| Summary: | LINE-1 (Long Interspersed Element-1/L1) and Alu are two active retrotransposon families in the human genome that have the potential to create genomic instability either during the insertion of new elements or through ectopic recombination. However, recent in vitro analyses have demonstrated that these elements also repair DNA double-strand breaks, hence contributing to the maintenance of genomic integrity. As such, the comprehensive role of mobile elements in either creating or mitigating instability in primate genomes remains unclear. The recent sequencing of the chimpanzee and rhesus macaque genomes uniquely facilitates the accurate resolution of this question, as three-way computational alignment of the human genome with two other hominoid genomes allows human lineage-specific changes (i.e., those younger than 5-6 million years) to be accurately dissected out. Here, using a combined computational and experimental approach, we have attempted to provide an unbiased picture of the contribution of the Alu and L1 families to human genomic stability. In the first analysis described herein, we assessed levels of genomic deletion associated with L1 retrotransposition and reported 50 deletions resulting in the loss of ~18 kb of human genomic sequence and ~15 kb of chimpanzee genomic sequence. We developed models to explain the observed bimodality of the deletion size distribution, and showed that overall, in vivo deletions are smaller than those observed in cell culture analyses. Next, we quantified Alu recombination-mediated deletion in the human genome subsequent to the human-chimpanzee divergence and described 492 deletions (totaling ~400 kb of human genomic sequence) attributable to this process. Interestingly, the majority of these deletions are located within known or predicted genes, opening the possibility that a portion of the phenotypic differences between humans and chimpanzees may be attributed to this mechanism. In the third analysis, we reported the in vivo existence of an endonuclease-independent insertion pathway for L1 elements and characterized twenty-one loci where L1 elements appear to have bridged genomic lesions. We show that these insertions are structurally distinguishable from classical L1 elements and suggest that this pathway may escape the purifying selection thought to be acting on endonuclease-dependent L1 loci in the genome.
|
|---|
