A repetitive DNA family (phBglII400) was characterized in the clam species Tapes philippinarum (Veneridae Tapetinae). The tandemly repeated sequences are AT-rich and show a mainly pericentromeric localization, as most satellite DNAs do. Sequence analysis of phBglII400 DNA family showed a high level of intraspecific homogeneity. Furthermore a 200 bp subunit motif within the 400 bp monomer was apparent as well as the existence of two main “open reading frames” along the 400 bp sequence.
In order to investigate the possible distribution of this DNA family among Veneridae, Southern blot analyses were performed on genomic DNAs of Tapes decussatus, Venerupis aurea and Paphia undulata (Tapetinae), Callista chione (Pitarinae), Chamelea gallina (Chioninae) and Venus verrucosa (Venerinae). The phBglII400 family has been found in two additional Tapetinae, namely V. aurea and P. undulata, but not in T. decussatus or other analyzed species. This pattern of sat-DNA distribution supports the high level of differentiation of T. decussatus observed in the previous gene-allozyme analysis. All of these suggest a better allocation of T. decussatus to a genus different from that of T. philippinarum.
INTRODUCTION
Tapes philippinarum, Tapes decussatus, Venerupis aurea and Paphia undulata (subfamily Tapetinae, Fischer, 1887) represent some of the economically most important bivalves. Their taxonomic position and phyletic relationships are questioned. On the basis of shell morphology, “decussatus” and “philippinarum” taxa have been invariably considered congeneric and ascribed to Tapes Von Muehlfeldt, 1811; “aurea” has been most often ascribed to Venerupis Lamarck, 1818 (Fischer-Piette and Metivier, 1971), but, recently, also to Paphia Roding, 1798 (Sabelli et al., 1990). Finally, “undulata” has been always ascribed to Paphia. On the other hand, morphological data sharply contrast allozyme and mitochondrial DNA analyses (Borsa and Thiriot-Quiévreux, 1990; Canapa et al., 1996; Passamonti et al., 1997). In particular, biochemical and mtDNA findings do not support the assignment of T. decussatus and T. philippinarum to the same genus.
A useful tool to a further understanding of the phyletic relationships within the genus Tapes and the Tapetinae sub-family, could be a molecular analysis of tandemly repeated DNA sequences in their genomes. Eukaryotic genomes embody large amounts of tandemly repeated, non-coding DNA sequences, so-called satellite DNA (John and Miklos, 1979). Sat-DNA is highly variable in complexity and copy number; the repeated units can range from 3, 4 to thousand base pairs and may represent from a few units to more than 50% of the genomic complement and can be arranged into either short interspersed or long tandemly repeated units (Singer, 1982; Miklos, 1985).
The general significance of satellite DNA is rather controversial and several hypotheses have been developed to assign this fraction of the eukaryote genome a role. It has been involved in heterochromatin constitution (Gershenson, 1933, 1940; Pardue, 1975; Brutlag, 1980), chromosome pairing (Salser et al., 1976; Fry and Salser, 1977), rearrangements (Hatch et al., 1976), tridimensional organization of the interphasic nucleus (Hilliker et al., 1980; Manuelidis, 1982), gene amplification events (Bostock and Clark, 1980) and chromosome - mitotic spindle interactions through a peculiar class of centromeric sat-DNA-binding proteins (CENP), that preferably interact with DNA curvatures caused by d(A · T)n≥5 stretches (Masumoto et al., 1989). Alternative hypotheses suggest that sat-DNA lacks any function except its own survival (Doolittle and Sapienza, 1980; Orgel and Crick, 1980). Since it often corresponds to a heterochromatic noncoding genomic fraction, some models have been produced to explain its maintenance and evolutionary processes, such as “molecular drive” and “concerted evolution” (Dover, 1982, 1986, 1989; Charlesworth et al., 1994).
Although a peculiar sat-DNA family is usually restricted to a single or a few closely related species, sometimes satDNAs show remarkable similarities in taxonomically distant taxa; highly conserved satellite DNAs can be spread over a whole group of phylogenetically related organisms (Arnason et al., 1984; Cremisi et al., 1988) or even be widely distributed among members of evolutionarily distant groups (Abad et al., 1992). It should be noted, however, that a higher level of homogeneity has been commonly found in intraspecific sequence analyses than in inter-specific comparisons. Therefore this approach could help when traditional investigations of taxonomy appear to be inadequate or even give controversial results at the specific level (Bachmann et al., 1993).
Therefore we started a genomic DNA screening of T. philippinarum, T. decussatus, V. aurea and P. undulata, all belonging to the Tapetinae subfamily. A sat-DNA family was first detected in T. philippinarum and its distribution in the remaining species was analyzed. Comparisons were afterwards made also with genomic DNAs of Callista chione (subfamily Pitarinae), Chamelea gallina (subfamily Chioninae) and Venus verrucosa (subfamily Venerinae).
MATERIALS AND METHODS
Nine individuals of seven Veneridae species from different localities were analyzed: two T. philippinarum (phSc - Scardovari Lagoon, Po river estuary; phGa - Ganzirri Lagoon, Sicily) and T. decussatus (deSf - Sfax, Tunisia; deCh - Chioggia, Venice Lagoon); one each of V. aurea (auCi - Civitanova Marche, Middle Adriatic Sea), P. undulata (unTh - Thailand), C. chione (chCh - Chioggia, Venice Lagoon), C. gallina (gaCh - Chioggia, Venice Lagoon) and V. verrucosa (veCh - Chioggia, Venice Lagoon).
Genomic DNAs were isolated by homogenizing a single foot-muscle in 160 mM sucrose, 80 mM EDTA and 100 mM Tris-HCl, pH 8 buffer and incubated at 65°C for 1 hr after addition of 0.5% SDS and 1 μg/ml of Proteinase K (Boehringer). Homogenates were then extracted for 3-5 times with phenol and/or chloroform. SDS separation was performed using potassium acetate at the final concentration of 1.2 M for 30′ on ice and centrifuging for 15′ at 15,000 rpm. The DNA was precipitated by the addition of 2 volumes of ethanol, centrifuging for 15′ at 12,000 rpm. The DNA was recovered in TE 1 × buffer (10 mM Tris-HCl pH 8 and 1 mM EDTA) and treated with 50 μg/ml of RNAse A (Boehringer) for 10′ at room temperature. The DNA was finally precipitated with 0.2 M sodium acetate and 2 volumes of ethanol. From 50 to 200 μg of genomic DNA were obtained through this procedure.
Genomic DNAs were digested with 5 units of specific endonucleases/μg of DNA for 7-8 hr, according to the manufacturer's instructions. The following endonucleases were used: Acc, AluI, ApaI, AvaI, BamHI, BclI, BglII, Dde, DraI, EcoRI, HaeIII, HindII, HindIII, Hpa, MspI, Nde, NsiI, PstI, RsaI, SacI, SalI, SmaI, StuI, TaqI. All restricted DNAs were separated on 1.2% agarose gels in 1 × TBE buffer (90 mM Tris-borate; 2 mM EDTA) and then transferred onto a Nylon Membrane (Boehringer) following the protocol of Southern (1975).
From the BglII digested genomic DNAs of T. philippinarum (Scardovari and Ganzirri), a band, at about 400 bp, was cut and eluted from a 1.2% agarose gel using the Genomed JETsorb Gel Extraction Kit (cat. n. 110150). Eluted DNAs were cloned into the BamHI sites of pGEM -7Zf(+) (Promega) or pUC19 (+) (Pharmacia) plasmid vectors.
V. aurea and P. undulata monomers, only observed in Southern blots, could not be cloned directly, owing to their extremely low quantity. Therefore, a PCR amplification was performed with 5′-AGATTTCCGCATGGC-3′ and 5′-GACGTGTTCCTTGGG-3′ primers, designed on the already obtained T. philippinarum sequence; these primers amplified the second part of the 400 bp sequence, since no good couple of primers could be designed to amplify the whole monomer. Primers were designed using PCR PLAN software in the PC/GENE 6.6 software package (© 1991, IntelliGenetics, 700 East El Camino Real, Mountain View, California). The obtained fragments were therefore cloned into a pCR-Script plasmid vector with the Stratagene pCR-Script™ Amp SK(+) Cloning Kit (Cat. n. 211189).
Labeling and Southern blot hybridizations were performed according the DIG-DNA kit instructions by Boehringer.
The nucleotide sequences of the recombinant clones were determined using the AutoRead™ Sequencing Kit (Pharmacia) with T7 and SP6 universal primers and a ALF™9000 automated sequencer (Pharmacia). Acrylamide gels were casted using ReadyMix Gels, ALF™ grade (Pharmacia).
Sequence alignments, consensus sequence, mutation distribution and codon analysis were performed using the Clustal algorithm and the Sequence Navigator software (Applied Biosystems). Genetic distances and cluster analyses were performed using the MEGA program (ver. 1.01, © S. Kumar, K. Tamura and M. Nei, 1993); in particular was utilized the “Kimura's two parameters” mutational model (Kimura, 1980) in computing distances.
Giemsa and NOR karyotypes and in situ hybridizations were performed on chromosome preparations from the gill tissue of T. philippinarum from Scardovari. Living clams were treated overnight with 0.005% colchicine in marine water. The dissected gills were hypotonically shocked with 1% w/v sodium citrate solution and then fixed with Carnoy's liquid (3:1, ethanol: acetic acid). Slides were prepared using a slightly modified air-drying technique by Crozier (1968).
Some slides were directly Giemsa stained, whereas others were treated with a silver impregnation technique, according to Marescalchi and Scali (1990), to evidence the active nucleolar organizing regions (NOR).
In situ hybridization was performed following Maluszynska and Heslop-Harrison (1993) with a DIG-labeled monomer obtained from the pphSc11 recombinant clone. Slides were examined under a Zeiss Axioskop fluorescence microscope with a fluorescein-specific filter set. Photomicrographs were taken on an Ektachrome Panther P1600 color-reversal film. The chromosomes were then printed at a final magnification of 1250x.
RESULTS
A restriction analysis of the genomic DNAs of T. philippinarum, T. decussatus, V. aurea and P. undulata was carried out. A ladder-like pattern was observed on ethidium bromide stained gels of the BglII restricted DNAs of T. philippinarum, its monomer band being localized at about 400 bp position. The cloning of the 400 bp monomer led to 8 recombinant clones (pphSc1, pphSc4, pphSc7, pphSc10-14) from Scardovari and 9 (pphGa1-5, pphGa7-10) from Ganzirri.
Southern blot hybridizations of restricted DNAs of all analyzed Veneridae species were also performed with the DIG-labeled pphSc11 clone. In T. philippinarum, V. aurea and P. undulata, they evidenced a main multimeric ladder with a 400 bp monomer and an intervening fainter ladder with a 200 bp monomer in the BglII restricted DNAs (Fig. 1). On the other hand, through this technique no related sequences were detected in T. decussatus (Fig. 1) as well as in the remaining Veneridae (C. chione, C. gallina and V. verrucosa - not shown). The multimeric pattern showed that the units were tandemly repeated and the family of related sequences was named phBglII400.
Sequences of 5 T. philippinarum randomly cloned monomers from Scardovari and Ganzirri were obtained (Table 1). The sequence was AT rich (about 60%) and embodied a few d(A·T)n≥5 stretches. Alignment analysis evidenced clear homology between all analyzed sequences and indicated that the monomer length of the family ranged from 400 to 405 bp, with small variation caused by either single insertion/deletions (bases 39, 62, 74, 242, 359, 368 and 372) or multiple ones (only observed between positions 365-370 in clones pphSc10 and pphSc4). The highest concentration of indels was localized in the 359-372 stretch of the sequence. As a consequence, very high sequence homology, ranging from 97.5 to 99.3%, resulted.
Table 1
Alignment of the phBglII400 DNA family cloned monomers of T. philippinarum from Scardovari (pphSc) and Ganzirri (pphGa). A consensus sequence, with BglII-like site (underlined) is shown. Asterisks (∗) indicate gaps introduced to increase sequence similarity. Last digits of numerals are aligned with the corresponding nucleotide.
The number of base-pair differences ascribable to point mutation was not evenly distributed throughout the sequence, ranging from 0 (positions 1-20, 51-60, 101-140, 221-230, 241-250, 281-320, 231-240, 401-410) to 4 (positions 141-150) every 10 bp.
Between bases 200 and 205 a BglII-like site (AGATTT, underlined in Table 1), was observed. Such a sequence might explain the secondary fainter ladder with a 200 bp monomer, observed in Southern blots, because a point mutation of the BglII site (AGATCT) into a BglII-like restriction site (AGATTT) in a subset of the sequences could result in both 400 bp (more numerous) and 200 bp (fewer) monomers. The mutated site has been used as a landmark to align the first part (bp 1-199) on the second one (bp 200-405) of the sequence of the cloned pphSc11 monomer. When aligned in this way, the two subunits (∼200 bp) showed a high level of similarity, although differed in length for deletions/insertions and in nucleotide sequence for point mutations (Table 2).
Table 2
Alignment between the 200 bp subunits of the pphSc11 monomer. #1 = first part (bp 1-199) of the pphSc11 sequence; #2 = second part (bp 200-405) of the pphSc11 sequence. Bars (|) indicate homologous sites. Asterisks (∗) indicate gaps introduced to increase sequence similarity.
Codon usage analysis of the whole 405 bp consensus sequence evidenced only three non-sense codons and two main open reading frames (ORF): the first one from base 2 to base 169 and the second one from 173 to 367 (56 and 65 codons in length, respectively).
“Kimura's two parameters” genetic distances were low, ranging from 0.037 to 0.059 (Table 3); furthermore no geographic differences between clones were found. None of neighbor-joining or maximum parsimony dendrograms (not shown) supported geographic differentiation.
Table 3
“Kimura two parameters” genetic distances between randomly cloned T. philippinarum monomers
PCR amplification and cloning allowed to obtain recombinant clones for both V. aurea (pauCiPCR5 and pauCiPCR26) and for P. undulata (punThPCR1 and punThPCR17) with a 200 bp monomer. Two PCR recombinant clones of T. philippinarum were also obtained for comparison (pphScPCR10 and pphScPCR11). Southern blot hybridization with the DIG-labeled pphSc11 clone (not shown) demonstrated the homology of the obtained sequences to the phBglII400 family.
Alignment of sequenced 200 bp stretches evidenced a high level of similarity among clones (0.990-0.995%): only 13 point mutations were observed (Table 4) and “Kimura's two parameters” distances ranged from 0.000 to 0.042 (Table 5). Chromosome analysis of T. philippinarum specimens from Scardovari constantly showed a 2n = 38 metaphase complement, with mainly meta- and submetacentric chromosomes. Moreover, silver staining technique invariably revealed a homozygous condition of active nucleolar organizing regions (NORs) on long arms of pair 17. In situ hybridization of T. philippinarum metaphases evidenced a mainly pericentromeric localization of the phBglII400 family on most chromosomes (Fig. 2). However, the rather homogeneous morphology of T. philippinarum chromosomes and their strong contraction in C-metaphases did not allow a very clear karyotype definition from FISH labeled sets. Giemsa and NOR-stained metaphases are available upon request.
Table 4
Alignment of the PCR-obtained 200 bp sequences from the phBglII400 DNA family of T. philippinarum from Scardovari (phSc), V. aurea from Civitanova Marche (auCi) and P. undulata from Thailand (unTh). A consensus sequence, with underlined primers' annealing sites, is shown.
Table 5
“Kimura two parameters” genetic distances between PCR-obtained 200 bp sequences of the phBglII400 family in three Tapetinae species (T. philippinarum, ph; V. aurea, au; P. undulata, un).
DISCUSSION
Previous allozyme analyses have evidenced high levels of genetic divergence between T. philippinarum and T. decussatus; T. philippinarum is genetically more similar to P. undulata and V. aurea than to T. decussatus (Passamonti et al., 1997). Analyses of 16S rRNA sequences in mtDNA (Canapa et al., 1996) appear to be in line with gene-enzyme findings.
Some features of the phBglII400 family are informative. Southern blot fits within the above reported phyletic frame, since the phBglII400 family appears to be shared by all the Tapetinae species presently analyzed with the only exception of T. decussatus. This reinforces the high level of differentiation of T. decussatus from the other Tapetinae taxa, which form a sound cluster. As a consequence, these data sets suggest that the two Tapes species could be separated to different genera.
While V. aurea is found only in the Mediterranean basin, T. philippinarum and P. undulata are Indo-Pacific clams in origin, their evolutionary divergence dating back, at least, to the closure of Tethys Ocean. Even if at present T. philippinarum and V. aurea have become artificially sympatric, the hypothesis of a recent “horizontal” transfer of these sequences appears very unlike, since neither natural hybridization has been observed, nor artificial hybrids have been obtained. The assumption seems to be inconsistent with the observation that intra-specific sequence divergence in the 405 bp monomers of T. philippinarum is somewhat higher than the inter-specific one, as obtained from the PCR sequences. It is to be noted, however, that this comparison is heavily biased by the non-random working of the PCR-cloning procedure: likely, the observed distances are an outcome of the primer design performed on T. philippinarum. As a consequence, PCR amplified sequences of V. aurea and P. undulata would have artificially increased similarity to those of T. philippinarum.
Very little is known about satellite DNA in the Mollusca; in the Bivalvia, only available for Mytilus edulis (Ruiz-Lara et al., 1992). No definite hypotheses on the function of these sequences can be put forward. The high inter- and intra-specific sequence similarity of the phBglII400 family may be attributed to the homogenization phenomena of the “concerted evolution”; these proceed through genomic turnover mechanisms (such as unequal crossing over, slippage replication and rolling circle amplification) and bisexual reproduction (Dover, 1982, 1986, 1989; Charlesworth et al., 1994). Even the existence of some sort of selective constraints, such as satellite DNA-centromeric protein (CENP) interactions or hypothetical ORF transcription, could play a role in the maintenance of a high similarity of the phBglII400 monomeres.
Acknowledgments
We greatly appreciated the skilful assistance of Dr. Ettore Randi (Istituto Nazionale per la Fauna Selvatica, Ozzano Emilia, Bologna) in the PCR approach. This work was supported by M.U.R.S.T. 40% and 60% funds.