Photocaging of Pyridinylimidazole-Based Covalent JNK3 Inhibitors Affords Spatiotemporal Control of the Binding Affinity in Live Cells

The concept of photocaging represents a promising approach to acquire spatiotemporal control over molecular bioactivity. To apply this strategy to pyridinylimidazole-based covalent JNK3 inhibitors, we used acrylamido-N-(4-((4-(4-(4-fluorophenyl)-1-methyl-2-(methylthio)-1H-imidazol-5-yl)pyridin-2-yl)amino)phenyl)benzamide (1) as a lead compound to design novel covalent inhibitors of JNK3 by modifying the amide bond moiety in the linker. The newly synthesized inhibitors demonstrated IC50 values in the low double-digit nanomolar range in a radiometric kinase assay. They were further characterized in a NanoBRETTM intracellular JNK3 assay, where covalent engagement of the target enzyme was confirmed by compound washout experiments and a loss in binding affinity for a newly generated JNK3(C154A)-NLuc mutant. The most potent compound of the series, N-(3-acrylamidophenyl)-4-((4-(4-(4-fluorophenyl)-1-methyl-2-(methylthio)-1H-imidazol-5-yl)pyridin-2-yl)amino)benzamide (13), was equipped with a photolabile protecting group leading to a nearly 10-fold decrease in intracellular JNK3 binding affinity, which was fully recovered by UV irradiation at a wavelength of 365 nm within 8 min. Our results highlight that photocaged covalent inhibitors can serve as a pharmacological tool to control JNK3 activity in live cells with light.


Introduction
The family of c-Jun N-terminal kinases (JNKs) consists of three kinases, namely JNK1, JNK2, and JNK3. The JNKs are serine threonine kinases which differ in their distribution pattern. In particular, JNK1 and JNK2 are ubiquitously expressed, while JNK3 is expressed predominantly in neuronal tissue and, to a lesser degree, in cardiac myocytes and testis [1][2][3]. Due to its unique expression pattern, JNK3 is a promising target for the treatment of various neurodegenerative disorders, such as Alzheimer's disease, Parkinson's disease, and Huntington's disease [4][5][6].
The amino acid sequence within the JNK family is highly conserved, and all three isoforms harbor a cysteine in position 154 (JNK3 numbering), which is unique within the kinome [7]. Two series of selective covalent pan-JNK inhibitors addressing this cysteine side chain with electrophilic Michael acceptor systems have been reported so far. The first series of covalent JNK3 inhibitors was disclosed by Gray and coworkers by modifying an Imatinib-derived lead structure in 2012. JNK-IN-8 represents the most promising covalent pan-JNK inhibitor of this series ( Figure 1A) [8]. In 2017, some of us reported a second series of covalent JNK3 inhibitors based on a pyridinylimidazole scaffold, including compound 1 as the most potent and most selective inhibitor within this series ( Figure 1A) [9]. Recently, the binding mode of 1 within the ATP binding site was resolved [10]. As shown in Figure 1B, compound 1 interacts with JNK3 via a bidentate hydrogen bond between the 2-aminopyridine moiety and Met149, and the imidazole ring forms an essential interaction in Figure 1B, compound 1 interacts with JNK3 via a bidentate hydrogen bond between th 2-aminopyridine moiety and Met149, and the imidazole ring forms an essential intera tion with Lys93. The 4-fluorophenyl moiety is located in the hydrophobic region I. A par meta-substitution pattern regarding the two benzene rings of the linker affords the highe target affinity and allows the acrylamide warhead to move into an optimal position fo covalent bond formation with JNK3. Moreover, this substitution pattern also allows th linker to interact with Asn152 and Gln155 with its two amide moieties. Photochemical reactions allow for precise spatiotemporal control of chemical and b ological processes as well as bioactive compounds [11,12]. Recently, Wu et al. highlighte photochemical approaches as a promising strategy to minimize off-target binding, esp cially regarding anti-cancer drugs [11].
The two most prominent approaches for applying photochemical reactivity to bioa tive compounds are photoswitching and photocaging [11,13]. Photoswitching refers to method where a molecule can be present in a trans or cis configuration that can be inte converted by light pulses of a specific wavelength. In the case of covalent kinase inhibito targeting non-catalytic cysteines, one of the two configurations orients the electrophil warhead into a suitable position for covalent bond formation ( Figure 2A). This step is r versible, as the molecule returns to the inactive configuration when the irradiation stopped [11,13].
In contrast, photocaging is a concept where a bioactive compound is rendered unab to bind to its target by the attachment of a photocleavable protecting group (PPG) at suitable position in the molecule [11,14,15]. If the photocleavable group is removed b irradiation, the binding affinity can be restored and, in the case of covalent inhibitors, th formation of the covalent bond to the target can take place ( Figure 2B). This process irreversible.
To date, photopharmacological approaches have been applied to several biological relevant molecules including a variety of kinase inhibitors [16][17][18][19][20]. In collaboration wit other research groups, some of us developed a diazocine-based photoswitchable JNK inhibitor 2 using compound 1 as the lead structure to acquire spatiotemporal control ov the inhibition of JNK3 (Figure 3) [10]. The introduction of a diazocine photoswitch int Photochemical reactions allow for precise spatiotemporal control of chemical and biological processes as well as bioactive compounds [11,12]. Recently, Wu et al. highlighted photochemical approaches as a promising strategy to minimize off-target binding, especially regarding anti-cancer drugs [11].
The two most prominent approaches for applying photochemical reactivity to bioactive compounds are photoswitching and photocaging [11,13]. Photoswitching refers to a method where a molecule can be present in a trans or cis configuration that can be interconverted by light pulses of a specific wavelength. In the case of covalent kinase inhibitors targeting non-catalytic cysteines, one of the two configurations orients the electrophilic warhead into a suitable position for covalent bond formation ( Figure 2A). This step is reversible, as the molecule returns to the inactive configuration when the irradiation is stopped [11,13].
In contrast, photocaging is a concept where a bioactive compound is rendered unable to bind to its target by the attachment of a photocleavable protecting group (PPG) at a suitable position in the molecule [11,14,15]. If the photocleavable group is removed by irradiation, the binding affinity can be restored and, in the case of covalent inhibitors, the formation of the covalent bond to the target can take place ( Figure 2B). This process is irreversible.
To date, photopharmacological approaches have been applied to several biologically relevant molecules including a variety of kinase inhibitors [16][17][18][19][20]. In collaboration with other research groups, some of us developed a diazocine-based photoswitchable JNK3 inhibitor 2 using compound 1 as the lead structure to acquire spatiotemporal control over the inhibition of JNK3 (Figure 3) [10]. The introduction of a diazocine photoswitch into the linker moiety of 1 allowed differentiation between an inactive (cis) and an active (trans) form, since only the trans configuration facilitates covalent engagement of Cys154. Light pulses at 390 nm proved sufficient to control the trans-cis switch, resulting in an increase in biochemical activity by a factor of 30. However, in the cellular NanoBRET TM assay, this effect was less pronounced, as light pulses at 400 nm increased the inhibitory potency by a factor of only two [10].
Pharmaceuticals 2023, 16, x FOR PEER REVIEW 3 the linker moiety of 1 allowed differentiation between an inactive (cis) and an active ( form, since only the trans configuration facilitates covalent engagement of Cys154. pulses at 390 nm proved sufficient to control the trans-cis switch, resulting in an inc in biochemical activity by a factor of 30. However, in the cellular NanoBRET TM assay effect was less pronounced, as light pulses at 400 nm increased the inhibitory poten a factor of only two [10].  The aim of the present study was to design novel pyridinylimidazole-based cov JNK3 inhibitors using compound 1 as the lead structure. First, the influence of the a bond between the two benzene rings was investigated ( Figure 4). Second, a photocl ble protecting group was introduced into the most promising modified covalent inhibitor as a proof-of-concept study of photocaging this class of covalent JNK3 inhib The hinge binding motif (pyridin-2-amino function) of 1 represents an essential stru feature for this inhibitor and its congeners to address the JNK3 ATP binding site (F 4). Consequently, we hypothesized this motif to be a suitable photocaging position, introduction of a bulky photocleavable protecting group is likely to impair the bind the inhibitor to the hinge region of the targeted kinase, resulting in a weak or in compound. pulses at 390 nm proved sufficient to control the trans-cis switch, resulting in an in biochemical activity by a factor of 30. However, in the cellular NanoBRET TM as effect was less pronounced, as light pulses at 400 nm increased the inhibitory pot a factor of only two [10].  The aim of the present study was to design novel pyridinylimidazole-based JNK3 inhibitors using compound 1 as the lead structure. First, the influence of th bond between the two benzene rings was investigated ( Figure 4). Second, a phot ble protecting group was introduced into the most promising modified covale inhibitor as a proof-of-concept study of photocaging this class of covalent JNK3 in The hinge binding motif (pyridin-2-amino function) of 1 represents an essential st feature for this inhibitor and its congeners to address the JNK3 ATP binding site 4). Consequently, we hypothesized this motif to be a suitable photocaging positio introduction of a bulky photocleavable protecting group is likely to impair the bin the inhibitor to the hinge region of the targeted kinase, resulting in a weak or compound. The aim of the present study was to design novel pyridinylimidazole-based covalent JNK3 inhibitors using compound 1 as the lead structure. First, the influence of the amide bond between the two benzene rings was investigated ( Figure 4). Second, a photocleavable protecting group was introduced into the most promising modified covalent JNK3 inhibitor as a proof-of-concept study of photocaging this class of covalent JNK3 inhibitors. The hinge binding motif (pyridin-2-amino function) of 1 represents an essential structural feature for this inhibitor and its congeners to address the JNK3 ATP binding site ( Figure 4). Consequently, we hypothesized this motif to be a suitable photocaging position, as the introduction of a bulky photocleavable protecting group is likely to impair the binding of the inhibitor to the hinge region of the targeted kinase, resulting in a weak or inactive compound.

Biological Evaluation
To evaluate the influence of the amide bond between the two benzene ring compound activity, we considered the inversion of this functionality as well as m tion of the nitrogen atom leading to compounds 8 and 13, respectively (Table 1) a two-fold decreased biological activity compared to parent compound 1 (IC50 these derivatives remained potent JNK3 inhibitors with IC50 values in the low dou nanomolar range. Interestingly, replacement of the acrylamide warhead in 8 and an unreactive propionamide moiety (inhibitors 10 and 14, respectively) was tole the target enzyme. These data suggested that the strong biochemical potency of t pound series is predominantly achieved by reversible interactions of the scaffold with a minor contribution from the covalent bond formation.
Next, the compound series was investigated in a NanoBRET TM target eng (TE) intracellular JNK3 assay. The data obtained from these experiments correla with the trends observed in the biochemical assay, as pyridinylimidazoles 8 an played a 4-fold and 5.5-fold decreased binding affinity, respectively, compared structure 1 (IC50 = 243 nM). Noteworthily, the propionamides 10 and 14 demo substantially weaker IC50 values in the NanoBRET TM assay than anticipated based biochemical potencies.
To study the putative covalent binding of inhibitors 8 and 13, we next app compound washout method in the NanoBRET TM assay [21]. Briefly, we establishe tocol where the compound dissociation of 8 and 13 was determined indirectly ability to inhibit binding of the ATP-competitive tracer. The covalent inhibitor 1 w as a positive control, and the propionamide 10 was used as a negative contro experiment, the saturated compound 10 was quickly displaced by the tracer, wh hibitors 8, 13, and lead compound 1 remained persistently bound to the target ( Figure 5).

Biological Evaluation
To evaluate the influence of the amide bond between the two benzene rings on the compound activity, we considered the inversion of this functionality as well as methylation of the nitrogen atom leading to compounds 8 and 13, respectively (Table 1). Despite a two-fold decreased biological activity compared to parent compound 1 (IC 50 = 6 nM), these derivatives remained potent JNK3 inhibitors with IC 50 values in the low double-digit nanomolar range. Interestingly, replacement of the acrylamide warhead in 8 and 13 with an unreactive propionamide moiety (inhibitors 10 and 14, respectively) was tolerated by the target enzyme. These data suggested that the strong biochemical potency of this compound series is predominantly achieved by reversible interactions of the scaffold to JNK3 with a minor contribution from the covalent bond formation.
Next, the compound series was investigated in a NanoBRET TM target engagement (TE) intracellular JNK3 assay. The data obtained from these experiments correlated well with the trends observed in the biochemical assay, as pyridinylimidazoles 8 and 13 displayed a 4-fold and 5.5-fold decreased binding affinity, respectively, compared to lead structure 1 (IC 50 = 243 nM). Noteworthily, the propionamides 10 and 14 demonstrated substantially weaker IC 50 values in the NanoBRET TM assay than anticipated based on their biochemical potencies.
To study the putative covalent binding of inhibitors 8 and 13, we next applied the compound washout method in the NanoBRET TM assay [21]. Briefly, we established a protocol where the compound dissociation of 8 and 13 was determined indirectly by their ability to inhibit binding of the ATP-competitive tracer. The covalent inhibitor 1 was used as a positive control, and the propionamide 10 was used as a negative control. In this experiment, the saturated compound 10 was quickly displaced by the tracer, whereas inhibitors 8, 13, and lead compound 1 remained persistently bound to the target enzyme ( Figure 5).
As these results indicated an irreversible binding mode of acrylamides 8 and 13 in the ATP binding site of JNK3, we further examined the covalent engagement of the targeted residue Cys154. To this end, we generated a JNK3(C154A)-NLuc mutant and determined the IC 50 values of 8 and 13 for the modified enzyme. As expected, this single point mutation led to a dramatic decrease in the binding affinity of inhibitors 8 and 13 ( Figure 6).        As these results indicated an irreversible binding mode of acrylamides 8 and the ATP binding site of JNK3, we further examined the covalent engagement of the geted residue Cys154. To this end, we generated a JNK3(C154A)-NLuc mutant and d mined the IC50 values of 8 and 13 for the modified enzyme. As expected, this single p mutation led to a dramatic decrease in the binding affinity of inhibitors 8 and 13 (Fi 6). Based on our findings, we selected 13 as the most promising covalent JNK3 inhi of our novel series for the photocaging approach. We introduced a photocleavable dimethoxy-2-nitrobenzyl protecting group on the pyridinyl amine nitrogen atom, re ing in compound 17 (Table 1), that, expectedly, led to a decrease in JNK3 binding by n  As these results indicated an irreversible binding mode of acrylamides 8 an the ATP binding site of JNK3, we further examined the covalent engagement of t geted residue Cys154. To this end, we generated a JNK3(C154A)-NLuc mutant and mined the IC50 values of 8 and 13 for the modified enzyme. As expected, this singl mutation led to a dramatic decrease in the binding affinity of inhibitors 8 and 13 ( 6). Based on our findings, we selected 13 as the most promising covalent JNK3 in of our novel series for the photocaging approach. We introduced a photocleavab dimethoxy-2-nitrobenzyl protecting group on the pyridinyl amine nitrogen atom, ing in compound 17 (Table 1), that, expectedly, led to a decrease in JNK3 binding by one order of magnitude in the NanoBRET TM assay.
Finally, this assay was repeated with compound 17, applying UV irradia achieve quantitative cleavage of the PPG. The selected conditions (light with a wave of 365 nm for 8 min) afforded full recovery of the binding affinity, as 17 and its tected parent compound 13 exhibited extremely similar IC50 values in this expe To assess the general reactivity of acrylamides 8 and 13 towards nucleophiles, we conducted a stability study in the presence of the physiological nucleophile glutathione (GSH) following a modified literature methodology (for details see Supplementary Materials) [22]. When exposed to a 200-fold excess of GSH at a physiological pH, acrylamides 8 and 13 demonstrated a similar degradation during 72 h compared to the chemical probe JNK-IN-8.
These results indicate a favorable stability profile for the pyridinylimidazole-based covalent JNK3 inhibitors against nucleophiles ( Figure S1, Supplementary Materials).
Based on our findings, we selected 13 as the most promising covalent JNK3 inhibitor of our novel series for the photocaging approach. We introduced a photocleavable 4,5dimethoxy-2-nitrobenzyl protecting group on the pyridinyl amine nitrogen atom, resulting in compound 17 (Table 1), that, expectedly, led to a decrease in JNK3 binding by nearly one order of magnitude in the NanoBRET TM assay.
Finally, this assay was repeated with compound 17, applying UV irradiation to achieve quantitative cleavage of the PPG. The selected conditions (light with a wavelength of 365 nm for 8 min) afforded full recovery of the binding affinity, as 17 and its unprotected parent compound 13 exhibited extremely similar IC 50 values in this experiment (Figure 7). These results confirmed the successful application of the photocaging strategy to our covalent JNK3 inhibitor 13.
The synthesis of the photocaged compound 17 started from intermediate 11 (Scheme 3). First, the PPG was installed under nucleophilic substitution conditions using sodium hydride as base and 4,5-dimethoxy-2-nitrobenzyl bromide as alkylating reagent according to a modified literature protocol [26]. In the next step, the ester function of 15 was saponified, providing carboxylic acid 16. Finally, the electrophilic warhead was introduced by amide coupling of 16 with N-(3-aminophenyl)acrylamide to obtain pyridinylimidazole 17. The synthesis of the photocaged compound 17 started from intermediate 11 (Scheme 3). First, the PPG was installed under nucleophilic substitution conditions using sodium hydride as base and 4,5-dimethoxy-2-nitrobenzyl bromide as alkylating reagent according to a modified literature protocol [26]. In the next step, the ester function of 15 was saponified, providing carboxylic acid 16. Finally, the electrophilic warhead was introduced by amide coupling of 16 with N- (3-aminophenyl)acrylamide to obtain pyridinylimidazole 17.

Conclusions
Starting from the potent covalent inhibitor acrylamido-N-(4-((4-(4-(4-fluorophenyl)-1-methyl-2-(methylthio)-1H-imidazol-5-yl)pyridin-2-yl)amino)phenyl)benzamide (1), we herein designed novel pyridinylimidazole-based inhibitors of JNK3 by modifying the amide bond moiety in the linker. The N-methylation, as well as the inversion of the amide bond, were both tolerated by the target enzyme, leading to inhibitors with IC50 values in the low double-digit nanomolar range in a radiometric JNK3 assay. The compounds were further evaluated in a NanoBRET TM JNK3 assay to investigate the target engagement in a cellular setting. Here, acrylamides 8 and 13 exhibited persistent binding to JNK3 in compound washout experiments, along with a loss in binding affinity for a newly generated JNK3(C154A)-NLuc mutant, confirming covalent bond formation to the targeted residue Cys154. Moreover, these inhibitors displayed a favorable stability against nucleophiles in a glutathione reactivity assay. The most promising candidate of the series, compound 13, was selected for the photocaging approach and equipped with a 4,5-dimethoxy-2-nitrobenzyl protecting group to impair its binding to the hinge region of JNK3. The introduction of this PPG in 17 dramatically decreased its intracellular JNK3 binding affinity, which was fully restored by UV irradiation at a wavelength of 365 nm within 8 min. Our results demonstrate that the concept of photocaging can be exploited to control kinase inhibitor binding affinity by nearly one order of magnitude in live cells.

General Information
All reagents and solvents were of commercial quality and were utilized without further purification. High-performance liquid chromatography (HPLC) was used to determine the retention times of all reported compounds, as well as the purity of the test compounds, which was >95%, if not stated otherwise under 3.1.3. The chromatographic

Conclusions
Starting from the potent covalent inhibitor acrylamido-N-(4-((4-(4-(4-fluorophenyl)-1-methyl-2-(methylthio)-1H-imidazol-5-yl)pyridin-2-yl)amino)phenyl)benzamide (1), we herein designed novel pyridinylimidazole-based inhibitors of JNK3 by modifying the amide bond moiety in the linker. The N-methylation, as well as the inversion of the amide bond, were both tolerated by the target enzyme, leading to inhibitors with IC 50 values in the low double-digit nanomolar range in a radiometric JNK3 assay. The compounds were further evaluated in a NanoBRET TM JNK3 assay to investigate the target engagement in a cellular setting. Here, acrylamides 8 and 13 exhibited persistent binding to JNK3 in compound washout experiments, along with a loss in binding affinity for a newly generated JNK3(C154A)-NLuc mutant, confirming covalent bond formation to the targeted residue Cys154. Moreover, these inhibitors displayed a favorable stability against nucleophiles in a glutathione reactivity assay. The most promising candidate of the series, compound 13, was selected for the photocaging approach and equipped with a 4,5-dimethoxy-2-nitrobenzyl protecting group to impair its binding to the hinge region of JNK3. The introduction of this PPG in 17 dramatically decreased its intracellular JNK3 binding affinity, which was fully restored by UV irradiation at a wavelength of 365 nm within 8 min. Our results demonstrate that the concept of photocaging can be exploited to control kinase inhibitor binding affinity by nearly one order of magnitude in live cells.

General Information
All reagents and solvents were of commercial quality and were utilized without further purification. High-performance liquid chromatography (HPLC) was used to determine the retention times of all reported compounds, as well as the purity of the test compounds, which was >95%, if not stated otherwise under Section 3.1.3. The chromatographic separation was carried out on an Agilent 1100 Series HPLC system from Agilent Technologies (Santa Clara, CA, USA), equipped with an ultraviolet diode array detector with detecting at 254 nm and an XBridge TM C18 column (150 × 4.6 mm, 5 µm) from Waters (Milford, MA, USA). The oven temperature was set to 30 • C, the injection volume was 5 µL, and the flow was 1.5 mL/min. Compounds were eluted using the following gradient: mobile phase A-0.01 M KH 2 PO 4 (pH 2.3); mobile phase B-MeOH; 40% B to 85% B in 8 min; 85% B for 5 min; 85% B to 40% B in 2 min; 40% B for 1 min. Purifications by preparative HPLC were carried out with a system from Knauer (Berlin, Germany). The system consisted of two Knauer K-1800 pumps, a Knauer K-2001 detector, and a Gemini ® 5 µm NX-C18 110 Å (250 × 21.2 mm) LC column from Phenomenex (Torrance, CA, USA). Compounds were eluted using the gradient as follows. Mobile phase C-0.1% aqueous TFA; mobile phase B-MeOH; 40% B to 90% B in 20 min; 90% B for 5 min. Compounds which were finally purified by preparative HPLC were obtained as hydrotrifluoroacetate salts as confirmed by 19 F NMR analysis. Column chromatography was performed on an Interchim PuriFlash XS520Plus automated flash chromatography system using Geduran Si 60 40-63 µm silica from Merck (Darmstadt, Germany) or commercial 50 µm silica columns from Interchim (Montluçon, France). The NMR spectra were measured on an Avance 300 (300 MHz for 1 H, 75 MHz for 13 C) or an Avance 400 (400 MHz for 1 H, 100 MHz for 13 C, 377 MHz for 19 F) NMR spectrometer from Bruker (Billerica, BA, USA). Chemical shifts are reported in parts per million (ppm) relative to tetramethylsilane. All spectra were calibrated against the (residual proton) peak of the deuterated solvent. Mass spectrometry was carried out on a Finnigan MAT 95, a Thermo Quest Finnigan TSQ 7000, a Finnigan MATSSQ 710 A, or an Agilent 6540 UHD Accurate-Mass Q-TOF liquid chromatography coupled electrospray ionization mass spectrometer (LC-ESI-MS) from Agilent Technologies (Santa Clara, CA, USA) at the analytical department of the University of Regensburg.
(2) General Procedure B (HATU-mediated Amide Coupling) The respective carboxylic acid (1.0 equiv.) and HATU (1.1 equiv.) were stirred in dry DCM (90 mL/mmol) at room temperature under an argon atmosphere. After 30 min, a solution of the appropriate aniline (1.1 equiv.) and DIPEA (3.0 equiv.) in 1-2 mL of dry DCM was added to the mixture. The solution was stirred at room temperature for 22-27 h. The reaction was quenched by the addition of saturated NaCl solution and extracted with EtOAc (3x). The combined organic phases were dried over MgSO 4 , filtered, and the solvents were removed under vacuum. The product was obtained after purification via flash chromatography, preparative HPLC, or recrystallization.

N-(4-Bromophenyl)-N-methyl-3-nitrobenzamide (4)
N-(4-Bromophenyl)-3-nitrobenzamide (3) (1.8 g, 5.76 mmol) [9] was dissolved in dry THF (50 mL) under an argon atmosphere, and the solution was cooled in an ice bath. Sodium hydride (461.0 mg of a 60% suspension on mineral oil, 11.5 mmol) was added in portions. After stirring at room temperature for 30 min the mixture was cooled with a MeOH-ice bath to −10 • C. Methyl iodide (0.9 mL, 14.4 mmol) was added, and the solution was left to warm to room temperature and stirred for 6 h, when full conversion was confirmed by TLC. All volatiles were removed under vacuum, and saturated NaCl solution as well as EtOAc were added to the residue. Phases were separated, and the aqueous phase was extracted with EtOAc (3x). The combined organic phases were dried over MgSO 4 , filtered, and the solvents were removed under vacuum. The brown oily residue was treated with a 3:1 mixture of PE and EtOAc with mild warming resulting 3-Amino-N-(4-bromophenyl)-N-methylbenzamide (5) Compound 4 (1.7 g, 5.07 mmol) and Sn(II)Cl 2 ·2H 2 O (5.7 g, 25.35 mmol) were dissolved in EtOH (50 mL). The solution was stirred at reflux temperature for 16 h. After cooling down, the solvent was removed under vacuum. The residue was carefully mixed with saturated NaHCO 3 solution and EtOAc and then filtered. The phases of the filtrate were separated, and the aqueous phase was extracted with EtOAc (3x). The combined organic phases were dried over MgSO 4 , filtered, and the solvents were removed under vacuum. The crude product (900 mg; 58%) was used without further purification. 1  3-Acrylamido-N-(4-bromophenyl)-N-methylbenzamide (6) Compound 5 (1.0 g; 3.28 mmol) and DIPEA (1.1 g, 8.19 mmol) were stirred in dry 1,4-dioxane (20 mL) under an argon atmosphere. The stirring mixture was cooled to 0 • C with an ice bath, and acryloyl chloride (326.2 mg, 3.60 mmol) was added. The solution was allowed to warm up to room temperature and stirred for 2 h when full conversion was confirmed via HPLC. The reaction was quenched by the addition of saturated NH 4 Cl solution and extracted with EtOAc (3x). The combined organic phases were dried over MgSO 4 , filtered, and the solvents were removed under vacuum. The crude product (520 mg; 44%) was used in the next step without further purification.  Synthesis of N-(4-Bromophenyl)-N-methyl-3-propionamidobenzamide (9) Compound 6 (450.0 mg, 1.47 mmol) and DIPEA (476.5 mg, 3.69 mmol) were stirred in dry 1,4-dioxane (15 mL) under an argon atmosphere. The mixture was cooled to 0 • C with an ice bath and propionic anhydride (230.3 mg, 1.77 mmol) was added. The solution was allowed to warm up to room temperature and was stirred for 3 h when full conversion was confirmed via HPLC. The reaction was quenched by the addition of saturated NH 4 Cl solution and extracted with EtOAc (3x). The combined organic phases were dried over MgSO 4 , filtered, and the solvents were removed under vacuum. The crude product was obtained as 280 mg of a brown oil (53%) and used in the next step without further purification. 4-((4-(4-(4-Fluorophenyl)-1-methyl-2-(methylthio)-1H-imidazol-5-yl)pyridin-2-yl)amino) benzoic acid (12) Compound 11 (110.0 mg, 0.25 mmol) was dissolved in a mixture of THF, MeOH, and 1.5 M aqueous LiOH solution (6:3:1) (9 mL). The mixture was stirred at room temperature for 40 h, when incomplete conversion was observed via HPLC. Additional 1.5 M aqueous LiOH solution (1 mL) was added, and then stirring continued at room temperature overnight. The solvents were removed under vacuum, and the residue was dissolved in water and acidified with acetic acid until a precipitate formed. The mixture was stored in the fridge