Abundant and species-specific DINE-1 transposable elements in 12 Drosophila genomes

Abstract

Background: Miniature inverted-repeat transposable elements (MITEs) are non-autonomous DNA-mediated transposable elements (TEs) derived from autonomous TEs. Unlike in many plants or animals, MITEs and other types of DNA-mediated TEs were previously thought to be either rare or absent in Drosophila. Most other TE families in Drosophila exist at low or intermediate copy number (around < 100 per genome).

Results: We present evidence here that the dispersed repeat Drosophila interspersed element 1 (DINE-1; also named INE-1 and DNAREP1) is a highly abundant DNA-mediated TE containing inverted repeats found in all 12 sequenced Drosophila genomes. All DINE-1s share a similar sequence structure, but are more homogeneous within species than they are among species. The inferred phylogenetic relationship of the DINE-1 consensus sequence from each species is generally consistent with the known species phylogeny, suggesting vertical transmission as the major mechanism for DINE-1 propagation. Exceptions observed in D. willistoni and D. ananassae could be due to either horizontal transfer or reactivation of ancestral copies. Our analysis of pairwise percentage identity of DINE-1 copies within species suggests that the transpositional activity of DINE-1 is extremely dynamic, with some lineages showing evidence for recent transpositional bursts and other lineages appearing to have silenced their DINE-1s for long periods of time. We also find that all species have many DINE-1 insertions in introns and adjacent to protein-coding genes. Finally, we discuss our results in light of a recent proposal that DINE-1s belong to the Helitron family of TEs.

Conclusion: We find that all 12 Drosophila species with whole-genome sequence contain the high copy element DINE-1. Although all DINE-1s share a similar structure, species-specific variation in the distribution of average pairwise divergence suggests that DINE-1 has gone through multiple independent cycles of activation and suppression. DINE-1 also has had a significant impact on gene structure evolution.

Figure 1

Flow chart of the strategy for identifying DINE-1 sequences in the 12 Drosophila genomes.

Figure 2

The generalized structure of DINE-1 sequences from 12 Drosophila genomes, and alignment of the DINE-1 consensus sequences from 12 species with each feature boxed. The element contains two conserved blocks, A and B. Within block A, the sequence can be further divided into two parts, A1 and A2, separated by a region of variable length containing the tandem repeats (CCGT)_n(CTGT)_n. Between blocks A and B is a region of central repeats, containing species-specific repeats. These central repeat sequences do not share homology among species; the length of the repeat unit can range from approximately 50 bp to approximately 500 bp, and the number of repeats is also variable within species. Locations of the subTIRs are shown as gray arrows; see Table 1 for precise designations of subTIR sequences. The 5' end also contains a second inverted repeat (IR) sequence that is partially complementary to the 5'-end terminal repeat and is shown as a gray arrowhead. An inverted repeat near the 3' end forms a potential stem-loop structure and is indicated by white arrowheads.

Figure 3

Analysis of ten sites that are polymorphic for DINE-1 insertions in natural populations of D. yakuba. For each site, the sequence from a strain containing a DINE-1 insertion is shown at the top, and the sequence from a strain lacking the insertion is shown at the bottom. Only the terminal sequences of DINE-1 and its flanking sequences are shown. The 5' subTIR is shown in bold italics. Insertions 1-3 were previously reported in [23]. The interpretation of these data depends on the designation of the DINE-1 termini and whether insertion causes a TSD. (a) Analysis using the annotation of DINE-1 structure presented in this paper. This annotation places the subTIR of D. yakuba 1 bp internal to the 5' end. It also assumes that no TSD is created, in accord with the proposed mechanism of Helitron-type replication [24]. Under these stipulations, all ten insertions occur between the dinucleotide TT in the consensus sequence WTT (where W = A or T), and eight of ten match a longer consensus sequence of insertion after the second nucleotide in the sequence of WTTTT. (b) Analysis assuming the DINE-1 termini of [28], and Helitron-type replication. Only sites 1 and 10 are shown. Under this annotation, DINE-1 would have an insertion preference for WT dinucleotides. (c) Analysis assuming that the DINE-1 5' end begins at its inverted repeat, inserts between the dinucleotide TT and causes a 2 bp TSD, as in MITE-like DNA transposons. The TSDs caused by DINE-1 are boxed.

Figure 4

The frequency distribution of sequence identity of DINE-1 in different species. The percentage identity was based on BLAST search, using consensus sequences of part A1 of block A from each species as the query. To exclude short and fragmented sequences from our analysis, only hits > 100 bp were used. Note that the y-axis scale differs among species.

Figure 5

Phylogenetic relationship of DINE-1 consensus sequences compared to their host species. (a) Phylogenetic tree based on pooled sequences of block A and B (Additional data file 1) and constructed using the neighbor-joining method with the Jukes-Cantor one parameter substitution model [55]. Bootstrap resampling percentages based on 500 replications are indicated. Scale bar represents the estimated number of substitutions. (b) The host species phylogeny is adapted from [29]. Myrs, million years.