Crystal structure and function of the isoniazid target of Mycobacterium tuberculosis.

Resistance to isoniazid in Mycobacterium tuberculosis can be mediated by substitution of alanine for serine 94 in the InhA protein, the drug's primary target. InhA was shown to catalyze the beta-nicotinamide adenine dinucleotide (NADH)-specific reduction of 2-trans-enoyl-acyl carrier protein, an essential step in fatty acid elongation. Kinetic analyses suggested that isoniazid resistance is due to a decreased affinity of the mutant protein for NADH. The three-dimensional structures of wild-type and mutant InhA, refined to 2.2 and 2.7 angstroms, respectively, revealed that drug resistance is directly related to a perturbation in the hydrogen-bonding network that stabilizes NADH binding.

inhA, a gene encoding a target for isoniazid and ethionamide in Mycobacterium tuberculosis.

Isoniazid (isonicotinic acid hydrazide, INH) is one of the most widely used antituberculosis drugs, yet its precise target of action on Mycobacterium tuberculosis is unknown. A missense mutation within the mycobacterial inhA gene was shown to confer resistance to both INH and ethionamide (ETH) in M. smegmatis and in M. bovis. The wild-type inhA gene also conferred INH and ETH resistance when transferred on a multicopy plasmid vector to M. smegmatis and M. bovis BCG. The InhA protein shows significant sequence conservation with the Escherichia coli enzyme EnvM, and cell-free assays indicate that it may be involved in mycolic acid biosynthesis. These results suggest that InhA is likely a primary target of action for INH and ETH.

Cell-free synthesis of mycolic acids in Mycobacterium aurum: radioactivity distribution in newly synthesized acids and presence of cell wall in the system.

Distribution of radiolabelling in different parts of the newly synthesized mycolic acids, by a cell-free system from Mycobacterium aurum previously described, is examined, [1-14C]acetate being the precursor. By oxidation cleavage of mycolic acids and examination of the fragments, it was shown that acetate was not uniformly incorporated into the molecule: the methyl terminal part was not labelled, while the central fragments--between unsaturations or between oxygenated functions (oxo or ester) and unsaturations--presented the major part of radioactivity, suggesting the elongation of a preformed compound that the cell-free extract was unable to synthesize. Moreover, the side-chain R2-CH2-COOH was only weakly labelled compared to the central fragments. Since non-hydroxylated fatty acids were not synthesized by the system, it is suggested that de novo C18 fatty acids may be elongated with C2 units by the cell-free extract into C22 fatty derivative, only a low level of labelling being recorded (two C2 units for all the molecule). A scheme is proposed to summarize the main results. Identification of meso-DAP which is a characteristic amino-acid of the peptidoglycan in Actinomycetes and analysis of the profiles of total fatty esters, demonstrated that the cell-free extract is partly constituted by fragments of the cell wall as has already been noticed by examination of micrographs of the extract.

Ethylenic mycolic acid biosynthesis: extension of the biosynthetic model using cell-free extracts of Mycobacterium aurum and Mycobacterium smegmatis.

The hypothetical schemes proposed for the biosynthesis of unsaturated mycolic acids (R1-CH(OH)-CH(R2)-COOH) of Mycobacteria cell walls were experimentally tested by using cell-free extracts either of Mycobacterium aurum or of Mycobacterium smegmatis which produce two kinds of unsaturated mycolic acids (mono and dialkene), [1-14C]acetate being the precursor. Examination of specific radioactivities, in the presence or in the absence of isoniazid, an antituberculous drug inhibiting mycolic acid synthesis, showed that saturated C22 and C24 acids play a role as precursors of two distinct parts of the mycolic acids. Moreover, determination of labelling distribution into mycolic acid fragments obtained by oxidative and pyrolytic cleavages showed first that the side chain R2 and the methyl end R1 both have these C22 and C24 saturated fatty acids as common precursors. Secondly, it is thought that the fragments located between the methyl end R1 and the side chain R2 mainly result from elongation steps (one or two successive additions of seven or eight C2 units according to the mycolic acid type) and a biosynthetic model is proposed for unsaturated mycolic acids extending the published models and illustrating the missing step in monoalkene formation.

Certain properties of isoniazid inhibition of mycolic acid synthesis in cell-free systems of M. aurum and M. avium.

In Mycobacterium tuberculosis isoniazid (INH)-susceptibility and the presence of a thermolabile catalase-peroxidase (T-catalase) are nearly always associated. It is shown in this study that an INH-susceptible strain of M. aurum had a T-catalase activity while its resistant mutants did not, but an in vitro susceptible strain of M. avium had a strong catalase activity without any detectable peroxidase properties. Synthesis of mycolic acids is a genus-specific target for INH and there is an excellent parallelism between INH-susceptibility of intact cells and that of a cell-free system synthesizing mycolic acids. We investigated whether the INH-inhibition of mycolic acid cell-free synthesis was dependent on a T-catalase activity in M. aurum and M. avium: no catalase activity was detectable in any of the cell-free systems tested, and addition of T-catalase from susceptible M. aurum strain to an INH-resistant system did not render it sensitive. So INH can inhibit mycolic acid synthesis independently of the presence of a T-catalase. An INH-susceptible cell-free system prepared from INH-treated (at the MIC) cells was progressively and irreversibly inhibited, while incubation of the same susceptible system in the presence of INH did not result in a significant irreversible inhibition. The possible participation of T-catalase in the irreversible effect of INH is discussed.

Isoniazid inhibition of mycolic acid synthesis by cell extracts of sensitive and resistant strains of Mycobacterium aurum.

Isonicotinic acid hydrazide (isoniazid; INH) inhibition of mycolic acid synthesis was studied by using cell extracts from both INH-sensitive and -resistant strains of Mycobacterium aurum. The cell extract of the INH-sensitive strain was inhibited by INH, while the preparation from the INH-resistant strain was not. This showed that the INH resistance of mycolic acid synthesis was not due to a difference in drug uptake or the level of peroxidase activity (similar in both extracts). As INH did not induce accumulation of any labeled intermediates, it is postulated that the drug acts either on production of labeled chain elongation precursors of mycolic acids or an early step of this elongation. The level of inhibition was not changed by addition of NAD or nicotinamide; thus, INH does not act on mycolic acid synthesis as an NAD antimetabolite. Benzoic or acetic acid hydrazides and known or postulated metabolites of INH (i.e., the corresponding acid, aldehyde, or alcohol) were not inhibitors of cell-free mycolic acid synthesis; the complete structure of INH was required, as already known for inhibition of mycobacterial culture growth. Extracts prepared from INH-treated cells showed reduced mycolic acid synthesis, and the inhibition level was not modified by either extensive dialysis or pyridoxal phosphate. This latter molecule efficiently antagonized INH action by reacting rapidly with INH, as shown by differential spectroscopy. Moreover, pyridoxal phosphate did not release inhibition of INH-treated extracts. It is proposed that INH may covalently react with an essential component of the mycolic acid synthesis system.

Mycolic acid synthesis: a target for ethionamide in mycobacteria?

Striking structural analogies exist between the two specific antimycobacterial drugs ethionamide (ETH) and isoniazid (INH), and they share several inhibitory properties in susceptible species of mycobacteria. The effect of ETH on mycolic acid synthesis was studied in whole cells and in cell extracts of various species, since this synthesis is one direct target for INH, as we recently demonstrated in cell extracts of Mycobacterium aurum. It was shown in the present study that there is not a direct relationship between ETH susceptibility and mycolic acid inhibition. This observation could explain the lack of cross-resistance between the two drugs. The presence of ETH disturbed mycolic acid synthesis in both resistant and susceptible mycobacteria. Synthesis of oxygenated species of mycolic acid was inhibited, while that of diunsaturated acids was either slightly altered or even increased. In contrast, INH inhibited the synthesis of all kinds of mycolic acids in the same way in all susceptible strains and had no effect on mycolic acid synthesis in resistant strains. In the presence of ETH, the unsaturated mycolic acid molecules presented a methyl end different from the usual one. These data strongly suggest that the normal unsaturated mycolic acid species are not the precursors of the oxygenated types. Moreover, they show that ETH probably acts early in the pathway leading to oxygenated mycolic acid.

Ethambutol inhibition of glucose metabolism in mycobacteria: a possible target of the drug.

The addition of D-arabinose, D-galactose, D-glucosamine, or D-mannose to the growth medium of Mycobacterium smegmatis suppressed the inhibitory effects of ethambutol both on acetate labeling of cell wall-linked mycolic acids and on the increase in the delipidated cell dry weight. The addition of D-glucose or D-fructose had no effect. It is proposed that ethambutol inhibits an early step of glucose conversion into the monosaccharides used for the biosynthesis of structurally and biologically important cell wall polysaccharides: arabinogalactan, arabinomannan, and peptidoglycan.

Effects of isoniazid on ultrastructure of Mycobacterium aurum and Mycobacterium tuberculosis and on production of secreted proteins.

Isoniazid (INH), one of the most effective antimycobacterial drugs, specifically inhibits, at an early stage of its action, the biosynthesis of mycolic acids, specific mycobacterial lipids which play a central role in the cell envelope architecture of mycobacteria. In the present study, the consequences of the action of INH on the cell morphology of Mycobacterium tuberculosis and Mycobacterium aurum were examined. Electron microscopy was used to observe bacilli which were previously treated with either subinhibitory concentrations of INH or the MIC of the drug, leading to a decrease of 20 to 35% (by weight) of their mycolic acid contents. The earlier effect of INH on the ultrastructure of mycobacteria, as revealed by negative staining of bacilli, was the alteration of the bacterial poles; this event was observed prior to the bacteriostatic action of the drug and was accompanied by a release of material from the poles into the extracellular medium. In a later stage of the drug's action, cell deformation occurred and more extracellular material was seen. The material released following the action of the drug on susceptible mycobacterial cells was identified as being almost exclusively composed of proteins. Labeling of amino acids with 35S prior to and during the action of INH on M. aurum and subsequent analysis of the labeled proteins led to the conclusion that they consisted of secreted proteins which were up to 20-fold oversecreted in the presence of the drug. Competitive enzyme-linked immunosorbent assay with the secreted 45/47-kDa antigen complex of M. tuberculosis demonstrated up to 20-fold oversecretion of these proteins. Taken together, the production of oversecreted proteins following the decrease of the cell envelope mycolate content by INH strongly suggests that mycolic acids may act as a barrier in the export of proteins secreted by mycobacteria.

Structure of a hydroxymycolic acid potentially involved in the synthesis of oxygenated mycolic acids of the Mycobacterium tuberculosis complex.

Mycolic acids are believed to play a crucial role in the architecture of the mycobacterial envelope. However, very few steps of their biosynthetic pathway have yet been elucidated. We previously isolated [Dubnau, E., Lanéelle, M. A., Soares, S., Bénichou, A., Vaz, T., Promé, D., Promé, J. C., Daffé, M. & Quémard, A. (1997) Mycobacterium bovis BCG genes involved in the biosynthesis of cyclopropyl keto- and hydroxy-mycolic acids, Mol. Microbiol. 23, 313-322] a gene cluster from Mycobacterium bovis BCG, cmaA-D, which confers upon M. smegmatis the ability to synthesize cyclopropanated ketomycolic acid, and a new type of mycolic acid which is hydroxylated. A meticulous analysis of all the mycolic-like fatty acids of M. bovis BCG and M. tuberculosis showed that these organisms produce small amounts of the hydroxymycolic acid. The structure of this molecule, determined by NMR spectroscopy, mass spectrometry and stereochemical studies, strongly suggests that there is a direct biosynthetic relationship between the keto- and the hydroxy-mycolic acids.

Enzymatic characterization of the target for isoniazid in Mycobacterium tuberculosis.

The inhA gene has been recently shown to encode a common protein target for isoniazid and ethionamide action in Mycobacterium tuberculosis. In this paper, we demonstrate that the M. tuberculosis InhA protein catalyzes the NADH-specific reduction of 2-trans-enoyl-ACP, essential for fatty acid elongation. This enzyme preferentially reduces long-chain substrates (12-24 carbons), consistent with its involvement in mycolic acid biosynthesis. Steady-state kinetic studies showed that the two substrates bind to InhA via a sequential kinetic mechanism, with the preferred ordered addition of NADH and the enoyl substrate. The chemical mechanism involves stereospecific hydride transfer of the 4S hydrogen of NADH to the C3 position of the 2-trans-enoyl substrate, followed by protonation at C2 of an enzyme-stabilized enolate intermediate. Kinetic and microcalorimetric analysis demonstrates that the binding of NADH to the S94A mutant InhA, known to confer resistance to both isoniazid and ethionamide, is altered. This difference can account for the isoniazid-resistance phenotype, with the formation of a binary InhA-NADH complex required for drug binding. Isoniazid binding to either the wild-type or S94A mutant InhA could not be detected by titration microcalorimetry, suggesting that this compound is a prodrug, which must be converted to its active form.

Oxygenated mycolic acids are necessary for virulence of Mycobacterium tuberculosis in mice.

Members of the Mycobacterium tuberculosis group synthesize a family of long-chain fatty acids, mycolic acids, which are located in the cell envelope. These include the non-oxygenated alpha-mycolic acid and the oxygenated keto- and methoxymycolic acids. The function in bacterial virulence, if any, of these various types of mycolic acids is unknown. We have constructed a mutant strain of M. tuberculosis with an inactivated hma (cmaA, mma4) gene; this mutant strain no longer synthesizes oxygenated mycolic acids, has profound alterations in its envelope permeability and is attenuated in mice.

Mycobacterium bovis BCG genes involved in the biosynthesis of cyclopropyl keto- and hydroxy-mycolic acids.

The resurgence of tuberculosis and the emergence of multidrug-resistant mycobacteria necessitate the development of new antituberculosis drugs. The biosynthesis of mycolic acids, essential elements of the mycobacterial envelope, is a good target for chemotherapy. Species of the Mycobacterium tuberculosis complex synthesize oxygenated mycolic acids with keto and methoxy functions. In contrast, the fast-growing Mycobacterium smegmatis synthesizes oxygenated mycolic acids with an epoxy function. We describe the isolation and sequencing of a cluster of four genes from Mycobacterium bovis bacillus Calmette-Guerin (BCG), coding for methyl transferases, and which, when transferred into M. smegmatis, allow the synthesis of ketomycolic acid, in addition to an as yet undescribed mycolic acid, hydroxymycolic acid. These oxygenated mycolic acids, unlike the regular mycolic acids of M. smegmatis, and similar to the mycolic acids of M. bovis, are highly cyclopropanated. Furthermore, there is a perfect match between the structures of the keto- and the hydroxy-mycolic acids. We propose a biosynthetic model in which there is a direct relationship between these two types of mycolic acid.