A cDNA encoding the calcitonin gene-related peptide type 1 receptor.

Calcitonin gene-related peptide (CGRP) is a neuropeptide with diverse biological effects including potent vasodilator activity. We report here the cloning of a complementary DNA (cDNA) encoding a human CGRP1 receptor, which shares significant peptide sequence homology with the human calcitonin receptor, a member of the G-protein-coupled receptor superfamily. Northern blot analysis revealed that the messenger RNA for this receptor is predominantly expressed in the lung and heart. In situ studies showed specific localization of the receptor mRNA to alveolar cells in the lung and to cardiac myocytes in the heart. Stable expression of the cDNA in human embryonic kidney 293 (HEK 293) cells produced specific, high affinity binding sites for CGRP that displayed pharmacological and functional properties very similar to native human CGRP1 receptor. Exposure of these cells to CGRP resulted in a 60-fold increase in cAMP production, which was inhibited in a competitive manner by the CGRP1 receptor antagonist, CGRP-(8-37).

Calcitonin gene-related peptide (CGRP) is a neuropeptide with diverse biological effects including potent vasodilator activity. We report here the cloning of a complementary DNA (cDNA) encoding a human CGRP 1 receptor, which shares significant peptide sequence homology with the human calcitonin receptor, a member of the G-protein-coupled receptor superfamily. Northern blot analysis revealed that the messenger RNA for this receptor is predominantly expressed in the lung and heart. In situ studies showed specific localization of the receptor mRNA to alveolar cells in the lung and to cardiac myocytes in the heart. Stable expression of the cDNA in human embryonic kidney 293 (HEK 293) cells produced specific, high affinity binding sites for CGRP that displayed pharmacological and functional properties very similar to native human CGRP 1 receptor. Exposure of these cells to CGRP resulted in a 60-fold increase in cAMP production, which was inhibited in a competitive manner by the CGRP 1

receptor antagonist, CGRP-(8 -37).
Calcitonin gene-related peptide (CGRP) 1 is 37-amino acid peptide that exists as highly homologous ␣ or ␤ isoforms in both human and rat (1,2). ␣and ␤-CGRP display very similar biological activities, including peripheral and cerebral vasodilation (3), cardiac acceleration (4), regulation of calcium metabolism (5), reduction of intestinal motility (6), regulation of glucose metabolism (reduction of insulin secretion and insulin sensitivity) (7), diminution of appetite (8), and reduction of growth hormone release (9). The two CGRP peptides differ by 3 amino acids in humans and 1 amino acid in rats. The amino acid sequences of CGRP peptides are well conserved among species and can be considered as members of a family of peptides including the related peptides amylin (46% homology), salmon calcitonin (32% homology), and adrenomedullin (24% homology). These peptides in general have N-terminal ring structures of 6 -7 amino acids involving a disulfide bridge and an amidated C-terminal end (10 -12).
CGRP peptides are localized predominantly in sensory afferent nerves and central neurons (11,12). When released from the cell, the peptides initiate their biological responses by binding to specific cell surface receptors which are predominantly coupled to the activation of adenylyl cyclase (3,13). CGRP receptors have been identified and pharmacologically evaluated in several tissues, including brain, cardiovascular, endothelial, and smooth muscle tissues (12). Multiple CGRP receptors have been observed: based on pharmacological properties they are divided into at least two subtypes and denoted as CGRP 1 and CGRP 2 , according to the classification of Dennis et al. (14). CGRP-(8 -37), which lacks 7 N-terminal amino acid residues, is a selective antagonist of CGRP 1 receptors, whereas the linear analog of CGRP, diacetoamidomethyl cysteine CGRP (Cys(ACM2,7)CGRP), is a selective agonist of CGRP 2 receptors (14).
A deeper understanding of the physiological and pathophysiological effects mediated by the CGRP peptides has been constrained for lack of a cloned cognate receptor. Here we describe the isolation by expressed sequence tag (EST) analysis (15)(16)(17) of a human CGRP receptor that exhibits ligand binding and functional properties of the CGRP 1 receptor. Curiously, this receptor had been cloned previously as an orphan calcitoninlike receptor receptor (rat (18) and, recently, human (19)), but the authors had been unable to demonstrate an interaction with CGRP or any other ligands. 10 ]Human (h) CGRP␣ (specific activity 2000 Ci/mmol) was obtained from Amersham. hCGRP␣, hCGRP␣-(8 -37), salmon calcitonin (CT), human CT, porcine vasoactive intestinal peptide, and angiotensin II were purchased from Bachem Biochemicals. Cys(ACM2,7)CGRP was obtained from Phoenix Pharmaceuticals (Belmont, CA). Salmon CT- (8 -32) was synthesized at SmithKline Beecham Pharmaceuticals. BCA protein assay kit was obtained from Pierce. All other reagents were obtained from Sigma.

Materials-[2-[ 125 I]iodohistidyl
cDNA Cloning-Expressed sequence tag (EST) analysis (15-17) of cDNA clones derived from a human synovial tissue cDNA library (oligo(dT)-primed and constructed in the ZAPII vector (Stratagene)) identified a 800-bp clone demonstrating significant homology to the human CT receptor. This cDNA clone encoding an incomplete hCGRP 1 receptor was used as a probe to screen an oligo(dT)-primed human lung cDNA library constructed in the ZAPII vector (Stratagene). cDNA library construction and screening were carried out essentially as described (20). Several positive clones were obtained, the longest of which was sequenced to completion by a ABI sequencer (15) (GenBank TM accession number L76380).
Stable Expression in HEK 293 Cells-The CGRP receptor has three potential in-frame ATG start codons; however, the most 3[prime-ATG codon resides in the most favorable Kozak consensus context. For this reason, and to potentially increase protein translation efficiency (21), we prepared a 1.4-kilobase cDNA fragment by PCR amplification that * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EMBL Data Bank with accession number(s) L76380.
ʈ To whom correspondence may be addressed: Dept encompassed the entire CGRP receptor coding region, beginning with the 3Ј-most ATG codon, and subcloned this fragment into the mammalian expression vector, pCDN (22). The oligonucleotide primers used for PCR amplification were 5Ј-G GGG TAC CCC ACC ATG GAG AAA AAG  TGT ACC TCG TAT TTT CTG G-3Ј and 5Ј-CGG GAT CCC GCA AAC  AGT GAG ACA ACC ATC CTT CTA TTT TCA AT-3Ј (the translation  start and stop codons are underlined). Human HEK 293 cells were grown in 100-mm culture dishes and transfected with 10 g of the CDN-CGRP-receptor cDNA using the LipofectAMINE transfection reagent (Life Technologies, Inc.) according to the manufacturer's instructions. After 2 weeks of G418 selection (0.8 mg/ml), colonies were picked and expanded. A single cell line (among many potential candidates) that functionally responded to CGRP treatment (by the generation of cAMP) was chosen for further analysis. A single cell line transfected with pCDN vector alone was selected for use as a control for functional and binding assays. Northern Blot Analysis-Total RNA was isolated from human tissues by TriZOL Reagent (Life Technologies, Inc.). 20 g of total RNA from each tissue was separated by formaldehyde, 1.0% agarose gel electrophoresis and transferred to a nylon membrane. The full-length human CGRP 1 cDNA was labeled and used as probe. Hybridization was done essentially as described (23).
In Situ Gene Amplification Analysis-Under anesthesia, a BALB/C mouse was perfused with 4% paraformaldehyde/phosphate-buffered saline, pH 7.5. Organs were dissected, fixed in 10% buffered formalin at room temperature for 3 days, then embedded in paraffin. Sections of 5 m were prepared. Reverse transcriptase in situ gene amplification was performed as described (24). Two primers used in this study were chosen from regions unique to the human CGRP 1 and very similar to the rat CT-like cDNA sequence (18) receptor, but not significantly conserved in the calcitonin receptor (20): the upstream primer sequence was 5Ј GAC ATC CAG CAA GCA ACA GA 3Ј; the downstream primer sequence was 5Ј CA ATG CCA AGC AAT GGC ACC 3Ј. Digoxigenin-dUTP was applied to the reaction of gene amplification. The products derived from gene amplification were detected by anti-digoxigenin antibody conjugated to alkaline phosphatase (Genius3, Boehringer Mannheim).
Binding Assays-Radioligand binding assays were performed as described (25). In saturation binding studies, increasing concentrations of 125 I-CGRP (8 pM to 150 pM) were added to membranes (ϳ25 g) and incubated in a total volume of 500 l for 60 min, at 25°C. In competition binding studies, the membranes (ϳ30 g of membrane proteins) were incubated with increasing concentrations (1 pM to 1 M) of competing ligands and 60 pM 125 I-CGRP for 60 min at 25°C. Nonspecific binding was defined in the presence of 1 M ␣CGRP and was usually ϳ20% for transfected cell membranes and Ͻ10% for human neuroblastoma SK-N-MC cell (ATCC HTB-10) membranes. No specific binding was observed in membranes of untransfected control 293 cells.
Functional Assays-Transformed or untransformed 293 cells were plated at 2.5 ϫ 10 3 cells/well in 6-well plates. On day 4, the medium was aspirated and the cells were washed with 1 ml of Dulbecco's phosphatebuffered saline containing 0.5 mM isobutylmethylxanthine for 10 min at room temperature. The cells were treated with various concentrations (1 pM to 1 M) of ␣CGRP or related peptides at 37°C for 10 min. The reaction was stopped by addition of 100 l of 100% ice-cold trichloroacetic acid to each well, and cAMP in each well was measured following the radioimmunoassay protocol as described (Advance Magnetics). Each experiment was performed in triplicate and repeated 2-3 times with different passages of cells.

RESULTS AND DISCUSSION
We hypothesized that the CGRP and recently cloned calcitonin (CT) receptors might display close homology since (i) like the G-protein-coupled CT receptor (26), CGRP signaling appears to be mediated through the activation of adenylyl cyclase, (ii) at high concentrations CT interacts with the CGRP binding site (27), and (iii) CGRP (the ␣, not ␤, isopeptide) and CT share the same mRNA transcript and are produced through alternate splicing (1), as are the substance P/K ligands that cross-interact with tachykinin receptors (28). Thus, when expressed sequence tag (EST) analysis (15)(16)(17) of cDNA clones derived from a human synovial tissue cDNA library identified a clone demonstrating significant homology to the human CT receptor, this cDNA was selected for further evaluation. Since the cDNA insert was incomplete (i.e. about 800 bp), it was used as a probe to screen a human lung cDNA library to isolate a clone with a complete open reading frame. Fig. 1 shows the 2995-bp nucleotide sequence and deduced protein of the cloned cDNA. Three potential in-frame ATG codons precede the open reading frame of the protein; however, translation from the second or third ATG codon will encode a protein with a size consistent with that predicted for the cloned Underlined amino acid sequences, putative seven transmembrane domains; potential sites for post-translational glycosylation are marked above the sequence by closed circle; circled amino acids are those that are highly conserved among the secretin/CT/PTH/PTH-RP/GLP1/GHRH/GLU G-protein-coupled receptor subfamily (20, 26, 30 -32); arrow, a putative secretory signal sequence cleavage site; double-underlined ATG codons, in-frame potential initiation codons; underlined AATAAA sequence, consensus polyadenylation signal. rat CT-like "orphan" receptor, with which it shares 91% amino acid sequence homology and thus likely represents a receptor ortholog (18). Also, since the third ATG most closely approximates a Kozak consensus translation initiation site (29), it is probably the translation initiation codon. Consequently, the cDNA encodes a protein of 461 amino acids, sharing several features in common with the G-protein-linked receptors (30,31). Most prominent is the existence of seven hydrophobic regions of 16 to 28 amino acids each, which are likely to be membrane-spanning domains that form a seven-transmembrane motif found among G-protein-coupled receptors. In addition, 52 amino acid residues highly conserved among a recently described subfamily of G-protein-coupled receptors including calcitonin (CT), secretin, parathyroid, glucagon, and other receptors (20,26,(32)(33)(34), are also present, interspersed within the sequence. Among this subfamily, the cloned protein shares its greatest sequence identity, 55.5%, with the human CT receptor. Furthermore, within the N-terminal domain there are several sites for post-translational modification including three asparagine residues within consensus sites for glycosylation and a potential cleavage site of an N-terminal hydrophobic sequence that may be a signal peptide.
Northern blot analysis using the full-length cDNA as a hybridization probe revealed mRNA species of approximately 7.5, 5.5, and 3.5 kb predominantly in the lung and to a lesser degree in the heart (Fig. 2A). RNA dot-blot analysis of over 20 different human tissues indicated widespread, but generally low levels, of receptor mRNA expression (data not shown). In situ analysis of mouse lung and heart tissues demonstrated specific expression of the CGRP 1 receptor mRNA in alveolar cells in the lung and cardiac myocytes in the heart (Fig. 2B). To determine the binding and functional properties of the receptor, we used HEK 293 cells stably transfected with the cDNA subcloned within an expression vector. Membranes prepared from untransfected 293 cells had very little specific 125 I-CGRP binding (data not shown). Membranes prepared from transfected 293 cells displayed high-affinity, low-density binding sites for 125 I-CGRP. The apparent dissociation constant (K d ) and maximum binding (B max ) were 19 Ϯ 3 pM and 86 Ϯ 14 fmol/mg protein, respectively (Fig. 3A). In competition binding studies, the rank order potency for a series of related peptides to inhibit 125 I-CGRP binding to these membranes was CGRP Ͼ CGRP-(8 -37) Ͼ Cys(ACM2,7)CGRP Ͼ adrenomedullin (ADM) Ͼ salmon CT-(8 -32) ϾϾϾ salmon CT Ͼ human CT Ͼ vasoactive intestinal peptide Ͼ angiotensin II (Fig. 3B). The binding affinity of 125 I-CGRP for the recombinant receptor as well as the pharmacological profile of the competing ligands were very similar to that observed for endogenous CGRP receptors present in membranes prepared from human neuroblastoma SK-N-MC cells (Fig. 3C).
Previous studies have shown that CGRP mediates its responses by activation of adenylyl cyclase and the generation of cyclic AMP (4,5). Treatment of vector-transfected 293 cells with CGRP induced less than a 2-fold increase in the accumulation of cAMP (Fig. 4A). Addition of CGRP to 293 cells expressing the recombinant receptor resulted in an increased accumulation of cAMP that was concentration-dependent (Fig. 4B). The threshold, half-maximal, and maximal concentrations of CGRP required to stimulate cAMP accumulation in these cells were 0.1, 0.9, and 10 nM, respectively. The maximal stimulation in response to agonist was 60-fold over basal levels of cAMP. Again, these results were comparable to those achieved using SK-N-MC cells containing endogenous CGRP receptors (35). Treatment of cDNA-transfected 293 cells with increasing concentrations of CGRP-(8 -37), a selective CGRP receptor antagonist, shifted the CGRP concentration response curve for cAMP accumulation to the right in a parallel manner, indicating competitive inhibition with a calculated pA 2 value of 7.57 (Fig.  4C). Except at high concentrations (1 M), the CGRP 2 receptorselective agonist Cys(ACM2,7)CGRP failed to stimulate cAMP in both receptor-transfected cells (Fig. 4B) and SK-N-MC cells (data not shown). The effect of Cys(ACM2,7)CGRP on the recombinant CGRP receptor in inhibiting only 125 I CGRP binding but not a functional cAMP response to CGRP is consistent with the properties of the native rat CGRP 1 receptor present in spleen membranes (36). Thus, the receptor described in the present study is a CGRP 1 receptor.
At high concentrations, ADM evoked a specific response (Fig.  4B), which was not surprising since this agonist has been reported to interact weakly with the CGRP receptor (37). In contrast, human CT does not interact with the recombinant receptor, since this ligand stimulated a nearly identical activity in both CGRP receptor-and vector control-transfected cells (Fig. 4, A and B). Collectively, the pharmacological and functional results confirm that the receptor cDNA encodes a human CGRP 1 receptor that is functionally coupled to the activation of adenylyl cyclase.
Although originally identified as an orphan rat CT-like receptor (18) and, very recently, classified as an orphan human CT-like receptor (despite testing for interaction with CGRP in addition to other ligands (19)), the results presented within this report clearly demonstrate that we have cloned a CGRP 1 receptor. As to the reason why previous investigators were unable to identify the reported CT-like receptor as the CGRP 1 receptor is only a matter of speculation. It is possible that the recombinant CGRP receptor can be functionally expressed only in certain cell types. In this study we used HEK 293 cells, whereas in other reports it appears that only COS and OK cells were used for receptor evaluation. We have recently cloned the porcine CGRP 1 receptor ortholog of the human receptor and found it to exhibit essentially identical pharmacological properties, which provides further supporting evidence that we have identified the CGRP 1 receptor. 2 The availability of the CGRP 1 receptor should facilitate the study of the physiology and pathophysiology of CGRP as a neurotransmitter, neuromodulator, local hormone, and inflammatory mediator.  (8 -37). The antagonist effect was studied by performing CGRP concentration response curves in the absence (f) and presence of 100 (å) and 300 (ç) nM CGRP- (8 -37).