Development and Pharmacological Characterization of Selective Blockers of 2-Arachidonoyl Glycerol Degradation with Efficacy in Rodent Models of Multiple Sclerosis and Pain.

: We report the discovery of compound 4a , a potent β -lactam-based monoacylglycerol lipase (MGL) inhibitor characterized by an irreversible and stereoselective mechanism of action, high membrane permeability, high brain penetration evaluated using a human in vitro blood − brain barrier model, high selectivity in binding and a ﬃ nity-based proteomic pro ﬁ ling assays, and low in vitro toxicity. Mode-of-action studies demonstrate that 4a , by blocking MGL, increases 2-arachidonoylglycerol and behaves as a cannabinoid (CB1/CB2) receptor indirect agonist. Administration of 4a in mice su ﬀ ering from experimental autoimmune encephalitis ameliorates the severity of the clinical symptoms in a CB1/CB2-dependent manner. Moreover, 4a produced analgesic e ﬀ ects in a rodent model of acute in ﬂ ammatory pain, which was antagonized by CB1 and CB2 receptor antagonists/inverse agonists. 4a also relieves the neuropathic hypersensitivity induced by oxaliplatin. Given these evidence, 4a , as MGL selective inhibitor, could represent a valuable lead for the future development of thera peutic options for multiple sclerosis and chronic pain. proteomics studies, pharmacokinetic pro ﬁ ling of 4a , analgesic e ﬀ ect of 4a in the mouse hot plate test, determination of the absolute con ﬁ guration of (3 R ,4 S )-6a , 1 H and 13 C NMR spectra of compounds ( ± )- 4a , ( ± )- 5a , and ( ± )- 6a , elemental analysis results for tested compounds (PDF)


* S Supporting Information
ABSTRACT: We report the discovery of compound 4a, a potent β-lactam-based monoacylglycerol lipase (MGL) inhibitor characterized by an irreversible and stereoselective mechanism of action, high membrane permeability, high brain penetration evaluated using a human in vitro blood−brain barrier model, high selectivity in binding and affinity-based proteomic profiling assays, and low in vitro toxicity. Mode-of-action studies demonstrate that 4a, by blocking MGL, increases 2-arachidonoylglycerol and behaves as a cannabinoid (CB1/CB2) receptor indirect agonist. Administration of 4a in mice suffering from experimental autoimmune encephalitis ameliorates the severity of the clinical symptoms in a CB1/CB2-dependent manner. Moreover, 4a produced analgesic effects in a rodent model of acute inflammatory pain, which was antagonized by CB1 and CB2 receptor antagonists/inverse agonists. 4a also relieves the neuropathic hypersensitivity induced by oxaliplatin. Given these evidence, 4a, as MGL selective inhibitor, could represent a valuable lead for the future development of therapeutic options for multiple sclerosis and chronic pain.

■ INTRODUCTION
Multiple sclerosis (MS) is a neuroinflammatory autoimmune disease characterized by demyelination and axonal damage leading to loss of cerebral functions. 1 The treatment landscape of MS includes several anti-inflammatory agents and monoclonal antibodies for immunotherapy. Besides demyelination and inflammation, thought to play the major role in MS progression, central neuronal defects also occur and have been associated with the onset of symptoms such as spasticity and pain.
We describe herein the development of a set of β-lactam-based inhibitors (4a−h, Chart 1) of human MGL which led to the identification of 4a as the prototypic compound of this class of analogues. 4a was characterized as a novel, ultrapotent hMGL inhibitor, with high selectivity toward the AEAinactivating enzyme, fatty acid amide hydrolase (FAAH), other serine hydrolases from the rat brain proteome, and CBRs. The unique profile of this prototypic inhibitor combines selectivity, potency, and favorable pharmacokinetic properties, highlighting its potential as a novel MGL inhibitor for further development ( Table 1). As expected for a potent and selective MGL inhibitor, in vivo administration of 4a reduced the clinical severity of the disease in mice suffering from experimental autoimmune encephalomyelitis (EAE), a rodent demyelinating disease model. 17 In vivo tests confirmed the analgesic profile of 4a in mouse models and evidenced a clear-cut dependence of its pharmacological efficacy upon the increased 2-AG levels and subsequent indirect CBRs modulation. Remarkably, the efficacy of 4a further supports the hypothesis 2 of a tight intersection between the endocannabinoid system and MS.

■ RESULTS AND DISCUSSION
Rational Design, Structure−Activity Relationships, and Molecular Modeling Studies. The MGL-2 cocrystal structure 22 defined the structural basis for MGL inhibition. This complex also shed light on the size of the lipophilic portion of the MGL binding pocket, which is of particular interest for designing novel and selective enzyme ligands. An in-depth examination of this X-ray structure, combined with the analysis of the structural elements of other diphenylmethane inhibitors, allowed us to develop a novel 3D pharmacophore model and to rationally design innovative MGL inhibitors based on a β-lactam skeleton (Chart 1, and Figure S1 of the Supporting Information). In particular, to improve the binding properties over current inhibitors such as 1, we chose the trans-3,4-diarylsubstituted β-lactam structural motif to obtain the correct distance and reciprocal orientation of the aromatic rings (see Supporting Information for details). In this frame, the combination of the 4-fluorophenyl and the methylene-3,4-dioxyphenyl moieties provides the key hydrophobic interactions with the enzyme (Supporting Information, Figure S1) and guarantees an improved  drug-like profile. The diaryl β-lactam system of compounds 4a−h represents a novelty in the field of MGL inhibition because the potent inhibitors so far described (compounds 1−3) all share a diphenylmethane-based structure.
The experimental data shown in Table 1 confirmed that the β-lactam is a key substructure for potent and selective MGL inhibition. When compound 4a was tested in vitro for its ability to inhibit human and rat MGL, it resulted in being a very potent inhibitor of hMGL (IC 50 = 7.4 nM, Table 1) and a picomolar inhibitor of rat MGL (rMGL, IC 50 = 250 pM). Selectivity toward human recombinant FAAH was measured, and 4a proved to be more than 380-fold selective over human FAAH (hFAAH) ( Table 1). Notably, the two enantiomers showed a stereoselective interaction with hMGL (Table 1), the eutomer (3R,4S)-(+)-4a being nearly 7 times more potent than the distomer (3S,4R)-(−)-4a. The selectivity versus hFAAH was maintained, with both enantiomers still being highly selective for hMGL (Table 1).
Further in vitro tests revealed that 4a does not significantly interact (IC 50 > 10 μM) with either CB1 or CB2 receptors in displacement assays carried out with a high affinity radiolabeled ligand of such receptors. A time-course experiment showed that 4a inhibits hMGL with an IC 50 = 195.8 nM (Table 2) when the assay was conducted without the standard 10 min preincubation, thus suggesting an irreversible mechanism of action.
We developed and tested a small set of analogues of 4a (4b−h, Table 1) in order to investigate: (i) the role played by the leaving group linked to the piperidine ring (4b−d, Table 1) and (ii) the effect of a small number of focused substitutions aimed at modifying steric hindrance and electronic properties of the aromatic systems placed on the β-lactam skeleton (4e−h, Table 1).
When studying the mechanism of action of irreversible MGL inhibitors, we applied the concept advanced by Aaltonen and coworkers 16 according to whom the effects of the leaving group are strictly dependent upon its physicochemical properties, although steric hindrance may also play a role. In fact, we found that the effect on potency seems to be related to the pK a of the leaving group's conjugated acid (ideal pK a ∼ 10). Accordingly, the best profile among compounds 4a−d was achieved with the triazole (pK a = 10), while the imidazole (pK a = 14) provided a 1500-fold  Experimental UV and ECD spectra of (+)-trans-6a (solid line) compared to the PBE0/TZ2P//B97D/TZ2P calculated spectra of (3R,4S)-6a (dashed line). Computational details are given in Experimental Section and Supporting Information.  reduction in potency (4b vs 4a, Table 1), the p-nitrophenoxy group (1) (pK a = 7) displayed a 200-fold drop of potency (4c vs 4a), and the hexafluoroisopropyl alcohol (pK a = 9) gave an analogue (4d) 17 times less potent than 4a.
The modification of steric hindrance and electronic properties of the aromatic systems placed on the β-lactam skeleton was also crucial for potency and selectivity of our MGL inhibitors. The replacement of the 4-fluoro of 4a with a methoxy group led to inhibitor 4e with a slight loss of inhibitory potency (IC 50 = 13.2 nM, Table 1). Compound 4f, in which the methylene-3,4dioxyphenyl moiety was replaced by a dimethoxyphenyl system, resulted in a 10 times less potent analogue (IC 50 = 61.4 nM, Table 1) and displayed the lowest degree of selectivity toward FAAH (IC 50 rFAAH/IC 50 hMGL = 12). Substitution (with methoxy groups) at para and meta positions on both the aromatic rings (4g) was detrimental and led to a sensible decrease of MGL inhibition potency (IC 50 = 251.3 nM). By contrast, a decreased steric hindrance, by substitution (with methoxy groups) at the para positions only (4h), triggered a partial recovery of inhibitory potency (IC 50 = 23.3 nM).
Taken together, these studies point out to the 4-F-phenyl system at C3 of the β-lactam skeleton and a methylene-3,4dioxyphenyl moiety at C4, combined to the triazole system, (4a) as the optimal structural arrangement for both potency and selectivity.
Further studies were dedicated to gain insight into the mechanism of action of 4a. We applied a molecular deconstruction procedure aimed at defining the essential structural determinants contributing to the high potency of 4a (Chart 2). To this aim, synthetic intermediates 5a and 6a were also tested in order to assess the importance of the leaving triazole group. In line with the hypothesized irreversible mechanism of action, both amines 5a and 6a lacked inhibitory activity toward MGL (Table 1). We also explored the role of the azetidinone system by synthesizing the azetidine 7. This compound, displaying a 2-fold drop in inhibitory potency (IC 50 = 376 nM), underlined the pivotal role played by the β-lactam amide carbonyl (Table 1). Finally, the isolated 3,4-trans-diaryl-β-lactam system (8) and the piperidine triazole urea disconnection unit 9 were found to be extremely weak MGL inhibitors (IC 50 > 10 μM). Overall, molecular deconstruction confirmed the original design hypothesis leading to 4a as a synergistic combination of key structural elements for the fine-tuning of potency.
The interaction of 4a with rMGL was also studied by combining top-down and bottom-up proteomics experiments ( Figure 2) (see Supporting Information for details). These studies, which were consistent with a covalent interaction of 4a with MGL, also indicated that the covalent MGL−4a adduct was achieved with 1:1 stoichiometry, under the assay conditions. Furthermore, the outcome of the bottom-up proteomic analyses, performed to map the binding site of 4a in MGL, indicated that the carbamoyl adduct of the tryptic peptide with 4a ( Figure 2C) contains six serine residues, including the catalytic Ser122. Tandem mass analysis allowed us to unequivocally assign Ser122 as the site of modification induced by the nucleophilic attack on the urea moiety of 4a (Figures S2−S5; see Supporting Information for details).
The stereoselective interaction of 4a with hMGL was investigated by docking studies (Glide, version 5.7, Schrodinger, LLC, New York, NY, 2011). The eutomer (3R,4S)-4a shows a different pattern of interactions with respect to the distomer  = −117.39 kcal/mol). Notably, the key interactions with His269, Ala51, and Met123, the two latter residues being constituents of the oxyanion hole, are relevant for MGL inhibition. 9 Our computational analysis also highlights the peculiar interactions of both enantiomers into the hMGL binding site, with the catalytic residues (Ser122 and His269 for (3R,4S)-4a and only Ser122 for (3S,4R)-4a) and with the oxyanion hole residues (Ala51 and Met123 for (3R,4S)-4a and only Ala51 for (3S,4R)-4a). Proteomic analysis proved the formation of a carbamoylated adduct of the enzyme with 4a. This data suggests the formation of an initial tetrahedral intermediate which was also probed, in silico, by a covalent docking approach. A tetrahedral complex generated by using GOLD software 23 for (3R,4S)-4a and (3S,4R)-4a in complex with hMGL ( Figure 3C) highlighted that the binding mode found from classical docking studies ( Figure 3A−C) is maintained and (3R,4S)-4a and (3S,4R)-4a similarly establish a H-bond with Gly50 (oxyanion hole residue), deeply positioned into the binding site. Consequently, we can assume that stereoselectivity of the interaction of 4a enantiomers with hMGL can rely on the different enantiomer/enzyme reciprocal recognition at the early stage of interactions, as explained by the docking studies ( Figures 3A,B).
hFAAH/hMGL selectivity ratio (Table 1)  Affinity-Based Proteomic Profiling (ABPP) Assay. Compound 4a was found to potently inhibit MGL with an IC 50 of ∼10 nM in a competitive ABPP assay carried out in the rat brain proteome and did not exhibit cross-reactivity with other rat brain serine hydrolase off-targets up to 500 nM, including FAAH, ABHD6, and ABHD12 ( Figure 5A). Only at concentrations ≥500 nM some inhibition of lysophopholipase A1/2 was observed ( Figure 5B).
Preliminary in Vitro ADME+T Profiling. Besides its in vitro profile, additional features that contribute to designate 4a as an excellent MGL inhibitor are the lack of mutagenic effect in the Salmonella typhimurium strains TA98 and TA100 at any of the concentrations tested (following the assay conditions of the Ames' test, Figure 6) along with its experimentally determined physicochemical features. 4a showed a favorable solubility and a chemical stability profile at both neutral and acidic pH (Supporting Information, Figure S7A). Metabolic stability to hCYP3A4 and parallel artificial membrane permeability assay confirmed the stability and high permeability through artificial membranes of 4a (Supporting Information, Figures S7B,C).
Brain penetration, evaluated by using a human in vitro blood− brain barrier model, is high (P app = 47 × 10 −6 cm/s ± 6.81), showing that 4a may easily cross the human blood−brain barrier. Notably, the measured P app for Lucifer yellow, the nonpermeant molecule used as marker of endothelium integrity and coincubated with 100 μM 4a (see Experimental Section for details), is low (P app = 9.48 × 10 −6 cm/s ± 1.68) and is in the same range found when Lucifer yellow is administered alone. This data indicates that transport of 4a across the blood−brain barrier is not due to any toxic effect on the endothelium. Cardiovascular adverse effects heavily contribute to drug withdrawals from the market and represent one of the major hurdles in the development of new drugs. To evaluate the potential cardiovascular toxicity of compound 4a, its effect on cardiac mechanical function and electrocardiogram (ECG) in Langendorff-isolated rat hearts was assessed, as previously described. 24 Under control conditions, left ventricle pressure (LVP) and coronary perfusion pressure (CPP) values of 60.07 ± 8.96 and 57.45 ± 2.58 mmHg (n = 4), respectively, were obtained. At the maximum concentration tested (1 μM), 4a significantly increased CPP to 75.19 ± 7.80 mmHg (n = 4, ** P < 0.01, repeated measures ANOVA and Dunnett's post test). Moreover, 4a, up to 1 μM, did not affect surface ECG (Table 3).
At the maximum concentration tested, 4a increases coronary perfusion pressure leaving unaltered heart contractility and ECG, thus allowing us to exclude cardiac toxicity for this compound.
In Vivo Efficacy of Compound 4a. Efficacy of 4a on the Experimental Autoimmune Encephalomyelitis (EAE) Model of MS in Rodents. Although no single animal model exists recapitulating all the features of MS, EAE is considered an animal model suitable for studying it 25 because it shares some important features with human MS, such as the chronic course and histopathological evidence of demyelination and neuroinflammation. 26 We therefore evaluated 4a in the myelin glycoprotein 35−55 (MOG 35−55 )-induced EAE rodent model (the disease was induced as detailed in the Experimental Section). Drug prophylactic treatment (4a, 3 mg/kg, ip) started at the asymptomatic stage of disease at day 6 postimmunization (dpi), up to the acute stage of the disease (21 ± 0.1 dpi, when the maximal score gravity is observed). Administration of 4a clearly influenced the disease progression, as assessed by the significant lower daily clinical score observed in 4a-administered EAE mice when compared to vehicletreated EAE mice starting from 8 dpi, as well as by the dramatic change in the cumulative score level ( Figure 7). Interestingly, we found that the reduction of the daily clinical score induced by   prophylactic 4a could involve the modulation of the endocannabinoid system. Actually, administration (starting from 6 dpi for 14 days) of the selective CB1 antagonist/inverse agonist N-(piperidin-1-yl)-5-(4-iodophenyl)-1-(2,4-dichlorophenyl)-4methyl-1H-pyrazole-3-carboxamide 20 (AM251) 27 and the CB2 antagonist/inverse agonist 6-iodo-2-methyl-1-[2-(4-morpholinyl)ethyl]-1H-indol-3-yl](4-methoxyphenyl)methanone 21 (AM630), 28 each used at 1 mg/kg, ip, both prevented the protecting effect of 4a starting at 12 dpi. Although activation of both CB1 and CB2 receptors by endogenous 2-AG could be beneficial for the course of the disease, 2 we found that the blockade of each one of the two receptors significantly reduced the positive effects exerted by 4a, leading us to conclude that in 4a-treated EAE mice both receptors did not compensate but synergize with each other mediating the 4a-induced amelioration of the clinical score. The respective contribution of CB1 and CB2 in ameliorating the course of the demyelinating disorder in EAE mice chronically administered 4a deserves further investigation to be clarified.
Changes in spontaneous locomotor activity were not observed in 4a-administered control (nonimmunized) mice when compared to control (not shown).
Following these encouraging results, we performed histological evaluation of myelin using the Luxol Fast Blue staining ( Figure 8). Pallor of Luxol Fast Blue stain is associated with a reduction in the density of staining for myelin. Representative images of spinal cord of each experimental group are shown in Figure 8. In the spinal cord white matter of the control group, a normal structural organization of myelin is evident. By contrast, in EAE mice demyelinated areas are present (arrows). The EAE mice treated with 4a exhibited myelin-density staining that was a Each value represents mean ± SEM (n = 4). HR, frequency; RR, cycle length; PQ, atrioventricular conduction time; QRS, intraventricular conduction time; QT, duration of ventricular depolarization and repolarization, i.e., the action potential duration; QTc, corrected QT. Figure 7. In vivo effects of compound 4a on EAE mice. Female C57BL/6J mice were immunized with the MOG 35-55 peptide and were randomly assigned to the following group: untreated EAE mice, 4a (3 mg/kg ip) treated EAE mice; 4a (3 mg/kg ip)/20 (1 mg/kg)-treated EAE mice; 4a (3 mg/kg ip)/21 (1 mg/kg) treated EAE mice. Mice were administered drugs daily. 20 and 21 were administered 10 min before 4a. Clinical score is expressed as average (media ± SEM) daily score (A−C) or as cumulative score (D). Data are means ± SEM from untreated EAE mice (A, white square, 7 animals), from 4a (3 mg/kg ip) treated EAE mice (A−C white circle, 9 animals); from 4a (3 mg/kg ip)/20 (1 mg/kg) treated EAE mice (B, black triangle, 4 animals); from 4a (3 mg/kg ip)/21 (1 mg/kg) treated EAE mice (C, black square, 6 animals). *P < 0.05 vs EAE + vehicle 1 dpi; ∧ P < 0.05 vs EAE + vehicle in the same day; ∧∧ P < 0.01 vs EAE + vehicle in the same day; ∧∧∧ P < 0.001 vs EAE + vehicle.
comparable to that of control animals. Treatment with 4a plus 20 or 21 revealed the presence of some demyelinated areas in comparison to control animals and to 4a-treated groups, but these areas were less extended than in EAE group ( Figure 8).
In Figure 9A, a panel of representative microglia staining in the spinal cord is shown. Cells were stained with Iba1 antibody. Immunofluorescence staining for microglia revealed an increase of Iba1-positive cells/optic field in the white matter anterior fasciculus of EAE mice (247.66 ± 11.50). The results of the quantitative measurements of microglia density are shown in Figure 9B. 4a administration in EAE mice caused a reduction in the number of Iba1 positive cells in the area evaluated that was, at least in part, reverted by coadministration of 20 or 21 ( Figure 9A,B).
Antinociceptive Effects of 4a in Animal Models of Inflammatory and Neuropathic Pain. The in vivo antinociceptive profile of 4a was also evaluated in two different animal models of pain. The compound inhibited both the first and second phase of the nocifensive response to formalin in mice (model of acute inflammatory pain). The maximum effect was reached at the dose of 0.3 mg/kg (ip) ( Figure 10A). MGL inhibition may also produce therapeutic actions by impeding the conversion of 2-AG into arachidonic acid and hence prostaglandins, thus acting in an endocannabinoid-and CBR-independent manner. 29 However, we found that the effect of 4a in the formalin test was mediated by both CB1 and CB2 receptors because both the selective CB1 antagonist/inverse agonist 20 and the CB2 antagonist/inverse agonist 21 (each used at 1 mg/kg, ip), administered 10 min prior to 4a, antagonized the antinociceptive actions of the compound ( Figure 10B). Importantly, 5 min after the administration of 4a (0.3 mg/kg, ip), the levels of 2-AG in the spinal cord and paw skin of mice coadministered with formalin, measured by LC-MS, were found to be elevated (Table 4, see Experimental Procedures for details). These results suggest that compound 4a alleviates the pain behavior in the formalin test by elevating endogenous 2-AG levels (Table 4), thus indirectly activating both CB1 and CB2 receptors, as previously reported for other MGL inhibitors. 10 Next we evaluated the effect of 4a in a mouse model of nociceptive behavior caused by the chemotherapeutic agent, oxaliplatin (OXP, Figures 7C,D) 30 (model of neuropathic pain which is also a common symptom in MS patients). After oral administration, the inhibitor dose-dependently reversed the lowering of the threshold to cold stimuli (cold plate test) induced by OXP, after acute administration on day 14 of OXP treatment, when changes in pain thresholds were fully established ( Figure 10C). As shown in Figure 10D, 4a maintained its efficacy after repeated treatments following a preventive protocol (daily oral administration with OXP). Moreover, when subchronically  administered, 4a was efficacious 24 h after the last administration. These antinociceptive properties of 4a are related to conditions of hypersensitivity, such as neuropathic pain, because 4a did not modify normal pain thresholds evaluated in the hot-plate test (Supporting Information, Table S1).
Collectively, the in vivo results are well consistent with the theory that the endocannabinoid signalosome 31 may play a role in the modulation of the symptoms of MS 32 as well as in the molecular events underlying nociception, thus pointing to selective inhibition of MGL as a potential strategy to produce beneficial effects against the clinical outcomes of MS.

■ CONCLUSIONS
Combining bioinformatics and molecular modeling efforts we generated an updated 3D pharmacophore model for the development of selective hMGL inhibitors. Synthetic accessibility, high potency and selectivity, and an interesting preliminary pharmacokinetic profile, including low geno/cardio-toxicity, characterize the novel β-lactam inhibitor 4a. We proved a stereoselective inhibition of MGL with the (3R,4S)-4a enantiomer (Figure 3), being 70 times and 8 times more potent than 1 and its distomer, respectively, against the human enzyme (4a IC 50 hMGL = 4 nM vs 1 IC 50 hMGL = 285 nM) ( Table 1). The mechanism of action (specific interaction with the catalytic Ser122) was assessed by top-down and bottom-up proteomics experiments while the high selectivity was assessed in binding experiments on CBRs and on a proteome-based assay of hydrolases including FAAH and ABHD ( Figure 5). 4a lacks mutagenic effect in the S. typhimurium strains TA98 and TA100 ( Figure 6) and cardiotoxicity as assessed by ECG measurements in Langendorff-isolated rat hearts. Additionally, 4a showed favorable solubility, chemical stability at both neutral and acidic pH, metabolic stability to hCYP3A4 and high permeability  through artificial membranes, and high brain penetration. Antinociceptive effects on inflammatory and neuropathic pain animal models were assessed after oral and ip administration of 4a. After ip administration of 0.3 mg/kg, a marked analgesic action was registered, with inhibition of both the first and second phase of the nocifensive response to formalin. Oral administration of 4a already at 1 mg/kg dose dependently reversed the lowering of the threshold to cold stimuli (cold plate test) induced by OXP, indicating its efficacy in the treatment of neuropathic pain.
In vivo experiments evidenced a clear-cut dependence of the pain reliever profile of 4a from endocannabinoid system activation ( Figure 7). 4a was also tested on the EAE animal model (ip administration, Figure 8), 4a elicited a striking beneficial effect on disease progression that could involve both CB1-and CB2-mediated signaling. Interestingly, in animals administered with either CBR antagonists/inverse agonists (20 and 21), the seriousness of the clinical symptoms was delayed when compared to control EAE mice, reaching however a maximum clinical score (at 21 dpi) that was comparable to that detected in untreated EAE mice (at 17 dpi). In line with these data, in EAE mice treated with 4a, the histological evaluation of myelin, by Luxol Fast Blue staining, demonstrated a significant reduction of the demyelinated areas which was partially counteracted by the coadministration of 20 and 21 ( Figure 8). Notably these outcomes were paralleled by those obtained when looking at microglia because 4a decreased the number of Iba1 positive cells, an effect that is in part prevented by coadministration of 20 or 21 ( Figure 9). Because EAE is a disease model widely applied to study MS, these data could suggest MGL inhibition as an innovative therapeutic approach for treating MS symptoms.

■ EXPERIMENTAL SECTION
Chemistry. Unless otherwise specified, materials were purchased from commercial suppliers and used without further purification. Reaction progress was monitored by TLC using silica gel 60 F254 (0.040−0.063 mm) with detection by UV. Silica gel 60 (0.040− 0.063 mm) or aluminum oxide 90 (0.063−0.200 mm) were used for column chromatography. 1 H NMR and 13 C NMR spectra were recorded on a Varian 300 MHz spectrometer by using the residual signal of the deuterated solvent as internal standard. Splitting patterns are described as singlet (s), doublet (d), triplet (t), quartet (q), quintet (p), and broad (br); the value of chemical shifts (δ) are given in ppm and coupling constants (J) in hertz (Hz). ESI-MS spectra were performed by an Agilent 1100 series LC/MSD spectrometer. Melting points were determined in Pyrex capillary tubes using an Electrothermal 8103 apparatus and are uncorrected. Optical rotation values were measured at room temperature using a PerkinElmer model 343 polarimeter operating at = 589 nm, corresponding to the sodium D line. Yields refer to purified products and are not optimized. All moisture-sensitive reactions were performed under argon atmosphere using oven-dried glassware and anhydrous solvents. ESI-MS spectra for exact mass determination were performed on a LTQ Orbitrap Thermo Fischer Scientific instrument. All final compounds were purified by flash column chromatography, and the purity of all compounds tested was ≥ 95% as determined by elemental analysis. Elemental analyses were performed in a PerkinElmer 240C elemental analyzer, and the results were within   [1,3]dioxol-5-yl)-3-(4-methoxyphenyl)azetidin-2-one (4e). Title compound was obtained starting from 6b as described for compound (±)-4a. The crude was purified by means of chromatography on silica gel (1:1 ethyl acetate/n-hexane) to afford pure title compound (30 mg, 57% yield) as an amorphous white solid. 1  trans-1-(1-(1H-1,2,4-Triazole-1-carbonyl)piperidin-4-yl)-3,4-bis(4methoxyphenyl)azetidin-2-one (4h). Title compound was obtained starting from 6d as described for compound (±)-4a. The crude was purified by means of chromatography on silica gel (1:1 ethyl acetate/ n-hexane) to afford title compound (24 mg, 47% yield) as an amorphous white solid. 1
Incubation of MGL with 4a. Purified rMGL was incubated with 4a at a molar ratio of 1:10, for 1 h at 37°C. The incubation buffer was the same used for the final MGL purification (buffer A plus imidazole). A reference incubation with DMSO was also included.
Intact MGL Analysis. After incubation, intact MGL was centrifuged (10000g for 5 min). A small aliquot of the supernatant was diluted 10-fold in water added with 0.1% formic acid and was analyzed by LC-MS using a Synapt G2 qTOF mass spectrometer coupled with an Acquity UPLC system. The protein was loaded on a BEH C4 column (1 mm × 100 mm) and eluted with a linear gradient of acetonitrile in water (both added with 0.1% formic acid). Column and LC-MS/MS system were purchased from Waters, Milford, USA. Mass spectra were acquired in positive ion mode in the 500−4000 m/z range. Leucine encephalin was infused in the ion source at 2 ng/mL as lock mass for spectra recalibration.
Bottom-Up Analysis of MGL. After the incubation process reported above, MGL was digested with trypsin (proteomic grade, Sigma, Milano, Italia) at 1:50 ratio (w/w) at 37°C for 12 h. After digestion, the sample was centrifuged at 10000g for 5 min. A small aliquot of the supernatant was diluted 10-fold in water added with 0.1% formic acid and was analyzed by LC-MS/MS using the same system described above. The column was a BEH C18 (1 mm × 100 mm) purchased from Waters. Peptides were eluted with a 12 min, 3−60% gradient of acetonitrile in water. Tandem mass data were acquired in data dependent acquisition mode, selecting multiple charged states as precursors. Collision energy was automatically set by the acquisition software (Supporting Information, Figure S3). We incubated the purified MGL with 4a, or control DMSO, at a molar ratio of 1:10 for 1 h at 37°C. Then, the proteins were digested with trypsin at 37°C for 12 h, and the tryptic fragments were analyzed by LC-MS/MS as described above (Supporting Information, Figure S4). We compared the mass spectra of the native peptide (m/z = 1067, red) and the modified peptide from the 4a-treated MGL (m/z = 1146, green). Compared to the fragments from naive MGL, 4a adducts display a mass increase of 395 Da (Supporting Information, Figure S5). The amino acid sequence shown on the top corresponds to MGL tryptic peptide 110−160, bearing six serine residues. MS/MS spectra of m/z 1067 (naïve, bottom) and m/z 1146 (4a adduct, top) both display the y fragment ion series for a part of the peptide, as indicated in Figure S5. From this analysis, three Ser residues are excluded from being the binding site of compound 4a (Supporting Information, Figure S6). Following a closer investigation of the tandem mass spectra, two internal acylium ions were found. These ions are unmatched between MS/MS spectra of m/z 1067 (naïve) and m/z 1146 (4a adduct). In particular, the presence of internal acylium ion EVPVFLLGHS(+395)MGG in the 4a−adduct tandem mass spectrum only indicates that compound 4a binds to the catalytic Ser122 as represented in Figure 2C.
Chiral Resolution of Amine (±)-6a. Racemic mixture (±)-6a was separated by using a cellulose-carbamate (OD) column (Daicel) eluted with a mixture of 50% of 2-propanol in n-hexane as eluent, flow 1 mL/min, injection volume 50 μL. This method allowed us to efficiently split the two enantiomers that were collected and analyzed before the calculation of the α-value. The first peak of the chromatogram corresponds to the (+)-enantiomer ([α] 20 D = +68.88), and the second one was the (−)-enantiomer ([α] 20 D = −69.05) (Supporting Information, Figure S2). Both the fractions were analyzed after collection to check their optical purity. The analytical conditions were identical to those above-described for the racemate separation. In Supporting Information, Figure S2 are reported the chromatograms obtained for respectively (+)-6a and (−)-6a.
Stereochemical Characterization of (+)-trans-6a. Experimental ECD and UV spectroscopic analysis was carried out at 25°C on a Jasco (Tokyo, Japan) J-810 spectropolarimeter equipped with a PTC-423S Peltier-type temperature control system, using a 2 nm spectral bandwidth, a 50 nm min −1 scanning speed, and a 2 s data integration time; spectra were averaged over three accumulation cycles. Quartz cells (Hellma, Milan, Italy) with a 1 mm path length were used to measure spectra in the 320−200 nm spectral range. Samples for electronic ECD and UV spectroscopic analysis on (+)-trans-6a were prepared using HPLC-grade 2-propanol (Sigma-Aldrich, Milan, Italy) at a 135 μM concentration.
Computational Spectroscopy. The theoretical chiroptical properties of (3R,4S)-6a (Supporting Information, Figure S8) were determined according to the standard protocol for stereochemical characterization by time-dependent density functional theory (TD-DFT) calculations. 33 Figure  S9, while thermodynamic data are reported in Supporting Information, Table S4; the optimized structure of the lowest-energy conformer of (3R,4S)-6a and (R,S)-6a13, is depicted in Supporting Information, Figure S10. TD-DFT calculations were also carried out using the Gaussian 09 software (Gaussian09, Revision C.01; Gaussian Inc.: Wallingford CT, 2010). The PBE0 functional 37,38 was used in combination with the TZ2P basis set and the IEFPCM solvation model for 2-propanol; calculations were performed on all optimized conformers. Theoretical values of oscillator strength ( f j ), rotational strength in dipole velocity formalism (R j ), and excitation energies (expressed as wavelengths, λ j ) were calculated for the 50 lowest-energy electronic transitions of each optimized conformer and are reported in Supporting Information, Tables S5 and S6. The theoretical spectra of optimized conformers were then derived by approximation of f j and R j values to Gaussian bands with a Δσ value of 0.25 eV. 39 The theoretical UV and ECD spectra of (3R,4S)-6a were finally derived as the weighted average of the contribution of all conformers, according to their equilibrium populations at 298.15 K and 1 atm (χ) as determined by Boltzmann statistics based on relative free energies (ΔG) and compared to the experimental spectra ( Figure 1).
Computational Studies. All calculations performed in this work were carried out on two Cooler Master Centurion 5 (Intel Core-i5 Quad CPU Q6600 @ 2.40 GHz) with Ubuntu 10.04 LTS (long-term support) operating system running Maestro 9.2 (Schrodinger, LLC, New York, NY, 2011) and GOLD 5.2 (The Cambridge Crystallographic Data Centre, Cambridge, UK). All the pictures presented in this study were generated by PyMOL (PyMOL v1.6-alpha; Schrodinger, LLC, New York, 2013).
Molecules Preparation. Three-dimensional structures of both enantiomers of compound (±)-4a were built by means of Maestro (Maestro, version 9.2, Schrodinger, LLC, New York, NY, 2011). Molecular energy minimizations were performed by means of MacroModel using the Optimized Potentials for Liquid Simulationsall atom (OPLS-AA) force field 2005. 40 The solvent effects are simulated using the analytical Generalized-Born/Surface Area (GB/SA) model, 41 and no cutoff for nonbonded interactions was selected. Polak−Ribiere conjugate gradient (PRCG) method with 1000 maximum iterations and 0.001 gradient convergence threshold was employed. Compound 4a was treated by LigPrep application (LigPrep, version 2.5, Schrodinger, LLC, New York, NY, 2011), implemented in Maestro suite 2011, generating the most probable ionization state of any possible enantiomers and tautomers at cellular pH value (7 ± 0.5).
Protein Preparation. The three-dimensional structures of the hMGL and hFAAH were taken from PDB (hMGL PDB ID, 3PE6; 42 hFAAH PDB ID, 3PPM 43 ) and imported into Schrodinger Maestro molecular modeling environment. The structures were submitted to Protein Preparation Wizard implemented in Maestro suite 2011 (Protein Preparation Wizard workflow 2011). This protocol, through a series of computational steps, allowed us to obtain a reasonable starting structure of the proteins for molecular docking calculations by a series of computational steps. In particular, we performed three steps to (1) add hydrogens, (2) optimize the orientation of hydroxyl groups, Asn, and Gln, and the protonation state of His, and (3) perform a constrained refinement with the impref utility, setting the max RMSD of 0.30. The impref utility consists of a cycles of energy minimization based on the impact molecular mechanics engine and on the OPLS_2005 force field. 40 Noncovalent Docking. Noncovalent docking studies were carried out by Glide (Grid-Based Ligand Docking with Energetics) (Glide, version 5.7, Schrodinger, LLC, New York, 2011) using the ligands and proteins prepared as above-mentioned, applying Glide extra precision (XP) method, as previously reported. 44 Energy grids were prepared using default value of protein atom scaling factor (1.0 Å) within a cubic box centered on the catalytic Ser122 (for hMGL) and Ser241 (for hFAAH, Figure 5). After grid generation, the ligands were docked into the enzymes with default parameters (no constraints were added). The number of poses entered to postdocking minimization was set to 50. Glide XP score was evaluated.
Covalent Docking. Concerning the covalent docking available in GOLD 5.2 program, 23 the representative poses of the most populated cluster of docked solutions obtained by Glide for both enantiomers of compound 4a into hMGL enzyme were redocked into binding site, through a covalent linkage between electrophilic carbon atom close to the triazole moiety and the hydroxyl side chain of the catalytic residue (Ser122). In fact, both protein and ligand files were set up with the link atom included. During docking runs, the link atom in the ligand is forced to fit onto the link atom in the protein. To ensure that the geometry of the bound ligand was correct, an angle-bending energy term for the link atom was included in the calculation of the fitness score. 45 Both enantiomers of compound 4a was docked 50 times with early termination if the top three poses were within 1.5 Å RMSD. The obtained hMGL-(3R,4S)-4a and hMGL-(3S,4R)-4a tetrahedral complexes was energy minimized using MacroModel using the Optimized Potentials for Liquid Simulations-All Atom (OPLS-AA) force field 2005. 40 The solvent effects are simulated using the analytical Generalized-Born/ Surface Area (GB/SA) model, 41 and no cutoff for nonbonded interactions was selected. Polak−Ribiere conjugate gradient (PRCG) method with 100000 maximum iterations and 0.001 gradient convergence threshold was employed.
Free-Binding Energies Calculation. The Prime/MM-GBSA method implemented in Prime software (Prime, version 3.0, Schrodinger, LLC, New York, 2011) consists in computing the change between the free and the complex state of both the ligand and the protein after energy minimization. The technique was used on the docking complexes (hMGL-(3R,4S)-4a and hMGL-(3S,4R)-4a; hFAAH-(3R,4S)-4a and hFAAH-(3S,4R)-4a) in order to calculate the free-binding energy (ΔG bind ). Prime/MM-GBSA was used employing the calculation of ligand strain energies and using the minimization as sampling method, defining as flexible residues, those comprised in 10 Å from the ligand.
Selectivity in an Affinity-Based Proteomic Profiling (ABPP) Assay. Rat brain membranes were prepared according to previously reported methods and diluted to 1 mg/mL prior to use. 46 Proteomes (50 μL) were preincubated with either DMSO or 1−1000 nM concentrations of inhibitor at 37°C. After 20 min, FP-Rh (1.0 μL, 50 μM in DMSO, a kind gift of Ben Cravatt) was added and the mixture was incubated for another 30 min at 37°C. Reactions were quenched with SDS loading buffer (12.5 μL, 5×) and run on SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis). 47 Following gel imaging, serine hydrolase activity was determined by measuring fluorescent intensity of gel bands corresponding to MAGL, ABHD6, and FAAH using ImageJ 1.43u software.
Mutagenicity Studies and Pharmacokinetic Profiling Procedures. Ames Test. Ames test was performed as previously described. 48 Solubility and Chemical Stability. Standard and sample solutions are prepared from a 10 mM DMSO stock solution using an automated dilution procedure. For compound 4a, three solutions are prepared; one to be used as standard and the other two as test solutions.
• rt under orbital shaking to achieve "pseudo-thermodynamic equilibrium" and to presaturate the membrane filter. Product suspensions/solutions are then filtered using centrifugation, diluted 1:2 with the same buffer solution, and analyzed by UPLC/UV/TOF-MS, using UV-detection at 254 nm for quantitation. An aliquot of the pH 7.4 solution is transferred in a microplate, left for 24 h at rt, and analyzed by UPLC/UV/TOF-MS for the chemical stability assay.
Metabolic Stability on Human CYP3A4. Compound 4a in 10 mM DMSO solution was added to an incubation mixture in a 96-well microplate containing 20 pmol/mL of hCYP3A4 (0.1−0.2 mg/mL protein). The mixture is split in two aliquots: one receiving a NADPH regenerating system, the other an equal amount of phosphate buffer. The final substrate concentration is 1 μM along with 0.25% of organic solvent. Incubation proceeds for 1 h at 37°C and is then stopped by addition of MeCN to precipitate proteins.
Permeability Assay (PAMPA). Compound 4a (10 μM in HBSS + Hepes buffer) are added to the donor chamber and incubated (together with a reference compound, warfarin, to verify membrane integrity) for 4 h at 37°C and 80% humidity. Concentrations of reference, donor, and acceptor solutions are measured by UPLC-MS-TOF (references, donor, and acceptor are injected in UPLC-MS in this order for compare the MS quantitative signal).
Brain Penetration. Transport across the blood−brain barrier was tested using a human in vitro blood−brain barrier model consisting of a coculture of endothelial cells derived from hematopoietic stem cells and pericytes. 49 After 6 days of coculture, compound 4a (100 μM in Ringer solution + Hepes) was added to the luminal side of endothelial cells (together with a nonpermeant compound, Lucifer yellow, to verify endothelium integrity) and incubated for 20, 40, and 60 min at 37°C and 80% humidity. Concentrations were measured at luminal and abluminal sides using fluorescence detection (synergy H1 multiplates reader; BioTek) for Lucifer yellow and LC-UV analysis for 4a. P app was determined according the equation P app = J/ACo, where J is the rate of appearance of the drug at the abluminal side, Co is the initial concentration at the luminal side, and A is the surface area of the filter. 50 Isolated Rat Heart Preparation and Perfusion. All animal care and experimental protocols conformed to the European Union Guidelines for the Care and the Use of Laboratory Animals (European Union Directive 2010/63/EU) and were approved by the Italian Department of Health (666/2015-PR). Male Sprague−Dawley rats (300 g; Charles River Italia, Calco, Italy) were anaesthetized (ip) with a mixture of Zoletil 100 (7.5 mg kg −1 tiletamine and 7.5 mg kg −1 zolazepam; Virbac Srl, Milano, Italy) and Xilor (4 mg/kg xylazine; Bio 98, San Lazzaro, Italy) containing heparin (5000 U/kg), decapitated, and bled. The hearts, spontaneously beating, were rapidly explanted and mounted on a Langendorff apparatus for retrograde perfusion via the aorta at a constant flow rate of 10 mL/min with a Krebs−Henseleit solution of the following composition (mM): NaCl 118, KCl 4.7, CaCl 2 2.5, MgSO 4 1.2, NaHCO 3 25, KH 2 PO 4 1.2, glucose 11.5, Na pyruvate 2, and EDTA 0.5, bubbled with a 95% O 2 −5% CO 2 gas mixture (pH 7.4), and kept at 37°C, as described elsewhere. 51 The hearts were allowed to equilibrate for at least 20 min before drug exposure.
Heart contractility was measured as left ventricle pressure (LVP) by means of latex balloon, inserted into the left ventricle via the mitral valve and connected to a pressure transducer (BLPR, WPI, Berlin, Germany). The balloon was inflated with deionized water from a microsyringe until a left ventricular end diastolic pressure of 10 mmHg was obtained.
Alteration in coronary perfusion pressure (CPP), arising from changes in coronary vascular resistance, were recorded by pressure transducer (BLPR, WPI, Berlin, Germany) placed in the inflow line.
A surface electrocardiogram (ECG) was recorded at a sampling rate of 1 kHz by means of two steel electrodes, one placed on the apex and the other on the left atrium of the heart. The ECG analysis included the following measurements: RR (cycle length), HR (frequency), PQ (atrioventricular conduction time), QRS (intraventricular conduction time), and QT (overall action potential duration). 52 LVP, CPP, and ECG were recorded with a digital PowerLab data acquisition system (PowerLab 8/30; ADInstruments, Castle Hill, Australia) and analyzed by using Chart Pro for Windows software (PowerLab; ADInstruments, Castle Hill, Australia).
LVP was calculated by subtracting the left ventricular diastolic pressure from the left ventricular systolic pressure. As the QT interval is affected by heart rate changes (e.g., it shortens with rapid heart rate), Bazett's formula (QTc = QT/(RR) 1/2 ) was routinely used to avoid confounding effects.
Compound 4a was dissolved in DMSO. Solvents failed to alter the response of the preparations (data not shown).
Statistical Analysis. Data are reported as mean ± SEM; n (indicated in parentheses) represents the number of rat hearts. Analysis of data was accomplished using GraphPad Prism version 5.04 (GraphPad Software, U.S.A.). Statistical analyses and significance as measured by repeated measures ANOVA (followed by Dunnett's post test) were obtained using GraphPad InStat version 3.06 (GraphPad Software, U.S.A.). In all comparisons, P < 0.05 was considered significant.
Pharmacological in Vivo Studies. Animal handling was carried out according to the European Community guidelines for animal care (DL 116/92, application of the European Communities Council Directive 86/609/EEC). The ethical policy of the University of Florence and the Second University of Naples conforms with the Guide for the Care and Use of Laboratory Animals of the U.S. National Institutes of Health (NIH Publication no. 85-23, revised 1996; University of Florence assurance number A5278-01). Formal approval to conduct the experiments described herein was obtained from the animal subjects review board of the University of Florence and Second University of Naples.
Formalin Experiments. For the formalin experiments, male CD-1 mice received formalin (1.25% in saline, 30 μL) in the dorsal surface of one side of the hind paw. Each mouse was randomly assigned to one of the experimental groups (n = 6) and placed in a plexiglass cage and allowed to move freely for 15−20 min. A mirror was placed at a 45°a ngle under the cage to allow full view of the hind paws. Lifting, favoring, licking, shaking, and flinching of the injected paw were recorded as nociceptive responses. 53 The total time of the nociceptive response was measured every 5 min and expressed in min (mean ± SEM). Recording of nociceptive behavior commenced immediately after formalin injection and was continued for 60 min. Mice received vehicle (0.5% DMSO in saline) or different doses of 4a (0.1 or 0.3 mg/kg, ip) 10 min before formalin injection. For the combination experiments, 20 or 21 (1 mg/kg, ip) were injected before 4a. For the oxyplatin experiment, male Swiss albino mice (23−25 g) were used. The animals were fed with a standard laboratory diet and tap water ad libitum and kept at 23 ± 1°C with a 12 h light/dark cycle, light on at 7 a.m.
Extraction, Purification, and Analysis of 2-AG Tissue Levels. Methods. Tissues (paw skin or spinal cords) from vehicle/saline, vehicle/formalin, compound 1a/saline, and compound 4a/formalintreated mice were homogenized in 5 volumes of chloroform/methanol/ Tris HCl 50 mM (2:1:1) containing 20 pmol each of d5−2-AG. Homogenates were centrifuged at 13000g for 16 min (4°C), and the aqueous phase plus debris was collected and extracted again twice with 1 volume of chloroform. The organic phases from the three extractions were pooled and the organic solvents evaporated in a rotating evaporator. Lyophilized samples were then stored frozen at −80°C under nitrogen atmosphere to be analyzed later. Lyophilized extracts were resuspended in chloroform/methanol 99:1 by volumes. The solutions were then purified by open bed chromatography on silica as described. 54 Fractions eluted with chloroform/methanol 9:1 by volume (containing 2-AG) were collected, the excess solvent evaporated with a rotating evaporator, and aliquots analyzed by isotope dilution liquid chromatography/atmospheric pressure chemical ionization/mass spectrometry (LC-MS) carried out under conditions described previously and allowing the gross purification of 2-AG. Mass spectrometric detection was carried out in the selected ion monitoring mode using m/z values of 384 and 379 (molecular ions +1 for deuterated and undeuterated 2-AG). The areas of the peaks corresponding to the 1(3)-and 2-isomers of 2-AG were added together. The amounts of the compounds were expressed as picomol/milligram of tissue.
Oxaliplatin Experiments. Oxaliplatin, a third-generation platinum analogue, has become a first-line chemotherapy in metastatic colorectal cancer and a valid option as adjuvant therapy in several types of cancer. 55 The major dose-limiting side effect is a painful neuropathy that persists between cycles 56 and correlated with characteristic alterations of the nervous system. Oxaliplatin neuropathy was induced in mice administering 2.4 mg/kg oxaliplatin intraperitoneally (ip) for 5 consecutive days every week for 2 weeks (we followed Cavaletti protocol 57 adapted to mice). Oxaliplatin was dissolved in 5% glucose solution. Control animals received an equivalent volume of 5% glucose ip (vehicle). Behavioral tests were performed on day 14. 4a (1−30 mg/kg) was suspended in the vehicle (1% carboxymethylcellulose) and per os (po) acutely administered on day 14. Alternatively, 4a (10 mg/kg) was daily administered starting from the first day of oxaliplatin treatment to day 13; behavioral measurements were performed 24 h after the last administration. For analgesia measurements performed by Hot plate test, animals were acutely po treated (4a, 3 and 10 mg/kg).
Cold and Hot Plate Tests. The cold plate test was performed placing animals in a stainless box (12 cm × 20 cm × 10 cm) with a cold plate as floor. The temperature of the cold plate was kept constant at 4°C ± 1°C. Pain-related behaviors (i.e., lifting and licking of the hind paw) were observed, and the time (s) of the first sign was recorded. The cutoff time of the latency of paw lifting or licking was set at 60 s.
Hot-plate test (Supporting Information, Table S1) was carried out accordingly with O'Callaghan and Holzman. 58 Mice were placed inside a stainless steel container, thermostatically set at 52.5 ± 0.1°C in a precision water-bath from KW Mechanical Workshop, Siena, Italy. Reaction times (s) were measured with a stop-watch before and at regular intervals up to a maximum of 60 min after treatment. The end point used was the licking of the fore or hind paws. Before treating animals with 4a, a pretest was performed: those mice scoring below 12 and over 18 s were rejected. An arbitrary cutoff time of 45 s was adopted.
Animals and EAE Induction. Female mice (C57BL/6J; 18−20 g, 6−8 weeks) were obtained from Charles River (Calco, Italy) and were housed in the animal facility of DIFAR, Section of Pharmacology and Toxicology (authorization no. 484 of June 8, 2004). Female mice were immunized according to a standard protocol previously described, 59 with minor modifications. Briefly, animals were subcutaneously injected with incomplete Freund's adjuvant containing 4 mg/mL Mycobacterium tuberculosis (strain H37Ra) and 200 μg of the MOG 35-55 peptide. Immunization with MOG 35-55 was followed by ip administration of 250 ng of pertussis toxin on day 0 and after 48 h. Clinical scores (0 = healthy, 1 = limp tail, 2 = ataxia and/or paresis of hindlimbs, 3 = paralysis of hindlimbs and/or paresis of forelimbs, 4 = tetraparalysis, 5 = moribund or death) were recorded daily. All efforts were made to minimize animal suffering and to use the minimal number of animals necessary to produce reliable results.
All the experimental procedures described here were in accordance with the Italian legislation (no. 26, March 4, 2014) and the European legislation (2010/63/UE, September 22, 2010) and the ARRIVE guidelines, and they were approved by the Italian Ministry of Health (protocol number no. 50/2011-B). Experiments were performed proteomics studies, pharmacokinetic profiling of 4a, analgesic effect of 4a in the mouse hot plate test, determination of the absolute configuration of (3R,4S)-6a, 1 H and 13 C NMR spectra of compounds (±)-4a, (±)-5a, and (±)-6a, elemental analysis results for tested compounds (PDF) Molecular formula strings (CSV) ■ AUTHOR INFORMATION Corresponding Author