The synthesis and evaluation of triazolopyrimidines as anti-tubercular agents

Graphical abstract


Introduction
Tuberculosis (TB) and its causative agent Mycobacterium tuberculosis present a serious threat to global health. There are over 1 million deaths each year and approximately 9 million new cases. 1 Taken together with the estimate that a third of the world's population is infected with M. tuberculosis and the existence of drugresistant strains of M. tuberculosis, it is apparent there is a pressing need for new therapies. To address these needs, there has been an increased effort directed towards TB drug discovery in recent years and a pipeline of new anti-TB drug candidates has started to emerge. The search for new molecular scaffolds with potentially novel mechanisms of action remains a priority.
Triazolopyrimidines (TZPs) are a well-known scaffold in medicinal chemistry, and their utility is exemplified by the discovery and development of novel agents to fight a wide range of diseases. For example, TZPs possess anticancer activity, 2 and have been used as phosphodiesterase inhibitors for diabetes treatment. 3 Recently, the first natural TZP, essramycin, was isolated and found to possess antibacterial activity. 4 A considerable effort has been made to develop TZPs with antimalarial activity. [5][6][7] Transition metal-containing TZPs have antiproliferative activity against Leishmania and Trypanosoma cruzi, the protozoa that cause leishmaniasis and Chagas disease, respectively. 8 In addition TZP acylsulfonamides with anti-mycobacterial activity target acetohydroxyacid synthase. 9 Similar compounds were also identified in a phenotypic screening campaign against Mycobacterium bovis BCG. 10 We identified a single TZP compound (1) from a whole-cell screen against M. tuberculosis which was active in liquid culture (Fig. 1)

Results and discussion
Our SAR investigation began with the design and synthesis of novel analogs based on modifications of the core structure of compound 1. We set out to explore modifications of the core by way of heteroatom replacement and the impact of chemical diversity at the C2, C5, and C7 positions.

Structure-activity relationship (SAR) studies
We determined the activity of all the compounds synthesized against both M. tuberculosis and eukaryotic cells (HepG2 cell line); we determined the MIC 90 against M. tuberculosis, defined as the concentration required to inhibit 90% growth in liquid medium, and the IC 50 against HepG2 cells, defined as the concentration required to reduce HepG2 viability by 50%. Selectivity index (SI) was calculated as IC 50 /MIC 90 . We completed a systematic evaluation by modification of the TZP ring at the C5, C7, C2 and core positions.
We first explored substitution of C5 phenyl with ortho-or paraelectron-donating groups, which resulted in a loss of activity (Table 1, compounds 7-10). The incorporation of fluorine on the para-or ortho-positions (compounds 13 and 14) of the C5 phenyl ring was tolerated and maintained good separation from cytotoxicity. This was not seen with chloride analogs 11 and 12 which had no activity. The strongly deactivating para-CF 3 phenyl moiety (compound 15) abrogated activity. Replacing the C5 phenyl with a polar 2-pyridyl gave compound 16 with slightly increased potency while having no effect on cytotoxicity, and provided a valuable point for further optimization. It is interesting to note that 3-or 4-pyridyl isomers at C5 (17 and 18) resulted in loss of activity (Table 1).
No advantage was gained by replacing the aromatic group at C5 with either a linear alkyl (20, 21) or a cyclohexyl group (25) and only small cyclic alkyl moieties such as the cyclopropyl (22), cyclopentyl 24 and cyclobutyl (23) analogs had good anti-tubercular activity ( Table 2).
We next explored modifications to the C7 position, while keeping a phenyl group at C5 (Table 3). Non-aromatic moieties at C7 (e.g. compounds 27-35) did not show any anti-tubercular activity, except for the ethyl tethered cyclohexyl analog (36). We next evaluated the effect of chain length at C7 (Table 4). Compounds 37 (with no tether), 38 (with a one carbon tether), and 40 (with a three carbon tether) were all inactive. A reduction in potency was observed for the unsubstituted terminal phenyl group as in compound 39 compared to the para-methoxy analog 1. This pointed to the importance of para-substitution for activity and selectivity. Polar aromatics at the terminal end of the C7 alkyl chain were not tolerated as demonstrated by the inactive pyridyl analogs 41-43. para-substituted analogs 44-48 were prepared, and the overall potency was rescued in all cases except for the para-fluoro analog 46. C7 amide analogs (49 and 50) retained activity and were not cytotoxic indicating that properties of the aniline nitrogen at C7 can be modified with little penalty to activity.
Thus far we demonstrated that optimal anti-tubercular activity was found in the 2-pyridyl analog 16. As with the C5 phenyl derivatives, we examined modifications to the terminal aromatic ring of the C7 side chain while keeping the 2-pyridyl at C5 constant (Table 5). A variety of para-substituted aromatic groups were tolerated at the terminal end of the C7 side chain resulting in analogs (55-60) with comparable or improved potency and good separation from cytotoxicity. It is worthy to note that analogs containing terminal pyridyl moieties on the C7 side chain (52-54) lost activ-ity, as observed in the C5 phenyl series, demonstrating that polar residues are not tolerated in this region for either series.
We investigated other SAR elements of the spacer at C7 ( Table 6). The N-methylated analog 61 had a loss of activity, an indication that H-bonding may be important for the TZP compounds binding to their target (this could also be due to unfavorable steric interactions). Introducing rigidity on the amine tether as in pyrrolidine 62, piperidine 63 and morpholino 64 was not tolerated. It has been previously reported that substitution on the C2 position of the TZP scaffold gave compounds with potent antimalarial activity and with good metabolic stability. 7 Based on these findings and with the goal of identifying active and metabolically stable compounds, we prepared and tested the 2-methyl (67) and 2-phenyl (68) analogs. Analog 67 had comparable activity to our original hit, but analog 68 showed an MIC > 20 lM. Pyrazolopyrimidine 11 and imidazopyridine 12,13 compounds have also been reported to possess potent anti-tubercular activity. Analog 69 with a pyrazolopyrimidine core demonstrated comparable activity to original hit 1, but with increased cytotoxicity, whereas the imidazopyrimidine-based analog 70 had excellent anti-tubercular activity as well as good separation from cytotoxicity.

Microsomal stability and in vivo pharmacokinetic (PK) studies
Based on in silico ADME predictions, three compounds were chosen to cover a range of cLogP values and were evaluated for their in vitro microsomal stability (Table 7). Rapid metabolism of compound 44 was observed, in rodent and human liver microsomes. The para-OCF 3 analog 48 was also rapidly metabolized. The amide (49) had improved in vitro microsomal stability, with only 12% loss after 30 min in mouse microsomes, 22% loss in rat microsomes and 31% loss in human microsomes. The difference in stability for these three compounds is probably due to presence of more than one oxidatively-labile carbon in 44 and 48 (both have a two carbon linker), while 49 has only one such soft spot, although these have not been confirmed with metabolite identification studies.
The oral exposure of the three compounds was evaluated in male mice by comparing exposures following oral (PO) and intravenous (IV) administration of 10 and 1 mg/kg doses of compounds, respectively (Table 7). Overall, the three compounds showed poor to moderate in vivo mouse PK properties. Compound 44 had an AUC = 676 nM⁄h, and rapid IV clearance (182 mL/min/kg) as predicted by mouse microsomes. Compound 48 had more promising mouse PK with greater oral exposure (AUC 2750 nM * h), a longer half-life, and slower IV clearance (33.5 mL/min/kg). Surprisingly, despite showing good in vitro microsomal metabolic stability, amide 49 had the lowest oral exposure and fastest IV clearance rates in vivo compared to the two other compounds.

Activity Spectrum
Three compounds (16, 48 and 49) were selected for testing against other organisms based on good activity in liquid and solid medium (MIC 99 < 20 mM) against M. tuberculosis (Table 8). We tested against a non-pathogenic mycobacterial species (Mycobacterium smegmatis), Escherichia coli (Gram negative), Pseudomonas aeruginosa (Gram negative), Bacillus subtilis (Gram positive) and yeast (Saccharomyces cerevisiae) ( Table 8). All three compounds had activity against M. tuberculosis, but no activity against any of the other species. Thus, the TZP compounds are selective for M. tuberculosis.

Activity against replicating and non-replicating M. tuberculosis
We also determined the effectiveness of the three compounds (16, 48 and 49) in killing replicating and non-replicating M. tuberculosis. Compounds exhibited static activity against replicating bacteria, preventing growth, but no kill was noted over 21 days. In contrast, we did note killing against non-replicating bacteria (starvation conditions) for all three compounds, with a 2-3 log kill over 14-21 days (see Fig. 2).

Conclusion
We conducted a systematic exploration of the triazolopyrimidine scaffold for activity against M. tuberculosis. Overall, the compounds in this series show good activity and selectivity. Our initial explorations suggest that non-substituted aromatic rings at C5, and a two-carbon chain connecting a terminal aromatic at C7 are preferred features for potency against M. tuberculosis and separation from cytotoxicity. The presence of NH at C7 and a lack of substituent at C2 are essential for potency of the molecule. Heteroatom replacement or homologation of the scaffold is well tolerated in the series. We have identified compounds with improved metabolic stability in rodent and human liver microsomes, however, oral exposure and clearance remains an issue for this series. We feel that these issues can be improved through further SAR exploration. Thus, there is substantial promise in developing the TZP series that warrants further investigation as novel tools in our drug arsenal to combat tuberculosis.

Determination of minimum inhibitory concentration (MIC)
MICs were determined against M. tuberculosis H37Rv as described in. 14 Briefly, M. tuberculosis was grown in Middlebrook 7H9 medium containing 10% OADC (oleic acid, albumin, dextrose, catalase) supplement (Becton Dickinson) and 0.05% w/v Tween 80 (7H9-Tw-OADC) under aerobic conditions. Bacterial growth was measured by OD 590 after 5 days of incubation at 37°C. Curves were fitted using the Levenberg-Marquardt algorithm. MIC 90 was defined as the minimum concentration required to inhibit growth of M. tuberculosis by 90%.

Determination of minimum inhibitory concentration on solid medium (MIC)
The serial proportion method was used to determine MIC 99 on solid medium for a variety of representative species. 15 M. tuberculosis H37Rv and Mycobacterium smegmatis mc 2 155 were grown in Middlebrook 7H9 medium + 10% v/v OADC (oleic acid, albumin, dextrose, catalase) supplement (Becton Dickinson) and on Middlebrook 7H10 medium + 10% v/v OADC. Plates were incubated at 37°C for 4 weeks and 3-4 days respectively; Escherichia coli DH5a and Staphylococcus aureus RN4220 were grown on LB agar and incubated at 37°C for 1 and 2 days respectively; Pseudomonas aeruginosa HER1018 (PAO1) was grown on tryptic soy agar and incubated at 37°C for 1 day; Bacillus subtilis Marburg was grown in on nutrient agar and incubated at 28°C for 3-4 days; Saccharomyces cerevisiae Y187 was grown on YPD agar plus 0.003% w/v adenine hemi-sulfate and incubated at 30°C for 3-4 days. The MIC 99 was the lowest concentration of compound, which yielded less than 1% growth relative to no-compound control.

Kill kinetics
To determine kill kinetics, replicating bacteria, a late log phase culture (OD 590 0.6-1.0) of M. tuberculosis was adjusted to an OD 590 of 0.1 in 7H9-Tw-OADC; 50 mL was used to inoculate 5 mL 7H9-Tw-OADC containing compounds. For non-replicating conditions, bacteria were incubated at 37°C for 14 days in PBS + 0.05% w/v Tyloxapol (PBS-Ty) at an OD 590 of 0.1 and then compounds were added (final DMSO concentration of 2%). Cultures were incubated standing at 37°C. Serial dilutions were plated on 7H10-OADC agar to determine CFU/mL.

Ethyl 3-oxo-3-(p-tolyl)propanoate (2d)
Sodium hydride (2.68 g, 112 mmmol) was taken in dry DMF (25 mL) in a 250 mL round bottom flask under N 2 and cooled it down to 0°C. To it was added a solution of 1-(p-tolyl)ethan-1one (5.0 g, 37.3 mmol) in DMF (5 mL). The reaction mixture was stirred at rt for 30 min followed by the addition of diethyl carbonate (18 mL, 149.2 mmol). The reaction mixture was then stirred at rt for 16 h. Ice-cooled water was added dropwise to quench the reaction. It was extracted with EtOAc (3 Â 100 mL). The combined organic layer was dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude was further purified by flash chromatography on silica gel (60-120 mesh) using 10% hexane-EtOAc as eluent to afford 2d as a colourless liquid (6.2 g, 81%). LCMS (ESI) m/z 207.32 [M+H + ]; 84.88% (purity). 4.5.7. Ethyl 3-(4-ethylphenyl)-3-oxopropanoate (2e) Sodium hydride (161 mg, 6.74 mmmol) was taken in dry THF (8 mL) in a 100 mL round bottom flask under N 2 and cooled it down to 0°C. To it was added a solution of 1-(4-ethylphenyl) ethan-1-one (500 mg, 3.37 mmol) in THF (2 mL). The reaction mixture was stirred at rt for 30 min followed by the addition of diethyl   To it was added a solution of 1-(2-fluorophenyl) ethan-1-one (5.0 g, 36.2 mmol) in THF (5 mL). The reaction mixture was stirred at rt for 30 min followed by the addition of diethyl carbonate (17.5 mL, 144.8 mmol). The reaction mixture was then stirred at rt for 12 h. Ice-cooled water was added dropwise to quench the reaction. It was extracted with EtOAc (3 Â 75 mL). The combined organic layer was dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude was triturated with Et 2 O afford 2i as a brown liquid (4.0 g, 52%). MS (ESI) m/z 211.17 [M+H + ].

Ethyl 3-cyclobutyl-3-oxopropanoate (2r)
Sodium hydride (2.44 g, 101.9 mmmol) was taken in dry THF (55 mL) in a 250 mL round bottom flask under N 2 and cooled it down to 0°C. To it was added a solution of 1-cyclobutylethan-1one (5.0 g, 50.9 mmol) in THF (5 mL). The reaction mixture was stirred at rt for 30 min followed by the addition of diethyl carbonate (24.7 mL, 203.6 mmol). The reaction mixture was then stirred at rt for 14 h. Ice-cooled water was added dropwise to quench the reaction. It was extracted with EtOAc (3 Â 200 mL). The combined organic layer was dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude was further purified by flash chromatography on silica gel (60-120 mesh) using 10% hexane-EtOAc as eluent to afford 2r as a colourless semi-solid (2.6 g, 30%). LCMS(ESI) m/z 169.07 [MÀH + ]; 80% (purity).
To it was added 1H-1,2,4-triazol-5-amine 3 (0.4 mL, 62.5 mmol). The reaction mixture was heated at 120°C for 14 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vaccum which affored a white solid (10.5 g, 95%). This was then used in the next step without any further purification. LCMS(ESI) m/z 213.09 [M+H + ]; 59.54% (purity). To it was added 1H-1,2,4-triazol-5-amine (0.4 mL, 62.5 mmol). The reaction mixture was heated at 120°C for 14 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded a white solid (10.5 g, 95%). This was then used in the next step without any further purification. LCMS(ESI) m/z 213.09 [M+H + ]; 59.54% (purity). Ethyl 3-(4-methoxyphenyl)-3-oxopropanoate (1.0 g, 4.5 mmol) was taken in AcOH (5 mL) in a 50 mL round bottom flask under N 2 . To it was added 1H-1,2,4-triazol-5-amine (416 mg, 4.9 mmol). The reaction mixture was heated at 120°C for 12 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4b as a brown solid (510 mg, 47%). This was then used in the next step without any further purification. LCMS(ESI) m/z 243.10 [M+H] + ; 80.54% (purity). (1.5 g, 6.7 mmol) was taken in AcOH (15 mL) in a 100 mL round bottom flask under N 2 . To it was added 1H-1,2,4-triazol-5-amine 2 (681 mg, 8.1 mmol). The reaction mixture was heated at 120°C for 14 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4c as an off-white solid (1.2 g, crude). This was then used in the next step without any further purification. LCMS(ESI) m/z 243.11 [M +H + ]; 95.10% (purity).
To it was added 1H-1,2,4-triazol-5-amine (2.9 g, 34.9 mmol). The reaction mixture was heated at 120°C for 14 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4d as a white solid (1.8 g, 27%). This was then used in the next step without any further purification. LCMS(ESI) m/z 227.22 [M+H + ]; 75.24% (purity).
To it was added 1H-1,2,4-triazol-5-amine (1.04 g, 12.4 mmol). The reaction mixture was heated at 110°C for 16 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4m as a yellow solid (1.2 g, 54%). This was then used in the next step without any further purification. LCMS(ESI) m/z 214.19 [M+H + ]; 56% (purity).
Ethyl 3-oxo-4-phenylbutanoate (1.5 g, 7.2 mmol) was taken in AcOH (8 mL) in a 50 mL round bottom flask under N 2 . To it was added 1H-1,2,4-triazol-5-amine (733 mg, 8.7 mmol). The reaction mixture was heated at 115°C for 16 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4n as a white solid (910 mg, 55%). This was then used in the next step without any further purification.
To it was added 1H-1,2,4-triazol-5-amine (2.9 g, 35.2 mmol). The reaction mixture was heated at 110°C for 14 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4r as a yellow semi-solid (3.8 g, crude). This was then used in the next step without any further purification. LCMS(ESI) m/z 191.30 [M+H + ]; 84.60% (purity).
To it was added 1H-1,2,4-triazol-5-amine (1.08 g, 13.0 mmol). The reaction mixture was heated at 110°C for 12 h. The reaction mixture was then evaporated to dryness using toluene as an azeo-tropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4s as a brown solid (1.1 g, 50%). This was then used in the next step without any further purification. LCMS(ESI) m/z 205.05 [M+H + ]; 66.67% (purity).
To it was added 1H-1,2,4-triazol-5-amine (407 mg, 4.84 mmol). The reaction mixture was heated at 115°C for 16 h. The reaction mixture was then evaporated to dryness using toluene as an azeotropic solvent and triturated with diethyl ether. This was finally dried under high vacuum which afforded 4t as an off-white solid (410 mg, 46%). This was then used in the next step without any further purification. LCMS(ESI) m/z 217.30 [MÀH + ]; 52.65% (purity).