Abstract
In DNA biodistribution studies, genomic DNA (gDNA) is commonly used as a normalizer for quantifying vector DNA (vDNA) to account for multiple variations including tissue type, sample input, DNA extraction efficiency, etc. Typically, the concentrations of both vDNA and gDNA in elution buffer after extraction are interpolated from standard curves prepared in neat solution, and the derived vector genome per diploid genome (vg/dg) ratio is reported to represent the relative amount of delivered vDNA in tissues. However, the accuracy of this interpolated vg/dg relies on the assumption that vDNA and gDNA have equal extraction efficiencies and are similarly affected by the tissue matrix during qPCR quantification. Despite its widespread use operationally, this assumption has yet to be scientifically examined. In this work, we tested this assumption by co-spiking vDNA and gDNA into the same tissue samples, then conducting DNA extraction and qPCR analysis to compare their extraction recovery and matrix effect. Our results revealed that vDNA and gDNA exhibit differential recovery even from the same sample and are affected differently by the tissue matrix. Consequently, this leads to bias in the interpolation of sample vg/dg values. We proposed and developed two strategies to mitigate the impact of recovery and matrix effect on DNA biodistribution studies: (1) using a matching matrix curve that closely resembles the study samples, and (2) optimizing the DNA extraction protocol to minimize the recovery and matrix effect differences. These strategies have been successfully applied to multiple biodistribution studies. We believe these findings provide valuable insights that enhance the accuracy and reliability of DNA biodistribution analysis in gene therapy research.