Abstract
In this study, we characterized a muscarinic acetylcholine receptor (mAChR) potentiator, LY2119620 (3-amino-5-chloro-N-cyclopropyl-4-methyl-6-[2-(4-methylpiperazin-1-yl)-2-oxoethoxy]thieno[2,3-b]pyridine-2-carboxamide) as a novel probe of the human M2 and M4 allosteric binding sites. Since the discovery of allosteric binding sites on G protein–coupled receptors, compounds targeting these novel sites have been starting to emerge. For example, LY2033298 (3-amino-5-chloro-6-methoxy-4-methyl-thieno(2,3-b)pyridine-2-carboxylic acid cyclopropylamid) and a derivative of this chemical scaffold, VU152100 (3-amino-N-(4-methoxybenzyl)-4,6-dimethylthieno[2,3-b]pyridine carboxamide), bind to the human M4 mAChR allosteric pocket. In the current study, we characterized LY2119620, a compound similar in structure to LY2033298 and binds to the same allosteric site on the human M4 mAChRs. However, LY2119620 also binds to an allosteric site on the human M2 subtype. [3H]NMS ([3H]N-methylscopolamine) binding experiments confirm that LY2119620 does not compete for the orthosteric binding pocket at any of the five muscarinic receptor subtypes. Dissociation kinetic studies using [3H]NMS further support that LY2119620 binds allosterically to the M2 and M4 mAChRs and was positively cooperative with muscarinic orthosteric agonists. To probe directly the allosteric sites on M2 and M4, we radiolabeled LY2119620. Cooperativity binding of [3H]LY2119620 with mAChR orthosteric agonists detects significant changes in Bmax values with little change in Kd, suggesting a G protein–dependent process. Furthermore, [3H]LY2119620 was displaced by compounds of similar chemical structure but not by previously described mAChR allosteric compounds such as gallamine or WIN 62,577 (17-β-hydroxy-17-α-ethynyl-δ-4-androstano[3,2-b]pyrimido[1,2-a]benzimidazole). Our results therefore demonstrate the development of a radioligand, [3H]LY2119620 to probe specifically the human M2 and M4 muscarinic receptor allosteric binding sites.
Introduction
Acetylcholine activates two families of receptors: the nicotinic ligand-gated ion channel receptors (nAChRs) and the G protein–coupled muscarinic receptors (mAChRs), classified initially based on their differential activation by nicotine (Lindstrom, 1997) and muscarine (Wess, 1996), respectively. The wide distribution of mAChRs in the central nervous system (CNS) and periphery support their involvement in physiologic processes such as arousal, cognition, pain, exocrine gland secretion, smooth muscle, and vascular contraction (Wess et al., 2003). Muscarinic acetylcholine receptors have long been viewed as viable targets for developing therapeutic agents to treat Alzheimer’s disease and other CNS disorders. The muscarinic agonist xanomeline, for example, was developed to treat Alzheimer’s disease (Bodick et al., 1994), but it was also found to induce improvements in positive, negative, and cognitive symptoms associated with schizophrenia (Shekhar et al., 2008). However, the lack of selectivity of xanomeline led to peripheral side effects that prohibited it from advancing in the clinic. Because of the highly conserved sequence within the acetylcholine binding domain (Heinrich et al., 2009), targeting the orthosteric site for small molecule development resulted in a number of muscarinic agonist compounds with poor selectivity.
However, the discovery of allosteric sites on G protein–coupled receptors (GPCRs) is allowing more selective small molecule modulators to emerge, offering a unique approach to treating CNS diseases (Christopoulos, 2002; May et al., 2007). Allosteric modulators bind to novel sites that are distinct from the natural transmitter orthosteric binding site. Positive allosteric modulators enhance the affinity and/or efficacy of the endogenous ligand and have a number of therapeutic advantages compared with direct-acting agonists such as xanomeline. Besides the observation of improved receptor selectivity, positive allosteric modulators offer physiologically relevant spatial and temporal signaling that may limit undesirable side effects compared with direct-acting agonists that can lead to desensitization and long-term changes in receptor densities (Christopoulos, 2002).
Several in vitro pharmacologic methods can be used to assess allosteric binding and functional signaling modulation. Typically, muscarinic allosteric agents have been identified by the effect they induce on radioligand competition experiments at the orthosteric site (Birdsall et al., 1997). For example, allosteric modulators can be revealed by their inability to block fully radiolabeled orthosteric probes such as [3H]NMS ([3H]N-methylscopolamine). Positive allosteric modulation is typically assessed by functional signal transduction studies using various assay formats, including measurement of cAMP generation and [35S]GTPγS [5′-O-(3-[35S]thio)triphosphate] binding. Because previous methods for studying allosteric binding mechanisms have been limited to indirect measurements using orthosteric binding, we describe here the development of a radiotracer from LY2119620 (3-amino-5-chloro-N-cyclopropyl-4-methyl-6-[2-(4-methylpiperazin-1-yl)-2-oxoethoxy] thieno[2,3-b]pyridine-2-carboxamide) that allows direct labeling of the muscarinic allosteric site. In addition, we discuss the possibility of being able to discern muscarinic allosteric binding sites in native tissue. Our data support the hypothesis that allosteric selectivity between M2 and M4 mAChR subtypes with LY2119620 is a result of differences in cooperativity and not affinity of the orthosteric agonist, similar to what has been previously described for thiochrome (Lazareno et al., 2004) and LY2033298 (3-amino-5-chloro-6-methoxy-4-methyl-thieno(2,3-b)pyridine-2-carboxylic acid cyclopropylamide) (Leach et al., 2010) at the M4 mAChR.
Materials and Methods
Chinese hamster ovary cell lines stably expressing human M1 (Bmax NMS = 4.4 pmol/mg membrane), M2 (Bmax NMS = 11.0 pmol/mg membrane), M3 (Bmax NMS = 7.64 pmol/mg membrane), M4 (Bmax NMS = 3.3 pmol/mg membrane), or M5 (Bmax NMS = 4.2 pmol/mg membrane) were purchased from PerkinElmer (Waltham, MA). Chemicals and ligands were purchased from the following sources: oxotremorine-M, VU152100 (3-amino-N-(4-methoxybenzyl)-4,6-dimethylthieno[2,3-b]pyridine carboxamide), and VU10010 (3-amino-N-[(4-chlorophenyl)methyl]-4,6-dimethylthieno[2,3-b]pyridine-2-carboxamide) from Tocris (Bristol, UK); acetylcholine, WIN 62,577 (17-β-hydroxy-17-α-ethynyl-δ-4-androstano[3,2-b]pyrimido[1,2-a]benzimidazole), and gallamine triethiodide (Sigma-Aldrich, St. Louis, MO); [3H]NMS (GE Healthcare, Piscataway, NJ); [3H]LY2119620, [35S]GTPγS (PerkinElmer), LY2119620, and LY2033298 (Lilly, Indianapolis, IN).
[3H]NMS Binding Assays.
[3H]NMS binding assays were performed in HEPES buffer [20 mM HEPES, 100 mM sodium chloride (NaCl), 10 mM magnesium chloride (MgCl2), pH 7.4] as described previously (Chan et al., 2008) with the following modifications. Briefly, frozen membrane preparations were thawed and resuspended in assay buffer, and approximately 25 µg protein was added to each well and incubated with [3H]NMS for 2 hours at room temperature in a total volume of 200 µl in polypropylene 96-deep well plates. Nonspecific binding was determined using 10 µM atropine. Membranes were collected by rapid filtration using a Tomtec cell harvester (Tomtec, Inc., Hamden, CT) through GF/A filters that had been presoaked in 0.3% polyethyleneimine. The filters were washed with 5 ml of ice-cold 50 mM Tris buffer (pH 7.4) and air-dried overnight. The dried filters were treated with MeltiLex A melt-on scintillator sheets, and the radioactivity retained on the filters was counted using a Wallac 1205 Betaplate scintillation counter (PerkinElmer). Displacement experiments for [3H]NMS were carried out in the presence of various concentrations of compounds for all five human muscarinic receptor subtypes. More specifically, in the potentiation experiments with [3H]NMS, various concentrations of orthosteric agonists were used to displace the radioligand but in the presence of 10.0, 1.0, 0.1, or 0 µM LY2119620. The dissociation kinetic binding assays were performed using a reverse-time protocol. For these experiments, P1 membrane preparations of a Chinese hamster ovary cell line stably expressing either the human M2 or M4 muscarinic mAChR were used. Membranes were added to approximately 1.0 nM [3H]NMS in the presence or absence of LY2119620 and allowed to equilibrate for 2 hours at room temperature. Once equilibrated, 100 μM oxotremorine-M was added in a time-staggered approach to allow 1- to 60-minute time point collection. In the statistical analyses, Ki values were determined from the Cheng-Prusoff relationship:where IC50 is determined from a four-parameter fit of displacement curves, [ligand] = 1 nM [3H]NMS, and Kd is the equilibrium dissociation constant of [3H]NMS for each mAChR subtype determined by saturation binding experiments carried out by the membrane supplier.
[3H]LY2119620 Binding Assays.
[3H]LY2119620 saturation binding assays were performed in HEPES buffer (20 mM HEPES, 100 mM NaCl, 10 mM MgCl2, pH 7.4). Saturation binding was initiated by incubating 15 μg of muscarinic-containing membranes (hM1-M5; PerkinElmer), orthosteric ligand (100 μM unless otherwise noted; oxotremorine-M or acetylcholine), and various concentrations of hot-ligand [3H]LY2119620 (0.2–60.0 nM) for 1 hour at room temperature, although equilibrium was achieved within 15 minutes (data not shown). [3H]LY2119620 displacement assays were performed in HEPES buffer as described previously. Muscarinic-containing membranes M2 or M4 receptors were incubated with 100 μM oxotremorine-M and [3H]LY2119620 at approximately the Kd concentration of the receptor, and varying concentrations (0.1 nM–10 μM) of allosteric ligands gallamine, VU152100, VU10010, WIN 62,577, LY2033298, and LY2119620. Incubations were carried out for 1 hour at room temperature. All reactions were stopped by rapid filtration on a Tomtec 96-well cell harvester. Nonspecific binding was determined using 10 μM LY2033298. Radioactivity retained on the filtermats was counted on a Wallac 1205 Betaplate. In the statistical analyses, the specific binding versus time data were fit to a one-site specific binding model using GraphPad Prism 6.7 and the Bmax and Kd for the allosteric molecule was calculated for each orthosteric ligand. Ki values were determined using the Cheng-Prusoff relationship for [3H]LY2119620 displacement studies.
Autoradiographic Studies Using [3H]LY2119620.
Male cynomolgus monkey brains were supplied from Covance (Greenfield, IN). Brains were rapidly removed, placed in ice-cold phosphate-buffered saline (PBS) for 5 minutes, frozen on dry ice, and then stored at −80°C. The brains were mounted onto chucks and sectioned at 12 µm using a cryostat (Zeiss, Thornwood, NY). Sagittal sections were thaw-mounted onto gelatin-coated slides and stored at −80°C until being assayed. Sections were initially preincubated for 10 minutes in PBS at room temperature. The sections were then placed into polypropylene containers with approximately 5.0 nM [3H]LY2119620 and either 100 µM acetylcholine or oxotremorine-M. Some near-adjacent sections were also incubated with 10 µM LY2033298 to define nonspecific binding. After a 1-hour incubation, the sections were rinsed with fresh ice-cold PBS on ice for 10 minutes each and dried rapidly. The labeled sections were exposed to FujiFilm Imaging Plate for 15 days. The plate was read in Fuji BAS-5000 and analyzed using MCID Software (Cambridge, UK).
Results
Previous studies have revealed that the mAChRs possess at least one allosteric site located extracellularly to the orthosteric site (Wess, 2005). This pocket is referred to as the “common” allosteric site because prototypical modulators, such as gallamine, alcuronium, and C7/3-phth, interact with all five mAChR subtypes, albeit with different degrees of affinity or selectivity (Christopoulos et al., 1999). In this study, we describe the identification of a novel positive allosteric modulator, LY2119620, for common allosteric site on the human M2 and M4 mAChR. For comparison’s sake, we also evaluated other muscarinic allosteric compounds, including LY2033298, VU10010, VU152100, gallamine, and WIN 62,577 (Fig. 1).
Binding Analysis of LY2119620 Using [3H]NMS.
Shown in Fig. 2, LY2119620 displays little to no binding affinity for all five human mAChRs to the orthosteric pocket when interacting with the nonselective antagonist radioligand [3H]NMS. In contrast, the orthosteric nonselective antagonist atropine caused a concentration-dependent inhibition in the binding of [3H]NMS for all five mAChRs (Fig. 2). The competition for atropine and LY2119620 for M1–M5 mAChRs is summarized in Table 1. To address whether LY2119620 interacts with the human mAChRs in an allosteric manner, radioligand dissociation experiments were performed to determine whether coincubating LY2119620 with the nonselective muscarinic agonist, oxotremorine-M, changed the off-rate (t1/2) of [3H]NMS because altered dissociation rates can be indicative of an allosteric interaction. Figure 3 illustrates the dissociation of [3H]NMS by oxotremorine-M in the presence of various concentrations of LY2119620 for both the human M2 and M4 mAChRs. The off-rate of [3H]NMS in the presence of oxotremorine-M was significantly reduced for both M2 (Fig. 3A) and M4 (Fig. 3B) by LY2119620 and could be fitted to a one-phase exponential decay model. The t1/2 for M2 mAChR alone was 2.5 minutes and in the presence 10 µM LY2119620 doubled to 5.8 minutes. The t1/2 of the radioligand at M2 mAChR decreased in a concentration-dependent manner as more LY2119620 was added. The t1/2 at 20 µM was 15.2 minutes and at 40 µM increased to 51.8 minutes. The t1/2 for the dissociation of [3H]NMS by oxotremorine-M for the M4 mAChR was 11.5 minute but, in the presence 10 µM LY2119620, nearly quadrupled to 44.6 minutes. Like the M2 mAChR, the t1/2 of [3H]NMS at the M4 decreases in a concentration-dependent manner. The t1/2 at 5 µM was 18.7 minutes and at 20 µM increased to 116.7 minutes. Clearly, the dissociation kinetic studies indicated that LY2119620 binds allosterically to the human M2 and M4 mAChRs and was positively cooperative with orthosteric ligand binding. To test an alternative approach to determine whether LY2119620 can be positively cooperative with orthosteric agonist binding, we measured the influence of LY2119620 on the ability of acetylcholine or oxotremorine-M to displace [3H]NMS (Fig. 4). LY2119620 was positively cooperative in its enhancement of orthosteric agonist competition for [3H]NMS binding as shown by a leftward shift in the binding curve for both agonists. LY2119620 was significantly more cooperative at the M4 mAChR compared with its cooperativity at M2 for both acetylcholine and oxotremorine-M. Application of an allosteric ternary complex model (Christopoulos and Kenakin, 2002) using the equation built in to the GraphPad Prism program yielded the logarithm of cooperativity factor (logα) for LY2119620 in the presence of acetylcholine, which was 0.7 and 1.9 for M2 and M4, respectively. LY2119620 increased oxotremorine-M affinity at the M2 receptor with a logα of 1.5, whereas M4 increased by a logα of 2.3.
Binding Analysis of [3H]LY2119620 as a Radioligand for Human Muscarinic Acetylcholine Receptors.
To address whether [3H]LY2119620 bound specifically to any of the five mAChRs, we performed saturation binding studies with membranes stably expressing the human M1–M5 mAChRs. A summary of these results can be found in Table 2. Confirming unlabeled binding studies, [3H]LY2119620 did not bind to the M1, M3, or M5 mAChRs. However, [3H]LY2119620 bound to the human M2 and M4 mAChRs with relativity high affinity. Depending on the orthosteric agonist used, both mAChRs bound with similar affinity (Kd) but very different Bmax values (number of binding sites). In the absence of orthosteric agonists, no specific binding of [3H]LY2119620 was detected further, indicating a robust cooperativity between the orthosteric and allosteric sites (data not shown). The Kd values for [3H]LY2119620 at the human M2 were not significantly different (P = 0.89, n = 3, Student’s t test): 12.9 ± 3.24 nM and 14.4 ± 2.7 nM in the presence of 100 µM acetylcholine or oxotremorine-M, respectively. The Bmax value for [3H]LY2119620 binding to the human M2 was 160 ± 34 fmol/mg protein in the presence of 100 µM acetylcholine. However, in the presence of 100 oxotremorine-M, the number of binding sites increased 17-fold (2700 ± 383 fmol/mg of protein). The M4 mAChR was similar to M2 in that the Kd values for 100 µM acetylcholine and oxotremorine-M were not significantly different (P = 0.57, n = 3, Student’s t test): 2.54 ± 0.39 nM and 2.73 ± 0.08 nM, respectively. The Bmax values for the M4 mAChR were significantly higher in the presence of 100 µM oxotremorine-M (1110 ± 157 fmol/mg of protein) compared with acetylcholine (468 ± 54 fmol/mg of protein) (P > 0.02, n = 3, Student’s t test). The concentrations of acetylcholine and oxotremorine-M were titered from 0–1000 µM, and 100 µM yielded a maximal response for both M2 and M4 mAChRs (data not shown). Displacement studies were conducted with [3H]LY2119620 at the human M2 or M4 mAChRs to elucidate whether this compound bound to a unique allosteric site on these receptors or to previously published sites using key allosteric tool compounds. Figure 5 illustrates that [3H]LY2119620 binding in the presence of 100 µM oxotremorine-M was potently displaced by unlabeled LY2119620 at both the M2 (Ki = 15.3 ± 1.36 nM) and M4 (Ki = 1.03 ± 0.08 nM) mAChRs. A structurally similar compound, LY2033298, displaced [3H]LY2119620 binding to the human M2 mAChR with a Ki of 87.1 ± 20.9 nM and M4 with a Ki of 2.14 ± 0.28 nM (Fig. 5). We also investigated whether LY2119620 bound to either the putative strychnine or staurosporine allosteric sites by displacing [3H]LY2119620 with gallamine and WIN 62,577, respectively. Neither gallamine nor WIN 62,577 displaced [3H]LY2119620 from the human M2 or M4 mAChRs under these assay conditions (Table 3). Other M4 allosteric modulators similar to LY2033298, VU152100, and VU10010 were also able to displace [3H]LY2119620 from the human M4 mAChR but not M2. This result was expected based on previous work that has demonstrated that these compounds are selective for the M4 receptor (Brady et al., 2008; Shirey et al., 2008)
Autoradiographic Localization of the M4 Allosteric Binding Sites in Cynomolgus Monkey Brain Using [3H]LY2119620.
A series of sagittal sections through cynomolgus monkey brains were incubated with approximately 5 nM [3H]LY2119620 and 100 µM acetylcholine to examine the distribution of labeling in different brain structures. Under certain assay conditions, one can favor M4 binding over M2. We took advantage of probe dependence to label mostly M4 receptors because [3H]LY2119620 does not label large numbers of M2 receptors in the presence of 100 µM acetylcholine (Supplemental Fig. 1). Because LY2119620 has relatively low affinity for rodent mAChRs similar to LY2033298, we relied on the distribution of the M4 allosteric binding sites in the cynomolgus monkey to give us insight into the distribution in higher species. Cynomolgus monkeys have nearly identical M2 and M4 mAChRs sequences as humans. In general, [3H]LY2119620 binding in monkeys was broadly distributed in the cortex and devoid in the cerebellum (Fig. 6). Some of the highest levels of binding with [3H]LY2119620 were observed in the caudate-putamen and the superficial (I–III) laminae of the cerebral cortex (Fig. 6A). [3H]LY2119620 binding was almost completely eliminated by the presence of 10 µM LY2033298 (Fig. 6B). Therefore, the amount of radioligand binding remaining in Fig. 6B represents nonspecific binding.
Discussion
The five subtypes of mAChRs are members of the superfamily of G protein–coupled receptors (Caulfield and Birdsall, 1998) and are now known to have allosteric binding sites that provide significant modulation of functional signaling (Christopoulos et al., 1998; Christopoulos, 2002; Christopoulos and Kenakin, 2002). In the present study, we showed that the allosteric modulator LY2119620 exerts its modulator effects through a common site on the M4 mAChR, similar to LY2033298 (Chan et al., 2008), which has been extensively studied using mutagenesis (Leach et al., 2010). In agreement with LY2033298, our findings clearly indicate that LY2119620 is a novel allosteric compound that does not interact with the orthosteric site, similar to the properties of known allosteric compounds such as staurosporine (Lazareno et al., 2000) or strychnine (Ellis et al., 1991). To investigate whether LY2119620 bound to mAChRs in a bitopic manner (a ligand engaging both the orthosteric and allosteric sites at the same time), we used [3H]NMS displacement studies. Unlike the M2 bitopic ligand, McN-A-343 (4-[[[(3-chlorophenyl)amino]carbonyl]oxy]-N,N,N-trimethyl-2-butyn-1-aminium chloride) (Valant et al., 2008), LY2119620 does not displace the classic orthosteric pocket labeled with [3H]NMS. In contrast, the orthosteric antagonist atropine readily displaces [3H]NMS from the M1–M5 mAChRs (Fig. 2). A radiolabeled allosteric modulator of mAChRs was first described for the M2 mAChR using [3H]dimethyl-W84 (N,N′-bis[3-(1,3-dihydro-5-methyl-1,3-dioxo-2H-isoindol-2-yl)propyl]-N,N,N′,N′-tetramethyl-1,6-hexanediamminium dibromide) (Tränkle et al., 1998). Prototype muscarinic allosteric agents alcuronium and gallamine displaced in a concentration-dependent manner the high-affinity site of [3H]dimethyl-W84 binding, These data led Tränkle and colleagues (1998) to conclude that this radioligand bound to the “common” allosteric site on M2. In contrast, [3H]LY2119620 was not displaced by previously described mAChR allosteric compounds such as gallamine or WIN 62,577. The greatest distinction between [3H]dimethyl-W84 and [3H]LY2119620 appears to be in how these radioligands interact with the allosteric site. [3H]Dimethyl-W84 negatively modulated the M2 allosteric site, whereas [3H]LY2119620 demonstrated positive cooperativity with this site.
The dissociation rate of [3H]NMS was significantly reduced for both the M2 and M4 mAChRs in the presence of LY2119620 (Fig. 3). We have taken the ability of LY2119620 to slow the off-rate of [3H]NMS as a measure of its allosteric effect (the binding of the allosteric ligand to the allosteric site that alters the affinity of the muscarinic orthosteric agonist to bind to the orthosteric binding pocket on the receptor). The cooperative effect was dependent on the muscarinic ligand it interacts with, which can be positive, negative, or neutral. In the [3H]NMS competition binding, the interaction of LY2119620 with either acetylcholine or oxotremorine-M was positively cooperative for both M2 and M4 mAChRs (Fig. 4). Interestingly, the affinity (Kd) of [3H]LY2119620 for the mAChR was similar, whereas the Bmax varied considerably whether acetylcholine or oxotremorine-M was used. This finding of probe dependence (the interaction between allosteric and orthosteric sites changing, depending on the orthosteric ligand used) was evident in these studies because we used saturating concentrations of acetylcholine or oxotremorine-M. Probe dependence was also shown for the structurally similar compound, LY2033298, at the mouse M4 mAChR (Suratman et al., 2011). In that study, it was speculated that the probe dependence was due to different cooperativities between modulator and orthosteric ligands because LY2033298 had similar affinities for both human and mouse M4 allosteric sites. The positive cooperativity between acetylcholine and LY2033298 was most evident at the human M4 and was lower at the human M2 and essentially neutral at the other mAChR subtypes (Chan et al., 2008). We see similar cooperativity differences between M2 and M4 with acetylcholine or oxotremorine-M using [3H]NMS binding (Fig. 4). Therefore, one could speculate that the difference in the Bmax values between orthosteric agonists with [3H]LY2119620 was due to a similar mechanism. That is, increased modulator binding was directly proportionate to the number of active state receptors since it is well known that GPCRs exist in two states, active (RG) and inactive (R). Although the exact mechanism remains unclear, a common interpretation is that this somehow reflects the coupling of the GPCR to the G protein(s) to promote the active state (Christopoulos and El-Fakahany, 1999). Not yet tested was whether the functional positive allosteric modulation by either LY2033298 or LY2119620 can be driven by increasing the cooperativity between orthosteric ligand and G protein binding, thus increasing the number of G protein–bound mAChRs, thereby increasing the functional output of the signaling being measured. This interaction could be inferred because functional [35S]GTPγS binding was positively modulated by LY2119620 (Croy et al., 2014). Furthermore, emerging crystal structure studies with allosteric compounds might shed some light on probe dependence. Recently, the M2 mAChR was crystalized in the active state with iperoxo docked in the orthosteric binding pocket in the presence of LY2119620 (Kruse et al., 2013). The M2 crystal structure revealed that LY2119620 induces additional, albeit subtle, structural changes compared with those seen with just the orthosteric agonist.
Like its predecessor LY2033298, LY2119620 is also subject to species variability. In the initial characterization of LY2033298, it was noted that this compound had reduced in vitro potency as a modulator in the rat compared with the human M4 mAChR (Chan et al., 2008) and later for the mouse (Suratman et al., 2011). However, the affinity for LY2033298 across species has been shown to be very similar (Leach et al., 2010; Suratman et al., 2011; Valant et al., 2012) in several assay formats. In contrast, any attempt to get [3H]LY2119620 to bind to either recombinantly expressed rodent muscarinic or native tissue in the presence of any orthosteric agonist was not successful. However, it should be pointed out that the concentrations of radioligand used were limited as a result of reagent costs as well as increasing nonspecific binding at higher concentrations. These direct-labeling experiments with a radiolabeled allosteric probe contradict previous reports using various functional assays that these allosteric modulators have similar affinity across species (Leach et al., 2010; Suratman et al., 2011; Valant et al., 2012). In addition, we used [3H]LY2119620 to probe the distribution of these allosteric sites in the brain. We used nonhuman primates as the gene sequences between them and humans are nearly identical. We found the distribution of [3H]LY2119620 to be similar to the distribution of M2 and M4 mAChRs using [3H]AF-DX 384 (N-[2-[2-[(dipropylamino)methyl]-1-piperidinyl]ethyl]-5,6-dihydro-6-oxo-11H-pyrido[2,3-b][1,4]benzodiazepine-11-carboxamide), a selective M2 and M4 antagonist of the muscarinic acetylcholine receptors (Quirion et al., 1993). [3H]AF-DX 384 binds preferentially to the striatum, cortex, thalamus, and cerebellum.
In summary, we have identified LY2119620 as an allosteric modulator of the human M2 and M4 mAChRs. Our data support the hypothesis that allosteric selectivity between M2 and M4 mAChR subtypes was the result of differences in cooperativity with the endogenous agonist, exemplifying probe dependence. Cooperativity governing selectivity, rather than affinity for a unique allosteric site, has been previously reported with thiochrome (Lazareno et al., 2004) and LY2033298 (Leach et al., 2010). This supports our hypothesis that the allosteric modulators can recruit G protein coupling because the number of high-affinity binding sites labeled with [3H]acetylcholine significantly increases in the presence of thiochrome, and LY2119620 increased the number of high-affinity binding sites using [3H]oxotremorine-M (Croy et al., 2014). Likewise, we report in this study that the affinity of [3H]LY2119620 was similar in the orthosteric agonists, but the total number of binding sites was significantly different between agonists. In conclusion, not all orthosteric agonists recruit high-affinity binding sites to the same extent, and allosteric binding governs G protein recruitment cooperatively with orthosteric agonist binding.
Authorship Contributions
Conducted experiments: Schober, Croy, Xiao.
Performed data analysis: Schober, Croy, Christopoulos, Felder.
Wrote or contributed to the writing of the manuscript: Schober, Croy, Felder, Christopoulos.
Footnotes
- Received January 10, 2013.
- Accepted May 7, 2014.
↵This article has supplemental material available at molpharm.aspetjournals.org.
Abbreviations
- [3H]AF-DX 384
- N-[2-[2-[(dipropylamino)methyl]-1-piperidinyl]ethyl]-5,6-dihydro-6-oxo-11H-pyrido[2,3-b][1,4]benzodiazepine-11-carboxamide
- CNS
- central nervous system
- [3H]dimethyl-W84
- N,N′-bis[3-(1,3-dihydro-5-methyl-1,3-dioxo-2H-isoindol-2-yl)propyl]-N,N,N′,N′-tetramethyl-1,6-hexanediamminium dibromide
- GPCR
- G protein–coupled receptor
- [35S]GTPγS
- 5′-O-(3-[35S]thio)triphosphate
- McN-A-343
- 4-[[[(3-chlorophenyl)amino]carbonyl]oxy]-N,N,N-trimethyl-2-butyn-1-aminium chloride
- [3H]NMS
- [3H]N-methylscopolamine
- LY2033298
- 3-amino-5-chloro-6-methoxy-4-methyl-thieno(2,3-b)pyridine-2-carboxylic acid cyclopropylamide
- LY2119620
- 3-amino-5-chloro-N-cyclopropyl-4-methyl-6-[2-(4-methylpiperazin-1-yl)-2-oxoethoxy] thieno[2,3-b]pyridine-2-carboxamide
- PBS
- phosphate-buffered saline
- VU10010
- 3-amino-N-[(4-chlorophenyl)methyl]-4,6-dimethylthieno[2,3-b]pyridine-2-carboxamide
- VU152100
- 3-amino-N-(4-methoxybenzyl)-4,6-dimethylthieno[2,3-b]pyridine carboxamide
- WIN 62,577
- 17-β-hydroxy-17-α-ethynyl-δ-4-androstano[3,2-b]pyrimido[1,2-a]benzimidazole
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics