Genetics, Vol. 163, 747-758, February 2003, Copyright © 2003

Genome-Wide Analysis of mariner-Like Transposable Elements in Rice Reveals Complex Relationships With Stowaway Miniature Inverted Repeat Transposable Elements (MITEs)

Cédric Feschottea, Lakshmi Swamy2,a, and Susan R. Wesslera
a Departments of Plant Biology and Genetics, The University of Georgia, Athens, Georgia 30602

Corresponding author: Cédric Feschotte, University of Georgia, Athens, GA 30602., cedric{at}dogwood.botany.uga.edu (E-mail)

Communicating editor: M. J. SIMMONS


*  ABSTRACT
*TOP
 *ABSTRACT
*MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

Stowaway is a superfamily of miniature inverted repeat transposable elements (MITEs) that is widespread and abundant in plant genomes. Like other MITEs, however, its origin and mode of amplification are poorly understood. Several lines of evidence point to plant mariner-like elements (MLEs) as the autonomous partners of the nonautonomous Stowaway MITEs. To better understand this relationship, we have taken advantage of the nearly complete genome sequences of two rice subspecies to generate the first inventory of virtually all MLEs and Stowaway families coexisting in a single plant species. Thirty-four different MLEs were found to group into three major clades and 25 families. More than 22,000 Stowaway MITEs were identified and classified into 36 families. On the basis of detailed sequence comparisons, MLEs were confirmed to be the best candidate autonomous elements for Stowaway MITEs. Surprisingly, however, sequence similarity between MLE and Stowaway families was restricted to the terminal inverted repeats (TIRs) and, in a few cases, to adjacent subterminal sequences. These data suggest a model whereby most of the Stowaway MITEs in rice were cross-mobilized by MLE transposases encoded by distantly related elements.

Tc1/mariner is a diverse and widespread superfamily of eukaryotic class 2 transposable elements (reviewed in CAPY et al. 1998 Down; PLASTERK et al. 1999 Down; PLASTERK and VAN LUENEN 2002 Down). One hallmark of the superfamily is insertion into the dinucleotide TA that is duplicated upon insertion and flanks the element as a target site duplication (TSD). Tc1/mariner elements are relatively short (1.2–3.5 kb) and are simple in structure with terminal inverted repeats (TIRs) and a single gene encoding the transposase. A common model for the transposition mechanism of Tc1/mariner elements has emerged from the functional study of a limited number of animal transposases (PLASTERK and VAN LUENEN 2002 Down). The N-terminal region of Tc1/mariner transposases contains DNA-binding domain(s) that bind specifically to the TIRs (PLASTERK et al. 1999 Down; LAMPE et al. 2001 Down; ZHANG et al. 2001 Down). A C-terminal domain is characterized by an amino acid signature called the DDE/D motif consisting of two aspartic acid residues and a glutamic acid residue (or a third D). This motif is required for catalysis of both the DNA cleavage and the strand transfer steps of the "cut and paste" transposition reaction (reviewed in HARTL et al. 1997 Down; PLASTERK and VAN LUENEN 2002 Down).

Tc1/mariner elements were recently found to be widespread in plants (reviewed in FESCHOTTE et al. 2002A Down). The first reported plant members were Soymar1, a mariner-like element (MLE) from soybean (JARVIK and LARK 1998 Down) and Lemi1, a pogo-like element from Arabidopsis thaliana (FESCHOTTE and MOUCHES 2000 Down). Three additional rice MLEs were subsequently identified by database searches, but none were characterized further (TARCHINI et al. 2000 Down; SHAO and TU 2001 Down; TURCOTTE et al. 2001 Down; FESCHOTTE and WESSLER 2002 Down). These five elements were used to derive plant-specific primers that successfully amplified MLE transposase genes in PCR assays with DNA from a wide spectrum of flowering plant genomes (FESCHOTTE and WESSLER 2002 Down). For the majority of genomes assayed, multiple divergent lineages of transposases were amplified from single species.

Demonstration that MLEs are widespread and diverse in plants provided support for the hypothesis that MLEs are the autonomous elements responsible for the origin and spread of Stowaway, a large group of miniature inverted repeat transposable elements (MITEs; BUREAU and WESSLER 1994 Down). MITEs are structurally reminiscent of class 2 nonautonomous elements with their small size (<600 bp), lack of coding capacity, and TIRs (reviewed in FESCHOTTE et al. 2002B Down). However, their high copy number and structural homogeneity have served to distinguish them from most of the previously described class 2 elements (WESSLER et al. 1995 Down). MITEs were first discovered in plants, where they are now recognized as the predominant type of transposable element associated with the noncoding regions of plant genes. This is particularly evident in the cereals, including rice, maize, barley, and wheat (BENNETZEN 2000 Down; FESCHOTTE et al. 2002A Down; GOFF et al. 2002 Down; YU et al. 2002 Down). Vast amounts of MITEs have also been discovered in many invertebrate and vertebrate genomes (reviewed in FESCHOTTE et al. 2002B Down).

Most of the tens of thousands of MITEs in plant genomes have been divided into two groups on the basis of the similarity of their TIRs and TSDs: Tourist-like MITEs and Stowaway-like MITEs (WESSLER et al. 1995 Down; FESCHOTTE et al. 2002B Down). That Stowaway-like MITEs and plant MLEs share similar terminal sequences (5'-CTCCCTCCRT-3', where R stands for A or G) and target site preference (TA) strongly suggested that Stowaway MITEs were mobilized in trans by transposases encoded by MLEs (TURCOTTE et al. 2001 Down; FESCHOTTE et al. 2002B Down). A model was formulated that hypothesized that Stowaway elements originated by internal deletion(s) from a larger autonomous element (like previously described nonautonomous DNA elements) and were amplified to very high copy number by the transposase encoded by the autonomous element (FESCHOTTE et al. 2002B Down). The diversity of Stowaway families observed in a single genome was explained by proposing that the families originated as deletion derivatives of distinct lineages of MLEs (FESCHOTTE et al. 2002A Down, FESCHOTTE et al. 2002B Down). If this model is correct, one should encounter Stowaway families that have extensive sequence similarity (i.e., not just in their termini) with MLEs present in the same genome. In addition, the diversity of Stowaway families should correspond with a similar diversity of MLEs in that same genome. Failure to match Stowaway families with MLEs would indicate that the model was incorrect or overly simplistic.

Comparison of all of the MLEs and Stowaway elements in a genome is possible only for Arabidopsis and rice for which entire genome sequences are available. Although remnants of MLE transposases are still recognizable in the sequence of A. thaliana, no full-length MLEs are identifiable (SHAO and TU 2001 Down; FESCHOTTE and WESSLER 2002 Down; C. FESCHOTTE, unpublished data). Furthermore, Stowaway MITEs are relatively scarce in this species (at least in the sequenced ecotype), with <250 copies organized into fewer than five families (LE et al. 2000 Down; C. FESCHOTTE, unpublished data). In contrast, previous searches of a limited amount of rice genomic sequence identified numerous families of Stowaway MITEs and full-length MLEs (BUREAU et al. 1996 Down; JIANG and WESSLER 2001 Down; SHAO and TU 2001 Down; TURCOTTE et al. 2001 Down; FESCHOTTE and WESSLER 2002 Down). For these reasons, the goal of this study was to characterize all MLE and Stowaway families in rice and determine the extent of sequence relatedness between these two groups.

A semiautomated computational approach was used to identify and compare MLEs and Stowaway MITEs in the two draft genome sequences of rice (GOFF et al. 2002 Down; YU et al. 2002 Down). In this way 34 MLEs were identified, with 22 considered full-length, as they contain a complete transposase coding region, TIRs, and TSD. Phylogenetic analysis and other criteria, such as the presence or absence of introns, led to their grouping into 25 distinct families falling into three major clades. In addition, up to 33,000 Stowaway MITEs were identified, with the high-copy-number elements grouping into at least 36 families. Surprisingly, none of the 25 MLE families could be associated by simple internal deletion with any of the Stowaway families. Instead, sequence similarity between Stowaway and MLE families was restricted to the TIRs and, in a few cases, to some adjacent subterminal sequence. These data have led us to conclude that most of the Stowaway MITEs in rice were probably cross-mobilized by MLE transposases encoded by distantly related elements.


*  MATERIALS AND METHODS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
*RESULTS
 *DISCUSSION
 *LITERATURE CITED

Semiautomated mining of full-length rice MLEs:
A series of Perl scripts was written to automate the process of identifying and fetching full-length elements related to a particular transposase. In a first step, the transposase amino acid sequence is used as a query in a local WU-TBLASTN search (http://blast.wustl.edu) against a genomic database. The output file is parsed and the significant hits (in this study, E values <10-5) are extracted from the database along with up to 10 kb of flanking DNA sequence. In a second step, the flanking sequences are searched for the possible ends of the elements using a subroutine called MATCH-TIR. This program scans the 5' and 3' flanking regions of each hit with a 16-mer sliding window for the presence of a consensus motif corresponding to the 5' and 3' ends of the element plus the expected target site duplications (user input). MATCH-TIR extracts 5' and 3' hits (sequences with >80% similarity to the motif) along with 50 nucleotides internal to the hits and produces pairwise alignments between 5' extended hits and the reverse complement of 3' extended hits. The alignments are inspected visually and the best matching pairs (usually fewer than four mismatches in the first 22 nucleotides) are considered as the TIRs of the element. In this study, the Osmar1 transposase sequence was used as the query in a WU-TBLASTN search against two databases. The first database contained ~360 Mb of bacterial artificial chromosome (BAC)/P1-derived artificial chromosome (PAC) sequences from Oryza sativa ssp. japonica cv. Nipponbare (downloadable at http://rgp.dna.affrc.go.jp/cgi-bin/statusdb/seqcollab-assign.pl). The second database contained ~430 Mb of contigs generated by whole-genome sequencing of O. sativa ssp. indica cv. 9311 (downloadable at http://btn.genomics.org.cn/rice/). The motif 5'-TACTCCCTCCAG-3' and its reverse complement were used for MATCH-TIR searches of the 5' and 3' ends of rice MLEs, respectively. Other searches were performed using WU-BLASTN against the two databases described above and a third database containing the whole-genome shotgun assembly of O. sativa ssp. japonica cv. Nipponbare produced by Syngenta (390 Mb; http://www.tmri.org).

Compilation of Stowaway families and copy number determinations:
Twenty-four Stowaway families analyzed in this study were previously published (from Stow-Os1-Os32; BUREAU and WESSLER 1994 Down; BUREAU et al. 1996 Down; JIANG and WESSLER 2001 Down; TURCOTTE et al. 2001 Down). These 24 families were compared to a collection of rice repeats identified de novo by the program RECON in ~30 Mb of BAC/PAC sequences of O. sativa ssp. japonica (BAO and EDDY 2002 Down). RECON identified and computed consensus sequences for the 24 previously recognized Stowaway families and for four newly identified families (from Stow-Os1-Os37; see Fig 3). An additional family (Stow-Os38) was identified through BLASTN searches with the subterminal regions of Osmar4. Jerzy Jurka and A. Drazkiewicz (Repbase; http://www.girinst.org) contributed 7 additional families (Stow-Os42-Os52). Copy numbers of Stowaway families were estimated for 360 Mb of BAC/PAC sequences from O. sativa japonica by two different methods and extrapolated to a genome size of 430 Mb. In the first method, the database was analyzed with RepeatMasker using the compilation of Stowaway consensus described above. Crude values obtained from this search were refined by dividing the number of hits by two for queries with TIRs >45 bp (these sequences will produce systematically two hits per position, one from each strand). To correct for multiple hits due to interfamily similarity, we combined families predicted to cross-hit (high level of similarity in their TIRs) and considered the highest value obtained for these families as the copy number of the combined families. This method gave values corresponding to the upper estimate in the range shown in Fig 3. The lower estimate was obtained by counting the number of hits produced in BLASTN and FASTA searches using each consensus as a query against the same database with default parameters. Each BAC was counted as a hit if it contained a sequence matching >50% of the length of the query with at least 85% similarity. This method gave a more conservative estimate partly because BAC/PAC sequences containing multiple family members produce a single hit.



View larger version (41K):
In this window
In a new window
Download PPT slide
  		Figure 1. Phylogenetic relationships and structures of rice MLEs. The neighbor-joining tree was generated from a multiple alignment of conceptually translated transposase sequences of 34 rice MLEs (Osmars; see supplemental table for accession numbers at http://www.genetics.org/supplemental) and Soymar1 from soybean (GenBank accession no. AF078934), which served as the outgroup. Bootstrap values >60 are shown as a percentage of 1000 replicates. Underlined names denote elements with potentially intact transposase genes. All Osmars are from the subspecies japonica, except those followed by i, which are from the subspecies indica. Capital letters and different colors emphasize different lineages and sublineages of Osmars. Arrowheads indicate the presence of a particular intron prior to the divergence of a lineage or a sublineage. The structure of the corresponding MLE is depicted on the right. Full-length elements are delimited by terminal inverted repeats (solid triangles). Other elements are incomplete due to secondary mutations or to an interruption in the indica contig sequence (this latter situation is shown by a double vertical bar). Transposase coding sequences are depicted as solid boxes and the position of the DD39D triad is shown. Soymar1 and Osmars of the sublineage A1 harbor an intronless transposase gene while other Osmars are predicted to contain one or two introns (shown as an open triangle below the element). Introns occur at four different positions ({alpha}, ß, {phi}, and {lambda}), which are specific for a lineage or sublineage of transposase. Insertions of other repeats in Osmars are shown as shaded triangles above the element along with the insertion size.



View larger version (65K):
In this window
In a new window
Download PPT slide
  		Figure 2. Classification of Osmars based on the TIR. Osmars are classified into four clades (A1, A2, B, and C) on the basis of the sequence of a 4-bp motif in their TIR (boxed). This division is supported by the phylogenetic analysis (see Fig 1). The diagnostic consensus motifs are shown as white letters on a black background. An asterisk indicates that the size of Osmar was calculated after removal of nested TE insertions (see Fig 1). The sign > is used for the size of incomplete Osmars, where only one terminus could be identified. The open triangle above the 5' TIR of Osmar8A and the 3' TIR of Osmar27i denotes an insertion of 5 and 2 bp, respectively, removed from the TIR sequence.



View larger version (66K):
In this window
In a new window
Download PPT slide
  		Figure 3. Classification of Stowaways based on the TIR. A consensus sequence was derived for each high-copy-number Stowaway family. Sources and methods of collecting these sequences are detailed in MATERIALS AND METHODS. Stowaway families were classified into four major clades on the basis of the same 4-bp motifs used to classify Osmars (see Fig 2). Copy numbers were estimated using two methods, producing a lower and upper estimate for each family or a group of closely related families (see MATERIALS AND METHODS).

Sequence and phylogenetic analyses:
Rice MLEs were conceptually translated in the six reading frames with MacVector (http://www.accelrys.com/products/macvector/). Transposase open reading frames (ORFs) were assembled by removing introns predicted with >85% confidence by NetGene2 (http://www.cbs.dtu.dk) and/or FGENESH (http://genomic.sanger.ac.uk/gf/gf.html). When necessary, frameshifts were judiciously introduced according to nucleotide alignments of closely related elements. Putative initiation codons were predicted by NetStart (http://www.cbs.dtu.dk). The resulting transposase sequences were aligned with ClustalW using default parameters in MacVector 7.0. Phylogenetic trees were generated with PAUP* version 4.0b8 (http://paup.csit.fsu.edu/) using the neighbor-joining and maximum parsimony methods with default parameters and rooted with the distantly related Soymar1 transposase from soybean. Sequence comparisons of Osmar and Stowaway elements were carried out using the LFASTA and BLAST2 servers available at http://www.infobiogen.fr.


*  RESULTS
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
*DISCUSSION
 *LITERATURE CITED

Extracting MLEs from rice genomic sequence:
Prior analyses of small fractions of the rice genome identified four MLEs and several Stowaway families (BUREAU et al. 1996 Down; MAO et al. 2000 Down; TARCHINI et al. 2000 Down; JIANG and WESSLER 2001 Down; SHAO and TU 2001 Down; TURCOTTE et al. 2001 Down; FESCHOTTE and WESSLER 2002 Down). The amount of rice sequences available in publicly accessible databases has increased dramatically since those studies and now encompasses nearly two complete genomes from two O. sativa subspecies: japonica (cv. Nipponbare) and indica (cv 93-11; see MATERIALS AND METHODS for details of the databases used in this analysis). This vast resource has been exploited to identify, classify, and compare MLEs and Stowaway MITEs coexisting within the rice genome.

To compare MLE and Stowaway families, it was first necessary to obtain full-length MLEs including complete ORFs and flanking TIRs and TSDs. The strategy employed is detailed in MATERIALS AND METHODS. The putative transposase sequence of Osmar1 was used as the query in TBLASTN searches against two different databases. The first database contained ~360 Mb of BAC/PAC sequences generated from O. sativa ssp. japonica (cv. Nipponbare) by the International Rice Genome Sequencing Project (IRGSP). The second database was the draft genome sequence of O. sativa ssp. indica (cv. 9311) recently released by the Beijing Genomics Institute (BGI; ~420 Mb of shotgun sequence; YU et al. 2002 Down). After manual filtering of redundant hits, a total of 39 sequences with significant similarity to Osmar1 transposase were identified (E values <10-5). To define the ends of the corresponding MLEs, 5 kb flanking each hit was searched for TIRs similar to those of previously identified rice MLEs and Stowaway MITEs. Elements with perfect or near perfect TIRs of >20 bp and large ORF(s) encoding the putative transposase were extracted from the database along with 50 bp of flanking genomic sequence and used as queries in BLASTN searches against the IRGSP and BGI databases and the whole-genome shotgun assembly of O. sativa japonica cv. Nipponbare produced by Syngenta (GOFF et al. 2002 Down). These BLASTN searches enabled us to isolate incomplete and/or noncoding copies and determine whether elements isolated from japonica and indica were present at orthologous positions (for this study orthologous MLEs are considered as the same insertion event). A list of all identified MLEs, their accession numbers, and coordinates are available in a supplemental table (available at http://www.genetics.org/supplemental).

Twenty-two MLEs, ranging in size from 3167 to 11290 bp, were classified as full length because they contained ORF(s) corresponding to the transposase, had TIRs ranging from 20 to 36 bp (with fewer than four mismatches in most cases), and were flanked by a TA target site duplication (Fig 1 and Fig 2). Searches with RepeatMasker and BLASTN revealed that other transposable elements had inserted into a few of the MLEs (Fig 1). For example, a 1795-bp Mutator-like element was found in Osmar4 while Osmar7 contained a 2708-bp insertion consisting of a Tourist-like MITE nested into a solo LTR from the retrotransposon RIRE1. By excluding secondary insertions in size determinations, full-length MLEs ranged from 3167 to 7072 bp. Ten additional MLEs appear to contain a full-length transposase gene and a substantial amount of subterminal sequence (see Fig 1). However, these elements were missing one or both termini due to either secondary mutations or rearrangements after insertion (such as large deletions or insertions) or gaps in the whole-genome sequence assembly from the BGI.

Phylogenetic analysis and classification of Osmar elements:
As a first level of classification, MLEs were grouped into the same family when they shared >85% similarity over their entire length. Using these criteria, 25 different families of MLEs were distinguished (Fig 1 and Fig 2). Consistent with the nomenclature introduced in animals, rice MLEs were designated Osmar (for O. sativa mariner) followed by the number of the family. Members of the same family were further designated by capital letters (for example, Osmar1A and Osmar1B; see Fig 1).

A phylogenetic analysis of transposase sequences was carried out to resolve evolutionary relationships among rice MLE families. Conceptual translation and multiple alignments of 34 Osmar transposases revealed that most (27/34) are corrupted by substitutions and small insertions/deletions (indels) that introduced premature stop codons in the protein sequence. However, several Osmars had intact ORFs and may encode active transposase (names underlined in Fig 1). After removal of predicted introns (see below) and, where necessary, introduction of frameshifts to restored ORFs, putative full-length Osmar transposases were found to range in size from 432 to 505 residues with pairwise amino acid identities that varied from 36% (Osmar13 vs. Osmar26i) to 99% (Osmar5A vs. Osmar5Bi). The most conserved region is a central domain of ~150 residues that is roughly delimited by the DD39D motif (see multiple alignment provided as supplemental data at http://www.genetics.org/supplemental). This motif is characteristic of plant MLE transposases (SHAO and TU 2001 Down; FESCHOTTE and WESSLER 2002 Down) and is found intact in 30 out of 34 putative Osmar proteins. Phylogenetic trees were generated from a multiple alignment of 34 Osmar transposases and the Soymar1 transposase using the neighbor-joining and maximum parsimony methods. Both methods produced trees with very similar topology that defined three major clades of Osmar transposase (A, B, and C in Fig 1). Clade A (10 families, 11 sequences) and clade B (13 families, 19 sequences) were more abundant and diverse than clade C (2 families, 4 sequences) and can be further divided into subclades with strong bootstrap values (A1, A2, etc.; see Fig 1).

An analysis of the positions of predicted introns in Osmar transposase genes provides additional support for the phylogenetic groupings. Four different introns (called {alpha}, ß, {phi}, and {lambda}) were associated with Osmar transposases and result in genes with zero, one, or two introns. When the distribution of these introns was superimposed on the transposase phylogeny, each type of intron was found to be specific to a clade or to a subclade of transposases (Fig 1). That is, intron-{alpha} was restricted to clade B, intron-ß to clade C, intron-{phi} to subclade A2, and intron-{lambda} to subclade B1.

The phylogenetic organization of Osmars is also supported by a comparison of their TIRs. For all Osmar elements, the first 10 bp of the TIRs are well conserved and match the consensus 5'-CTCCCTCCRT-3' (Fig 2). Adjacent to this motif is a 4-bp sequence that serves to define a subset of Osmars. There is a striking correspondence between these groupings and those defined by the phylogenetic groupings of transposases (compare groups in Fig 2 and phylogeny in Fig 1). Indeed, all Osmars in subclade A1 have a TTCG motif in their TIRs while Osmars in subclade A2 display a consensus ACTC motif. Osmars clustered in clade B are characterized by a CCCA motif and those falling in clade C are characterized by TCCT. That each motif is diagnostic of an Osmar transposase clade (or subclade) suggests coevolution between transposase and TIR sequences.

Classification of Stowaway MITEs and sequence relationship with Osmars:
Although numerous Stowaway families were previously identified in rice, analysis of repeats was limited to a small fraction of the genomic sequence (<50 Mb; MAO et al. 2000 Down; TARCHINI et al. 2000 Down; JIANG and WESSLER 2001 Down; TURCOTTE et al. 2001 Down). For this reason, a more comprehensive search for Stowaway MITEs was undertaken in ~360 Mb of BAC/PAC sequence from the IRGSP.

Searches were carried out with BLASTN and RepeatMasker using a collection of previously characterized Stowaway elements (JIANG and WESSLER 2001 Down; Repbase Update, http://www.girinst.org) in addition to elements identified de novo by RECON (BAO and EDDY 2002 Down; N. JIANG, Z. BAO, S. EDDY and S. R. WESSLER, unpublished data). Depending on the stringency of these searches (see MATERIALS AND METHODS), a total of 22,000–33,000 Stowaway elements are estimated to populate the Nipponbare genome. Sequence comparisons led to the grouping of most of these elements into 36 high-copy-number families (Fig 3). As with Osmar families, members of the same Stowaway family share at least 85% similarity over their entire length. Stowaway families are represented by consensus sequences that range in size from 96 to 312 bp with TIRs of 21 to 94 bp (Fig 3; consensus sequences were deposited in Repbase Update, http://www.girinst.org).

With the Osmar and Stowaway elements organized into families, it was of interest to determine whether any correspondence existed that would indicate a clear-cut relationship between autonomous (Osmar) and nonautonomous (Stowaway) elements. Two complementary approaches were used to compare the sequences in the terminal regions of Osmars with Stowaway families. First, each rice MLE was used as a query in BLASTN searches against the three rice genomic databases (IRGSP, BGI, and Syngenta). These searches revealed that Osmars were associated with few, if any, deletion derivatives (see supplemental table at http://www.genetics.org/supplemental and Fig 4). Furthermore, these deleted copies were heterogeneous in size (Fig 4) and were usually larger than Stowaway elements (280 bp–~2 kb vs. 94–350 bp for Stowaway consensus). Although a few MLE families, such as Osmar10, include a small homogeneous group of short deletion derivatives (Fig 4), none of the derivatives have attained the high copy number that is a hallmark of MITE families.



View larger version (25K):
In this window
In a new window
Download PPT slide
  		Figure 4. Structure of some Osmar families. Shown is the structure of Osmar families representative of the three major clades of rice MLEs (clade A, Osmar8; clade B, Osmar5; and clade C, Osmar10). Each family includes members of variable size that either contain or do not contain transposase sequences (solid boxes). Average percentages of pairwise similarity within and between the two subsets of family members are shown. Hatched and dotted fragments in Osm5m1 and Osm5m2 represent portions that are unrelated to each other and to other Osmar5 sequences. Otherwise, smaller members resemble internal deletion derivatives of the larger members. Asterisks indicate the positions of premature stop codons caused by nucleotide substitution or small indels in the transposase sequence.

In a second approach, each Stowaway consensus was used as a query in BLASTN searches against a database containing all full-length rice MLEs. These searches revealed that when significant sequence similarity existed, it was restricted to the terminal nucleotides (usually <50 bp; see example of Osmar1 and Stow-Os6, Fig 5). The most extensive matches were found between Osmar4 and Stow-Os10b, Osmar13 and Stow-Os16, and Osmar11 and Stow-Os49 (Fig 5).



View larger version (80K):
In this window
In a new window
Download PPT slide
  		Figure 5. Selected examples of pairwise sequence comparisons between the terminal sequences of Osmar and Stowaway elements. Examples of Osmar-Stowaway matches were selected to illustrate the range and extent of sequence similarity that can be found between the two groups. Significant similarities are usually restricted to the TIRs (Osmar1 vs. Stow-Os6 and Osmar4 vs. Stow-Os10b) but, in a few cases, can be extended to the subterminal regions (Osmar13 vs. Stow-Os16 and Osmar11 vs. Stow-Os49). TIRs are shown as arrows above Osmar sequences and below Stowaway.

Most of the Stowaway families can be assigned to one of the four major Osmar clades on the basis of similarities in their TIRs. In fact, 34 out of 36 Stowaway consensus sequences display one of the four TIR motifs diagnostic of Osmar clades (compare Fig 2 and Fig 3). For example, group B of Stowaway and Osmar are characterized by the same CCCA motif in the TIRs. Guided by these groupings, we generated consensus TIR sequences for each major clade of Osmar and Stowaway in the form of pictograms (Fig 6). Comparison of the pictograms further revealed the similarities in the TIRs of corresponding clades of Osmar and Stowaway.



View larger version (54K):
In this window
In a new window
Download PPT slide
  		Figure 6. Comparison of Osmar and Stowaway TIRs. Osmar and Stowaway can be classified into corresponding clades (A1, A2, B, and C) on the basis of the presence of a 4-bp diagnostic motif in their TIR sequences (see Fig 2 and Fig 3). To further illustrate the TIR similarity in corresponding clades of Osmar and Stowaway, a pictogram was constructed using the first 25 nucleotides and the reverse complement of the last 25 nucleotides of all clade members (see Fig 2 and Fig 3). In this representation, the size of the letter is proportional to its frequency at a given position. A thick black line underscores the clade-specific 4-bp motif used to classify Osmar and Stowaway families. Pictograms were generated at http://genes.mit.edu/pictogram.html, using default parameters.


*  DISCUSSION
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
*LITERATURE CITED

Several lines of evidence point to plant MLEs as the autonomous partners of Stowaway MITEs. In this study, we have taken the next step in testing this hypothesis by generating an inventory of virtually all MLEs and Stowaway families coexisting in a single genome and analyzing in detail their sequence relationships.

The first whole-genome picture of plant MLEs:
A total of 39 MLE transposases and 22 potentially full-length MLEs were identified from the genomes of the two rice subspecies. On the basis of the phylogenetic analysis of transposases, the intron/exon structure of the transposase gene, and a comparison of terminal sequences, rice MLEs could be divided into 25 families that group into three major clades (Fig 1 and Fig 2). These clades correspond to the three lineages of MLE transposases that were recently isolated by PCR using plant-specific MLE primers and genomic DNA from a wide range of plant species (FESCHOTTE and WESSLER 2002 Down; results not shown). A conclusion of this prior study was that three major lineages of MLE transposases had diversified prior to the divergence of the Poaceae family (~70 MYA) and have been maintained in the genomes of most extant grass species (FESCHOTTE and WESSLER 2002 Down). The fact that the three MLE clades identified in rice correspond to these three major lineages indicates that no other, more divergent, lineages are in rice. Thus, the rice genome is representative of the diversity of MLE transposases in the grasses and as such should serve as a suitable model for the evolutionary analysis of plant MLEs.

Despite the ancient origin of the three MLE lineages, all appear to include families recently active in rice. This is reflected by the high level of sequence similarity among members of several Osmar families and the presence of copies with intact coding capacity (see Fig 1 and examples in Fig 4). On the basis of these criteria, one of the most recently active MLE family is Osmar5: the three full-length members are >99.5% identical to each other and harbor intact transposase ORFs. It is therefore possible that one or more active MLEs may still reside in the rice genome.

Full-length Osmars are heterogeneous in size, ranging from 3.2 to 7.1 kb, and there is also extensive size variation within Osmar families (see Fig 1, Fig 2, and Fig 4). Full-length Osmars harbor a single gene corresponding to the putative transposase, which generally occupies a central position in the element (but see Fig 2 for the few exceptions) and has a similar size among rice MLEs. Thus, most of the size variation among Osmars is due to the variable length of the subterminal regions. These regions do not display any obvious structural features, such as direct or inverted motifs, like those of some other plant DNA transposons, including hAT or CACTA superfamily members (KUNZE and WEIL 2002 Down).

In contrast to rice MLEs, there is a remarkable conservation in the size of full-length MLEs described from a wide range of metazoan species. The dozens of elements described from species as diverse as planarians, hydra, nematodes, insects, or humans vary in size from only 1.2 to 1.4 kb, despite extreme variation in sequence (ROBERTSON et al. 1998 Down). Furthermore, MLE families seem to be mainly represented by full-length copies in these species (ROBERTSON et al. 1998 Down). One notable exception is Hsmar1 in humans, which is present in 200 full-length copies (1.3 kb), but is responsible for the spread of >2000 80-bp MITEs (MORGAN 1995 Down). In rice, there is an overwhelming copy number excess of Stowaway MITEs over Osmars (22,000–33,000 vs. <40). Together, these differences may reflect differences in the cis- and trans-requirements of animal and plant MLE transposases and/or their evolutionary dynamic.

A comprehensive inventory of Stowaway MITEs in rice:
A comprehensive collection of Stowaway families was obtained by combining data gathered from previous studies with those generated de novo by the program RECON for ~30 Mb of rice sequences (see MATERIALS AND METHODS). Searches of ~360 Mb of Nipponbare BAC/PAC sequences with this collection indicate that this genome contains from 22,000 to 33,000 Stowaway elements that group into 36 families (Fig 3). These values are in the range of those reported in previous studies (JIANG and WESSLER 2001 Down; TURCOTTE et al. 2001 Down; GOFF et al. 2002 Down; YU et al. 2002 Down) and confirm that Stowaway is one of the most abundant classes of interspersed repeats in rice, contributing up to ~2% of the total genomic DNA.

Like most previously described MITE families, Stowaway families are characterized by relatively high numbers of copies (for class 2 transposons) and a remarkable conservation in size (standard deviation from consensus size is typically <2% per family; data not shown). There are, however, large variations in copy number among families, ranging from several dozen to a few thousand (Fig 3). Interestingly, the most expansive families are also those with the longest TIRs (e.g., Stow-Os1, Stow-Os23, and Os-Stow46; see Fig 3). It is tempting to speculate that the palindromic structure of these Stowaways may have contributed to their success.

The complex relationship of Osmar and Stowaway elements:
Multiple alignments and phylogenetic analyses of hundreds of family members show that most Stowaway families are made of multiple subfamilies of variable age (not shown). This phylogenetic structure indicates that there have been multiple waves of amplification of a limited number of Stowaway progenitors. Having identified and characterized Stowaway and MLE families, we are now in a position to address two key questions: What are the enzymatic sources responsible for the bursts of Stowaway transposition? How do Stowaway progenitors originate?

Osmars as the transposase sources for Stowaway MITEs: Comparison of Osmar and Stowaway sequences shows that similarity is primarily restricted to the first 20–30 bp of the elements (Fig 5 and Fig 6). In some pairwise comparisons, significant similarities could be extended to the subterminal regions, associating a given Stowaway family with an Osmar family (see Fig 5). However, the level of similarity in these comparisons (<85%) is below the value of a typical intrafamily relationship. Nevertheless, these are the closest matches that can be established in the rice genome between a high-copy-number Stowaway family and an element encoding a transposase. Therefore, Osmar elements are the best candidates as the autonomous partners of Stowaway families.

Our comparative analysis of the TIRs of Osmar and Stowaway provides further evidence for a functional relationship between these two groups of transposons. We showed that Osmar and Stowaway families could be placed into corresponding groups on the basis of characteristic motifs in their TIRs (Fig 2, Fig 3, and Fig 6). In turn, each of these motifs was found to be diagnostic of a distinct group of MLE transposase. Thus, similarities of Osmar and Stowaway in TIRs were used to connect almost every Stowaway family with one of four distinct clades of Osmar transposase (see Fig 6). Coevolution of TIR and transposase sequences is expected because transposase molecules recognize and bind specifically to the TIRs during the transposition reaction of most class 2 transposons, including Tc1/mariner elements (LAMPE et al. 2001 Down; ZHANG et al. 2001 Down; PLASTERK and VAN LUENEN 2002 Down). Hence, changes in the transposase sequences are likely to be accompanied by changes in the TIR sequences and vice versa (LAMPE et al. 2001 Down; NAUMANN and REZNIKOFF 2002 Down). The conservation of TIR sequences for different element families should thus reflect the use of the same or a very similar source of transposase. For these reasons, we believe that the correspondence of Osmar and Stowaway TIRs is functionally significant and supports the notion that different Stowaway families have amplified by using distinct MLE transposases (FESCHOTTE et al. 2002A Down).

Origin of Stowaway MITEs: Although evidence for a functional relationship between Stowaway MITEs and Osmar transposases is accumulating, there were very few cases of clear-cut sequence relationship between Stowaway and Osmar elements (i.e., where the MITE resembles an internal deletion derivative of the larger element; see examples in Fig 4). In fact, such cases are restricted to Stowaway elements that have not amplified to high copy numbers and represent a negligible fraction of the 22,000–33,000 Stowaways present in rice. Thus, the origin of high-copy-number Stowaway families remains enigmatic.

One possible explanation for this situation is the differential retention of Stowaway and Osmar elements in the rice genome over evolutionary time. Assuming that the loss of transposons is primarily a stochastic process (HARTL et al. 1997 Down), MITEs may simply have a greater chance to persist because they outnumber their autonomous partners. MITEs may also have a selective advantage over Osmars, because their insertions are less likely than those of larger elements to be deleterious. While differential retention may explain some Osmar-Stowaway situations, it cannot explain all of them. Among the dozens of Osmar and Stowaway families described in this study, one would have expected to find at least a few cases of direct association. After all, clear-cut relationships between large MITE families and full-length Tc1/mariner transposons were previously found in the Arabidopsis and human genomes (e.g., MORGAN 1995 Down; SMIT and RIGGS 1996 Down; FESCHOTTE and MOUCHES 2000 Down; reviewed in FESCHOTTE et al. 2002B Down).

Instead of differential retention, we propose two, not mutually exclusive, alternative hypotheses. First, some Stowaway families may not be derived from Osmar, but may originate de novo following the fortuitous association and recognition of TIRs flanking unrelated segments of DNA. The creation of a new DNA transposon by capture of flanking sequence has been reported for the P element in Drosophila (TSUBOTA and HUONG 1991 Down) and a similar scenario was proposed for the origin of Ds1 elements in maize (MACRAE and CLEGG 1992 Down). Support for this hypothesis comes from the fact that 13 high-copy-number Stowaway families are characterized by elements with long TIRs (48–94 bp; Fig 3), whereas there are no Osmars with TIRs longer than 36 bp (Fig 2). Long arrays of palindromic DNA are frequently encountered in eukaryotic genomes (e.g., CAVALIER-SMITH 1974 Down; DEININGER and SCHMID 1976 Down) and may provide the raw material for the de novo origin of some MITE families.

De novo origins are unlikely for other Stowaway families that have extended regions of similarity with coexisting MLEs (Fig 5). These Stowaway families may have originated by internal deletion of Osmars, but amplification to higher copy numbers could be a secondary event mediated by a transposase encoded by a distantly related element (see Fig 7). In this model, the origin and amplification of MITEs are considered as two different steps that may be separated by a long period of time. The more time elapsed between these two steps, the more difficult it will be to recognize the filiation between a MITE family and an autonomous element.



View larger version (20K):
In this window
In a new window
Download PPT slide
  		Figure 7. Model for the amplification of Stowaway MITEs. The first three steps of this model are based on the life cycle of MLEs proposed by HARTL et al. 1997 Down. An autonomous element, newly introduced in the genome of a species by either vertical inheritance or horizontal transmission, first transposes at relatively high frequency (step 1). This may lead to a rapid increase in copy number if the double-strand gap left after excision is repaired using a locus containing the same transposon (for example, the homologous chromosome). Many newly synthesized transposons are internally deleted versions of the autonomous copy because of frequent interruption and/or slippage during gap repair (step 2). Copy numbers may increase until host defense mechanisms (homology-dependent silencing) and auto-regulatory processes (transposase titration, overproduction inhibition, etc.) act to repress transposition and stabilize copy numbers (HARTL et al. 1997 Down; OKAMOTO and HIROCHIKA 2001 Down; HANNON 2002 Down; PLASTERK and VAN LUENEN 2002 Down). Over time, both active and inactive copies are progressively degraded by point mutations (vertical inactivation) and are stochastically lost or fossilized in the genome (step 3). By chance, some of the decayed elements might preserve (or evolve de novo) sequences recognized by the transposase of a newly introduced autonomous element (step 4). The new autonomous element might be introduced from another species by horizontal transfer or genetic introgression, but it may also emerge "vertically" by diversifying evolution of a previously inactivated full-length element (as discussed by LAMPE et al. 2001 Down). The newly expressed transposase will thus be able to mobilize its own family members and distantly related MITEs (step 4). We propose that sequence divergence between the MITE and its autonomous partner may favor the propagation of MITEs and allow their amplification to high copy numbers (step 5; see text for details).

MITE amplification via cross-mobilization: Regardless of the origin of MITEs (de novo or ancient deletion derivatives), our results suggest that cross-mobilization is one of the major mechanisms operating in the rice genome to amplify MITEs to high copy numbers. There are previous examples of cross-mobilization of short DNA transposons by distantly related autonomous elements. In maize, Ds1 elements (~400 bp) have only the 5' terminal 13 bp and the 3' terminal 26 bp in common with Ac elements, but they can be mobilized by the Ac transposase (e.g., SHEN et al. 1998 Down). In Caenorhabditis elegans, the nonautonomous Tc7 elements are mobilized in vivo and in vitro by the Tc1 transposase, even though Tc7 and Tc1 share only their 36 terminal nucleotides (REZSOHAZY et al. 1997 Down). Interestingly, Tc7, like many other MITE-like families in this species, has no parental autonomous copies recognizable in the C. elegans genome. For example, the MITE families CeleTc2, Cele11, and Cele12 all have Tc2-like TIRs while their internal sequences have little similarity to each other or to other Tc2 sequences (OOSUMI et al. 1996 Down). To explain this and related instances, the authors hypothesized that a single Tc family can cross-mobilize a variety of highly divergent sequences (OOSUMI et al. 1996 Down). The cross-mobilization model also gains support from the recent discovery that another rice MITE family, mPing, is comobilized in cell culture with a closely, but not directly related, autonomous Pong element (JIANG et al. 2003 Down). Interestingly, mPing elements are Tourist-like MITEs, the other principal MITE group in plants. Together with our study of the Osmar-Stowaway relationships in rice, these data converge toward a model where cross-mobilization plays a major role in the amplification of MITEs.

How could cross-mobilization contribute to MITE amplification? Recent studies have shown that the activity of many transposable element families is repressed by epigenetic mechanisms that act at the transcriptional or post-transcriptional level to repress the expression of the transposon gene product (OKAMOTO and HIROCHIKA 2001 Down; FESCHOTTE et al. 2002A Down; HANNON 2002 Down; PLASTERK and VAN LUENEN 2002 Down). All of these mechanisms are based on recognition of nucleic acid sequence homology and triggered by multiple copies of the target sequence. As a result, there is usually an inverse correlation between the copy number of a transposable element family, the expression levels of their gene products, and/or the transpositional activity of the family (e.g., CHANDLER and WALBOT 1986 Down; CHABOISSIER et al. 1998 Down; HIROCHIKA et al. 2000 Down). We speculate that a MITE lacking extensive sequence homology with an active autonomous element, but retaining the short cis-sequences (TIRs) recognized by the corresponding transposase, may be able to multiply without triggering homology-dependant mechanisms of transposon silencing (see model in Fig 7). This would ensure the maintenance of a high level of transposase expression, which would allow MITE families to quickly spread and attain high copy numbers.


*  FOOTNOTES

2 Present address: Tri-Institutional MD/PhD Program, Weill Medical College, Cornell University, New York, NY 10021. Back


*  ACKNOWLEDGMENTS

We are grateful to N. Jiang and Z. Bao for their help in the mining and analysis of Stowaway families and for sharing unpublished information. We also thank J. Jurka for providing access to data on Stowaway families prior to their publication in Repbase. We thank N. Jiang, M. Osterlund, E. Pritham, and X. Zhang for critical reading of the manuscript and helpful discussions. This work was supported by grants from the National Science Foundation Plant Genome Initiative, the National Institutes of Health, and the University of Georgia Research Foundation to S.R.W.

Manuscript received September 26, 2002; Accepted for publication November 11, 2002.


*  LITERATURE CITED
*TOP
 *ABSTRACT
 *MATERIALS AND METHODS
 *RESULTS
 *DISCUSSION
 *LITERATURE CITED

BAO, Z. and S. R. EDDY, 2002  Automated de novo identification of repeat sequence families in sequenced genomes. Genome Res. 12:1269-1276.[Abstract/Free Full Text]

BENNETZEN, J. L., 2000  Transposable element contributions to plant gene and genome evolution. Plant Mol. Biol. 42:251-269.[Medline]

BUREAU, T. E. and S. R. WESSLER, 1994  Stowaway: a new family of inverted-repeat elements associated with genes of both monocotyledonous and dicotyledonous plants. Plant Cell 6:907-916.[Abstract]

BUREAU, T. E., P. C. RONALD, and S. R. WESSLER, 1996  A computer-based systematic survey reveals the predominance of small inverted-repeat elements in wild-type rice genes. Proc. Natl. Acad. Sci. USA 93:8524-8529.[Abstract/Free Full Text]

CAPY, P., C. BAZIN, D. HIGUET and T. LANGIN, 1998 Dynamics and Evolution of Transposable Elements. Springer-Verlag, Austin, TX.

CAVALIER-SMITH, T., 1974  Long palindromes in eukaryotic DNA. Nature 262:255-256.

CHABOISSIER, M. C., A. BUCHETON, and D. J. FINNEGAN, 1998  Copy number control of a transposable element, the I factor, a LINE-like element in Drosophila. Proc. Natl. Acad. Sci. USA 95:11781-11785.[Abstract/Free Full Text]

CHANDLER, V. L. and V. WALBOT, 1986  DNA modification of a maize transposable element correlates with loss of activity. Proc. Natl. Acad. Sci. USA 83:1767-1771.[Abstract/Free Full Text]

DEININGER, P. L. and C. W. SCHMID, 1976  An electron microscope study of the DNA sequence organization of the human genome. J. Mol. Biol. 106:773-790.[Medline]

FESCHOTTE, C. and C. MOUCHÈS, 2000  Evidence that a family of miniature inverted-repeat transposable elements (MITEs) from the Arabidopsis thaliana genome has arisen from a pogo-like DNA transposon. Mol. Biol. Evol. 17:730-737.[Abstract/Free Full Text]

FESCHOTTE, C. and S. R. WESSLER, 2002  Mariner-like transposases are widespread and diverse in flowering plants. Proc. Natl. Acad. Sci. USA 99:280-285.[Abstract/Free Full Text]

FESCHOTTE, C., N. JIANG, and S. R. WESSLER, 2002a  Plant transposable elements: where genetics meets genomics. Nat. Rev. Genet. 3:329-341.[Medline]

FESCHOTTE, C., X. ZHANG and S. WESSLER, 2002b Miniature inverted-repeat transposable elements (MITEs) and their relationship with established DNA transposons, pp. 1147–1158 in Mobile DNA II, edited by N. L. CRAIG, R. CRAIGIE, M. GELLERT and A. M. LAMBOWITZ. American Society for Microbiology Press, Washington, DC.

GOFF, S. A., D. RICKE, T. H. LAN, G. PRESTING, and R. WANG et al., 2002  A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296:92-100.[Abstract/Free Full Text]

HANNON, G. J., 2002  RNA interference. Nature 418:244-251.[Medline]

HARTL, D. L., A. R. LOHE, and E. R. LOZOVSKAYA, 1997  Modern thoughts on an ancyent marinere: function, evolution, regulation. Annu. Rev. Genet. 31:337-358.[Medline]

HIROCHIKA, H., H. OKAMOTO, and T. KAKUTANI, 2000  Silencing of retrotransposons in Arabidopsis and reactivation by the ddm1 mutation. Plant Cell 12:357-369.[Abstract/Free Full Text]

JARVIK, T. and K. G. LARK, 1998  Characterization of Soymar1, a mariner element in soybean. Genetics 149:1569-1574.[Abstract/Free Full Text]

JIANG, N. and S. R. WESSLER, 2001  Insertion preference of maize and rice miniature inverted repeat transposable elements as revealed by the analysis of nested elements. Plant Cell 13:2553-2564.[Abstract/Free Full Text]

JIANG, N., Z. BAO, X. ZHANG, H. HIROCHIKA, and S. R. EDDY et al., 2003  An active DNA transposon family in rice. Nature 421:163-167.[Medline]

KUNZE, R., and C. F. WEIL, 2002 The hAT and CACTA superfamilies of plant transposons, pp. 565–610 in Mobile DNA II, edited by N. L. CRAIG, R. CRAIGIE, M. GELLERT and A. M. LAMBOWITZ. American Society for Microbiology Press, Washington, DC.

LAMPE, D. J., K. K. WALDEN, and H. M. ROBERTSON, 2001  Loss of transposase-DNA interaction may underlie the divergence of mariner family transposable elements and the ability of more than one mariner to occupy the same genome. Mol. Biol. Evol. 18:954-961.[Abstract/Free Full Text]

LE, Q. H., S. WRIGHT, Z. YU, and T. BUREAU, 2000  Transposon diversity in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 97:7376-7381.[Abstract/Free Full Text]

MACRAE, A. F. and M. T. CLEGG, 1992  Evolution of Ac and Ds1 elements in select grasses (Poaceae). Genetica 86:55-66.[Medline]

MAO, L., T. C. WOOD, Y. YU, M. A. BUDIMAN, and J. TOMKINS et al., 2000  Rice transposable elements: a survey of 73,000 sequence-tagged-connectors. Genome Res. 10:982-990.[Abstract/Free Full Text]

MORGAN, G. T., 1995  Identification in the human genome of mobile elements spread by DNA-mediated transposition. J. Mol. Biol. 254:1-5.[Medline]

NAUMANN, T. A. and W. S. REZNIKOFF, 2002  Tn5 transposase with an altered specificity for transposon ends. J. Bacteriol. 184:233-240.[Abstract/Free Full Text]

OKAMOTO, H. and H. HIROCHIKA, 2001  Silencing of transposable elements in plants. Trends Plant. Sci. 6:527-534.[Medline]

OOSUMI, T., B. GARLICK, and W. R. BELKNAP, 1996  Identification of putative nonautonomous transposable elements associated with several transposon families in Caenorhabditis elegans.. J. Mol. Evol. 43:11-18.[Medline]

PLASTERK, R. H. A., and H. G. VAN LUENEN, 2002 The Tc1/mariner family of transposable elements, pp. 519–532 in Mobile DNA II, edited by N. L. CRAIG, R. CRAIGIE, M. GELLERT and A. M. LAMBOWITZ. American Society for Microbiology Press, Washington, DC.

PLASTERK, R. H. A., Z. IZSVÁK, and Z. IVICS, 1999  Resident aliens: the Tc1/mariner superfamily of transposable elements. Trends Genet. 15:326-332.[Medline]

REZSOHAZY, R., H. G. A. M. VAN LUENEN, R. M. DURBIN, and R. H. A. PLASTERK, 1997  Tc7, a Tc1-hitch hiking transposon in Caenorhabditis elegans.. Nucleic Acids Res. 25:4048-4054.[Abstract/Free Full Text]

ROBERTSON, H. M., F. N. SOTO-ADAMES, K. O. WALDEN, R. M. AVANCINI and D. J. LAMPE, 1998 The mariner transposons of animals: horizontally jumping genes, pp. 268–284 in Horizontal Gene Transfer, edited by M. SYVANEN and C. I. KIDO. Chapman & Hall, London.

SHAO, H. and Z. TU, 2001  Expanding the diversity of the IS630-Tc1-mariner superfamily: discovery of a unique DD37E transposon and reclassification of the DD37D and DD39D transposons. Genetics 159:1103-1115.[Abstract/Free Full Text]

SHEN, W. H., C. RAMOS, and B. HOHN, 1998  Excision of Ds1 from the genome of maize streak virus in response to different transposase-encoding genes. Plant Mol. Biol. 36:387-392.[Medline]

SMIT, A. F. A. and A. D. RIGGS, 1996  Tiggers and DNA transposon fossils in the human genome. Proc. Natl. Acad. Sci. USA 93:1443-1448.[Abstract/Free Full Text]

TARCHINI, R., P. BIDDLE, R. WINELAND, S. TINGEY, and A. RAFALSKI, 2000  The complete sequence of 340 kb of DNA around the rice adh1-adh2 region reveals interrupted colinearity with maize chromosome 4. Plant Cell 12:381-391.[Abstract/Free Full Text]

TSUBOTA, S. I. and D. V. HUONG, 1991  Capture of flanking DNA by a P element in Drosophila melanogaster: creation of a transposable element. Proc. Natl. Acad. Sci. USA 88:693-697.[Abstract/Free Full Text]

TURCOTTE, K., S. SRINIVASAN, and T. BUREAU, 2001  Survey of transposable elements from rice genomic sequences. Plant J. 25:169-179.[Medline]

WESSLER, S. R., T. E. BUREAU, and S. E. WHITE, 1995  LTR-retrotransposons and MITEs: important players in the evolution of plant genomes. Curr. Opin. Genet. Dev. 5:814-821.[Medline]

YU, J., S. HU, J. WANG, G. K. WONG, and S. LI et al., 2002  A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296:79-92.[Abstract/Free Full Text]

ZHANG, L., A. DAWSON, and D. J. FINNEGAN, 2001  DNA-binding activity and subunit interaction of the mariner transposase. Nucleic Acids. Res. 29:3566-3575.[Abstract/Free Full Text]




This article has been cited by other articles:


Home page
 Genome ResHome page
H. Kuang, C. Padmanabhan, F. Li, A. Kamei, P. B. Bhaskar, S. Ouyang, J. Jiang, C. R. Buell, and B. Baker
Identification of miniature inverted-repeat transposable elements (MITEs) and biogenesis of their siRNAs in the Solanaceae: New functional implications for MITEs
Genome Res., January 1, 2009; 19(1): 42 - 56.
[Abstract] [Full Text] [PDF]

Home page
 Genetics