A Workflow for In Vivo Evaluation of Candidate Inputs and Outputs for Cell Classifier Gene Circuits

therapeutic outputs to cancer cells, have the potential to result in efficacious and safe cancer

therapies. Preclinical translation of the hitherto developments requires creating the conditions

where the animal model, the delivery platform, in vivo expression levels of the inputs, and the

efficacy of the output, all come together to enable detailed evaluation of the fully-assembled

circuits. Here we show an integrated workflow that addresses these issues and builds the

framework for preclinical classifier studies using the design framework of microRNA (miRNA,

miR)-based classifier gene circuits. Specifically, we employ HCT-116 colorectal cancer cell

xenograft in an experimental mouse metastatic liver tumor model together with Adenoassociated

virus (AAV) vector delivery platform. Novel engineered AAV-based constructs are

used to validate in vivo the candidate inputs miR-122 and miR-7 and, separately, the

cytotoxic output HSV-TK/ganciclovir. We show that while the data are largely consistent with

expectations, crucial insights are gained that could not have been obtained in vitro. The

results highlight the importance of detailed stepwise interrogation of the experimental

parameters as a necessary step toward clinical translation of synthetic gene circuits.