Restriction fragment analysis of the ribosomal DNA of Paratelmatobius and Scythrophrys species (Anura, Leptodactylidae)

Physical maps of the ribosomal RNA gene 28S of species belonging to the genera Paratelmatobius and Scythrophrys were constructed, using five restriction endonucleases. The restriction sites for Bam HI, Bgl II, Bst EII, and Eco RI had similar positions in all species, although there were interspecific differences in the size of the restriction fragments obtained. An additional Pvu II site was found in Scythrophrys specimens from Piraquara (State of Paraná, Brazil) and from São Bento do Sul (State of Santa Catarina, Brazil), but not in the Scythrophrys specimens from Rancho Queimado (State of Santa Catarina, Brazil). This finding is in agreement with the hypothesis regarding the existence of two species in the genus Scythrophrys. On the other hand, the extra Bst EII site considered in the literature to be a synapomorphy for the subfamilies Leptodactylinae and Telmatobiinae was not observed in the genera Paratelmatobius and Scythrophrys, which brings new questions about some taxonomic classifications that include Paratelmatobius in Leptodactylinae and Scythrophrys in Telmatobiinae. Interspecific variation was observed in the size of the restriction fragments analyzed and, in the case of group I Scythrophrys, there was also a variation between the individuals of the two populations. These data suggest that sequencing of the rDNA segments studied here may be useful in phylogenetic studies of the genera Paratelmatobius and Scythrophrys.


Introduction
The nuclear ribosomal DNA (rDNA) in the NOR (nucleolar organizer regions) of eukaryotes consists of tandemly repeated copies of the transcriptional unit for the precursor of 18S, 5.8S, and 28S ribosomal RNA (rRNA), separated by an intergenic spacer (IGS).Two internal transcribed spacers, known as ITS 1 and ITS 2, are located between the regions coding for 18S and 5.8S rRNA, and between the latter and the 28S coding region.In addition, an external transcribed spacer (ETS) occurs upstream to the 18S gene (Long and Dawid, 1980;Miller, 1981).The copies of rDNA genes in a species evolve in concert through coordinated mechanisms (Arnheim et al., 1980;Dover and Coen, 1981;Krystal et al., 1981;Coen et al., 1982a,b;Arnheim, 1983).As a result, intra-and interindividual variations in rDNA are small and, when present, affect particularly the intergenic spacer (see Hillis and Dixon 1991 for a review).
The low rate of polymorphism in the rDNA transcription unit allows characterization of the rDNA of each spe-cies using only a few specimens, and makes this DNA useful for interspecific comparisons.In addition, the different coding regions of the rDNA repeats usually show distinct evolution rates.As a result, this DNA can provide information about almost any systematic level (see Hillis and Dixon, 1991).
In anurans, restriction mapping of the 28S gene in 54 species identified regions of variability that could be useful in phylogenetic studies (Hillis and Davis, 1987).Additionally, an interesting Bst EII site was reported by Hillis and Davis (1987) as a possible synapomorphy for the leptodactylid subfamilies Leptodactylinae and Telmatobiinae, as they are considered by Duellman and Trueb (1986).
The leptodactylid genera Paratelmatobius and Scythrophrys are closely related, as suggested by morphological, ecological (Lynch, 1971;Heyer, 1975;Garcia, 1996) and cytogenetical (Lourenço et al., in press) data.However, many authors have considered them as belonging to two different subfamilies, Leptodactylinae and Telmatobiinae, respectively (see Lynch, 1971;Duellman and Trueb, 1986;Frost, 2000).According to the most recent taxonomic review of amphibians, both Paratelmatobius and Scythrophrys belong to the same subfamily, named Cycloramphinae (Frost, 2002).Actually, the taxonomic relationships of Paratelmatobius and Scythrophrys with other leptodactylids are still not clear, and further data are needed for their correct allocation in subfamilies.
The available information about Paratelmatobius and Scythrophrys includes cytogenetic analyses showing interspecific variation in the NOR location in three karyotypes of Paratelmatobius (Lourenço et al., 2000;Lourenço et al., 2003) and two of Scythrophrys (Lourenço et al, in press).However, the organization of the rDNA genes in these genera is not known yet.
In this paper, we present a preliminary analysis of the 28S ribosomal RNA gene in Paratelmatobius and Scythrophrys.

Southern blot
Genomic DNA was extracted from fresh or frozen liver and muscle tissue, using a standard phenol/chloroform method (Sambrook et al., 1989).About 1-2 µg of genomic DNA from each specimen were digested with the restriction enzymes Eco RI, Bam HI, Bgl II, Bst EII, and Pvu II, used singly or in combination.The resulting fragments were separated on 1.2% agarose gels at about 1.5 V/cm in 1x TBE, and transferred onto positively charged nylon membranes (Hybond) (Sambrook et al., 1989).

28S rDNA probe
An Eco RI-Bam HI fragment (~2,5 kb) of the Xenopus laevis 28S gene extracted from the plasmid HM 123 (Meunier-Rotival et al., 1979) was subcloned into the phagemid pBlueScript (pBS 28) (Figure 1) and used as probe, after labeling with digoxigenin by the random primer method, according to the manufacturer's instructions (Boehringer Mannheim).

Hybridization
The membranes were pre-hybridized for 2-3 h in 5x SSC, 50% formamide, 0.1% lauroylsarcosine, 0.02% SDS, and 2% blocking reagent (Boehringer Mannheim), and then hybridized overnight with the pBS 28 probe, at 42 °C in pre-hybridization solution.After hybridization, the membranes were washed twice in 2x SSC and 0.1% SDS solution at room temperature, and twice in 0.1x SSC and 0.1% SDS solution at 60 °C.Each wash lasted 15 min.The probe was detected by using a chemiluminescence system from Boehringer Mannheim, with CSPD as the chemiluminescent substrate.After incubation with the substrate, the membranes were exposed to X-ray films (Hyperfilm ECL, Amersham).The bands were documented, and the length of the restriction fragments was inferred using a Kodak Digital Science System.

Results
Restriction site maps of the 28S ribosomal gene were constructed for P. poecilogaster, P. cardosoi, Paratelmatobius sp (aff.cardosoi), and for three Scythrophrys populations (Figure 2).The restriction sites for Bam HI, Bgl II, Bst EII and Eco RI had similar positions in all species, although there were interspecific differences in the size of the restriction fragments obtained (Figures 2-4).The digestions using Pvu II showed that the individuals from the two group I Scythrophrys populations had an extra site for this enzyme, that was absent in the Scythrophrys specimens from Rancho Queimado and in the other species analyzed.Hylodes asper had three Pvu II sites (Figures 5 and 6).
Weak bands were detected after double digestion with Bgl II and Pvu II, and were considered to be the product of partial digestions (Figure 5).This conclusion agreed with the pattern obtained in other experiments and by some 140 Lourenço et al. single digestions not shown here.Comparison of these data with the restriction sites in the rDNA sequence of Xenopus laevis (GenBank/X02995) indicated that the ~1 kb fragment in lanes 1 and 5 of Figure 5 was a Bgl II-Pvu II segment derived from partial digestion of Pvu II, since this hypothetical site is present in X. laevis.The total digestion of this fragment probably gave rise to two fragments, one of ~900 bp and another of about 100 bp, not detected in our hybridization experiments.Since partial digestions were not done for all of the species studied, the hypothetical Pvu II site mentioned above was not considered in interspecific comparisons.
There was no intraindividual or intrapopulational variation either in the number or in the size of the fragments generated in any of the experiments (Figures 3-6 and data not shown).In contrast, an interpopulational variation was observed in the size of the digested fragments of Scythrophrys group I (Figure 3: lanes 10-12; Figure 4: lanes 6-8; Figure 5: lanes 5-7).Interspecific variation in the size of the digested fragments was also found.The three Paratelmatobius species, for example, could easily be distinguished from each other by variations in the size of the Bgl II-Pvu II fragment (Figure 5: lanes 2-4) and of the Bst EII-Bam HI fragment (Figure 2), whose size was inferred from the analysis of the experiments shown in Figures 3  (lanes 7-9) and 4 (lanes 3-5).

Discussion
The restriction sites of Bam HI, Bgl II, and Eco RI found in all the specimens studied here were also detected in the anurans analyzed by Hillis and Davis (1987).The same applies to the Pvu II site located between the Bgl II and the Eco RI sites.On the other hand, the Bst EII site observed in all Paratelmatobius and Scythrophrys species was absent in some of the anurans investigated by Hillis and Davis, although it was present in the three leptodactylids analyzed by them (Ceratophrys ornata, Leptodactylus wagneri and Telmatobius niger).
Moreover, Hillis and Davis (1987) detected an interesting extra Bst EII site between the Bgl II and the Pvu II sites mentioned earlier.It was found only in the leptodactyline Leptodactylus wagneri and in the telmatobine Telmatobius niger and was therefore considered to be a possible synapomorphy for these Leptodactylidae subfamilies.None of the Paratelmatobius and Scythrophrys species studied here had this extra Bst EII site.So, if we consider the classification adopted by Duellman and Trueb (1986) and followed by Hillis and Davis (1987), our results refute the hypothesis of Hillis and Davis about the interpretation of the extra Bst EII site, since, by that classification, Scythrophrys belonged to the RFLP analysis of rDNA of some anurans 141   Telmatobiinae and Paratelmatobius to the Leptodactylinae subfamily.
The monophyly of Leptodactylinae and Telmatobiinae (sensu Duellman and Trueb, 1986) was already refuted in a study by Ruvinsky and Maxson (1996), in which Eleutherodactylus cuneatus (currently allocated to the Eleutherodactylinae subfamily) was the representative of Telmatobiinae, and Lithodytes lineatus the representative of Leptodactylinae.In that paper, the results also suggested Leptodactylidae polyphyly.
On the other hand, if we consider the current classification presented by Frost (2002), in which Paratelmatobius and Scythrophrys are allocated together and apart from the Leptodactylinae and Telmatobiinae species, our results do not confront with the hypothesis brought by Hillis and Davis (1987).So, although a study including various Leptodactylidae genera and several characters is necessary for a conclusive analysis of their taxonomy and intergeneric relationships, our results suggest that an investigation of the Bst EII site in the 28S gene may be useful.
The other informative region was the additional Pvu II site found in the two group I Scythrophrys populations, which was absent in the Scythrophrys specimens from Rancho Queimado, in the Paratelmatobius species, and in other anurans of this study and others (Hillis and Davis, 1987).Since intraspecific variation in the 28S gene is not common (see Hillis and Dixon, 1991), and the existence of two species in the genus Scythrophrys has already been proposed based on studies of the same populations sampled here (Lourenço et al., in press), it is possible that the variation in the Pvu II restriction site described above represents an interspecific divergence.
It is also interesting to note that there was no intraindividual variability in P. poecilogaster, despite the two NOR-bearing chromosome pairs in this karyotype (chromosome pairs 8 and 10; Lourenço et al., 2000).Moreover, all the species studied, as well as the individuals from the two group I Scythrophrys populations, could be distinguished from each other by the size of the digested fragments.This finding and the restriction site variation discussed above suggest that sequencing of the rDNA segment focused in this study could provide useful data for phylogenetic studies of Paratelmatobius and Scythrophrys.
, eight Scythrophrys from the karyotype group I (five from Piraquara, State of Paraná, and three from São Bento do Sul, State of Santa Catarina, Brazil), and six Scythrophrys specimens from the karyotype group II (from Rancho Queimado, State of Santa Catarina, Brazil) (see Lourenço et al, in press) were studied.The P. poecilogaster and P. cardosoi specimens were collected at Paranapiacaba (State of São Paulo, Brazil), and the Paratelmatobius sp (aff.cardosoi) specimens came from Piraquara (State of Paraná, Brazil).For comparison, specimens of two Cycloramphinae genera (sensu Frost, 2002), Cycloramphus izecksohni (previously in the subfamily Leptodactylinae) (one specimen from Corupá, State of Santa Catarina, Brazil) and Hylodes asper (previously in the subfamily Hylodinae) (one specimen from Paranapiacaba) were also used.All the specimens studied were deposited at the Célio F.B. Haddad collection (CFBH), Departament of Zoology, Universidade Estadual Paulista, Rio Claro, SP, Brazil, and at the "Prof.Adão José Cardoso" Museum of Natural History (ZUEC), Universidade Estadual de Campinas, Campinas, SP, Brazil.

Figure 1 -
Figure 1 -Representation of the Xenopus laevis rDNA fragment in the plasmid HM 123 (Meunier-Rotival et al., 1979).The arrows and arrowheads indicate the restriction sites of Eco RI and Bam HI, respectively.The solid block represents the fragment of about 2.5 kb subcloned in pBueScript and used as probe.

Figure 2 -
Figure 2 -Restriction maps of Paratelmatobius and Scythrophrys 28S rDNA for the enzymes Bst EII, Bam HI, Bgl II, Pvu II and Eco RI.(The presumed Pvu II site of Scythrophrys from group II experiments is not shown).