Molecular Characterization and Regulation of the Human Endothelin Receptors*

Endothelin receptors (ETRs) are distributed throughout a variety of tissues. Two human cDNAs were identified which encode distinct ETR proteins. One cDNA encoded a 427-amino acid protein that shared 91% identity to rat ETAR. The second cDNA encoded a 442-amino acid protein that was 88% identical to rat ETeR. Ligand binding studies of the cloned receptors expressed in COS cells confirmed that they were pharmacologically ETAR and ETBR subtypes; al- though the selective antagonist BQ123 showed a potency similar to ET-3 in displacing "'1-ET-1 binding to ETAR. This observation contrasts with rat ETAR pharmacology where BQ123 has a 100-fold higher affinity than ET3. Chinese hamster cells ex- pressing the human ETAR equal potencies in displacing "'I-ET-l binding, which indicates that rat and human ETAR are pharmacologically distinct. Electrophysiological studies both that they functional. are differentially


Endothelin receptors (ETRs)
are distributed throughout a variety of tissues. Two human cDNAs were identified which encode distinct ETR proteins. One cDNA encoded a 427-amino acid protein that shared 91% identity to rat ETAR. The second cDNA encoded a 442-amino acid protein that was 88% identical to rat ETeR. Ligand binding studies of the cloned receptors expressed in COS cells confirmed that they were pharmacologically ETAR and ETBR subtypes; although the selective antagonist BQ123 showed a potency similar to ET-3 in displacing "'1-ET-1 binding to ETAR. This observation contrasts with rat ETAR pharmacology where BQ123 has a 100-fold higher affinity than ET3. Chinese hamster ovary cells expressing the human ETAR displayed equal potencies in displacing "'I-ET-l binding, which indicates that rat and human ETAR are pharmacologically distinct. Electrophysiological studies of both ETRs expressed in Xenopus oocytes revealed that they are functional. Northern analysis indicated that the two ETRs are differentially expressed in many tissues. Marmosets maintained on a high fat/high cholesterol diet exhibited 3-fold increase in ETBR mRNA levels with little change in ETAR mRNA levels. Availability of cDNA clones for ETR subtypes can open avenues for future analysis of their role in pathophysiology of various diseases.
Endothelins are a family of peptide hormones having profound cardiovascular, mitogenic, and potential neuroregulatory functions. In mammals, the endothelin peptide family is composed of three members, endothelin 1, endothelin 2, and endothelin 3 (ET-1, ET-2 and ET-3),l which are encoded by three separate genes that are differentially expressed in various tissues (for reviews see Refs. 1 and 2). Mammalian endothelins are 21 amino acids in length and share a high degree of sequence and structural similarity to the sarafotoxin * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
5 To whom correspondence should be addressed.
Receptors for the ET-related peptides are known to occur in at least two major subtypes, denoted ETA and ETB (4), as defined by rank order potency of ET-1, ET-2, ET-3, and SRT (5,6). ETA receptors are defined by high and equal affinity for ET-1 and ET-2, approximately 70-100-fold lower affinity for ET-3, and a 1000-fold lower affinity for SRT. In contrast, the ETB receptor subtype displays high and similar affinity for all endothelin-related peptides. Receptors for ET and SRT are differentially expressed in a wide variety of tissues and cell types of the periphery and central nervous system (5)(6)(7)(8)(9)(10). ETs and SRTs bind to a common receptor and initiate a common signal transduction pathway, principally a G-protein-mediated activation of phospholipase C and subsequent inositol triphosphate-mediated increase in intracellular Ca2+ levels (1,11).
The physiological effects of E T are diverse and include potent vasoconstrictive effects (E?), vasodilative effects (13), induction of c-fos transcription (14), activation of DNA synthesis and mitogenesis in a variety of cells (15,16), stimulation of P I hydrolysis in granular cell neurons (16,17), depolarization of spinal neurons (18), and stimulation of the release of substance P (19), vasopressin (20), prolactin (21), aldosterone (22), and the glycoprotein hormones, FSH, LH, and TSH, (23,24). These diverse and complex physiological effects mediated by ET in conjunction with the molecular heterogeneity and differential tissue distribution of the ET-related peptides and their receptors underscores the importance of utilizing a molecular biological approach to dissect the components of E T physiology. By expression cloning we have recently isolated and characterized an ETB receptor subtype cDNA clone from porcine cerebellum (25). Here we report the cloning, functional characterization, and regulation of the human ETA and ETB receptor subtypes.

Construction and Screening
of the Human cDNA Libraries-A human lung cDNA library was obtained from Stratagene. A human heart cDNA library was prepared as follows: total RNA was isolated from human heart left ventricle tissue as described (26). Polyadenylated RNA was isolated by two cycles on oligo(dT)-cellulose (Collaborative Research (27)) chromatography, and a XZAPII cDNA library was prepared by priming the poly(A)+ RNA with oligo(dT) (28). Plaques were screened by hybridization to nitrocellulose replicates using the 32P-labeled porcine ETBR cDNA coding sequence as a probe in 20% formamide, 5 X ssc (SSC is 150 mM NaCl, 15 mM sodium citrate), 5~ Denhardt's, 0.1% SDS, and 0.2 mg/ml of Escherichia coli tRNA at 42 "C (28). Filters were washed with 2 X SSC, 0.1% SDS at 42 "C. Several putative recombinant XZAP clones were isolated from human heart and human lung libraries, and two of them (pHETR21.1 and pHETR11) were further characterized. Preliminary sequence C ETAk 3 TLCLRASFWLALVGCVISDNPERYSTNLSNHVDDFTTFRGTELSFLVTTH 5 2    analysis indicated that pHETR21.1 contained deletions in the coding sequence of the gene (see "Results and Discussion"). The complete coding sequence was isolated using a nested polymerase chain reaction. Based on the sequence information obtained from the 3' and 5' noncoding sequence of pHETR21.1, oligonucleotide primers were synthesized corresponding to the amino (ATT CTC GAG CCA CCC ACC CTC GCC GGC TCC (outside) and (GCC ACC ATG GAA ACC C T T TGC CTC AGG GCA TCC (inside)) and carboxyl (TTG CTC CAG GGA GTG AAT TAA GGA AGA AGG (outside)) and TTA CTC GAG TAC CGA GGA GTG CTT CTA AGG (inside)) termini and used to obtain the full length clone pHETRld5 from human heart left ventricle RNA using the PCR reaction (29).

L S V D R Y R A V A S W S R V Q G I G I P L V T A I E I V S I W l L S F I L A l P E A I G f~~~
Nucleotide Sequence Analysis-The inserts of the two positive clones pHETRld5 and pHETRll were subcloned into the pBluescript vector (Stratagene), and a series of EnoIII deletions were made. Nucleotide sequence of both strands was determined by a modification of the dideoxy chain termination method (30) using the Sequenase I1 kit (United States Biochemicals). The Wisconsin Genetics Computer Group Software package (31) was used to assemble composite sequences from the various fragments and for further sequence analysis. RNA Blot Analysis-For Northern analysis, total RNA was isolated from various human or marmoset tissues using the guanidinium thiocyanate acid-phenol method (32). Fifteen pg of each RNA was fractionated on 1% agarose formaldehyde gels (33) and transferred to nitrocellulose membranes. Northern hybridization reactions were performed a t 42 "C in 50% formamide, 5 X SSPE, 5 X Denhardt's reagent, 0.1% SDS, and 100 pg/ml yeast tRNA. The blots were washed with 0.1 X SSC, 0.1% SDS at 50 "C and exposed to x-ray film for 48 h at -70 "C.
Dot blot analysis was performed with a template manifold apparatus (Schleicher & Schuell, Keene, NH) to assure uniform dot size. Total cellular RNA was applied a t four different concentrations (3, 2, 1, and 0.5 pg) (34). The RNA samples were denatured by adjusting them to 1 M formaldehyde and heating them at 55 "C for 15 min. The samples were diluted into 20 volumes of 3 M NaCl containing 0.3 M trisodium citrate and applied to nitrocellulose filters under a gentle vacuum. The filters were washed with additional diluent, baked a t 80 "C for 2 h and then hybridized and washed under high stringency as described above. Autoradiograms of filters were analyzed by quantitative scanning densitometry.
Stable Expression and Amplification of Human ET Receptor in CHO Cells-A fragment containing the entire human ETAR cDNA coding sequence was sublconed into the mammalian expression vector RLDN which contains both the G418 resistance and dihydrofolate reductase genes (35). CHO cells were transfected with RLDN/ETAR using electroporation (36) and plated onto 96-well plates containing F-12 medium (GIBCO) supplemented with 5% fetal calf serum. After 48 h, the cells were grown in the same media containing neomycin (400 pg/ml) for 2 weeks. Surviving cells were assayed for the production of ETA receptor. Cells that expressed the highest level of receptors were selected, then grown in Dulbeceo's modified Eagle's medium without nucleosides and supplemented with 10 nM methotrexate and 5% dialyzed fetal calf serum. After 2 weeks, cells surviving the methotrexate selection were allowed to grow and assayed for ETA receptor expression using a binding assay.
Membrane Preparation-COS cells grown in 245 X 245-mm tissue culture plates were transfected with 100 pg of pHETRld5 or p H E T R l l DNA and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum for 2 days as previously described (25). The cells (CHO, COS) were washed with Dulbecco's phosphate-buffered saline containing a protease inhibitor mixture (5 mM EDTA, 0.5 mM phenylmethlysulfonyl fluoride, 5 pg/ml leupeptin, and 0.1 p/ml aprotinin) and scraped in the same buffer. After centrifugation a t 800 X g, the supernatant was aspirated, the cells were lysed by freezing in liquid N z and thawing on ice, and then resuspended in hypotonic lysis buffer containing 20 mM Tris, p H 7.5, and the protease inhibitor mixture. The cell lysate was homogenized 30 times using a glass dounce homogenizer and centrifuged a t 800 x g for 10 min to remove unbroken cells and nuclei. The supernatant was centrifuged at 40,000 X g for 15 min, and the pellet was resuspended in 50 mM Tris-HC1, pH. 7.5, and 10 mM MgC12, divided into small aliquots and stored at -70 "C after freezing in liquid NP. These membrane preparations were stable at -70 "C for up to 3-4 months. Protein was determined by the bicinchoninic acid method using bovine serum albumin as the standard.
Binding S t~d i e s --'~~I -E T -l binding to COS or CHO cell membranes was initiated by the addition of 25 pl of 1251-ET-1 (0.2 nM) in 0.05% bovine serum albumin to membranes in the absence (total binding) or presence (nonspecific binding) of 100 nM unlabeled ET-1. The assay volume was 100 pl, and the concentrations of membrane proteins were 0.5 and 0.05 pg per assay tube for ETA and ETB receptors, respectively. The incubations (30 "C 60 min) were stopped by dilution with cold buffer (20 mM Tris-HC1, pH 7.5, and 10 mM MgC12) and filtering through Whatman GF/C filters presoaked in 0.1% bovine serum albumin. After 3 washes (5 ml each time) with the same buffer using a Brandel cell harvester, the filters were counted using a y counter a t 75% efficiency.
I n Vitro Transcription, Microinjection into Xenopus laeuis Oocytes, and Electrophysiology-RNA transcripts were synthesized (42) from linearized pHETRld5 or pHETRll DNA using T 7 RNA polymerase (Stratagene), and digested with DNase1 (1 unit/pg DNA) to remove template DNA. The reaction mixture was extracted with phenol/ CHC13/isoamyl alcohol (50:49:1) and ethanol-precipitated after addition of 3.75 M ammonium acetate. For microinjection, X. laeuis females (NASCO) were anesthetized by hypothermia, and the ovaries were surgically removed. Follicle cells were dispersed, and individual oocytes were released by incubation with 2 mg/ml collagenase (Worthington) in modified Barth's medium as described (37). After collagenase treatment and washing, the oocytes were allowed to recover overnight at 18 "C in Barth's medium. Stage V-VI oocytes were selected, and the follicular membranes were manually removed.  apparatus) and maintained in modified Barth's medium at 18 "C until electrophysiological measurements were made. Electrophysiology was performed using the voltage clamp technique using an oocyte voltage clamp apparatus (Warner Instruments). Oocytes were clamped at -60 mV, and the Ca2+ activated C1-channel activity was recorded in Bnrth's medium at room temperature as described (38).

RESULTS AND DISCUSSION
Cloning cDNAs Encoding the Human ETa and ETB Receptors-The porcine ETB receptor cDNA, previously cloned in our laboratory (25), was used to probe human heart and lung cDNA libraries. Two cDNA clones were isolated from the lung library that encoded the ETB receptor. One clone, termed pHETRl1, was chosen for further analysis. A single clone, designated pHETR21.1, was isolated from the heart cDNA library that encoded a full length ETAR with 448 bp of 5' and 420 bp of 3' untranslated regions, respectively. Nucleotide sequence analysis revealed that this clone had two deletions of 250 and 70 bp within the coding region. The receptor encoded by this cDNA was not functional as determined by both COS cell transfection and Xenopus oocytes electrophysiological studies. In order to acquire the complete coding sequence, we designed oligonucleotide primers corresponding to the 3' and 5' untranslated regions of the pHETR21.1 clone and used these to obtain cDNA encoding the complete coding region from human heart mRNA using nested PCR (see "Experimental Procedures"). Several clones from the PCR reaction were subcloned into the PCR 1000 vector and analyzed by nucleotide sequence. One clone, pHETRld5, was used for further ligand binding and functional analysis.
Both pHETRld5 and pHETR11 have one open reading frame, beginning at a methionine codon ATG (nucleotide 1) and ending a t a stop codon TGA) (Fig. 1, A and €3). The deduced polypeptides consist of 427 and 442 amino acid residues, respectively, with an expected molecular mass of approximately 47 and 49 kDa. The hydropathy profile of the ETR determined by the method of Kyte and Doolittle (39) indicated the presence of seven hydrophobic regions (22-27 amino acid residues in length) which are likely transmembrane domains typical of receptors coupled to G-proteins.
The deduced amino acid sequence of the human heart ETR (pHETRld5) was 94 and 91% identical to the bovine and rat ETAR, respectively (44,45). Similarly, the human lung ETR (pHETR11) protein was 90 and 88% identical to the porcine and rat ETBR (25,46). This high level of homology suggests that pHETRld5 and pHETR11 encode ETA and ETB receptor subtypes, respectively. The NH2-terminal region preceding the putative first transmembrane domain for each clone contains one potential N-glycosylation site. Comparison of the primary protein sequence of the human ETA and ETB receptor subtypes indicates a high degree of homology in the transmembrane domains with greater sequence divergence in the putative extracellular domains (Fig. IC). The overall amino acid sequence identity between the two receptors was 58.9% ( Fig. IC) (47,48). Future studies will delineate the importance of these domains in ligand binding, signal transduction, and receptor regulation.
In addition, the human ETRs have several amino acid residues and sequence motifs shared by other members of the G-protein superfamily including cysteines a t position 158 and 239 (pHETRld5) and 174 and 255 (pHETRll), which are postulated to cross-link extracellular loops 1 and 2 by a disulfide bond (40). As expected, the regions of greatest sequence similarity between the ETRs and other G-protein receptors are concentrated in the hydrophobic segments, whereas the amino-and carboxyl-terminal regions and the loops between the hydrophobic segments IV, V, and VI differ widely in length and amino acid composition (38-43). ETRs have several other notable features with respect to other members of the G-protein superfamily. First, the E T receptors are predicted to have a leader sequence, a structural feature shared only by the thyrotropin (TSH) and lutropin choriogonadotropin (LH-CG) receptors (49)(50)(51). Second, all E T receptors have a large extracellular NH2-terminal region composed of between 90 to 100 residues; a feature qualitatively shared with the receptors for TSH or LH-CG (49-51). Third, Abundance of ETAR and ETBR mRNA in tissues of marmoset as function of diet. Marmoset was maintained in two groups on either control or high fat high cholesterol diet containing 15% coconut, 5% cholesterol for 3 months. The level of serum cholesterol were 170 and 350 mg/dl in control and cholesterol-treated animals, respectively. Total cellular RNA was prepared from the control as well as the cholesterol-treated animals. These were then examined by Northern as well as dot blot analysis using 32P-labeled ETAR and ETBR cDNAs as probe as described in Fig. 5. Various exposures of the resulting autoradiograms were examined by quantitative densitometry, and the densitometric absorbance over the linear range of the exposures was compared with known standards of total heart or lung RNA. The normalized densitometric absorbances are plotted (average of two different replicates with less than 15% differences).
Ligand Binding Properties-Binding of IZ51-ET-l to membranes prepared from COS cells transfected with ETA and ETB receptors was specific, saturable, and of high affinity as shown in Fig. 2, A and B. The nonspecific binding was between 8-25% and 2-5% for ETA and ETB receptors, respectively. The Scatchard transformation of the specific binding from saturation binding experiments indicated the presence of a single class of high affinity binding sites with apparent dissociation constant (&) and maximum binding (BmaX) of 83 k 8 PM and 3 k 0.3 pmol/mg protein and 52 +-4 pM and 60 -C 10 pmol/mg protein for ETA and ETB receptors, respectively (Fig. 2, A and B, inset). Although the affinity of ET-1 for ETA and ETB receptors were very similar, the density of ETB receptors was 10-20 times higher than the density of ETA receptors when expressed in COS cells. The reason for this differential expression of ETA and ETB receptors in COS cells is not known; however, Aramori and Nakanishi (61) have also observed differential expression of ETA (13 pmol/mg protein) and ETB (0.5 pmol/mg protein) in CHO cells.
Competition of lZ5I-ET-l binding to ETAR and ETBR transfected COS cell membranes by unlabeled ET-1, ET-3, and subtype-selective ligands such as S6c (>1000-fold more selective for ETB receptors) and the recently developed ETAselective antagonist BQ123 (>lOOO-fold more selective for

!man Endothelin Receptors
ETA receptors) (62) are shown in Fig. 3, A and B. While ET-1 displayed similar affinities for ETA and ETB receptors (IC50 = 0.3 and 0.2 nM, respectively), ET-3 was 300-fold more potent for ETB with of 70 and 0.2 nM for ETA and ETB, respectively (Fig. 3, A and B ) . S6c displayed an IC5o value of 0.3 nM for ETB and >lo00 nM for ETA receptors, whereas BQ123 displayed an ICs0 value of 100 nM and >lo00 nM for ETA and ETB receptors, respectively. As shown in Fig. 3A, BQ123 and ET-3 displayed very similar potency in displacing lZ5I-ET-l binding to ETA receptors. This was very different from the data reported on rat aortic smooth muscle cells (A-10 cells) which display E T receptors of ETA subtype where BQ123 was much more potent (100-fold) than ET-3 (63). It was important to know whether this difference in potency observed in rat aortic smooth cells and human ETA receptor cloned and expressed in COS cells was due to the transient expression of the receptors in COS cells or difference in the species. To answer this question, stable expression of ETA receptors in CHO cells was obtained, and the results of the saturation and competition binding experiments are shown in Fig. 4, A and B. The binding parameters obtained with membranes prepared from ETAR-expressed CHO cells were same as that obtained with COS cell membranes. These data suggest that the pharmacological difference between A-10 cells and human ETA receptors might be due to species difference. Further experiments are required to confirm this.
Functional Studies-The ability of ETAR and ETBR to trigger a functional response was demonstrated by ET-1mediated electrophysiological response in Xenopus oocytes injected with RNA derived from pHETRld5 and pHETRl1. Oocytes injected with RNA derived from pHETRld5 display strong and rapid activation of Ca2+-activated C1currents upon addition of ET-1 and weak response with ET-3. However, oocytes injected with RNA from pHETRl l illicits strong activation of Ca2+-activated C1-currents upon addition of ET-1 and ET-3 (Fig. 5). These data suggests that pHETRld5 and pHETRll encode functional receptors capable of hydrolyzing polyphosphoinositides resulting in intracellular Ca2+ mobilization and are consistent with previous reports of the ETR signaling pathway (1,11).
mRNA Abundance as a function of Organ and Diet-RNA dot analysis was used to examine the pattern of expression of the ETA and ETB receptor mRNA derived from human and marmoset tissues. As shown in Fig. 6, mRNA expressed by each ETR gene was specifically detected in all the RNA samples. The amount of ETR mRNA present in each sample did not vary drastically, although human lung and heart tissues and marmoset heart tissues had significantly greater levels of ETA receptor mRNA. Among the tissues analyzed, ETB receptor mRNA levels appeared to be higher in lung, liver, and placenta in humans and lung tissue in marmoset.
Endothelin has been proposed to play a role in the development of hypertension since elevated plasma levels of ET have been detected in patients with hypertension (64-65). In addition, hypertensive patients with impaired renal function showed a &fold increase in plasma ET-1 levels (66). We have been interested in the association between hypercholesterolemia and the development and maintenance of hypertension. Accordingly we investigated the effect of hypercholesterolemia and on mRNA levels of endothelin receptors in marmosets. The ETAR and ETBR mRNA steady-state levels were measured in tissues of marmosets fed a high fat/high cholesterol diet. This diet resulted in an increase in the level of serum cholesterol to over 350 mg/dl (data not shown). Interestingly, such a treatment also resulted in a 2-3-fold increase in ETB mRNA in most of the tissues examined when com-Characterization of the Human Endothelin Receptors 3879 pared with control-fed animals, whereas there was little change in the level of ETAR mRNA (Fig. 7). The specific increase in the level of ETBR mRNA might result in increased levels of ETBR protein. Whether the increase in the number of ETBR contributes to, or is a cause of hypercholesterolemia is not known. Interestingly, ETR receptor number has been shown to increase in other disorders; for example, ETR levels increased in the kidneys of rats with cyclosporine-induced nephrotoxicity failure (67). Further studies will illuminate the importance of up-regulation of E T receptors in the development and maintenance of hypercholesterolemia in humans and the role of E T receptors in other human disease states.