Abstract
Adrenomedullin (ADM) and calcitonin gene-related peptide (CGRP) receptors and their respective ligands play important roles in cardiovascular (patho-)physiology. Functional expression of ADM and CGRP receptors requires the presence of the calcitonin receptor-like receptor (CRLR) together with receptor-activity-modifying proteins (RAMPs). We have characterized the expression patterns of CRLR and RAMP1 to RAMP3 in human cardiovascular-related tissues by quantitative polymerase chain reaction. We could identify high expression levels of CRLR, RAMP1, and RAMP2 in human heart and various blood vessels. RAMP3 expression in these tissues, however, was detectable at significantly lower levels. In addition, we describe here a novel, aequorin luminescence-based G protein-coupled receptor reporter assay that enables the real-time detection of receptor activation in living cells. In the assay system, intracellular cAMP levels are monitored with high sensitivity by using a modified, heteromultimeric cyclic nucleotide-gated channel mediating calcium influx. Gq-coupled receptor activation is detected via aequorin luminescence stimulated by calcium release from intracellular stores. Using this novel reporter assay, we established and characterized stable ADM1 and CGRP1 receptor cell lines. The peptide ligands ADM, CGRP1, and CGRP2 were characterized as potent agonists at their respective receptors. In contrast, intermedin acted as a weak agonist on both receptors and showed only partial agonism on the ADM1 receptor. Agonist activities were effectively antagonized by the receptor antagonists ADM(22-52) and CGRP(8-37). Various vasoactive ADM fragments were also characterized but showed no activity on the ADM1 receptor cell line. In addition, luminescence signal kinetics after activation of Gs- and Gq-coupled receptors were found to be markedly different.
Adrenomedullin (ADM) and the calcitonin gene-related peptides (CGRP1 and CGRP2) are members of the calcitonin family of peptides (Muff et al., 1995; Wimalawansa, 1997). These peptides play a pivotal role in cardiovascular physiology and pathophysiology and are involved in the regulation of the vascular tone, cardiac output, smooth muscle cell proliferation, and fluid and electrolyte homeostasis (Kurihara et al., 2003; Brain and Grant, 2004; Muff et al., 2004; Ishimitsu et al., 2006). Functional ADM and CGRP receptors are heterodimeric complexes and require coexpression of the calcitonin-receptor-like receptor (CRLR) together with associated receptor activity-modifying proteins (RAMPs), which regulate CRLR transport to the plasma membrane and determine ligand specificity. Activation of these receptors leads to the stimulation of intracellular cAMP synthesis (McLatchie et al., 1998; Poyner et al., 2002; Conner et al., 2004). In recent years, ADM and CGRP and their respective receptors have gained considerable attention and have become targets for new drug development (Doggrell, 2001; Ishimitsu et al., 2006). Intermedin (IMD) has been identified as a novel member of the calcitonin peptide family. IMD also shows cardiovascular activity, which might be related to the activation of ADM and CGRP receptors (Roh et al., 2004).
Various functional assays to monitor GPCR activation and signaling have been developed and are used for the characterization of GPCR pharmacology and drug discovery (Williams, 2004; Jacoby et al., 2006; Kostenis, 2006). Novel fluorescence-based assays using cyclic nucleotide-gated (CNG) channels as biosensors to detect intracellular cAMP levels and GPCR activity have been introduced (Fagan et al., 2001; Rich et al., 2001; Reinscheid et al., 2003). In these reports, homomeric CNG channels were described as biosensors to monitor intracellular cAMP levels. However, native CNG channels usually form heterotetrameric complexes of two or three different types of subunits, and the ligand sensitivity and selectivity are determined by their particular subunit composition (Kaupp and Seifert, 2002). Native olfactory channels are composed of three different subunits: CNGA2, CNGA4, and CNGB1b. Upon heterologous coexpression of these subunits, functional CNG channels with increased cAMP sensitivity are observed (Bradley et al., 1994; Liman and Buck, 1994; Sautter et al., 1998; Bönigk et al., 1999). In addition, the ligand specificity and sensitivity can also be shifted by mutations in the cyclic nucleotide-binding domain. Thereby, a CNGA2 mutant, CNGA2(T537A), with increased cAMP and decreased cGMP sensitivity could be identified (Altenhofen et al., 1991).
We have shown previously that the homomeric, olfactory CNGA2 channel is well suited for the detection of intracellular cGMP generation in an ultra-high-throughput screening (uHTS) assay format (Wunder et al., 2005a). We describe in this report the development of a novel, highly sensitive, luminescence-based cAMP and GPCR reporter assay. To achieve optimal cAMP sensitivity, three different subunits, CNGA2(T537A), CNGA4, and CNGB1b, were coexpressed in an aequorin reporter cell line. In addition, we describe here the pharmacological and kinetic characterization of newly established ADM1 and CGRP1 receptor cell lines using this novel reporter technology.
Materials and Methods
Quantitative Real-Time RT-PCR Analysis. Quantitative TaqMan analysis was performed using the Applied Biosystems PRISM 7900 sequence detection system (Applied Biosystems, Foster City, CA). Human tissue mRNA probes were obtained from Ambion, Inc. (Austin, TX), Analytical Biological Services, Inc. (Wilmington, DE), Clontech Laboratories (Mountain View, CA), and Stratagene (La Jolla, CA) and were reverse-transcribed using random hexamers. Probes were carefully designed to cross exon boundaries, and comparable probe efficiencies were assured by titration of corresponding plasmid constructs. Normalization was performed using β-actin as control, and relative expression was calculated using the following formula: relative expression = 2(15 - (Ctprobe - Ctactin)). The parameter Ct is defined as the threshold cycle number at which the amplification plot passed a fixed threshold above baseline. The resulting expression is given in arbitrary units. The primers and fluorescent probes used are shown in Table 1.
Generation of the Parental GPCR Reporter Cell Line. A recombinant CHO cell line expressing cytosolic apoaequorin was cotransfected with a pcDNAI plasmid construct encoding the bovine CNGA2 channel (accession number X55010) with one amino acid substitution [CNGA2(T537A); Altenhofen et al., 1991], a pcDNAI construct containing the rat CNGA4 cDNA (accession number U12623; Bradley et al., 1994) and pZeoSV (zeocin resistance). Positive clones were identified by 8-bromoadenosine-3′,5′-cyclic monophosphate (8-Br-cAMP) and 8-bromoguanosine-3′,5′-cyclic monophosphate (8-Br-cGMP) stimulation (data not shown) and were purified by the limited dilution technique. One purified clone was then cotransfected with a pcDNA1.1/Amp plasmid construct encoding the CNGB1b channel (accession number AJ000515; Sautter et al., 1998) and a plasmid providing hygromycin resistance. Stably transfected clones were characterized by 8-Br-cAMP and forskolin stimulation (data not shown) and were again purified by the limited dilution technique. One clonal cell line, referred to here as the GPCR reporter cell line, was selected for further experiments. All plasmid vectors were purchased from Invitrogen (Carlsbad, CA).
Generation of ADM1 and CGRP1 Receptor Cell Lines. The parental GPCR reporter cell line was cotransfected with a pcDNA3 plasmid construct encoding the human CRLR receptor (accession number U17473) and pcDNA1.1/Amp constructs containing either the human RAMP1 cDNA (accession number NM_005855) or the human RAMP2 cDNA (accession number NM_005854) according to McLatchie et al. (1998). Stably transfected cell clones were obtained by G-418 (Geneticin) selection and were characterized by ADM and CGRP1 stimulation (data not shown). Positive clones were purified by the limited dilution technique, and two clonal cell lines were selected for further characterization, referred to here as the ADM1 receptor (CRLR/RAMP2) and CGRP1 receptor (CRLR/RAMP1) cell lines.
Cell Culture Conditions and Aequorin Luminescence Measurements. Cells were cultured at 37°C and 5% CO2 in Dulbecco's modified Eagle's medium/NUT mix F-12 with l-glutamine, supplemented with 10% (v/v) inactivated fetal calf serum, 1 mM sodium pyruvate, 0.9 mM sodium bicarbonate, 50 U/ml penicillin, 50 μg/ml streptomycin, 2.5 μg/ml amphotericin B, 0.6 mg/ml hygromycin B, and 0.25 mg/ml zeocin. In addition, 1 mg/ml G-418 was added to the cell culture medium used for the ADM1 and CGRP1 receptor cell lines. Cells were passaged using enzyme-free/Hanks'-based cell dissociation buffer. All cell culture reagents were obtained from Invitrogen.
Luminescence measurements were performed on opaque 384-well microtiter plates (MTPs). Either 2500 or 1500 cells/well were plated and cultured for 24 or 48 h, respectively. After the removal of the cell culture medium, cells were loaded for 3 h with 0.6 μg/ml coelenterazine in Ca2+-free Tyrode solution (130 mM NaCl, 5 mM KCl, 20 mM HEPES, 1 mM MgCl2, and 4.8 mM NaHCO3, pH 7.4) at 37°C and 5% CO2. Agonists and antagonists were added for 6 min in Ca2+-free Tyrode containing 0.1% bovine serum albumin. 3-Isobutyl-1-methylxanthine (IBMX) (0.2 mM) was used to prevent cAMP degradation by endogenous phosphodiesterases. Immediately before adding Ca2+ ions (final concentration 3 mM), measurement of the aequorin luminescence was started by using a charge-coupled device camera (Hamamatsu Corporation, Shizuoka, Japan) in a light-tight box. Alternatively, a conventional luminometer may also be used (Wunder et al., 2005a). Luminescence was monitored continuously for 60 s. For the real-time detection of cAMP generation and calcium release from intracellular stores, coelenterazine loading and agonist stimulation were performed in Tyrode solution containing 2 mM calcium ions (Wunder et al., 2005a).
Transient Transfections. Transient transfections were performed using Lipofectamine 2000 from Invitrogen according to the manufacturer's standard protocol. Luminescence measurements were performed on 384-well MTPs 24 or 48 h after transfection of plasmid constructs encoding β2 adrenergic (accession number NM_000024), bradykinin B2 (accession number NM_000623), vasopressin receptor (V1A); accession number NM_000706), or endothelin A receptors (accession number NM_001957).
Detection of cAMP by Radioimmunoassay. For the measurement of intracellular cAMP, the ADM1 receptor cell line (20,000 cells/well) was seeded onto 24-well MTPs, and cells were cultured for 2 days. The medium was removed, the cells were washed once, and agonist stimulation was performed in Tyrode supplemented with 0.2 mM IBMX for 15 min at 37°C. After removal of the supernatant, cAMP was extracted overnight with 70% EtOH at -20°C. Intracellular cAMP was measured by using a commercially available radioimmunoassay (RIA) kit (IBL, Hamburg, Germany). Measurements were performed in triplicate.
Compounds. ADM(1-21) and ADM(16-21) were purchased from Phoenix Pharmaceuticals (Burlingame, CA). All other human peptide agonists and antagonists were obtained from Bachem (Bubendorf, Switzerland). Forskolin, IBMX, ATP, 8-Br-cGMP, and 8-Br-cAMP were obtained from Sigma-Aldrich (Taufkirchen, Germany).
Statistics. The data are presented as mean values with standard deviation errors. The GraphPad Prism software (version 4.02; Graph-Pad Software Inc., San Diego, CA) was used for curve-fitting and calculation of the half-maximal effective and inhibitory concentrations (EC50 and IC50, respectively). For the determination of pEC50 and pIC50 values, three to six independent experiments were performed in quadruplicate. pEC50 and pIC50 values are given as means ± S.E.M.
Results
Analysis of CRLR and RAMP Expression in Human Cardiovascular-Related Tissues. Because we were interested in the cardiovascular pharmacology of ADM and CGRP receptors, we analyzed the expression patterns of CRLR and RAMPs 1 to 3 by quantitative RT-PCR in human heart, various blood vessels, and kidney. As shown in Fig. 1, we were able to detect very high expression levels of CRLR, RAMP1, and RAMP2 in heart ventricle and atrium. In addition, we could verify medium to high expression of these transcripts in all blood vessels studied. RAMP3 expression was also detected in heart and blood vessels but at significantly lower levels. In primary endothelial cells from coronary arteries, high levels of CRLR and RAMP2 mRNA were found, whereas RAMP1 and RAMP3 expression could not be detected. Similar results were obtained with primary endothelial cells from aorta and pulmonary arteries and with HUVEC cells (data not shown). Low expression levels of CRLR, RAMP1, and RAMP2 could also be detected in primary smooth muscle cells from aorta, pulmonary, and coronary artery. In contrast, RAMP3 expression could not be detected in these cells (data not shown). In human kidney, high expression of CRLR and RAMP2 and medium levels of RAMP1 and RAMP3 transcripts were detected.
Generation of the Recombinant GPCR Reporter Cell Line. A CHO cell line expressing cytosolic apoaequorin was cotransfected with a plasmid construct encoding a mutated CNGA2 channel with one amino acid substitution [CNGA2(T537A); Altenhofen et al., 1991] and a second construct encoding the CNGA4 channel (Bradley et al., 1994). Positive clones were identified by 8-Br-cAMP and 8-Br-cGMP stimulation (data not shown) and were purified by the limited dilution technique. One purified clone was subsequently transfected with a plasmid construct encoding the CNGB1b channel (Sautter et al., 1998). Stably transfected clones were characterized by 8-Br-cAMP and forskolin stimulation (data not shown) and were again purified by the limited dilution technique. In our reporter system, ligand-mediated activation of Gs-coupled GPCRs can be monitored in real-time via soluble adenylyl cyclase activation and calcium influx through a cAMP-gated cation channel, acting as the intracellular cAMP sensor. In addition, activation of Gq-coupled receptor and stimulation of the phospholipase C/inositol 1,4,5-triphosphate (IP3) pathway is detected via aequorin luminescence stimulated by calcium release from intracellular stores (Fig. 2).
As shown in Fig. 3, we compared different stable CNG channel-expressing cell lines for their ability to detect intracellular cAMP synthesis. Therefore, we tested forskolin-induced luminescence signals using our cGMP reporter cell line expressing the CNGA2 subunit only (Wunder et al., 2005a) compared with our newly established cell lines expressing a combination of CNGA2(T537A) and CNGA4, or a combination of CNGA2(T537A), CNGA4, and CNGB1b subunits. The cGMP reporter cell line showed only minor stimulation by forskolin, significant stimulation could only be observed at very high concentrations starting from 10 μM (Fig. 3A). In contrast, the CNGA2(T537A)/CNGA4 cell line already showed an increased forskolin response (pEC50 = 5.73 ± 0.01; Fig. 3B). As shown in Fig. 3C, high cAMP sensitivity was achieved by heterologous expression of a combination of CNGA2(T537A), CNGA4, and CNGB1b subunits. Using this heteromultimeric CNG channel, forskolin stimulated luminescence signals with a pEC50 value of 6.62 ± 0.04. Therefore, one clonal cell line expressing this CNG channel subunit combination [CNGA2(T537A)/CNGA4/CNGB1b] was used for further experiments (referred to here as the GPCR reporter cell line).
Generation and Characterization of ADM1 and CGRP1 Receptor Cell Lines. According to the results of our TaqMan expression analysis, the parental GPCR reporter cell line was cotransfected with CRLR and RAMP2 plasmid constructs encoding the ADM1 receptor. For the generation of the CGRP1 receptor cell line, cotransfection with CRLR and RAMP1 encoding plasmids was performed. Active clones were identified by ADM and CGRP1 stimulation and were purified once by the limited dilution technique. One clonal cell line of both transfection series was selected and used for further characterization (referred to here as the ADM1 and CGRP1 receptor cell lines).
We characterized these newly established cell lines by testing the effects of different members of the human calcitonin family of peptides. Stimulation of the ADM1 receptor cell line with ADM, CGRP2, and CGRP1 resulted in concentration-dependent luminescence signals with pEC50 values of 9.45 ± 0.08, 7.95 ± 0.03, and 6.87 ± 0.01, respectively (Fig. 4A). As expected, stimulation of the CGRP1 reporter cell line showed that CGRP1 and CGRP2 are potent agonists with pEC50 values of 9.50 ± 0.12 and 9.53 ± 0.10, respectively (Fig. 4B). ADM stimulated the CGRP1 receptor cell line with lower potency (pEC50 = 8.44 ± 0.03). In contrast, calcitonin and amylin did not stimulate any luminescence signals on both receptor cell lines at concentrations up to 1 μM. We also tested the agonists ADM, CGRP1, and CGRP2 on the parental GPCR reporter cell line. However, the agonists did not induce significant luminescence signals (data not shown). Next, we correlated agonist-stimulated luminescence signals with intracellular cAMP accumulation measured by a cAMP radioimmunoassay. As shown in Fig. 4C, stimulation of the ADM1 receptor cell line by ADM resulted in intracellular cAMP accumulation (pEC50 = 9.48 ± 0.06). The ADM-mediated cAMP increase was inhibited (pEC50 = 7.84 ± 0.04) by the addition of the ADM receptor antagonist ADM(22-52).
We also tested various shorter C-terminally amidated ADM peptides, which were shown previously to possess cardiovascular activities (Watanabe et al., 1996; Champion et al., 1999) for their ability to stimulate the ADM1 receptor cell line. ADM(1-21), ADM(16-21), ADM(16-31), and ADM(22-52) did not stimulate any signals when tested up to 1 μM. In contrast, ADM(13-52) stimulated the ADM1 receptor cell line with the same potency as the full-length agonist ADM (data not shown). The ADM1 receptor cell line was also simultaneously stimulated with a combination (up to 1 μM each) of ADM(1-21) and ADM(22-52) or a combination of ADM(16-21) and ADM(22-52). However, no luminescence stimulation was observed. We also tested the activity of pro-ADM(1-20) on the ADM1 receptor cell line. However, pro-ADM(1-20) was not able to mediate receptor stimulation (data not shown).
Modulation by Receptor Antagonists. We next sought to determine whether the agonist-stimulated luminescence signals could be inhibited by the addition of the receptor antagonists ADM(22-52) and CGRP(8-37). As shown in Fig. 5A, ADM(22-52) and CGRP(8-37) concentration-dependently inhibited 3 nM ADM-stimulated signals on the ADM1 receptor cell line with pIC50 values of 7.57 ± 0.05 and 5.67 ± 0.11, respectively. In contrast, on the CGRP1 receptor cell line, CGRP(8-37) inhibited 3 nM CGRP1-stimulated luminescence signals more potently (pIC50 = 7.31 ± 0.12), whereas ADM(22-52) showed only minor effects (pIC50 < 5.52; Fig. 5B). As shown in Fig. 5C, the antagonist ADM(22-52), when applied to the ADM1 receptor cell line at concentrations of 0.1 and 1 μM, shifted the pEC50 values of ADM-mediated luminescence signals from 9.16 ± 0.02 to 8.34 ± 0.01 and 7.76 ± 0.01, respectively. Using the CGRP1 receptor cell line, the antagonist CGRP(8-37), at concentrations of 0.1 and 1 μM, shifted the pEC50 values for CGRP1-mediated luminescence signals from 9.23 ± 0.01 to 8.20 ± 0.04 and 7.51 ± 0.04, respectively (Fig. 5D). In addition, we tested the truncated ADM fragments ADM(1-21), ADM(16-21), and ADM(16-31) for their antagonistic activity on the ADM1 receptor cell line. However, none of the shorter peptides was able to antagonize 3 nM ADM-stimulated luminescence signals (data not shown).
Characterization of Intermedin. Next, we characterized the activity of human IMD, a newly discovered member of the calcitonin family of peptides (Roh et al., 2004). As shown in Fig. 6A, IMD showed only partial agonism on the ADM1 receptor cell line and stimulated luminescence signals with a pEC50 value of 8.27 ± 0.06. The activity of IMD was potently inhibited (pEC50 < 6) in the presence of the antagonist ADM(22-52). On the CGRP1 receptor cell line, IMD acted as a full agonist and luminescence signals were stimulated with a pEC50 value of 7.74 ± 0.09. In the presence of the antagonist CGRP(8-37), IMD activity on the CGRP1 receptor was potently antagonized (pEC50 < 6; Fig. 6B).
Real-Time Detection of cAMP Synthesis. To monitor Gs-coupled receptor activation and cAMP synthesis in real time, increasing concentrations (0.1-100 nM) of the agonists ADM and CGRP1 were added to the reporter cell lines in the presence of 2 mM calcium ions (“kinetic mode”; Wunder et al., 2005a). Luminescence measurements were started immediately before agonist addition. As shown in Fig. 7A, ADM concentration-dependently stimulated luminescence signals on the ADM1 receptor cell line with slow signal kinetics. Lag times in the range of 22 s (at 0.1 nM) to 8 s (at 10 and 100 nM) were observed. CGRP1 induced luminescence signals with similar kinetics on the CGRP1 receptor cell line, with lag times ranging from 18 s (0.1 nM) to 7 s (100 nM), respectively (Fig. 7B). At the highest agonist concentration of 100 nM, maximal signals stimulated by ADM and CGRP1 were reached after 31 and 34 s, respectively.
Next, we measured the luminescence signal kinetics mediated by endogenous, Gq-coupled P2Y receptors after stimulation with the P2Y agonist UTP (Burnstock, 2004). Compared with ADM- and CGRP1-mediated signals, UTP (1-1000 μM) stimulated luminescence signals with much faster activation kinetics (Fig. 7C). At the lowest UTP concentration used, a lag time of 4 s was observed. At higher UTP concentrations, the signals started to increase with a lag time of 1 s, and maximal signals were reached after 4 to 5 s. Similar results were obtained after stimulation of purinergic receptors with UDP and ATP (data not shown). Simultaneous application of UTP (100 μM) and ADM (100 nM) to the ADM1 receptor cell line resulted in luminescence signals with two maxima, which were reached after 4 and 36 s, respectively (Fig. 7D).
To further corroborate these findings, we studied the luminescence signals generated by activation of transiently transfected human bradykinin B2 (BK2) receptors. Stimulation of the Gq-coupled BK2 receptors (Zhang et al., 2001) with bradykinin (BK) resulted in luminescence signals with fast kinetics (Fig. 7E). Similar results were obtained after (Arg8)-vasopressin stimulation of transiently transfected, Gq-coupled human arginine V1A receptors (Liu and Wess, 1996; data not shown).
Stimulation of transiently transfected endothelin A (ETA) receptors, shown previously to couple to both the IP3 and the cAMP pathway (Aramori and Nakanishi, 1992), with endothelin-1 (ET-1) resulted in luminescence signals that peaked after 3 and 37 s, respectively (Fig. 7F). The agonists isoproterenol, BK, (Arg8)-vasopressin, and ET-1 did not stimulate significant luminescence signals on the GPCR reporter cell line in the absence of transiently transfected receptors (data not shown).
In additional experiments, we characterized the luminescence signal decrease after ADM and CGRP1 stimulation. By sequential addition of ADM or Tyrode control, followed by UTP stimulation, we could show that the observed signal decrease is not due to aequorin consumption (Supplementary Fig. S1A). However, by sequential application of ADM or Tyrode control, followed by isoproterenol stimulation of transiently transfected, Gs-coupled β2 adrenoceptors, we could show that the signal decrease is probably caused by CNG channel inactivation (Supplementary Fig. S1B).
Discussion
CNG channels have properties that stimulated us to characterize them as tools for the sensitive detection of the intracellular second-messenger molecules cGMP and cAMP. We have shown previously that reporter cell lines expressing the homomeric, cGMP-sensitive CNGA2 channel and the calcium-sensitive photoprotein aequorin are well suited for the identification and characterization of modulators of the cGMP/NO signaling pathway. Using this reporter assay platform, we have identified and characterized activators of soluble guanylyl cyclase, phosphodiesterase inhibitors, and modulators of nitric-oxide synthesis (Wunder et al., 2005a,b, 2007).
In this report, we described a novel approach using a recombinant aequorin reporter cell line expressing a modified, heteromultimeric CNG channel. The reporter cell line enables the monitoring of Gs-coupled receptor activation and stimulation of cAMP production, which is linked to Ca2+ influx through the cAMP-gated cation channel. In addition, activation of Gq-coupled receptors can be detected via aequorin luminescence stimulated by calcium release from intracellular stores.
Similar fluorescence-based assays using homomeric CNG channels to monitor GPCR activity have been introduced. However, the observed forskolin responses, reflecting the assay sensitivity and suitability of these CNG channels as cAMP biosensors, and the reported dynamic ranges are usually low, despite the fact that modified versions of the CNGA2 channel were used (Fagan et al., 2001; Rich et al., 2001; Reinscheid et al., 2003).
Native CNG channels of olfactory sensory neurons display high cAMP sensitivity and are composed of three different subunits (Bönigk et al., 1999). Therefore, we tested heteromultimeric CNG channels composed of two or three olfactory channel subunits for their suitability as cAMP biosensors. To achieve maximal cAMP sensitivity, we used a mutated CNGA2 subunit with increased cAMP and decreased cGMP sensitivity (Altenhofen et al., 1991) instead of the wild-type CNGA2. We found that the combination of CNGA2(T537A), CNGA4, and CNGB1b shows superior forskolin sensitivity and, therefore, is best suited for the detection of Gs-coupled GPCR activation. In addition, very high signal-to-background ratios (>100) were observed. The cAMP sensitivity of the GPCR reporter assay was tested by a comparison of agonist-stimulated luminescence signals and intracellular cAMP formation measured by RIA. The results show that the cAMP sensitivity of the luminescence assay is comparable with the sensitivity of the RIA measurements. However, a high cAMP background signal was measured by RIA, which was not seen in the luminescence measurements. This might be related to intracellular cAMP compartmentalization and to the fact that CNG channels only detect local changes in cAMP levels near the cell membrane (Rich et al., 2000).
ADM, CGRP1, and CGRP2, three members of the calcitonin family of peptides, play pivotal roles in cardiovascular physiology, possess potent vasodilatory activities, and have been implicated in the pathophysiology of hypertension, heart and renal failure, circulatory shock, and migraine (Doggrell, 2001; Kurihara et al., 2003; Brain and Grant, 2004; Muff et al., 2004; Ishimitsu et al., 2006). The pioneering work of McLatchie et al. (1998), who showed that the so-called RAMP proteins regulate CRLR receptor membrane transport and determine receptor pharmacology, has enabled functional ADM and CGRP receptor expression and characterization. It has now been widely accepted that functional ADM receptors with similar pharmacology (ADM1 and ADM2 receptors) are generated by coexpression of CRLR with RAMP2 or RAMP3, respectively. In addition, the CGRP1 receptor is generated by coexpression of CRLR with RAMP1 (Poyner et al., 2002; Conner et al., 2004).
Because we were interested in the establishment of reporter cell lines expressing cardiovascular ADM and CGRP receptors, we studied the expression patterns of CRLR and RAMPs 1 to 3 in human cardiovascular-related tissues by quantitative polymerase chain reaction. We were able to detect high expression levels of CRLR, RAMP1, and RAMP2 in human heart and all blood vessels studied. Although we were able to detect RAMP3 in these tissues, the expression levels were found to be significantly lower. It is interesting that in cultured vascular endothelial cells, high levels of CRLR and RAMP2 transcripts were found, whereas RAMP1 and RAMP3 expression could not be detected. Therefore, it seems likely that ADM and CGRP peptides exert their cardiovascular activities primarily by stimulation of ADM1 and CGRP1 receptors.
According to the results of our expression analysis, we used our novel GPCR reporter cell line to establish the ADM1 (CRLR/RAMP2) and CGRP1 (CRLR/RAMP1) receptor cell lines. We characterized both cell lines using the various members of the calcitonin family of peptides. The results show that ADM acted as a potent agonist on the ADM1 receptor cell line, whereas CGRP2 and CGRP1 were approximately 30-fold and 350-fold less potent, respectively. In contrast, CGRP1 and CGRP2 were nearly equipotent agonists on the CGRP1 receptor cell line, with ADM being approximately 10-fold less potent. IMD, a novel member of this peptide family (Roh et al., 2004), activated both receptors with similar potencies. However, IMD was characterized as a partial agonist on the ADM1 receptor and as a full agonist on the CGRP1 receptor. In addition, IMD potency on both receptor cell lines was approximately 15- to 60-fold lower compared with ADM and CGRP potencies on their respective receptors. Given that IMD and ADM possess comparable vasodilatory potencies in vivo (Roh et al., 2004), we speculate that the observed IMD in vivo activity cannot be attributed solely to activation of ADM1 and CGRP1 receptors. Therefore, we speculate that additional, specific IMD receptors might exist. We also characterized the activity of amylin and calcitonin. However, these peptides did not stimulate any luminescence signals.
In addition, we characterized the competitive antagonists ADM(22-52) and CGRP(8-37) on both receptor cell lines. As anticipated, ADM(22-52) was an effective antagonist of the ADM1 receptor but had very weak activity on the CGRP1 receptor. In contrast, CGRP(8-37) was effective at antagonizing agonist action on both receptors, with approximately 40-fold higher potency at the CGRP1 receptor. Both antagonists showed competitive behavior and induced rightward shifts of the ADM and CGRP1 concentration-response curves at their respective receptors. These results correspond well to data reported previously (McLatchie et al., 1998; Aiyar et al., 2001; Poyner et al., 2002; Bailey and Hay, 2006).
We also studied the activity of shorter ADM peptides, which have been shown to possess vasodilator or vasopressor activities (Watanabe et al., 1996; Champion et al., 1999). However, none of the shorter ADM peptides, with the exception of ADM(13-52) stimulated the recombinant ADM1 receptor. In addition, none of these peptides, except for ADM(22-52), antagonized ADM responses on the ADM1 receptor. We therefore speculate that the reported cardiovascular activities are due to stimulation of receptors different from the ADM1 receptor. In addition, we could show that the simultaneous addition of ADM(1-21) or ADM(16-21) with ADM(22-52) did not stimulate any luminescence signals. The six-membered ring structure of ADM(16-21) has been implicated in receptor activation, whereas the C-terminal tail was shown to be necessary for ADM binding to its receptor (Eguchi et al., 1994; Champion et al., 1999). Therefore, the N-terminal ADM ring structure must be physically coupled to the C-terminal tail of the peptide for being able to activate the receptor. Taken together, the results of the pharmacological characterization of the ADM1 and CGRP1 receptor cell lines are in good agreement with literature data and imply that the heteromultimeric CNG channel used in our reporter cell lines represents an optimized cAMP biosensor suitable to monitor GPCR activity.
As an additional interesting application, the CNG channel assay technology allows the real-time detection of cAMP synthesis within living cells. Therefore, we characterized the luminescence signal kinetics of our receptor cell lines in the presence of calcium ions. Under these conditions, luminescence signals stimulated by activation of Gs-coupled ADM1 and CGRP1 receptors started to increase after prolonged lag times, whereas UTP-mediated activation of endogenous, Gq-coupled purinergic P2Y receptors stimulated signals with fast kinetics. To further corroborate this finding, we studied the luminescence signal kinetics after stimulation of transiently transfected BK2 and V1A receptors, which were previously characterized as Gq-coupled receptors (Liu and Wess, 1996; Zhang et al., 2001). As expected, activation of both receptors stimulated luminescence signals with fast kinetics and a single peak. In addition, we also studied luminescence signal kinetics mediated by ETA receptor activation, previously shown to positively couple to both the IP3 and the cAMP pathway (Aramori and Nakanishi, 1992). In accordance with this finding, fast and slow luminescence peaks were observed after ETA receptor stimulation. Therefore, activation of Gs- and Gq-coupled receptors is characterized by different signal kinetics, which can be used to identify intracellular receptor coupling and signaling pathways. We speculate that the difference in signal kinetics is due to compartmentalized cAMP synthesis and restricted diffusion, as described for cardiomyocytes (Fischmeister et al., 2006). This hypothesis is further supported by the observation of slow signal kinetics measured in CHO cells by fluorescence resonance energy transfer-based, cytosolic cAMP sensors after Gs-coupled receptor activation (Nikolaev et al., 2004).
The luminescence signals induced by ADM1 and CGRP1 receptor activation were found to be transient, which might be due to aequorin consumption or CNG channel inactivation. By sequential application of different agonists we could show that the signal decrease is probably caused by Ca2+/calmodulin-dependent CNG channel inactivation. A Ca2+/calmodulin-binding site has been identified in the N-terminal region of the CNGA2 channel (Liu et al., 1994).
The ADM1 and CGRP1 receptor reporter assays have now been successfully transferred to the 1536-well MTP format with minor variations (F. Wunder, unpublished data). Therefore, these reporter cell lines provide further examples for the implementation of the CNG channel assay technology in a uHTS format (Wunder et al., 2005a,b, 2007). In summary, the results presented in this report show that our novel reporter assay is well suited for the characterization of GPCR pharmacology but can also be used for uHTS purposes.
Acknowledgments
We thank Guido Buehler, Friederike Jung, and Jasmin Dischinger for excellent technical assistance. In addition, we thank Jonathan Bradley, Jürgen Franz, Stefan Golz, Franz Hofmann, Ulrich Benjamin Kaupp, and Peter Kolkhof, who kindly provided the necessary plasmid constructs, and Walter Born, for providing the CRLR receptor and for critical comments on the manuscript.
Footnotes
-
ABBREVIATIONS: ADM, adrenomedullin; CGRP, calcitonin gene-related peptide; CNG, cyclic nucleotide-gated; CRLR, calcitonin receptor-like receptor; GPCR, G protein-coupled receptor; IBMX, 3-isobutyl-1-methylxanthine; IMD, intermedin; MTP, microtiter plate; RAMP, receptor activity-modifying protein; uHTS, ultra-high-throughput screening; RT-PCR, reverse-transcriptase polymerase chain reaction; CHO, Chinese hamster ovary; RIA, radioimmunoassay; BK, bradykinin; BK2, bradykinin B2; V1A, vasopressin 1A; ETA, endothelin A; ET-1, endothelin-1; 8-Br-cAMP, 8-bromoadenosine-3′,5′-cyclic monophosphate; 8-Br-cGMP, 8-bromoguanosine-3′,5′-cyclic monophosphate; IP3, inositol 1,4,5-triphosphate.
-
↵ The online version of this article (available at http://molpharm.aspetjournals.org) contains supplemental material.
- Received September 28, 2007.
- Accepted December 27, 2007.
- The American Society for Pharmacology and Experimental Therapeutics