Molecular characterization of a novel human endothelin receptor splice variant.

Endothelin receptors are widely distributed throughout a number of tissues. A novel ETB receptor splice variant (ETB-SVR) was identified from a human placental cDNA library. Sequence analysis indicated that the ETB-SVR is 436 amino acids long and shares 91% identity to the human ETB-R. Northern blot analysis indicated an mRNA species of 2.7 kilobases, which is expressed in the lung, placenta, kidney, and skeletal muscle. Ligand binding studies of the cloned ETB-SVR and ETB-R receptors expressed in COS cells showed that ET peptides exhibited similar potency in displacing 125I-ET-1 binding. Functional studies showed that ET-1, ET-3, and sarafotoxin 6c displayed similar potencies for inositol phosphates accumulation in ETB-R-transfected COS cells, whereas no increase in inositol phosphate accumulation was observed in ETB-SVR-transfected cells. In addition, exposure of ETB-R-transfected cells to ET-1 caused an increase in the intracellular acidification rate whereas ETB-SVR-transfected cells did not respond to ET-1. These data suggest that the ETB-SVR and ETB-R are functionally distinct and the difference in the amino acid sequences between the two receptors may determine functional coupling. Availability of cDNA clones for endothelin receptors can facilitate our understanding of the role of ET in the pathophysiology of various diseases.

Endothelins are a family of peptide hormones having profound cardiovascular, mitogenic, and potential neuroregulatory functions. In mammals, the ET 1 peptide family is composed of three members, ET-1, ET-2, and ET-3, that are encoded by three separate genes, which are differentially expressed in the tissues of the periphery and central nervous system (for reviews, see Refs. 1 and 2). Mammalian ETs share high sequence homology and structural similarity with a family of 21 amino acid peptide toxins from the snake Atractaspis engaddensis, the sarafotoxins (3).
Two major subtypes of ET receptors (ET A and ET B ) (4) have been identified based on the rank order potency of ET-1, ET-2, ET-3, and S6c (5,6). ET A receptor is 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 S6c. In contrast, the ET B receptor subtype displays equal high affini-ties for all ET-related peptides. Two additional receptors have been cloned and characterized from Xenopus melanophores (ET C ) and heart (ET AX ) (7,8). While ET C receptors display high affinity for ET-3 compared to ET-1, ET AX receptors displayed extremely weak affinity for BQ123 as well as S6c (ET A -and ET B -selective ligands, respectively). Receptors for ET are differentially expressed in a wide variety of tissues and cell types (9,10). ETs and sarafotoxins 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 Ca 2ϩ levels (1,11).
ET mediates a number of physiological effects including vasoconstriction mitogenesis, and induction of c-fos transcription (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23). 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 molecular biological approaches to dissect the components of ET physiology. Several laboratories have postulated the presence of additional ET receptors to account for the diverse biochemical and physiological activities of various ETs (24). This hypothesis has been supported by binding as well as functional studies. We have previously reported the cloning, functional characterization, and regulation of the human ET A and ET B receptor subtypes (25). In this report, we describe the cloning and functional characterization of a novel ET B receptor splice variant from human placenta.

EXPERIMENTAL PROCEDURES
Construction and Screening of the cDNA Libraries-The porcine cerebellum cDNA library (26) in pcDNA vector was screened by hybridization to nitrocellulose replicates using 32 P-labeled porcine ET B -R cDNA coding sequence as a probe in 20% formamide, 5 ϫ SSC (SSC is 150 mM NaCl, 15 mM sodium citrate), 5 ϫ Denhardt's, 0.1% SDS, and 0.2 mg/ml Escherichia coli tRNA at 42°C (27). Filters were washed with 2 ϫ SSC, 0.1% SDS at 42°C. Several positive recombinant clones were isolated from the porcine cerebellum library and characterized. Preliminary sequence analysis of these clones showed that 6 of these clones encode the ET B -R clones except two clones which contain the same 5Ј-coding region of ET B -R and a diverge 3Ј-coding sequences. In order to obtain the human homologue and eliminate the possibility that this variant is due to a recombination artifact, primers were synthesized corresponding to the amino termini of ET B -R (5Ј-TCTGGAGCAGGATCCAGCAT-GCAGCCGCCT-3Ј) and the 3Ј-untranslated region of the porcine ET B -SVR (5Ј-CCCGTGATATCTAAA TAGAATCCATATGGTGTG-3Ј) and used to obtain the human full-length splice variant cDNA clone from human placental RNA using polymerase chain reaction (28).
Nucleotide Sequence Analysis-The inserts of the porcine ET B -SVR560 and the human ET B -SVR cDNAs were sequenced on both strands using a modification of the dideoxy chain termination method (29) using the Sequenase II Kit (U. S. Biochemical Corp.). The Wisconsin Genetics Computer Group Software package (30) was used to assemble composite sequences from the various fragments and for future sequence analysis.
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Expression of Human ET B Receptors in COS Cells-Fragments containing the entire human ET B -SVR cDNA and ET B -R cDNA coding sequences was subcloned into the mammalian expression vector pRLDN (25). COS cells grown in 245 ϫ 245-mm tissue culture plates were transfected with 75 g of human ET B -SVR or human ET B -R cDNA and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum for 2 days as described previously (26).
Membrane Preparation-The COS cells were washed with Dulbecco's phosphate-buffered saline containing a protease inhibitor mixture (5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, and 0.1 g/ml aprotinin) and scraped in the same buffer. The membranes were prepared as described previously (25). Protein content was determined by the bicinchoninic acid method using bovine serum albumin as the standard.
Radioligand Binding Studies-125 I-ET-1 binding to membranes prepared from COS cells transfected with ET B -R or ET B -SVR was performed as explained (25) with the following modifications. The assay volumes were 50 l, and the protein was 50 -100 ng/tube. The procedure for 125 I-ET-3 and 125 I-IRL-1620 binding was exactly the same as that for 125 I-ET-1. Each experiment was done three times with separate transfections. The data presented are from one experiment which is repre-sentative of the others. Intra-experimental variation was Ͻ10%. 125 I-ET-1, 125 I-ET-3, and 125 I-IRL-1620 (specific activities 2200 Ci/mmol) were obtained from DuPont NEN.
Phosphoinositide Turnover-COS cells transfected with human ET B -R or ET B -SVR receptor clones were treated with 1 Ci/ml myo-[ 3 H]inositol for 24 h in serum-free medium. At the end of treatment, medium was removed, cells were washed with Dulbecco's phosphatebuffered saline, and then exposed to indicated concentrations of various agonists for 10 min at 37°C. The inositol phosphates were separated using ion exchange chromatography following the procedure of Aiyar et al. (34).
Microphysiometry-The cytosensor microphysiometer was used to measure the intracellular pH change with a pH-sensitive silicon sensor which is part of a microvolume flow chamber (35). COS cells transfected with vector or human ET B R or ET B -SVR were allowed to attach in capsule cups (Molecular Devices, Sunnyvale, CA) for 24 h before trans- ferring to the sensor chamber in the microphysiometer (Molecular Devices). After establishing a steady base line, acidification rate in response to ET was measured as a change in pH over time (35).

RESULTS AND DISCUSSION
Cloning of a cDNA Encoding the Human ET B Splice Variant-The porcine ET B -R cDNA, previously cloned in our laboratory (26) was used to probe porcine cerebellum cDNA library. Several positive clones were identified. Nucleotide sequence analysis revealed that several of these positive clones encoded for ET B -R except for two novel ET B -R clones which differed in the length and the amino acid sequence at the 3Ј-coding region. In order to determine the presence of this novel ET B -R form in human tissues, oligonucleotide primers corresponding to the 5Ј-coding region of ET B -R and the 3Ј-noncoding region of the novel ET B receptor form were used to polymerase chain reaction the human homologue using human placental cDNA as a template. Agarose gel electrophoresis identified a polymerase chain reaction product of 1400 base pairs. Sequence analysis of this product revealed a sequence that was identical to ET B receptor from the 5Ј-untranslated region through the putative seventh transmembrane domain, and then completely differed in the cytoplasmic domain and 3Ј-untranslated region. The deduced polypeptide consisted of 436 amino acid residues with a calculated molecular mass of approximately 48 kDa (Fig. 1, A  and B). This is different from the size of ET B receptor which is 442 amino acids with a calculated molecular mass of 49 kDa. Several independent clones in our library were found to contain this sequence which appears to represent alternative splicing of the carboxyl-terminal tail of the ET B receptor. The carboxylterminal region of the ET B -SVR clone bore no significant homology with other known proteins. Several lines of evidence support the fact that the ET B -SV and the ET B receptors represent alternatively spliced variants of a single gene. First, the 3Ј terminal tail of the ET B -SVR clone was used to search the data base for identical sequences. Interestingly this sequence was found to correspond to the extreme 3Ј-untranslated region of ET B -R. Further analysis of the ET B -R genomic structure (36) revealed that exons 1 through 6 are identical in ET B -SVR and the ET B -R in terms of their nucleotide sequence composition and their splice site. However, the 3Ј terminal tail of the ET B -SVR represents a putative splice site and an A(G/G)T sequence is found at the 5Ј donor junction which showed that this region is a splice site for the ET B receptor (Fig. 2). Therefore, two putative splice sites were identified at exon 7 in the human ET B receptor gene: the first splice site at nucleotide 1194 to produce a 2855-base pair exon which encodes the normal ET B -R; and the second splice site at nucleotide 2970 to produce a 997-base pair exon which encodes the ET B -SVR. Second, an mRNA transcript of 2.7 kb corresponds to the size of the ET B -SV receptor was identified in various human tissues (Fig. 3).
Receptor subtypes can arise through divergent genes, e.g. ET A and ET B and in the case of intron-containing genes, additional variants within a subtype can arise by alternative RNA splicing. Recent studies have identified an ET B receptor variant which contains an additional 10 amino acids in the second cytoplasmic domain of the ET B receptor (37). This sequence was part of the ET B receptor intron that separates the second and the third exons and therefore arises by alternative RNA splicing of a single gene. The identification of splice variants among the seven transmembrane receptors has been increasing at a phenomenal rate following the initial observation of two variant forms for dopamine D2 receptor (38). Recent splice variants have been identified for thyrotropin stimulating hormone, TRH (39), neurokinin receptors (40), prostaglandin EP3 receptor (41,42), pituitary adenylyl cyclase-activating polypeptide, PACAP receptor (43), and monocyte chemoattractant protein (MCP-1) receptor (44). It is interesting to note that in all these examples the cytoplasmic domain of the receptor is altered indicating that alternative RNA splicing plays an important role in the generation of physiologically divergent receptor activity for the same ligand.
mRNA Size and Tissue Distribution-Northern hybridization analysis of RNA from various human tissues with the human ET B -SV receptor cDNA clone indicated the presence of two major mRNA species at 4.4 and 1.7 kb and one minor band at 2.7 kb. The two major bands correspond to the full-length ET B receptor clones resulted from two polyadenylation signals which are 32 and 29 base pairs upstream of the polyadenylation sites of the two mRNA species. The minor band of 2.7 kb corresponds exactly to the size of the splice variant ET B receptor.
Northern blot was also used to examine the pattern of expression of the ET B and the ET B -SV receptor mRNA derived from different human tissues. As shown in Fig. 3, mRNA expressed by each ET B -R species were specifically detected in all the RNA samples. The total amount of ET B -R mRNA levels appeared to be higher in the lung, liver, kidney, and placenta. However, the relative ratio of expression of each individual ET B -SVR mRNA does not represent more than 10% of the total ET B -R expression, except the skeletal muscle where ET B -SVR contributes to more than 39% of the total ET B receptor expression (Fig. 3). In addition, the mRNA levels of normal as well as splice variant ET B -R were examined by Northern blot as well as reverse transcriptase-polymerase chain reaction methods in smooth muscle and endothelial cells. While both displayed normal ET B -R mRNA, the ET B -SVR mRNA was not detectable in both cells (data not shown).
Ligand Binding Properties-Addition of increasing concentrations of 125 I-ET-3 or 125 I-IRL-1620 to membranes prepared from COS cells transfected with ET B -R or ET B -SVR clone resulted in specific, saturable and high affinity binding as shown in Figs. 4 and 5. The nonspecific bindings were 5-25 and 5-35% for 125 I-ET-3 and 125 I-IRL-1620, respectively (Figs. 4, A and B,  and 5, A and B, respectively). The Scatchard transformations of the specific binding from the saturation binding experiments are shown as Figs. 4C and 5C for ET B -R and ET B -SVR, respectively. The apparent dissociation constants (K d values) for 125 I-ET-3 and 125 I-IRL-1620 were 53 and 145 pM for ET B -R and 42 and 68 pM for ET B -SVR. The maximum density of the receptors were comparable for both receptors with both radioligands (Figs. 4C and 5C), indicating that the expression of these two receptors in COS cells is comparable.
Competition binding data using 125 I-ET-1 and increasing concentrations of unlabeled ET-1, ET-3, S6c, and BQ123 (ET Aselective antagonist) obtained with ET B -R and ET B -SVR are shown in Fig. 6, A and B. While ET-1, ET-3, and S6c gave superimposable monophasic competition curves with both receptors, BQ123 was totally inactive up to 10 M in both receptors (Fig. 6), indicating that the binding properties of ET B -SVR are very similar to those observed with ET B -R. Similar binding profiles were obtained when the competition binding experiments were done with the ET B -selective radioligand, 125 I-IRL-1620, and the ET B -selective peptide antagonist, BQ788, and the nonselective nonpeptide antagonists, SB 209670 and SB 222802 (Fig. 6, C and D). Thus, the data presented so far clearly indicate that the modifications present at the COOH-terminal end of the receptor did not have any effect on the binding of the radioligands ( 125 I-ET-1, 125 I-ET-3, and 125 I-IRL-1620) or ET Bselective peptide antagonist, ET A -selective peptide antagonist, as well as nonselective-nonpeptide antagonists.
Since the intracellular regions of seven transmembrane G protein-coupled receptors have been implicated in the coupling of these receptors to the signal transduction pathway, it was of interest to test whether there were any differences in the coupling of these two variants to signal transduction pathways. ET receptors have been shown to activate phospholipase C, resulting in the generation of diacylglycerol and inositol triphosphate which, in turn, releases intracellular calcium. Exposure of myo-[ 3 H]inositol-prelabeled COS cells expressing ET B -R to increasing concentrations of ET-1, ET-3, or S6c resulted in a concentration-dependent increase in inositol phosphate accumulation as shown in Fig. 7A. The maximum stimulation (50 -60% over basal) obtained with the three agonists as well as the EC 50 values (0.2-1.0 nM) of these three agonists for stimulation of inositol phosphate accumulation were very similar. In contrast, the COS cells transfected with ET B -SVR and prelabeled with myo-[ 3 H]inositol did not respond to any of the three agonists tested (Fig. 7B). Since ET peptides were ineffective in stimulating inositol phosphates accumulation in ET B SVR-transfected COS cells, a more sensitive method, such as cytosensor microphysiometer, was used to test whether these receptors were capable of inducing a change in intracelluar pH in response to ET. As shown in Fig. 7C, COS cells

FIG. 7. Inositol phosphate accumulation in ET B -R (A) and ET B -SVR (B) transfected COS cells in response to ET-1, ET-3, and
S6c. myo-[ 3 H]Inositol-prelabeled cells were exposed to increasing concentrations of ET-1, ET-3, or S6c for 10 min at 37°C, and the reactions were stopped with trichloroacetic acid. The inositol phosphates were quantitated as explained under "Experimental Procedures." C, intracellular acidification rate in response to ET-1. COS cells transfected with vector alone (M.T.) or human ET B -R or ET B -SVR were challenged with ET-1 and the change in intracellular acidification rate was measured using microphysiometer. transfected with human ET B -R gave a nice increase in the acidification rate, whereas COS cells transfected with vector alone or ET B -SVR gave very little response. These data agree well with the data obtained from the inositol phosphates accumulation and indicate that even though the splice variant displayed the same binding properties as the wild type, this receptor was incapable of stimulating a functional response. Similar results have been reported by Kuang et al. (45) for chemokine receptors MCP-1Ra and MCP-1Rb. These two receptors are alternately spliced receptors, and the differences between the two is the COOH-terminal intracellular domain. While MCP-1Rb coupled to both G␣16 and G␣14, MCP-1Ra did not couple to either of these G proteins. Thus, the data presented here and in the literature strongly suggest that the intracellular COOH-terminal domain of the 7TM receptors is critical for G protein coupling and functional responses. It is clear from the amino acid sequence comparison of the two receptors that the only difference between the ET B -R and ET B -SVR is the proximal 52 amino acids of the COOH-terminal end. The amino acid composition of the wild type has 5 cysteines, 9 serines, 2 tyrosines, and no threonine, whereas the splice variant has 1 cysteine, 2 serines, 2 threonines, and no tyrosines. The importance of these key amino acids (involved in palmitoylation as well as phosphorylation) in the coupling to signal transduction pathway is not clear. In an elegant study to identify the functional domains of human ET A receptor, Adachi et al. (46) have reported that the regions involved in signal transduction for human ET A receptor are amino acids 296 -305 and 373-385. Comparison of human ET A and ET B in these regions indicate that these regions are Ͼ95% identical in both subtypes; suggesting the significance of this region in the signaling mechanism. Further analysis of the 52 amino acids located at the COOH-terminal tail of ET B -R identify the sequence SCLC which is part of the 373-385 conserved sequence of ET B -R and ET A -R. Interestingly, this sequence in the ET B -SVR was replaced by AGPH. These data suggest that these four amino acids are critical for the signal transduction of these receptors.
The main unanswered question is what is the physiological role of the ET B -SVR receptor in the cell? It is tempting to hypothesize that this receptor might function as a clearance receptor for the ET peptides, since there was no observed difference in the binding characteristics between the two receptors. Another possibility is that ET exhibits many diverse physiological effects, and it is possible that the ET B -SVR could mediate another, yet unidentified, response. It has been demonstrated recently that prostaglandin E 2 receptor has four alternatively spliced carboxyl-terminal tails (37)(38). Interestingly, the four isoforms couple to different G proteins to activate different second messengers.
Several laboratories have demonstrated that cells expressing seven transmembrane receptors could undergo homologous desensitization in response to repeated challenges with agonists (47,48). The mechanism of homologous desensitization to ET-1 is not well understood because the binding of ET-1 to its receptor is irreversible which makes it rather difficult to study down-regulation after prolonged exposure to agonists. One commonly used mechanism for inactivation of seven transmembrane receptors is phosphorylation of the serines and/or threonines in the carboxyl-terminal tails by the ␤-adrenergic receptor kinase, which belongs to a multigene family containing six known subtypes and are called G-protein-coupled receptor kinases (49 -51). It is of interest in this regard that the ET B -R has a total of 9 serines, whereas the ET B -SVR has a total of 2 serines and 2 threonines. Differences in receptor phosphorylation by ␤-adrenergic receptor kinase-like enzyme might lead to differences in the pattern of deactivation between the two forms of the ET B -receptors, providing another mechanism for increasing the diversity of cellular responses to ETs. The availability of these clones will facilitate investigation of the potential roles of ET B -SVR and its possible participation in physiological and pathophysiological processes.