Abstract
The cysteinyl leukotrienes (CysLTs) are potent biological mediators in the pathophysiology of inflammatory diseases, in particular of airway obstruction in asthma. Pharmacological studies have suggested the existence of at least two types of CysLT receptors, designated CysLT1 and CysLT2. The CysLT1receptor has been cloned recently. Here we report the molecular cloning, expression, localization, and functional characterization of a human G protein-coupled receptor that has the expected characteristics of a CysLT2 receptor. This new receptor is selectively activated by nanomolar concentrations of CysLTs with a rank order potency of LTC4 = LTD4 ≫ LTE4. The leukotriene analog BAY u9773, reported to be a dual CysLT1/CysLT2 antagonist, was found to be an antagonist at CysLT1 sites but acted as a partial agonist at this new receptor. The structurally different CysLT1 receptor-selective antagonists zafirlukast, montelukast, and MK-571 did not inhibit the agonist-mediated calcium mobilization of CysLT2 receptors at physiological concentrations. Localization studies indicate highest expression of CysLT2 receptors in adrenal glands, heart, and placenta; moderate levels in spleen, peripheral blood leukocytes, and lymph nodes; and low levels in the central nervous system and pituitary. The human CysLT2 receptor gene is located on chromosome 13q14.12–21.1. The new receptor exhibits all characteristics of the thus far poorly defined CysLT2 receptor. Moreover, we have identified BAY u9773 as a CysLT2 selective agonist, which could prove to be of immediate use in understanding the functional roles of the CysLT2 receptor.
Cysteinyl leukotrienes (CysLTs) are the products of the 5-lipoxygenase pathway in arachidonic acid metabolism. They are predominantly produced by myeloid cells associated with the inflammatory responses (Samuelsson et al., 1987) and are potent constrictors of pulmonary smooth muscles (Dahlén et al., 1980), trachea, and parenchyma in human airways, where they induce microvascular permeability (Dahlén et al., 1981) and mucus secretion (Marom et al., 1982). Leukotrienes have been implicated in a number of pathological inflammatory diseases including asthma, allergic rhinitis, inflammatory bowel disease, and psoriasis (Busse and Gaddy, 1991). The effects of CysLTs are mediated via specific plasma membrane receptors belonging to the superfamily of G protein-coupled receptors. There is evidence for the existence of two CysLT receptor subtypes (Fleisch et al., 1982; Labat et al., 1992;Coleman et al., 1995; Metters, 1995): CysLT1 and the CysLT2 receptors, the latter of which encompasses all receptors that cannot be inhibited by CysLT1-specific antagonists (Coleman et al., 1995). The CysLT1 receptor has been studied intensively because of the availability of CysLT1-specific antagonists and of the existence of cell lines expressing it endogenously (Saussy et al., 1989). Recently, a cDNA encoding a CysLT1 receptor was cloned (Lynch et al., 1999; Sarau et al., 1999). The pharmacological profile of the cloned CysLT1 showed a rank order potency of LTD4>LTC4>LTE4and was potently inhibited by the CysLT1antagonists pranlukast, montelukast, zafirlukast, and pobilukast. The CysLT2 receptor, on the other hand, is pharmacologically less defined, mainly because of the lack of selective agonists and antagonists. In man, CysLT2receptors have been indirectly shown to be responsible for contracting pulmonary veins, contractions that were resistant to a number of CysLT1-selective antagonists (Labat et al., 1992).
In our quest to identify the natural ligands of orphan GPCRs, we cloned in silico a receptor, HPN321, that exhibited moderate similarity to the CysLT1 receptor. We describe here the pharmacological characterization of this novel receptor and conclude that we cloned a receptor exhibiting the expected characteristics of the CysLT2 receptor.
Experimental Procedures
Materials.
LTB4, LTC4, LTD4, and LTE4 were from Cayman Chemical (Ann Arbor, MI). BAY u9773 and MK-571 were from BIOMOL Research Laboratories (Plymouth Meeting, PA). LY-17883 was from Sigma (St. Louis, MO). Zafirlukast (ICI 204,219; Accolate) and montelukast (MK-476; Singulair) were purchased from the local pharmacy. All other standard chemicals used were either from Fisher or Sigma.
Identification, Cloning, and Sequencing of HPN321 and CysLT1 receptors.
A human chromosomal clone (AL137118) was identified during our systematic computational queries of novel DNA sequences coding for GPCR-related proteins. The corresponding open reading frame was identified using TBLAST algorithm and known GPCR protein sequences as queries. The DNA sequence was used to design two sets of nested primers (first set 5′-GATAGTATTGCTCCCTGTTTCATT-3′ [−112 to −89] and 5′-GAAGATGGACACAAGGATACAAGA-3′[1064 to 1087] and second set 5′-ATGTAATCAGTAAGCAAGAAGGA-3′[−75 to −53] and 5′-ACAGGTCTCATCTAAG AGCTCCTT-3′ [1063 to 1040]). The first polymerase chain reaction (PCR) was carried out with the ExpandLong Template PCR system (Hoffman La Roche, Nutley, NJ) on 0.1 μg of human genomic DNA (CLONTECH, Palo Alto, CA) according to the manufacturer's recommendation. Second PCR was performed with 2 μl of first-round PCR product and 1.0 U of Taq polymerase (Promega, Madison, WI) in 50 μl of the buffer supplied by the manufacturer containing 2.5 mM MgCl2 and was carried out for 30 cycles at 94°C for 30 s, 50°C for 30 s, and 72°C for 1.5 min. The resulting product was analyzed by agarose gel electrophoresis and subcloned into pCDNA3.1 (Invitrogen, San Diego, CA). The accuracy of the sequence was confirmed by DNA sequencing using a dideoxy termination kit from Amersham Pharmacia Biotech (Piscataway, NJ). CysLT1 was cloned from an expressed sequence tag available through the I.M.A.G.E. consortium (clone 1357580). The insert that contained the whole open reading frame, was excised byEcoRI/NotI digestion and directionally cloned into pCDNA3.1 (Invitrogen).
Expression in HEK 293T Cells.
HPN321/CysLT1 receptor plasmids were transiently expressed in human embryonic kidney cells stably expressing the simian virus 40 large T antigen (HEK 293T) cells using LipofectAMINE PLUS (Life Technologies, Gaithersburg, MD) according to the manufacturer's protocol. Transfections were done in 100-mm tissue culture dishes and seeded after 24 h into microtiter plates for subsequent assays. Assays (see below) were performed 48 h after cell transfections. Cell lines stably expressing HPN321 were established by selecting hygromycin-resistant clones.
Measurement of Agonist-Induced Calcium Release.
Calcium mobilization assays were carried out using transiently transfected HEK 293T-HPN321 or CysLT1 cells loaded with Fluo-4 AM fluorescent indicator dye (Molecular Probes, Eugene, OR) in the fluorescent imaging plate reader system (Molecular Devices, Sunnyvale, CA). Briefly, cells were seeded in poly-d-lysine-treated black microtiter plates at 8 × 104cells/well and grown to confluence. The cells were loaded with 2 μM Fluo-4 in growth medium (Dulbecco's modified Eagle's medium, 10% fetal bovine serum) supplemented with 2.5 mM probenecid for 1 h at 37°C, 5% CO2. The cells were washed thrice with Hanks' balanced salt solution containing 20 mM HEPES and 2.5 mM probenecid. Calcium transient curves were generated by reading baseline fluorescence values for 10 s, followed by addition of test compounds. For the first minute, fluorescence values were collected in 1-s intervals; for the next 2 min, data were collected in 3-s intervals. For calculation of dose-response curves, the peak fluorescence values for each agonist concentration were determined and analyzed by nonlinear regression using PRISM software (GraphPad, San Diego, CA). Antagonist studies were performed under the same conditions but test compounds were added 2 to 5 min before addition of agonist (at EC50 value). EC50 is defined as the agonist concentration generating 50% of the peak fluorescence values. IC50 is the concentration required to inhibit 50% of the peak fluorescence.
Radioligand Binding Studies.
Transiently transfected HEK 293T cells were grown and harvested, and crude membranes were prepared as described (Nothacker et al., 1999). For competition binding studies, membranes (200 μg of total membrane protein) were incubated with 1.5 nM [3H]LT D4 (NEN Life Science, Boston, MA) and variable concentrations of competitors in 250 μl of 50 mM Tris-HCl pH 7.4, 20 mM CaCl2, 25 mM MgCl2, 10 mM glycine, and 10 mM cysteine for 1 h at 22°C. The membranes were filtered over Whatman GF/C filters using a Brandel cell harvester, washed thrice with ice-cold binding buffer, and counted in the presence of 5-ml CytoScint (ICN, Irvine, CA) in a Beckman LS 3801 liquid scintillation counter. Nonspecific binding was determined in the presence of 2 μM unlabeled LTD4. Under these conditions, specific binding accounted for ∼60% of total binding. Data were analyzed by nonlinear regression analysis using PRISM software (GraphPad).
RNA-Array and Northern Blot Analysis.
Human multiple tissue arrays (CLONTECH) were analyzed by hybridization to an [α-32P]dCTP-labeled, 600-bpBamHI-fragment of HPN321. Hybridization was conducted according to the manufacturer's recommendation and with final washes of 0.1× standard saline citrate, 0.5% SDS at 55°C for 30 min before exposition to X-OMAT film (Kodak, Rochester, NY) at −80°C. Lymphoma-derived U-937 cell lines and leukemia-derived HL-60 cell lines were differentiated with 1.3% dimethyl sulfoxide in RPMI 1640 (Life Technologies) supplemented with 10% fetal bovine serum for three days at 37°C, 5% CO2. The cells were harvested and total RNA prepared using TRIZOL reagent (Life Technologies). Northern blots were prepared from 5 or 15 μg of total RNA according to standard molecular biology techniques (Sambrook et al., 1989) and hybridized to the probe described above. Control hybridization was done with a CysLT1 specific probe composed of a 3′-HincII/NotI fragment of I.M.A.G.E. clone 1357580.
Results
Identification and Molecular Characterization of the CysLT1-Like Receptor, HPN321.
Computational screenings of public expressed sequence tag and genomic databases were carried out to identify novel sequences displaying structural characteristics common to GPCRs. One genomic clone (GenBank accession no.AL137118) contained an intronless open reading frame of 347 amino acids (Fig. 1) exhibiting seven putative membrane-spanning helices characteristic of GPCR sequences and sharing 36% amino acid identity with the recently cloned CysLT1 receptor (Lynch et al., 1999; Sarau et al., 1999). This sequence, named HPN321, was amplified by PCR from human genomic DNA, inserted into a mammalian expression vector under the control of a CMV promotor, and transiently expressed in HEK 293T cells. The transfected cells were challenged with either LTC4 or LTD4 and simultaneously monitored for changes of intracellular calcium. Both induced a strong calcium mobilization at low concentrations, whereas no changes could be observed in mock-transfected control cells.
Pharmacological Characterization of the HPN321 Receptor.
To further characterize the pharmacological profile of the receptor, the responses to four different CysLTs were determined. The HPN321 receptor transfected HEK 293T cells responded to LTC4, LTD4, and LTE4 in a dose-dependent manner, with respective EC50values of 8.9, 4.4, and 293 nM (n = 3, Fig.2). No activation could be observed by addition of the non-peptido-leukotriene LTB4, even at concentrations up to 5 μM. The highest concentrations of LTC4 and LTD4 produced similar maximal responses, whereas LTE4 behaved as a partial agonist, attaining only 60% of the maximal response of that elicited by LTC4 and LTD4. Consequently, the rank order of potency was LTC4 = LTD4 ≫ LTE4, distinct from that of the CysLT1 receptor (Table1). We also tested several structurally different and selective CysLT1 antagonists (LY-17883, MK-571, montelukast and zafirlukast) to assess their ability to block HPN321 activation. All of the CysLT1antagonists tested were practically inactive, showing only weak inhibition at concentrations of >1 μM (Table 1).
Pharmacological Properties of BAY u9773 at HPN321 and CysLT1.
To further characterize this novel receptor, we tested BAY u9773, a compound reported to behave as a dual CysLT1/CysLT2 antagonist. When applied simultaneously with the agonist LTD4(10 nM, in presence of 5 μM BAY u9773) BAY u9773 exhibited no inhibition. We then administered BAY u9773 alone and found that BAY u9773 was able to activate HPN321 with an EC50 of 100 nM, thus more potently than LTE4, but less potently than LTC4 or LTD4. The response was concentration dependent and reached 67% of the maximal LTC4 and LTD4 responses (Fig. 3A), indicating that BAY u9773 acts as a partial agonist. We next tested whether BAY u9773 would also antagonize the effects of full agonists such as LTC4/LTD4. We therefore monitored LTD4 responses obtained 2 min after the addition of BAY u9773 at variable concentrations. BAY u9773 exhibited its intrinsic agonistic effect but in addition inhibited that elicited by LTD4 in a dose-dependent manner (Fig. 3, B and C). When BAY u9773 was tested on CysLT1transfected cells, it had no intrinsic agonistic effect but elicited an antagonistic effect on a subsequent LTD4challenge (Fig. 3D). The IC50 values of BAY u9773 on LTD4 stimulation were very similar for both receptors (Fig. 3B; Table 1) albeit the antagonistic mechanism at both receptors seemed to be different. The most probable explanation for the antagonistic properties of BAY u9773 at HPN321 is that the drug desensitizes the system in a manner similar to that seen by repeated challenge with LTC4. To provide additional data demonstrating that BAY u9773 is a partial agonist at HPN321 receptor and that all of its effects in HEK 293T cells can be explained by desensitization of [Ca2+]i mobilization responses, we carried out a series of mutual desensitization experiments (Fig. 4). Sequential addition of 100 nM BAY u9773 leads to a desensitization of the HPN321 specific response, leaving other calcium mobilization agonists unaltered as shown by the normal response to UTP (Fig. 4B). LTC4 and LTD4 applied at an equipotent concentration desensitized a secondary BAY u9773 challenge (Fig. 4, C and D). The observed homologous desensitization events strongly suggest a similar mode of action for the natural agonists as well for BAY u9773.
Displacement of [3H]LTD4
Competition binding experiments carried out on membranes of transiently transfected HEK 293T cells (Fig. 5) revealed the same rank order of potency for the competition of [3H]LTD4 binding sites as obtained in the functional assays. LTC4 and LTD4 showed high binding affinity, with IC50 values of 4 and 7 nM, respectively. In contrast, LTE4 competed rather weakly, as reflected by its IC50 value of 0.7 μM. BAY u9773 fully inhibited [3H]LTD4 binding with an IC50 value of 0.4 μM, very similar to the values obtained for the calcium mobilization assay (Table 1).
Tissue Distribution of HPN321.
Multiple human tissues were examined by Northern hybridization and dot blot analysis to determine sites of HPN321 expression (Fig. 6A). Using a 600-base pair 5′-probe that did not cross-hybridize with CysLT1-DNA, we found the strongest expression in the adrenal gland, the heart, and the placenta. Cardiac expression could be detected throughout the entire heart, including ventricles, atrium, septum, and apex. Moderate expression could be detected in the immune system, in particular spleen, lymph nodes, and peripheral blood leukocytes. No signals were found in HL-60, a cell line known to express CysLT1 receptors. We also investigated U-973 myeloid leukemia cells for HPN321 expression because this cell line has been used to develop CysLT1 antagonists (Sarau et al., 1999). However, in neither undifferentiated nor differentiated U-973 cells could HPN321 expression be detected, whereas CysLT1 mRNA was easily detected (Fig. 6B). HPN321-expression was present in lower levels throughout the central nervous system, with highest levels in pituitary and spinal cord. This expression pattern is distinct from that of the CysLT1 receptor (Lynch et al., 1999; Sarau et al., 1999) yet overlapping in some tissues.
Discussion
While searching in silico for novel GPCRs, we discovered a DNA sequence that could encode a putative GPCR sharing moderate sequence similarity (36%) with the recently described CysLT1 receptor. When expressed in HEK 293T cells, this new receptor responded to low concentrations of CysLTs by inducing intracellular calcium mobilization. Furthermore, this receptor was not inhibited by antagonists specific to the CysLT1receptor, fitting the postulated pharmacological profile of CysLT2 receptor.
Interestingly, this receptor could be partially activated by the leukotriene analog BAY u9773, originally described as a dual CysLT1/CysLT2 antagonist. In this report we demonstrate that BAY u9773 acts as a partial agonist at the new receptor and that this property may have led to its classification as a CysLT2 antagonist. BAY u9773 has been a useful tool to define CysLT2 receptors pharmacologically. It has been shown to inhibit physiological responses insensitive to CysLT1 antagonists in several species including man (Labat et al., 1992; Tudhope et al., 1994;Bäck et al., 1996), thus allowing the detection of the CysLT2 sites.
A functional model of CysLT1 and CysLT2 receptors in human lung has been developed (Gorenne et al., 1996). CysLT1 receptors, mainly located in bronchial smooth muscle, mediate the contractions evoked by CysLTs. In contrast, CysLT2 receptors located mainly in the vascular smooth muscles of pulmonary veins, induce contractions that cannot be blocked by CysLT1specific antagonists. In addition, the vascular endothelium contains both receptor subtypes, a CysLT1 type associated with contractions and a CysLT2 type associated with relaxation. We were able to demonstrate an important pharmacological difference in the action of BAY u9773 at CysLT1 and CysLT2receptors, which enables us now to explain previous findings. Labat et al. had already reported that BAY u9773 elicits small contractile responses in human pulmonary veins and speculated that this effect was caused by a partial agonist activity (Labat et al., 1992). We show that BAY u9773 acts as a partial agonist at HPN321 receptor and thus suggest that it represents the target of BAY u9773 in the human pulmonary venous preparation. In addition, the same authors observed a BAY u9773-induced relaxation in human lung tissues. We can therefore speculate that the same receptor exists in human lung coupled to a relaxation effect, probably mediated through stimulation of endothelial nitric-oxide synthase. BAY u9773 was also reported to contract various tissue preparations from guinea pig (Tudhope et al., 1994;Wikström Jonsson et al., 1998). It must be emphasized, however, that the molecular targets of BAY u9773 in this species might not be identical to the CysLT receptors found in the human lung (Gorenne et al., 1996). One recent study in guinea pig lung parenchyma found that contractions evoked by BAY u9773 were antagonized by a CysLT1 specific antagonist (Wikström Jonsson et al., 1998). This result suggests the existence of a CysLT1-like molecular target for BAY u9773 in this guinea pig tissue. However, in our hands, the human CysLT1 receptor (Lynch et al., 1999; Sarau et al., 1999) could not be activated by BAY u9773 and was sensitive to CysLT1 antagonists. Our results thus support the existence of a pharmacological heterogeneity of CysLT receptors in different species.
Our finding that BAY u9773 acts as a subtype selective agonist offers a unique opportunity to study HPN321 receptor selective physiological activities, particularly in tissues in which HPN321 is dominantly expressed over CysLT1, such as the adrenal gland and the heart. The antagonistic effects seen by pretreatment with BAY u9773 at CysLT2 sites might be caused by receptor desensitization. Calcium mobilization responses mediated by HPN321 can be desensitized by repeated challenges of LTC4, LTD4, LTE4, and BAYu9773 as well. Because BAY u9773 is structurally related to LTE4, a partial agonist at both receptors, it probably competes with the full agonists LTC4 and LTD4 for the same binding site. We are currently studying the mechanism of BAY u9773 in greater detail, particularly the surprising agonistic selectivity toward HPN321.
The distribution data of HPN321 suggests major role(s) for this receptor in endocrine and cardiovascular systems. CysLTs are well known for their modulatory effects in cardiovascular functions, where they reduce myocardial contractility and coronary blood flow (Letts and Piper, 1982) and have vasoactive effects (Drazen et al., 1980). They are thus considered to be important players in cardiovascular diseases (for review, see Folco et al., 2000). The strong expression of HPN321 in adrenal gland points at a new tissue where to study the influence of CysLTs on endocrine circuits. Finally, leukotrienes have also been found to act on the pituitary to modulate the release of the pituitary hormones (Hulting et al., 1984; Saadi et al., 1990). Our discovery of the existence of the HPN321 message in pituitary adds a molecular credence to this concept. The HPN321 receptor may thus modulate a variety of different physiological functions, which can now be tested using BAY u9773.
During the preparation of this manuscript, two groups also reported the pharmacological characterization of a second human CysLT receptor (Heise et al., 2000; Takasaki et al., 2000). The sequences of the reported receptors (GenBank accession nos. AF254664 andAB038269) are identical with the one described herein. Relative potencies of the CysLTs are very similar, yet none report the selective activity of BAY u9773.
In summary, we present the molecular and pharmacological characterization of a novel human CysLT receptor subtype that we have named the CysLT2 receptor and report the tissue distribution of its expression. The receptor shows a selective rank order of potency toward CysLTs and is the most preferred target for LTC4. We also identified BAY u9773 as a subtype selective agonist for the CysLT2 receptor and suggest the use of BAY u9773 as a selective tool in studies on the physiological roles of the CysLT2 receptor in cardiac, neuronal, endocrine, and inflammatory circuits.
Acknowledgments
We thank Dr. Yumiko Saito for support. We also thank Dr. Fred Ehlert for helpful discussion in preparing this manuscript.
Footnotes
- Received August 4, 2000.
- Accepted August 24, 2000.
-
Send reprint requests to: Dr. Olivier Civelli, UC Irvine, Department of Pharmacology, 355 Med Surge II, Irvine, CA 92697-4625. E-mail: ocivelli{at}uci.edu
-
This work was supported in part by grants from Neotherapeutics, the Eric and Lila Nelson Chair in Neuropharmacology, and from BioStar (S97–107).
Abbreviations
- CysLT
- cysteinyl leukotrienes
- PCR
- polymerase chain reaction
- GPCR
- G protein-coupled receptor
- HEK 293T
- human embryonic kidney cells stably expressing the simian virus 40 large T antigen
- The American Society for Pharmacology and Experimental Therapeutics