In the chicken genome putative PTH1R (ENSGALG00000005476) and PTH3R (ENSGALG00000019797) were identified. The predicted mature transcripts of chicken PTH1R and PTH3R are 1614 bp and 1626 bp in length that correspond to deduced proteins of 538 and 542 amino acids, respectively, both containing putative signal peptide sequences and sharing 54% overall amino acid sequence identity. The chicken PTH1R mRNA (Accession number FR746109, Additional file 1) was confirmed by RT-PCR on cDNA from whole chicken embryos at stage 26HH and from 3 ESTs found in database from stage 20-21HH whole chick embryos (BU219643), from stage 36HH limbs (BU401969) and from growth plate chondrocytes (BU419888). The chicken PTH3R (Accession number FR746110, Additional file 2) was also confirmed by RT-PCR on the same cDNA sample but no EST was found. Despite thorough database searches and attempts to amplify a chicken PTH2R receptor using degenerate primers on genomic DNA and cDNA a homologue was not identified.

Multiple sequence alignment of the deduced amino acid sequence of chicken PTHRs with mammalian and teleost homologues (Figure 1) revealed the chicken PTH1R shares at least 76% similarity to other vertebrate PTH1Rs and that the chicken PTH3R is 71% similar with the zebrafish PTH3R. The seven transmembrane domains (TM) and the six cysteine residues at the N-ted are fully conserved across vertebrates (Figure 1) as are four putative N-glycosylation sites (Nx[TS]) in PTH1R two of which are also found in PTH3R. The chicken PTHR N-ted also contains the residues L¹³, T³³, Q³⁷, F¹⁸⁴, R¹⁸⁶, L¹⁸⁷ and I¹⁹⁰ previously identified to be involved in the interaction of mammalian PTH1R with PTH(1–34) and PTHrP(1–34) 13 29 . The only amino acid residues in N-ted that are not conserved are L¹³ which is replaced by I¹³ in PTH1R and I¹⁹⁰ which substituted by M¹⁹⁰ in PTH3R (Figure 1). The residues D¹¹³, W¹¹⁸, P¹³² and W¹⁵⁴ which are involved in the structural conformation of the ligand-binding pocket of family 2 B1 GPCRs 11 30 are also conserved in the chicken receptor sequences.

To place in context the evolution of chicken PTHRs, potential PTH/PTHrP homologues from other vertebrate and invertebrate genomes were retrieved from public databases (Table 1). In Xenopus, and lizard genomes, partial sequences for putative PTH1R, PTH2R and PTH3R were identified. In the bird genomes - zebra finch, turkey and duck - only homologues of the chicken PTH1R (ENSTGUG00000000188, ENSMGAG00000002767 and ENSAPLG00000005598, respectively) and PTH3R (ENSTGUG00000001924, ENSMGAG00000001866 and ENSAPLG00000010124, respectively) genes were retrieved. The deduced mature peptide sequence of the bird PTHRs identified are highly conserved with the chicken homologues and share at least 75% and 83% with PTH1R and PTH3R, respectively. In the recently available genome of the cartilaginous fish, elephant shark, three potential receptors (homologues of the vertebrate PTH1R, PTH2R and PTH3R) were identified and in lamprey two PTH/PTHrP receptors seem to be present (Table 1). Among the invertebrate chordates two potential PTH/PTHrP receptors were identified in the Ciona genome (Ciona_a, ENSCING00000002669 on scaffold_67 and Ciona_b, ENSCING00000006282 in scaffold_162) and in the amphioxus Branchiostoma floridae genome (XM_002599399) a PTH/PTHrP receptor gene also appears to be present. Amino acid sequence similarity of the predicted Ciona_a and Ciona_b receptors is 46% and 49%, respectively, with the chicken PTH1R and 45% and 47% from amphioxus with the paralogue PTH3R. The deduced PTH receptor from amphioxus shares 49% and 47% sequence similarity with the chicken PTH1R and PTH3R, respectively.

Species names of published genomes and database accession numbers of identified PTH1R, PTH2R and PTH3R.

The chicken PTH1R and PTH3R share a complex gene organisation and are composed of 13 exons and are identical with the predicted gene structures of their homologues in Xenopus and zebrafish (Additional file 1 and Additional file 2). This contrasts to human PTH1R that contains 15 exons within the mature receptor region. For both chicken PTH1R and PTH3R, the signal peptide region is encoded within the 1^st exon and the TM regions are distributed between the 5^th and 13^th exon.

Gene synteny was maintained for both PTH1R and PTH3R and linked genes were found in chicken, zebrafish, Xenopus and human (Figure 2). The PTH1R gene maps to chicken chromosome 2, to human chromosome 3, to Xenopus scaffold 479, and to zebrafish chromosome 2 and the gene TMIE (transmembrane inner ear-like) was found to be conserved within the compared homologue genome regions. The chicken PTH3R gene maps to chromosome 27 and is homologous to Xenopus scaffold_155 and zebrafish chromosome 12. The gene TACO1 (translational activator of mitochondrially encoded cytochrome c oxidase I) was found to be common in the teleost and tetrapod genome regions analysed.

Despite the apparent absence of a putative PTH2R gene, chicken homologue genes that are conserved in the vertebrate PTH2R gene environment were also found in a conserved cluster in the chicken genome (Figure 2). The genes IDH1 (isocitrate dehydrogenase) and PIP5K3 (1-phosphatidylinositol-4-phosphate 5-kinase) found in close proximity to the zebrafish, Xenopus and human PTH2R were identified on chicken chromosome 7. Similarly, the human genome homologues of the conserved vertebrate PTH3R gene environment were also identified on chromosome 17.

Comparative analysis with the homologue genome regions in the Ciona genome indicates that Ciona PTHR (Ciona_a) located in scaffold_67 shares gene environment conservation with the vertebrate PTH1R. Putative homologues of the vertebrate ALS2CL and JMJD4 are also predicted in the Ciona homologue genome region, which contrasts with Ciona_b gene environment in scaffold_162, where there was no gene synteny. In the amphioxus genome a homologue of the vertebrate TMIE is predicted in close proximity with the putative PTHR locus (scaffold_98).

Phylogenetic analysis of the chicken PTH1R and PTH3R with the retrieved vertebrate and invertebrate homologues is shown in Figure 3A. The consensus tree obtained suggests that the PTH/PTHrP receptors emerged early during the deuterostome radiation. The vertebrate PTHR members shared a common ancestor gene with Ciona and amphioxus and have evolved via gene or genome duplications. Two major clades are present one containing the vertebrate PTH2R and the other clusters PTH1R and PTH3R. This suggests that after the initial gene duplication they have been under different evolutionary pressure and that the vertebrate PTH1R and PTH3R are the result of a recent duplication event. The chicken and other avian PTH/PTHrP receptors cluster within the PTH1R/PTH3R group, confirming their homology (Figure 3A). The three Xenopus and lizard PTH/PTHrP receptors group with the teleost homologues. In Ciona the existence of two PTHRs appears to be the result of a specific duplication event.

The absence of PTH2R in chicken raises questions about the presence in the genome of its putative ligand TIP39. Searches performed on the chicken genome and EST databases failed to identify the homologue of human and teleost TIP39. However, a putative TIP39 gene is predicted in the Xenopus (ENSXETG00000027477) genome and also in non-annotated regions of the lizard (Scaffold GL343355), Platypus (SuperContig Contig8806) and Opossum (chromosome 4) genomes. In addition, in lamprey (GL477315) and elephant shark (AAVX01305587.1) a putative TIP39 gene also seems to be present. In elephant shark, four PTH-family genes were identified clustering with PTH, PTHrP, TIP39 and PTH-L (Figure 3B). No potential TIP39 gene was identified in the tunicate Ciona and amphioxus genomes, suggesting that this gene is specific to vertebrates. Phylogenetic analysis of the deduced TIP39 mature peptides with the PTH/PTHrP members confirmed that the TIP39 clade is basal and shared common origin (Figure 3B). Characterization of the vertebrate TIP39 gene environment revealed that gene order and gene synteny is only conserved in Xenopus and human and no conserved genes were found in teleost and lamprey (Additional file 3).

Tissue expression of the chicken PTHRs transcripts was characterised in several adult tissues and during embryo development. The chicken PTH1R was detected in all adult tissues and embryo stages analysed (Figure 4). During embryo development PTH3R was expressed from stage 4HH (19 hours of incubation) onwards, but was absent or down-regulated in the head of 31HH and 36HH. The chicken PTH3R was also detected in most of the adult tissues despite its low level of expression, but highest transcript expression was found in intestine (hindgut), lung, liver and cartilage (Figure 4).

The cAMP accumulation of a stable cell line expressing chicken PTH1R and PTH3R in the presence of chicken PTH(1–34), PTHrP(1–34) and PTH-L(1–34) is shown in Figure 5. All chicken peptides were able to activate the two receptors in a dose-dependent manner with different half maximal concentrations (EC₅₀). Furthermore, the accumulation of cAMP in cells transfected with PTH3R was almost one order of magnitude larger compared to those with PTH1R. PTHrP produced the highest stimulation of PTH1R (p < 0.05) and together with PTH-L was the most efficient peptide (Table 2). Stimulation of PTH3R was highest for PTHrP and PTH, while PTH-L (100nM) only produced a small cAMP accumulation above basal levels and was approximately 13 times lower than PTH and PTHrP (p < 0.05). The EC₅₀ followed the same pattern with PTH-L in the μM range (Table 2). Human PTH was also able to stimulate chicken PTH1R (21.4 ± 10 pmol/well at 10^-7 M) (Additional file 4). Human TIP39 failed to induce accumulation of cAMP with PTH1R and only negligible cAMP production was obtained for PTH3R (Additional file 4).

Peptides tested, cAMP production, 95% confidence intervals and maximal cAMP production E_max (mean ± SEM). Each result is an average of 3 experiments carried out in triplicate. * indicates statistical significance compared to other peptides (p < 0.05).

A peptide screen using 1 μM and 100 nM of each PTH and PTHrP peptide revealed that irrespective of the peptide PTH3R activation did not cause release of intracellular Ca²⁺ (iCa²⁺). However, PTHrP efficiently stimulated iCa²⁺ accumulation with PTH1R with an EC₅₀ of 2.6 nM and PTH-L (1 μM) caused only a slight stimulation and PTH had no effect (Figure 6).

All procedures with animals were performed in accordance with Portuguese legislation under a “Group-1” licence from the Direcção-Geral de Veterinária, Ministério da Agricultura, do Desenvolvimento Rural e das Pescas, Portugal. Tissues for gene expression analysis were obtained from adult white leghorn chickens (Gallus gallus), anesthetized with diethyl ether (Merck, Spain) before sacrifice by decapitation. Fertile chicken eggs to produce embryos were obtained from Quinta da Freiria (Serpa, Portugal) and kept at 37.6°C under high humidity in an automatic incubator (Brinseca OCTAGON 40) with gentle rotation. Developmental stages 58 selected for analysis coincided with the presence of prominent calcified structures: 4 HH (definitive primitive streak process); 6 HH (head and neural folds); 17 HH (leg bud formation); 26 HH (toes formation); 31 HH (feather germs; emergence of the interdigital membrane) and 36 HH (labial groove; uropygial gland). The chorionallantois membrane (CAM) from 44 HH which increases calcium transport in response to PTH-family peptides 2 was also collected. Tissues were snap frozen in liquid nitrogen and stored at −80°C until use.

Genomes and EST databases were interrogated for PTH/PTHrP receptor homologues using the human (PTH1R, AAR18076; PTH2R, AAH36811) and zebrafish (PTH1R, NP_571432; PTH2R, NP_571452; PTH3R, NP_571453) deduced protein sequences with tBLASTn and default settings 59 . The following databases were searched: 1) the Ensembl genome and pre_ensembl (http://www.ensembl.org) genome assemblies of chicken, opossum (Monodelphis domestica), platypus (Ornithorhynchus anatinus), zebra finch (Taeniopygia guttata), turkey (Meleagris gallopavo), duck (Anas platyrhynchos), toad (Xenopus tropicalis), lizard (Anolis carolinensis), sea lamprey (Petromyzon marinus), 2) the the elephant shark (Callorhinchus milii) genome assembly at http://esharkgenome.imcb.a-star.edu.sg, 3) the chicken EST database (dbEST) of the Biotechnology and Biological Sciences Research Council (http://www.chick.manchester.ac.uk/), the dbEST subsets Aves (taxid:8782), lamprey (taxid:7745) and cartilaginous fishes (taxid:7777) of the National Center for Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov/). Due to the short size of contigs of the shark genome assembly searches were carried out using the coding sequence of the human and zebrafish PTHRs transmembrane (TM) domains retrieved from PRINTs database (http://www.bioinf.manchester.ac.uk/dbbrowser/PRINTS/).

The presence of a putative TIP39 and PTH/PTHrP genes in early vertebrate genomes of the elephant shark, lamprey and early deuterostomes Ciona intestinalis and Branchiostoma floridae (JGI genome assembly, http://www.jgi.doe.gov/) and in protostomes were also investigated using a similar search strategy using the human (Q96A98) and zebrafish (AAI64665) TIP39 sequences and the Fugu PTHrPA (Q9I8E9), PTHrPB (CAG26459), PTH1 (CAG26460) and PTH2 (CAG26461) and PTH-L (CAG26462) sequences.

The deduced mature peptide sequences were obtained from BCM search launcher (http://searchlauncher.bcm.tmc.edu/seq-util) and the localisation of the putative signal peptide region predicted using SignalP (http://www.cbs.dtu.dk/services/SignalP). TM domain regions were deduced using TMHMM (http://www.cbs.dtu.dk/services/TMHMM) and were manually edited according to the PRINTS annotation (http://www.bioinf.manchester.ac.uk/dbbrowser/PRINTS/). Multiple sequence alignments of the amino acid sequence of chicken PTH/PTHrP receptors were performed using ClustalX [version 1.83 60 ] with the following parameters: Gonnet series matrix, Gap opening penalty 10, Gap extension 0.2. The alignments were displayed and manually edited and percentages of sequence similarity and identity calculated using GeneDoc 61 . Phylogenetic trees were constructed using the deduced amino acid sequences of the full-length receptor and were built with the maximum likelihood (PhyML 3.0) 62 63 and Distance methods using Neighbour Joining (BioNJ and Neighbour) 64 . A substitution JTT model gave the best fit for the protein dataset in PROTTEST and was used for phylogenetic tree construction 65 . Maximum likelihood analysis was performed with 500 bootstrap replicates with a discrete gamma distribution of rates among sites with 4 categories. Distance methods performed 1000 bootstrap replicates and ProtDist 66 . Both maximum likelihood (ML) and distance methods generated similar tree topologies. The human secretin receptor (HsaSCTR, AAA64949) was used as an out-group.

An unrooted phylogenetic tree of PTH-family members was also carried out using the deduced mature protein of the vertebrate genes. Consensus trees were constructed following a similar strategy to that described for the PTHRs and both maximum likelihood and distance methods generated similar tree topologies.

Characterization of the chicken PTH/PTHrP receptor gene structures was performed on the basis of Ensembl gene predictions and by searching the genome with the nucleotide sequences of the mature receptor precursor using the NCBI Spidey interface (http://www.ncbi.nlm.nih.gov/spidey/). Intron-exon boundary splice sites (AG/GT) were manually confirmed. The immediate gene environment of the chicken PTHRs was characterised using the NCBI genome chromosome annotations and compared with the homologue genome regions in Human, Xenopus and zebrafish and also with Ciona. The human and zebrafish gene environment was accessed via the NCBI Mapview (http://www.ncbi.nlm.nih.gov/mapview/) and the Xenopus and Ciona retrieved from Ensembl. Genes and gene order flanking PTHRs in the zebrafish and human genomes were confirmed using the Vertebrate Genome Annotation (VEGA) database (http://vega.sanger.ac.uk/). The gene environment of vertebrate TIP39 and PTH was also compared in lamprey, zebrafish, Xenopus and human using a similar strategy.

The methodology for production of cDNA from reverse transcription of total RNA and polymerase chain reaction (PCR) methodologies have been described by Pinheiro et al. 2 . The PTH1R PCR used primers PTH1Rfw atgggatcatatctggtttat and PTH1Rrv ggccagcagacaatacca with denaturation at 94°C for 3 min, 30 cycles of 94°C 30 sec; 55°C 30 sec; 72°C 2 min and a final elongation step at 72°C for 10 min. PTH3R was amplified using the primers PTH3Rfw atggggtctgtgggcagg and PTH3Rrv gttgaagtcgtagatgtagtc and 35 cycles with annealing at 57°C. Since genome searches were negative for PTH2R and to confirm it was not a genome sequence gap, primers PTH2Rfw₁ caaagtagttcatacacatataggagt, PTH2Rfw₂ tgcctcacacatttactgg and PTH2Rrv ggactggctgctggtgct were designed using the conserved regions of known vertebrate sequences and PCR reactions were performed as for PTH1R using the same panel of tissues. The ribosomal subunit 18 S was used as reference gene with primers 18Sfw tcaagaacgaaagtcggagg and 18Srv ggacatctaagggcatcaca and 22 thermocycles in the same conditions as for PTH1R. The PCR products were analysed on 1.5% agarose gel and products sequenced to confirm identity.

The complete coding regions, including the stop codon, of the chicken PTH1R and PTH3R were amplified from embryo limbs (36 HH) cDNA with the primers PTH1Rfw atgggatcatatctggtttat and PTH1Rfinalrv ttacatcactgtctctctttc for PTH1R and PTH3Rfw atggggtctgtgggcagg and PTH3Rfinalrv tcatagcatcgtctccagct for PTH3R. Taq DNA proofreading polymerase (Advantage® 2, polymerase mix, Clontech, France) was used in the PCR reactions according to the manufacturer’s instructions with 10X Advantage 2 PCR Buffer, 0.2 mM dNTP’s (GE Healthcare, Spain) and 0.25 μM of each specific primer for a final volume of 25 μl. The PCR reaction was carried out using the following thermocycle, 94°C 2 min, 30 cycles of 94°C 30 sec, 57°C for PTH1R and 59°C for PTH3R for 1 min, 72°C for 2 min, followed by 72°C for 10 min. The PCR products obtained were gel extracted using a GFX -PCR DNA and Gel Band Purification kit (Amersham Biosciences, Spain) and cloned into a pcDNA3.1/V5-His-TOPO expression vector (Invitrogen, USA) according to the manufacturer’s instructions. The bacterial clones obtained were PCR screened using vector and receptor specific primers and colonies that contained the coding regions in frame with the promoter vector pCMV were selected and plasmid DNA was extracted using the DNA Midi-Prep kit (Roche, Germany). To facilitate receptor integration in the genome, approximately 5 μg of the recombinant vectors were linearized with the restriction enzymes ApaI for PTH1R and BglII for PTH3R (Promega, Spain). The digested products were purified using phenol:clorophorm, ressuspended in 20 μl of distilled DNAse-free water and 5 μl was used to double transfect human embryonic kidney cell line 293 (HEK293 from European Collection of Cell Cultures; Salisbury, UK).

HEK293 cells were maintained in complete Dulbecco’s modified Eagle’s medium (DMEM, Sigma, Spain) with 4.5 g/L glucose, 110 mg/L sodium pyruvate and L-glutamine supplemented with 10% sterile foetal bovine serum and 0.1% penicillin: streptomycin antibiotic mixture (10.000 U:10 mg/ml, Sigma) in a humid 5% CO₂ incubator (Sanyo) at 37°C. On the day prior to transfection, 2–3 x10⁵ cells were seeded on 6 well-plates (Starsted, Portugal). Linearized constructs (1 μg) and Fugene 6 (Roche, Germany) cell transfection reagent were used to transfect cells according to the manufactures protocol. The transfection complex was incubated for 40 min at room temperature before adding to the cells which were left to grown for 72 hours with daily changes of complete medium. Selection of the stably transfected clones was performed with complete medium supplemented with 800 μg/ml of Geneticin (G418 sulphate, GibcoBRL) and 250 μg/ml sterile filtered 1:100 amphotericin B solution (Sigma, Spain). Cell recovery was monitored daily by constant changes of antibiotic selective medium and the success of gene integration and expression was confirmed by RT-PCR with receptor specific primers.

The capacity of chicken (1–34) PTH-family peptides to activate the stably transfected PTH1R and PTH3R cell lines was analysed by measuring intracellular cAMP production. Cell line bearing chicken receptors PTH1R and PTH3R were assayed in triplicate in the same assay in 96 well plates (Starstedt). Three independent experiments were performed. Cells were incubated at 37°C in a CO₂ incubator and two days prior to the experiments, 2–3 x 10⁵ cells were plated and peptide assays were carried out by adding to the recombinant cells decreasing concentrations of test peptide: 100 nM to 0.1 nM of chicken PTH(1–34), PTHrP(1–34) and PTH-L(1–34) or 100 nM of human PTH(1–34) and human TIP39 (Sigma, Spain). Prior to peptide addition, cells were incubated for 40 min with 1 mM 3-isobutyl-1-methylxantine (IBMX, Sigma), which was replaced with fresh medium containing the test peptide in the presence of 1 mM IBMX for a further 40 min. Forskolin (10 μM, Sigma) was used as positive control. Negative control assays were carried out using non-transfected cells in the presence or absence of 100 nM of each peptide. At the end of the assays, cells were washed, ressuspended in 100 μl of 1 x PBS/0.5 M EDTA and immediately frozen at −80°C to promote cell lysis for later quantification of intracellular cAMP production.

For quantification of intracellular cAMP, cells were lysed using 3 consecutive thaw (42°C)/freeze (−80°C) cycles and the supernatant and cell debris were sonicated for 20 sec on ice. Cells were boiled at 100°C for 10 min to denature proteins and the supernatant was collected after centrifugation for 10 min at 4°C and 19000 g. cAMP was quantified in duplicate by radioimmunoassay using the TRK432 kit (GE Healthcare, UK) following the manufacturer's instructions. Data was normalized by subtracting basal cAMP accumulation in the negative controls and production above basal levels per well (cAMP/well) was plotted against peptide concentration. cAMP was also quantified in HEK293 cells stably transfected with the pCMV-GFP (vector expressing green fluorescent protein) to confirm receptor expression and the results for cAMP accumulation were equivalent to the peptide assay negative controls indicating that the vector construct in the cells did not interfere with cAMP production.

Intracellular Ca²⁺ (iCa²⁺) release was measured using the Ca²⁺ sensitive fluorescent dye Fluo-4 NW (Molecular Probes, Invitrogen) according to manufacturer’s instructions. Prior to the assay, plates (96 well black/plates, μClear bottom, Greiner, Germany) were treated with sterile poly-L-lysine (0.1 mg/ml, Sigma, Spain) to avoid cell release. Approximately 1 x 10⁵ cells in 100 μl of selective medium were plated per well and allowed to attach for 2 days. Medium was removed and cells were washed once with 2.5 mM probenecid (Molecular Probes, Invitrogen, Spain) in 1 x PBS and incubated for 30 min at 37°C with 100 μl of the Fluo-4 NW dye followed by an additional 30 min incubation period at room temperature. Incubations were carried out with 100 μl of the different peptides concentrations (from 1 μM to 100 nM) and cell fluorescence was measured every 10 sec over 2 minutes after addition of reagent using a Synergy4 (Biotek, USA) plate reader. Carbachol (100 nM; Sigma, Spain) was used as a positive control for the assay and the background signal was determined by measuring fluorescence in wells containing non-transfected cells. An assay negative control was carried out using non-transfected cells incubated with 100 nM peptide. Data was analysed by subtracting the background fluorescent values and plotted against peptide concentration.

The results are presented as the mean ± SEM of three independent experiments carried out in triplicate for cAMP experiments and two independent experiments carried out in quadruplicate for Ca²⁺ experiments. Data was plotted as the output of cAMP or Ca²⁺ at different peptide concentrations for each PTHR using SigmaPlot 9 (Systat, Inc., San Jose, CA). EC₅₀ and confidence limits were calculated using a sigmoidal curve fitting feature within the Single Ligand Binding routine of the Pharmacology module of Sigmaplot. Unfortunately, the commercial radioimmunoassay for cAMP was withdrawn from the market while the study was already advanced and it was not possible to find an equivalent assay for tests at higher peptide concentrations to ensure saturation was achieved for all peptides. To improve curve fitting the maximum empirical stimulation achieved for each receptor was assumed to be at 100 nM for all assays and the curve fitted accordingly.