The EMBO Journal vol. 1 0 no. 7 pp. 1 635 - 1641, 1991

Molecular cloning and expression of a cDNA encoding the

Takeshi lshiharal,2, Shun Nakamura3, Yoshito they act not only as hormones, but also as Kaziro4, Takayuki Takahashi2, Kenji Takahashi2 or neuromodulators (Gozes and Brenneman, 1989). and Shigekazu Nagatal The 27 amino acid rat secretin is synthesized as a precursor protein of 134 amino acids in specific endocrine cells, S cells, 'Osaka Bioscience Institute, 6-2-4 Furuedai, Suita, Osaka 565, in the mucosa of the small intestine (Kopin et al., 1990). 2Department of Biophysics and Biochemistry, Faculty of Science, The for the secretin was University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, 3National transcript precursor also detected Institute of Neuroscience, NCNP, 4-1-1 Ogawa-Higashi-machi, in rat hypothalamus, brain stem and cortex, although the Kodaira, Tokyo 187, Japan and 4DNAX Research Institute of expression level was far less than that in small intestine Molecular and Cellular Biology, 901 California Avenue, Palo Alto, (Kopin et al., 1990). In gut, secretin works as hormone CA 94304, USA which stimulates secretion of enzymes and electrolytes. In Communicated by C.Weissmann brain and adrenal pheochromocytoma PC-12 cells, secretin regulates tyrosine hydroxylase which is the rate-limiting Secretin is a 27 amino acid peptide which stimulates the enzyme in the biosynthesis of catecholamines, suggesting secretion of bicarbonate, enzymes and potassium ion secretin may also work as a neuromodulator (Roskoski et from the pancreas. A complementary DNA encoding the al., 1989). rat secretin receptor was isolated from a CDM8 Diverse functions of peptide hormones of the secretin expression library of NG108-15 cell line. The secretin family are mediated by the interaction with their respective receptor expressed in COS cells could specifically bind receptors (Rosselin, 1986). Since secretin, VIP and glucagon the iodinated secretin with high and low affinities. Co- increase intracellular cAMP, and GTP regulates the expression of the secretin receptor with the a-subunit of interaction of hormones with their receptors, it was rat Gs protein increased the concentration of the high postulated that receptors for these hormones are coupled to affinity receptor in the membrane fraction of the the cyclase-stimulated G protein (Gs) (Roth et al., 1984; transfected COS cells. Secretin could stimulate Fremeau et al. ,1986). However, except for the VIP receptor accumulation of cAMP in COS cells expressing the cloned (Couvineau et al., 1990), the purification and biochemical secretin receptor. The nucleotide sequence analysis of the characterization of receptors for these peptide receptors were cDNA has revealed that the secretin receptor consists of hampered by the low abundance of the receptors. In the case 449 amino acids with a calculated Mr of 48 696. The of secretin receptor, high and low affinity-secretin binding secretin receptor contains seven putative transmembrane sites were found in the membrane fraction from rat pancreatic segments, and belongs to a family of the G protein- acini (Bissonnette et al., 1984). Chemical cross-linking coupled receptor. However, the amino acid sequence of analyses of the receptor with [1251]secretin indicated that the the secretin receptor has no significant similarity with that secretin receptor in rat gastric glands (Bewab et al., 1988) of other G protein-coupled receptors. A 2.5 kb mRNA and pancreatic acini (Gossen et al., 1989) have Mr values coding for the secretin receptor could be detected in of 62 000 and 51 000, respectively. NG108-15 cells, and rat heart, stomach and pancreatic In this report, we have isolated a complementary DNA tissue. encoding rat secretin receptor from NG108-15 cells. Secretin Key words: cAMP/expression cloning/G protein/receptor/ bound specifically to COS cells transfected with the cloned secretin cDNA and stimulated the intracellular accumulation of cAMP. The nucleotide sequence analysis of the cDNA has revealed that the secretin receptor is a new type of the G Introduction protein-coupled receptor with seven transmembrane segments. Secretin was discovered in 1902 by Bayliss and Starling in the duodenal mucosa as a hormone which stimulates the secretion of bicarbonate, enzymes and potassium ions from Results the pancreas (Bayliss and Starling, 1902). Vasoactive Expression cloning of the secretin receptor intestinal peptide (VIP), peptide histidine isoleucine (PHI), To identify the cDNA clone encoding secretin receptor, we glucagon, growth hormone releasing factor (GRF) and utilized a direct expression cloning strategy, which has been gastric inhibitory peptide (GIP) are related to secretin, and used to isolate various cytokine receptor cDNAs (Munro and constitute a family of peptide hormones (Rosselin, 1986; Maniatis, 1989; Fukunaga et al., 1990). In this method, a Gozes and Brenneman, 1989). They have various cDNA library constructed with a mammalian expression pharmacological effects on a variety of tissues including vector was expressed in COS cells, and the transfected cells pancreas, liver, heart, intestine and kidney (Robberecht et were subsequently assayed for the ability to bind radioactive al., 1990). In addition, since these hormones can be found ligands. Secretin stimulates adenylate cyclase through the in the central and peripheral nervous system, it is likely that interaction of its receptor with Gs protein (Roth et al., 1984;

© Oxford University Press 1635 T.ishihara et al.

Fremeau et al., 1986). In f3-adrenergic systems, association a of the receptor with Gs results in an increase in the affinity of the receptor for ligands (Cerione et al., 1984). We screened the secretin receptor cDNA by its direct expression w in a COS cell line (COSGsl) which overexpresses the oa- w subunit of the Gs-protein (Gsa), in the hope that the receptor would bind [1251]secretin with a higher affinity than in LL. normal COS cells. zm A cDNA library was constructed with the CDM8 vector 0 (Seed, 1987) using poly(A) RNA from NG108-15 cells that express a relatively large number of the secretin receptor (Roth et al., 1984). Plasmid DNA from pools of 750-1000 individual transformants was transfected into COSGsl cells. Expression of the secretin receptor was screened by 0.00 L- incubating cells with [1251]secretin, followed by exposure to o 50 100 150 200 an IP plate (Fuji BAS 2000 system) and image analysis. This BOU ND(pM) method reduced the exposure time about 50 times compared with the normal autoradiography using an X-ray film. After b screening 400 pools ( - 350 000 clones) in this manner, we identified a pool which consistently gave a positive signal. About 3000 independent clones from the positive pool were then subjected to sib selection to isolate the pQ17 clone. uJlU Binding characteristics of the cloned receptor To confirm that the cloned cDNA encoded the secretin receptor, the pQ17 cDNA was expressed in COS cells with z or without rat Gsa cDNA (Itoh et al., 1986). [1251]secretin 0 bound to the membranes prepared from COS cells co- m transfected with pQ17 and a mammalian expression vector pEF-BOS (Mizushima and Nagata, 1990) in a saturating manner, whereas no specific binding of secretin was observed with the membranes from untransfected COS cells. 0 20 40 60 80 A Scatchard analysis of the binding data revealed the presence of two classes of binding sites with apparent BOUND(pM) dissociation constants (Kd) of 0.57 nM and 20.1 nM (Figure la), which were similar to the values obtained using the Fig. 1. Scatchard analysis of [1251]secretin binding to membranes from membranes from NG108-15 cells (0.44 nM and 14.5 nM) COS cells transfected with the secretin receptor cDNA and NG108-15 (Figure lb). The concentration of high and low affinity cells. Membrane fractions from COS7 cells or NG108-15 cells were binding sites in membranes from COS cells was 0.022 pmol, incubated with various concentrations of [1251]secretin with or without an excess of unlabeled secretin as described in Materials and methods. and 1.22 pmol/mg protein, respectively. When COS cells The binding data were analyzed using the MacLigand Program were co-transfected with pQ17 and pEF-BOS-Gs, the (Munson and Rodbard, 1980). (a) Scatchard plot of [1251]secretin concentration of high affinity binding sites for secretin in binding data with membranes from COS cells. COS cells were the membranes increased - 10-fold (0.21 pmol/mg protein), cotransfected with pQ17 and pEF-BOS (0) or with pQ17 and pEF- BOS-Gs (O). (b) Scatchard plot of [1251]secretin binding data with while low affinity binding sites remained at approximately membranes from NG108-15 cells. the same level (Figure la). These results indicate that the secretin receptor associates with the Gsa protein to form the high affinity binding site. Intracellular accumulation of cAMP mediated by the The binding specificity of the cloned secretin receptor was cloned secretin receptor then examined. Figure 2a shows the ability of various To examine whether the cloned secretin receptor expressed unlabeled peptides to compete with the binding of in COS cells could transduce the signal, stimulation of the [1251]secretin to the transfected COS cell membrane. The adenylate cyclase activity by secretin was studied (Figure results indicated that secretin is the most potent agent in 3). Secretin stimulated the accumulation of cAMP in displacing the binding of [1251]secretin. The C-terminal COSGsl cells transfected with pQ17, whereas COSGsl cells peptide of secretin (secretin5-27) and VIP were - 1000 times did not respond to secretin in order to accumulate less potent than secretin in inhibiting the binding of intracellular cAMP. The response was dose-dependent, and [125I]secretin to the membrane. Glucagon had a very weak the half-maximal response was obtained with 1 nM secretin, ability to compete with the binding of secretin to the secretin which was very close to the value obtained with NG108-15 receptor expressed in COS cells. The apparent half-maximal cells (data not shown). VIP functioned as an agonist to concentrations for the inhibition (IC50) of secretin, stimulate adenylate cyclase activity in the pQ17-transfected secretin5-27 and VIP were 4.5 nM, 9.5 ,tM and 4.5 jiM, COSGsl cells, but -60-fold molar excess of VIP than respectively. These values agreed well with the results secretin was needed to show the half-maximal response obtained with membranes from NG108-15 cells (Figure 2b) (Figure 3). These results agree with the previous report for and guinea pig pancreatic acini (Jensen et al., 1983). the native secretin receptor using rat gastric glands (Gespach 1636 Molecular cloning of a secretin receptor cDNA

400 0

0 E 300 Z 80 0., 0~ 60 w 200 Q 40 -J Li- -J 100 0 20 I- Lz

07' 10 1010 109 0-8 0'7 -6 10-5 10 4 z PEPTIDE CONCENTRATION(M) 1 o-10 109 1 o-8 10'7 1 o-6 10'5 10'4 PEPTIDE CONCENTRATION(M)

b 120 l o * Fig. 3. Stimulation of cAMP production in COSGsl cells expressing the rat secretin receptor by secretin, VIP and glucagon. COSGsl cells (D 100 transfected with pQ17 were incubated for 45 min with various z concentrations of secretin (0), VIP (E) or glucagon (A). As a control, COSGsl cells transfected with the CDM8 vector were treated z 80 with 10 yM secretin (0), VIP (-) or glucagon (A). The intracellular amount of cAMP was measured as described in Materials and methods and is expressed per 1 x 105 cells. 60

40 Heijne, 1986), the 23rd amino acid (Ala) from the initiation codon was tentatively assigned as the N-terminal amino acid LL 0 of the mature protein. Thus, the mature secretin receptor 20 may consist of 427 amino acids with a calculated Mr of 48 696, which is 2-13 kd smaller than the value estimated for 07/ 10110 i09 10-8 10-7 10-6 10-5 1i04 the secretin receptor in rat gastric glands or pancreatic acini by cross-linking studies (Bewab et al., 1988; Gossen et al., PEPTIDE CONCENTRATION(M) 1989). The difference is probably due to the glycosylation at some of the 5 potential N-glycosylation sites (Asn-X- Fig. 2. Displacement of [125l]secretin binding by secretin, secretin5-27, Thr/Ser) found in the putative extracellular domain of the VIP and glucagon. The binding of [1251]secretin to membranes was secretin determined in the presence of indicated concentrations of secretin (0), receptor (Figure 4a). secretin5-27 (0), VIP (O) and glucagon (A). (a) Displacement of The NG108-15 cell line used to prepare the cDNA library [1251]secretin binding to membranes from COS7 cells cotransfected is a hybrid cell line between mouse N18TG neuroblastoma with pQ17 and pEF-BOS-Gs. (b) Displacement of ['251]secretin and rat C6Bu-1 glioma cell lines (Klee et al., 1974). To binding to membranes from NG108-15 cells. determine the origin of the secretin receptor cDNA (pQ17), a set of oligonucleotides (nucleotides from positions et al., 1986), and confirm that the pQ17 cDNA codes for 1476-1495 for the forward primer and 1705-1724 for the a functional secretin receptor. reverse primer) was synthesized, and genomic DNAs from rat and mouse were analyzed by polymerase chain reaction Primary structure of secretin receptor (PCR). Amplification of rat and mouse DNAs gave a single Figure 4a shows the complete nucleotide sequence (1796 discrete band of about 250 nucleotides (data not shown), and nucleotides long) and the deduced amino acid sequence (449 the partial nucleotide sequences of these DNA fragments amino acids) for the secretin receptor. There are two were determined after subcloning into pUC 119. The DNA potential AUG initiation codons at amino acid positions -22 fragment amplified from the rat genomic DNA had a and -18 (nucleotide positions 213-215 and 225-227, completely identical nucleotide sequence to that of pQ17 respectively) in the same reading frame. Although the cDNA while the nucleotide sequence of the DNA fragment neighboring sequence of the second AUG agrees better with from mouse genomic DNA had a 17% mismatch with that the consensus sequence (CCA/GCCAUGG) proposed by of pQ17. These results indicate that the secretin receptor Kozak (1987), the first AUG from the 5' terminus was mRNA corresponding to the pQ17 cDNA was derived from tentatively assigned as the initiation codon for the secretin the gene on rat genome present in the NG108-15 hybrid cell receptor. A hydropathy plot of the amino acid sequence of line. the secretin receptor (Figure 4b) indicated the presence of a signal sequence at the N-terminal and predicted seven Tissue distribution of secretin receptor mRNA membrane-spanning segments each consisting of 20-24 Secretin has pharmacological effects not only in the pancreas amino acids (Figure 4a). By comparing the 5' portion of but also in other tissues including kidney, intestine, heart the sequence with the typical signal peptide cleavage site (von and brain (Christophe et al., 1984; Roth et al., 1984; 1637 T.Ishihara et al.

50 a TTTGTTAACAGTCTTGCCACCTCCTTCAGGGAGCAGCAGCCTCAAGCCTGGCTGTGGATCTGATCGGACTGGGTAGGTCCAGCCATGGTCTCT

100 150 200 CTGGGCTTGGAGGTCGCCCAGGGTCCTGGACCACGTGCAGGTCGCTGAGCTGGACGCAGAACCGGCCACCAAGCTCCTGAAGGGGCCAGAACGCAAAGGGGAGCAACAGTGGGCGGAGC 250 300 ATG CTC AGC ACC ATG AGA CCC CGG CTG TCC CTG CTT CTG CTG AGG CTG TTG CTG CTC ACC AAA GCA GCG CAC ACC GTT GGA GTT CCG CCC Met Leu Ser Thr Met Arg Pro Arg Leu Ser Leu Leu Leu Leu Arg Leu Leu Leu Leu Thr Lys Ala Ala His Thr Val Gly Val Pro Pro -20 -1 1 350 CGA CTG TGT GAC GTG CGG CGT GTG CTG CTT GAG GAA CGG GCC CAC TGC CTA CAG CAA CTC TCC AAA GAG AAG AAA GGA GCA CTG GGC CCG Arg Leu j Asp Val Arg Arg Val Leu Leu Glu Glu Arg Ala His M Leu Gln Gln Leu Ser Lys Glu Lys Lys Gly Ala Leu Gly Pro 20 400 450 GAG ACA GCA TCA GGC TGT GAG GGG CTT TGG GAC AAC ATG AGC TGC TGG CCC TCC TCT GCA CCA GCA CGG ACA GTG GAA GTG CAG TGT CCG Glu Thr Ala Ser Gly M Glu Gly Leu Trp Asp Asn Met Ser E Trp Pro Ser Ser Ala Pro Ala Arg Thr Val Glu Val Gln M Pro 40 * 60 500 550 AAG TTC CTC CTG ATG CTC TCA AAC AAA AAT GGG TCT CTG TTC CGG AAC TGC ACA CAA GAT GGA TGG TCA GAA ACC TTC CCC AGG CCT GAC Lys Phe Leu Leu Met Leu Ser Asn Lys Asn Gly Ser Leu Phe Arg AsnM Thr Gln Asp Gly Trp Ser Glu Thr Phe Pro Arg Pro Asp * 80 * 600 650 CTG GCC TGT GGG GTC AAC ATA AAC AAC TCC TTT AAT GAG AGA CGG CAT GCG TAC CTG CTG AAA CTC AAG GTC ATG TAC ACT GTA GGC TAC Leu Ala M Gly Val Asn Ile Asn Asn Ser Phe Asn Glu Arg Arg His Ala Tyr Leu Leu Lys Leu Lys Val Met Tyr Thr Val Gly Tyr 100 * 120 700 750 AGT TCC TCC TTG GCC ATG CTC CTG GTC GCC CTC AGC ATC CTG TGC TCT TTC CGG AGG CTG CAC TGC ACT CGC AAC TAC ATC CAC ATG CAC Sor Ser Ser Lou Ala Met Lou Lou Vol Ala Lou Ser 11e Lou Cys Ser Phe Arg Arg Leu His Cys Thr Arg Asn Tyr Ile His Met His 1 40 800 CTG TTC GTG TCC TTC ATC CTG CGC GCC CTG TCC AAC TTC ATC AAG GAC GCC GTA CTC TTC TCC TCG GAC GAC GTC ACC TAC TGT GAT GCC Leu Phe Val Ser Phe Ile Leu Arq Ala Leu Ser Asn Phe Ile Lys Asp Ala Val Leu Phe Ser Ser Asp Asp Val Thr Tyr Cys Asp Ala 160 180 850 900 CAT AAG GTG GGC TGC AAG CTG GTC ATG ATC TTC TTC CAG TAC TGC ATC ATG GCC AAC TAC GCA TGG CTG CTG GTG GAG GGC CTC TAC CTT His Lys Val Gly Cys Lys Leu Val Met Ile Phe Phe Gln Tyr Cys Ile Met Ala Asn Tyr Ala Trp Leu Leu Val Glu Gly Leu Tyr Leu 200 950 1000 CAC ACA CTC CTT GCC ATC TCC TTC TTC TCA GAA AGG AAG TAC CTG CAG GCC TTT GTG CTC CTC GGA TGG GGT TCT CCA GCC ATT TTT GTT His Thr Leu Leu Ala Ile Ser Phe Phe Ser Glu Arg Lys Tyr Leu Gln Ala Phe Val Leu Leu GlY Trp Gly Ser Pro Ala Ile Phe Val 220 24U 1050 1000 GCT TTG TGG GCT ATC ACC AGG CAC TTT CTA GAA AAT ACT GGA TGC TGG GAC ATC AAC GCC AAT GCT TCC GTC TGG TGG GTC ATT CGA GGG Ala Leu Trp Ala Ile Thr Arg His Phe Leu Glu Asn Thr Gly Cys Trp Asp Ile Asn Ala Asn Ala Ser Val Trp Trp Val Ile Arcg GlY 260 * 1150 1200 CCT GTG ATT CTG TCC ATC CTG ATC AAT TTC ATC TTC TTC ATA AAC ATT CTA AGA ATC CTG ATG AGG AAA CTT AGA ACA CAA GAA ACA AGA Pro Val Ile Leu Ser Ile Leu Ile Asn Phe Ile Phe Phe Ile Asn Ile Leu Arg Ile Leu Met Arg Lys Leu Arg Thr Gln Glu Thr Arg Z80 300 1250 GGA AGT GAA ACA AAC CAT TAT AAG CGC CTG GCC AAG TCC ACC CTC CTG CTG ATC CCG CTC TTT GGC ATC CAC TAC ATT GTC TTC GCC TTC Gly Ser Glu Thr Asn hlis Tyr Lys Arg Leu Ala Lys Ser Thr Leu Leu Leu Ile Pro Leu Phe Gly Ile His Tyr Ile Val Phe Ala Phe 320 1300 1350 TCC CCA GAG GAC GCC ATG GAG GTC CAG CTG TTC TTC GAA CTG GCC CTG GGC TCC TTC CAG GGT CTG GTG GTA GCT GTC CTT TAC TGC TTC Ser Pro Glu Asp Ala Met Glu Val Gln Leu Phe Phe Glu Leu Ala Leu Gly Ser Phe Gln Gly Leu Val Val Ala Val Leu Tyr CYs Phe 3Ub 1400 1450 CTC AAT GGT GAG GTG CAG TTG GAA GTT CAG AAA AAA TGG CGC CAG TGG CAC CTT CAA GAG TTC CCG CTG CGC CCT GTG GCC TTC AAT AAC Leu Asn Gly Glu Val Gln Leu Glu Val Gln Lys Lys Trp Arg Gln Trp His Leu Gln Glu Phe Pro Leu Arg Pro Val Ala Phe Asn Asn 380 - 1500 1550 TCC TTC AGC AAC GCC ACC AAC GGC CCC ACC CAC AGC ACC AAG GCT AGC ACT GAG CAA AGC CGG AGC ATC CCC AGG GCC AGT ATC ATC TGA Ser Phe Ser Asn Ala Thr Asn Gly Pro Thr His Ser Thr Lys Ala Ser Thr Glu Gln Ser Arg Ser Ile Pro Arg Ala Ser Ile Ile End 400 420 1600 1650 GGCTGCAGCAAGGCGCCAGCGGACGAAGGGTTGGATATTCGGTAGGACGGACATCTGTCTGATAGCCATCTCGTGCTCGACTGGCTGAACAGGCTTGTCTCTCGGGTCCGAGCTTCCTC 1 700 1750 GGAGATCAAGCAAGAGGAGAC'TACTAGGTCATCGCGGAGCAGGAAACGGACCCATGACTTGGTTTTTCTTGTTCTTCTAAGAAGAGCAATTCAGCAACTCCAGAAGGGGCAGGGGdo

b 100 200 300 400

3 3 2 2 1 11

0 10

-1 -1 -2 -2 -3 -3

100 200 300 400

Fig. 4. The primary structure of the rat secretin receptor. a. The cDNA sequence for the rat secretin receptor and its deduced amino acid sequence. Amino acid numbering begins at the N-terminal sequence postulated for the mature receptor with negative numbers for the encoded signal sequence. Positions of the putative transmembrane segments I-VII are underlined. The potential N-glycosylation sites are marked by stars, while the cysteine residues in the first extracellular domain are boxed. Two oligonucleotides used as primers in PCR are indicated by arrows above the nucleotide sequence. b. Hydropathy plot of the amino acid sequence of rat secretin receptor. The hydropathy plot was obtained by the method of Kyte and Doolittle (1982). The numbers indicate positions of the amino acid residues of the precursor protein.

Fremeau et al., 1986; Gespach et al., 1986). As shown in the stomach and pancreas, but was not found in the lung, Figure 5, Northern hybridization analysis of the poly(A) liver, kidney and brain. Secretin receptor mRNA was RNAs from various rat tissues using pQ17 cDNA as a probe detected in RNA from the hybrid NG108-15 and its parental showed a hybridizing band of about 2.5 kb. The secretin mouse N18 cells but not from the parental rat C6 glioma receptor mRNA was most abundant in the heart, less so in cells, which agreed with the results obtained from 1638 Molecular cloning of a secretin receptor cDNA

- 20-30 %. The amino acid sequence of the secretin -c receptor has litfle similarity to any members of the G protein- coupled receptors. Furthermore, the amino acid residues c-x r: conserved in the superfamily such as the Arg residue in the

C- 1 Z: in the , and the Asn-Pro residues >- kl- co second intracellular domain, C C-- f = , seventh transmembrane segment were not found in the .11 I O -, secretin receptor (Figure 4). These results suggest that the -5 -, = ClD Y~< m -. CD3 Z Z secretin receptor is a novel type receptor with seven transmembrane segments. Another notable feature of the secretin receptor is that it contains a signal sequence and a relatively large extracellular region (121 amino acids) -4128S preceding the transmembrane segment I. These properties were found in lutropin-choriogonadotropin (McFarland et 18S al., 1989) and endothelin receptors (Arai et al., 1990; Sakurai et al., 1990), but not other members of G protein- coupled receptors. The extracellular domain of the secretin receptor is rich in cysteine residues (7 residues) and may be involved in the binding of secretin (Robberecht et al., 1984). VIP could weakly compete with [125I]secretin for binding Fig. 5. Northern hybridization analysis of the secretin receptor rnRNA. to the secretin receptor (Figure 2) and stimulated the cAMP Poly(A) RNA (0.5 tzg) from rat tissues and cell lines was receptor electrophoresed on a 1.2% agarose gel containing 6.6% formaldehyde production in COS cells expressing the secretin and analyzed by Northern hybridization using the 32P-labeled 0.7 kb (Figure 3). Previously, Couvineau et al.(1990) have shown NcoI-SphI fragment of pQ17 as a probe. that secretin weakly competes with VIP for the VIP receptor purified from porcine liver. These results suggest that the [1251]secretin binding (Roth et al., 1984). However, as secretin and VIP receptors are structurally related. The pointed out above, the secretin receptor cDNA cloned from availability of the secretin receptor cDNA will facilitate the the hybrid NG108-15 was cDNA for rat mRNA. Therefore, isolation and characterization of cDNAs for the VIP receptor the rat secretin receptor gene from the C6 cell line appears and also the receptors for other secretin-related hormones to be transcriptionally activated in the rat/mouse hybrid such as glucagon, GRF and PHI. NG108-15 cell line. Although the previous reports with [125I]labeled secretin have suggested the existence of the secretin receptor in Discussion membranes of rat brain (Fremeau et al. ,1983), the Northern hybridization analysis did not show the secretin receptor Previously, Munro and Maniatis have developed a method mRNA in brain (Figure 5). As previously suggested to isolate cDNA clones for by using (Fremeau et al., 1983), the secretin receptor may be localized a radioactive ligand to probe the COS cell monolayer in the specific regions of brain. The in situ hybridization transfected with a expression cDNA library (Munro and with the labeled DNA or RNA probe for the secretin receptor Maniatis, 1989). In this report, we have successfully used may reveal the specific localization of the receptor in brain this method with a slight modification to clone the secretin in more detail, and clarify the possibility that secretin works receptor. Because of the low specific activity of as a or a neuromodulator. [1251]secretin and the low affinity of the receptor to [1251]secretin, the autoradiography of the COS cell monolayer with the ordinary X-ray film was impracticable Materials and methods in exposure times. This problem was overcome by using the Cells and peptides IP plate instead of X-ray film. Namely, the 18 h exposure The cell line NG108-15 (Klee et al., 1974), which was a hybrid cell line to the IP plate gave an almost identical signal to that obtained of mouse N18TG2 neuroblastoma and rat C6Bu-1 glioma, was obtained from Dr H.Higashida (University of Kanazawa), and was cultured in with the 40 day exposure to the X-ray film. This modified Dulbecco's modified Eagle's medium (DMEM, Nissui) containing 10% method may be applied to screen cDNAs for other cytokine fetal calf serum (FCS, Hyclone), and 0.1 mM hypoxanthine, 0.6 AM and peptide hormone receptors. aminopterin and 5 AM thymidine (HAT). Monkey COSGsl and COS7 cells We have shown that the secretin receptor can interact with were maintained in DMEM supplemented with 10% FCS with or without of Gs and activate the 600 A.g/ml G-418, respectively. the cr-subunit protein (Figure 1) Porcine secretin (Mutt et al., 1970) was synthesized using a peptide adenylate cyclase in COS cells (Figure 3). Accordingly, the synthesizer (Applied Biosystems, model 431A), and purified by HPLC. secretin receptor contains seven putative transmembrane The authenticity of the synthesized secretin was confirmed by Edman segments (Figure 4) which are topologically similar to the degradation using a gas-phase protein sequenator (Applied Biosystems, model photoreceptor and other G protein-coupled 477A). The synthetic porcine secretin5-27 fragment was purchased from Bachem Fine Chemicals, Inc., while VIP and glucagon were from Peptide receptors (O'Dowd et al., 1989). The structure of many G Institute Inc. protein-coupled receptors such as Secretin was labeled at the N-terminal histidine residue by the chloramine (O'Dowd et al., 1989), tachykinin receptors (Shigemoto et T method and purified essentially according to Chang and Chey (1980). al., 1990), and endothelin receptors (Arai et al., 1990; The specific radioactivity of [1251]secretin, determined by self-displacement 500-1000 Ci/mmol. Sakurai et al., 1990) have been elucidated, showing that they analysis, was routinely constitute a large superfamily. Each subfamily consists of Construction of the CDM8 cDNA library several members of receptors with 50-70% similarity Total RNA was prepared from NG108-15 cells as described (Chomczynski among them, whereas the similarity between the family is and Sacchi, 1987), and poly(A) RNA was selected by oligo(dT) -cellulose 1639 T.Ishihara et al.

column chromatography. Double-stranded cDNA primed with random oligotex-dT30 (Takara Shuzo, Japan). Northern hybridization was carried hexamer oligonucleotides (pdN6) was synthesized as described previously out as described (Sambrook et al., 1989). As a probe, the 0.7 kb NcoI-SphI (Fukunaga et al., 1990) except that AMV reverse transcriptase was replaced DNA fragment of pQ17 was labeled with 2p by the random primer by M-MLV RNase H- reverse transcriptase (BRL). After addition of the labeling method (Sambrook et al., 1989). BstXI adapter, cDNA larger than 1.0 kb was recovered from the agarose DNA sequencing was performed by the dideoxy nucleotide chain gel, ligated with BstXI-digested CDM8 vector (Seed, 1987), and introduced termination method using T7-DNA polymerase (Pharmacia) and into E.coli MC1061/p3 cells by electroporation. A total of 350 000 bacterial [a1-35S]dATPaS (Amersham). colonies were prepared as 400 pools of750-1000 clones, and plasmid DNAs were prepared from each pool by the alkaline lysis method (Sambrook et al., 1989). Acknowledaments We thank T.Kozasa for rat Gsa cDNA and rabbit anti-rat Gs antibody, Establishment of COSGs 1 cells H.Higashida for NG108-15 cells and P.J.Munson for the MacLigand Rat cDNA for Gsca (Itoh et al., 1986) was inserted at the XbaI site of pdEF- program. We are grateful to R.Fukunaga, S.Mizushima and K.Arai for BOS, which was generated by deleting the 0.25 kb of HindIII fragment helpful advice. This work was supported in part by Grants-in-Aid from the containing the SV40 replication origin from pEF-BOS (Mizushima and Ministry of Education, Science, and Culture of Japan. Nagata, 1990). The resultant expression plasmid (pEFGs1) was introduced into COS7 cells by the Ca-phosphate coprecipitation method in the presence of 0.1 mM chloroquine (Luthman and Magnusson, 1983) together with References pEFneo which carried the neomycin phosphotransferase gene under the human EF-Ia promoter. After selection in medium containing 1 mg/ml Arai,H., Hori,S., Aramori,I., Ohkubo,H. and Nakanishi,S. (1990) Nature, G-418 for 14 days, one of transformants (COSGsl) expressed a reasonable 348, 730-732. amount of Gsa, identified by Western blotting analysis using rabbit anti- Bayliss,W.M. and Starling,E.H. (1902) J. Physiol., 28, 325-353. rat Gsa antibody (a kind gift from Dr T.Kozasa). Bewab,W., Gespach,C., Marie,J.-C., Chastre,E. and Rosselin, G. (1988) Life Sci., 42, 791-798. Transfection and screening in COSGs 1 cells Bissonnette,B.M., Collen,M.J., Adachi,H., Jensen,R.T. and Gardner,J.D. 1.5 x 105 COSGsl cells per well in 6 well plate were transfected with 1.5 (1984) Am. J. Physiol., 246, G710-G717. /tg of plasmid DNA by the DEAE-dextran method (Fukunaga et al., 1990). Cerione,R.A., Codina,J., Benovic,J.L., Lefkowitz,R.J., Birnbaumer,L. and At 60 h after transfection, cells were incubated for 90 min at 37°C with Caron,M.G. (1984) Biochemistry, 23, 4519- 4525. 3.0x 105 c.p.m. of [125I]secretin in 1 ml of DMEM containing 10 mg/mi Chang,T.-M. and Chey,W.Y. (1980) Dig. Dis. Sci., 25, 529-552. BSA, 2 mg/mi bacitracin and 20 mM HEPES (pH 7.4). Cells were washed Chomczynski,P. and Sacchi,N. (1987) Anal. Biochem., 162, 156-159. four times with phosphate-buffered saline containing 0.5 mM MgCl2 and Christophe,J., Waelbroeck,M., Chatelain,P. and Robberecht,P. (1984) 0.7 mM CaCl2 (PBS+), and fixed for 15 min at room temperature with Peptides, 5, 341-353. 2% glutaraldehyde in PBS+. After washing once with PBS+, cells were Couvineau,A., Voisin,T., Guijarro,L. and Laburthe,M. (1990) J. Biol. air-dried, and vertical edges of the plates were removed. The plates were Chem., 265, 13386-13390. exposed to an IP plate (Fuji Photo Film Co. Ltd) for 18 h, and the bound Fremeau,R.T.Jr, Jensen,R.T., Chrlton,C.G., Miller,R.L., O'Donohue,T.L. [1 51]secretin was visualized by autoradiography using the Fuji BAS 2000 and Moody,T.W. (1983) J. Neurosci., 3, 1620- 1625. Bio-image analyzer (Fuji Photo Film Co. Ltd). Fremeau,R.T.Jr, Korman,L.Y. and Moody,T.W. (1986) J. Neurochem., 46, 1947-1955. Binding of [1251]secretin to membranes from COS cells and Fukunaga,R., Ishizaka-Ikeda,E., Seto,Y. and Nagata,S. (1990) Cell, 61, NG108-15 cells 341 -350. COS7 cells grown in 15 cm plates were transfected by the Ca-phosphate Gespach,C., Bataille,D., Vauclin,N., Moroder,L., Wunsch,E. and coprecipitation method with 10 ug each of plasmid pQ17, and the pEF- Rosselin,G. (1986) Peptides, 7, suppl. 1, 155-163. BOS (Mizushima and Nagata, 1990) or pEF-BOS-Gs which carries rat Gsax Gossen,D., Poloczek,P., Svoboda,M. and Christophe,J. (1989) FEBS Lent., cDNA in the pEF-BOS vector. The crude cell membrane fractions were 243, 205-208. prepared from COS cells or NG108-15 cells as described by Katada et al. Gozes,I. and Brenneman,D.E. (1989) Mol. Neurobiol., 3, 201-236. (1982) except that DTT was omitted from all buffers. 20 jtg of membrane Itoh,H., Kozasa,T., Nagata,S., Nakamura,S., Katada,T., Ui,M., Iwai,S., protein was incubated at 37°C for 90 min with [1251]secretin in 100 ,lA of Ohtsuka,E., Kawasaki,H., Suzuki,K. and Kaziro,Y. (1986) Proc. Natl. a solution containing 25 mM HEPES (pH 7.4), 5 mM MgCl2, 1 mM Acad. Sci. USA, 83, 3776-3780. EGTA, 50 mM NaCl, 10 mg/ml BSA, 2 mg/mi bacitracin, 0.1 mM Jensen,R.T., Charlton,C.G., Adachi,H., Jones,S.W., O'Donohue,T.L. and (p-amidinophenyl)methanesulfonyl fluoride hydrochloride and 100 jLg/mil Gardner,J.D. (1983) Am. J. Physiol., 245, G186-G195. leupeptin. The reaction was terminated by centrifugation at 14 000 g for Katada,T., Amano,T. and Ui,M. (1982) J. Biol. Chem., 257, 3739- 3746. 3 mmn. The pellet was suspended in 1 ml of ice-cold PBS, and the suspension Klee,W.A. and Nirenberg,M. (1974) Proc. Natl. Acad. Sci. USA, 71, was filtered through a Whatman GF/C filter which had been presoaked for 3474-3477. 24 h in 0.3% polyethyleneimine at room temperature. The radioactivity Kopin,A.S., Wheeler,M.B. and Leiter,A.B. (1990) Proc. Natl. Acad. Sci. retained on the filter was counted using a -y-counter. To determine the non- USA, 87, 2299-2303. specific binding of [1251]secretin to the membrane, a large excess of Kozak,M. (1987) Nucleic Acids Res., 15, 8125-8148. unlabeled secretin (10 /AM) was included in the assay mixture, and subtracted Kyte,J. and Doolittle,R.F. (1982) J. Mol. Biol., 157, 105-132. from the total binding to yield the specific binding. Luthman,H. and Magnusson,G. (1983) Nucleic Acids Res., 11, 1295-1308. McFarland,K.C., Sprengrel,R., Phillips,H.S., Kohler,M., Rosemblit,N., cAMP assay Nikolics,K., Segaloff,D.L. and Seeburg,P.H. (1989) Science, 245, The intracellular cAMP levels were assayed essentially as described by Zhou 494-499. et al. (1990). COSGsl grown in 75 cm3 flask were transfected with 6 yg Mizushima,S. and Nagata,S. (1990) Nucleic Acids Res., 18, 5322. of pQ17 or CDM8 by the DEAE dextran method. Cells were split into Munro,S. and Maniatis,T. (1989) Proc. Natl. Acad. Sci. USA, 86, 6 well microtiter plates 24 h after glycerol shock and grown for 40 h in 9248-9252. DMEM supplemented with 10% FCS. Cells were washed twice with Munson,P.J. and Rodbard,D. (1980) Anal. Biochem., 107, 220-239. incubation buffer (DMEM containing 1 mg/ml BSA and 0.5 mM Mutt,V., Jorpes,J.E. and Magnusson,S. (1970) Eur. J. Biochem., 15, 1-methyl-3-isobutylxanthine), and incubated at 37°C for 45 min with various 513-519. concentrations of peptide hormone dissolved in incubation buffer. After O'Dowd,B.F., Lefkowitz,R.J. and Caron,M.G. (1989) Annu. Rev. aspirating the buffer, the reaction was terminated by adding 1 ml ofethanol, Neurosci., 12, 67-83. and transferred to 1.5 ml Eppendorf tubes. After centrifugation at 14 000 Robberecht,P., Waelbroeck,M., Camus,J.-C., De Neef,P. and Christophe,J. g for 3 min, the supernatants were dried, and the level of cAMP was (1984) Biochim. Biophys. Acta., 773, 271-278. quantified using a cAMP assay system from Amersham. Robberecht,P., Cauvin,A., Gourlet,P. and Christophe,J. (1990) Arch. Int. Pharmacodyn., 303, 51-66. Northern hybridization and nucleotide sequence analysis Roskoski,Jr.R., White,L., Knowlton,R. and Roskoski,L.M. (1989) Mol. Total RNA was prepared from various rat tissues as described previously Pharmacol., 36, 925-931. (Chomczynski and Sacchi, 1987), and poly(A) RNA was selected using Rosselin,G. (1986) Peptides, 7, Suppl 1, 89-100.

1640 Molecular cloning of a secretin receptor cDNA

Roth,B.L., Beinfeld,M.C. and Howlett,A.C. (1984) J. Neurochem., 42, 1145-1152. Sakurai,T., Yanagisawa,M., Takuwa,Y., Miyazaki,H., Kimura,S., Goto,K. and Masaki,T. (1990) Nature, 348, 732-735. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York, 2nd edition. Seed,B. (1987) Nature, 329, 840-842. Shigemoto,R., Yokota,Y., Tsuchida,K. and Nakanishi,S. (1990) J. Biol. Chem., 265, 623-628. von Heijne,G. (1986) Nucleic Acids. Res., 14, 4683-4690. Zhou,Q.-Y., Grandy,D.K., Thambi,L., Kushner,J.A,. Van Tol,H.H.M., Cone,R., Pribnow,D., Salon,J., Bunzow,J.R. and Civelli,O. (1990) Nature, 347, 76-80. Received on March 14, 1991

Note added in proof The nucleotide sequence data reported here will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession number X59132.

1641