Pancreatic cancer models exhibit interferon-independent resistance mechanisms to a Morreton hybrid oncolytic vesiculovirus

The five-year survival rate for pancreatic ductal adenocarcinoma (PDAC) remains only 10%, urging the development of novel therapeutics and combination therapies. Immune checkpoint blockade (ICB) and virotherapy have improved survival for a subset of cancer patients in recent years. However, extending the benefit to patients with PDAC has lagged. Prior studies have highlighted the role of intact type I interferon (IFN) signaling in mediating PDAC resistance to oncolytic vesicular stomatitis virus (VSV). Despite this, ongoing clinical trials utilize oncolytic VSV encoding interferon beta (IFNβ) to increase safety, while simultaneously limiting viral efficacy. Due to safety concerns regarding wild-type VSV’s ability to infect neurons, we engineered a hybrid vesiculovirus replacing the viral surface glycoprotein (G protein) with that of Morreton virus, shown to be well-tolerated in immunocompetent mice. We characterized the oncolytic properties of the hybrid virus - VSV expressing the Morreton G protein (VMG) - in PDAC cell lines, patient-derived organoids, and a syngeneic murine model. Cell lines were variably susceptible to VMG-induced cell death in vitro, independent of type I IFN or infectious viral particle production. We evaluated resistance signatures using Significance Analysis of Microarrays and RNA-Seq (SAMseq) and quantitative proteomics. Differential gene expression of sensitive and resistant cell lines showed higher expression of HAS3 at baseline in resistant cell lines. We also performed quantitative proteomics of infected versus control sensitive (MIA-PaCa2) and resistant (CFPAC-1) PDAC cell lines. Overall, fewer proteins were differentially expressed in CFPAC-1 after infection, but we detected all five structural proteins of VMG in infected samples. IFN-related proteins were not among the most upregulated within either cell line after infection with VMG. The highest upregulated proteins in CFPAC-1 after VMG infection were PLEKHA2, PALM, and TMEM30B. Notably, PLEKHA2 has also been shown to have a role in Zika virus resistance. We observed variability in response to VMG treatment between patient-.derived organoids, although the cause remains to be determined. In the syngeneic murine model of PDAC Panc.02, we observed no tumor regression after two intratumoral injections of VMG. However, VMG administration was well tolerated, and flow cytometric analysis showed a significant increase of CD8 T cells within oncolytic virus treated tumors. These results demonstrate oncolytic VMG shows promise in some models of PDAC, but responses can be expected to be variable in a broad patient base. It is vital to identify genetic or cellular signatures which can indicate whether a tumor will be responsive to oncolytic VMG therapy, and how to rationally combine with other treatment modalities such as ICB to overcome the elucidated mechanisms. No conflict of interest.