Analgesic and Anticancer Activity of Benzoxazole Clubbed 2-Pyrrolidinones as Novel Inhibitors of Monoacylglycerol Lipase

Ten benzoxazole clubbed 2-pyrrolidinones (11–20) as human monoacylglycerol lipase inhibitors were designed on the criteria fulfilling the structural requirements and on the basis of previously reported inhibitors. The designed, synthesized, and characterized compounds (11–20) were screened against monoacylglycerol lipase (MAGL) in order to find potential inhibitors. Compounds 19 (4-NO2 derivative) and 20 (4-SO2NH2 derivative), with an IC50 value of 8.4 and 7.6 nM, were found most active, respectively. Both of them showed micromolar potency (IC50 value above 50 µM) against a close analogue, fatty acid amide hydrolase (FAAH), therefore considered as selective inhibitors of MAGL. Molecular docking studies of compounds 19 and 20 revealed that carbonyl of 2-pyrrolidinone moiety sited at the oxyanion hole of catalytic site of the enzyme stabilized with three hydrogen bonds (~2 Å) with Ala51, Met123, and Ser122, the amino acid residues responsible for the catalytic function of the enzyme. Remarkably, the physiochemical and pharmacokinetic properties of compounds 19 and 20, computed by QikProp, were found to be in the qualifying range as per the proposed guideline for good orally bioactive CNS drugs. In formalin-induced nociception test, compound 20 reduced the pain response in acute and late stages in a dose-dependent manner. They significantly demonstrated the reduction in pain response, having better potency than the positive control gabapentin (GBP), at 30 mg/kg dose. Compounds 19 and 20 were submitted to NCI, USA, for anticancer activity screening. Compounds 19 (NSC: 778839) and 20 (NSC: 778842) were found to have good anticancer activity on SNB-75 cell line of CNS cancer, exhibiting 35.49 and 31.88% growth inhibition (% GI), respectively.


Introduction
Endocannabinoids (endogenous ligands), cannabinoid (CB) receptors, and proteins for their biological synthesis and degradation constitute the endocannabinoid system (ECS) [1]. Endocannabinoids are biosynthesized from the membrane phospholipids [2]. Endocannabinoid, N-arachidonoyl ethanolamine (AEA, Anandamide) functions as partial agonist on CB 1 and CB 2 receptors. It has low affinity for CB 2 and moderate affinity for CB 1 . Endocannabinoids, 2-arachidonoylglycerol (2-AG) function as full agonist and have moderate affinity for both the receptors. Interestingly, 2-AG is the major endocannabinoid and is found to be approximately 170-fold higher in concentration than AEA, in the brain [3]. AEA and 2-AG hydrolysis and degradation are facilitated by fatty acid amide hydrolase moderate affinity for both the receptors. Interestingly, 2-AG is the major endocannabinoid and is found to be approximately 170-fold higher in concentration than AEA, in the brain [3]. AEA and 2-AG hydrolysis and degradation are facilitated by fatty acid amide hydrolase (FAAH) and monoacylglycerol lipase (MAGL) enzymes, correspondingly [1]. MAGL (an α/β-hydrolase) hydrolyzes 2-AG into glycerol and free fatty acid (FFA), by the action of the catalytic triad organized with Ser122, Asp239 and His269 amino acids in the active site of the enzyme [4]. MAGL hydrolyzes approximately 85% of total 2-AG in the CNS [5] and generates arachidonic acid (AA), giving rise to neuro-inflammatory PGE2 and PGD2 prostaglandins [6]. MAGL inhibitors ameliorates neuropathic pain by increasing 2-AG and decreasing neuro-inflammatory prostaglandins in the CNS (Figure 1) [7]. In addition, 2-AG demonstrated analgesic activity by acting on CB1 receptors in the CNS and periphery [8][9][10][11]. It is reported that JJKK-048 (MAGL, IC50 363 pM) exhibited analgesic activity in tail immersion and writhing test [12]. MAGL inhibition has shown significant neuroprotective and anti-inflammatory potential in Parkinson's and Alzheimer's disease [13,14]. PF-06795071 (IC50 3 nM), a MAGL inhibitor, is reported to have considerable antineuroinflammatory potential [15]. Another MAGL inhibitor, ABX-1431 (IC50 14 nM) is under clinical trials for broad range of CNS disorders like Tourette syndrome [16]. Many more research findings encouraged that the inhibitors of MAGL has therapeutic potentials in pain and CNS disorders [17][18][19][20]. The governing importance of MAGL in abnormal lipolysis in cancer has been demonstrated [21]. MAGL was originally known for its lipolytic action on monoacylglycerols from stored triacylglycerols into glycerol and free fatty acids (FFA) [22]. Cancer cells utilizes this lipolytic pathways for their hastened proliferation [23]. The cancer-supporting action of MAGL is due to elevated FFA levels. This MAGL-FFA pathway promotes in vivo tumor growth by increasing FFA-derived oncogenic signaling lipids (PA, LPA, S1P, and PGE2) [18]. These protumorigenic lipid mediators encourage tumor growth, angiogenesis, The governing importance of MAGL in abnormal lipolysis in cancer has been demonstrated [21]. MAGL was originally known for its lipolytic action on monoacylglycerols from stored triacylglycerols into glycerol and free fatty acids (FFA) [22]. Cancer cells utilizes this lipolytic pathways for their hastened proliferation [23]. The cancer-supporting action of MAGL is due to elevated FFA levels. This MAGL-FFA pathway promotes in vivo tumor growth by increasing FFA-derived oncogenic signaling lipids (PA, LPA, S1P, and PGE 2 ) [18]. These protumorigenic lipid mediators encourage tumor growth, angiogenesis, and metastasis in cancer ( Figure 1) [24]. MAGL is reported to expressed vastly in aggressive type of cancer cells and is associated with pathogenesis, proliferation, and in vivo tumor growth. MAGL inhibition disrupts cancer cell proliferation, growth and metastasis [25][26][27]. The anticancer effect of MAGL inhibition in prostate cancer was totally abolished by cotreat-ment with SR141716 (rimonabant; CB 1 receptor antagonist) and fatty acids, signifying that amplified endocannabinoid action and reduced stock of FFA from MAGL inhibition is the reason behind antitumor effect [26].
In continuation of our work on MAGL inhibitors, we have designed novel 2-pyrrolidinone linked benzoxazole derivatives and screened them for analgesic and anticancer effects.

Molecular Docking Study
Most active and selective compounds identified by MAGL and FAAH inhibition assay (19 and 20), were docked at the catalytic center of MAGL, in order to get an insight of their binding pattern, with the help of XP Glide docking using Maestro (Schrodinger). Compounds 19 and 20, showed comparable docking scores of −9.87 and −9.83, respectively. The binding of compounds 19 and 20 in the active site of MAGL revealed that the carbonyl group of pyrrolidinone is located exactly in the oxyanion hole and stabilized by three hydrogen bonds (~2Å) with alanine 51, serine 122, and methionine 123. Serine 122 is one of the critical amino acid residues of the catalytic triad of MAGL. The benzoxazole moiety is found to be positioned in the amphiphilic pouch, having π-π stacking contact with the amino acid Tyr194. The 4-NO 2 (19) and 4-SO 2 NH 2 (20) phenyl ring of the ligands were involved in hydrophobic (van der Waals) attractions with the amino acids, leucine 148, 213, and 241. In addition, the 3D and 2D binding pattern of compounds 19 and 20 in the catalytic location of MAGL is depicted in Figure 4. the value reported (IC50 = 4.6 nM) [48]. The outcomes of the experiments are presented in Table 1.

Molecular Docking Study
Most active and selective compounds identified by MAGL and FAAH inhibition assay (19 and 20), were docked at the catalytic center of MAGL, in order to get an insight of their binding pattern, with the help of XP Glide docking using Maestro (Schrodinger). Compounds 19 and 20, showed comparable docking scores of −9.87 and −9.83, respectively. The binding of compounds 19 and 20 in the active site of MAGL revealed that the carbonyl group of pyrrolidinone is located exactly in the oxyanion hole and stabilized by three hydrogen bonds (~2Å) with alanine 51, serine 122, and methionine 123. Serine 122 is one of the critical amino acid residues of the catalytic triad of MAGL. The benzoxazole moiety is found to be positioned in the amphiphilic pouch, having π-π stacking contact with the amino acid Tyr194. The 4-NO2 (19) and 4-SO2NH2 (20) phenyl ring of the ligands were involved in hydrophobic (van der Waals) attractions with the amino acids, leucine 148, 213, and 241. In addition, the 3D and 2D binding pattern of compounds 19 and 20 in the catalytic location of MAGL is depicted in Figure 4.

Pharmacokinetic and Physicochemical Characteristics
To investigate the potential of the identified derivatives (19 and 20) to cross the selectively permeable membranes of hematoencephalic barrier (BBB), to develop orally active CNS drugs, their pharmacokinetic and physicochemical features were computed by QikProp (ADMET predictor) of Schrodinger. Guidelines, concerning the validation and optimization of orally active CNS compounds, were developed by Ghose et. al., by analyzing 35 characteristic features of orally bioavailable 317 CNS and 626 non-CNS drugs [49]. This guideline states that in order to design high-quality CNS drugs, the molecule must qualify by the following parameters: TPSA less than 76 Å 2 (ideally 25-60 Å 2 ), number of N atoms between 1-2, comprising 1 aliphatic amine, 2-4 side chains on/outside rings, number of polar H atoms < 3 (ideally 0-1), SASA 460-580 Å 2 , molecular volume 740-970 Å 3 , and must have +ve QikProp CNS property. Remarkably, most of the properties of compounds 19 and 20, computed by QikProp, were found to be in the qualifying range as per the proposed guideline ( Table 2). The properties of compound 19 were found to be within the qualifying range except dipole moment. For compound 20, 5 out of 35 properties is just slightly above the upper qualifying limit. Most importantly, qualifying limits for CNS active drugs in terms of TPSA is from 3.8 to 109, and the calculated TPSA for compound 20 was found to be 119.08. Therefore, the designing of more potent MAGL inhibitors having physicochemical and pharmacokinetic properties within the preferred CNS limits is required.

In Silico Absorption and Toxicity Profile
The selected compounds (19 and 20) were evaluated for their absorption and toxicity profile by a bioinformatics tool admetSAR [50]. The results suggested that both the compounds have high blood-brain barrier (BBB) penetration properties as well as high chance of human intestinal absorption. In AMES test, compound 19 was found to be mutagenic, while 20 was non-mutagenic. Carcinogenicity test revealed that both the compounds were non-carcinogens. The LD50 values in rat were also evaluated, a compound with high value is considered as less lethal. The LD50 for compounds 19 and 20 were found to be 2.30 and 2.21 mol/kg, respectively. Overall, compound 20 has better toxicity profile as compared to compound 19 (Table 3). Table 3. In-silico absorption and toxicity profile of compounds 19 and 20 obtained from admetSAR server [50].

Analgesic Activity
The formalin-induced analgesic test is an extensively acknowledged animal nociception model. In order to evaluate both central and peripheral effects of the compound (20), formalin-induced nociception model was selected for analgesic activity. The formalin induced behavioral response comprises two typical phases, stage I and II. Stage I persists up to five minutes after formalin injection and is characterized by acute pain with vigorous licking and biting of the injected site. Stage I consists the action formalin on afferent C-fiber nociceptors. While Stage II starts 10-30 min after formalin injection and persists till 60 min, characterized by periodic licking and biting of the injected site. Stage II imitates the action of central sensitization of the spinal dorsal horn neurons [51,52]. Compound 20 was selected for analgesic activity due to its higher potency (IC 50 7.6 nM). Compound 20 (suspensions prepared with 0.5% CMC) were administered per oral (p.o, in doses of 5, 10, 30, and 50 mg/kg body weight, 4 h prior to the formalin injection. Gabapentin (GBP), (dissolved in 0.9% normal saline), was chosen as positive control (reference drug) and administered intraperitoneal (i.p) in 100 mg/kg dose. GBP exhibited little analgesic effects in Stage I (acute nociception), in comparison to the control (0.5% CMC). Though, in Stage II, it displayed significant reduction of paw licking and biting, endorsing GBP central effects. However, compound 20, reduced the pain response significantly both in acute (Stage I) and late (Stage II) phases, in a dose-dependent manner. They significantly demonstrated the reduction in pain response, having better potency than the positive control GBP at 30 mg/kg. The duration (in seconds) of paw licking and paw biting throughout Stage I and II is provided in Figure 5. C-fiber nociceptors. While Stage II starts 10-30 min after formalin injection and persists till 60 min, characterized by periodic licking and biting of the injected site. Stage II imitates the action of central sensitization of the spinal dorsal horn neurons [51,52]. Compound 20 was selected for analgesic activity due to its higher potency (IC50 7.6 nM). Compound 20 (suspensions prepared with 0.5% CMC) were administered per oral (p.o, in doses of 5, 10, 30, and 50 mg/kg body weight, 4 h prior to the formalin injection. Gabapentin (GBP), (dissolved in 0.9% normal saline), was chosen as positive control (reference drug) and administered intraperitoneal (i.p) in 100 mg/kg dose. GBP exhibited little analgesic effects in Stage I (acute nociception), in comparison to the control (0.5% CMC). Though, in Stage II, it displayed significant reduction of paw licking and biting, endorsing GBP central effects. However, compound 20, reduced the pain response significantly both in acute (Stage I) and late (Stage II) phases, in a dose-dependent manner. They significantly demonstrated the reduction in pain response, having better potency than the positive control GBP at 30 mg/kg. The duration (in seconds) of paw licking and paw biting throughout Stage I and II is provided in Figure 5.

Anticancer Activity
Compounds 19 and 20 were supplied to National Cancer Institute (USA), for sulforhodamine B (SRB) assay and anticancer screening [53,54]. Single-dose (10 µM) assay results for compounds 19 and 20 were provided as a mean of percent growth (% G) and growth inhibition (% GI) against 60 cell lines of nine types of cancers and are tabulated in Table 4.

Discussion
Ten benzoxazole clubbed 2-pyrrolidinone derivatives (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) as the inhibitors of monoacylglycerol lipase were designed on the criteria fulfilling the structural requirements and on the basis of previously reported inhibitors [36,[42][43][44][45]. The designed, synthesized, and characterized compounds (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) were screened against monoacylglycerol lipase (MAGL) in order to find potential inhibitors. The substituted phenyl derivatives (13)(14)(15)(16)(17)(18)(19)(20) were established to reduce the MAGL activity at 100 µM concentration below 50%. Compound 19 (4-NO 2 derivative) and compound 20 (4-SO 2 NH 2 derivative) were the most potent, with IC 50 of 8.4 and 7.6 nM, correspondingly. The benzoxazole derivatives having 4-NO 2 phenyl (19) and 4-SO 2 NH 2 phenyl (20), with an FAAH IC 50 value greater than 50 µM, were considered selective MAGL inhibitors. In molecular docking studies, compounds 19 and 20 showed comparable docking scores of −9.87 and −9.83, respectively. The binding of compounds 19 and 20 in the active site of MAGL revealed that the carbonyl group of pyrrolidinone is located exactly in the oxyanion hole and stabilized by three hydrogen bonds (~2 Å) with alanine 51, serine 122, and methionine 123. Serine 122 is one of the critical amino acid residues of the catalytic triad of MAGL. The benzoxazole moiety is found to be positioned in the amphiphilic pouch, having π-π stacking contacts with the amino acid tyrosine 194. The 4-NO 2 phenyl (19) and 4-SO 2 NH 2 phenyl (20) part of the ligand was engaged in hydrophobic (Van der Waals) attractions with the amino acids leucine 148, 213, and 241. The binding patterns of compounds 19 and 20 in the catalytic site of MAGL were found to be similar as those of the reported inhibitors bound crystal structures [32,36,39,41]. Remarkably, the physiochemical and pharmacokinetic properties of compounds 19 and 20 computed by QikProp were found to be in the qualifying range as per the proposed guideline for good orally bioactive CNS drugs. Moreover, compound 20 showed better toxicity profile than compound 19, as predicted by admetSAR [50]. In formalin-induced analgesic test, compound 20 reduced the pain response significantly both in acute (stage I) and late (stage II) phases in a dose-dependent manner. It significantly demonstrated the reduction in pain response, having better potency than the positive control GBP, at the dose of 30 mg/kg. Moreover, in one dose (10 µM), anticancer screening by SRB assay, compounds 19 (NSC: 778839) and 20 (NSC: 778842) were found to have good anticancer activity towards SNB-75 cell line of CNS cancer, having % growth inhibition (% GI) of 35.49 and 31.88, respectively. Therefore, the present work concluded that compound 20 is the potential lead compounds that can be further manipulated at points 1 and 4 of the 2-pyrrolidinone moiety for the discovery and development of more selective and potent inhibitors of MAGL for neuropathic pain and CNS disorders including cancers.

Human FAAH Assay
The screening of the selected compounds (15, 16, 18, 19 and 20) for their potential to inhibit hFAAH was performed according to the information leaflet provided with Cayman's assay kit (Cayman Chemical, Ann Arbor, Michigan, USA) by the reported method [47], as discussed in detail in our previous publication [45]. The results were compared with standard FAAH inhibitor, URB597, and are provided in Table 1.

Molecular Docking Study
Glide executed on Maestro 9.4 (Schrödinger Inc., New York, NY, USA) was utilized for Glide XP docking of active compounds (19 and 20). The .pdb file of hMAGL X-ray crystal structure was downloaded from protein data bank having ID 5ZUN (crystal structure resolution 1.35 Å) for molecular docking study [36]. The protein structure was refined, optimized, and energy-minimized with the help of preparation wizard in Maestro. A docking grid of 20 × 20 × 20 Å, was created around the catalytic site by defining the cocrystallized ligand. Ligand (compounds 19 and 20) structures were prepared with the help of LigPrep 2.6 with Epik 2.4 at pH 7.0 ± 2.0. The methodology was validated by docking the cocrystallized ligand with Glide XP docking protocol [55].

Physicochemical and Pharmacokinetic Characteristics
Guidelines, concerning the validation and optimization of orally active CNS compounds, were developed by Ghose et. al. by analyzing 35 characteristic features of orally bioavailable 317 CNS and 626 non-CNS drugs [49]. For computations of these properties of the selected compounds (19 and 20), QikProp 3.6 module of Schrodinger was utilized. The generated data was then matched with the qualifying range as per the suggested guideline for good orally bioactive CNS drugs.

In Silico Absorption and Toxicity Profile
The selected compounds (19 and 20) were evaluated for their absorption and toxicity profile by a bioinformatics tool admetSAR [50]. Oral bioavailability, intestinal absorption, and BBB penetration properties were calculated. AMES test for mutagenicity, carcinogenicity test, and the calculation of LD50 for both the compounds (19 and 20) were also evaluated.

Analgesic Activity
Formalin-induced analgesic test was executed by the procedure described by Coderre and Laughlin [51,52] as discussed in detail in our previous publication [45]. Male Wistar rats (180-200 g) were obtained with the permission of IAEC (proposal number 1048) from Jamia Hamdard, New Delhi, India. The results of test compound 20 and reference drug, Gabapentin (GBP), were statistically compared with the control group.

Anticancer Screening: Sulforhodamine B Assay
Compounds 19 and 20 were supplied to National Cancer Institute (Bethesda, Maryland, USA), for in vitro sulforhodamine B (SRB) assay, anticancer screening on 60 cell lines of cancers of leukemia, melanoma, and tumors of the kidney, brain, breast, lung, colon, ovary, and prostate, as per their standard protocol [53,54]. One dose anticancer results (NCI, USA) of compounds 19 and 20 are provided in the Supplementary Materials.

Statistical Analysis
The statistical study of the data was accomplished by GraphPad Prism (version 8.0.2; GraphPad Software, San Diego, CA, USA). The dose response of the test compounds was compared with that of control, in formalin-induced analgesic test, by analysis of variance (ANOVA) followed by Dunnett's test. Outcomes are communicated as mean ± SEM.
The physiochemical and pharmacokinetic properties of compounds 19 and 20 were found to be almost in the qualifying range as per the proposed guideline for good orally bioactive CNS drugs. Compound 20 significantly demonstrated the reduction in pain response, having better potency than the positive control GBP, at the dose of 30 mg/kg. The present work concluded that compound 20 is the potential lead compounds that can be further studied and optimized at points 1 and 4 of the 2-pyrrolidinone moiety for the discovery and development of more selective and potent inhibitors of MAGL for neuropathic pain.

Data Availability Statement:
The data presented in this study are available on request from the corresponding authors.