Neuromedin U (NMU), through its cognate receptor NMUR2 in the central nervous system, regulates several important physiological functions, including energy balance, stress response, and nociception. By random screening of our corporate compound collection with a ligand binding assay, we discovered (R)-5′-(phenylaminocarbonylamino)spiro[1-azabicyclo[2.2.2]octane-3,2′(3′H)-furo[2,3-b]pyridine] (R-PSOP), a highly potent and selective NMUR2 antagonist. R-PSOP is a nonpeptidic small-molecule with the chemical composition C20N4O2H22. In competition binding experiments, this compound was found to bind to NMUR2 with high affinity; the Ki values were determined to be 52 and 32 nM for the human and rat NMUR2, respectively. Moreover, in functional assays measuring phosphoinositide turnover or intracellular calcium mobilization, R-PSOP strongly inhibited the responses stimulated by peptide agonists NMU-25, NMU-23, and NMU-8 in human embryonic kidney 293 cells expressing NMUR2. From Schild analyses, the functional Kb values for R-PSOP were determined to be 92 and 155 nM at human and rat NMUR2, respectively. Highly selective for NMUR2, R-PSOP exhibited low affinity to the other subtype of NMU receptor, NMUR1, with a Ki value >10 μM. R-PSOP in vivo attenuated NMU-23-evoked nociceptive responses in a rat spinal reflex preparation. To our knowledge, this is the first antagonist ever reported for NMU receptors. This compound could serve as a valuable tool for further understanding the physiological and pathophysiological roles of NMU system, while providing a chemical starting point that may lead to development of new therapeutics for treatment of eating disorders, obesity, pain, and stress-related disorders.
Neuromedin U (NMU) is a brain-gut peptide originally isolated from porcine spinal cord and named for its activity to contract rat uterine smooth muscle (Minamino et al., 1985). NMU has now been found in almost all vertebrate species and is implicated in a variety of physiological processes, such as energy homeostasis, nociception, stress, inflammation, reproduction, bone remodeling, and cardiovascular and gastrointestinal (GI) functions (Brighton et al., 2004). Different molecular forms of NMU have been identified in various species: the 25-mer (NMU-25) in human and pig, the 23-mer (NMU-23) in rodents, and NMU-8 in other species. These peptides have similar biological activity and share a common C-terminal seven-amino acid segment that is essential for biological activity. NMU is widely distributed in the body, with the highest abundance in the central nervous system (CNS), adrenal gland, GI tract, and genitourinary tract (Domin et al., 1987; Ballesta et al., 1988). Recently, a 36-mer peptide, neuromedin S (NMS), was identified in rat brain. Its structure and biological activity is similar to NMU, but it is exclusively expressed in suprachiasmatic nucleus in the brain and in testes and spleen at the periphery (Mori et al., 2005; Vigo et al., 2007). NMS acts at the same receptors as NMU and may play a role in the mammalian circadian oscillator system.
Two high-affinity receptors, NMUR1 and NMUR2, have been identified for NMU and NMS. They are G protein-coupled receptors (GPCRs) formerly known as “orphan” receptor FM3/GPR66 and FM4 (Fujii et al., 2000; Howard et al., 2000; Raddatz et al., 2000). The tissue distribution of the two receptors is quite distinct, yet complementary to each other. The expression of NMUR1 seems to be confined to peripheral tissues, particularly the GI tract, pancreas, uterus, and testes (Fujii et al., 2000). NMUR2, in contrast, is predominantly expressed in the CNS, with the highest level in the arcuate and paraventricular nuclei (PVN) of the hypothalamus, nuclei accumbens, hippocampus, thalamus, medulla oblongata, and spinal cord (Guan et al., 2001).
Over the past several years, significant progress has been made toward understanding the mechanisms and physiological roles of NMU system. It is increasingly clear that NMUR2 mediates the central effects of NMU for the regulation of food intake, energy balance, stress response, and nociception. In the brain, NMUR2 is prominent in the hypothalamic regions known to be associated with regulation of food intake and energy balance. Acute central administration of NMU suppresses food intake and increases energy expenditure in rodents (Howard et al., 2000). Long-term NMU treatment reduces body weight and adiposity in mice. Genetic ablation of NMUR2 abolishes the effects of both acute and chronic central NMU injection on food intake and body weight in mice (Zeng et al., 2006; Peier et al., 2009). The high level NMUR2 expression in PVN, a major site for corticotrophin-releasing hormone release, also suggests that the NMU system may play a role in mediating stress responses. Indeed, NMU stimulates corticotrophin-releasing hormone release from hypothalamic explants in vitro, and direct NMU administration into PVN significantly elevates plasma adrenocorticotropin and corticosterone level in rodents (Thompson et al., 2004). Central NMU injection induces stress-related behaviors (Hanada et al., 2001). Conversely, NMU-knockout mice exhibit blunted behavioral response to environmental stress (Nakahara et al., 2004). NMU and NMUR2 are also abundantly expressed in nociceptive sensory pathways, including spinal dorsal horn, dorsal root ganglia, thalamus, and brain stem (Cao et al., 2003; Yu et al., 2003). Acute spinal or supraspinal NMU administration increases pain sensitivity in rodents. NMUR2-null mice exhibit reduced pain sensitivity in hot plate and formalin test (Zeng et al., 2006; Torres et al., 2007).
Based on its physiological roles, NMUR2 receptor has been proposed to be a potential target for developing new therapeutics for treatment of eating disorders, obesity, stress-related disorders, and pain. However, the existing NMU analogs, antibodies, and small interfering RNAs have limited clinical value due to their poor pharmacokinetic properties. Extensive efforts to generate a CNS-penetrable small-molecule agonist or antagonist at NMUR2 (or NMUR1) by peptidomimetic approaches have been unsuccessful thus far.
In this article, we describe the identification of a highly potent and selective small-molecule NMUR2 antagonist by screening our corporate compound collection using a novel electrochemiluminescent (ECL) ligand binding assay. We characterized the pharmacological properties of this first ever-reported NMUR2 antagonist and demonstrated its potential as a tool compound for further understanding the physiological/pathophysiological roles of NMU system and as a chemical starting point that may lead to development of new therapeutics for many CNS disorders.
Materials and Methods
Materials and Instruments. NMU-8, NMU-23, human NMU-25, bovine serum albumin (BSA), probenecid, EGTA, HEPES, and Freund's complete adjuvant (FCA) were purchased from Sigma-Aldrich (St. Louis, MO). Nonwash calcium assay kit and fluorometric imaging plate reader (FLIPR) were products from Molecular Devices (Sunnyvale, CA). Ru(bpy)2+3-NMU-23 and MSD Multi-Array High Bind 384-well black plates were custom made by MSD, Inc. (Gaithersburg, MD). Meso Scale Discovery Sector high-throughput screening (HTS) reader and ECL assay kit were also from MSD, Inc. myo-[3H]Inositol, 125I-NMU-25, and wheat germ agglutinin-coated PVT SPA beads were purchased from GE Healthcare (Chalfont St. Giles, Buckinghamshire, UK). The compound 5′-(phenylaminocarbonylamino)spiro[1-azabicyclo[2.2.2]octane-3,2′(3′H)-furo[2,3-b]pyridine] (PSOP) and its two enantiomers, R-PSOP and S-PSOP, were prepared according to the experimental procedure described in Phillips et al. (1999).
Cell Culture and Membrane Preparation. HEK293S cells stably transfected with NMUR1 or NMUR2 (human and rat) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 0.25 mg/ml hygromycin and maintained at 37°C in a CO2 incubator. For membrane preparation, the cells were harvested and resuspended in lysis buffer (50 mM Tris, pH 7.0, 2.5 mM EDTA, and 0.5 mM phenylmethylsulfonyl fluoride) at 5 × 106 cells/ml. Cells were then homogenized with a Polytron homogenizer (Brinkmann Instruments (Westbury, NY) and centrifuged at 1000g for 10 min at 4°C. Supernatants were centrifuged at 46,000g for 30 min. Membrane pellets were resuspended in membrane buffer (50 mM Tris-HCl and 0.32 M sucrose, pH 7.0), aliquoted, frozen in dry ice/ethanol, and stored at -70°C until use. The protein concentration was determined by Bradford method using BSA as the standard (Bradford, 1976).
ECL Ligand Binding Assay and HTS. ECL membrane binding assay was carried out according to the instructions for the assay kit from MSD, Inc. In brief, NMUR2 membranes were diluted in assay buffer (25 mM HEPES, 5 mM MgCl2, and 1 mM CaCl2, pH 7.4) and dispensed at 1 μg protein/well in 2 μl into assay plates (MSD Multi-Array High Bind 384-well black wall). The immobilization of the receptors to the bottom surfaces of the plates occurred when plates were incubated at room temperature for 1 h. Thirteen microliters of a blocking agent (containing BSA) was added to minimize nonspecific binding. After addition of test compounds to each well, Ru(bpy)2+3-NMU-23 was added to the assay plates to start the binding reaction. Plates were incubated at room temperature for 1 h. At the end of the incubation, reading buffer from the assay kit containing tripropylamine was added to each well, and the ECL readings were immediately taken on Sector HTS reader (MSD, Inc.). The total reaction volume of the binding reaction was 25 μl/well, and the final labeled ligand concentration was 0.5 nM.
To eliminate false positives from color quenching and chemical reactivity, the active samples were counter-screened in an assay measuring the binding of biotinylated Ru(bpy)2+3-IgG to the surface of avidin-coated plates.
125I-NMU-25 Competition Binding. Cell membranes containing NMU receptors were diluted in binding buffer (25 mM HEPES, pH 7.4, 5 mM MgCl2, 1 mM CaCl2, and 0.2% BSA) and incubated with wheat germ agglutinin-coated SPA beads for 1 h on ice. The mixture of membranes and SPA was dispensed into each well of a 96-well white opaque plate containing test compounds or peptides in binding buffer. After addition of 125I-NMU-25, the plate was put on a shaker and incubated at room temperature for 1 h. The radioactivities of the bound ligand were measured by a TopCount reader (PerkinElmer Life and Analytical Sciences, Boston, MA). Nonspecific binding was defined by the binding in the presence of 1 μM NMU-8.
Intracellular Calcium Mobilization. HEK293S cells stably expressing NMU receptors were plated onto a poly-d-lysine-coated 384-well black plate with a clear bottom. After overnight culture, the cells were loaded with nonwash calcium-sensitive fluorescent dye (in calcium 4 assay kit from Molecular Devices) for 30 min and incubated with test compounds in Hanks' buffer for 15 min. The plate was then transferred onto a FLIPR, and the peptide agonist was applied to the cells. The fluorescence response, as an indicator of changes in intracellular calcium concentration triggered by agonist stimulation, was recorded with excitation wavelength at 488 nm and emission wavelength at 510 nm, and the data were analyzed by FLIPR software.
Phosphoinositide Turnover. HEK293S cells expressing NMUR2 receptor were plated onto 24-well plates and grown in medium containing 5 μCi/ml myo-[3H]inositol for 2 days. After removal of the medium, the cells were preincubated with test compounds in Hanks' buffer (BD Bioscience, San Jose, CA) containing 10 mM LiCl for 10 min. The peptide agonist was added to each well, and the plate was incubated at 37°C for 45 min. The amount of [3H]inositol phosphate that accumulated in the cells, as a result of the stimulated phosphoinositide turnover, was measured by an ion exchange column chromatography method, as described previously (Liu et al., 2003).
Broad in Vitro Pharmacological Selectivity Profiling. R-PSOP was evaluated in vitro for its activities at a broad spectrum of biological targets, including receptors, ion channels, enzymes, and transporters. The studies were mainly performed in MDS Pharma Service, Inc. (Toronto, ON, Canada) and BioSignal Packard, Inc. (Montréal, QC, Canada) (a subsidiary of PerkinElmer Life and Analytical Sciences) using standard methods and assay conditions according to the literature and as described on the websites of the two outsourcing companies.
In Vivo Studies. In vivo studies on nociceptive responses were carried out on adult male Sprague-Dawley rats (Charles River Canada, St. Constant, QC, Canada) by experimenters blinded to the treatment that the animals received. All experiments were approved by the AstraZeneca R&D Montréal Animal Care Committee in accordance with the Canadian Council on Animal Care and the Association for the Assessment and Accreditation of Laboratory Animal Care guidelines. Rats were tested naive or 24 h after intraplantar injection of FCA into one hind paw.
The heat sensitivity of the plantar surface of the paw was assessed as described previously (Yu et al., 2003). In brief, rats were placed on a transparent plastic enclosure on a glass surface maintained at 30°C and habituated for at least 30 min. The pain thresholds were measured using a radiant heat source connected to an automatic timer. When the hind paw was lifted, the heat automatically shut off and the latency was recorded. Two readings were taken at 5-min intervals. The cut-off time of 20 s was established to prevent tissue injury. R-PSOP in 10 μl of saline was delivered by intrathecal injection under brief isoflurane anesthesia.
Electrophysiological recording of flexor reflex responses were performed as described previously (Yu et al., 2003). In brief, extracellular recordings of flexor α-motoneuron activity were made from a filament of the nerve to the posterior biceps femoris-semitendinosus muscles in decerebrate-spinal rats. The excitability of the flexor reflex was determined by measuring efferent activities evoked by touch, pinch, and heat (water, 52°C) stimuli applied to the hind paw. Drugs were delivered via an intrathecal cannula in a volume of 10 μl, followed by a 10-μl saline flush.
Pharmacokinetic Profiling. R-PSOP was dissolved in saline and administered to adult male Sprague-Dawley rats via oral or intravenous route. Blood samples were collected at different time points after dosing (5 min–11 h), and the plasma was isolated by centrifugation. Plasma proteins were precipitated with ice-cold acetonitrile containing 0.1% formic acid. After centrifugation, R-PSOP in the resulting supernatants was quantitated by high-performance liquid chromatography-mass spectrometry. The pharmacokinetic parameters of R-PSOP were estimated using standard noncompartmental methods. For measurement of CNS penetration, the rats were sacrificed after intravenous dosing, and the distribution of R-PSOP in the brain, spinal cord, and plasma was analyzed by high-performance liquid chromatography-mass spectrometry.
Data Analysis. HTS data were analyzed with ActivityBase (IDBS, Inc., Guildford, UK) and Spotfire (TIBCO, Somerville, MA) software. The Z′-factor, as an indicator of assay performance in a HTS setting, was calculated based on the method described in Zhang et al. (1999). The binding Kd, Ki, and functional Kb values were calculated using Prism (GraphPad Software Inc., San Diego, CA). The statistical analyses of in vivo data were also performed with Prism.
ECL Ligand Binding Assay and HTS. Radioligand binding assay is a sensitive and reliable technology for detection of compounds that act at membrane receptors. However, the utility of a radioactive assay in a large-scale HTS is limited due to safety and waste concerns. Several nonradioactive alternatives have been developed. Among them, ECL is the simplest homogeneous (nonseparation) method with sensitivity comparable with the radioactive assays. The principle and applications of ECL technology have been thoroughly described elsewhere (Marquette and Blum, 2008; Miao, 2008). To establish an ECL binding assay for NMUR2 receptor, the peptide NMU-23 was labeled at the N terminus with Ru(bpy)2+3 and allowed to bind to the human NMUR2 receptor in the cell membranes immobilized on the electrode on the bottom of each assay plate. Upon application of an electric current, the receptor-bound Ru(bpy)2+3-ligand undergoes an oxidation-reduction cycle in the presence of a coreactant tripropylamine and emits light. Signal is only generated when the Ru(bpy)2+3 label is in proximity to the electrode, thus discriminating the bound label from the unbound and enabling a no wash, homogenous assay format. In this assay, the active compounds that compete with Ru(bpy)2+3-NMU23 for binding to NMUR2 would be detected by the decrease in ECL intensity.
Figure 1 shows the structure of the ligand, Ru(bpy)2+3-NMU23, and the saturation and competition binding curves from the ECL assay on human NMUR2. From the saturation binding experiment, the Kd value and the total number of binding sites (Bmax) were determined to be 0.8 nM and 4 pmol/mg membrane protein, respectively. From competition binding experiments, the Ki values for NMU-25 (human), NMU-23, and NMU-8 (porcine) were determined to be 3.2, 4.7, and 3.0 nM, respectively. The Ki values from ECL assay for were very close to those derived from a parallel radioactive binding assay with 125I-NMU-25 (reported below) and in general agreement with literature values (Hosoya et al., 2000; Howard et al., 2000; Kojima et al., 2000; Raddatz et al., 2000; Shan et al., 2000; Szekeres et al., 2000).
The ECL-based NMUR2 binding assay was used to screen our corporate compound collection. Approximately 670,000 compounds with diverse structures were tested at 10 μM for their ability to displace the binding of 0.5 nM Ru(bpy)2+3-NMU-23 to human NMUR2 receptors. The assay performance was strong during the entire HTS, with a signal/background window >11 and a Z′-factor >0.7. Figure 2A shows the scattergram of the data from a typical 384-well plate. In total, 355 compounds showed activities >50% and were identified as putative hits. After retest in triplicates, 57 hits were confirmed. Among them, the most active compound was PSOP. PSOP is a nonpeptidic small molecule with the chemical composition C20N4O2H22. Because PSOP is racemic, the two enantiomers of PSOP were synthesized, and their binding affinities at NMUR2 were evaluated. R-PSOP was found to be the active enantiomer, whereas S-PSOP was inactive. The structures of PSOP (top) and R-PSOP (bottom) are shown in Fig. 2B.
High-Affinity Binding of R-PSOP to NMUR2. The affinities of R-PSOP binding to human and rat NMUR2 were evaluated in both ECL and radioactive membrane binding assays, and similar results were obtained from the two different assays. Figure 3 shows that R-PSOP inhibits the binding of 125I-NMU-25 to human NMUR2 membranes in a concentration-dependent manner and with a Hill coefficient close to unity, suggesting that R-PSOP binds competitively to the same site on the receptor as NMU-25. From radioactive competition binding assays, the Ki values were determined to be 52 and 32 nM for the human and rat NMUR2, respectively. The reference peptide NMU-8 had Ki values at 3 and 4.4 nM in the same assay. From ECL binding assay, similar Ki values were obtained for R-PSOP binding to human and rat NMUR2 (78 and 44 nM, respectively).
Functional Antagonist Activity of R-PSOP at NMUR2. Through the Gq/11 signaling pathway NMU stimulates the classical response of PI turnover and intracellular calcium mobilization in HEK293 cells expressing NMUR2. Pretreatment of the cells with R-PSOP strongly inhibited the second messenger response induced by agonist NMU-25, NMU-23, or NMU-8, although the compound itself has no intrinsic agonist activity in either PI turnover or intracellular calcium mobilization assay.
As shown in Fig. 4, R-PSOP concentration-dependently inhibited the PI turnover response in human NMUR2-expressing cells stimulated by 10 nM NMU-25 (EC50 of ∼5 nM). The IC50 value was determined to be 86 nM (n = 2). At 1 μM, R-PSOP completely blocked the agonist response. In a separate experiment, it was found that R-PSOP treatment had no significant effect on the basal level of PI turnover in the cells treated with LiCl, indicating that R-PSOP did not have inverse agonist activity, rather, it behaved as a pure antagonist at NMUR2.
To further characterize the mode of action, we performed Schild analysis on the effects of R-PSOP on the intracellular calcium mobilization response induced by NMU-25 in HEK293 cells expressing human or rat NMUR2. The response was measured by FLIPR using a calcium-sensitive dye. Figure 5A shows the curves of the agonist concentration-dependent response in the presence of increasing concentrations of R-PSOP. As the concentration of the inhibitor increases, the curves were parallel shifted to the right, consistent with a competitive mode of inhibition. From Schild analysis (Fig. 5B), the Kb values were calculated to be 92 and 155 nM at human and rat NMUR2, respectively.
A 5-min wash of the cells with Hanks' buffer after R-PSOP treatment almost completely restored the agonist-induced response, suggesting that the interaction of this antagonist with NMUR2 is reversible as well as competitive.
Selectivity of R-PSOP at NMUR2 over Other Biological Targets. For selectivity evaluation, PSOP and R-PSOP was tested at the other subtype of NMU receptor, NMUR1, in 125I-NMU-25 binding assay. The binding activity of PSOP and R-PSOP was very weak at both human and rat NMUR1, each with a Ki value >10 μM. Moreover, in NMUR1 functional assay PSOP and R-PSOP were inactive at concentrations up to 10 μM in inducing calcium mobilization response or in inhibiting the response induced by NMU-23. The data indicate that R-PSOP has a >200-fold selectivity at NMUR2 over NMUR1.
In addition, R-PSOP was tested in vitro for activities at a panel of >100 various biological targets, including receptors, ion channels, transporters, and enzymes. R-PSOP was inactive at all the targets, except the ionotropic 5-HT3 and α7 nicotinic receptors. R-PSOP bound to human 5-HT3 and α7 receptors, with Ki values at 220 and 460 nM, respectively, and activated the channels at >5 μM concentrations. The racemic compound PSOP showed a similar activity. Table 1 provides an abbreviated list of the targets tested, the assays used, and the activities of R-PSOP in those assays. A complete list is also provided in Supplemental Data. Overall, R-PSOP has a reasonably clean in vitro pharmacological profile.
Physicochemical and Pharmacokinetic properties of R-PSOP. To assess the usefulness of R-PSOP as a tool compound for in vivo studies, we have evaluated its physicochemical and pharmacokinetic properties. R-PSOP has a molecular weight of 350 and excellent physicochemical properties: polar surface area, 67 A2; log P, 2.2; log D at pH 7.4, 0.99; and aqueous solubility, 6.5 mM. The intrinsic clearance was low; 8 and 13 μl/min/mg when measured with human and rat liver microsomal preparations. R-PSOP also showed good permeability in CaCo2 assays with A-to-B apparent membrane permeability at 3 and B-to-A apparent membrane permeability at 5.6 μM · cm/s at pH 7.4, respectively. Pharmacokinetic evaluation of R-PSOP in rats (Table 2) indicated that R-PSOP had low oral bioavailability, moderate half-life, and high clearance. The CNS penetration of PSOP was moderate, as evidenced by the brain/plasma ratio of approximately 0.2 and spinal cord/plasma ratio of approximately 0.4. The physicochemical and pharmacokinetic properties of R-PSOP support its utility in vivo for interrogating the effects of NMUR2 inhibition in animal models.
In Vivo Effect of R-PSOP on Nociceptive Responses Evoked by NMU-23 or by Paw Inflammation. To establish whether R-PSOP acted as a functional antagonist of NMUR2 in their native environment, we tested R-PSOP in vivo on spinal nociceptive reflexes, because NMUR2s are expressed by intrinsic neurons in spinal cord. As reported previously (Yu et al., 2003), NMU-23 applied to the surface of the spinal cord via an intrathecal cannula in a decerebrate spinalized rat preparation increased nociceptive spinal reflex responses to pinch of the paw. R-PSOP injected intrathecally in this preparation 15 min after NMU-23 application significantly attenuated the enhanced responses to pinch (Fig. 6). A similar attenuation of enhanced responses to touch of the paw was also observed. When R-PSOP was given intrathecally in the absence of NMU pretreatment, it had no effect on the reflex responses, suggesting there is no NMU tone in the rat spinal cord under normal conditions (Fig. 6).
We further assessed the effect of R-PSOP given intrathecally in rats with acute paw inflammation induced by FCA, both electrophysiologically and behaviorally. R-PSOP had no marked effect on the enhanced reflex responses to pinch, touch, or heat in FCA-treated rats. Likewise, R-PSOP had no effects on the reflex withdrawal latency to heat stimuli applied to the paw in awake, intact rats. There was no significant difference in response latency to a standard heat stimulus between vehicle-treated rats (latency, 7.0 ± 0.4 s; n = 7) and those treated with 5 nmol of R-PSOP in 10 μl of saline (latency, 6.9 ± 0.4 s; n = 7). These data suggest that there is no elevated NMU tone in the rat spinal cord induced by acute inflammation.
Taken together, we have identified R-PSOP from HTS and demonstrated that it is a highly potent and selective NMUR2 antagonist. This compound binds competitively to NMUR2 with high affinity, and it reversibly blocks the activation of the receptor by peptide agonist. R-PSOP also acts as a functional antagonist in vivo attenuating NMUR2-mediated nociceptive responses.
A sensitive and HTS-compatible nonradioactive homogeneous ligand binding assay for membrane receptors, especially GPCRs, has always been challenging. Using the ECL technology, we developed an HTS assay to screen for compounds that bind to NMUR2 receptors. The assay is simple, fast, and highly sensitive. The major caveat in the use of this assay technology is the interference produced by compounds with strong colors or high oxidation-reduction potential. This shortcoming can be overcome by counter-screening with biotinylated Ru(bpy)2+3-IgG binding to avidin-coated plates. In addition, because the labeled ligand starts to dissociate from the receptor upon addition of the reading buffer, the plate reading has to be made within an inconvenient short time frame of 2 min. Nonetheless, an HTS campaign was successfully completed with the ECL membrane binding assay, and a highly potent compound, R-PSOP, was discovered.
The binding of R-PSOP to NMUR2 seems to be competitive with NMU-25 and NMU-23. It is possible that R-PSOP binds to the same or overlapping site as the peptide agonist. Comparing the structure of R-PSOP with the highly conserved C-terminal seven-amino acid segment of NMU peptide, there are two apparent common features that may be critical for receptor binding: a positively charged amine group and a lipophilic aromatic group. The tertiary amine group in the azabicyclic octane ring and the terminal phenyl group in R-PSOP could correspond to the arginine and phenylalanine residues, respectively, in the peptide. Similar compounds without the phenyl group were inactive in the binding assay. Correspondingly, changing the phenylalanine residue destroys the activity of the peptide (Brighton et al., 2004). The amphiphilic nature of the ligand binding site is corroborated by the structures of two recently disclosed nonselective small-molecule NMUR agonists, (2-(trityloxy)ethyl)guanidine and (3-(trityloxy)propyl)guanidine. It was reported that when the guanidine group was replaced by nonbasic substituents, such as 2-(thiophen-2-yl)ethanyl and 2-(pyridin-4-yl)ethanyl, the bioactivity of the resulting compounds was lost (Meng et al., 2008).
The binding site for peptide agonist is generally believed to be well conserved between NMUR1 and NMUR2. It is surprising that R-PSOP achieves more than >200-fold selectivity at NMUR2 over NMUR1, while binding to the same site as the peptide agonist. Sequence alignment and molecular modeling based on the known GPCR crystal structures might provide insight into mechanisms of R-PSOP interaction with NMUR2 and the subtype selectivity. Site-directed mutagenesis could also help pinpoint the residues on the receptor responsible for the selectivity. It is likely that some steric hindrance from certain divergent residues on NMUR1 prevents R-PSOP from reaching the binding pocket. Even though R-PSOP per se is inactive at NMUR1, it might be possible to generate NMUR1-active compounds by modifying the structure of R-PSOP. Indeed, R-PSOP could be the chemical starting point for development of compounds with various pharmacological properties, including agonists, antagonists, or inverse agonists, at NMUR1 and NMUR2.
R-PSOP has no intrinsic agonist or inverse agonist activity at NMUR2. Probably, it exerts its effect by simply blocking the natural agonist from reaching to the binding pocket of the receptor, thus preventing receptor activation. Therefore, the effects of R-PSOP in vivo on various physiological processes are very likely to be dependent on the tone of the NMU system involved. The results we observed in the nociceptive system reinforce this hypothesis. Thus, R-PSOP had no effect on nociceptive responses when given intrathecally either in normal, naive rats, or in animals with an acute inflammation of the paw. In contrast, R-PSOP was effective at attenuating the effects of exogenously applied NMU in the same system, indicating that it reached the spinal neurons expressing NMUR2 receptors. The reversal of the effects of NMU was not complete, probably due to failure to block the effects on NMUR1, which is also expressed in the pain pathway in the spinal cord, although on a different cell type, primary afferent terminals (Yu et al., 2003).
NMU and its receptors have been implicated in an ever-increasing number of physiological processes, including food intake, energy balance, stress response, inflammation, nociception, circadian rhythms, reproduction, and GI and cardiovascular functions (Brighton et al., 2004). In general, three lines of evidence have been used to support the claims: tissue distribution data, the pharmacological effects of peptides, and the phenotypes of genetically modified animals. However, exogenously applied NMU peptide may not reflect true physiological conditions, and the phenotypic analysis of knockout mice might be limited by gene compensation and developmental effects. Potent and selective antagonists at NMU receptors, such as R-PSOP, will provide pharmacological tools to elucidate further the physiological and pathophysiological roles of NMU system.
NMUR2 has been proposed to be a drug target for developing new therapeutics for several major unmet medical needs (Howard et al., 2000; Guan et al., 2001; Brighton et al., 2004; Sato et al., 2007). For example, NMUR2 agonists could potentially be used to treat obesity and metabolic disorders, whereas NMUR2 antagonists were hypothesized to be beneficial for anorexia, chronic pain, osteoporosis and stress-related disorders. The discovery of R-PSOP as a potent and selective antagonist at NMUR2 represents a breakthrough toward the realization of the potential of NMUR2 as a drug target. Further evaluation of the effects of systemic or local administration of R-PSOP or its improved analogs in various animal models will help address many unanswered questions and provide insight into future directions.
In conclusion, we have identified from HTS a highly potent and selective NMUR2 antagonist. This compound could be a valuable tool for further understanding the physiological and pathophysiological roles of the NMU system and provides a chemical starting point that may lead to the development of new therapeutics for several CNS disorders.
We thank Dr. Manon Valiquette for providing all the cell lines, and Dr. Clay Scott, Dr. James Campbell, Kenneth Mitchell, Dr. Norman Ledonne, and Linda Sygowski for critically reading the manuscript and for fruitful discussions.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: NMU, neuromedin U; GI, gastrointestinal; CNS, central nervous system; NMS, neuromedin S; NMUR, neuromedin U receptor; GPCR, G protein-coupled receptor; PVN, paraventricular nuclei; ECL, electrochemiluminescent; BSA, bovine serum albumin; FCA, Freund's complete adjuvant; FLIPR, fluorometric imaging plate reader; HTS, high-throughput screen; PSOP, 5′-(phenylaminocarbonylamino)spiro[1-azabicyclo[2.2.2]octane-3,2′(3′H)-furo[2,3-b]pyridine]; HEK, human embryonic kidney; PI, phosphoinositide; 5-HT, 5-hydroxytryptamine; MK-912, (2S-trans)-1,3,4,5′,6,6′,7,12b-octahydro-1′,3′,-dimethyl-spiro[2H-benzofuro[2,3-a]quinolizine-2,4′(1′H)-pyrimidin]-2′-(3′H)-one hydrochloride hydrate; SR141716A, N-(piperidin-1-yl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboximide hydrochloride; SR140333, nolpitantium; GR-65630, 3-(5-methyl-1H-imidazol-4-yl)-1-(1-methyl-1H-indol-3-yl)-1-propanone; WIN55212-2, R-(+)-[2,3-dihydro-5-methyl-3-[(morpholinyl)methyl] pyrrolo[1,2,3-de]-1,4-benzoxazinyl]-(1-naphthalenyl)-methanone mesylate; LY341495, 2S-2-amino-2-(1S,2S-2-carboxycyclopropyl-1-yl)-3-(xanth-9-yl) propanoic acid; L-703606, cis-2-(diphenylmethyl)-N-[(2-iodophenyl)methyl]-1 azabicyclo[2.2.2]octan-3-amine; MDL-72222, 3-tropanyl-3,5-dichlorobenzoate.
- Received February 27, 2009.
- Accepted April 13, 2009.
- The American Society for Pharmacology and Experimental Therapeutics