Synthesis and evaluation of triazole congeners of nitro-benzothiazinones potentially active against drug resistant Mycobacterium tuberculosis demonstrating bactericidal efficacy

Abstract

With growing concerns regarding target residue mutation hovering over established anti-TB pharmacophores, it is imperative to have reserve chemotypes at our disposal to curb unrestrained spread of tuberculosis. In this context, we herein present the synthesis and bio-evaluation of a library of new nitrobenzothiazinone (BTZ) congeners comprising 2-mercapto/amino-benzothiazinone tethered 1,2,3-triazole hybrids as antitubercular agents. In preliminary screening, 10 out of 37 compounds displayed substantial in vitro potency against Mtb H37Rv (MIC 0.5–8 μg mL⁻¹). Structural optimization of the initial hit 5o (MIC 0.5 μg mL⁻¹) led to identification of linker variants 9a, 9b, 9c, and 9d exhibiting potent anti-TB activity (MIC 0.03–0.12 μg mL⁻¹). When tested against Vero cells to determine their selectivity index (SI), these compounds displayed no appreciable cytotoxicity (SI >80). Further studies on activity against drug resistant (DR) Mtb indicated these compounds to be equally potent (MIC 0.03–0.25 μg mL⁻¹). The in silico covalent docking study suggested a similar polar interaction to that of PBTZ169 with an additional and contrasting side chain interaction at the active site of Mtb DprE1 target protein. Further, the time kill kinetic study found compounds 9a and 9d to be demonstrating bactericidal efficacy, completely eliminating bacilli in 7 days at 10× MIC. The most promising compound 9d, considering its potent anti-TB activity (MIC 0.06 μg mL⁻¹ against drug susceptible Mtb and MIC 0.06–0.25 μg mL⁻¹ against DR Mtb) along with a broad therapeutic index (SI >640) demonstrating a comparable concentration dependent bactericidal efficacy to that of RIF, holds a significant edge to be translated into a potent anti-Mtb agent.

Lead compound was identified to be a selective inhibitor of Mtb H37Rv with no appreciable cytotoxicity, demonstrating quite comparable bactericidal efficacy to RIF.

Introduction

Health is a fundamental human right. However, infectious disease, more particularly tuberculosis (TB), an air-borne contagious disease caused by Mycobacterium tuberculosis (Mtb), continues to jeopardize public health owing to the startling increase in extensive or untreatable drug resistance (DR) cases. Global control efforts in containing this menacing disease have severely been challenged by DR to existing drugs.¹ In addition, burdensome treatment protocols with currently available drugs, associated side effects,² drug–drug interaction not being conducive for combination therapy,³ escalating incidences of patient non-adherence to treatment regimen,⁴ growing concerns of latent TB,⁵ and prevalence of co-infection⁶ are some of the other factors hampering efficient tackling of TB worldwide.

Despite improvement in healthcare measures and resurgence in drug discovery efforts as evidenced by the approval of bedaquiline, delamanid, pretomanid, etc. in the last decade, TB is still the leading cause of death from a single infectious disease until the coronavirus pandemic. However, cardiac toxicity with these approved drugs along with drug–drug interaction^2,3 begs for urgent acceleration in the development of potent and safer new chemical entities (NCE) effective against TB. Therefore, constant feeding of the clinical drug pipeline is urgently needed to make TB a ‘disease of the past’ as in low-burden countries, to aid in achieving the WHO's end TB strategy of 80% reduction in TB incidence cases by 2030.⁷

Considering the quest for potent and selective anti-TB agents, research on nitrobenzothiazinone (BTZ) based suicidal inhibitors of DprE1 has been gaining traction in recent times due to their impressive activity and unmatched selectivity.⁸ Decaprenylphosphoryl-β-d-ribose 2′-epimerase (DprE1) is a vital flavoprotein in the cell wall of Mtb that plays a significant role in the biosynthetic pathway for synthesis of Mtb cell-wall polysaccharide arabinogalactan. By covalently binding to the Cys-387 residue of DprE1, BTZ based compounds halt cell wall biosynthesis, ultimately leading to growth inhibition of Mtb.⁹ Since the development of the BTZ lead compound BTZ043 by Makarov et al. which exhibited a MIC of 0.001 μg mL⁻¹ against Mtb H37Rv,¹⁰ several modifications have been implemented targeting improvement in the pharmacokinetic (PK) profile without compromising potency (Fig. 1). Replacement of piperidine with the sufficiently lipophilic but more hydrophilic piperazine led to the 2-piperazino-benzothiazinone derivative PBTZ169, also called macozinone (MCZ), which demonstrated comparatively improved safety, potency, and efficacy (MIC 0.00003 μg mL⁻¹).¹¹ Both of these compounds are in phase 2 clinical development due to their exceptional activity and wide therapeutic window. Subsequently, a close analogue of PBTZ169 featuring an oxime moiety,¹² N-4 of piperazine inserted with a sulphonyl group¹³ (SPBTZ) and a carbonyl group,¹⁴ a novel hybrid combining the active anti-TB chemotype imidazo[1,2-a]pyridine¹⁵ with BTZ have been reported to display potent anti-TB activity although faring no better than MCZ. Since the 6-trifluoromethyl-8-nitrobenzothiazinone heterocycle is indispensable to the potency of BTZ, quite a number of alterations have been directed at the more tolerant C-2 position of the BTZ core. Consequently, replacement of the piperazine of PBTZ with various linkers like piperidine,¹⁶ spiro-heterocycles bearing a 2-benzyl-2,7-diazaspiro[3.5]nonane moiety,^17,18 and different spiro and bicyclic motifs¹⁹ has been reported to result in a better PK profile without significantly modulating potency with respect to MCZ. Apart from these classical BTZ containing hetero-alicyclic C-2 modified moieties, 2-amino substituted BTZ I has been found to demonstrate potent inhibition.²⁰ Recently, BTZ bearing a mesyl group at C-6 with C-2 substituted by 5-membered hetero-aryl like triazole and oxadiazole II linked through a methylamino group has been reported to exhibit an anti-TB MIC of 0.014 μg mL⁻¹.²¹ Lipophilicity plays a pivotal role in the inhibitory activity displayed by BTZ as can be inferred from the fact that superior potency is directly proportional to inclination towards the C-2 alicyclic motif. While the anti-TB potency achieved so far with the BTZ chemotype is unparalleled, the need to improve solubility cannot be denied since absorption and permeability play a significant role in eliciting in vivo efficacy.²² In this respect, the triazole moiety has been known to be hydrophilic in nature owing to its dipolar character and H-bond forming ability.²³ Moreover, its synthetic versatility, stable nature, and anchoring property and being a bio-isostere make triazole an attractive scaffold to medicinal chemists.²⁴ An abundant number of 1,2,3-triazole bearing scaffolds have been established in the literature as promising anti-TB agents, some of which are depicted in Fig. 1. Patpi et al. integrated 1,2,3-triazole with dibenzo[b,d]thiophene to form a potent series of anti-TB agents with improved pharmacophoric features; the best compound III displayed a MIC of 0.78 μg mL⁻¹.²⁵ The first line anti-TB drug INH was clubbed with N-phenyl-1,2,3-triazole resulting in the potent hybrid IV exhibiting MIC 0.62 μg mL⁻¹ with SI >1612.²⁶ Targeting the virulence factor is a novel strategy in the anti-TB drug discovery process that disrupts the Mtb survival mechanism and compound V epitomises this phenomenon, acting as an efficient MptpB inhibitor.²⁷ Combination of triazoles with privileged versatile scaffolds like benzothiazole²⁸VI has demonstrated proven anti-TB activity.

Fig. 1. Literature reported representative anti-TB compounds featuring 8-nitrobenzothiazinone and 1,2,3-triazole motifs.

Rationale

Target based drug discovery is the most validated and efficient method of potent molecule identification. However, the probability of a single mutation in the binding site potentially turning the inhibitor ineffective is quite high, which makes these approaches vulnerable to resistance. BTZ chemotypes suffer a similar fate, wherein the target Cys-387 residue is predisposed to mutation.²⁹ Therefore, moving forward, new generation BTZs not only need improvement in PK and physicochemical parameters but also necessitate exploration of possible/additional interactions in the target binding site, so as to curtail emergence of resistance. One way of accomplishing this is through the scaffold hybridization strategy that could possibly inhibit multiple targets in Mtb while retaining the innate anti-TB activity of integrating pharmacophores.³⁰ In this context, this study aspires to combine a physiochemically compatible and flexible triazole pharmacophore with BTZ (Fig. 2) in an aim to find out whether the molecular hybridization could lead to a promising anti-Mtb agent.

Fig. 2. Illustration of the design rationale and figurative representation of the SAR strategy: dotted square boxes indicate part of literature cited anti-TB compounds altered by molecular hybridization involving bio-isosteric replacement of piperazine (PBTZ169) with triazole and scaffold morphing of benzothiazole (VI) to benzothiazinone to generate the designed compounds; solid square boxes refer to the modification type along with the corresponding compound number.

Synthetic planning and SAR strategy

In order to efficiently explore and find out the optimum combination of chemical features, the target molecule was categorized into three segments, as indicated in Fig. 2. Preliminary investigation was focused on identification of a suitable substituent for the RHS (right hand side/segment) for which 37 compounds (5a–ak) were synthesized and bio-evaluated for their anti-Mtb potency. By keeping the substituent with optimum activity on the RHS fragment fixed, variation in the substituent pattern of the core BTZ scaffold or the LHS (left hand side/segment) was carried out to obtain compounds 6a–d. Lastly, modification of the spacer and replacement of the linker in the middle region of the target compound generated 6e–f and 9a–d, respectively.

Results and discussion

Chemistry

Following the click cycloaddition protocol, 37 (5a–ak) analogues followed by 10 (6a–f and 9a–d) systematically modified derivatives as new benzothiazinone-triazole hybrids were synthesized, as depicted in Scheme 1. Synthesis of key intermediates 4a–e was achieved by adopting a previously reported modified literature procedure.³¹ Briefly, nitration of 1a–e using a conc. nitric acid–sulphuric acid mixture under appropriate conditions led to the formation of nitro-substituted 2-chlorobenzoic acid derivatives 2a–e in good to excellent yield. Refluxing 2a–e with thionyl chloride in chloroform under an inert atmosphere generated the corresponding acyl chloride derivatives that were subsequently converted to their respective benzamide analogues 3a–e by treatment with aqueous ammonia under cooling conditions. Intermediates 3a–e were reacted with carbon disulphide in a NaOH/DMSO system followed by coupling with propargyl bromide that led to the formation of key starting materials 4a–e (BTZ-S propargyl).

Scheme 1. Synthetic route to 2-mercapto/aminobenzothiazinone tethered 1,2,3-triazole conjugates.

Meanwhile, a cluster of aryl/heteroaryl azides were synthesized from the corresponding amines following a literature reported procedure³² utilizing sodium nitrite in aqueous HCl to afford diazotization prior to being displaced by sodium azide. Cycloaddition of intermediate 4a with substituted azides in the presence of copper sulphate and sodium ascorbate in a DMF–water solvent system at room temp. generated a primary series of target compounds 5a–ak. Structurally diversified analogues 6a–d were obtained by employing a similar procedure to the above from 4b–e and 1-azido-4-chlorobenzene. Spacer modified analogues 6e and 6f were obtained by a similar cycloaddition procedure from the reaction of 4a with 4-chlorobenzyl azide and 2-azido-N-(4-chlorophenyl) acetamide, respectively. The aforementioned azides were synthesized by a literature reported method³³ from the corresponding bromo-derivatives and sodium azide in a DMF–water mixture using triethylamine as base upon heating. Intermediates 7a–d (BTZ-S Me) were prepared from 3a–d by using similar synthetic procedure to that of 4a–e wherein propargyl bromide was replaced by methyl iodide in the penultimate step. 7a–d, upon coupling with propargyl amine in ethanol under heating conditions, afforded 8a–d (BTZ-NH propargyl). Synthesis of linker modified derivatives 9a–d was achieved by utilizing a similar cycloaddition method to the one described above from 8a–d and 1-azido-4-chlorobenzene.

Anti-TB activity

The anti-Mtb activity of the synthesized compounds was evaluated by resazurin microtiter plate assay (REMA). The concentration of the newly synthesized compounds was tested from 64–0.03 μg mL⁻¹ in a 2-fold serial dilution manner along with isoniazid (INH), rifampicin (RIF), streptomycin (STR), and ethambutol (ETB) as reference standards. Preliminary screening revealed that the synthesized compounds inhibited Mtb H37Rv ATCC 27294 while showing no activity against non-tuberculous mycobacteria (NTM), i.e., M. abscessus ATCC 19977, M. fortuitum ATCC 6841, and M. chelonae ATCC 35752, proving their selectivity towards Mtb. The complete antimycobacterial screening results of 5a–ak are presented in Table S1 in the ESI.† For convenience in discussion, the compounds were categorized as potent (MIC ≤0.5 μg mL⁻¹), highly active (MIC 1–4 μg mL⁻¹), and moderately active (MIC 8–16 μg mL⁻¹) while MIC values of 32–64 and >64 μg mL⁻¹ indicated weak and inactive compounds, respectively. The anti-TB activities of all the compounds are listed in Table 1. Compounds 5a–ak were also screened against a panel of bacteria (Table S2 in the ESI†) to determine their selectivity. The anti-TB screening results in Table 1 indicated that 10 out of the 37 evaluated compounds exhibited significant activity against Mtb H37Rv (MIC 0.5–8 μg mL⁻¹). The screening results against the bacterial pathogen panel demonstrated poor activity (MIC 32 ≥ 64 μg mL⁻¹) against S. aureus while showing no inhibition on other bacteria, indicating their specificity towards Mtb.

MIC (μg mL⁻¹) of 2-mercaptobenzothiazinone linked 1,2,3-triazole analogues (5a–ak) against Mtb H37Rv ATCC 27294.

Compound code	Structure R	MIC (μg mL−1) Mtb H37Rv ATCC 27294	Compound code	Structure R	MIC (μg mL−1) Mtb H37Rv ATCC 27294
5a	Phenyl	32	5v	4-Hydroxyphenyl	64
5b	2-Fluorophenyl	16	5w	4-Carboxyphenyl	>64
5c	2-Chlorophenyl	32	5x	4-Aminosulfonylphenyl	>64
5d	3-Fluorophenyl	8	5y	4-(1-Piperidinyl) phenyl	64
5e	3-Chlorophenyl	16	5z	2,4-Dimethylphenyl	32
5f	3-Trifluoromethylphenyl	8	5aa	3,4-Difluorophenyl	8
5g	m-Tolyl	8	5ab	3-Chloro-4-fluorophenyl	8
5h	3-Nitrophenyl	32	5ac	3,4-Dimethoxyphenyl	32
5i	3-Cyanophenyl	64	5ad	3,5-Dimethoxyphenyl	>64
5j	3-Acetylphenyl	32	5ae	3-Fluoro-4-(1-piperidinyl) phenyl	64
5k	3-Isopropoxyphenyl	64	5af	3-Fluoro-4-morpholinophenyl	64
5l	3-Carboxyphenyl	>64	5ag	3-Hydroxy-4-carboxyphenyl	>64
5m	3-Aminosulfonylphenyl	>64	5ah	3,4,5-Trimethoxyphenyl	64
5n	4-Fluorophenyl	8	5ai	Naphthalen-2-yl	64
5o	4-Chlorophenyl	0.5	5aj	Pyridin-3-yl	32
5p	4-Bromophenyl	4	5ak	2-Oxo-2H-chromen-6-yl	32
5q	p-Tolyl	8	Std.	Isoniazid	0.03
5r	4-Methoxyphenyl	32	Std.	Rifampicin	0.06
5s	4-Hexyloxyphenyl	8	Std.	Streptomycin	1
5t	4-Phenoxyphenyl	32	Std.	Ethambutol	1
5u	4-Nitrophenyl	16	—

A detailed overview of the anti-TB screening results revealed that most of the halogen substituted compounds exhibited better anti-TB activity. More particularly, the compound with 4-chloro substitution 5o was found to be the potent compound of the series (MIC 0.5 μg mL⁻¹) followed by the 4-bromo derivative 5p (MIC 4 μg mL⁻¹). On the other hand, the mono-fluoro substituted analogues (5b, 5d, and 5n) displayed moderate potency along with the trifluoromethyl derivative 5f. Similarly, m- and p-tolyl (5g and 5q) bearing analogues exhibited moderate potency (MIC 8 μg mL⁻¹). Coincidentally, the di-halogen substituted derivatives (5aa and 5ab) also displayed moderate potency. In contrast, compounds with –COOH (5l and 5w), –SO₂NH₂ (5m and 5x), and –OH (5v) substitution were found to show weak anti-TB activity. These facts revealed that non-polar electron withdrawing groups (EWGs) are optimal for potency while polar EWGs are detrimental for anti-Mtb activity. The electron donating methyl group tends to increase potency. However, methoxy 5r and phenoxy 5t derivatives displayed poor activity while the n-hexyloxy derivative 5s exhibited moderate potency, further indicating that alkyl appendages increase anti-TB potency. Likewise, dimethyl 5z and di-methoxy 5ac analogues displayed poor activity. Nitro derivatives (5h and 5u) were found to exhibit weak to moderate potency. Surprisingly, compounds bearing piperidino and morpholino moieties along with fluoro-substitution (5ae and 5af) were found inclined towards diminished anti-Mtb activity indicating that bulky substituents are not favoured. The weak activity exhibited by trimethoxy 5ah and naphthalene 5ai analogues confirmed the same. Heterocyclics (5aj and 5ak) were found to be not tolerated, displaying poor potency. In general, non-polar substituents were found to be optimal for anti-TB potency while polar EWG and bulky substituents tended to abolish anti-TB activity. All these key findings are summarized in Fig. 3. The initial hit compound 5o was subjected to modification (hit optimization) in search of a more potent compound.

Fig. 3. Overview of SAR: highlighted colours of rectangular boxes indicate identical colours of part of the structure being referred to.

Structural optimization of initial hit 5o

Variation at the LHS part of compound 5o was carried out in order to identify the most suitable substitution as well as its impact on anti-TB potency. This was achieved by altering –Cl at C-6 with –F, –Br, and –CF₃ to obtain compounds 6a–c, respectively. Similarly, the position of the nitro group was changed from C-8 to C-6 (6d), in an aim to ascertain its role. Furthermore, spacing between the triazole moiety and 4-chlorophenyl was introduced to study its effect (6e and 6f) on anti-Mtb activity. Replacing the ‘thio’ linker with ‘amino’ was carried out to understand whether the presence of a H-bond donor could improve the anti-TB potency (9a–d). The summary of hit optimization is presented in Fig. 4. All these compounds were screened against a panel of bacteria and were found to be inactive (Table S2 in the ESI†). Antimycobacterial screening revealed these compounds to be selective towards Mtb while displaying no inhibition against NTMs (Table S1 in the ESI†).

Fig. 4. Illustration of the structural modifications of the most potent compound 5o and their outcome: analogues obtained from the variation at C-6 (6a–c), the analogue generated through alteration of the NO₂ position (6d), spacer inserted derivatives (6e–f), and linker modified variants (9a–d).

Preliminary inspection of anti-TB activity revealed that compound 6b exhibited 2-fold reduction in potency compared with 5o while compounds 6a and 6c displayed a 4-fold decrease in potency, indicating –Cl at C-6 to be optimal for anti-TB activity. Surprisingly, this result is in contrast to previously reported BTZ based anti-TB agents wherein more EWGs like –CF₃ and –NO₂ were most active at C-6. Compound 6d was found to be inactive, confirming the essential nature of –NO₂ at C-8, which is in agreement with a previous literature report.¹¹ The increase in spacing between the triazole and 4-chlorophenyl moieties completely abolished the anti-TB activity (6e and 6f). Interestingly, compounds 9a–d exhibited a remarkable increment in anti-TB activity compared to 5o. Coincidentally, compounds with –Cl and –Br substitution (9a and 9c) were found to display 2-fold better potency than compounds with –F and –CF₃ substitution (9b and 9d), the result of which is in agreement with ‘thio’ linked analogues (5o and 6a–c).

Cytotoxicity assay against Vero cells

MTT assay was employed for cytotoxicity evaluation of potent compounds (5o and 9a–d) having MIC ≤0.5 μg mL⁻¹. Doxorubicin was taken as positive control. The lowest concentration of the compound causing 50% reduction in cell viability is considered as CC₅₀ and the selectivity index (SI) was then determined as the concentration of the compound that inhibits 50% of mammalian Vero cell growth (CC₅₀)/MIC of the corresponding compound. As presented in Table 2, the ‘thio’ linked derivative 5o was found to be nontoxic to Vero cells with SI >100. Amongst ‘amino’ linked analogues 9a–d, compound 9c demonstrated a lower therapeutic index (CC₅₀ >2.5 μg mL⁻¹), while the rest of the compounds were found to be devoid of any significant toxicity. Compound 9d was revealed to be the most selective (CC₅₀ >40 μg mL⁻¹ & SI >640) among the compounds tested. To sum up, the –CF₃ (9d) substituent conferred better selectivity followed by –Cl (9a), while the –Br (9c) analogue displayed a relatively narrow therapeutic window.

Cytotoxicity assay and MIC against clinical isolates of DR-Mtb strains of selected compounds.

Compound code	MIC (μg mL−1)					Cytotoxicity
	Mtb H37Rv ATCC 27294	Mtb INH-resa ATCC 35822	Mtb ETB-resa ATCC 35837	Mtb STR-resa ATCC 35820	Mtb RIF-resa ATCC 35838	Vero cell CC50 (μg mL−1)	SIb
5o	0.5	2	2	1	2	>50	>100
9a	0.03	0.25	0.03	0.03	0.03	>10	>320
9b	0.12	0.25	0.06	0.25	0.06	>20	>160
9c	0.03	0.12	0.03	0.03	0.03	>2.5	>80
9d	0.06	0.25	0.06	0.06	0.06	>40	>640
INH	0.03	>64	0.03	0.03	0.03	n.d.	n.d.
RIF	0.06	0.03	0.03	0.03	64	n.d.	n.d.
STR	1	1	1	>64	1	n.d.	n.d.
ETB	2	4	32	2	2	n.d.	n.d.

^aINH-res: isoniazid-resistant; ETB-res: ethambutol-resistant; STR-res: streptomycin-resistant; RIF-res: rifampicin-resistant.

^bSelectivity index = CC₅₀/MIC; n.d.: not determined.

Determination of activity against clinical isolates of DR-Mtb strains

All compounds subjected to the cytotoxicity assay and cleared were further studied for their activity against clinical strains of DR-Mtb. The results obtained are tabulated in Table 2. Compound 5o (thio-linked derivative) was found to inhibit DR strains at a relatively higher concentration compared to susceptible Mtb H37Rv (MIC 1–2 μg mL⁻¹). In contrast, amino-linked analogues 9a–d displayed potent activity (MIC 0.03–0.25 μg mL⁻¹). Amongst them, 9c fared better, followed by 9a and 9d. Notably, all these compounds demonstrated potent inhibitory activity against all DR strains (MIC 0.03–0.25 μg mL⁻¹) barring INH-res strain (MIC ≥0.12 μg mL⁻¹), wherein they were found to be comparatively less potent. Taken together, compound 9d, considering its potent anti-TB activity against both susceptible as well as DR strains and with an impressive SI, was found to be the most active compound of the series.

Molecular docking studies

Since the evaluated compounds are BTZ congeners with identical pharmacophoric features, molecular docking was carried out on DprE1 in order to understand the plausible mode of binding and to get insight into the observed SAR. Based on the concept of covalent binding previously reported, nitroso metabolic forms of 5o and 9a–d were subjected to docking at the active site of the DprE1 enzyme (PDB ID: 4NCR). The docking results along with major interactions obtained for all these compounds as well as for the co-crystal ligand PBTZ-169 are depicted in Table S3 in the ESI.† Predictable anchoring of the nitroso group to Cys-387 through covalent bonding at the active site of DprE1 was observed for all the compounds tested. However, subsequent formation of a H-bond by a semi-mercaptal adduct to Gln-336 as found with the co-crystal ligand was only exhibited by 9b and 9d. This is indicative of the superior selectivity displayed by these compounds over other amino-linked derivatives. The major interactions observed for the representative compound 9d are illustrated in Fig. 5. The OH group of the semi-mercaptal adduct acts as a H-bond donor in the interaction with the Gln-336 residue (H-bond distance 1.88 Å). Interestingly, an additional H-bond interaction involving the –OH group of the adduct as the H-bond acceptor (H-bond distance 2.18 Å) and the Lys-418 residue was observed specifically in compound 9d, possibly explaining its higher selectivity over other evaluated compounds. Although both the compounds tested and PBTZ-169 displayed similar interactions through the BTZ core, a major deviation was observed involving the tail part of the corresponding scaffolds. In contrast to the cyclohexylmethyl-piperazine fragment of the co-crystal exhibiting hydrophobic interaction with Tyr-314, Arg-325, Ala-326, Tyr-327, and Gly-328 residue, the (4-chlorophenyl)-1H-1,2,3-triazole motif of compound 9d was found involved in non-polar interaction with Tyr-111, Phe-199, Ser-228, Ala-229, Ala-244, Ser-246, and Phe-313 residue (Fig. 6).

Fig. 5. Docking pose of the active form of compound 9d (brown coloured stick) in the active site of DprE1: covalently bonded residue Cys-387, residue involved in H-bonding (Lys-134, Gln-336, and Lys-418), and residue involved in π–π interaction. Trp-230 is shown as cyan sticks; the yellow coloured dashes represent H-bonds while FAD-502 is in the background as green coloured thin stick; the H-bond distances are indicated in Å.

Fig. 6. Overlay of the active form of compound 9d (brown coloured stick) and PBTZ169 (cyan coloured stick): the green and magenta coloured labels represent residues involved in the hydrophobic interaction with compound 9d and PBTZ169, respectively, the residues interacting through H-bonding are shown as blue coloured labels while the red colour label indicates the covalently bonded residue Cys-387. The π–π interaction displaying residue Trp-230 is represented as a purple label.

Apart from displaying hydrophobic interaction with a significantly different residue than the co-crystal, the triazole moiety of the compounds tested exhibited additional H-bonding with Lys-134. Also, the terminal phenyl moiety was found involved in π–π stalking interaction with Trp-230, conferring stability to the molecule at the active site of DprE1. These observations suggested that replacement of the aliphatic moiety in BTZ based anti-TB scaffolds with suitable pharmacophores is a feasible option, as proposed in the rationale of design that allows exploring additional binding sites without drastically reducing the anti-TB potency.

Time-kill kinetics study

Potent non-toxic compounds of the series displaying inhibition of both DS and DR-Mtb strains were subjected to time kill analysis to determine the rate at which they reduce the bacterial load. The time-kill kinetics of RIF along with the test compounds 9a and 9d are shown in Fig. 7(A and B), respectively. As expected, RIF demonstrated a concentration-dependent killing activity, rapidly reducing Mtb log₁₀ CFU mL⁻¹ in 6–7 days in both 1x and 10× MIC. By comparison, all the compounds tested at 10× MIC completely eliminated the bacilli within 7 days while at 1× MIC, 9a eliminated bacilli much more efficiently than 9d. Taken together, compound 9a was identified to be the most potent compound of the series with the best time-kill kinetics which is quite comparable to RIF.

Fig. 7. (A and B). Time kill kinetics graph of compounds 9a (A) and 9d (B).

Conclusions

In summary, we successfully accomplished the design and synthesis of new BTZ congeners featuring 2-mercapto and 2-aminobenzothiazinone linked 1,2,3-triazole derivatives. Preliminary anti-TB evaluation of 2-mercapto linked analogues (5a–ak) found the initial hit 5o displaying MIC 0.5 μg mL⁻¹ against Mtb H37Rv. Subsequent optimization of hit 5o led to the identification of 2-amino linked derivatives 9a–d exhibiting potent anti-TB activity (MIC 0.03–0.12 μg mL⁻¹). Further, the cytotoxicity study suggested these compounds to display a high SI. Moreover, the activity against DR clinical isolates revealed potent inhibition, notably for 9a–d (MIC 0.03–0.25 μg mL⁻¹). The SAR study indicated that non-polar EWGs at the triazole functionality in general and halogens in particular are optimal for potency while polar, heteroaryl, bulky substituents are detrimental for anti-TB activity. The lipophilic methyl group tended to increase the anti-TB activity while 2-amino linkers were found to be significantly superior to the mercapto linker. Although –Cl and –Br substituents at C-6 in the BTZ core exhibited 2–4-fold superior potency compared to –F and –CF₃ groups, the latter were found to exhibit better selectivity. The docking studies confirmed the same and indicated the feasibility of replacement of alicyclic motifs in the BTZ core with hetero-aryl pharmacophores with the purpose of exploring an alternate mechanism and target. The time kill kinetics study revealed bactericidal efficacy with compounds 9a and 9d eliminating the bacilli load within 7 days at 10× MIC. 9d was identified to be the lead with a SI >640 and a potency of MIC 0.06 μg mL⁻¹ against susceptible Mtb and MIC 0.06–0.25 μg mL⁻¹ against DR-Mtb strains. With the threat of mutation lingering over the target residue causing resistance to classical BTZ analogues, combination of suitable anti-TB pharmacophores with the BTZ core could possibly lead to eliciting additional interactions. Moreover, this is deemed to be a plausible strategy in circumventing solubility issues currently grappling BTZ analogues. In this aspect, the present study attempts to provide a possible tactic in averting inadequacies with BTZ derivatives.

Author contributions

Santosh Kumar Sahoo: conceptualization, investigation, methodology, software, visualization, writing – original draft. Mohammad Naiyaz Ahmad: investigation, methodology. Grace Kaul: investigation, methodology. Srinivas Nanduri: supervision, resources. Arunava Dasgupta: supervision, resources, writing – review & editing. Sidharth Chopra: supervision, writing – review & editing. Venkata Madhavi Yaddanapudi: project administrator, supervision, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.