Genome-wide comparison of Asian and African rice reveals high recent activity of DNA transposons

Abstract

Background: DNA (Class II) transposons are ubiquitous in plant genomes. However, unlike for (Class I) retrotransposons, only little is known about their proliferation mechanisms, activity, and impact on genomes. Asian and African rice (Oryza sativa and O. glaberrima) diverged approximately 600,000 years ago. Their fully sequenced genomes therefore provide an excellent opportunity to study polymorphisms introduced from recent transposon activity.

Results: We manually analyzed 1,821 transposon related polymorphisms among which we identified 487 loci which clearly resulted from DNA transposon insertions and excisions. In total, we estimate about 4,000 (3.5% of all DNA transposons) to be polymorphic between the two species, indicating a high level of transposable element (TE) activity. The vast majority of the recently active elements are non-autonomous. Nevertheless, we identified multiple potentially functional autonomous elements. Furthermore, we quantified the impacts of insertions and excisions on the adjacent sequences. Transposon insertions were found to be generally precise, creating simple target site duplications. In contrast, excisions almost always go along with the deletion of flanking sequences and/or the insertion of foreign 'filler' segments. Some of the excision-triggered deletions ranged from hundreds to thousands of bp flanking the excision site. Furthermore, we found in some superfamilies unexpectedly low numbers of excisions. This suggests that some excisions might cause such large-scale rearrangements so that they cannot be detected anymore.

Conclusions: We conclude that the activity of DNA transposons (particularly the excision process) is a major evolutionary force driving the generation of genetic diversity.

Keywords: DNA transposon activity; Proliferation mechanism; Rice.

Figure 1

The abundance of Mariner (DTT) and Mutator (DTM) families in O. sativa and O. glaberrima. (A) Overview of Mariner (DTT) abundance. Copy numbers of individual families show large differences within the Mariner (DTT) superfamily. For example, in O. sativa, the most successful DTT_SB is represented 4′702 times while we only identified 25 copies of the DTT_SR family. (B) Overview of Mutator (DTM) superfamily. Despite an overall similar distribution, we found one exception for the Mutator (DTM) superfamily DTM_MA, where we found slightly more elements in the O. glaberrima genome (302 copies in O. sativa and 336 copies in O. glaberrima).

Figure 2

Examples of DNA transposon polymorphisms in O. sativa (Osat) and O. glaberrima (Ogla). The alignments show the polymorphic TE plus some of the genomic flanking sequences. Diagnostic sequence motifs are highlighted with colors. (A) Insertion. (B) Perfect excision. (C) Excision with deletion. (D) Excision with deletion and filler sequence.

Figure 3

An example of a transposon excision that caused a large deletion in its flanking region. The transposon DTT_SC is indicated by a gray box. Solid lines represent the genomic sequences of O. sativa and O. glaberrima. The excision is precise at the left border of the DTT_SC element while a 2,479-bp segment was deleted at its right border.

Figure 4

The relative activity and relative abundance of the TE families in O. sativa. We compared the relative activity with the relative abundance of all TE families in O. sativa. Group I consists of families with high activity and high abundance. The CACTA family DTC_Calvin, which is the overall most abundant family, also shows remarkable activity. Group II contains elements with high activity but low copy numbers. We found that Mutator (DTM) and hAT (DTA) families are relatively active despite their poor abundance. Finally, Group III consists of families with high abundance but relatively little activity. This class is dominated by families of the Harbinger (DTH) and Mariner (DTT) superfamilies. The Harbinger family DTH_TO seems to be still relatively active despite its high abundance, whereas the most abundant Mariner and Mutator families DTT_SB and DTM_MAF, respectively, show no activity at all.

Figure 5

The schematic representation of the copies of the Mutator family DTM_MK which were polymorphic in O. sativa and O. glaberrima. The family includes three copies which contain intact transposase ORFs (top three copies). One of these putative mother elements additionally carries a fragment of a second ORF which was also found in other derivatives. Presumed non-autonomous copies have partially deleted or disrupted reading frames containing stop codons or frameshifts in the transposase ORF. Additionally, we found six copies of non-autonomous elements which consist only of TIRs plus an internal sequence that has no homology to that of larger elements (bottom). The fact that all six are very similar to each other indicates that they originate from the same deletion event and are multiplied later.