Back

Manufacturing DNA in E. coli yields higher fidelity DNA than in vitro enzymatic synthesis

Hersch, S. J.; Chandrasekaran, S.; Lam, J.; Nafissi, N.; Slavcev, R.

2023-09-12 synthetic biology
10.1101/2023.09.12.557453 bioRxiv
Show abstract

The rise of biotechnologies such as gene therapy have brought DNA vectors to the forefront of pharmaceutical development. The quality of the genetic starting material plays a pivotal role in determining the quality of the final product. In this study we examined the fidelity of DNA replication using enzymatic methods (in vitro) compared to plasmid DNA produced in vivo in E. coli. Next-generation sequencing approaches predominantly rely on in vitro polymerases, which have inherent limitations in sensitivity. To address this challenge, we introduce a novel assay based on loss-of-function (LOF) mutations in the conditionally toxic sacB gene. Our findings show that DNA production in E. coli results in significantly fewer LOF mutations (approximately 80-to 3000-fold less) compared to various enzymatic DNA synthesis methods. This includes the most accurate PCR polymerase (Q5) and a commonly employed rolling circle amplification (RCA) DNA polymerase (Phi29). These results suggest that using low-fidelity starting material DNA synthesized in vitro by PCR or RCA may introduce a substantial number of impurities, potentially affecting the quality and yield of final pharmaceutical products. In summary, our study underscores that DNA synthesized in vitro has a significantly higher mutation rate than DNA produced traditionally in E. coli. Therefore, utilizing in vitro enzymatically-produced DNA in biotechnology and biomanufacturing may entail considerable fidelity-related risks, while DNA starting material derived from E. coli substantially mitigates this risk, enhancing overall quality in the production processes.

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.