Background
During the past several years, many reports have confirmed that intratumoral as well as systemic delivery of a variety of virus strains leads to viral replication in tumors accompanied by oncolysis of tumor cells [
1‐
3]. Most of these replicating oncolytic viruses specifically target solid tumors [
4], which is a significant advantage over the use of conventional chemo- and radiotherapy. Although oncolytic viruses are successfully used as tumor-targeting agents in animal models, the modulation of the tumor microenvironment by the viruses as well as the virus-host interaction dynamics are not well understood and therefore, the exact underlying mechanism leading to tumor elimination is less clear [
5‐
8].
Malignant tumors are complex organ-like tissues composed of ever-evolving neoplastic cells and non-neoplastic cellular components, including fibroblasts, endothelial cells and immune cells, surrounded by an extracellular matrix [
9]. These stromal components have an important function in maintaining and supporting solid tumor growth and viral infection could theoretically interfere with all of them. Moreover, viruses induce local inflammation at sites of infection leading to local remodeling of the infected tissue such as activation of the vasculature and local recruitment of immune cells. Up to date, the long-term VACV-infected tumor microenvironment is not described in the literature and the mechanism of VACV-mediated tumor regression is less clear. Theoretically, three possible mechanisms may explain virus-mediated tumor elimination - tumor cell specific oncolysis [
10], destruction of the tumor vasculature [
11,
12] followed by oxygen and nutrients deprivation, an anti-tumoral immune response [
7,
13], or a combination of these mechanisms [
14,
15]. For optimization of oncolytic virus therapy it is desired to determine which factors contribute to most optimal virus-mediated tumor regression.
Recently, Zhang et al. [
16,
17] have introduced a novel attenuated recombinant vaccinia virus GLV-1h68 and described its improved safety profile in comparison to the parental wild-type LIVP strain. Furthermore, they documented the successful application as an oncolytic agent in therapy of human breast tumor xenografts in nude mice.
In this study, we used the GLV-1h68 vaccinia virus strain to investigate the factors that may contribute to VACV-mediated tumor regression, with the final aim of improving therapeutic outcomes. We found that GLV-1h68 infection of GI-101A human breast tumor xenografts in nude mice leads to specific oncolytic destruction of the tumor tissue accompanied by tumor shrinkage. Interestingly, endothelial cells were uninfected and the vasculature remained functional. However, the tumor vasculature in infected areas strongly resembled the activated endothelium in wounded tissue, characterized by vessel dilatation, hyperpermeability and the increased expression of adhesion molecules. Furthermore, viral infection triggered increased expression of genes involved in leukocytes recruitment in the late regression phase leading to massive MHCII-positive leukocytes infiltration via the activated tumor vasculature. However, immunosuppression (MHCII+-cell depletion) of tumor-bearing, VACV-infected animals as well as the use of T-, B-, and NK-deficient mouse models for tumor growth analysis revealed that none of these immune cells are a prerequiste for VACV-mediated GI-101A tumor regression. Our results suggested that viral oncolysis is the critical factor for tumor elimination in the late regression phase mediated by VACV. We therefore propose that the most beneficial way to improve therapeutic outcomes with the oncolytic vaccinia virus GLV-1h68 strain is to enhance viral replication and spread within the tumor tissue.
Methods
Cell lines
GI-101A human ductual breast adenocarcinoma cells were kindly provided by A. Aller (Rumbaugh-Goodwin Institute for Cancer Research, Inc., FL, USA) and cultured in RPMI 1640 supplemented with 5 ng/ml β-estradiol and 5 ng/ml progesterone (Sigma Aldrich, Taufkirchen, Germany), 1 mM sodium pyruvate, 10 mM HEPES, 20% FBS, 100 Units/ml penicillin, and 100 μg/ml streptomycin (PAA Laboratories, Cölbe, Germany). African green monkey kidney fibroblasts (CV-1) were obtained from the American Type Culture Collection (ATCC-No. CCL-70) and cultured in DMEM supplemented with 10% FBS. The murine endothelial cell line 2H-11 (ATCC-No. CRL-2163) as well as mouse brain endotheliomas bEnd.3 (kindley provided by G. J. Hämmerling, Deutsches Krebsforschungszentrum, Heidelberg, Germany) were obtained in DMEM with 10% FBS. Human umbilical vein endothelial cells (HUVEC) were obtained from PromoCell (Heidelberg, Germany) and cultured in M199 medium supplemented with 10% FBS, 10 ng/ml human EGF and 50 μg/ml endothelial cell growth supplement (Sigma Aldrich). The human kidney cell line 293FT was obtained from Invitrogen GmbH (Karlsruhe, Germany) and cultured in DMEM supplemented with 10% FBS, 0.1 mM non-essential amino acids, 6 mM L-glutamine, and 1 mM sodium pyruvat. All cells were maintained at 37°C and 5% CO2.
Viruses and plasmids
The construction of the attenuated vaccinia virus strain GLV-1h68 was described previously by Zhang et al. [
16]. Briefly, three expression cassettes (encoding for
Renilla luciferase-GFP fusion protein, β-galactosidase and β-glucuronidase) were recombined into the
F14.5L,
J2R and
A56R loci, respectively, of the LIVP strain viral genome. Viruses were propagated in CV-1 cells and purified through sucrose gradients.
The RFP-expressing GI-101A cell line was constructed using the ViraPower™ Gateway Cloning and Lentiviral Expression System Kit (Invitrogen GmbH, Germany) in accordance with the manufacture's instructions. The mRFP-encoding plasmid pCR-TK-Sel-mRFP was provided by Q. Zhang (Genelux Corporation, San Diego) and used as a template for PCR amplification of the mRFP gene using primers containing attB recombination sites for gateway cloning (forward-attB1-mRFP: 5'-GGGGACAAGTTTGTACAAAAAAGCAGGCTGCCACCATGGCCTCCTCCGAGG-3', reverse-attB2-mRFP: 5'-GGGGACCACTTTGTACAAGAAAGCTGGGTCAGAATTCGCCCTTTCATTAGG-3'). The mRFP-containing lentiviral vectors were generated by gateway recombination between the pDONR-221-mRFP entry vector and the pLenti6/V5-DEST destination vector. The mRFP-containing replication-incompetent Lentiviruses for transduction of GI-101A cells were produced in 293FT cells using Lipofectamine™2000 for transfection with the ViraPower™ Packaging Mix and the pLenti6/V5-DEST-mRFP expression plasmid. Stable-expressing GI-101A-RFP clones were selected using 10 μg/ml blasticidin.
Tumor inoculation and administration of the virus
All animal experiments were carried out in accordance with protocols approved by the Regierung von Unterfranken, Germany (permit number: 55.2-2531.01-17/08).
Six-week-old female athymic nude Foxn1
nu
mice were obtained from Harlan Winkelmann GmbH (Borchen, Germany). Six-week-old female B6.12956-Rag2
tm1Fwa
N12 mice and Tac:NIHS-Lyst
bg
Foxn1
nu
Btk
xld
mice were ordered from Taconic Inc. (Hudson, NY, USA). GI-101A breast cancer cells (5 × 106/100 μl PBS) were subcutaneously (s.c.) injected into the abdominal right flank and tumor volume was calculated as (length × width2)/2. For all experiments, tumors were grown up to 200-400 mm3 in size (4-6 weeks) before viral administration. A single viral dose of 1 × 106 or 5 × 106 plaque forming units (p.f.u.) in 100 μl PBS was injected either intraveneously (i.v.) via the tail vein or via the retro-orbital (r.o.) sinus vein. For r.o. injection, animals were anesthetized using 75 mg/kg ketamine (Pfizer, Karlsruhe, Germany) and 20 mg/kg xylazine (Bayer, Leverkusen, Germany).
Immunohistochemistry
For histological studies, tumors were excised and snap-frozen in liquid N
2, followed by fixation in 4% paraformaldehyde/PBS pH 7.4 for 16 h at 4°C. Fixed tumors were rinsed in PBS and embedded in 5% (w/v) low-melt agarose (AppliChem, Darmstadt, Germany). Tissue-sectioning (100 μm) was performed using the Leica VT1000S Vibratome (Leica, Heerbrugg, Switzerland) and the labelling procedures were previously described in detail elsewhere [
18].
Fluorescence microscopy
The fluorescence-labelled preparations were examined using the MZ16 FA Stereo-Fluorescence microscope (Leica) equipped with the digital DC500 CCD camera and the Leica IM1000 4.0 software (1300 × 1030 pixel RGB-color images) as well as the Leica TCS SP2 AOBS confocal laser microscope equipped with an argon, helium-neon and UV laser and the LCS 2.16 software (1024 × 1024 pixel RGB-color images). Digital images were processed with Photoshop 7.0 (Adobe Systems, Mountain View, CA) and merged to yield overlay images.
Fluorescence intensity measurements
Fluorescence intensity of the CD31- and MHCII-labelling in 100-μm-thick Vibratome sections of control tumors and infected areas of GLV-1h68-colonized tumors was measured on digital images (× 50 objective, × 1 ocular, tissue region 2700 μm by 2150 μm) of specimens stained for CD31 or MHCII immunoreactivity. On the fluorescence microscope, the background fluorescence was set to a barely detectable level by adjusting the gain of the CCD camera before all the images were captured with identical settings. RGB-images were converted into 8-bit gray scale images (intensity range 0 - 255) using Photoshop 7.0. The fluorescence intensity of the CD31-labelling represented the average brightness of all vessel-related pixels and was measured using Image J software
http://rsbweb.nih.gov/ij. For CD31-labelling the mean value was calculated for nine images (three images of three different control and GLV-1h68-infected tumors) and presented with standard deviation.
The extent of the viral distribution in GLV-1h68-colonized tumors was measured by the GFP fluorescence signal on digital images (× 10 objective, × 1 ocular, image size 14 mm by 11.1 mm) of two whole tumor cross-sections (100 μm) of five or six different tumors. The whole area of the tumor cross-section was determined by Hoechst-labelling of cell nuclei. Both, GFP and Hoechst fluorescence images were converted into 8-bit gray scale images (intensity range 0 - 255) using Photoshop 7.0. The background fluorescence of GFP images was set to the fluorescence intensity of < 20 using Image J software. A fluorescence intensity of 20 was thus established as the threshold for distinguishing pixels of the GFP signal from those of the background. The area of pixels (inch2) on GFP images (fluorescence intensity > 20) as well as on Hoechst images (fluorescence intensity > 0) was measured by Image J and the proportion of infected tissue was calculated for two images from each tumors (n = 6). Mean values + standard deviations are shown.
Measurements of microvessel density and vessel diameter
The vascular density was determined in microscopic images (× 200 objective, × 1 ocular, tissue region 680 μm by 540 μm) of CD31-labelled tumor sections. On the fluorescence microscope, for each image the CD31 fluorescence was set to a clearly detectable level by individually adjusting the gain of the CCD camera before the images were captured. All images were decorated with five horizontal lines at identical positions using Photoshop 7.0 and all vessels which intersected these lines were counted to yield the vascular density. The vascular density was calculated for nine images (three images of three different control and GLV-1h68-infected tumors) and presented as mean values with standard deviations.
The vessel diameter was measured on digital images (× 200 objective, × 1 ocular) of CD31-labelled 100-μm-thick tumor cross-sections using Leica IM1000 4.0 software. Images of control and infected tumors (GLV-1h68-infected area) were obtained with individual exposure times to get optimal CD31 signals and exclude signal-dependent variability of vessel diameter. Seven horizontal lines were drawn across each image and the diameter of all blood vessels that intersected these lines was measured (5 images per tumor). Mean values + standard deviations are shown.
Antibodies, reagents and treatment of animals
Endothelial cells were labelled with monoclonal rat anti-mouse CD31 antibody (BD PharMingen, San Diego, CA) or hamster anti-mouse CD31 antibody (Chemicon, International, Temecula, CA). Pericytes were labelled with Cy3-conjugated monoclonal mouse anti-mouse α-smooth muscle actin (SMA) (Sigma Aldrich). Basement membrane was labeled using polyclonal rabbit anti-mouse collagen IV antibody (Abcam, Cambridge, UK). Immune cells were labeled using rat anti-mouse MHCII antibody (B, dendritic cells, monocytes, macrophages) and rat-anti mouse CD45 antibody (common leukocyte antigen) (eBioscience, San Diego, CA).
The Cy3- or Cy5-conjugated secondary antibodies (donkey) were obtained from Jackson ImmunoResearch (West Grove, PA).
Phalloidin-TRITC (Sigma Aldrich) was used to label actin and Hoechst 33342 to label nuclei in tissue sections.
For the labelling of functional blood vessels in tumors, mice were anesthetized using 75 mg/kg ketamine and 20 mg/kg xylazine, followed by the injection of 100 μg of biotinylated-Lycopersicum esculentum lectin (Vector Laboratories, Burlingame, CA) via the tail vein of the mice. Two minutes later the chest was opened, and the vasculature was perfused at a pressure of 120 mmHg with fixative (4% paraformaldehyde/PBS pH 7.4) from a cannula inserted into the left ventricle. After fixation, tumors were removed and prepared for histology. Tumor cross-sections (100 μm) were labelled with Cy3-conjugated streptavidin (Sigma Aldrich) to visualize the lectin-labelled tumor vasculature.
Nonspecific rat-IgG from Jackson ImmunoResearch was used in extravasation studies and injected intravenously into tumor-bearing mice (11 mg/kg body weight). After 6 h incubation, the treated tumors were excised and used for histological analysis. Surface plot profiles of the IgG extravasation pattern were prepared using ImageJ software.
For immunosuppression a stock solution of cyclophosphamide monohydrate (42 mg/ml) (Sigma Aldrich) was prepared in water and sterile filtered. Immediately before use, the stock solution was diluted 1:1 in 1.8% NaCl to yield a final concentration of 21 mg/ml. CPA was administered by intraperitoneal injection twice per week throughout the entire duration of the study. The treatment was started 10 days p.i. with an initial dose of 140 mg/kg body weight followed by 100 mg/kg body weight. The dose and schedule was based on previously published studies of CPA immunosuppression in mice and hamsters [
19].
Viral replication in vitro
For viral replication assays, tumor cells as well as endothelial cells were seeded in triplicates into 24-well plates to reach a confluency of 80% after a culture period of 12-16 h. Before infection, cell layers were starved with individual starvation media containing 1% FBS for 24 h and were finally infected with GLV-1h68 at m.o.i. of 0.01. After 1 h of incubation, the infection medium was replaced by fresh starvation medium and cells were cultured for further 6, 24 and 48 h, respectively. At the indicated time points, cells and supernatants were harvested and after three thaw-freeze cycles, serial dilutions of the lysates were titered by standard plaque assays on CV-1 cells.
Co-culture experiments
To mimic in vivo conditions, we cultured endothelial cells on growth factor reduced Matrigel Matrix (BD Biosciences, Heidelberg, Germany), which is a soluble basement membrane extract. For co-culture experiments, we coated 24 well plates with 100 μl of Matrigel for 30 min at 37°C. Endothelial cells (1 × 104 cells/well) were seeded into 24 well plates and allowed to assemble into tube-like structures. Three hours later GI-101A-RFP tumor cells were seeded into these wells and co-cultures were incubated for 12-15 h. Co-cultures were infected with GLV-1h68 at m.o.i. of 0.5 for 1 h, before the medium was replaced with virus-free medium. The degree of infection was microscopically determined after 24 h.
Total RNA from both infected and uninfected GI-101A xenografts at days 21 and 42 post VACV infection was extracted using Trizol reagent (Sigma Aldrich) according to the manufacturer's instructions. Total RNA was amplified into anti-sense RNA (aRNA) as previously described [
20,
21] and the quality of both, total RNA and secondarily amplified RNA was tested with the Agilent Bioanalyzer 2000 (Agilent Technologies, Palo Alto, CA). Confidence about array quality was based on the principle of reference concordance as previously described [
22]. Mouse reference RNA was prepared by homogenization of the following mouse tissues (lung, heart, muscle, kidneys and spleen) and RNA was pooled from 4 mice. Pooled reference and test aRNA was isolated and amplified in identical conditions during the same amplification/hybridization procedure to avoid possible inter-experimental biases. Both, reference and test aRNA was directly labeled using ULS aRNA Fluorescent labeling Kit (Kreatech, Netherlands) with Cy3 for reference and Cy5 for test samples.
Whole genome mouse 36 k oligo arrays were printed in the Infectious Disease and Immunogenetics Section of the Department of Transfusion Medicine (IDIS), Clinical Center, National Institute of Health, Bethesda using oligos purchased from Operon (Huntsville, AL). The Operon Array-Ready Oligo Set (AROS™) V 4.0 contains 35,852 longmer probes representing 25,000 genes and about 38,000 gene transcripts and also includes 380 controls. The design is based on the Ensembl Mouse Database release 26.33b.1, Mouse Genome Sequencing Project, NCBI RefSeq, Riken full-length cDNA clone sequence, and other GenBank sequence. The microarray is composed of 48 blocks and one spot is printed per probe per slide. Hybridization was carried out in a water bath at 42°C for 18-24 hours and the arrays were then washed and scanned on a Gene Pix 4000 scanner at variable PMT to obtain optimized signal intensities with minimum (< 1% spots) intensity saturation.
Resulting data files were uploaded to the mAdb databank
http://nciarray.nci.nih.gov and further analyzed using BRBArrayTools developed by the Biometric Research Branch, National Cancer Institute [
23]
http://linus.nci.nih.gov/BRB-ArrayTools.html and Cluster and Treeview software [
24]. The global gene-expression profiling consisted of 16 experimental samples. Global expression data were filtered using automated filtering option of BRBArray software. Therefore, genes involved in pathways such as "adhesion molecules on lymphocytes, B lymphocytes cell surface molecule, cytokines and inflammatory response, monocytes and its surface molecules, neutrophiles and its surface molecules, T cytotoxic cell surface molecules, T helper cell cytotoxic molecules" as listed by the Biocarta database were included. Genes that belonged to at least one of those pathways, that were present in more than 10 experimental samples (≥ 60%) and with a fold change of two in at least one sample passed the filter. Gene ratios were average corrected across experimental samples. Subsequent cluster analysis applying uncentered correlation algorithm with genes involved in selected pathways as listed above allowed experimental samples to cluster according to their biological similarity. Treeview program was used for visualization of array data [
25].
Statistics
A two-tailed Student's t test was used for statistical analysis. P values of < 0.05 were considered statistically significant.
Discussion
The mechanisms involved in tumor regression during oncolytic therapy are still a matter of heated debate. They may naturally vary among different tumor models and/or be dependent on different oncolytic virus strains used. Predominantly, oncolytic viruses are effective therapeutic agents due to their ability to efficiently infect and destroy cancer cells. In addition, replicating viruses may interfere with components of the tumor microenvironment such as the tumor vasculature and the immune system of the host. Therefore, oncolytic tumor destruction may be a multi-step process, in which the different components work with or against each other. In this study, we show that the oncolytic VACV GLV-1h68 drastically interfered with host components such as the tumor vasculature and induced also a massive innate immune response. But the predominant mechanism which leads to regression of tumors, in this model at late stage of tumor regression, was found to be direct viral-infection-mediated tumor cell destruction.
In contrast, Breitbach et al [
32] showed that a significant portion of the tumor killing activity of vesicular stomatitis virus and VACV in the murine CT-26 colon cancer model is caused by indirect killing of uninfected tumor cells. In this tumor model the authors suggest that massive neutrophil activation followed by vascular damage and apoptosis of uninfected tumor cells one day after infection are the main cause of tumor cell destruction. However, the tumor killing may be caused by the high viral titer (1 × 10
9 pfu) used in the study to infect tumor-bearing animals. Further, the large amount of activated neutrophils may occlude the abnormal tortuous tumor vessels and the oxitative burst may directly destroy endothelial cells leading to vascular shutdown followed by apoptosis of surrounding tumor cells. In contrast we showed, that tissue necrosis exactly colocalizes with GFP-expressing VACV-infected tumor cells in human breast tumor xenografts. Interestingly, the infected, necrotic tumor areas remain highly vascularized in late infection stages (42 dpi). Therefore, neither vascular shutdown nor other indirect killing activities seem to occur and tumor destruction parallels with the site of viral replication and spreading.
Destruction of the tumor vasculature seems to be, in general, a promising strategy to induce tumor shrinkage by deprivation of nutrients and oxygen. In this regard, the oncolytic viruses itself may target, infect and destroy the tumor vasculature. For example, Kirn et al. [
14] showed that an oncolytic vaccinia virus mutant strain with
B18R deletion infects and destroys tumor-associated vascular endothelial cells. In contrast, we did not find any infected endothelial cells in tumors nor did the GLV-1h68 strain show general replication in murine or in human endothelial cell lines. On the other hand, combination of oncolytic viruses with anti-angiogenic therapy also appeared to enhance virotherapy, possibly due to the stabilization of the tumor vasculature or reducing the neovascular responses associated with viral replication [
33‐
35]. Numerous therapy approaches are described using oncolytic viruses in combination with anti-angiogenic molecules to potentiate tumor destruction [
11,
12,
34,
36]. Beside the widely discussed starvation effects exhibited by angiogenesis-inhibitors, the destruction or normalization of the tumor vasculature may also have effect on the host immune response, due to a reduced infiltration of innate immune cells in the infected tumor tissue, which in turn affects virus survival.
In general, the vascular endothelium regulates innate and adaptive immune responses by controlling the extravasation of leukocytes from the blood into the inflamed tissue [
27,
29]. During inflammation, these patroling leukocytes extravasate into the tissue via a sequential process that is initiated by their adhesion to endothelial cells, which in turn leads to endothelial cell activation [
29]. The activated endothelium is characterized by vascular hyperpermeability and increased tissue edema, which in turn facilitates perivascular inflammatory cell infiltration. The tumor vasculature in GLV-1h68-infected areas strongly resembles the activated endothelium in inflamed tissue, which is supported by the increased expression of CD31, the up-regulation of genes involved in leukocytes recruitment as well as the observed vasodilatation and hyperpermeability. However, the inflammatory response of the endothelium remains mainly localized in and directly around viral patches within the tumor tissue. At least, the encapsulation of viral patches by CD45- as well as MHCII-positive leukocytes indicated that the viral-induced immune response is mainly restricted to the invader and seems not to be a general anti-tumoral response in the late tumor regression phase.
Detailed confocal analysis of the tumor vasculature in late-stage infected tumors revealed a cell population, not yet described in the context of oncolytic tumor therapy, which coexpresses endothelial (CD31) and dendritic cell markers (CD45, MHCII) [
37]. Recently, Conejo-Garcia et al. [
37] described this novel leukocyte subset within ovarian carcinomas and attributed this cells to have the capacity to generate functional blood vessels in tumors. During oncolytic tumor therapy, this cell population may stabilze or partly restore the tumor vasculature within the infected, necrotic tumor areas, which represent truly unfavourable conditions for cell survival.
The observed phenomenon of vasodilatation as well as the increased permeability could also be enhanced by necrosis of the vessel-surrounding area, including destruction of pericytes, which reduces mechanical stress on the tumor vasculature leading to decompression of tumor vessels. Recently, Padera et al [
38] showed, that tumor-specific cytotoxic therapy results in more efficient drug delivery by decompressing collapsed vessels. Therefore, we suppose that oncolytic VACV could be used as a "natural enhancer" of chemotherapy by improving the intratumoral dissemination of chemotherapeutics.
The relative importance of direct oncolysis versus immune-mediated tumor regression remains in most of the animal models uncertain. In general, xenografts are considered as chronically inflamed and do not by themselves provide sufficient signals to induce an acute inflammation leading to immune-mediated rejection of the tumor [
39]. The generation of an effective anti-tumoral immune response depends on danger signals within the tumor, and infectious agents inherently offer these sufficient signals [
40,
41]. Therefore, viral infection of the tumor tissue may generate the essential trigger to alter the immune milieu of the tumor microenvironment. To elucidate whether the initiated immune response is directed against the viral invader or against the tumor tissue, we used the previously reported immunosuppressive agent CPA [
19,
31] to deplete virus-induced intratumoral immune cell recruitment. The tumor growth curve analysis showed no significant difference in the tumor growth characteristics between CPA-treated and CPA-untreated VACV-infected GI-101A tumors. However, we propose here that in general, factors such as viral distribution/degree of oncolysis or extent of tissue destruction/necrosis and correlation e.g. to the recruitment of immune cells offer much more information about therapy success than the tumor volume alone, because the tumor volume that was measured could be either necrotic debris or host cells of the tumor microenvironment instead of cancer cells. Our study showed that the extent of viral distribution/necrotic tissue destruction increased in CPA-treated, VACV-infected tumors and correlated with reduced intratumoral MHCII-positive immune cell infiltration compared to untreated, VACV-infected tumors. Although there is no significant change in the tumor growth characteristic upon immunosuppression, the significant higher extent of viral distribution/necrosis in these animals indicates that MHCII-positive immune cells impede viral-mediated tissue destruction. According to data shown here, immunosuppression increases oncolytic effects of a herpes simplex virus-derived OV [
42] as well as an oncolytic adenovirus [
19] via enhancing intratumoral viral spread. Further, Fulci et al [
42] showed via clodronate liposome-dependent depletion of phagocytic cells, that mainly cells of the monocytic origin are responsible for clearance of intratumoral viral particles.
The immunolgical response against GLV-1h68, however, is not effective enough to eliminate the virus from the tumor tissue. This may be due to the wealth of immune evasion mechanisms presented by vaccinia virus [
43]. Furthermore, the recruited immune cells may not as cytotoxic as usual due to the local immunosuppressing tumor microenvironment and can only slow down viral spread but not eliminate the viral infection focus. Alternatively, these recruited immune cells, which encapsulate viral patches, form an anatomical barrier, which may not be overcome by the virus due to their lesser susceptibility to VACV infection.
Previously, we have shown by transcriptional profiling of different tumor models that oncolytic GLV-1h68 infection does induce strong pro-inflammatory signatures [
44,
45]. Recently, Wang et al. [
39] postulated that viral infection of the tumor tissue generates the necessary trigger that activates an acute immune response against the tumor tissue. However, the here described data revealed, that the immunological response in GLV-1h68-mediated GI-101A tumor destruction represents indeed an acute inflammation, but this response is only directed against the "pathogen" and not against the tumor tissue. We could show here that neither MHCII-positive immune cells nor T-, B-, or NK cells contribute significantly to VACV-mediated tumor regression. Collectively, these data support the primarily oncolytic character of VACV-mediated tumor destruction and demonstrate that the activation of a direct anti-tumoral immune response was not critical for tumor regression in this tumor model.
Competing interests
This work was supported by grants from Genelux Corporation (R&D facility in San Diego, CA, USA). SW received a postdoctoral fellowship, VR and AW received a graduate fellowship by Genelux Corporation awarded to the University of Wuerzburg, Germany. YAY and AAS are salaried employees of Genelux Corporation and have personal financial interests in Genelux Corporation.
Authors' contributions
SW conceived and designed the study, performed experiments, analyzed the data, and wrote the manuscript. VR helped to perform some experiments and analyzed data. YAY conceived, designed and carried out experiments with immunodeficient mouse strains, and helped to draft the paper. AW, EW, FMM conceived, designed and carried out the microarray analysis. AAS conceived the study, and participated in the design and coordination and helped to draft the manuscript.
All authors read and approved the final manuscript.