Nature Genetics 33, 123 - 124 (2003)
Published online: 13 January 2003; | doi:10.1038/ng1074
Retroelements containing introns in diverse invertebrate taxaIrina R. Arkhipova1, 4, Konstantin I. Pyatkov2, 4, Matthew Meselson1
& Michael B. Evgen'ev2, 31 Department of Molecular and Cellular Biology, Harvard University, Cambridge, Massachusetts 02138, USA. 2 Institute of Cell Biophysics, Pushino, Russia. 3 Institute of Molecular Biology, Moscow, Russia. 4 These two authors contributed equally to this work.
Correspondence should be addressed to Matthew Meselson msm@wjh.harvard.edu We report that two structurally similar transposable elements containing reverse transcriptase (RT), Penelope in Drosophila virilis and Athena in bdelloid rotifers, have proliferated as copies containing introns. The ability of Penelope-like elements (PLEs) to retain introns, their separate phylogenetic placement and their peculiar structural features make them a novel class of eukaryotic retroelements.
Both known classes of eukaryotic retroelements (LTR and non-LTR retrotransposons) and retropseudogenes are generally devoid of introns because they pass through an RNA intermediate that would be subject to splicing (spliced subgenomic RNAs coding for gag- or env-like proteins do not serve as transposition intermediates). The involvement of an RNA intermediate in transposition is usually indicated by the absence of introns in transposed copies of elements into which they are artificially introduced1,
2,
3.
Unexpectedly, the unusual transposon Penelope contains RT and is a causative agent of hybrid dysgenesis in Drosophila virilis4. Its structure and the amino-acid sequence of its conserved RT domains differ greatly from those of the two known classes of eukaryotic retrotransposons, yet it shares structural features with both classes. As in non-LTR retrotransposons without LTRs, Penelope has frequent 5' truncations and its target-site duplications, when present, are of variable length. Also, as in LTR retrotransposons, Penelope may have LTR-like structures; when present, however, these may be in either direct or inverted orientation. The Penelope RT bears a remote resemblance to the RT of retrons and bacterial or organellar group II introns. Its single open reading frame contains a C-terminal Uri (or GIY-YIG) domain with homology to homing endonucleases of group I introns and to UvrC bacterial DNA-repair endonucleases, which has not previously been encountered in retroelements5.
Comparison of Penelope genomic and cDNA sequences showed that the 5' untranslated region contains a functional intron of 75 bp bounded by GT/AG donor/acceptor splice sites (Fig. 1a and Supplementary Fig. A online). To distinguish between splicing and the existence of a transcribed intronless copy, we examined D. virilis genomic DNA by PCR (see Supplementary Note A online), but detected no signals corresponding to intronless copies (see Supplementary Fig. A online).
 | | |
Recently, it has become apparent that Penelope is a member of a widely distributed group of similar elements with hosts as diverse as flatworms, roundworms, crustaceans, echinoderms, fish and amphibians5,
6,
7. We found PLEs in rotifers of the class Bdelloidea, an anciently asexual taxon in which PCR screens had not detected representatives of either of the two known classes of retroelements, although these were readily detected in all other animal phyla tested8,
9. We identified amplification products corresponding to PLEs in three distantly related bdelloid species, Adineta vaga, Habrotrocha constricta and Philodina roseola (Fig. 1b and Supplementary Fig. A online). Their RT coding sequences could not be aligned to those of other PLEs without removing a variable-length (53–70 bp) region containing in-frame stop codons and frameshifts and surrounded by GT/AG donor/acceptor splice sites. This indicated the presence of a functional intron, which was confirmed by the finding of spliced RNA molecules by RT–PCR (Fig. 1c and Supplementary Fig. A online). As in D. virilis, PCR with primers flanking the intron region showed no intronless copies in the total genomic DNA of A. vaga and H. constricta.
The bdelloid RT sequences, especially the introns, showed a high degree of diversity (Fig. 1b,c). But the exon sequences immediately adjacent to the donor and acceptor splice sites were conserved and, in most introns, a pyrimidine tract could be distinguished between the putative branch point and the 3' splice site (Fig. 1c). Southern-blot analysis and genomic library screens with diverged RT fragments as probes (data not shown) indicated that each RT sequence was present at very low copy number (1–2 copies). Thus, bdelloid PLEs either have been inactive since their expansion in copy-number tens of millions of years ago or have been active but kept at low copy-number. Both possibilities are consistent with the expectation that active deleterious vertically transmitted transposons will not persist in the genomes of ancient asexual organisms9.
Phylogenetic analysis of the seven conserved domains of numerous RT coding sequences (Fig. 2 and Supplementary Fig. B online) indicated that the bdelloid PLEs, which we named Athena, form a clade distinct from the PLEs of other organisms. Moreover, the entire PLE clade is clearly different from LTR and non-LTR retrotransposons and constitutes a third, phylogenetically distinct, group of eukaryotic retroelements. Notably, the high clade credibility value for the branch containing PLEs and telomerase reverse transcriptases indicates that the RT of telomerases (the only known non-selfish RT with introns) may be a sister clade to the PLEs, adding a new twist to discussions regarding the origin of telomerases10,
11.
| | |
The differing placement of introns in D. virilis and bdelloid PLEs and the apparent absence of intron-like interruptions in the open reading frame in PLEs from other organisms indicate that introns are not a necessary feature of PLEs. The presence of an intron at the same site in all genomic copies of Athena suggests that it was already present before the element proliferated, possibly before the bdelloid radiation. As functional RT cannot be translated from the unspliced Athena transcript, proliferation seemingly occurred either by an RNA-independent pathway or by reverse transcription of an unspliced transcript. In the latter case, the existence of an intron in the RT coding region of Athena would rule out the cis-preference in transposition that is characteristic of non-LTR retrotransposons, whereby the RT preferentially acts on the mRNA template that produced it12. If RNA is not an intermediate, one may speculate that the Uri endonuclease domain found in all PLEs enables them to use a DNA-dependent transposition pathway similar to that used by group I and occasionally group II introns13. Regardless of whether transposition occurs by an RNA-dependent or RNA-independent pathway, it is clear that the ability to retain introns after transposition and their distinct placement in the phylogeny of elements containing RT, as well as the variable structure of their termini4,
5,
6,
7 and notable parallels with both group I and group II introns, support assignment of PLEs to a third class of eukaryotic retroelements, distinct from LTR and non-LTR retrotransposons.
GenBank accession numbers.
Penelope cDNA, AF418571; Athena elements, AY179351 through AY179364; Cercyon, TPA BK000685.
Note: Supplementary information is available on the Nature Genetics website.
Received 16 September 2002; Accepted 19 November 2002; Published online: 13 January 2003.
REFERENCES
-
Boeke, J.
,
Garfinkel, D.
,
Styles, C.
&
Fink, G.
Cell 40, 491–500 (1985). | Article | PubMed | ISI | ChemPort |
-
Heidmann, O.
&
Heidmann, T.
Cell 64, 159–170 (1991). | Article | PubMed | ISI | ChemPort |
-
Moran, J.
et al. Cell 87, 917–927 (1996). | Article | PubMed | ISI | ChemPort |
-
Evgen'ev, M.
et al. Proc. Natl. Acad. Sci. USA 94, 196–201 (1997). | Article | PubMed | ChemPort |
-
Lyozin, G.
et al. J. Mol. Evol. 52, 445–456 (2001). | PubMed | ISI | ChemPort |
-
Volff, J.
,
Hornung, U.
&
Schartl, M.
Mol. Genet. Genomics 265, 711–720 (2001). | Article | PubMed | ISI | ChemPort |
-
Dalle Nogare, D.
,
Clark, M.
,
Elgar, G.
,
Frame, I.
&
Poulter, R.
Mol. Biol. Evol. 19, 247–255 (2002). | PubMed | ChemPort |
-
Mark Welch, D.
&
Meselson, M.
Science 288, 1211–1215 (2000). | Article | PubMed |
-
Arkhipova, I.
&
Meselson, M.
Proc. Natl. Acad. Sci. USA 97, 14473–14477 (2000). | Article | PubMed | ChemPort |
-
Nakamura, T.
&
Cech, T.
Cell 92, 587–590 (1998). | Article | PubMed | ISI | ChemPort |
-
Eickbush, T.
Science 277, 911–912 (1997). | Article | PubMed | ISI | ChemPort |
-
Wei, W.
et al. Mol. Cell. Biol. 21, 1429–1439 (2001). | Article | PubMed | ISI | ChemPort |
-
Belfort, M.
,
Derbyshire, V.
,
Parker, M.
,
Cousineau, B.
&
Lambowitz, A.
in Mobile DNA II (eds. N. Craig, R. Craigie, M. Gellert & A. Lambowitz) (ASM Press, Washington, DC, 2002).
-
Arkhipova, I.
Genetics 139, 1359–1369 (1995). | PubMed | ISI | ChemPort |
-
Moore, M.
,
Query, C.
, &
Sharp, P.
in The RNA World (eds. R. Gesteland & J. Atkins) (Cold Spring Harbor Laboratory Press, New York, 1993).
Acknowledgments
We thank M. Cummings and D. Mark Welch for comments. This work was supported by the National Science Foundation (M.M.), the Wellcome Trust and the Russian Foundation for Basic Research (M.B.E.).
Competing interests statement:
The authors declare that they have no competing financial interests. |
