Previous Article |
Table of Contents
| Next Article 
* Department of Molecular and Cellular Biology, Harvard University,
7 Divinity Avenue, Cambridge, MA 02138; and Communicated by M. S. Meselson, Harvard University,
Cambridge, MA, September 19, 2001 (received for review July 27, 2001)
Transposable elements inhabiting eukaryotic genomes are generally
regarded either as selfish DNA, which is selectively neutral to the
host organism, or as parasitic DNA, deleterious to the host. Thus far,
the only agreed-upon example of beneficial eukaryotic transposons is
provided by Drosophila telomere-associated
retrotransposons, which transpose directly to the chromosome
ends and thereby protect them from degradation. This article reports
the transposon content of the genome of the protozoan Giardia
lamblia, one of the earliest-branching eukaryotes. A total of
three non-long terminal repeat retrotransposon families have been
identified, two of which are located at the ends of chromosomes, and
the third one contains exclusively dead copies with multiple internal
deletions, nucleotide substitutions, and frame shifts. No other reverse
transcriptase- or transposase-related sequences were found. Thus, the
entire genome of this protozoan, which is not known to reproduce
sexually, contains only retrotransposons that are either confined to
telomeric regions and possibly beneficial, or inactivated and
completely nonfunctional.
According to the model of
transposon proliferation presented by Hickey (1), deleterious
transposons are not expected to persist in long-term asexuals. Results
consistent with this expectation were obtained in recent experiments
testing 24 eukaryotic phyla for the presence of known transposons by
nested PCR (2). There are two classes of autonomous eukaryotic
transposons encoding conserved enzymes necessary for transposition:
(i) retrotransposons, which transpose by means of an RNA
intermediate copied into DNA by an element-encoded reverse
transcriptase (RTase); and (ii) DNA transposons, which
transpose as DNA by a cut and paste mechanism, using an element-encoded
transposase (3, 4). In PCR assays, the only group that tested negative
for the presence of RTase-related sequences, although positive for
mariner-like DNA transposases, were rotifers of the class Bdelloidea, a
monophyletic group which apparently lost sexual reproduction many
millions of years ago (5). Of those tested in ref. 2, the only other
species in which no sexual process is known is Giardia
lamblia (or G. intestinalis), which tested positive for
RTases. Cloning and sequencing of the corresponding PCR products
suggested the existence of two non-long terminal repeat [also called
long interspersed nuclear element (LINE)-like] retrotransposon
families in the G. lamblia genome, a finding seemingly at
odds with the indication from the Bdelloidea that such elements would
not persist in long-term asexuals. Only in sexuals, but not long-term
asexuals, can deleterious transposons be expected to go to fixation
(6).
G. lamblia is a protozoan, parasitizing the intestines of
mammals and birds, which has two morphologically identical nuclei in
each cell and a polyploid genome (reviewed in refs. 7-9). Its genome
can be divided into five major groups displaying physical linkage of
markers, and five chromosome-like bodies can be visualized in each
nucleus (7, 8). Each linkage group, however, can be represented by
several size variants detectable by pulse-field gel electrophoresis,
with an invariant central core and a significant degree of variability
toward the ends of the chromosomes (7-11). The variable ends undergo
frequent rearrangements, but the central regions do not.
Because all of the eukaryotic genomes previously sequenced to
completion are those of sexually reproducing organisms, it was of
particular interest to evaluate the G. lamblia genome for
abundance and activity status of these retrotransposons, in light of
their apparent absence from bdelloids. An ongoing G. lamblia
genome sequencing project (12) is approaching completion and is
currently at the gap-closure stage, making it possible to assess the
transposon content of this genome and to analyze internal and flanking
sequences of all identified transposon copies with respect to their
degrees of divergence, intactness of ORFs, and insertional specificities.
Primers, DNA sources, and amplification conditions for PCR
reactions were as described (2). Single-pass sequencing reads from the
G. intestinalis (strain WB) genome-sequencing project were
obtained from the high-throughput genomic sequence (HTGS) subdivision
of GenBank (see www.mbl.edu/Giardia). Consensus sequences were
assembled from 216, 66, and 588 individual sequencing reads, averaging
800 bp in length, for GilM, GilT, and GilD, respectively. For gap
closure in GilT, additional sequencing of G. lamblia genomic clones was performed with the Big Dye Terminator Cycle Sequencing kit
(Applied Biosystems) and analyzed on an Applied Biosystems Prism 310 genetic analyzer. Sequence assembly and analysis was done with
WISCONSIN PACKAGE VERSION 10.0 (GCG). RTase sequences chosen for phylogenetic analysis represent a subset of the seed alignment PF00078 rvt (PFAM RELEASE 6.4),
which includes only the seven conserved domains that are common to all
RTases. After removal of the most prominent gaps, a total of 304 amino
acids was included in the analysis. Inference of phylogenetic
relationships was performed by using MRBAYES2.01
(13), using the JTT substitution matrix and a mixed (invariable plus
gamma) model of rate heterogeneity, with rates inferred from the data
set. Four Markov chains were initiated at random, and the program was
allowed to run for 100,000 generations with sample frequency of 10. On
average, 30,000 generations were required for likelihood convergence,
with the first 3,000 less likely trees discarded as burn-in, and the
remaining 7,000 trees used to build a consensus tree.
PCR Experiments.
Nested PCR amplification of G. lamblia genomic DNA with
highly degenerate primers specific for the superfamily of LINE-like RTases typically yielded a single band of high intensity (Fig. 1). This result indicates that
representatives of multiple LINE-like clades, which usually have
different characteristic distances between B and C motifs of the RTase
gene, are not likely to be present in the G. lamblia genome,
in contrast to most other eukaryotes (ref. 2; Fig. 1), as was confirmed
by analysis of genome sequence (see below). Cloning and sequencing of
120-bp PCR products from G. lamblia revealed that their
sequences fell into two different groups, which were assigned to two
transposon families hereafter named GilM and GilD (G.
intestinalis LINEs). It was not possible,
however, to assign these fragments with confidence to the superfamily
of LINE elements, because their homology with known RTases in GenBank
was insufficient. Therefore, full-length RTase sequences for each
family had to be determined to reliably establish their relationship to
known RTases.
From the Cover
Evolution
Three retrotransposon families in the genome of Giardia
lamblia: Two telomeric, one dead
and
Bay Paul
Center for Comparative Molecular Biology and Evolution, Marine
Biological Laboratory, Woods Hole, MA 02543
Abstract
Top
Abstract
Introduction
Methods
Results
Discussion
References
Introduction
Top
Abstract
Introduction
Methods
Results
Discussion
References
Methods
Top
Abstract
Introduction
Methods
Results
Discussion
References
Results
Top
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (68K):
[in a new window]
Fig. 1.
PCR amplification of total genomic DNA isolated from representatives of
20 animal phyla, plus G. lamblia and Escherichia
coli, with nested primers specific for LINE-like RTases. An
arrow corresponds to the position of prominent sequence-specific
amplification products about 120 bp in length, typically representing
members of the most abundant CR1 clade; larger products correspond to
members of other clades such as L1, jockey, etc. (2).
Amplification products are not visible in E. coli and in
Habrotrocha constricta, a bdelloid rotifer (2).
Sequence Assembly and Database Analysis.
Genomic shotgun reads were used to assemble full-length consensus
sequences for GilM and GilD by extending sequence homology in both
directions from the RTase fragments obtained in PCR experiments. Conceptual translations of ORFs from the consensus sequences were used
in BLASTX searches to detect any other related elements in the G. lamblia genome. This search revealed a third
family of LINE-like elements, designated GilT. This family has a much
lower copy number than the other two, and the sequences present in the database could not be assembled into a single consensus without gaps,
which were closed by targeted sequencing. Analysis of the resulting
full-length ORFs from the three families demonstrates that they can be
unambiguously assigned to the LINE-like superfamily of RTases, with
E values of BLASTX matches to the
conserved PFAM00078 RTase domain ranging
from 10
19 to 1022.
Telomere-Associated LINE Families. GilT and GilM are potentially active elements, because they are mostly represented by intact sequences. The nucleotide sequence identity within each family exceeds 99%, which is indicative of recent retrotransposition activity. Remarkably, both families are confined to immediate subtelomeric regions, because any GilT or GilM sequence is flanked at its 5' end either by reverse complement of G. lamblia telomeric repeats (TACCC)n (14) or by another copy of the same element in the same orientation. Members of such a tandem array are separated from each other by the (A)n stretch (n = 10-16) and arranged in a strictly head to tail orientation, with no target site duplications. The most distal member in the array is truncated at its 5' end and capped by telomeric repeats. Interestingly, (A)n stretches are always followed by an intact 5' end of another copy and never by a 5'-truncated copy (Fig. 2B).
|
Most Telomeric Repeats Are Joined to LINE Elements. Telomeric repeat-containing clones were extracted from the database in a BLASTN search, and the junctions between telomeric repeats and the rest of chromosomal DNA were inspected by BLAST comparisons or obtained by targeted sequencing. All of the (TAGGG)n stretches (n > 3) were joined to chromosomal DNA only on one side, indicating that there were no interstitial TAGGG sequences repeated more than three times. This finding agrees with results of Upcroft et al. (24), who reported no hybridization of the telomeric probe to internal chromosome fragments.
Remarkably, the analysis of junction DNA revealed that 8 of 11 (TAGGG)n sequences were adjacent to 5'-truncated copies of LINE elements (Table 1). Thus, these elements are the predominant components of telomeric junctions, in contrast to previous studies identifying sequences adjoining G. lamblia telomeric repeats mostly as rDNA (10, 11, 14). In the remaining three clones, telomeric repeats are joined to a variant-specific surface protein or to rDNA (Table 1).
|
A Retrotransposon Inactivated by Multiple Deletions. Assembly of a full-length consensus sequence of the high-copy-number family, GilD, was not a trivial task: it is represented exclusively by dead copies, none of which have preserved an intact ORF, and the process is complicated by the inability of common sequence-assembly programs to combine sequences with long deletions and a high degree of divergence into a multiple-sequence alignment. Therefore, in contrast to the easily defined ORFs of GilT and GilM, the ORF of GilD is a result of a multistep reconstruction, which included restoration of many deleted regions disrupting the ORF and substitution of frameshifts and stop codons with sequences present in other copies wherever necessary for reconstitution of the reading frame. The restored consensus ORF occupies most of the 3-kb element and can be aligned with the ORFs of the other two elements with an overall 25% identity and 39% similarity, except for the truncated N terminus lacking one of the C2H2 fingers and a short 3' UTR.
There are 12 deletion tracts throughout the entire length of the element (Fig. 2A), and very few (only the shortest) clones have no internal deletions. Any of these deletions introduced into an intact element would abolish its function by disrupting conserved protein motifs. Deletions are apparently mediated by 3-5-bp repeats at the boundaries and range in size from 8 to 60 bp. Many deletions are shared between several copies but are not present in the others containing deletions in other places, suggesting that some of the deleted copies continued to proliferate in trans, using the enzymes provided by still-intact copies. The presence of deletions in some copies but not in the others implies that deletions are not induced by sequence per se, and also indicates that proliferation of deleted copies took place before accumulation of single-nucleotide polymorphisms, most of which are not shared between copies. No tandem arrangement, as for GilT and GilM, is observed nor are there any junctions with telomeric repeats. GilD sequences with shared single-nucleotide polymorphisms (SNPs) are typically repeated 3-6 times in the database, each version representing a diverged unique copy and reflecting the current degree of genome coverage. The copy-number estimate for this family (about 30 per genome, as determined by the number of sequence clusters with shared SNPs) is about 2-fold higher than that of the two telomere-associated families combined. The divergence of GilD sequences from the consensus ranges from 6% to 13%, indicating its inactivation in the distant past. Single-nucleotide substitutions and indels are distributed uniformly and without any bias toward synonymous sites. Adjacent to the coding region at the 5' and 3' ends are
200-bp
imperfect inverted terminal repeats with a low degree of homology, alignments of which also exhibit a mosaic appearance indicative of
frequent recombination or gene conversion. Such repeats are not typical
of LINE-like elements and may not constitute an integral part of the
element but might be present at both ends as a result of
sequence-specific GilD insertion in either direction. A polyadenylation signal [but not the (A)n stretch] can be
identified in at least some copies of the repeat, favoring the
explanation of GilD insertion into its own UTR. GilD is often located
near variant-specific surface protein genes/pseudogenes or other
repetitive genes such as ankyrins. Recombination in these regions might
contribute to generation of antigenic diversity, as described in other
parasitic protozoans (26, 27).
Phylogenetic Placement of Giardia LINEs. In the world of RTases (Fig. 3), LINE families from G. lamblia seem to be phylogenetically closest to those containing the REL-ENDO domain, such as the NeSL and R2 clades (15, 17). The dead GilD family occupies a more basal position than the two functional ones, and all of them form a distinct clade, indicating that they established residence in the G. lamblia genome a long time ago, with the active ones maintaining a high degree of sequence homogeneity.
|
| |
Discussion |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Transposons and Telomeres. Telomeric and subtelomeric regions are believed to represent a particularly suitable environment for harboring transposon insertions, because the latter would not cause much damage to the host by interfering with the function of nuclear genes, and might even confer benefits by expanding the buffer zone between the end of the chromosome and the nearby single-copy genes. Indeed, transposons with insertional specificities for telomeric or subtelomeric regions have been identified in such diverse organisms as Saccharomyces cerevisiae (Ty5), Bombyx mori (SART, TRAS), Chlorella vulgaris (Zepp), and Allium cepa (MP7) (28-31).
Only in Drosophila melanogaster (HeT-A and TART) (32, 33) have retrotransposons completely taken over the function of telomere maintenance. There are no telomeric repeats in this species, and the full genome sequence (34) does not contain coding regions homologous to telomerases. Associated with this function is the ability for terminal transposition, in which the RNA transcript of a retrotransposon attaches directly to the end of the chromosome by means of a poly(A) tail and undergoes reverse transcription in situ, initiated at the 3' hydroxyl end of the chromosome (Fig. 2B). The process repeats itself, resulting in a chain of HeT-A and/or TART retrotransposons (often interspersed), with the same polarity for all members of the chain. Healing of chromosome breaks has been shown to occur by telomere formation associated with terminal transposition of HeT-A and TART (32, 35, 36). The mechanisms underlying direct attachment to chromosome termini and array formation are unknown. Another distinctive feature of HeT-A and TART, found also in GilM and GilT, is an unusually long 3' UTR. Long 3' noncoding regions are highly atypical for retroelements, and their presence in the two Drosophila telomere-associated LINEs prompted speculations that they are required for telomeric chromatin structure and/or terminal transposition (33, 35, 37). HeT-A and TART can also carry tandemly duplicated segments at their 3' ends (36). Interestingly, the ORFs of telomere-associated retrotransposons from Giardia and Drosophila retain a coding capacity for an endonuclease-like protein, which is expected to provide recognition and cleavage of the target sequence. TART has an endonuclease domain belonging to the apurinic-apyrimidinic (AP)-like category (38), whereas the Giardia elements possess the REL-ENDO domain. It is unclear why these domains would be retained in otherwise unrelated families of telomeric transposons. Sequence-specific insertion into telomeric repeats may be excluded, because such repeats are never found at their 3' flanks. Perhaps insertions into internal sites, if they do occur, are rapidly eliminated. Also required for transposition in cis is the nucleic acid-binding capacity, which serves the nucleic acid chaperone function (39-41). This capacity is conserved not only in the autonomous TART element but also in HeT-A, which is nonautonomous because it does not code for its own RTase (22, 36). GilM and GilT are highly homologous in the putative NA-binding region, whereas only one of the C2H2 fingers can be identified in the reconstructed GilD. Overall, structural comparisons show that GilT and GilM bear profound functional resemblance to the telomere-associated retrotransposons of Drosophila, including the ability to form tandem polar arrays at the chromosome ends, the coding capacity for RTase and nucleic acid-binding proteins, and an exceptionally long UTR with the 3'-most region prone to tandem duplications (Fig. 2B). It is quite remarkable that telomere-associated retrotransposons are found in Giardia, considered to be one of the earliest-branching eukaryotes on the basis of numerous molecular phylogenetic studies (e.g., refs. 42-45). It is also notable that in a tandem head to tail arrangement, which is characteristic of terminally transposing HeT-A as well as GilM/GilT, the transposon may no longer be regarded as selfish DNA when the promoter is located in the 3' UTR, because it provides transcription of its downstream neighbor but not itself (25). Studies of GilM/GilT promoter activity will therefore be of significant interest. A single-copy coding sequence for the telomerase catalytic subunit, which is a specialized RTase, has been identified in the G. lamblia genome (46). Its structure differs from other eukaryotic telomerase RTase genes because it lacks the conserved T motif. Its absence might interfere with proper telomerase function by affecting interaction with telomerase RNA (47, 48). It is possible that terminal transposition of LINE elements can to some extent compensate for such deficiencies in telomerase function, so that 5' truncation occurs when LINEs are exposed to terminal degradation until telomerase starts adding TAGGG repeats to their ends. An alternative but less likely possibility is that GilM and GilT may insert sequence-specifically into their own 3' UTRs in the same orientation, somehow losing oligo(A) after such insertion. The integration process does not proceed to completion in its usual sense, however; thus, target-site duplication is not observed and incomplete reverse transcription is followed not by template joining to the other end of the target, but by telomeric repeat addition to the truncated 5' end. This explanation would imply direct coupling of telomerase action, and therefore telomere formation, with the transposition process. It is only after complete assembly of the G. lamblia genome that we will know the degree of variability of its chromosome ends. The minimum number of telomeric repeat junction fragments would be 10; additional junctions may be present in minor chromosome variants, seen in many G. lamblia isolates. A total of 10 restriction fragments was reported to hybridize to the telomeric repeat probe in at least one G. lamblia isolate (24). Telomeric probe also hybridizes to two NotI fragments of chromosome 4 (11). Eleven telomeric junctions, eight of them being LINE elements, were identified in this study; the exact relationship between size variants of different chromosomes and terminal sequences remains to be determined as the assembly progresses.Complete Inactivation of a High-Copy-Number LINE-Like Element. The main structural differences between GilT/GilM and the reconstructed GilD are the absence of the N-terminal C2H2 finger and of the extended 3' UTR region with the (A)n stretch in GilD. If these features are required for terminal transposition and they were initially present in the ancestral GilD transposon, their loss could have resulted in the inability to attach to the chromosome termini in a chain-like fashion. An alternative possibility is that GilM and GilT both acquired this ability, perhaps by means of addition of the long 3' UTR, and therefore persisted in an active state, whereas GilD did not.
Short deletions involving regions of microhomology are known to occur during the error-prone Ku-independent nonhomologous end-joining double-strand break (DSB) repair backup pathway (49). In light of the apparent absence of Ku homologs from the G. lamblia genome, it seems plausible that such deletions could be generated as a result of such error-prone DSB repair. Interestingly, two or three deletion tracts of different sizes can occur in the same limited region, indicating possible DSB hotspots repaired independently by using different microhomologies in the same region.Is Giardia Asexual? Several unicellular organisms, once considered to be entirely asexual, have been found to undergo meiosis or at least form synaptonemal complexes (SC) after more careful examination (50, 51). A recent example is the pathogenic yeast Candida albicans, for which whole-genome sequencing data in combination with experimental stimulation revealed the potential to undergo at least part of a sexual cycle (52-54). Is it possible that Giardia, although thought to be asexual (7-9), might also have some form of sexual process? Its genome sequence does contain coding regions with homology to several genes known in other organisms to be involved in meiotic recombination (RAD51, RAD52, rec14, SPO11, and DMC1) or meiosis initiation and regulation (SME1/IME2, MEK1, and ran1 +) (www.mbl.edu/Giardia/Giardia-Total-BlastX/blastreport.html). ORFs similar to SC proteins (HOP1, ZIP1/SCP1) can also be identified; these, however, are most similar to other coiled-coil proteins (e.g., myosins and kinesins) and their function cannot be established at this time on the basis of sequence similarity alone. In addition, SC formation is not always required for a sexual cycle (55). It is worth noting that Candida, even if it has a potential to undergo sexual reproduction, is still mostly asexual, because its populations are primarily clonal in structure and genetic exchange is infrequent (56). This finding seems to be correlated with low active transposon content. In contrast to S. cerevisiae, which has mostly intact retrotransposons belonging to a few families with multiple members, the C. albicans genome contains 35 retrotransposon families, each having only a few highly rearranged and defective members. Only two or three copies appear intact (57). Thus, even when sexual reproduction is not completely excluded from the lifestyle of the organism, its prolonged absence may influence the transposon content of the genome. Because Giardia populations seem to be mostly clonal (58), it is likely that sexual reproduction, if any, did not play a major role in shaping its genome structure.
In conclusion, our previous studies of transposon content in sexual and anciently asexual rotifers and numerous other eukaryotes strongly suggested that the genomes of ancient asexuals do not retain RTase-related sequences detectable in PCR assays. The present analysis of transposon content of G. lamblia at the level of the entire genome sequence demonstrates that its two intact retroelements are confined to telomeric regions and therefore neither cause deleterious insertional mutations nor serve as sites for ectopic rearrangements in internal chromosomal locations. Because they are found on the majority of the chromosome ends, they could be beneficial to the host by providing additional protection from terminal degradation, even though they have not entirely replaced telomeric repeats, as has happened in Drosophila. A complete inactivation of the third family has occurred, as evidenced by multiple deletions and point mutations. Together with the absence of other autonomous transposon-related sequences, it may be concluded that the genome of this protozoan is free of active deleterious transposons. In combination with our earlier PCR experiments in bdelloid rotifers, this study establishes a connection between the mode of reproduction and the abundance, activity, and role of transposable elements in eukaryotic genomes.| |
Acknowledgements |
|---|
We thank M. Meselson for encouragement and support; D. Mark Welch, J. Mark Welch, and M. Meselson for valuable comments; A. McArthur for critical reading and for performing hidden Markov searches of G. lamblia contigs to verify the absence of transposases and Ku homologs; M. Sogin for permission to use unpublished contig data and clones from the library; the National Science Foundation for supporting studies of transposable elements (MCB-9905998); and the National Institutes of Health for supporting the Giardia genome project (AI43273).
| |
Abbreviations |
|---|
LINE, long interspersed nuclear element; RTase, reverse transcriptase; UTR, untranslated region.
| |
Footnotes |
|---|
To whom reprint requests should be addressed. E-mail:
arkhipov{at}fas.harvard.edu.
Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. AF433875-AF433877).
See commentary on page 14195.
| |
References |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
| 1. |
Hickey, D.
(1982)
Genetics
101,
519-531 |
| 2. |
Arkhipova, I.
& Meselson, M.
(2000)
Proc. Natl. Acad. Sci. USA
97,
14473-14477 |
| 3. | Arkhipova, I. , Lyubomirskaya, N. & Ilyin, Y. (1995) Drosophila Retrotransposons (Landes, Austin, TX). |
| 4. | Capy, P. , Bazin, C. , Higuet, D. & Langin, T. (1998) Dynamics and Evolution of Transposable Elements (Landes, Austin, TX). |
| 5. |
Mark Welch, D.
& Meselson, M.
(2000)
Science
288,
1211-1215 |
| 6. | Bestor, T. (1999) Genetica 107, 289-295[CrossRef][ISI][Medline] . |
| 7. | Adam, R. (2000) Int. J. Parasitol. 30, 475-484[CrossRef][ISI][Medline] . |
| 8. |
Adam, R.
(2001)
Clin. Microbiol. Rev.
14,
447-475 |
| 9. | Upcroft, J. & Upcroft, P. (2000) Protist 150, 17-23. |
| 10. |
Hou, G.
, Le Blancq, S., E. Y.
, Zhu, H.
& Lee, M.
(1995)
Nucleic Acids Res.
23,
3310-3317 |
| 11. | Le Blancq, S. & Adam, R. (1998) Mol. Biochem. Parasitol. 97, 199-208[CrossRef][ISI][Medline] . |
| 12. | McArthur, A. , Morrison, H. , Nixon, J. , Passamaneck, N. , Kim, U. , Hinkle, G. , Crocker, M. , Holder, M. , Farr, R. , Reich, C. , et al. (2000) FEMS Microbiol. Lett. 189, 271-273[CrossRef][ISI][Medline] . |
| 13. |
Huelsenbeck, J.
& Ronquist, F.
(2001)
Bioinformatics (Oxford)
17,
754-755 |
| 14. |
Adam, R.
, Nash, T.
& Wellems, T.
(1991)
Mol. Cell. Biol.
11,
3326-3330 |
| 15. | Malik, H. , Burke, W. & Eickbush, T. (1999) Mol. Biol. Evol. 16, 793-805[Abstract]. |
| 16. |
Yang, J.
, Malik, H.
& Eickbush, T.
(1999)
Proc. Natl. Acad. Sci. USA
96,
7847-7852 |
| 17. |
Malik, H.
& Eickbush, T.
(2000)
Genetics
154,
193-203 |
| 18. |
Aksoy, S.
, Williams, S.
, Chang, S.
& Richards, F.
(1990)
Nucleic Acids Res.
18,
785-792 |
| 19. |
Gabriel, A.
, Yen, T.
, Schwartz, D.
, Smith, C.
, Boeke, J.
, Sollner-Webb, B.
& Cleveland, D.
(1990)
Mol. Cell. Biol.
10,
615-624 |
| 20. |
Villanueva, M.
, Williams, S.
, Beard, C.
, Richards, F.
& Aksoy, S.
(1991)
Mol. Cell. Biol.
11,
6139-6148 |
| 21. | Hatfield, D. & Oroszlan, S. (1990) Trends Biochem. Sci. 15, 186-190[CrossRef][Medline] . |
| 22. | Pardue, M. , Danilevskaya, O. , Lowenhaupt, K. , Wong, J. & Erby, K. (1996) J. Mol. Evol. 43, 572-583[CrossRef][ISI][Medline] . |
| 23. |
Moran, J.
, DeBerardinis, R.
& Kazazian, H.
(1999)
Science
283,
1530-1534 |
| 24. |
Upcroft, P.
, Chen, N.
& Upcroft, J.
(1997)
Genome Res.
7,
37-46 |
| 25. | Danilevskaya, O. , Arkhipova, I. , Traverse, K. & Pardue, M. (1997) Cell 88, 647-655[CrossRef][ISI][Medline] . |
| 26. |
Borst, P.
& Rudenko, G.
(1994)
Science
264,
1872-1873 |
| 27. | Freitas-Junior, L. , Bottius, E. , Pirrit, L. , Deitsch, K. , Scheidig, C. , Guinet, F. , Nehrbass, U. , Wellems, T. & Scherf, A. (2000) Nature (London) 407, 1018-1021[CrossRef][Medline] . |
| 28. | Noutoshi, Y. , Arai, R. , Fujie, M. & Yamada, T. (1998) Mol. Gen. Genet. 259, 256-263[CrossRef][ISI][Medline] . |
| 29. | Pich, U. & Schubert, I. (1998) Chromosome Res. 6, 315-321[CrossRef][ISI][Medline] . |
| 30. |
Zhu, Y.
, Zou, S.
, Wright, D.
& Voytas, D.
(1999)
Genes Dev.
13,
2738-2749 |
| 31. |
Anzai, T.
, Takahashi, H.
& Fujiwara, H.
(2001)
Mol. Cell. Biol.
21,
100-108 |
| 32. | Biessmann, H. , Mason, J. , Ferry, K. , d'Hulst, M. , Valgeirsdottir, K. , Traverse, K. & Pardue, M. (1990) Cell 61, 663-673[CrossRef][ISI][Medline] . |
| 33. | Levis, R. , Ganesan, R. , Houtchens, K. , Tolar, L. & Sheen, F. (1993) Cell 75, 1083-1093[CrossRef][ISI][Medline] . |
| 34. |
Adams, M.
, Celniker, S. E.
, Holt, R. A.
, Evans, C. A.
, Gocayne, J. D.
, Amanatides, P. G.
, Scherer, S. E.
, Li, P. W.
, Hoskins, R. A.
, Galle, R. F.
,
et al.
(2000)
Science
287,
2185-2195 |
| 35. |
Biessmann, H.
, Valgeirsdottir, K.
, Lofsky, A.
, Chin, C.
, Ginther, B.
, Levis, R.
& Pardue, M.
(1992)
Mol. Cell. Biol.
12,
3910-3918 |
| 36. |
Sheen, F.
& Levis, R.
(1994)
Proc. Natl. Acad. Sci. USA
91,
12510-12514 |
| 37. | Biessmann, H. & Mason, J. (1997) Chromosoma 106, 63-69[CrossRef][ISI][Medline] . |
| 38. | Martin, F. , Olivares, M. , Lopez, M. & Alonso, C. (1996) Trends Biochem. Sci. 21, 283-285[Medline] . |
| 39. | Feng, Q. , Moran, J. , Kazazian, H. & Boeke, J. (1996) Cell 87, 905-916[CrossRef][ISI][Medline] . |
| 40. | Dawson, A. , Hartswood, E. , Paterson, T. & Finnegan, D. (1997) EMBO J. 16, 4448-4455[CrossRef][ISI][Medline] . |
| 41. |
Martin, S.
& Bushman, F.
(2001)
Mol. Cell. Biol.
21,
467-475 |
| 42. |
Sogin, M.
, Gunderson, J.
, Elwood, H.
, Alonso, R.
& Peattie, D.
(1989)
Science
243,
75-77 |
| 43. | Henze, K. , Morrison, H. , Sogin, M. & Muller, M. (1998) Gene 222, 163-168[CrossRef][ISI][Medline] . |
| 44. | Roger, A. , Morrison, H. & Sogin, M. (1999) J. Mol. Evol. 48, 750-755[CrossRef][ISI][Medline] . |
| 45. | Bouzat, J. , McNeil, L. , Robertson, H. , Solter, L. , Nixon, J. , Beever, J. , Gaskins, H. , Olsen, G. , Subramaniam, S. , Sogin, M. & Lewin, H. (2000) J. Mol. Evol. 51, 532-543[ISI][Medline] . |
| 46. | Malik, H. , Burke, W. & Eickbush, T. (2000) Gene 251, 101-108[CrossRef][ISI][Medline] . |
| 47. |
Friedman, K.
& Cech, T.
(1999)
Genes Dev.
13,
2863-2874 |
| 48. |
Lai, C.
, Mitchell, J.
& Collins, K.
(2001)
Mol. Cell. Biol.
21,
990-1000 |
| 49. |
Feldmann, E.
, Schmiemann, V.
, Goedecke, W.
, Reichenberger, S.
& Pfeiffer, P.
(2000)
Nucleic Acids Res.
28,
2585-2596 |
| 50. | Raikov, I. (1995) Eur. J. Protistol. 31, 1-7. |
| 51. | Dacks, J. & Roger, A. (1999) J. Mol. Evol. 48, 779-783[CrossRef][ISI][Medline] . |
| 52. |
Hull, C.
, Raisner, R.
& Johnson, A.
(2000)
Science
289,
307-310 |
| 53. |
Magee, B.
& Magee, P.
(2000)
Science
289,
310-313 |
| 54. |
Tzung, K.
, Williams, R.
, Scherer, S.
, Federspiel, N.
, Jones, T.
, Hansen, N.
, Bivolarevic, V.
, Huizar, L.
, Komp, C.
, Surzycki, R.
,
et al.
(2001)
Proc. Natl. Acad. Sci. USA
98,
3249-3253 |
| 55. | Kohli, J. & Bahler, J. (1994) Experientia 50, 295-306[CrossRef][Medline] . |
| 56. |
Graser, Y.
, Volovsek, M.
, Arrington, J.
, Schonian, G.
, Presber, W.
, Mitchell, T.
& Vilgalys, R.
(1996)
Proc. Natl. Acad. Sci. USA
93,
12473-12477 |
| 57. |
Goodwin, T.
& Poulter, R.
(2000)
Genome Res.
10,
174-191 |
| 58. |
Tibayrenc, M.
, Kjellberg, F.
, Arnaud, J.
, Oury, B.
, Breniere, S.
, Darde, M. L.
& Ayala, F.
(1991)
Proc. Natl. Acad. Sci. USA
88,
5129-5133 |
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg What's this?
Related Commentary in PNAS:
This article has been cited by other articles in HighWire Press-hosted journals:
|
|
 |
|
  S. Shpiz, D. Kwon, A. Uneva, M. Kim, M. Klenov, Y. Rozovsky, P. Georgiev, M. Savitsky, and A. Kalmykova Characterization of Drosophila Telomeric Retroelement TAHRE: Transcription, Transpositions, and RNAi-based Regulation of Expression Mol. Biol. Evol., November 1, 2007; 24(11): 2535 - 2545. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Omilian, M. E. A. Cristescu, J. L. Dudycha, and M. Lynch Ameiotic recombination in asexual lineages of Daphnia PNAS, December 5, 2006; 103(49): 18638 - 18643. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Rehmeyer, W. Li, M. Kusaba, Y.-S. Kim, D. Brown, C. Staben, R. Dean, and M. Farman Organization of chromosome ends in the rice blast fungus, Magnaporthe oryzae Nucleic Acids Res., October 18, 2006; 34(17): 4685 - 4701. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Savitsky, D. Kwon, P. Georgiev, A. Kalmykova, and V. Gvozdev Telomere elongation is under the control of the RNAi-based mechanism in the Drosophila germline Genes & Dev., February 1, 2006; 20(3): 345 - 354. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Berriman, E. Ghedin, C. Hertz-Fowler, G. Blandin, H. Renauld, D. C. Bartholomeu, N. J. Lennard, E. Caler, N. E. Hamlin, B. Haas, et al. The Genome of the African Trypanosome Trypanosoma brucei Science, July 15, 2005; 309(5733): 416 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Ullu, H. D. Lujan, and C. Tschudi Small Sense and Antisense RNAs Derived from a Telomeric Retroposon Family in Giardia intestinalis Eukaryot. Cell, June 1, 2005; 4(6): 1155 - 1157. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Upcroft, M. Abedinia, and P. Upcroft Rearranged Subtelomeric rRNA Genes in Giardia duodenalis Eukaryot. Cell, February 1, 2005; 4(2): 484 - 486. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Fillingham, T. A. Thing, N. Vythilingum, A. Keuroghlian, D. Bruno, G. B. Golding, and R. E. Pearlman A Non-Long Terminal Repeat Retrotransposon Family Is Restricted to the Germ Line Micronucleus of the Ciliated Protozoan Tetrahymena thermophila Eukaryot. Cell, February 1, 2004; 3(1): 157 - 169. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. K. Mandal, A. Bagchi, A. Bhattacharya, and S. Bhattacharya An Entamoeba histolytica LINE/SINE Pair Inserts at Common Target Sites Cleaved by the Restriction Enzyme-Like LINE-Encoded Endonuclease Eukaryot. Cell, February 1, 2004; 3(1): 170 - 179. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Wickstead, K. Ersfeld, and K. Gull Repetitive Elements in Genomes of Parasitic Protozoa Microbiol. Mol. Biol. Rev., September 1, 2003; 67(3): 360 - 375. [Abstract] [Full Text] [PDF]
|
|
|
|
