The main part of the project used time-resolved RNA sequencing to measure how quickly newly produced RNAs are exported from the nucleus or degraded there. Early experiments revealed that commonly used RNA-labelling conditions introduced a technical bias during sequencing. Further experiments identified a second bias caused by the gradual accumulation of the labelled molecule inside cells.
To address this problem, the project developed RoSA-MoSA, a computational method that estimates this accumulation directly from sequencing data and includes it in the analysis. The computational estimate closely matched an independent measurement made using metabolomics. This provides confidence that the method is correcting a real biological effect rather than simply improving the fit of the model.
Using optimised conditions, approximately 5,000 genes were modelled. The approach was also applied to a dataset covering approximately 50,000 introns, allowing different stages of RNA splicing to be measured. These analyses showed that completing splicing accounts for a substantial part of the time an RNA spends in the nucleus and can be the rate-limiting step for export.
The project also examined the effects of depleting EIF4A3, a protein deposited on RNA during splicing, and NXF1, a central RNA-export factor. Depleting EIF4A3 reduced the average export rate of intron-containing RNAs by approximately 50% and also reduced their nuclear stability. Super-resolution RNA imaging was used as an independent way to investigate these findings.
The planned genome-wide ORF-tag screen could not be completed. The reporter worked in human cells, whereas the available screening system worked in mouse embryonic stem cells, where the reporter did not behave as required. Attempts to transfer the screening system to human cells did not achieve sufficient efficiency. The remaining effort was therefore redirected towards the productive sequencing, modelling, protein-depletion and imaging work. Manuscripts are now in preparation, and the associated software, analysis code and sequencing data are being prepared for public release.
