Design, Synthesis, Antinociceptive and Anti-Inflammatory Activities of Novel Piroxicam Analogues

In this paper we report the design, synthesis, antinociceptive and anti-inflammatory activities of a series of benzothiazine N-acylhydrazones 14a–h, planned by structural modification of piroxicam (1), a non steroidal anti-inflammatory drug. Among the synthesized analogues, compounds 14f (LASSBio-1637) and 14g (LASSBio-1639) were identified as novel antinociceptive and anti-inflammatory prototypes, active by oral administration, acting by a mechanism of action that seems to be different from that of piroxicam, since they were inactive as an inhibitor of cyclooxygenase (COX-1 and COX-2) at concentrations of 10 μM.


Introduction
Inflammation is an adaptive response triggered by trauma, tissue injury or infection, characterized by symptoms such as pain, redness, heat and swelling. Acute inflammation is mediated mainly by the release of prostaglandins and other pro-inflammatory mediators, like leukotrienes, bradykinin and histamine, which promote local vascular alterations and leukocyte recruitment and activation. This process is normally self limited and resolution occurs with elimination of the noxious stimuli, removal of inflammatory cells and tissue repair. However, under certain conditions the inflammatory process persists, leading to a chronic inflammatory process, which additionally involves interleukins, tumor necrosis α, interferons, among other mediators, and underlines a variety of chronic diseases, such as cancer, atherosclerosis and reumathoid arthiritis [1][2][3]. Currently drug management of inflammation includes non steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids, slow-acting disease-modifying anti-rheumatic drugs (DMARDs), immunosuppressants and biologics that specifically target inflammatory cytokines [3].

NAH
Very recently, some antioxidant and antimicrobial N-acylhydrazones bearing a 4-hydroxy-1,2benzothiazine-1,1-dioxide scaffold were reported [24][25][26] (Figure 3). Nevertheless, to the best of our knowledge, the antinociceptive and anti-inflammatory potential of this molecular pattern has not been investigated. Besides, some structural features that seems to be related to the oxicams' anti-inflammatory activity, such as heteroaromatic amide subunit [4,27], have not been fully explored in these previous works.  In this paper, we report the design, synthesis and pharmacological evaluation of a novel series of antinociceptive and anti-inflammatory N-acylhydrazones derivatives (all compounds are original, with the exception of 14a, which was first published by Zia-ur-Rehman et al. [24]). Target compounds 14a-h were designed (Figure 4) to include a 4-hydroxy-1,2-benzothiazine-1,1-dioxide scaffold, a pharmacophore present in oxicam drugs, such as pioxicam, and the NAH framework, a privileged structure which is also encountered in many anti-inflammatory compounds [21]. The nature of the aryl and heteroaryl groups attached to the imine subunit was varied based on ring isosteric replacement to give 14a-f. In addition, we have synthesized compound 14g, which bears a biphenyl group, a pharmacophore found in some recently reported mPGES-1 inhibitors [28] and also considered a privileged structure [13,14]. In order to confer antioxidant and radical scavenging properties we designed compound 14h, featuring a 3,5-di-tert-butyl-4-hydroxyphenyl group, a subunit found in many dual cyclooxygenase/5-lipoxygenase (COX/5-LOX) inhibitors [29].

Chemistry
N-acylhydrazones 14a-h were synthesized from commercial available ethyl 4-hydroxy-2H-1,2benzothiazine-3-carboxylate 1,1-dioxide [30] (15)  The novel N-acylhydrazone derivatives 14a-h were fully characterized by spectroscopic methods, such as IR, 1 H and 13 C-NMR, their purities were established by HPLC and determined by elemental analysis (CHN). Their molecular formulae were confirmed by mass spectroscopy (ESI-MS). All compounds were obtained as a single tautomer, which was found to be the enolic form, as indicated by signals at δ 8.  [24][25][26]. Besides, the presence of only one peak in HPLC chromatogram and a single imine-hydrogen signal in the 1 H-NMR spectra suggest that condensation reaction was diastereoselective. Carefully comparison of amide and imine hydrogens chemical shift of compounds 14a-h with previous published NMR data for N-acylhydrazones whose stereochemistry had been established by crystallography [17,32,33] suggested the presence of the E-diastereoisomer, which was confirmed by performing a NOE experiment on compound 14b. Thus, on irradiating at the frequency of the amide hydrogen (CONH) in 14b, a positive NOE was observed for the signal at δ 8.71 ppm, assigned to imine hydrogen (C=N-H), and a positive NOE was also observed for the signal at δ 2.89 ppm (CH 3 ) on irradiating imine hydrogen (N=C-H), thus indicating the E-configuration for C=N double bond. The predominance of the E diastereoisomer of N-acylhydrazone derivatives is consistent with previous works [17,[31][32][33][34][35][36][37] and has been ascribed to the greater thermodynamic stability of the E in respect to the Z diastereoisomer [36,38].
In addition, an intense NOE observed for amide hydrogen on irradiating the methyl hydrogens suggests that a close flat conformation, featuring an S-cis conformation for C8-C9 bond and amide hydrogen antiperiplanar to carbonyl oxygen (14b-I), predominates over an open or folded conformation, featuring an S-trans C8-C9 bond (14b-II) or an amide hydrogen antiperiplanar to carbonyl oxygen (14b-III) or both (14b-IV) ( Figure 5). The predominance of conformer 14b-I is likely to be due to additional stability provided by hydrogen bonds between the enolic and carboxyl groups and amide hydrogen and benzothiazine nitrogen lone electron pair ( Figure 5). These results are in agreement with previous works [24,32] and are likely to reflect also the stereochemical and conformational features of the other compounds of the series. In order to confirm the results mentioned above and to determine unambiguously the relative configuration of compounds 14a-h, the single crystal of compound 14a (LASSBio-1606) was obtained for X-ray diffraction. The ORTEP [39] view of compound 14a is shown in Figure 6. Crystallographic analysis confirmed the E configuration and revealed a flat conformation, described by the least squares plane through the atoms N2/C3/C4/O16/O15/C14/N17 with mean deviation from the plane of 0.07 Å, featuring the intra-molecular hydrogen bonds involving N17-H…N2 and O16-H…O15, with donor-acceptor distances 2.723(3) Å and 2.541(3) Å, respectively, corroborating the conclusions suggested by NOE experiment performed with 14b. Figure 6. ORTEP view of compound 14a (LASSBio-1606) with atom displacement ellipsoids drawn at 30% probability.
A dose-response curve was obtained for derivatives 14e, 14g and 14h ( Table 2). Considering that compounds 14b (LASSBio-1617) and 14f (LASSBio-1637) present the major structural similarity to the standard drug piroxicam, they were also selected for the dose-response study. Compounds 14b and 14e-h produced dose-related inhibition of acetic acid-induced abdominal constrictions in mice and their ID 50 and maximum efficacy are summarized in Table 2. None of the NAH derivatives were found to be more potent than piroxicam (ID 50 = 0.40 µmol/kg) in the acetic acid-induced writhing test, since the most potent antinociceptive in this series was 14h (LASSBio-1639, ID 50 of 2.87 µmol/kg). Nevertheless, all compounds showed maximum efficacy in the 80-90% range, like the standard drug piroxicam (95.4%). In order to better understand the antinociceptive effect showed by NAH derivatives, the formalin-induced pain test in mice [41] was carried out (Table 3). Formalin is known to produce a biphasic pain behavior. The first transient phase, or neurogenic phase, is ascribed to the direct effect of formalin on sensory C fibers, and the second prolonged phase, also called inflammatory phase, is associated to the development of an inflammatory response and the release of analgesic mediators [41][42][43]. With the exception of compound 14b (58%), none of NAH derivatives were shown to be active in the neurogenic phase of the formalin test (data not shown). On the other hand, at 100 µmol/kg, compounds 14b, 14e-h were able to inhibit the pain response in the inflammatory phase, with compounds 14f (LASSBio-1637; 60.1%) and 14g (LASSBio-1638; 54.2%), which presented similar antinociceptive effect than piroxicam (53.9%) standing out. Their lack of effect in the neurogenic phase and expressive inhibition of the inflammatory phase of the formalin test, comparable with the response of piroxicam and NSAIDs in general, suggest that these new NAH derivatives are likely to produce their antinociceptive effect by acting on the inflammatory process. In order to determine any eventual central antinociceptive activity for compounds 14a-h, the hot plate test [31] was carried out in mice. This test is used to evaluate antinociceptive activity mediated by a central mechanism and it is selective for opioid drugs, although sensitive to some sedatives, hypnotics and muscle relaxants [44][45][46][47]. None of NAH derivatives were found to be active in this test (data not shown), even 14b (LASSBio-1617), which produced a weak inhibition of pain response in the neurogenic phase of formaline test. Piroxicam was also inactive, as expected. These results reinforce that compounds 14a-h do not have any central effect and that their antinociceptive effect is mediated peripherically, thus confirming the results found in the formalin-induced pain test. To confirm the anti-inflammatory profile of NAH derivatives, as suggested by the results obtained in the formalin test, zymosan A [48,49] and carrageenan-induced peritonitis test in mice [50] were performed. As shown in Table 4, all NAH derivatives significantly inhibited cell recruitment in the zymosan-induced peritonitis test and, with exception of 14e (37.1%) and 14a (43.7%), all compounds were found to be similar or more active than piroxicam (57.4%) at the dose of 100 µmol/kg, highlighting compounds 14g (LASSBio-1638) and 14h (LASSBio-1639), which inhibited 81.3% and 82.6% of cell migration, respectively. 6.6 ± 1.3 ** 82.6% The readings represent the mean ± S.E.M. The asterisks denote the significance levels in comparison with control groups (* p < 0.05, ** p < 0.01).
In order to investigate if the anti-inflammatory effect of NAH derivatives were COX inhibitiondependent, like piroxicam, compounds 14b and 14f were selected for in vitro assay for their ability of inhibiting COX-1 and COX-2 isoforms [51]. Despite their antinociceptive and anti-inflammatory activities, in vivo, none of the compounds were able to inhibit COX at concentration of 10 μM, suggesting that they may act in a different way than NSAIDs (data not shown) or would be significantly less potent than the prototype piroxicam.

General
Reagents and solvents were obtained from commercial sources (Sigma-Aldrich, Alfa Aesar) and used as received. Reaction progress was monitored by thin-layer chromatography (TLC) on commercial silica gel plates (KieselGel 60 F245 on aluminum sheets, Merck) and visualized by UV-light (254 and 365 nm). Column chromatography purifications were carried out using silica gel Merck 230-400 Mesh. Melting points were determined on a Quimis Q240.23 apparatus and are uncorrected. 1 H-and 13 C-nuclear magnetic resonance spectra were recorded in DMSO-d 6 (16). A mixture of compound 15 (2.000 g, 7.06 mmol), and hydrazine hydrate 80% (6.8 mL, 137.4 mmol) in ethanol (40,0 mL) was stirred under reflux for 2h, when completion of reaction was indicated by TLC. The mixture was partially concentrated under vacuum, followed by addition of water and HCl 37% until precipitation. The solid was filtered and washed with water and cold ethanol to furnish compound 22 (

General Procedure for Preparation of Compounds 14a-h
A mixture of compound 16 (0.37 mmol) and the corresponding aromatic or heteroaromatic aldehydes (0.37 mmol) in absolute ethanol (10 mL) containing one drop of 37% hydrochloric acid was stirred at room temperature for ca 30 min, until reaction completion (as indicated by TLC). Then the mixture was poured in cold water and filtered. The residue was washed with water and hot hexane and dried under vacuum to produce the desired N-acylhydrazone derivatives 14a-h. When necessary, further purification was performed by silica gel column chromatography to give compounds 14a-h with purity of over 98% by HPLC.        3-carbohydrazide 1,1-dioxide (14h). The title compound was obtained by condensation of 16 with 3,5-di-tert-buthyl-4-hydroxyphenylcarboxaldehyde as a light green powder (161 mg, 90%). Mp 249-251 °C; R f = 0.61 (CH 2 Cl 2 /MeOH, 9:1). IR (KBr): 3287, 1645, 1617, 1348, 1183, 954

X-ray Crystallography
A colorless prismatic single crystal of compound 14a (LASSBio-1606) suitable for x-ray study was obtained by slow evaporation of a solution of dichloromethane-dimethyl sulfoxide (1:15) at room temperature 295(2) K. Data collection was performed using the Enraf-Nonius CAD-4 diffractometer operating with Cu-Kα radiation at room temperature. 3718 data points were collected of what 3004 are symmetry independent (R int = 0.0267). The molecule crystallizes in the C2/c space group, having Z = 8. Structure solution was obtained using Direct Methods implemented in SHELXS [52] and the model refinement was performed with full matrix least squares on F 2 using SHELXL [52], with final residuals R1 = 0.056, wR2 = 0.148 for 2633 observed data with I > 2σ(I), and R1 = 0.071, wR2 = 0.299 for all data. The crystal packing is mediated by a pair of intermolecular hydrogen short contact of type N17−H17…O12 i with donor-acceptor distance 3.156(3) Å and DHA angle 149.3°, forming dimmers about an inversion center, and further stabilization is due to weak interactions of types C10-H…O12 ii , C13-H…O15 iii and C22-H…O11 iv . Hydrogen interactions geometry is given in Table 6. The programs ORTEP-3 [39], SHELXS/SHELXL [52] were used within WinGX [53] software package.

Animals
Swiss mice weighing 20-30 g (from the BIOCEN-UFAL) were housed in group cages and maintained on a 12 h light/12 h dark cycle. Animals had free access to food and water at all times. Experiments were carried out according to a protocol approved by the Animal Welfare Committee of Federal University of Alagoas (UFAL) (Number: 026681/2009-23), and in according with the ethical guidelines for investigation of experimental pain in conscious animals

Reagents
Acetic acid (Merck), arabic gum (Sigma Aldrich), morphine sulphate (Dimorf-Cristalia-BR) and piroxicam (Merck) were obtained from commercial sources. A solution of formalin 2.5% was prepared with formaldehyde (Merck) in saline (NaCl 0.9%). Piroxicam and compounds 14a-h were used as suspension in arabic gum in all the experiments and oral administrations.

Formalin-induced Nociception
The procedure used was essentially the same as that described previously [41]. Animals received 20 mL of 2.5% formalin solution (0.92% formaldehyde in saline) in the ventral surface of the right hind paw. Animals were observed from 0 to 5 min (neurogenic phase) and from 15 to 30 min (inflammatory phase) and the time that they spent licking the injected paw was recorded and considered as indicative of nociception. Animals received piroxicam or compounds 14a-h (100 μmol/kg, oral administration) 40 min beforehand. Control animals received vehicle (arabic gum).

Hot-plate Test
Mice were treated according to the method described by Kuraishi et al. [46]. Animals (n = 6) were placed on a hot-plate set at 55 ± 1 °C. Reaction time was recorded when the animals licked their fore and hind-paws and jumped at 30, 60, 90 and 120 min after oral administration of 100 µmol/kg of piroxicam or compounds 14a-h or reference drug (morphine, 15 µmol/kg. i.p.). Baseline was considered as the mean of reaction time obtained at 30 and 60 min before administration of derivatives or piroxicam or morphine and was defined as normal reaction of animal to the temperature.

Zymosan-induced Peritonitis
Peritoneal inflammation was induced according to the method described by Leite et al. [49]. A solution of Zymosan A (Sigma-Aldrich) (2 mg/mL) was prepared in saline (NaCl 0.9%) and injected into the peritoneal cavity of mice (0.5 mL). Six hours after injection of Zymosan A, the animals were killed by cervical dislocation and the peritoneal cavity was washed with 3 mL of cold Hank's. Compounds 14a-h or piroxicam were administered at the dose of 100 µmol/kg (p.o.), 40 min before Zymosan A injection. Control group received 10 mL/kg of vehicle (arabic gum, p.o.). The number of cells was quantified by optical microscope, using the 100 lens.

Carrageenan-induced Peritonitis
Peritoneal inflammation was induced according to the method described by Ferrandiz and Alcaraz [50]. A solution of carrageenan 1% (Sigma-Aldrich) was prepared in saline (NaCl 0.9%) and injected into the peritoneal cavity of mice (250 µL/animal). Four hours after injection of carrageenan, the animals were killed by cervical dislocation and the peritoneal cavity was washed with 3 mL of cold Hank's. Compounds 14a-h and piroxicam were administered at the dose of 100 µmol/kg (p.o.), 30 min before carrageenan injection. Control group received 10 mL/kg of vehicle (arabic gum, p.o.). The number of cells was quantified by optical microscope, using 100 lens.

Evaluation of Human COX-1/COX-2 Inhibition
Evaluation of piroxicam and compounds 14b,e for COX-1 and COX-2 inhibition were performed by CEREP Laboratories (Celle L'Evescault, France) using, respectively, assay catalog reference no. 0726 and 0727, and both were carried out as described by Glaser et al. [51]. In these assays, the inhibitory effect of compounds, at concentration of 10 μM, on activity of human recombinant COX-1 or COX-2, isolated from Sf-9 cells, was quantified by measuring the formation of PGE 2 , detected by homogeneous time resolved fluorescence (HTRF), from arachidonic acid. The results are expressed as a percent inhibition of the control enzyme activity.

Test Evaluation of the Inhibition of LOX
Compounds 14b, 14e-h and piroxicam were evaluated for their ability to inhibit LOX, using the kit for determining the inhibition of LOX (Lipoxygenase Inhibitor Screening Assay Kit, Cayman Chemical Company, Ann Arbor, MI, USA) according to manufacturer's instructions. The test compound solutions were prepared using DMSO as solvent and reaction buffer (supplied in kit) to a concentration of 10 µM.

Statistical Analysis
Data obtained from animal experiments are represented by mean ± standard error of the mean (Mean ± S.E.M.). Statistical differences between the treated and the control groups were evaluated by test t of Student or ANOVA in the tutorial Prisma®. Values were considered significant if * p < 0.05 and ** p < 0.01. The ID 50 values (i.e., the dose of derivatives N-acylhydrazone which reduced the pain response by 50% in relation to control group values) were determined by linear regression from individual experiments using the GraphPad software (GraphPad Software, San Diego, CA, USA) and are reported as geometric means accompanied by their respective 95% confidence limits. Maximal inhibition values were calculated at the more effective dose used.

Conclusions
In general, compounds 14a-h presented antinociceptive and anti-inflammatory activities in vivo, by oral administration, at the screening dose of 100 µmol/kg. Pharmacological evaluation suggest that the NAH derivatives reported herein present a better pharmacological profile than standard drug piroxicam, especially for their markedly activity in acute inflammation models. Moreover, we were able to identify LASSBio-1637 (14f) and LASSBio-1639 (14g) as new antinociceptive and anti-inflammatory prototypes, which are able to inhibit cell recruitment in carrageenan and zymosan-induced peritonitis in more than 70% and 80% (dose = 100 μmol/kg, p.o), respectively, through a mechanism of action that seems to be distinct of piroxicam and remains to be elucidated.

Supplementary Material
The structural results for 14a were deposited with the Cambridge Crystallographic Data Centre as Supporting Information, CCDC number: 908020. Copies of the data can be obtained free of charge upon application to CCDC, 12 Union Road, Cambridge CB2 1EZ, United Kingdom (Fax: (44) 1223 336-033; E-mail: deposit@ccdc.cam.ac.uk).