Molecular Cloning and Functional Expression of a cDNA Encoding the Human v 1 b Vasopressin Receptor*

Arginine vasopressin modulates the release of adreno-corticotropic hormone, p-endorphin, and prolactin from the anterior pituitary. Release is mediated by the V,, receptor through the mobilization of intracellular Cas+ by phosphatidylinositol hydrolysis. In contrast to its well characterized peripheral actions, such as antidiuresis, contraction of vascular smooth muscle, and stimulation of hepatic glycogenolysis, the exact site and mechanism of vasopressin action in the pituitary re-main unclear. This is largely due to a lack of information on the molecular identity and exact localization of the V,, receptor. This lack prompted us to try to isolate this receptor subtype. Here we report the molecular cloning and functional expression of a complementary DNA encoding the human V,, receptor. The deduced 424-amino acid sequence of the receptor has highest overall homology with the Via, V,, and oxytocin receptors, with homologies of 45, 39, and 45%, respectively. The receptor expressed in COS-1 cells has a single binding site for arginine vasopressin with a Kd of 0.17 * 0.04 m. It binds various agonists and antagonists of vasopressin with affinities distinct from those of VI, and V, receptors but consistent with those anticipated for the V,, receptor on the basis of the pharmacological studies. Furthermore, arginine antagonist, were tested on AVP-induced current responses at -70 mV in VI, (A) and V,, ( B ) receptor-expressed Xenopus oocytes. A, application of M AVP (closed bar, 1 min) elicited a current response that became desensitized with repeated application in the same oocyte. The current response was not affected with co-application of 5 x lo-' M ANT (open bar, 2 mid. B, the current response to low7 M AVP was abolished by eo-application of 5 x M ANT. The response to AVP was partially restored following washout of ANT. Experiments were repeated several times (n = 6 for A and n = 5 for B ) in different oocytes, and similar results were obtained.

sified into at least three subtypes, the V,,, V,,, and V, AVP receptors. The V, receptor is exclusively expressed in the kidney and regulates water reabsorption in the collecting duct and medullary thick ascending limb of the loop of Henle via the activation of adenylate cyclase (1). This receptor has recently been cloned from the human (2) and rat (3) kidney, and it has been shown that expression of the receptor gene is consistent with previous observations using combinations of various agonists and antagonists. The V,, receptor has been localized in the liver (11, vascular smooth muscle (11, brain (4-61, mesangial cells (71, and platelets (1). AVP action through the VI, receptor is mediated by activating phospholipase C, which in turn stimulates phosphatidylinositol turnover to increase intracellular Ca2+ (1). Distribution of the V,, receptor and coupling to phosphatidylinositol turnover are confirmed by recent molecular cloning of human and rat VI, receptors (8, 9). In the central nervous system, AVP is synthesized by several brain nuclei including the supraoptic, paraventricular, and suprachiasmatic hypothalamic nuclei, the medial amygdala, and the bed nucleus of the stria terminalis (10, ll). These nuclei have extensive vasopressinergic projections to hypothalamic and extrahypothalamic regions where AVP may act as a neurotransmitter or neuromodulator influencing blood pressure, memory, body temperature, and brain development (1, 6, 10, 11). AVP receptors in the central nervous system seem most similar t o the V,, receptor; this similarity includes the second messenger system involved (1, 4 -6 , [10][11][12]. However, the adenohypophysial AVP receptor involved in AVP-induced corticotropin release seems to differ from the vascular, hepatic, and other brain V,, receptors and has therefore been classed as a VI, receptor (13,14). However, the site and mechanism of action of AVP in the brain, particularly its pituitary action, are not precisely known. This is largely due to a lack of information on the molecular identity and localization of the V,, receptor and to a lack of potent specific agonist or antagonist for this receptor. To overcome these difficulties, the V,, receptor must be characterized at the molecular level. This should enable the differentiation of the sites and actions ofAVP mediated through the V,, receptor from those mediated through the V,, receptor and the screening of specific agonists and antagonists for this receptor. Here, we report the molecular cloning of the putative V,, receptor from the cDNA library of the human pituitary. The function of this receptor expressed in COS-1 cells andXenopus oocytes was also studied.
Reverse Dunscription Polymerase Chain Reaction (RT-PCR)--TWo degenerate oligonucleotides corresponding to transmembrane domains 111 and VI (the 5' primer 5'-TGGCCATG(Afl')C(A/G)(C/G)(C/T)(C/ G)GACCG-3' and the 3' primer 5 ' -( A / G ) ( A / C ) T G~G ) G G ( C / T)G(CPT)CCAGCA-3') were synthesized based on cDNA sequences of the cloned human vasopressin and oxytocin receptors. RT-PCR was performed using human pituitary cDNA as a template for 30 cycles at 94 "C for 1 min, 60 "C for 2 min, and 72 "C for 2 min. The PCR product was gel-purified with a Geneclean I1 kit (BIO 101, Inc.), subcloned into pCR I1 plasmid, and partially sequenced. The human pituitary cDNA for RT-PCR was synthesized from 1 pg of human pituitary mRNA using a cDNA synthesis kit (Boehringer Mannheim, Germany).
Human Pituitary cDNA Library Screening-The A gtl0 human pituitary cDNA library was plated on twenty 150-mm Petri dishes at a density of 40,000 plaque-forming unitldish. Each dish was lifted onto duplicate colony/plaque nylon filters (DuPont NEN). Replicate filters were prehybridized at 42 "C for 2 h in a solution containing 50% formamide, 6 x SSC, 0.1% SDS, 5 x Denhardt's solution, 0.25 M sodium phosphate, and 100 pg/ml denatured salmon sperm DNA. The prehybridization solution was replaced by the fresh prehybridization solution containing the 32P-labeled PCR fragment for further incubation at 42 "C for 24 h. After hybridization, the filters were washed with 4 x SSC, 0.1% SDS at room temperature for 15 min, with 2 x SSC, 0.1% SDS at 42 "C, and with 0.5 x SSC, 0.1% SDS at 52 "C followed by autoradiography with an intensification screen at -80 "C overnight. Positive clones were subcloned into pCR I1 plasmids.
Expression of VI, Receptors in COS-1 Cells-The 1.5-kb SpeI-Xbal fragment of the human VI, receptor gene was inserted into eukaryotic expression vector pEF-BOS containing human polypeptide chain elongation factor 1 a promoter and SV40 replication origin (15). COS-1 cells, grown in 150-mm diameter dishes, were transfected with 50 pg of plasmid by the DEAE-dextran method (16). The transfected cells were harvested within 48 h, resuspended in a solution comprising 50 m M Tris-HC1,3 nm MgSO,, and 1 m M EGTA (pH 7.41, and homogenized by 50 strokes in a glass homogenizer, followed by centrifugation at 38,000 x g for 60 min to obtain membrane fraction. Receptor Binding Studies-Aliquots of membrane preparation were added to tubes containing 250 pl of binding assay mixture (50 m M Tris-HC1, pH 7.4,3 m M MgSO,, 1 m EGTA, 0.1 mg/ml bacitracin, 0.1% bovine serum albumin, VHIAVF' , and indicated drugs). Incubations were carried out at 22 "C for 1 h and stopped by filtration through glass filters (Pharmacia, Sweden) and immediate washing with 3 x 2 ml of ice-cold buffer containing 50 m M Tris-HC1 and 0.01% bovine serum albumin (pH 7.4). Experiments were carried out at least three times, in duplicate for the displacement study and in triplicate for the saturation study. 1 y AVF' was added to define nonspecific binding, which was subtracted from total binding to calculate specific binding. IC,, values were derived by nonlinear least squares analysis, from which K, values were calculated by the method of Cheng and Prusoff (28).
Expression of V,, and VI, Receptor in Xenopus Oocyte-The cDNA of human liver V,, receptor (9) was cloned using RT-PCR from human liver mRNA. Each plasmid encoding V,, or V,, receptor was transcribed in vitro in the presence of the cap analogue m7G(5')ppp(5')G by T7 RNA polymerase after digestion with XhoI and HindIII, respectively, as previously described (17). Xenopus oocytes were injected with 50 nl (approximately 50 ng) of the transcribed RNA and incubated for 2-5 days. Defolliculated oocytes were voltage clamped at -70 mV and monitored for a response caused by activation of calcium-dependent chloride channel elicited by the application of AVP to the bath, under continuous perfusion, in the presence or absence of a V,,Nz antagonist ([des-

Glyg,d(CH,),',Tyr(Et)2,Va141AVF') (17).
Northern Hybridization-Five pg of each mRNA (pituitary, liver, and kidney) was resolved on a formaldehyde-agarose gel, transferred to Hybond-N membrane ( h e r s h a m , U. K.), and hybridized with 32P-labeled V,, receptor cDNA. The hybridization was performed at 42 "C for 24 h in the same solution used for plaque hybridization. The membrane was finally washed with 0.2 x SSC, 0.1% SDS at 62 "C for 20 min, and then autoradiography was performed with an intensification screen at -80 "C for 4 days.
RT-PCR Southern Hybridization-One pg of each human mRNA was converted to cDNA with a cDNA synthesis kit according to the protocol. A part of this cDNA was subjected to PCR using a forward primer (127-146 nt) and a reverse primer (627-646 nt) under the following condition: 94 "C for 1 min, 60 "C for 2 min, and 72 "C for 2 min for 22 cycles. Each PCR product was run on 1% TAE-agarose gel and stained with ethidium bromide. After a picture was taken, the gel was transferred to Hybond-N membrane and hybridized with 32P-labeled V,, receptor cDNA. The hybridization procedure was the same as Northern blot hybridization. The membrane was autoradiographed with an intensification screen at -80 "C for 1 day.
Data Analysis-Nucleotide and amino acid sequences were analyzed and compared with the computer program Gene Works on a Macintosh computer (IntelliGenetics, Inc.).

RESULTS AND DISCUSSION
Two degenerate oligonucleotides corresponding to the third and sixth transmembrane domains of G protein-coupled receptors were synthesized using sequence data reported for the human V,, (8), V, (2), and oxytocin (18) receptors. Using these primers and human pituitary cDNA as a template, PCR was performed, and the product was subcloned. Since sequencing showed that this PCR product had an identity of 50, 39, and 49% with human Vla, V,, and oxytocin receptors at the nucleotide level, respectively (data not shown), and no other homologous genes were found in the data bases, this fragment was used to screen a human pituitary cDNA library to clone the full-length VI, receptor. Five positive clones were obtained from 800,000 plaques screened. These turned out to be the same gene products, one of which (H-5 clone) contained a 1824-base pair insert that covers the full length of the putative V,, receptor molecule.
The H-5 has an open reading frame (ORF) of 1272 nt (Fig.   LA), which contains a nucleotide sequence identical to that of the used fragment. The nucleotide sequence of H-5 is unique in that it lacks the Kozak consensus sequence and has a short ORF of octapeptide at -184 to -161 nt upstream of the second initiation codon of the receptor gene (Fig. lA). Such findings have also been reported for other G protein-coupled receptors, including hamster p2 adrenergic receptor (19). The short ORF might regulate the translation rate of the receptor gene (20,21). The deduced 424-amino acid sequence (calculated relative molecular mass is 47,034) has seven putative transmembrane domains and an overall sequence homology of 45,39, and 45% with sequences of Vla, V,, and oxytocin receptors, respectively (Fig. 1B). All other known members of the G protein-coupled receptor family have homologies with this H-5 clone of less than 21% (data not shown). Motif analysis of H-5 amino acid residues revealed the presence of one N-linked glycosylation site (N-X-(S/T)) at Asn-21 in the N-terminal region, one consensus sequence for the protein kinase A phosphorylation site ((FUK)-(wK)-X-(SpT)) at Ser-368, and 5 consensus sequences for the protein kinase C phosphorylation site ((Sfl)-X-(R/K)) at Thr-258, Thr-279, Ser-373, Ser-398, and Ser-406 (Fig. LA). An aspartic acid residue in the second transmembrane domain is highly conserved among the G protein-coupled receptors and is functionally crucial for agonist binding (22,23). Indeed, Asp-80 of the H-5 clone is conserved among members of the vasopressidoxytocin receptor family (Fig. lA). Two clusters of positively charged residues, namely Lys-Val-Lys at 231-233 and Lys-Ile-Arg at 27G278, in the third intracellular loop of H-5 a r e well conserved i n h u m a n V,, and oxytocin receptors but absent in the human V, receptor (Fig. 1B). These regions have been suggested to be important for the coupling of receptor and phospholipase C (24, 251, and the results shown here are in good agreement with previous reports showing V,, receptor-phospholipase C coupling in the anterior pituitary (1, 2, 12).
To determine the binding characteristics of the putative Vlb receptor, we used COS-1 cells transiently expressed with the H-5 clone. The membrane of transfected cells bound [3H]AVP in a saturable fashion. Scatchard analysis revealed that the receptor has a single binding site for the ligand with a dissocia-     (Fig. 2B).

CTCAGCCTCAGCCTAACCCTCAGTGGGAGGCCCAGGCCTGAAGAGTCACC~GGGACT~GAGCTGGCAGATGGGGAAGGCACCGCTGAG L S L S L T L S G R P R P E E S P R D L E L A D G E G T A E
The Ki values for these ligands, in the same order as above, were 0.15 -c 0.01,6.5 2 1.0,9.9 2 1.0,29 2 2,506 2 21, 1430 2 70, and 30,000 2 9,700 m, respectively. These data are mostly consistent with previous pharmacological findings in rat tissues (26,27). hybridization probed with the human V,, receptor fragments. Arrow-H, heart; B, brain; P1, placenta; Lu, lung; Li, liver; Sk, skeletal muscle; K, kidney; Pa, pancreas; Pi, pituitary. antagonist than that necessary to affect the hepatic and renal action of AVP is necessary to affect AVP-induced corticotropin release from the pituitary (27). The affinities of pituitary membrane and the H-5 receptor for the VI, agonist [deamino-(~-3pyridyl)Ala21AVP are similar and are 10 times higher than that of kidney membrane (V, receptor). Some differences among the reports may be explained by the different species involved and the purity of the receptor preparations used. Thus, binding studies clearly indicate that the cloned H-5 receptor is distinct from the V,, and V, receptors and is classified as the V,, receptor.
To examine the functional response of the receptor protein, the cloned H-5 cDNA was transcribed in vitro by T7 RNA po-lymerase, and the cRNA obtained was microinjected into Xenopus oocytes (17). When the H-5-expressed oocytes were stimulated with AVP, typical inward calcium-dependent chloride currents, mediated through G protein-coupled phospholipase C activation, were recorded in a dose-dependent manner. Fig.  3A depicts the current activated by M AVP. Water-injected oocytes showed no response (data not shown). Oocytes microinjected with human V,, receptor cRNA similarly responded to AVP; this response was inhibited by the addition of the V,,N,  antagonist, [des-Glyg,d(CH,),',Qryr(Et)2,Va14]AVP, in a dose-dependent manner, and the chloride current was abolished by this antagonist at 5 x M (Fig. 3B). The inhibition was partly restored after washing (Fig. 3B), suggesting that it is specific to the antagonist. In contrast, this antagonist failed to affect the AVP-evoked current observed in oocytes expressed with H-5 (Fig. 3a). These data clearly demonstrate that the cloned receptor has functional characteristics of the V,, receptor (26,27).
Northern blot analysis demonstrated that V,, receptor expression was detected as a 5.2-kb transcript in the pituitary but was undetectable in the liver and kidney (Fig. 4A). RT-PCR Southern hybridization using a V,, receptor-specific primer failed to reveal detectable expression of this receptor in any other tissue examined (Fig. 4b). These tissues included the heart, brain, lung, placenta, skeletal muscle, pancreas, kidney, and liver. Using a specific primer for V,, receptor, we were able to demonstrate the expression of this receptor subtype in the pituitary in addition to that of the V,, receptor (data not shown).
Taken together, our data clearly establish the molecular identity of the human pituitary V,, receptor and provide an opportunity to extend our studies to the precise elucidation of the localization and metabolic regulation of the receptor molecule. Such elucidation may lead to an understanding of the role and mechanism of vasopressin action in the pituitary. Studies to this end are currently in progress in our laboratory.