The L1 Retrotransposition Cycle

Human L1 retrotransposition begins with transcription from an internal promoter located within its 5' UTR (Fig. 2). After transcription, the bicistronic L1 RNA is transported to the cytoplasm, where ORF1 and ORF2 undergo translation. Recent studies suggest that ORF2p is translated by an unconventional mechanism and that it is made at much lower levels than ORF1p (Alisch, Garcia Perez, and Moran, unpublished data). Indeed, it is possible that as few as one molecule of ORF2p is synthesized from a single L1 RNA (74).

Genetic, biochemical, and phylogenetic studies indicate that ORF1p and ORF2p exhibit a strong cis-preference (7,9), and preferentially associate with the RNA that encoded them to form an RNP, which is a proposed retrotransposition intermediate (19,75). The L1 RNP enters the nucleus either by active import or passively, perhaps during mitotic nuclear envelope breakdown. The only non-LTR retrotransposon studied in this respect is the Tad element from Neurospora crassa, which gains access to the nucleus by active transport (76).

Once in the nucleus, L1 retrotransposition likely occurs by target-site primed reverse transcription (TPRT), a mechanism first demonstrated for the R2 retrotransposon from Bombyx mori (R2Bm) (77). During TPRT, L1 EN is thought to cleave genomic DNA, liberating a 3' hydroxyl, which then serves as a primer for reverse transcription of L1 RNA by the L1 RT, generating the first strand of L1 cDNA (22,67,78). Recent in vitro biochemical data suggest that besides functioning in RNP assembly, the nucleic acid chaperone activity of ORF1p also may facilitate early steps of TPRT (60). Second-strand synthesis and integration of the nascent L1 cDNA into

Fig. 2. Model of L1 retrotransposition cycle. A retrotransposition competent L1 is transcribed and the L1 RNA is transported to the cytoplasm. Here, ORFlp (blue circle) and ORF2p (green oval) are translated and bind back to the L1 RNA from which they were transcribed to form a ribonucleoprotein particle (RNP). This L1 RNP then translocates to the nucleus where target-site primed reverse transcription (TPRT) occurs to integrate the L1 at a new location in the genome. The newly integrated L1 exhibits characteristic structures including 5' truncation, a poly (A) tail, and variable length target site duplications (black arrows). A variation on TPRT is endonuclease independent retrotransposition, in which L1 integrates at pre-existing nicks in DNA. Endonuclease independent retrotransposition events sometimes lack a poly (A) tail, lack target sited duplications, and may be 3' truncated. TPRT is the prevalent route of integration into the genome (indicated by bold arrow), whereas endonuclease independent retrotransposition probably occurs at a much lower frequency in vivo (indicated by the dashed arrow).

Fig. 2. Model of L1 retrotransposition cycle. A retrotransposition competent L1 is transcribed and the L1 RNA is transported to the cytoplasm. Here, ORFlp (blue circle) and ORF2p (green oval) are translated and bind back to the L1 RNA from which they were transcribed to form a ribonucleoprotein particle (RNP). This L1 RNP then translocates to the nucleus where target-site primed reverse transcription (TPRT) occurs to integrate the L1 at a new location in the genome. The newly integrated L1 exhibits characteristic structures including 5' truncation, a poly (A) tail, and variable length target site duplications (black arrows). A variation on TPRT is endonuclease independent retrotransposition, in which L1 integrates at pre-existing nicks in DNA. Endonuclease independent retrotransposition events sometimes lack a poly (A) tail, lack target sited duplications, and may be 3' truncated. TPRT is the prevalent route of integration into the genome (indicated by bold arrow), whereas endonuclease independent retrotransposition probably occurs at a much lower frequency in vivo (indicated by the dashed arrow).

genomic DNA completes retrotransposition and results in L1 structural hallmarks (i.e., frequent 5' truncations, a 3' A-tail, and variable-length target site duplications) (24). However, the initiation of second strand synthesis and the completion of L1 integration remains a mystery.

TPRT: VARIATIONS ON A THEME

Approximately 99% of L1s present in the human genome working draft (HGWD) contain random 5' truncations and approx 20% of those L1s contain internal rearrangements (i.e., they have inversion/deletion structures; Fig. 3). It has been proposed that 5' truncation is owing to reduced processivity of the L1 RT (79). However, it is counterintuitive that L1 would evolve a poorly processive RT when it requires this enzyme for its survival. Furthermore, studies on the related R2Bm RT show that it is a highly processive enzyme in vitro (80). Thus, incomplete L1 cDNA synthesis may arise from host repair processes that act to suppress retrotransposition by disassociating the L1 RT from its nascent cDNA (74).

Alterations to L1 structure

Alterations to target site

Fig. 3. Ll-mediated changes on insertion. Insertion into a new site in the genome (gray bar) can result in 5' truncation of the L1 or inversion/deletion structures (breakpoints denoted by arrows) when compared with its progenitor L1 (black bar). L1 insertion can also result in changes to the target site (gray bar with three black triangles), such as target site duplications, deletions of target site nucleotides, and the insertion of small "filler" DNAs (checked box).

Fig. 3. Ll-mediated changes on insertion. Insertion into a new site in the genome (gray bar) can result in 5' truncation of the L1 or inversion/deletion structures (breakpoints denoted by arrows) when compared with its progenitor L1 (black bar). L1 insertion can also result in changes to the target site (gray bar with three black triangles), such as target site duplications, deletions of target site nucleotides, and the insertion of small "filler" DNAs (checked box).

The "twin-priming" model can explain the formation of inversion/deletion Lis (81). Twin priming evokes the use of the 3' OH on the top strand of target site DNA (after its cleavage) as a second primer for reverse transcription on the Li RNA. Resolution of the convergent cDNAs followed by microhomology-mediated recombination then leads to the formation of the inversion/deletion. The twin-priming model is supported by both analyses of genome insertions from databases and cell culture experiments (28,29).

In general, TPRT leads to minor alterations of target site DNA, including the generation of short target site duplications or small deletions, yet in some instances there is neither a gain or loss of target site nucleotides (Fig. 3). How TPRT can lead to these various outcomes requires further study, but it has been proposed that the variable placement of top strand cleavage can account for all the observed scenarios (28). In fact, though there is a clear Li EN consensus "bottom strand" cleavage site, there is little or no target site preference for top strand cleavage (28,82). Thus, Li EN either is an unusual enzyme that only displays cleavage specificity for a single DNA strand, or an undiscovered activity (encoded by either Li or the host) is required for second (i.e., top) strand cleavage.

Various forms of genetic instability are associated with another approx 10% of L1 retrotransposition events in transformed cultured cells (Fig. 3). These alterations include the addition of "filler DNA" at the 5' genomic DNA/L1 junction, the generation of target site deletions or intrachromosomal duplications, the formation of chimeric Lis, and the possible generation of chromosomal inversions and interchromosomal translocations (28,29). Although the relatively high incidence of these unusual rearrangements may reflect the transformed status of the cells used in these experiments, comparative genomic studies have identified rare deletion events associated with both Li and Alu insertions in humans (46,83,84). Moreover, deletions are associated with two of eight mutagenic Li insertions in the mouse (85,86). Thus, it is probable that genetic instability occasionally accompanies Li retrotransposition in vivo and it is intriguing to speculate that these events may have impacted genome evolution (87).

There also are variations in the standard TPRT model of retrotransposition. For example, it is hypothesized that pre-existing lesions in genomic DNA may be used as primers in place of the Li endonuclease-generated sites, referred to as endonuclease-independent retrotransposition (Fig. 2) (30,88). This phenomenon is most obvious in Chinese Hamster Ovary cells defective for nonhomologous end joining and can be considered a type of RNA-mediated DNA repair (i.e., a chromosomal "Band-Aid") that is a remnant from the "RNA world."

Was this article helpful?

0 0

Post a comment