Synthesis and biological evaluation of aryl-oxadiazoles as inhibitors of Mycobacterium tuberculosis

Graphical abstract


a b s t r a c t
Despite increased research efforts to find new treatments for tuberculosis in recent decades, compounds with novel mechanisms of action are still required. We previously identified a series of novel aryl-oxadiazoles with anti-tubercular activity specific for bacteria using butyrate as a carbon source. We explored the structure activity relationship of this series. Structural modifications were performed in all domains to improve potency and physico-chemical properties. A number of compounds displayed sub-micromolar activity against M. tuberculosis utilizing butyrate, but not glucose as the carbon source. Compounds showed no or low cytotoxicity against eukaryotic cells. Three compounds were profiled in mouse pharmacokinetic studies. Plasma clearance was low to moderate but oral exposure suggested solubilitylimited drug absorption in addition to first pass metabolism. The presence of a basic nitrogen in the linker slightly increased solubility, and salt formation optimized aqueous solubility. Our findings suggest that the 1,3,4-oxadiazoles are useful tools and warrant further investigation. Tuberculosis is the leading cause of death from infectious disease. 1 Although the number of tuberculosis cases decreased during the 20th century, the emergence of HIV and the incidence of multiple-drug resistance have increased the difficulty of treating many new cases. 1,2 Despite efforts to improve the outcome of tuberculosis care, the discovery of new antibiotics against the causative agent Mycobacterium tuberculosis has been insufficient to eradicate the disease. 3 New and more effective drugs with novel mechanisms of action are required to shorten treatment, improve patient adherence, and reduce the appearance of resistance.
Mycobacterium tuberculosis can adapt metabolically to host environments and can catabolize multiple carbon sources simultaneously. 4 Fatty acids are the major carbon source available during infection, 5 although carbohydrates, lipids, and carbon dioxide can also be utilized as carbon sources. 6 We recently reported the identification of a family of oxadiazoles 1-5 ( Fig. 1) 7 from a whole cell screen against M. tuberculosis using butyrate as the carbon source. The compounds were active in medium containing butyrate, but not glucose and lacked mammalian cytotoxicity. 7,8 The lack of cytotoxicity and the low molecular weight prompted us to undertake structure activity relationship (SAR) investigations around this series.
Aryl-oxadiazoles, the common structural motif in compounds 1-5, have been widely applied in medicinal chemistry for the development of new drugs. Compounds containing the 1,2,4-and 1,3,4-oxadiazole motif have been evaluated against a broad spectrum of pharmacological activities, with special attention to their properties as antimicrobial and antitubercular agents. [9][10][11][12] Synthetic methods for the preparation of differently functionalized 1,3,4-oxadiazoles have been recently reviewed. 13   In order to prepare thiadiazole 10, reaction of furan-2-carbohydrazide 10a with chloroacetyl chloride in the presence of N-methylmorpholine produced the intermediate acylsemicarbazide 10b (Scheme 2). Acylsemicarbazide 10b was refluxed with Lawesson's reagent in THF to obtain the intermediate chloride 10c. Chloride replacement by 2-chloro-6-fluorobenzylamine at reflux in the presence of DIPEA and sodium iodide generated the secondary amine 10d which was treated with sodium hydride and methyl iodide to give compound 10. Compounds 11, 12 and 19 were prepared from the corresponding oxadiazole analogue to chloride 10c by reaction with the appropriate primary amine followed by methylation.
The synthesis of the oxadiazole 21 containing an ether group in the linker was achieved by cyclization of acylsemicarbazide 10b using phosphoryl chloride to create 21a followed by the reaction of the intermediate 21a with (2-chloro-6-fluorophenyl)methanol (Scheme 3). Amide 22 was prepared from chloride 21a in two steps, via 22a. Substitution of chloride 21a by methylamine followed by amide formation using 2-chloro-6-fluorobenzoic acid in the presence of HATU and triethylamine provided the amide 22 (Scheme 3). Compound 20 was prepared by reaction of intermediate 21a with 2-chloro-6-fluorobenzylamine using DIPEA and sodium iodide.
The synthesis of compound 23 utilized the three-step route shown in Scheme 4. Commercially available 1,3,4-oxadiazol-2-one 23a was refluxed with 2-chloro-6-fluoro-phenethylamine 23b in ethanol to obtain intermediate 23c that was cyclized by heating with phosphoryl chloride. Methylation of secondary amine 23d using methyl iodide in the presence of sodium hydride in DMF afforded compound 23.
We began by evaluating mouse microsomal metabolism and thermodynamic solubility for aryl-oxadiazole hits from the screen (Fig. 1, Table 1), and learned that these compounds were all highly metabolized, and solubility varied. We then focused our attention on the most potent hit from the screen, compound 1 (previously reported MIC was 0.4 ± 0.1 mM in butyrate medium) 7 and synthesized compounds 6-9 ( Fig. 2) to answer specific structure-activity questions. In order to compare 2-thienyl-with 2-furyl-oxadiazole, the thiophene ring in compound 1 was replaced with furan 6 (Fig. 2, Table 2). These two compounds have comparable activity   in butyrate medium (7H9-Ty-BT) ( Table 2). Replacements for 1naphthyl in compound 1 were also examined, detecting that 2naphthyl 7 was less potent while 5-benzodioxine 8 or phenyl 9 retained activity Compounds 6-9 showed neither activity under glucose conditions (7H9-Tw-OADC) nor cytotoxicity in Vero cells ( Table 2). Compound 1 (Fig. 1, Table 1) and the analogs 6-9 (Table 2) suffered from very high microsomal metabolism. Although aqueous solubility was low for the most lipophilic compounds 1, 3, 4, 6 and 7, a trend to improve solubility was observed when lowering calculated logP (clogP), as in compounds 2, 5, 8 and 9.
The focus of SAR evaluation became improving solubility and metabolic stability while maintaining potency. With this goal, we explored the SAR around compound 2 (Fig. 1), which had a previously reported MIC of 0.8 ± 0.3 mM in butyrate medium 7 based on the improved solubility with respect to compounds 1, 3-5. The 1,3,4-thiadiazole 10 (Fig. 3, Table 3) was slightly less potent and less soluble than the 1,3,4-oxadiazole (2), so we continued the SAR evaluation using 1,3,4-oxadiazole as the heterocyclic core. Replacement of the furan ring in 2 by thiophene 11 or phenyl 12 led to equipotent compounds with similar solubility. The lack of both substituents at the ortho-positions, as in compounds 13-14, or the presence of a single substituent, as in compound 15, did not change the antimicrobial activity relative to compound 2, suggesting that substitution at ortho-positions is not required for activity. Nevertheless, the lack of one or both substituents at the ortho-positions decreased clogP and significantly increased solubility. Compound 16, having a fluoro group at the 4-position of the benzyl substituent, had similar potency in comparison to compound 12. However, the trifluoromethyl group at 4-position in compounds 17 and 18 substantially decreased the activity with respect to compounds 11 and 12 and negatively impacted on solubility. Replacement of the benzyl group with a cyclohexylmethyl group as in compound 19 was also well tolerated. With respect   to the linker, the N-methyl group in compound 2 appears important for activity. Replacement by the secondary amine 20 or oxygen 21 decreased the activity, while the tertiary amide 22 resulted in complete loss of activity. As expected, the lower lipophilicity of compounds 20-22 compared to compound 2 was beneficial for solubility. Compound 23, with the nitrogen directly attached to the oxadiazole ring, reduced activity and the absence of a basic nitrogen resulted in reduced solubility. As previously seen, there was no or low cytotoxicity against Vero cells for compounds 10-23 (Fig. 3, Table 3). Compound 3 (Fig. 1), which had a previously reported MIC of 2.2 ± 0.4 mM in butyrate medium 7 , is a structurally constrained analogue containing a 1,2,4-oxadiazole core linked to a 3-thiophene instead of the 2-thiophene. Compound 24 (Fig. 4), a hybrid structure between compounds 3 and 5 containing the tetrahydroquinoline piece and the 1,3,4-oxadiazole core, was active (Table 4). A representative subset of tetrahydroquinoline compounds replacing the electron-rich thiophene by phenyl 25 or electron-deficient phenyls 26, 27 was prepared. In this set, phenyl 25 failed to increase potency over 24 but, in contrast, 4-fluorophenyl 26 achieved a small improvement in activity and 4-trifluoromethylphenyl 27 retained potency. Interestingly, the tetrahydroisoquinoline regioisomers 28-31 were much less potent, showing that the site of fusion between the phenyl and the piperidine ring had a significant effect on activity. The clear preference for the tetrahydroquinoline orientation also indicates that the basicity of the nitrogen is not critical for maintaining potency although it influences solubility. 1,4-Benzoxazines 32-35 were less potent than the corresponding tetrahydroquinolines 24-27. The size of the heterocyclic ring also impacted the activity. Indolines 36-37 lowered potency while isoindolines 38-39 lowered it even more, as expected based on the position of the nitrogen atom. When a nitrogen atom is part of the 2-phenylpyrrolidine ring as in 40-41, compounds displayed minimal activity. As expected, compounds 24-41 were not cytotoxic and did not have activity against M. tuberculosis in medium containing glucose (Fig. 4, Table 4).
All the compounds we synthesized suffered from poor microsomal stability in mice (Tables 1-4) and rats (data not shown). One of the first strategies to increase metabolic stability is to reduce the overall lipophilicity. However, a correlation between microsomal metabolism and clogP could not be observed as a few compounds that lowered calculated lipophilicity did not reduce oxidative metabolism (Tables 3 and 4). On the positive side, lipophilicity apparently plays a secondary role in potency because correlation between clogP and activity was not observed (Tables 1-4).
To further investigate the impact of microsomal metabolism on clearance, three of the structurally different screening hits were selected for mouse pharmacokinetic (PK) studies ( Table 5). The PK evaluation of 1, 2 and 3 was performed after intravenous (1 mg/kg) and oral (10 mg/kg) administration to mice. In order to confirm that clearance was due to oxidative metabolism, compound 2 was co-administered with 1-aminobenzotriazole (ABT), a non-selective cytochrome P450 inhibitor. 15 Compound 2 was rapidly eliminated from the body, but clearance decreased significantly when co-administered with ABT, demonstrating that the high clearance was due to CYP-mediated metabolic oxidation. This result is consistent with the very low microsomal stability. Oral exposure was low and increased >10 fold when co-administered with ABT, proving that first pass metabolism limits exposure. Blood  clearance for compound 1 was low but oral exposure was much lower than expected, suggesting solubility-limited drug absorption. Compound 3 had moderate clearance and the very low oral exposure may be due to first-pass metabolism and/or solubilitylimited absorption. We tried several things to improve aqueous solubility for this series. As a general trend, the presence of a basic nitrogen in the linker increased solubility, and solubility was further increased with additional reduction in lipophilicity as seen in Table 3. We also explored how compounds with basic character could benefit from forming salts to optimize solubility and biopharmaceutical properties. As expected, compound 2 as free base had much lower solubility (0.049 mg/mL at pH = 6 and 0.077 mg/mL at pH = 7.4) than the maleate salt (0.687 mg/mL at pH = 6 and 0.717 mg/mL at pH = 7.4). Based on this result, salt formation should be considered the preferred approach to increase aqueous solubility for this structural class containing a basic center.
In summary, we completed the SAR evaluation of a 1,3,4-oxadiazole series of compounds with activity against M. tuberculosis. Compounds had good anti-tubercular activity when tested against bacteria utilizing butyrate as a carbon source, but not with glucose as a carbon source. Although this series will require optimization of molecular properties to improve oral exposure, several 1,3,4-oxadiazoles are valuable tools that will facilitate further study including target identification.

A. Supplementary data
Supplementary data (these data include experimental procedures and analytical data for compounds 8, 10, 21-23, as representative examples, protocols for the phenotypic assays using butyrate  Cl, clearance; Cl,u, Unbound clearance (Total Cl/unbound fraction in plasma); Vd, volume of distribution; t 1/2 , plasma elimination half-life; AUC, area under the curve AUC,u, unbound area under the curve (total AUC x unbound fraction in plasma). See supplementary information for the unbound fraction in plasma. For unbound concentration calculations, blood:plasma ratio was assumed to be 1. a Compound 2 was co-administered with ABT (100 mg/kg, 1 h pre-dose), a non-selective CYP-P 450 inhibitor. and glucose as carbon source, methods for microsomal stability and PK studies, as well as the PK profiles and a table of unbound fraction for compounds 1, 2, and 3) associated with this article can be found in the online version at https://doi.org/10.1016/j. bmcl.2018.04.028.