Crystal structures of lipoglycopeptide antibiotic deacetylases: implications for the biosynthesis of A40926 and teicoplanin

The lipoglycopeptide antibiotics teicoplanin and A40926 have proven efficacy against Gram-positive pathogens. These drugs are distinguished from glycopeptide antibiotics by N-linked long chain acyl-D-glucosamine decorations that contribute to antibacterial efficacy. During the biosynthesis of lipoglycopeptides, tailoring glycosyltransferases attach an N-acetyl-D-glucosamine to the aglycone, and this N-acetyl-glucosaminyl pseudoaglycone is deacetylated prior to long chain hydrocarbon attachment. Here we present several high-resolution crystal structures of the pseudoaglycone deacetylases from the biosynthetic pathways of teicoplanin and A40926. The cocrystal structure of the teicoplanin pseudoaglycone deacetylase with a fatty acid product provides further insights into the roles of active-site residues, and suggests mechanistic similarities with structurally distinct zinc deacetylases, such as peptidoglycan deacetylase and LpxC. A unique, structurally mobile capping lid, located at the apex of these pseudoaglycone deacetylases, likely serves as a determinant of substrate specificity.     

Genetic characterization of the chlorothricin gene cluster as a model for spirotetronate antibiotic biosynthesis

The biosynthetic gene cluster for chlorothricin (CHL) was localized to a 122 kb contiguous DNA from Streptomyces antibioticus DSM 40725, and its involvement in CHL biosynthesis was confirmed by gene inactivation and complementation. Bioinformatic analysis of the sequenced 111.989 kb DNA region revealed 42 open reading frames, 35 of which were defined to constitute the CHL gene cluster. An assembly model for CHL biosynthesis from D-olivose, 2-methoxy-5-chloro-6-methylsalicyclic acid, and chlorothricolide building blocks was proposed. This work represents cloning of a gene cluster for spirotetronate antibiotic biosynthesis and sets the stage to investigate the unusual macrolide biosynthesis including tandem Diels-Alder cyclizations, Baeyer-Villiger oxidation, and incorporation of an enoylpyruvate unit.     

Biosynthetic gene cluster of the glycopeptide antibiotic teicoplanin: characterization of two glycosyltransferases and the key acyltransferase

The gene cluster encoding biosynthesis of the clinically important glycopeptide antibiotic teicoplanin has been cloned from Actinoplanes teichomyceticus. Forty-nine putative open reading frames (ORFs) were identified within an 89 kbp genetic locus and assigned roles in teicoplanin biosynthesis, export, resistance, and regulation. Two ORFs, designated orfs 1 and 10*, showed significant homology to known glycosyltransferases. When heterologously expressed in Escherichia coli, these glycosyltransferases were shown to catalyze the transfer of UDP-(N-acetyl)-glucosamine onto, respectively, 3-chloro-beta-hydroxytyrosine-6 (3-Cl-6betaHty) and 4-hydroxyphenylglycine-4 (4Hpg) of the teicoplanin heptapeptide aglycone. The product of another ORF, orf11*, was demonstrated in vitro to transfer n-acetyl-, n-butyryl-, and n-octanoyl-groups from acyl-CoA donors either to a free UDP-aminosugar or to an aminosugar moiety in the teicoplanin pseudoaglycone, thus identifying Orf11* as the key acyltransferase in teicoplanin maturation. These findings should accelerate the combinatorial engineering of new and improved glycopeptide drugs.     

The biosynthetic gene cluster for a monocyclic beta-lactam antibiotic, nocardicin A

The monocyclic beta-lactam antibiotic nocardicin A is related structurally and biologically to the bicyclic beta-lactams comprised of penicillins/cephalosporins, clavams, and carbapenems. Biosynthetic gene clusters are known for each of the latter, but not for monocyclic beta-lactams. A previously cloned gene encoding an enzyme specific to the biosynthetic pathway was used to isolate the nocardicin A cluster from Nocardia uniformis. Sequence analysis revealed the presence of 14 open reading frames involved in antibiotic production, resistance, and export. Among these are a two-protein nonribosomal peptide synthetase system, p-hydroxyphenylglycine biosynthetic genes, an S-adenosylmethionine-dependent 3-amino-3-carboxypropyl transferase (Nat), and a cytochrome P450. Gene disruption mutants of Nat, as well as an activation domain of the NRPS system, led to loss of nocardicin A formation. Several enzymes involved in antibiotic biosynthesis were heterologously overproduced, and biochemical characterization confirmed their proposed activities.     

A glycopeptide antibiotic chiral stationary phase for the enantiomer resolution of hydroxy acid derivatives by capillary electrochromatography

Separation of hydroxy acid enantiomers was achieved by using capillary electrochromatography (CEC) employing a chiral stationary phase (CSP) based on MDL 63,246 (Hepta-Tyr), a macrocyclic antibiotic of the teicoplanin family. The chiral selector was chemically bonded to 5 num diol-modified silica particles and the CSP mixed with amino silica (3:1 w/w) was packed into a 75 num ID fused-silica capillary. The CEC experiments were carried out by using an aqueous reversed-phase mode for the enantiomeric resolution of hydroxy acid compounds. Good enantioresolution was achieved for mandelic acid (MA), m-hydroxymandelic acid (m-OH-MA), p-OH-MA, and 3-hydroxy-4-methoxymandelic acid (3-OH-4-MeO-MA). The CEC system was less enantioselective towards 2-phenyllactic acid (2-PhL) and 3-PhL while mandelic acid methyl ester (MA-Et-Est) enantiomers were not resolved. Several experimental parameters, such as organic solvent type and concentration, buffer pH, capillary temperature, on enantioresolution factor, retention time, and retention factor were studied.     

The gene cluster for fluorometabolite biosynthesis in Streptomyces cattleya: a thioesterase confers resistance to fluoroacetyl-coenzyme A

A genomic library of Streptomyces cattleya was screened to isolate a gene cluster encoding enzymes responsible for the production of fluorine-containing metabolites. In addition to the previously described fluorinase FlA which catalyzes the formation of 5'-fluoro-5'-deoxyadenosine from S-adenosylmethionine and fluoride, 11 other putative open reading frames have been identified. Three of the proteins encoded by these genes have been characterized. FlB was determined to be the second enzyme in the pathway, catalyzing the phosphorolytic cleavage of 5'-fluoro-5'-deoxyadenosine to produce 5-fluoro-5-deoxy-D-ribose-1-phosphate. The enzyme FlI was found to be an S-adenosylhomocysteine hydrolase, which may act to relieve S-adenosylhomocysteine inhibition of the fluorinase. Finally, flK encodes a thioesterase which catalyzes the selective breakdown of fluoroacetyl-CoA but not acetyl-CoA, suggesting that it provides the producing strain with a mechanism for resistance to fluoroacetate.     

Optimization of fermentation conditions for production of glycopeptide antibiotic vancomycin by Amycolatopsis orientalis

Glycopeptides produced by Streptomyces species are the drugs used against β-lactam drug-resistant staphylococcal infections, and vancomycin is important among them. Increased prevalence of resistant strains increased the usage of vancomycin worldwide and also promoted attempts for indigenous production. The optimum process conditions pH, temperature, inoculum size, agitation, and aeration for vancomycin production by Amycolatopsis orientalis were evaluated, statistically analyzed, and the response surface curves were constructed. The optimum process conditions were a pH of 7.6, a temperature of 29°C, an inoculum size of 4.5%, an agitation of 255 rpm, and an aeration of less than 1:10 medium-to-air ratio.     

A gene cluster from a marine Streptomyces encoding the biosynthesis of the aromatic spiroketal polyketide griseorhodin A

The telomerase inhibitor griseorhodin A is probably the most heavily oxidized bacterial polyketide known and features a unique epoxyspiroketal moiety crucial for its activity. To gain insight into which tailoring enzymes generate this pharmacophore, we have cloned and fully sequenced the griseorhodin biosynthesis gene cluster. Among other unusual features, this aromatic polyketide synthase (PKS) system encodes an unprecedented number of functionally diverse oxidoreductases, which are involved in the oxidative modification of a polyaromatic tridecaketide precursor by cleavage of three carbon-carbon bonds. The cluster was highly unstable on a variety of shuttle plasmids but could finally be functionally expressed in its entirety in Streptomyces lividans using a novel integrative cosmid vector. The availability of the tailoring system now opens up the possibility of engineering nonnatural biosynthetic pathways yielding novel pharmacologically active analogs with a similar pharmacophore.     

Characterization of the alnumycin gene cluster reveals unusual gene products for pyran ring formation and dioxan biosynthesis

Alnumycin is closely related to the benzoisochromanequinone (BIQ) polyketides such as actinorhodin. Exceptional structural features include differences in aglycone tailoring that result in the unique alnumycin chromophore and the existence of an unusual 4-hydroxymethyl-5-hydroxy-1,3-dioxan moiety. Cloning and sequencing of the alnumycin gene cluster from Streptomyces sp. CM020 revealed expected biosynthesis genes for polyketide assembly, but several genes encoding subsequent tailoring enzymes were highly atypical. Heterologous expression studies confirmed that all of the genes required for alnumycin biosynthesis resided within the sequenced clone. Inactivation of genes aln4 and aln5 showed that the mechanism of pyran ring formation differs from actinorhodin and granaticin pathways. Further inactivation studies identified two genes, alnA and alnB, involved in the synthesis and attachment of the dioxan moiety, and resulted in the production of the polyketide prealnumycin.     

The complete gene cluster of the antitumor agent gilvocarcin V and its implication for the biosynthesis of the gilvocarcins

Gilvocarcin V, an antitumor agent produced by the bacterium Streptomyces griseoflavus G? 3592, is the most studied representative of the distinct family of benzo[d]naphtho[1,2-b]pyran-6-one aryl C-glycoside antibiotics, which show excellent antitumor activity and a remarkably low toxicity. Its biosynthesis contains many intriguing steps, including an oxidative rearrangement, the C-glycosylation, and the generation of a vinyl side chain. These steps all contribute to structural elements of the drug, which are essential for its biological activity, but only poorly understood. Herein we report the cloning and characterization of the gilvocarcin (gil) gene cluster from S. griseoflavus G? 3592, and its heterologous expression in a foreign host (S. lividans). This is the first reported gene cluster encoding the biosynthesis of a benzo[d]naphtho[1,2-b]pyran-6-one aryl C-glycoside antibiotic, which not only provides insights regarding the biosynthesis of gilvocarcin V but also lays the foundation for the detailed studies of its intriguing biosynthetic steps and possibly for the generation of gilvocarcin analogues with improved biological activities through combinatorial biosynthesis.     

Identification of a gene cluster that directs putrebactin biosynthesis in Shewanella species: PubC catalyzes cyclodimerization of N-hydroxy-N-succinylputrescine

Putrebactin is a dihydroxamate iron chelator produced by the metabolically versatile marine bacterium Shewanella putrefaciens. It is a macrocyclic dimer of N-hydroxy-N-succinyl-putrescine (HSP) and is structurally related to desferrioxamine E, which is a macrocyclic trimer of N-hydroxy-N-succinyl-cadaverine (HSC). We recently showed that DesD, a member of the NIS synthetase superfamily, catalyzes the key step in desferrioxamine E biosynthesis: ATP-dependent trimerisation and macrocylization of HSC. Here we report identification of a conserved gene cluster in the sequenced genomes of several Shewanella species, including Shewanella putrefaciens, which is hypothesized to direct putrebactin biosynthesis from putrescine, succinyl-CoA and molecular oxygen. The pubC gene within this gene cluster encodes a protein with 65% similarity to DesD. We overexpressed pubC from Shewanella species MR-4 and MR-7 in E. coli. The resulting His6-PubC fusion proteins were purified by Ni-NTA affinity and gel filtration chromatography. The recombinant proteins were shown to catalyze ATP-dependent cyclodimerization of HSP to form putrebactin. The uncyclized dimer of HSP pre-putrebactin was shown to be an intermediate in the conversion of two molecules of HSP to putrebactin. The data indicate that pre-putrebactin is converted to putrebactin via PubC-catalyzed activation of the carboxyl group by adenylation, followed by PubC-catalyzed nucleophilic attack of the amino group on the carbonyl carbon of the acyl adenylate. This mechanism for macrocycle formation is very different from the mechanism involved in the biosynthesis of many other macrocyclic natural products, where already-activated acyl thioesters are converted by thioesterase domains of polyketide synthases and nonribosomal peptide synthetases to macrocycles via covalent enzyme bound intermediates. The results of this study demonstrate that two closely related enzymes, PubC and DesD, catalyze specific cyclodimerization and cyclotrimerization reactions, respectively, of structurally similar substrates, raising intriguing questions regarding the molecular mechanism of specificity.     

Characterization of the mupirocin biosynthesis gene cluster from Pseudomonas fluorescens NCIMB 10586

The polyketide antibiotic mupirocin (pseudomonic acid) produced by Pseudomonas fluorescens NCIMB 10586 competitively inhibits bacterial isoleucyl-tRNA synthase and is useful in controlling Staphylococcus aureus, particularly methicillin-resistant Staphylococcus aureus. The 74 kb mupirocin biosynthesis cluster has been sequenced, and putative enzymatic functions of many of the open reading frames (ORFs) have been identified. The mupirocin cluster is a combination of six larger ORFs (mmpA-F), containing several domains resembling the multifunctional proteins of polyketide synthase and fatty acid synthase type I systems, and individual genes (mupA-X and macpA-E), some of which show similarity to type II systems (mupB, mupD, mupG, and mupS). Gene knockout experiments demonstrated the importance of regions in mupirocin production, and complementation of the disrupted gene confirmed that the phenotypes were not due to polar effects. A model for mupirocin biosynthesis is presented based on the sequence and biochemical evidence.     

A complete gene cluster from Streptomyces nanchangensis NS3226 encoding biosynthesis of the polyether ionophore nanchangmycin

The PKS genes for biosynthesis of the polyether nanchangmycin are organized to encode two sets of proteins (six and seven ORFs, respectively), but are separated by independent ORFs that encode an epimerase, epoxidase, and epoxide hydrolase, and, notably, an independent ACP. One of the PKS modules lacks a corresponding ACP. We propose that the process of oxidative cyclization to form the polyether structure occurs when the polyketide chain is still anchored on the independent ACP before release. 4-O-methyl-L-rhodinose biosynthesis and its transglycosylation involve four putative genes, and regulation of nanchangmycin biosynthesis seems to involve activation as well as repression. In-frame deletion of a KR6 domain generated the nanchangmycin aglycone with loss of 4-O-methyl-L-rhodinose and antibacterial activity, in agreement with the assignments of the PKS domains catalyzing specific biosynthetic steps.     

Characterization of protein encoded by spnR from the spinosyn gene cluster of Saccharopolyspora spinosa: mechanistic implications for forosamine biosynthesis

d-Forosamine is a 4-N,N-(dimethylamino)-2,3,4,6-tetradeoxy-alpha-d-threo-hexopyranose found in spinosyn produced by Saccharopolyspora spinosa. Studies of spinosyn biosynthesis in S. spinosa led to the isolation of the entire biosynthetic gene cluster. Heterologous expression of spnR, one putative gene in forosamine biosynthesis, in E. coli and purification of the SpnR protein identified it as an aminotransferase catalyzing the conversion of the 4-keto-2,3,6-trideoxy sugar intermediate to the corresponding 4-amino sugar product. Identification of SpnR function relied on the use of a stable TMP-phosphonate sugar in place of TDP-sugar substrate to determine the function of SpnR. This strategy may find general applicability for designing probes to study enzymes which catalyze the transformation of labile deoxysugar intermediates.     

The biosynthesis of teicoplanin-type glycopeptide antibiotics: assignment of p450 mono-oxygenases to side chain cyclizations of glycopeptide a47934

Streptomyces toyocaensis produces A47934, a teicoplanin-like type-IV glycopeptide with antibiotic activity against methicillin-resistant Staphylococcus aureus. A47934 differs from the type-I vancomycin glycopeptides, which possess a tricyclic peptide backbone, by the presence of an additional ring closure between the aromatic amino acids 1 and 3. To elucidate the order of crosslinking reactions, P450 mono-oxygenase-inactivation mutants (DeltastaF, DeltastaG, DeltastaH, and DeltastaJ) of the A47934 producer were generated, and the accumulated intermediates were analyzed. Thus, the formation of each crosslink could unambiguously be assigned to a specific oxygenase. The structure of the released intermediates from the wild-type nonribosomal peptide synthetase assembly line facilitated the determination of the cyclization order. Unexpectedly, the additional ring closure in A47934, catalyzed by StaG, is the second oxygenase reaction.