
nLTR-RTs have been a source of evolutionary novelties, yet they can also be a source of deleterious alleles. Since nLTR-RTs are rarely transmitted horizontally in vertebrates (with the exception of elements of the RTE clade ), the interaction between nLTR-RTs and the genome of their host is among the most intimate and long-lasting co-evolutionary processes found in nature. The mode of mobilization of nLTR-RTs has not been elucidated for all clades but it is likely that, considering their structural similarities, all these elements transpose via a target-primed reverse transcription reaction, as experimentally demonstrated for the R2Bm and L1 elements. They can be classified into 28 clades that differ in the number of open-reading frames (ORFs - one or two) and the presence of functional motifs. nLTR-RTs constitute a diverse and ancient group of transposable elements whose origin predates the diversification of the main eukaryotic lineages. Non-LTR retrotransposons (nLTR-RTs) are ubiquitous in vertebrate genomes and have profoundly affected the size, structure and function of these genomes. Finally, the coexistence of elements with drastically different base composition suggests that these elements may be using different strategies to persist and multiply in the genome of their host.

Our results suggest that base composition is evolving under selection and may be reflective of the long-term co-evolution between non-LTR retrotransposons and their host. We showed that nucleotide content remains constant within the same host over extended period of evolutionary time, despite mutational patterns that should drive nucleotide content away from the observed base composition. We found that non-LTR retrotransposons differ in base composition among clades within the same host but also that elements belonging to the same clade differ in base composition among hosts. We also investigated if elements belonging to the same clade evolved towards different base composition in different genomes or if elements from different clades evolved towards similar base composition in the same genome.

We examined if the A-richness of L1 is a general feature of non-LTR retrotransposons or if different clades of elements have evolved different nucleotide content. It is plausible that the A-richness of mammalian L1 is a self-regulatory mechanism reflecting a trade-off between transposition efficiency and the deleterious effect of L1 on its host.

For instance, the mammalian L1 element is AT-rich with a strong A bias on the positive strand, which results in a reduced transcription.

Non-LTR retrotransposons often exhibit base composition that is markedly different from the nucleotide content of their host’s gene.
