Bioorganic & Medicinal Chemistry Letters 20 (2010) 175–179 Contents lists available at ScienceDirect Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl Scaffold hopping from pyridones to imidazo[1,2-a]pyridines. New positive allosteric modulators of metabotropic glutamate 2 receptor Gary Tresadern a,*, Jose María Cid b, Gregor J. Macdonald c, Juan Antonio Vega b, Ana Isabel de Lucas b, Aránzazu García b, Encarnación Matesanz b, María Lourdes Linares b, Daniel Oehlrich c, Hilde Lavreysen d, Ilse Biesmans d, Andrés A. Trabanco b,* a Research Informatics, Johnson & Johnson, Pharmaceutical Research & Development, Janssen-Cilag S.A., Calle Jarama 75, Poligono Industrial, Toledo 45007, Spain Medicinal Chemistry, Johnson & Johnson, Pharmaceutical Research & Development, Janssen-Cilag S.A., Calle Jarama 75, Poligono Industrial, Toledo 45007, Spain Medicinal Chemistry, Johnson & Johnson, Pharmaceutical Research & Development, Janssen Pharmaceutica N.V., Turnhoutsweg 30 B-2340, Beerse, Belgium d Neuroscience, Johnson & Johnson, Pharmaceutical Research & Development, Janssen Pharmaceutica N.V., Turnhoutsweg 30 B-2340, Beerse, Belgium b c a r t i c l e i n f o Article history: Received 7 October 2009 Revised 3 November 2009 Accepted 3 November 2009 Available online 10 November 2009 a b s t r a c t Imidazo[1,2-a]pyridines were identified via their shape and electrostatic similarity as novel positive allosteric modulators of the metabotropic glutamate 2 receptor. The subsequent synthesis and SAR are described. Potent, selective and metabolically stable compounds were found representing a promising avenue for current further studies. Ó 2009 Elsevier Ltd. All rights reserved. The authors dedicate this Letter to the memory of Dr. Hassan Imogai, Addex Pharmaceuticals SA Keywords: mGluR2 Metabotropic Chemotype hopping Glutamate Allosteric PAM The metabotropic glutamate type 2 receptor (mGluR2) is a Gprotein coupled receptor (GPCR) expressed on presynaptic nerve terminals where it negatively modulates glutamate and GABA release.1 Mixed mGluR2/mGluR3 agonists such as LY354740 (1) have shown activity in a range of preclinical animal models of anxiety and schizophrenia (see Fig. 1).2,3 Early clinical work with LY354740 demonstrated activity in a CO2 inhalation study suggesting the usefulness for the treatment of anxiety related disorders.4 Subsequently, a related prodrug LY2140023 (2) demonstrated improvements in positive and negative symptoms in patients suffering from schizophrenia.5 These molecules exhibit combined mGluR2/mGluR3 activity although there is evidence from knockout studies that preclinical anti-psychotic effects may be mediated via the mGluR2 receptor.6 An alternative avenue for modulating GPCRs is to act via allosteric mechanisms, binding at a different site from the orthosteric agonist.7 Positive allosteric modulators (PAMs) of mGluR2 have been claimed and reported, Figure 2.8 PAMs may offer benefits over orthosteric agonists. The orthosteric site is highly conserved across the mGluR family therefore targeting alternative sites assists the identification of selective compounds.9 Allosteric ligands will in general be more brain penetrant as they would not be amino acid analogues. The allosteric modulator exerts its effect when glutamate is present allowing the receptor to respond to physiological CO2H H NH2 O O S H H HO2C * Corresponding authors. Tel.: +34 925 245792 (A.T.); +34 925 245782; fax: +34 925 245771 (G.T.). E-mail addresses: gtresade@its.jnj.com (G. Tresadern), atrabanc@its.jnj.com (A.A. Trabanco). 0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.bmcl.2009.11.008 1, LY354740 HO2C CO2H H N H O SMe NH2 2, LY2140023 Figure 1. Structures of mixed mGluR2/3 agonists LY354740 and LY2140023. 176 G. Tresadern et al. / Bioorg. Med. Chem. Lett. 20 (2010) 175–179 OH O Br O N O S O N O O N CF3 N N N H 3, 2,2,2-TEMPS 4 O Cl O O N N O N N N 5, BINA CO2H 6 O NC N N N N O N N H N N N 7 O O H N F 9 8 Cl 10 Cl Figure 2. Reported mGluR2 positive allosteric modulators. changes in endogenous agonist levels. In addition, it is expected that PAMs will not produce the same degree of receptor desensitization.10 The first allosteric modulators of mGluR2 were a series of sulfonamides developed by Lilly that include the 2,2,2-TEMPS 3 which inhibited ketamine evoked norepinephrine release in rats.11 BINA 5 has shown anxiolytic and antipsychotic effects in behavioral models in mice.12 Finally, compound 10 was active in a methamphetamine induced hyperlocomotion model in mice.8h In this Letter we present the discovery of imidazo[1,2-a]pyridines as a new series of mGluR2 PAMs. With the need to identify new hit matter for this target we applied computational techniques based on 3D shape and electrostatic similarity. Such approaches are well suited to scaffold hopping as similarity is assessed using properties important for biological recognition not the underlying atom connectivity. We demonstrate that examples from the imidazopyridine series, show good in vitro mGluR2 PAM activity, selectivity and metabolic stability. An overlay hypothesis for mGluR2 PAMs 4–8 is shown in Figure 3. This was derived with flexible alignment in MOE, the MMFF94x force field, Born solvation, internal dielectric equal to 4 and all other settings as defaults.13,14 We do not present a fully validated mGluR2 PAM pharmacophore here yet the plausible alignment serves to highlight shared features. Firstly, the molecules have a central scaffold with an often exocyclic H-bond acceptor, adjacent to which is a small substituent such as –Cl, –Br, –Me or –CN. Secondly, a larger lipophilic group is pendant from the scaffold, examples include isopentyl, cyclopentyl, phenyl and benzyl. Finally, there is a scaffold substituent with greater extent of structural variation, ranging from pyridyl and phenylpiperidine to larger and more flexible substituted ethers as in 4 and 5. In summary, inspection of the molecules in Figure 2 and overlay in Figure 3 helped identify conserved features of mGluR2 PAM molecules and suggested that alternative scaffolds with suitable features and decoration would yield new hits. Within our labs a process of scaffold hopping using annotated fragment databases and 3D shape and electrostatic similarity has Figure 3. Overlay of molecules 4–8. Selected conserved features are highlighted with dotted spheres, H-bond acceptor, small scaffold substituent and larger lipophilic scaffold group. Molecule color-coding: 4—yellow, 5—magenta, 6—green, 7—black, 8—blue. been implemented.15 In this project a similar approach was applied. Search queries were defined as scaffold fragments of molecules 6–8 with phenyl, ethyl and chloro decoration to mimic the complete molecules, see pyridone query in Figure 4. The search database was constructed from two sources: ideas from the project team and scaffolds generated in an automated fashion via fragmentation of in-house and external compound collections. Database fragments were capped in an analogous manner to the queries using phenyl, ethyl and chloro at each open valence. The database contained 115,647 fragments and 3D conformations were generated with Omega.16 The lowest energy conformer was used 177 G. Tresadern et al. / Bioorg. Med. Chem. Lett. 20 (2010) 175–179 Figure 4. Electrostatic fields of (a) pyridone query and (b) imidazopyridine hit fragment. Negative field is red and positive blue. for each query whilst default settings generated multiple 3D conformers of each database fragment. Firstly, searches were performed with ROCS17 using comboscore ranking to align and identify shape and feature similar database fragments. Subsequently the top 500 hits were re-ranked by their electrostatic field similarity to the query using EON.18 From the search using the pyridone query the imidazopyridine was identified among the best ranked hits. The Poisson–Boltzmann electrostatic Tanimoto similarity was 0.87. The query and hit are shown in Figure 4. Overall they have similar shape, size and distribution of substituent groups. The negative region of the carbonyl acceptor is well reproduced by the imidazo nitrogen. The phenyl groups are both twisted yielding very similar regions of negative charge above and below the ring. In addition the positive regions produced by –CH moieties on the scaffold are well matched. Before synthesis a virtual library of imidazopyridines was enumerated and target compounds including examples from Tables 1 and 2 with the best overall shape and feature similarity compared to molecule 8 were prioritized. The synthesis of the final imidazo[1,2-a]pyridines19 is depicted in Schemes 1 and 2. Microwave promoted thermal condensation of commercially available pyridine derivative 11 with an a-bromoaldehyde 12 afforded in moderate to good yields the 8-cyano-7hydroxyimidazo[1,2-a]pyridines 13. Subsequent reaction of 13 with P(O)Cl3 led to the corresponding chloro derivatives 14 which were then transformed into the final compounds 16 by nucleophilic substitution of the chlorine atom with secondary amine 15 in the presence of diisopropylethylamine (DIPEA) under microwave irradiation (Scheme 1). In the case of aryl substituted imidazo[1,2-a]pyridines 18 the target compounds were obtained by Suzuki coupling of the corresponding boronic esters 17 with the chloroimidazopyridine 14a. Table 1 Functional activity and metabolic stability in rat (RLM) and human liver microsomes (HLM) of representative mGluR2 PAMs (general structure 16)a Compds R1- R2- mGluR2 pEC50a mGluR2 EMAXa (%) RLMb,c (%) HLMb,c (%) 3 5 8 — — — — — — 7.51 7.56 6.20 124 213 235 nd nd nd nd nd nd 16a 6.01 241 88 71 16b <5.5 73 nd nd <5.5 214 nd nd <5.5 248 nd nd 5.96 229 nd 57 6.07 231 26 32 N 5.64 142 nd 70 N 6.19 179 nd 70 N 6.35 128 nd 57 N 16c N N 16d N N 16e N 16f CF3 N N N 16g N 16h N 16i a b c CF3 Values are means of three experiments. RLM and HLM data refer to % of compound metabolized after 15 min at 5 lM concentration. nd: not determined. 178 G. Tresadern et al. / Bioorg. Med. Chem. Lett. 20 (2010) 175–179 Table 2 Functional activity and metabolic stability in rat (RLM) and human liver microsomes (HLM) of representative mGluR2 PAMs (general structure 18)a R1- Compds Ar- 18a CF3 18b CF3 18c CF3 N mGluR2 pEC50a mGluR2 EMAXa (%) RLMb,c (%) HLMb,c (%) 6.35 272 21 32 6.63 123 nd nd 6.85 187 38 28 6.73 200 23 23 6.80 222 nd nd Cl N H 18d CF3 O CF3 18e a b c O Values are means of three experiments. RLM and HLM data refer to % of compound metabolized after 15 min at 5 lM concentration. nd: not determined. NH2 NC N MeO N Br + (i) R1 OHC 11 NC 13 N NH N NC N R1 N HO 12 (ii) R1 R2 NC N N R1 15 N Cl (iii) R2 14 N 16 Scheme 1. Reagents and conditions: (i) EtOH, lW, 150 °C, 40 min, 50–82%; (ii) P(O)Cl3, lW, 150 °C, 15 min, 70–85%; (iii) DIPEA, CH3CN, mW, 180 °C, 15 min, 67– 97%. N NC CF3 N Cl O B O Ar 17 (i) 14a N NC CF3 N Ar 18 Scheme 2. Reagents and conditions: (i) Pd(PPh3)4, NaHCO3 (aq)/1,4-dioxane, 150 °C, 15 min 71–88%. The functional activity20 and microsomal stability data for a set of selected imidazo[1,2-a]pyridines are listed in Tables 1 and 2. Two activity data are reported for each molecule: the maximum % effect increase in glutamate response (EMAX) and the pEC50 for the modulatory effect. Based on data from our previous hit 88f we initially decided to use different phenylpiperazinyl and pyrimidylpiperazinyl groups to study the influence of substitution on the imidazole ring of the bicyclic imidazopyridine core (Table 1). Thus lipophilic alkyl and aryl groups such as n-propyl (16a, 16d and 16h), phenyl (16b), ethyl (16c and 16g), cyclopropylethyl (16e) and 2,2,2-trifluoroethyl (16f and 16i) were introduced. It was pleasing that amongst the first molecules synthesized with the new scaffold, mGluR2 PAM activity was found. This suggested that the scaffold replacement strategy was successful. However, not all combinations were well tolerated. The n-propyl group was active in compounds 16a and 16h yet 16d, although delivering an increase in glutamate response, EMAX 248%, did not have a pEC50 greater than the concentration limit. The highest pEC50 was observed for the more lipophilic groups as seen in the trend within the two subseries 16c–f and 16g–i.21 The 2,2,2-trifluoroethyl was the most active group in combination with both pyrimidylpiperazinyl groups in compounds 16f and 16i with pEC50 values of 6.07 and 6.35, respectively. For compound 16c having the less lipophilic ethyl residue a pEC50 could not be measured despite showing an increase in glutamate signal. In mGlu2 binding experiments 16f displaced a tritiated analogue from the pyridone series with a pIC50 of 6.4 offering further experimental support for the scaffold hopping replacement of pyridone by imidazopyridine. Overall interesting functional activity was found in a similar range to the analogous pyridone 8, pEC50 = 6.20 and EMAX 235%. However, the compounds were less active than reference compounds 3 and 5 in the same assay, 7.51 and 7.56, respectively. Some of the compounds shown in Table 1 were tested for single point microsomal stability in rat and human liver microsomes. Examples containing linear alkyl chains such as n-propyl 16a suffered from extensive metabolism 88% and 71% metabolized after 15 min in rat and human liver microsomes, respectively. Compound 16f with the 2,2,2-trifluoroethyl substituent was better, being 26% metabolized in rat and 32% in human microsomes. With the aim of expanding the SAR and the hope of improving in vitro activity and metabolic stability, the amines present in the structures of 16a–i were replaced by different substituted aryl groups leading to the arylimidazo[1,2-a]pyridines 18a–e shown in Table 2. The 2,2,2-trifluoroethyl substituent on the imidazole G. Tresadern et al. / Bioorg. Med. Chem. Lett. 20 (2010) 175–179 ring was fixed as it had delivered improved pEC50 activity and metabolic stability. In general a diverse range of substituents were well tolerated in the aromatic ring (chloro-, alkyl-, amino-, aryloxygroups) and a good level of activity was retained. As observed in the case of the aminoimidazo[1,2-a]pyridines 16 the 2,2,2-trifluoroethyl substituent led to improved metabolic stability in both HLM and RLM (18a and 18c,d). These compounds were more active than the examples in Table 1 and in all cases more active than the pyridone compound 8. Increase in the EMAX glutamate response was also good. As mentioned earlier, one advantage of allosteric modulation is to facilitate the identification of selective compounds. To examine the selectivity of this new series, examples from both the aminoimidazo[1,2-a]pyridines 16 and arylimidazo[1,2-a]pyridines 18 were profiled in a panel of mGluR assays.22 Examples such as 16f, 16h and 18d showed no activity in either agonism or antagonism assays of mGluR1, mGluR3, mGluR4, mGluR5, mGluR6, mGluR7 or mGluR8 at 10 lM. In summary, a series of imidazo[1,2-a]pyridines with mGluR2 PAM activity have been presented. The scaffold was identified via shape and electrostatic similarity to a known pyridone family previously reported by our team.8f Such approaches can be of value for providing new chemical series in medicinal chemistry programs. Incorporating side chains from our other pyridone series led to the rapid identification of hits. The reported compounds show comparable in vitro activity to reference compounds and examples displayed favorable metabolic stability. Imidazo[1,2-a]pyridines represent a promising avenue for current exploration. Acknowledgements The authors thank collaborators from Addex pharmaceuticals; members of the purification and analysis group; and the in vitro metabolism team from the discovery ADME/Tox group. References and notes 1. Cartmell, J.; Schoepp, D. D. J. Neurochem. 2000, 75, 889. 2. Swanson, C. J.; Bures, M.; Johnson, M. P.; Linden, A. M.; Monn, J. A.; Schoepp, D. D. Nat. Rev. Drug Disc. 2005, 4, 131. 3. Schoepp, D. D.; Marek, G. J. Curr. Drug Targets: CNS Neurol. Disord. 2002, 1, 215. 4. Levine, L.; Gaydos, B.; Sheehan, D.; Goddard, A.; Feighner, J.; Potter, W.; Schoepp, D. D. Neuropharmacology 2002, 43, 294. 5. Patil, S. T.; Zhang, L.; Martenyi, F.; Lowe, S. L.; Jackson, K. A.; Andreev, B. V.; Avedisova, A. S.; Bardenstein, L. M.; Gurovich, I. Y.; Morozova, M. A.; Mosolov, S. N.; Neznanov, N. G.; Reznik, A. M.; Smulevich, A. B.; Tochilov, V. A.; Johnson, B. G.; Monn, J. A.; Schoepp, D. D. Nat. Med. 2007, 13, 1102. 179 6. Fell, M. J.; Svensson, K. A.; Johnson, B. G.; Schoepp, D. D. J. Pharmacol. Exp. Ther. 2008, 326, 209. 7. Jeffrey Conn, P.; Christopoulos, A.; Lindsley, C. W. Nat. Rev. Drug Disc. 2009, 8, 41. 8. Compound 3: (a) Barda, D. A.; Wang, Z. Q.; Britton, T. C.; Henry, S. S.; Jagdmann, G. E.; Coleman, D. S.; Johnson, M. P.; Andis, S. L.; Schoepp, D. D. Bioorg. Med. Chem. Lett. 2004, 14, 3099; compound 4: (b) Pinkerton, A. B.; Cube, R. V.; Hutchinson, J. H.; Rowe, B. A.; Schaffhauser, H.; Zhao, X.; Daggett, L. P.; Vernier, J. M. Bioorg. Med. Chem. Lett. 2004, 14, 5329; compound 5: (c) Pinkerton, A. B.; Cube, R. V.; Hutchinson, J. H.; James, J. K.; Gardner, M. F.; Rowe, B. A.; Schaffhauser, H.; Rodriguez, D. E.; Campbell, U. C.; Daggett, L. P.; Vernier, J. M. Bioorg. Med. Chem. Lett. 2005, 15, 1565; (d) compound 6: Empfield, J.; Folmer, J. WO2007/095024 A1; (e) compound 7: Balestra, M.; Bunting, H; Chen, D.; Egle, I.; Forst, J.; Frey, J.; Isaac, M.; Ma, F.; Nugiel, D.; Slassi, A.; Steelman, G.; Sun, G.; Sundar, B.; Ukkiramapandian, R.; Urbanek, R.; Walsh, S. WO2006/071730 A1; (f) compound 8: Cid-Nuñez, J. M.; Andrés-Gil, J.; Trabanco-Suárez, A. A., Oyarzábal-Santamarina, J.; Dautzenberg, F.; Macdonald, G.; Pullan, S.; Lutjens, R.; Duvey, G.; Nhem, V.; Finn, T.; Melikyan, G. WO2007/104783 A1; compound 9: (g) Zhang, L.; Rogers, B. N.; Duplantier, A. J.; McHardy, S. F.; Efremov, I.; Berke, H.; Qian, W.; Zhang, A. Q.; Maklad, N.; Candler, J.; Doran, A. C.; Lazzaro, J.; Ganong, A. H. Bioorg. Med. Chem. Lett. 2008, 18, 5493; compound 10: (h) Duplantier, A. J.; Efremov, I.; Candler, J.; Doran, A. C.; Ganong, A. H.; Haas, J. A.; Hanks, A. N.; Kraus, K. G.; Lazzaro, J.; Lu, J.; Maklad, N.; McCarthy, S. A.; O’Sullivan, T. J.; Rogers, B. N.; Siuciak, J. A.; Spracklin, D. K.; Zhang, L. Bioorg. Med. Chem. Lett. 2009, 19, 2524. 9. Recasens, M.; Guiramand, J.; Aimar, R.; Abdulkarim, A.; Barbanel, G. Curr. Drug Targets 2007, 8, 651. 10. Gjoni, T.; Urwyler, S. Neuropharmacology 2008, 55, 1293. 11. Lorrain, D. S.; Schaffhauser, H.; Campbell, U. C.; Baccei, C. S.; Correa, L. D.; Rowe, B.; Rodriguez, D. E.; Anderson, J. J.; Varney, M. A.; Pinkerton, A. B.; Vernier, J. M.; Bristow, L. J. Neuropsychopharmacology 2003, 28, 1622. 12. Galici, R.; Jones, C. K.; Hemstapat, K.; Nong, Y.; Echemendia, N. G.; Williams, L. C.; de Paulis, T.; Conn, P. J. J. Pharmacol. Exp. Ther. 2006, 318, 173. 13. Chemical Computing Group Inc. Molecular Operating Environment (MOE), 1010 Sherbrooke St. W, Suite 910, Montreal, QC, Canada; available from Chemical Computing Group Inc. at http://www.chemcomp.com. 2009. 14. mGluR2 PAM binding data generated in house via displacement of a tritiated compound from our internal pyridone series (structure not disclosed) provides support that some examples share a common binding site. Example 5 had a pIC50 of 6.8 and another member of the isoindolone series displayed pIC50 of 6.6. 15. Oyarzábal, J.; Howe, T.; Alcazar, J.; Andrés, J. I.; Alvarez, R. M.; Dautzenberg, F.; Iturrino, L.; Martínez, S.; Van der Linden, I. J. Med. Chem. 2009, 52, 2076. 16. OMEGA. OpenEye Scientific Software, Inc.; http://www.eyesopen.com. 2006. 17. Grant, J. A.; Gallardo, M. A.; Pickup, B. T. J. Comput. Chem. 1996, 17, 1653. 18. Nicholls, A.; MacCuish, N. E.; MacCuish, J. D. J. Comput. Aided Mol. Des. 2004, 18, 451. 19. Trabanco-Suárez A. A.; Tresadern G. J.; Vera-Ramiro J. A.; Cid-Nuñez J. M. WO2009/062676 A2. 20. The effect of these compounds on the [35S]-GTPcS binding induced by 4 lM glutamate (EC20) was characterized using a CHO cell line expressing the human mGluR2 receptor. See Ref. 19 for a detailed description of this assay. 21. The calculated Alog P for 16c–f were 1.2, 1.6, 2.0 and 1.6, respectively, and 1.8, 2.2 and 2.2 for 16g–i. 22. Compounds were tested for agonist or antagonist activity on mGlu receptors in fluorescent Ca2+ assays using HEK293 cells expressing human mGluR1, mGluR5, mGluR3, mGluR7 or mGluR8. Effects on the human mGluR4 and rat mGluR6, expressed in L929 or CHO cells, were assessed in [35S]-GTPcS functional assays.