Hereditary breast and ovarian cancers (HBOC) represent 5-10% of breast and 10-15% of ovarian cancer cases. These cancers tend to be aggressive and curative treatment strategies are scarce. Poly(ADP-ribose) polymerase inhibitors (PARPi), a family of drugs that inhibit DNA repair, are a promising therapy for cancers harbouring mutations in their DNA repair machinery, such as HBOC. Unfortunately, nearly all patients ultimately become resistant to PARPi, leaving limited options for definitive treatment. Oncolytic or “cancer-killing” viruses are an innovative immunotherapeutic platform capable of selectively targeting cancer cells, leaving healthy tissues unharmed. Our group has demonstrated that oncolytic rhabdoviruses may be used to deliver therapeutic payloads by encoding targeting sequences to act on genes via RNA interference. In the present work, I have engineered the oncolytic virus, vesicular stomatitis virus (VSV), to express a variety of microRNA (miRNA) sequences that target genes essential for DNA repair, sensitizing resistant cancer cells to PARPi therapy. After initial experiments revealed hurdles concerning the functionality of artificial miRNAs which specifically target BRCA1 and BRCA2 I encoded the naturally occurring hsa-miR-182 into VSV to knockdown BRCA1 and additional genes essential for DNA repair. Using a 3D spheroid model, I have demonstrated sensitization of initially resistant MDA-MB-231 breast cancer cells to the PARPi, rucaparib. Complementary work exploring the shuttling of miRNAs into small extracellular vesicles, or EVs, has also shown that we can take advantage of the EV packaging facilities in infected cells, inducing the packaging of miRNAs over-expressed by VSV (EV-miRNAs) into EVs. Future work will address the functionality of these EV-miRNAs, testing their ability to knockdown targets in uninfected cancer cells.

Survival for patients with advanced cancer has remained dismal, and there is a need for new treatments. In this context viral immune therapy is a promising novel strategy. Intravenous delivery confers advantages as it enables simultaneous treatment of primary tumour and any metastatic deposits but host defences limit Vaccinia virus's (VV) ability to infect tumour after systemic administration. Although Vaccinia virus can potentially be delivered systemically as it can evade both complement and neutralising antibodies, our investigations have revealed that VV cannot effectively infect tumour cells in immunocompetent mice after systemic delivery. Strikingly, we observed that if macrophages were depleted in the mice using clodronate liposomes, VV infection of tumours was dramatically enhanced. However, clodronate liposomes non-selectively deplete macrophages and potentially diminish any beneficial macrocytic activity in the tumour microenvironment unrelated to viral clearance. Consequently, a more clinically appropriate agent is needed. Macrophages recognise and ingest pathogenic microorganisms through phagocytosis, a process for which several lines of evidence have highlighted an important role for phosphatidylinositol 3-kinases. Accordingly, in these investigations I have evaluated the effect of selective PI3K inhibitors on macrophage phagocytosis in vitro and demonstrated that IC87114 (a PI3 kinase delta inhibitor) is effective at reducing uptake of VV by macrophages, confirming this finding in transgenic macrophages with a mutant of the PI3 kinase delta isoform knocked in. Subsequently, it was confirmed that IC87114 affects attachment of the virus to macrophages but plays no role in internalisation of the virus. In cancer cells cultured in isolation, the inhibitor has no direct cytotoxic effect and when combined with VV, in the same in vitro system, there is no change in the amount of cell death compared to VV alone treated controls. Biodistribution studies have established that IC87114 combined with VV results in statistically significantly higher levels of virus detected in tumours compared to the groups treated with VV alone, with similarly limited off-target effects. Finally, three different efficacy studies have demonstrated statistically significantly superior tumour responses in the VV+IC87114 group. In conclusion, PI3k delta blockade is an effective strategy for enhancing systemic delivery of VV in a preclinical model and could be a useful adjuvant in VV clinical trials.

Oncolytic virotherapy with the adenovirus mutant dl 1520 in combination with chemotherapy has shown clinical response. Approaches to cancer gene therapy involving delivery of enzymes to activate the prodrugs CB 1954 and 5-FC are currently being tested in clinical trials. We hypothesised that the combination of an adenoviral vector equivalent to dl1520 with activation ofCB 1954or 5-FC and the combination of CB 1954 activation with 5-FU may further improve the antitumour effects for colorectal cancer therapy. The initial in vitro data showed that the combination of dl 1520 with CB 1954 activation or 5-FU (metabolite of 5-FC activation) and the combination of CB 1954 activation with 5-FU led to an additive or synergistic cytotoxicity. Subsequent data showed that the incorporation of Ntr or CD-UPRT genes into replicating oncolytic adenoviruses (ROAds) resulted in enhanced Ntr expression or CD-UPRT activity and augmented cytotoxic effects in tissue culture, surpassing the levels and cytotoxic effects mediated by the corresponding replication-defective vectors. When tested in subcutaneous human colon cancer xenografis, Ntr expression mediated by the ROAd was apparently higher than the level mediated by replication-defective CTLI02. Importantly, the antitumoural efficacy of CB 1954 activation mediated by ROAd is significantly superior to that mediated by CTLI 02 (p = 0.01). The ROAds displayed viral replication and oncolysis in vitro and in vivo and these attributes can contribute to the increased gene expression level and enhanced efficacy. Overall, the data suggested that the use of ROAds improved Ntr or CD-UPRT expression and antitumoural efficacy in the presence of corresponding prodrugs and may have the potential to achieve clinical significance in the treatment for colorectal cancer.

Conditionally-replicating adenoviruses (CRAds) have generally demonstrated only modest therapeutic efficacy in human clinical trials, in part due to their poor ability to spread throughout a tumor mass. In these studies, I first examined whether inclusion of an intact early region 3 (E3) and the p14 fusion-associated small transmembrane (FAST) protein in a CRAd vector can enhance oncolytic efficacy by improving viral spread. E3 encodes the adenovirus death protein (ADP), which enhances virus progeny release from infected cells, while p14 FAST can allow spread of the virus through cell-cell fusion. I generated viruses with (CRAdRC109) or without (CRAdRC111) an intact E3 region, which encoded the p14 FAST gene between the fiber coding sequence and E4 region of their viral genomes. In the A549 human lung cancer cell line, both CRAdRC109 and CRAdRC111 expressed p14 FAST at very low levels when compared to CRAdFAST, a similar virus that expressed the protein from within the E3 deletion, and thus had a relatively poor ability to mediate cell-cell fusion. Although inclusion of E3/ADP in CRAdRC109 did result in larger plaques and increased virus spread relative to CRAdRC111, neither virus showed improved oncolytic activity relative to CRAdFAST. I subsequently developed CRAdRC116, in which the E3 region of the viral genome was replaced with a bicistronic expression cassette containing the p14 FAST and ADP coding sequences separated by a self-cleaving 2A peptide sequence. This virus co-expressed p14 FAST and ADP and caused extensive cell-cell fusion in A549 cells. However, expression of ADP from CRAdRC116 did not increase cancer cell killing nor virus spread, and thus did not enhance oncolytic efficacy relative to CRAdFAST. These studies suggest that p14 FAST and ADP do not exhibit synergy when co-expressed from a CRAd vector. Future studies should instead focus on combining other methods of improving viral spread in conjunction with expression of ADP or FAST proteins from CRAd.

Oncolytic viruses (OVs) have begun to show their promise in the clinical setting, however these results have been predominantly associated with loco-regional administration of virus. The treatment of metastatic disease necessitates a systemic approach to virus delivery. The circulatory system, though, is a hostile environment for viruses and the advantages associated with intravenous (IV) delivery come at a heavy cost that must be understood and brokered. Pre-existing immunity, specifically through the function of antibody and complement, poses a significant hurdle to the IV delivery of infectious virus to dispersed tumor beds. This is of particular importance for therapeutic vaccinia viruses as a majority of today’s cancer patients were vaccinated during the smallpox eradication campaign. In vitro neutralization assays of oncolytic vaccinia virus demonstrated that the antibodies elicited from smallpox vaccination, and also the anamnestic response in patients undergoing Pexa-Vec treatment, was minimally neutralizing in the absence of functional complement. Accordingly, in a Fischer rat model, complement depletion stabilized virus in the blood of pre-immunized hosts and correlated with improved delivery to mammary adenocarcinoma tumors. Complement depletion additionally enhanced infection of tumors following direct intratumoral injection of virus. The feasibility and safety of using a complement inhibitor, CP40, was tested in a cynomolgus macaque model. Immune animals saw an average 10-fold increase in infectious virus titer at an early point after the infusion, and a prolongation of the time during which infectious virus was still detectable in the blood. We have also demonstrated that vaccinia virus engages in promiscuous interactions with cells in the blood and that these interactions may be partially complement-dependent. Additionally, we have translated this complement inhibition approach to other OV candidates and found that reovirus, measles virus and a virus pseudo typed with the LCMV glycoprotein all elicit antibodies, that to some degree, are dependent on complement activation to neutralize their target viruses. We show here that capitalizing on the complement dependence of anti-viral antibody with adjunct complement inhibitors may increase the effective dose to enable successful delivery of multiple rounds of OV in immune hosts.

Oncolytic viruses (OVs) are an emerging cancer therapy that use an oncotropic virus to selectively infect and kill cancer cells, as well as stimulate long-lasting anti-tumor immune responses. In order to achieve high therapeutic efficacy, OVs need sufficient replication within the tumor tissue to mediate these effects. However, OV’s infectivity varies between different tumors and the host’s immune system can rapidly clear the virus, hampering treatment efficiency. Oncolytic virus sensitizers are chemical compounds that specifically enhance OV’s infectivity and efficacy. In our lab, I found that treatment of various cancer cell lines with BI-D1870, a pan-RSK (ribosomal S6 kinase) inhibitor, resulted in augmented Herpes Simplex Virus-1 (HSV1) and Vesicular Stomatitis Virus (VSVΔ51) infectivity. I also demonstrated that the effects of BI-D1870 on viral infection are virus-specific, and that RSK inhibition is not the primary target causing the enhancement of HSV1 and VSVΔ51 infection. Finally, BI-D1870 structural analogs were generated in an attempt to enhance the efficacy and selectivity of BI-D1870-based OV sensitizers. One of the analogs synthesized, KA-019, showed an improvement in the augmentation of OV infection over BI-D1870. As a genetically engineered strain of HSV1 has been approved by FDA for treatment of melanoma, the results of my project propose a novel viral sensitizer to improve viral replication within tumour cells with the hope of improving therapeutic efficacy.

Acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) are blood cancers that are often treated with stem cell transplantation (SCT). Since SCT treatments have variable success, especially in adults with AML whose disease frequently relapses, novel and more effective solutions must be considered. In this thesis, I will explore one type of immunotherapy in murine models for ALL (L1210) and AML (C1498) using in vitro and in vivo techniques such as flow cytometry and transcriptomics. In my approach, I am attempting to enhance the immunogenicity of whole cell vaccines by pre-infecting the leukemia cells with oncolytic virus (OV) and thus producing leukemia infected cell vaccines (ICVs). While it has been previously shown that L1210-ICV pre-treatment works well in protecting mice from ALL challenge, I have found that pre-immunization with C1498-ICV has a limited efficacy in protecting animals from AML progression. By investigating the downstream effects of ICV, I was able to show that unlike C1498 cells, L1210 cells produce previously unknown immunogenic factors following OV infection.

Vaccinia virus has a large and still incompletely understood genome although several strains of this virus are already in clinical development. For the most part, clinical candidates have been attenuated from their wild type vaccine strains through deletion of metabolic genes like the viral thymidine kinase gene.In the present work, we thoroughly examined the genetic elements of vaccinia which could be modulated to tailor the virus as a cancer therapeutic. Using a variety of cancer cell lines and primary tumor explants, we performed a fitness assay that directly compares multiple wild-type Vaccinia strains to identify the genetic elements that together create an optimal “oncolytic engine”. Using a transposon insertion strategy and deep sequencing of viral populations we systematically examined Vaccinia genes that do or do not play a role in the therapeutic activity of the virus. Our studies allowed us to identify a variety of genes in the vaccinia genome that when deleted, augment the oncolytic activity of a newly engineered Vaccinia virus. In the context of this thesis, I define enhanced oncolytic activity as superior therapeutic activity, increased immunogenicity and an improved safety profile, all aspects which we used to compare this novel virus to Vaccinia viruses currently in the clinic.

Oncolytic viruses (OV) are a promising alternative cancer therapy due to their specificity and lack of debilitating side effects, such as those which typically accompany conventional therapeutics such as chemotherapy and radiation. Bovine herpesvirus type 1 (BHV-1) is an alphaherpesvirus with the ability to infect and kill multiple human tumor cell types. In comparison to other species-specific viruses, for which deficiencies in type I interferon signalling pathways dictates cellular sensitivity to infection, mutations in KRAS were found to correlate with high levels of BHV-1 replication. Interestingly, BHV-1 is able to induce cellular cytotoxicity in the absence of a productive infection. In contrast to current breast cancer (BC) treatments, which are largely based on receptor expression status, BHV-1 is able to infect and kill BC cells and breast cancer initiating cells (BCICs) from luminal and basal subtypes. Furthermore, BHV-1-infected BC cells are significantly diminished in their capacity to form tumors in vivo, suggesting that BHV-1 reduces the tumor forming capacity of BCICs. Combination therapy involving OVs has been used to exploit differences in the mechanism of tumor cell death elicited by individual treatments. Treatment with epigenetic modifiers such as 5-Azacytidine (5-Aza), a DNA methyltransferase inhibitor, has been shown to increase the antitumor activity of OVs. Our data indicates that 5-Aza strongly synergises with BHV-1, increasing virus replication and cytotoxicity in vitro. In vivo, BHV-1 monotherapy did not significantly impact tumor growth or survival of CR bearing subcutaneous breast tumors; however, combination therapy with 5-Aza significantly decreased the number of secondary lesions compared to BHV-1 monotherapy. Overall, the data presented in this dissertation indicate that BHV-1 is a promising broad spectrum OV with a unique mechanism of tumor cell targeting, and the ability to infect and kill tumor cells independent of a productive infection. / Thesis / Doctor of Philosophy (Medical Science)

Oncolytic virotherapy is emerging as a new treatment option for cancer patients. At present, there are relatively few oncolytic virus clinical trials that are underway or have been conducted, however one virus that shows promise in pre-clinical models is Vesicular Stomatitis Virus (VSV). VSV is a naturally occurring oncolytic rhabdovirus that has the ability to preferentially replicate in and kill malignant versus normal cells. VSV also has a low seroprevalence, minimal associated morbidity and mortality in humans, and simple non-integrating genome that can be genetically manipulated, making it an optimal oncolytic vector. Currently, many labs are using a variety of different strategies including inserting trans genes that can modulate the innate and adaptive immune response. VSV can also be retargeted by altering its surface glycoprotein (G) or be made replication incompetent by deleting the G protein. Currently, our lab has engineered a series of new recombinant VSVs, incorporating either the murine p53 (mp53), IPS-1, or TRIF transgene. mp53, IPS-1 and TRIF were incorporated into the normal VSV-XN2 genome and mp53 was also incorporated into the mutated VSV-ΔM vector generating VSV-mp53, VSV-IPS-1, VSV-TRIF and VSV-ΔM-mp53. Our data using these new viruses indicate that these viruses preferentially replicate in and kill transformed versus non-transformed cells and efficiently express the transgene. However, despite the ability for VSV-IPS-1 and VSV-TRIF to induce a robust type 1 IFN response, VSV-ΔM-mp53 was the only construct that had reduced toxicity and elicited an increased anti-tumor response against a syngeneic metastatic mammary tumor model. VSV- ΔM-mp53 treatment lead to a reduction in IL-6 and IP-10 production, an increase in tumor specific CD8+ T cells, and immunologic memory against the tumor. Collectively these studies highlight the necessity for additional VSV construct development and the generation of new clinically relevant treatment schema.