Structural elucidation of the predominant motifs of the major cell wall arabinogalactan antigens from the borderline species Tsukamurella paurometabolum and Mycobacterium fallax

Tsukamurella paurometabolum and Mycobacterium fallax are members of the suprageneric actinomycete group Corynebacterineae that possesses a cell wall skeleton composed of a peptidoglycan to which an arabinogalactan is covalently attached. This polysaccharide is further modified by esterification with C60-C80 mycolic acid residues in mycobacteria and T. paurometabolum. However, M. fallax and T. paurometabolum produce polyenoic (up to six double bonds) mycolic acids whereas the most common type of mycobacterial mycolates, called alpha-mycolates, are mono- and di-enoic or -cyclopropanated mycolic acids. To determine whether this difference also applied to the structures of cell wall arabinogalactans, competitive inhibition experiments using antibodies raised against the cell wall from Mycobacterium bovis and the arabinogalactans from T. paurometabolum and M. fallax were performed. They demonstrated the structural identity between the polysaccharide of M. fallax and those of mycobacteria and showed a strong similarity between the latter polysaccharides and that of T. paurometabolum. Structural analyses of the per-O-alkylated alditol fragments derived from the polysaccharides by gas chromatography-mass spectrometry (GC-MS) and 13C nuclear magnetic resonance (NMR) spectroscopy of the intact solubilized polysaccharides demonstrated that the polysaccharides from the two species analyzed contained all the major structural features previously characterized in mycobacterial arabinogalactans. These include (1) the homogalactan of alterning 5-linked galactofuranosyl (Galf) and 6-linked Galf residues, (2) a linear 5-linked arabino furanosyl (Araf), (3) a beta-Araf-(1-->2)-alpha-Araf disaccharide branched on both position 3 and position 5 of an alpha-Araf unit, and (4) a 5-linked-alpha-Araf unit branched on both position 3 and position 5 of an alpha-Araf residue. The polysaccharide from T. paurometabolum possesses additional structural domains composed of a terminal (t) Araf directly linked to either a 5-linked-alpha-Araf or to both position 3 and position 5 of a 3,5-linked alpha-Araf unit. Both the remarkable similarity of arabinogalactans from Corynebacterineae and their genus- and/or species-specificities are reflected in their 13C NMR spectra that may be used as a valuable help in the identification of members of the actinomycete group.     

Identification of a novel arabinofuranosyltransferase AftB involved in a terminal step of cell wall arabinan biosynthesis in Corynebacterianeae, such as Corynebacterium glutamicum and Mycobacterium tuberculosis

Arabinofuranosyltransferase enzymes, such as EmbA, EmbB, and AftA, play pivotal roles in the biosynthesis of arabinogalactan, and the anti-tuberculosis agent ethambutol (EMB) targets arabinogalactan biosynthesis through inhibition of Mt-EmbA and Mt-EmbB. Herein, we describe the identification and characterization of a novel arabinofuranosyltransferase, now termed AftB (Rv3805c), which is essential in Mycobacterium tuberculosis. Deletion of its orthologue NCgl2780 in the closely related species Corynebacterium glutamicum resulted in a viable mutant. Analysis of the cell wall-associated lipids from the deletion mutant revealed a decreased abundance of cell wall-bound mycolic acids, consistent with a partial loss of mycolylation sites. Subsequent glycosyl linkage analysis of arabinogalactan also revealed the complete absence of terminal beta(1 --> 2)-linked arabinofuranosyl residues. The deletion mutant biochemical phenotype was fully complemented by either Mt-AftB or Cg-AftB, but not with muteins of Mt-AftB, where the two adjacent aspartic acid residues, which have been suggested to be involved in glycosyltransferase activity, were replaced by alanine. In addition, the use of C. glutamicum and C. glutamicumDeltaaftB in an in vitro assay utilizing the sugar donor beta-D-arabinofuranosyl-1-monophosphoryl-decaprenol together with the neoglycolipid acceptor alpha-D-Araf-(1 --> 5)-alpha-D-Araf-O-C(8) as a substrate confirmed AftB as a terminal beta(1 --> 2) arabinofuranosyltransferase, which was also insensitive to EMB. Altogether, these studies have shed further light on the complexities of Corynebacterianeae cell wall biosynthesis, and Mt-AftB represents a potential new drug target.     

Arabinan-deficient mutants of Corynebacterium glutamicum and the consequent flux in decaprenylmonophosphoryl-D-arabinose metabolism

The arabinogalactan (AG) of Corynebacterianeae is a critical macromolecule that tethers mycolic acids to peptidoglycan, thus forming a highly impermeable cell wall matrix termed the mycolyl-arabinogalactan peptidoglycan complex (mAGP). The front line anti-tuberculosis drug, ethambutol (Emb), targets the Mycobacterium tuberculosis and Corynebacterium glutamicum arabinofuranosyltransferase Mt-EmbA, Mt-EmbB and Cg-Emb enzymes, respectively, which are responsible for the biosynthesis of the arabinan domain of AG. The substrate utilized by these important glycosyltransferases, decaprenylmonophosphoryl-D-arabinose (DPA), is synthesized via a decaprenylphosphoryl-5-phosphoribose (DPPR) synthase (UbiA), which catalyzes the transfer of 5-phospho-ribofuranose-pyrophosphate (pRpp) to decaprenol phosphate to form DPPR. Glycosyl compositional analysis of cell walls extracted from a C. glutamicum::ubiA mutant revealed a galactan core consisting of alternating beta(1-->5)-Galf and beta(1-->6)-Galf residues, completely devoid of arabinan and a concomitant loss of cell-wall-bound mycolic acids. In addition, in vitro assays demonstrated a complete loss of arabinofuranosyltransferase activity and DPA biosynthesis in the C. glutamicum::ubiA mutant when supplemented with p[14C]Rpp, the precursor of DPA. Interestingly, in vitro arabinofuranosyltransferase activity was restored in the C. glutamicum::ubiA mutant when supplemented with exogenous DP[14C]A substrate, and C. glutamicum strains deficient in ubiA, emb, and aftA all exhibited different levels of DPA biosynthesis.     

BC10, a DUF266-containing and Golgi-located type II membrane protein, is required for cell-wall biosynthesis in rice (Oryza sativa L.)

Glycosyltransferases (GTs) are one of the largest enzyme groups required for the synthesis of complex wall polysaccharides and glycoproteins in plants. However, due to the limited number of related mutants that have observable phenotypes, the biological function(s) of most GTs in cell-wall biosynthesis and assembly have remained elusive. We report here the isolation and in-depth characterization of a brittle rice mutant, brittle culm 10 ( bc10 ). bc10 plants show pleiotropic phenotypes, including brittleness of the plant body and retarded growth. The BC10 gene was cloned through a map-based approach, and encodes a Golgi-located type II membrane protein that contains a domain designated as 'domain of unknown function 266' (DUF266) and represents a multiple gene family in rice. BC10 has low sequence similarity with the domain to a core 2 β-1,6- N- acetylglucosaminyltransferase (C2GnT), and its in vitro enzymatic activity suggests that it functions as a glycosyltransferase. Monosaccharide analysis of total and fractioned wall residues revealed that bc10 showed impaired cellulose biosynthesis. Immunolocalization and isolation of arabinogalactan proteins (AGPs) in the wild-type and bc10 showed that the level of AGPs in the mutant is significantly affected. BC10 is mainly expressed in the developing sclerenchyma and vascular bundle cells, and its deficiency causes a reduction in the levels of cellulose and AGPs, leading to inferior mechanical properties.     

Corynebacterium glutamicum as a model bacterium for the bioremediation of arsenic.

Arsenic is an extremely toxic metalloid that, when present in high concentrations, severely threatens the biota and human health. Arsenic contamination of soil, water, and air is a global growing environmental problem due to leaching from geological formations, the burning of fossil fuels, wastes generated by the gold mining industry present in uncontrolled landfills, and improper agriculture or medical uses. Unlike organic contaminants, which are degraded into harmless chemical species, metals and metalloids cannot be destroyed, but they can be immobilized or transformed into less toxic forms. The ubiquity of arsenic in the environment has led to the evolution in microbes of arsenic defense mechanisms. The most common of these mechanisms is based on the presence of the arsenic resistance operon (ars), which codes for: (i) a regulatory protein, ArsR; (ii) an arsenite permease, ArsB; and (iii) an enzyme involved in arsenate reduction, ArsC. Corynebacterium glutamicum, which is used for the industrial production of amino acids and nucleotides, is one of the most arsenic-resistant microorganisms described to date (up to 12 mM arsenite and >400 mM arseniate). Analysis of the C. glutamicum genome revealed the presence of two complete ars operons (ars1 and ars2) comprising the typical three-gene structure arsRBC, with an extra arsC1 located downstream from arsC1 (ars1 operon), and two orphan genes (arsB3 and arsC4). The involvement of both ars operons in arsenic resistance in C. glutamicum was confirmed by disruption and amplification of those genes. The strains obtained were resistant to up to 60 mM arsenite, one of the highest levels of bacterial resistance to arsenite so far described. Using tools for the genetic manipulation of C. glutamicum that were developed in our laboratory, we are attempting to obtain C. glutamicum mutant strains able to remove arsenic from contaminated water.     

Heterologous expression of the Mycobacterium tuberculosis gene encoding antigen 85A in Corynebacterium glutamicum.

By using appropriate Corynebacterium glutamicum-Escherichia coli shuttle plasmids, the gene encoding the fibronectin-binding protein 85A (85A) from Mycobacterium tuberculosis was expressed in C. glutamicum, also an actinomycete and nonsporulating gram-positive rod bacterium, which is widely used in industrial amino acid production. The 85A gene was weakly expressed in C. glutamicum under the control of the ptac promoter from E. coli, but it was produced efficiently under the control of the promoter of the cspB gene encoding PS2, one of the two major secreted proteins from C. glutamicum. The 85A protein was produced in various forms, with or without its own signal sequence and with or without the signal sequence and the NH2-terminal (18-amino-acid) mature sequence of PS2. Western blot analysis with monoclonal antibodies raised against the M. tuberculosis antigen 85 complex showed that recombinant 85A protein was present in the corynebacterial cell wall extract and also released in extracellular culture medium. NH2-terminal microsequencing of recombinant 85A secreted by C. glutamicum showed that signal peptide was effectively cleaved off at the predicted site. The recombinant 85A protein was biologically active in vitro, inducing significant secretion of Th1 T-cell cytokines, particularly interleukin-2 and gamma interferon, in spleen cell cultures from mice vaccinated with live Mycobacterium bovis BCG. Heterologous expression of mycobacterial antigens in C. glutamicum now offers a potent tool for further immunological characterization and large scale preparation of these recombinant proteins.     

Regulation of nitrogen metabolism in Mycobacterium tuberculosis: a comparison with mechanisms in Corynebacterium glutamicum and Streptomyces coelicolor

The mechanisms governing the regulation of nitrogen metabolism in Corynebacterium glutamicum and Streptomyces coelicolor have been extensively studied. These Actinomycetales are closely related to the Mycobacterium genus and may therefore serve as a models to elucidate the cascade of nitrogen signalling in other mycobacteria. Some factors involved in nitrogen metabolism in Mycobacterium tuberculosis have been described, including glutamine synthetase and its adenylyltransferase, but not much data concerning the other components involved in the signalling cascade is available. In this review a comparative study of factors involved in nitrogen metabolism in C. glutamicum and S. coelicolor is made to identify similarities with M. tuberculosis on both a genomic and proteomic level. This may provide insight into a potential global mechanism of nitrogen control in Mycobacterium tuberculosis.     

Mycobacterium thermoresistibile as a source of thermostable orthologs of Mycobacterium tuberculosis proteins

The genus Mycobacterium comprises major human pathogens such as the causative agent of tuberculosis, Mycobacterium tuberculosis (Mtb), and many environmental species. Tuberculosis claims ~1.5 million lives every year, and drug resistant strains of Mtb are rapidly emerging. To aid the development of new tuberculosis drugs, major efforts are currently under way to determine crystal structures of Mtb drug targets and proteins involved in pathogenicity. However, a major obstacle to obtaining crystal structures is the generation of well-diffracting crystals. Proteins from thermophiles can have better crystallization and diffraction properties than proteins from mesophiles, but their sequences and structures are often divergent. Here, we establish a thermophilic mycobacterial model organism, Mycobacterium thermoresistibile (Mth), for the study of Mtb proteins. Mth tolerates higher temperatures than Mtb or other environmental mycobacteria such as M. smegmatis. Mth proteins are on average more soluble than Mtb proteins, and comparison of the crystal structures of two pairs of orthologous proteins reveals nearly identical folds, indicating that Mth structures provide good surrogates for Mtb structures. This study introduces a thermophile as a source of protein for the study of a closely related human pathogen and marks a new approach to solving challenging mycobacterial protein structures.     

The Corynebacterium glutamicum gene pmt encoding a glycosyltransferase related to eukaryotic protein-O-mannosyltransferases is essential for glycosylation of the resuscitation promoting factor (Rpf2) and other secreted proteins

Two-dimensional gel electrophoresis and immunoassays revealed several proteins of the secretory subproteome of Corynebacterium glutamicum to be glycosylated. By genome-wide searches for genes involved in glycosylation, the C. glutamicum gene cg1014 was found to exhibit significant similarity to eukaryotic protein-O-mannosyltransferases (PMTs) and to a recently identified orthologue of Mycobacterium tuberculosis, Rv1002c, which is responsible for protein-O-mannosylation. The putative membrane protein Cg1014 showed the same predicted transmembrane topology as Saccharomyces cerevisiae PMT1 and M. tuberculosis Rv1002c along with conserved amino acid residues responsible for catalytic activity. Deletion of the C. glutamicum pmt gene (cg1014) caused a complete loss of glycosylation of secreted proteins including the resuscitation promoting factor 2 (Rpf2), which is involved in intercellular communication and growth stimulation of C. glutamicum. Because the gene pmt as well as rpf genes are present in the genomes of all actinobacteria sequenced so far, this work provides new insights into bacterial protein glycosylation and new opportunities to elucidate the molecular mechanisms of Rpf activity in pathogenic growth and infection.     

NrdH-redoxin of Mycobacterium tuberculosis and Corynebacterium glutamicum Dimerizes at High Protein Concentration and Exclusively Receives Electrons from Thioredoxin Reductase

NrdH-redoxins are small reductases with a high amino acid sequence similarity with glutaredoxins and mycoredoxins but with a thioredoxin-like activity. They function as the electron donor for class Ib ribonucleotide reductases, which convert ribonucleotides into deoxyribonucleotides. We solved the x-ray structure of oxidized NrdH-redoxin from Corynebacterium glutamicum (Cg) at 1.5 Å resolution. Based on this monomeric structure, we built a homology model of NrdH-redoxin from Mycobacterium tuberculosis (Mt). Both NrdH-redoxins have a typical thioredoxin fold with the active site CXXC motif located at the N terminus of the first α-helix. With size exclusion chromatography and small angle x-ray scattering, we show that Mt_NrdH-redoxin is a monomer in solution that has the tendency to form a non-swapped dimer at high protein concentration. Further, Cg_NrdH-redoxin and Mt_NrdH-redoxin catalytically reduce a disulfide with a specificity constant 1.9 × 10⁶ and 5.6 × 10⁶ m⁻¹ min⁻¹, respectively. They use a thiol-disulfide exchange mechanism with an N-terminal cysteine pK_a lower than 6.5 for nucleophilic attack, whereas the pK_a of the C-terminal cysteine is ∼10. They exclusively receive electrons from thioredoxin reductase (TrxR) and not from mycothiol, the low molecular weight thiol of actinomycetes. This specificity is shown in the structural model of the complex between NrdH-redoxin and TrxR, where the two surface-exposed phenylalanines of TrxR perfectly fit into the conserved hydrophobic pocket of the NrdH-redoxin. Moreover, nrdh gene deletion and disruption experiments seem to indicate that NrdH-redoxin is essential in C. glutamicum.     

Substitution of the N-glycan function in glycosyltransferases by specific amino acids: ST3Gal-V as a model enzyme

The sialyltranferase ST3Gal-V transfers a sialic acid to lactosylceramide. We investigated the role of each of the N-glycans modifying mouse ST3Gal-V (mST3Gal-V) by measuring the in vitro enzyme activity of Chinese hamster ovary (CHO) cells transfected with ST3Gal-V cDNA or its mutants. By examining mutants of mST3Gal-V, in which each asparagine was replaced with glutamine (N180Q, N224Q, N334Q), we determined that all three sites are N-glycosylated and that each N-glycan is required for enzyme activity. Despite their importance, N-glycosylation sites in ST3Gal-V are not conserved among species. Therefore, we considered whether the function in the activity that is performed in mST3Gal-V by the N-glycan could be substituted for by specific amino acid residues selected from the ST3Gal-V of other species or from related sialyltransferases (ST3Gal-I, -II, -III, and -IV), placed at or near the glycosylation sites. To this end, we constructed a series of interspecies mutants for mST3Gal-V, specifically, mST3Gal-V-H177D-N180S (medaka or tetraodon type), mST3Gal-V-N224K (human type), and mST3Gal-V-T336Q (zebrafish type). The ST3Gal-V activity of these mutants was quite similar to that of the wild-type enzyme. Thus, we have demonstrated here that the N-glycans on mST3Gal-V are required for activity but can be substituted for specific amino acid residues placed at or near the glycosylation sites. We named this method SUNGA (substitution of N-glycan functions in glycosyltransferases by specific amino acids). Furthermore, we verified that the ST3Gal-V mutant created using the SUNGA method maintains its high activity when expressed in Escherichia coli thereby establishing the usefulness of the SUNGA method in exploring the function of N-glycans in vivo.     

Genetic and biochemical analysis of the aspartokinase from Corynebacterium glutamicum

The lysC/asd gene cluster of Corynebacterium glutamicum ATCC 13032 was cloned and sequenced. The lysC locus coding for aspartokinase consists of two in-frame overlapping genes, lysC alpha encoding a protein of 421 amino acids (Mr 44,300) and lysC beta encoding a protein of 172 amino acids (Mr 18,600). The C. glutamicum aspartokinase was purified and found to contain two proteins of Mr 47,000 and Mr 18,000. A C. glutamicum mutant expressing a feedback-resistant aspartokinase was shown to be changed in a single base pair of the lysC beta gene, leading to an amino acid exchange in the beta-subunit of the aspartokinase. In addition, the identified mutation was found to be responsible for the enhanced expression of the asd gene located downstream of lysC.     

Nucleotide sequence and fine structural analysis of the Corynebacterium glutamicum hom-thrB operon

The complete nucleotide sequence of the Corynebacterium glutamicum hom-thrB operon has been determined and the structural genes and promoter region mapped. A polypeptide of Mr 46,136 is encoded by hom and a polypeptide of Mr 32,618 is encoded by thrB. Both predicted protein sequences show amino acid sequence homology to their counterparts in Escherichia coli and Bacillus subtilis. The promoter region has been mapped by S1-nuclease and deletion analysis. Located between -88, RNA start site and -219 (smallest deletion clone with complete activity) are sequence elements similar to those found in E. coli and B. subtilis promoters. Although there are no obvious attenuator-like structures in the 5'-untranslated region, there is a dyad-symmetry element, which may act as an operator.     

Computational structural analysis of proteins of Mycobacterium tuberculosis and a resource for identifying off-targets

Advancement in technology has helped to solve structures of several proteins including M. tuberculosis (MTB) proteins. Identifying similarity between protein structures could not only yield valuable clues to their function, but can also be employed for motif finding, protein docking and off-target identification. The current study has undertaken analysis of structures of all MTB gene products with available structures was analyzed. Majority of the MTB proteins belonged to the α/β class. 23 different protein folds are used in the MTB protein structures. Of these, the TIM barrel fold was found to be highly conserved even at very low sequence identity. We identified 21 paralogs and 27 analogs of MTB based on domains and EC classification. Our analysis revealed that many of the current drug targets share structural similarity with other proteins within the MTB genome, which could probably be off-targets. Results of this analysis have been made available in the Mycobacterium tuberculosis Structural Database (http://bmi.icmr.org.in/mtbsd/MtbSD.php/search.php) which is a useful resource for current and novel drug targets of MTB.     

Cloning and analysis of the aroB gene encoding dehydroquinate synthase from Corynebacterium glutamicum

The aroB gene encoding dehydroquinate synthase of Corynebacterium glutamicum has been cloned by complementation of an aro auxotrophic mutant of Escherichia coli with the genomic DNA library. The recombinant plasmid contained a 1.4-kb fragment that complemented the Escherichia coli dehydroquinate-synthase-deficient mutant. The nucleotide sequences of the subcloned DNA has been determined. The sequences contain an open reading frame of 360 codons, from which a protein with a molecular mass of about 38 kDa could be predicted. This is consistent with the size of the AroB protein expressed in E. coli. Alignment of different prokaryotic and eukaryotic aroB gene products reveals an overall identity ranging from 29 to 57% and the presence of several highly conserved regions.     

Acyl-CoA carboxylases (accD2 and accD3), together with a unique polyketide synthase (Cg-pks), are key to mycolic acid biosynthesis in Corynebacterianeae such as Corynebacterium glutamicum and Mycobacterium tuberculosis

The Corynebacterianeae such as Corynebacterium glutamicum and Mycobacterium tuberculosis possess several unique and structurally diverse lipids, including the genus-specific mycolic acids. Although the function of a number of genes involved in fatty acid and mycolic acid biosynthesis is known, information relevant to the initial steps within these biosynthetic pathways is relatively sparse. Interestingly, the genomes of Corynebacterianeae possess a high number of accD genes, whose gene products resemble the beta-subunit of the acetyl-CoA carboxylase of Escherichia coli, providing the activated intermediate for fatty acid synthesis. We present here our studies on four putative accD genes found in C. glutamicum. Although growth of the accD4 mutant remained unchanged, growth of the accD1 mutant was strongly impaired and partially recovered by the addition of exogenous oleic acid. Overexpression of accD1 and accBC, encoding the carboxylase alpha-subunit, resulted in an 8-fold increase in malonyl-CoA formation from acetyl-CoA in cell lysates, providing evidence that accD1 encodes a carboxyltransferase involved in the biosynthesis of malonyl-CoA. Interestingly, fatty acid profiles remained unchanged in both our accD2 and accD3 mutants, but a complete loss of mycolic acids, either as organic extractable trehalose and glucose mycolates or as cell wall-bound mycolates, was observed. These two carboxyltransferases are also retained in all Corynebacterianeae, including Mycobacterium leprae, constituting two distinct groups of orthologs. Furthermore, carboxyl fixation assays, as well as a study of a Cg-pks deletion mutant, led us to conclude that accD2 and accD3 are key to mycolic acid biosynthesis, thus providing a carboxylated intermediate during condensation of the mero-chain and alpha-branch directed by the pks-encoded polyketide synthase. This study illustrates that the high number of accD paralogs have evolved to represent specific variations on the well known basic theme of providing carboxylated intermediates in lipid biosynthesis.     

Differential immunogenicity and protective efficacy of DNA vaccines expressing proteins of Mycobacterium tuberculosis in a mouse model

BALB/c mice were vaccinated three times (2-week intervals) with plasmid DNA separately encoding antigen Ag85B, ESAT-6 or Ag85A from Mycobacterium tuberculosis. The protective efficacy of these DNA vaccines against intravenous M. tuberculosis H37Rv challenge infection was measured by counting bacterial loads in spleen and lung and recording changes in lung pathology. The splenocyte proliferative response to the corresponding antigens and antigen-specific interferon (IFN)-γ secreted by splenocytes of the vaccinated mice were also detected. We found a clear hierarchy of protective efficacies among the three DNA vaccines tested in this study. Plasmid DNA encoding Ag85A provided the strongest protection and showed the least change in lung pathology, followed by plasmid DNAs encoding Ag85B and ESAT-6. However, DNA-85B reduced comparative bacterial load in lung tissue, as did DNA-85A. Compared to the control group, protective efficacies conferred by different DNA vaccines were consistent with the lymphoproliferative responses to the corresponding antigens as well as the secretions of antigen-specific IFN-γ. Our study demonstrates that both Ag85A and Ag85B are the most promising of the candidate antigens tested for future TB vaccine development.     

Inactivation of polyketide synthase and related genes results in the loss of complex lipids in Mycobacterium tuberculosis H37Rv

AIMS: Phthiocerol dimycocerosate (PDIM) waxes and other lipids are necessary for successful Mycobacterium tuberculosis infection, although the exact role of PDIM in host-pathogen interactions remains unclear. In this study, we investigated the contribution of tesA, drrB, pks6 and pks11 genes in complex lipid biosynthesis in M. tuberculosis. METHODS AND RESULTS: Four mutants were selected from M. tuberculosis H37Rv transposon mutant library. The transposon insertion sites were confirmed to be within the M. tuberculosis open reading frames for tesA (a probable thioesterase), drrB (predicted ABC transporter), pks11 (putative chalcone synthase) and pks6 (polyketide synthase). The first three of these transposon mutants were unable to generate PDIM and the fourth lacked novel polar lipids. CONCLUSIONS: Mycobacterium tuberculosis can be cultivated in vitro without the involvement of certain lipid synthesis genes, which may be necessary for in vivo pathogenicity. SIGNIFICANCE AND IMPACT OF THE STUDY: The use of transposon mutants is a new functional genomic approach for the eventual definition of the mycobacterial 'lipidome'.