NO EVIDENCE OF DOUBLY UNIPARENTAL INHERITANCE IN THE BROWN MUSSEL Perna perna FROM THE RFLP ANALYSES OF THE MITOCHONDRIAL 16 S rDNA

Weber, L.I.; Hildebrand, C.G. & Clezar, C. 2009. No evidence of doubly uniparental inheritance in the brown mussel Perna perna from the RFLP analyses of the mitochondrial 16S rDNA. Braz. J. Aquat. Sci. Technol. 13(2):51-57. ISSN 1808-7035. The doubly uniparental mode of inheritance, characteristic of the Mytilidae, consists of an F-type mitochondrial lineage transmitted only through females and an M-type lineage present only in male gonads and therefore transmitted only through males. In this work, we search for evidence of two mitochondrial lineages in Perna perna by studying the same 16S rRNA region that allowed for the discovery of doubly uniparental inheritance in the Mytilus group and in one venerid clam. The region was screened for substitutions using eight restriction enzymes to analyze two kinds of tissues (somatic and gonadal) from 20 males and 20 females from the south of Brazil. A restriction map was constructed after confirming restriction sites with sequence analysis. After amplification with the same primers used in previous studies, a fragment of 517 bp was obtained, which was 10 bp shorter than the one from Mytilus species. No variation was found among individuals or between sexes or kinds of tissue (gonadal and somatic). The absence of variation in this region was confirmed by sequence analysis, and this result left us unable to reject the hypothesis of common maternal mitochondrial inheritance. The 16S rDNA sequence obtained for Perna perna was aligned with twenty-three representative sequences from fifteen mytilid species available in GenBank, and a neighborjoining tree was constructed. The phylogenetic analysis showed all Perna species clustered together in a single branch, which was supported 100% by bootstrap analysis and was more closely related to the branch of the Mytilus group than the other representative genera of the Mytilidae.


INTRODUCTION
The mussels of the family Mytilidae have unique mitochondrial genome when compared with other metazoans (Hoffman et al., 1992;Boore et al., 2004;Mizi et al., 2005), as shown by studies of Mytilus edulis Linnaeus, 1758 and M. galloprovincialis (Lamarck1819).The absence of ATPase 8 and the presence of two tRNA Met genes in the mitochondrial genome of the Mytilidae were characteristics only seen before in Nematoda (Okimoto et al., 1991).The mitochondrial genome of Mytilids shows the highest evolutionary rate within the metazoans (Hoeh et al., 1996).These authors estimated, for the genus Mytilus, about 5.23 x 10 -4 substitutions/nucleotide-site/Myr, more than twice the values estimated for Drosophila (2.31 x 10 -4 substitutions/nucleotide-site/Myr) and for humans (2.48 x 10 -4 substitutions/nucleotide-site/Myr).This is attributed to the high rates of evolution of the male lineage characteristic of this group, which shows another very interesting feature of the mitochondrial genome: doubly uniparental mitochondrial inheritance (Hoeh et al., 1996).
Most metazoans inherit the mitochondrial genome in a unique matrilineal way, explained by the interruption of the passage of male mitochondria or by their destruction within the first 24 h after the recognition of a factor found on the male mitochondrial surface by a nuclear factor (Sutherland et al., 1998;Zouros, 2000).It is believed that this recognition system is broken in some bivalves, in which both male and female lineages have been observed, showing the exceptional doubly uniparental inheritance mode of the mitochondrial genome (DUI;Fisher & Skibinski, 1990;Zouros et al., 1992;Zouros et al., 1994b) already observed in the Mytilidae (Fisher & Skibinski, 1990;Zouros et al., 1992;Skibinski et al., 1994a,b;Zouros et al., 1994a,b;Rawson & Hilbish, 1995), in the freshwater bivalves of the Unionidae family (Liu et al., 1996), and, more recently, in the Veneridae (Passamonti & Scali, 2001).The doubly uniparental mode of inheritance is based on the existence of heteroplasmic males (Skibinski et al., 1994b;Zouros et al., 1994b), which have in their somatic tissues mainly the F (Female) mtDNA lineage, while having in their gonads exclusively the male lineage (M) (Garrido-Ramos et al., 1998); females, meanwhile, are normally homoplasmic, with few exceptions, and have only the F lineage in both kinds of tissues (Ladoukakis et al., 2002;OEmietanka et al., 2004).Therefore, the F lineage is only transmitted through females, while the M lineage is only transmitted through the males (Fisher & Skibinski, 1990).
It is believed that the doubly uniparental mode of inheritance arose via the production of a new factor, which acts by interrupting the recognition of the male gamete factor localized on the external surface of the mitochondria.It is believed that this new factor may also be involved in sex determination in the Mytilidae (Zouros, 2000).When this recognition fails in the Mytilidae, the lack of recombination between lineages should lead to their divergence with time.Studies have demonstrated that the M lineages in the Mytilidae evolve at higher rates than the maternal ones (Skibinski et al., 1994b;Rawson & Hilbish, 1995;Hoeh et al., 1996;Stewart et al., 1996;Quesada et al., 1998;Skibinski et al., 1999).This fact cannot be explained by the hypothesis of evolution driven by males as in mammals (Shimmen et al., 1993) because the number of cell divisions during gametogenesis in the Mytilidae is very similar in both sexes (Rawson & Hilbish, 1995).It is believed that selection is more relaxed in the male mitochondrial genome than in the female because the male mitochondrial genome is less functionally constricted due to its expression only in male gonads and not in somatic tissues (Skibinski et al., 1994b;Rawson & Hilbish, 1995;Hoeh et al., 1996;Stewart et al., 1996).Rawson & Hilbish (1995) first found evidence of the doubly uniparental mode of inheritance of the mitochondrial genome in Mytilus edulis, M. galloprovincialis, and M. trossulus Gould, 1850 by studying 527 bp of the 16S rRNA gene using restriction enzymes.Nothing is known about the mode of inheritance of the mitochondrial genome in Perna perna (Linnaeus, 1758); therefore, our aim was to investigate this process using the same mitochondrial genome region of the 16S rDNA used for studying the Mytilus group.

Sampling and DNA extraction
Forty adult individuals (20 females and 20 males) of P. perna were sampled from the locality of Penha, Santa Catarina, South Brazil.The sex was determined for all individuals, and samples from the gonads and somatic adductor muscle tissues were taken.DNA extraction from these tissues was performed according to a Proteinase-k/Phenol/Chloroform protocol (Hoelzel, 1998).

RFLP pattern of the 16S rDNA region
The 16S rDNA region was amplified by the Polymerase Chain Reaction (PCR) using the primers described by Palumbi et al. (1991): AR 5'-CGCCTGTTTATCAAAAACAT-3' and BR 5'-CCGGTCTGAACTCAGATCACGT-3'.The 25 µL PCR reaction included PCR buffer (1x), Triton-X-100 (0.1%), dNTPs (0.4 µM), MgCl 2 (1 mM), primers (0.4 µM each), 1 U of Taq DNA polymerase, and 4-7.5 µL of extracted DNA.The reaction was cycled according to the following program: 1 cycle at 94ºC for 3 min; 30 cycles at 94ºC for 20 s, 51ºC for 20 s, and 72ºC for 45 s; and a final step at 72ºC for 3 min.PCR products were cleaved with the following restriction enzymes: EcoR V, Hae III, Spe I, Hind III, EcoR I, Taq I, Rsa I, and Alu I. Cleavage reactions were performed in a final volume of 20 µl, containing 5 µL of PCR product, 2 µL of reaction buffer (10x), and 5 U (0.5 U in the case of Spe I) of the restriction enzyme.Reactions were incubated overnight (after ruling out star activity) at 37ºC and at 65ºC for Taq I. First, cleavage was verified for all individuals in 1.5% agarose gel electrophoresis.Then, the fragments obtained after digestion with single enzymes and different combinations of enzymes were separated by denaturing 6% polyacrylamide vertical gel electrophoresis, following the Brown (1994) protocol: 33.6 g urea (6M), 8 mL TEB 10x (Tris-Borate-EDTA buffer), 12 mL 40% Polyacrylamide (19:1), 800 µL APS (ammonium persulfate, 0.1 g/1 mL), and 60 µL Temed.A pre-run step was done at 75 Watts until the gel reached 70°C.Samples were denatured for 5 min at 95°C in 1:1 denaturing loading buffer (95% deionized formamide; 0.05% Bromophenol Blue; 0.05% Xylene Cyanol FF; and 20 mM EDTA, pH 8.0).Then, the samples were placed on ice and loaded into the wells of the gel.The electrophoresis was maintained at 50ºC for 1.5 h at 45 Watts.DNA fragments were visualized by silver staining following the protocol of Hoelzel (1998).The restriction map was confirmed with sequence analysis, therefore some fragments were sent to Macrogen Inc, Korea for sequencing by the DNA analyzer ABI 3700.
Perna perna sequences were compared with twenty-three representative sequences from fifteen mytilid species available in GenBank, and these sequences were used to construct a tree.Alignment of sequences was done with ClustalW (1.82) software.The evolutionary tree was constructed by neighbor-joining algorithms using Saitou & Nei's (1987) genetic distance, and bootstrap analysis was done over 500 replications.MEGA 3.1 (Kumar et al. 2004) was used for this analysis.

RESULTS
The 16S rDNA region in Perna perna was found to be 517 bp in length, a measurement that was confirmed with sequence analysis (GeneBank Access Numbers: DQ923878, DQ923879, DQ923880, DQ923881, and DQ923882).The 16S rDNA fragment was cleaved by only five of the restriction enzymes used (Alu I, EcoR I, Hae III, Hind III, and Taq I).No variation was observed among individuals or between sexes or kinds of tissues (somatic and gonadal).Therefore, it was not possible to recognize different gender lineages of the mitochondrial genome in the examined 517 bp of the 16S rDNA (Figure 1).The lack of variation was confirmed by the sequence analysis.Alu I cleaved the 517 bp product into four fragments: 252 bp, 108 bp, 99 bp, and 58 bp; EcoR I cleaved it into two fragments: 166 bp and 351 bp; Hae III also cleaved it into two fragments: 47 bp and 470 bp; Hind III cleaved it into fragments of 250 bp and 267 bp; and Taq I cleaved it into fragments of 424 bp and 93 bp.The constructed restriction map of the 517 bp 16S rDNA region in P. perna showed a total of seven restriction sites, recognized by five different enzymes (Figure 2).
The 16S rDNA sequences of other Mytilidae available in GenBank allowed comparisons to only 439 bp of our sequence.The constructed tree (Figure 3) showed Perna perna from the south of Brazil clustered together with a sample of the same species from the USA.All Perna species included in the analysis were clustered together in a single branch with high certainty (100% bootstrap value) and were shown to be more closely related to the Mytilus group than to the other representative genera of the Mytilidae.

DISCUSSION
The universal primers AR and BR amplified a fragment of 517 bp corresponding to part of the 16S rRNA gene in Perna perna; this fragment is 10 bp shorter than the region obtained with the same primers for the Mytilus group (Rawson & Hilbish, 1995).While analyzing this region using the restriction enzymes EcoR V and Spe I, Rawson & Hilbish (1995) found variation within species in the Mytilus group, allowing the authors to   detect the doubly uniparental inheritance in these species.Surprisingly, we were unable to detect any variation between genders and kinds of tissues; therefore, we were unable to reject the hypothesis of single maternal inheritance.This result suggests that the 16S rDNA region analyzed is highly conserved in the population of Perna perna from the south of Brazil.
Studying the bivalve Tapes philippinarum (Adans & Reeve, 1850) from the Veneridae, Passamonti & Scali (2001) found a high divergence (75 to 78 nucleotide substitutions) between gender mitochondrial lineages at the same region studied for Perna perna.The finding that DUI is present in divergent groups of bivalve leads to the hypothesis that doubly uniparental inheritance emerged early in the bivalve and should represent an ancestral character in the group.Occasionally, the paternal route is invaded by the maternal lineage (Hoeh et al., 1997;Saavedra et al., 1997), producing a heterogeneous pool transmitted by males, so that recombination, like gene conversion, could eventually happen (Ladoukakis & Zouros, 2001), as has been shown in M. galloprovincialis.Another event that may happen is the complete replacement of the M lineage by the F, so that the M becomes the same as the F, an event called the "masculinization of the F".The masculinization of the F leads to the resetting of differences between both lineages to zero (Hoeh et al., 1997).Differences between the F-and M-type have already been estimated in more than 20% for the Mytilus group (Skibinski et al., 1994b).
Could Perna perna have suffered one of the events that reset differences between lineages?Is it a species that recently invaded South America with a few founders lacking variation at this region, and their populations have not had enough time to evolve differences between gender lineages at the highly conserved 16S rDNA region?Or is the genus Perna just an exception to the doubly uniparental inheritance found in the Mytilidae?
Perna perna is the most common mussel of the south and south-east of Brazil.Although it is believed that P. perna was introduced into South America by ships coming from South Africa (Holland, 2001;Souza et al., 2003Souza et al., , 2004)), this species is well-established on rocky shores, where it is predominant.No P. perna fossils were found in Brazilian prehistoric strata such as Sambaquis, where the native peral-oyster Pinctada imbricata Röding, 1798 is abundant (Souza et al., 2003).This oyster at present is almost excluded from rocky shores by P. perna (Souza et al., 2004;Weber & Silva, 2008).Therefore, the lack of variation at this mitochondrial region might be related to a recent invasion.
Nonetheless, we cannot rule out the possibility that those primers were unable to amplify the male lineage or when isolating the male gonads, somatic tissue passed together.More extensive studies of more variable mitochondrial regions, as well as to use highly purified spermatic cells, will be necessary to clarify this point for Perna perna.Including samples from the full geographic range of the species may also be helpful to clarify this point.Non-coding regions in the mitochondrial genome also may be used, which have already been described for the Mytilidae (Hoffman et al., 1992;Cao et al., 2004;Mizi et al., 2005).

Figure 2 -
Figure 2 -Restriction map of the 517 bp 16S rDNA region from Perna perna.

Figure 3 -
Figure 3 -Neighbor-joining tree for fifteen mytilid species, constructed from twenty-three representative sequences of the 16S rDNA region obtained from GenBank.Values at branches represent the bootstrap confidence level.