Characterization of the Human Cysteinyl Leukotriene 2 Receptor*

The contractile and inflammatory actions of the cysteinyl leukotrienes (CysLTs), LTC4, LTD4, and LTE4, are thought to be mediated through at least two distinct but related CysLT G protein-coupled receptors. The human CysLT1 receptor has been recently cloned and characterized. We describe here the cloning and characterization of the second cysteinyl leukotriene receptor, CysLT2, a 346-amino acid protein with 38% amino acid identity to the CysLT1 receptor. The recombinant human CysLT2 receptor was expressed in Xenopusoocytes and HEK293T cells and shown to couple to elevation of intracellular calcium when activated by LTC4, LTD4, or LTE4. Analyses of radiolabeled LTD4 binding to the recombinant CysLT2 receptor demonstrated high affinity binding and a rank order of potency for competition of LTC4 = LTD4 ≫ LTE4. In contrast to the dual CysLT1/CysLT2 antagonist, BAY u9773, the CysLT1 receptor-selective antagonists MK-571, montelukast (SingulairTM), zafirlukast (AccolateTM), and pranlukast (OnonTM) exhibited low potency in competition for LTD4 binding and as antagonists of CysLT2receptor signaling. CysLT2 receptor mRNA was detected in lung macrophages and airway smooth muscle, cardiac Purkinje cells, adrenal medulla cells, peripheral blood leukocytes, and brain, and the receptor gene was mapped to chromosome 13q14, a region linked to atopic asthma.

The cysteinyl leukotrienes (CysLTs), 1 LTC 4 , LTD 4 , and LTE 4 , previously known as slow reacting substance of anaphylaxis, or SRS-A, are derived from arachidonic acid via oxygenation and dehydration by 5-lipoxygenase followed by specific glutathione addition by LTC 4 synthase (1). The CysLTs medi-ate their biological actions through two pharmacologically defined G-protein-coupled receptors (GPCRs), named the CysLT 1 and CysLT 2 (2,3). The recent cloning and characterization of the human CysLT 1 receptor confirmed the previous pharmacological data (Refs. 4 and 5; GenBank TM accession nos. AF 119711 and AF 133266). LTD 4 is the preferred endogenous ligand for the CysLT 1 receptor, and activation of the receptor results in an elevation of intracellular calcium (4,5). The gene for the CysLT 1 receptor has been mapped to human chromosome Xq13-q21 (4,5). The CysLT 1 receptor is the molecular target of the anti-asthmatic drugs montelukast (Singulair TM ), zafirlukast (Accolate TM ), and pranlukast (Onon TM ) that have both anti-bronchoconstrictive and anti-inflammatory actions (6 -8). All known CysLT receptor antagonists, except BAY u9773 (a nonselective antagonist at CysLT 1 and CysLT 2 receptors) selectively antagonize activation of the CysLT 1 receptor (6 -9). CysLT 1 receptor mRNA and protein are expressed on human lung smooth muscle cells and tissue macrophages and on peripheral blood monocytes and eosinophils (4). 2 The CysLT 2 receptor has been documented pharmacologically to be expressed in guinea pig trachea and ileum, ferret trachea and spleen, sheep bronchus, and human pulmonary and saphenous vein preparations (2,10,11). At the CysLT 2 receptor subtype, the agonist potency rank order is LTC 4 ϭ LTD 4 Ͼ Ͼ LTE 4 , and LTE 4 is a partial agonist (2). We describe here the molecular cloning and characterization of the human CysLT 2 receptor. 4 and LTC 4 were from Cayman (Ann Arbor, MI); LTB 4 , LTE 4 , and BAY u9773 were from BIOMOL (Plymouth Meeting, PA); 1-oleoyl lysophosphatidic acid was from Avanti Polar Lipids (Alabaster, AL). MK-571, montelukast, pranlukast, and zafirlukast were synthesized by the Department of Medicinal Chemistry at Merck Frosst, and [ 3 H]LTD 4 (146 Ci/mmol) was from NEN Life Science Products.

Materials-LTD
Cloning of HG57, the CysLT 2 Receptor-A partial rat EST (accession no. ai178926) encoding a GPCR fragment, with 40% identity to the human CysLT 1 receptor, was found during a routine search of updates to the GenBank TM data base of expressed sequence tags using the FAST_PAN program (12). The cognate cDNA was retrieved from the American Type Culture Collection and was used to screen a rat brain cDNA library as reported previously (13). Two cDNAs were obtained that had a translational open reading frame of 981 nucleotides. This putative receptor, designated by the expressed sequence tags clone name, RSPBT32, was 40% identical to the human CysLT 1 receptor but 16 and 21 amino acids shorter at the N-and C-terminal ends, respec-* 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  tively. The RSPBT32 protein sequence was added to the FAST_PAN GPCR query set, and during a subsequent search of GenBank TM releases, we found a human genome survey sequence deposition (accession no. aq001459) encoding a peptide fragment that was about 70% identical to RSPBT32. The genome survey sequence was used to design oligonucleotide primers for the polymerase chain reaction, and ultimately we isolated a human genomic DNA fragment that contained a 1041-nucleotide open reading frame encoding a 346-amino acid protein that we named HG57. The amplified product was subcloned into the TA cloning site of the mammalian cDNA expression vector pCR3.1 (Invitrogen, Carlsbad, CA), and the sequence of the inserted DNA was determined.
Xenopus laevis Oocyte Expression-HG57 (CysLT 2 ) cRNA was produced by in vitro transcription (T7 mMessage mMachine, Ambion) using T7 RNA polymerase in the presence of a capping analogue. 20 -40 ng of the capped cRNA was injected into stage V-VI X. laevis oocytes as described previously (14,15). Compounds were delivered as a 30-l aliquot over a period of 1-2 s, and the recording chamber was washed with 30% methanol between oocyte assays to remove residual ligand. X. laevis were purchased from Xenopus I (Ann Arbor, MI).
Aequorin Luminescence Functional Assay-Human embryonic kidney (HEK) 293 cells stably expressing the SV40 large T antigen, designated HEK 293T cells, were transfected with HG57 (CysLT 2 )-pCR3.1 or pCR3.1 and AEQ-pCDM plasmids (5 g of each DNA per 75-cm 2 culture flask), using the LipofectAMINE™ PLUS reagent (Life Technologies, Inc.) following the manufacturer's instructions. Cells were prepared subsequently for use in the aequorin luminescence functional assay as described previously (16). Briefly, agonists in dimethyl sulfoxide or ethanol were serially diluted, in duplicate, in a white 96-well cliniplate FB (Labsystems) using a Biomek in a final volume of 100 l in phosphate-buffered saline (with 70 mg of CaCl 2 ) so that the final solvent concentration was Յ1%. The plate was then loaded into the Luminoskan RS plate-reading luminometer (Labsystems, Needham Heights, MA), and wells were tested sequentially. Cells (ϳ2-5 ϫ 10 4 ) in 100 l of Ham's F12 medium were injected into the well, and light emission was recorded over 30 s (peak 1). The cells were then lysed by injection of 25 l of 0.9% Triton X-100 solution in H 2 O, and light emission was measured for an additional 10 s (peak 2). For the antagonist, the plate was divided into four series. Series 1 contained a leukotriene C 4 or leukotriene D 4 control curve into which transfected cells were injected in the absence of the antagonist, while series 2-4 contained leukotriene C 4 or leukotriene D 4 curves, into which cells that had been preincubated for 15 min in the presence of a given concentration of antagonist were injected.
Fractional luminescence for each well was determined by dividing the area under peak 1 by the total area under peaks 1 and 2 (P1/(P1 ϩ P2)). These calculations were performed using the Lskan Controller program, and data files were analyzed using the LDAM software employing a modified version of the Levenberg-Marquardt four-parameter curve-fitting algorithm to calculate EC 50 values (14). For the antagonist tests, the EC 50 values were then used to generate a K B value using Schild plot analysis.
Radioligand Binding Assays-COS-7 cell transfection, harvesting, and membrane preparation were done as previously reported (17). Tritiated LTD 4 binding assays were performed as described (17) with the exception that the reaction was initiated by the addition of 250 g of membrane protein and that 4 mM acivicin replaced serine-borate in equilibrium competition assays with LTC 4 . LTD 4 -specific binding was calculated by subtracting nonspecific binding, determined in the presence of 1 M LTD 4 , from total binding. Specific binding accounted for 50 -60% of the total binding and was linear with respect to the concentrations of radioligand and protein present in the incubation. Total binding represented less than 10% of the radioligand added to the incubation. In these conditions, average total and nonspecific binding of [ 3 H]LTD 4 for 250 g of HG57 (CysLT 2 ) receptor membrane protein was 2600 and 1200 dpm, respectively.
Northern Blot Analysis-RNAs from several human tissues were extracted as described previously (18). Total RNA was extracted by the method of Chomczynski and Sacci (19), and poly(A) ϩ RNA was isolated using oligo(dT)-cellulose spin columns (Amersham Pharmacia Biotech). RNA was denatured and size-fractionated on a 1% formaldehyde-agarose gel, transferred onto nylon membrane, and immobilized by UV irradiation. The blots were hybridized with a 32 P-labeled DNA fragment encoding the CysLT 2 receptor, washed with 2ϫ SSPE and 0.1% SDS at room temperature for 15 min, washed again with 0.5ϫ SSPE and 0.1% SDS at 50°C for 45 min, subjected to four final washings with 0.2ϫ SSPE and 0.1% SDS at 50°C each for 1 h, and exposed to x-ray film at Ϫ70°C in the presence of an intensifying screen for 3-7 days.
In Situ Hybridization Analysis and Immunohistochemistry-The oligonucleotide antisense probes used for the in situ studies were 5Ј-C-CAGGCACTCCTGATGCT-3Ј and 5Ј-CCCACCACCAAGGAATA-3Ј, and their complementary sequences were used as sense probes. Tailing of oligonucleotides with biotin-16-dUTP (Roche Molecular Biochemicals) was carried out as described by the manufacturer except for substitution of digoxigenin-dUTP with biotin-16-dUTP in the tailing reaction. In situ hybridization was carried out on 6-m cryostat sections of surgical human lung specimens (National Disease Research Interchange, Philadelphia, PA) using 2 pmol/ml labeled oligonucleotides for 18 h at 37°C. Bound probe was visualized using Texas Red Tyramide Signal Amplification detection reagent (NEN Life Science Products) according to the manufacturer's instructions. Peripheral blood mononuclear cells were isolated from buffy coat preparations by centrifugation over lymphocyte separation medium (ICN). T cells were rosetted by incubation of the peripheral blood mononuclear cells with neuraminidase-treated sheep red blood cells and pelleted through lymphocyte separation medium. The sheep red blood cells were removed by lysis with ACK lysis buffer (Life Technologies). T cell-depleted peripheral blood mononuclear cells accumulated at the interface of the lymphocyte separation medium. Eosinophils were prepared from peripheral blood from a donor known to have elevated peripheral blood eosinophils but no history of asthmatic disease. Erythrocytes were removed by hypotonic lysis of the pelleted cells, followed by negative selection with anti-CD16 microbeads (Miltenyi Biotech) according to the manufacturer's instructions. The purity of eosinophil preparations was Ͼ90% with some contaminating monocytes and neutrophils. All preparations of cells were resuspended in O.C.T. compound (Miles Scientific) and fresh frozen prior to in situ hybridization. In situ hybridization on these cell preparations was carried out exactly as described above for tissue sections. Sections and cells were counterstained with 4,6-diamidino-2phenylindole (Molecular Probes, Inc., Eugene, OR) to visualize cell nuclei, and images were digitally acquired and reassembled using a MicroMax CCD camera (Princeton Instruments) and the Metamorph imaging program (Universal Imaging). Cell nuclei are seen in teal blue pseudocolor or in conventional 4,6-diamidino-2-phenylindole blue. Immunohistochemistry was performed after in situ hybridization, with antibodies LN5 (Zymed Laboratories Inc.), CD14 (Serotec), and Protein Gene Product 9.5 (Biomeda Corp.) at the manufacturer's specified dilutions for 2 h at room temperature. Primary antibody markers were detected with either a fluorescein isothiocyanate-conjugated donkey anti-mouse IgG (Jackson Immunoresearch) or a fluorescein isothiocyanate-conjugated donkey anti-rat IgG (Jackson Immunoresearch). Colocalization of red (in situ) and green (antibody) staining is seen as yellow fluorescence.
Chromosomal Localization-Chromosomal mapping of the CysLT 2 receptor gene was performed using the GeneBridge 4 panel, consisting of 93 radiation hybrid clones (Research Genetics, Huntsville, AL). Two primer pairs were designed from the HG57 sequence to amplify each of two different regions of the gene. Primer pair 1 sequences (5Ј-AATGG-CACCTTCAGCAATAACA-3Ј (forward) and 5Ј-TGGACAACCCATTTC-CCAAGAC-3Ј (reverse)) produced a band of ϳ100 base pairs; primer pair 2 sequences (5Ј-AGACTGCATAAAGCTTTGGTTATC-3Ј (forward) and 5Ј-ATACTCTTGTTTCCTTTCTCAACC-3Ј (reverse)) produced a band of ϳ180 base pairs. Polymerase chain reaction was performed with AmpliTaq Gold (Perkin-Elmer) using the following cycling parameters: 94°C for 9 min; 94°C for 2 min, 62°C for 30 s, 72°C for 1 min (32 cycles), and 72°C for 7 min. Results were submitted to the Whitehead Institute Genome center server (available on the World Wide Web) and confirmed by QUANTUM Somatic Cell Hybrid PCRable Panel (QUAN-TUM) and Stanford G3 Radiation Hybrid Panel (Research Genetics).

RESULTS AND DISCUSSION
Sequence Comparison-Phylogenetic analysis (Fig. 1A) showed that the human orphan GPCR HG57 (identified here as the CysLT 2 receptor) was 73% identical to the rat orphan GPCR RSBPT32 sequence, 38% identical to the human CysLT 1 receptor sequence, and 33-35% identical to the human orphan GPCRs GPR17 and GPR23. Despite its high similarity to HG57, we have been unable to demonstrate CysLT activation of the rat RSBPT32 receptor (data not shown). The deduced amino acid sequence of HG57 (the CysLT 2 receptor), portrayed as a putative seven-transmembrane domain protein in threedimensional helical representation, is shown in Fig. 1B (Gen-Bank TM sequence AF254664). The primary sequence of the CysLT 2 receptor is quite dissimilar to another leukotriene re-ceptor, the BLT high affinity LTB 4 receptor (Ref. 20; Fig. 1A). Although there is no record of HG57 in the expressed sequence tags division of GenBank TM , an identical open reading frame is contained as an HTGS deposition (accession no. AL137118).
Xenopus Oocyte Functional Activation-The CysLTs LTD 4 and LTC 4 produced calcium-dependent chloride flux in HG57 (CysLT 2 ) receptor cRNA-injected X. laevis oocytes and were desensitized to subsequent challenge by CysLTs ( Fig. 2A and data not shown). Control oocytes injected with saline or other GPCR cRNAs, including those encoding RSPBT32 and GPR17, showed no response to LTD 4 or LTC 4 but were able to respond to lysophosphatidic acid challenge through an endogenous lysophosphatidic acid receptor (Fig. 2B). LTD 4 and LTC 4 produced dose-dependent activation of calcium-dependent chloride flux in HG57 (CysLT 2 ) receptor cRNA-injected Xenopus oocytes, and the maximal response was not significantly reduced by preincubation with pertussis toxin (Fig. 2C), indicative that CysLT 2 did not couple to G i ␣ in this system. The LTD 4 or LTC 4 activation of calcium-dependent chloride flux in the HG57 (CysLT 2 )-expressing oocytes was not blocked by the CysLT 1 receptor-selective antagonist, MK-571, but was blocked by the dual CysLT 1 /CysLT 2 receptor antagonist BAY u9773 (Fig. 2D and data not shown).
Radioligand Binding Characterization-Saturation analysis of [ 3 H]LTD 4 -specific binding to COS-7 cell membranes transiently expressing HG57 (CysLT 2 ) was performed in two separate experiments with two different membrane preparations. In one case (Fig. 4A), data analysis using Prism (GraphPad Software Inc.) revealed the presence of high and low affinity binding sites (K d ϭ 0.4 and 51 nM; B max ϭ 0.135 and 1.415 pmol/mg of membrane protein, respectively) as illustrated in the Scatchard representation of the deduced specific binding isotherm (Fig. 4A, inset). However, in the second case, saturation analysis showed the presence of a single population of binding sites (K d ϭ 4.8 nM; B max ϭ 0.338 pmol/mg of membrane protein). This difference is potentially due to variations between membrane preparations inherent in using a transient expression system. We are currently developing clonal cell lines expressing CysLT 2 that should permit a more accurate assessment of ligand:receptor affinity and receptor number. In equilibrium competition assays, the rank order of potency of leukotriene agonists to compete with [ 3 H]LTD 4 for binding to the CysLT 2 receptor was LTD 4 ϭ LTC 4 Ͼ Ͼ LTE 4 with no competition up to 10 M by LTB 4 (Fig. 4B). CysLT 1 receptor antagonists were either weak (zafirlukast and pranlukast) or inactive Procedures." Data are from a representative of three experiments, performed in duplicate. (montelukast) at competing for radiolabeled LTD 4 binding to the CysLT 2 receptor (Fig. 4B and Table I). However, full competition was observed with the dual CysLT 1 /CysLT 2 antagonist BAY u9773 with an IC 50 value of 0.6 M (Fig. 4B and Table I). COS-7 cell membranes transfected with vector alone showed no specific LTD 4 binding, as previously reported (4).
Human Tissue RNA Northern Blot and in Situ Analyses- The human CysLT 2 receptor was shown to be expressed in peripheral blood leukocytes, lymph node, spleen, heart, and several central nervous system regions (Fig. 5). In human RNA dot blot analyses, the receptor was also shown to be expressed in the adrenal gland (data not shown). The CysLTs are known to have potent contractile and inflammatory effects in the human lung, so we investigated the in situ expression of FIG. 5. Northern blot analysis of human tissue RNA. RNA was prepared from human tissues as described under "Experimental Procedures" and hybridized with a 32 P-labeled fragment of the CysLT 2 receptor. X-ray films were exposed for 3 days for all tissues except brain, which was exposed to film for 7 days. CysLT 2 receptor mRNA in human lung (Fig. 6, A-E). In contrast to the CysLT 1 receptor, which was most highly expressed in human lung smooth muscle (4), the strongest lung expression of the CysLT 2 receptor was seen in interstitial macrophages (Fig. 6, A-C) with distinctly weaker expression in smooth muscle cells (Fig. 6, D and E). It was of interest that we noted particularly elevated expression of the CysLT 2 receptor in macrophages when in close proximity to smooth muscle cells (Fig. 6D). Peripheral blood leukocytes were another abundant source for CysLT 2 receptor transcripts (Fig. 5). We carried out in situ analysis of the CysLT 2 receptor in partly purified normal human peripheral blood monocytes and demonstrated abundant expression in greater than 20% of these cells (Fig.  6F). In addition, we found strong expression of the receptor in purified human eosinophils, a cell type in which we had previously shown expression of the CysLT 1 receptor (Ref. 4; Fig. 6, G and H). 2 Tissues other than myeloid cell containing organs, such as the spleen and lymph nodes, that showed expression of the mRNA for the CysLT 2 receptor included regions of the central nervous system, the adrenal gland, and the heart. The synthesis and activities of cysteinyl leukotrienes in the brain of various animal species has been well documented (21)(22)(23). Studies on expression of both the CysLT 1 and CysLT 2 receptors in the central nervous system are under way in our laboratories. In situ analyses of the CysLT 2 receptor in adrenal gland showed expression in medullary pheochromocytes and ganglion cells (Fig. 7A). Pheochromocytes were identified by their morphology and granular nonhomogenous cytoplasm. CysLT 2 receptor expression was also noted in the zona reticularis adjacent to the corticomedullary junction. The expression of the CysLT 2 receptor in the adrenal gland was unexpected, and the biological roles of CysLTs in the adrenal medulla are unknown. In the human heart, CysLT 2 receptor expression was concentrated in Purkinje fiber cells, identified by their characteristic morphology as transitional cells with large nuclei in histologically inhomogenous areas, with few myofibrils (Fig. 7, C-E). Consistent with Purkinje cell identity, these CysLT 2 receptorexpressing cells co-expressed PGP9.5 (protein gene product 9.5) (data not shown). In contrast, the CysLT 1 receptor has not been detected in the heart (4,5). A considerable pharmacological literature exists on the synthesis and actions of CysLTs in human vasculature (24 -26). The restricted expression of CysLT 2 in the heart is intriguing.
The human CysLT 2 receptor was localized to chromosome 13q14 by radiation hybrid and somatic cell hybrid analyses. The marker closest to the CysLT 2 receptor gene was D13S153. This is of particular interest, since this marker has been associated with atopic asthma in both a United Kingdom population (27) and in a Japanese population study (28). We are investigating the possible linkage of polymorphisms within the CysLT 2 receptor gene to the asthmatic phenotype. The recent deposition of a bacterial artificial chromosome in the high through put genome sequencing division of GenBank TM (clone_id_RP11-108P5) containing the precise HG57 (CysLT 2 ) receptor open reading frame reports a map position of 13q14.2-21.1, which is in agreement with our experimentally determined localization. The cloning and characterization of the human CysLT 2 receptor will allow a detailed molecular characterization of its activation in pulmonary, hematologic, cardiovascular, endocrine, and central nervous system tissues.