Slipins: ancient origin, duplication and diversification of the stomatin protein family

Abstract

Background: Stomatin is a membrane protein that was first isolated from human red blood cells. Since then, a number of stomatin-like proteins have been identified in all three domains of life. The conservation among these proteins is remarkable, with bacterial and human homologs sharing 50 % identity. Despite being associated with a variety of diseases such as cancer, kidney failure and anaemia, precise functions of these proteins remain unclear.

Results: We have constructed a comprehensive phylogeny of all 'stomatin-like' sequences that share a 150 amino acid domain. We show these proteins comprise an ancient family that arose early in prokaryotic evolution, and we propose a new nomenclature that reflects their phylogeny, based on the name "slipin" (stomatin-like protein). Within prokaryotes there are two distinct subfamilies that account for the two different origins of the eight eukaryotic stomatin subfamilies, one of which gave rise to eukaryotic SLP-2, renamed here "paraslipin". This was apparently acquired through the mitochondrial endosymbiosis and is widely distributed amongst the major kingdoms. The other prokaryotic subfamily gave rise to the ancestor of the remaining seven eukaryotic subfamilies. The highly diverged "alloslipin" subfamily is represented only by fungal, viral and ciliate sequences. The remaining six subfamilies, collectively termed "slipins", are confined to metazoa. Protostome stomatin, as well as a newly reported arthropod subfamily slipin-4, are restricted to invertebrate groups, whilst slipin-1 (previously SLP-1) is present in nematodes and higher metazoa. In vertebrates, the stomatin family expanded considerably, with at least two duplication events giving rise to podocin and slipin-3 subfamilies (previously SLP-3), with the retained ancestral sequence giving rise to vertebrate stomatin.

Conclusion: Stomatin-like proteins have their origin in an ancient duplication event that occurred early on in the evolution of prokaryotes. By constructing a phylogeny of this family, we have identified and named a number of orthologous groups: these can now be used to infer function of stomatin subfamilies in a meaningful way.

Figure 1

Maximum likelihood (ML) phylogeny of stomatin family members. Neighbour Joining (NJ) and ML trees were constructed with the same 149 amino acid alignment. 100 ML and 1000 NJ bootstraps were performed. Percentage bootstrap values are shown for the major groups that were recovered in either the ML or NJ (in parentheses) phylogeny. The virus sequence comes from the Acanthamoeba polyphaga mimivirus. The scale bar indicates the number of amino acid substitutions per site. The length of each triangle corresponds to longest branch length, whilst the height represents the number of taxa. Names in square brackets represent previous nomenclature. The full tree is available as additional file 1: Figure A1. Note the partitioning of the family into at least four major groups termed here paraslipins, eoslipins, slipins and alloslipins, with a long internal branch between paraslipins and the other subfamilies where we have tentatively located the root. Accession numbers available in Additional file 2.

Figure 2

Stomatin subfamilies. This diagram illustrates the various subfamilies of stomatin-like proteins recovered by phylogenetic analysis in Figures 1-5, and our proposed nomenclature. Names in [] represent previous nomenclature if a new term is being proposed, whilst * indicates a newly reported subfamily.

Figure 3

Maximum likelihood tree of paraslipin proteins. The phylogeny was based on a 272 amino acid alignment. Numbers on the branches show percentage bootstrap occurrence of nodes in 100 replicates. Only values > 70 are shown. The scale bar indicates the number of amino acid substitutions per site. indicates the clade uniting Chlorobi, Gammaproteobacteria and Spirochetes with two parallel phylogenies. Note the position of Rickettsiales as the sister group to the eukaryotic clade. Accession numbers available in Additional file 2.

Figure 4

Maximum likelihood tree of protostome slipins. Phylogenies were based on a 239 amino acid alignment. The scale bar indicates the number of amino acid substitutions per site. 100 bootstraps were performed and values ≥ 50 are shown on each branch. Cnidarian sequences were included to root the phylogeny whilst the sea urchin Ciona intestinalis sequences were included to allow comparison with Fig. 5. Accession numbers available in Additional file 2.

Figure 5

Maximum likelihood phylogenies of metazoan slipins. The full tree is based on a 227 amino acid alignment of chordate, echinoderm and protostome sequences. 100 bootstraps were performed and values ≥ 50 are shown in black on each branch. A separate phylogeny, shown by the green box, was constructed from a longer, 252 amino acid alignment of only chordate and echinoderm sequences. 100 bootstraps were performed and values ≥ 50 are shown in red on each branch. Accession numbers available in Additional file 2.

Figure 6

Alignment of stomatin subfamily consensus sequences. Consensus sequences (100-70 %) were generated by selecting and aligning (ClustalX) subfamily members identified in monophyletic groups in Figures 1-4. The shading threshold was set to 0.6. Identical residues have a black background and similar residues have a grey background. A 60 % slipin (pink) and paraslipin (green) consensus sequence is also displayed. p2_para refers to prokaryotic paraslipins present in the upper clade in Figure 3, whilst p1_para refers to prokaryotic paraslipins present in the lower clade with eukaryotic paraslipins (euPara). The dotted line above the sequences shows the region shared by all stomatin family members, whilst the solid red line indicates regions shared by slipins and eoslipins to the exclusion of alloslipins. n = the number of sequences used to generate each consensus. – indicates the position of a gap, indicates an unconserved amino acid.

Figure 7

Simplified hypothesis for the origin, duplication and divergence of vertebrate stomatin subfamilies. An ancestral stomatin gene, possibly present in the last universal common ancestor, duplicated to give rise to eoslipin and prokaryotic paraslipin. Prokaryotic paraslipin was transferred into eukaryotes during the acquisition of the mitochondrion. Eukaryotic slipins probably evolved from eoslipin, which we assume was present in the last common ancestor of all eukaryotes. Within metazoa, slipin-1 arose from a gene duplication (GD) event involving a stomatin-like gene which subsequently fused (GF) with a sterol carrier domain (Edqvist and Blomqvist 2006). Podocin and slipin-3 arose from two further duplications of an ancestral stomatin-like gene that might have occurred during the two whole genome duplications in early vertebrate evolution.