Abstract
The ability to deliver transgenes into the human genome using viral vectors is a major enabler of the gene-modified cell therapy field. However, the control of viral transduction is difficult and can lead to product heterogeneity, impacting efficacy and safety, as well as increasing the risk of batch failure during manufacturing. To address this, we generated a novel analytical method to measure vector copy distribution at the single-cell level in a gene-modified, lentiviral-based immunotherapy model. As the limited amount of genomic DNA in a single cell hinders reliable quantification, we implemented a preamplification strategy on selected lentiviral and human gene targets in isolated live single cells, followed by quantification of amplified material by droplet digital PCR. Using a bespoke probability framework based on Bayesian statistics, we show that we can estimate vector copy number (VCN) integers with maximum likelihood scores. Notably, single-cell data are consistent with population analysis and also provide an overall measurement of transduction efficiency by discriminating transduced (VCN ≥ 1) from nontransduced (VCN = 0) cells. The ability to characterize cell-to-cell variability provides a powerful high-resolution approach for product characterization, which could ultimately allow improved control over product quality and safety.