Fragment-Based Approach to Targeting Inosine-5′-monophosphate Dehydrogenase (IMPDH) from Mycobacterium tuberculosis

Tuberculosis (TB) remains a major cause of mortality worldwide, and improved treatments are needed to combat emergence of drug resistance. Inosine 5′-monophosphate dehydrogenase (IMPDH), a crucial enzyme required for de novo synthesis of guanine nucleotides, is an attractive TB drug target. Herein, we describe the identification of potent IMPDH inhibitors using fragment-based screening and structure-based design techniques. Screening of a fragment library for Mycobacterium thermoresistible (Mth) IMPDH ΔCBS inhibitors identified a low affinity phenylimidazole derivative. X-ray crystallography of the Mth IMPDH ΔCBS–IMP–inhibitor complex revealed that two molecules of the fragment were bound in the NAD binding pocket of IMPDH. Linking the two molecules of the fragment afforded compounds with more than 1000-fold improvement in IMPDH affinity over the initial fragment hit.


Tuberculosis (TB) is a contagious infectious disease caused by
Mycobacterium tuberculosis (Mtb), which can be transmitted through the air as droplets. The infection predominantly affects the lungs, but it can spread to other parts of the body, especially in patients with a suppressed immune system.
The World Health Organization (WHO) has estimated that nearly one-third of the world's population is infected with Mtb, leading to 1.8 million TB deaths in 2015. 1 Although there has been a slow decline in new TB cases and TB-related deaths in recent years, the emergence and spread of multidrug resistant (MDR) and extensively drug resistant (XDR) strains of Mtb has increased the threat that this disease poses for global public health. According to the WHO, approximately 480,000 cases of MDR-TB emerged in 2015, and the cure rate of those patients was only 50%. 1 Current TB treatments require combinations of four first-line drugs, isoniazid, rifampicin, ethambutol, pyrazinamide, and streptomycin, which must be taken for six months or longer. 2 Resistant strains are not susceptible to the standard drugs, and although MDR-TB is treatable using second-line drugs, such treatments have a number severe side effects. 3 Consequently, there is an urgent need for the development of novel and more effective drugs for the treatment of drug resistant TB.
Inosine-5′-monophosphate dehydrogenase (IMPDH, E.C. 1.1.1.205) has received considerable interest in recent years as an important target enzyme for immunosuppressive, 4 anticancer, 5,6 and antiviral drugs. 7 Most recently, IMPDH has emerged as a promising antimicrobial drug target. 8−11 IMPDH catalyzes the first unique step in the de novo synthesis of guanine nucleotides, the oxidation of inosine 5′-monophosphate (IMP) to xanthosine 5′-monophosphate (XMP) with the concomitant reduction of the cofactor nicotinamide adenine dinucleotide (NAD + ) to NADH (Figure 1). 12 XMP is then subsequently converted to guanosine 5′-monophosphate (GMP) by a GMP synthetase. 13 IMPDH has been deemed essential in every pathogen analyzed to date, including Mtb, Staphylococcus aureus, and Pseudomonas aeruginosa, which are three of the most serious bacterial threats. 14 −16 However, this has been somewhat contradictory, 17 in comparing cell versus animal work. IMPDH is an ubiquitous enzyme present in several eukaryotes, bacteria, and protozoa. 18 The IMPDH reaction involves two chemical transformations. The first step of the IMPDH catalyzed reaction involves the attack of catalytic Cys on substrate IMP followed by hydride transfer to NAD + , forming the covalent enzyme intermediate E-XMP*. In the second step, E-XMP* is hydrolyzed to XMP. 11 The enzyme exists in two different conformations, an open form that accommodates both the substrate and cofactor during the first step and a closed form where the active site flap moves into the NAD + -binding site for the E-XMP* hydrolysis. 19 In recent years, there has been considerable effort aimed at identifying small molecule inhibitors of IMPDH as potential antitubercular agents. X-ray crystal structures of a truncated form of the Mtb enzyme in complex with some of these compounds have been reported. 20 −26 In antibacterial drug discovery, and especially in TB drug discovery, high-throughput screening (HTS) typically identifies a number of leads that show high potency in vitro, but most did not show any translation to an in vivo effect. It is also inevitable that the HTS libraries represent only a small fraction of possible chemical space and so limit confidence in finding a good starting point for subsequent development. Phenotypic screens can potentially lead to the identification of a molecule that modifies a disease phenotype by acting on a previously undescribed target or by acting simultaneously on more than one target. 27 However, for many of these hits the relevant target or targets has not yet been identified, thus preventing further target-based optimization of the compounds. 28,29 The previously reported IMPDH inhibitors, such as compounds 7759844 (1), MAD1, P41, VCC234718, and DDD00079282 (Figure 2), were identified by phenotypic screening or target based HTS of compound libraries. 21, 23−25 We have sought to develop IMPDH inhibitors using a fragment-based approach. Fragment-based drug discovery (FBDD) is now established in both industry and academia as an alternative approach to high-throughput screening for the generation of hits or chemical tools for drug targets. 30 We have previously reported the discovery of several series of novel and potent inhibitors using FBDD to target Mtb. Previously we have reported the fragment elaboration strategies that we have applied which have included fragment growing, merging, and linking. Although fragment linking is conceptually the most appealing strategy for fragment elaboration, in practice, Figure 1. Purine nucleotide biosynthesis. The commonly occurring guanine nucleotide biosynthetic and salvage reactions are shown, as is the adenine nucleotide biosynthetic pathway. The IMPDH reaction is boxed. NK, nucleoside kinase; HPRT, hypoxanthine phosphoribosyl transferase; XPRT, xanthine phosphoribosyl transferase; GPRT, guanine phosphoribosyl transferase; GMPS, guanosine 5′-monophosphate synthetase; GMPR, guanosine 5′-monophosphate reductase; ADSS, adenylosuccinate synthetase; ADSL, adenylosuccinate lyase.

Journal of Medicinal Chemistry
Article this strategy can be challenging where the choice of the optimal fragment linker can be crucial. 31,32 In the fragment-based approach, biophysical techniques are usually used to identify small chemical compounds (fragments) that bind with low affinity to the drug target. X-ray crystallography is then usually employed to establish the binding mode of the fragment and to facilitate the design of elaborated fragments. The availability of high-resolution X-ray crystal structures of a truncated form of the IMPDH, 21,24,25 in both the substrate-free and substrate/ligand-bound forms, makes this enzyme attractive for a fragment-based approach.
In this work, the discovery of a new class of potent nM inhibitors of IMPDH using a FBDD approach is reported. A library of 960 fragments was screened against Mth IMPDH ΔCBS using a biochemical assay. The fragment hits from this assay were examined using X-ray crystallography, and an X-ray crystal structure of one of the fragment complexes was solved to a resolution of 1.45 Å. Examination of the X-ray crystal structure suggested a strategy of fragment-linking for optimization of this fragment hit.
■ RESULTS AND DISCUSSION Fragment Screening. An in-house fragment library composed of 960 fragments was screened using a biochemical assay against Mth IMPDH ΔCBS. Mth IMPDH, which shares 85% sequence identity with Mtb IMPDH and is 100% identical in the active site, 24,25 was chosen for the fragment screening and structural studies because it gave higher protein expression yields and better diffracting crystals than the Mtb orthologue. IMPDH activity was monitored spectrophotometrically by measuring the formation of NADH at 340 nm. The biochemical assay was performed at a fragment concentration of 1 mM, and hits were retested in triplicate. Compound 1 (7759844) previously

Journal of Medicinal Chemistry
Article reported as IMPDH inhibitor was used as a positive control in assays (Table 1). 23 The screen resulted in 18 hits (1.9% hit rate), where a hit was defined as a compound that gave greater than 50% inhibition at a concentration of 1 mM. A complete list of fragment hits identified is included in Table 1. A number of common scaffolds were observed, in particular phenylimidazole (fragments 2−4), aminopyrazole (fragments 5−6), and naphthol (fragments 7− 11), and the remaining compounds contained a substituted phenyl or a heterocyclic five membered ring (fragments 12−19).
The IC 50 values of six of the most active fragments were measured and the IC 50 and ligand efficiency (LE) values of these fragments are summarized in Table 2. The fragment screen provided an array of hits with IC 50 ranging from 325 μM to 674 μM and ligand efficiencies from 0.31 to 0.42. The inhibition constant [K i ] with respect to both substrates IMP and NAD + was determined by assaying various concentrations of each inhibitor with five different concentrations of substrate and a fixed saturating concentration of the cosubstrate. The inhibition data for these fragments are summarized in Table 2. All compounds yielded an uncompetitive inhibition pattern with respect to NAD + with K i values ranging from 262 to 525 μM. Fragments 14, 16, and 19 yielded a mixed inhibition with respect to IMP with K i values ranging from 126 to 398 μM, and compounds 1, 2, 11, and 18 yielded an uncompetitive inhibition with K i values ranging from 361 to 609 μM.
Inhibition constants of compound 2 toward full-length Mtb IMPDH were also determined. Compound 2 inhibited fulllength Mtb IMPDH enzyme with a K i value of 572 ± 14 μM with IMP as the substrate and a K i value of 534 ± 18 μM with NAD as the substrate, which are similar to the K i values observed for the Mth IMPDH ΔCBS enzyme.
X-ray Structure of Compounds 1 (7759844) and 2. Compound 1 and the six fragment hits (2, 11, 14, 16, 18, and 19) were selected for structural characterization using X-ray crystallography by soaking into preformed crystals of Mth IMPDH ΔCBS as previously described. 20 After molecular replacement, clear electron density was observed in the 2F 0 − F c difference map (σ = 3.0) for IMP, and in addition density was observed for one molecule of compound 1 ( Figure 3A) and two molecules of compound 2 ( Figure 3B), which were partially occupying the NAD + binding site. None of the other fragments

Journal of Medicinal Chemistry
Article (11, 14, 16, 18, and 19) showed any electron density in the X-ray crystal structures. Although the kinetic studies of compounds 1 and 2 suggested that these two compounds are uncompetitive with respect to NAD + , the binding mode of compounds 1 and 2 closely resembles that of other previously reported uncompetitive inhibitors of Mtb IMPDH. 21,24,25 It has been proposed that the uncompetitive mode of inhibition of IMPDH inhibitors with respect to NAD + is consistent with their binding preferentially to the covalent IMPDH-XMP* intermediate after NADH has been released. 24, 25,33 The structure of Mth IMPDH ΔCBS with compound 1 showed that the inhibitor binds in the NAD pocket in a near identical manner to our recently described IMPDH inhibitors. 24, 25 Compound 1 formed strong π interactions with the hypoxanthine group of IMP, P45′, Y471′, and polar interactions with G409, E442, P45′, and G470′. The electron density revealed two molecules of compound 2 within the NAD binding pocket of IMPDH. One molecule of compound 2 stacked with IMP, forming extensive π interactions with the hypoxanthine group of IMP. This fragment was further stabilized through polar interactions, hydrogen bonds, and π interactions to surrounding residues in the active site pocket, including A269, G318, and E442. The other molecule of compound 2 sits closer to the opening of the active site, making polar interactions with N273 and E442, and π interactions with H270 and Y471′.
A comparison of the structure of Mth IMPDH ΔCBS with compound 1 with the fragment 2 structure shows that the two molecules of 2 mimic the position of the larger inhibitor 1 ( Figure 3C). Fragment Elaboration. Fragment 2 was selected as the starting point for exploration because of the ease of synthetic modification and the availability of a X-ray crystal structure to guide the optimization. For chemical elaboration of 2, fragment linking as well as fragment growing were considered. As the two molecules of fragment 2 are found to bind in adjacent regions of the target protein, the fragment-linking approach was the more attractive option. However, before fragment linking, fragment 2 was further optimized with the aim of improving the binding affinity. The structures and inhibitory activities of these compounds against Mth IMPDH ΔCBS are summarized in Tables 3 and 4. The corresponding data for fragment 2 have also been included for comparative purposes.
All compounds were evaluated at a concentration of 100 μM with Mth IMPDH ΔCBS.
Fragment Growing. The fragment-growing strategy involves using structure-based drug design to form additional interactions by growing out from the starting fragment. Fragment 2 was modified at the 2-position of the imidazole ring to explore the introduction of various aromatic rings linked by a thioacetamide (20−22) to form π interactions with the hypoxanthine group of IMP ( Figure S1). Such modifications gave compounds with improved Mth IMPDH ΔCBS inhibition. The phenyl and benzofuran derivatives (20 and 21) showed 13 and 31% inhibition, respectively. Mth IMPDH ΔCBS inhibition was shown to be sensitive to minor modifications of the phenyl substituent groups; for example, the 4-iodo substituted 22 showed 30% inhibition at 100 μM, whereas the nonsubstituted compound, 20, showed 13% inhibition at the same concentration. The effect of the removal of the 4-bromo group was investigated and compounds 23−25 were synthesized. Removal of the bromo substituent in compounds 20−22 (13−31% inhibition at 100 μM) was tolerated (23−25, 3−46% inhibition at 100 μM). The importance of the aromatic amide linked by a thioacetamide was subsequently examined. Replacing the phenyl with isopropyl (26) resulted in complete loss of activity. Substitution of the thioacetamide by a thioacetic acid also led to a complete loss in activity (compounds 27 and 28).
Fragment Linking and SAR. Examination of the X-ray crystal structure of the previously reported inhibitor 1 when overlaid with fragment 2 revealed that the distance between the 4-position of the phenyl ring of fragment 2 and the 2-position of the imidazole ring represents the closest approach of the molecules ( Figure S1). On the basis of this structural

Journal of Medicinal Chemistry
Article information, three different linkers were designed to connect the two copies of the fragment 2 at these positions (compounds 29− 31). Initially, a thioacetamide and urea linker moieties were examined. Compounds 29 and 30 showed 20% and 56% of Mth IMPDH ΔCBS inhibition, respectively, at 100 μM (Table 4). Interestingly, compound 30 showed a 12-fold improvement in Mth IMPDH ΔCBS inhibitory activity with an IC 50 of 58 μM, compared to the fragment 2.

Journal of Medicinal Chemistry
Article bromophenyl)-1H-imidazole linked with a lactate analogue, as in compound 1. Compound 31 (Table 4) showed markedly improved Mth IMPDH ΔCBS inhibition with a LE of 0.29 and IC 50 of 0.52 μM, which is 1300-fold more potent compared to the fragment 2. Although compound 31 binds in the cofactor site, the mechanism of inhibition can vary depending on its relative affinities for the E·IMP and E-XMP* complexes. 8,11 Kinetic evaluation of compound 31 showed the mode of Mth IMPDH ΔCBS inhibition was uncompetitive with respect to both IMP and the NAD + cofactor (see Figure S2, Supporting Information) with a K i value of 0.30 ± 0.02 μM with IMP as the substrate and a K i value of 0.20 ± 0.01 with NAD as the substrate.
Inhibition constants of compound 31 toward full-length Mtb IMPDH were also determined. Compound 31 inhibited fulllength Mtb IMPDH enzyme with a K i value of 0.61 ± 0.05 μM with IMP as the substrate and a K i value of 0.39 ± 0.02 μM with NAD as the substrate. The inhibition constants were consistent with the data using the Mth IMPDH ΔCBS enzyme.
Removal of the bromo substituent in compound 29 to give compound 32 was well tolerated ( Table 4). The importance of the imidazole group for the inhibitory activity against Mth IMPDH ΔCBS was confirmed by replacing of the 4-(4bromophenyl)-1H-imidazole substituent of compound 31 with 4-(4-bromophenyl)oxazole (33) which resulted in a 4-fold loss of activity (Table 4). Replacing the 4-(4-bromophenyl)-1Himidazole of 31 by a phenyl (34) resulted in complete loss of activity (Table 4). However, the 4-iodophenyl derivative 35 demonstrated slightly improved activity (IC 50 = 0.47 μM, LE = 0.34) compared to compounds 31 (IC 50 = 0.52 μM, LE = 0.29) and 1 (IC 50 = 0.77 μM, LE = 0.40). It is noteworthy that LE of compound 35 was comparable to that of the original fragment hit 2 (LE = 0.36) and other reported IMPDH inhibitors. 34 The importance of the 4-bromo substituent on the phenyl ring in compound 31 was also explored (compounds 36−39, Table  5). Removal of the bromo substituent (36) resulted in a 5-fold loss in activity, whereas replacing this group with an iodine (39) or a morpholine ring (37) resulted in less than 1.3-fold loss in activity. The electronic nature of the substituent in this position had little effect on inhibitory activity. For example, an electrondonating methoxy (38) retained activity comparable to that of the bromine derivative (31). Among them, imidazoles 31, 37− 38, and 39 were found to be potent inhibitors of Mth IMPDH ΔCBS with IC 50 values ranging from 520 to 690 nM.
The (S)-isomer of 31 was found to bind preferentially (Table  5), with the racemate 31 having approximately half the Mth IMPDH ΔCBS inhibition of the (S)-isomer 31. This observation accords with the results previously reported for other series of IMPDH inhibitors. 21,35 X-ray Structure of Compound 31. From crystals soaked with compound 31, the 2F 0 − F c difference map (σ = 3.0) revealed strong density for the inhibitor. The structure of compound 31 ( Figure 4A) showed that it bound in a nearly identical manner to compounds 1 and 2 in the NAD binding pocking ( Figure 4B), stacking with IMP, and maintaining the interactions with H270, N273, E442, and P45 and Y471 from the neighboring subunit. Compound 31 made additional interactions in the binding pocket, including polar interactions with D267 and N297.
Whole-Cell Activity against Mtb. The whole-cell activity of the most potent analogues 31, 33−39, and (S)-31 in vitro was determined against Mtb H37Rv (see Table S1, Supporting Information). No significant inhibition of bacterial growth was detected for any of the compounds (MIC 90 ≥ 50 μM) over the tested concentration range (0−100 μM). There are currently ongoing efforts to explain the lack of efficacy of the potent Mth IMPDH ΔCBS inhibitors described which could be caused by low membrane permeability, poor metabolic stability, and/or drug-efflux mechanisms.

Journal of Medicinal Chemistry
Article zol-2-yl)methyl)carbamate, 36 followed by the deprotection of the benzyloxycarbonyl group under acidic condition.
Imidazole derivatives 31, 36−38, and 39 were prepared following the synthetic procedure outlined in Scheme 3. 2-Aminoimidazoles 50−54 were synthesized according to a published microwave-assisted protocol. 37 In brief, 2-aminoimidazoles 50−54 were prepared by reaction of the commercially available α-haloketones and N-acetylguanidine, followed by deacetylation (Scheme 3). Acid 59 was synthesized starting with imidazole 55, which was prepared by reaction of 2-bromo-4′hydroxyacetophenone with formamide as reported previously. 38 The phenol 57 was synthesized by alkylation of imidazole 55, followed by deprotection of the methyl ether with BBr 3 . Substituted phenol 57 was converted to the ether 58 upon treatment with methyl 2-bromopropionate in the presence of Cs 2 CO 3 . Enantiomerically pure phenyl ethers were synthesized by using Mitsunobu reaction conditions with ethyl D-lactate (Scheme 4). After the hydrolysis of the ester group, the resulting carboxylic acid 59 was treated with thionyl chloride to give the acid chloride 60, which was reacted with 2-aminoimidazoles 50− 54 to afford imidazole derivatives 31, 36−38, and 39.
The syntheses of 2-acylaminooxazole 33 and amides 34−35 were achieved as shown in Scheme 5. 2-Acylaminooxazole derivative 33 was obtained by coupling the acid chloride derivative 60 with 2-aminooxazole derivative 61, which was prepared by reaction of 2,4′-dibromoacetophenone with urea. Similarly, amides 34−35 were prepared by coupling the corresponding anilines with the acid chloride derivative 60.

■ CONCLUSIONS
FBDD has emerged as a robust approach to identify small molecules that bind to a wide range of therapeutic targets. Fragment elaboration strategies have resulted in the development of a number of compounds that have progressed into clinical trials. Within the area of TB drug discovery, a number of HTS and phenotypic screens have been performed during the past decade. Although HTS identified a number of leads that show high potency in vitro, the translation to an in vivo effect has proven challenging.

Article
This study illustrates the successful application of a fragmentbased approach followed by fragment optimization to obtain nanomolar affinity ligands of IMPDH. A library of 960 fragments were screened against Mth IMPDH ΔCBS, and from the screen the phenylimidazole fragment hit 2 (IC 50 = 674 μM) was identified. Kinetic experiments showed that 2 was an uncompetitive inhibitor of Mth IMPDH ΔCBS with respect to NAD + and IMP. Two molecules of the fragment 2 were shown to bind at the NAD binding site of the enzyme. The X-ray crystal structure also revealed that one molecule of fragment 2 makes π interactions with IMP and the other molecule sits closer to the opening of the active site, making polar interactions with N273 and E442 and π interactions with H270 and Y471′. This provides potential for further optimization of fragment 2. To explore better the possibilities given by fragment 2, fragment-linking and fragment-growing strategies were employed, resulting in low micromolar to nanomolar affinity compounds. Among them, compounds 31, 35, and 37− 39 were the most potent IMDPH inhibitors of the series described in this work with IC 50 values between 0.47 and 0.69 μM, which represent >1000-fold improvement in Mth IMPDH ΔCBS potency over the initial fragment hit. Compound 31 was shown to bind at the NAD binding site of the enzyme, and the X-ray crystal structure also revealed that it makes π interactions with IMP, maintaining the interactions with H270, N273, E442, and P45 and Y471 from the neighboring subunit. Moreover, compound 31 made additional interactions in the binding pocket, including polar interactions with D267 and N297. A comparison of this structure with the fragment 2 structure shows that the two molecules of 2 mimic the position of the larger inhibitor 31. This is the first example of   Reactions were monitored by TLC and LCMS to determine consumption of starting materials. Flash column chromatography was performed using an Isolera Spektra One/Four purification system and the appropriately sized Biotage SNAP column containing KP-silica gel (50 μm). Solvents are reported as volume/volume eluent mixture where applicable.
High resolution mass spectra (HRMS) were recorded using a Waters LCT Premier Time of Flight (TOF) mass spectrometer or a Micromass Quadrapole-Time of Flight (Q-TOF) spectrometer.
Liquid chromatography mass spectrometry (LCMS) was carried out using an Ultra Performance Liquid Chromatographic system (UPLC) Waters Acquity H-class coupled to a Waters SQ Mass Spectrometer detector. Samples were detected using a Waters Acquity TUV detector at 2 wavelengths (254 and 280 nm). Samples were run using an Acquity UPLC HSS column and a flow rate of 0.8 mL/min. The eluent consisted of 0.1% formic acid in water (A) and acetonitrile (B); gradient, from 95% A to 5% A over a period of 4 or 7 min.
All final compounds had a purity greater than 95% by LCMS analysis. A solution of 4-phenyl-1H-imidazole-2-thiol or 4-(4-bromophenyl)-1Himidazole-2-thiol (2.80 mmol) and NaOH (5.0 mmol) in EtOH (5 mL) was reflux for 1 h. After cooling to rt, a solution of 2-chloroacetic acid (2.80 mmol) in EtOH (2 mL) was added. The reaction was stirred at reflux for an additional 3 h and then cooled to 0°C. The reaction mixture was diluted with cold water (5 mL) and acidified with 1 M HCl. The precipitated product was collected by filtration and washed with DCM (2 × 2 mL).
General Method C: Synthesis of Compounds 31, 33, and 36−39. A mixture of acid 59 (0.20 mmol) and SOCl 2 (2 mL) was heated at 80°C for 2 h. The solvent was removed under reduced pressure to give the acid chloride 60 as a white solid. The resulting solid was immediately dissolved in anhydrous DCM, and the resulting solution was added slowly dropwise at 0°C to a solution of the corresponding substituted 2aminoimidazoles (0.20 mmol) and TEA (0.80 mmol) in anhydrous DCM (5 mL). The reaction mixture was stirred at 40°C for 36 h and then diluted with DCM (20 mL) and washed with saturated aqueous NaHCO 3 . The aqueous phase layer was then extracted with DCM (2 × 20 mL), and the combined organic layers were dried over MgSO 4 and filtered, and the solvent was removed under reduced pressure to afford a yellow oil, which was purified by flash chromatography, eluting with the solvent system specified.
General Method D: Synthesis of Amides 34−35. A mixture of acid 59 (0.20 mmol) and SOCl 2 (2 mL) was heated at 80°C for 2 h. The solvent was removed under reduced pressure to give 60 as a white solid. The resulting solid was immediately dissolved in anhydrous DCM, and the resulting solution was added slowly dropwise at 0°C to a solution of the corresponding aniline (0.20 mmol) and triethylamine (0.80 mmol) in anhydrous DCM (5 mL). The reaction mixture was stirred at rt for 4 h and then diluted with DCM (20 mL) and washed with saturated aqueous NaHCO 3 . The aqueous phase layer was then extracted with DCM (2 × 20 mL), and the combined organic layers were dried over MgSO 4 and filtered, and the solvent was removed under reduced pressure to afford a yellow oil, which was purified by flash chromatography, eluting with the solvent system specified.
General Method E: Synthesis of α-Chloroacetamides 40−43. Et 3 N (4.77 mmol) followed by a solution of chloroacetyl chloride (4.77 mmol) in DCM (3 mL) were added to a stirred solution of the corresponding aniline (4.38 mmol) in DCM (5 mL) at rt. The reaction mixture was stirred at rt for 2−4 h. The reaction was then diluted with DCM (20 mL) and washed with saturated aqueous NaHCO 3 , 1 M HCl, and brine and dried over anhydrous MgSO 4 , and the solvent was removed under reduced pressure. Compound 43 was purified by flash chromatography eluting with the solvent system specified, although other analogues were used in subsequent reactions without further purification.

Journal of Medicinal Chemistry
Article General Method F: Synthesis of Substituted N-(1H-Imidazol-2yl)acetamides 45−49. A mixture of the corresponding 2-bromoacetophenone derivative (0.38 mmol) and acetylguanidine (1.13 mmol) in anhydrous acetonitrile (3 mL) was heated at 100°C using microwave irradiation for 15 min. The solvent was removed, and the residue was taken in H 2 O (3 mL) and filtered, and the solid was washed with H 2 O (2 mL × 2) and DCM (2 mL). The solid obtained was used in the next step without further purification.
General Method G: Synthesis of Substituted 2-Aminoimidazoles 50−54. To a solution of the corresponding substituted N-(1H-imidazol-2-yl)acetamides (0.31 mmol) in a 1:1 v/v mixture of MeOH and H 2 O (2.4 mL) was added concentrated H 2 SO 4 (0.6 mL), and the reaction mixture was heated at 100°C under microwave irradiation for 15 −30 min. The reaction mixture was concentrated, and the resulting residue was resuspended in H 2 O (5 mL), and a saturated aqueous Na 2 CO 3 was added until pH 8. The product was extracted into EtOAc (3 × 40 mL). The combined organic fractions were dried over MgSO 4 , and the solvent was removed under reduced pressure. The resulting solid was used in the next reaction without further purification.

Journal of Medicinal Chemistry
Article (76 mg, 0.47 mmol). The mixture was stirred at rt overnight, and then it was filtered. The filter was washed with THF (2 × 3 mL) and the resulting solid (30 mg, 0.12 mmol) was dissolved in DMF (3 mL), and 44 (42 mg, 0.12 mmol) and N,N-diisopropylethylamine (74 μL, 0.48 mmol) were added. The reaction mixture was stirred at rt for 14 h, and then it was poured into water, extracted with EtOAc, dried over Na 2 SO 4 , and concentrated. After cooling to 0°C, a 1:5 v/v mixture of MeOH and DCM (12 mL) was added and the suspended solid was collected by filtration and dried at vacuum to yield 30 (35 mg, 0.08 mmol, 26% yield) as a white solid. 1 H NMR (500 MHz, DMSO-d 6 37 4-(4-Iodophenyl)-1H-imidazol-2-amine (54). Following general method G, from N-(4-(4-iodophenyl)-1H-imidazol- 4-(1H-Imidazol-4-yl)phenol (55). 2-Bromo-1-(4-hydroxyphenyl)ethan-1-one (1.0 g, 4.67 mmol) was dissolved in formamide (5 mL), and the reaction mixture was heated at 150°C for 24 h. After cooling to rt, the resulting mixture was diluted with EtOAc (40 mL) and washed with saturated aqueous NaHCO 3 (40 mL 41 4-(1-Methyl-1H-imidazol-4-yl)phenol (57). A solution of 4-(4methoxyphenyl)-1-methyl-1H-imidazole 56 (1.21 g, 6.42 mmol) in anhydrous DCM (30 mL) at −78°C was treated with BBr 3 (16 mL, 16.05 mmol, 1 M in DCM). The reaction mixture was stirred at −78°C for 10 min and 2 h at rt. The reaction mixture was then cooled to −78°C and quenched with MeOH (4 mL). The solvents were then removed under reduced pressure, and the crude product was redissolved in EtOAc (50 mL) and washed with saturated NaHCO 3 solution. The aqueous phase was extracted with EtOAc (2 × 50 mL), and the combined organic fractions were dried with anhydrous MgSO 4 and filtered off, and the solvent was removed under vacuum to give 57 (1.30 g, 6.17 mmol, 96% yield) as a brown solid, which was used in the next step without further purification. 1  2. Enzyme Assay. The activity of Mth IMPDH ΔCBS was determined using a plate reader by monitoring the production of NADH in absorbance at 340 nm and corrected for noncatalyzed chemical reactions in the absence of Mth IMPDH ΔCBS. All the measurements were done in the assay buffer (50 mM Tris HCl pH 8, 1 mM DTT, 1 mM EDTA, and 100 mM KCl) at 37°C with 20 nM Mth IMPDH ΔCBS, 2.8 mM NAD + , and 1 mM IMP in a total of 150 μL volume in a 96 well platebased format, and data were collected for 32 min. The reaction was initiated by the addition of the substrate, IMP, at a concentration of 1 mM. All reactions were performed in triplicate. Prior to reaction initiation, the compounds were preincubated in a buffer with enzyme for 5 min. The inhibitors were dissolved in DMSO-d 6 and diluted to a final concentration of 1% v/v in experimental reactions.
IC 50 values were calculated by plotting the percentage of inhibition against the logarithm of inhibitor concentration, and dose−response curves were fitted using Prism software (GraphPad).
The K i value for NAD +  The initial velocities at various inhibitor concentrations were determined based on the slope in the linear part of each reaction containing the inhibitor and the uninhibited reaction. To determine the inhibition constant (K i values), the initial rate data versus substrate concentration at different inhibitor concentrations were fit using Prism software (GraphPad) to equations for uncompetitive or mixed inhibition. For each inhibitor concentration, the reciprocal of enzyme reaction velocity versus the reciprocal of the substrate concentration was plotted in a Lineweaver−Burk plot to determine the pattern of inhibition.
3. Protein Purification, Crystallization, and Data Collection of Mth IMPDH ΔCBS. Mth IMPDH was expressed, purified, and crystallized as previously described. 24,25 Briefly, hexahistidine tagged Mth IMPDH ΔCBS in pHat2 was expressed overnight in BL21 DE3 (NEB) cells at 18°C by addition of 500 μM IPTG. Cells were lysed in 50 mM Hepes, pH 8.0, 500 mM NaCl, 5% glycerol, 10 mM βmercaptoethanol, and 20 mM imidazole, and the recombinant protein purified using a Hi-Trap IMAC FF column (GE Healthcare) charged with nickel and an elution gradient of up to 300 mM imidazole. The hexahistidine tag was cleaved by TEV protease, and the purified Mth IMPDH ΔCBS was obtained by negative nickel gravity-flow purification 42 and size exclusion chromatography on a Superdex 200 gel filtration column equilibrated in 20 mM Hepes pH 8.0, 500 mM NaCl, 5% glycerol, and 1 mM TCEP step. The recombinant Mth IMPDH ΔCBS was then concentrated to 12.5 mg/mL for crystallization.
Mth IMPDH ΔCBS protein crystallized in 1 μL + 1 μL hanging drops with 100 mM sodium acetate, pH 5.5, 200 mM calcium chloride, and 8− 14% isopropanol. Crystals were soaked overnight in drops of well solution + 5 mM IMP and either 5 mM Fragment 2 or Compound 1 or 1 mM Compound 31. Crystals were cryoprotected by passing through drops containing well solution + 25% glycerol and flash-frozen in liquid nitrogen. Data were collected from the crystals at Diamond Light Source beamline.
4. Structure Solution, Ligand Fitting, and Refinement. Data were processed using XDS 43 and Pointless (CCP4). To solve the structure, molecular replacement was performed with Phenix Phaser 44 using a previously solved IMP-bound Mth IMPDH ΔCBS structure as a probe (PDB IDs: 5J5R; 5K4X; 5K4Z). Refinement was performed using Phenix.refine and manually in Coot. 45 IMP and the inhibitors were sequentially fitted into the density using the LigandFit function of Phenix, and the structures were manually refined further using Coot. Information regarding the crystallographic statistics can be found in Table S2. Protein−ligand interactions were analyzed using Arpeggio 46 and CSM-Lig. 47,48 Compound properties were evaluated using pkCSM. 49 All figures made using Pymol (Schrodinger).

Drug Susceptibility Testing against
Mtb. An Alamar Blue fluorescence-based broth microdilution assay was used to assess the minimum inhibitory concentration (MIC) of compounds against Mtb H37Rv, as described previously. 25,50 Briefly, Mtb H37Rv was grown in standard Middlebrook 7H9 broth (BD) supplemented with OADC (BD), 0.2% glycerol, and 0.05% Tween-80 to midexponential phase. Compounds dissolved in DMSO (1%) were tested in clear-bottomed, round-well 96-well microtiter plates at eight different concentrations using the standard anti-TB drugs, rifampin and isoniazid, as positive controls. An inoculum of ∼10 5 bacteria was added to each well, and the plates were incubated at 37°C for 7 d. On day 7, 10 μL of Alamar Blue (Invitrogen) was added to each well, and plates were further incubated at 37°C for 24 h. The fluorescence (excitation 544 nm; emission 590 nm) was measured in a FLUOstar OPTIMA plate reader (BMG LABTECH, Offenberg, Germany). Data were normalized to the minimum and maximum inhibition controls to generate a dose response curve (% inhibition) from which the MIC 90 was determined.  (Table S1), and X-ray data collection and refinement statistics (Table  S2)