Identification of a Novel Neuropeptide S Receptor Antagonist Scaffold Based on the SHA-68 Core

Activation of the neuropeptide S receptor (NPSR) system has been shown to produce anxiolytic-like actions, arousal, and enhance memory consolidation, whereas blockade of the NPSR has been shown to reduce relapse to substances of abuse and duration of anesthetics. We report here the discovery of a novel core scaffold (+) N-benzyl-3-(2-methylpropyl)-1-oxo-3-phenyl-1H,3H,4H,5H,6H,7H-furo[3,4-c]pyridine-5-carboxamide with potent NPSR antagonist activity in vitro. Pharmacokinetic parameters demonstrate that 14b reaches pharmacologically relevant levels in plasma and the brain following intraperitoneal (i.p.) administration, but is cleared rapidly from plasma. Compound 14b was able to block NPS (0.3 nmol)-stimulated locomotor activity in C57/Bl6 mice at 3 mg/kg (i.p.), indicating potent in vivo activity for the structural class. This suggests that 14b can serve as a useful tool for continued mapping of the pharmacological functions of the NPS receptor system.


Introduction
The neuropeptide S (NPS) system was first described publicly more than 15 years ago [1]; however, its therapeutic potential has not yet been achieved due to the lack of high-affinity drug-like ligands. Central administration of NPS (20 AA peptide) into mice produces anxiolytic-like actions, arousal, and enhances memory consolidation [1][2][3]. These wide-ranging effects are mediated by NPS receptors (NPSR) distributed throughout the CNS, e.g., cortex, amygdala, parasubiculum, and hypothalamus [4]. Expression of NPS is highly localized to discrete nuclei, e.g., peri-locus coerelus [1], although there are established interactions with the CRF (corticotropin releasing factor) system which support a role for NPS in the modulation of stress [5][6][7]. The exact connectivity of the NPSR-expressing and NPS-expressing neurons is still under investigation, and so are a number of other fundamental elements of this neuromodulatory system.
The discovery of small-molecule NPSR agonist molecules would represent a new class of non-sedating fast-acting anxiolytics, with memory enhancement properties [3]. However, NPSR antagonists have shown promise for treating relapse to substances of abuse [5,8] and may enhance the duration of mechanistically distinct anesthetics [9]. Improved NPSR agonists would be specifically useful for elucidating aspects of the NPS-system. In addition, more drug-like NPSR antagonists would be useful as tool compounds for target validation. However, currently available NPSR antagonists (Figure 1, Compounds A-D) have good potency but tend to be highly lipophilic and have a high number of sp2 hybridized atoms. These properties result in compounds that are challenging to formulate for in vivo work [10][11][12][13][14][15]. of sp2 hybridized atoms. These properties result in compounds that are challenging to formulate for in vivo work [10][11][12][13][14][15]. For a number of known NPSR antagonists, only limited structure-activity relationship (SAR) studies have been reported. The SHA-66/68 class of molecules, which were disclosed by Okamura et al. in 2008, possess poor drug-like properties, particularly aqueous solubility, and several structural modifications have been undertaken by our group and others, resulting in a number of improved analogs such as RTI-118 [10,12]. One structural feature of the SHA class that has not been thoroughly investigated is the diaryl moiety. The chemical synthesis of SHA makes differential modification of the diaryl moiety cumbersome and introduces an additional chiral center. Considering the high sp2 count of SHA and the metabolic liabilities associated with unsubstituted aryl rings, we hypothesized that modification of this feature could lead to improved NPSR antagonists. We envisioned that replacement of the carbamate nitrogen in SHA with an alkene would provide analogs that retain the overall geometry of SHA and keep in place a number of the critical pharmacophoric elements ( Figure 2). SHA already contains one chiral center; therefore, differentially substituting the diaryl core would result in a complex mixture of diastereomers that could be challenging to resolve. The addition of a double bond at the bicyclic ring junction eliminates one chiral center, leaving the remaining chiral center at the differentially substituted aryl rings of the oxazolidinone core. Therefore, our group explored structural modifications that retained the overall shape of SHA-68 yet allowed chemically tractable and differential modification of the diaryl ring system. Our resulting core, 4,5,6,7-tetrahydrofuro [3,4-c]pyridin-1(3H)-one, has produced a series of novel NPSR antagonists with similar antagonist potency in vitro, and favorable activity in vivo.  For a number of known NPSR antagonists, only limited structure-activity relationship (SAR) studies have been reported. The SHA-66/68 class of molecules, which were disclosed by Okamura et al. in 2008, possess poor drug-like properties, particularly aqueous solubility, and several structural modifications have been undertaken by our group and others, resulting in a number of improved analogs such as RTI-118 [10,12]. One structural feature of the SHA class that has not been thoroughly investigated is the diaryl moiety. The chemical synthesis of SHA makes differential modification of the diaryl moiety cumbersome and introduces an additional chiral center. Considering the high sp2 count of SHA and the metabolic liabilities associated with unsubstituted aryl rings, we hypothesized that modification of this feature could lead to improved NPSR antagonists. We envisioned that replacement of the carbamate nitrogen in SHA with an alkene would provide analogs that retain the overall geometry of SHA and keep in place a number of the critical pharmacophoric elements ( Figure 2). SHA already contains one chiral center; therefore, differentially substituting the diaryl core would result in a complex mixture of diastereomers that could be challenging to resolve. The addition of a double bond at the bicyclic ring junction eliminates one chiral center, leaving the remaining chiral center at the differentially substituted aryl rings of the oxazolidinone core. Therefore, our group explored structural modifications that retained the overall shape of SHA-68 yet allowed chemically tractable and differential modification of the diaryl ring system. Our resulting core, 4,5,6,7-tetrahydrofuro [3,4-c]pyridin-1(3H)-one, has produced a series of novel NPSR antagonists with similar antagonist potency in vitro, and favorable activity in vivo.
Pharmaceuticals 2021, 14,1024 2 of 29 of sp2 hybridized atoms. These properties result in compounds that are challenging to formulate for in vivo work [10][11][12][13][14][15]. For a number of known NPSR antagonists, only limited structure-activity relationship (SAR) studies have been reported. The SHA-66/68 class of molecules, which were disclosed by Okamura et al. in 2008, possess poor drug-like properties, particularly aqueous solubility, and several structural modifications have been undertaken by our group and others, resulting in a number of improved analogs such as RTI-118 [10,12]. One structural feature of the SHA class that has not been thoroughly investigated is the diaryl moiety. The chemical synthesis of SHA makes differential modification of the diaryl moiety cumbersome and introduces an additional chiral center. Considering the high sp2 count of SHA and the metabolic liabilities associated with unsubstituted aryl rings, we hypothesized that modification of this feature could lead to improved NPSR antagonists. We envisioned that replacement of the carbamate nitrogen in SHA with an alkene would provide analogs that retain the overall geometry of SHA and keep in place a number of the critical pharmacophoric elements ( Figure 2). SHA already contains one chiral center; therefore, differentially substituting the diaryl core would result in a complex mixture of diastereomers that could be challenging to resolve. The addition of a double bond at the bicyclic ring junction eliminates one chiral center, leaving the remaining chiral center at the differentially substituted aryl rings of the oxazolidinone core. Therefore, our group explored structural modifications that retained the overall shape of SHA-68 yet allowed chemically tractable and differential modification of the diaryl ring system. Our resulting core, 4,5,6,7-tetrahydrofuro [3,4-c]pyridin-1(3H)-one, has produced a series of novel NPSR antagonists with similar antagonist potency in vitro, and favorable activity in vivo.

Chemistry
Compounds 1, 2, 8-19, 21, 22, and 24-28 (Tables 1-3; Figures S1-S28, Supplementary Materials) were synthesized in a manner analogous to that outlined in Scheme 1. Previous research by Epsztajn et al. [16] demonstrated that the ortho lithiation of diiso- propyl nicotinamides proceeded smoothly and the resulting anion could be condensed with various ketones and aldehydes to afford the precursor to the bicyclic ring structures needed for our investigation. N,N-diisopropylisonicotinamide 30 was treated with lithium di-isopropylamide to form the anion of 30 which was condensed with the appropriate ketone to yield an alcohol intermediate, such as compound 31, in moderate yield. Upon treatment with aqueous acid (6N HCl), the alcohol readily cyclizes into lactone 32. A series of reduction conditions were evaluated for the pyridyl ring. We discovered reduction of the pyridine ring with 1 equivalent of acid, hydrogen, and platinum oxide as a catalyst partially reduced ring 33, leaving the tetrasubstituted double bond intact. These reduction conditions afforded the 4,5,6,7-tetrahydrofuro [3,4-c]pyridin-1(3H)-one core 33 in good yield. Final treatment of amine 33 with the desired isocyanate in dichloromethane afforded target compounds 14a,b in good yield. by roughly 50-fold. This suggested that either the overall conformation of 1 was less tolerated by the receptor or the oxazolidinone nitrogen was a critical pharmacophore element. In order to investigate this further, reduction of the double bond to both the cisisomers 3 and 4 and trans-isomers 5 and 6 were undertaken. The cis-isomers with a significantly different three-dimensional shape lacked antagonist activity, whereas the transisomers 5 and 6 were moderately potent with Ke values of 170 nM and 260 nM, respectively. We then evaluated whether shifting the urea substituent within the ring system, as shown in compound 7, could restore activity, but this resulted in a significant loss of potency. Importantly, this new core scaffold allows more efficient access to modification of the 3,3-diphenyl ring system without the introduction of a second chiral center (Table 2). Conformational control is an important aspect of ligand design. Reducing conformational freedom was explored by linking the two aryl rings together (compounds 8 and 9) as well as removing an aryl ring (compound 10) and activity was lost. In order to reduce the sp2 count for SHA, we envisioned the replacement of one aryl ring with an aliphatic group. Replacement of one phenyl ring into various aliphatic substituents yielded compounds 11-15. The n-butyl compound 11 was moderately potent (Ke = 165 nM), but the iso-butyl compound 14 (Ke = 118 nM) was the most potent and was also anticipated to have an improved metabolic profile. Integrating various heteroatoms with hydrogen bond acceptors or donors into this portion of the molecule, such as a dimethylamino group or a hydroxy group (17)(18)(19)(20)(21), resulted in a loss of activity. In order to potentially reduce oxidative metabolism on one of the remaining unsubstituted aryl rings, we explored a series of substituted aryl rings (22-25) and found that the incorporation of naphthyl, hydroxy or benzyl ether groups (22-24) resulted in a reduction in potency. However, the 3-chlorophenyl group 25 only had a modest reduction in potency, at 136 nM compared to 14. by roughly 50-fold. This suggested that either the overall conformation of 1 was less tolerated by the receptor or the oxazolidinone nitrogen was a critical pharmacophore element. In order to investigate this further, reduction of the double bond to both the cisisomers 3 and 4 and trans-isomers 5 and 6 were undertaken. The cis-isomers with a significantly different three-dimensional shape lacked antagonist activity, whereas the transisomers 5 and 6 were moderately potent with Ke values of 170 nM and 260 nM, respectively. We then evaluated whether shifting the urea substituent within the ring system, as shown in compound 7, could restore activity, but this resulted in a significant loss of potency. Importantly, this new core scaffold allows more efficient access to modification of the 3,3-diphenyl ring system without the introduction of a second chiral center (Table 2). Conformational control is an important aspect of ligand design. Reducing conformational freedom was explored by linking the two aryl rings together (compounds 8 and 9) as well as removing an aryl ring (compound 10) and activity was lost. In order to reduce the sp2 count for SHA, we envisioned the replacement of one aryl ring with an aliphatic group. Replacement of one phenyl ring into various aliphatic substituents yielded compounds 11-15. The n-butyl compound 11 was moderately potent (Ke = 165 nM), but the iso-butyl compound 14 (Ke = 118 nM) was the most potent and was also anticipated to have an improved metabolic profile. Integrating various heteroatoms with hydrogen bond acceptors or donors into this portion of the molecule, such as a dimethylamino group or a hydroxy group (17)(18)(19)(20)(21), resulted in a loss of activity. In order to potentially reduce oxidative metabolism on one of the remaining unsubstituted aryl rings, we explored a series of substituted aryl rings (22-25) and found that the incorporation of naphthyl, hydroxy or benzyl ether groups (22-24) resulted in a reduction in potency. However, the 3-chlorophenyl group 25 only had a modest reduction in potency, at 136 nM compared to 14. by roughly 50-fold. This suggested that either the overall conformation of 1 was less tolerated by the receptor or the oxazolidinone nitrogen was a critical pharmacophore element. In order to investigate this further, reduction of the double bond to both the cisisomers 3 and 4 and trans-isomers 5 and 6 were undertaken. The cis-isomers with a significantly different three-dimensional shape lacked antagonist activity, whereas the transisomers 5 and 6 were moderately potent with Ke values of 170 nM and 260 nM, respectively. We then evaluated whether shifting the urea substituent within the ring system, as shown in compound 7, could restore activity, but this resulted in a significant loss of potency. Importantly, this new core scaffold allows more efficient access to modification of the 3,3-diphenyl ring system without the introduction of a second chiral center (Table 2). Conformational control is an important aspect of ligand design. Reducing conformational freedom was explored by linking the two aryl rings together (compounds 8 and 9) as well as removing an aryl ring (compound 10) and activity was lost. In order to reduce the sp2 count for SHA, we envisioned the replacement of one aryl ring with an aliphatic group. Replacement of one phenyl ring into various aliphatic substituents yielded compounds 11-15. The n-butyl compound 11 was moderately potent (Ke = 165 nM), but the iso-butyl compound 14 (Ke = 118 nM) was the most potent and was also anticipated to have an improved metabolic profile. Integrating various heteroatoms with hydrogen bond acceptors or donors into this portion of the molecule, such as a dimethylamino group or a hydroxy group (17)(18)(19)(20)(21), resulted in a loss of activity. In order to potentially reduce oxidative metabolism on one of the remaining unsubstituted aryl rings, we explored a series of substituted aryl rings (22-25) and found that the incorporation of naphthyl, hydroxy or benzyl ether groups (22-24) resulted in a reduction in potency. However, the 3-chlorophenyl group 25 only had a modest reduction in potency, at 136 nM compared to 14. by roughly 50-fold. This suggested that either the overall conformation of 1 was less tolerated by the receptor or the oxazolidinone nitrogen was a critical pharmacophore element. In order to investigate this further, reduction of the double bond to both the cisisomers 3 and 4 and trans-isomers 5 and 6 were undertaken. The cis-isomers with a significantly different three-dimensional shape lacked antagonist activity, whereas the transisomers 5 and 6 were moderately potent with Ke values of 170 nM and 260 nM, respectively. We then evaluated whether shifting the urea substituent within the ring system, as shown in compound 7, could restore activity, but this resulted in a significant loss of potency. Importantly, this new core scaffold allows more efficient access to modification of the 3,3-diphenyl ring system without the introduction of a second chiral center ( Table 2). Conformational control is an important aspect of ligand design. Reducing conformational freedom was explored by linking the two aryl rings together (compounds 8 and 9) as well as removing an aryl ring (compound 10) and activity was lost. In order to reduce the sp2 count for SHA, we envisioned the replacement of one aryl ring with an aliphatic group. Replacement of one phenyl ring into various aliphatic substituents yielded compounds 11-15. The n-butyl compound 11 was moderately potent (Ke = 165 nM), but the iso-butyl compound 14 (Ke = 118 nM) was the most potent and was also anticipated to have an improved metabolic profile. Integrating various heteroatoms with hydrogen bond acceptors or donors into this portion of the molecule, such as a dimethylamino group or a hydroxy group (17)(18)(19)(20)(21), resulted in a loss of activity. In order to potentially reduce oxidative metabolism on one of the remaining unsubstituted aryl rings, we explored a series of substituted aryl rings (22-25) and found that the incorporation of naphthyl, hydroxy or benzyl ether groups (22-24) resulted in a reduction in potency. However, the 3-chlorophenyl group 25 only had a modest reduction in potency, at 136 nM compared to 14.     Our lab has previously shown that replacement of the benzyl urea with a piperidino ethyl urea on the SHA core provides analogs with good brain permeability, enhanced solubility, and favorable in vivo activity [8,10,18]. With lead compound 14 in hand, we moved forward with exploring further modifications to the urea group using the 3-isobutyl-3-phenyl substituted core (Table 3). Incorporating n-butyl urea (26), piperidino ethyl urea (27) and piperidino ethyl thiourea (28) substituents resulted in a reduction in potency. The two enantiomers of compound 14 were separated by chiral HPLC, and we determined that the plus isomer (14b) was the active enantiomer (Ke = 71.4 nM, Table 3) and showed a significant right shift to the agonist activity of NPS at 3 µM concentration in calcium mobilization assay ( Figure 3). In order to evaluate if this new series was interacting with the NPSR in a similar manner as SHA, we synthesized the SHA analog with isobutyl substituted for one of the aryl rings. Compound 29 was completely inactive, suggesting that this new series binds to the NPSR in an overlapping but distinct manner.  Our lab has previously shown that replacement of the benzyl urea with a piperidino ethyl urea on the SHA core provides analogs with good brain permeability, enhanced solubility, and favorable in vivo activity [8,10,18]. With lead compound 14 in hand, we moved forward with exploring further modifications to the urea group using the 3-isobutyl-3-phenyl substituted core (Table 3). Incorporating n-butyl urea (26), piperidino ethyl urea (27) and piperidino ethyl thiourea (28) substituents resulted in a reduction in potency. The two enantiomers of compound 14 were separated by chiral HPLC, and we determined that the plus isomer (14b) was the active enantiomer (Ke = 71.4 nM, Table 3) and showed a significant right shift to the agonist activity of NPS at 3 µM concentration in calcium mobilization assay ( Figure 3). In order to evaluate if this new series was interacting with the NPSR in a similar manner as SHA, we synthesized the SHA analog with isobutyl substituted for one of the aryl rings. Compound 29 was completely inactive, suggesting that this new series binds to the NPSR in an overlapping but distinct manner.    One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double bond of 14b (Scheme 4). Compound 14b was incubated in a solution of acetonitrile and  One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double bond of 14b (Scheme 4). Compound 14b was incubated in a solution of acetonitrile and  One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double bond of 14b (Scheme 4). Compound 14b was incubated in a solution of acetonitrile and  One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double bond of 14b (Scheme 4). Compound 14b was incubated in a solution of acetonitrile and  One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double with aqueous acid (6N HCl), the alcohol readily cyclizes into lactone 32. A series of reduction conditions were evaluated for the pyridyl ring. We discovered reduction of the pyridine ring with 1 equivalent of acid, hydrogen, and platinum oxide as a catalyst partially reduced ring 33, leaving the tetrasubstituted double bond intact. These reduction conditions afforded the 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core 33 in good yield. Final treatment of amine 33 with the desired isocyanate in dichloromethane afforded target compounds 14a,b in good yield.

Scheme 1. General synthetic route to alkene-based NPS antagonists.
Reduction of the 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core 36 with nickel acetate and sodium borohydride resulted in the reduced cis-intermediate 37, which was then alkylated with benzyl or butyl isocyanate to yield compounds 3 and 4, respectively. Prior to alkylation, the cis-intermediate 37 was also treated with sodium hydride to isomerize the alpha proton to the trans-isomer. Although this resulted in an inseparable mixture of the cis-and trans-diastereomers, after alkylation with benzyl or butyl isocyanate, the diastereomers could be separated to yield the pure trans-analogs 5 and 6 (Scheme 2). Scheme 1. General synthetic route to alkene-based NPS antagonists.
Reduction of the 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core 36 with nickel acetate and sodium borohydride resulted in the reduced cis-intermediate 37, which was then alkylated with benzyl or butyl isocyanate to yield compounds 3 and 4, respectively. Prior to alkylation, the cis-intermediate 37 was also treated with sodium hydride to isomerize the alpha proton to the trans-isomer. Although this resulted in an inseparable mixture of the cisand trans-diastereomers, after alkylation with benzyl or butyl isocyanate, the diastereomers could be separated to yield the pure trans-analogs 5 and 6 (Scheme 2). reduced ring 33, leaving the tetrasubstituted double bond intact. These reduction conditions afforded the 4,5,6,7-tetrahydrofuro [3,4-c]pyridin-1(3H)-one core 33 in good yield. Final treatment of amine 33 with the desired isocyanate in dichloromethane afforded target compounds 14a,b in good yield. Scheme 1. General synthetic route to alkene-based NPS antagonists.
Reduction of the 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core 36 with nickel acetate and sodium borohydride resulted in the reduced cis-intermediate 37, which was then alkylated with benzyl or butyl isocyanate to yield compounds 3 and 4, respectively. Prior to alkylation, the cis-intermediate 37 was also treated with sodium hydride to isomerize the alpha proton to the trans-isomer. Although this resulted in an inseparable mixture of the cis-and trans-diastereomers, after alkylation with benzyl or butyl isocyanate, the diastereomers could be separated to yield the pure trans-analogs 5 and 6 (Scheme 2). Scheme 2. Synthesis of saturated diaryl analogs using the alkene core.
Previous work by Niiyama and colleagues [17] demonstrated that the addition of aryl Grignard reagents to pyridine 3,4-dicarboxylic acid anhydride affords compounds with selective addition of the Grignard to the carbonyl para to the pyridyl nitrogen. Compound 7 was obtained from the addition of excess phenyl magnesium bromide to pyridine 3,4dicarboxylic acid anhydride to yield pyridine intermediate 39. The sequence depicted in Scheme 1 was then followed to yield target compound 7.
Oxazolidinone analog 29, which mimics SHA with one phenyl group replaced with an isobutyl group, was synthesized from N-t-butoxy carbonyl, N-benzyl piperidine. Deprotonation of piperidine 92 was accomplished with sec-butyl lithium and tetramethylene diamine. The anion was quenched with the addition of isovalerophenone followed by cyclization to the oxazolidinone ring 93. Catalytic reduction of the benzyl ring was unsuccessful due to hydrogenolysis of the oxazolidinone ring. Removal of the benzyl group was undertaken with FMOC-chloride (94) followed by deprotection of the FMOC group with an isobutyl group, was synthesized from N-t-butoxy carbonyl, N-benzyl piperidine. Deprotonation of piperidine 92 was accomplished with sec-butyl lithium and tetramethylene diamine. The anion was quenched with the addition of isovalerophenone followed by cyclization to the oxazolidinone ring 93. Catalytic reduction of the benzyl ring was unsuccessful due to hydrogenolysis of the oxazolidinone ring. Removal of the benzyl group was undertaken with FMOC-chloride (94) followed by deprotection of the FMOC group with piperidine to yield the free amine 95. Treatment of the free amine 95 with benzyl isocyanate yielded the desired compound 29 (Scheme 3; Figure S29. Supplementary Materials).

Results and Discussion
We hypothesized that a 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core, which is similar to that of SHA-68, would allow access to broader functional group modifications and, ultimately, allow access to a larger SAR study than those previously reported for SHA-related antagonists. Compound 1 maintained the same substituents as SHA-68 but incorporated a fused double bond in place of the oxazolidinone nitrogen atom.
A direct comparison of compound 1 (Ke = 522 nM) with SHA-66 (Ke = 10 nM) indicated that replacement of the nitrogen atom with a double bond was detrimental for antagonist activity by roughly 50-fold (Table 1). We have previously shown that replacement of the benzyl urea on SHA with an n-butyl group resulted in compounds with equipotent antagonist activity but more conformational flexibility [10]. Therefore, we evaluated compound 2 with n-butyl urea; however, this was also reduced in activity compared to SHA Scheme 3. Synthesis of 3-N analogs and isobutyl replacement for phenyl in SHA.

Results and Discussion
We hypothesized that a 4,5,6,7-tetrahydrofuro[3,4-c]pyridin-1(3H)-one core, which is similar to that of SHA-68, would allow access to broader functional group modifications and, ultimately, allow access to a larger SAR study than those previously reported for SHA-related antagonists. Compound 1 maintained the same substituents as SHA-68 but incorporated a fused double bond in place of the oxazolidinone nitrogen atom.
A direct comparison of compound 1 (Ke = 522 nM) with SHA-66 (Ke = 10 nM) indicated that replacement of the nitrogen atom with a double bond was detrimental for antagonist activity by roughly 50-fold (Table 1). We have previously shown that replacement of the benzyl urea on SHA with an n-butyl group resulted in compounds with equipotent antagonist activity but more conformational flexibility [10]. Therefore, we evaluated compound 2 with n-butyl urea; however, this was also reduced in activity compared to SHA by roughly 50-fold. This suggested that either the overall conformation of 1 was less tolerated by the receptor or the oxazolidinone nitrogen was a critical pharmacophore element. In order to investigate this further, reduction of the double bond to both the cis-isomers 3 and 4 and trans-isomers 5 and 6 were undertaken. The cis-isomers with a significantly different three-dimensional shape lacked antagonist activity, whereas the trans-isomers 5 and 6 were moderately potent with Ke values of 170 nM and 260 nM, respectively. We then evaluated whether shifting the urea substituent within the ring system, as shown in compound 7, could restore activity, but this resulted in a significant loss of potency.
Importantly, this new core scaffold allows more efficient access to modification of the 3,3-diphenyl ring system without the introduction of a second chiral center ( Table 2). Conformational control is an important aspect of ligand design. Reducing conformational freedom was explored by linking the two aryl rings together (compounds 8 and 9) as well as removing an aryl ring (compound 10) and activity was lost. In order to reduce the sp2 count for SHA, we envisioned the replacement of one aryl ring with an aliphatic group. Replacement of one phenyl ring into various aliphatic substituents yielded compounds 11-15. The n-butyl compound 11 was moderately potent (K e = 165 nM), but the iso-butyl compound 14 (K e = 118 nM) was the most potent and was also anticipated to have an improved metabolic profile. Integrating various heteroatoms with hydrogen bond acceptors or donors into this portion of the molecule, such as a dimethylamino group or a hydroxy group (17)(18)(19)(20)(21), resulted in a loss of activity. In order to potentially reduce oxidative metabolism on one of the remaining unsubstituted aryl rings, we explored a series of substituted aryl rings (22-25) and found that the incorporation of naphthyl, hydroxy or benzyl ether groups (22-24) resulted in a reduction in potency. However, the 3-chlorophenyl group 25 only had a modest reduction in potency, at 136 nM compared to 14.
Our lab has previously shown that replacement of the benzyl urea with a piperidino ethyl urea on the SHA core provides analogs with good brain permeability, enhanced solubility, and favorable in vivo activity [8,10,18]. With lead compound 14 in hand, we moved forward with exploring further modifications to the urea group using the 3-isobutyl-3-phenyl substituted core (Table 3). Incorporating n-butyl urea (26), piperidino ethyl urea (27) and piperidino ethyl thiourea (28) substituents resulted in a reduction in potency. The two enantiomers of compound 14 were separated by chiral HPLC, and we determined that the plus isomer (14b) was the active enantiomer (K e = 71.4 nM, Table 3) and showed a significant right shift to the agonist activity of NPS at 3 µM concentration in calcium mobilization assay ( Figure 3). In order to evaluate if this new series was interacting with the NPSR in a similar manner as SHA, we synthesized the SHA analog with isobutyl substituted for one of the aryl rings. Compound 29 was completely inactive, suggesting that this new series binds to the NPSR in an overlapping but distinct manner.    One of the primary concerns for compound 14b was the potential to function as a Michael acceptor and non-specifically form covalent bonds to nucleophilic proteins. We therefore evaluated the ability of glutathione to add to the alpha beta unsaturated double bond of 14b (Scheme 4). Compound 14b was incubated in a solution of acetonitrile and water with glutathione, according to the method of Cevher et al. [19]. Over the course of 1, 2, 4, 6, 24, and 48 h, the solution was evaluated for the appearance of molecular ions at 711 and 713 AMU. No glutathione adduct was identified over the course of the study, indicating that the tetrasubstituted double bond is most likely too sterically crowded and does not serve as an electrophilic acceptor for reactive proteins. Considering that 14b functions as a potent NPSR antagonist and should not serve as an electrophile in vivo, its ability to access the CNS and metabolic stability were evaluated. A pharmacokinetic study of 14b was undertaken in C57Bl6 mice. Over the four hours following an i.p. dose of 30 mg/kg, blood and brain were collected at 15 and 30 min, followed by 1, 2, and 4 h ( Figure 4). A Cmax of 1938 ng/mL and 186 ng/mL was obtained in the plasma and brain, respectively. Clearance in plasma was 288 mL/min/kg, indicating a rapid clearance most likely due to oxidative metabolism. Although the clearance was higher than desired, brain levels for the duration of behavioral experiments up to 1 h were sufficiently above the Ke of 14b. Further optimization of metabolic stability will be addressed in future studies. After establishing that sufficient levels of compound 14b reached the CNS after systemic injection, it was evaluated in vivo for antagonist activity in a well-established NPSRmediated behavior; NPS-mediated hyperlocomotion [1]. In this behavioral paradigm, mice are habituated to an open arena and centrally administered NPS, which has been shown to increase locomotion in a dose-dependent and NPSR-dependent manner. In the current study, this NPS-mediated effect was attenuated with the pretreatment of systemically administered 3 mg/kg compound 14b ( Figure 5; one-way ANOVA found a significant effect, p = 0.004, with Tukey's post hoc analysis showing significant differences in the following comparisons: Veh-aCSF vs. Veh-NPS p = 0.0004, Veh-NPS vs. 14b-aCSF p = 0.0020, Veh-NPS vs. 14b-NPS p = 0.0389). However, compound 14b did not decrease locomotion on its own (3 mg/kg; Veh-aCSF vs. 14b-aCSF p = 0.8199). The data suggest that compound 14b is highly effective in blocking the effects of exogenously administered Considering that 14b functions as a potent NPSR antagonist and should not serve as an electrophile in vivo, its ability to access the CNS and metabolic stability were evaluated. A pharmacokinetic study of 14b was undertaken in C57Bl6 mice. Over the four hours following an i.p. dose of 30 mg/kg, blood and brain were collected at 15 and 30 min, followed by 1, 2, and 4 h ( Figure 4). A Cmax of 1938 ng/mL and 186 ng/mL was obtained in the plasma and brain, respectively. Clearance in plasma was 288 mL/min/kg, indicating a rapid clearance most likely due to oxidative metabolism. Although the clearance was higher than desired, brain levels for the duration of behavioral experiments up to 1 h were sufficiently above the Ke of 14b. Further optimization of metabolic stability will be addressed in future studies. Considering that 14b functions as a potent NPSR antagonist and should not serve as an electrophile in vivo, its ability to access the CNS and metabolic stability were evaluated. A pharmacokinetic study of 14b was undertaken in C57Bl6 mice. Over the four hours following an i.p. dose of 30 mg/kg, blood and brain were collected at 15 and 30 min, followed by 1, 2, and 4 h ( Figure 4). A Cmax of 1938 ng/mL and 186 ng/mL was obtained in the plasma and brain, respectively. Clearance in plasma was 288 mL/min/kg, indicating a rapid clearance most likely due to oxidative metabolism. Although the clearance was higher than desired, brain levels for the duration of behavioral experiments up to 1 h were sufficiently above the Ke of 14b. Further optimization of metabolic stability will be addressed in future studies. After establishing that sufficient levels of compound 14b reached the CNS after systemic injection, it was evaluated in vivo for antagonist activity in a well-established NPSRmediated behavior; NPS-mediated hyperlocomotion [1]. In this behavioral paradigm, mice are habituated to an open arena and centrally administered NPS, which has been shown to increase locomotion in a dose-dependent and NPSR-dependent manner. In the current study, this NPS-mediated effect was attenuated with the pretreatment of systemically administered 3 mg/kg compound 14b ( Figure 5; one-way ANOVA found a significant effect, p = 0.004, with Tukey's post hoc analysis showing significant differences in the following comparisons: Veh-aCSF vs. Veh-NPS p = 0.0004, Veh-NPS vs. 14b-aCSF p = 0.0020, Veh-NPS vs. 14b-NPS p = 0.0389). However, compound 14b did not decrease locomotion on its own (3 mg/kg; Veh-aCSF vs. 14b-aCSF p = 0.8199). The data suggest that compound 14b is highly effective in blocking the effects of exogenously administered After establishing that sufficient levels of compound 14b reached the CNS after systemic injection, it was evaluated in vivo for antagonist activity in a well-established NPSRmediated behavior; NPS-mediated hyperlocomotion [1]. In this behavioral paradigm, mice are habituated to an open arena and centrally administered NPS, which has been shown to increase locomotion in a dose-dependent and NPSR-dependent manner. In the current study, this NPS-mediated effect was attenuated with the pretreatment of systemically administered 3 mg/kg compound 14b ( Figure 5; one-way ANOVA found a significant effect, p = 0.004, with Tukey' s post hoc analysis showing significant differences in the following comparisons: Veh-aCSF vs. Veh-NPS p = 0.0004, Veh-NPS vs. 14b-aCSF p = 0.0020, Veh-NPS vs. 14b-NPS p = 0.0389). However, compound 14b did not decrease locomotion on its own (3 mg/kg; Veh-aCSF vs. 14b-aCSF p = 0.8199). The data suggest that compound 14b is highly effective in blocking the effects of exogenously administered NPS. The dose of Pharmaceuticals 2021, 14, 1024 9 of 29 NPS used in this study was one-third of that required to produce maximal effects in this paradigm; therefore, it is likely that compound 14b had adequate potency and affinity to antagonize endogenous NPS. NPS. The dose of NPS used in this study was one-third of that required to produce maximal effects in this paradigm; therefore, it is likely that compound 14b had adequate potency and affinity to antagonize endogenous NPS.

Experimental Section
All standard reagents were commercially available. Compounds were purified by column chromatography on a Teledyne Isco Rf chromatography unit and by HPLC on an Agilent-Varian HPLC system equipped with Prostar 210 dual pumps, a Prostar 335 Diode UV detector and a SEDEX75 (SEDERE, Olivet, France) ELSD detector. The HPLC solvent system was binary, water containing 0.1% trifluoroacetic acid (TFA) and solvent B (acetonitrile containing 5% water and 0.1% TFA). A semi-preparative Synergy Hydro ® RP 80A C18 column (4 µm 250 × 21.2 mm column; Phenomenex) was used to purify final compounds at 15 mL/min using a linear gradient from 5% to 50 or 60% B over 30 min. The purity of final compounds was determined using an analytical Synergy Hydro ® RP80A C18 (4 µm 250 × 4.60 mm column; Phenomenex, Torrance, CA, USA) with a linear gradient of 5-95% solvent B over 20 or 30 min at a flow rate of 1 mL/min. Absorbance was monitored at 220 nm. Cis/Trans diastereomers were separated using a semi-preparative RP HPLC YMC ODS-A (S5µm, 120Å, 20 × 250 mm; 10 mL/min) with isocratic conditions (35:65 acetonitrile/water) at 204 nm. Diastereomeric purity was determined using an analytical RP HPLC YMC ODS-A (S5µm, 120Å, 4.6 × 250 mm; 1 mL/min). Enantiomers of compound 14 were separated by analytical and semi-preparative chiral HPLC performed using a dual-pump system (Dynamax SD-300 solvent system delivery system with 25 mL pump heads), a Rheodyne injector and a Varian ProStar 330 diode-array detector (DAD) controlled by Varian Star Workstation software.
The molecular ion of final compounds was determined using a PE Sciex API 150 EX LC/MS system from Perkin Elmer (San Jose, California). Reactions were monitored by thin-layer chromatography (TLC) carried out on pre-coated 60 Å 250 mm silica gel TLC plates with F-254 indicator visualized under UV light, and developed using ceric ammonium molybdate. 1 H NMR spectra were recorded at 300 MHz on a Bruker Avance 300 Spectrospin instrument and are reported as follows: chemical shift δ in ppm (multiplicity, coupling constant (Hz), and integration. The following abbreviations were used to explain

Experimental Section
All standard reagents were commercially available. Compounds were purified by column chromatography on a Teledyne Isco Rf chromatography unit and by HPLC on an Agilent-Varian HPLC system equipped with Prostar 210 dual pumps, a Prostar 335 Diode UV detector and a SEDEX75 (SEDERE, Olivet, France) ELSD detector. The HPLC solvent system was binary, water containing 0.1% trifluoroacetic acid (TFA) and solvent B (acetonitrile containing 5% water and 0.1% TFA). A semi-preparative Synergy Hydro ® RP 80A C18 column (4 µm 250 × 21.2 mm column; Phenomenex) was used to purify final compounds at 15 mL/min using a linear gradient from 5% to 50 or 60% B over 30 min. The purity of final compounds was determined using an analytical Synergy Hydro ® RP80A C18 (4 µm 250 × 4.60 mm column; Phenomenex, Torrance, CA, USA) with a linear gradient of 5-95% solvent B over 20 or 30 min at a flow rate of 1 mL/min. Absorbance was monitored at 220 nm. Cis/Trans diastereomers were separated using a semi-preparative RP HPLC YMC ODS-A (S5µm, 120 Å, 20 × 250 mm; 10 mL/min) with isocratic conditions (35:65 acetonitrile/water) at 204 nm. Diastereomeric purity was determined using an analytical RP HPLC YMC ODS-A (S5µm, 120 Å, 4.6 × 250 mm; 1 mL/min). Enantiomers of compound 14 were separated by analytical and semi-preparative chiral HPLC performed using a dual-pump system (Dynamax SD-300 solvent system delivery system with 25 mL pump heads), a Rheodyne injector and a Varian ProStar 330 diode-array detector (DAD) controlled by Varian Star Workstation software.
The molecular ion of final compounds was determined using a PE Sciex API 150 EX LC/MS system from Perkin Elmer (San Jose, California). Reactions were monitored by thin-layer chromatography (TLC) carried out on pre-coated 60 Å 250 mm silica gel TLC plates with F-254 indicator visualized under UV light, and developed using ceric ammonium molybdate. 1 H NMR spectra were recorded at 300 MHz on a Bruker Avance 300 Spectrospin instrument and are reported as follows: chemical shift δ in ppm (multiplicity, coupling constant (Hz), and integration. The following abbreviations were used to explain multiplicities: s, singlet; d, doublet; t, triplet; q, quartet; quin, quintet; m, multiplet; br,   Compound 32 (1.12 g, 4.19 mmol, 1 equiv.) was dissolved in 1.25 M HCl in ethanol (3.35 mL, 1 equiv.). A catalytic amount of PtO 2 was added and the mixture was put under hydrogen on a parr hydrogenator at 40 psi for 4 h. The PtO 2 was removed by filtration and the solvent was removed under reduced pressure. EtOAc was added to the residue, and it was basified with aqueous sodium bicarbonate and extracted. The organic layer was dried with sodium sulfate and concentrated under reduced pressure, and the residue was purified by silica gel column chromatography to yield alkene 33 (0.90 g, 3.32 mmol, 79% yield). 1  Benzyl isocyanate (0.029 g, 0.22 mmol, 1.2 equiv.) was added to compound 33 (0.05 g, 0.184 mmol, 1 equiv.) in CH 2 Cl 2 (10 mL). The reaction was stirred at room temperature for 1 h. The reaction mixture was concentrated under reduced pressure and the resulting reside was purified via silica gel chromatography (20% EtOAc/hexane) to yield compound 14 (0.051 g, 0.13 mmol, 69% yield). 1

Conclusions
This study was designed to evaluate a new series of NPSR antagonists and provide a deeper understanding of the key features of the SHA scaffold that lead to high potency at the NPSR. The diaryl system of SHA appears to be required for high antagonist potency when the oxazolidinone core is intact. Once the oxazolidinone nitrogen is replaced with a double bond, the binding mode appears to shift in the NPSR binding site and aliphatic groups are now tolerated for one of the aryl rings. Comparison of compound 29 with 14b demonstrates that the binding sites for these two series are distinct. It was anticipated that the replacement of one aryl ring would enhance the solubility and drug-like properties of this class. Compound 14b was selected for further study based on antagonist potency and the potential to exhibit differing physiochemical properties in vivo. Pharmacokinetic evaluation of 14b indicated that it accesses the CNS in sufficient quantities but is cleared rapidly from plasma, suggesting the need to reduce oxidative metabolism. Behavioral evaluation of 14b in a mouse model of NPS-stimulated locomotor activity demonstrated that a relatively low dose of 14b (3 mg/kg i.p.) was sufficient to significantly decrease NPS (0.3 nmol)-mediated locomotor stimulation. Although 14b is approximately sevenfold less potent than SHA-68, it possessed enhanced in vivo activity compared to published reports. Typically, SHA-68 is dosed at 50 mg/kg to significantly block 0.1 nmol of NPSstimulated locomotor activity [12,20]. In this study, a significant reduction in NPS-mediated locomotor activity with a reduced dose of 14b (3 mg/kg) and a threefold higher dose of NPS (0.3 nmol) was achieved. Although it is difficult to compare behavioral results across different labs, our data suggest that further improvements in potency and clearance of 14b should result in substantially improved antagonists for the NPSR. Overall, we have now identified a novel NPSR antagonist scaffold with enhanced properties in vivo.