Quantitative analysis of clinically relevant mutations occurring in lymphoid cells harboring gamma-retrovirus-encoded hsvtk suicide genes.

The in vivo regulation of T lymphocyte activity by the activation of a suicide mechanism is an essential paradigm for the safety of adoptive cell therapies. In light of reports showing that gamma-retroviral vector-encoded herpes simplex virus thymidine kinase (hsvtk) undergoes recombination, we undertook a thorough investigation of the genomic stability of SFG-based vectors using two variants of the wild-type hsvtk gene. In a large panel of independent clones, we demonstrate that both hsvtk genes undergo recombination with molecular signatures indicative of template switching in GC-rich regions displaying homology at the deletion junctions or RNA splicing. In the absence of ganciclovir selection, the frequency of recombination is 3% per retroviral replication cycle. Our results underscore the importance of the five nucleotide difference between the two hsvtk genes that account for the presence of recombinogenic hot spots in one variant and not the other, indicating that the probability of RNA splicing is influenced by minute nucleotide changes in sequences adjacent to the splice donor and acceptor sites. Furthermore, our mutational analysis in an unbiased panel of human lymphoid cells (that is, without immune or ganciclovir-mediated selective pressure) provides a robust in vitro assay to predict and quantify clinically relevant mutations in hsvtk suicide genes, which can be applied to studying and improving the stability of any transgene expressed in gamma-retroviral or lentiviral vectors.

Production scale-up and validation of packaging cell clearance of clinical-grade retroviral vector stocks produced in cell factories.

The clinical implementation of gene therapy requires large-scale production of viral vector stocks (VS) derived from packaging cell lines. Upon scaling-up, maintenance of high viral titers and filtration of the VS become significantly challenging. Thus, production schemes amenable to straightforward validation must be developed. To this end, we have established a semi-closed process to manufacture batches of 7 l or more of clinical-grade oncoretroviral VS using 10-tray Cell Factories. Using a peristaltic pump, the VS are collected on 3 consecutive days, filtered, pooled and stored frozen. To ensure the absence of viable vector-producing cells (VPCs) from each VS unit-dose, we undertook an orthogonal log-removal validation study to demonstrate the ability of both the filtration system to remove viable cells and the VS freezing process to inactivate them. We demonstrate a total VPC-reduction of 11.6 log, thus insuring the absence of contaminating VPCs in transduced clinical samples. We also show that this production process generates stable VS that can be stored at -80 degrees C for more than 3 years. Importantly, this relatively simple and affordable process can be customized to generating large volume of VS for small animal or non-human primate studies. This methodology is not limited to the generation of cell-free clinical oncoretroviral VS, and can be applied to other types of vectors produced in packaging cell lines, such as lentiviral vectors.