1. Introduction
Transposable elements (TEs) are a significant component of almost all genomes studied thus far. They are greatly variable, having different transposition mechanisms, and a large sequence diversity. They are a source of genetic diversity for their hosts because TE mobilization is known to promote a repertory of different mutations. For example, disruption of coding sequences or gene regulatory elements can generate new coding sequences or even establish new regulatory gene networks (Biémont & Vieira, Reference Biémont and Vieira2006; Feschotte, Reference Feschotte2008). They are also associated with chromosome rearrangements and genome restructuring, thus, TEs are known to play an important role in genome evolution (Pritham, Reference Pritham2009).
The copia retrotransposon belongs to the copia superfamily, which is characterized by the order of the open reading frames (ORFs) of the enzymes integrase, reverse transcriptase and RNase H in a polyprotein domain (Wicker et al., Reference Wicker, Sabot, Hua-Van, Bennetzen, Capy, Chalhoub, Flavell, Leroy, Morgante, Panaud, Paux, SanMiguel and Schulman2007). This element was first identified in Drosophila melanogaster and inserted into the white locus, promoting a white-apricot mutation. Copia is 5·4 kb long, has long terminal repeats (LTRs) of 276 bp and a single ORF of 4227 nucleotides, which codify a polyprotein of 1409 amino acids. This polyprotein is similar to the products of the gag and pol genes of retroviruses (Mount & Rubin, Reference Mount and Rubin1985). Sequences showing similarity to the copia retrotransposon were identified by Southern blot analyses of 52 species in the Drosophila genus. Twenty-two of these species belong to the melanogaster group, seven to the willistoni group, seven to the obscura group, six to the saltans group, two to the immigrans group and one to the mesophragmatica and pinicula groups (Martin et al., Reference Martin, Wiernasz and Schedl1983; Stacey et al., Reference Stacey, Lansman, Brock and Grigliatti1986). Complete or partial sequences of copia were described in 13 species (reviewed in Biémont & Cizeron, Reference Biémont and Cizeron1999 and Almeida & Carareto, Reference Almeida and Carareto2006; and FlyBase section on natural transposons, see Supplementary Table 1 available at http://journals.cambridge.org/GRH).
In a phylogenetic analysis of the copia element in the genus Drosophila, Jordan & McDonald (Reference Jordan and McDonald1998) demonstrated that this TE diverged into two different families, and that three subfamilies can be found within the species more related to D. melanogaster, a taxon formed by nine species called the melanogaster subgroup. In a subsequent study, Jordan et al. (Reference Jordan, Matyunina and McDonald1999) showed that D. melanogaster and D. willistoni, which belong to different species groups, share copia LTRs with 99% sequence identity, whereas the sibling species D. melanogaster and Drosophila simulans share only 90% sequence identity. Since the sibling species possess a far more recent common ancestor, estimated at 2–3 million years ago (MYA) (Lachaise & Silvain, Reference Lachaise and Silvain2004) and D. melanogaster and D. willistoni diverged about 62 MYA (Tamura et al., Reference Tamura, Subramanian and Kumar2004), the authors suggested horizontal transposon transfer (HTT) as a possible explanation for the incongruence observed in the sequence similarity between copia elements in these species.
Only two occurrences of HTT between D. willistoni and D. melanogaster have been reported (Loreto et al., Reference Loreto, Carareto and Capy2008), namely that of the P element, which was the first well-documented observation of HTT in eukaryotes (Daniels et al., Reference Daniels, Peterson, Strausbaugh, Kidwell and Chovnick1990; Quesneville & Anxolabéhère, Reference Quesneville and Anxolabéhère1998) and of the copia element. In the first case, it was shown that the P element invaded the D. melanogaster genome and quickly spread worldwide. According to Jordan et al. (Reference Jordan, Matyunina and McDonald1999), the HTT of copia element occurred in the opposite direction, namely, from D. melanogaster to D. willistoni. Another important difference between both HTT events is that the P element is a class II TE, whereas copia is a class I TE (Wicker et al., Reference Wicker, Sabot, Hua-Van, Bennetzen, Capy, Chalhoub, Flavell, Leroy, Morgante, Panaud, Paux, SanMiguel and Schulman2007); it was suggested that the elements of different classes behave differently during HTT (Loreto et al., Reference Loreto, Carareto and Capy2008; Schaack et al., Reference Schaack, Gilbert and Feschotte2010). As pointed out by Schaack et al. (Reference Schaack, Gilbert and Feschotte2010), this difference can be attributed to the fact that the DNA HTT was reported first and was more closely investigated. In contrast, no study has been conducted to understand the dynamics of the copia element invasion in the D. willistoni species in other locations in its wide geographical distribution. To address this point, we performed a population analysis of the copia element distribution in South American samples of D. willistoni, as well as two strains originally used by Jordan et al. (Reference Jordan, Matyunina and McDonald1999) . Other species of the D. willistoni group were also screened.
2. Materials and methods
Nucleotide sequences with similarity to the 5′LTR-URL region of the copia retroelement were screened by PCR and Southern blot analyses in species and strains of the willistoni group (Table 1). The identification of the cryptic species of the willistoni group was confirmed by isozyme patterns of acid phosphatase (Acph1) (Garcia et al., Reference Garcia, Rohde, Audino, Valente and Valiati2006). Table 1 also shows the species used as positive and negative controls for copia presence in the molecular assays.
Table 1. List of analysed species and strains, with their respective collection sites and the results of Southern blot and PCR analyses in relation to the presence (+), absence (−) or weak signal (?) for 5′LTR-URL sequence of the copia retrotransposon
a White apricot mutant.
cf White-coffee mutant.
Genomic DNA was prepared from adult flies as previously described (Oliveira et al., Reference Oliveira, Wallau and Loreto2009). PCR analyses were performed in 50 μl reactions using 50 ng of genomic DNA, 1 U of Taq DNA Polymerase (Invitrogen), 1×reaction buffer, 200 μM of NTPs, 20 pmol of each primer and 2·5 mM of MgCl2. The primers used are specific to the 5′LTR-URL region of copia retrotransposon from D. melanogaster, amplifying a 440 bp fragment (Jordan & McDonald, Reference Jordan and McDonald1998). The amplification conditions were 94°C for 5 min, 30 cycles at 94°C for 45 s, 52°C for 60 s and 72°C for 60 s, and the final step at 72°C for 5 min. To exclude possible PCR contamination, a 786 bp fragment of gene COII was sequenced using primers TL2J3037 and TKN3785 (Simon et al., Reference Simon, Frati, Beckenbach, Crespi, Liu and Flook1994). PCRs with these primers were also used as controls for DNA quality (Fig. 2 b).
For the Southern blot analyses, approximately 6 μg of genomic DNA was digested with EcoRI. The DNA fragments were fractioned by agarose gel electrophoresis (0·8%) and transferred to a nylon membrane (Hybond N+, GE Healthcare). The 440 bp fragment corresponding to the copia 5′LTR-URL region of the PTZ18 plasmid produced by PCR was used for the hybridization probe, as previously described (Almeida & Carareto, Reference Almeida and Carareto2006). The membranes were hybridized at 60°C. In order to label and detect the copia sequences, the AlkPhos® kit and the CPD Star Detection kit (GE Healthcare) were used according to the manufacturer's instructions.
The membranes were rehybridized with a second probe corresponding to a single copy gene, the white gene. In this case, the primers used in probe amplification were forward: GCGCCACGAAAACATTTACT and reverse: ACATCGAGCCTGCATCTCTT. These rehybridizations have been used as a control for copia hybridization.
Sequence searches for copia element were carried out for the D. willistoni genome in the FlyBase BLAST database (http://flybase.bio.indiana.edu/blast/) (Altschul et al., Reference Altschul, Gish, Miller, Myers and Lipman1990) using the complete copia sequence for query (Accession number X02599; Mount & Rubin, Reference Mount and Rubin1985). Hits with scores above 100 were selected for further analysis. The sequenced strain of D. willistoni (Gd-H4-1 strain, Stock Center number: 14030-0811.94, of Drosophila 12 Genomes Consortium, 2007) comes from the Guadalupe Island (west Coast of Mexico's Baja California Peninsula).
3. Results and discussion
The Southern blot analyses revealed a strong hybridization signal of copia in D. melanogaster and the absence of signal in Drosophila immigrans and Drosophila paramediostriata, as expected. In the willistoni group, a few weak signals were obtained in Drosophila nebulosa, Drosophila paulistorum, Drosophila insularis, Drosophila equinoxialis and in some strains of D. willistoni. Strains previously shown by Jordan et al. (Reference Jordan, Matyunina and McDonald1999) as possessing copia did not present a hybridization signal in our studies (Table 1, Fig. 1). These faint signals were weaker than those obtained by hybridization with the D. willistoni white gene probe, which was used as a hybridization control (data not shown). These results indicate that the similarity between the canonical copia sequence, used as a probe, and the copia-related sequences occurring in the genomes of the investigated species was very low.
Fig. 1. Southern blot using a probe of 440 bp 5′LTR-URL region of copia retrotransposon from D. melanogaster. (A) (1) PTZ18 plasmid, (2) D. paramediostriata, (3) D. melanogaster, (4) D. willistoni Wip4, (5) D. willistoni 17A2, (6) D. willistoni Tucson Stock Center, (7) D. willistoni Morro Santana, (8) D. paulistorum POA, (9) D. paulistorum Andino-brasileira, (10) D. paulistorum Orinocana, (11) D. insularis and (12) D. equinoxialis. (B) (1) PTZ18 plasmid, (2) D. immigrans, (3) D. melanogaster, (4) D. nebulosa, (5) D. willistoni Wip4, (6) D. willistoni EM1.00, (7) D. willistoni Q14.F11, (8) D. willistoni Ey10.00, (9) D. willistoni EM1.00, (10) D. willistoni Q14.F1 and (11) D. willistoni TB46.02. (C) (1) D. melanogaster, (2) D. willistoni Royal Palm Park, Florida, (3) D. willistoni Santa Maria de Ostuna, Nicaragua, (4) D. willistoni Serra Talhada, (5) D. willistoni Montevideo, (6) D. willistoni 17A2 and (7) D. willistoni Wip4.
In agreement with the Southern blot analysis, only D. melanogaster produced the expected 440 bp fragment in the PCR assays. No amplification was obtained in the other species, even in the D. willistoni strains previously studied by Jordan et al. (Reference Jordan, Matyunina and McDonald1999) (Fig. 2 a).
Fig. 2. (A) PCR using the primers copPCS and copLTR (Jordan & McDonald, Reference Jordan and McDonald1998), specific to the 5′LTR-URL region of the copia retrotransposon from D. melanogaster, amplifying a 440 bp fragment. (1) negative control, (2) PTZ18 plasmid, (3) D. melanogaster, (4) D. willistoni Royal Palm Park, Florida, (5) D. willistoni Santa Maria de Ostuna, Nicaragua, (6) D. willistoni Montevideo, (7) D. willistoni Serra Talhada, (8) D. willistoni 17A2 and (9) D. willistoni Wip4. (B) PCR control using primers TL2J3037 and TKN3785 (Simon et al., Reference Simon, Frati, Beckenbach, Crespi, Liu and Flook1994) for amplification of a fragment of 786 bp of gene COII. (1) negative control, (2) D. melanogaster, (3) D. willistoni Royal Palm Park, Florida, (4) D. willistoni Santa Maria de Ostuna, Nicaragua, (5) D. willistoni Montevideo, (6) D. willistoni Serra Talhada, (7) D. willistoni 17A2 and (8) D. willistoni Wip4.
The in silico analysis for the available D. willistoni genome produced 25 hits. However, all of these hits were degenerate sequences, and the longest hit was 1360 bp long. The similarity level was also variable, with the highest observed similarity being 89·5% (see Supplementary Figure 1 available at http://journals.cambridge.org/GRH). These results suggest that these sequences correspond to old copia-like sequences that have been active for a long in the D. willistoni genome because the obtained sequences are degenerated and do not correspond to canonical copia elements.
In light of these results, the presence of the copia element in species of the willistoni group, as detected through Southern blot analysis with medium stringency hybridization (Martin et al., Reference Martin, Wiernasz and Schedl1983; Stacey et al., Reference Stacey, Lansman, Brock and Grigliatti1986), has been verified and are in agreement with the faint signals we have observed in this study. In fact, the D. willistoni genome showed sequences that have enough similarity to be detected by Southern blot analyses. For this reason, the sequences previously hybridized are probably from a copia related element, but not the canonical copia element.
The copia element has been reported to be involved in other HTT events, as observed between flies of the melanogaster group and between species of this group with Zaprionus indianus (Almeida & Carareto, Reference Almeida and Carareto2006). Also, Jordan et al. (Reference Jordan, Matyunina and McDonald1999) described a recent HTT of the copia element from D. melanogaster to some D. willistoni populations of Florida and Nicaragua. They showed, also, some populations from these geographic regions did not harbour the element, showing a polymorphic pattern among D. willistoni populations. We have enlarged these analyses to many South American populations, one Mexican population, and two of the strains previously studied by Jordan et al. (Reference Jordan, Matyunina and McDonald1999) : D. willistoni Royal Palm Park, Florida and D. willistoni Santa Maria de Ostuna, Nicaragua. We found that all these populations are void of canonical copia sequences.
Jordan et al. (Reference Jordan, Matyunina and McDonald1999) conducted some controls to exclude the possibility of PCR contamination (ITS region of 18S and 28S rDNA and Adh genes) and showed specific amplifications for each species. Also, a high stringency Southern blot was performed and showed the expected fragment in the D. willistoni strains, presenting a strong hybridization signal for D. melanogaster, a weaker signal for D. simulans and an even weaker one for D. willistoni. A possible explanation for the disagreement between the results reported here and those obtained by Jordan et al. (Reference Jordan, Matyunina and McDonald1999) is that the D. willistoni strains used in both studies present a large temporal separation. In fact, these strains have been maintained under laboratory conditions for a long time and could have lost their original canonical copia elements.
It is interesting to compare the copia and the P element HTT events between D. melanogaster and D. willistoni. In a very short period of time, which is hypothesized to be approximately 40 years, the P element invaded the D. melanogaster genome and spread worldwide (Bregliano & Kidwell, Reference Bregliano, Kidwell and Shapiro1983). On the contrary, the copia element, which invaded the D. willistoni genome, was only found in a very restricted geographic area of the D. willistoni distribution. These differences can be related to the TEs themselves, for example, variations in characteristics, such as transposition rates, requirements for host specific factors, mechanisms for transposition control, or even the population's structure of the host species. Populations that are more structured are less prone to disseminate genetic material, including TEs (Vieira et al., Reference Vieira, Fablet and Lerat2009). D. willistoni has a large geographical distribution, which extends to areas from Florida to Argentina (Spassky et al., Reference Spassky, Richmond, Perez-Salas, Pavlovsky, Mourao, Hunter, Hoenigsberg, Dobzhansky and Ayala1971; Dobzhansky & Powell, Reference Dobzhansky, Powell and King1975). Strains from any part of this distribution do not show incipient reproductive isolation, except for samples from Lima, Peru, which are considered a subspecies called D. willistoni quechua (Ayala & Tracey, Reference Ayala and Tracey1973; Robe et al., Reference Robe, Cordeiro, Loreto and Valente2010). However, the existence of endemic chromosome inversions suggests some level of population restructuring in this species (Rohde et al., Reference Rohde, Garcia, Valiati and Valente2006).
Theoretical models predicting the events following a genome invasion by a TE highlight the fact that transposition rate is a critical factor for TE distribution, without being lost due to genetic drift (Le Rouzic & Capy, Reference Le Rouzic and Capy2005). However, the transposition rate is not the only property contributing to the geographic distribution of TEs. It is a complex trait that involves host regulatory mechanisms and perhaps environmental influences. The transposition rate of copia in D. willistoni and the relationship of this TE with the host genome are not as well characterized as those of the P element.
The P element in D. melanogaster is one of the best characterized TEs, and the biological mechanism for its wide-spread distribution is well understood (Engels, Reference Engels, Berg and Howe1989). Our data show that even if the copia element was able to invade the genome of some populations of D. willistoni as suggested by Jordan et al. (Reference Jordan, Matyunina and McDonald1999), it was not able to spread to the D. willistoni populations studied here or even be maintained in the original populations under laboratory conditions. Further studies focusing on the mobilization rates and the mechanism of copia transposition in D. willistoni are required to understand the reasons why copia has failed to spread to the genomes of D. willistoni populations.
We are grateful to Ms C Cristina Parada, Dr Beatriz Goñi and Dr Yanina Panzera for strains of Drosophila willistoni; to Luiz F. V. Oliveira for valuable help in several aspects of this research and to Dr Lizandra Robe for suggestions. We also thank the reviewers for idea and suggestions. This study was supported by grants from CNPq, Fapergs, Proape-CAPES-PRPGP/UFSM.
