Intracellular and in vivo evaluation of imidazo[2,1-b]thiazole-5-carboxamide anti-tuberculosis compounds

Summary

The imidazo[2,1-b]thiazole-5-carboxamides (ITAs) are a promising class of anti-tuberculosis agents shown to have potent activity in vitro and to target QcrB, a key component of the mycobacterial cytochrome bcc-aa3 super complex critical for the electron transport chain. Herein we report the intracellular macrophage potency of nine diverse ITA analogs with MIC values ranging from 0.0625–2.5 μM and mono-drug resistant potency ranging from 0.0017 to 7 μM. The in vitro ADME properties (protein binding, CaCo-2, human microsomal stability and CYP450 inhibition) were determined for an outstanding compound of the series, ND-11543. ND-11543 was tolerable at >500 mg/kg in mice and at a dose of 200 mg/kg displayed good drug exposure in mice with an AUC(0-24h) >11,700 ng·hr/mL and a >24 hr half-life. Consistent with the phenotype observed with other QcrB inhibitors, compound ND-11543 showed efficacy in a chronic murine TB infection model when dosed at 200 mg/kg for 4 weeks. The efficacy was not dependent upon exposure, as pre-treatment with a known CYP450-inhibitor did not substantially improve efficacy. The ITAs are an interesting scaffold for the development of new anti-TB drugs especially in combination therapy based on their favorable properties and novel mechanism of action.

Keywords:

tuberculosis – Mycobacterium tuberculosis – blood plasma – Spleen – Mouse models – High performance liquid chromatography – Macrophages – Animal models of infection

Introduction

Tuberculosis (TB) is among the worst diseases to plague humanity. TB kills about 5,000 people each day. As an insidious air borne pathogen, Mycobacterium tuberculosis, the causative agent of TB, simultaneously infects millions more each year [1]. The current treatment for drug sensitive TB is a dated combination regimen that includes rifampicin, isoniazid, ethambutol and pyrazinamide, which together cause a myriad of side-effects leading to reduced patient compliance [2]. Despite decades of focused effort to eradicate TB, this disease persists and is becoming more drug resistant. It is estimated that more than 550,000 people have drug resistant TB and only 54% of these people are successfully treated for various reasons, including-economics, drug availability, and lack of effective drugs [3]. It has taken over 40 years for newer agents, such as bedaquiline and pretomanid, to become conditionally approved for the treatment of drug resistant TB [4–6] Bedaquiline validated that targeting energy generation via inhibition of respiratory ATP synthesis is a viable approach, particularly within mycobacteria where respiration is so essential for survival [7]. Our labs have been pursuing new anti-tuberculosis compounds that target energy generation and have discovered and endeavored to optimize the imidazo[1,2-a]pyridine-3-carboxamide (IPA) class [8–15]. The IPAs, including Q203 (1, Fig 1), target the electron transport chain, specifically at QcrB, a component of the terminal cytochrome oxidase [16, 17]. Various scaffolds have emerged that target QcrB (Fig 1) [16–31], two classes, imidazo[2,1-b]thiazole-5-carboxamides (ITAs) (2) [26–28] and pyrazolo[1,5-a]pyridine-3-carboxamides (3) [28–31], bear the greatest structural homology and potency relative to the IPAs (Fig 1). We have disclosed the impressive in vitro properties of various ITAs, including low nanomolar potency against replicating and drug-resistant Mtb strains and low cyctotoxicity [26, 27], as exemplified by ND-11543 (2). However, the ITAs have not been evaluated ex vivo or in vivo. The pyrazolo[1,5-a]pyridine-3-carboxamides, exemplified by the lead compound TB47 (3), have demonstrated low nanomolar potency against replicating M. tuberculosis and a panel of 56 isolates (including multi- and extensively drug-resistant strains). TB47 itself has MIC values between 0.03–0.93 μM (0.016 and 0.500 μg/mL) [30, 31]. When co-dosed with rifampicin or pyrazinamide, TB47 showed synergistic bactericidal efficacy within the acute murine infection model of tuberculosis and comparable potency to Q203 within the chronic murine infection model [31]. Compounds 4, 5 and 6 (Fig 1) show some of the diverse chemical structures that can inhibit QcrB. Indeed, the gastric proton pump inhibitor lansoprazole (4) undergoes sulfoxide reduction to an active anti-tubercular compound (LPZS) against replicating Mtb (IC₉₀ = 1.1–1.7 μM, against three family strains) and drug-resistant Mtb (0.5–1.4 μM, against six strains) as well as efficacy within the acute murine infection model [20]. The arylvinylpiperazines, like AX-35 (5), were identified through a large phenotypic screening campaign conducted GlaxoSmithKline. AX-35 has an MIC of 0.14 μM (0.05 μg/mL), has demonstrated good ex vivo and in vivo efficacy, and is believed to interact with QcrB with a binding mode different than that of Q203 [21]. The morpholino thiophenes represent another point of diversity in structures that can inhibit QcrB. Compound 6 has an MIC of 0.2 μM, overlays well with Q203 by 2D comparison modeling, and has shown efficacy within the acute murine infection model of tuberculosis [22].

In this study, we have selected a focused set of ITAs and evaluated their potency against M. tuberculosis through ex vivo screening in the macrophage, which more closely mimics the physiological conditions of the disease state [32, 33]. Evaluation of ATP synthesis inhibitors within the macrophage, particularly QcrB inhibitors like the ITA class, might be especially informative as this assay takes into consideration the contribution of host cells in eradicating Mtb. Herein we report the microsomal stability and in vivo pharmacokinetic properties of an outstanding compound, ND-11543. We also demonstrate that ND-11543 is active when evaluated in the chronic murine model of TB.

Materials and methods

Determination of minimum inhibitory concentration (MIC)

MICs were determined in liquid media in 96-well, black, clear-bottom plates as described [14]. A 10-point, 2-fold serial dilution was run for each compound and bacterial growth was measured by OD₅₉₀ after 5 days of incubation at 37°C. Growth inhibition curves were fitted using the Levenberg–Marquardt algorithm [34]. The MIC was defined as the minimum concentration at which growth was completely inhibited and was calculated from the inflection point of the fitted curve to the lower asymptote.

Mtb non-replicating (low oxygen recovery assay—LORA)

The Low Oxygen Recovery Assay was carried out in 96-well plates as described [35]. Bacteria (M. tuberculosis strain H37Rv-LUX) were cultured in Dubos media with supplement (DTA) in the Wayne Model of hypoxia [36] for 18 days to enter hypoxia and used to seed 96-well plates containing compounds. Plates were incubated for 9 days under anaerobic conditions followed by 28h outgrowth under aerobic conditions. As a comparator, plates were incubated for 6 days under aerobic conditions. Growth was measured by luminescence. Growth inhibition curves were fitted using the Levenberg–Marquardt algorithm. The MIC was defined as the minimum concentration at which growth was completely inhibited and was calculated from the inflection point of the fitted curve to the lower asymptote.

Determination of eukaryotic cytotoxicity

Cytotoxicity was tested against the African green monkey adult kidney cell line (VERO, ATCC CRL-1586) as described [37]. Compounds were tested as 10-point, 3-fold serial dilutions in DMSO (final assay concentration of 1% DMSO) and viability was measured using CellTiter-Glo (Promega). Inhibition curves were fitted using the Levenberg–Marquardt algorithm. IC₅₀ was defined as the concentration required to reduce cell viability by 50% after 2-day incubation.

Intracellular mycobacterial replication assay

Raw 264.7 macrophages (ATCC^® TIB-71, purchased from American Type Culture Collection) were cultivated in RPMI-1640 Glutamax medium (Gibco) containing 10% heat-inactivated fetal bovine serum (FBS, Gibco; RPMI-FBS). Cells were cultured in a controlled 5% CO₂ atmosphere at 37°C. A recombinant strain of M. tuberculosis GC1237 of the W/Beijing type constitutively expressing a green fluorescent protein (Mtb-GFP) was used as a reporter for the intracellular mycobacterial replication assay [38]. This strain was grown in Middlebrook 7H9 media supplemented with 10% oleic acid-albumin-dextrose-catalase (OADC; Difco), 0.05% Tween 80 (Sigma-Aldrich), 0.5% glycerol (Euromedex), and 50 μg/ml of hygromycin B (Invitrogen). Cultures were maintained at 37°C under static conditions for up to 14 days to reach the exponential phase of bacterial growth before being used for the intracellular mycobacterial replication assay.

Raw 264.7 macrophage (4 x 10⁵ cells/mL) suspensions were infected with Mtb-GFP (MOI = 5) in RPMI-FBS for 2 h at 37°C under mild stirring (120 rpm). Cells were washed twice in RPMI-FBS and treated with 50 μg/mL of amikacin (Sigma-Aldrich) for 1 hr at 37°C under mild stirring (120 rpm). The infected cells were washed twice and 50 μL of cell suspension (2 x 10⁴ cells) were added per well in 384-well assay plates (Greiner) containing a 14 point 2-fold dilution series of ITA analog compounds, as well as positive (isoniazid MIC x100) and negative (DMSO 1%) controls. The MICs of tested compounds were assessed in comparison with MICs of two positive controls, BTZ043 (ND-10379) and Q203 (ND-11496). Microplates were incubated for 5 days at 37°C with 5% CO₂.

Macrophages were then stained with 5 μM Syto60 (Invitrogen) for 30 min at 37°C with 5% CO₂ before imaging. Image acquisitions were performed on an automated confocal microscope (PerkinElmer) using a 20X water objective. GFP-bacteria and Syto60 signal were detected using 488 and 630 nm laser excitation with a 540/75 and 690/70 nm emission filter, respectively. A series of six images were taken by well, and dedicated image-based analyses were performed using Columbus software (version 2.5.1, PerkinElmer) (Fig 2 and S1 Table in Supplementary data) [32, 33]. Briefly, the two images were first segmented to remove background and to normalize pixel intensities between images. Then, cell nuclei and cytoplasm were detected using an intensity detection algorithm applied on the Syto60 channel. Properties of intensity and morphology were filtered to correctly select the cell population. A spot detection algorithm based on the GFP channel was applied for the detection of Mtb-GFP in cells. The bacterial intensity and area in pixels were measured. The image-based parameters that correlate to bacterial growth are the percentage of infected cells and the bacterial area (pixel) per cell and per well. The number of cells informed on the cytotoxicity of the compound for effective concentrations. Typically, drugs demonstrated a dose-dependent decrease on both bacterial area and percentage of infected macrophages. In this study, the parameter used as a read-out was the area of bacteria present per cell and per well (expressed in px) for infected Raw264.7. Normalizations were performed based on the average values obtained for the negative (DMSO 1%) and positive (isoniazid 10 μg/mL) controls. Inhibition of bacterial replication was determined for each compound. Percentage of inhibition effect was plotted against the log10 of compound concentrations. Fitting was performed by Prism software using the sigmoidal dose-response (variable slope) model. The minimum inhibitory concentration (MIC) was determined and corresponded to the first concentration on the top of the curve, corresponding to the obtained maximum inhibition. Cell viability was measured by comparison of the cell number obtained for each compound concentration to the average value obtained for the positive control (isoniazid 10 μg/mL). Data of two replicates were averaged.

1: Typical 2-color image; 2 and 3: Circled objects correspond respectively to detected nuclei and detected cells, 4: Filled white objects correspond respectively to total intracellular bacteria area, 5: Circled green cells correspond to infected cells.

Solubility determination

The relative solubility of compounds was determined in microbiological medium using turbidity as a measure [27]. A serial dilution of compounds was prepared in DMSO and then transferred to microbiological medium at pH 6.8. Turbidity was measured and compared to a control (diethylstilbestrol, 125 μM); the lowest concentration at which compounds were insoluble (defined as turbidity >300% of control) was recorded.

Plasma protein binding

Plasma protein binding for each test compound was determined by equilibrium dialysis [27]. Compounds were tested using a semi-permeable membrane which separates two compartments containing protein (human plasma) and buffer. Molecules can penetrate freely, but proteins cannot pass through the membrane. Test compounds were mixed with human plasma and applied to the device; after equilibration at 37°C with PBS, the test compound in each compartment was quantified by LC-MS/MS. Human pooled plasma used in this experiment was purchased from Bioreclamation Inc. (catalog number HMPLNAHP).

Caco-2

The permeability of test compounds was assessed using a Caco-2 cell monolayer [27]. Compound permeability was measured in both directions. For A-B permeability, test compound was added to the apical side of the Caco-2 monolayer and the transport of compound to the basal side was monitored. For B-A permeability, test compound was added to the basal side of the Caco-2 monolayer and the transport of the compound to the apical side was monitored. Assays were run for 2h in duplicate. The amount of compound present in each compartment was quantified by LC-MS/MS. The Caco-2 cell line (ATCC HTB-3) used in this experiment was purchased from American Type Culture Collection.

Cytochrome P450 enzyme inhibition

Compounds were tested for inhibition of six cytochrome P450 enzyme isoforms—CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6 and CYP3A4 [27]. For each assay, human liver microsomes were incubated with a probe substrate for each CYP isoform in the presence of compound. Compounds were prepared as a 7-point dilution series in acetonitrile:DMSO (9:1). The final DMSO content in the reaction mixture was equal in all solutions used within an assay and was < 0.2%. Samples were run in duplicate.

Human microsomal stability

Compounds were tested for microsomal stability using pooled human liver S9 microsomes [27]. Microsomes were incubated with the test compound at 37°C in the presence of the co-factor NADPH. The reaction was terminated, the supernatant recovered, and test compounds quantified by LC-MS/MS. A fixed concentration of test compound was tested in duplicate at 5 timepoints and compound stability was expressed as a function of time. Human liver microsomes were purchased from Celsis (catalog number X008068).

Animal welfare and procedures

All animal studies were performed at Colorado State University in a certified bio-safety level III animal facility in strict accordance to the regulations and recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and the Centers of Disease Control. All procedures and performed protocols for infecting mice with M. tuberculosis and subsequent drug treatments in the described mouse infection studies were approved by the Colorado State University Institutional Animal Care and Use Committee (IACUC) (Reference numbers of approved protocols: 09-1367A, 12-3723A, 15-5942A, 18-8006A). All experiments were approved by IACUC prior to initiation of the mouse studies and conducted following the relevant guidelines and regulations. Mice were euthanized by CO₂ inhalation, a method approved by the CSU IACUC.

When animals are infected with M. tuberculosis, there is no physical discomfort when inside the aerosol machine. The animals can move freely inside large wire baskets, and this process is brief [less than 25 minutes]. The veterinarians at the Laboratory Animal Resources at Colorado State University have procedures in place to control animal pain, as defined by the Karnofski scale. The parameters used to monitor pain and severity of disease of M. tuberculosis infected mice are as follows: 1) parameters monitored daily to determine the condition of the mice (activity and temperament, feeding behavior, appearance), 2) parameters, which when evident, require that the mice be humanely euthanized within the day (severe physiologic changes, 20% weight loss). Analgesics, anesthesia, or sedation

In vivo safety study

In this acute toxicity test, 3 healthy female mice (6–8 week old Balb/c from Charles River) were given three consecutive daily doses of ND-11543 and were observed at regular times for any adverse effects. ND-11543 was dissolved in 20% TPGS (d-α tocopheryl polyethylene glycol 1000 succinate) to give a fine white suspension which was dosed at 100, 300 and 500 mg/kg orally on three consecutive days, 3 mice per dose group. Mice were observed 10 min, 1, 2, 4 hr after dosing and then daily afterwards. During the observations, humane handling practices and animal welfare regulations were strictly followed.

Single dose plasma PK

Basic pharmacokinetic analysis was determined after a single dose of 200 mg/kg with and without the administration of 100 mg/kg of 1-aminobenzotriazole (ABT) (Sigma, #A3940) given 2 hr by gavage prior to ND-11543 administration. ABT (100 mg/kg dose) was administered by oral gavage in 0.5% (w/v) methyl cellulose 400 solution. ND-11543 (200 mg/kg dose) was prepared in 20% TPGS and dosed orally 2 hr after ABT administration. Female mice (6–8 week old Balb/c from Charles River) were rested one to two weeks before dosing. Blood was collected from the mice at time points of 0.5, 2, 4, 8, 12 and 24 hr after drug administration. Blood samples (< 100 μL each) for all time points were collected via the mandibular vein in plasma collector tubes, centrifuged and plasma (< 50 μL each) was frozen until further use. Plasma samples were analyzed by LC/MS to determine drug plasma levels.

In vivo chronic Balb/c model of TB infection

A chronic Balb/c infection model with M. tuberculosis Erdman was carried out as previously described [39]. Female Balb/c mice, aged 6−8 weeks (from Charles River), were infected by the aerosol route on day 0 using the inhalation exposure system (Glas-col Inc., Terre Haute, IN) to give a lung bacterial load of an average of ~50–100 colony forming units (CFU) per mouse. Three mice were sacrificed day 1 post-infection to determine bacterial uptake. Whole lungs were extracted, homogenized in PBS buffer and plated on 7H11/OADC agar plates. The plates were placed in a 37^o C dry-air incubator for ~3 weeks. At day 28 post-infection, 5 mice were sacrificed to determine bacterial load in the lungs and spleens at the start of therapy. Lung and spleen homogenate homogenized in PBS buffer were plated on 7H11/OADC agar plates. Each compound was administered via oral gavage on day 28 post-infection with and without the administration of 100 mg/kg of 1- aminobenzotriazole ABT (Sigma, #A3940) given 2 hr prior to test dose and continued five days a week for 30 days until day 58. A control group, isoniazid (INH), was administered via oral gavage at 25 mg/Kg/day, at 200 μL/mouse. Day 58 post-infection, 5 mice of each group were sacrificed, and bacterial loads were determined. Lung and spleen homogenate in PBS buffer were plated on 7H11/OADC agar plates. Statistical analysis was performed by first converting CFU to logarithms, which were then evaluated by a one-way ANOVA followed by a multiple comparison analysis of variance by a one-way Tukey test (SigmaStat software program). Differences were considered significant at the 95% level of confidence.

Chemical syntheses

The syntheses and characterization of imidazo[2,1-b]thiazole-5-carboxamide (ITA) ND-11543 has been previously reported [26] and was profiled in many additional assays described herein.

Chemistry

All anhydrous solvents, reagent grade solvents for chromatography and starting materials were purchased from either Aldrich Chemical Co. (Milwaulkee, WI) or Fisher Scientific (Suwanee, GA) unless otherwise noted. Water was distilled and purified through a Milli-Q water system (Millipore Corp., Bedford, MA). General methods of purification of compounds involved the use of silica cartridges purchased from Practichem, LLC. (www.practicachem.com) and/or recrystallization. The reactions were monitored by TLC on precoated Merck 60 F254 silica gel plates and visualized using UV light (254 nm). All compounds were analyzed for purity by HPLC and characterized by ¹H and ¹³C NMR using a Bruker DPX Avance I NMR Spectrometer (300MHz) and/or a Bruker Ascend Avance III HD Spectrometer (500 MHz). Chemical shifts are reported in ppm (δ) relative to the residual solvent peak in the corresponding spectra; chloroform δ 7.27 and δ 77.23, methanol δ 3.31 and δ 49.00 and coupling constants (J) are reported in hertz (Hz) (where, s = singlet, bs = broad singlet, d = doublet, dd = double doublet, bd = broad doublet, ddd = double doublet of doublet, t = triplet, tt–triple triplet, q = quartet, m = multiplet) and analyzed using ACD NMR data processing. ¹⁹F NMR were run without a standard and are uncorrected. Mass spectra values are reported as m/z. Melting points were measured on a Thomas-Hoover capillary melting point apparatus and are uncorrected.

LC-MS method

Liquid Chromatography-Mass Spectrometry method was performed on an Agilent 1290 infinity coupled to Agilent 6538 Ultra High Definition Quadrupole Time of Flight (UHD-QToF) instrument. A separation was achieved by using reverse phase Waters Acquity UPLC HSS T3 1.8μm (2.1 X 100mm) column from Waters (Milford, USA). All solvents were purchased from Fischer Scientific LCMS Optima grade solvents. Water containing 0.1% formic acid was used as mobile phase A and acetonitrile containing 0.1% formic acid was used as mobile phase B. The injection volume was set at 1 μL. Samples were injected in a gradient of 95% mobile phase A and 5% mobile phase B in the initial condition to 5% mobile phase A and 95% mobile phase B in 9 min. The eluent was held at that composition for an additional 3 min and switched back to the initial condition at 12 min.

The MS data acquisition was performed from 50-1000m/z at 1.0 spectra/sec scan rate. The source gas temperature was set at 350°C with a flow of 8 l/min. The nebulizer gas was set at 55 psig. The capillary voltage was set at 3500 volts with fragmentor at 100, skimmer at 45 and octopole RF 500 volts. Prior to sample runs, the instrument was calibrated using Agilent low mass calibrant solution.

Data analysis

The data collected in Agilent LC-MS was analyzed using Agilent Mass Hunter software for HRMS calculation.

Representative example of EDC-mediated coupling used to prepare analogs

Synthesis of N-(3-Fluoro-4-(4-(5-(trifluoromethyl)pyridin-2-yl)piperazin-1-yl)benzyl)-2,6-dimethylimidazo[2,1-b]thiazole-5 carboxamide (ND-11543) [26].

2,6-Dimethylimidazo[2,1-b]thiazole-5-carboxylic acid (CAS #: 1007875-19-5, 1.17 g, 5.9 mmol) and (3-fluoro-4-(4-(5-(trifluoromethyl)pyridin-2-yl)piperazin-1-yl)phenyl)methanamine (CAS #: 1897429-40-1, 2.29 g, 7.2 mmol) were dissolved in dry CH₃CN. EDC-HCl (1.38 g, 7.2 mmol) and 4-(dimethylamino)pyridine (DMAP, 965 mg, 7.2 mmol) were added and the reaction was stirred for 12 h at room temperature. The reaction was concentrated in vacuo. The residue was taken up in CH₂Cl₂ and washed with 5% aqueous acetic acid (2x), saturated aqueous NaHCO₃ (2x), brine and then dried over sodium sulfate. The drying agent was removed by filtration and the organic layer was concentrated in vacuo. The resulting residue was purified by recrystallization with hot CH₃CN to give N-(3-fluoro-4-(4-(5-(trifluoromethyl)pyridin-2-yl)piperazin-1-yl)benzyl)-2,6-dimethylimidazo[2,1-b]thiazole-5-carboxamide (ND-11543) as an off white solid, 2.61 g (62% yield). ¹H NMR (500 MHz, CDCl₃) δ ppm 8.40 (d, J = 2.5 Hz, 1H), 7.99–7.95 (m, 1H), 7.63 (dd, J = 9.0, 2.6 Hz, 1H), 7.09–7.03 (m, 2H), 6.92 (t, J = 8.4 Hz, 1H), 6.67 (d, J = 9.0 Hz, 1H), 5.99 (bt, J = 5.8 Hz, 1H, NH), 4.57 (d, J = 5.8 Hz, 2H), 3.79 (t, J = 5.0 Hz, 4H), 3.15 (t, J = 5.0 Hz, 4H), 2.56 (s, 3H), 2.41 (s, 3H). ¹³C (126 MHz, CDCl₃) δ ppm 160.43, 160.30, 156.69, 154.72, 145.71 (aq, J = 4.0 Hz), 144.01, 139.12 (d, J = 8.6 Hz), 134.57 (d, J = 2.6 Hz), 134.38 (d, J = 7.1 Hz), 126.75, 123.53 (d, J = 2.7 Hz), 122.33 (q, J = 271.6 Hz), 119.28 (d, J = 2.7 Hz), 118.10, 115.65, 115.52 (q, J = 33.4 Hz), 115.49, 105.64, 50.29 (d, J = 2.4 Hz), 44.84, 42.56, 16.48, 13.93. ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -61.12 (s, 3F), -121.84 (dd, J = 12.9, 8.7 Hz, 1F). HRMS (EI), M + 1 calcd. for C₂₅H₂₅F₄N₆OS, 533.1759; found 533.1741. HPLC t_R = 5.2–5.3 min, mp = 234–235 ^oC.

N-((2,3-Dihydrobenzofuran-5-yl)methyl)-2,6-dimethylimidazo[2,1-b]thiazole-5-carboxamide (ND-11459) was prepared by the method described for ND-11543 except 5-(aminomethyl)-2,3-dihydrobenzofuran hydrochloride (CAS #: 635309-62-5) was used.

¹H NMR (500 MHz, CDCl₃) δ ppm 8.00 (q, J = 1.2 Hz, 1H), 7.21 (s, 1H), 7.09 (d, J = 8.2 Hz, 1H), 6.76 (d, J = 8.2 Hz, 1H), 5.92 (bt, J = 6.0 Hz, 1H, NH), 4.57 (t, J = 8.6 Hz, 2H), 4.56 (d, J = 6.3 Hz, 2H), 3.20 (t, J = 8.9 Hz, 2H), 2.55 (s, 3H), 2.44 (d, J = 1.5 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ ppm 160.37, 159.70, 150.38, 143.94, 130.23, 127.71, 127.62, 126.51, 124.59, 118.22, 118.12, 109.35, 71.34, 43.24, 29.65, 16.43, 13.91. HRMS (EI), M + 1 calcd. for C₁₇H₁₈N₃O₂S, 328.1120; found 328.1105. HPLC t_R = 6.3 min, mp = 129–130°C.

N-((2,3-Dihydrobenzofuran-5-yl)methyl)-6-ethyl-2-methylimidazo[2,1-b]thiazole-5-carboxamide (ND-11503) was prepared by the method described for ND-11543 except 5-(aminomethyl)-2,3-dihydrobenzofuran hydrochloride (CAS #: 635309-62-5) was used.

¹H NMR (600 MHz, CDCl₃) δ ppm 8.01 (s, 1H), 7.20 (s, 1H), 7.09 (d, J = 7.7 Hz, 1H), 6.76 (d, J = 8.5 Hz), 5.93 (bt, J = 6.2 Hz, 1H, NH), 4.58 (t, J = 8.8 Hz, 2H), 4.57 (d, J = 5.2 Hz, 2H), 3.21 (t, J = 8.4 Hz, 2H), 2.85 (q, J = 7.7 Hz, 2H), 2.44 (s, 3H), 1.35 (t, J = 7.7 Hz, 3H). ¹³C NMR (151 MHz, CDCl₃) δ ppm 160.38, 159.71, 150.50, 149.64, 130.18, 127.72, 127.59, 126.59, 124.57, 118.14, 117.49, 109.36, 71.35, 43.27, 29.65, 23.42, 13.95, 13.35. HRMS (EI), M + 1 calcd. for C₁₈H₂₀N₃O₂S, 342.1276; found 342.1269. HPLC t_R = 6.8 min, mp = 108–109°C.

2,6-Dimethyl-N-(4-(3-(trifluoromethyl)phenoxy)benzyl)imidazo[2,1-b]thiazole-5-carboxamide (ND-11564) was prepared by the method described for ND-11543 except [4-[(3-trifluoromethylphenyl)oxy]benzyl]amine (CAS #: 864263-13-8) was used.

¹H NMR (500 MHz, CDCl₃) δ ppm 8.00 (q, J = 1.2 Hz, 1H), 7.44 (dd, J = 7.7, 7.7 Hz, 1H), 7.40–7.34 (m, 3H), 7.25 (dd, J = 1.8, 1.8 Hz, 1H), 7.04–7.00 (m, 2H), 6.04 (bt, J = 5.5 Hz, 1H, NH), 4.67 (d, J = 5.8 Hz, 2H), 2.60 (s, 3H), 2.44 (d, 1.5 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ ppm 160.49, 157.65, 155.66, 150.51, 144.04, 134.16, 132.27 (q, J = 33.3 Hz), 130.32, 129.57, 126.71, 123.56 (q, J = 273 Hz), 121.64, 119.78 (q, J = 3.1 Hz), 119.65, 118.12, 115.31 (q, J = 3.9 Hz), 42.81, 16.48, 13.92. ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -66.66 (s, 3H). HRMS (EI), M + 1 calcd. For C₂₂H₁₉F₃N₃O₂S, 446.1150; found 446.1136. HPLC t_R = 8.6 min, mp = 103–104°C.

2,6-Dimethyl-N-(4-(3-(trifluoromethyl)phenoxy)phenyl)imidazo[2,1-b]thiazole-5-carboxamide (ND-11566) was prepared by the method described for ND-11543 except 4-(3-(trifluoromethyl)phenoxy)aniline (CAS #: 41605-31-6) was used.

¹H NMR (600 MHz, CDCl₃) δ ppm 7.99 (s, 1H), 7.60 (d, J = 8.8 Hz, 2H), 7.47 (s, 1H), 7.44 (t, J = 8.0 Hz, 1H), 7.34 (d, J = 7.7 Hz, 1H), 7.24 (s, 1H), 7.17 (d, J = 8.4 Hz), 7.07 (d, J = 8.8 Hz, 2H), 2.73 (s, 3H), 2.46 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ ppm 158.54, 158.12, 152.55, 151.09, 144.46, 133.80, 132.26 (q, J = 33.8 Hz), 130.32, 127.14, 124.59 (q, J = 272 Hz), 122.24, 121.19, 120.40, 119.56 (q, J = 4.4 Hz), 118.32, 118.07, 114.83 (q, J = 3.3 Hz), 16.61, 13.98. ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -62.65 (s, 3F). HRMS (EI), M + 1 calcd. For C₂₁H₁₇F₃N₃O₂S, 432.0994; found 432.0974. HPLC t_R = 8.9 min, mp = 149–150°C.

N-(4-(1,1-Dioxidothiomorpholino)-3-fluorobenzyl)-2,6-dimethylimidazo[2,1-b]thiazole-5-carboxamide (ND-11903) was prepared by the method described for ND-11543 except 4-(4-(aminomethyl)-2-fluorophenyl)thiomorpholine 1,1-dioxide (CAS #: 1155537-01-1) was used.

¹H NMR (500 MHz, CDCl₃) δ ppm 7.98 (q, J = 1.6 Hz, 1H), 7.11–7.07 (m, 2H), 6.96 (dd, J = 8.3, 8.3 Hz, 1H), 6.04 (bt, J = 5.5 Hz, 1H, NH), 4.59 (d, J = 6.1 Hz, 2H), 3.62–3.59 (m, 4H), 3.23–3.18 (m, 4H), 2.59 (s, 3H), 2.44 (d, J = 1.3 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ ppm 160.49, 153.62 (d, J = 247.0 Hz), 150.60, 144.13, 137.62 (d, J = 10.0 Hz), 135.03 (d, J = 7.5 Hz), 126.80, 123.57 (d, J = 3.7 Hz), 120.84 (d, J = 2.7 Hz), 118.08, 117.99, 115.80 (d, J = 23.1 Hz). ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -121.90 (dd, J = 12.2, 9.2 Hz, 1F). HRMS (EI), M + 1 calcd. For C₁₉H₂₂FN4₃O₃S₂, 437.1117; found 437.1099. HPLC t_R = 5.98 min, mp = 230–232°C.

N-(4-(1,1-Dioxidothiomorpholino)benzyl)-2,6-dimethylimidazo[2,1-b]thiazole-5-carboxamide (ND-12015) was prepared by the method described for ND-11543 except 4-(4-(aminomethyl)phenyl)thiomorpholine 1,1-dioxide (CAS #: 1152880-11-9) was used.

¹H NMR (600 MHz, CDCl₃) δ ppm 8.00 (s, 1H), 7.31 (d, J = 8.4 Hz, 2H), 6.91 (d, J = 8.8 Hz, 2H), 5.96 (bt, J = 5.0 Hz, 1H, NH), 4.59 (d, J = 5.5 Hz, 2H), 3.86 (dd, J = 4.7, 4.7 Hz, 4H), 3.11 (dd, J = 5.2, 5.2 Hz, 4H), 2.58 (s, 3H), 2.44 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ ppm 160.41, 150.39, 147.05, 143.85, 130.57, 129.23, 126.73, 118.16, 118.11, 116.58, 50.46, 47.64, 42.77, 16.44, 13.92. HRMS (EI), M + 1 calcd. For C₁₉H₂₃N₄O₃S₂, 419.1212; found 419.1203. HPLC t_R = 5.6 min, mp = 214–215°C.

6-Ethyl-2-methyl-N-((6-(3-(trifluoromethyl)phenoxy)yridine-3-yl)methyl)imidazo[2,1-b]thiazole-5-carboxamide (ND-12024) was prepared by the method described for ND-11543 except (6-(3-(trifluoromethyl)phenoxy)yridine-3-yl)methanamine (CAS #: 197459-39-5) was used.

¹H NMR (500 MHz, CDCl₃) δ ppm 8.18 (s, 1H), 7.98 (d, J = 1.3 Hz, 1H), 7.79 (dd, J = 8.6, 2.2 Hz, 1H), 7.52 (dd, J = 7.9, 7.9 Hz, 1H), 7.46 (d, J = 7.7, 1H), 7.41 (s, 1H), 7.33 (d, J = 7.7 Hz, 1H), 6.98 (d, J = 8.6 Hz, 1H), 6.08 (bt, J = 5.5 Hz, 1H, NH), 4.64 (d, J = 5.8 Hz, 2H), 2.88 (q, J = 7.7 Hz, 2H), 2.44 (s, 3H), 1.38 (t, J = 7.7 Hz, 3H). ¹³C NMR (125 MHz, CDCl₃) δ ppm 162.55, 160.58, 154.22, 150.77, 149.94, 146.67, 139.65, 132.08 (q, J = 35.4 Hz), 130.11, 129.45, 126.91, 124.54, 123.69 (q, J = 270 Hz), 121.35 (q, J = 5.6 Hz), 118.24 (q, J = 3.1 Hz), 118.06, 117.19, 112.04, 40.31, 23.51, 13.93, 13.35. ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -62.59 (s, 3F). HRMS (EI), M + 1 calcd. For C₂₂H₂₀F₃N₄O₂S, 461.1259; found 461.1250. HPLC t_R = 8.3 min, mp = 104–106°C.

2,6-Dimethyl-N-((6-(3-(trifluoromethyl)phenoxy)yridine-3-yl)methyl)imidazo[2,1-b]thiazole-5-carboxamide (ND-12025) was prepared by the method described for ND-11543 except (6-(3-(trifluoromethyl)phenoxy)pyridine-3-yl)methanamine (CAS #: 197459-39-5) was used.

¹H NMR (600 MHz, CDCl₃) δ ppm 8.18 (s, 1H), 7.98 (s, 1H), 7.80 (dd, J = 8.5, 2.2 Hz, 1H), 7.51 (t, J = 8.0 Hz, 1H), 7.46 (d, J = 7.7 Hz, 1H), 7.41 (s, 1H), 7.33 (d, J = 8.0 Hz, 1H), 6.97 (d, J = 8.5 Hz, 1H), 6.07 (bt, J = 5.4 Hz, 1H, NH), 4.63 (d, J = 5.9 Hz, 2H), 2.59 (s, 3H), 2.45 (s, 3H). ¹³C NMR (151 MHz, CDCl₃) δ ppm 162.56, 160.55, 154.21, 150.61, 146.69, 144.17, 139.68, 132.03 (q, J = 31.3 Hz), 130.11, 129.47, 126.89, 123.71 (q, J = 272 Hz), 124.55, 121.36 (q, J = 3.3 Hz), 118.25 (q, J = 4.5 Hz), 118.05, 117.95, 112.04, 40.28, 116.49, 13.92. ¹⁹F NMR (282 MHz, CDCl₃) δ ppm -62.59 (s, 3F). HRMS (EI), M + 1 calcd. For C₂₁H₁₈F₃N₄O₂S, 447.1103; found 447.1091. HPLC t_R = 7.9 min, mp = 105–106°C.

Results and discussion

Nine imidazo[2,1-b]thiazole-5-carboxamides were prepared and evaluated in this study (ND-11543, Figs 1 and 3) along with clinical candidates Q203 [16] (Fig 1) and BTZ043 [40] (Fig 3) as assay controls.

The in vitro potency against replicating H37Rv-Mtb, non-replicating H37Rv-Mtb (LORA), toxicity to mammalian cells (VERO) and aqueous solubility were determined for these compounds and a positive control, Q203 (Table 1). The ITA class showed very similar SAR trends to those observed with very structurally similar imidazo[1,2-a]pyridine-3-carboxamines [9,10,15,17,18] and pyrazolo[1,5-a]pyridine-3-carboxamides [29–31]. For instance, the most potent compounds were often the most lipophilic compounds as exemplified by ND-11543 and ND-11564 (clogP of 5.4 and 6.6, respectively) whereas polar compounds like ND-11903 and ND-12015 (clogP of 2.4 and 2.1, respectively) were >100-fold less active by comparison (Table 1). Benzylic amides were superior to anilides as demonstrated by ND-11564 and ND-11566 having a >50-fold difference in potency (MIC = 0.160 μM vs. 9.1 μM). Two compounds (ND-11459 and ND-11503) obey the “rule of five” [41] rubric and had good potency (MIC = 0.25–0.54 μM) and high aqueous solubility ≥200 μM at pH 6.8 (Table 1). All compounds showed greatly diminished activity in anerobic, non-replicating conditions (MIC ranging from 1.1 to 200 μM). This could be due to the mechanism of action of these compounds, targeting inhibition of cytochrome C oxidase [26] and the respiratory flexibility of that target [25]. All compounds showed no cellular toxicity (IC₅₀ >50 μM).

ITAs are potent against intracellular M. tuberculosis

The ability of compounds to inhibit M. tuberculosis ex vivo is a desirable characteristic. We tested a focused set of four ITAs to determine their activity in an intracellular model of M. tuberculosis infection along with two positive controls BTZ043 (ND-10379) and Q203 (ND-11496). Experimental data from the intracellular M. tuberculosis growth assay are shown in Fig 4. As anticipated, the positive controls BTZ043 and Q203 displayed very potent activity as reflected by an MIC of 0.0078 μM for each in these experiments (Fig 4(A) and 4(B), respectively) which is consistent with previous reports of about <0.098 μM [27]. The four selected ITA compounds were active against M. tuberculosis in the macrophage assay. ND-11543 and 11564 (Fig 4(E) and 4(F); respectively) were shown to exhibit lower MIC values of 0.0625 and 0.25 μM, respectively, than those obtained for the 2 other compounds (ND-11503 and 11459) ranging from 1.25–2.5 μM (Fig 4(D) and 4(C); respectively). All compounds were devoid of cytotoxicity on murine Raw 264.7 macrophages. Effect on the cell viability was shown for ND-11564, ND-11459 and ND-11543 which, respectively, showed macrophage toxicity at concentrations 4 and 8-fold higher than their MIC. These experiments showed that imidazo[2,1-b]thiazole-5-carboxamide compounds are active against pathogens residing intracellularly in macrophages, particularly against Mycobacterium tuberculosis, and highlight two more potent ITA compounds, ND-11543 and 11564.

ITAs are potent against drug resistant strains

With the rapid rise in drug resistant TB, finding new compounds which can potentially inhibit strains resistant to first line drugs is of great interest. To that end, we screened four compounds and four positive controls (Q203, isoniazid, rifampicin, and levoflaxacin) against a panel of five drug resistant strains (Table 2). Screening reported here revealed that even moderately active compounds like ND-11903 and ND-11529 remain active even against hard to kill drug resistant strains and highlights the low nanomolar potency of ND-11543 particularly against RIF and INH resistant strains.

In vitro ADME of ND-11543

ND-11543 was a good candidate for ADME evaluation based upon its impressive in vitro potency against replicating and drug-resistant Mtb. Table 3 relates the plasma protein binding, Caco-2 permeability, cytochrome P450 inhibition, and human microsomal stability of ND-11543. ND-11543 was completely protein bound (100%) consistent with compounds with higher clogP values like Q203 [27]. Caco-2 values give an approximation of compound absorption and compounds with an efflux ratio of greater than 2 might be subject to active efflux. However, this is not expected to be an issue for ND-11543 as talinolol (a known P-glycoprotein substrate with blood brain transport) has an efflux ratio of 93.9 in this assay. Another desired characteristic of new TB therapeutics would be to have low drug-drug interactions. ND-11543 has high IC₅₀ values ranging from >10 μM to > 20 μM against all the major cytochrome P450 isoforms coupled with low MIC values against H37Rv Mtb (Table 3). Lastly, ND-11543 demonstrated moderate stability with a half-life of 28.4 min in human liver microsomes.

In vivo tolerability of ND-11543

Based upon its encouraging in vitro ADME properties, ND-11543 was assessed in an acute safety and tolerability mouse model to determine its Maximum Tolerated Dose, MTD. No acute toxicity was observed resulting in an MTD of >500 mg/kg.

In vivo PK of ND-11543

Based on MIC and acute toxicity data, ND-11543 was selected for evaluation in a single dose pharmacokinetic study with and without pre-dosing of a CYP-inhibitor (ABT) in Balb/c mice (Table 4). ND-11543 had reasonable drug exposure (AUC = 11,704 ng·hr/mL over 24 hr) and a Cmax (645 ng/mL) that remained above the MIC for 24 hr. The absolute AUC could not be calculated due to drug being present after 24 hr. When ABT was added, with a three-fold increased drug exposure (AUC = 37,250 ng·hr/mL over 24 hr) and Cmax (2,411 ng/mL). The large shift in tmax in presence of ABT (from 2 to 24 hr) resulted in a long half-life. This compound appeared to be slowly absorbed in presence of ABT as drug levels increased over 24 hr.

ND-11543 is active in a chronic model of TB

ND-11543 was evaluated in a chronic Balb/c murine model of TB at either 100 or 200 mg/kg oral dose with or without ABT (Tables 5 and 6). The dose of 200 mg/kg ND-11543 with ABT was selected as this dose results in similar plasma drug exposure over 24 hr compared to Q203 (AUC 0–24 = 37,250 ng·hr/mL vs. 44,100 ng·hr/mL; respectively) [16]. As observed with other QcrB inhibitors, Q203 and TB47, ND-11543 was found to be active in the chronic murine model [30].

Mice treated with 200 mg/kg of ND-11543 showed a slight reduction in CFU in both lung and spleen (0.30 and 0.36 log10 CFU, respectively), but this was not statistically significant relative to control (Table 5) (P > 0.05). However, ND-11543 (200 mg/kg) + ABT (100 mg/kg) treatment resulted in a 0.59 log10 CFU reduction in the lung, and a 0.77 log10 CFU reduction in the spleen, both of which were statistically significant (p = 0.03, p<0.001 in lung and spleen, respectively). Two animals in this group showed adverse effects late in the first week of dosing. One animal was euthanized and the other animal, which showed minor lethargy, recovered. There were no other adverse effects noted in this treatment group over the course of the study. Due to the slight adverse effects at 200 mg/kg, ND-11543 (100 mg/kg) + ABT (100 mg/kg) treatment was evaluated in a follow up study (Table 6). This treatment resulted in a 0.29 log10 CFU reduction in the lung, and a 0.82 log10 CFU reduction in the spleen, both of which were statistically significant (p = 0.03, p<0.001 in lung and spleen, respectively).

Conclusions

Nine imidazothiazoles (ITAs) were prepared and evaluated for their in vitro and in vivo properties revealing many potent compounds against replicating, drug resistant and intracellular Mtb. ND-11543 displayed good in vivo pharmacokinetics and tolerability. To the best of our knowledge, this is the first disclosure of an ITA that could halt TB disease progression in the chronic murine model of TB infection. Recent literature has shown that other QcrB inhibitors (namely, Q203 and TB47) have better efficacy within the acute model of TB infection and when used in combination with first line drugs such as rifampicin and pyrazinamide [30]. Due to the respiratory flexibility of Mtb there have been many reports that the efficacy of QcrB inhibitors might be greatly enhanced by tandem inhibition of the cytochrome bd oxidase based upon in vitro data [21, 22, 43–45]. However, this dual inhibition concept was strengthened by the impressive in vivo efficacy observed when QcrB inhibitors were evaluated against Mycobacterium ulcerans, a disease-causing strain which lacking cyt-bd oxidase [46–48]. Overall, ND-11543 presents a proof of principle compound which sets a new foundation for the design of TB drugs based on the imidazo[2,1-b]thiazole-5-carboxamide class.

Supporting information

S1 Table [docx]
Analysis sequence using Columbus system (version 2.3.1, PerkinElmer).