Vertebrate extracellular calcium-sensing receptor evolution: Selection in relation to life history and habitat

Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

Contents lists available at SciVerse ScienceDirect

Comparative Biochemistry and Physiology, Part D journal homepage: www.elsevier.com/locate/cbpd

Vertebrate extracellular calcium-sensing receptor evolution: Selection in relation to life history and habitat Amanda L. Herberger ⁎, Christopher A. Loretz Department of Biological Sciences, University at Buffalo, Buffalo, NY 14260-1300, USA

a r t i c l e

i n f o

Article history: Received 26 October 2012 Received in revised form 7 December 2012 Accepted 9 December 2012 Available online 20 December 2012 Keywords: Extracellular calcium-sensing receptor Evolutionary selection Vertebrate skeleton Bone

a b s t r a c t Ionic calcium (Ca2+) supports essential functions within physiological systems, and consequently its concentration is homeostatically regulated within narrow bounds in the body fluids of animals through endocrine effects at ion-transporting osmoregulatory tissues. In vertebrates, extracellular Ca2+ is detected at the cell surface by the extracellular calcium-sensing receptor (CaSR), a member of the G protein-coupled receptor (GPCR) superfamily. Interestingly, the taxonomic distribution of CaSRs is restricted to vertebrates, with some CaSR-like receptors apparently present in non-vertebrate chordates. Since bone is a known Ca2+ storage site and is characteristically restricted to the vertebrate lineage, we hypothesized a functional association of CaSR with vertebrate skeleton that may have an ancient origin. Protein sequence alignment and phylogenetic analysis of vertebrate CaSRs and related GPCRs of the glutamate receptor-like family expose similarities and indel differences among these receptors, and reveal the evolutionary history of CaSRs. Evolutionary selection was tested statistically by evaluating the relationship between non-synonymous (replacement, dN) versus synonymous (silent, dS) amino acid substitution rates (as dN/dS) of protein-coding DNA sequences among branches of the estimated protein phylogeny. On a background of strong purifying selection (dN/dSb 1) in the CaSR phylogeny, statistical evidence for adaptive evolution (dN/dS >1) was detected on some branches to major clades in the CaSR phylogeny, especially to the tetrapod vertebrate CaSRs and chordate CaSR-like branches. Testing also revealed overall purifying selection at the codon level. At some sites relaxation from strong purifying selection was seen, but evidence for adaptive evolution was not detected for individual sites. The results suggest purifying selection of CaSRs, and of adaptive evolution among some major vertebrate clades, reflecting clade specific differences in natural history and organismal biology, including skeletal involvement in calcium homeostasis. © 2012 Elsevier Inc. All rights reserved.

1. Introduction Since the initial description of the extracellular calcium-sensing receptor (CaSR) from bovine parathyroid gland nearly twenty years ago (Brown et al., 1993), the understanding and sense of CaSR as a physiological detector for ionic calcium (Ca2+) have dramatically expanded. The receptor was first seen as the sensor of extracellular Ca2+ for endocrine and other cells involved in Ca2+ homeostasis. So, in addition to receptor expression in endocrine cells that secrete calcium-homeostatic hormones (parathyroid hormone-secreting parathyroid cells and calcitonin-secreting thyroidal C-cells in mammals, and other endocrine cell types in fishes, for example), CaSR is expressed in kidney and Abbreviations: CaSR, calcium-sensing receptor; CRD, cysteine-rich domain; ECD, extracellular domain; GABABR, γ-aminobutyric acid type B receptor; GPCR, G protein-coupled receptor; ICD, intracellular domain; LRT, likelihood ratio test; MAP, mitogen-activated protein (kinase); mGluR, metabotropic glutamate receptor; OdorR, odorant receptor; PherR, pheromone receptor; T1R, taste type 1 receptor; V2R, vomeronasal type 2 receptor; VFT, Venus flytrap; 7TM, seven-transmembrane. ⁎ Corresponding author at: Department of Biological Sciences, 109 Cooke Hall, University at Buffalo, Buffalo, NY 14260-1300, USA. Tel.: +1 716 645 4963; fax: +1 716 645 2975. E-mail address: [email protected] (A.L. Herberger). 1744-117X/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cbd.2012.12.004

intestine where it may directly influence Ca2+ absorption and/or secretion, and in bone where it may affect Ca2+ storage and mobilization (Brown et al., 1993; Brown and MacLeod, 2001; Chang and Shoback, 2004; Loretz, 2008; Loretz et al., 2009; Brown, 2010; Loretz et al., 2012). Broader ligand sensitivity and CaSR expression in a variety of other tissues have since been reported, suggesting roles for the receptor outside the traditionally-recognized calcium homeostatic domain, too. The CaSR is a G protein-coupled receptor (GPCR). The GPCR superfamily of proteins is large, and the 7TM domain molecular signature of the superfamily is found in all eukaryotic species (Strotmann et al., 2011). The superfamily can be divided into five major families according to phylogenetic and structural and ligand-binding properties demonstrated by the GRAFS scheme: G (glutamate receptor-like), R (rhodopsin receptor-like), A (adhesion receptor-like), F (frizzled/taste2 receptor-like), and S (secretin receptor-like) (Strotmann et al., 2011). Current thinking pegs the evolutionary emergence of GPCRs and G protein-coupled transductional signaling in eukaryotes to about 1.2 billion years ago (Römpler et al., 2007; Strotmann et al., 2011). The CaSR is included in the glutamate receptor-like family of GPCR receptors (also widely referred to as GPCR family C or family 3 in some alternate GPCR classification schemes) together with pheromone receptors

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

(PherRs) and vomeronasal type 2 receptors (V2Rs), taste type 1 receptors (T1Rs) and some amino acid and odorant (OdorR) and orphan receptors, metabotropic glutamate receptors (mGluRs), and γ-aminobutyric acid type B receptors (GABABRs) (Pin et al., 2003; Loretz, 2008). This family is characterized, and distinguished from other GPCRs, by having a large extracellular domain (ECD) that functions in ligand detection. The ECD comprises a Venus flytrap (VFT) module that includes the Ca2+ binding sites, and a cysteine-rich domain (CRD) that couples the VFT module to the heptahelical seven-transmembrane (7TM) domain (Pin et al., 2003). The CaSRs possess these structural features, and, additionally, have a substantial C-terminal intracellular domain (ICD). Although the length of the ICD tail varies among vertebrate classes, it is, nevertheless, critically important in transductional signaling by the mammalian and piscine receptors in which it has been experimentally tested (Chang and Shoback, 2004; Loretz et al., 2004; Loretz, 2008). However, there is another deeper and more complicated evolutionary story for GPCRs, as evidenced by the modular nature of the receptors. Relevant to our interest in CaSRs, this modularity is clearly seen in the case of glutamate receptor-like, where the VFT (Pfam designation: ANF_receptor, for atrial natriuretic factor receptor; Finn et al., 2006) of the ECD is a structural motif that is shared with natriuretic peptide receptors and bacterial periplasmic binding proteins (Felder et al., 1999; Loretz, 2008). Similarly, the CRD (Pfam: NCD3G, for Nine Cysteines Domain of family 3 GPCRs) is represented in all glutamate receptor-like GPCRs (with the exception of GABABRs, in which the module is conspicuously lacking) where it links the VFT module with the core 7TM domain (Pfam: 7tm_3) that is characteristic of GPCRs (Liu et al., 2004). Given the ancient nature of the superfamily and the apparent early emergence of the receptor families, it is not surprising that the evolutionary history of the GPCRs is incompletely resolved (Römpler et al., 2007; Strotmann et al., 2011). Among the many GPCRs, the extracellular calcium-sensing receptor stands out based on its rather special role in recognizing primarily and importantly a cation ligand, relative to the detecting capabilities of other GPCRs for molecular ligands. Our searching of publicly-accessible genomes for Casr gene and Casr-like gene sequences points to a relatively recent emergence of CaSRs that appears restricted in distribution to the chordate–vertebrate lineage (Loretz, 2008). In this report, we present the results of our molecular phylogenetic analysis of CaSRs and CaSR-like molecules, including tests for evidence of selection, and we interpret our findings in the contexts of receptor protein structure and vertebrate calcium homeostasis, and of mineralized bone skeleton as a key characteristic of vertebrates.

2. Materials and methods 2.1. Collection and assembly of CaSR nucleotide and protein sequences Nucleotide sequences of CaSR, and CaSR-like receptors were retrieved by BLAST search from GenBank, from annotated Ensembl genome resources, or by BLAST search of unannotated genomes at Ensembl or Pre Ensembl or other publically-available genome resources (Ensembl release 69 and Pre Ensembl release 66 at the time of submission; The Neanderthal Genome). Similarly, nucleotide and inferred protein sequences from representative PherRs, V2Rs, T1Rs, and other related glutamate receptor-like GPCRs were gathered. Manual editing of some annotated genome-derived sequences was applied for proper exon assembly and nucleotide alignment. In total, 138 sequences were collected: 42 sequences comprising the CaSR subgroup, where sequence data from all vertebrate classes are represented; and 8 sequences of non-vertebrate chordates representing a CaSR-like group, some of which resemble PherRs (in the case of Petromyzon marinus variation 2 sequence) or mGluRs (in the cases of Caenorhabditis elegans, Strongylocentrotus purpuratus, and Drosophila melanogaster variant A and B sequences) based on subsequent phylogenetic analysis. The

87

sequence data set is tabulated in Supplementary File 1 together with corresponding accession numbers and genome assembly versions. 2.2. Phylogenetic analysis Receptor nucleotide sequences were preliminarily aligned using the Clustal algorithm implemented in MEGA5 (version 5.04) in order to confirm proper exon identification and assembly (Tamura et al., 2011). Then the inferred protein sequences were subjected to alignment using Clustal, followed by a final round of manual editing where needed (see Supplementary File 2 for complete data set sequence alignment). Finally, based on initial phylogenetic analysis, a data subset comprising only CaSRs, CaSR-like, and PherRs and V2Rs was isolated for selection tests (see Section 2.3 for details). CaSR protein phylogenetic tree estimates were constructed by Bayesian analysis (MrBayes 3.1.2; Ronquist and Huelsenbeck, 2003) and by neighbor-joining method (MEGA5; Tamura et al., 2011). Phylogenetic trees were visualized and displayed using the Tree Explorer module of MEGA5 or Dendroscope 2.7.4 (Huson et al., 2007). The un-rooted, fully-expanded phylogenetic tree estimate is presented in Supplementary File 3. 2.3. Maximum likelihood tests for evolutionary selection Testing for selection was conducted using the ‘codeml’ module of the Phylogenetic Analysis by Maximum Likelihood (PAML, version 4) software package (Yang, 2007) and using Selecton (version 2.4) via its web server (Stern et al., 2007). Selection testing was applied to several subsets of the full data set as specified below. Using PAML, selection was assessed using a calculated omega (ω) value that is a ratio of non-synonymous codon changes (dN, substitutions resulting in amino acid replacement) to synonymous codon changes (dS, nucleotide substitutions that do not alter the amino acid), where ω = dN/dS. By this analysis, ω values greater than 1 are evidence for adaptive evolution, while values less than 1 are evidence for purifying evolution (and where ω = 1 reflects neutral evolution, or the absence of selection pressure). For PAML branch analyses, data sets included: (1) the full-length aligned codon sequences of CaSR, CaSR-like, and PherR/V2R molecules; (2) a refined data set that included only the codon sequences representing the VFT module for the same molecules; (3) and a second refined (“trimmed”) data set where indels and gaps among the full-length aligned sequences were removed (see Supplementary File 4 for trimmed sequence alignment). PAML site analysis tested a smaller data set of vertebrate CaSR sequences only. For branch analysis, comparisons were made between the “one-ratio” model (Model 0, M0, a single ω value for all tree branches) and the “free ratio” model (Model 1, M1, independent ω values for each branch of the tree). The better-supported models of these test pairs were chosen based on the likelihood ratio test (LRT), as interpreted by the chi-square statistic. For site analysis, PAML NSsite models M7 (beta) and M8 (beta and ω) were employed, with statistical analysis by LRT to distinguish purifying versus adaptive selection. Selecton was applied as a test for selection at the single amino acid scale (Stern et al., 2007). The test statistic for Selecton was Ka/Ks (the calculated rate of non-synonymous codon changes relative to the rate of synonymous changes). For Selecton analysis, a smaller data set comprising only CaSR and CaSR-like sequences was used. 2.4. Amino acid conservation scoring WebLogo (version 3.0; Crooks et al., 2004) was used to visualize amino acid conservation in the five predicted Ca 2+ binding sites of CaSRs according to Huang et al. (2009). Sequence conservation by position was assessed by bit score calculation for each amino acid residue position in the five binding sites, with relative frequency of amino acid usage displayed in pictorial fashion.

88

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

2.5. Feature mapping to receptor structure

3. Results

Distinguishing receptor features, relative to phylogenetic clade and evident selection, were viewed in the context of threedimensional receptor protein structure. Since there is no available crystal structure for the CaSR, for visualization purposes, 3-dimensional homology models of CaSRs were built by comparative protein homology modeling using SWISS-MODEL Workspace (Arnold et al., 2006; Loretz, 2008). The template for ECD modeling was the rat metabotropic glutamate receptor 3 (mGluR3, complexed with the agonist (2S,2′,3′)-2-(2′ 3′ dicarboxycyclopropyl) glycine; PDB: 2E4V). The modeled molecular structures were visualized using DeepView-Swiss-Pdb Viewer (version 4.1; Guex and Peitsch, 1997).

3.1. Protein molecular phylogeny analysis By protein molecular phylogeny estimation using Bayesian analysis, CaSRs are one subfamily of glutamate receptor-like GPCRs (Fig. 1). The close relationship of CaSRs to pheromone, odorant and taste receptors is clearly evident. The tree has overall strong statistical support. Phylogeny estimation by neighbor-joining method generated a tree that shared topology with the Bayesian tree, and likewise had strong statistical support by bootstrap analysis (data not shown). The expanded view of the CaSR and CaSR-like clades in Fig. 1 shows the topology of major taxonomic groups. The phylogenetic tree

Fig. 1. The extracellular Ca2+-sensing receptor (CaSR) and glutamate receptor-like G protein-coupled receptor (GPCR) protein phylogeny estimate from Bayesian inference. Bayesian analysis of CaSR protein sequences from 42 vertebrates, CaSR-like protein sequences from 4 non-vertebrate chordates in expanded view, and 92 related glutamate receptor-like GPCR protein sequences in collapsed view (comprising: pheromone receptors, PherRs; vomeronasal type 2 receptors, V2Rs; taste type 1 receptors, T1Rs; some amino acid and odorant and orphan receptors; metabotropic glutamate receptors, mGluRs; and γ-aminobutyric acid type B receptors, GABABRs; see Supplementary File 1 for complete list with accession numbers). Vertebrate and non-vertebrate chordate clades are in shaded blocks. These clades exhibit subtle, but clearly identifiable sequence differences and indel features. Posterior probabilities are indicated on the phylogram as percent values; where not indicated, the values were 100%. The number of sequences in collapsed branches is shown in parentheses.

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

89

placement in the tree estimates of some additional adaptive evolutionary events differed. Specifically, in the full-sequence and the VFT subset, selection was evidenced on the branch to vertebrate CaSRs, whereas in the trimmed data subset, ω evidence for selection was seen at the basal branch of the CaSR/CaSR-like clade and on the branches to cartilaginous fishes and to the birds/mammals.

estimates of CaSR evolution by both methods match the currently accepted evolutionary relationships among vertebrates (Delsuc et al., 2006; Maddison and Schulz, 2007) and among glutamate receptorlike GPCRs (Cao et al., 2009; Conigrave and Hampson, 2010). The sub-group features of CaSR and CaSR-like protein sequences are in Fig. 2 in comparison with the PherR/V2R sister group. Sequence alignment reveals a distinctive indel in the ECD and striking variation in ICD length among clades as a consequence of multiple apparent indel events. The ECD indels that are seen among the tetrapod, lobe-finned and cartilaginous fish clades are distinctly different among groups and conserved within each group. The agnathan representative (P. marinus) has an unresolved genome sequence that spans this ECD indel. Within the ICD, the indel difference near the ICD-7TM junction and the overall length of the ICD are highly variable. An insertion of significant size is only seen in the tetrapod and cephalochordate clades, and these insertions are dissimilar. Even within the tetrapod clade there is length variability of this insertion, with mammal receptors displaying a longer indel sequence relative to that seen in receptors of bird and reptile (showing smaller insertion among the examined members of the tetrapod group).

3.2.3. Selection analysis by site Using both statistical approaches (PAML ‘codeml’ and Selecton) to test evolutionary selection by site, we again found evidence broadly supporting strong purifying selection. Omega values were consistently b1, and near to zero in fact, at all sites reported in the analysis. However, there are some deviations away from the near-zero values that were consistently seen between each analysis, and these exceptions may represent codon sites experiencing relaxed selection pressure (Fig. 4). Displayed against a reference human CaSR protein sequence for position numbering, statistical evidence as PAML ω and Selection Ka/Ks by amino acid position shows purifying selection widely. Differences from the general pattern are seen within the signal peptide sequence, within the VFT, and within the ICD.

3.2. Evolutionary selection

3.3. Amino acid conservation at Ca 2+ binding sites

Testing for evidence of evolutionary selection by PAML and Selecton methods yielded a similar overall finding of purifying selection, with strong consensus among replicate analyses. There were some notable deviations, however, from this general pattern of purifying selection, as specifically mentioned below (in Sections 3.2.2 and 3.2.3).

Compared with all collected glutamate receptor-like GPCR sequences, WebLogo revealed high bit scores at the predicted residues of Huang et al. (2009), involved in Ca2+ binding (Fig. 5) in the vertebrate CaSR subgroup corresponding to strong amino acid conservation, and for which Selecton analysis reported evidence for purifying selection (Fig. 4).

3.2.1. Overall estimate of selection pressure Overall, there is evidence for strong purifying selection under the PAML ‘codeml’ one-ratio model (branch model 0), with ω values close to zero in all replicate analyses (Table 1).

3.4. Feature mapping of Ca 2+ binding sites

3.2.2. Selection analysis by branch Branch analysis for selection by PAML, comparing branch model 0 with the free-ratio model (branch model 1) in a likelihood ratio test points to model 1 (different ω by branch) as the more probable. This was a consistent finding across the data subsets in multiple replicate analyses. Generally, there was overall strong evidence for purifying selection, with ω values close to zero within major phylogenetic subgroups among CaSRs and CaSR-like molecules. But, PAML branch analysis also revealed clear statistical evidence for adaptive evolution at some branches (Fig. 3). In the CaSR and CaSR-like lineages, on a background of purifying selection, adaptive evolution is strongly evidenced by statistical analysis on the basal branch for non-vertebrate chordate CaSR-like proteins and also on the tetrapod branch, distinguishing this latter clade from the other vertebrates. Although consistent overall, the

Glutamate receptor-like GPCRs have a characteristically large extracellular domain which functions in ligand binding to trigger receptor activation and downstream signaling cascades. Our SWISS-MODEL homology model of the CaSR ECD, has the expected VFT and CRD elements and the expected placement of the five predicted Ca 2+ binding sites of Huang et al. (2009) within the VFT interlobe cleft (Fig. 6). 4. Discussion 4.1. Phylogenetic analysis: CaSR as a subgroup of glutamate receptor-like GPCRs Our consensus molecular phylogeny estimate mirrors the accepted evolutionary history of chordate and vertebrate taxa (Delsuc et al., 2006; Maddison and Schulz, 2007), and on a finer scale, the reported topologies of molecular phylogenies for GPCRs and CaSRs in the fishes (Hashiguchi and Nishida, 2006; Cao et al., 2009). Our phylogenetic

Fig. 2. Phylogenetic tree estimate of extracellular Ca2+-sensing receptor (CaSR) and CaSR-like proteins, with pheromone and vomeronasal type 2 receptors (PherR/V2R) as the outgroup. Sequence architecture is diagramed in line fashion next to each corresponding clade. The seven-transmembrane (7TM) domain is aligned among all clades. Both extracellular domain (ECD) and intracellular domain (ICD) have indel differences among clades, represented by either a block (inserted sequence) or a dotted line (deleted or absent sequences). Variations in shading represent group-specific differences among the inserted protein sequence. Question mark indicates unknown amino acid residues in the inferred protein sequences as a consequence of gaps, or poor or ambiguous reads in the underlying genomic sequence. Asterisks indicate indel sequence length variation among vertebrate classes within the clade (see Sections 3.1 and 4.2 for details).

90

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

Table 1 Selection analysis results. Maximum likelihood selection analysis by PAML ‘codeml’ was conducted on extracellular Ca2+-sensing receptor (CaSR), CaSR-like and outgroup pheromone and vomeronasal receptor (PherR/V2R) clades. Analysis was repeated for the full-length sequences, the Venus flytrap (VFT) module sequences only, and the trimmed sequences (see Section 2.3 for details). Average ω values, likelihood, and probabilities (as calculated by the X2 test) supporting branch model M1 (where ω differs among branches) are listed. For each data set, 6–7 replicate analyses were performed. Data set

Model

Full-sequence

Branch Branch Branch Branch Branch Branch

VFT module only Trimmed sequence

M0 M1 M0 M1 M0 M1

Average dN/dS

Likelihood

Probability

ω = 0.03 ω = by branch ω = 0.03 ω = by branch ω = 0.03 ω = by branch

−49,853.4 −49,695.7 −27,263.9 −27,146.3 −49,861.6 −49,700.3

1.8E−21 6.1E−11 9.4E−23

scheme clearly places the CaSRs (including the CaSR-like molecules) as a subgroup of glutamate receptor-like GPCRs, with PherR/V2Rs as the sister group. The physiological sensitivity of CaSRs to aromatic amino acids in addition to Ca2+ (Conigrave et al., 2007) is consistent with the phylogenetic placement of CaSRs among PherRs and V2Rs, and other odorant and taste receptors. Interestingly, CaSRs appear restricted in distribution to the chordate–vertebrate lineage. Our overall study aim is to understand better the evolutionary history of the CaSR in relation to vertebrate biology and the natural history of major vertebrate groups. And we interpret on this scaffold our current findings from phylogenetic tree building and tests of selection. 4.2. CaSR clade characteristics and features The CaSR expresses sensitivity to extracellular Ca 2+ in the lowmillimolar range, a range that is tuned to typical extracellular Ca 2+ concentrations in vertebrates (Chang et al., 1998). CaSR functionality

Fig. 4. PAML ‘codeml’ and Selecton analysis of selection at the codon level. The graph displays both PAML ‘codeml’ ω values and Selecton Ka/Ks values for codons mapped against the human CaSR (hCaSR) protein numbering sequence. Receptor domains and features are annotated at the top of the graph. Due to gaps in sequence alignments, ω values for some positions were not returned from PAML analysis. The positions of the ECD and ICD indels are indicated by asterisks (see Sections 3.1 and 4.2 for details).

has been confirmed in mammals, bony fishes and cartilaginous fishes (Brown et al., 1993; Nearing et al., 2002; Loretz et al., 2004). High sequence similarity across all vertebrate classes suggests strong structural and functional conservation of the receptor among these groups. The predicted CaSR-like genes and their inferred proteins for the two urochordate tunicates (Ciona intestinalis and Ciona savignyi) and the cephalochordate amphioxus (Branchiostoma floridae) share sequence similarity with vertebrate CaSRs, but the functional status of these receptor proteins is untested to date. Among and within the CaSR and CaSR-like subgroups, taxonomic subdivisions display subtle, but characteristic, sequence differences involving indel changes and C-terminal truncation, for example. The functional consequences, if any, of these featural differences are

Fig. 3. Selection test results mapped onto phylogenetic tree estimates of CaSR and CaSR-like molecules, with the collapsed PherR/V2R sister outgroup. Some tetrapod classes are collapsed (mammals, birds, and bony fishes), with the number of sequences in each group indicated. Selection tests by PAML ‘codeml’ analyses support purifying selection on almost all branches in the CaSR and CaSR-like clades, where ω was calculated to be close to zero. Dots indicate branches with ω evidence for adaptive selection (ω > 1), based on a consensus of multiple data subsets and analyses (see Sections 2.3, 3.2.2 and 4.3 for details). With consensus among data sets, the marked branches show strong evidence leading to adaptive evolution for the chordate CaSR-like clade and on the branch giving rise to the tetrapods from the piscine vertebrate clades. (A) Consensus protein phylogeny estimate for the full-length and VFT module data sets. There is statistical evidence of adaptive evolution at the root of the vertebrate CaSR clade. (B) Consensus protein phylogeny estimate for the trimmed data subset showing slight variation from that in panel A. Statistical support for high ω is evidenced within the tetrapods at the basal branch of the bird and mammalian classes, as well as at the base of the cartilaginous fish clade. Support for adaptive evolution is seen at a basal branch for the CaSR and CaSR-like clades (this apparent clade includes two mouse vomeronasal receptors, Mus musculus Vm2r1 and V2R2, that may have this placement because of long-branch attraction; see Section 4.3 for details). See text and supplementary files for keys to species identifiers.

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

91

Fig. 5. WebLogo analysis of the five predicted binding sites of the Ca2+-sensing receptor (CaSR) and of corresponding, aligned amino acid positions in glutamate receptor-like GPCRs (including CaSRs). Conservation at each amino acid position in the binding site is assessed as the WebLogo bit score, with amino acid usage frequencies at that position reflected in the relative size of the single-letter amino acid identifiers within the bit score bars. (A) WebLogo bit scores and consensus amino acid residues calculated for the glutamate receptor-like GPCR sequence data set. (B) Bit scores and site consensus calculated for the vertebrate CaSR sequence subset, revealing overall higher bit scores and reflecting stabilization and conservation of binding sites for Ca2+ ligand recognition in CaSRs. The amino acid residues contributing to predicted binding sites of Huang et al. (2009) (according to the numbering scheme for human CaSR) are: site 1, S147, S170, D190, Y218, E297; site 2, D215, L242,S244, D248, Q253; site 3, S224, E228, E229, E231, E232; site 4, E350, E353, E354, N386, S388; and site 5, E378, E379, T396, D398, E399. These sites are positioned in the receptor protein as shown in Fig. 6.

largely unknown at present. In fact, our comparative functional studies on the bovine and piscine CaSRs found no substantial differences in either Ca2+-sensing by the receptor or in signal transductional capabilities through two major pathways, specifically the inositol phosphate and MAP kinase systems (Loretz et al., 2004). As depicted in Fig. 2, an indel distinguishes the ECD region among major clades represented in the CaSR phylogeny. CaSRs of both the sarcopterygians (the lobe-finned fishes and the tetrapods) and the cartilaginous fishes have sequence insertions relative to the CaSR sequence of bony fishes. The agnathan representative in our phylogenetic tree has a sequence gap in the genomic record that, unfortunately, stretches

Fig. 6. Protein homology model visualization of the Ca2+-sensing receptor (CaSR) extracellular domain (ECD). The model was constructed using the tilapia CaSR sequence and a rat metabotropic glutamate receptor template (see Section 2.5 for details). The wire-frame ECD illustration displays the bi-lobed Venus flytrap (VFT) module and the nine-cysteine domain (CRD) of glutamate receptor-like G protein-coupled receptor module. The two lobes of the VFT surround the ligand-binding cleft. Amino acids contributing to the five predicted Ca2+ binding sites of Huang et al. (2009) (labeled 1–5) are displayed in space-filling mode.

over this ECD indel region. Similar to the CaSRs of bony fishes, the more primitive CaSR-like representatives of tunicates and amphioxus lack the inserted ECD segment, as do the other PherRs, V2Rs, OdorRs and T1R1s that are included in our phylogenetic analysis. Comparison of ECD indels reveals sequences that are similar within individual vertebrate clades, but that are different among clades. Therefore, multiple insertion events probably occurred. Since the location of the inserted segments is conserved among clades, the ancestral sequence may have possessed a “hot spot” susceptible to insertions and for which there might have been some functional consequence. The ICD domain is highly length-variable among CaSR clades but, nevertheless, includes some conserved secondary structural features (Loretz, 2008). A second prominent group-specific indel follows the 7TM domain as seen in Fig. 2 and separates predicted secondary protein structures within the ICD. Specifically, the C-terminal tail contains a predicted helix structure (nearer the 7TM domain) and a β-strand structure (farther from the 7TM domain) that bound the indel locus (Loretz, 2008). These two secondary structures are linked experimentally to phospholipase C and MAP kinase activation and to filamin-A interaction, respectively. Relative to the PAML and Selecton analyses, these two structured sequences exhibit low ω values in selection analysis, indicative of strong conservation (Fig. 4). The amniote tetrapods and the cephalochordates both have sequence insertions of significant length. Among the amniotes, the insertion is longest in mammals and shorter in the birds (represented by Gallus gallus, Meleagris gallopavo, Anas platyrhynchos, and Taeniopygia guttata), and reptile (Anolis carolinensis). These insertions are markedly dissimilar in sequence between the amniotes and cephalochordates, suggesting that insertion events occurred independently in these groups, and not as a single indel event in an ancestral glutamate receptor-like GPCR. The function(s) of the insertion remain(s) unknown, but it may contribute to positioning of the two structured regions in functionally-significant clade-specific ways (Loretz, 2008), or it may itself have functional activity related to receptor trafficking and degradation (Zhuang et al., 2012). Moreover, and adding complexity to signal transducing potential, the clade-specific length differences at the C-terminus are of unknown functional significance still.

92

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

4.3. Evolutionary selection of CaSR

4.4. CaSR ligand specificity and conservation of binding sites

The CaSR is a critical detector in calcium homeostasis in vertebrates, and because its presence is apparently limited to this taxonomic lineage, we propose a possible functional association with the vertebrate pattern of organismal calcium balance. Evidence for evolutionary selection within and among the CaSR, CaSR-like, and the closely related PherR/V2R subgroups points to strong overall purifying selection. Within the collapsed clades of Fig. 3, strong support for purifying selection is reported for almost all branches (among 36 taxa in the collapsed clades there are 4 minor branches that deviate from purifying selection). This finding is confirmed by our analysis of the VFT module and trimmed data sets. This evidentiary finding of purifying selection is not surprising since the CaSR is specialized for recognizing Ca2+, and that functional role is maintained and exhibited through a strongly conserved structure. However, a finer-scale test of selection by branch yielded quite interesting results. The evidence suggests that the CaSR subfamily of glutamate receptor-like GPCRs has undergone at times bouts of adaptive evolution, to perhaps “tune” the receptor to particular roles in the biology of some vertebrate groups. Adaptive evolution testing by Cao et al. (2009) at a deeper level of VFT module phylogeny supports the divergence of glutamate receptor-like GPCRs into ligand-specific subtypes. For the CaSRs, the statistical testing evidence supports adaptive evolution along some branches of our phylogenetic tree reconstruction. These branches are located at or near the roots of major taxonomic groups. Specifically, and confirmed in all data sets, the root branch of the tetrapod clade shows evidence of adaptive evolution. In addition, analysis of both the full-length and VFT module subset supports adaptive evolution with high confidence at the root of the CaSR clade (see Fig. 3A). As further interpreted below, these bouts of adaptive evolution may be related to the physiology and natural history of vertebrates. And adaptive evolution is evidenced for the non-vertebrate chordate CaSR-like molecules. The trimmed data subset focuses statistical analysis on conserved amino acid sequences, excluding sequence differences arising by indel events. Similar statistical findings are generated with this subset, but with adaptive evolution being indicated for some branches. Specifically, within the CaSR subgroup, a high ω value is evidenced at the branch separating the birds and mammals from other tetrapod classes, and at the branch leading to the cartilaginous fish clade (see Fig. 3B). Analysis of the trimmed sequence subset generates the anomaly of two closely related mouse vomeronasal sequences (Mus musculus Vm2R1 and V2R2) being placed apart from the PherR/V2R group, to which they likely belong. This finding probably results from long-branch attraction, due to somewhat unusual, and distinctive, sequences of these two receptors (Yang and Rannala, 2012). The high ω values, as evidence for adaptive selection, on these branches within the CaSR and CaSR-like subgroups of the tree provide a focus for interpreting the biological functions of this receptor and the possible significance of evolutionary changes. Based on the multiple ligand sensitivities of CaSRs (Wellendorph et al., 2009; Conigrave and Hampson, 2010) and the phylogenetic position of CaSRs, the ancestral CaSR was likely a PherR, V2R, or OdorR. Through adaptive evolution, the proto-CaSR became specialized, and achieved a narrower sensing repertoire with greatest sensitivity and selectivity to Ca 2 +. At the individual codon level, our PAML and Selecton tests for evolutionary selection were unable to detect evidence of adaptive evolution, and instead supported a conclusion of purifying selection within the CaSR and CaSR-like groups. At some amino acid sites, however, the PAML and Selecton test statistics displayed deviations away from purifying selection as seen in Fig. 4. These amino acid residues are found within the signal peptide sequence (which probably has little functional significance for the mature, fully-processed receptor in the plasma membrane), and at the indel positions within the ECD and ICD that characterize the major clades according to the molecular architecture shown in Fig. 2.

Examination of the five predicted Ca 2+ binding sites of the CaSR VFT (Huang et al., 2009) indicates strong amino acid conservation as reflected in the high WebLogo bit scores for each of these residues. The CaSR five predicted Ca 2+ binding sites are characterized by an abundance of negatively charged amino acid residues positioned within the cleft region of the VFT motif in the ECD (Huang et al., 2009). Interestingly, the predicted binding site 1 that is located within interlobe cleft is conserved within all glutamate receptor-like GPCRs as illustrated in the WebLogo analyses. Perhaps, the specific configuration of amino acids at this site, pre-adapted an ancestral receptor for calcium sensing. And, apparently, this sensing capability is expressed in other, related GPCRs such as GPRC6A (Pi et al., 2005). Although other related glutamate receptor-like GPCRs display Ca 2+ sensitivity, such as the GPCR6A that is expressed in several tissue types including bone, the status of such orphan receptors as physiological sensors of Ca 2+ remains uncertain (Pi et al., 2005; Conigrave and Hampson, 2006). In comparison with our larger data set of sequences, these amino acid sites have higher bit scores among an exclusive data set of CaSRs. This conservation is supported by the Selecton test result of purifying selection at these binding sites. The surface exposure within the VFT cleft on the homology model of these strongly conserved residues (the highlighted residues in Fig. 6) is consistent with the binding site prediction of Huang et al. (2009) and with a strong functional role in vertebrate calcium homeostasis. There is consensus agreement that the endogenous ligands of glutamate receptor-like GPCRs bind in the VFT, eliciting conformational changes that trigger downstream signaling (Wellendorph et al., 2009; Conigrave and Hampson, 2010). But complicating the understanding of CaSR evolution as a Ca2+ sensor is the knowledge that other surfaces of the receptor are apparently capable of recognizing Ca2+ and other allosteric compounds. These surfaces are present at the interface between the homodimeric receptor, at the CRD, or within pockets of the 7TM domain (Wellendorph et al., 2009; Conigrave and Hampson, 2010). 4.5. Selection of CaSR relative to vertebrate natural history The subtle but distinct differences in CaSR receptor protein structure among vertebrate clades may be interpreted relatively, or even linked evolutionarily, to important natural historical and physiological characteristics. These include, importantly, element of calcium homeostasis supporting successful lifestyles in a variety of habitats, including marine and fresh waters and land, and perhaps even the presence of a vertebrate-type skeleton based on cartilage and hydroxyapatite bone matrix (Fig. 7). Vertebrates arose in the oceans and coastal marine waters (Witten and Huysseune, 2009), and as some migrated into fresh waters, they certainly required mechanisms to detect and respond to the lower-calcium milieu. In this natural historical context, CaSRs seem to have become an important vertebrate detector of internal and external medium composition (Hubbard et al., 2000, 2002; Hubbard and Canario, 2007). In their natural bathing waters, aquatic vertebrates have continual access to dissolved Ca 2+. Supported by CaSR immunohistochemisty in tilapia (Loretz et al., 2009), receptor expression in ion-transporting gill tissue which interfaces with the external medium, may directly influence Ca2+ transport and homeostasis. In all vertebrates, internal Ca2+ concentration is detected by ion-transporting intestine and kidney that could functionally respond to Ca 2+ concentration changes in an independent manner, and by the endocrine system that indirectly regulates Ca2+ homeostasis through hormone-signaling pathways (Nearing et al., 2002; Loretz et al., 2009). With terrestrial life that left behind an aquatic environment with its always-available source of dissolved Ca2+, we hypothesize that in sarcopterygians the CaSR evolved to have special significance in the terrestrial lifestyle where bone took on the critically important function of calcium storage. In the context of the vertebrate radiation, clade-specific variations in receptor structure and function may assume

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

93

zebrafish and advanced Mozambique tilapia, respectively). Altogether, the different patterns of tissue composition and bone organization may confer advantages in Ca 2+ storage and lability to specific vertebrate groups according to their habitat and life history. Future work to explore further these patterns will be informative and interesting. 5. Conclusion

Fig. 7. Matrix of vertebrate and non-vertebrate chordate clades indicating their characteristic biological and natural historical features. Clade divisions from the estimated phylogenetic tree reflect differences in protein sequence and architecture, and in a biological context, suggest group-specific roles for CaSR in Ca2+ homeostasis that have been selected over evolutionary time.

added importance in increasingly complex calcium homeostatic systems. Our findings from branch analysis for selection would support such adaptive evolutionary fine-tuning of receptor functionality according to life histories. 4.6. CaSR and vertebrate skeleton The vertebrate skeleton is complex in organization. It consists of four main tissue types, namely bone, dentine, enamel, and cartilage, that are variously developed as dermoskeleton and endoskeleton (Donoghue et al., 2006). Our data set comprises CaSR sequences from vertebrate groups representing different patterns of skeletal development and organization. These include, specifically the jawless lamprey P. marinus (the most primitive living vertebrate with simple cartilaginous skeletal elements), the chondrichthyan fishes (with largely cartilaginous skeleton and dermal denticles), the bony fishes (with well-developed bone endoskeleton and dermal scales), and terrestrial vertebrates (with bone endoskeleton and lacking scales). Cartilage, bone, and dermal scale development depends on the regulated activities of chondrocytes, osteocytes (including both osteoblasts and osteoclasts), scleroblasts, and other cell types to lay down the protein matrix and to deposit and (re)organize the mineral components of the vertebrate skeleton. And, in fact, the development of the various expressions of vertebrate skeleton may depend strongly on the nature of the protein matrix (Kawasaki and Weiss, 2003; Kawasaki et al., 2004; Donoghue et al., 2006). In this regard, and as a component of overall calcium homeostasis, the evolution of vertebrate skeleton likely involved a complex interplay among various cells expressing extracellular matrix and calcium-binding proteins, and perhaps CaSR as the calcium-detector. The critical importance of CaSR in normal mammalian bone development has been elegantly demonstrated by conditional gene knockout in mouse (Chang et al., 2008). CaSR is expressed in skeletal tissues of the bony fishes Mozambique tilapia (Oreochromis mossambicus) and zebrafish (Danio rerio), and our preliminary studies show an altered skeletal phenotype consequent to morpholino oligonucleotide knockdown of CaSR expression in zebrafish (Loretz et al., 2012; Herberger and Loretz, in press). Therefore, from these experimental studies on mouse and zebrafish, an important role for CaSR in vertebrate skeletogenesis (among other roles in calcium homeostasis) seems to be evolutionarily old. Most living vertebrates have osteocytes that reside within the bone. In contrast, some advanced teleosts have bone from which osteocytes are excluded (Witten and Huysseune, 2009; Witten et al., 2010; Loretz et al., 2012). Our collected sequence data includes representatives of both cellular and acellular teleost bone types (primitive

The CaSR is tightly associated with the regulation of calcium balance in vertebrates through endocrine and other mechanisms (Brown, 2007; Loretz, 2008; Loretz et al., 2009). In addition, its restricted distribution and expression in the vertebrate animals (and perhaps in some evolutionarily-related chordates as CaSR-like proteins of unconfirmed function) may have evolved contemporaneously (at an early time) with matrix proteins and skeletal forms, including mineralized tissues, that became characteristic of the vertebrate lineage (Donoghue et al., 2006). The phylogenetic analysis of CaSR places this receptor as a subgroup of glutamate receptor-like GPCRs. In a biological context, clade relationships within the CaSR phylogeny can be related to distinguishing life history features. Strong purifying selection is evidenced overall within the CaSR subfamily, with notable bouts of adaptive evolution separating the CaSRs of terrestrial tetrapods from those of aquatic vertebrates, and the CaSR-like molecules of non-vertebrate chordates. Analysis at the codon level also reveals purifying selection and strong amino acid conservation, specifically at the residues predicted for Ca2+ binding activity. With the addition in the future of more CaSR and related glutamate receptor-like GPCR sequences, and knowledge of corresponding receptor functionality, we hope to refine our understanding of adaptive evolution at the scale of individual codons in functional and threedimensional structural frameworks. Conflicts of interest The authors declare no conflicts of interests, financial or otherwise. Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.cbd.2012.12.004. Acknowledgments Portions of the work by ALH were supported by an East Asia and Pacific Summer Institute (EAPSI) fellowship from the U.S. National Science Foundation and the Japan Society for the Promotion of Science. The advice and guidance of Professor Yoshio Takei and Dr. Susumu Hyodo at The University of Tokyo are also recognized. References Arnold, K., Bordoli, L., Kopp, J., Schwede, T., 2006. The SWISS-MODEL Workspace: a web-based environment for protein structure homology modeling. Bioinformatics 22, 195–201 (http://swissmodel.expasy.org/). Brown, E.M., 2007. The calcium-sensing receptor: physiology, pathophysiology and CaRbased therapeutics (Physiology and pathophysiology of CaR). Subcell. Biochem. 45, 139–167 (No URL). Brown, E.M., 2010. Clinical utility of calcimimetics targeting the extracellular calciumsensing receptor (CaSR). Biochem. Pharmacol. 80, 297–307. http://dx.doi.org/ 10.1016/j.bcp.2010.04.002. Brown, E.M., MacLeod, R.J., 2001. Extracellular calcium sensing and extracellular calcium signaling. Physiol. Rev. 81, 239–297 (http://physrev.physiology.org/cgi/content/full/ 81/1/239). Brown, E.M., Gamba, G., Riccardi, D., Lombardi, M., Butters, R., Kifor, O., Sun, A., Hediger, M.A., Lytton, J., Hebert, S.C., 1993. Cloning and characterization of an extracellular Ca2+-sensing receptor from bovine parathyroid. Nature 366, 575–580. http:// dx.doi.org/10.1038/366575a0. Cao, J., Huang, S., Qian, J., Huang, J., Jin, L., Su, Z., Yang, J., Lui, J., 2009. Evolution of the class C GPCR Venus flytrap modules involved positive selected functional divergence. BMC Evol. Biol. 9, 67. http://dx.doi.org/10.1186/1471-2148-9-67. Chang, W., Shoback, D., 2004. Extracellular Ca2+-sensing receptors—an overview. Cell Calcium 35, 183–196. http://dx.doi.org/10.1016/j.ceca.2003.10.012. Chang, W., Pratt, S., Chen, T.H., Nemeth, E., Huang, Z., Shoback, D., 1998. Coupling of calcium receptors to inositol phosphate and cyclic AMP generation in mammalian cells and Xenopus laevis oocytes and immunodetection of receptor protein by

94

A.L. Herberger, C.A. Loretz / Comparative Biochemistry and Physiology, Part D 8 (2013) 86–94

region-specific antipeptide antisera. J. Bone Miner. Res. 13, 570–580. http:// dx.doi.org/10.1359/jbmr.1998.13.4.570. Chang, W., Tu, C., Chen, T.H., Bikle, D., Shoback, D., 2008. The extracellular calciumsensing receptor (CaSR) is a critical modulator of skeletal development. Sci. Signal. 1 (35), ra1. http://dx.doi.org/10.1126/scisignal.1159945. Conigrave, A.D., Hampson, D.R., 2006. Broad-spectrum L-amino acid sensing by class 3 G-protein-coupled receptors. Trends Endocrinol. Metab. 17, 398–407. http:// dx.doi.org/10.1016/j.tem.2006.10.012. Conigrave, A.D., Hampson, D.R., 2010. Broad-spectrum amino acid-sensing class C Gprotein coupled receptors: molecular mechanisms, physiological significance and options for drug development. Pharmacol. Ther. 127, 252–260. http://dx.doi.org/ 10.1016/j.pharmthera.2010.04.007. Conigrave, A.D., Mun, H.-C., Brennan, S.C., 2007. Physiological significance of L-amino acid sensing by extracellular Ca2+-sensing receptors. Biochem. Soc. Trans. 35, 1195–1198. http://dx.doi.org/10.1042/BST0351195. Crooks, G.E., Hon, G., Chandonia, J.M., Brenner, S.E., 2004. WebLogo: a sequence logo generator. Genome Res. 14, 1188–1190 (http://weblogo.threeplusone.com/). Delsuc, F., Brinkmann, H., Chourrout, D., Philippe, H., 2006. Tunicates and not cephalochordates are the closest living relatives of vertebrates. Nature 439, 965–968. http://dx.doi.org/10.1038/nature04336. Donoghue, P.C.J., Sansom, I.J., Downs, J.P., 2006. Early evolution of vertebrate skeletal tissues and cellular interactions, and the canalization of skeletal development. J. Exp. Zool. (Mol. Dev. Evol.) 306B, 278–294. http://dx.doi.org/10.1002/jez.b.21090. Felder, C.B., Graul, R.C., Lee, A.Y., Merkle, H.-P., Sadee, W., 1999. The Venus flytrap of periplasmic binding proteins: an ancient protein module present in multiple drug receptors. AAPS PharmSci. 1 (2). http://dx.doi.org/10.1208/ps010202 (article 2). Finn, R.D., Mistry, J., Shuster-Böckler, B., Griffiths-Jones, S., Hollich, V., Lassmann, T., Moxon, S., Marshall, M., Khanna, A., Durbin, R., Eddy, S.R., Sonnhammer, E.L.L., Bateman, A., 2006. Pfam: clans, web tools and services. Nucleic Acids Res. 34, D247–D251. http://dx.doi.org/10.1093/nar/gkj149. Guex, N., Peitsch, M.C., 1997. SWISS-MODEL and the Swiss-Pdb Viewer: an environment for comparative protein modeling. Electrophoresis 18, 2714–2723 (http:// www.expasy.org/spdbv/). Hashiguchi, Y., Nishida, M., 2006. Evolution and origin of vomeronasal-type odorant receptor gene repertoire in fishes. BMC Evol. Biol. 6, 76 (http://www.biomedcentral. com/1471-2148/6/76). Herberger, A.L., Loretz, C.A., in press. Morpholino oligonucleotide knockdown reveals extracellular calcium-sensing receptor role in zebrafish skeletal development. FASEB J. 27. Huang, Y., Zhou, Y., Castiblanco, A., Yang, W., Brown, E., Yang, J., 2009. Multiple Ca2+binding bites in the extracellular domain of the Ca2+-sensing receptor corresponding to cooperative Ca2+ response. Biochemistry 48, 388–398. http://dx.doi.org/ 10.1021/bi8014604. Hubbard, P.C., Canario, A.V.M., 2007. Evidence that olfactory sensitivities to calcium and sodium are mediated by different mechanisms in the goldfish Carassius auratus. Neurosci. Lett. 414, 90–93 (http://dx.doi.org/10.1016/j.neulet.2006.12. 007). Hubbard, P.C., Barata, E.N., Canario, A.V.M., 2000. Olfactory sensitivity to changes in environmental [Ca2+] in the marine teleost Sparus aurata. J. Exp. Biol. 203, 3821–3829 (http://jeb.biologists.org/cgi/reprint/203/24/3821). Hubbard, P.C., Ingleton, P.M., Bendell, L.A., Barata, E.N., Canário, A.V.M., 2002. Olfactory sensitivity to changes in environmental [Ca2+] in the freshwater teleost Carassius auratus: an olfactory role for the Ca2+-sensing receptor? J. Exp. Biol. 205, 2755–2764 (http://jeb.biologists.org/cgi/content/full/205/18/2755). Huson, D.H., Richter, D.C., Rausch, C., Dezulian, T., Franz, M., Rupp, R., 2007. Dendroscope: an interactive viewer for large phylogenetic trees. BMC Bioinforma. 8, 460. http://dx.doi.org/10.1186/1471-2105-8-460. Kawasaki, K., Weiss, K.M., 2003. Mineralized tissue and vertebrate evolution: the secretory calcium-binding phosphoprotein gene cluster. Proc. Natl. Acad. Sci. U. S. A. 100, 4060–4065. http://dx.doi.org/10.1073/pnas.0638023100. Kawasaki, K., Suzuki, T., Weiss, K.M., 2004. Genetic basis for the evolution of vertebrate mineralized tissue. Proc. Natl. Acad. Sci. U. S. A. 101, 11356–11361. http://dx.doi.org/ 10.1073/pnas.0404279101.

Liu, X., He, Q., Studholme, D.J., Wu, Q., Liang, S., Yu, L., 2004. NCD3G: a novel ninecysteine domain in family 3 GPCRs. Trends Biochem. Sci. 29, 458–461. http:// dx.doi.org/10.1016/j.tibs.2004.09.007. Loretz, C.A., 2008. Extracellular calcium-sensing receptors in fishes. Comp. Biochem. Physiol. A: Mol. Integr. Physiol. 149, 225–245. http://dx.doi.org/10.1016/j.cbpa.2008.01.037. Loretz, C.A., Pollina, C., Hyodo, S., Takei, Y., Chang, W., Shoback, D., 2004. cDNA cloning and functional expression of a Ca2+-sensing receptor with truncated carboxyterminal tail from the Mozambique tilapia (Oreochromis mossambicus). J. Biol. Chem. 279, 53288–53297 (http://www.jbc.org/content/279/51/53288). Loretz, C.A., Pollina, C., Hyodo, S., Takei, Y., 2009. Extracellular calcium-sensing receptor expression in osmoregulatory and endocrine tissues of the tilapia. Gen. Comp. Endocrinol. 161, 216–228. http://dx.doi.org/10.1016/j.ygcen.2008.12.020. Loretz, C.A., Pollina, C., Herberger, A.L., Hyodo, S., Takei, Y., 2012. Skeletal tissues in Mozambique tilapia (Oreochromis mossambicus) express the extracellular calciumsensing receptor. Comp. Biochem. Physiol. A: Mol. Integr. Physiol. 163, 311–318. http://dx.doi.org/10.1016/j.cbpa.2012.07.015. Maddison, D.R., Schulz, K.-S. (Eds.), 2007. The Tree of Life Web Project (http://tolweb. org). Nearing, J., Betka, M., Quinn, S., Hentschel, H., Elger, M., Baum, M., Bai, M., Chattopadhyay, N., Brown, E.M., Herbert, S.C., Harris, H.W., 2002. Polyvalent cation receptor proteins (CaRs) are salinity sensors in fish. Proc. Natl. Acad. Sci. U. S. A. 99, 9231–9236 (http://www.pnas.org/cgi/reprint/99/14/9231). Pi, M., Faber, P., Ekema, G., Jackson, P.D., Ting, A., Wang, N., Fontilla-Poole, M., Mays, R.W., Brunden, K.R., Harrington, J.J., Quarles, L.D., 2005. Identification of a novel extracellular cation-sensing G-protein-coupled receptor. J. Biol. Chem. 280, 40201–40209. http://dx.doi.org/10.1074/jbc.M505186200. Pin, J.-P., Galves, T., Prézeau, L., 2003. Evolution, structure, and activation mechanism of family 3/C G-protein-coupled receptors. Pharmacol. Ther. 98, 325–354. http:// dx.doi.org/10.1016/S0163-7258(03)00038-X. Römpler, H., Stäubert, C., Thor, D., Schulz, A., Hofreiter, M., Schöneberg, T.G., 2007. protein-coupled time travel: evolutionary aspects of GPCR research. Med. Interv. 7, 17–25. http://dx.doi.org/10.1124/mi.7.1.5. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574. http://dx.doi.org/10.1093/ bioinformatics/btg180. Stern, A., Doron-Faigenboim, A., Erez, E., Martz, E., Bacharach, E., Pupko, T., 2007. Selecton 2007: advanced models for detecting positive and purifying selection using a Bayesian inference approach. Nucleic Acids Res. 35 (Suppl. 2), W506–W511. http:// dx.doi.org/10.1093/nar/gkm382. Strotmann, R., Schröck, K., Böselt, I., Stäubert, C., Russ, A., Schöneberg, T., 2011. Evolution of GPCR: change and continuity. Mol. Cell. Endocrinol. 331, 170–178. http:// dx.doi.org/10.1016/j.mce.2010.07.012. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kuman, S., 2011. MEGA5: Molecular Evolutionary Genetics Analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739. http:// dx.doi.org/10.1093/molbev/msr121. Wellendorph, P., Johansen, L.D., Bräuner-Osborne, H., 2009. Molecular pharmacology of promiscuous seven transmembrane receptors sensing organic nutrients. Mol. Pharmacol. 76, 453–465. http://dx.doi.org/10.1124/mol.109.055244. Witten, P.E., Huysseune, A., 2009. A comparative view on mechanisms and functions of skeletal remodeling in teleost fish, with special emphasis on osteoclasts and their function. Biol. Rev. 84, 315–346. http://dx.doi.org/10.1111/j.1469-185X.2009.00077.x. Witten, P.E., Huysseune, A., Hall, B.K., 2010. A practical approach for the identification of the many cartilaginous tissues in teleost fish. J. Appl. Ichthyol. 26, 257–262. http://dx.doi.org/10.1111/j.1439-0426.2010.01416.x. Yang, Z., 2007. PAML 4: Phylogenetic Analysis by Maximum Likelihood. Mol. Biol. Evol. 24, 1586–1591. http://dx.doi.org/10.1093/molbev/msm088. Yang, Z., Rannala, B., 2012. Molecular phylogenetics: principles and practice. Nat. Rev. Genet. 13, 303–314. http://dx.doi.org/10.1038/nrg3186. Zhuang, X., Norhup, J.K., Kausik, R., 2012. Large putative PEST-like sequence motif at the carboxyl tail of human calcium receptor directs lysosomal degradation and regulates cell surface receptor level. J. Biol. Chem. 287, 4165–4176. http:// dx.doi.org/10.1074/jbc.M111.271528.