Discovery of Novel Allosteric Modulators of the Muscarinic Receptor M5
Screening Center Name & PI: Vanderbilt Screening Center for GPCRs, Ion Channels and Transporters, C. David Weaver ..more
Depositor Specified Assays
Assigned Assay Grant Number: MH077606-1
Screening Center Name & PI: Vanderbilt Screening Center for GPCRs, Ion Channels and Transporters, C. David Weaver
Chemistry Center Name & PI: Vanderbilt Specialized Chemistry Center for Accelerated Probe Development, Craig W. Lindsley
Assay Submitter & Institution: P. Jeffrey Conn, Vanderbilt University
Muscarinic acetylcholine receptors are family A GPCRs comprised of five distinct mammalian subtypes (mAChR1-5 or M1-M5), which are expressed differentially throughout the body and play an important role in a variety of physiological processes. Among the mAChRs, M1 and M4 have been historically considered attractive targets for small molecule treatments of numerous CNS disorders such as Alzheimer's disease and schizophrenia due to their respective localization and involvement in regulation of certain aspects of learning, memory, sleep, motor control, reward, and pain, among others. However, discovery of subtype-selective small molecules has proven highly difficult due to the conservation of the orthosteric binding-site across the mAChRs. This has contributed to the failure of muscarinic agonists in clinical trials and has also hampered pharmacological investigation into the role(s) of each mAChR in basic neurobiology.
Among the mAChRs, M5 has remained perhaps the most challenging to investigate pharmacologically due in part to its extremely low expression level and a complete lack of M5-selective ligands. Interestingly, studies using M5-KO mice suggest that M5 is the sole mediator of acetylcholine-induced cerebrovasodilation, which has led to the hypothesis that an M5 activator would have therapeutic efficacy in treatment of cerebrovascular dementias and ischemic stroke. Furthermore, M5-KO mice show dramatically reduced reward responses to drugs of abuse, consistent with its putative localization on midbrain dopaminergic neurons of the nigrostriatal and mesolimbic pathways. This suggests that M5 antagonism or negative modulation may have utility in treatment of illicit drug addiction and withdrawal. Despite these and other related findings from M5-KO mice, there remains a strong need for small molecule tools to probe M5 function and test M5-related hypotheses in order to advance the state of the mAChR research field and provide critical proof-of-concept studies for drug discovery aims.
The M5 probe molecule, SID 85285464 , represents a unique, and never before seen, pharmacological profile in that it was a PAM of all the Gq-coupled mAChRs (M1 EC50 =6.1 uM, M3 EC50 =6.4 uM and M5 EC50 = 4.1 uM), but devoid of activity at the Gi/o-coupled M2 and M4 (Marlo, 2009). This led us to predict it would be possible to dial-in or dial-out different mAChR sub-types and potentially develop an M1 selective PAM, an M3 selective PAM and/or an M5 selective PAM from this non-selective lead through chemical optimization (Bridges, 2009).
For the lead optimization of SID 85285464 , we employed a 2-dimensional library approach, evaluating 8 diverse isatins in combination with 12 functionalized benzyl halides to afford 92/96 analogs of SID 85285464 (Bridges, 2009). Yields were generally good providing the desired products in yields ranging from 20-95%. In order to evaluate the analogs expeditiously, we screened them in a single point format at a 30 uM final concentration in Ca2+ mobilization assays using M5 and M1 cells receiving a fixed submaximal concentration (~EC10) of ACh (AID 2198, AID 2206). This method efficiently triaged analogs displaying high M1 vs. M5 or M5 vs. M1 preference. Interestingly, some analogs displayed robust potentiation effects at M5 (i.e. elevation of ACh ~EC10 to >50-60% of maximum ACh response) with absent or weak potentiation at M1, thus exhibiting strong preference for M5 versus M1 activity.
In terms of maximal potentiation efficacy, SAR from the initial 30 uM screen of the entire analog library suggested that 5-OCF3 substitution of the isatin core(R1) was generally favored for increased M5 versus M1 activity, while halogen substitutions at the 4- position conferred a more dual M1/M5 activity (Bridges, 2009). Indeed, most of the 5-OCF3 substituted compounds chosen from the initial analog screen possessed 1-5 uM potencies at M5 and >30 uM potencies at M1 (AID 2198, AID 2206). Numerous benzylic substitutions (R2) were tolerated for M1 and M5 activity, depending on the isatin core. Methoxybenzyls were generally favored for M5 versus M1 activity whereas trifluoromethylbenzyls were generally favored for dual M1/M5 potentiation activity (AID 2198, AID 2206). Most other simple congeners, including those with various methyl substitutions at R1 or R2 displayed weak or negligible potentiation at either receptor. In light of the high M5 versus M1 potentiation preference displayed by SID 85285486 (1.16 uM M5 EC50 and >30 uM M1 EC50) in these assays, the compound was further profiled across the remaining subtypes in similar Ca2+ assays using M2, M3, and M4 cells (AID 2204). SID 85285486 had no effect on M2 and M4, and afforded only modest activation of M1 and M3 at 30 uM (~30%); thus, SID 85285486 provides >30x selectivity for M5 versus the other four subtypes, thus representing the first highly M5-preferring muscarinic ligand ever reported (Bridges, 2009).
Subsequent experiments (AID 2194, AID 2192, AID 2186) were performed to determine if SID 85285486 was a true PAM. SID 85285486 had no effect on M5 cells alone, but caused a dose-dependent increase in M5 receptor activation in the presence of a submaximal (EC20) concentration of ACh. At a fixed 30 uM concentration, SID 85285486 elicited a 14-fold leftward shift of the ACh concentration-response-curve (CRC) (AID 2186). In [3H]-NMS binding experiments, SID 85285486 did not displace the orthosteric radio-ligand whereas atropine, an orthosteric antagonist, afforded complete displacement (AID 2194, AID 2192). In order to address the mechanism of M5 PAM activity, we measured the effect of fixed 30 uM SID 85285486 on ACh competition with [3H]-NMS binding, and found a 10-fold increase in ACh affinity for M5 - a value close to the leftward fold-shift (14-fold). Thus, the mechanism of M5 potentiation by SID 85285486 is due in part to the enhancement of ACh binding. Given these data, SID 85285486 is a true PAM of M5, devoid of intrinsic agonist activity (Bridges, 2009).
We then evaluated pharmacokinetics and brain penetration for the M5 PAM to evaluate its ability to serve as an in vivo probe. Unfortunately, SID 85285486 is characterized by poor systemic absorption after intraperitoneal administration with maximum concentration in plasma (161.7 ng /mL) being achieved within 1 hour. However, it is slowly eliminated from systemic circulation and has elimination half life of 4.5 hours. Although quickly taken up in the brain, it exhibits poor brain penetration with AUCbrain/AUCplasma value of 0.25 (Bridges, 2009). Future lead optimization will focus on improving brain penetration. At this point, the Lead Profiling Screen (68 GPCRs, ion channels and transporters) from MDS Pharma was performed on SID 85285486 to determine a broader ancillary pharmacology profile for this MLPCN probe. In addition to selectivity versus the mGluR family, SID 85285486 possessed reasonably clean ancillary pharmacology in a 68 Target MDS GPCR, Ion Channel and Transporter Lead Profiling Panel, displaying no significant activity (no inhibition >50% at 10 uM) for 59 of the 68 targets. SID 85285486 did have activity at 9 targets: A3A (65%@10 uM), A2A (99%@10 uM), NET (66%@10 uM), CB1 (84%@10 uM), D4 (61%@10 uM), H1 (62%@10 uM), H2 (55%@10 uM), MOP (70%@10 uM) and hERG (60%@10 uM). Of these, only two possessed Kis below 10 uM: A2A (760 nM) and CB1 (2.2 uM) (Bridges, 2009).
Thus, SID 85285486 is a first-in-class M5 PAM, the first M5-selective ligand (>30-fold versus M1-M4) with generally clean ancillary pharmacology. CNS penetration is poor however, restricting SID 85285486 for use as an in vitro probe (Bridges, 2009).
The second generation of M5 PAM probes resulted in SID 87352032. This probe possesses improved muscarinic selectivity over our previous M5 PAM (SID 85285486). The compound can be used for in vitro molecular pharmacology and electrophysiology experiments to study the role of selective M5 receptor activation. This probe possesses unprecedented selectivity versus M1, M2, M3 and M4 (an M5 EC50 of 1.9 uM with a 75% ACh Max and no elevation of an ACh EC20 at M1-4 at concentrations as high as 30 #M.) SID 8735203 is not readily CNS penetrant with poor brain exposure (AUCBrain /AUCPlasma ca. 0.25), and so would need to be administered i.c.v. to study the role of central M5 activation in vivo. SID 8735203 also displays reasonable solubility in acceptable vehicles (>5 mg/mL) in 20% cyclodextrin and >100 uM in DMSO.