Novel c-Jun N-Terminal Kinase (JNK) Inhibitors with an 11H-Indeno[1,2-b]quinoxalin-11-one Scaffold

c-Jun N-terminal kinase (JNK) plays a central role in stress signaling pathways implicated in important pathological processes, including rheumatoid arthritis and ischemia-reperfusion injury. Therefore, inhibition of JNK is of interest for molecular targeted therapy to treat various diseases. We synthesized 13 derivatives of our reported JNK inhibitor 11H-indeno[1,2-b]quinoxalin-11-one oxime and evaluated their binding to the three JNK isoforms and their biological effects. Eight compounds exhibited submicromolar binding affinity for at least one JNK isoform. Most of these compounds also inhibited lipopolysaccharide (LPS)-induced nuclear factor-κB/activating protein 1 (NF-κB/AP-1) activation and interleukin-6 (IL-6) production in human monocytic THP1-Blue cells and human MonoMac-6 cells, respectively. Selected compounds (4f and 4m) also inhibited LPS-induced c-Jun phosphorylation in MonoMac-6 cells, directly confirming JNK inhibition. We conclude that indenoquinoxaline-based oximes can serve as specific small-molecule modulators for mechanistic studies of JNKs, as well as potential leads for the development of anti-inflammatory drugs.

Pharmacological and genetic evidence suggests that inhibition of JNK signaling may represent a promising therapeutic strategy [5,24]. To date, numerous medicinal chemistry efforts and high throughput screens have been initiated in efforts to identify selective and nontoxic JNK inhibitors with high therapeutic potential ( Figure 1). For example, the JNK inhibitor AS602801 (Bentamapimod) has been shown to be effective and safe in clinical trials for the treatment of inflammatory endometriosis [25]. However, this compound inhibits all three JNK isoforms with similar efficacy [26]. Since JNK isoforms have been demonstrated to function independently and have distinct tissue expression patterns [17], the development of JNK isoform-specific inhibitors may have greater clinical benefit. Notably, JNK3 is expressed in the brain and the heart [26], so the development of selective inhibitors for this isoform may be promising in treating neurodegenerative diseases and other inflammatory disorders [27][28][29][30][31]. Furthermore, the therapeutic efficacy of lead compounds should be evaluated in appropriate ischemia/reperfusion models. rheumatoid arthritis, osteoarthritis, insulin resistance, inflammatory bowel diseases, cancer, stroke, renal ischemia, essential hypertension, Alzheimer's disease, and Parkinson's disease [10][11][12][13][14][15][16][17][18][19][20][21][22][23].
Pharmacological and genetic evidence suggests that inhibition of JNK signaling may represent a promising therapeutic strategy [5,24]. To date, numerous medicinal chemistry efforts and high throughput screens have been initiated in efforts to identify selective and nontoxic JNK inhibitors with high therapeutic potential ( Figure 1). For example, the JNK inhibitor AS602801 (Bentamapimod) has been shown to be effective and safe in clinical trials for the treatment of inflammatory endometriosis [25]. However, this compound inhibits all three JNK isoforms with similar efficacy [26]. Since JNK isoforms have been demonstrated to function independently and have distinct tissue expression patterns [17], the development of JNK isoform-specific inhibitors may have greater clinical benefit. Notably, JNK3 is expressed in the brain and the heart [26], so the development of selective inhibitors for this isoform may be promising in treating neurodegenerative diseases and other inflammatory disorders [27][28][29][30][31]. Furthermore, the therapeutic efficacy of lead compounds should be evaluated in appropriate ischemia/reperfusion models. Previously, we identified a new class of JNK inhibitors based on the 11H-indeno [1,2b]quinoxalin-11-one and indolo [2,1-b]quinazoline-6,12-dione scaffolds [35,36]. Specifically, compound IQ-1 (11H-indeno [1,2-b]quinoxalin-11-one oxime) and its analogs inhibited JNK activity and, consequently, proinflammatory cytokine production by human myeloid cells [35][36][37][38][39]. These JNK inhibitors required the presence of an oxime group (Table  1), and compound C containing a Cl atom at position R 3 had greater selectivity for JNK1/3 compared to JNK2 [36]. Based on these results, we suggested that modification of IQ-1 could increase potency and/or a selectivity of the resulting analogs toward the three JNK isoforms.
In the present study, novel 11H-indeno[1,2-b]quinoxalin-11-one analogs with various substituents on the tetracyclic moiety (mainly at the R 3 position or a combination of R 3 with substituents at R 1 /R 2 ) were synthesized and evaluated for their effects on JNK1-3. Compound design was based on introducing different types of substituents (i.e., strong and moderate electron donors/acceptors) into the quinoxaline moiety of IQ-1 using correspondingly substituted 1,2-diaminobenzenes as the starting reagents. We also evaluated the anti-inflammatory potential of these compounds using in vitro cell-based assays. Previously, we identified a new class of JNK inhibitors based on the 11H-indeno[1,2b]quinoxalin-11-one and indolo[2,1-b]quinazoline-6,12-dione scaffolds [35,36]. Specifically, compound IQ-1 (11H-indeno [1,2-b]quinoxalin-11-one oxime) and its analogs inhibited JNK activity and, consequently, proinflammatory cytokine production by human myeloid cells [35][36][37][38][39]. These JNK inhibitors required the presence of an oxime group (Table 1), and compound C containing a Cl atom at position R 3 had greater selectivity for JNK1/3 compared to JNK2 [36]. Based on these results, we suggested that modification of IQ-1 could increase potency and/or a selectivity of the resulting analogs toward the three JNK isoforms.
In the present study, novel 11H-indeno[1,2-b]quinoxalin-11-one analogs with various substituents on the tetracyclic moiety (mainly at the R 3 position or a combination of R 3 with substituents at R 1 /R 2 ) were synthesized and evaluated for their effects on JNK1-3. Compound design was based on introducing different types of substituents (i.e., strong and moderate electron donors/acceptors) into the quinoxaline moiety of IQ-1 using correspondingly substituted 1,2-diaminobenzenes as the starting reagents. We also evaluated the anti-inflammatory potential of these compounds using in vitro cell-based assays. Table 1. Chemical structure and activity of previously identified JNK1-3 inhibitors with an 11Hindeno [1,2-b]quinoxalin-11-one scaffold containing different substituents in the tetracyclic moiety [35,36].
Oximes 4a-4l were synthesized from the corresponding ketones by condensation with hydroxylamine hydrochloride in a pyridine/ethanol mixture under heating due to the high solubility of 3a-3l in pyridine. Oxime 4m was synthesized by treating 3m with a 5-fold excess of hydroxylamine hydrochloride in ethanol in the presence of NaOH in equimolar quantities.
Nucleophilic attack of the NH 2 group is directed preferably to the C-2 atom of the ninhydrin molecule [45,46]. Thus, regioselective formation of ketones 3a-k and 3n (Scheme 1) is in general agreement with the charge distribution calculated for the precursor substituted 1,2-diaminobenzene molecules 2a-k, 2n using the DFT method with a B3LYP/aug-cc-pVDZ level of theory. The nitrogen atom Mulliken charges obtained for these diamines are shown in Table 2. Oximes 4a-4l were synthesized from the corresponding ketones by condensation with hydroxylamine hydrochloride in a pyridine/ethanol mixture under heating due to the high solubility of 3a-3l in pyridine. Oxime 4m was synthesized by treating 3m with a 5-fold excess of hydroxylamine hydrochloride in ethanol in the presence of NaOH in equimolar quantities.
Nucleophilic attack of the NH2 group is directed preferably to the C-2 atom of the ninhydrin molecule [45,46]. Thus, regioselective formation of ketones 3a-k and 3n (Scheme 1) is in general agreement with the charge distribution calculated for the precursor substituted 1,2-diaminobenzene molecules 2a-k, 2n using the DFT method with a B3LYP/aug-cc-pVDZ level of theory. The nitrogen atom Mulliken charges obtained for these diamines are shown in Table 2.  According to the Mulliken charge distribution, most of the precursor diamines should react with ninhydrin and form the ketones shown in Scheme 1, except for the difluorodiamines used in the synthesis of compounds 3e and 3g. In these cases, the observed regioselectivity results from the strong ortho-effect of the fluorine atom located next to one of the NH 2 groups in the diamine (e.g., specific solvation of the fluorine atom by acetic acid used as the solvent). In diaminobenzene 2c, the values of nitrogen charges are very close to each other. Thus, steric effects of the methyl substituent next to one of the NH 2 groups leads to formation of ketone 3c as the major regioisomer.
The observed regioselectivity of ketone formation on the first stage (Scheme 1) was additionally confirmed for one of the synthesized ketones (3b) using single-crystal X-ray diffraction analysis. This compound formed crystals of high quality and size sufficient for this analysis. The crystal structure of 3b is formed by two crystallographically-independent molecules. Molecular geometry of 3b is shown in Figure 2 (one crystallographicallyindependent molecule is shown). The crystal structure was analyzed for short contacts between non-bonded atoms using the PLATON [47] and MERCURY programs [48]. The bond lengths and bond angles are very close for the two crystallographically-independent molecules and correspond to the statistical means [49]. According to the of X-ray diffraction data, molecules of 3b are perfectly planar in the crystal. According to the Mulliken charge distribution, most of the precursor diamines should react with ninhydrin and form the ketones shown in Scheme 1, except for the difluorodiamines used in the synthesis of compounds 3e and 3g. In these cases, the observed regioselectivity results from the strong ortho-effect of the fluorine atom located next to one of the NH2 groups in the diamine (e.g., specific solvation of the fluorine atom by acetic acid used as the solvent). In diaminobenzene 2c, the values of nitrogen charges are very close to each other. Thus, steric effects of the methyl substituent next to one of the NH2 groups leads to formation of ketone 3c as the major regioisomer.
The observed regioselectivity of ketone formation on the first stage (Scheme 1) was additionally confirmed for one of the synthesized ketones (3b) using single-crystal X-ray diffraction analysis. This compound formed crystals of high quality and size sufficient for this analysis. The crystal structure of 3b is formed by two crystallographically-independent molecules. Molecular geometry of 3b is shown in Figure 2 (one crystallographicallyindependent molecule is shown). The crystal structure was analyzed for short contacts between non-bonded atoms using the PLATON [47] and MERCURY programs [48]. The bond lengths and bond angles are very close for the two crystallographically-independent molecules and correspond to the statistical means [49]. According to the of X-ray diffraction data, molecules of 3b are perfectly planar in the crystal. The crystal structure of 3b is composed of infinite π-stacks. Neighboring molecules are arranged in zero-inclined slipped-parallel fashion with head-to-head orientations. The separation of molecular planes within the 3.44 and 3.47 Å stacks are typical for π-stacking interactions [47]. The stacks are additionally linked by H-bonds of C-H…N and C-H…O types. These π-stacks and H-bond interactions lead to the formation of 3D networks in the crystal packing of 3b (see Supplementary Materials Table S1 and Figure S1).

Binding Affinity of Novel Compounds toward JNK1-3
All compounds were evaluated for their ability to bind to JNK1-3 and compared with previously reported compound IQ-1 [35]. Four compounds (4a, 4b, 4e, and 4l) had high affinity for JNK, with Kd values in the submicromolar range for all three JNK isoforms, whereas 4f was more selective and had high affinity for JNK1 and JNK3 but did not bind to JNK2 (Table 3). Introduction of t-Bu at R 2 led to compound 4d, which had relatively low affinity for JNK. On the other hand, introduction of a fluoride atom as R 2 or a chlorine substituent at R 1 led to compounds 4a and 4b, respectively, which had submicromolar Kd values for all JNK isoforms. Introduction of a morpholinomethyl substituent at R 1 led to The crystal structure of 3b is composed of infinite π-stacks. Neighboring molecules are arranged in zero-inclined slipped-parallel fashion with head-to-head orientations. The separation of molecular planes within the 3.44 and 3.47 Å stacks are typical for π-stacking interactions [47]. The stacks are additionally linked by H-bonds of C-H . . . N and C-H . . . O types. These π-stacks and H-bond interactions lead to the formation of 3D networks in the crystal packing of 3b (see Supplementary Materials Table S1 and Figure S1).

Binding Affinity of Novel Compounds toward JNK1-3
All compounds were evaluated for their ability to bind to JNK1-3 and compared with previously reported compound IQ-1 [35]. Four compounds (4a, 4b, 4e, and 4l) had high affinity for JNK, with K d values in the submicromolar range for all three JNK isoforms, whereas 4f was more selective and had high affinity for JNK1 and JNK3 but did not bind to JNK2 (Table 3). Introduction of t-Bu at R 2 led to compound 4d, which had relatively low affinity for JNK. On the other hand, introduction of a fluoride atom as R 2 or a chlorine substituent at R 1 led to compounds 4a and 4b, respectively, which had submicromolar K d values for all JNK isoforms. Introduction of a morpholinomethyl substituent at R 1 led to compound 4m, which had high water solubility and relatively high binding affinity for all JNKs. The enhanced solubility of 4m is clearly due to the presence of the morpholine moiety, which can easily be protonated in aqueous solution. The other synthesized oximes had low water solubility. To increase their bioavailability for potential pharmaceutical applications, it will be necessary to apply special techniques (e.g., see [50]). Note that 4m has been reported previously to be a potential antiviral agent and DNA intercalator [51]. Simultaneous substitution at R 3 (Br or CF 3 ) and R 1 (CH 3 , CF 3 , Cl, or Br) led to inactive compounds 4c, 4h, 4i, and 4j for all JNK. Simultaneous substitution with two F atoms at R 3 and R 2 led to compound 4f with high selectivity for JNK1/3 in comparison with JNK2. Simultaneous substitution at R 1 and R 3 with two F atoms led to compound 4e, which also had high binding affinity (submicromolar K d ) for JNK1 and JNK3. However, according to the 1 H NMR data, the intermediate ketone for this compound is a mixture of two regioisomers of 3e in a ratio of 63:37. There was no reason to assume that the corresponding oxime differed in isomeric composition, so the real affinity of the active isomer may be 1.5-3 times higher than that which we found. been reported previously to be a potential antiviral agent and DNA intercalator [51]. Simultaneous substitution at R 3 (Br or CF3) and R 1 (CH3, CF3, Cl, or Br) led to inactive compounds 4c, 4h, 4i, and 4j for all JNK. Simultaneous substitution with two F atoms at R 3 and R 2 led to compound 4f with high selectivity for JNK1/3 in comparison with JNK2. Simultaneous substitution at R 1 and R 3 with two F atoms led to compound 4e, which also had high binding affinity (submicromolar Kd) for JNK1 and JNK3. However, according to the 1 H NMR data, the intermediate ketone for this compound is a mixture of two regioisomers of 3e in a ratio of 63:37. There was no reason to assume that the corresponding oxime differed in isomeric composition, so the real affinity of the active isomer may be 1.5-3 times higher than that which we found. Note that Z,E-isomerization of oximes can occur readily [52], hence isomerization is possible on interaction with a kinase binding site, as discussed previously [36]. Our calculation by the DFT method gives an estimated ΔG°298 value of 2.26 kJ/mol for Z E isomerization of oxime 4f (i.e., the E-isomer is just slightly less thermodynamically stable than the Z-isomer), and the E-configuration can be adopted by the molecule upon interaction with the JNK1-3 binding sites. Note that Z,E-isomerization of oximes can occur readily [52], hence isomerization is possible on interaction with a kinase binding site, as discussed previously [36]. Our calculation by the DFT method gives an estimated ∆G • 298 value of 2.26 kJ/mol for Z E isomerization of oxime 4f (i.e., the E-isomer is just slightly less thermodynamically stable than the Z-isomer), and the E-configuration can be adopted by the molecule upon interaction with the JNK1-3 binding sites.

Evaluation of Compound Biological Activity
Prior to biological evaluation, we analyzed cytotoxicity of the compounds at various concentrations up to 50 µM in human monocytic THP-1Blue and MonoMac-6 cells during a 24-h incubation with the compounds. Compounds 4c, 4j, and 4l exhibited cytotoxicity in THP1-Blue cells, and compounds 4c, 4g, and 4h were cytotoxic in MonoMac-6 cells (Table 4). Thus, all six of these compounds were excluded from further biological evaluation. The remaining non-cytotoxic oxime derivatives were evaluated for their ability to inhibit LPS-induced NF-κB/AP-1 reporter activity and interleukin (IL)-6 production in THP-1Blue and MonoMac-6 cells, respectively. As shown in Table 4, most of these compounds inhibited NF-κB/AP-1 reporter expression with activity that exceeded that of IQ-1, including a few compounds that did not have affinity (4h and 4i) or only had very low affinity (4k) for JNK (see Table 3). Similarly, all compounds inhibited IL-6 production in MonoMac-6 cells (Table 4), including a few compounds that did not have affinity (4i and 4j) or only had very low affinity (4k) for JNK (see Table 3). As an example, the dosedependent inhibition of LPS-induced IL-6 production by compounds 4f and 4m is shown in Figure 3. The inhibitory activity of 4h-4k may be due to interaction with non-JNK targets in these cells.
To confirm that the active compounds were inhibiting JNK activity, we selected two compounds (4f and 4m) and evaluated their effects on c-Jun phosphorylation. MonoMac-6 monocytic cells were pretreated with the compounds, stimulated with LPS, and the level of phospho-c-Jun (Ser63) was determined. As expected, both compounds inhibited c-Jun phosphorylation in LPS-treated cells (Figure 4), verifying that the active compounds did indeed inhibit c-Jun phosphorylation at a similar concentration range as their inhibitory effects on IL-6 production in MonoMac-6 cells.  To confirm that the active compounds were inhibiting JNK activity, we selected two compounds (4f and 4m) and evaluated their effects on c-Jun phosphorylation. MonoMac-6 monocytic cells were pretreated with the compounds, stimulated with LPS, and the level of phospho-c-Jun (Ser63) was determined. As expected, both compounds inhibited c-Jun phosphorylation in LPS-treated cells (Figure 4), verifying that the active compounds did indeed inhibit c-Jun phosphorylation at a similar concentration range as their inhibitory effects on IL-6 production in MonoMac-6 cells.

ADME Predictions
The ADME properties determining either the access of a potential drug to the target or its elimination by the organism are necessary during initial stages of drug discovery [53]. We evaluated ADME characteristics of the most potent compounds (4b, 4f, and 4m) using the SwissADME online tool [54]. As shown in Table 5, ADME assessment predicts

ADME Predictions
The ADME properties determining either the access of a potential drug to the target or its elimination by the organism are necessary during initial stages of drug discovery [53]. We evaluated ADME characteristics of the most potent compounds (4b, 4f, and 4m) using the SwissADME online tool [54]. As shown in Table 5, ADME assessment predicts that all three of these compounds would permeate through the blood-brain barrier, which is important since JNK3 is primarily located in the central nervous system. We also created bioavailability radar plots that display assessment of the drug-likeness of compounds 4b, 4f, and 4m. Six physicochemical properties were considered: lipophilicity (LIPO), size, polarity (POLAR), insolubility (INSOLU), flexibility (FLEX), and unsaturation (INSATU). The physicochemical range on each axis is depicted as a pink area in which the radar plot of the molecule must fall entirely to have high predicted bioavailability [54]. We found that these heterocyclic oximes in general had satisfactory ADME properties, and the plots suggest high bioavailability ( Figure 5). The only unfavorable property was a high unsaturation score in compounds 4b and 4f, which is common for most 11H-indeno [1,2-b]quinoxalin-11-one derivatives. Notably, the anthrapyrazolone JNK inhibitor SP600125 [32] also has high unsaturation and a bioavailability radar very similar to compounds 4b and 4f. The morpholine derivative 4m contains more aliphatic CH 2 groups and thus has a 'better' bioavailability radar ( Figure 5) and lower lipophilicity (Table 5). Additionally, the morpholine moiety can easily be N-protonated with salt formation, which usually improves drug delivery properties. most 11H-indeno [1,2-b]quinoxalin-11-one derivatives. Notably, the anthrapyrazolone JNK inhibitor SP600125 [32] also has high unsaturation and a bioavailability radar very similar to compounds 4b and 4f. The morpholine derivative 4m contains more aliphatic CH2 groups and thus has a 'better' bioavailability radar ( Figure 5) and lower lipophilicity (Table 5). Additionally, the morpholine moiety can easily be N-protonated with salt formation, which usually improves drug delivery properties.

Conclusions
Synthesis and analysis of novel 11H-indeno[1,2-b]quinoxalin-11-one oxime analogs demonstrated that several of these compounds had high affinity for JNK1-3. These analogs also inhibited LPS-induced nuclear NF-κB/AP-1 activation and IL-6 production in human monocytic cells. In silico evaluation of ADME characteristics indicated that 4f can cross the blood-brain barrier and thus has potential for use in the treatment of neuroinflammation and ischemia-reperfusion injury. Finally, the identified oximes represent new chemical tools that may be useful in further development of JNK inhibitors and could find application in the treatment of inflammatory diseases and neurodegenerative pathologies.

Chemistry
Reaction progress was monitored by thin-layer chromatography (TLC) with UV detection using pre-coated silica gel F254 plates (Merck) or a Silufol UV-254. The synthesized structures were confirmed on the basis of analytical and spectral data. The melting points (m.p.) were determined using an electrothermal Mel-Temp capillary melting point apparatus. LC-HRMS analysis was performed on an Agilent G6538A 6538 UHD Accurate-Mass Q-TOF LC/MS (800-3728) instrument. NMR spectra were recorded on Bruker spectrometers (600 MHz for 1 H, 150 MHz for 13 C, and 280 MHz for 19 F NMR). 1 H, 13 C, and 19 F NMR

Conclusions
Synthesis and analysis of novel 11H-indeno[1,2-b]quinoxalin-11-one oxime analogs demonstrated that several of these compounds had high affinity for JNK1-3. These analogs also inhibited LPS-induced nuclear NF-κB/AP-1 activation and IL-6 production in human monocytic cells. In silico evaluation of ADME characteristics indicated that 4f can cross the blood-brain barrier and thus has potential for use in the treatment of neuroinflammation and ischemia-reperfusion injury. Finally, the identified oximes represent new chemical tools that may be useful in further development of JNK inhibitors and could find application in the treatment of inflammatory diseases and neurodegenerative pathologies.

Chemistry
Reaction progress was monitored by thin-layer chromatography (TLC) with UV detection using pre-coated silica gel F254 plates (Merck) or a Silufol UV-254. The synthesized structures were confirmed on the basis of analytical and spectral data. The melting points (m.p.) were determined using an electrothermal Mel-Temp capillary melting point apparatus. LC-HRMS analysis was performed on an Agilent G6538A 6538 UHD Accurate-Mass Q-TOF LC/MS (800-3728) instrument. NMR spectra were recorded on Bruker spectrometers (600 MHz for 1 H, 150 MHz for 13 C, and 280 MHz for 19 F NMR). 1 H, 13 C, and 19 F NMR spectra for the most potent compounds 4a, 4b, 4e, 4f, and 4m are shown in Supplementary Materials.

General Procedures for Indenoquinoxalin-11-one (3a-3o) Synthesis
A hot solution of 1 mmol of ninhydrin in 0.5 mL of acetic acid was added to a hot solution of 1 mmol of the appropriate substituted 1,2-diaminobenzene in 0.2 mL of acetic acid, stirred, and left for 1 h at 100 • C. Hot H 2 O (0.7 mL) was added to the reaction mixture, stirred, and left overnight. The precipitate was separated by centrifugation, washed three times with 10% acetic acid until neutral and dried under vacuum over NaOH. The dry crude product was dissolved in CHCl 3 , the solution was filtered through a layer of silica gel G (2 × 4 cm), the sorbent was washed with CHCl 3 while controlling the composition of the eluate by TLC, and the filtrate was evaporated under vacuum to dryness to obtain the corresponding substituted indenoquinoxalin-11-ones.   [39], 1 g (3.6 mmol) of 3n was suspended in 5 mL of dry dimethylformamide, and 0.88 g (5.4 mmol) 1,1carbonyldiimidazole was added at 40 • C. The mixture was stirred at this temperature for 30 min and cooled to 20 • C. A solution of 0.390 g (7.2 mmol) of sodium methylate in 1 mL of methanol was added, and the mixture was stirred for 15 min. Most of the solvent was removed under reduced pressure, 50 mL of a 5% solution of Na 2 CO 3 was added, and the mixture was extracted with CHCl 3 . The extract was washed (3 × 10 mL) with a 5% solution of Na 2 CO 3 , three times with H 2 O, dried over magnesium sulfate, and filtered through a column of silica gel (2 × 4 cm) while controlling the composition of the eluate by TLC.

General Procedure for Oximation of the Indenoquinoxaline Ketones
A solution of 1.5 mmol of hydroxylamine hydrochloride in a mixture of 0.3 mL of pyridine and 0.3 mL of ethanol was added to a solution of 0.3 mmol of the appropriate substituted indenoquinoxalin-11-one in 0.5 mL of pyridine and heated at 100 • C for 6 h. Next, 9 mL of 10% acetic acid solution was added, and the mixture was left overnight. The precipitate was centrifuged, triple washed with a 5% solution of acetic acid until neutral and dried under vacuum over NaOH. The crude products were recrystallized from dimethylacetamide, ethanol, or a mixture thereof. 100 mg (0.3 mmol) 3m in 3 mL of ethanol. The mixture was heated in a boiling water bath, adding ethanol as necessary to maintain homogeneity of the reaction mixture until the 3m spot on the sample chromatogram (silica gel plate, benzene: triethylamine 10:1) was absent. The reaction mixture was diluted with 10 mL of 0.2 M (NH 4 ) 2 CO 3 buffer solution (pH = 9). After cooling, the precipitate was centrifuged, washed with the same buffer solution (

X-ray Diffraction Analysis
X-ray crystallography of 3b crystals was performed on a Bruker Kappa Apex II CCD diffractometer using ϕ,ω-scans of narrow (0.5 • ) frames with Mo Kα radiation (λ = 0.71073 Å) and a graphite monochromator. The structures were solved by direct methods using the SHELX-97 program (University of Göttingen, Germany, 1997) and refined by full-matrix least-squares method against all F2 in anisotropic approximation using the SHELXL-2014/7 program [55]. Absorption corrections were applied using the empirical multiscan method with the SADABS program ( Selected compounds were submitted for dissociation constant (K d ) determination using KINOMEscan (Eurofins Pharma Discovery, San Diego, CA, USA), as described previously [56]. In brief, kinases were produced and displayed on T7 phage or expressed in HEK-293 cells. Binding reactions were performed at room temperature for 1 h, and the fraction of kinase not bound to test compound was determined by capture with an immobilized affinity ligand and quantified by quantitative polymerase chain reaction. Primary screening at fixed concentrations of compounds was performed in duplicate. For dissociation constant K d determination, a 12-point half-log dilution series (a maximum concentration of 33 µM) was used. Assays were performed in duplicate, and their mean value is displayed.

Analysis of AP-1/NF-κB Activation
Activation of AP-1/NF-κB was measured using an alkaline phosphatase reporter gene assay in THP1-Blue cells. Human monocytic THP-1 Blue cells are stably transfected with a secreted embryonic alkaline phosphatase gene that is under the control of a promoter inducible by NF-κB/AP-1. THP1-Blue cells (2 × 10 5 cells/well) were pretreated with test compound or DMSO for 30 min, followed by addition of 250 ng/mL lipopolysaccharide (LPS) for 24 h, and alkaline phosphatase activity was measured in cell supernatants using QUANTI-Blue mix (InvivoGen) as absorbance at 655 nm and compared with positive control samples (LPS). For selected compounds, the concentrations of inhibitor that caused 50% inhibition of the NF-κB reporter activity (IC 50 ) were calculated.

IL-6 Analysis
A human IL-6 ELISA kit (BD Biosciences, San Jose, CA, USA) was measure IL-6 production. MonoMac-6 cells were plated in 96-well plates at a density of 2 × 10 5 cells/well in culture medium supplemented with 3% (v/v) endotoxin-free FBS. The cells were pretreated with test compound or DMSO for 30 min, followed by addition of 250 ng/mL LPS for 24 h. IC 50 values for IL-6 production were calculated by plotting percentage inhibition against the logarithm of inhibitor concentration (at least five points).

Cytotoxicity Assay
Cytotoxicity was analyzed with a CellTiter-Glo Luminescent Cell Viability Assay Kit from Promega (Madison, WI, USA), according to the manufacturer's protocol. Cells were treated with compounds under investigation and incubated for 24 h. After treatment, the cells were equilibrated to room temperature for 30 min, substrate was added, and the samples were analyzed with a Fluoroscan Ascent FL (Thermo Fisher Scientific, Waltham, MA, USA). The IC 50 values were calculated by plotting percentage inhibition against the logarithm of inhibitor concentration (at least five points).

Western Blotting
MonoMac-6 monocytic cells were pretreated with different concentrations of the compounds under investigation for 30 min and treated with LPS (250 ng/mL) or vehicle for another 30 min. Cells were washed twice with Hanks' balanced salt solution, and cell lysates were prepared using lysis buffer from the JNK kinase assay kit (Cell Signaling Technology, Danvers, MA). Cell lysates (from 5 × 10 6 cells) were separated on ExpressPlus 4-20% PAGE Gels (GenScript, Piscataway, NJ, USA) using TRIS-MOPS running buffer (GenScript) and transferred to nitrocellulose membranes. The blots were probed with antibodies against c-Jun, phospho-c-Jun (Ser63), and total c-Jun (Cell Signaling Technology, Danvers, MA, USA), followed by horseradish peroxidase-conjugated secondary antibody (Cell Signaling Technology). The blots were developed using SuperSignal West Femto chemiluminescent substrate (Thermo Fisher Scientific) and visualized with a FluorChem FC2 imaging system (Alpha Innotech Corporation, San Leandro, CA, USA). Quantitation of the chemiluminescent signal was performed using AlphaView software (ver. 3.0; Alpha Innotech).

Molecular Modeling
DFT calculations were performed with the Gaussian 16 program (Gaussian, Inc., Wallingford CT) for compounds 2a-k, 2n, and Z, E-isomers of oxime 4f. The hybrid B3LYP functional [57,58] and correlation consistent aug-cc-pVDZ basis set [59,60] were used. Vibrational frequency analysis was done for all the optimized geometries in order to ensure attaining energy minima for the molecules and to calculate thermodynamic properties of 4f isomers.
Supplementary Materials: The following are available online, Figure S1: Crystal packing of compound 3b along crystallographic axis b. Figures S2-S13: NMR spectral data for compounds 4a, 4b, 4e, 4f, and 4m. Table S1: Bond lengths and angles of H-bonds for compound 3b in the crystal.