Gene Editing Thumbs a Ride with Oncolytic Virotherapy

Main Text

The reigning paradigm of cancer drug development is a relatively straight road, beginning with single agent discovery and testing, with demonstration of the required safety, but often only small hints of efficacy. With rare exceptions, the true power of a therapeutic lies in its collaborative nature: that is, how good is it as a team player? Is it compatible with, or, better yet, synergistic with other effective therapeutics? Most successful cancer treatment protocols have been honed over decades with refinements of such combinations given in specific orders and schedules. For example, the current Children’s Oncology Group study for standard-risk leukemia (AALL1731) explores the use of no fewer than 13 different drugs in various sequential blocks, a strategy with a long track record of success.¹

While biologic molecular therapies are decades behind classical cytotoxic chemotherapy and targeted small molecules, the idea of merging distinct yet complementary approaches will likely continue to serve us well. A major attraction of oncolytic viruses is that different biologic mechanisms can be packaged into a single therapeutic, limited only by the cloning capacity of a given vector. In this issue of Molecular Therapy, Yoon et al.² merge the lytic (and presumably immunostimulatory) effects of such a virus with clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein (Cas) in situ gene editing to selectively mutate an oncogenic driver of lung cancer, epidermal growth factor receptor (EGFR), thereby eliciting more potent antitumor efficacy. In contrast to the more commonly used type II Cas9, they created an oncolytic adenovirus expressing the type V Cas12a and the specific guide RNA. Cas12a has inherent advantages over Cas9, the most important being higher fidelity with less off-target nuclease activity. It also causes staggered double-stranded breaks, increasing the likelihood of inactivating mutations, and it has a shorter coding region, facilitating packaging in a vector. Engineering an oncolytic virus with CRISPR/Cas9 might a priori seem redundant, as infected cells are likely to be killed by the lytic infection anyway. The combination might also be predicted to be antagonistic if the CRISPR/Cas9 targets a gene important for cell survival, because hastening cell death might reduce virus spread. Yet the authors found enhanced cytotoxicity in vitro and near complete inhibition of tumor growth in their models, with increased numbers of apoptotic cells. They also found no evidence of off-target mutations in tumor cells or bystander-type nuclease activity in co-cultured normal human dermal fibroblasts. Moreover, the tumor-selective nature of oncolytic viruses (via direct intratumoral injection and selective amplification in tumors) likely limits systemic off-target effects, and, indeed, they found no nuclease-induced mutations in other mouse organs tested.

This report actually extends and validates prior work by Phelps et al.,³ who engineered an oncolytic myxoma virus to express a CRISPR/Cas9 that targeted NRAS in embryonal rhabdomyosarcoma. Unfortunately, in their system, tumors rapidly developed resistance, mostly due to mutations at the guide RNA cut sites. Why resistance did not occur in the current report is unclear, but some target sites may be more prone to mutation than others. Also, the mice were not followed long-term, and most still had small tumors present at the end of the experiment, suggesting later regrowth or relapse is a possibility.² While the Phelps et al.³ study was arguably more pioneering and targeted a protein considered relatively “undruggable,” the current work demonstrates that the approach can be used across different virus backbones, different driver oncogenes, and higher fidelity Cas proteins. As Yoon et al.² point out, future versions will likely further extend the idea of combinatorial therapy, targeting multiple genes in various pathways to elicit disparate mechanistic effects. Synergy with the immunostimulatory properties of oncolytic viruses could be sought by altering immunoregulatory genes, for example, by downregulating expression of immune checkpoints.⁴ To the extent strategies are developed that enhance bystander killing, the effects of virotherapy-mediated gene editing may not just be limited to infected cells.

One potential pitfall for the clinical implementation of CRISPR/Cas gene editing is the immune response to the bacterial Cas protein, which can eliminate gene edited cells and thwart attempts at gene therapy.⁵ While that issue may become a significant challenge for correction of genetic disorders, anti-Cas immunity to help destroy cancer cells would be a welcome addition to the carpool.