Elongation Factor 1-Alpha Sequences Alone Do Not Assist in Resolving the Position of the Acoela Within the Metazoa

Abstract

Following years of debate, and thanks largely to molecular systematic studies, the long-held idea that the phylum Platyhelminthes is the most basal branch of the bilaterian Metazoa has been widely abandoned (e.g., Haszprunar 1996 ; Carranza, Baguñà, and Riutort 1997 ; Hausdorf 2000 ). This new status quo has been upset by a recent paper that has suggested that the Platyhelminthes are polyphyletic and that a single “flatworm” clade, the Acoela, is the most basal extant bilaterian lineage distinct from the other platyhelminths (Ruiz-Trillo et al. 1999 ). The potential importance of the acoels being confirmed as the outgroup to all other bilaterians is that they would then be instrumental in determining the character states of the ancestral Bilateria through outgroup comparison and thus greatly further our understanding of the evolution of the animals. The currently recognized outgroup—the diploblasts—are generally too evolutionarily and phenotypically distant for this sort of comparison to be informative. Although recent data support the molecular results (Henry, Martindale, and Boyer 2000 ), some systematists distrust the findings based on 18S rDNA and have claimed that long-branch attraction (LBA; Felsenstein 1978 ) affected the final analysis of these data (e.g., Adoutte et al. 2000 ). Furthermore, according to the same 18S rDNA study (Ruiz-Trillo et al. 1999 ), the supposed sister group of the acoels, the nemertodermatids (Smith and Tyler 1985 ; Smith, Tyler, and Rieger 1986 ; Tyler 2001), do not unite with the acoels at the base of the Bilateria, but instead fall within the turbellarian flatworms. This has suggested that the more controversial position of their sister group, the acoels, was erroneous. Recently, it has been proposed that the published putative nemertodermatid sequence used by Ruiz-Trillo et al. (1999) (Nemertinoides elongatus, accession number U70083; Carranza, Baguñà , and Riutort 1997) that grouped among the Rhabditophora may have been misidentified (Jondelius et al. 2000 ). Berney, Pawloski, and Zaninetti (2000) directly contradict the basal position of the Acoela based on analyses of elongation factor 1-alpha (EF1a). This gene was chosen with the aim of avoiding LBA. Comparing EF1a sequences derived from a single acoel species, Convoluta roscoffensis, with those of other animals and fungi, they forward two lines of evidence suggesting that acoels are not basal metazoans, but rather that they are flatworms, as historically regarded. Their evidence comes not only from a standard phylogenetic comparison of the amino acid and nucleotide sequences of EF1a that places the acoels within the turbellarian flatworms (allied either to the triclads or to the polyclads, depending on the analysis), but also from the presence of a putative insertion of four amino acids in common between the acoels and the triclads that are not found in other EF1a protein sequences compared. Importantly, they must, according to the common possession of the four-amino-acid signature, be derived flatworms allied among the rhabditophorans to the Tricladida. It is in part the derived position of the acoels within, rather than basal to, the Rhabditophora (the more commonly espoused position according to morphological analyses, e.g., Ehlers 1984 ; Littlewood, Rohde, and Clough 1999 ) based on EF1a that has prompted us to examine these claims more closely. In addition to contradicting the results based on 18S rDNA (Ruiz-Trillo et al. 1999 ), the conclusions of Berney, Pawloski, and Zaninetti (2000) are in conflict with a study (Telford et al. 2000 ) that has lent support to the idea of a monophyletic Rhabditophora (Ehlers 1984 ). This analysis of the mitochondrial genetic codes of the flatworms shows that all rhabditophoran flatworms have two differences in their genetic code compared with most other invertebrates: the codons AAA, coding for asparagine as opposed to lysine, and AUA, coding for isoleucine as opposed to methionine (Telford et al. 2000 ). Acoela, Nemertodermatida, and Catenulida all shared the plesiomorphic condition and were hence excluded from the monophyletic Rhabditophora. Two such convincing synapomorphies are in direct conflict with the interpretation of the EF1a data of Berney, Pawloski, and Zaninetti (2000) ; clearly, the acoels cannot be both excluded from the Rhabditophora and a sister group of the rhabditophoran triclads or polyclads. Consideration of the morphology casts further doubt on this derived rhabditophoran position of the acoels, as they lack all rhabditophoran or triclad morphological synapomorphies (Tyler 2001 ). Consequently, we looked carefully at the analyses of Berney, Pawloski, and Zaninetti (2000) to assess their robustness. We provided new EF1a sequences from three additional species of acoels and seven platyhelminths and combined these with 21 previously published flatworm Ef1a sequences to reassess the phylogenetic content of the gene and the homology of the amino acid insertion