On the origins of arrestin and rhodopsin

Abstract

Background: G protein coupled receptors (GPCRs) are the most numerous proteins in mammalian genomes, and the most common targets of clinical drugs. However, their evolution remains enigmatic. GPCRs are intimately associated with trimeric G proteins, G protein receptor kinases, and arrestins. We conducted phylogenetic studies to reconstruct the history of arrestins. Those findings, in turn, led us to investigate the origin of the photosensory GPCR rhodopsin.

Results: We found that the arrestin clan is comprised of the Spo0M protein family in archaea and bacteria, and the arrestin and Vps26 families in eukaryotes. The previously known animal arrestins are members of the visual/beta subfamily, which branched from the founding "alpha" arrestins relatively recently. Curiously, we identified both the oldest visual/beta arrestin and opsin genes in Cnidaria (but not in sponges). The arrestin clan has 14 human members: 6 alphas, 4 visual/betas, and 4 Vps26 genes. Others recently showed that the 3D structure of mammalian Vps26 and the biochemical function of the yeast alpha arrestin PalF are similar to those of beta arrestins. We note that only alpha arrestins have PY motifs (known to bind WW domains) in their C-terminal tails, and only visual/betas have helix I in the Arrestin N domain.

Conclusion: We identified ciliary opsins in Cnidaria and propose this subfamily is ancestral to all previously known animal opsins. That finding is consistent with Darwin's theory that eyes evolved once, and lends some support to Parker's hypothesis that vision triggered the Cambrian explosion of life forms. Our arrestin findings have implications on the evolution of GPCR signaling, and on the biological roles of human alpha arrestins.

Figure 1

Arrestin family tree. Annotated arrestin proteins from select sequenced genomes were used to construct a neighbor-joining phylogenetic tree. The visual/beta arrestin proteins branch from the remainder of the tree with a bootstrap confidence score of 100 (arrow). The scale bar shows the number of substitutions per site. Taxonomy abbreviations follow: str_protist, stramenopiles 9STRA (Protista, Stramenopiles); par_protist, Paramecium tetraurelia (Protista, Alveolata); sl_mold, slime mold Dictyostelium discoideum (Protista, Mycetozoa); fis_yeast, fission yeast Schizosaccharomyces pombe; asc_fungus, Emericella nidulans (Fungi/Ascomycota); hydra, Hydra magnipapillata (Cnidaria); worm, nematode C. elegans (Nematoda); fly, Drosophila m. (Arthropoda); protochord, Ciona intestinalis (Urochordata); fish, Danio rerio (zebrafish; Vertebrata). Vertebrate visual/beta arrestins are given in HUGO nomenclature [see Additional file 1]: S-antigen, SAG (aka, rod arrestin, arrestin 1); arrestin 3, ARR3 (aka, cone arrestin, X-arrestin, arrestin 4); arrestin, beta 1, ARRB1 (aka, arrestin 2); and arrestin, beta 2, ARRB2 (aka, arrestin 3).

Figure 2

Arrestin protein family: multiple sequence alignment of phylogenetically diverse members. The alpha and beta classes of arrestins are distinct. Positions that may be widely conserved by common descent are yellow, beta/visual-specific are red and alpha-specific are blue (ambiguous in gray). Black positions in the Tail are the PPXY (or (P/L)PXY or "PY") motifs. PPXY motifs can have alternative residues in the first position. Notably, the sea urchin alpha arrestin has two PPXY and seven QPXY motifs. The sea urchin and nematode alphas also share the PY-like sequence (Y/F)APXYP(Y/F)Y. The arrestin domains are given below the alignment, n for N domain, c for C and t for Tail; italics show sequence not considered as part of the N and C domains according to Pfam. Shading on that line maps secondary structure elements on cone arrestin (symbol ARR3, HUGO nomenclature; aka arrestin 4, X-arrestin) – beta strands in gray and the one alpha helix in black [78]. Underlining highlights regions involved in receptor specificity, as described in Ref. [78] and references within. Two sets of intra-molecular interactions are important for keeping visual/beta arrestins in their basal conformation (see text): number symbols (#) mark residues that make up the "polar core"; asterisks (*) show residues involved in the "three-element interaction". Identifiers follow: fly a, D. melanogaster alpha arrestin CG18745-PA; nem worm a, nematode C. elegans alpha T12D8.4; nem worm b, beta F53H8.2; sea urchin, Strongylocentrotus purpuratus alpha XP_001175756; others are listed by gene name [see Additional file 2].

Figure 3

Bacterial Spo0M and Eukaryotic Vps26 proteins are members of the arrestin clan. (A) Sequence alignment of B. subtilis Spo0M and Arrestin N domain consensus (Pfam HMMER, statistical significance score E = 9.9e-5 [see Additional file 2]). (B) Alignment of Dictyostelium discoideum PepA (Vps26) and Arrestin N domain (E = 0.02). The Pfam motif pattern includes weakly (lower case) and highly conserved positions (bold upper case). Conserved sequence is yellow.

Figure 4

Novel Hydra opsin aligned with rag-worm ciliary opsin. As expected, the sequence conservation is highest in the seven transmembrane regions (underlined). The signature residue perfectly conserved in all opsins is the lysine that retinal forms a Schiff base with. This conserved position is shown in red. The position of the well-characterized glutamate counterion of some ciliary opsins, is instead a tyrosine in several ciliary opsins and all rhabdomeric opsins. Hydra ciliary opsin has a serine at that position. The cytoplasmic regions shown by mutagenesis to be critical for G protein interaction are underlined by gray boxes. The first of those cytoplasmic regions contains the highly conserved DRY element, which contains an arginine critical for G protein activation. That sequence is more similar in Hydra (YRY) than in rag-worm (VRC), shown by gray highlighting.