Nucleosomes are essential components of eukaryotic cells. A physical model of nucleosome formation showed that, for all vertebrate genomes analyzed, the positions of nucleosomes are in part encoded in the DNA sequence by Nucleosome Inhibitory Energy Barriers (NIEBs). This has been confirmed using experimental data from human (Drillon et al. 2016) and mouse (Tartour et al. 2023). In all cases, average GC-content profiles at NIEB borders display oscillations in phase with positioned nucleosomes, regions covered by nucleosomes being more GC-rich than NIEB and linker regions between nucleosomes.
We analyzed systematically all Dfam annotated TEs (Hubley 2015) in the human, mouse and pig genomes. We show that, for TEs having a polyA at their 3'end, this polyA is preferentially positioned on the interior side of a NIEB border. NIEB-associated TEs present GC-content oscillations that are similar to non-TE GC oscillations, excepted for the linker in human L1 retrotransposons and B2 mouse SINEs. We reconstructed the genome of the last common ancestors of human and chimpanzee, Mus musculus and Mus caroli, Sus scrofa and Sus cebifrons. We calculated the mutation rates since the divergence within all pairs of species, which was used to predict the GC content at equilibrium of TEs associated with NIEBs. For most NIEB-associated highly repeated TEs, we found that the current GC oscillation and the GC equilibrium are both correlated and in accord with nucleosome positioning around NIEBs. Interestingly, for both the B2 and L1 elements, GC at equilibrium at linker position showed a clear decrease that was not observed in current GC, indicating that these sequences are evolving towards the common GC oscillation pattern associated with NIEBs.
Taken together, our results suggest that TEs at NIEB borders have been shaped by mutations to present nucleosome-compatible GC-content oscillations, indicating a general link between chromosome structure and TE evolution.