|Home | About | Journals | Submit | Contact Us | Français|
We report the synthesis and evaluation of a series of heterobiaryl amides as positive allosteric modulators of mGluR4. Compounds 9b and 9c showed submicromolar potency at both human and rat mGluR4. In addition, both 9b and 9c were shown to be centrally penetrant in rats using nontoxic vehicles, a major advance for the mGluR4 field.
The metabotropic glutamate receptors (mGluRsa) are members of the GPCR family C, characterized by a large extracellular amino-terminal binding domain (agonist) along with a 7-transmembrane spanning (7TM) domain which is the binding site for most known mGluR allosteric modulators.1-3 The eight cloned mGluRs have been assigned to three groups (Groups I,II,III) based on their structural similarity, ligand specificity, and preferred coupling mechanisms.4 The Group I subfamily is composed of mGluR1 and mGluR5; Group II receptors include mGluR2 and mGluR3; and the Group III receptors are represented by mGluR4, mGluR6, mGluR7 and mGluR8. Among the mGluRs, the Group III receptors have thus far received less attention in terms of their therapeutic potential due to the paucity of selective ligands. However, recently there have been numerous reports detailing the potential benefits of mGluR4 activation in several disease models, most notably rodent models of Parkinson's disease.5,6 It has been shown that activation of mGluR4 decreases GABAergic transmission at the inhibitory striato-pallidal synapse with the basal ganglia, a mechanism that is expected to provide palliative benefit for the treatment of Parkinson's disease.7 In addition, there have been recent reports detailing the neuroprotective effects of an mGluR4 PAM in cultured neurons and in vivo.8,9
The most well characterized mGluR4 PAM for many years has been the compound PHCCC, 1, a partially selective mGluR4 potentiator.8,10,11 More recently, our laboratory and others have expanded the list of novel probes for positive allosteric modulation of mGluR4 (Figure 1).12-16 Unfortunately, all of the disclosed mGluR4 PAMs are deficient in their penetration into the CNS and their effects have only been demonstrated via intracerebroventricular (icv) injection or with the use of toxic vehicles such as a 50% DMSO solution. Herein, we report a class of centrally penetrant mGluR4 PAMs which can be administered in a nontoxic vehicle.
A high-throughput screening (HTS) campaign was initiated at Vanderbilt to identify novel mGluR4 PAMs.12-14 In addition to the ligands shown in Figure 1, there were a number of small aryl amide compounds identified as having mGluR4 PAM activity (Figure 2). These were attractive hits due to their favorable calculated properties (MW <300, cLogP <3.50, total Polar Surface Area (tPSA) <40, Ligand Efficiency (LE)17 >0.30). From these lead compounds we initiated an optimization program in order to further profile this chemical series as novel mGluR4 PAMs.
Initial effort was directed at the left-side heteroaryl amide portion for which the synthesis is outlined in Scheme 1. A small library utilizing the furyl amide revealed that the 3,4-dichloroaniline displayed good potency and efficacy at the hmGluR4 receptor. Next, a more detailed SAR evaluation of this structural class was undertaken.
Keeping the right-hand aniline constant as the 3,4-dichlorophenyl, we first evaluated different aryl, cycloalkyl, and heteroaryl amide replacements (Table 1). One issue that has been observed for many allosteric ligands is the noted intractable SAR14,19,20; this property was also observed for the amide modifications. While the furyl amide, 8a, was active in the micromolar range with good efficacy (>5 μM, 135% GluMax), and bromo substitution was tolerated (8b, 4.1 μM, 62% GluMax), further modifications led to loss of activity (see 8c-d). An additional survey of 5-membered heterocycles was undertaken but all compounds were inactive as mGluR4 PAMs (8e-g). Substituting a cyclohexyl, 8h, or phenyl, 8i, amide for the furyl also led to inactive compounds. However, upon introduction of the 2-pyridyl group, good potency and GluMax values were restored (8j, 1.4 μM; 80% GluMax). The 2-pyrazine compound, 8k, lost activity, while the 4-pyrimidine, 8l, retained potency, again highlighting this feature of intractable SAR around allosteric modulators. Based on these results, the 2-pyridyl amide was chosen to further analyze for SAR.21
Starting from the 3,4-dichloroaniline derivative, 8j, a number of halogenated and/or oxygenated compounds were synthesized and evaluated (Table 2). Compounds 9a-9p were synthesized and tested and a clear improvement on the HTS lead was identified. Compounds 9a-9c were the best compounds evaluated in this series and presently are some of the best compounds for mGluR4 PAM activity that have been disclosed to date. These compounds have excellent potency at hmGluR4 (240 nM - 780 nM) and rmGluR4 (80 nM - 370 nM) receptors and exhibit excellent efficacy (hGluMax%, 182 - 235). The SAR shows that oxygen in the 3-position is favored; however a 3-Cl substituent also imparts activity. As has been seen with other allosteric modulators, the SAR is very narrow with very minor modifications leading to compounds with reduced or no activity. A number of modifications highlight this fact - see compounds 9c (340 nM) vs. 9d (>5 μM), where a simple modification of a difluoromethoxy to a trifluoromethoxy imparts a >10-fold loss of activity. Substituting the phenyl ring of the aniline with either a pyridine (9n,o) or pyrimidine (9p) also led to compounds with much reduced activity (9b (240 nM) vs. 9n (>5 μM)) or inactive compounds (9o,p). A number of compounds were synthesized to examine the 3-methoxy substitution by making longer alkyl chain substitutions or by cyclization (data not shown). However, each of these modifications led to either inactive or weakly active compounds.
A number of the more active compounds were next evaluated for their ability to shift the glutamate response curve to the left. The fold shift for both the human and rat receptors are shown in Table 2. Many of the compounds that were analyzed showed a robust shift of the glutamate response (Figure 3, 9b). The SAR for this series of compounds was also mirrored in the fold shift data. As was observed in the potency data (Table 2), 9a-c were the most potent compounds and they also produced the largest fold shift in both the human and rat cell lines. In addition to these compounds, 9f and 9j also produced robust fold shifts, giving this series several compounds with fold shifts >15. In addition, the selectivity of a few compounds (9a-c) was determined among the various mGluR subtypes. Although these compounds are very active at both the human and rat mGluR4 receptor, they showed little activity against the other mGluRs, with the exceptions being weak PAM activity at mGluR5 and mGluR8. For example, compounds 9b and 9c showed weak PAM activity at both mGluR5 and 8 (Supplementary Material, Table 1); however, there is more than a 20-fold separation in potencies compared to mGluR4.
Based on the potency and efficacy of compounds 9a-e, and the very favorable calculated properties, we next looked at metabolic stability and protein binding (PPB) (Table 3). These compounds were not stable in human or rat liver microsomes (HLM, RLM) with three compounds (9a-c) having less than 10% of the parent remaining after the incubation period and two compounds (9d-e) having less than 25% remaining. However, it is worth noting that 9d and 9e are significantly more stable; neither compound possesses the metabolically unstable methoxy group. All of these compounds (except 9d) possess favorable % free fraction in human and rat protein binding experiments. Due to these findings and the potency of these compounds, 9b and 9c were advanced further for in vivo DMPK analysis.
The hydrochloride salts of these compounds (9b,9c) were synthesized and dosed intraperitoneally (10mg/kg) as an aqueous microsuspension containing 10% tween 80 and the amount of compound present in brain and plasma was determined at 0.5, 1 and 8 h after administration (Table 4). Consistent with the poor in vitro microsomal stability, these compounds showed high clearance and low plasma exposure in rats. However, the brain levels for compounds 9b and 9c, were significant when compared to the total amount of compound, indicative of good brain penetration for these amides.
In summary, we report a series of small molecule mGluR4 positive allosteric modulators. These compounds represent a series of 2-pyridyl amide compounds that possess excellent calculated properties, making them ideal candidates for tool compounds. In addition, a number of compounds have excellent in vitro potency and efficacy at both the human and rat mGluR4 receptor, and many possess the ability to robustly shift the glutamate response to the left (>15). Selected compounds were further profiled for selectivity, in vitro PK and ultimately in vivo PK. Two compounds, 9b and 9c, although possessing less than ideal in vitro PK parameters, show sufficient brain penetration to enable further evaluation in anti-Parkinsonian in vivo rodent models which will be reported in due course.
The authors would like to thank the assistance of members of the Vanderbilt HTS facility, Miranda Nolan for the in vitro PK, as well as Matt Mulder, Chris Denicola and Sichen Chang for the purification of compounds utilizing the mass-directed HPLC system. This work was supported by the National Institute of Mental Health, the Michael J. Fox Foundation, the Vanderbilt Department of Pharmacology and the Vanderbilt Institute of Chemical Biology.
aAbbreviations: mGluR: metabotropic glutamate receptor; PAM: positive allosteric modulator; HTS: high-throughput screening; tPSA: total polar surface area; LE: ligand efficiency; HLM: human liver microsomes; RLM: rat liver microsomes; PPB: protein binding.
Supporting Information Available: Experimental procedures, spectroscopic data, and NMR data for select compounds, along with biological procedures. This material is available free of charge via the Internet at http://pubs.acs.org.