated during OV therapy (see Supplementary Figure S3). A particularly interesting finding was that the transcripts encoding the neutrophil chemo-attractants CXCL1 and CXCL5 were increased 127- and 38-fold respectively, thereby suggesting that recruitment of inflammatory cells could be a key event in the process leading to loss of intra-tumor blood flow (Table 1). Quantitative real-time polymerase chain reaction experiments showed that VSV infection of CT-26 cells in vitro can also result in up-regulation of pro-inflammatory genes (e.g., CXCL1) (data not shown).

Table 1 - Genes up-regulated in CT-26 tumor 24 hours after infection with VSV, as compared with genes expressed in tumor treated with UV-inactivated VSV.

Full table

Neutrophil depletion abrogates vascular shutdown and allows for more persistent virus infection in tumors

Neutrophils are known to be one of the first cell types recruited to the sites of infections,²³ and have been implicated in causing substantial tissue damage following exposure to double-stranded RNA²⁴ or after ischemia/reperfusion.^24,25 Indeed, immunohistochemical analysis of VSV-treated CT-26 tumors demonstrated neutrophil infiltration after virus infection (Figure 5a). Others have shown that activated neutrophils can inhibit blood flow in some pathological situations.^26,27 From mice with tumors, we therefore selectively depleted this cell type by injecting rat mAb RB6-8C5 (refs. ^28,29) intraperitoneally 24 hours prior to initiating OV therapy. In these experiments we were able to demonstrate that the RB6-8C5 antibody effectively depleted neutrophils (see Supplementary Figure S4). In animals that had a normal complement of neutrophils, OV therapy resulted in loss of tumor perfusion and massive induction of apoptosis (Figures 2a and 5b). By contrast, in neutrophil-depleted animals blood flow in the tumor was not significantly interrupted, the killing of uninfected cells did not take place, and the replication and spread of the virus within the tumor were enhanced (Figure 5b). Also, in neutrophil-depleted animals, the active caspase 3 staining (signaling apoptosis) in the tumors seemed to be restricted to areas undergoing active viral infection, whereas in neutrophil-replete animals, the staining was found in areas ahead of the infection, suggesting that killing of uninfected tumor cells was in progress (compare Figures 2a and 5b). The absence of blood flow (triggered by neutrophil infiltration) is likely to restrict the spread of virus within the tumor, and emerging viruses would remain confined to the perfused tumor rim where "fertile ground" for virus replication would be quickly exhausted. These results suggest that neutrophil depletion prior to therapy may lead to a greater number of tumor cells becoming infected with virus.

Figure 5.

The killing of uninfected tumor cells requires the presence of neutrophils. (a) BALB/c mice with subcutaneous CT-26 tumors were treated intravenously with ultraviolet (UV)-inactivated vesicular stomatitis virus (VSV) particles or VSV expressing green fluorescent protein (GFP) and euthanized 24 hours later. Representative scans of tumors treated as indicated and stained for neutrophils (NIMP-R14). Scale bars, 100 m. (b) BALB/c mice with subcutaneous CT-26 tumors were pre-treated with 100 l 50:50 rat serum:phosphate-buffered saline or 7.5 g RB6 8C5 antibody intra-peritoneally 24 hours prior to intravenous treatment with VSV expressing GFP. Twenty-four hours after virus treatment, mice were perfused with fluorescent microspheres for 5 minutes and euthanized. Five mice were tested per group, and representative images are shown.

Full figure and legend (61K)

Next, mice with tumors were infected with VSV engineered to express firefly luciferase upon infection, and we carried out imaging over time using the IVIS imaging system (Xenogen, Alameda, CA). Mice with tumors were given intraperitoneal injections of RB6-8C5 antibody every other day (starting 1 day prior to VSV infection) and were imaged every day for the following 5 days. Mice that received normal rat serum showed VSV replication until day 3 after infection, while mice that were depleted of neutrophils demonstrated tumor-specific viral replication until at least 5 days after infection (Figure 6). Together, these results demonstrate that OV infection within tumors triggers the release of cytokines that attract neutrophils, leading to localized acute ischemia on account of the loss of perfusion, and resulting in tumor cell death.

Figure 6.

Neutrophil depletion allows for more persistent replication of vesicular somatitis virus (VSV). BALB/c mice with subcutaneous CT-26 tumors were pre-treated with 100 l 50:50 rat serum: phosphate-buffered saline or with 100 g RB6 8C5 antibody intra-peritoneally 24 hours prior to intravenous treatment with VSV expressing a green fluorescent protein luciferase fusion protein (Day 0). Neutrophil depletion was ensured by subsequent injection of 100 g anti-Ly6G antibody on Day 1, Day 3, and Day 5. Mice were imaged using the IVIS200 system (Xenogen, Alameda, CA) starting 24 hours after infection (Day 1) and every following day for 5 days.

Full figure and legend (100K)

Discussion

It is widely believed that OVs are effective therapeutic agents, given their ability to directly infect and kill tumor cells. There is accumulating evidence that an adaptive anti-tumor immune response triggered by virus infection contributes to in vivo efficacy. We show here that targeted activation of localized inflammation is an as-yet unexplored third component involved in tumor killing. Indeed, within 24 hours of VSV administration in vivo, viral replication can be severely restricted to the periphery of the tumor, and yet it causes massive cell death in uninfected cells. In order to generate a better understanding of the tumor environment following oncolytic virus infection, transcriptional profiling was performed, which revealed a pro-inflammatory gene signature. Because OVs are engineered to selectively infect tumor tissues they direct an inflammatory response to the tumor bed and initiate a cascade of events that culminate in tumor destruction. The impressive tumor-killing effect is the result of "a perfect storm" within the malignancy requiring the co-incident replication of the OV, secretion of pro-inflammatory cytokines, and recruitment of inflammatory cells to the tumor microenvironment.

In this study, we clearly show that neutrophils play a critically important role in the shutdown of blood flow within the tumor. During natural infections, neutrophils are the first line of defense against invading pathogens, and are often responsible for collateral tissue damage following exposure to double-stranded RNA²⁴ or hypoxia,²⁵ probably through the production of reactive oxygen species,^30,31 secretion of proteases,³² or impairment of micro-vascular perfusion.^26,27 Like other immune cells, neutrophils must make their way through the vast network of vasculature including the smallest capillaries and, in their normal state, can negotiate through microvasculature by distorting their shape. However, during inflammatory reactions, activated neutrophils adopt a "rigid" phenotype which can result in the clogging of small capillaries.³³ As tumor vasculature possesses inherently different properties from those of normal vessels,³⁴ a tortuous network of capillaries within the tumor may be more likely to act as a sink for activated neutrophils. Co-incident with the accumulation of neutrophils in the tumor, we observed a decrease in the number of neutrophils in the peripheral blood of treated mice. Although it is not unusual to see a transient loss of neutrophils from peripheral blood following infection,^35,36,37 it is reasonable to suggest that in these animals it is partly due to the sequestration of neutrophils to tumor beds. Parenthetically, in an ongoing phase I trial, human patients treated with oncolytic vaccinia virus experienced a similar transient decrease in neutrophil counts in their peripheral blood (D.K., Jennerex Biotherapeutics, personal communication, 1 November 2006) but it remains to be determined whether this was due to recruitment of neutrophils to the tumor.

The results obtained from mouse models and reported here are consistent with the observation that adoptive transfer of tumor-reactive CD8 T cells triggers granulocyte infiltration and hypoxia within tumors.³⁸ This would suggest that the described mechanism of tumor cell death may not be specific to oncolytic virus therapeutics but is applicable to all therapies that involve immune activation and rapid apoptosis in solid tumors. Interestingly, in a clinical study of an oncolytic adenovirus it was demonstrated, by means of a computed tomography scan, that virus infection triggered extensive necrosis in a liver metastasis of a colon adenocarcinoma.³⁹ In addition positron emission tomography scan images showed large areas that were metabolically inactive,⁴⁰ consistent with the idea that virus infection triggers inhibition of intra-tumor blood flow. In addition, another clinical trial studying herpes simplex virus-1 [IN HUMAN SUBJECTS?] also reported extensive necrosis in an OV-treated tumor, demonstrating histologic characteristics similar to the OV-treated tumors from mouse models in this report.⁴¹

We have found thattargeted inflammation to tumor beds leads also to the killing of uninfected cells in the tumor, probably through acute oxygen deprivation caused by a rapid loss of intra-tumor perfusion. While the stoppage of intra-tumor blood flow and the initiation of killing of uninfected tumor cells seem to provide a significant therapeutic advantage, these phenomena may also limit the spread and persistence of the virus within the tumor. The concept that immune cells tend to limit the effects of OV therapy is consistent with reports from others that chemotherapeutic agents can enhance the outcomes of viral therapy by eliminating both the adaptive and innate immune responses.⁴² However, generalized immune suppression poses risks to the patient and would prevent the generation of anti-tumor immune responses during OV therapy, of the kind noted by our group and others.^5,43 Grote et al.⁴⁴ correlated enhanced tumor regression following treatment of a granulocyte/macrophage colony stimulating factor-expressing measles virus with enhanced neutrophil recruitment to tumors. This study, however, attributed the neutrophil-mediated tumor regression to expression of the transgene, rather than to OV replication. The results presented here suggest that inflammatory immune cells can be beneficial for OV therapy, and that more refined approaches to manipulating individual cell populations will be required in order to maximize positive outcomes from oncolytic virus therapeutic regimes. Perhaps neutrophil activity could be first down-regulated so as to promote initial viral replication in the tumor, and then enhanced so as to capitalize on the ensuing tumor cell death secondary to absence of intra-tumor blood flow. In addition, measuring the blood flow in the tumor during treatment may provide an opportunity to observe tumor response to OVs. This might help in planning the dose and timing of OV administration in a multiple-dose regimen. Overall, we describe mechanisms by which inflammatory cells can be manipulated to cause enhanced and targeted tumor cell death. By investigating early events following OV infection in tumors, we have discovered a novel, targeted mechanism of tumor destruction, requiring the correct combination of virus infection, tumor microenvironment, and inflammation. It is clear that a better understanding of the direct and indirect mediators of OV tumor lysis will allow us to fine-tune therapeutic protocols to achieve the best possible outcomes for patients.

Materials and Methods

Viruses. The Indiana serotype of VSV was used throughout this study and was propagated in vero cells (American Type Culture Collection). AV1 VSV is a naturally occurring interferon-inducing mutant of VSV⁵ while 51 VSV expressing GFP⁵ and GFP-firefly luciferase fusion⁴⁵ are recombinant interferon inducing mutants of the heat-resistant strain of wild-type VSV Indiana. TP3,⁴⁶ herein referred to as AV2, was inactivated by UV light. Doubled deleted vaccinia virus expressing GFP²¹ was also propagated in vero cells. Virions were purified from cell culture supernatants by passage through a 0.2 m Steritop filter (Millipore, Billerica, MA) and centrifugation at 30,000g before resuspension in PBS (HyClone, Logan, UT).

Cell lines. CT-26 (murine colon adenocarcinoma)- and SW620 (human colon carcinoma)-derived cells were purchased from American Type Culture Collection and cultured in HyQ Dulbecco's modified Eagle medium (High glucose) (HyClone) supplemented with 10% fetal calf serum (CanSera, Etobicoke, Canada).

Tumor models. Female 6–8-week-old BALB/c mice were obtained from Charles River Laboratories (Wilmington, MA). Syngeneic subcutaneous tumors were established by injection of 3 10⁵ cells in 100 l PBS (CT-26) or 1 10⁶ cells in 100 l PBS (SW620) in the left and right hind flanks. When tumors reached a palpable size, mice were treated with VSV by tail vein injection and double-deleted vaccinia virus intra-peritoneally. Mice were sacrificed at the indicated time points by cervical dislocation and tumors were frozen in Shandon Cryomatrix freezing medium (ThermoElectron, Waltham, MA) on dry ice. Five or ten micromolar sections were cut using a Microm HM500 OM cryostat. For fluorescence imaging of intact tumors, mice were euthanized by halothane overdose. Skins were removed, and the tumors were visualized in a dissecting fluorescent microscope (Leica MZFLIII) with a standard GFP filter set. Pictures were captured with a Nikon Coolpix 100 digital camera. Fluorescent and light images were overlayed in Adobe Photoshop 7.0. Exact magnifications cannot be provided because this microscope has a continuous gradient of magnifications. All experiments were conducted with the approval of the University of Ottawa Animal Care and Veterinary Service.

Titration of VSV from mouse tissues. Tissues were removed at the indicated time points, weighed and homogenized in 1 ml of PBS using a homogenizer (Kinematica AG-PCU-11). Serial dilutions of tissue preparations were prepared in serum free media and applied to confluent Vero cells for 45 minutes. Subsequently, the plates were overlayed with 0.5% agarose in media and the plaques were grown overnight. Plaques were counted by visual inspection (between 50 and 200 plaques/plate).

Immunohistochemistry, immunofluorescence and TUNEL staining. Immunohistochemistry was performed using the Vectastain ABC kit for rabbit primary antibodies (Vector Labs, Burlingame, CA), according to instructions provided. The following primary antibodies were utilized: VSV (gift of Earl Brown), active caspase 3 (BD Pharmingen, Rockville, MD), and anti-neutrophil NIMP-R14 (Abcam, Cambridge, MA). Horseradish peroxidase activity was visualized with a Diaminobenzene-HRP kit (KPL Biosciences). Nuclei were counterstained in hematoxylin. For assessment of cell morphology, sections were stained with hematoxylin and eosin according to standard protocols. Whole tumor images were obtained with an Epson Perfection 2450 Photo Scanner while magnifications were captured using a Xeiss Axiophot HBO 50 microscope. Bound anti-VSV antibody was detected with a Cy3 conjugated donkey anti-rabbit antibody (Jackson Immunoresearch Laboratories, West Grove, PA). TUNEL staining was carried out according to manufacturer's instructions (In Situ Cell Death Detection Kit—FITC; Roche, Mississauga, Canada). Nuclei were counterstained with Hoechst #33242 (2.5 g/ml). VSV and TUNEL staining was visualized in a Zeiss Axiocam HRM Inverted fluorescent microscope and analyzed using Axiovision 4.0 software. Negative control sections were used for setting exposures, which were kept constant throughout. Neutrophil staining was captured using Aperio ScanScope (Axiovision Technologies) and analyzed using Aperio ImageScope software.

Analysis of tumor perfusion. Mice were injected intravenously with 100 l of a 50% solution of 100 nm diameter orange fluorescent microspheres (Molecular Probes, Burlington, Canada). Five minutes later, animals were sacrificed and the tumors were immediately snap frozen as previously described. Tumor perfusion was analyzed by visualizing fluorescent microspheres in the vasculature of 10 m unfixed frozen sections using a ScanArray Express microarray scanner with a standard Cy3 laser (Packard Bioscience). For in vivo Hoechst staining, mice were injected with 200 l 10 mg/kg Hoechst #33242 with 0.25 microspheres and sacrificed by cervical dislocation after 5 minutes. Tumors were immediately snap frozen as previously described and 6 m sections prepared and visualized using a Zeiss Axiocam HRM Inverted fluorescent microscope and analyzed using Axiovision 4.0 software.

Microarray. BALB/c mice with subcutaneous CT-26 tumors were treated with 5 10⁸ plaque forming units 51 VSV GFP or UV-inactivated VSV control, and tumors were harvested 24 hours later for total RNA extraction. Total RNA was isolated using the Qiagen RNeasy kit (as per manufacturer's instructions; Qiagen, Mississauga, Canada) followed by sodium acetate/ethanol precipitation to concentrate each sample. Twenty micrograms of each RNA sample was processed according to manufacturer's standard protocol (Affymetrix) and hybridized to an Affymetrix Mouse430_2 chip. Signals were normalized to the UV-inactivated VSV control sample on a gene-per-gene basis. Signals below 600 were considered absent and changes less than twofold were considered insignificant. Data were analyzed using Genespring software (SiliconGenetics).

Reverse transcription and quantitative polymerase chain reaction. CAT RNA was made by in vitro transcription with the RiboMAX Large Scale RNA Production Systems (Promega, Madison, WI) using the pCAT plasmid as template. Total tumor-or cell-derived RNA was reverse transcribed (1 or 2 g RNA) using Superscript II reverse transcriptase (Invitrogen, Mississauga, Canada) with a spike of 5 ng of CAT RNA, an exogenous control used for quantitation and normalization for reverse transcription efficiency. Quantitative real-time polymerase chain reaction was performed in triplicate on all the samples, using the Roche LightCycler rapid thermal cycler system (according to the manufacturer's instructions) and the FastStart DNA Master SYBR Green I kit (Roche Diagnostics, Laval, Canada). Standard curves were initially generated by standard dilutions and used for finding absolute values for each reaction. All values obtained were normalized to CAT values in order to normalize the reverse transcription efficiency. Primers were designed with Primer3 software. Primers used for each gene are as listed in Supplementary Table S1.

In vivo neutrophil depletion. Mice were injected intra-peritoneally with 100 g purified RB6 8C5 rat monoclonal antibody, clone RB6-8C5²⁹ (BD Pharmingen, Rockville, MD) in order to systemically deplete granulocytes. 150 l non-immune rat serum was utilized as a negative control. Twenty four hours later, mice were treated intravenously with 5 10⁸ pfu 51 VSV GFP, perfused with fluorescent microspheres after a further 24 hours, and sacrificed by cervical dislocation. Tissues were collected and stained as described above. For in vivo imaging, mice VSV GFP-luciferase-expressing mice were anesthetized with 3% isofluorane (Baxter, Deerfield, IL), injected intra-peritoneally with 2 mg luciferin (Sigma, Oakville, Canada), and imaged using IVIS 200 Imaging System (Xenogen, Alameda, CA). Data acquisition analysis was performed using Living Image v2.5 software. All images were captured under identical exposures and aperture and pixel binning settings, and bioluminescence was plotted on identical color scales.

Statistical analysis. All statistical analysis was performed using Graphpad Prism 3.0 software. Data are represented as means standard error. Virus quantifications from plaque assays were log transformed before statistical analysis and plotting.

References

REFERENCES

1. Parato, KA, Senger, D, Forsyth, PA and Bell, JC (2005). Recent progress in the battle between oncolytic viruses and tumours. Nat Rev Cancer 5: 965–976. | Article | PubMed | ISI | ChemPort |
2. Bell, JC, Lichty, B and Stojdl, D (2003). Getting oncolytic virus therapies off the ground. Cancer Cell 4: 7–11. | Article | PubMed | ChemPort |
3. Ries, SJ and Brandts, CH (2004). Oncolytic viruses for the treatment of cancer: current strategies and clinical trials. Drug Discov Today 9: 759–768. | Article | PubMed | ISI | ChemPort |
4. Smith, ER and Chiocca, EA (2000). Oncolytic viruses as novel anticancer agents: turning one scourge against another. Expert Opin Investig Drugs 9: 311–327. | Article | PubMed | ISI | ChemPort |
5. Stojdl, DF, Lichty, B, tenOever, BR, Paterson, JM, Power, AT, Knowles, S et al. (2003). VSV strains with defects in their ability to shutdown innate immunity are potent systemic anti-cancer agents. Cancer Cell 4: 263–275. | Article | PubMed | ISI | ChemPort |
6. Bauerschmitz, GJ, Guse, K, Kanerva, A, Menzel, A, Herrmann, I, Desmond, RA et al. (2006). Triple-targeted oncolytic adenoviruses featuring the cox2 promoter, E1A transcomplementation, and serotype chimerism for enhanced selectivity for ovarian cancer cells. Mol Ther 14: 164–174. | Article | PubMed | ISI | ChemPort |
7. Ye, X, Jerebtsova, M and Ray, PE (2000). Liver bypass significantly increases the transduction efficiency of recombinant adenoviral vectors in the lung, intestine, and kidney. Hum Gene Ther 11: 621–627. | Article | PubMed | ISI | ChemPort |
8. Chen, Y, Yu, DC, Charlton, D and Henderson, DR (2000). Pre-existent adenovirus antibody inhibits systemic toxicity and antitumor activity of CN706 in the nude mouse LNCaP xenograft model: implications and proposals for human therapy. Hum Gene Ther 11: 1553–1567. | Article | PubMed | ISI | ChemPort |
9. Ikeda, K, Wakimoto, H, Ichikawa, T, Jhung, S, Hochberg, FH, Louis, DN et al. (2000). Complement depletion facilitates the infection of multiple brain tumors by an intravascular, replication-conditional herpes simplex virus mutant. J Virol 74: 4765–4775. | Article | PubMed | ISI | ChemPort |
10. Jiang, H, Wang, Z, Serra, D, Frank, MM and Amalfitano, A (2004). Recombinant adenovirus vectors activate the alternative complement pathway, leading to the binding of human complement protein C3 independent of anti-ad antibodies. Mol Ther 10: 1140–1142. | Article | PubMed | ISI | ChemPort |
11. Lyons, M, Onion, D, Green, NK, Aslan, K, Rajaratnam, R, Bazam-Peregrino, M et al. (2006). Adenovirus type 5 interactions with human blood cells may compromise systemic delivery. Mol Ther 14: 118–128. | Article | PubMed | ISI | ChemPort |
12. Fisher, K (2006). Striking out at disseminated metastases: the systemic delivery of oncolytic viruses. Curr Opin Mol Ther 8: 301–313. | PubMed | ISI | ChemPort |
13. Shen, BH, Bauzon, M and Hermiston, TW (2006). The effect of hypoxia on the uptake, replication and lytic potential of group B adenovirus type 3 (Ad3) and type 11p (Ad11p). Gene Ther 13: 986–990. | Article | PubMed | ISI | ChemPort |
14. Parkins, CS, Stratford, MR, Dennis, MF, Stubbs, M and Chaplin, DJ (1997). The relationship between extracellular lactate and tumour pH in a murine tumour model of ischaemia-reperfusion. Br J Cancer 75: 319–323. | PubMed | ISI | ChemPort |
15. Liu, TC and Kirn, D (2007). Systemic efficacy with oncolytic virus therapeutics: clinical proof-of-concept and future directions. Cancer Res 67: 429–432. | Article | PubMed | ISI | ChemPort |
16. Alemany, R, Suzuki, K and Curiel, DT (2000). Blood clearance rates of adenovirus type 5 in mice. J Gen Virol 81: 2605–2609. | PubMed | ISI | ChemPort |
17. Shayakhmetov, DM, Li, ZY, Ni, S and Lieber, A (2004). Analysis of adenovirus sequestration in the liver, transduction of hepatic cells, and innate toxicity after injection of fiber-modified vectors. J Virol 78: 5368–5381. | Article | PubMed | ISI | ChemPort |
18. Springer, ML, Ip, TK and Blau, HM (2000). Angiogenesis monitored by perfusion with a space-filling microbead suspension. Mol Ther 1: 82–87. | Article | PubMed | ISI | ChemPort |
19. Martinive, P, De Wever, J, Bouzin, C, Baudelet, C, Sonveaux, P, Gregoire, V et al. (2006). Reversal of temporal and spatial heterogeneities in tumor perfusion identifies the tumor vascular tone as a tunable variable to improve drug delivery. Mol Cancer Ther 5: 1620–1627. | Article | PubMed | ISI | ChemPort |
20. Goertz, DE, Yu, JL, Kerbel, RS, Burns, PN and Foster, FS (2002). High-frequency Doppler ultrasound monitors the effects of antivascular therapy on tumor blood flow. Cancer Res 62: 6371–6375. | PubMed | ISI | ChemPort |
21. McCart, JA, Ward, JM, Lee, J, Hu, Y, Alexander, HR, Libutti, SK et al. (2001). Systemic cancer therapy with a tumor-selective vaccinia virus mutant lacking thymidine kinase and vaccinia growth factor genes. Cancer Res 61: 8751–8757. | PubMed | ISI | ChemPort |
22. Bonizzi, G and Karin, M (2004). The two NF-B activation pathways and their role in innate and adaptive immunity. Trends Immunol 25: 280–288. | Article | PubMed | ISI | ChemPort |
23. Nathan, C (2006). Neutrophils and immunity: challenges and opportunities. Nat Rev Immunol 6: 173–182. | Article | PubMed | ISI | ChemPort |
24. Londhe, VA, Belperio, JA, Keane, MP, Burdick, MD, Xue, YY and Strieter, RM (2005). CXCR2 is critical for dsRNA-induced lung injury: relevance to viral lung infection. J Inflamm (Lond) 2: 4. | PubMed |
25. Eltzschig, HK and Collard, CD (2004). Vascular ischaemia and reperfusion injury. Br Med Bull 70: 71–86. | PubMed | ISI | ChemPort |
26. Korthuis, RJ and Gute, DC (1997). Postischemic leukocyte/endothelial cell interactions and microvascular barrier dysfunction in skeletal muscle: cellular mechanisms and effect of Daflon 500 mg. Int J Microcirc Clin Exp 17 (suppl. 1): 11–17. | PubMed | ISI | ChemPort |
27. Harris, AG and Skalak, TC (1996). Effects of leukocyte capillary plugging in skeletal muscle ischemia-reperfusion injury. Am J Physiol 271: H2653–H2660. | PubMed | ISI | ChemPort |
28. Maus, U, von Grote, K, Kuziel, WA, Mack, M, Miller, EJ, Cihak, J et al. (2002). The role of CC chemokine receptor 2 in alveolar monocyte and neutrophil immigration in intact mice. Am J Respir Crit Care Med 166: 268–273. | Article | PubMed | ISI |
29. Tepper, RI, Coffman, RL and Leder, P (1992). An eosinophil-dependent mechanism for the antitumor effect of interleukin-4. Science 257: 548–551. | Article | PubMed | ISI | ChemPort |
30. Babior, BM (1984). Oxidants from phagocytes: agents of defense and destruction. Blood 64: 959–966. | PubMed | ISI | ChemPort |
31. Henson, PM and Johnston, RB Jr. (1987). Tissue injury in inflammation. Oxidants, proteinases, and cationic proteins. J Clin Invest 79: 669–674. | PubMed | ISI | ChemPort |
32. Carrell, RW (1986). 1-Antitrypsin: molecular pathology, leukocytes, and tissue damage. J Clin Invest 78: 1427–1431. | PubMed | ISI | ChemPort |
33. Burns, AR, Smith, CW and Walker, DC (2003). Unique structural features that influence neutrophil emigration into the lung. Physiol Rev 83: 309–336. | PubMed | ISI | ChemPort |
34. Hashizume, H, Baluk, P, Morikawa, S, McLean, JW, Thurston, G, Roberge, S et al. (2000). Openings between defective endothelial cells explain tumor vessel leakiness. Am J Pathol 156: 1363–1380. | PubMed | ISI | ChemPort |
35. Harlan, JM (1985). Leukocyte-endothelial interactions. Blood 65: 513–525. | PubMed | ISI | ChemPort |
36. Schleiffenbaum, B, Fehr, J, Odermatt, B and Sperb, R (1998). Inhibition of leukocyte emigration induced during the systemic inflammatory reaction in vivo is not due to IL-8. J Immunol 161: 3631–3638. | PubMed | ISI | ChemPort |
37. Werfel, T, Oppermann, M, Schulze, M, Krieger, G, Weber, M and Gotze, O (1992). Binding of fluorescein-labeled anaphylatoxin C5a to human peripheral blood, spleen, and bone marrow leukocytes. Blood 79: 152–160. | PubMed | ISI | ChemPort |
38. Blohm, U, Potthoff, D, van der Kogel, AJ and Pircher, H (2006). Solid tumors "melt" from the inside after successful CD8 T cell attack. Eur J Immunol 36: 468–477. | Article | PubMed | ISI | ChemPort |
39. Reid, TR, Freeman, S, Post, L, McCormick, F and Sze, DY (2005). Effects of Onyx-015 among metastatic colorectal cancer patients that have failed prior treatment with 5-FU/leucovorin. Cancer Gene Ther 12: 673–681. | Article | PubMed | ISI | ChemPort |
40. Au, T, Thorne, S, Korn, WM, Sze, D, Kirn, D and Reid, TR (2007). Minimal hepatic toxicity of Onyx-015: spatial restriction of coxsackie-adenoviral receptor in normal liver. Cancer Gene Ther 14: 139–150. | Article | PubMed | ISI | ChemPort |
41. Hu, JC, Coffin, RS, Davis, CJ, Graham, NJ, Groves, N, Guest, PJ et al. (2006). A phase I study of OncoVEXGM-CSF, a second-generation oncolytic herpes simplex virus expressing granulocyte macrophage colony-stimulating factor. Clin Cancer Res 12: 6737–6747. | Article | PubMed | ISI | ChemPort |
42. Fulci, G, Breymann, L, Gianni, D, Kurozomi, K, Rhee, SS, Yu, J et al. (2006). Cyclophosphamide enhances glioma virotherapy by inhibiting innate immune responses. Proc Natl Acad Sci USA 103: 12873–12878. | Article | PubMed | ChemPort |
43. Markert, JM, Medlock, MD, Rabkin, SD, Gillespie, GY, Todo, T, Hunter, WD et al. (2000). Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther 7: 867–874. | Article | PubMed | ISI | ChemPort |
44. Grote, D, Cattaneo, R and Fielding, AK (2003). Neutrophils contribute to the measles virus-induced antitumor effect: enhancement by granulocyte macrophage colony-stimulating factor expression. Cancer Res 63: 6463–6468. | PubMed | ISI | ChemPort |
45. Power, AT, Wang, J, Falls, TJ, Paterson, JM, Parato, KA, Lichty, BD et al. (2007). Carrier cell-based delivery of an oncolytic virus circumvents antiviral immunity. Mol Ther 15: 123–130. | Article | PubMed | ISI |
46. Desforges, M, Charron, J, Berard, S, Beausoleil, S, Stojdl, DF, Despars, G et al. (2001). Different host-cell shutoff strategies related to the matrix protein lead to persistence of vesicular stomatitis virus mutants on fibroblast cells. Virus Res 76: 87–102. | Article | PubMed | ISI | ChemPort |

Acknowledgments

Caroline Breitbach (Ottawa Health Research Institute) and Jennifer Paterson (Ottawa Health Research Institute) are recipients of scholarships from NSERC. This work was funded by grants from CIHR and NCIC awarded to John C. Bell (Ottawa Health Research Institute). David Kirn, Harold Atkins, John Bell, and David Stojdl are co-founders of Jennerex Biotherapeutics, a company involved in the development of oncolytic virus therapeutics.