Sylvia Hagemann

The evolutionary life history of P transposons: from horizontal invaders to domesticated neogenes

Abstract. P elements, a family of DNA transposons, are known as aggressive intruders into the hitherto uninfected gene pool of Drosophila melanogaster. Invading through horizontal transmission from an external source they managed to spread rapidly through natural populations within a few decades. Owing to their propensity for rapid propagation within genomes as well as within populations, they are considered as the classic example of selfish DNA, causing havoc in a genomic environment permissive for transpositional activity. Tracing the fate of P transposons on an evolutionary scale we describe different stages in their evolutionary life history. Starting from horizontal transfer events, which now appear to be rather a common phenomenon, the initial transpositional burst in the new host is slowed down by the accumulation of defective copies as well as host-directed epigenetic silencing. This leads to the loss of mobility and, finally, to molecular erosion by random mutations. Possible escape routes from genomic extinction are the reactivation within the original host genome by recombination or suspension of the repressing regime, horizontal emigration to a virgin gene pool, or genomic integration and acquisition of a novel function as a domesticated host gene.

Repeated horizontal transfer of P transposons between Scaptomyza pallida and Drosophila bifasciata

Two distinct P element subfamilies, designated M-type and O-type, reside in the genome of D. bifasciata. PCR-screening of 65 Drosophila species revealed that only D. bifasciata and its closest relative D. imaii possess O-type elements. Outside the genus, O-type elements were detected in Scaptomyza pallida. Restriction analyses show that the general structure of the O-type elements from S. pallida and D. bifasciata is the same. Sequence divergence turned out to be extremely low (0.43%). These results suggest that the O-type subfamily of D. bifasciata has been received by horizontal transfer from an external source, most probably from the genus Scaptomyza, as has been previously suspected for the M-type family. Since the sequence divergence between M-type elements from S. pallida and D. bifasciata is eighteenfold higher than that between O-type elements, two independent intergeneric transfer events have to be postulated. In order to re-examine the taxonomic status of S. pallida, a partial sequence (489 bp) of the Adh gene was analysed. The data clearly prove that S. pallida has to be placed far outside the D. obscura group.

Two distinct P element subfamilies in the genome of Drosophila bifasciata

The genome of Drosophila bifasciata harbours two distinct subfamilies of P-homologous sequences, designated M-type and O-type elements based on similarities to P element sequences from other species. Both subfamilies have some general features in common: they are of similar length (M-type: 2935 bp, O-type: 2986 bp), are flanked by direct repeats of 8 by (the presumptive target sequence), contain terminal inverted repeats, and have a coding region consisting of four exons. The splice sites are at homologous positions and the exons have the coding capacity for proteins of 753 amino acids (M-type) and 757 amino acids (O-type). It seems likely that both types of element represent functional transposons. The nucleotide divergence of the two P element subfamilies is high (31%). The main structural difference is observed in the terminal inverted repeats. Whereas the termini of M-type elements consist of 31 by inverted repeats, the inverted repeats of the O-type elements are interrupted by non-complementary stretches of DNA, 12 by at the 5′ end and 14 by at the 3′ end. This peculiarity is shared by all members of the O-type subfamily. Comparison with other P element sequences indicates incongruities between the phylogenies of the species and the P transposons. M-type and O-type elements apparently have no common origin in the D. bifasciata lineage. The M-type sequence seems to be most closely related to the P element from Scaptomyza pallida and thus could be considered as a more recent invader of the D. bifasciata gene pool. The origin of the O-type elements cannot be unequivocally deduced from the present data. The sequence comparison also provides new insights into conserved domains with possible implications for the function of P transposons.

Ancient and Recent Horizontal Invasions of Drosophilids by P Elements

Abstract.P elements of two different subfamilies designated as M- and O-type are thought to have invaded host species in the Drosophila obscura group via horizontal transmission from external sources. Sequence comparisons with P elements isolated from other species suggested that the horizontal invasion by the O-type must have been a rather recent event, whereas the M-type invasion should have occurred in the more distant past. To trace the phylogenetic history of O-type elements, additional taxa were screened for the presence of O- and M-type elements using type-specific PCR primers. The phylogeny deduced from the sequence data of a 927-bp section (14 taxa) indicate that O-type elements have undergone longer periods of regular vertical transmission in the lineages of the saltans and willistoni groups of Drosophila. However, starting from a species of the D. willistoni group they were transmitted horizontally into other lineages. First the lineage of the D. affinis subgroup was infected, and finally, in a more recent wave of horizontal spread, species of three different genera were invaded by O-type elements from the D. affinis lineage: Scaptomyza, Lordiphosa, and the sibling species D. bifasciata/D. imaii of the Drosophila obscura subgroup. The O-type elements isolated from these taxa are almost identical (sequence divergence <1%). In contrast, no such striking similarities are observed among M-type elements. Nevertheless, the sequence phylogeny of M-type elements is also not in accordance with the phylogeny of their host species, suggesting earlier horizontal transfer events. The results imply that P elements cross species barriers more frequently than previously thought but require a particular genomic environment and thus seem to be confined to a rather narrow spectrum of host species. Consequently, different P element types acquired by successive horizontal transmission events often coexist within the same genome.

Repetitive sequences in the genome ofAnemone blanda: Identification of tandem arrays and of dispersed repeats

A tandemly repetitive sequence family (AbS1) and a repetitive sequence (Hd) forming part of a larger dispersed element (dorf-1) ofAnemone blanda were characterised. TheAbS1 satellite sequence family is located in all 4′,6-diamidino-2-phenylindole (DAPI) positive intercalary heterochromatic bands and in the DAPI positive heterochromatic terminal region of chromosome 3, while the dispersedHd homologous sequences are preferentially associated with euchromatic chromosome regions. The major component of theAbS1 satellite isAbS1-H1 with a basic repeat unit of 1640 bp; a minor fraction (AbS1-H5) consists of 320 bp units. A subsection of theAbS1-H1 repeat unit exhibits homologies to the 25S rRNA gene of flowering plants suggesting that the 1.64 kb satellite was generated by amplification of a precursor satellite and/or single copy sequence together with an rDNA fragment. The rDNA homologous region is considered to evolve at a rate similar to pseudogenes and thus the age of this satellite DNA fraction can be roughly estimated as about 27 million years.The dispersed repeated sequenceHd (about 1300 bp) is associated with the 8 kb elementdorf-1. A. blanda dorf-1 constitutes about 0.2% of the genome (3×104 copies), is bounded by identical long terminal repeats, and exhibits partial homology to theLilium gypsy-type elementdell, but has yet to be confirmed as a retrotransposon. In contrast to theAbS1 satellite sequence family,Hd homologous sequences were found not only inA. apennina, the closest relative ofA. blanda, but also inA. nemorosa andA. ranunculoides indicating that a progenitor sequence ofdorf-1 was present in a common ancestor before speciation ocurred.

Different evolutionary behaviour of P element subfamilies: M-type and O-type elements in Drosophila bifasciata and D. imaii

Distribution and variation of two P-element subfamilies designated M-type and O-type elements were investigated in Drosophila bifasciata (Db) and its relatives. PCR screening revealed that full-sized and internally deleted elements of both types occur in three geographic Db strains and in the related species, D. imaii (Di). Molecular analyses indicate differences in the evolutionary behaviour of the two P-element types. Internally deleted M-type elements fall into two size classes present in all three Db strains. In contrast, internally deleted O-type elements vary between the strains in number and length. With respect to genomic location, M-type elements seem to be restricted to conserved euchromatic sites, whereas the positions of O-type elements appear to be geographically variable. In one strain of Db (Italy), O-type elements seem to accumulate in the heterochromatin. Sequencing of a 397-bp segment shows intra- and interspecific divergence of M-type elements. In a 452-bp segment of the O-type elements, no substitutions were found, neither within nor between species. This finding suggests recent introgression of O-type elements via hybridization between Db and Di. Sequence identity and variation in chromosomal locations among different copies imply that O-type elements are transpositionally active. For M-type elements, genomic mobility cannot be proved. In a survey of several other taxa, no O-type-related sequences were detected so far. Therefore, the origin of the O-type subfamily remains unknown, whereas the source of M-type elements can be traced back to the genus Scaptomyza.

P-related sequences inDrosophila bifasciata: A molecular clue to the understanding of P-element evolution in the genusDrosophila

SummaryTwo P-elements (bif1 and bif2) were isolated from a genomic library ofDrosophila bifasciata. Both elements are internally deleted and have lost the coding capacity for a functional transposase. One of the elements (bif2) contains an insert consisting of a repetitive sequence. The terminal inverted repeats and the segments necessary for passive mobility are well conserved. Element bif2 has retained rudiments of the coding sequence of exon 0 and exon 3, but the reading frame is destroyed by insertions and deletions. The comparison of theD. bifasciata P-elements with P-elements ofDrosophila melanogaster andDrosophila nebulosa reveals that the two latter sequences are more similar to each other than either of them is to theD. bifasciata elements. This finding contradicts the phylogenetic relationship of the species and can be taken as an indirect but unequivocal evidence for recent horizontal gene transfer from a relative ofD. nebulosa to the gene pool ofD. melanogaster. The P-elements ofD. bifasciata are phylogenetically ancient and have evolved independently for about 50 million years. A higher substitution rate at the third codon position as well as a predominance of conservative replacements at the amino acid level indicates that the P-elements ofD. bifasciata have been under selective constraint over a long period and that immobilization has occurred only recently.

Transcription and splicing patterns of M- and O-type P elements in Drosophila bifasciata, D. helvetica, and Scaptomyza pallida

Abstract. RT-PCR was applied to analyze the splicing patterns of P-element-derived mRNAs in Drosophila bifasciata, D. helvetica, and Scaptomyza pallida. D. melanogaster was used as a control. The experiments revealed that P elements are transcribed in all species investigated. However, there are differences in the splicing patterns of IVS3, which has to be removed in order to produce transposase mRNA instead of repressor mRNA. These differences are observed among species as well as between the P element subfamilies, the M and the O type, which coexist in the genomes of D. bifasciata and S. pallida. In D. helvetica M-type transposase mRNA was found in the germline and repressor mRNA in the soma, as has been previously described for the canonical (M-type-related) P element of D. melanogaster. In contrast, in S. pallida only repressor mRNA of M-type elements was detected in all tissues. In D. bifasciata, M-type IVS3, although activated both in the soma and the germline, is never completely excised. Instead, two alternative double-spliced variants occur in which two small introns are removed within the IVS3 region. One of these variants codes for a protein 12 aa longer than the regular transposase. Taking these findings together, transposase production and transpositional activity of M-type elements seem to be limited to D. helvetica and D. melanogaster, whereas M-type elements have become immobile in D. bifasciata and S. pallida. Unlike the M type, the splicing of O-type transcripts in D. bifasciata and S. pallida follows the classical rules of tissue-specific P element regulation: transposase mRNA is produced exclusively in the germline whereas repressor mRNA is formed in somatic cells. Thus O-type elements are thought to be still transpositionally active in both species. This finding is in accordance with the postulated recent transfer of O-type elements between the gene pools of D. bifasciata and S. pallida. In addition, we were able to show that the IVS3 double-spliced variants of both P element types are produced regularily in all species of the genus Drosophila investigated so far, but not in S. pallida.