Calcitonin Receptor-stimulating Peptide, a New Member of the Calcitonin Gene-related Peptide Family

We isolated a novel biologically active peptide, designated calcitonin receptor-stimulating peptide (CRSP), from the acid extract of the porcine brain by monitoring cAMP production in the porcine kidney cell line LLC-PK1. Determination of the amino acid sequence and cDNA analysis encoding a CRSP precursor showed that this peptide has ∼60% identity in the amino acid sequence with human calcitonin gene-related peptide type-α (αCGRP), type-β (βCGRP), and porcine CGRP. Northern blot analysis and radioimmunoassay demonstrated that CRSP is expressed mainly in the thyroid gland and the central nervous system, in which the calcitonin receptor was abundantly expressed. Synthetic CRSP elicited a potent stimulatory effect on the cAMP production in LLC-PK1 cells. Although it shows significant sequence similarity with CGRPs, this peptide did not elicit cAMP elevation in cells that endogenously expressed a CGRP receptor or an adrenomedullin receptor or were transfected with either of these recombinant receptors. Administration of CRSP into anesthetized rats did not alter the blood pressure but induced a transient decrease in the plasma calcium concentration. In fact, this peptide potently increased the intracellular cAMP concentration in COS-7 cells that expressed the recombinant calcitonin receptor. These unique properties indicate that CRSP is not a porcine counterpart of βCGRP and probably elicits its biological effects via the calcitonin receptor.

Isolation and sequence determination of new biologically active peptides have greatly advanced our understanding of the communication between cells and tissues. Various biologically active peptides, such as natriuretic peptides (1-3) and neuromedins (4), were isolated by monitoring smooth muscle contraction and relaxation. This strategy was useful, but relatively large amounts of tissue extracts were required to monitor the biological activities. Another strategy involving the measure-ment of intracellular second messenger levels has often been adopted in recent purification studies, and two methods have been successful to date. The first method is to monitor the second messenger levels in intact cultured cells obtained from a particular organ. Pituitary adenylate cyclase-activating peptide and adrenomedullin (AM) 1 were identified by monitoring the adenylyl cyclase activity in the anterior pituitary cells (5) and platelets (6), respectively. The second method is to monitor the second messenger levels in cells expressing a particular orphan receptor. Orexins and ghrelin were purified by monitoring a transient intracellular calcium elevation in HEK293 cells expressing OX 1 R or OX 2 R receptor (7) and growth hormone secretagogue receptor (8), respectively.
We employed the first method to screen new biologically active peptides and measured cAMP production in LLC-PK 1 cells, a cell line prepared from porcine kidney epithelial cells, because this cell line is known to express several peptide receptors that stimulate adenylyl cyclase activity (9). Vasopressin, for example, strongly stimulates adenylyl cyclase activity in LLC-PK 1 cells (10). Among three isotypes of the vasopressin receptor, members of the G-protein-coupled receptor family (11,12), V 2 receptor expressing in LLC-PK 1 cells (10,13) as well as in the kidney is coupled to adenylyl cyclase (13,14) and participates in physiological events such as anti-diuresis. Calcitonin (CT) also potently increases the cAMP production in LLC-PK 1 cells (15). CT is secreted mainly from the thyroid gland and is involved in calcium resorption in the bone (16,17) as well as calcium reabsorption in the kidney (18). The CT receptor is also a member of the G-protein-coupled receptor family, and its cDNA clone was first isolated from a cDNA library constructed from LLC-PK 1 cells (19).
Although CT receptor is expressed in the central nervous system (CNS), expression of CT has not been identified in the CNS (20,21). The CT in the systemic circulation secreted from the thyroid gland is not considered to be incorporated into the CNS beyond the blood brain barrier. Although some investigators report that brain extracts contain biologically active CTlike materials (22)(23)(24), the existence of a cognate ligand to CT receptor in the CNS remains controversial. This was another reason we chose LLC-PK 1 cells to screen biologically active peptides by monitoring the adenylyl cyclase activity.
In this paper, we report the isolation and structural determination of a novel biologically active peptide from porcine brain extract by monitoring the cAMP production in LLC-PK 1 cells. Although the purified peptide showed higher sequence homology to CGRP, this peptide bound to CT receptor with a higher affinity and stimulated cAMP production at a potency 350-fold greater than CT. Thus, we designated this peptide as calcitonin receptor-stimulating peptide (CRSP). The findings obtained in this study strongly suggest that CRSP is a candidate for the unidentified ligand to CT receptor in the CNS and may participate in physiological events by activating the central as well as peripheral CT receptor.
Measurement of Intracellular cAMP Production-Cells were harvested and cultured on 48-well plates for 2 days. The cells were washed twice with DMEM/Hepes (20 mM, pH 7.4) containing 0.5 mM 3-isobutyl-1-methyl xanthine (Sigma) and 0.05% bovine serum albumin (BSA) and incubated in the same medium for 30 min at 37°C. The incubation medium was then replaced with 150 l of medium in which the sample of interest was dissolved and further incubated at 37°C for another 30 min. Aliquots (100 l) of the incubation media were succinylated, evaporated, and then submitted to radioimmunoassay (RIA) for cAMP as reported previously (6).
Isolation of CRSP-Approximately 20 kg of porcine brain was minced and boiled for 10 min in 2 volumes (v/w) of water to inactivate the intrinsic proteases. After cooling, glacial CH 3 COOH was added to a final concentration of 1 M, and the boiled tissue was homogenized with a Polytron homogenizer. The homogenate was centrifuged at 17,700 ϫ g for 25 min, and the supernatant was condensed with a Pellicon cassette (PLAC No. 000 -05, Millipore). The condensate was subjected to acetone precipitation (final concentration ϭ 66%), and the resulting supernatant was evaporated. The resulting solution was divided into four equal volumes, and each was loaded onto a LC-SORB SPW-C-ODS column (1.5 liter, Chemco, Osaka, Japan). After washing with 0.5 M CH 3 COOH, the materials that had adsorbed on the column were eluted with H 2 O:CH 3 CN:trifluoroacetic acid at 40:60:0.1 (v/v). The column eluates were pooled and lyophilized. The dried materials were dissolved in 1 M CH 3 COOH and subjected to batch chromatography on an SP-Sephadex C-25 column (H ϩ -form, 3 ϫ 28 cm; Amersham Biosciences). Successive elutions with 1 M CH 3 COOH, 2 M pyridine, and 2 M pyridine acetate (pH 5.0) yielded three respective fractions of SP-I, SP-II, and SP-III. After lyophilization, the SP-III fraction was separated by gel filtration on a Sephadex G-50 column (fine, 75 ϫ 1450 mm). Fractions corresponding to a relative molecular mass (M r ) of 1000 -5000 were collected and subjected to a second gel filtration on a Sephadex G-25 column (fine, 75 ϫ 1500 mm). Fractions containing peptides of M r 3000 were pooled and subjected to carboxymethyl (CM) ion exchange chromatography (CM-52, 24 ϫ 450 mm; Whatman) eluting with a linear gradient elution of HCOONH 4 (pH 6.5) from 10 mM to 1 M in the presence of 10% CH 3 CN. The fractions eliciting cAMP-producing activity were pooled and re-purified by CM ion exchange high performance liquid chromatography (HPLC) (TSKgel CM-2SW, 7.8 ϫ 300 mm; Tosoh, Japan), eluting with a linear gradient elution of HCOONH 4 (pH 3.8) from 10 mM to 1 M in the presence of 10% CH 3 CN. The biologically active fractions were then separated by reverse phase HPLC on a C 18 column (218TP54, 4.6 ϫ 250 mm; Vydac), and the peptide was finally purified on a diphenyl column (219TP5215, 2.1 ϫ 150 mm; Vydac) using a linear gradient elution of CH 3 CN from 10 to 60% in 0.1% trifluoroacetic acid. The amino acid sequence was analyzed with a Procise cLC protein sequencer (492, Applied Biosystems). After the gel filtration step, aliquots (1/2000) of all fractions in the course of purification were lyophilized, dissolved in 200 l of incubation medium for the cAMP assay, and submitted to the cAMP assay using LLC-PK 1 cells.
Molecular Cloning and Determination of cDNA Sequence of CRSP-Both the N and C termini of CRSP were chosen to synthesize degenerated primers, and the PCR reaction was performed with those primers on porcine genomic DNA. The amplified DNA fragment was sequenced and radiolabeled as a probe for screening. Porcine hypothalamus cDNA was synthesized from 3 g of poly(A) ϩ RNA using a Timesaver cDNA synthesis kit (Amersham Biosciences), inserted into ZAP II bacterio-phage vector (Stratagene), and packaged in vitro using Giga Pack Gold III packaging extract (Stratagene). Approximately 10 5 plaques were screened by the standard protocol. The hybridization buffer contained 50% formamide, 6ϫ SSC (1ϫ SSC ϭ 0.15 M NaCl and 0.015 M sodium citrate), 5ϫ Denhardt's solution, 0.5% SDS, and 10 6 cpm/ml of the radiolabeled probe. After hybridization at 42°C for 16 h, the filters were washed twice with 2ϫ SSC containing 0.1% SDS for 5 min at room temperature and washed twice with 0.2ϫ SSC containing 0.1% SDS at 65°C for 1 h. The filters were exposed to RX-U films (Fuji, Tokyo, Japan), and 12 positive clones were obtained. The isolated positive clones were rescued as pBluescript SK Ϫ using Escherichia coli strain SOLR and helper phage. DNA sequences of these clones were determined by the dideoxynucleotide chain termination method using a DNA sequencer (373A, Applied Biosystems).
Peptide Synthesis-Synthetic CRSP, CRSP-(24 -38), Tyr-CRSP- (24 -38), and Cys-CRSP- (24 -38) were prepared by solid phase techniques using N-(9-fluorenyl)methoxycarbonyl chemistry using a peptide synthesizer (431A, Applied Biosystems). An intramolecular disulfide linkage was formed by the action of K 3 [Fe(CN) 6 ]. Porcine CGRP was prepared by the American Peptide Company. Synthetic peptides were purified by reverse phase HPLC and ion exchange HPLC, and correct synthesis was confirmed by amino acid analysis and sequencing. Salmon and porcine CT were purchased from Peninsula. Human AM was kindly donated by Shionogi & Co. (Osaka, Japan).
Generation of Antisera-All experimental procedures were approved by the local animal experiments and care committee. Rabbit antibodies were raised to a peptide sequence (CRSNLLPTKMGFKVFG-NH 2 ) that corresponds to residues 24 -38 of CRSP (CRSP- (24 -38)) to which an N-terminal cysteine was added to facilitate cross-linking to maleimideactivated keyhole limpet hemocyanin (Pierce). New Zealand White rabbits (Japan SLC, Hamamatsu, Japan) were immunized by injecting 1 mg of the peptide-keyhole limpet hemocyanin antigen conjugate in complete Freund's adjuvant, and antibody production was boosted by 5 additional injections of the antigen in complete Freund's adjuvant at 3-week intervals.
Northern Blot Analysis-A CRSP probe was prepared by a PCR reaction performed on 1 ng of full-length CRSP cDNA using KOD plus polymerase and primers (CTCTCTGAGGAGGAATCACG and GAGT-TCAGAGTCATAGTAACC) for 30 cycles (each cycle consisted of 15 s at 94°C, 15 s at 55°C, and 1 min at 72°C). Total RNAs (30 g) obtained from various tissues were separated on a 1.0% agarose gel containing 2.2 M formaldehyde, 20 mM Mops (pH 7.0), 8 mM CH 3 COONa, and 1 mM EDTA. After electrophoresis, the RNAs were transferred to a Hybond XL nylon membrane (Amersham Biosciences) and baked at 80°C for 2 h. The filter was prehybridized in Ultrahyb hybridization solution (Ambion) for 2 h at 37°C and then hybridized at 42°C for 16 h with the 32 P-labeled CRSP cDNA probe. The filter was washed twice with 2ϫ SSC containing 0.1% SDS at room temperature and twice with 0.1ϫ SSC containing 0.1% SDS at 65°C for 1 h and exposed to RX-U film at Ϫ80°C for 10 days with an intensifying screen.
RIA for CRSP-N-Tyr-CRSP- (24 -38) was radioiodinated by the lactoperoxidase method, and the monoiodinated peptide was purified by reverse phase HPLC. Approximately 1 g of porcine tissue was minced and boiled for 10 min in 5 ml of water. After cooling, water and glacial CH 3 COOH were added to a final volume and concentration of 10 ml and 1 M, respectively, and the boiled tissues were homogenized with a Polytron homogenizer. The homogenate was then centrifuged, and the resulting supernatant was lyophilized and dissolved in the same volume of RIA buffer (50 mM sodium phosphate (pH 7.4) containing 80 mM NaCl, 25 mM EDTA, 0.05% NaN 3 , 0.5% BSA treated with N-ethylmaleimide, and 0.5% Triton X-100). Standard CRSP- (24 -38) or sample dissolved in 100 l of RIA buffer was incubated with 100 l each of monoiodinated N-Tyr-CRSP-(24 -38) (18,000 cpm) and anti-CRSP antiserum at 4°C for 48 h. Then 100 l of 1% ␥-immunoglobulin and 500 l of 23% polyethylene glycol 6000 solution were added to the reaction tubes, and the mixture was incubated on ice for 15 min. After centrifugation at 2,000 ϫ g for 15 min, the supernatant was aspirated, and the radioactivity in the pellet was counted with a ␥ counter (ARC-1000, Aloka, Tokyo, Japan). Aliquots of all the reserved fractions at each purification step were lyophilized, dissolved in the RIA buffer for CRSP, and submitted to RIA for CRSP to evaluate the recovery yield.
Mass Spectrometry-Mass spectra were measured using a single quadrupole mass spectrometer equipped with an electrospray ionization (ESI) source (SSQ 7000, Finnigan). Peptides (10 pmol) were dissolved in a mobile solvent (water:methanol ϭ 50:50, 1% acetic acid) and infused using a syringe pump (5 l/min) coupled directly to the ionization source via a fused silica capillary. Spectra were measured in a multi-channel acquisition mode in the mass range m/z 300 -2000 with scan durations of 3.5 s.

Measurement of Blood Pressure-Eight-week-old male Sprague-
Dawley rats were purchased from Japan SLC. All study protocols were approved by the local animal experiments and care committee. All animals were provided free access to tap water and standard chow. The rats were anesthetized by an intraperitoneal injection of thiobutabarbital sodium salt (60 mg/kg; Wako, Osaka, Japan). Rectal temperature was monitored and maintained at 37 Ϯ 1°C during the entire experimental procedure, and a tracheotomy was performed to aid spontaneous breathing. The mean arterial blood pressure and heart rate were monitored through the right carotid artery using a PE 50 catheter connected to a pressure transducer (model P231D, Gould) and a polygraph (7758 B System, Hewlett-Packard), and the left jugular vein was cannulated using a PE 10 catheter for the bolus administration of peptides or vehicle. The rats were left for at least 30 min after surgery to allow for equilibration. Then 100 l of saline containing the peptide and 0.1% BSA (n ϭ 3) or the same amount of saline containing 0.1% BSA alone as a vehicle (n ϭ 4) was injected through the left jugular vein. Mean arterial blood pressure and heart rate were measured 15 min before and 0, 2, 5, 10, and 30 min after the administration.
Measurement of Plasma Calcium Concentration-The rats were anesthetized as described above. The right carotid artery was cannulated using a PE 50 catheter for blood collection and for the bolus administration of peptides (16 nmol/kg) or vehicle. The rats were left for at least 30 min after surgery to allow for equilibration. Then porcine CRSP, salmon CT, or porcine CT dissolved in 100 l of saline containing 0.1% BSA (n ϭ 4) or vehicle (n ϭ 4) was injected through the right carotid artery. Arterial blood (125 l) was collected in heparinized syringes 15 min before and 15, 30, 60, and 180 min after administration, and the blood volume was replaced with saline (200 l) on each occasion.
Blood samples were analyzed for ionized calcium with an automatic analyzer (ABL 555, Radiometer, Copenhagen, Denmark) immediately after the blood was drawn. At the same time, arterial blood gas parameters (PaO 2 , PaCO 2 , and pH) and hematocrit levels were measured.
Expression of CT Receptor-A partial cDNA clone of porcine CT receptor was amplified by PCR with porcine hypothalamus cDNA. Full-length porcine CT receptor was isolated from a porcine hypothalamus cDNA library synthesized as described above. The full-length clone was ligated into pcDNA 3.1 expression vector (Promega). The CT receptor cDNA ligated into pcDNA (pcDNA/CTR) was transfected into COS-7 cells with LipofectAMINE Plus (Invitrogen) according to the manufacturer's protocol. The transfected cells were used for cAMP assay and competitive binding experiments 24 h after transfection.
Co-expression of Receptor and Receptor Activity-modifying Protein (RAMP) cDNA-Porcine CT receptor, calcitonin-like receptor (CL receptor), and three isoforms of RAMP (RAMP1, -2, and -3) cDNAs encoding their complete open reading frames were isolated from porcine lung and hypothalamus cDNA libraries and ligated into pcDNA 3.1 expression vector (Promega) as described previously (25,31). All combinations of the two cDNAs of one of the two receptors and one of the three RAMPs and insert-free vector (cDNA ratio 1:4) were co-transfected into COS-7 cells with LipofectAMINE Plus. The transfected cells were then used for cAMP assay.
Competitive Binding Experiments-COS-7 cells were plated in 48well plates for transfection of pcDNA/CTR. 125 I-Labeled porcine CT was prepared by the lactoperoxidase method described above. The pcDNA/CTR-transfected cells were washed twice with DMEM/Hepes (20 mM, pH 7.4) and incubated with a binding medium (DMEM supplemented with 20 mM Hepes (pH 7.4), 0.05% BSA, and 0.01% Triton X-100) containing 100,000 cpm of 125 I-labeled porcine CT in the absence or presence of unlabeled peptides at concentrations ranging from 10 pM to 10 M. After incubation for 1 h at room temperature, the cells were washed once with the ice-cold DMEM and lysed and suspended with 0.5 M NaOH. The radioactivity in the lysate was counted with a ␥ counter (ARC-1000).
Statistical Analysis-Statistical analysis was performed by using one way analysis of variance with repeated measurements combined with multiple comparison (Scheffe's F test). These analyses were carried out using SPSS for Windows, Version 6.0.1 (SPSS Inc.). The data are expressed as the mean Ϯ S.E.

RESULTS
Isolation of CRSP-Crude basic peptides (SP-III fraction) extracted from porcine brain were first subjected to Sephadex G-50 gel filtration to remove any remaining proteins. The peptide fraction thus obtained was next separated by Sephadex G-25 gel filtration, and each fraction was assayed by monitoring the cAMP production in LLC-PK 1 cells. Elevation of cAMP  production in LLC-PK 1 cells was detected in the fractions corresponding to M r 2000 -4000. These fractions were pooled and separated by CM-52 cation exchange chromatography, and an aliquot of each fraction was subjected to the cAMP production assay in LLC-PK 1 cells (Fig. 1a). At least five major peaks of stimulatory activity in the cAMP production assay were observed. Among them peaks 1 and 2 were identified as CGRP (peak 2) and its methionine sulfoxide form (peak 1). Although one peptide was purified from peak 3 and found to have an N-terminal sequence identical to that of CGRP, we could not completely determined the structure of this peptide. Peaks 4 and 5 showed different chromatographic behavior from those of known biologically active peptides, and peak 5 was used for the present isolation of CRSP. The biologically active fractions corresponding to peak 5 were further subjected to CM ion exchange HPLC on a TSKgel CM-2SW column, and only one fraction elicited significant biological activity (data not shown). We next applied the biologically active fraction to reverse phase HPLC on a C 18 column, and a single minute peak with absorbance at 210 nm showed compatible cAMP-increasing activity (Fig. 1b, arrow). Final purification was performed by another reverse phase HPLC on a diphenyl column, and the peptide was purified to a homogenous state (Fig. 1c). The yield of each purification step was retrospectively determined by RIA for CRSP and is summarized in Table I. At the CM-52 ion exchange chromatography step, the purification yield was greatly reduced. This reduction was probably due to the long term storage of these fractions.
Amino Acid Sequence of CRSP-Intact and tryptic fragments of the peptide purified from peak 5 were subjected to N-terminal sequence analysis using a protein sequencer (Fig. 2, arrows). Based on the sequence analysis data, its amino acid sequence was deduced up to the 37th residue (Fig. 2), except for the 2nd and 7th residues. The molecular mass of this peptide was estimated to be 4042.0 daltons (Da) from the sequence, assuming that two unidentified residues were cysteines forming an intramolecular disulfide bond. To determine the precise molecular mass of CRSP, the purified peptide was analyzed with an ESI mass spectrometer, and its molecular mass was determined to be 4130.6 Ϯ 0.7 Da. The difference between the molecular mass estimated from the peptide sequence and that determined by mass spectrometry was assumed to be due to oxygens from two methionine sulfoxides (16.0 Da ϫ 2) and a C-terminal glycine amide (57.0 Da), which were hard to identify by protein sequencing.
To determine the cDNA sequence of a precursor for CRSP, we synthesized degenerated primers corresponding to the N and C termini of the sequence, and a PCR reaction was performed with porcine genomic DNA. An amplified fragment was radiolabeled for screening of a hypothalamus cDNA library, FIG. 2. Nucleotide and deduced amino acid sequences of porcine CRSP precursor. Nucleotide and amino acid numbers are shown on the right. Mature amino acid sequence of CRSP is boxed. Arrows indicate the amino acid sequence determined by a protein sequencer. The donor glycine of the C-terminal amide is shaded, and the terminal codon is marked with asterisk. Putative prohormone convertase cleavage sites are double-underlined. Signal sequence (italics) was predicted by PSORT II, a computer program that predicts protein localization sites in animal cells (26).

FIG. 3. Alignment of the amino acid sequences of porcine CRSP and its related peptides.
The amino acid sequence of porcine CRSP (pCRSP) is compared with equine CGRP-I (eCGRPI), porcine CGRP (pCGRP), equine CGRP-II (eCGRPII), human ␣CGRP (h␣CGRP), human ␤CGRP (h␤CGRP), human amylin (hAmylin), porcine CT (pCT), and human adrenomedullin (hAM). Identical residues to CRSP are boxed. Two cysteines in the N-terminal region of CRSP form an intramolecular disulfide linkage. and 12 positive clones were obtained out of 10 5 clones. Fig. 2 shows the nucleotide and deduced amino acid sequences of a CRSP precursor. Analysis of the amino acid sequence of the precursor for CRSP showed that it had typical features of a secretory precursor protein for a biologically active peptide including a signal sequence (Fig. 2, italics), dibasic cleavage sites (double-underlined), and a cleavage/amidation site (gray box). This cDNA sequence supported the existence of glycine amide at the C-terminal of CRSP.
We next chemically synthesized CRSP based on the sequence and structure (C-terminal amidation and intramolecular disulfide linkage) thus determined. Synthetic CRSP or its methionine sulfoxide form elicits chromatographic and mass spectro-metric behavior quite similar to that of the native peptide.
Both mature and prepro-CRSP sequences were utilized in a BLAST search of the DDBJ/GenBank TM /EBI protein database, and this peptide and its precursor were found to show sequence similarity with CGRP and its related peptides (Fig. 3). To further find out a corresponding peptide and its precursor in other species, we searched the EST database and identified one candidate gene in horse in addition to horse CGRP. The candidate gene had 77% identity with porcine CRSP, and the amino acid sequence from the C terminus was highly conserved (see "Discussion"). In human and mouse, no corresponding gene has been identified that has significant similarity with CRSP except for ␣and ␤CGRP and amylin in both databases.
Tissue Distribution of CRSP-Messenger RNA levels for CRSP in various porcine tissues were estimated by Northern blot analysis (Fig. 4). Approximately 1.0 kilobase of CRSP mRNA was detected predominantly in the CNS, particularly in the hypothalamus and pons/medulla oblongata, but a faint band was observed in the spinal cord, where CGRP was abundantly expressed (24). Another strong band was detected in the thyroid gland, although no band was observed in other peripheral tissues.
To measure the concentrations of CRSP in various tissues, we raised a polyclonal antibody against CRSP- (24 -38). Both CRSP-(24 -38) and CRSP-(1-38) elicited comparable affinity to this antibody. Half-maximal inhibition of radioiodinated ligand binding to the antibody by these peptides was observed at 10 fmol/tube, and no significant cross-reactivity was observed with porcine CT and CGRP (Ͻ0.001%). Table II shows the concentrations of immunoreactive (IR) CRSP in porcine tissues. The posterior pituitary was found to contain the highest concentration of IR-CRSP (96 Ϯ 15 pmol/g wet weight). IR-CRSP was also abundant in the CNS, particularly in the midbrain, thalamus and hypothalamus (Ͼ 3.0 pmol/g wet weight). In the peripheral tissues, the thyroid gland contained an extremely high concentration of IR-CRSP (68 Ϯ 39 pmol/g wet weight), but other peripheral tissues contained low levels of IR-CRSP (Ͻ0.5 pmol/g wet weight).
Effects of CRSP on cAMP Production in Various Cells-We next measured the effects of CRSP on cAMP production in LLC-PK 1 cells along with porcine CT, porcine CGRP, and hu-  man AM (Fig. 5a). This cell line was reported to express both CT receptor and vasopressin V 2 receptors that elevate the cAMP production (13,14,19). CRSP stimulated the production of cAMP in LLC-PK 1 cells in a dose-dependent manner and more potently than any other peptides. To examine the effect of CRSP on other cultured cells, this peptide was administered to Hs68 and Swiss 3T3 fibroblasts, in which cAMP production was reported to be stimulated predominantly by CGRP and AM, respectively (27). Although CGRP and AM increased cAMP concentration in Hs68 and Swiss 3T3 cells, CRSP did not alter the cAMP level in either cell line (Figs. 5, b and c). Systemic Pharmacological Effects of CRSP-Because CRSP shares structural similarity with CGRP, CRSP was expected to elicit a vasodilatory effect comparable with that of CGRP. CRSP was injected into anesthetized rats up to 16 nmol/kg, and the blood pressure was measured. However, no significant decrease in the blood pressure was observed by bolus administration of CRSP. In the same system, CGRP induced a potent depressor effect (Fig. 6a).
We next examined the effects of CRSP on the plasma calcium levels and compared them to those of salmon and porcine CT (Fig. 6b). Bolus administration of CRSP to anesthetized rats at a dose of 16 nmol/kg resulted in a significant decrease in the plasma calcium level. The plasma calcium-reducing effect of CRSP was detectable up to 1 h and almost diminished 3 h after injection. On the other hand, injection of salmon or porcine CT caused a sustained reduction in the plasma calcium level even 3 h after injection.
Effects of CRSP and Its Related Peptides on Recombinant CT Receptor-Porcine CT receptor was transiently expressed in COS-7 cells, and we evaluated the effect of CRSP and its related peptides on these cells. In Fig. 7a, a dose-dependent production of cAMP was observed by stimulation of CRSP, salmon CT, porcine CT, porcine CGRP, and human AM. The production of cAMP was most strongly enhanced with salmon CT (a median effective dose (ED 50 ) ϭ 0.03 nM), which is known to stimulate mammalian CT receptor more potently than the respective endogenous CT. On the other hand, CRSP stimulated the CT receptor (ED 50 ϭ 0.20 nM) with a potency more than 350 times higher than that of porcine CT (ED 50 ϭ 71 nM), although its effect was weaker than that of salmon CT. None of these peptides stimulated the cAMP production in mock-transfected COS-7 cells (data not shown).
To investigate the binding activity of CRSP to CT receptor, 125 I-labeled porcine CT and 10 pM to 10 M CRSP, salmon CT, or porcine CT as a competitor were incubated with COS-7 cells that transiently expressed the CT receptor (Fig. 7c). The binding of 125 I-labeled porcine CT to CT receptor in the COS-7 cells was abolished by the addition of an excess amount of each peptide. The median inhibition concentration (IC 50 ) of CRSP in this binding assay system was found to be 1.3 nM, which was ϳ2-fold greater than that of salmon CT (0.61 nM).
Effect of CRSP on CT Receptor or CL Receptor with or without Co-transfection of RAMP-A recent report revealed that AM and CGRP receptors are composed of CL receptor and RAMPs with a single transmembrane domain protein, although CL receptor belongs to the family B of G-protein-coupled receptors. The CGRP receptor consisted of CL receptor and RAMP1, whereas the AM receptor consisted of CL receptor and RAMP2 or RAMP3 (28). High affinity amylin receptor was also shown to be constituted by co-expression of CT receptor and RAMPs (29,30). On the basis of these findings, CT receptor or CL receptor was expressed in COS-7 cells with one of the RAMPs or a blank expression vector and stimulated with 100 nM CRSP, porcine CT, CGRP, or human AM to examine through which receptor-RAMP complex CRSP can stimulate cAMP production. Porcine CT stimulated the cAMP production more than 10-fold in the COS-7 cells expressing CT receptor with or without RAMP but not in those expressing CL receptor (data not shown). Porcine CGRP or human AM stimulated the cAMP production ϳ5-fold in the COS-7 cells expressing CL receptor with RAMP1, -2, or -3, respectively (data not shown). These results were compatible with those reported previously  (25). On the other hand, ϳ10-fold enhancement of the cAMP production was observed by stimulation of CRSP in the COS-7 cells expressing CT receptor with or without RAMPs, whereas this peptide did not alter the cAMP level in the COS-7 cells expressing CL receptor with or without RAMPs (Fig. 7b). DISCUSSION In the present study, we isolated a novel peptide, CRSP, from a porcine brain extract by monitoring the cAMP production in LLC-PK 1 cells. Amino acid sequence analysis and molecular cloning of CRSP and its precursor protein showed that this peptide shows significant sequence similarity to CGRP. CGRP is a potent vasodilatory peptide, and two isoforms have to date been reported in humans and rodents (31)(32)(33)(34). CGRP type-␣ was generated from tissue-specific alternative splicing of mRNA transcribed from the CT gene (32). On the other hand, ␤CGRP gene encodes a CT-like sequence that is not transcribed into mRNA (34). Both genes have striking sequence similarity to each other. Because one form of CGRP corresponding to ␣CGRP has been identified in the pig, we first considered that CRSP was a porcine counterpart of the ␤CGRP. However, this peptide has a distinct structure, function, and distribution compared with those of ␣and ␤CGRP as described below. Compared with the amino acid sequence of porcine CGRP, this peptide has 90% identity in the 10-amino acid sequence from the N terminus, and two cysteines that are deduced to form an intramolecular disulfide bond are conserved. On the other hand, only 40% identity between CRSP and CGRP was observed in the 10-amino acid sequence from the C terminus. In contrast, the amino acid sequences of human and porcine CGRPs are highly conserved throughout the molecule (Fig.  3). CRSP stimulated cAMP production in LLC-PK 1 cells about 300-fold more potently than porcine CGRP (Fig. 5a). CRSP also stimulated the cAMP production in COS-7 cells expressing recombinant CT receptor in a manner similar to that in LLC-PK 1 cells. On the other hand, this peptide did not augment the cAMP production at all in either Hs68 or Swiss 3T3 cells, in which it was stimulated by CGRP and adrenomedullin. Both ␣and ␤CGRP are reported to be abundantly expressed in the rat spinal cord (24). On the other hand, Northern blot analysis data shows that CRSP is expressed predominantly in the hypothalamus and poorly in the spinal cord (Fig. 4). Although CRSP has structural similarity to CGRP, several lines of evidence listed above indicate that CRSP has characteristics distinct from those of CGRP and elicits its effects via the CT receptor. Thus, we designated this peptide as calcitonin receptor-stimulating peptide (CRSP).
By searching CRSP or related peptides in the EST database, one candidate peptide, named CGRP-I, was found in horse. In horse, another peptide named CGRP-II has been identified, which has 97, 87, and 89% identity with porcine and human ␣and ␤CGRP, respectively, and this peptide is considered to be a counterpart of CGRP. The equine CGRP-I gene encodes a peptide having 77% identity with porcine CRSP, which is ϳ10% higher than that with porcine CGRP (67%). In particular, the sequence identity between equine CGRP-I and porcine CRSP is 80% in the 10-amino acid sequence from the C terminus, whereas that between equine CGRP-I and porcine CGRP is only 30%. The high identity in the N-terminal region and low identity in the C-terminal region is one of the features that differentiates CRSP from CGRP as described above. Because no corresponding gene has so far been found in humans and rodents by database searching, the peptides having high identity with CRSP might be limited to species evolutionarily close to pig. Further biochemical and pharmacological analyses using synthetic equine CGRP-I should be performed to determine whether it actually elicits CRSP-like effects.
CGRP, CT, AM, and amylin, belonging to the CT and CGRP superfamily, are widely distributed in the CNS as well as in the peripheral tissues and induce multiple biological effects, including vasodilation, calcium resorption in bone, and reduction in nutrient intake (31). However, their receptor systems are more complex than other biologically active peptides. Recent studies have verified that the functional AM receptor and CGRP receptor consist of two membrane proteins, CL receptor, and one of three RAMPs (30). CGRP is considered to stimulate cAMP production via a CL receptor-RAMP1 complex (30), whereas AM is shown to increase the intracellular signal via a CL receptor-RAMP2 or a CL receptor-RAMP3 complex (28). Furthermore, later studies showed that the expression of RAMP3 facilitated amylin to bind and activate CT receptor (29,30). Although CRSP was found to stimulate CT receptor, these previous data indicated that the CRSP receptor should be characterized in detail in endogenous as well as recombinant systems. As shown in Fig. 5, CGRP and AM predominantly stimulated cAMP production in Hs68 and Swiss 3T3 cells, respectively, but CRSP did not alter the cAMP concentration in either of these cell lines. As clearly shown in Fig. 7b, CRSP did not stimulate cAMP production at all via porcine CL receptor in the presence or absence of porcine RAMPs in the COS-7 cell expression system. These findings demonstrate that CRSP does not stimulate endogenous and recombinant AM/CGRP receptors, although this peptide has high structural similarity to CGRP. However, CRSP stimulated cAMP production in the COS-7 cells expressing the recombinant CT receptor more potently than porcine CT (Fig. 7a), and co-expression of RAMPs did not alter the cAMP production level induced by CRSP stimulation (Fig. 7b). Binding of 125 I-labeled porcine CT to the COS-7 cells expressing the recombinant CT receptor was prominently displaced by CRSP (Fig. 7c). The evidence described above indicate that both CRSP and CT can stimulate the CT receptor, although a detailed analysis of ligand-receptor interaction is required to understand how two structurally distinct peptides can bind to the same receptor.
Bolus injection of CRSP into rats transiently reduced the plasma calcium concentration, although its effect was weaker than that of salmon and porcine CT (Fig. 6). At least two mechanisms have been elucidated for the plasma calcium reduction. 1) CT receptor is abundantly expressed on osteoclast cells (35). Treatment of resorbing bones with CT increases loss of the ruffled border of the osteoclast cells and decreases the release of lysosomal enzymes from them, which results in a reduction in the plasma calcium level (16,17). 2) Renal tubular epithelia express CT receptor. The excretion of plasma calcium through the kidney is stimulated by CT by inhibiting renal tubular calcium reabsorption (18). As both osteoclast and renal epithelial cells express CT receptor, CRSP is deduced to activate CT receptor on these cells and decrease the plasma calcium concentration. In contrast to the potent effects of CRSP on cAMP production on LLC-PK 1 cells and COS-7 cells expressing CT receptor, the plasma calcium-reducing activity of CRSP is much weaker than that of porcine CT. The difference in the biological potency of CRSP in the in vivo and in vitro system may indicate the presence of unidentified physiological effects and mechanisms distinct from those of CT.
The CT receptor is also expressed in the CNS at a high level (36) in addition to the peripheral system. Although CT receptor in the CNS has been recognized to be involved in the regulation of appetite (37), gastric acid secretion (38), and analgesic effect (39), endogenous ligands for the CT receptor in the brain remain unidentified. In fact, CT is mainly synthesized and secreted from the thyroid gland but is almost undetectable in the CNS. The CT secreted from the thyroid and circulating in the blood is not considered to be incorporated into the CNS beyond the blood brain barrier. Several attempts have been made to identify an endogenous ligand that stimulates cAMP production via CT receptor in the CNS. Fischer et al. (22) reported that human calcitonin-and C-terminal adjacent peptide-like immunoreactivities were found in extracts of the human periventricular mesencephalic region. Sexton and Hilton (23) detect salmon CT-like immunoreactivity in rat brain using an anti-salmon CT antibody. This finding was supported by CT bioassays including cAMP production and receptor binding to culture cells and brain membranes. In 1998, these authors reported the successful purification of brain CT-like peptide with a molecular mass of 3267 Da, which was N-terminally blocked but had a unique 6-amino acid sequence of EKSQSP in the molecule (24). Unfortunately, no subsequent study has been reported for this peptide. In our study, the structure of CRSP was determined by direct analysis of the peptide purified from porcine brain extracts, and the sequence of prepro-CRSP was deduced from cDNA cloning and analysis. The CT-like biological activity was confirmed by the data showing that synthetic CRSP stimulated cAMP production via recombinant CT receptor and reduced the rat plasma calcium concentration by bolus administration. These findings suggest that this pep- tide with low amino acid sequence similarity with CT is a cognate ligand for CT receptor in the CNS, although it is necessary to examine whether a unique CRSP receptor that is more specific than CT receptor is present or not. Identification of CRSP in other species as well as the elucidation of its physiological effects and expression profiles in the CNS and peripheral system will clarify the relationship between CRSP and the previously reported CT-like peptide.