Prediction of polyadenylation site usage from epigenetic modifications
In this computational study, the authors explore several links between epigenetic factors (including DNA methylation, nucleosomal association, and histone modifications) and the usage of polyadenylation (poly(A)) sites in transcribed messages.
While the authors make reference to some existing data sets that indicate the role of CpG methylation in the usage of proximal poly(A) sites, most of the new findings stem from a careful study of the interaction between RNA conformers and their free energies, and how this dynamic aligns with annotated epigenomic regulators. One of the most important conclusions is that higher nucleosomal occupancy downstream from poly(A) sites demonstrates a high correlation with lower energy, thus more stable, mRNA structure at these poly(A) sites. This allows for better exposure of cis-elements, such as the polyadenylation signal (PAS), for trans-factor recognition, and thus higher frequency utilization of these poly(A) sites. Nucleosomal occupancy also correlates closely with enhanced accumulation of RNA polymerase II at poly(A) sites, and the authors suggest that this accumulation contributes to more favorable mRNA folding near poly(A) sites.
In addition to factors stemming from ‘tightness’ of the chromatin, covalent modifications of histone tails, including lysine methylation and acetylation, demonstrate consistent patterns of modification near poly(A) sites. In the case of most types of histone lysine methylation, loss of the modifications was observed near poly(A) sites. However, for some positions of lysine di- and tri-methylation, results showed consistent elevation of these marks in positions flanking poly(A) sites. The authors were then able to utilize these epigenetic correlations to write algorithms to predict and distinguish active poly(A) sites from unused poly(A) sites with higher accuracy than predictions based on well-annotated cis-elements. Further, these algorithms also demonstrated the ability to differentiate high-use from low-use poly(A) sites.
Khaladkar M, Smyda M, Hannenhalli S. (2011) Epigenomic and RNA structural correlates of polyadenylation. RNA Biol. 8(3):529-37.
Full article freely available at: http://www.landesbioscience.com/journals/rnabiology/KhaladkarRNA8-3.pdf