Homologous sequences to the P transposable element in Drosophila subobscura

Summary - The distribution of P homologous sequences among different strains of D subobscura is described. Southern analysis of genomic DNAs from the strains restricted with different enzymes showed rather homogeneous hybridization patterns. P homologous sequences are ubiquitous in D subobscura although they are present in a very low copy number. They are strongly divergent from that of D melanogaster and are probably not functional. These results are consistent with the idea that P elements are very old components of the D subobscura genome. D subobscura / molecular evolution / P transposable element


INTRODUCTION
Transposable elements are widely distributed among the genomes of living organisms. The role of mobile elements as a cause of spontaneous mutagenesis is well established. In spite of their evolutionary significance as a source of genetic variability, the screening and identification of the different element families from the different species of eukaryotes is just starting. In this respect, D melnnogaster is one of the species whose transposable elements are best understood. Up to now, approximately 30 different families of transposable elements in this species have been described (Finnegan and Fawcett, 1986). Among them are the hybrid dysgenesis inducing families, P, I and hobo, which are especially interesting. The phenomenon of hybrid dysgenesis occurs when males bearing complete copies of these elements (P, I or hobo) are crossed with females lacking these elements (M, reactive and empty strains respectively). P, I and hobo families share a common feature. They show a temporally and spatially discontinuous distribution among D melanogaster populations (Anxolab6h6re et al, 1985b(Anxolab6h6re et al, , 1988Bucheton et al, 1986;Boussy et al, 1988;P6riquet et al, 1989). An even more patchy distribution among Drosophila species has been described (Stacey et al, 1986). This pattern of distribution of these sequences strongly supports the rapid-invasion hypothesis proposed by Kidwell (1979Kidwell ( , 1983, based on the analysis of dysgenic potential of P strains. The P element family is composed of a group of sequences heterogeneous in size: a complete 2.9 kb functional element (O'Hare and Rubin, 1983) and various smaller defective nonfunctional elements (Spradling and Rubin, 1982;O'Hare and Rubin, 1983;Voelker et al, 1987). There have been analyses of the distribution of these sequences among Drosophila species other than D melanogaster. Although in early studies no evidence for complete or deleted sequences in other species of the melanogaster group was reported (Brookfield et al, 1984;Lansman et al, 1985), recent studies demonstrate the presence of these sequences in some of the species of the mor,tiu!n subgroup (Daniels et al, 1990). Among the other 3 Drosophila species groups: willistoni, saltans and obscura, P homologous sequences are widely distributed. Extensive Southern analyses of the distribution of P sequences among the wildistoni and saltans groups (Daniels and Strausbaugh, 1986) show that complete element sequences are present in most species of these groups. Recent reports on sequences from D nebulosa (Lansman et al, 1987) and D williston,i (Daniel et al, 1990) indicate that a complete highly homologous element is carried by both species. While the P sequence from D willistoni is completely functional, no transposition ability is shown by D nebulosa P sequences. Because of this distribution, the widlistoni species group has been proposed as the source of the sequences invading the D medanogaster genome (Daniels and Strausbaugh, 1986;Daniels et al, 1990).
Bearing in mind the phylogenetic relationships among the 4 groups of Drosophila species, it seems especially interesting to analyse the distribution of sequences amongst the obscura group. As in the willistoni and saltan groups, P homologous sequences were found to be widely distributed among obscura group species as detected by blotting analysis but only in low stringency conditions (Anxolab6h6re et al, 1985a;Stacey et al, 1986;de Frutos et al, in preparation). While recent horizontal transfer seems to have occurred from some widlistoni group species to the D melanogaster genome, P homologous sequences among the obscura species group seem to be old components of the genome. The studies reported here were designed to analyse the distribution of P homologous sequences among natural populations of a species of the obscura group: D subobscura. This species, which has been extensively analysed from cytogenetic and evolutionary points of view (Krimbas and Loukas, 1980), was originally Palearctic but has recently invaded North and South America (Brncic et al, 1981;Prevosti et al, 1988). Thus, it would be especially interesting to analyse the transposable elements in an expanding species, which may be subject to rapid changes in its genome. Harwich is a strong P strain that contains many P element copies, and it was used as a positive control. Gruta and Canton-S are true M strains devoid of P elements, and they were used as negative controls.

Probes
The !!r25.7 BWC plasmid was used as a probe. This plasmid was constructed by K O'Hare and is a derivative of p 7 r25.7 (Spradling and Rubin, 1982). This clone contains a nearly complete P element from D rnelar!ogaster which lacks 39 pb from its 5'end and 23 pb from its 3'end.

Experimental procedures
Plasmid DNA was extracted following standard methods (Maniatis et al, 1982). Drosophila genomic DNAs were extracted following the method described by Junakovic et al (1984). Nick translated probes labeled with 32 P were hybridized to the genomic DNAs from D subobscura by squash blotting (an especial dot blot designed by Tchen et al, 1985) and Southern blotting (Southern, 1975). Genomic DNAs were restricted with enzymes which cut twice inside the P element: AccI, PvuII and PstI. Complete P elements of D melanogaster digested with these enzymes yield 2.4 kb, 0.9 kb and 0.7 kb fragments respectively (fig 1). In addition, genomic DNAs were digested with AvaII which generates specific discrete 0.63 kb and 0.48 kb fragments from deleted KP elements (Black et al, 1987).

RESULTS
Filters with DNA from flies from the 38 strains of D subobscura used were probed with the P element of D melanogaster. Weak positive signals were detected in all strains analysed. No differences in the intensity of the signals were detected depending on geographic origin of the samples. As an example we show in figure 2 an autoradiogram from North-European samples.
More information about the molecular structure of homologous P sequences of D subobscura can be obtained from Southern analysis. Genomic DNAs from different D subobscura strains restricted with AccI, PvulI, PstI and AvaII were probed with complete p!r25.7BWCplasmid. AccI, Pvv,II and PstI produce 2.4 kb, 0.9 kb and 0.7 kb specific internal fragments from the complete P element of D melanogaster. The 2.4 kb fragment produced by AccI digestion encompasses most of the P element sequence which has been used to determine whether or not complete P elements are present in the D subobscura genome. Figure 3 shows an autoradiogram of some samples of D subobscura genomic DNAs restricted with AccI. Some features of the P homologous sequences in D subobscura can be seen in this picture. First, sequences homologous to P element of D medanogaster are present in the genome of D subobscura. Positive hybridization signals were found in all the 38 analysed strains (data not shown). Secondly, the strongly hybridizing 2.4 kb fragment, indicative of the presence of a complete P element (lane 1 fig 3), is missing in D subobscura.
A very weak band of similar size is detected in lanes 3, 4, 7 and 8, but these signals can be due to the low stringency washes. We have no information about the size of P sequences in D subobscura. It is possible that no complete P elements are present in this species. Alternatively, it is possible that an element of approximately 2.9 kb is present in D subobscura but 1 or both target sites for AccI are missing in this sequence. Instead of the 2.4 kb fragment, 2 bands of 4 kb and 2 kb tend to be present in all the analysed strains. However, some variability in the banding pattern among strains was detected. This variability is not correlated with the geographical distribution of the strains. Thus, identical patterns are shown by BA and MA strains (lanes 3 and 6, fig 3) which come from Argentina and from the Madeira Islands respectively, LC and PC strains (lanes 8 and 4, fig 3) which come from Spain and British Columbia respectively and so on.
Similar results were obtained from samples digested with Pvtdl and PstI (figs 4 and 5). The presence of a complete P element of D melanogaster produces a 0.9 kb PvuII fragment. As can be observed in figure 4, the 0.9 kb fragment (lane 10) is not present in the D subobscura genome. Genomic DNAs of D subobscura digested with PvulI and probed with the P element exhibit bands of 3.1 kb and 2.7 kb (lanes 1 to 8, fig 4, and lanes 1, 3 and 5, fig 5). A highly homogeneous banding pattern is generated by this enzyme in all the strains analysed. Similarly, the 0.7 kb PstI fragment does not appear in the D subobscura genome. Instead, a strong 1.8 kb band and some other weak signals were found in all the strains analysed (examples of this pattern are shown in lanes 2, 4 and 6, fig 5). From these patterns we can deduce that variation exists between P sequences of D melanogaster andD subobscura. PvulI, PstI and probably Accl target sites were missing. It can also be concluded that only one or very few P elements are present in D subobscura genome. Figure 6 shows the results of digesting genomic DNAs from some samples with AvaII. The digestion of the D melanogaster DNA with this enzyme produces a 1.8 kb fragment and 2 smaller 0.54 kb and 0.48 kb fragments from complete P elements as can be seen in lane 9. AvaII also produces 2 specific 0.64 kb and 0.48 kb small fragments from the deleted KP element. A homogeneous hybridization pattern is shown by D subobscura strains, but fragments of a similar size of those of the D medanogaster complete P element and not of the KP element were found in this species (lanes 1 to 7, fig 6). DISCUSSION Sequences homologous to D melanogaster P elements were found in D subobscura. The distribution pattern of P sequences amongst the strains analysed shows strong differences from that of D melanogaster. All the strains bear P sequences, independently of their geographic origin. This differs from the temporally and geographi-cally patchy distribution found among populations of D melanogaster. We have no information about temporal changes in the presence or absence of P elements in D subobscura. Most of the strains used in this analysis came from wild individuals captured recently. However, currently, D subobscura is undergoing an interesting new evolutionary process. Until 1978 D subobscura was a Palearctic species. In that year it was detected in Puerto Montt, Chile (Brncic et al, 1981;Prevosti et al, 1988). Since then this species has invaded extensive areas in South and North America. This interesting evolutionary process has been accompanied by a rapid genomic reconstruction, eg new clines of inversions have been found (Prevosti et al, 1988). Similarly, changes in transposable elements could be expected. However, no differences have been found between European and American populations.
Not only are P sequences ubiquitous in D subobscura but there are also strong similarities between the elements. The intensity of the hybridization signals is very similar in all strains and no strong homologies with the D rnelanogaster P sequences were found. Positive hybridization signals appeared exclusively at low stringency washes. This agrees with other results of the distribution of P homologous sequences among the obscura group species reported by Anxolab6h6re et al (1985a) and de Frutos et al (in preparation).
We can draw some conclusions about the structure of P homologous sequences in D subobscura from the Southern analysis. Firstly, samples from the 38 D subobscura strains, coming from different geographic locations, show a rather homogeneous hybridization pattern, which indicates that no active elements are present in the D subobscura populations. Secondly, a very low number of bands are generated by the restriction enzymes used in this study. We can deduce that very few homologous P sequences occur in the genome of D subobscura. Thirdly, depending on the restriction enzyme used we can obtain some additional information about the structure of these sequences. PvulI and PstI produce 0.9 kb and 0.7 kb internal fragments, respectively, from the complete P sequence of D melanogaster. Neither the 0.9 kb nor the 0.7 kb fragment has been found in any strain of D subobscura.
In addition from AccI and AvaII digestions we can deduce that neither complete P nor deleted KP elements exist in the D subobscura genome. Alternatively, it is possible that both complete and deleted elements are kept in the D subobscura genome but target sequences from AccI and AvaII enzymes have been lost in this species. To summarize, no targets for any of the cited enzymes are conserved in the P homologous sequences of D subobscura.
Thus, P homologous sequences are present in the D subobscura genome, but they are probably not functional and they are strongly divergent from those of D melanogaster. A very low number of sequences exist in the genome of this species. Although we are using the P element of D melanogaster as a reference, it is highly probable that this species has recently been invaded by P (Kidwell, 1979(Kidwell, , 1983 and that this sequence comes from D wildistoni or any other species of the wildistoni group (Daniels and Strausbaugh, 1986;Daniels et al, 1990). In this context we can conclude that P sequences from D subobscura are strongly diverged from those of D willistoni.
Simple models based on vertical transmission do not account for the distribution of P sequences among Sophophora subgenus species. In a strictly vertical model, P elements must have been present among Drosophilidae family members before Sophophora radiation. In this model we would have to postulate that after their radiation, the wildistoni-sadtans lineages kept active elements. Meanwhile, the P sequence among the obscura group probably lost their function quickly and diverged strongly. Most of the species of the melanogaster group including D melanogaster would have had to lose these elements. If the Sophophora radiation occurred about 50 My ago (Throckmorton, 1985) we would have to postulate that, during this time P elements disappeared completely in most of the species. Nevertheless, P sequences reappeared recently in D melanogaster.
Obviously, horizontal transference of genetic elements between reproductive isolated species is not the usual method of transposable element dispersion. However, some evidence supports the hypothesis of some kind of transport of P elements from some species of the widlistoni group to the D melanogaster genome (Daniels and Strausbaugh, 1986;Daniels et al, 1990). It is not known whether this is a recurrent phenomenon. Other apparent transposable element homologies could be explained by a mechanism of horizontal transmission (eg the close relationship between the copia element of D rraelanogaster and Tyl of yeast, or gypsy group also from D melanogaster and plant viruses, Doolittle et al, 1989). Assuming that horizontal transfer can occur, the distribution of P sequences in the 4 lineages of Sophophora can be explained in the following way. P elements are widely distributed among the willistoni-saltans lineages, but not to such an extent among the melanogasterobscura lineages. Thus, there were probably P sequences before Sophophora radiation. Since P related sequences have also been detected in other Diptera outside the Drosophilidae family (Anxolab6h!re and P6riquet, 1987), they should have been present from more ancient times. P sequences seem to have evolved differently in the 4 Sophophora lineages. While in willistoni-saltans lineages these P sequences have remained from these times (and moreover they are active in D willistoni), in most of the species of the melanogaster group they have been lost and D melanogaster should have been recently invaded by these sequences. In the obscura group strong changes seem to have occurred. Few elements remain in the D subobscura genome, and some of them seem to be located in the heterochromatin. Preliminary in situ hybridization results, using as probe a D subobscura clone (homologous to the P element of D !nelanogaster), indicate the presence of P related sequences at the 68D locus on E chromosome and at centromeric region (Paricio et al, in preparation).
If P sequences were transposed early from the euchromatic to the heterocromatic region it is possible to expect that they were submitted to an early inactivation, and that they diverged quickly.
Although horizontal transmission could be invoked to explain the distribution of P elements among Sophophora species, our results suggest that no such evolutionary event has recently occurred in D subobscura.