First insights on the retroelement Rex1 in the cytogenetics of frogs

Background While some transposable elements (TEs) have been found in the sequenced genomes of frog species, detailed studies of these elements have been lacking. In this work, we investigated the occurrence of the Rex1 element, which is widespread in fish, in anurans of the genus Physalaemus. We isolated and characterized the reverse transcriptase (RT)-coding sequences of Rex1 elements of five species of this genus. Results The amino acid sequences deduced from the nucleotide sequences of the isolated fragments allowed us to unambiguously identify regions corresponding to domains 3–7 of RT. Some of the nucleotide sequences isolated from Physlaemus ephippifer and P. albonotatus had internal deletions, suggesting that these fragments are likely not active TEs, despite being derived from a Rex1 element. When hybridized with metaphase chromosomes, Rex1 probes were revealed at the pericentromeric heterochromatic region of the short arm of chromosome 3 of the P. ephippifer karyotype. Neither other heterochromatin sites of the P. ephippifer karyotype nor any chromosomal regions of the karyotypes of P. albonotatus, P. spiniger and P. albifrons were detected with these probes. Conclusions Rex1 elements were found in the genomes of five species of Physalaemus but clustered in only the P. ephippifer karyotype, in contrast to observations in some species of fish, where large chromosomal sites with Rex1 elements are typically present. Electronic supplementary material The online version of this article (doi:10.1186/s13039-015-0189-5) contains supplementary material, which is available to authorized users.


Background
Eukaryotic genomes contain large amounts of repetitive DNA sequences, many of which are interspersed repeats derived from transposable elements (TEs) (reviewed in [1][2][3]). With the ability to integrate and occupy a large portion of the eukaryotic genome, TEs greatly influence genomic architecture (reviewed in [1,4]). TEs are also involved in karyotype evolution because these mobile sequences can induce chromosomal rearrangements, including deletions, duplications, inversions and translocations (reviewed in [1,5]). Therefore, the identification of this type of repetitive sequence may be valuable for evolutionary cytogenetic studies.
TE sequences are grouped in two large families, class I elements (retrotransposons) and class II elements (transposons), which are characterized by the intermediate molecule used in the transposition process. Class I elements have an RNA molecule as an intermediary for transposition, while class II elements move in the genome using DNA copies as intermediates or without any intermediate [2,6,7]. The eukaryotic transposons are further classified as "cut-and-paste" transposons, Helitrons and Politrons, which are, respectively, nonreplicative, rolling-circle replicative and self-synthesizing (reviewed in [2,3]). Among the eukaryotic retrotransposon, two principal groups are recognized, according to the presence or absence of long terminal repeats (LTR) flanking their open ready frames (ORFs): the LTR retrotransposons and the non-LTR retrotransposons, also known as LINEs (long interspersed nucleotide element). Additionally, Penelope and DIRS (Dictyostelium intermediate repeat sequence) retrotransposons have been identified (reviewed in [2]).
The retrotransposons families Rex (Retroelement of Xiphophorus) 1, Rex2, Rex3 and Rex6 are non-LTR retrotransposons, and they were first isolated from the fish genus Xiphophorus [8][9][10]. The Rex1 element encodes a reverse transcriptase and an apurinic/apyrimidinic endonuclease, which is frequently lost [9]. The 3′ untranslated region of several Rex1 elements is followed by tandem repeat oligonucleotides that are variable in length (5-7 nt) and sequence. Based on an analysis of RT amino acid sequence, Volff and colleagues [9] reasoned that the Rex1 sequences and Babar elements (for Battrachocottus baikalensis retrotransposon) cannot be assigned to any other known family of TE and suggested a moderately close relationship between Rex1/ Babar elements and members of the CR1 family of non-LTR retrotransposons.
The Rex family is widespread in fishes and was already mapped to a number of fish karyotypes through in situ hybridization, providing valuable markers for karyotype comparisons (examples in [8][9][10][11][12][13][14][15][16][17][18][19][20]; for review, see [21]). For Anura, the only available reports on Rex sequences arose from studies not designed specifically for the analysis of this TE but rather from the sequencing of the whole genomes of the pipid Xenopus tropicalis [22,23] and the dicroglossid Nanorana parkeri [24; GenBank accession number: JYOU00000000.1], which are representatives of the non-neobatrachian and Ranoidea, respectively. In addition, in contrast to fish, the karyotype organization of this or any other TE has not been explored using cytogenetic studies.
In this study, we aimed to evaluate i) whether Rex1 is also present in Hyloidea, the Anuran superfamily that, together with Ranoidea, composes Neobatrachia and ii) whether Rex1 sequences are sufficiently clustered in Anuran genomes to be used as chromosomal markers in Anuran cytogenetics. To assess these goals, we elected the leptodactylid genus Physalaemus, which comprises 46 species [25] that are currently arranged in two major clades: the P. cuvieri Clade with six species groups (P. biligonigerus group, P. cuvieri group, P. gracilis group, P. henselii group, P. olfersii group) and the P. signifer Clade [26]. Twenty of the species of Physalaemus have already been karyotyped [27][28][29][30][31][32][33][34][35][36][37], and the results show 2n = 22. Physalaemus is attractive for cytogenetic and genomic studies, particularly because of the high interspecific variation in the number and/or distribution of nucleolus organizer regions (NOR) [29,30,[32][33][34][35][36][37]) and because heteromorphic sex chromosomes are only recognized in P. ephippifer [34]. We searched for sequences related to the Rex1 family in five species belonging to different species groups of Physalaemus and used the isolated sequences as probes for in situ hybridization assays.

Rex1 sequences of Physalaemus species
We obtained 23 clones containing fragments of Rex1 isolated from genomic DNA of Physalaemus ephippifer, 18 from P. albifrons, 16 from P. albonotatus and one from Physalaemus aff. cuvieri. One sequence of P.
henselii and one of P. spiniger were isolated and sequenced directly without cloning.
Among the 16 fragments isolated from Physalaemus albonotatus, three types of sequences were found, which differed in nucleotide sequence and size. In the 339 bp sequence isolated from the P. albonotatus genome (Pab-Rex1C12), a 221 bp segment (positions 190-410 in Fig. 1) is missing, which was present in all of the remaining sequences isolated from the Physalaemus species. The first 190 bp and the last 160 bp of this sequence were highly similar to the corresponding regions of the other isolated sequences (Fig. 1). A codon analysis of this truncated sequence of P. albonotatus revealed an in-frame stop codon at the beginning of the isolated sequence (at positions 85-87 as shown in Fig. 1). The other two sequences isolated from P. albonotatus were 86 % similar to each other and were 547 bp and 571 bp in length. One segment with 6 nucleotides and another with 17 nucleotides that were present in the 571 bp fragments were absent in the 547 bp sequence (segments from positions 269-274 and from positions 421-437 as shown in Fig. 1). Two sequences that were 95 % similar to each other were also recognized among the fragments isolated from P. ephippifer (Fig. 1).
The Rex1 fragments isolated from all the Physalaemus species, except for the 339 bp sequence of P. albonotatus, were very similar to each other (average similarity = 92 %) and were 68 and 73 % similar to the Rex1/Babar sequences already described for Anguilla japonica and Battrachocottus baikalensis, respectively (Fig. 1). We note that the sequence isolated from Physalaemus aff. cuvieri differed significantly at positions 29-52 with regard to the other sequences of Physalaemus ( Fig. 1). When the Rex1 sequences of Physalaemus (except for the 339 bp sequence of P. albonotatus) were compared with the sequence REX1-5_XT, which was obtained from the anuran Xenopus tropicalis and previously recognized as Rex1 [22,23] (sequence available at Repbase database http://www.girin st.org/censor/index.php), lower similarity values were found (from 54 to 58 %) (Additional file 1). Comparison of the Rex1 sequences isolated from Physalaemus with the element CR1 of Gallus gallus, which Volff and colleagues consider to be distantly related to Rex1/Babar elements [9], showed no similarity (Additional file 1).
Amino acid sequence translation of all fragments of Rex1 ( Fig. 2) allowed us to clearly identify the regions corresponding to the conserved domains 3 to 7 of the RT, as identified by Malik and colleagues [37] and Volff and colleagues [8,9]. When compared to Battrachocottus baikalensis (GenBank accession number U18939.1) and Anguilla japonica (GenBank accession numbers AJ288466.1) sequences, the Rex1 sequences isolated from Physalaemus ephippifer, P. albifrons, P. henselli, P. spiniger, P. albonotatus and P. aff. cuvieri from Alenquer-PA failed to show an AAC nucleotide triplet at positions 416-418 of the sequences shown in Fig. 1, as well as of the sequences isolated from Xiphophorus helleri [9] (GenBank accession numbers AJ288450.1, AJ288444.1 and AJ288442.1). The absence of this ACC triplet potentially affects the translation of two codons. In addition, deletions of one or two nucleotides could be detected in the fragments Pep-Rex1C1, Pep-Rex1C17 and Pep-Rex1C19 of P. ephippifer after comparing these sequences to the others (Fig. 1).

Mapping of Rex1 sequences on metaphase chromosomes
When used as probes and mapped to metaphase chromosomes, the Rex1 sequences did not reveal any chromosomal sites at all (not even small dots) in the karyotypes of Physalaemus albonotatus, P. albifrons, P. aff. cuvieri or P. spiniger. In contrast, the probe Pep-Rex1C17 localized to the pericentromeric region of the short arm of chromosome pair 3 (3p per) of P. ephippifer (Fig. 3), suggesting accumulation of Rex1 sequences at this heterochromatic chromosomal region (Fig. 3inset). No hybridization signal was observed at any other site of this karyotype (Fig. 3), not even at regions previously revealed to be heterochromatic by Nascimento and colleagues [34]. The region at 3p per also hybridized with type I 5S ribosomal DNA (5S rDNA), as previously reported by Nascimento and colleagues [34].
Rex1 and 5S rDNA sequences isolated from microdissected 3p per of P. ephippifer Polymerase chain reaction (PCR) using primers for Rex1 sequences resulted in the isolation of~550-bp fragments from the microdissected 3p per band of the Physalaemus ephippifer karyotype, and three of these fragments were cloned and sequenced. Two cloned fragments were 547 bp (P. ephippifer Pep-Rex1C11P3 and P. ephippifer Pep-Rex1C12P3 in Fig. 1), and one was 554 bp (P. ephippifer Pep-Rex1C13P3 in Fig. 1). All of these fragments were very similar to the major Rex1 sequences isolated from genomic DNA of Physalaemus species (Fig. 1). The 547-bp fragments and the sequence Pab-Rex1C6 of P. albonotatus had a deletion of two segments compared to   [8,9] and Malik and colleagues [37]. Asterisks represent premature stop codons  Fig. 1).
PCR using primers for 5S rDNA resulted in the amplification of~300-bp fragments from the microdissected 3p per band. Their nucleotide sequencing proved the presence of the 5S rRNA gene in this chromosome band (Fig. 4).

Discussion
The results described here suggest that the elements we isolated from the genus Physalaemus might represent part of the retroelement Rex1, or at least sequences derived from it. We have formed this hypothesize because the elements' nucleotide sequences and presumed amino acid sequences are highly similar to the previously described RT coding region sequences of Rex1, especially those from Anguilla japonica [9] and Battrachocottus baikalensis (GenBank accession number U18939.1see comment in Volff and colleagues [9]). Although Volff and colleagues [9] affirm that Rex1/Babar elements are somewhat related to the CR1 element isolated from Gallus gallus, we could not detect any similarity between this element and the sequences isolated from species of Physalaemus.
The nucleotide sequences isolated from chromosome 3 of Physalaemus ephippifer and two of the three fragments isolated from P. albonotatus were similar to the RT-coding sequences of the Rex1 elements; however, these sequences have deletions that may affect their presumed reading frame. For the sequence Pab-Rex1C12 of P. albonotatus, a premature stop codon was also detected. It is likely, therefore, that these sequences are no longer active retroelements. Interestingly, the truncated sequences isolated from P. ephippifer were isolated from the heterochromatic block at 3p per, and this chromosomal region was detected using Rex1 probes in fluorescence in situ hybridization (FISH) assays. It is therefore reasonable to suggest that the loss of transposition ability may be related to mechanisms of accumulation of Rex1-derived sequences in this heterochromatic region. The accumulation of TEs in heterochromatin has been reported and discussed by some authors [4,19,[38][39][40][41]. Charlesworth & Langley [38] report that the suppression of heterochromatin crossing-over and gene inactivity are two factors that influence the preferential distribution of transposable sequences in such regions of the genome. For P. ephippifer, the molecular mechanisms involved in Rex1-derived sequences in 3p per were not investigated, but colocalization with 5S rDNA in FISH assays was an intriguing finding.
The association between TEs and rDNA sequences was already reported for many organisms. In arthropods, for example, R1 and R2 retrotransposable elements are components of the major rDNA [42,43]. In the fish Erythrinus erythrinus, sequences belonging to the Rex3 family were mapped at 5S rDNA sites and proposed to be involved in the spreading of 5S rDNA sequences in the genomes of some karyomorphs [43]. No evidence that supports the involvement of Rex1 sequences in spreading of 5S rDNA in Physalaemus or any other anuran is available to date. The molecular nature of the association between these sequences (whether they are interspersed or form independent clusters that are sufficiently close to co-locate in FISH assays) as well as the evolutionary implications of this association are open questions for future studies.
Although in situ hybridization revealed a cluster of Rex1 sequences at the heterochromatic band in the short arm of chromosome 3 of Physalaemus ephippifer, no other heterochromatic site in the karyotype of this species was detected by Rex1 probes in FISH experiments. Rex1 probes were unable to detect any chromosomal site in the karyotypes of P. albonotatus, P. spiniger, or P. albifrons. These findings diverge from those found for several species of fish, in which Rex1, Rex3 and Rex6 elements have been mapped to several heterochromatin sites [13,14,16,19]. The FISH experiments suggest that in Physalaemus species, Rex1 sequences are less abundant than those found for fish species. It is noteworthy that in the genome of anuran Xenopus (Silurana) tropicalis, non-LTR retrotransposons, including Rex1, CR1 and L2 elements, are estimated to correspond to only 0.6 % of the genome [23].

Conclusions
Our findings show that the Rex1 family of retrotransposons is not restricted to Xenopus tropicalis and Nanorana parkeri but is also present in the leptodactylid genus Physalaemus. The occurrence of these TEs both in two basal lineages of Anura (i.e., pipids and dicroglossids) and in a derived genus of Neobatrachia (i.e., Physalaemus) suggests that this element could be largely distributed in anuran genomes. Although Rex1 sequences were not highly clustered in the Physalaemus karyotypes, which differed from observations in fish species, we provided evidence for the accumulation of Rex1 sequences in a heterochromatin site of the karyotype of P. ephippifer that associated with 5S rDNA sequences. Our results shed new light for further investigation into the evolutionary dynamics of both types of sequences in anuran genomes.

Samples
Individuals were used that belonged to the two major clades of Physalaemus (i.e., P. cuvieri Clade and P. signifer Clade [26]). Samples of liver and muscle of P. albonotatus (P. cuvieri group), P. albifrons (P. cuvieri group), P. aff. cuvieri (P. cuvieri group), P. ephippifer (P. cuvieri group), P. spiniger (P. signifer Clade) and P. henselii (P. henselii group of the P. cuvieri Clade) were obtained from the tissue collection deposited at the Laboratory of Chromosomal Studies, at the Department of Structural and Functional Biology of the Institute of Biology, at the University of Campinas (UNICAMP) ( Table 1). Metaphase chromosome preparations of P. ephippifer individuals from Belém-PA (ZUEC 13734♂; ZUEC 13740♂, ZUEC13741♀ and ZUEC 13705♂) were obtained from cell suspensions available at the same laboratory, previously obtained by Nascimento and colleagues [34]. The experiments were approved by the ethics committee CIBio/IB-UNICAMP (#2005/03).
Isolation of partial sequence of the retrotransposon Rex1 from genomic DNA Samples of genomic DNA were isolated according to a procedure previously reported by Medeiros and colleagues [44] and subjected to PCR to isolate Rex1 sequences. PCR assays were performed using the primers RTX1-F1 (TTCTCCAGTGCCTTCAACACC) and RTX1-R3 (TCC CTCAGCAGAAAGAGTCTGCTC) [9], which are specific for isolating a region of the gene encoding the RT (ORF 2) of the TE Rex1. The products of these reactions were analyzed after electrophoresis in a 1 % agarose gel. Bands corresponding to fragments of approximately 550 bp were observed in each case. For Physalaemus albonotatus, an additional band of fragments of approximately 350 bp was obtained. All of these bands were cut from the gel with sterile blades, and the DNA fragments were purified with the GFX PCR and Gel Band DNA Purification kit (GE Healthcare).

Cloning of fragments of the retrotransposon Rex1 isolated from genomic DNA
The Rex1 fragments amplified as described above were each inserted into the plasmid vector pGEM-T (Promega), following the manufacturer's instructions. The recombinant vectors were used to transform JM109 Escherichia coli competent cells using the cloning kit Transformaid Bacterial Transformation (Fermentas), following the manufacturer's directions. The positive clones were selected and used for extraction of plasmids, according to the mini-prep method described by Sambrook and colleagues [45]. For amplification of the inserts, samples of the isolated plasmids were used in PCR with T7 and SP6 universal primers. After purification with the GFX PCR and Gel Band DNA Purification kit (GE Healthcare), samples of the amplified inserts were sequenced using the BigDye Terminator kit (Applied Biosystems), according to the manufacturer's instructions. The sequences obtained were edited using BioEdit software [46], and compared to sequences available at the GenBank and Repbase databases. For comparison to sequences in the Repbase database, we used the CENSOR engine with default parameters, searching within the Xenopus (Silurana) tropicalis collection for Rex1.

Metaphase chromosome analyses
To obtain metaphase chromosome preparations of Physalaemus ephippifer (ZUEC 13740 and ZUEC 13741), cell suspensions available at the Laboratory of Chromosomal Studies, at the Department of Structural  [47] used an anti-biotin antibody (Vector) and a fluorescein isothiocyanate (FITC)-conjugated secondary antibody (Vector). The chromosomes were stained with DAPI (0.5 μg/mL).
Isolation of Rex1 and 5S rDNA sequences from the microdissected 3p per band of Physalaemus ephippifer To demonstrate that Rex1 sequences are present in the pericentromeric C-band of the short arm of chromosome 3 of Physalaemus ephippifer (as suggested by FISHsee Results for details), which also bears 5S rDNA (as reported previously [34]), we isolated Rex1 and 5S rDNA sequences from the microdissected 3p per band by PCR. For microdissection, cell suspensions of the P. ephippifer specimen ZUEC 13734 were dropped onto a membrane strip containing polyethylene-naphthalate (PEN) that was previously exposed to UV and incubated at −20°C for 30 min. The material was subjected to C-banding according to a previously reported procedure [48]. The best metaphase examples were used for UV laser-microdissection of 18 copies of the heterochromatic pericentromeric region of the short arm of chromosome 3 of P. ephippifer with the MicroBeam 4.1 system (Zeiss). The cuts in the PEN membrane were made with UV at 0.5-0.6 μJ/pulse, and isolated regions were catapulted using a pulse of 0.2 μJ to the lid of a microtube containing 9 μL of TE buffer. The collected material was maintained in TE for at least 16 h and then subjected to PCR using the primers RTX1-F1 and RTX1-R3 [9] for the amplification of Rex1 sequences and the primers 5S-A (TACGCCCGATCTCGTCCGATC) and 5S-B (5′-CAGGCTGGTATGGCCGTAAGC-3′) [49] for the amplification of 5S rDNA sequences. The amplified Rex1 fragments were cloned and sequenced as described above. Two rounds of PCR with the primers 5S-A and 5S-B were performed, and the amplified 5S rDNA fragments were directly sequenced using the primers 5S-A [49] and 5S120T1-R (AGCTTACAGCACCTGGTATTC) [50] (see the annealing sites of the primers in Fig. 4) and the BigDye Terminator kit (Applied Biosystems), according to the manufacturer's instructions.
To ensure that the microdissected regions correspond to the 3p per band, we used the captured DNA as probes in FISH assays. Therefore, the microdissected material was first amplified using GenomePlex Single Cell WGA4 (Sigma-Aldrich) and labeled with dUTP-biotin (Roche®) using GenomePlex Single Cell WGA3 (Sigma-Aldrich) (Fig. 5). The hybridization and probe detection protocols were the same as used with the Rex1 probes (reported above).

Additional file
Additional file 1: Alignment of the fragments of the retroelement Rex1 isolated from species of Physalaemus with corresponding sequences available in GenBank and Repbase. Note that the sequence CR1 of Gallus gallus (GenBank accession number U88211.1) significantly differs from all other sequences. The primers used to isolate the sequences are indicated in gray. Black areas indicate identical sites, while variable sites are colored white. Premature stop codons are shown in blue. U18939.1 is the GenBank accession number of a Babar sequence (a Rex1-related element) of Battrachocottus baikalensis. AJ288466.1, AJ288450.1, AJ288444.1 and AJ288442.1 are GenBank accession numbers of retroelement Rex1 sequences isolated from Anguilla japonica (AJ288466.1) or Xiphophorus helleri. Xenopus tropicalis REX1-5, REX1-2 and REX1-3 are sequences isolated from Xenopus tropicalis and available at the Repbase database (http://www.girinst.org/censor/index.php). (DOCX 33 kb)

Competing interests
The authors declare that they have no competing interests.
Authors' contributions JN obtained the sequences and chromosomal data and drafted the manuscript. DB collected some of the specimens and helped with the analyses. LBL designed and coordinated the study and drafted the manuscript. All authors have read and approved the final manuscript.