SAGA-Q: Gene Expression Differences Enable Detection of Retroviral Vectors with Mutagenic Potential

Abstract

Although the safety of retroviral vector (RV) gene therapy has been improved over the last years, insertional mutagenesis is still a risk factor, as seen in some of the clinical trials targeting hematopoietic stem cells. This highlights the necessity of appropriate preclinical genotoxicity assays. Our group previously developed the In Vitro Immortalization Assay (IVIM) and Surrogate Assay for Genotoxicity Assessment (SAGA) to evaluate the risk of side effects by integrating vectors. In this study, murine hematopoietic stem and progenitor cells are transduced with RVs, and genotoxicity can be detected by a proliferation advantage under limiting dilution conditions (IVIM) or the activation of genes associated with oncogenesis and stem cell-like properties (SAGA). A limitation of SAGA is the costly microarray technology. In this study, we present the digital droplet-based SAGA-Quantification (SAGA-Q) as a cost-efficient and faster alternative. Murine samples transduced with known mutagenic vector designs consistently showed upregulation of genotoxicity predictor genes. Based on a training set of 140 IVIM samples (including untransduced controls and samples transduced with long terminal repeat-driven γRV, SIN-LV.SF, SIN-LV.EFS, SIN-LV.PGK.RAG2, SIN-LV.MND.RAG1, and SIN-LV.MND.RAG2), we used random forest prediction for reliable and fast identification of genotoxic vector designs. The relevance of the predictor genes for the immortalization process was further highlighted by an elevated expression in immortalized clones. By simplifying SAGA to SAGA-Q, we aim to increase the accessibility of genotoxicity assessment and, thus, support the safer translation of gene therapy products to clinical trials.