**2. The Mycobacterial Cell Wall**

The biosynthesis of the mycobacterial cell wall components has been extensively described recently [20–22], and only a summary of the most pertinent points for this article is presented here. The proteins implicated or targeted by TB drugs discussed below are underlined.

The three main components of the cell wall are the peptidoglycan (PG), the lipopolysaccharides (arabinogalactan, lipoarabinomannan, and lipomannan) and the outer membrane, which contains mycolic acids, various glycolipids and phthiocerol dimycocerosates (Figure 1).

The *M. tuberculosis* peptidoglycan is located outside of the mycobacterial inner membrane, conferring rigidity, integrity and shape to the cell [22]. The peptidoglycan is a polysaccharide composed of alternating *N*-acetylglucosamine and muramic acid (either *N*-acetylated or *N*-glycolylated) residues linked by β (1→4) bonds [23–26]. Strands of polysaccharides are acylated on the muramic acid residues by the pentapeptide L-alanyl-D-isoglutaminyl-*meso*-diaminopimelyl-D-alanyl-D-alanine synthesized by the Mur ligases (MurC/D/E/F). The acylation reaction is performed by the D-Ala:D-Ala ligase DdlA. The pentapeptides are cross-linked to form the peptidoglycan. Two types of cross-linkages are observed in *M. tuberculosis*: (i) the D,D-transpeptidase activity of the penicillin binding proteins PonA1 and PonA2 cross-links meso-diaminopimelic acid and D-alanine to form a 3→4 linkage; (ii) the

L,D-transpeptidases (LdtMt1 to 5) cross-link two meso-diaminopimelate residues to form a 3→3 linkage [22].

**Figure 1.** Illustration of the mycobacterial cell wall. The outermost layer of the cell wall, the capsule, is omitted. The cell wall components/sites inhibited by the tuberculosis (TB) drugs discussed herein are indicated. Abbreviations: AG, arabinogalactan; LAM, lipoarabinomannan; LM, lipomannan; MA, mycolic acids; PG, peptidoglycan; P, phospholipid; PIMs, phosphatidyl-*myo*-inositol mannosides; INH, isoniazid; ETH, ethionamide; PTH, prothionamide; TAC, thiacetazone; PMD, pretomanid; EMB, ethambutol.

The arabinogalactan is a branched polysaccharide composed of arabinose (Ara*f*) and galactose (Gal*f*) residuesin the furanose configuration [27]. The first stepin the arabinogalactan biosynthesisis the formation of thelinker that anchors the arabinogalactan complex to the peptidoglycan via the*N*-glycolylated-muramic acid residues [28]. This linker is a decaprenyl-diphospho-*N*-acetylglucosamine-rhamnosyl molecule produced by the successive transfer of *N*-acetylglucosamine-1-phosphate to decaprenyl-phosphate by WecA followed by the transfer of L-rhamnose by WbbL. On this linker, 30 linear Gal*f*residues are added by the galactofuranosyl transferases GlfT1 and GlfT2. An Ara*f*unit is then transferred to the galactan chain using the arabinose donor decaprenylphosphoryl-D-arabinose (DPA). DPA is formed through several steps, starting with phospho-α-D-ribosyl-1-pyrophosphate (pRpp). UbiA adds a decaprenyl group to form decaprenol-1-monophosphate 5-phosphoribose (DPPR), which is dephosphorylated and epimerized by DprE1 and DprE2 to form DPA. The arabinofuranosyltransferase AftA catalyzes the transfer of the first unit of Ara*f* to the galactan chain. The arabinosyltransferases EmbA and EmbB catalyze the further addition of Ara*f* to form the arabinan. The final product, the arabinogalactan, is a linear galactan to which highly branched arabinans are attached. The arabinan anchors the mycolic acids forming the mycolyl-arabinogalactan-peptidoglycan (mAGP) complex.

Mycolic acids are the hallmark of the *M. tuberculosis* cell wall, an essential component regulating the permeability, acid-fast staining, viability and virulence of *M. tuberculosis* [29]. As stated above, mycolic acids are found attached to the arabinose part of the arabinogalactan complex but can also be found as a free form or bound to other saccharides to form trehalose mono/dimycolates (TMM/TDM) and glucose monomycolate. Mycolic acids are long-chain fatty acids composed of a meromycolate chain containing up to 62 carbons with various modifications (*cis*/*trans* cyclopropane, keto, or methoxy groups) to which a saturated C26 alkyl chain is attached at the α position (Figure 2). The biosynthesis of mycolic acids starts with two fatty acid synthases (FAS): the eukaryotic-like type I (FAS-I) and the prokaryotic-like type II (FAS-II) (Figure 2). FAS-I is a multidomain polypeptide [30] that synthesizes the α-C26 alkyl chain and also provides a C14/<sup>16</sup> fatty acyl-CoA to be elongated into the meromycolate chain by the FAS-II system. This elongation reaction starts with the condensation of C14/<sup>16</sup> fatty acyl-CoA with malonyl-Acyl Carrier Protein (ACP) by the β-ketoacyl-ACP synthase III FabH. The resulting β-ketoacyl-ACP intermediate is delivered to FAS-II, which performs cycles of elongation using independent enzymes: the reductase MabA, the heterodimer dehydratases HadAB and HadBC, the enoyl-reductase InhA and the condensases KasA and KasB (Figure 2). The modifications (*cis*/*trans* cyclopropane, keto, or methoxy groups) of the meromycolate chains are introduced either during the elongation by FAS-II or when the meromycolate chain is fully formed [31]. The meromycolate chain is activated by the fatty acid AMP synthase FadD32 to a meromycolyl-AMP, which is coupled to the carboxylated α-C26 fatty acyl-CoA (from FAS-I) by the polyketide synthase Pks13. Reduction of the resulting mycolic β-ketoester by the mycolyl reductase CmrA yields a mature mycolic acid. The biosynthesis of mycolic acids takes place in the cytoplasm. Transfer of the mycolic acids to the cell envelope occurs via the formation of a trehalose monomycolate (TMM) which is translocated by the efflux pump MmpL3 (Mycobacterial membrane protein Large). The mycolyltransferase Antigen 85 complex (*fbpA*, *fbpB*, *fbpC*) then condenses mycolic acids to the arabinogalactan releasing a molecule of trehalose.

**Figure 2.** Schematic representation of the biosynthesis of mycolic acids. The three families of mycolic acids in *Mycobacterium tuberculosis* are the α-, methoxy- and keto-mycolic acids. Only the cis configuration of the cyclopropane group of the methoxy- and keto-mycolic acids is shown, but *M. tuberculosis* also produces the trans-methoxy- and trans-keto-mycolic acids. The known targets of the mycolic acid inhibitors in clinical use or in clinical trial are InhA for isoniazid (INH) and ethionamide (ETH); HadAB for thiacetazone (TAC); and Mmpl3 for SQ109.

Besides the mAGP complex, the mycobacterial cell wall also includes lipomannan (LM), lipoarabinomannan (LAM) and phosphatidyl-*myo*-inositol mannosides (PIMs). These lipids are anchored in the inner membrane and play an important role in *M. tuberculosis* growth, survival and virulence [32–34]. In the outer membrane, trehalose-containing glycolipids and phthiocerol dimycoserates are found interlaced with the mycolic acids. These lipids are crucial for their interactions with the host and its immune system [35].

Specific components of the cell wall are essential for *M. tuberculosis* survival in the host. It is, therefore, not surprising that two of the four first-line TB drugs target the cell wall. In the past 20 years, potent new TB drugs have been discovered that target the peptidoglycan, arabinogalactan, LAM, and mycolic acids or their transfer to the cell wall. This article focuses on the cell wall inhibitors included in the TB pharmacopeia since the 1950s/1960s (old TB drugs) and the new ones in clinical development.
