Structural basis of O-methylation of (2-heptyl-)1-hydroxyquinolin-4(1H)-one and related compounds by the heterocyclic toxin methyltransferase Rv0560c of Mycobacterium tuberculosis.

The S-adenosyl-L-methionine-dependent methyltransferase Rv0560c of Mycobacterium tuberculosis belongs to an orthologous group of heterocyclic toxin methyltransferases (Htm) which likely contribute to resistance of mycobacteria towards antimicrobial natural compounds as well as drugs. HtmM.t. catalyzes the methylation of the Pseudomonas aeruginosa toxin 2-heptyl-1-hydroxyquinolin-4(1H)-one (also known as 2-heptyl-4-hydroxyquinoline N-oxide), a potent inhibitor of respiratory electron transfer, its 1-hydroxyquinolin-4(1H)-one core (QNO), structurally related (iso)quinolones, and some mycobactericidal compounds. In this study, crystal structures of HtmM.t. in complex with S-adenosyl-L-homocysteine (SAH) and the methyl-accepting substrates QNO or 4-hydroxyisoquinoline-1(2H)-one, or the methylated product 1-methoxyquinolin-4(1H)-one, were determined at < 1.9 Å resolution. The monomeric protein exhibits the typical Rossmann fold topology and conserved residues of class I methyltransferases. Its SAH binding pocket is connected via a short tunnel to a large solvent-accessible cavity, which accommodates the methyl-accepting substrate. Residues W44, F168, and F208 in connection with F212 form a hydrophobic clamp around the heteroaromatic ring of the methyl-accepting substrate and likely play a major role in substrate positioning. Structural and biochemical data suggest that H139 and T136 are key active site residues, with H139 acting as general base that activates the methyl-accepting hydroxy group. Our structural data may contribute to the design of Htm inhibitors or of antimycobacterial drugs unamenable for methylation.

Modification of the Pseudomonas aeruginosa toxin 2-heptyl-1-hydroxyquinolin-4(1H)-one and other secondary metabolites by methyltransferases from mycobacteria.

The opportunistic pathogen Pseudomonas aeruginosa, one of the most prevalent species in infections of the cystic fibrosis lung, produces a range of secondary metabolites, among them the respiratory toxin 2-heptyl-1-hydroxyquinolin-4(1H)-one (2-heptyl-4-hydroxyquinoline N-oxide, HQNO). Cultures of the emerging cystic fibrosis pathogen Mycobacteroides abscessus detoxify HQNO by methylating the N-hydroxy moiety. In this study, the class I methyltransferase MAB_2834c and its orthologue from Mycobacterium tuberculosis, Rv0560c, were identified as HQNO O-methyltransferases. The P. aeruginosa exoproducts 4-hydroxyquinolin-2(1H)-one (DHQ), 2-heptylquinolin-4(1H)-one (HHQ), and 2-heptyl-3-hydroxyquinolin-4(1H)-one (the 'Pseudomonas quinolone signal', PQS), some structurally related (iso)quinolones, and the flavonol quercetin were also methylated; however, HQNO was by far the preferred substrate. Both enzymes converted a benzimidazole[1,2-a]pyridine-4-carbonitrile-based compound, representing the scaffold of antimycobacterial substances, to an N-methylated derivative. We suggest that these promiscuous methyltransferases, newly termed as heterocyclic toxin methyltransferases (Htm), are involved in cellular response to chemical stress and possibly contribute to resistance of mycobacteria toward antimicrobial natural compounds as well as drugs. Thus, synthetic antimycobacterial agents may be designed to be unamenable to methyl transfer. ENZYMES: S-adenosyl-l-methionine:2-heptyl-1-hydroxyquinolin-4(1H)-one O-methyl-transferase, EC 2.1.1.

Efficient modification of the Pseudomonas aeruginosa toxin 2-heptyl-1-hydroxyquinolin-4-one by three Bacillus glycosyltransferases with broad substrate ranges.

Glycosylation of natural and synthetic products can alter the physical, chemical and pharmacological properties of the aglycon. Conversion of 2-heptyl-1-hydroxyquinolin-4-one (HQNO), a potent respiratory inhibitor produced by Pseudomonas aeruginosa, to the less toxic 2-heptyl-1-(ß-D-glucopyranosydyl)-quinolin-4-one, was recently demonstrated for Bacillus subtilis strain 168. In this study, we compared the genomes of several Bacillus spp. to identify candidate enzymes for HQNO glucosylation. All three (putative) UDP-glycosyltransferases (GT) of B. subtilis 168 tested, YjiC, YdhE and YojK, were capable of HQNO glucosylation, with YjiC showing the highest turnover rate (kcat) of 4.6 s-1, and YdhE exhibiting the lowest Km value for HQNO of 9.1 µM. All three GT predominantly utilized UDP-glucose, but YdhE was similarly active with TDP-glucose. Among the aglycons tested, HQNO was the preferred substrate of all three GT, but they also showed activities toward the P. aeruginosa exoproducts pyocyanin, 2-heptyl-3-hydroxyquinolin-4(1H)-one (the Pseudomonas quinolone signal) and 2,4-dihydroxyquinoline, the plant derived antimicrobials vanillin and quercetin, and the macrolide antibiotic tylosin A. Our results underline the promiscuity and substrate flexibility of YjiC, YdhE and YojK, and suggest a physiological role in natural toxin resistance of B. subtilis. Especially YdhE appears to be an attractive biocatalyst for the glycoengineering of natural products.