LINE-1 preTa elements in the human genome

The preTa subfamily of long interspersed elements (LINEs) is characterized by a three base-pair "ACG" sequence in the 3' untranslated region, contains approximately 400 members in the human genome, and has low level of nucleotide divergence with an estimated average age of 2.34 million years old suggesting that expansion of the L1 preTa subfamily occurred just after the divergence of humans and African apes. We have identified 362 preTa L1 elements from the draft human genomic sequence, investigated the genomic characteristics of preTa L1 insertions, and screened individual elements across diverse human populations and various non-human primate species using polymerase chain reaction (PCR) assays to determine the phylogenetic origin and levels of human genomic diversity associated with the L1 elements. All of the preTa L1 elements analyzed by PCR were absent from the orthologous positions in non-human primate genomes with 33 (14%) of the L1 elements being polymorphic with respect to insertion presence or absence in the human genome. The newly identified L1 insertion polymorphisms will prove useful as identical by descent genetic markers for the study of human population genetics. We provide evidence that preTa L1 elements show an integration site preference for genomic regions with low GC content. Computational analysis of the preTa L1 elements revealed that 29% of the elements amenable to complete sequence analysis have apparently escaped 5' truncation and are essentially full-length (approximately 6kb). In all, 29 have two intact open reading frames and may be capable of retrotransposition.     

Genomic rearrangements by LINE-1 insertion-mediated deletion in the human and chimpanzee lineages

Long INterspersed Elements (LINE-1s or L1s) are abundant non-LTR retrotransposons in mammalian genomes that are capable of insertional mutagenesis. They have been associated with target site deletions upon insertion in cell culture studies of retrotransposition. Here, we report 50 deletion events in the human and chimpanzee genomes directly linked to the insertion of L1 elements, resulting in the loss of ~18 kb of sequence from the human genome and ~15 kb from the chimpanzee genome. Our data suggest that during the primate radiation, L1 insertions may have deleted up to 7.5 Mb of target genomic sequences. While the results of our in vivo analysis differ from those of previous cell culture assays of L1 insertion-mediated deletions in terms of the size and rate of sequence deletion, evolutionary factors can reconcile the differences. We report a pattern of genomic deletion sizes similar to those created during the retrotransposition of Alu elements. Our study provides support for the existence of different mechanisms for small and large L1-mediated deletions, and we present a model for the correlation of L1 element size and the corresponding deletion size. In addition, we show that internal rearrangements can modify L1 structure during retrotransposition events associated with large deletions.     

Acquisition of endonuclease specificity during evolution of L1 retrotransposon

L1 is the most proliferative autonomous retroelement that comprises about 20% of mammalian genomes. Why L1s have proliferated so extensively in mammalian genomes is an important yet unsolved question. L1 copies are amplified via retrotransposition, in which the DNA cleavage specificity by the L1-encoded endonuclease (EN) primarily dictates sites of insertion. Whereas mammalian L1s show target preference for 5'-TTAAAA-3', other L1-like elements exhibit various degrees of target specificity. To gain insights on diversification of the EN specificity during L1 evolution, ENs of zebrafish L1 elements were analyzed here. We revealed that they form 3 discrete clades, M, F, and Tx1, which is in stark contrast to a single L1 clade in mammalian species. Interestingly, zebrafish clade M elements cluster as a sister group of mammalian L1s and show target-site preference for 5'-TTAAAA-3'. In contrast, elements of the clade F, the immediate outgroup of the clade M, show little specificity. We identified certain clade-specific amino acid residues in EN, many of which are located in the cleft that recognizes the substrate, suggesting that these amino acid alterations have generated 2 types of ENs with different substrate specificities. The distribution pattern of the 3 clades suggests a possibility that the acquisition of target specificity by the L1 ENs improved the L1 fitness under the circumstances in mammalian hosts.