The major advantage of mRNA vaccines over more conventional approaches is their potential for rapid development and large-scale deployment in pandemic situations. In the current COVID-19 crisis the two mRNA COVID-19 vaccines have been conditionally approved and broadly applied, while others are still in clinical trials. However, there is no previous experience with the use of mRNA vaccines on the large scale in general population. This warrants a careful evaluation of mRNA vaccine safety properties by considering all available knowledge on the mRNA molecular biology and evolution. Here, I discuss the pervasive claim that mRNA-based vaccines cannot alter genomes. Surprisingly, this notion is widely stated in the mRNA vaccine literature, but never supported by referencing any primary scientific papers that would specifically address this question. This discrepancy becomes even more puzzling if one considers previous work on the molecular and evolutionary aspects of retroposition in murine and human populations that clearly documents the frequent integration of mRNA molecules into genomes, including clinical contexts. By performing basic comparisons, I showed that the sequence features of mRNA vaccines meet all known requirements for retroposition by L1 elements — the most abundant autonomously active retrotransposons in the human genome. In contrast, I found an evolutionary bias in the set of known retrocopy generating genes — a pattern that might help in the future development of retroposition-resistant therapeutic mRNAs.
I conclude that is unfounded to a priori assume that mRNA-based therapeutics do not impact genomes, and that the route to genome integration of vaccine mRNAs via endogenous L1 retroelements is easily conceivable.
This implies that we urgently need experimental studies that would rigorously test for the potential retroposition of vaccine mRNAs. At present, the insertional mutagenesis safety of mRNA-based vaccines should be considered unresolved.