Novel 1,4-substituted-1,2,3-triazoles as anti-tubercular agents.

Abstract

A series of 1,4-substituted-1,2,3-triazoles have been synthesized and evaluated as potential anti-tubercular agents. These compounds were assembled via click chemistry in high crude purity and in moderate to high yield. Of the compounds tested, twelve compounds showed promising anti-tubercular activity with six possessing MIC values <10 μg/mL, and total selectivity for Mycobacterium tuberculosis (Mtb) growth inhibition. A second set of 21 compounds bearing variations on ring C were synthesised and evaluated. This second library gave an additional six compounds displaying MIC values ≤10 μg/mL and total selectivity for Mtb growth inhibition. These compounds serve as an excellent starting point for further development of anti-tubercular therapies.

Graphical Abstract

Tuberculosis (TB) has plagued humanity for thousands of years and has claimed millions of lives over this time. In 2012 approximately 8.6 million people developed TB and 1.3 million died.^[1] TB infection is caused by Mycobacterium tuberculosis (Mtb), which predominantly causes disease in the lungs resulting in cough, fever and weight loss. Despite the preventability of this disease, it is still present in first-world countries although more than 95% of all deaths attributed to TB are from middle-to-low economically developed countries.^[2] Individuals with compromised immune systems, such as people who are HIV-positive, are at a great risk of contracting TB. Indeed, over 30% of people with HIV are co-infected with M. tuberculosis and are approximately 30 times more likely to develop active TB which, if untreated, is almost always fatal in this population. In addition to this, the development of multi-drug resistant tuberculosis (MDR-TB) is becoming more common due to the extensive use of isoniazid and rifampicin, with over 450,000 people contracting MDR-TB in 2012 and the WHO estimates that ~10% of these cases were extensively drug resistant TB (XDR-TB). As such, the development of novel molecular scaffolds targeting M. tuberculosis is an area of increasingly active investigation globally.

Over the past decade, a range of approaches to the development of anti-tubercular compounds has been undertaken, though TB remains a challenge to treat due to its extremely thick hydrophobic cell wall. A useful review regarding anti-tubercular drug candidates is provided here.^[3] One strategy, which is becoming more common is the use of 1,2,3-triazoles incorporated into the molecular scaffold, either as a means to link several molecular portions together or as a key part of the pharmacophore.^[4] The surge of enthusiasm for the inclusion of this moiety within medicinal scaffolds most likely has arisen from several factors including; (i) ease of installation via ‘click chemistry’, (ii) general high yields, (iii) amide isosterism, and (iv) broad functional group tolerance.^{[3a, 5]}

Recent work by Boechat et al.^[6] and Kantevari et al.^{[4a, 4k, 4m]} investigated the use of rapidly accessed 1,2,3-triazoles (Figure 1, 1 and 2, respectively) as novel anti-tubercular agents. These compounds showed good minimum inhibitory concentration (MIC) values (typically < 10 μg/mL) against the H37Rv strain of M. tuberculosis, and the former also showed lack of toxicity against a human liver cell line. In a similar approach Kim and co-workers^[7] synthesized a range of 1,2,3-triazole derived econazole/miconazole derivatives displaying minimum inhibitory concentration (MIC) values as low as 8 μg/mL.

Figure 1.

Previously described triazole containing anti-tubercular compounds 1 and 2 and the focus of this study 3-5

In this study, we sought to determine whether the 1,2,3-triazole moiety can be used as a N-phenylamide isostere in the development of antitubercular agents. In this manuscript, we present the rapid synthesis and evaluation of 1,4-substituted-N-aryl-1,2,3-triazoles with substitution variation on the 1-phenyl and triazole-4-position substituents. While the majority of these compounds do not possess potent anti-tubercular activity, several had promising potency against M. tuberculosis with MIC values <10 μg/mL and form the basis of future development for more potent compounds.

We recently reported the rapid synthesis and evaluation of novel 1,4-substituted N-phenyl-1,2,3-triazoles, of general structure 3, as androgen receptor antagonists.^[8] The primary motivation of that work was to use the 1,4-substituted triazole as an amide isostere for electron deficient N-phenylamides found in common anti-androgenic compounds. A search of the literature showed that similar N-phenyl amides have shown antitubercular activity,^[9] and based on this, we further explored this premise. The aryl azides for click chemistry were previously synthesized based on an optimized set of reaction conditions using typical copper mediated azide-alkyne cycloaddition (CuAAC).^[8] The chosen alkyne reaction partners were the propargyl alcohols 8, which were synthesised in-house via ethynyl Grignard addition to the corresponding aldehyde.

Using copper(II)sulfate/ascorbic acid combination under microwave irradiation to react phenyl azides 7 with propargyl alcohols 8, gave the desired triazoles 6 in good to excellent yield. Presumably the lowest yield (49%, Table 1, Entry 4) was due to poor solubility of the propargylic alcohol in the aqueous system. Nevertheless, all compounds were isolated in good crude purity and high enough yield to allow for biological evaluation. Note that despite the high crude purity all compounds were purified to >95% prior to biological evaluation.

Table 1.

Formations of triazoles 5.

Entry	R1	R2	R3	R4	Product	Yield (%)
1	H	CN	CF3	H	5a	53
2	H	NO2	CF3	H	5b	68
3	H	CN	CF3	2-pyrene	5c	77
4	H	NO2	CF3	2-pyrene	5d	49
5	H	CN	CF3	C6F5	5e	67
6	H	NO2	CF3	C6F5	5f	69
7	H	CN	CF3	3,5-Br	5g	96
8	H	NO2	CF3	3,5-Br	5h	94
9	H	CN	CF3	4-Br	5i	89
10	H	NO2	CF3	4-Br	5j	73
11	H	CN	CF3	4-F	5k	81
12	H	NO2	CF3	4-F	5l	83

Compounds 6a-l as well as 10a-g, 11a-g, 12a and 12b which were on hand from our previously mentioned anti-androgenic study, were submitted for biological evaluation as they are still suitable compounds to investigate under the hypothesis of this study.^[8] This gave a total of 33 compounds for preliminary anti-tubercular evaluation.

It was thought that by pooling the compounds together a greater idea of potential favourable/non-favourable interactions could be gleaned thus informing the development of a second, more focussed library of compounds to be examined (vide infra).

These compounds were evaluated against a panel of bacteria including M. tuberculosis H37Rv (ATCC 27394), Staphylococcus aureus, Bacillus subtilis, Pseudomonas aeruginosa, and Escherichia coli as previously described, using isoniazid as the positive control.^[10] Of the compounds tested, twelve showed some degree of anti-tubercular activity, and showed full inhibition of growth for Mtb only. Note that in some cases the other bacterial strains on the panel showed slowed growth in the presence of high concentrations of these compounds but not inhibition. The structures of the compounds synthesized in this study (e.g. triazoles of general structure 5 or 8) and their corresponding MIC values are shown in figure 3, with compounds with substantial potency (MIC < 10 μg/mL) framed.

Figure 3.

Compounds displaying anti-tubercular activity from series 5

Compounds series 5 gave six compounds with some anti-tubercular activity, two of which, 5d and 5l, possessed MIC values < 10 μg/mL. An interesting trend was observed here, in that the aryl substituent on the alcohol seemed to determine activity, as the aryl triazole portion was consistent – effectively giving 3 ‘sets’ of active compounds. Another trend was observed where the compounds bearing the NO₂ functionality were always more active that the analogous compound with a CN group at the 4-position. The degree of improvement was variable between compounds of each ‘set’, e.g 5j was 1.5 times better than 5i, whereas 5d was 3.7 times more potent than 5c. It is perhaps not surprising that the nitro-aromatic compounds were the most potent compounds since Mtb can reductively activate such molecules.^[10–11]

Of the compounds from series 8 and 9, five were active which are shown below in figure 4. In direct contradiction to the previous set of compounds (Fig 3.), only one nitro-aromatic compound 8b was active against Mtb and it was the weakest of this series by far. Intriguingly, the most potent compounds were structurally very simple, such as 8g, 8f, and 8e posessing MIC values of 2.1, 3.4, and 9.6 μg/mL, respectively. The contrast in structurally active compounds between those in Figure 3 and Figure 4 suggest a different mode of action.

Figure 4.

Compounds displaying anti-tubercular activity from series 8 and 9.

The nitroaromatic compounds have generally been associated with a very high frequency of mutation in Mtb due to the ease of inactivation of the various processes required for nitro-reduction^[10] making compounds 8g, 8f, and 8e particularly attractive for follow-up since these are unlikely pro-drugs and may inhibit a specific target.

From this preliminary study it was seen that compounds consisting of three aromatic rings (e.g. 8a-g) were the better class of potential anti-tubercular compounds. Addtionally, the N-aryl triazole moiety which bore either; 4-NO₂-3-CF₃, 3,5-CF₃, or phenyl group showed the most promise. Therefore, considering that 8b, 8f, and 8g all have unsubsititued aryl ring substituents on ring C (see 13, below), our next objective was to vary this moiety. In doing this we chose the most promising azide units which correspond to 8b, f, and g cross-sectioned with a range of phenyl acetylenes which possess various substitution on ring C.

The synthesis of these compounds proceeded smoothly and in poor to good yield (20–70%). The anilinic triazoles 13d, 13h, and 13l represent an interesting opportunity to derivatise these compounds by acylation of the aniline to give various amides. Amide formation proceeded in high yield under mild conditions for all compounds 14a-i. We chose a simple acetate amide as we thought this unit imparts amide character without additional lipophilicity. Whereas, the octanoyl amide was chosen to impart lipophilicity – as it was thought that this may assist the passage of these compounds through the thick Mtb cell wall, and finally a simple benzoyl amide was chosen.

With compounds 13a-l and 14a-i in hand, our attention focussed on evaluation of these compounds for their antitubercular activity, in addition to selectivity against a representative panel of other bacteria (S. aureus, B. subtilis, P. aeruginosa, and E. coli).

These compounds showed complete selectivity towards M. tuberculosis compared to the other strains investigated. The structural elaborations of N-phenyl triazoles (13a-d) showed that all modifications made to the parent scaffold resulted in a decrease in MIC value relative to the parent compound 8g. This was consistent with the observed activities for 13e-h, whereby all modifications made to the parent scaffold 8f resulted in a small increase of MIC. Interestingly, the 2-fluoro analogue 13e and anilinic derivative 13h were not tolerated at all, with no bactericidal properties observed for either compound. We were pleased to observe that all of our modifications made to parent scaffold 8b, resulted in improved MIC values for all compounds. Indeed, 13j and 13k possess Mtb toxicity three fold more potent than the parent.

To our surprise the addition of amide groups to the phenyl ring of triazoles 13d, 13h, and 13l resulted in vastly decreased MIC values or total abolition of antimicrobial properties. The reason for this is unknown and will be the subject of ongoing studies.

In conclusion, we have synthesised a series of compounds bearing 1,4-substituted triazoles and have evaluated them for their anti-tubercular activity. These compounds were rapidly accessed via a simple 2 step synthetic sequence which is very amenable to library generation. From the initial 35 compounds evaluated in this study, twelve possessed some anti-tubercular activity with half of these displaying MIC values <10 μg/mL and selectivity of Mtb across a panel of gram-positive and negative bacteria. We then synthesised a second library of 21 compounds which possessed structural variation on ring C of the triazole. These analogues demonstrated that an amide unit on ring C is not well tolerated. Nevertheless, we have shown that this simple tri-aromatic molecular scaffold is an excellent starting point for the development of more potent anti-tubercular agents and mode of action studies.

Experimental Section

General procedures for CuAAc reaction

Alkyne (1 mmol), azide (1 mmol), Copper(II) Sulphate (10 mol%) and ascorbic acid (20 mol%) were stirred in water (3 mL) under microwave irradiation at 100 °C for 30 min, to the solution was added dichloromethane (3 mL) and the solution stirred for 3 min. The product was then extracted using dichloromethane (2 × 10 mL) and the organic layer washed with water (10 mL), 4M HCl (5 mL) and water (5 mL). The resulting solution was dried over MgSO₄ and filtered before the final solution was concentrated under vacuum to afford the product which was purified by dissolving the crude product in the minimal amount chloroform (2 mL) and then adding cold pet spirits (15 mL) before filtering under vacuum to afford the product as a solid in >95% purity.

General procedure for the synthesis of propargylic alcohols 7

Ethynyl Grignard (3 eq) was added to a solution of aldehyde (1 mmol) in THF at 0 °C. The solution was then stirred overnight and allowed to come to room temperature. The reaction was quenched with 1M HCl, and the crude material extracted with ethyl acetate. The crude material was purified by flash chromatography to give the title alcohol.

General procedure for the synthesis of amides 14a-i

Anilinic triazole (1 eq) and NEt₃ (3 eq) was dissolved in CH₂Cl₂, the corresponding acid chloride (2.0 eq) was added and the solution stirred overnight at room temperature. The reaction was quenched with the addition of HCl (10 mL, 1M) and the aqueous phase extracted with ethyl acetate (3 × 10 mL). The combined organic layers were washed with brine (1 × 20 mL) followed by drying with MgSO₄. The crude materials were dissolved in CHCl₃ (1 mL/ 100 mg) then precipitated with petroleum spirits (30 mL), filtered under vacuum and the product was obtained as a powder in >95% purity as determined by ¹H NMR.

Antimicrobial susceptibility testing

By visual inspection all compounds were soluble at 100 μM in 0.5% DMSO/water (refer to ESI for details). MIC assays against Mtb were set up as previously described.¹² MIC determinations for Staphylococcus aureus ATCC 13801, Bacillus subtilis, Pseudomonas aeruginosa PA01, and Escherichia coli HB101, were similarly set up using Luria Bertani broth with growth evaluated after overnight incubation at 37°C.^[12] Evaluation was measured in duplicate from two independent experiments. Note that it is not unusual to observe a 100% variation in MIC value.^[13]

Figure 2.

Previously synthesised triazoles incorporated into this study.

Table 2.

Second library of compounds 13 bearing aryl substitution.

Entry	R1	R2	R3	R4	Product	Yield (%)
1	H	H	H	2-F	13a	20
2	H	H	H	4-Cl	13b	40
3	H	H	H	4-OEt	13c	55
4	H	H	H	3-NH2	13d	50
5	CF3	H	CF3	2-F	13e	63
6	CF3	H	CF3	4-Cl	13f	64
7	CF3	H	CF3	4-OEt	13g	52
8	CF3	H	CF3	3-NH2	13h	54
9	H	NO2	CF3	2-F	13i	70
10	H	NO2	CF3	4-Cl	13j	40
11	H	NO2	CF3	4-OEt	13k	67
12	H	NO2	CF3	3-NH2	13l	64

Table 3.

Formation of amides 14

Entry	R1	R2	R3	R4	Product	Yield (%)
1	H	H	H	CH3	14a	86
2	H	H	H	(CH2)6CH3	14b	75
3	H	H	H	Ph	14c	77
4	CF3	H	CF3	CH3	14d	85
5	CF3	H	CF3	(CH2)6CH3	14e	91
6	CF3	H	CF3	Ph	14f	89
7	H	NO2	CF3	CH3	14g	78
8	H	NO2	CF3	(CH2)6CH3	14h	45
9	H	NO2	CF3	Ph	14i	77

Table 4.

MIC values for 13a-l and 14a-i for Mtb.

Compound	MIC (μg/mL)	Compound	MIC (μg/mL)
13a	5 ±2	14a	10 ±5
13b	7 ±3	14b	26 ±13
13c	>100	14c	28 ±14
13d	9 ±4	14d	>100
13e	>100	14e	>100
13f	5 ±2	14f	>100
13g	15 ±7	14g	>100
13h	>100	14h	>100
13i	13 ±6	14i	>100
13j	7 ±3	Isoniazid	0.02 ±0.01
13k	8 ±4
13l	13 ±6