3.1. Cell Wall
The cell wall of Mtb is the primary host-pathogen interaction spot, and a major determinant of the bacillus durability and robustness. The complex and dynamic structure of the cell wall (
Figure 1) is essential for maintaining cellular integrity, enabling adaptation of the bacilli to the host conditions, and plays a crucial role in long-term infection and virulence. It comprises three essential substructures: a peptidoglycan (PG) inner layer, a mycolic acid (MA) outer layer, and an arabinogalactan polysaccharide (AGP) middle layer. The inhibition of key enzymes responsible for the biosynthesis of these substructures are excellent targets for novel drug development due to the absence of homologous characteristics in the host [
37].
Peptidoglycan layer. The peptidoglycan is composed of
N-acetylglucosamine, GlcNAc, and
N-acetylmuramic acid, MurNAc, that are cross-linked with short peptides. The biosynthesis of peptidoglycan is a complex sequence of reactions, starting with the synthesis of Lipid II, in which a hydrophobic polyisoprene tail embedded in the membrane is connected to a monomer of cell-wall peptidoglycan through a pyrophosphate linker. This step is followed by translocation of Lipid II bound to the membrane formation, Lipid II polymerization and cross-linking by penicillin-binding proteins, PBPs, including L,D-transpeptidases [
38]. Lipid II is targeted by the antibiotics ramoplanin and teixobactin, inhibiting the transglycosylation process and affecting peptidoglycan formation.
Figure 1.
Composition of the cell wall of Mycobacterium tuberculosis.
Figure 1.
Composition of the cell wall of Mycobacterium tuberculosis.
Mtb also produces
β-lactamase, an enzyme that catalyses the hydrolysis of
β-lactam antibiotics, a reason why the use of these antibiotics is not included in TB treatment. However, carbapenems (
Figure 2A) are resistant to inactivation by
β-lactamases, and thus are included in the treatment of multidrug-resistant TB, since they target the biosynthesis of peptidoglycan by inhibiting L,D-transpeptidases.
Mycolic acid layer. Mycolic acids are very long-chain (C60-90)
α-alkyl
β-hydroxy fatty acids that contribute to the hydrophobic, impermeable, and rigid structure of the outer membrane [
39,
40]. Mycolic acids are synthesised from acetyl-CoA, by at least two elongation systems, the type I and type II fatty acid synthases, FAS-I and FAS-II. The FAS-II system can only be found in bacteria, turning this system a potential selective antibacterial target. The NADH-dependent enzyme 2-
trans-enoyl-acyl carrier protein reductase, InhA, is involved in the FAS-II system and is targeted by isoniazid (
Figure 2B), a first-line agent used for treatment of TB. Other small molecules including ethionamide (
Figure 2B), which is structurally related to isoniazid, and triclosan, have their mode of action related with InhA inhibition.
Subsequent cycles of fatty acid elongation are carried out by
β-ketoacyl synthase KasA, which completes chain elongation via condensation of FAS-I-derived acyl-CoAs with malonyl-ACP (acyl carrier protein). KasA is the only essential member of three
β-ketoacyl synthases encoded in the Mtb genome [
41], and has been reported as a validated target for the treatment of TB [
42]. A structure-based approach was used to optimize existing KasA inhibitor DG167 [
43,
44] to afford indazole JSF-3285 (
Figure 2B) with a 30-fold increase in mouse plasma exposure. Biochemical, genetic, and X-ray studies further confirmed that JSF-3285 targets KasA.
Mycolic acids are transported to the outer membrane by bacterial membrane proteins called Mycobacterial membrane protein large (MmpL), which are part of the RND (Resistance, Nodulation and Cell Division) family. The primary role of RND proteins is to translocate a broad range of compounds across the plasma membrane to the periplasmic space, including virulence-associated envelope lipids and siderophores. Mtb genome encodes 13 MmpL proteins, of which MmpL3 has been reported in the biosynthesis of the mycobacterial outer membrane. The ethylenediamine derivative SQ109 (
Figure 2B) is a MmpL3 inhibitor and has completed phase IIb-III clinical trials. SQ109 also accumulates in the lungs, the site of infection, increasing the drug efficacy [
6]. Other promising MmpL3 inhibitors include indolocarboxamides and adamantylureas. As part of a drug scaffold repurposing program, the cannabinoid receptor modulator rimonabant (
Figure 2B) and its diaryl pyrazole analogs, was reported to display potent anti-TB activity [
45,
46].
Figure 2.
Selected compounds that target the cell envelope of Mycobacterium tuberculosis. (A) carbapenems as inhibitors of L,D-transpeptidases; (B) inhibitors of the biosynthesis of mycolic acids; (C) inhibitors of the biosynthesis of arabinogalactan.
Figure 2.
Selected compounds that target the cell envelope of Mycobacterium tuberculosis. (A) carbapenems as inhibitors of L,D-transpeptidases; (B) inhibitors of the biosynthesis of mycolic acids; (C) inhibitors of the biosynthesis of arabinogalactan.
Arabinogalactan polysaccharide layer. The branched-chain arabinogalactan (AG) is the major cell wall polysaccharide, representing
ca 35% of the cell wall, being composed by arabinose and galactose residues, both in the furanose configuration. This middle layer is covalently attached to peptidoglycan and mycolic acid layers requiring several enzymes that are potential targets for the design of novel inhibitors to block the formation of arabinogalactan polysaccharide [
47,
48,
49,
50]. An example are the enzymes arabinosyltransferases (EmbA, EmbB, and EmbC), which are a known targets for the drug ethambutol [
51]. Another target to block the arabinogalactan polysaccharide formation is the enzyme arabinofuranosyltransferase (Aft), responsible for the polymerization of arabinofuranyl residues in decaprenylphosphoryl-D-arabinose (DPA), the lipid donor of D-arabinofuranosyl residues of AG. The DPA synthetic pathway is a potential drug target, and several arabinosyltransferases are essential in the growth of Mtb, as AftA, AftB, AftC and AftD. While AftA and AftB are responsible for the transference of the arabinofuranosyl residue, AftC and AftD introduce the α-1,3-branching in the segments of α-1,5-linked D-Araf residues [
47,
52].
Attachment of the arabinogalactan to the peptidoglycan structure is performed via an essential linker, the disaccharide L-rhamnose-D-
N-acetylglucosamine. The enzyme
N-acetylglucosamine-1-phosphate transferase, GlcNAc-1-P transferase (WecA) catalyses the first step of this linker biosynthesis. For this reason, WecA inhibitors such CPZEN-45 (
Figure 2C), a caprazamycin derivative, prevents the growth of Mtb [
53,
54].
The enzymes decaprenylphosphoryl-
β-D-ribose 2’-oxidase (DprE1) and decaprenylphosphoryl-D-2-keto erythropentose reductase (DprE2) are involved in the two-step epimerization of decaprenylphosphoryl-β-D-ribofuranose (DPR) into decaprenylphosphoryl-β-D-arabinofuranose [
55]. Diverse chemical scaffolds as azaindoles, aminoquinolones, benzothiazinones, benzothiazoles, dinitrobenzamides, nitrobenzamides, pyrazolopyridines, quinoxalines, triazoles, and thiadiazoles demonstrated DprE1 inhibition. The benzothiazinone derivatives BTZ-043 and PBTZ169 (
Figure 2C) are currently in phase II clinical trials and demonstrated high efficacy against
M. tuberculosis. Additionally, the non-covalent inhibitors, azaindole TBA-7371 and OPC-167832, currently in phase II and phases I/II clinical trials, respectively, have shown promising results [
56].
3.2. Energy Metabolism
Mtb operates its energetic metabolism in a modular and compartmentalized mode to support distinct and key cellular functions [
2,
19].
Electron Transport Chain. Mtb relies on oxidative phosphorylation (OxPhos) via the electron transport chain (ETC) to produce energy for growth and division. During the OxPhos process, electrons are transferred from electron donors produced in the central metabolic pathways to molecular oxygen through the ETC. The energy released in this process is conserved by proton-pumping transmembrane proteins that establish a proton gradient and thus generate an electrochemical gradient, called proton motive force (PMF). This bioenergetic pathway generates ATP from the phosphorylation of ADP [
57,
58].
The Mtb ETC is a highly conserved collection of membrane-bound and membrane-associated enzymes and co-factors. It is comprised by five main primary dehydrogenases, which fuel the ETC as electron donors; two main terminal oxidoreductases, which catalyse the transfer of electrons to terminal electron acceptor; and an ATPsynthase, that produces ATP through the dissipation of the PMF. A schematic representation of the Mtb ETC is presented in
Figure 3 [
59,
60,
61].
The respiratory flexibility of Mtb, that allows the bacilli to vary the ETC enzyme composition in response to environmental conditions, and the existence of human homologs to most ETC enzymes, had hampered the development of selective inhibitors. However, the discovery of bedaquiline, an ATPsynthase inhibitor, leads to an increase of research focused on targeting OxPhos. Currently, more than 30% of all new antimycobacterial drugs in clinical trials target the OxPhos, and more than 65% of Phase III trial regimens include a OxPhos inhibitor [
59].
Cytochrome bc1-aa3. The Mtb Cyt bc
1-aa
3 supercomplex is comprised by two tightly associated protein complexes: a menaquinol-cytochrome c oxidoreductase, or cyt bc
1, and an aa
3 oxidase, or cyt aa
3. This supercomplex acts as the primary terminal oxidase under normoxia and during exponential growth and its inhibition results in growth arrest. However, cyt bc
1-aa
3 is not essential for cell survival and as long as the alternate cyt bd is expressed, bc
1-aa
3 inhibitors do not induce bactericidal effects. The central role of the bc
1-aa
3 complex in the ETC and the significant differences to the mammalian counterpart make the supercomplex a good therapeutical target [
58,
59,
60,
61,
62,
63,
64,
65,
66].
Imidazopyridine derivatives are examples of inhibitors that have shown to be particularly potent, the most prominent example being Q203, which is currently in phase II clinical trials and is capable of inhibiting multidrug resistant (MDR) and extensively drug resistant (XDR) Mtb strains [
59,
67,
68,
69]. Structurally similar to Q203, TB-47, is currently in pre-clinical studies and is active against drug sensitive (DS) and drug resistant (DR) Mtb strains, both active and latent bacilli [
70,
71]. Lansoprazole, a gastric proton-pump inhibitor, was found to be a potent hit compound in a screen of FDA-approved drugs. Lansoprazole acts as prodrug and is converted in vivo into lansoprazole sulphide, that was identified to be a cyt bc
1-aa
3 inhibitor on a distinct site from the one targeted by imidazopyridines (
Figure 4A) [
61,
66,
72].
Cytochrome bd. Cytochrome bd-type menaquinol (MKH
2) oxidase, or cyt bd, is a non-proton pumping, less energetically efficient terminal oxidase that transfer electrons from MKH
2 to molecular oxygen. Cyt bd is exclusive to prokaryotic ETC and unlike cyt bc
1-aa
3, the enzyme is more versatile, with multiple functions reported. The terminal oxidase is capable of detoxifying ROS and antibacterials, and protects the bacilli against hypoxia, and is capable of compensating the inactivation of cyt bc
1-aa
3. Cyt bd may also play a role into the Mtb’s natural drug tolerance, namely to drugs that directly target the ETC [
59,
60,
73,
74]. Thus, this cytochrome contributes to Mtb virulence, and since the enzyme is not encoded in animal genomes, it can serve as an attractive promising therapeutical target for new, selective anti-TB drugs [
67,
72,
75].
Inhibition of cyt bd alone does not have any antimycobacterial effects. However, cyt bd inhibitors have synergetic effects with isoniazid, quicken the bactericidal activity of ATPsynthase inhibitors, and turn bc
1-aa
3 inhibitors bactericidal [
59,
60,
61]. Thus, cyt bd inhibitors appear to be particularly attractive in combination therapy, namely in combination with cyt bc
1-aa
3 inhibitors, as the simultaneous inhibition of both terminal oxidases is highly bactericidal in a short period of time, and successful at killing both active and latent bacilli. The non-essentiality of cyt bd represents a challenge to identify its inhibitors, and thus not many cyt bd inhibitors are known. To this date, only few were identified and only one, aurachin D (
Figure 4B), is characterized
. Aurachins are isoprenoid quinoline alkaloids, originally extracted from myxobacteria. Further development of aurachin D is complicated by its toxicity and lack of selectivity, but optimized derivatives of aurachin D have great potential as anti-TB drugs [
67,
72,
74,
75,
76].
Delamanid (DLM) and pretomanid (PTM) (
Figure 4B) are two structurally related nitroimidazoles recently approved for the treatment of MDR-TB that were found to inhibit the biosynthesis of mycolic acid. However, the observance that these drugs were bactericidal against both active and latent bacilli suggested an alternative mechanism of action, as mycolic acid biosynthesis is downregulated in latency. Both DLM and PTM are pro-drugs that require activation by an F420 nitroreductase, an enzyme which produces des-nitro metabolites, with the release of NO. The putative additional mechanism of action is that the intracellular release of NO poisons the cytochrome oxidases, resulting in respiration arrest, and consequent cell death. DLM and PTM treatment results on a quick decrease in intracellular ATP levels, an increased menaquinol/menaquinone (MKH
2/MK) ratio, and upregulation of cyt bd and nitrate reductase, which further support the terminal oxidases as a target of these nitroimidazoles [
59,
60,
69].
Figure 4.
Selected compounds that target the energy metabolism of Mycobacterium tuberculosis. (A) inhibitors of cytochrome bc1-aa3; (B) inhibitors of cytochrome bd; (C) inhibitors of ATP synthase.
Figure 4.
Selected compounds that target the energy metabolism of Mycobacterium tuberculosis. (A) inhibitors of cytochrome bc1-aa3; (B) inhibitors of cytochrome bd; (C) inhibitors of ATP synthase.
ATPsynthase. Bedaquiline (BDQ,
Figure 4C), approved by the FDA in 2012, was the first approved drug specifically for TB in more than 40 years, is an inhibitor of the ATPsynthase. BDQ is a potent bactericidal, efficacious against MDR and latent bacilli and currently, is conditionally administrated for the MDR-TB treatment [
59,
68,
77]. The activity is proposed to work through binding to two subunits on the ATPsynthase, thus inhibiting ATP synthesis and leading to a depletion of intracellular ATP levels. Additionally, bedaquiline is capable of acting as a protonophore leading to the uncoupling of the ETC via collapse of the PMF. Inhibition by BDQ depletes intracellular ATP levels, activates respiration, and induces a metabolic remodelling that upregulates ATPsynthase, NDH-2 and cyt bd. Interestingly, the bacterial activity of BDQ is delayed,
i.e., does not occur immediately upon the ATP depletion, explained by the metabolic remodelling Mtb experiences upon BDQ exposure [
57,
59,
67,
77].
The toxicity associated with the drug and the emergence of bedaquiline-resistant Mtb strains restrain its use to MDR- and XDR-TB patients. Thus, to address its shortcomings, a medicinal chemistry approach was conducted to study the chemical space of diarylquinolines to find the next generation equivalents with superior safety profiles. In this context, two 3,5-diakoxy-4-pyridyl derivatives, TBAJ-587 and TBAJ-876 (
Figure 4C), showed to be particularly interesting and are currently in Phase I clinical trials [
58,
61,
68,
77].
A number of recent studies have identified new ATPsynthase inhibitors with novel mechanisms of action. A family of squaramide derivatives showed to be particularly interesting, with its lead compound (
Figure 4C) currently being evaluated on pre-clinical trials. These compounds target ATPsynthase through a different binding site, meaning they do not show cross-resistance to BDQ, and have shown to be active against BDQ-resistant strains [
57,
59,
67].
Other targets on the ETC. The PMF consists of an electrical potential due to charge separation across the membrane and a chemical potential of protons. The generation and maintenance of a PMF is essential for the Mtb energy production and consequent bacterial growth and survival, in every metabolic state. PMF uncouplers generally act as protonophores and uncouple OxPhos from the ETC, thus inhibit ATP synthesis, leading to cell death [
78]. Generally, this kind of compounds are not sufficiently selective to be used as antimycobacterial agents and thus, the development of specific PMF uncouplers remains an area of interest. However, there are some examples of anti-TB drugs in clinical use that act as PMF uncouplers in addition to an alternative mode of action, such as bedaquiline (
Figure 4C), pyrazinamide (PZA), nitazoxanide (NTZ), (
Figure 5) and SQ109 (
Figure 2B) [
57,
78].
Although the mechanism of action of PZA is not still fully understood, current knowledge is that it acts as a multitarget drug that dissipates the PMF, inhibits ATP synthesis, inhibits membrane transport, and reduces the activity of other proteins, such as aspartate decarboxylase, a protein involved in the Coenzyme A biosynthetic pathway. Evidence of PZA’s uncoupling activity first arise with its ability to target latent bacilli. Additionally, PZA showed to synergize with other PMF uncouplers to deplete ATP depletion and enhance mycobacterial killing, implying its anti-TB activity relies substantially on its uncoupling activity [
57,
61].
Figure 5.
Proton motive force uncouplers currently in clinical settings.
Figure 5.
Proton motive force uncouplers currently in clinical settings.
Initially, SQ109 was reported to interfere with the assembly of mycolic acids in the mycobacterial cell wall through inhibition of membrane transporter MmpL3, but recently, it was demonstrated that SQ109 interferes with respiration, through the ability to act as protonophore and dissipate the PMF [
59,
68].
Nitazoxanide is an FDA-approved repurposed drug with broad-spectrum antiparasitic and antiviral activity. NTZ is proposed to promote Mtb killing by enhancing autophagy through the inhibition of the human mTORC1, and additionally disrupt the PMF, through activity as a protonophore. NTZ inhibits potently both active and latent Mtb bacilli but has poor pharmacokinetic and pharmacodynamic proprieties. Thus, there is some interest in the development of NTZ derivatives with improved bioavailability [
59,
79,
80].