Rolling-circle transposons in eukaryotes

Abstract

All eukaryotic DNA transposons reported so far belong to a single category of elements transposed by the so-called "cut-and-paste" mechanism. Here, we report a previously unknown category of eukaryotic DNA transposons, Helitron, which transpose by rolling-circle replication. Autonomous Helitrons encode a 5'-to-3' DNA helicase and nuclease/ligase similar to those encoded by known rolling-circle replicons. Helitron-like transposons have conservative 5'-TC and CTRR-3' termini and do not have terminal inverted repeats. They contain 16- to 20-bp hairpins separated by 10--12 nucleotides from the 3'-end and transpose precisely between the 5'-A and T-3', with no modifications of the AT target sites. Together with their multiple diverged nonautonomous descendants, Helitrons constitute approximately 2% of both the Arabidopsis thaliana and Caenorhabditis elegans genomes and also colonize the Oriza sativa genome. Sequence conservation suggests that Helitrons continue to be transposed.

Figure 1

Reconstruction of the Helitron1 (A), Helitron2 (B), and Helitron1_CE (C) consensus sequences. The consensus sequences are schematically depicted as rectangles. Contiguous copies of Helitrons that we used for reconstruction of the consensus sequences are shown as bold lines beneath the rectangles. Gaps in the lines mark deletions of corresponding regions of the consensus sequences. GenBank accession nos. and sequence coordinates are indicated. Genes and their coordinates in the consensus sequences are indicated above the rectangles. Genes coding for proteins composed of the Rep and helicase domains are shaded in gray. The AT target sites are encircled.

Figure 2

Multiple alignment of helicases encoded by the Helitron1 and Helitron1_CE transposons with a set of eukaryotic and prokaryotic DNA helicases. Domains I–VI that are conservative in DNA helicases from the SF1 superfamily (20) and distances between these domains are indicated. Domain IV/V has not been reported previously. Invariable positions are shaded in black, and those conserved in more than 60% of the sequences are shaded in gray. The following are names of helicases: PIF1 (GenBank protein identification no. 130196, yeast), BACULOVIRUS (7460536, the dsDNA Lymantia dispar nucleopolyhedrovirus), CHILO (5725645, the dsDNA chilo iridescent virus), TRAA_RHISN (2499024, a Ti-like plasmid from Rhizobium), TRAI_EC (136208, the F plasmid from Escherichia coli), EXOV_EC (2507018, the RecD subunit from the E. coli exodeoxyribonuclease V), TRWC (1084124, the R388 conjugative plasmid from E. coli), and HEL_T4 (416895, the dsDNA T4 bacteriophage).

Figure 3

Termini of Helitrons. Conserved 5′ and 3′ termini are in bold capital letters, 3′ terminal hairpins are shaded in gray, and inverted repeats are underlined.

Figure 4

Precise integration of Helitrons into the host AT target sites. Six insertion cases of Helitrons (red) into different transposons (green) are shown separately. Two copies of ATREP1, two copies of Helitrony3_CE, and single copies of Helitrony2_CE and ATREP9 are inserted into copies of the ATTIRX1D, ATHATN1, PAL5A_CE, LTR2_CE, PALTTAA1_CE, and ATREP10 transposons, respectively. The consensus sequences of the elements harboring Helitrons are marked by the bold black letters and are described in the A. thaliana and C. elegans sections of Repbase Update at www.girinst.org/Repbase_Update.html. Consensus sequences of the corresponding Helitrons are marked in blue. Asterisks, semicolons, and dots indicate identical nucleotide positions, transitions, and transversions, respectively. Only termini of Helitrons are shown. Black, green, and blue numbers show positions in the consensus sequences of the harboring transposons, GenBank, and Helitron consensus sequences, respectively.

Figure 5

Distributions of protein lengths predicted in 100 random DNA sequences. Every random sequence was 61% identical to ATRPA2H, without insertions or deletions. Black marks all proteins predicted by genscan; gray marks proteins similar to ATRPA2Hp.

Figure 6

Alignment of the RC motifs in the Helitrons. Following is a list of the RCR initiator-like proteins: SVTS2 (GenBank accession no. AAF18310, the SVTS2 ssDNA spiroplasma plectrovirus); Rep_SC (BAA34784, the pSA1.1 conjugative plasmid from Streptomyces cyaneus); Rep_BB (BAA07788, the pHT926 Bacillus borstelensis cryptic plasmid); Rep_AA (AAC37125, the pVT736–1 RCR plasmid from Actinobacillus actinomycetemcomitans); and Pf3 (AAA88392, the Pf3 ssDNA bacteriophage from Pseudomonas aeruginosa). Color shading shows different physicochemical properties of conserved amino acids (17). Conserved tyrosines, corresponding to the RCR nicking/ligation catalytic center, are marked by dots.