Expression Cloning of a Human B, Bradykinin Receptor*

A cDNA clone encoding a human B, bradykinin receptor was isolated from a human embryonic lung fibro- blast cDNAlibrary by expression cloning. The photoprotein aequorin was utilized as an indicator of the ability of the B, receptor agonist [de~-Arg'~]kallidin to mediate Ca2+ mobilization in Xenopus luevis oocytes injected with RNA. A clone was isolated with a 1307-nucleotide insert which contains an open reading frame encoding a 353-amino acid protein with the characteristics of a G-protein-coupled receptor. The amino acid sequence of the B, bradykinin receptor is 36% identical to the amino acid sequence of the B, bradykinin receptor. The cloned B, bradykinin receptor expressed in mammalian cells exhibits high affinity binding for 3H-labeled [des- Arg"1kallidin low affinity for bradykinin. The B, receptor 3H-labeled cloned receptor, whereas the B, receptor antagonist Hoe-140 Thi is assays as described previously (27, 35). Displacement studies were done with 1 nM [des-Arg'01,[3,4-3Hlkal-lidin (DuPont NEN) in the presence of varying concentrations of com- petitor compounds. Binding assays were performed at room tempera-ture for 45 min. Reactions were terminated by filtration using either an Inotech cell harvester or a Tomtech cell harvester onto glass fiber filters that had been briefly soaked in 0.3% polyethyleneimine. The filters were washed with cold phosphate-buffered saline and counted either in an LKB Betaplate or Beckman liquid counter.

Two mammalian bradykinin receptor subtypes, B, and B,, have been defined based on their pharmacological properties (1, 2). The B, receptor is synthesized de novo following tissue injury and has recently been shown to mediate hyperalgesia in animal models of chronic inflammation (1). The B, bradykinin receptor is normally present in smooth muscle and certain neurons, where activation of B, receptors causes pronounced hypotension, bronchoconstriction, pain, and inflammation (1, 2). The agonists for the B, and B, bradykinin receptors are generated by the proteolytic action of kallikreins which release the nonapeptide bradykinin (BK)' and the decapeptide Lys-BK (kallidin) from large protein precursors, low and high molecular weight kininogen. BK and kallidin are equipotent agonists at the B, receptor. In contrast, BK is inactive at the B, bradykinin receptor subtype. Degradation of the B, receptor agonists by a carboxypeptidase produces the B, receptor agonists, [des-* 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.
to the GenBankTMIEMBL Data Bank with accession number(s) U12512.
Argg]BK and [de~-Arg'~lkallidin. The phenomenon of proteolytic transformation of a peptide from B, to B, selectivity has been observed not only for the endogenous kinin agonists but also for several synthetic peptide antagonists (3, 4).
The B, receptor was originally discovered through a contractile response to [des-ArgIBK that was observed in rabbit aortic strips only after a prolonged in vitro incubation (5-7). The de novo synthesis of B, receptors has been reported in vivo following treatment with bacterial lipopolysaccharide (8) and in animal models of antigen arthritis (9). In vitro studies have implicated a number of cytokines, most notably interleukin-1 (IL-1) and IL-2, as mediators that induce the expression of B, receptors (6, 10-12). Furthermore, the activation of a B, bradykinin receptor on mouse macrophages causes the release of cytokines (13,14). Significantly, the B, bradykinin receptor antagonist [des-Ar$,Leu']BK was recently found to alleviate hyperalgesia in animal models of persistent inflammation (1,15,16). Thus, a body of evidence implicates the B, bradykinin receptor in the pathophysiology of chronic inflammation. Relatively little is known about the role of the B, receptor in healthy tissues, although both B, and B, receptors may play a physiological role in renal function (17,18).
The cloning of the B, bradykinin receptor has revealed that this receptor is a member of the superfamily of G-proteincoupled receptors (19-221, definitive evidence that the B, receptor couples to G-proteins has not been forthcoming. The rat B, bradykinin receptor was cloned (19) using a Xenopus oocyte expression system that exploited the ability of the B, receptor t o act through G-proteins to activate phospholipase C and mobilize Ca2+ (23,24). Recently, the B, bradykinin receptor has also been shown t o activate phospholipase C in primary cultures of rabbit aorta smooth muscle cells, rabbit mesenteric artery smooth muscle cells, and rat mesangial cells (25)(26)(27). Furthermore, both B, and B, bradykinin receptor activities were detected when mRNA from the human fibroblast cell line W1-38 was injected into X. laevis oocytes (28,291. Although the similarity of ligands for the two bradykinin receptor subtypes suggests a similarity between the B, and B, receptor genes, the results of genomic Southern analyses indicated that these two receptors are not highly homologous (19,30). Therefore, to clone the human B, receptor, we pursued an expression cloning strategy inXenopus oocytes utilizing the photoprotein aequorin as an indicator of Ca2+ mobilization (31, 32). We isolated a cDNA clone that encodes a G-protein-coupled receptor with an amino acid sequence that is 36% identical to that of the B, bradykinin receptor. The pharmacological properties of this cloned receptor expressed in mammalian cells demonstrate that it is a B, bradykinin receptor.

MATERIALS AND METHODS
Oocyte Injections-Injection of mRNA or cRNA into Xenopus oocytes was performed by a modification of established protocols (33, 34). The excised ovarian lobes were teased apart with jeweler's forceps and then placed into OR-2 medium (82.5 mM NaC1,2 m M KCl, 1 mM MgCl,, 5 mM 21583 HEPES, pH 7.4) containing 2 mg/ml collagenase B (Boehringer Mannhiem) for 2 h a t room temperature. Oocytes were selected and cultured overnight in supplemented OR-2 (OR-2 containing 1.8 mM CaCl,, 0.5 mg/ml gentamycin, and 0.5 II~M theophylline). Initially, oocytes were injected with 46 nl of RNA at a concentration of 1 or 2 mg/ml in H,O. Once the pool size became less than 30 clones, the cRNA concentration was decreased to 40 ng/ml. RNA was injected using a Nanoject automatic oocyte injector (Drummond Scientific), and injection needles were pulled from 3.5-inch Drummond capillaries using a FlamingBrown Micropipette puller (Sutter Instruments). Two to three days after the RNA injection, oocytes were injected with 92 ng of aequorin (Friday Harbor Photoproteins) resuspended in 46 nl of 1 mM EDTA, as described previously (31, 32). The following day, individual oocytes placed in wells of a microtiter dish containing 225 pl OR-2 were challenged with peptide agonists, and the aequorin photo response was measured using a ML3000 microtiter plate luminometer (Dynatech).
RNA Fractionation-IMR-90 cells (ATCC CCL 186) were grown in minimal essential medium supplemented with 10% fetal calf serum, glutamine, nonessential amino acids, sodium pyruvate, penicillin, and streptomycin (Life Technologies, Inc.). Two and a half hours prior to mRNA extraction, IMR-90 cells were exposed to 200 pg/ml IL-lp (R & D Systems). mRNA was purified from these cells using the poly(A) tract mRNA isolation system (Promega) and resuspended in H,O at a concentration of 2 mg/ml. IMR-90 mRNA was size-fractionated on a continuous 6 2 0 % sucrose gradient in 15 mM PIPES, pH 6.5, 5 mM EDTA, and 0.25% N-lauroylsarcosine. 480 pg of mRNA from the IL-lp-induced IMR-90 cells was heat-denatured and size-fractionated by centrifugation a t 18 "C for 19 h a t 77,000 x g. Fractions (450 pl) from each gradient were collected from the bottom of the tube. Fractions were ethanol-precipitated twice and resuspended to a final concentration of 1 pg/pl. RNA size determination was based on the migration pattern of 80 pg of 9.49-0.24-kb RNA markers (Life Technologies, Inc.) loaded on a parallel gradient.
Library Construction-First strand cDNA synthesis of 3 pg of approximately 1.7-kb size-selected mRNA was primed with 50 ng of random hexamers and 400 ng of a NotI oligo(dT1 oligonucleotide and synthesized with the Life Technologies, Inc. Superscript I1 reverse transcriptase. Following second strand synthesis, BstXUEcoRI adaptors (Invitrogen) were ligated onto the ends, and the cDNA was passed over a Life Technologies, Inc. cDNA sizing column. The cDNA was cloned into pcDNA3 (Invitrogen) cut with BstXI. Plasmid DNA was transformed into XL-1 Blue cells (Stratagene).
Colonies were plated on ColonyPlaque screen filters (DuPont NEN) that were placed on Luria-Bertani (LB) agar plates supplemented with 100 pg/ml ampicillin (Sigma). Plasmid DNA was linearized with NotI, and cRNA was synthesized using T7 RNA polymerase with the mCAP RNA capping kit (Stratagene).
The DNA sequence of both strands of clone 3339 was determined by a combination of manual sequencing using Sequenase version 2.0 (U. S. Biochemical Corp.) and automated sequencing using a n A B 1 373A (Perkin-Elmer).
Mammalian Cell Expression and Pharmacological Characterization-COS-7 cells were transfected by electroporation using a Bio-Rad gene pulser. Three days post-transfection, cells were processed for either whole cell or membrane binding assays as described previously (27,35). Displacement studies were done with 1 nM [des-Arg'01,[3,4-3Hlkallidin (DuPont NEN) in the presence of varying concentrations of competitor compounds. Binding assays were performed at room temperature for 45 min. Reactions were terminated by filtration using either a n Inotech cell harvester or a Tomtech cell harvester onto glass fiber filters that had been briefly soaked in 0.3% polyethyleneimine. The filters were washed with cold phosphate-buffered saline and counted either in a n LKB Betaplate 1205 or a Beckman liquid scintillation counter.

RESULTS AND DISCUSSION
The human embryonic fibroblast cell line IMRDO had been shown previously to express the B, bradykinin receptor subtype (36). A more detailed pharmacological characterization revealed the presence of approximately 5000 high affinity bind- strated using Fura-2 as an indicator.' Based on these data, we chose IL-lp-induced IMR-90 cells as the source of mRNA for expression cloning. Injection of mRNA prepared from IL-10-induced IMR-90 cells into X. Zaevis oocytes resulted in aequorin-mediated luminescence in response to either the B, agonist [de~-Arg'~]kallidin or the B, agonist BK. The mRNA was size-fractionated over a sucrose gradient, fractions were injected into oocytes, and the oocytes were assayed for their ability to respond to either BK or [de~-Arg'~]kallidin. The B, and B, receptor transcripts were clearly separated by the size fractionation (Fig. 1). The mRNA mediating the response to [de~-Arg'~]kallidin exhibited an apparent size of 1.6-1.8 kb, whereas the mRNA mediating the response to BK had an apparent size of 4.4-4.6 kb. The apparent size of the mRNA enabling the BK response is consistent with the previously determined size of the B, bradykinin receptor transcript (19, 30). The RNA fraction from the sucrose gradient which gave the greatest response to [de~-Arg'~]kallidin was utilized to generate a cDNA library. The library contained greater than 90% inserts, with an average insert size of 1.9 kb. The library was plated in pools of approximately 5000 clones that were used to synthesize cRNA. Of the 25 pools of cRNA that were injected into Xenopus oocytes, 11 exhibited aequorin-mediated luminescence in response to [de~-Arg'~]kallidin. The pool that gave the most robust response was replated and fractionated into 25 pools of approximately 800 clones. Eight pools exhibited a response to [des-Arg'"]kallidin. The strongest positive pool was further examined using electrophysiology to monitor activation of the Ca2+ activated C1-channel (data not shown). [des-Arg'OIKallidin (20 m) produced a response that was blocked by prior incubation with the B, receptor antagonist [des-A~-g'~,Leu~]kallidin (20 nM). This pool was then subdivided into 32 pools of approximately 25 individual clones. Two positive pools, containing 14 and 34 clones, respectively, were identified. cRNA was prepared from individual clones and analyzed in Xenopus oocytes. Three individual clones were found to elicit a [de~-Arg'~]kallidin response. One clone, 3339 (Fig. 21, was chosen for further DNA sequence analysis, expression, and pharmacological characterization. R. W. Ransom

TALILTLVVAFLVCWAPYHFFAFLEFLFQVQAVRGCFWEDFIDLGLQLAN T V L V L V V L L L F I I C W L P F O I S T F L D T L H R L G I L S S C Q D E R I I D V~
1 , l : l . : : : 1 : : l l I::: . I I : I .:. : . : I I : I I : I : I .

FIG. 3. Comparison of the amino acid sequence of the human B, bradykinin receptor and the human B, bradykinin receptor.
The alignment was performed using the GAP program in the GCG software package. The seven putative transmembrane domains are underlined. The symbol * indicates a potential N-linked glycosylation site, A indicates potential protein kinase C phosphorylation sites, and V indicates potential CAMP-dependent protein kinase sites. The highly conserved cysteine residues that are proposed to be involved in a disulfide bond are connected by a dotted line.
Clone 3339 contains an insert of 1307 nucleotides with an open reading frame of 1059 nucleotides. We isolated several different clones that encompassed the same DNA sequence as clone 3339 but began and ended at different locations, indicating that they were independently derived. The sequence surrounding the proposed initiator methionine codon at nucleotide 209 conforms to the Kozak consensus sequence in the +4 position but not at the -3 position (37). The open reading frame encodes a protein that is 36% identical to the B, bradykinin receptor (Fig. 3). The sequence identity at the nucleotide level, 54%, probably explains the failure to clone this receptor by low stringency hybridization with DNA encoding the B, receptor. A homology search of the Swiss Protein data base indicates that the B, receptor is 30% identical to the angiotensin type 2 receptor and 29% identical to the angiotensin type 1 receptor (38-40) and less homologous to other G-protein-coupled receptors. A Kyte and Doolittle hydrophobicity plot of the amino acid sequence reveals the potential for the seven transmembrane domains that are characteristic of G-protein-coupled receptors (41,421. Two conserved Cys residues that are proposed to form a disulfide bond between the second and third extracellular domains in nearly all G-protein-coupled receptors are also present in this sequence (Fig. 3). There are two consensus sites for N-linked glycosylation in the NH,-terminal domain and one in the third extracellular domain. Potential phosphorylation sites for protein kinase C and CAMP-dependent protein kinase are present in intracellular domains 2 and 3 and the carboxylterminal domain. Similar potential phosphorylation sites in other G-protein-coupled receptors have been implicated in short term desensitization of the receptor following agonist stimulation (41,42).
Clone 3339 was transfected into COS-7 cells, and the pharmacological properties of the expressed receptor were determined. Scatchard analysis of saturation binding data with 3Hlabeled [de~-Arg~~]kallidin indicated a Kd of 0.4 nM and a B,, of approximately 100 fmoVmg of protein (Fig. 4A). Mock-transfected COS-7 cells did not demonstrate any specific binding for 3H-labeled [des-Arglo]kallidin (data not shown).
Both the cloned human B, receptor and the B, receptor in IMR-90 cells exhibit a relatively low affinity for the "classical" B, receptor agonist [des-Ar$]BK, which has an affinity of 42 nM for the B, receptor in rabbit aorta (27). The B, bradykinin receptor present in human aorta also has a relatively low affinity of 1 PM for [des-Ar$]BK.' However, like the cloned human B,, the B, receptor in rabbit aorta has a lower affinity for [des-ArglBK than [des-ArglO]kallidin (7, 27). Thus the most potent natural ligand for the B, receptor appears to be [des-Arg'o]kallidin. Based on the pharmacological profile outlined in Table I, we believe it is likely that the B, receptor isolated here is the human homolog of the B, receptor present in rabbit aorta and that the lower affinity of the human receptor for [des-Ar$lBK may be a consequence of species differences.
The ability of several bradykinin receptor antagonists to displace 1 n M 3H-labeled [de~-Arg'~]kallidin from the cloned receptor was also analyzed. The cloned B, receptor has relatively high affinity binding for the B,-specific antagonists [des-Arglo,Leug]kallidin and [des-Ar$,LeuslBK (Table I). By contrast, the cloned receptor has a very low affinity for the potent B,-specific antagonist Hoe-140. Significantly, the removal of the COOH-terminal Arg from Hoe-140 results in a dramatic increase in affinity (Table I), as would be expected for a B, receptor (3). Therefore, the interaction of the cloned receptor with bradykinin antagonists is consistent with the B, receptor subtype classification. In summary, we have utilized an expression cloning strategy to isolate a clone encoding a human B, bradykinin receptor. This receptor was isolated by its ability, when expressed in Xenopus oocytes, to functionally respond to the B, receptor agonist [de~-Arg'~]kallidin. The cloned receptor is a G-proteincoupled receptor that is most similar in amino acid sequence to the B, bradykinin receptor. The pharmacological properties of the cloned receptor expressed in mammalian cells are characteristic of the B, bradykinin receptor classification. The B, bradykinin receptor has been implicated in chronic inflammation and hyperalgesia, whereas the B, receptor appears to mediate acute inflammatory and algesic responses. The availability of cloned human B, and B, receptors should lead to a greater understanding of the role of these receptors in both normal and pathophysiological conditions.