|Home | About | Journals | Submit | Contact Us | Français|
A high-throughput cell-based screen identified a series of 6-substituted-4-anilinoquinazolines as non-competitive antagonists of metabotropic glutamate receptor 5 (mGlu5). This communication describes the SAR of this series and the profile of selected compounds in selectivity and radioligand binding assays.
Glutamate is the major excitatory transmitter in the mammalian CNS, exerting its effects through both ionotropic and metabotropic glutamate receptors. The metabotropic glutamate receptors (mGlus) belong to family C of the G-protein-coupled receptors (GPCRs). These receptors are characterized by a seven transmembrane (7TM) α-helical domain that is connected via a cysteine-rich region to a large bi-lobed extracellular amino-terminal domain. The eight mGlus discovered to date have been further divided according to their structure, preferred signal transduction mechanisms, and pharmacology (Group I: mGlu1 and mGlu5; Group II: mGlu2 and mGlu3; Group III: mGlu4, mGlu6, mGlu7, and mGlu8).1
Whereas orthosteric ligands of mGlus bind in the amino-terminal domain of the receptor, known allosteric binding sites are located in the 7TM domain. Orthosteric ligands often suffer from poor selectivity among the mGlus due to a highly conserved binding site. The discovery of non-competitive antagonists, also known as negative allosteric modulators (NAMs), has offered a potential solution to such selectivity issues.2 The mGlu5 NAMs 2-methyl-6-(phenylethynyl) pyridine (MPEP)3 and 3-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine (MTEP)4 (Figure 1) have demonstrated efficacy in numerous preclinical models of disease, including pain5, anxiety6, gastroesophageal reflux disease (GERD)7, and fragile X syndrome8. In addition, there have been recent positive disclosures from phase II clinical studies with two small molecule mGlu5 NAMs, ADX10059 in GERD9 and acute migraine10 and fenobam (Figure 1) in fragile X syndrome11. With such a large body of compelling evidence, the search for new and improved mGlu5 NAMs remains an attractive and active area for drug discovery research.12
We have recently reported our initial results from an effort to identify mGlu5 antagonists from multiple diverse chemotypes.13 A functional cell-based high-throughput screen of a collection of 160,000 compounds identified 624 mGlu5 antagonists. The confirmation of hits using full concentration response curves left 345 verified non-competitive antagonists of the target. Among that set of confirmed hits were a few 6-bromo-4-anilinoquinazolines. 3-Chloroaniline analog 1 (Figure 2) represented the most potent compound in our functional assay, which measures the ability of the compound to block the mobilization of calcium by an EC80 concentration of glutamate in HEK293A cells expressing rat mGlu5.14 A binding affinity determination measuring the ability of the compound to compete with the equilibrium of [3H]3-methoxy-5-(pyridin-2-ylethynyl)pyridine15, a close structural analog of MPEP, confirmed the interaction of 1 with the known allosteric binding site.16
The quinazoline scaffold as a chemotype for mGlu5 antagonists has been disclosed by Grüenenthal GmbH in the patent literature in the form of 6-aryl-4-aminoquinazolines.17 Quinazolines were also used by Yamanouchi Pharmaceutical Company18, Eli Lilly19, and Pfizer20 to design mGlu1 antagonists. Nonetheless, an investigation of the SAR of 6-substituted-4-anilinoquinazolines as non-competitive antagonists of mGlu5 has yet to be disclosed. The development of such SAR is the subject of this communication.
Quinazoline derivatives of interest were readily accessible through SNAr reaction of the appropriate nucleophiles with commercially available 6-substituted-4-chloroquinazolines using microwave-assisted organic synthesis (MAOS)21 (Scheme 1). Such chemistry was amenable to our preferred iterative library synthesis approach, which in combination with our custom mass-directed HPLC purification system allows for rapid evaluation of new screening hits.22 Prior to biological testing, all compounds were analyzed by LCMS and determined to be ≥ 95% pure, and selected compounds were further characterized by proton NMR.23 Initially, we decided to conduct a small scan with commercially available 3-substituted anilines while maintaining the 6-bromoquinazoline functionality (Table 1). Substitution at the 3-position of the aniline ring appeared to improve potency as unsubstituted aniline 2 was only weakly active. 3-Fluoroaniline 3 exhibited improved potency relative to 2, while 3-bromoaniline 4 and 3-methylaniline 5 were comparable in activity to the hit compound 1. The sensitivity of this position to subtle modifications was evident as 3-trifluoromethylaniline 6 was 7-fold less potent than 5. 3-Methoxyaniline 7 was only weakly active, similar to 2.
We also decided to examine various substituents at the 6-position of the quinazoline ring while maintaining the 3-chloroaniline substituent (Table 2). Other halogens at this position (8 and 9) were similar in potency to the hit compound (1). 6-Nitroquinazoline 10 had comparable activity to the 6-halogen compounds. 6-Methoxy (11) and 6-cyano (12) quinazolines were less potent. Other substituents, including larger aryl and heteroaryl groups (data not shown) were not tolerated and resulted in a complete loss of activity.
Another area of interest was the quinazoline core of the template. As such, we prepared a few modified cores (Table 3). 6-Chloroquinoline analog 14 was essentially inactive in our assay, which was a dramatic change from the potent antagonist activity observed with comparator 6-chloroquinazoline 13. Modification of the template to afford 7-bromoisoquinoline 15 reduced the activity by approximately 25-fold relative to 6-bromoquinazoline comparator 4. Such results further illustrate how small structural modifications within this chemotype can profoundly impact the observed mGlu5 activity.
We also explored the concept of extending the linker between the 4-amine substituent and the aryl ring (Table 4). Both benzyl amines (17-20) and phenethylamines (21-24) were investigated. Only unsubstituted phenethylamine analog 21 exhibited any antagonist activity, albeit weak. Chloro-substituted analogs failed to demonstrate any improved activity, which was a significant contrast to the enhanced potency of 3-chloroaniline analog 9 relative to unsubstituted analog 16.
An additional area of interest was a survey of potential alternatives to the secondary amine linker. Such work was accomplished in the context of the 6-chloro and 6-bromoquinazolines (Table 5). The N-methyl (25 and 28) and thiol (27 and 30) analogs proved inactive in our assay. Only the ether analogs (26 and 29) demonstrated weak antagonist activity. The consequences of the secondary amine (13 and 4) to ether (26 and 29) modifications were severe, reducing potency by more than 50-fold in each case.
Once we concluded that secondary anilines were likely optimal in this series, we sought to further evaluate the SAR around that portion of the scaffold. One of the potential issues with this series was the relatively high lipophilicity of the initial screening hit 1 (cLogP = 5.49).24 Having achieved a potent hit with 6-fluoroquinazoline 8 (cLogP = 4.77), we decided to examine additional substituted anilines in the context of this core (Table 6). As was the case with the 6-bromoquinazolines, unsubstitued aniline 31 and 3-trifluoromethylaniline 34 were weak antagonists. 3-Bromo aniline 32 was a potent antagonist, while 3-methyl aniline 33 was only moderately potent. The moderate potency of 33 represented a departure from the SAR observed with the 6-bromoquinazolines, where 3-methylaniline 5, 3-chloroaniline 1, and 3-bromoaniline 4 were essentially equipotent. The 3-cyanoaniline 35 was moderately potent, similar to 33. Interestingly, 3-ethylaniline 36 was a weak potentiator, or positive allosteric modulator (EC50 > 10 μM; % Glu max = 52 ± 5 in the presence of an EC20 concentration of glutamate). Such a switch in mGlu5 pharmacology has been noted and reported before in our laboratory in the context of other chemotypes.13,16 3-Methoxyaniline 37 was a weak antagonist. We also prepared a few analogs (38-41) in order to evaluate the effect of fluoro substitution in the context of the 3-chloroaniline. In every case, fluoro substitution proved detrimental to potency. 2-Fluoro-3-chloroaniline 39 demonstrated moderate potency, while other analogs were only weak antagonists.
Having identified some molecules with good potency in our cell-based functional assay, we decided to further profile these analogs with regard to binding affinity and selectivity (Table 7). Data for MPEP is shown as a comparator. Binding affinity determinations with [3H]3-methoxy-5-(pyridin-2-ylethynyl)pyridine confirmed the interaction of 13 and 32 with the known MPEP allosteric binding site, as was the case with hit compound 1. The same three compounds were also examined for their selectivity versus additional mGlus and were determined to be inactive against mGlu2-4 and mGlu7-8. On the other hand, each compound demonstrated moderate antagonist activity of mGlu1 in a calcium mobilization assay.25 Recent publications indicate possible links between mGlu1 and anxiety6a,26 as well as pain27 and fragile X28. Such an overlap with regard to therapeutic applications makes the concept of dual antagonism of mGlu1 and mGlu5 potentially interesting.
It is worth commenting on the fact that 4-anilinoquinazolines are well established as an effective template for the design of inhibitors of epidermal growth factor receptor (EGFR) and other members of the ErbB kinase family.29 In fact, dichloro analog 9 has been reported to moderately inhibit EGFR autophosphorylation (IC50 = 2.7 μM) in A431 cells.30 While this level of potency is markedly less than that observed toward mGlu5 in our functional cell based assay, monitoring potential off-target activity against such kinases will be necessary in the further development of this series as mGlu5 non-competitive antagonists.
In summary, we have identified a series of non-competitive antagonists of mGlu5 within the 6-substituted-4-anilinoquinazoline chemotype. Although the series was chemically unrelated to MPEP, selected potent compounds were confirmed to inhibit binding of a radioligand at the MPEP allosteric binding site. While the SAR in this series was somewhat shallow, several compounds demonstrated moderate to good potency. Of further interest was the dual activity of this series with respect to mGlu1.
The authors thank NIDA (RO1 DA023947-01) and Seaside Therapeutics (VUMC33842) for their support of our programs in the development of non-competitive antagonist of mGlu5. Matt Mulder, Chris Denicola, and Sichen Chang are also thanked for the purification of compounds using the mass-directed HPLC system.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.