Expanding the scope of aromatic polyketides by combinatorial biosynthesis.

Combinatorial biosynthesis is a technology for mixing genes responsible for the biosynthesis of secondary metabolites, in order to generate products for compound libraries serendipitously or to cause desired modifications to natural products. Both of these approaches are extremely useful in drug discovery. Streptomyces and related species are abundant in bioactive secondary metabolites and were therefore the first microbes to be used for combinatorial biosynthesis. Polyketides are the most abundant medicinal agents among natural products. Structural diversity and a wide scope of bioactivities are typical of the group. However, the common feature of polyketides is a biosynthetic process from simple carboxylic acid residues. In molecular genetics, polyketides are sub-classified as types I and II, called modular and aromatic polyketides respectively. The best-known bioactivities of aromatic polyketides are their antibacterial and antitumor effects. Genetic analysis of aromatic polyketides has resulted in almost 30 cloned and identified biosynthetic gene clusters. Several biosynthetic enzymes are flexible enough to allow their use in combinatorial biosynthesis to create high diversity compound libraries. This review describes the state of the art of combinatorial biosynthesis, giving anthracyclines as examples. Contiguous DNA sequences for antibiotics, cloned from four different anthracycline producers, provide tools for rapid lead optimization or other structural modification processes, and not only for anthracyclines. Two gene cassettes enabling fast and flexible structural modification of polyketides are introduced in this paper.

The entire nogalamycin biosynthetic gene cluster of Streptomyces nogalater: characterization of a 20-kb DNA region and generation of hybrid structures.

Fragments spanning 20 kb of Streptomyces nogalater genomic DNA were characterized to elucidate the molecular genetic basis of the biosynthetic pathway of the anthracycline antibiotic nogalamycin. Structural analysis of the products obtained by expression of the fragments in S. galilaeus and S. peucetius mutants producing aclacinomycin and daunomycin metabolites, respectively, revealed hybrid compounds in which either the aglycone or the sugar moiety was modified. Subsequent sequence analysis revealed twenty ORFs involved in nogalamycin biosynthesis, of which eleven could be assigned to the deoxysugar pathway, four to aglycone biosynthesis, while the remaining five express products with unknown function. On the basis of sequence similarity and experimental data, the functions of the products of the newly discovered genes were determined. The results suggest that the entire biosynthetic gene cluster for nogalamycin is now known. Furthermore, the compounds obtained by heterologous expression of the genes show that it is possible to use the genes in combinatorial biosynthesis to create novel chemical structures for drug screening purposes.

A gene cluster from Streptomyces galilaeus involved in glycosylation of aclarubicin.

We have cloned and characterized a gene cluster for anthracycline biosynthesis from Streptomyces galilaeus. This cluster, 15-kb long, includes eight genes involved in the deoxyhexose biosynthesis pathway, a gene for a glycosyltransferase and one for an activator, as well as two genes involved in aglycone biosynthesis. Gene disruption targeted to the activator gene blocked production of aclacinomycins in S. galilaeus. Plasmid pSgs4, containing genes for a glycosyltransferase (aknS), an aminomethylase (aknX), a glucose-1-phosphate thymidylyltransferase (akn Y) and two genes for unidentified glycosylation functions (aknT and aknV), restored the production of aclacinomycins in the S. galilaeus mutants H063, which accumulates aklavinone, and H054, which produces aklavinone with rhodinose and deoxyfucose residues. Furthermore, pSgs4 directed the production of L-rhamnosyl-epsilon-rhodomycinone and L-daunosaminyl-epsilon-rhodomycinone in S. peucetius strains that produce epsilon-rhodomycinone endogenously. Subcloning of the gene cluster was carried out in order to further define the genes that are responsible for complementation and hybrid anthracycline generation.

Modifications of aclacinomycin T by aclacinomycin methyl esterase (RdmC) and aclacinomycin-10-hydroxylase (RdmB) from Streptomyces purpurascens.

The genes rdmB and rdmC of Streptomyces purpurascens encoding aclacinomycin modifying enzymes RdmB and RdmC were expressed in Streptomyces lividans TK24. In contrast to the earlier suggestion that RdmC may be an esterase that causes the removal of the carbomethoxy group from the 10 position of aclacinomycins, RdmC functions as an aclacinomycin methyl esterase and catalyzes the removal of the methoxy group from the C-15 position of aclacinomycin T producing 15-demethoxyaclacinomycin T. RdmB acts upon C-10 of 15-demethoxyaclacinomycin T and is able to remove the carboxylic group from the C-10 position. It functions also as an aclacinomycin-10-hydroxylase being able to add a hydroxyl group at the same, C-10 position in vitro. Aclacinomycin methyl esterase was purified to apparent homogeneity from S. lividans carrying the rdmC and aclacinomycin-10-hydroxylase as a glutathione S-transferase fusion construct from Escherichia coli carrying the rdmB gene, respectively. Aclacinomycin methyl esterase functions as a monomer and aclacinomycin-10-hydroxylase as a tetramer. Aclacinomycin methyl esterase has an exceptionally high temperature stability and has an apparent K(m) for aclacinomycin T of 15.5 microM. The introduction of rdmC and rdmB in a Streptomyces galilaeus mutant HO38 produced the same modifications of aclacinomycin T in vivo as aclacinomycin methyl esterase and aclacinomycin-10-hydroxylase in vitro.

Hybrid anthracyclines from a genetically engineered Streptomyces galilaeus mutant.

The genetic engineering of antibiotic-producing Streptomyces strains is an approach that is emerging and ready to become established as a successful methodology in developing analogues of the original, pharmaceutically important, natural products obtained from the organisms. The current report highlights this succes by demonstrating the high-level production of novel anthracyclines. The biosynthetic pathways of the nogalamycin-producing Streptomyces nogalater and the aclacinomycin-producing S. galilaeus were combined by transferring the genes of S. nogalater polyketide synthetase into a nonproducing S. galilaeus mutant. The resulting anthracycline antibiotics that were produced possessed structural features characteristic of compounds from both of the undoctored Streptomycesstrains.

Identification of a cyclase gene dictating the C-9 stereochemistry of anthracyclines from Streptomyces nogalater.

Nogalamycin is an anthracycline antibiotic produced by Streptomyces nogalater. Its aglycone has a unique stereochemistry (7S, 9S, 10R) compared to that of most other anthracyclines (7S, 9R, 10R). The gene snoaL, encoding a nogalonic acid methyl ester cyclase for nogalamycin, was used to generate nogalamycinone, demonstrating that the single cyclase dictates the C-9 stereochemistry of anthracyclines.

Crystal structure of Bacillus subtilis YabJ, a purine regulatory protein and member of the highly conserved YjgF family.

The yabJ gene in Bacillus subtilis is required for adenine-mediated repression of purine biosynthetic genes in vivo and codes for an acid-soluble, 14-kDa protein. The molecular mechanism of YabJ is unknown. YabJ is a member of a large, widely distributed family of proteins of unknown biochemical function. The 1.7-A crystal structure of YabJ reveals a trimeric organization with extensive buried hydrophobic surface and an internal water-filled cavity. The most important finding in the structure is a deep, narrow cleft between subunits lined with nine side chains that are invariant among the 25 most similar homologs. This conserved site is proposed to be a binding or catalytic site for a ligand or substrate that is common to YabJ and other members of the YER057c/YjgF/UK114 family of proteins.

An efficient approach for screening minimal PKS genes from Streptomyces.

Degenerated oligonucleotide primers were designed to amplify fragments of ketosynthase genes from polyketide antibiotics producing Streptomyces spp. and bacterial strains enriched from soil samples. Cell lysates were used as templates in amplification, so time-consuming DNA purification was avoided. A phylogenetic tree constructed from the amino acid sequences of the amplified fragments shows a distribution of spore pigments and antibiotics in separate classes. In addition, several different subgroups form within the antibiotics group. Anthracyclines were divided into separate branches according to the starter unit used in biosynthesis.

Incorrectly folded aromatic polyketides from polyketide reductase deficient mutants.

Compounds produced by the polyketide ketoreductase deficient Streptomyces mutants HO61 and P67 are described. The structures of the compounds indicate that ketoreductase activity is required for correct condensation of the polyketide chain in the biosynthesis of aromatic polyketides.

A role for a highly conserved protein of unknown function in regulation of Bacillus subtilis purA by the purine repressor.

Regulation of the purine biosynthetic gene purA was examined by using a transcriptional fusion to a luciferase reporter gene. Transcription was repressed about 10-fold by the addition of adenine and increased approximately 4.5-fold by the addition of guanosine. This regulation is mediated by a purine repressor (PurR). In a purR mutant, basal expression was increased 10-fold, and there was no further stimulation by guanosine or repression by adenine. An open reading frame, yabJ, immediately downstream from purR was found to have a role in the repression of purA by adenine. Repression by adenine was perturbed in a purR+ yabJ mutant, although guanosine regulation was retained. Mutations in the PurR PRPP binding motif abolished guanosine regulation in the yabJ mutant. Thus, PRPP appears to be required for upregulation by guanosine. The amino acid sequence of YabJ is homologous to the YER057c/YjgF protein family of unknown function.

Isolation and characterization of a purC(orf)QLF operon from Lactococcus [correction of Lactobacillus] lactis MG1614.

We have isolated genes encoding enzymes of the de novo purine nucleotide biosynthesis pathway from Lactococcus lactis MG1614 by colony hybridization using DIG-labeled DNA probes. The organization of the genes needed for the de novo biosynthesis of purine nucleotides in L. lactis differs from that found in other organisms. In L. lactis there is a gene cluster, which contains five out of the 11 genes needed for the de novo biosynthesis of IMP, namely purC, orf, purQ, purL and purF. These genes were shown to be transcribed as a single transcription unit by Northern hybridization analysis. The 5' end of the transcript of the purC(orf)QLF operon was determined by primer extension analysis using fluorescently end-labeled probes. The purC(orf)QLF operon of L. lactis is transcribed in Escherichia coli, and the gene product of the purF gene, glutamine phosphoribosylpyrophosphate amidotransferase (glutamine PRPP ATase, EC 2.4.2.14), can functionally complement the E. coli purF mutant strain TX158. We also show that the promoter of the purC(orf)QLF operon is regulated in response to exogenously added purines.

Characterization of aklavinone-11-hydroxylase from Streptomyces purpurascens.

Aklavinone-11-hydroxylase (RdmE) is a FAD monooxygenase participating in the biosynthesis of daunorubicin, doxorubicin and rhodomycins. The rdmE gene encodes an enzyme of 535 amino acids. The sequence of the Streptomyces purpurascens enzyme is similar to other Streptomyces aromatic polyketide hydroxylases. We overexpressed the gene in Streptomyces lividans and purified aklavinone-11-hydroxylase to apparent homogeneity with four chromatographic steps utilizing a kinetic photometric enzyme assay. The enzyme is active as the monomer with a molecular mass of 60 kDa; it hydroxylates aklavinone and other anthracyclinones. Aklavinone-11-hydroxylase can use both NADH and NADPH as coenzyme but it is slowly inactivated in the presence of NADH. The apparent Km for NADPH is 2 mM and for aklavinone 10 microM. The enzyme is inactivated in the presence of phenylglyoxal and 2,3-butanedione. NADPH protects against inactivation of aklavinone-11-hydroxylase by phenylglyoxal.

Engineering the steroid-specificity of an anti-17beta-estradiol Fab by random mutagenesis and competitive phage panning.

We have employed random mutagenesis and phage display to improve the steroid-specificity of an anti-17beta-estradiol Fab fragment. The VH domain was mutated using error-prone PCR; the mutation rate was controlled by adjusting the number of effective duplications. A phage library of 2 x 10(6) independent mutants was generated, each mutant containing on average 24 amino acid changes. We selected for decreased testosterone (TES) cross-reactivity by adding a large excess TES as a competitor to the panning reactions. After four panning rounds, the cross-reactivities of the individual mutant clones ranged from 19 to 4%, showing up to 20-fold improvement over the original value (78%). Estradiol affinities were mainly unchanged. Sequencing of the VH regions revealed two hot spots, one located around Ser32 in CDR1 and the other around Thr52A in CDR2, while no mutations were found in CDR3. Although most clones had multiple mutations, it was possible to deduce the residues relevant to the improved specificity by comparing the sequences and binding data of the mutants. We demonstrated that controlled error-prone PCR mutagenesis is a rapid method to identify such key residues, lending itself to the scanning of 'lead' positions for further mutagenesis by other methods.

Oxidation of Crocein Orange G by lignin peroxidase isoenzymes. Kinetics and effect of H2O2.

The ligninolytic enzyme system of Phanerochaete chrysosporium is able to decolorize several recalcitrant dyes. Three lignin peroxidase isoenzymes, LiP 3.85, LiP 4.15, and LiP 4.65, were purified by preparative isoelectric focusing from the carbon-limited culture medium of P. chrysosporium. Based on amino terminal sequences, the purified isoenzymes correspond to the isoenzymes H8, H6, and H2, respectively, from the N-limited culture. The purified isoenzymes were used for decolorization of an azo dye, Crocein Orange G (COG). According to the kinetic data obtained, the oxidation of COG by lignin peroxidase appeared to follow Michaelis-Menten kinetics. Kinetic parameters for each isoenzyme were determined. The inactivating effect of ascending H2O2 concentrations on COG oxidation is shown to be exponential within the used concentration range. The best degree of decolorization of 100 microM COG was obtained when the H2O2 concentration was 150 microM. This was also the lowest H2O2 concentration for maximal decolorization of 100 microM COG, regardless of the amount of lignin peroxidase used in the reaction.

Folding of the polyketide chain is not dictated by minimal polyketide synthase in the biosynthesis of mithramycin and anthracycline.

BACKGROUND: Mithramycin, nogalamycin and aclacinomycins are aromatic polyketide antibiotics that exhibit antitumour activity. The precursors of these antibiotics are formed via a polyketide biosynthetic pathway in which acetate (for mithramycinone and nogalamycinone) or propionate (for aklavinone) is used as a starter unit and nine acetates are used as extender units. The assembly of building blocks is catalyzed by the minimal polyketide synthase (PKS). Further steps include regiospecific reductions (if any) and cyclization. In the biosynthesis of mithramycin, however, ketoreduction is omitted and the regiospecificity of the first cyclization differs from that of anthracycline antibiotics (e.g. nogalamycin and aclacinomycins). These significant differences provide a convenient means to analyze the determinants for the regiospecificity of the first cyclization step. RESULTS: In order to analyze a possible role of the minimal PKS in the regiospecificity of the first cyclization in polyketide biosynthesis, we expressed the mtm locus, which includes mithramycin minimal PKS genes, in Streptomyces galilaeus, which normally makes aclacinomycins, and the sno locus, which includes nogalamycin minimal PKS genes, in Streptomyces argillaceus, which normally makes mithramycin. The host strains are defective in the minimal PKS, but they express other antibiotic biosynthesis genes. Expression of the sno minimal PKS in the S. argillaceus polyketide-deficient strain generated mithramycin production. Auramycins, instead of aclacinomycins, accumulated in the recombinant S. galilaeus strains, suggesting that the mithramycin minimal PKS is responsible for the choice of starter unit. We also describe structural analysis of the compounds accumulated by a ketoreductase-deficient S. galilaeus mutant; spectroscopic studies on the major polyketide compound that accumulated revealed a first ring closure which is not typical of anthracyclines, suggesting an important role for the ketoreductase in the regiospecificity of the first cyclization. CONCLUSIONS: These experiments clearly support the involvement of ketoreductase and a cyclase in the regiospecific cyclization of the biosynthetic pathway for aromatic polyketides.

Isolation and characterization of 8-demethoxy steffimycins and generation of 2,8-demethoxy steffimycins in Streptomyces steffisburgensis by the nogalamycin biosynthesis genes.

Streptomyces steffisburgensis (NRRL 3193, ATCC 27466) is described as a steffimycin producer. Steffimycin belongs to the anthracycline group of aromatic polyketide antibiotics. The structural analysis of the products accumulated by the wild type ATCC 27466 strain revealed three different forms of 8-demethoxy steffimycin suggesting the loss of C-8 hydroxylation/methylation activity. In our approach to generate new anthracycline molecules, we used this strain as a host in gene cloning. The genes encoding the polyketide ketoreductase and aromatase enzymes of nogalamycin biosynthesis caused the production of 2-demethoxy steffimycins in S. steffisburgensis.

Cloning, sequencing, expression and characterization of three anti-estradiol-17beta Fab fragments.

In order provide data for a basic understanding of the mechanisms of antibody specificity and for the design of antibodies with desired properties, we have sequence-analysed three high affinity anti-estradiol-17beta monoclonal antibodies. All three monoclonal antibodies to estradiol-17beta had been raised by conjugation of the 6-carboxymethyloxime derivative to protein carrier. The genes encoding heavy (Fd) and light (L) chains of these three antibodies were cloned and sequenced. The sequenced antibody chains were found to be from 46.0 to 89.7% sequence identical to a monoclonal antibody (DB3) binding a related steroid, progesterone. The Fd and L chains were paired with all possible Fd-L combinations and the corresponding proteins were expressed in Escherichia coli and characterized for their binding (immunoreactivity) to estradiol-17beta. Under the lac promoter and using the pelB signal sequences the production levels of the soluble (total) heavy and light chain Fab fragment combinations in periplasm and in supernatant varied from 115 to 2207 microg/l, while the immunoreactivity percentages (IR%) varied from < 1 to 45%. The production levels and IR% were dependent on the first constant domain subclasses of the heavy chain as well as the Fd-L chain combination expressed.

Characterization of Streptomyces nogalater genes encoding enzymes involved in glycosylation steps in nogalamycin biosynthesis.

The sno gene cluster in Streptomyces nogalater ATCC 27451 contains the nogalamycin biosynthesis genes. A set of plasmid constructions carrying fragments of the sno cluster that lie downstream of snoD were used to complement the S. galilaeus mutant H039, which is blocked in rhodosamine and 2-deoxyfucose biosynthesis in the aclacinomycin pathway. Sequence analysis of this cluster revealed three contiguous open reading frames (ORFs) that were designated snoF, snoG, and snoH. Only those plasmid constructs that expressed SnoG were able to complement H039. SnoG shows similarity to GalE, a UDP-glucose-4-epimerase catalyzing the epimerization of UDP-glucose to UDP-galactose. The putative SnoF protein is similar to 3,5-epimerases involved in rhamnose biosynthesis. The deduced product of snoH is a 489-amino acid polypeptide. It is similar to the product of dau ORF3 found in the daunomycin cluster. However its function is still unclear. Based on the complementation experiments and sequence analysis, this part of the sno cluster is suggested to be involved in the biosynthesis of the sugar portion of nogalamycin. Interestingly, SnoA, a transcriptional activator for the sno minimal polyketide synthase, is also needed to express this cluster.

Production of hybrid anthracycline antibiotics by heterologous expression of Streptomyces nogalater nogalamycin biosynthesis genes.

A cluster of anthracycline biosynthetic genes isolated from Streptomyces nogalater was expressed in Streptomyces lividans and in Streptomyces galilaeus. A 12 kb DNA fragment cloned from this cluster in pIJ486 caused the production of a novel compound when introduced into S. lividans. The compound is derived from nogalonic acid methyl ester, an early intermediate in nogalamycin biosynthesis. Complementation with the cloned 12 kb fragment of S. galilaeus mutants blocked in aclacinomycin biosynthesis caused the production of hybrid anthracyclines. Cloning of the nogalamycin gene cluster should make possible a detailed study of the biosynthesis of this interesting antibiotic, as well as the production of novel anthracyclines of potential value as cytostatic drugs.

A gene cluster involved in nogalamycin biosynthesis from Streptomyces nogalater: sequence analysis and complementation of early-block mutations in the anthracycline pathway.

We have analyzed an anthracycline biosynthesis gene cluster from Streptomyces nogalater. Based on sequence analysis, a contiguous region of 11 kb is deduced to include genes for the early steps in anthracycline biosynthesis, a regulatory gene (snoA) promoting the expression of the biosynthetic genes, and at least one gene whose product might have a role in modification of the glycoside moiety. The three ORFs encoding a minimal polyketide synthase (PKS) are separated from the regulatory gene (snoA) by a comparatively AT-rich region (GC content 60%). Subfragments of the DNA region were transferred to Streptomyces galilaeus mutants blocked in aclacinomycin biosynthesis, and to a regulatory mutant of S. nogalater. The S. galilaeus mutants carrying the S. nogalater minimal PKS genes produced auramycinone glycosides, demonstrating replacement of the starter unit for polyketide biosynthesis. The product of snoA seems to be needed for expression of at least the genes for the minimal PKS.