Discovery of N-cyclobutylaminoethoxyisoxazole derivatives as novel sigma-1 receptor ligands with neurite outgrowth efficacy in cells

Herein we reported a series of 14 novel derivatives based on the N-cyclobutylaminoethoxyisoxazole scaffold. In vitro binding studies of these compounds demonstrated their low nanomolar to subnanomolar potencies as σ1 receptor ligands, with moderate to excellent selectivity over the σ2 receptor as represented by compounds 17–30. The majority of the derivatives scored high (>4.7) in the CNS MPO appraisal system, indicating their high likelihood in penetrating the blood–brain barrier. A number of these compounds exhibited significant neurite outgrowth efficacy in N1E-115 neuronal cells and displayed excellent selectivity for σ1 receptors over the selected endogenous neurotransmitter transporters, such as DAT, NET and SERT. Among the mini-series, compound 28 (Ki σ1 = 0.2 nM, Ki σ2 = 198 nM, CNS MPO score = 5.4) emerged as a promising selective σ1 receptor ligand that warrants its further evaluation as a potential therapeutic for neurodegenerative diseases.


Introduction
2][3] Current therapeutic strategies for neurodegenerative diseases are mainly aimed to decrease CNS neuron damage or brain dysfunction by conferring neuroprotection and neurogenesis. 4,5Consequently, the development of effective therapeutic medicine or discovery of new biological targets with neurogenesis activities remains an urgent need in the treatment of neurodegenerative diseases.
During the past decades, various kinds of receptors, ion channels and signaling pathways associated with neurogenesis activities have been identied as potential therapeutic targets for neurodegenerative diseases.Among these, the s1 receptor has attracted wide attention. 6s1 receptor is one of the subtype belonging to the s receptors family which was rst discovered by Martin in 1976. 7The s1 receptor has been cloned and encoded as a protein of 223 amino acids, and specically localized at the mitochondrial-associated endoplasmic reticulum membrane. 8,9In 2016, the crystal structure of the human s1 receptor was reported in complex with two divergent ligands.The structure indicated that the human s1 receptor has a trimeric architecture with a single transmembrane domain. 102][13] In particular, neurogenesis efficacy of brain penetrant s1 receptor ligands is of a considerable interest, which has been attributed to the s1 receptor's role in the modulation of cellular trafficking. 14,157][18] Furthermore, neurogenesis function was reported to be subdued in the hippocampus of s1 receptorknockout animals. 19n increasing number of s1 receptor ligands that stimulate neurite outgrowth in cellular screening models have been reported to date.Pentazocine (1) is a classical s1 receptor ligand, capable of enhancing the neurite outgrowth in PC-12 cellular model caused by the activation of s1 receptor and this activity was antagonized by NE-100 (2), a s1 receptor antagonist. 20uscher et al. reported that the cortical neurons outgrowth upon treatment with SA4503 (3) could be prevented with s1 receptor silencing. 16In addition, some reports in the literature suggested that the neuronal outgrowth activity by s1 receptor activation is synergistic with some growth factors.PRE-084 ( 4) is an example of high affinity s1 receptor ligand that can enhance neurite outgrowth and elongation induced by NGF or EGF in PC12 cells. 21Subsequently, PRE-084 could enhance the survival of the motoneurons and improve locomotor function in murine models of ALS. 224][25] The (R)-enantiomer of RC-33 (6) has been selected as a lead compound for studies in multiple sclerosis. 26Therefore, the development of novel, brain-penetrant s1 receptor ligands with neurogenesis efficacy could nd some therapeutic use for neurodegenerative diseases (Fig. 1).
In our preliminary work in the s receptors drug discovery, we developed a series of alkoxyisoxazole derivatives that were found to be s1 receptor ligands with high affinity and selectivity over the s2 receptor. 27Based on this work, 27 we further screened some of the most potent derivatives in a cellular model of neurite outgrowth.During the course of this investigation, alkoxyisoxazole 7 exhibited synergistic neurite outgrowth effects at a level similar to that of the positive control RC-33 (Fig. 2).Encouraged by this nding, we performed further structure-activity relationship (SAR) studies on this scaffold, incorporating N-cyclobutylaminoethoxy linker that was hypothesized to be a crucial structural element for neurite outgrowth.With 7 as the starting compound, herein we reported our continued efforts in the SAR studies of the N-cyclobutylaminoethoxyisoxazole as s1 receptor ligands and their cellular evaluation in neurite outgrowth assay.

Chemistry
The synthesis of the new compounds was accomplished as shown in Scheme 1.The starting material 8 or 9 underwent a Mitsunobu reaction with previously reported methyl 3hydroxyisoxazole-5-carboxylate (10) to obtain the intermediate esters 11 and 12. Lithium borohydride reduction afforded the alcohols 13 and 14, respectively, which were converted to the corresponding iodides 15 and 16 using iodine, triphenylphosphine and imidazole as a base.Subsequent reaction of the iodides with phenol or substituted phenols under alkaline conditions gave the Boc-protected isoxazole intermediates.Finally, treatment with hydrogen chloride and reductive amination with cyclobutanone and sodium cyanoborohydride resulted in nal compounds 17-30.All nal compounds were recrystallized by HCl/ethyl acetate as hydrochloride salts.
Radioligand binding studies at s1 and s2 receptors. 28 this study, [ 3 H] pentazocine and [ 3 H]DTG (6) were used as the radioligands at s1 and s2 receptors.The binding affinity assays of compounds 17-30 are shown in Table 1.Firstly, we examined the effects of changing the substituents on the benzene ring.3-Methyl substituted compound 27 retained high affinity to parent compound 7 at s1 receptor but a 3-fold increase in the binding affinity for s2 receptor.Compound 26 with a 3-tri-uoromethyl substituent showed similar affinity for s1 receptor and signicantly decreased selectivity over the s2 receptor when compared with compound 7. Mono-halogen substitution on the benzene ring exemplied by compounds 17-20 generally retained affinity for s1 receptor compared with compound 7.However, all of these mono-halogen substituted derivatives also exhibited subdued selectivity over the s2 receptor.Notably, the 3-bromo substituted compound 19 displayed a better selectivity (K i s2/K i s1 ¼ 131) than the 4-bromo substituted compound 20.The derivatives 21-25 also maintained similar high affinities for the s1 receptor upon incorporation of two halogen atoms on the benzene ring.However, this is also accompanied with higher binding affinity to the s2 receptor.Among the disubstituted compounds, 3,5-dichloro substituted compound 24 showed a 4-fold decreased affinity for s1 receptor and more than 10-fold increased binding affinity for s2 receptor.Next, we examined the effects of removing the methyl group on the basic nitrogen atom.Gratifyingly, compound 28 showed a high affinity for s1 receptor (K i s1 ¼ 0.2 nM) and excellent selectivity on s2 receptor (K i s2 ¼ 198 nM).Compound 29 with a 3-chloro substituted benzene ring also exhibited high binding affinity for s1 receptor (K i s1 ¼ 0.4 nM) but slightly lower selectivity over s2 receptor (K i s2 ¼ 64.3 nM).Compound 30 was a by-product obtained during the synthesis of compound 28, which showed a decreased binding affinity and selectivity indicating an unfavourable steric repulsion by the cyclobutyl group.In summary, except for compound 30, most of the derivatives  exhibited high binding affinities for s1 receptor (K i s1 < 10 nM).However, all the benzene ring substituted derivatives showed various degree of decreased selectivity over the s2 receptor, implying that the non-substituted benzene ring was a crucial structural element for selectivity.The results for the secondary amines 28 and 29 indicate that the hydrogen bond donor on basic nitrogen atom was benecial for increased binding affinity to the s1 receptor.

Blood-brain barrier permeability studies by CNS MPO
In this study, all of the novel s1 receptor ligands were tested their blood-brain barrier permeability by using the central nervous system multiparameter optimization (CNS MPO) desirability tool which was reported by Travis T. Wager et al. 29,30 In CNS MPO appraisal system, six structural properties of candidates would be used such as Clog P, Clog D, MW, TPSA, HBDs and pK a that were determined as important factors in the BBB permeability.Each property was valued between 0 to 1 and weighted equally.Final collective score ranged from 0 to 6, with the higher scores of CNS MPO correlating with desirable brain permeability.Through the analysis of CNS MPO scores on more than hundreds of the drugs and original candidates acting on the CNS, a majority of them had CNS MPO desirability scores greater than 4. 30 As showed in Table 2, majority of the newly    synthesized alkoxyisoxazoles obtained CNS MPO scores greater than 4. In particular, compound 28 scored the highest CNS MPO scores (5.4), in addition to the observed high binding affinities at s1 receptor (Table 1).

Neuritogenesis studies in N1E-115 cells
Following the radioligand binding studies and CNS MPO calculations, all of the synthesized derivatives were further tested in cellular neuritogenesis studies.In this study, N1E-115 neuronal cells were used as a screening platform for assessing the neuritogenesis effects of candidate compounds. 31,32All the derivatives and positive control (compound 7 and RC-33) were assessed at 10 mM concentration in DMSO, and vehicle group using DMSO alone.The test results are shown in Fig. 3. Gratifyingly, a large proportion of the drug administration signicantly increased the average neurite length of N1E-115 cells at 10 mM concentration vs. vehicle group, suggestive of a neuritogenesis effect.Through the analysis of test results, most of the disubstituted derivatives such as compound 22-25 exhibited greater neuritogenesis efficacy when compared with vehicle control.The 4-uoro substituted compound 18 and 3-triuoromethyl substituted compound 26 displayed better efficacy than other mono-substituted compounds.The higher efficacies observed for compounds 28 and 29 indicated that the secondary amine moiety was benecial for promoting neuritogenesis efficacy within this series.Particularly, we obtained six compounds (18, 22, 24, 25, 28 and 29) that showed preferable neurite outgrowth effects compared to the parent compound 7 and positive control RC-33 in N1E-115 neuronal cells.

Selectivity studies at selected neurotransmitter transporters
In our previous work, we found that some derivatives based on the alkoxyisoxazole scaffold not only bind to s1 receptor but also to common neurotransmitter transporters, such as DAT, NET and SERT. 27,33In order to exclude the potential interference of other receptor pathways to the observed neuritogenesis efficacy, we assessed the binding affinities of compounds 7, 18, 22, 24, 25, 28 and 29 at the biogenic amine transporters (NET, DAT, and SERT).Gratifyingly, no appreciable binding affinity was observed for these transporters using 10 mM concentration of tested compounds.This set of results indicated that the selected compounds had excellent selectivity for s1 receptor over DAT, NAT and SERT (Table 3).

Conclusions
In summary, 14 novel derivatives based on the N-cyclobutylaminoethoxyisoxazole scaffold were designed and synthesized.Most of the derivatives showed potent binding affinities to the s1 receptor (K i < 10 nM) and good selectivity over the s2 receptor in radioligand binding tests.CNS MPO scores were calculated for predicting their blood-brain barrier permeability and the results showed most of derivatives obtained favourable CNS MPO scores greater than 4. All of the compounds were assessed for their neurite outgrowth efficacy in N1E-115 cells.Compounds 18, 22, 24, 25, 28 and 29 exhibited signicant neuritogenesis effects at 10 mM concentration vs. vehicle group and superior efficacy of increasing the average neurite length compared with compound 7 and RC-33.Moreover, compounds 18, 22, 24, 25, 28 and 29 showed great selectivity for s1 receptor over DAT, NET, and SERT transporters.Among these derivatives, compound 28 was found as the most promising compound with excellent in vitro binding prole (K i s1 ¼ 0.2 nM, K i s2 ¼ 198 nM), high CNS MPO score (5.4) and signicant neuritogenesis efficacy, which will be further developed as potential therapeutics for neurodegenerative diseases.

Experimental section
Chemistry General methods.Unless otherwise specied, commercial reagents and solvents were all of analytical grade or of chemical purity (>99%).Anhydrous THF was obtained by distillation over sodium wire, respectively.All reactions were run under a nitrogen atmosphere and all reaction vessels were oven-dried.The TLC was performed on silica gel GF254.Column chromatographic purication was carried out using silica gel (200-300 mesh). 1 H and 13 C NMR spectra were recorded on a Bruker Advance III 400 spectrometer at 400 MHz ( 1 H) and 101 MHz ( 13 C).Chemical shis are reported in d (ppm) using the d 0 signal of tetramethylsilane (TMS) as internal standards.High resolution mass spectra were performed using a Bruker ESI-TOF high-resolution mass spectrometer.Purities of nal compounds were established by Agilent 1200 HPLC system with a ZORBAX Eclipse XDB-C18 column, with detection at 220 or 254 nm on a variable wavelength detector G1365D; ow rate ¼ 1.4 mL min À1 ; gradient of 0 to 100% methanol in water (both containing 0.05 vol% of TFA) in 25 min.
General procedure for the Mitsunobu reaction (method A).Diisopropyl azodicarboxylate (2.0 mmol) was added dropwise into a stirred solution of 8 or 9 (1.0 mmol), 10 (1.2 mmol), and PPh 3 (2.0 mmol) in anhydrous THF (20 mL) at 0 C under the nitrogen.Aer stirring overnight at rt, the solvent was evaporated, and the residue was dissolved in EtOAc (30 mL).The organic layer was washed with water (20 mL) and brine (15 mL), dried over Na 2 SO 4 and evaporated in vacuo.The residue was puried by ash chromatography to give the product.
General procedure for the reduction reaction (method B).LiBH 4 (4 mmol) was added into a solution of compound 11 or 12 (0.8 mmol) in anhydrous THF (20 mL) with ice cooling under nitrogen.Aer stirring overnight at rt, the reaction was quenched by saturated aqueous NH 4 Cl solution with ice cooling.The mixture was extracted with EtOAc (3 Â 30 mL).The organic layer was washed with water (20 mL) and brine (15 mL), dried over Na 2 SO 4 and evaporated in vacuo.The residue was puried by ash chromatography to give the product.
General procedure for the preparation of iodides (method C).I 2 (1.2 mmol) was added into a stirred solution of compound 13 or 14 (0.7 mmol), imidazole (2.0 mmol), and PPh 3 (1.2mmol) in anhydrous DCM (10 mL) with ice cooling under nitrogen.Aer reacting completely at rt, the solvent was evaporated.The residue was puried by the ash chromatography to give the product.
General procedure for the preparation of phenyl ethers (method D).K 2 CO 3 (6.0mmol) was added into a stirred solution of compound 15 or 16(1.0 mmol) and phenol or substituted phenol (2.0 mmol) in anhydrous DMF (4 mL) under nitrogen.
Aer stirring overnight at rt, water (30 mL) was added.The mixture was extracted with EtOAc (2 Â 30 mL), and the combined organic layers were washed with water (3 Â 20 mL) and brine (20 mL), dried over Na 2 SO 4 and evaporated in vacuo.The residue was puried by the ash chromatography to give the phenyl ether product.
General procedure for the deprotection (method E).HCl/ EtOAc (4 mol L À1 , 2 mL) was added into a stirred solution of the N-Boc protected intermediates (1.0 mmol) in CH 2 Cl 2 (5 mL) under nitrogen with ice cooling.The mixture was stirred overnight at rt. Aer the solvent was evaporated, the residue was by ether (20 mL).The resulting solid was ltered to give the HCl salt.
General procedure for the reductive amination (method F).NaCNBH 4 (0.6 mmol) was added into a stirred solution of compound primary or secondary amine (0.4 mmol), cyclobutanone (0.5 mmol) and acetic acid (0.05 mL) in CH 3 OH (10 mL) at rt under nitrogen.Aer stirring overnight at rt, the solvent was evaporated and the residue was extracted with EtOAc (3 Â 10 mL).The combined organic layer was washed with water (20 mL) and brine (15 mL), dried over Na 2 SO 4 and evaporated in vacuo.The residue was puried by ash chromatography.To a solution of the N-alkyl compound in CH 2 Cl 2 (3 mL) was added HCl/EA (4 mol L À1 , 1 mL) under N 2 with ice cooling.The mixture was stirred overnight at rt.The solvent was evaporated to give the HCl salt.
For experimental details please refer to the PDSP website http://pdsp.med.unc.edu/. 28neral procedures for cellular evaluation NGF-induced neurite outgrowth in PC-12 cells.PC-12 cells, from rat pheochromocytoma cell lines, were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).Undifferentiated PC12 cells were cultured in DMEM complete medium with 10% heat-inactivated horse serum, 5% fetal bovine serum, 100 U mL À1 penicillin, and 100 mg mL À1 streptomycin and maintained at 37 C in a humidied incubator supplemented with 5% CO 2 .To induce differentiation, PC12 cells were seeded at a density of 1 Â 10 4 cells per mL on a poly-L-lysine-coated 96-well plate and cultured in complete medium containing 2.5 ng mL À1 NGF or different tested compounds (10 mM) concomitant with 2.5 ng mL À1 NGF administration.Four days aer incubation, morphometric analysis was performed on digitized images of live cells taken under phase contrast illumination with an inverted microscope linked to a camera.Images of ve elds per well were taken with an average of 20 cells per eld.The number of differentiated cells was determined by counting cells that had at least one neurite with a length equal to the cell body diameter and was expressed as a percentage of the total cells in the eld.The counting was performed in a blinded manner.All experiments were performed at least ve times.
Neurite outgrowth in N1E-115 cells.N1E-115 cells, the mouse neuroblastoma cell lines, were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).Undifferentiated N1E115 cells were cultured in DMEM complete medium with 10% fetal bovine serum, 100 U mL À1 of penicillin, and 100 mg mL À1 of streptomycin and maintained at 37 C in a humidied incubator supplemented with 5% CO 2 .N1E-115 cells were seeded at a density of 1 Â 10 4 cells per mL on poly-L-lysine-coated 96-well plates and grown with DMSO or different tested compounds (10 mM) in DMSO solution for four days.Four days aer incubation, morphometric analysis was performed on digitized images of live cells taken under phase contrast illumination with an inverted microscope linked to a camera.Images of ve elds per well were taken with an average of 100 cells per eld.Cells that had at least one neurite with a length that was twice as long as the body diameter were expressed as a percentage of the total cells in the eld.The counting was performed in a blinded manner.All experiments were performed at least ve times.

Fig. 2
Fig. 2 The potential effect of compound 7 (10 mM) and RC-33 (10 mM) on NGF (2.5 ng mL À1 )-induced neurite outgrowth model in PC12 cells.a See Experimental section.Each histograms represent the mean AE SEM of at least five different experiments.Statistically significant differences: **p < 0.01 vs. control using NGF alone.

1 a
See Experimental section.DAT: dopamine transporter; NET: norepinephrine transporter; SERT: serotonin transporter.Radioligands: DAT: [ 3 H] WIN35428; NET: [ 3 H] nisoxetine; SERT: [ 3 H] citalopram.bThe K i values for GBR12909, desipramine, amitriptyline and compound 7 are cited from the literature. 27c NA: not active, dened as <50% binding in the primary assay at 10 mM.K i values were determined for those targets where the binding efficacy at 10 mM was greater than 50%.d Data are expressed as average of K i values and each performed from three independent experiments.

Table 1
Binding affinities of derivative at s1 and s2 receptors

Table 2
30S MPO scores of compound 7 and all the derivatives a The data sources and explanations.Clog P: calculated partition coefficient; Clog D: calculated distribution coefficient at pH 7.4; MW: molecular weight; TPSA: topological polar surface area; HBDs: number of hydrogen-bond donors: pK a : most basic center.The calculated physicochemical properties of the derivatives were obtained using standard commercial soware: ACD/Laboratories, version 6.0 for Clog P, pK a , HBDs and Clog D at PH 7.4, ChemBioDraw, version 14.0 for MW and TPSA.bLast CNS MPO scores were calculated by the tools reported by Travis T. Wager et al.30 a