Bioorganic & Medicinal Chemistry 23 (2015) 3040–3058

Contents lists available at ScienceDirect

Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc

Insights into the interaction of negative allosteric modulators with the metabotropic glutamate receptor 5: Discovery and computational modeling of a new series of ligands with nanomolar affinity Andrew Anighoro a, , Davide Graziani b, , Ilaria Bettinelli b, Antonio Cilia b, Carlo De Toma b, Matteo Longhi b, Fabio Mangiarotti b, Sergio Menegon b, Lorenza Pirona b, Elena Poggesi b, Carlo Riva b, Giulio Rastelli a,⇑ a b

Life Sciences Department, University of Modena and Reggio Emilia, Via Campi 103, 41125 Modena, Italy Drug Discovery Department, Recordati S.p.A. Via Civitali 1, 20148 Milan, Italy

a r t i c l e

i n f o

Article history: Received 26 February 2015 Revised 28 April 2015 Accepted 4 May 2015 Available online 12 May 2015 Keywords: Metabotropic glutamate receptor 5 Negative allosteric modulators Homology modeling Docking

a b s t r a c t Metabotropic glutamate receptor 5 (mGlu5) is a biological target implicated in major neurological and psychiatric disorders. In the present study, we have investigated structural determinants of the interaction of negative allosteric modulators (NAMs) with the seven-transmembrane (7TM) domain of mGlu5. A homology model of the 7TM receptor domain built on the crystal structure of the mGlu1 template was obtained, and the binding modes of known NAMs, namely MPEP and fenobam, were investigated by docking and molecular dynamics simulations. The results were validated by comparison with mutagenesis data available in the literature for these two ligands, and subsequently corroborated by the recently described mGlu5 crystal structure. Moreover, a new series of NAMs was synthesized and tested, providing compounds with nanomolar affinity. Several structural modifications were sequentially introduced with the aim of identifying structural features important for receptor binding. The synthesized NAMs were docked in the validated homology model and binding modes were used to interpret and discuss structure–activity relationships within this new series of compounds. Finally, the models of the interaction of NAMs with mGlu5 were extended to include important non-aryl alkyne mGlu5 NAMs taken from the literature. Overall, the results provide useful insights into the molecular interaction of negative allosteric modulators with mGlu5 and may facilitate the design of new modulators for this class of receptors. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Glutamate, the major excitatory neurotransmitter in the central nervous system, acts on two distinct classes of receptors: ionotropic glutamate receptors to elicit fast excitatory responses, and metabotropic glutamate receptors (mGlu) to modulate synaptic transmission.1 Ionotropic glutamate receptors have been classified

Abbreviations: DMAC, N,N-dimethylacetamide; DMF, N,N-dimethylformamide; LiHMDS, lithium bis(trimethylsilyl)amide; mGlu5, metabotropic glutamate receptor subtype 5; NAM, negative allosteric modulator; GPCR, G-protein coupled receptor; TFA, trifluoroacetic acid; TLC, thin layer chromatography; MD, molecular dynamics; rt, room temperature. ⇑ Corresponding author. Tel.: +39 059 2058564; fax: +39 059 2055131. E-mail address: [email protected] (G. Rastelli).   These authors contributed equally. http://dx.doi.org/10.1016/j.bmc.2015.05.008 0968-0896/Ó 2015 Elsevier Ltd. All rights reserved.

into NMDA, AMPA and kainic acid receptor subtypes, according to the ability of ligands to bind to the different receptor sites and to produce distinct physiological effects.2 Since the mid-1980s, evidences of the existence of a large family of metabotropic glutamate receptors coupled to effector systems through GTP-binding proteins have been reported. The mGlus belong to class C G-protein coupled receptors (GPCRs), and are characterized by an unusually large extracellular amino-terminal domain (ATD) (500–600 amino acids) with no sequence homology to other families of GPCRs. Although the sequence homology among class C members is low (about 20% amino acid identity), these receptors are structurally related. O’Hara et al. observed that the ATD of mGlu1 shows some degree of similarity with a family of bacterial periplasmic amino acid-binding proteins,3 in particular the leucine-, isoleucine- and valine binding protein (LIVBP).4 Based on the crystal

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

structure of LIVBP, they proposed a bilobal structure for the agonist-binding pocket of mGlu receptors in which the glutamate is bound in a ‘venus flytrap’ mechanism. To date, eight metabotropic glutamate receptor subtypes have been cloned and classified into three groups on the basis of sequence similarity, agonist and antagonist binding profiles, and preferred coupling to signal transduction pathways. Group I receptors (mGlu1 and mGlu5) preferentially couple via Gq/11 proteins to stimulate phospholipase C activity but are differentially localized and mediate distinct physiological functions.5–8 Activation of these receptors elicits highly distinct Ca2+ responses at cellular level. mGlu1 primarily elicits a peak-plateau type Ca2+ response, whereas mGlu5 leads to oscillatory changes in intracellular Ca2+ concentration in both recombinant and native (e.g., astrocyte) cell environment.9–13 Group II receptors (mGlu2 and mGlu3) are negatively coupled to cAMP production and are not stimulated by L-(±)-2-amino-4-phosphono butyric acid (L-AP4), and finally group III receptors (mGlu4, mGlu6, mGlu7, and mGlu8) are negatively coupled to cAMP production but are activated by L-AP4.1,14 Because of their critical role as modulators of synaptic transmission, ion channel activity, and synaptic plasticity,15,16 mGlu receptors are implicated in major neurological disorders such as Alzheimer’s and Parkinson’s disease as well as depression, schizophrenia, anxiety, and pain.17,18 The ongoing interest in mGlu5 as a drug target is borne out by the large number of companies focused on the discovery of new ligands and new chemotypes acting as negative allosteric modulators of this receptor. Clinical studies that received major consideration include drugs for the treatment of the L-DOPA induced dyskinesia in Parkinson’s disease (PD-LID), anxiety, depressive disorders, and the treatment of Fragile X syndrome (FXS). mGlu5 antagonists involved in phase II clinical trial studies include AFQ056/mavoglurant (Novartis) for PD-LID and FXS, ADX48621/Dipraglurant (Addex Therapeutics) for PD-LID, and RG7090/Basimglurant (Roche) for FXS and major depressive disorders. Very recently, Novartis and Roche announced the suspension of the development of mavoglurant and basimglurant for Fragile X syndrome. The other trials are still on-going, but so far none of the mGlu5 candidates have progressed to phase III. In conclusion, despite the extensive efforts, the identification of potent, selective and clinically efficacious candidates does still represent a highly desirable goal. Recently, the crystal structure of the transmembrane domain of mGlu1 in complex with the negative allosteric modulator 4Fluoro-N-[4-[6-(isopropylamino)pyrimidin-4-yl]-1,3-thiazol-2-yl]N-methylbenzamide (FITM)19 has been reported. The structure, which was the first of a class C GPCR, provided important insights into the architecture of the seven-transmembrane (7TM) domain of mGlu1 and revealed the location of the allosteric modulator binding site, providing a key framework for understanding molecular recognition and enabling the structure-based design of new modulators for this class of receptors. Here we report a homology model of the 7TM domain of mGlu5 based on the crystal structure of the high sequence identity (77%) mGlu1 template. Moreover, we describe the complexes of mGlu5 with MPEP and fenobam (Scheme 1), two established negative allosteric modulators (NAMs), predicted by means of docking and molecular dynamics simulations. The structural models were validated using mutagenesis data available in the literature, and subsequently corroborated by the recently described crystal structure of mGlu5 in complex with mavoglurant.20 Finally, the validated models were used to predict the binding mode of a newly synthesized series of NAMs based on arylpropiolic acid piperazine amides and 4-(3-arylprop-2-ynylidene)piperidine scaffolds21 (Scheme 1) and to discuss structure–activity relationships within this class. The structures of these compounds were elaborated

3041

starting from the piperazine propionamide scaffold (compound 5 in Scheme 1) reported by Euro-Celtique22 (now Purdue Pharma) by replacing the amide group with a double bond, followed by terminal aryl optimization. To this aim, a subset of NAMs synthesized during our medicinal chemistry efforts21 was carefully selected by including molecules with key structural variations and different activities. In particular, key compounds were selected in consideration of their suitability to investigate and probe the effects of the substitution pattern, conformational requirements, and other features of the mGlu5 allosteric pocket relevant for receptor binding. The compounds were docked into the receptor structure and the resulting binding modes and structure–activity relationships were used to further elaborate and probe the model of the interaction of NAMs with the allosteric pocket. The reduction of the acetoamido group of compound 5 (rat Ki 0.6 nM, human Ki 0.5 nM) provided compound 12 with significantly reduced binding affinity in both rat and human species (rat Ki 117.4 nM, human Ki 60.7 nM). Reestablishing planarity with the introduction of a double bond, such as in compound 13 (rat Ki 0.6 nM, human Ki 0.4 nM), restored high affinity for mGlu5 and excluded the possibility that a hydrogen bond acceptor mapped on the carbonyl was strictly necessary for binding. After terminal aryl optimization potent (nanomolar) and selective NAMs of mGlu5 were obtained. Finally, the model of the interaction of NAMs with mGlu5 was further extended by docking clinically-important non-aryl alkyne mGlu5 NAMs taken from the literature.

2. Results and discussion 2.1. Putative MPEP and fenobam binding modes in the mGlu5 homology model A refined homology model of the 7TM domain of mGlu5 was built on the basis of the recently reported mGlu1 crystal structure template19 (see experimental methods for details). The two receptors share high sequence identity (77%). Then, docking of MPEP and fenobam in the homology model was performed, and the predicted binding poses were compared with the mutagenesis data available for these two ligands. Mutagenesis studies suggest that fenobam, N-(3-chlorophenyl)-N0 -(4,5-dihydro-1-methyl-4-oxo-1H-imidazole-2-yl)urea (compound 3 in Scheme 1) and MPEP, 2-methyl-6(phenylethynyl)pyridine, (compound 1) share the same or at least a partially overlapping binding site.23–27 The Y659V, W785A, F788A, and A810V mutations abolished MPEP and fenobam binding. In addition, P655S, S658C, and T781A mutations severely impaired fenobam binding but left MPEP binding almost unaffected. Remarkably, in the generated model the residues mentioned above clearly define the walls of a pocket having a size and shape able to accommodate the two NAMs. Interestingly, superposition of the mGlu1 crystal structure with the mGlu5 homology model shows that the area encompassed by these key residues overlaps only partially with the binding pocket of FITM, i.e. the NAM co-crystallized with mGlu1 (Fig. 1A). For instance, the fluorophenyl ring of FITM, which points to the bottom of the pocket toward the intracellular side, is found in proximity of W785, F788 and T781 but is 8 Å distant from A810 and S658, and the substituted pyrimidine ring, which extends toward the opposite extracellular side, is >5 Å distant from other residues identified as important by mutagenesis of mGlu5. Compared to mGlu1, mGlu5 exhibits an additional sub-pocket recently proposed to become accessible in mGlu5 because of the serine to proline 655 replacement.19,20,28 Interestingly, we observed that A810, one of the residues lining this intracellular sub-pocket, is a valine in mGlu1. As a consequence of the bulkier side chain, the pocket is

3042

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

Scheme 1. Chemical structures of the investigated compounds.

not accessible in mGlu1. In agreement with this observation, the A810V mutation in mGlu5 completely abolished both MPEP and fenobam binding to mGlu5,24–27 confirming that the sub-pocket is unique to mGlu5.

MPEP and fenobam were docked in the identified binding site using Induced Fit Docking (Schrödinger Suite 2014). The resulting top-ranking orientations were visually inspected and structures were analyzed for their consistency with the available mutagenesis

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

3043

Figure 1. Predicted MPEP and fenobam binding site in the 7TM region of mGlu5. (A) The key residues identified by mutagenesis are highlighted in orange. The location of the negative allosteric modulator FITM (shown in cyan) taken from the mGlu1-FITM crystal structure (PDB code 4OR2) is shown for comparison. (B) MPEP, binding mode A found with induced fit docking. (C) MPEP, binding mode B found with induced fit docking. (D) Fenobam, binding mode found with induced fit docking. (E) Fenobam, binding mode found by molecular dynamics. (F) Fenobam, alternative hydrogen bond pattern explored by molecular dynamics.

data. In order to explore the conformational variability of the ligand-mGlu5 complexes and to highlight possible time-dependent structural features, the selected docking orientations were also investigated with molecular dynamics simulations. Two possible docking orientations were identified for MPEP (Fig. 1B and C). In both orientations MPEP interacts with the four key residues indicated as important by mutagenesis, that is, Y659, W785, F788, and A810, the latter residue being located at the more buried end of the intracellular sub-pocket described above. The two binding modes correspond to head to tail orientations in which the A810 sub-pocket is alternatively occupied by the methyl-pyridine (Fig. 1B) or by the phenyl (Fig. 1C) rings. In both cases, it is interesting to observe that the A810V mutation would cause significant steric clashes with MPEP, explaining why this NAM is unable to bind the receptor mutant. Hydrogen bonds between Y659 and the pyridine nitrogen were observed in both orientations. However, since mutation of this tyrosine into phenylalanine (Y659F) does not affect MPEP binding,24 such hydrogen bond appears to be dispensable. On the contrary, a Y659V mutation would significantly alter the hydrophobicity and shape of the binding site in this position, explaining why a valine in this position is detrimental for MPEP and fenobam binding.24,25 One of the two MPEP binding modes (Fig. 1C) displays an additional hydrogen bond with S809, an interaction already hypothesized by Mølck et al.26 Even if the S809A mutation does not destroy MPEP binding, potency is >50 fold decreased, suggesting that this hydrogen bond helps stabilizing the complex.26 W785 and F788 make hydrophobic contacts with the phenyl (Fig. 1B) or with the methyl-pyridine moiety (Fig. 1C) of MPEP. Mutations of these two residues into alanine are detrimental for NAM binding. Finally, other hydrophobic contacts are formed with P655, V806, and I784. The proposed binding modes were generally conserved during a 20 ns molecular dynamics simulation, as indicated by small average root mean squared deviations with respect to the docked orientation. Hydrogen bonding between the MPEP nitrogen and Y659 was characterized by a rather low occupancy (8% and 18% in binding modes of Fig. 1B and C, respectively). Hydrogen bonds between Y659 and

the nearby T781 side chain were observed for only 19% of the simulation time in the binding mode of Figure 1B. These findings may explain why the Y659F mutant binds MPEP with an affinity similar to that of wild type.25 According to these data, the binding of MPEP is almost entirely driven by hydrophobic interactions. Docking of MTEP (compound 2) and its analogue 4 (Scheme 1) provided binding modes similar to that of MPEP (Fig. S1). With respect to MPEP and MTEP, the bulkier compound 4 protrudes towards M802 and Y792 and required a rotation of the F788 side chain to be accommodated in the binding site. Docking of fenobam into the mGlu5 homology model suggested that the molecule binds in the same pocket identified for MPEP (Fig. 1D). In the complex, the dihydro-imidazol-4-one ring occupies the bottom of the sub-pocket and is in proximity of A810 and S658. Mutation of serine 658 into cysteine (S658C) is detrimental for the binding of fenobam,24 suggesting the presence of a hydrogen bond. Indeed, in the docking complex the carbonyl group of the dihydroimidazol-4-one ring is in proximity of S658, and a hydrogen bond was clearly detected by molecular dynamics (see below). The urea carbonyl forms a hydrogen bond with Y659, even though this interaction seems dispensable for the reasons described above. As for hydrophobic interactions, residues A810, P655, W785, and F788 are in close contact with the ligand. In the modeled complex, the chlorophenyl ring of fenobam is almost superimposed with the fluorophenyl ring of FITM in the mGlu1-FITM crystal structure (Fig. 2B). In this orientation, an A810V mutation would provide significant steric clashes with the dihydro-imidazol-4-one ring of fenobam, explaining the observed loss of binding for this mutant. The molecular dynamics simulation of the mGlu5-fenobam complex revealed a highly dynamic hydrogen bonding network involving several surrounding polar residues, in which the urea carbonyl hydrogen bonds Y659 (23% occupancy, Fig. 1E) or S809 (32% occupancy, Fig. 1F). As shown in Figure 1E and F, residue T781 participates to this hydrogen bond network by contacting either S809 (20% occupancy) and Y659 (37% occupancy). Interestingly, the hydrogen bond between the carbonyl of the dihydro-imidazol-4one ring of fenobam and S658 was fairly populated during

3044

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

Figure 2. Predicted binding mode of compound 14. (A) binding mode found with induced fit docking. (B) Superposition of compound 14 (shown in cyan), MPEP (binding modes A and B, represented in light green and dark green, respectively), fenobam (most populated binding mode during molecular dynamics, shown in orange), FITM (shown in magenta), and mavoglurant (shown in grey).

molecular dynamics (43% occupancy, Fig. 1E and F). Such interaction with S658 is consistent with the key role of this residue as indicated by the mutagenesis data mentioned above. Overall, the alternative hydrogen bonding configurations identified by molecular dynamics can help rationalize the experimentally observed importance of residues S658 and T781, and predict an important role also for S809, for which mutagenesis data are not yet available. In conclusion, the modeling results nicely explain the mutagenesis data and suggest that a dynamic network of hydrogen bonds with a number of polar residues of the cavity can be established. As for hydrophobic interactions, residues A810, P655, W785, and F788 were found to be in close contact with fenobam, in agreement with the observed lack of binding to receptors in which these residues were mutated to alanine. These findings further validate the generated homology model and the proposed binding modes. A superposition between the binding modes of MPEP, fenobam, FITM, mavoglurant, and compound 14 (described below) is shown in Figure 2B for comparison. 2.2. Chemical synthesis The structures of the synthesized compounds along with reference molecules are reported in Scheme 1. MPEP, MTEP and fenobam were purchased from Tocris Bioscience, while compound 4 was prepared as described by Roppe et al.29 (Scheme 2). Compounds 5–8 were synthesized by condensing 3-phenylpropiolic acid or quinoxaline-2-carboxylic acid with the corresponding piperazine derivatives using diethyl cyanophosphonate or benzotriazol-1-ol (HOBt), 3-(ethyliminomethyleneamino)-N,Ndimethylpropan-1-amine (EDC) as condensing agent in DMF as solvent (Scheme 3). Compounds 9–11, in which the alkyne moiety was replaced with a phenyl ring, were obtained by reacting 2-, 3and 4-biphenylcarboxylic acid with 1-(3-nitro-2-pyridyl)piperazine or 1-(-6-methyl-3-nitro-2-pyridyl)piperazine using EDC as condensing agent in DMF as solvent (Scheme 3). The preparation of 12 was carried out by a CuI catalyzed Mannich reaction in DMSO of ethynylbenzene as terminal alkyne (Scheme 4). The synthesis of compounds bearing an exocyclic double bond at position 4 of the piperidine ring (13–25) was based on different strategies according to the stability and reactivity of the intermediates to Sonogashira reaction conditions. Two key intermediates were used for the synthesis of most compounds. 3-Nitro-2-[4(prop-2-ynyliden)piperidin-1-yl]pyridine (13c) was particularly useful for the introduction of several aromatic rings as substituents at the triple bond, for example, phenyl (13), 2-pyridyl (14), 6-methyl-2-pyridyl (15), 3-tolyl (18), 3-chlorophenyl (19), 3-chloro-5-cyanophenyl (20), 1-[4-(3,3-dimethyl-2-oxoazetidin1-yl)phenyl (22) (Scheme 5). The second key synthetic intermediate was the 2-methyl-6-[3-(piperidin-4-ylidene)prop-1-yn-1-

yl]pyridine (16d). The Horner-Emmons reaction of diethyl (3trimethylsilylprop-2-ynyl)phosphonate (13a) in THF, lithium bis(trimethylsilyl)amide (LiHMDS) and t-butoxycarbonyl protected piperidin-4-one led to the intermediate tert-butyl 4-[3(trimethylsilyl)prop-2-yn-1-ylidene]piperidine-1-carboxylate (16a). The trimethylsilyl protection was removed by tetra-n-butylammonium fluoride (TBAF) and the resulting intermediate (16b) was reacted with 2-bromo-6-methylpyridine to obtain compound 16c. The removal of BOC protection using trifluoroacetic acid (TFA) in CHCl3 yielded the key intermediate 2-methyl-6-[3-(piperidin-4-ylidene)prop-1-yn-1-yl]pyridine (16d) which was reacted following a Büchwald–Hartwig or nucleophilic aromatic substitution (SNAr) methodology with bromobenzene, 1-bromo-2-nitrobenzene and 2-chloro-5-phenylnicotinonitrile, to obtain compounds 16, 17 and 23, respectively (Scheme 6). Compound 21 was synthesised by reacting intermediate 16b with 2-chloro-5-cyanopyridine according to Sonogashira procedure in the presence of bis(triphenylphosphine)palladium(II) dichloride (PdCl2(PPh3)2) and CuI, to obtain after purification the tert-butyl 4-[3-(5-cyano-2-pyridyl)prop-2-ynylidene]piperidine1-carboxylate (21a). The latter compound was deprotected with TFA (21b) and condensed with 2-chloro-3-nitro-6-methylpyridine to obtain compound 21 (Scheme 7). The synthesis of compounds 24 and 25 was accomplished through reaction of the ylide obtained from compound 13a on the BOC protected piperidin-3-one (Scheme 8). Both the stereoisomers E (24a) and Z (25a) were obtained and separated by chromatography. Afterwards, trimethylsilyl protection was removed by TBAF, and the terminal alkynes 24b and 25b were reacted by Sonogashira coupling with 2-bromo-6-methylpyridine to obtain, respectively, the BOC protected (E) and (Z) 3-[3-(6-methyl-2-pyridyl)prop-2-ynylidene]piperidine 24c and 25c. The usual TFA deprotection followed by condensation with 2-chloro-6-methyl3-nitropyridine led to the two isomers E (24) and Z (25). 2.3. Biological results The ability of compounds 1–25 (Scheme 1) to displace the allosteric antagonist [3H]MPEP was used to determine the binding affinity of these compounds for mGlu5. Membranes were prepared from rat frontal cortex or from human and rat cloned receptors expressed in Chinese Hamster Ovary (CHO) T-REx cells. The binding affinities, expressed as inhibition constants (Ki), are reported in Table 1. Figure 3 shows the displacement curve of [3H]MPEP by means of compound 15 (Ki = 0.5 nM) on membranes of CHO T-REx cells stably transfected with human cloned mGlu5 receptors. Table 1 reports the measured binding affinities and functional activities of compounds 1–25. First, compounds were tested using rat native mGlu5 receptor, then the binding data of the most active

3045

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

N O

Cl

N N

a

S

S

S

b

c N

Si N

Si

N

Cl

4a (50.4%)

4b (67.3%)

4c (66.6%)

4 (5.7%)

Scheme 2. Synthetic scheme of compound 4. Reagents and conditions: (a) thioacetamide, DMF, 25 °C. (b) 2-chloro-5-iodopyridine, CuI, TBAF, Pd(PPh3)4, TEA, DMF, 70 °C. (c) 3-pyridylboronic acid, PdCl2(PPh3)2, K2CO3, DME, 80 °C.

NH N

Ar

+

O

a

O

N R

HO

R

N

Ar

5-11 (13.8 - 93%) Number 5 6 7 8 9 10 11

R PhC#C PhC#C PhC#C quinoxalin-2-yl 4-phenylphenyl 3-phenylphenyl 2-phenylphenyl

Ar 3-nitropyrid-2-yl 1-(2-nitrophenyl) 3-nitro-6-vinylpyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl 6-methyl-3-nitropyrid-2-yl 3-nitropyrid-2-yl

Scheme 3. Synthetic scheme of compounds 5–11. Reagents and conditions: (a) DEPC and TEA, DMF, 0 °C to 25 °C, or HOBt, EDC, DMF, 0 °C to 25 °C.

NH N

N O N+ – O

N

a

+

N

N O N+ – O

12 (99%)

Scheme 4. Synthetic scheme of compound 12. Reagents and conditions: (a) HCHO, DMSO, CuI (cat.) 25 °C.

O N

N +O N – O

a

O P O O

+ Si

N

N +O N – O

13a

b

N

N O N+ – O

13c

Number 13 14 15 18 19 20 22

SiMe3

13b

c

N

N

Ar

O N+ – O

13-15, 18-20, 22 (41.5 - 95%)

Ar Ph 2-pyridyl 6-methylpyrid-2-yl 3-methylphenyl 3-chlorophenyl 3-chloro-5-cyanophenyl 4-(3,3-dimethyl-2-oxo-azetidinyl)phenyl

R 3-nitropyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl 3-nitropyrid-2-yl

Scheme 5. Synthetic scheme of compounds 13–15, 18–20, 22. Reagents and conditions: (a) LiHMDS, THF, 60 °C, 25 °C. (b) TBAF, THF, 25 °C. (c) aryl halide, Pd(PPh3)4, CuI, TEA, 80 °C.

3046

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

O P Si

O

a

O

+

N

O

O

O

N O

O

16b

16a (92%)

d

N

N

O

O

13a

c

b

Si

N

O

e

N

N

Ar1

N

N

O

16c

16d

Number 16 17 23

16, 17, 23 (13.5%, 27.6%, 9.5%)

Ar1 phenyl 1-(2-nitrophenyl) 5-phenyl-3-cyanopyridin-2-yl

Scheme 6. Synthetic scheme of compounds 16, 17, 23. Reagents and conditions: (a) LiHMDS, THF, 60 °C then 25 °C. (b) TBAF, THF, 25 °C. (c) 2-bromo-6-methylpyridine, Pd(PPh3)4, CuI, TEA, 80 °C. (d) TFA, CHCl3, 70 °C. (e) PhBr, Cs2CO3, BINAP, Pd(OAc)2, PhCH3, 110 °C; 1-Br-2-NO2Ph, 110 °C; 2-chloro-5-phenylnicotinonitrile, (i-Pr)2EtN, NMP, 150 °C 15 min (mw).

N

N

N

N

N N

b

a

c

N O

N

O O

16b

N O

21a (51.3%)

N

N

21b (89%)

O N+ – O

21 (90%)

Scheme 7. Synthetic scheme of compound 21. Reagents and conditions: (a) 2-chloro-5-cyanopyridine, PdCl2(PPh3)2, CuI, TEA, 80 °C. (b) TFA, CHCl3, 70 °C. (c) 2-chloro-3nitro-6-methylpyridine, TEA, NMP, 25 °C.

compounds were confirmed by binding assay on CHO human cloned mGlu5 receptor obtaining comparable outcomes. Compounds 20, 22 and 25 were directly tested on human mGlu5 receptor (Table 1). The similar activities are in agreement with the fact that the sequences of the 7TM regions of rat and human mGlu5 are 100% identical. For the most potent compounds (binding affinity lower than 100 nM), Table 1 also reports binding affinities for rat cloned mGlu1, functional activity for human mGlu5 and rat mGlu1 subtypes determined by calcium mobilization assay. Table 1 shows how human mGlu5 binding affinity and functional activity assay results are in agreement. The limited number of compounds tested on mGlu1 assays display larger difference in binding and functional results in comparison with mGlu5. MPEP and MTEP are known to be highly selective30 for mGlu5 subtype; this selectivity was confirmed in our experiments. Arylpropiolic acids (13–21, 23) showed remarkable increase in mGlu5 affinity and meaningful functional activity in calcium assay with respect to MPEP and MTEP, although an increased affinity for mGlu1 was also noted. Binding selectivity ratios were in favour of mGlu5, (compound 5 mGlu1/mGlu5 selectivity ratio: 161). The same comparison made on functional data resulted in even better outcomes (compound 5 mGlu1/mGlu5 selectivity ratio: 311). Both classes of compounds showed this trend in selectivity, compounds 14 and 15 bore mGlu1/mGlu5 binding selectivity ratios, respectively, of 124 and 382, and functional selectivity ratios of 517 and 563 in assays.

Figure 4 shows the ability of compounds 14 and 15 to inhibit the quisqualate-induced [Ca2+]i response in human mGlu5 receptors. Under these conditions, the two compounds displayed potent IC50 values of 0.4 nM and 0.8 nM, respectively. Owing to the non-competitive nature of these NAMs, an increase of the concentration of the tested compound shifted the concentration-response curves of quisqualate to the right, and reduced the maximal response to the agonist. Compound 15 (Fig. 5) reduced the agonist response of almost 50% at a 0.3 nM concentration, and completely abolished quisqualate response at a 1 nM concentration. Starting from compound 5, the relevance of substituents on the two distal parts of the molecules was evaluated using human mGlu5 binding affinity data. Condensation of 3-phenylpropiolic acid and 1-(2-nitrophenyl)piperazine led to compound 6 having a binding affinity (Ki 0.4 nM) similar to that of compound 5. One of the most intriguing features of mGlu5 NAMs was the crucial importance of the triple bond. In these compounds, the triple bond connects an aromatic ring with N-carboxypiperazine group. The attempts to substitute the triple bond with a linear para-phenyl (9, Ki >10,000 nM), meta-phenyl (10, Ki >10,000 nM), or ortho-phenyl ring moiety (11, Ki >10,000 nM) resulted in loss of binding affinity. The incorporation of the triple bond into a condensed bicyclic aromatic ring also resulted in a loss of binding affinity (8, Ki >10,000 nM).

3047

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

Table 1 In vitro binding affinities (Ki, nM) for rat native and human cloned mGlu5, in vitro binding affinities for rat cloned mGlu1, and in vitro Ca2+ functional mobilization assay activities (IC50, nM) for rat cloned mGlu1 and human cloned mGlu5

O N O O

a Si

25a (9.5%)

b

N

O O

O

O

24a (10.3%)

N

O

O

24b (42.5%)

25b (60.5%) c O N

N

N

O

N

O

24c (42.5%)

O

25c (61%)

d

N

N

O

N

O

24d

O

25d

e

N N O

24 (11%)

r-mGlu1 Kic (nM)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

15.3 38.5 312.7 11.2 0.6 1.8 1.3 >1000 >1000 >10,000 >10,000 117.4 0.6 0.1 0.03 92 0.2 0.4 0.4

10.4 32.7 162.2 4.7 0.5 0.4

11.2 47.3 36.3 6.1 0.9

>10,000 >10,000

0.4 2.9 111

>10,000 96.8 69.3 54.2

d

42 >10,000 >10,000 >10,000 >10,000 60.7 0.4 0.3 0.5 29.2 0.3 0.3 0.5 283.1 0.3 >1000 0.5 41.8 264.3

>1000 4.8 39.1 184.2

2.8 0.4 0.8 55 10.4 1.8 3.8

9.7 91

271

39.3 206.7 447.5

70.7 18.6 18.6 >1000 189.6

1.3

r-mGlu1 Ca2+ IC50c (nM)

550.7 543.4 1337

443.1 >1000 >1000

N

N N

h-mGlu5 Ca2+ IC50b (nM)

The results were obtained by three independent experiments and each assay point was determined at least as duplicate. SEM was less than 10% of the mean and omitted. a rat forebrain mGlu5. b CHO human cloned mGlu5 receptor. c CHO rat cloned mGlu1 receptor. d Compound 6 showed interference with the fluorescence assay, therefore its activity was impossible to determine.

O N

h-mGlu5 Kib (nM)

N

Si N

O

r-mGlu5 Kia (nM)

N

+

O

–

N

O N+ – O

25 (18%)

Scheme 8. Synthetic scheme of compounds 24 and 25. Reagents and conditions: (a) 13a, LiHMDS, THF, 60 °C then 25 °C. (b) TBAF, THF, 25 °C. (c) 2-bromo-6methylpyridine, Pd(PPh3)4, CuI, TEA, 80 °C. (d) TFA, CHCl3, 70 °C. (e) 2-chloro-6methyl-3-nitropyridine, TEA, DMAC, 25 °C.

The reduction of the amido group of compound 5 (rat Ki 0.6 nM, human Ki 0.5 nM) led to compound 12 with a reduced affinity for rat and human receptor (rat Ki 117.4 nM, human Ki 60.7 nM). The lower affinity of 12 could be due to the introduction of a strong basic group, with concomitant loss of a potential hydrogen bond interaction of the carbonyl group and/or introduction of a tetrahedral carbon atom, which likely affects conformation. One of the key modifications made to the starting scaffold was the substitution of the amido group of compound 5 with a double bond, to obtain compound 13. The double bond restored sp2 geometry, providing high human mGlu5 binding affinity (Ki of 0.4 nM). Compound 13 emphasizes the crucial importance of conformation for achieving high mGlu5 antagonist affinity. It is well known from previous compounds that 6-methylpyridine (present in MPEP) or 3-chlorophenyl (present in fenobam) moieties are generally important for activity. This result was confirmed also in our series of compounds, where phenyl (13), 2-pyridyl (14), 6-methyl-pyridyl (15), m-methylphenyl (18),

[3H]MPEP binding at T-REx human cloned mGlu5 receptors 120

Specific binding

O

Compound

100 80 60 40 20 0 -20 -12

-11

-10

-9

-8

-7

-6

Log [M], compound 15 Figure 3. [3H]MPEP displacement of compound 15 on membranes of CHO cells stably transfected with human cloned T-REx mGlu5 receptors. The plot was obtained by three independent experiments performed in duplicate. Error bars represent ±SD. The data were analyzed by nonlinear regression analysis using GraphPad 4.0 software.

m-chlorophenyl (19) derivatives were equipotent (Table 1). On the contrary, a double meta substitution (20, 3-chloro-5-cyanophenyl) resulted in a large decrease in mGlu5 binding affinity (Ki of 283.1 nM). While compound 22, bearing a bulky 3,3-dimethyl-azetidin-2-one substituent in para-position of phenyl ring, was

3048

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

Inhibition of quisqualate-induced [Ca2+] response in CHO-T-REx cell expressing human mGlu5. 8000

RFU

6000 4000 2000 quisqualate 10 nM

0 -15 -14 -13 -12 -11 -10

-9

-8

-7

Log [M], compound 14 Inhibition of quisqualate-induced [Ca2+] response in CHO-T-REx cell expressing human mGlu5. 25000

on the piperidine nitrogen were more tolerated and could be used to fine tune affinity for the receptor. The affinity of the unsubstituted phenyl ring derivative (16, Ki of 29.2 nM) significantly increased after introduction of a hydrogen bond acceptor such as the nitro group (17, Ki of 0.3 nM). 2-Nitrophenyl (17) and 3-nitropyridine (15) showed almost identical binding affinities (Ki of 0.3 nM and 0.5 nM, respectively). Likewise, a cyano group (23) in place of the nitro group showed appreciable affinity, confirming that a hydrogen bond acceptor in this position is important. An additional phenyl ring linked to the cyanopyridine (compound 23) did not impair binding, suggesting that the binding site tolerates bulky substitutions in this position. Finally, double bond insertion on piperidine ring turned out to be important to achieve high binding affinity. Using 4-piperidone as starting material, we obtained compound 15 showing low nanomolar affinity. Using 3piperidone as intermediate we obtained the E and Z isomers (compounds 24 and 25, respectively), which were significantly less potent. The affinity of 24 (Ki of 41.8 nM), whose structure is more superimposable to that of compound 15, had 6-fold higher affinity compared to 25.

RFU

20000

2.4. Binding modes of the newly discovered NAMs and analysis of structure–activity relationships

15000 10000 5000 quisqualate 10 nM

0 -15 -14 -13 -12 -11 -10 -9

-8

-7

-6

Log [M], compound 15 Figure 4. Comparison of compound 14 and compound 15 antagonism on quisqualate-evoked [Ca2+]i in the CHO T-REx cells stably expressing human mGlu5 receptors. The plots show the concentration-dependent inhibition of quisqualate-induced mGlu5 activation, measured by Ca2+ mobilization, of compounds 14 and 15 and expressed as relative fluorescence units (RFU). The concentration of quisqualate, corresponding to the EC80 determined the same day and used for the experiments is reported. The plot was obtained by performing three independent experiments. Error bars represent ±SD. The data were analyzed by nonlinear regression analysis using GraphPad 4.0 software.

% of agonist maximal response

Quisqualate-induced [Ca2+] response in CHO-T-REx cell expressing human mGlu5. 120 quisqualate

100

+ 0.3 nM compound 15

80

+

1 nM compound 15

60 40 20 0 -20 -11 -10

-9

-8

-7

-6

-5

-4

The compounds from the newly discovered series of NAMs were docked in the validated mGlu5 homology model, assuming that they share at least a partially overlapping binding site with MPEP and fenobam, as suggested by the competition experiments performed with [3H]MPEP described above. Docking complexes were then used to rationalize structure–activity relationships. Compound 14, one of the most potent compound (Ki of 0.3 nM), was taken as representative of the series. Other compounds of the series were docked to validate the results and the interpretation of the SAR. The binding modes of the other ligands are available as supporting information (Fig. S1). Most ligands feature a propynyl benzene or propynyl pyridine moiety that resembles the corresponding moiety of MPEP. In agreement with their structural similarity, the propynyl pyridine moiety of compound 14 is positioned at the bottom of the sub-pocket, with the triple bond and the pyridine ring closely superimposed to those of MPEP (Fig. 2A). At the extracellular side, the nitropyridine ring is predicted to bind to the same region occupied by the thiazole ring of FITM in the mGlu1/FITM crystal structure (Fig. 2B). In comparison with MPEP and fenobam, compound 14 protrudes toward the extracellular side of the receptor and makes additional interactions. The nitro group establishes a hydrogen bond with N747. Molecular dynamics of the complex between mGlu5 and 14 showed that the binding mode and molecular interactions predicted by docking were mostly conserved throughout the simulation. The predicted binding modes were used to interpret and discuss structure–activity relationships within the series of compounds under investigation:

-3

Log [quisqualate] Figure 5. Progressive fold shifts showing noncompetitive antagonism mode of compound 15 at CHO T-REx mGlu5 receptors. Quisqualate-concentration responses in the absence (control) and presence of various concentrations of 15 were generated using Ca2+sensitive dye, Fluo-4 and a Fluorimetric Plate Reader (FlexStation III, molecular devices). Data were expressed as % of agonist maximal response. Each data point is the mean ±SD (bars) of three individual experiments performed in triplicate. Data were analyzed by nonlinear regression analysis using GraphPad 4.0 software.

completely inactive, the compound with a smaller and linear cyano substituent (21) was still highly potent (Ki of 0.3 nM). While modifications at the triple bond were spatially restricted, substitutions

i) The pyridine of 14 binds in proximity of A810 and tolerates meta substitutions with small groups, such as the cyano of 21, the methyl of 17 or 18, or the chlorine group of 19, without steric hindrance. On the other hand, double substitutions on the phenyl ring, such as those of compound 20 that carries both a cyano and a chlorine, caused a 1000-fold decrease in potency. Likewise, larger para substituents such as the dimethyl-azetidin-2-one moiety of compound 22 resulted in loss of binding, and in fact the compound could not be docked in the binding site because of steric repulsion. These data indicate that the SAR in this region is largely dependent on the small volume of the intracellular mGlu5 sub-pocket that accommodates this portion of the ligands.

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

ii) Affinity data indicate that the nitrogen atom of the pyridine of compound 14 is not required for binding, since 13, 6, and 5, that have a phenyl in place of the pyridine, retain high affinity. This finding is fully consistent with the fact that MPEP docked the sub-pocket with either the pyridine or the phenyl ring (Fig. 1B and C). Worth of note is that hydrogen bonds involving the pyridine nitrogen atom of 14 were not detected during molecular dynamics of the complex, further supporting the conclusion that the nitrogen atom is dispensable. iii) As mentioned above, bulky groups in place of the alkyne substituent such as those present in 8, 9, 10, 11 are detrimental for biological activity. Induced fit docking of the latter ligands required very remarkable receptor conformational changes and rotation of side chains to accommodate the ligands in the pocket, thus explaining their significantly lower affinity. Hence, both activity data and docking suggest that the triple bond is ideal for fitting the narrow cavity present in the lower end of the pocket.20,28 iv) The most active NAMs of the series, such as 14, displayed a ‘V’ shaped geometry with a nearly 120° angle formed by the two branches of the ligand, one extending toward the intracellular and the other toward the extracellular side of the receptor binding pocket. Structural modifications that significantly alter the ‘V’ shape typical of the 14 template, such as those present in 25, 8, 10, and 11, spoil the complementarity with the binding site and result in significantly lower affinity. This observation is rationalized by the docking results, in which the ‘V’ shape of the ligands nicely fits the mGlu5 binding site in the proposed binding mode as shown in Figure 2A. v) The sp2 carbon atom at the vertex of the ‘V’ shape can either be an alkene carbon (14) or a carbonyl, such as in 6 and 5. The carbonyl group is well tolerated, because it maintains the ‘V’-shaped conformation and according to our model it forms additional hydrogen bonds with Y659, T781, and S809 (Fig. S1). On the contrary, substitution of the sp2 alkene or carbonyl with a sp3 methylene carbon, such as in 12, causes a relevant decrease in potency. The resulting altered conformation of the molecule was predicted to fit less ideally in binding site. vi) The nitro group of the nitropyridine moiety is important for affinity. In fact, compound 16 is about 100 fold less potent than 17, the two molecules differing only for the presence of the nitro group. In the docking complexes the nitro group was predicted to form a hydrogen bond with N747. Moreover, we found that the nitro group could be replaced with a cyano without loss of activity, such as in 23 (Ki of 0.5 nM), and that the cyano group still formed a hydrogen bond with N747 (Fig. S1). These data suggest that a hydrogen bond acceptor in this position is important for binding affinity. Finally, compounds having a nitrobenzene in place of the nitropyridine such as compound 17 display high affinity (Ki of 0.3 nM), suggesting that the pyridine nitrogen is dispensable. In agreement with this observation, in the docking complexes this nitrogen is not involved in hydrogen bonds with the receptor. vii) Affinity data show that the nitropyridine ring can tolerate substituents such as a methyl at the 6-position (21), without loss of binding. A vinyl group in this position (7) resulted in only two-fold reduced affinity with respect to 5. Standing on the predicted binding mode of compound 14, larger groups in this position would cause steric conflicts with Y792 and M802. Position 3 is more amenable to larger substituents, since this position is directed toward the opening of the

3049

extracellular side of the pocket. In agreement with this observation, the phenyl ring of compound 23 was predicted to fit well in the binding pocket (Fig. S1).

2.5. Binding modes of representative non-aryl alkyne NAMs In order to investigate whether the generated model of the interaction of NAMs with mGlu5 could be successfully extended to include non-aryl alkyne compounds, seven known structurally-diverse and clinically important NAMs lacking the arylalkyne moiety31–37 (compounds 26–32 in Scheme 1) were docked in the homology model and the results were compared with those described for the other NAMs. Compounds were selected on the basis of structural diversity, potency, clinical relevance, and their ability to displace MPEP in competition experiments. In the docking complexes, all compounds were predicted to bind in the buried intracellular sub-pocket occupied by MPEP and fenobam (Fig. 6). Hydrogen bonds between key residues and MPEP, fenobam and the newly synthesized NAMs reported in this work were observed also for these compounds. In particular, S809 forms a hydrogen bond with the 2-methylquinoline moiety of 26 (Fig. 6A) and the oxazole moiety of 32 (Fig. 6G). S658 forms a hydrogen bond with the methylpyridine moiety of 27, 30, and 31 (Fig. 6B, E, and F, respectively). Interestingly, the flexible ethyloxy moiety of 27 (Fig. 6B) allows this compound to hydrogen bond with S658 even when the pyridine nitrogen is positioned in para instead of ortho or meta as in the other compounds. In agreement with the binding modes predicted for MPEP and fenobam, Y659 was found to interact with all investigated non-aryl alkyne NAMs. Finally, N747 establishes a hydrogen bond with compounds 29–32 (Fig. 6D–G). Interestingly, the same residue was involved in a hydrogen bond with the nitro group of compound 14 (Fig. 2A). The fact that these NAMs place a hydrogen bond acceptor in correspondence of the nitro group of 14 confirm the importance of this interaction. Overall, the results suggest that the non-aryl alkyne NAMs under investigation, although structurally diverse, are able to compete for the same site and to establish similar molecular interactions with key residues discussed above. Therefore, the proposed models can be of more general use. Very recently, the crystal structure of mGlu5 in complex with the NAM mavoglurant (methyl (3aR,4S,7aR)-4-hydroxy-4-[(3methylphenyl)ethynyl]octahydro-1H-indole-1-carboxylate) was reported (PDB code: 4OO9).20 Encouragingly, the homology model described in this work is closely superimposable with this crystal structure, root mean square deviations (RMSD) of residues 66 Å from the co-crystallized ligand and of all residues of the 7TM helices being only 0.6 Å and 0.7 Å, respectively. Superimpositions between the homology model and the crystal structure are shown in Figure S4 panels A and B. This finding confirms that the homology model based on the mGlu1 crystal structure template was of good quality. Notably, superposition of the docking complexes reported in this work with the crystal structure of the mGlu5mavoglurant complex provides further confidence on the accuracy of the predicted binding modes (Fig. S2). In fact, the ethynylmethyl-benzene moiety of mavoglurant closely matches the ethynyl-methyl-pyridine moiety of MPEP, and the same holds for the ethynyl-pyridine or ethynyl-methyl-pyridine moieties of compounds 14 and 17 (Fig. S2). Moreover, in the crystal structure the carbonyl of mavoglurant forms a hydrogen bond with N747, which is the same residue predicted to form a hydrogen bond with the nitro group of 14 and other derivatives, further supporting the importance of this interaction as discussed in the SARs. As a final validation, docking of MPEP, 5, and 14 into the newly released

3050

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

Figure 6. Predicted binding modes of compounds 26–32 (panels A–G, respectively) non-aryl alkyne NAMs found by induced fit docking.

mGlu5 crystal structure using the same docking protocol provided binding modes that were almost identical to those obtained using the homology model (RMS deviations being 0.6 Å, 1.3 Å, and 1.1 Å for MPEP, 5, and 14, respectively). Lastly, mavoglurant was docked into the homology model of mGlu5. The superposition between the docked and the x-ray orientation of mavoglurant (RMS deviation of 0.9 Å) is graphically represented in Figure 4S, to further demonstrate that the two structures closely match. 3. Conclusions In this work we described a homology model of mGlu5 built on the basis of mGlu1 template. Structural insight into the interaction of negative allosteric modulators with the transmembrane region of mGlu5 were obtained by performing an extensive docking and molecular dynamics analysis of MPEP and fenobam, two wellknown NAMs. The wealth of mutagenesis data information available in the literature for these two ligands was used to validate the models and to provide evidence that the predicted binding poses are reliable. The results were further corroborated by docking selected compounds into the recently described mGlu5 crystal structure. Moreover, a new series of potent NAMs of mGlu5 was synthesized and tested. Several structural modifications were sequentially introduced aiming at probing the effects of the substitution pattern, conformational requirements, and other structural features important for high binding affinity to the allosteric pocket. The most potent compound of the series (14) displayed a binding affinity of 0.3 nM for human mGlu5 receptor, and several other

compounds with sub-nanomolar potency and selectivity with respect to mGlu1 were reported. In order to rationalize the SAR within this new series of compounds, molecular docking studies in the validated homology model were performed using the same protocol used for MPEP and fenobam. Remarkably, the molecular interactions established by these NAMs in the transmembrane binding pocket were able to explain the observed SAR, providing useful indications for achieving high affinity. The interactions deemed important for binding of these NAMs to the mGlu5 allosteric pocket were confirmed by docking a set of chemically diverse and clinically important non alkyne-based NAMs reported in the literature, suggesting that the model of the interaction of NAMs can be of more general utility. Overall, the obtained results provide useful insights into the molecular interaction of negative allosteric modulators with mGlu5 and may facilitate the design of new modulators for this class of receptors. 4. Materials and methods 4.1. Homology modeling of mGlu5 The structure of the transmembrane domain of mGlu5 was obtained via homology modeling, using the recently determined crystal structure of mGlu119 (PDB code 4OR2) as a template. Sequence alignment of the transmembrane domains of mGlu5 and mGlu1 was made with ClustalW38 and resulted in high (77%) sequence identity, therefore accurate homology models can be obtained. Chain B of 4OR2, containing fewer missing residues,

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

was used. Homology models were built by keeping the co-crystallized ligand FITM19 as a ‘block’ residue in the binding pocket. Although the affinity of FITM is higher for mGlu1 rather than mGlu5 (micromolar affinity), the presence of a ligand with an experimentally validated binding mode is valuable, because it keeps the binding site open and accessible to ligands. Homology models were obtained with Modeller 9v11.39 A thousand homology models were produced, and the best-scoring model according to DOPE (normalized DOPE score of 0.44) was saved for visual inspection and further analysis. Five residues in the second intracellular loop (ICL2) were missing in the crystal structure due to disorder. Therefore, ICL2 residues of the best-scoring homology model were subjected to loop refinement with the loop refinement routine available in Modeller (1000 models requested), and the bestscoring model (normalized DOPE score of 0.66) was saved. This model was then refined with 10,000 steps of molecular mechanics conjugate gradient energy minimization, using Amber10.40 Minimization was performed using GBSA with the Tsui and Case parameters (igb=1),41 employing a cutoff of 18 Å for non-bonded interactions. The final model had a normalized DOPE score of 0.87, 92% of residues in most favoured regions, and 8% of residues in additional allowed regions (Ramachandran plots). 4.2. Protein and ligand preparation for docking The refined homology model was prepared for docking with the Schrödinger Protein Preparation Wizard.42,43 The structure was preprocessed to assign bond orders, to add hydrogen atoms, and to create the conserved disulfide bond between C644 and C733. Hydrogen atoms were minimized using the force field OPLS_2005, with default settings. The recently published crystal structure of mGlu5 in complex with the antagonist mavoglurant was taken from the Protein DataBank44 (PDB code: 4OO9).20 All the co-crystallized molecules present in the buffer and all the water molecules were removed from the structure, except for the buried water 4126 which participates to a hydrogen bond network in the binding site. The structure was then prepared as described above for the homology model. All the ligands were prepared for docking with LigPrep42 to obtain an initial three-dimensional conformation and to assign possible tautomers and ionization states. 4.3. Docking Molecular docking calculations were performed with the Induced Fit Docking protocol available in the Schrödinger 2014 suite, using default settings for Glide docking (receptor and ligand van der Waals scaling of 0.5, maximum number of poses 20), Prime refinement (residues within 5 Å from the ligand), and Glide redocking (redocking with SP precision structures within 30 Kcal/mol from the best scoring and within the overall top 20).42,45 For docking in the mGlu5 homology model, the grid was centered on the residues crucial for MPEP and fenobam binding as indicated by mutagenesis experiments available in the literature (i.e., P655, S658, Y659, T781, W785, F788, A810). For docking in the mGlu5 crystal structure, the grid was centered on the co-crystallized ligand mavoglurant. In all cases, up to 20 poses for each compound were generated, and a representative binding mode was selected among the best scoring solutions after visual inspection and comparison with available experimental data (mutagenesis, SAR) as described in the results section. 4.4. Molecular dynamics Energy minimization and molecular dynamics (MD) simulations were performed with the sander module of AMBER 10.40 Docking complexes investigated with molecular dynamics were

3051

initially minimized with 2000 step of molecular mechanics conjugate gradient GBSA energy minimization with the Tsui and Case parameters (igb = 1) and a cutoff of 18 Å for non-bonded interactions. 20 ns (1 ns equilibration and 19 ns production) molecular dynamics simulations were then performed on the minimized complexes by allowing the ligand and the residues within a radius of 6 Å from the ligand free to move. MD was performed at 300 K, with SHAKE turned on for bonds involving hydrogens, allowing a time-step of 2.0 fs, and a distance-dependent dielectric constant e = 4r. This set of parameters proved to perform well for MD simulations of other GPCRs.46 In the first 20 ps the system was gradually heated to 300 K, after which 1 ns of equilibration were performed. In the 19 ns production run coordinates were collected every 2 ps and then analyzed and averaged with ptraj.40 Average structures were generated for fenobam, MPEP, and 14. These structures were then minimized with 2000 steps of conjugated gradient energy minimization. The occupancy values of the hydrogen bonds formed during the 19 ns simulation were obtained with ptraj, by imposing a hydrogen bonding distance cut-off of 3.5 Å between the donor and acceptor heavy atoms, and an angle cut-off of P120°. Convergence of the MD simulation was monitored by collecting root mean squared deviation (RMSD) values with respect to the minimized docking complex. This analysis showed that the simulation converged after 1 ns with RMSD values that always remained below 2 Å. RMSD fluctuation graphs, average values, and standard deviations are reported as Supplementary data (Fig. S5 and Table S1). 4.5. Chemical synthesis and analytical procedures All reactions were carried out employing standard chemical techniques. Unless otherwise noted, reactions were carried out in anhydrous solvents, under N2 atmosphere. Solvents for extraction, washing, and chromatography were HPLC grade and were used without further purification. Petroleum ether (PE) is a mixture of low boiling hydrocarbon (>90% between 40 and 60 °C). All NMR spectra were recorded on a 200 MHz Bruker or a 400 MHz Bruker Avance Spectrometers. 1 H NMR chemical shifts were reported as d (ppm) relative to residual solvent peaks. Low-resolution mass spectra were obtained on a 2525 Waters pump fitted with a CFO and a 2767 fraction collector/auto injector, the flow was split at the exit of the column to a 2996 DAD and to a Micromass ZQ detector fitted with an electrospray source, and equipped with a Xterra MS C18 4.6  50 mm, running a gradient of 20–100% acetonitrile in a buffer solution of 20 mM NH4HCO3 at pH8. Analytical thin-layer chromatography was carried out on pre-coated silica gel 60 F254 Merck plates. Flash chromatography was performed using silica gel 60 (35– 70 lm) purchased by Merck, or using Biotage SP1Ò or Biotage HorizonÒ with related flash chromatography cartridges. The purity of the biologically tested compounds was determined by an analytical HPLC method and was found to be greater than or equal to 95%. The analytical HPLC were performed on a Waters instrument comprising a separation module (Waters 2695), an automatic injector, a photodiode array detector (Waters 2996), and a software system controller (Empower). The reverse-phase column Xterra C18 4.6  100 mm was used at flow rate of 1.3 mL/min and UV detection at 220 nm, running a gradient of 20–100% acetonitrile in a buffer solution of 20 mM NH4HCO3 at pH 8. Reagents were purchased from commercial sources at the highest commercial quality. Compounds 1 (MPEP hydrochloride), 2 (MTEP hydrochloride) and 3 (fenobam) were purchased from Tocris Bioscience. The compound R21412747 was synthesized by Recordati Drug Discovery Department following the reported procedures. [3H]-MPEP and [3H]-R214127 were obtained from American Radiolabelled Chemicals Inc. (Saint Louis, MO, USA).

3052

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

4.5.1. 5-[(2-Methyl-1,3-thiazol-4-yl)ethynyl]-2,30 -bipyridine (4) 4.5.1.1. 1-Chloro-4-trimethylsilylbut-3-yn-2-one (4a). A solution of bis(trimethylsilyl)acetylene (4 g, 23.5 mmol) and 2chloroacetyl chloride (2.06 mL, 25.9 mmol) in 10 mL of DCM was added dropwise to a cooled suspension of AlCl3 (4.39 g, 32.9 mmol) in 30 mL of DCM, keeping the temperature below 5 °C. The reaction mixture was stirred for 1 h at 0 °C; afterwards it was kept under stirring at 25 °C for 1 h. The reaction was quenched with 60 mL of aqueous 1 M HCl. The aqueous layer was extracted with DCM and the combined organic phases were washed sequentially with aqueous NaHCO3, H2O, a saturated solution of NaCl, dried over anhydrous Na2SO4 and evaporated to dryness to afford 2.6 g of a red crude oil which was distilled under vacuum at 62 °C yielding 2.07 g (50%) of compound 4a as pale oil. 1H NMR (400 MHz, CDCl3) d: 4.24 (s, 2H), 0.29 (s, 9H). 4.5.1.2. 2-Methyl-4-[(trimethylsilyl)ethynyl]-1,3-thiazole (4b). To a solution of 4a (2 g, 11.4 mmol) in 20 mL of DMF was added thioacetamide (1.1 g, 14.8 mmol). The reaction mixture was stirred at 25 °C overnight. Afterwards, it was diluted with EtOAc and washed with H2O. The organic layer was dried over anhydrous Na2SO4, filtered and evaporated to dryness. The crude was purified via flash chromatography over silica gel eluting with EtOAc/PE (2:98). The combined fractions were evaporated to afford 1.5 g of compound 4b. Yield: 67%. 1H NMR (200 MHz, CDCl3) d: 7.33 (s, 1H), 2.71 (s, 3H), 0.25 (s, 9H). MS (ESI) m/z 196.12 [M+H]+. HPLC 92.3% (UV). 4.5.1.3. 2-Chloro-5-[(2-methyl-1,3-thiazol-4-yl)ethynyl]pyridine (4c). A mixture of compound 4b (700 mg, 3.58 mmol), 2chloro-5-iodopyridine (857 mg, 3.58 mmol), CuI (136 mg, 0.7 mmol), TBAF.3H2O (1 g, 3.58 mmol), tetrakis(triphenylphosphine)palladium (Pd(PPh3)4), 827 mg, 0.7 mmol), TEA (998 lL, 7.16 mmol) in 20 mL of DMF was stirred at 70 °C for 5 h. The reaction was cooled at 25 °C, diluted with EtOAc and washed with H2O. The organic layer was dried over anhydrous Na2SO4, filtered and evaporated to dryness. The crude was purified via flash chromatography eluting with PE/EtOAc 3:7. The combined fractions were evaporated to afford 560 mg of compound 4c. Yield: 67%. 1H NMR (400 MHz, CDCl3) d: 8.59 (d, J = 2.30 Hz, 1H), 7.81 (dd, J = 2.31, 8.27 Hz, 1H), 7.46 (s, 1H), 7.35 (d, J = 8.26 Hz, 1H), 2.78 (s, 3H). MS (ESI) m/z 235.20 [M+H]+. HPLC 90.0% (UV). 4.5.1.4. 5-[(2-Methyl-1,3-thiazol-4-yl)ethynyl]-2,30 -bipyridine (4). A mixture of compound 4c (560 mg, 2.39 mmol), 3pyridylboronic acid (323 mg, 2.63 mmol), PdCl2(PPh3)2 (336 mg, 0.479 mmol), K2CO3 (336 mg, 2.39 mmol) in 20 mL of DME and 5 mL of H2O was stirred at 80 °C for 8 h, then it was cooled at 25 °C, diluted with H2O and extracted with EtOAc. The combined organic layers were washed with H2O, dried over anhydrous Na2SO4, filtered and evaporated to dryness. The crude was purified via flash chromatography eluting with 1.4 N NH3 solution in MeOH/EtOAc gradient from 1:99 to 2:98. The collected combined fractions were evaporated to afford 38 mg of compound 4. Yield: 5.7%. 1H NMR (400 MHz, CDCl3) d: 9.25 (d, J = 2.3 Hz, 1H), 8.90 (dd, J = 0.9, 2.1 Hz, 1H), 8.69 (dd, J = 1.6, 4.8 Hz, 1H), 8.39 (td, J = 2.3, 8.0 Hz, 1H), 7.96 (dd, J = 2.1, 8.2 Hz, 1H), 7.79 (d, J = 8.2 Hz, 1H), 7.48 (s, 1H), 7.43–7.48 (m, 1H), 2.79 (s, 3H). MS (ESI) m/z 278.03 [M+H]+. HPLC 95.5% (UV). 4.5.2. 1-[4-(3-Nitropyridin-2-yl)piperazin-1-yl]-3-phenylprop2-yn-1-one (5) To a solution of 3-phenylpropiolic acid (250 mg, 1.71 mmol) in anhydrous DMF (20 mL) stirred at 0 °C was added EDC (361 mg, 1.88 mmol) and HOBT (259 mg, 1.88 mmol). The mixture was

stirred at 0 °C for 3 h. Afterwards, 1-(3-nitro-2-pyridyl)piperazine (392 mg, 1.88 mmol) was added and the reaction mixture was stirred at 0 °C for 1 h followed by 2 h at 25 °C. After overnight resting, it was poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with H2O, dried (Na2SO4) and evaporated to dryness in vacuo. The crude was purified by flash chromatography eluting with EtOAc/PE gradient from 4:6 to 6:4 affording 345 mg of the title compound. Yield: 60%. 1H NMR (400 MHz, CDCl3) d: 8.40 (dd, J = 1.7, 4.6 Hz, 1H), 8.21 (dd, J = 1.7, 8.0 Hz, 1H), 7.56–7.65 (m, 2H), 7.36–7.51 (m, 3H), 6.88 (dd, J = 4.6, 8.0 Hz, 1H), 3.97–4.08 (m, 2H), 3.82–3.92 (m, 2H), 3.49– 3.61 (m, 4H). MS (ESI) m/z 337.16 [M+H]+. HPLC 98.7% (UV). 4.5.3. 1-[4-(2-Nitrophenyl)piperazin-1-yl]-3-phenylprop-2-yn1-one (6) To a solution of 3-phenylpropiolic acid (60 mg, 0.411 mmol) and 1-(2-nitrophenyl)piperazine (85.2 mg, 0.411 mmol) in 3 mL of DMF was added diethyl cyanophosphonate (DEPC, 0.074 mL, 1.1 mmol) and TEA (0.063 mL, 0.452 mmol) stirring at 0 °C. After 5 min the cooling bath was removed and the solution was stirred at 25 °C for 2 h. After overnight resting, the reaction mixture was quenched with H2O, alkalinized with 3 N NaOH to pH 9–10, extracted with EtOAc (3). The crude was purified by flash chromatography eluting with EtOAc/PE 4:6 to afford 77 mg of compound 6 as an orange solid. Yield: 56%. 1H NMR (400 MHz, CDCl3) d: 7.84 (d, J = 8.0 Hz, 1H), 7.52–7.63 (m, 3H), 7.36–7.50 (m, 3H), 7.20 (d, J = 8.0 Hz, 1H), 7.12–7.19 (m, 1H), 3.98–4.11 (m, 2H), 3.84–3.93 (m, 2H), 3.07–3.24 (m, 4H). MS (ESI) m/z 336.14 [M+H]+. HPLC 98.9% (UV). 4.5.4. 1-[4-(6-Ethenyl-3-nitropyridin-2-yl)piperazin-1-yl]-3phenylprop-2-yn-1-one (7) 4.5.3.1. tert-Butyl 4-(6-chloro-3-nitropyridin-2-yl)piperazine-1carboxylate (7a). A solution of 92% 2,6-dichloro-3-nitropyridine (2 g, 9.53 mmol), 1-tert-butoxycarbonylpiperazine (1.86 g,10.1 mmol), 10 mL of 98% diisopropylethylamine (DIPEA, 57.2 mmol) in 50 mL of toluene was stirred at 25 °C overnight. Afterwards, it was poured into H2O, extracted three times with EtOAc and the combined extracts washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo. The crude was purified by flash chromatography eluting with EtOAc/PE 1:9 to afford 2.53 g of 7a as yellow solid. Yield: 77%. 1H NMR (400 MHz, CDCl3) d: 8.13 (d, J = 8.3 Hz, 1H), 6.75 (d, J = 8.3 Hz, 1H), 3.55–3.65 (m, 4H), 3.41–3.52 (m, 4H), 1.50 (s, 9H). MS (ESI) m/z 343.19 [M+H]+. HPLC 95.2% (UV). 4.5.3.2. tert-Butyl 4-(6-ethenyl-3-nitropyridin-2-yl)piperazine1-carboxylate (7b). To a solution of 7a (567 mg, 1.65 mmol) in 38.6 mL of DMF was added 97% tributylvinyltin (0.748 mL, 2.48 mmol), PdCl2(PPh3)2 (0.386 mg, 1.75 mmol), 2,6-ditertbutyl4-methylphenol (386 mg, 1.75 mmol). The reaction mixture was heated at 50 °C under inert gas atmosphere for 6 h until LC–MS showed complete conversion of the starting material. Dilution with Et2O, washing the organic layer with H2O, drying over anhydrous Na2SO4 and evaporation afforded a crude, which was purified by flash chromatography eluting with EtOAc/PE 5:95 to afford 529 mg of 7b. Yield: 96%. 1H NMR (400 MHz, CDCl3) d: 8.18 (d, J = 8.2 Hz, 1H), 6.80 (d, J = 8.2 Hz, 1H), 6.73 (dd, J = 10.5, 17.3 Hz, 1H), 6.33 (dd, J = 1.4, 17.3 Hz, 1H), 5.63 (dd, J = 1.4, 10.5 Hz, 1H), 3.56–3.66 (m, 4H), 3.43–3.53 (m, 4H), 1.50 (s, 9H). MS (ESI) m/z 335.19 [M+H]+. HPLC 96.4% (UV). 4.5.3.3. 1-[4-(6-Ethenyl-3-nitropyridin-2-yl)piperazin-1-yl]-3To a solution of 7b (150 mg, phenylprop-2-yn-1-one (7). 0.449 mmol) in 3 mL of DCM was added TFA (0.346 mL,

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

4.49 mmol). After overnight stirring at 25 °C, the solvent was repeatedly evaporated adding over time DCM. The resulting yellow oil was dissolved in DMF (5 mL) and slowly dripped into a cooled solution of phenylpropiolic acid (98.4 mg, 0.674 mmol), DEPC (0.116 mL, 0.763 mmol) and TEA (0.376 mL, 2.69 mmol) in 5 mL of DMF. Stirring was continued at 25 °C for 4 h. Afterwards, H2O was added and extraction with EtOAc (3) was carried out. The combined extracts were washed with brine, dried over anhydrous Na2SO4, evaporated to dryness in vacuo to afford 170 mg of yellow crude, which was purified by preparative HPLC giving 36 mg of 7 as yellow solid. Yield: 22%. 1H NMR (400 MHz, CDCl3) d: 8.23 (d, J = 8.2 Hz, 1H), 7.55–7.67 (m, 2H), 7.36–7.51 (m, 3H), 6.86 (d, J = 8.2 Hz, 1H), 6.74 (dd, J = 10.5, 17.3 Hz, 1H), 6.35 (dd, J = 1.8, 17.3 Hz, 1H), 5.66 (dd, J = 1.8, 10.5 Hz, 1H), 4.03 (m, 2H), 3.88 (m, 2H), 3.50–3.66 (m, 4H). MS (ESI) m/z 363.3 [M+H]+. HPLC 97.3% (UV). 4.5.5. [4-(3-Nitro-2-pyridyl)piperazin-1-yl]quinoxalin-2-ylmethanone (8) To a suspension of quinoxaline-2-carboxylic acid (48.2 mg, 0.25 mmol), 0.02 mL of DMF in 5 mL of DCM stirred at 25 °C was added SOCl2 (0.05 mL, 0.688 mmol) and the reaction mixture was stirred at 25 °C for 2 h. After overnight resting, TEA (0.3 mL, 2.09 mmol) was dripped into the solution, followed by a solution of 1-(3-nitro-2-pyridyl)piperazine (52.1 mg, 0.25 mmol) in 5 mL of DCM. After overnight stirring, the mixture was diluted with DCM, washed with 1 N NaOH, H2O, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo giving a brown oil. This oil was purified by flash chromatography eluting with 3.2 N NH3 in MeOH/EtOAc/PE 2:20:80 affording 82 mg of 8 as orange oil. Yield: 90%. 1H NMR (400 MHz, CDCl3) d: 9.27 (s, 1H), 8.38–8.47 (m, 1H), 8.07–8.26 (m, 3H), 7.82–7.95 (m, 2H), 6.84–6.91 (m, 1H), 3.98–4.13 (m, 4H), 3.68–3.75 (m, 2H), 3.56–3.63 (m, 2H). MS (ESI) m/z 365.2 [M+H]+. HPLC 98.3% (UV). 4.5.6. [4-(3-Nitro-2-pyridyl)piperazin-1-yl]-(4-phenylphenyl)methanone (9) To a solution of 4-biphenylcarboxylic acid (50 mg, 0.252 mmol) in 5 mL of DMF was added HOBt (55 mg, 0.399 mmol) and EDC (75 mg, 0.391 mmol) and the mixture was stirred at 0 °C for 3 h. 1-(3-Nitro-2-pyridyl)piperazine (67.2 mg, 0.323 mmol) was added. After overnight stirring, the reaction mixture was poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with H2O, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo giving a brown oil. This oil was purified by flash chromatography (HorizonÒ Biotage) eluting with PE/diethyl ether 4:6 affording 13.5 mg of 9. Yield: 14%. 1H NMR (400 MHz, CDCl3) d: 8.40 (dd, J = 1.7, 4.5 Hz, 1H), 8.20 (dd, J = 1.7, 8.0 Hz, 1H), 7.65– 7.70 (m, 2H), 7.60–7.65 (m, 2H), 7.52–7.57 (m, 2H), 7.46–7.52 (m, 2H), 7.38–7.44 (m, 1H), 6.86 (dd, J = 4.5, 8.0 Hz, 1H), 3.25– 4.17 (m, 8H). MS (ESI) m/z 389.3 [M+H]+. HPLC 98% (UV). 4.5.7. [4-(6-Methyl-3-nitro-2-pyridyl)piperazin-1-yl]-(3-phenylphenyl)methanone (10) To a solution of 3-biphenylcarboxylic acid (150 mg, 0.757 mmol) and 1-(6-methyl-3-nitropyridin-2-yl)piperazine (185 mg, 0.832 mmol) in DMF (8 mL) was added DEPC (0.126 mL, 0.833 mmol), TEA (0.127 mL, 0.908 mmol) and the resulting mixture was stirred at 25 °C for 3 h. Afterwards, it was poured into H2O and extracted with EtOAc (3). The combined organic layers were dried over anhydrous Na2SO4, filtered and evaporated to dryness in vacuo. The crude was purified by flash chromatography (HorizonÒ Biotage) eluting with EtOAc/PE 3:7 affording 285 mg of 10 as yellow solid. Yield: 94%. 1H NMR (400 MHz, CDCl3) d: 8.14 (d, J = 8.2 Hz, 1H), 7.66–7.72 (m, 2H), 7.59–7.64 (m, 2H), 7.45–7.56 (m, 3H), 7.37–7.44 (m, 2H), 6.70 (d, J = 8.2 Hz, 1H),

3053

3.27–4.22 (m, 8H), 2.52 (s, 3H). MS (ESI) m/z 403.04 [M+H]+. HPLC 99.0% (UV). 4.5.8. [4-(3-Nitro-2-pyridyl)piperazin-1-yl]-(2-phenylphenyl)methanone (11) Compound 11 was synthesised following the method described for compound 9 but using 2-biphenylcarboxylic acid (100 mg, 0.504 mmol). After the usual work-up procedure the crude was purified by flash chromatography eluting with PE/EtOAc 3:7 to afford 103 mg of 11. Yield: 53%. 1H NMR (400 MHz, CDCl3) d: 8.26–8.35 (m, 1H), 8.08–8.15 (m, 1H), 7.42–7.57 (m, 8H), 7.35– 7.41 (m, 1H), 6.78 (dd, J = 4.4, 8.3 Hz, 1H), 3.83–3.98 (m, 1H), 3.46–3.64 (m, 2H), 2.99–3.18 (m, 3H), 2.87–2.99 (m, 1H), 2.17– 2.31 (m, 1H). MS (ESI) m/z 389.12 [M+H]+. HPLC 99.4% (UV). 4.5.9. 1-(3-Nitropyridin-2-yl)-4-(3-phenylprop-2-yn-1-yl)piperazine (12) To a solution of ethynylbenzene (0.22 mL, 2 mmol), 35% aqueous formaldehyde (0.8 mL, 24.8 mmol), 1-(3-nitro-2-pyridyl)piperazine (458 mg, 2.2 mmol) in 4 mL of DMSO was added CuI (8 mg) and the resulting mixture was stirred at 25 °C for 2 h. After overnight resting, the reaction mixture was poured into H2O, extracted with EtOAc (3) and the combined extracts washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo. The crude was purified by flash chromatography eluting with EtOAc/PE 1:95 to afford 645 mg of 12. Yield: 99%. 1H NMR (400 MHz, CDCl3) d: 8.35 (dd, J = 1.7, 4.5 Hz, 1H), 8.14 (dd, J = 1.7, 8.0 Hz, 1H), 7.42–7.55 (m, 2H), 7.29–7.36 (m, 3H), 6.77 (dd, J = 4.5, 8.0 Hz, 1H), 3.62 (s, 2H), 3.58 (t, J = 4.9 Hz, 4H), 2.79 (t, J = 4.9 Hz, 4H). MS (ESI) m/z 323.15 [M+H]+. HPLC 98.8% (UV). 4.5.10. 3-Nitro-2-[4-(3-phenylprop-2-yn-1-ylidene)piperidin-1yl]pyridine (13) 4.5.10.1. Diethyl (3-trimethylsilylprop-2-ynyl)phosphonate (13a). To a solution of LiHMDS (1 M in THF, 63.8 mL, 63.8 mmol) in anhydrous THF (162 mL) under stirring at 10 °C in nitrogen atmosphere, diethyl phosphite (7.4 mL, 63.8 mmol) was added dropwise. The obtained solution was stirred at the same temperature for 20 min. Afterwards, 3-bromo-1-trimethylsilyl-1propyne (10 mL, 63.8 mmol) was dripped into the reaction mixture and was stirred at 10 °C for 2 h, then quenched with H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried on Na2SO4 and evaporated to dryness in vacuo to afford 14.86 g of 13a. MS (ESI) m/z 266.15 [M+NH4]+. 4.5.10.2. 3-Nitro-2-{4-[3-(trimethylsilyl)prop-2-yn-1-ylidene]piperidin-1-yl}pyridine (13b). Into a solution of 13a (0.68 g, 2.74 mmol) in anhydrous THF (15 mL) stirred at –60 °C under N2 stream, was added dropwise a solution of LiHMDS (1 M in THF, 2.74 mL, 2.74 mmol) and the mixture was stirred at 60 °C for 15 min. To the resulting solution was added dropwise a solution of 1-(3-nitropyridin-2-yl)piperidin-4-one (0.55 g, 2.49 mmol) in anhydrous THF (12 mL). The reaction mixture was stirred at –60° for 15 min; then it was allowed to warm up to 25 °C over 2 h. Afterwards, it was quenched with H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford 0.79 g of the title product that was used in the following step without any further purification. MS (ESI) m/z 316.16 [M+H]+. 4.5.10.3. 3-Nitro-2-[4-(prop-2-yn-1-ylidene)piperidin-1-yl]pyriA solution of compound 13b (0.57 g, 1.81 mmol), dine (13c). TBAF.3H2O (0.57 g, 2.03 mmol) in 38 mL of THF was stirred at 25 °C for 2 h. The reaction mixture was poured into H2O and extracted with EtOAc. The combined organic layers were washed

3054

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 1:9) giving the title product (0.21 g) used in the following step without any further purification. MS (ESI) m/z 244.13 [M+H]+. 4.5.10.4. 3-Nitro-2-[4-(3-phenylprop-2-yn-1-ylidene)piperidin1-yl]pyridine (13). A mixture of compound 13c (200 mg, 0.822 mmol), iodobenzene (103 lL, 0.92 mmol), TEA (20 mL), Pd(PPh3)4 and CuI (15.6 mg, 0.082 mmol) was stirred at 80 °C in a closed vessel for 2 h. Afterwards, it was poured into H2O, extracted with EtOAc (3), dried over anhydrous Na2SO4 and purified by flash chromatography of the crude with EtOAc/PE 35:65 affording149 mg of the title product. Yield: 57%. 1H NMR (400 MHz, CDCl3) d: 8.38 (d, J = 4.6 Hz, 1H), 8.19 (d, J = 7.8 Hz, 1H), 7.43–7.47 (m, 2H), 7.30–7.36 (m, 3H), 6.80 (dd, J = 4.6, 7.8 Hz, 1H), 5.65 (s, 1H), 3.52–3.60 (m, 4H), 2.76–2.81 (m, 2H), 2.50–2.55 (m, 2H). MS (ESI) m/z 320.24 [M+H]+. HPLC 95.3% (UV). 4.5.11. 3-Nitro-2-{4-[3-(pyridin-2-yl)prop-2-yn-1-ylidene]piperidin-1-yl}pyridine (14) The title compound was obtained as described for compound 13, reacting compound 13c (93 mg, 0.38 mmol) with 2-iodopyridine (61 lL, 0.57 mmol) instead of iodobenzene. After the usual work-up, the crude was purified by flash chromatography (EtOAc/PE 1:1) affording the title compound (93 mg). Yield: 76%. 1 H NMR (400 MHz, CDCl3) d: 8.60–8.65 (m, 1H), 8.36 (dd, J = 1.7, 4.2 Hz, 1H), 8.16 (dd, J = 7.8 Hz, 1.7 Hz, 1H), 7.81 (t, J = 4.2 Hz, 1H), 7.48 (d, J = 7.84 Hz, 1H), 7.27–7.32 (m, 1H), 6.78 (dd, J = 4.6, 7.8 Hz, 1H), 5.68 (s, 1H), 3.50–3.60 (m, 4H), 2.80–2.85 (m, 2H), 2.50–2.55 (m, 2H). MS (ESI) m/z 321.10 [M+H]+. HPLC 96.8% (UV). 4.5.12. 2-{4-[3-(6-Methylpyridin-2-yl)prop-2-yn-1-ylidene]piperidin-1-yl}-3-nitropyridine (15) A mixture of compound 13c (0.21 g, 0.86 mmol), 2-bromo-6methylpyridine (0.11 mL, 0.95 mmol), Pd(PPh3)4 (70 mg, 0.06 mmol), CuI (16 mg, 0.09 mmol) in anhydrous and degassed TEA (10 mL) was heated at 80 °C under a nitrogen atmosphere for 2 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 3.5:6.5) affording the title product (0.20 g). Yield: 69%. 1H NMR (400 MHz, CDCl3) d: 8.35– 8.38 (m, 1H), 8.14–8.19 (m, 1H), 7.55 (t, J = 7.5 Hz, 1H), 7.24–7.29 (m, 1H), 7.09 (d, J = 7.5 Hz, 1H), 6.75–6.80 (m, 1H), 5.66 (s, 1H), 3.50–3.58 (m, 4H), 2.80–2.85 (m, 2H), 2.61 (s, 3H), 2.48–2.55 (m, 2H). MS (ESI) m/z 335.12 [M+H]+. HPLC 97.4% (UV). 4.5.13. 2-Methyl-6-[3-(1-phenylpiperidin-4-ylidene)prop-1-yn1-yl]pyridine (16) 4.5.13.1. tert-Butyl 4-[3-(trimethylsilyl)prop-2-yn-1-ylidene]piperidine-1-carboxylate (16a). To a solution of 13a (4.861 mg, 19.573 mmol) in dry THF (60 mL) stirred at 60 °C under nitrogen atmosphere, was added dropwise LiHMDS as a 1 M solution in THF/ethylbenzene (1.3 equiv, 19.6 mL) and the resulting mixture was stirred at 60 °C for 1 h. A solution of tertbutyl 4-oxopiperidine-1-carboxylate (3 g, 15.06 mmol) in dry THF (20 mL) was added dropwise and the mixture was stirred at 60 °C for 20 min, then heated to 25 °C over 2 h, quenched with H2O and extracted with EtOAc (3). The organic layers were dried over anhydrous Na2SO4 and the solvent was evaporated to dryness to give a crude product which was purified via automated flash chromatography (Biotage SP1Ò, SNAP100 Cartridge), eluting with a EtOAc/PE gradient from 2:98 to 15:85 to give 4.08 g of 16a as a white powder. Yield: 92%. 1H NMR (400 MHz, CDCl3) d: 5.40 (s,

1H), 3.40–3.51 (m, 4H), 2.48–2.53 (m, 2H), 2.21–2.27 (m, 2H), 1.50 (s, 9H), 0.21 (s, 9H). MS (ESI) m/z 294.29 [M+H]+. 4.5.13.2. tert-Butyl 4-prop-2-ynylidenepiperidine-1-carboxylate (16b). A solution of 16a (1 g, 3.41 mmol), TBAF.1H2O (993 mg, 3.56 mmol) in 30 mL of THF was stirred at 25 °C for 2 h. Afterwards, it was poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried (Na2SO4) and the solvent was evaporated to dryness in vacuo to give a crude which was purified by flash chromatography eluting with EtOAc/PE 1:9 affording 900 mg of the title compound. 1H NMR (400 MHz, CDCl3) d: 5.48 (s, 1H), 3.40–3.51 (m, 4H), 3.02 (s, 1H), 2.48–2.53 (m, 2H), 2.20–2.27 (m, 2H), 1.50 (s, 9H). MS (ESI) m/z 222.23 [M+H]+. 4.5.13.3. tert-Butyl 4-[3-(6-methylpyridin-2-yl)prop-2-yn-1-yliThe title compound dene]piperidine-1-carboxylate (16c). was obtained following the procedure described for compound 15 but starting from compound 16b instead of compound 13c. After the usual work-up procedure, evaporation of the combined EtOAc extracts afforded a crude which was purified by flash chromatography (EtOAc/PE 3.5:6.5) affording the title product. 1H NMR (400 MHz, CDCl3) d: 7.55 (t, J = 7.5 Hz, 1H), 7.26 (d, 7.5 Hz, 1H), 7.09 (d, J = 7.5 Hz, 1H), 5.60 (s, 1H), 3.45–3.55 (m, 4H), 2.59–2.65 (m, 2H), 2.58 (s, 3H), 2.29–2.35 (m, 2H), 1.50 (s, 9H). MS (ESI) m/z 313.27 [M+H]+. HPLC 93.1% (UV). 4.5.13.4. 2-Methyl-6-[3-(piperidin-4-ylidene)prop-1-yn-1yl]pyridine (16d). To a solution of compound 16c (17 g, 54.4 mmol) in CHCl3 (840 mL), TFA (60 mL, 779 mmol) was added and the reaction mixture was stirred at 70 °C until the complete conversion of the reagent was observed by LC–MS (about 15 min). After cooling to 25 °C, H2O was added followed by aq. NaOH (2 N) to make the pH alkaline. Separation of the organic layer and extraction of the aqueous layer with DCM, washing with brine and drying over anhydrous Na2SO4 afforded 11.6 g of the title compound which was used in the following step without any further purification. 1H NMR (400 MHz, CDCl3) d: 7.53 (t, J = 7.5 Hz, 1H), 7.24 (d, 7.5 Hz, 2H), 7.07 (d, J = 7.5 Hz, 1H), 5.52 (s, 1H), 2.91–2.99 (m, 4H), 2.59–2.65 (m, 2H), 2.58 (s, 3H), 2.29–2.35 (m, 2H). MS (ESI) m/z 213.25 [M+H]+. HPLC 98.1% (UV). 4.5.13.5. 2-Methyl-6-[3-(1-phenylpiperidin-4-ylidene)prop-1yn-1-yl]pyridine (16). A mixture of compound 16d (0.22 g, 1.04 mmol), bromobenzene (0.17 g, 1.04 mmol), Cs2CO3 (0.68 g, 2.1 mmol), BINAP (0.031 g, 0.05 mmol), palladium(II)acetate (0.01 mg, 0.05 mmol), in anhydrous and degassed toluene (10 mL) was heated at 110 °C under a nitrogen atmosphere for 12 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted with EtOAc. The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 2:8) affording 40 mg of the title compound. Yield: 14%. 1H NMR (400 MHz, CDCl3) d: 7.58 (t, J = 7.6 Hz, 1H), 7.32 (d, J = 7.6 Hz, 1H), 6.8–7.20 (m, 6H), 5.63 (s, 1H), 3.38 (br d, 4H), 2.86 (br d, 2H), 2.40–2.65 (m, 5H). MS (ESI) m/z 289.3 [M+H]+. HPLC 95.0% (UV). 4.5.14. 2-Methyl-6-{3-[1-(2-nitrophenyl)piperidin-4-ylidene]prop-1-yn-1-yl}pyridine (17) A well homogenised mixture of 16d (20 mg, 0.09 mmol) and 1bromo-2-nitrobenzene (22.8 mg, 0.11 mmol) was stirred at 90 °C for 0.5 h, then at 110 °C for 1 h. The reaction crude was purified by flash chromatography (EtOAc/PE gradient from 3:7 to 4:6) affording 8.3 mg of the title product. Yield: 27.6%. 1H NMR (400 MHz, CDCl3) d: 7.81 (d, J = 7.5 Hz, 1H), 7.52–7.64 (m, 1H),

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

7.48 (t, J = 7.5 Hz, 1H), 7.26–7.30 (m, 1H), 7.19 (d, J = 7.5 Hz, 1H), 7.03–7.16 (m, 2H), 5.62 (s, 1H), 3.16–3.21 (m, 2H), 3.11–3.16 (m, 2H), 2.82–2.87 (m, 2H), 2.61 (s, 3H), 2.50–2.55 (m, 2H). MS (ESI) m/z 334.30 [M+H]+. HPLC 99.4% (UV). 4.5.15. 2-{4-[3-(3-Methylphenyl)prop-2-yn-1-ylidene]piperidin1-yl}-3-nitropyridine (18) A mixture of compound 13c (66 mg, 0.271 mmol), 3-iodotoluene (36.8 lL, 0.285 mmol), PdCl2(PPh3)2 (9.51 mg, 0.014 mmol), CuI (6.16 mg, 0.027 mmol) in anhydrous and degassed TEA (3 mL) was heated at 80 °C under a nitrogen atmosphere for 2 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 1:9) affording 55 mg of the title compound. Yield: 69%. 1H NMR (400 MHz, CDCl3) d: 8.33–8.42 (m, 1H), 8.14–8.23 (m, 1H), 7.19–7.30 (m, 3H), 7.13 (d, J = 8.1 Hz, 1H), 6.74–6.84 (m, 1H), 5.64 (s, 1H), 3.49–3.61 (m, 4H), 2.77 (t, J = 5.7 Hz, 2H), 2.50 (t, J = 5.7 Hz, 2H), 2.36 (s, 3H). MS (ESI) m/z 334.30 [M+H]+. HPLC 95.0% (UV). 4.5.16. 2-[4-[3-(3-Chlorophenyl)prop-2-ynylidene]-1-piperidyl]3-nitro-pyridine (19) A mixture of 13c (56 mg, 0.23 mmol), 1-chloro-3-iodobenzene (30.5 lL, 0.242 mmol), PdCl2(PPh3)2 (8.07 mg, 0.012 mmol), CuI (4.38 mg, 0.023 mmol) in anhydrous and degassed TEA (3 mL) was heated at 80 °C under a nitrogen atmosphere for 2 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 5:95) affording 77 mg of the title compound. Yield: 95%. 1H NMR (400 MHz, CDCl3) d: 8.38 (dd, J = 1.7, 4.5 Hz, 1H), 8.18 (dd, J = 1.7, 8.0 Hz, 1H), 7.44 (t, J = 1.6 Hz, 1H), 7.20–7.36 (m, 3H), 6.80 (dd, J = 4.5, 8.0 Hz, 1H), 5.63 (s, 1H), 3.43–3.68 (m, 4H), 2.76 (t, J = 5.8 Hz, 2H), 2.51 (t, J = 5.8 Hz, 2H). MS (ESI) m/z 354.2 [M+H]+. HPLC 95.7% (UV). 4.5.17. 3-Chloro-5-[3-[1-(3-nitro-2-pyridyl)-4-piperidylidene]prop-1-ynyl]benzonitrile (20) A mixture of 13c (60 mg, 0.247 mmol), 3-chloro-5-iodobenzonitrile (65.1 mg, 0.247 mmol), PdCl2(PPh3)2 (8.68 mg, 0.0124 mmol), CuI (4.7 mg, 0.0247 mmol) in anhydrous and degassed TEA (3.6 mL) was heated at 80 °C under a nitrogen atmosphere for 2 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography (EtOAc/PE 1:9) affording 68 mg of compound 20. Yield: 73%. 1H NMR (400 MHz, CDCl3) d: 8.38 (dd, J = 1.9, 4.5 Hz, 1H), 8.17 (dd, J = 1.9, 8.0 Hz, 1H), 7.63 (t, J = 1.6 Hz, 1H), 7.60 (t, J = 1.6 Hz, 1H), 7.56 (t, J = 1.6 Hz, 1H), 6.81 (dd, J = 4.5, 8.0 Hz, 1H), 5.63 (s, 1H), 3.46–3.66 (m, 4H), 2.75 (t, J = 5.8 Hz, 2H), 2.52 (t, J = 5.8 Hz, 2H). MS (ESI) m/z 379.10 [M+H]+. HPLC 98.5% (UV). 4.5.18. 6-[3-[1-(6-Methyl-3-nitro-2-pyridyl)-4-piperidylidene]prop-1-ynyl]pyridine-3-carbonitrile (21) 4.5.18.1. tert-Butyl 4-[3-(5-cyano-2-pyridyl)prop-2-ynylidene]piperidine-1-carboxylate (21a). A mixture of 16b (100 mg, 0.452 mmol), 2-chloro-5-cyanopyridine (76.7 mg, 0.542 mmol), PdCl2(PPh3)2 (15.9 mg, 0.023 mmol), CuI (8.61 mg, 0.045 mmol) in anhydrous and degassed TEA (10 mL) was heated at 80 °C under a nitrogen atmosphere for 6 h in a sealed vessel. The reaction mixture was cooled, poured into H2O and extracted

3055

with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography in EtOAc/PE gradient from 5:95 to 40:60 affording 75 mg of the title product. Yield: 51%. 1H NMR (400 MHz, CDCl3) d: 8.84 (d, J = 2.2 Hz, 1H), 7.90 (dd, J = 2.2, 8.2 Hz, 1H), 7.50 (d, J = 8.2 Hz, 1H), 5.62 (s, 1H), 3.43–3.58 (m, 4H), 2.61 (t, J = 5.8 Hz, 2H), 2.34 (t, J = 5.8 Hz, 2H), 1.49 (s, 9H). 4.5.18.2. 6-[3-(4-Piperidylidene)prop-1-ynyl]pyridine-3-carbonitrile (21b). The title compound was obtained following the procedure described for compound 16d, but starting from compound 21a instead of compound 16c. After the usual work-up procedure, evaporation of the combined EtOAc extracts afforded a crude of 40 mg used in the following step without any further purification. Yield: 89%. 4.5.18.3. 6-[3-[1-(6-Methyl-3-nitro-2-pyridyl)-4-piperidylidene]prop-1-ynyl]pyridine-3-carbonitrile (21). A mixture of 21b (40 mg, 0.179 mmol), 2-chloro-3-nitro-6-methylpyridine (35.5 mg, 0.206 mmol), TEA (38.6 lL, 0.269 mmol), 4.55 mL of NMP were reacted overnight at 25 °C under a nitrogen atmosphere. The reaction mixture was poured into H2O and extracted with EtOAc (3). The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue, which was purified by flash chromatography eluting with EtOAc/PE gradient from 5:95 to 40:60 affording 58 mg of the title product as pale yellow oil. Yield: 90.2%. 1H NMR (400 MHz, CDCl3) d: 8.84 (d, J = 2.1 Hz, 1H), 8.10 (d, J = 8.2 Hz, 1H), 7.90 (dd, J = 2.1, 8.2 Hz, 1H), 7.50 (d, J = 8.2 Hz, 1H), 6.63 (d, J = 8.2 Hz, 1H), 5.67 (s, 1H), 3.46–3.64 (m, 4H), 2.79 (t, J = 5.8 Hz, 2H), 2.52 (t, J = 5.8 Hz, 2H), 2.48 (s, 3H). MS (ESI) m/z 360.24 [M+H]+. HPLC 98.6% (UV). 4.5.19. 3,3-Dimethyl-1-(4-{3-[1-(3-nitropyridin-2-yl)piperidin4-ylidene]prop-1-yn-1-yl}phenyl)azetidin-2-one (22) A mixture of compound 13c (45 mg, 0.185 mmol), 1-(4-iodophenyl)-3,3-dimethylazetidin-2-one (55.7 mg, 0.185 mmol), PdCl2(PPh3)2 (7.23 mg, 0.012 mmol), CuI (3.92 mg, 0.206 mmol) in anhydrous and degassed TEA (3 mL) was heated at 80 °C under a nitrogen atmosphere for 2 h in a sealed vessel. The reaction mixture was cooled, filtered on Celite, poured into H2O and extracted with EtOAc. The combined organic layers were washed with brine, dried over anhydrous Na2SO4 and evaporated to dryness in vacuo to afford a residue. Purification was carried out by automated flash liquid chromatography (HorizonÒ Biotage) eluting with EtOAc/PE in gradient from 0:100 to 80:20 to afford 32 mg of title compound. Yield: 42%. 1H NMR (400 MHz, CDCl3) d: 8.36 (d, J = 4.2 Hz, 1H), 8.15 (d, J = 8.2 Hz, 1H), 7.38–7.44 (m, 2H), 7.31 (m, 2H), 6.77 (dd, J = 4.2, 8.2 Hz, 1H), 5.62 (s, 1H), 3.47–3.60 (m, 4H), 3.46 (s, 2H), 2.66–2.84 (m, 2H), 2.41–2.55 (m, 2H), 1.43 (s, 6H). MS (ESI) m/z 417.5 [M+H]+. HPLC 96.1% (UV). 4.6. 2-[4-[3-(6-Methyl-2-pyridyl)prop-2-ynylidene]-1piperidyl]-5-phenyl-pyridine-3-carbonitrile (23) A solution of compound 16d (86 mg, 0.405 mmol), 2-chloro-5phenylnicotinonitrile (95.6 mg, 0.446 mmol) and DIPEA (142 mL) in 6 mL of NMP was heated in a microwave oven (Biotage) for 15 min at 150 °C giving a conversion of about 60%. A second heating for 10 min was carried out. After cooling, the reaction was poured into H2O and extracted with EtOAc (3), which was dried over anhydrous Na2SO4 and evaporated to dryness. Purification of the crude by flash chromatography with EtOAc/PE gradient from 1:9 to 6:4 afforded 85 mg of product which underwent further LCPREP purification giving 13.8 mg of the pure title product. Yield:

3056

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

9.5%. 1H NMR (400 MHz, CDCl3) d: 8.61 (s, 1H), 8.00 (s, 1H), 7.35– 7.78 (m, 6H), 7.07–7.35 (m, 2H), 5.68 (s, 1H), 3.74–4.06 (m, 4H), 2.85 (t, J = 5.7 Hz, 2H), 2.62 (s, 3H), 2.55 (t, J = 5.7 Hz, 2H). MS (ESI) m/z 391.4 [M+H]+. HPLC 98.3% (UV). 4.6.1. 6-Methyl-2-{(3E)-3-[3-(6-methylpyridin-2-yl)prop-2-yn1-ylidene]piperidin-1-yl}-3-nitropyridine (24) and 6-methyl-2{(3Z)-3-[3-(6-methylpyridin-2-yl)prop-2-yn-1-ylidene]piperidin1-yl}-3-nitropyridine (25) 4.6.1.1. tert-Butyl (3E)-3-[3-(trimethylsilyl)prop-2-yn-1-ylidene]piperidine-1-carboxylate (24a) and tert-Butyl (3Z)-3-[3(trimethylsilyl)prop-2-yn-1-ylidene]piperidine-1-carboxylate (25a). The title compounds were synthesized following the procedure described for 16a but using 1-(t-butoxycarbonyl)piperidin-3-one instead of 1-(3-nitro-2-pyridyl)piperidin4-one. After the work-up, the residue was purified by automated flash liquid chromatography (SP1Ò Biotage) eluting withEtOAc/PE gradient from 5:95 to 20:80 affording the title product. The less polar collected fractions afforded after evaporation 145 mg of compound 25a. Yield: 9.5%. The last eluted collected fractions (157 mg) were attributed to compound 24a. Yield: 10%. Both the group of fractions were used in the following step without further purification. 24a MS (ESI) m/z 294.26 [M+H]+. 25a MS (ESI) m/z 294.26 [M+H]+. 4.6.1.2. tert-Butyl (3E)-3-prop-2-ynylidenepiperidine-1-carboxylate (24b). Starting from compound 24a (53 mg, 0.180 mmol) the title compound was prepared following the procedure reported for compound 16b. After the usual work-up procedure, evaporation of the combined EtOAc extracts afforded a crude residue of 24 mg used in the next step without any further purification. Yield: 43%. 4.6.1.3. tert-Butyl (3Z)-3-(prop-2-yn-1-ylidene)piperidine-1carboxylate (25b). Starting from compound 25a (45 mg, 0.153 mmol) the title compound was prepared following the procedure reported for compound 16b. After the usual work-up procedure, evaporation of the combined EtOAc extracts afforded 29 mg of crude residue used in the next step without further purification. Yield: 61%. 4.6.1.4. tert-Butyl (3E)-3-[3-(6-methyl-2-pyridyl)prop-2-ynylidene]piperidine-1-carboxylate (24c). The title compound was obtained following the procedure described for compound 15, but starting from compound 24b (24 mg, 0.122 mmol), instead of compound 13c, and 2-bromo-6-methylpyridine (36.29 mg, 0.210 mmol). After the work-up, the residue was purified by automated flash liquid chromatography (SP1Ò Biotage) eluting with EtOAc/PE gradient from 15:85 to 50:50 affording 24 mg of title compound. Yield: 43%. MS (ESI) m/z 313.35 [M+H]+. 4.6.1.5. tert-Butyl (3Z)-3-[3-(6-methyl-2-pyridyl)prop-2-ynyliThe title compound dene]piperidine-1-carboxylate (25c). was obtained following the procedure described for compound 15, but starting from compound 25b (29 mg, 0.147 mmol), instead of compound 13c, and 2-bromo-6-methylpyridine (30.24 mg, 0.176 mmol). After the work-up, the residue was purified by automated flash liquid chromatography (SP1Ò Biotage) eluting with EtOAc/PE gradient from 15:85 to 50:50 affording 29 mg of title product. Yield: 61%. MS (ESI) m/z 313.35 [M+H]+. 4.6.1.6. 2-Methyl-6-[(3E)-3-(3-piperidylidene)prop-1-ynyl]pyriThe title compound was obtained following the dine (24d). procedure described for compound 16d, but starting from compound 24c instead of 16c. After usual work-up, the separation of the organic layer and extraction of the aqueous layer with DCM,

washing with brine and drying over anhydrous Na2SO4 afforded the title compound that was used as crude in the following step without further purification. MS (ESI) m/z 213.33 [M+H]+. 4.6.1.7. 2-Methyl-6-[(3Z)-3-(3-piperidylidene)prop-1-ynyl]pyridine (25d). The title compound was obtained following the procedure described for compound 16d, but starting from compound 25c instead of 16c. After usual work-up, the separation of the organic layer and extraction of the aqueous layer with DCM, washing with brine and drying over Na2SO4 afforded the title compound that was used as crude in the following step without further purification. MS (ESI) m/z 213.33 [M+H]+. 4.6.1.8. 6-Methyl-2-[(3E)-3-[3-(6-methyl-2-pyridyl)prop-2ynylidene]-1-piperidyl]-3-nitro-pyridine (24). A mixture of compound 24d (16 mg, 0.075 mmol) and 2-chloro-6-methyl-3-nitropyridine (18.2 mg, 0.105 mmol), TEA (32.5 lL, 0.226 mmol) in 3 mL of N,N-dimethylacetamide was stirred at 25 °C for 24 h. The reaction mixture was poured into H2O, extracted with EtOAc and washed with H2O and brine. The combined organic layers were dried over anhydrous Na2SO4 and the solvent was removed in vacuo. After the work-up, the residue was purified by automated flash liquid chromatography (SP1Ò Biotage) eluting with EtOAc/PE gradient from 3:7 to 4:6 affording 3 mg of the title product. Yield: 11%. 1H NMR (400 MHz, CDCl3) d: 8.09 (d, J = 8.2 Hz, 1H), 7.58 (t, J = 8.2 Hz, 1H), 7.29 (d, J = 7.8 Hz, 1H), 7.12 (d, J = 7.8 Hz, 1H), 6.60 (d, J = 8.2 Hz, 1H), 5.69 (s, 1H), 4.10 (s, 2H), 3.43–3.59 (m, 2H), 2.77 (t, J = 6.3 Hz, 2H), 2.61 (s, 3H), 2.48 (s, 3H), 1.84– 2.02 (m, 2H). MS (ESI) m/z 349.34 [M+H]+. HPLC 95.4% (UV). 4.6.1.9. 6-Methyl-2-{(3Z)-3-[3-(6-methylpyridin-2-yl)prop-2yn-1-ylidene]piperidin-1-yl}-3-nitropyridine (25). A mixture of compound 25d (24 mg, 0.113 mmol) and 2-chloro-6methyl-3-nitropyridine (27.3 mg, 0.158 mmol), TEA (22.7 lL, 0.158 mmol) in 3 mL of N,N-dimethylacetamide was stirred at 25 °C for 24 h. The reaction mixture was poured into H2O, extracted with EtOAc and washed with H2O and brine. The combined organic layers were dried over anhydrous Na2SO4 and the solvent was removed under vacuum. After the work-up, the residue was purified by automated flash liquid chromatography (SP1Ò Biotage) eluting with EtOAc/PE gradient 3:7 to 4:6 affording 7 mg of the title product. Yield: 18%. 1H NMR (400 MHz, CDCl3) d: 8.07 (d, J = 8.2, 1H), 7.61 (bs, 1H), 7.35–7.37 (s, 1H), 7.11–7.13 (m, 1H), 6.57–6.59 (d, J = 8.2, 1H), 5.62 (s, 1H), 4.35 (s, 2H), 3.61 (m, 2H), 2.60 (s, 3H), 2.46–2.53 (m, 5H), 1.87 (m, 2H). MS (ESI) m/z 349.41 [M+H]+. HPLC 96.2% (UV). 4.7. Biology CHO T-REx h-mGlu5 and CHO T-REx r-mGlu1. Cell lines stably transfected were generated using inducible expression vectors encoding human mGlu5 and rat mGlu1 receptor using the Tetracycline-Regulated Expression system (T-REx™ system, Invitrogen, Life Technologies). Human mGlu5 and rat mGlu1 open reading frame (ORF), comprehensive of the stop codon, were cloned into the pcDNA4/TO/myc-HisTM A vector, carrying the TetO2. The insertion site was, respectively, HindIII-PstI for mGlu5 and BamH1-XhoI for mGlu1 receptors. The obtained constructs were then transfected into the T-REx CHO™ cell line using the FuGENE protocol (Roche); the CHO T-REx™ cell line stably expresses the Tet repressor (from the pcDNA6/TR plasmid) under the selection of blasticidin 10 lg/mL. Stable clones were obtained selecting with zeocine 1 mg/mL and maintaining in ULTRA CHO medium (LONZA) supplemented with dialyzed FBS, zeocin, blasticidine, at 37 °C, in an atmosphere of 5% CO2. The expression of hmGlu5 and r-mGlu1 receptors was derepressed with 1 lg/mL

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

tetracycline 18 h before binding experiment, while the expression of h-mGlu5 and r-mGlu1 receptors was derepressed, respectively, with 3 ng/mL and 10 ng/mL tetracycline 18 h before calcium fluorescence experiment. Radioligand binding assay at native mGlu5 and cloned mGlu1 and mGlu5 receptor subtypes. Male Sprague Dawley rats (Crl:CD(SD)BR, 200–300 g bw) from Charles River Italy were used. Animals were housed with free access to food and water and maintained on a forced 12 h light-dark cycle at 22–24 °C for at least one week before experiments were carried out. The animals were handled according to internationally accepted principles for care of laboratory animals (E.E.C. Council Directive 86/609, O. J. no L358, 18/12/86). Affinity at native or cloned transmembrane glutamate metabotropic receptor subtypes, was evaluated according to the methods of Lavreysen47 and Anderson48 respectively, with some modifications. Briefly, male Sprague Dawley rats were killed by cervical dislocation and their brain rapidly removed. Forebrain (cortex, striatum and hippocampus) and cerebellum were dissected and homogenized (2  20 s) in 50 volumes of cold 50 mM Tris buffer pH 7.4, using a Politron homogenizer (Kinematica). Homogenates were centrifuged at 48,000g for 15 min, resuspended in 50 volumes of the same buffer, incubated at 37 °C for 15 min and centrifuged and resuspended two more times. The final pellets were frozen and stored at 80 °C until use. Native mGlu5 membranes from rat forebrain were resuspended in 100 volumes of 20 mM HEPES, 2 mM MgCl2, 2 mM CaCl2, pH 7.4, then were incubated in a final volume of 1 mL for 60 min at 25 °C with 4 nM [3H]MPEP in absence or presence of competing drugs. Non-specific binding was determined in the presence of 10 lM MPEP. Cloned mGlu1 were obtained resuspending CHO T-REx rmGlu1cells (60 lg/well) in 50 mM Tris, 1.2 mM MgCl2, 2 mM CaCl2, pH 7.4, that then were incubated in a final volume of 1 mL for 30 min at 0 °C with 2 nM [3H]R214127 in absence or presence of competing drugs. Non-specific binding was determined in the presence of 1 lM R214127. Cloned mGlu5 was obtained resuspending CHO T-REx h-mGlu5 cells (50 lg/well) in 20 mM HEPES, 2 mM MgCl2, 2 mM CaCl2, pH 7.4, that then were incubated in a final volume of 1 mL for 60 min at 25 °C with 4 nM [3H]MPEP in absence or presence of competing drugs. Non-specific binding was determined in the presence of 10 lM MPEP. The incubation was stopped by addition of cold Tris buffer pH 7.4 and rapid filtration through 0.2% polyethyleneimine pretreated Filtermat 1204401 (Perkin Elmer) filters. The filters were then washed with cold buffer and the radioactivity retained on the filters was counted by liquid scintillation spectrometry (Betaplate 1204 BS-Wallac). For binding studies, the compounds were dissolved in DMSO or demineralized water according to their solubility. All the reported doses were those of the corresponding salts or bases. Statistical analysis. The inhibition curves of the tested compounds at native and cloned mGlu1 and mGlu5 subtypes were determined by nonlinear regression analysis using software Prism 4.0 (Graphpad, San Diego, CA). The IC50 values and pseudoHill slope coefficients were estimated by the software. The values for the inhibition constant, Ki, were calculated according to the equation Ki = IC50/(1 + [L]/Kd), where [L] is the concentration of radioligand and Kd is the equilibrium dissociation constant of the radioligand-receptor complex.49 Calcium Fluorescence Measurements. Cells were seeded into black-walled, clear-bottom, 96-well plates at a density of 80,000 cell/well, in RPMI (without Phenol Red, without L-glutamine; Gibco LifeTechnologies, CA) supplemented with 10% dialyzed FBS. Following 18 h incubation with tetracycline, the cells were loaded with 2 lM Ca2+-sensitive fluorescent dye Fluo-4/AM (Molecular Probes) in Hanks’ balanced saline solution (HBSS, Gibco LifeTechnologies, CA) with 20 mM Hepes (Sigma) and 2.5 mM probenecid (Sigma), for 1 h at 37 °C. The cells were washed three

3057

times with HBSS to remove extracellular dye. Fluorescence signals were measured by using the fluorescence microplate reader Flexstation III (Molecular Devices) at sampling intervals of 1.5 s for 60 s. All antagonists were dissolved in 100% DMSO, and diluted in assay buffer to a 5 stock (2.5% DMSO). This stock was then applied to the cells at a final DMSO concentration of 0.5%. The antagonist potency was determined using EC80 of the quisqualate used as agonist (daily determined). The NAMs were applied 10 min before the application of the agonist. Responses were measured as peak increase in fluorescence minus basal fluorescence. Curves fitting of concentration-dependent data was performed using software Prism 4.0 (Graphpad, San Diego, CA), and half-maximal (IC50) were determined using a 4 parameter logistic nonlinear regression analysis available in the software. Acknowledgements ISCRA (Italian SuperComputing Resource Allocation), Project BEAR 2010, Class A # HP10A6E2CI is gratefully acknowledged. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.bmc.2015.05.008. References and notes 1. Conn, P. J.; Pin, J. P. Annu. Rev. Pharmacol. Toxicol. 1997, 37, 205. 2. Erreger, K.; Chen, P. E.; Wyllie, D. J. A.; Traynelis, S. F. Crit. Rev. Neurobiol. 2004, 16, 187. 3. O’Hara, P. J.; Sheppard, P. O.; Thøgersen, H.; Venezia, D.; Haldeman, B. A.; McGrane, V.; Houamed, K. M.; Thomsen, C.; Gilbert, T. L.; Mulvihill, E. R. Neuron 1993, 11, 41. 4. Sack, J. S.; Saper, M. A.; Quiocho, F. A. J. Mol. Biol. 1989, 206, 171. 5. Hermans, E.; Challiss, R. A. Biochem. J. 2001, 359, 465. 6. Mannaioni, G.; Marino, M. J.; Valenti, O.; Traynelis, S. F.; Conn, P. J. J. Neurosci. 2001, 21, 5925. 7. Ferraguti, F.; Shigemoto, R. Cell Tissue Res. 2006, 326, 483. 8. Kumar, V.; Jong, Y.-J. I.; O’Malley, K. L. J. Biol. Chem. 2008, 283, 14072. 9. Kawabata, S.; Tsutsumi, R.; Kohara, A.; Yamaguchi, T.; Nakanishi, S.; Okada, M. Nature 1996, 383, 89. 10. Nakahara, K.; Okada, M.; Nakanishi, S. J. Neurochem. 1997, 69, 1467. 11. Nash, M. S.; Young, K. W.; Challiss, R. A.; Nahorski, S. R. Nature 2001, 413, 381. 12. Nash, M. S.; Schell, M. J.; Atkinson, P. J.; Johnston, N. R.; Nahorski, S. R.; Challiss, R. A. J. J. Biol. Chem. 2002, 277, 35947. 13. Atkinson, P. J.; Young, K. W.; Ennion, S. J.; Kew, J. N. C.; Nahorski, S. R.; Challiss, R. A. J. Mol. Pharmacol. 2006, 69, 174. 14. De Blasi, A.; Conn, P. J.; Pin, J.; Nicoletti, F. Trends Pharmacol. Sci. 2001, 22, 114. 15. Nakanishi, S. Neuron 1994, 1031, 13. 16. Hölscher, C.; Gigg, J.; O’Mara, S. M. Neurosci. Biobehav. Rev. 1999, 23, 399. 17. Caraci, F.; Battaglia, G.; Sortino, M. A.; Spampinato, S.; Molinaro, G.; Copani, A.; Nicoletti, F.; Bruno, V. Neurochem. Int. 2012, 61, 559. 18. Nicoletti, F.; Bockaert, J.; Collingridge, G. L.; Conn, P. J.; Ferraguti, F.; Schoepp, D. D.; Wroblewski, J. T.; Pin, J. P. Neuropharmacology 2011, 1017, 60. 19. Wu, H.; Wang, C.; Gregory, K. J.; Han, G. W.; Cho, H. P.; Xia, Y.; Niswender, C. M.; Katritch, V.; Meiler, J.; Cherezov, V.; Conn, P. J.; Stevens, R. C. Science 2014, 344, 58. 20. Doré, A. S.; Okrasa, K.; Patel, J. C.; Serrano-Vega, M.; Bennett, K. A.; Cooke, R. M.; Errey, J. C.; Jazayeri, A.; Khan, S.; Tehan, B.; Weir, M.; Wiggin, G. R.; Marshall, F. H. Nature 2014, 511, 557. 21. Leonardi, A.; Motta, G.; Riva, C.; Poggesi, E.; Graziani, D.; Longhi, M. M. Patent No. WO 2009/015897 A1, 2009. 22. Gharagozloo, P.; Islam, K.; Kyle, D. J.; Sun, Q.; Tafesse, L.; Whitehead, J. W. F.; Yang, J.; Zhou, X. Patent No. WO 2003/093236 A1, 2003. 23. Pagano, A.; Ruegg, D.; Litschig, S.; Stoehr, N.; Stierlin, C.; Heinrich, M.; Floersheim, P.; Prezèau, L.; Carroll, F.; Pin, J. P.; Cambria, A.; Vranesic, I.; Flor, P. J.; Gasparini, F.; Kuhn, R. J. Biol. Chem. 2000, 275, 33750. 24. Malherbe, P.; Kratochwil, N.; Mühlemann, A.; Zenner, M.-T.; Fischer, C.; Stahl, M.; Gerber, P. R.; Jaeschke, G.; Porter, R. H. P. J. Neurochem. 2006, 98, 601. 25. Malherbe, P.; Kratochwil, N.; Zenner, M.-T.; Piussi, J.; Diener, C.; Kratzeisen, C.; Fischer, C.; Porter, R. H. P. Mol. Pharmacol. 2003, 64, 823. 26. Mølck, C.; Harpsøe, K.; Gloriam, D. E.; Clausen, R. P.; Madsen, U.; Pedersen, L. Ø.; Jimenez, H. N.; Nielsen, S. M.; Mathiesen, J. M.; Bräuner-Osborne, H. Mol. Pharmacol. 2012, 82, 929. 27. Gregory, K. J.; Nguyen, E. D.; Reiff, S. D.; Squire, E. F.; Stauffer, S. R.; Lindsley, C. W.; Meiler, J.; Conn, P. J. Mol. Pharmacol. 2013, 83, 991. 28. Bennett, K. A.; Doré, A. S.; Christopher, J. A.; Weiss, D. R.; Marshall, F. H. Curr. Opin. Pharmacol. 2015, 20, 1.

3058

A. Anighoro et al. / Bioorg. Med. Chem. 23 (2015) 3040–3058

29. Roppe, J. R.; Wang, B.; Huang, D.; Tehrani, L.; Kamenecka, T.; Schweiger, E. J.; Anderson, J. J.; Brodkin, J.; Jiang, X.; Cramer, M.; Chung, J.; Reyes-Manalo, G.; Munoz, B.; Cosford, N. D. P. Bioorg. Med. Chem. Lett. 2004, 14, 3993. 30. Cosford, N. D. P.; Tehrani, L.; Roppe, J.; Schweiger, E.; Smith, N. D.; Anderson, J.; Bristow, L.; Brodkin, J.; Jiang, X.; McDonald, I.; Rao, S.; Washburn, M.; Varney, M. A. J. Med. Chem. 2003, 46, 204. 31. Milbank, J. B. J.; Knauer, C. S.; Augelli-Szafran, C. E.; Sakkab-Tan, A. T.; Lin, K. K.; Yamagata, K.; Hoffman, J. K.; Zhuang, N.; Thomas, J.; Galatsis, P.; Wendt, J. A.; Mickelson, J. W.; Schwarz, R. D.; Kinsora, J. J.; Lotarski, S. M.; Stakich, K.; Gillespie, K. K.; Lam, W. W.; Mutlib, A. E. Bioorg. Med. Chem. Lett. 2007, 17, 4415. 32. Büttelmann, B.; Peters, J.-U.; Ceccarelli, S.; Kolczewski, S.; Vieira, E.; Prinssen, E. P.; Spooren, W.; Schuler, F.; Huwyler, J.; Porter, R. H. P.; Jaeschke, G. Bioorg. Med. Chem. Lett. 1892, 2006, 16. 33. Rodriguez, A. L.; Grier, M. D.; Jones, C. K.; Herman, E. J.; Kane, A. S.; Smith, R. L.; Williams, R.; Zhou, Y.; Marlo, J. E.; Days, E. L.; Blatt, T. N.; Jadhav, S.; Menon, U. N.; Vinson, P. N.; Rook, J. M.; Stauffer, S. R.; Niswender, C. M.; Lindsley, C. W.; Weaver, C. D.; Conn, P. J. Mol. Pharmacol. 2010, 78, 1105. 34. Pilla, M.; Andreoli, M.; Tessari, M.; Delle-Fratte, S.; Roth, A.; Butler, S.; Brown, F.; Shah, P.; Bettini, E.; Cavallini, P.; Benedetti, R.; Minick, D.; Smith, P.; Tehan, B.; D’Alessandro, P.; Lorthioir, O.; Ball, C.; Garzya, V.; Goodacre, C.; Watson, S. Bioorg. Med. Chem. Lett. 2010, 20, 7521. 35. Spanka, C.; Glatthar, R.; Desrayaud, S.; Fendt, M.; Orain, D.; Troxler, T.; Vranesic, I. Bioorg. Med. Chem. Lett. 2010, 20, 184. 36. Zhang, L.; Balan, G.; Barreiro, G.; Boscoe, B. P.; Chenard, L. K.; Cianfrogna, J.; Claffey, M. M.; Chen, L.; Coffman, K. J.; Drozda, S. E.; Dunetz, J. R.; Fonseca, K. R.; Galatsis, P.; Grimwood, S.; Lazzaro, J. T.; Mancuso, J. Y.; Miller, E. L.; Reese, M. R.; Rogers, B. N.; Sakurada, I.; Skaddan, M.; Smith, D. L.; Stepan, A. F.; Trapa, P.; Tuttle, J. B.; Verhoest, P. R.; Walker, D. P.; Wright, A. S.; Zaleska, M. M.; Zasadny, K.; Shaffer, C. L. J. Med. Chem. 2014, 57, 861.

37. Burdi, D. F.; Hunt, R.; Fan, L.; Hu, T.; Wang, J.; Guo, Z.; Huang, Z.; Wu, C.; Hardy, L.; Detheux, M.; Orsini, M. A.; Quinton, M. S.; Lew, R.; Spear, K. J. Med. Chem. 2010, 53, 7107. 38. Larkin, M. A.; Blackshields, G.; Brown, N. P.; Chenna, R.; McGettigan, P. A.; McWilliam, H.; Valentin, F.; Wallace, I. M.; Wilm, A.; Lopez, R.; Thompson, J. D.; Gibson, T. J.; Higgins, D. G. Bioinformatics 2007, 23, 2947. 39. Sali, A.; Blundell, T. L. J. Mol. Biol. 1993, 234, 779. 40. Case, D. A.; Darden, T. A.; Cheatham, T. E. I. I. I.; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang, W.; Merz, K. M.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.; Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wu, X.; Brozell, S. R.; Steinbrecher, T.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin, V.; Kollman, P. A. AMBER 10; University of California: San Francisco, 2008. 41. Feig, M.; Onufriev, A.; Lee, M. S.; Im, W.; Case, D. A.; Brooks, C. L., III J. Comput. Chem. 2004, 25, 265. 42. Schrödinger Release 2014-1; Schrödinger LLC, New York, 2014. 43. Sastry, G. M.; Adzhigirey, M.; Day, T.; Annabhimoju, R.; Sherman, W. J. Comput. Aided Mol. Des. 2013, 27, 221. 44. Bernstein, F. C.; Koetzle, T. F.; Williams, G. J.; Meyer, E. F.; Brice, M. D.; Rodgers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. J. Mol. Biol. 1977, 112, 535. 45. Sherman, W.; Day, T.; Jacobson, M. P.; Friesner, R. A.; Farid, R. J. Med. Chem. 2006, 49, 534. 46. Anighoro, A.; Rastelli, G. J. Chem. Inf. Model. 2013, 53, 739. 47. Lavreysen, H.; Janssen, C.; Bischoff, F.; Langlois, X.; Leysen, J. E.; Lesage, A. S. J. Mol. Pharmacol. 2003, 1082, 63. 48. Anderson, J. J.; Rao, S. P.; Rowe, B.; Giracello, D. R.; Holtz, G.; Chapman, D. F.; Tehrani, L.; Bradbury, M. J.; Cosford, N. D. P.; Varney, M. A. J. Pharmacol. Exp. Ther. 2002, 1044, 303. 49. Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099.