The project’s published results already demonstrate why tissue context matters for medicine. First, we discovered that an mRNA vaccine can be stabilized in immune cells through re-adenylation: a cellular enzyme extends the vaccine’s poly(A) tail after delivery, helping the mRNA persist longer. In a mouse model, this stabilization mechanism increased vaccine effectiveness, showing that the body’s own RNA-tail machinery can directly influence therapeutic outcomes—and suggesting a route to improve therapeutic mRNAs by designing them to benefit from cell-type-specific stabilization (DOI: 10.1038/s41586-025-08842-1).
Second, ViveRNA delivered a practical tool that makes these effects measurable at scale. We developed Ninetails, a neural-network method that directly profiles “non-A” letters within poly(A) tails from nanopore direct RNA data. Using this approach, we showed that mixed poly(A) tails are widespread across species and mRNA classes, and we demonstrated its usefulness for both endogenous RNAs and therapeutic mRNAs, including vaccines. The work also highlights that poly(A) tail composition can change during an mRNA’s cellular lifetime and that manufacturing-related features can influence poly(A) “purity”—information that is relevant for understanding and quality-assessing mRNA medicines (DOI: 10.1038/s41467-025-57787-6).
Third, we improved the quantitative understanding of how poly(A) shortening (deadenylation) relates to mRNA decay. By modeling tail-length dynamics, we estimated a transcriptome-wide deadenylation rate and showed that the relationship between deadenylation and decay is often linked but can be more complex for specific groups of transcripts. This refines simplified “one-pathway” views of mRNA turnover and supports better predictive models of mRNA lifetime (DOI: 10.1038/s44318-024-00258-3).
Publicly available publications: 10.1038/s41586-025-08842-1; 10.1038/s41467-025-57787-6; 10.1038/s44318-024-00258-3.
