Type II thioesterase from Streptomyces coelicolor A3(2)

Magdalena Kotowska1, Krzysztof Pawlik1, Andrew R. Butler2, Eric Cundliffe2, Eriko Takano3 and Katarzyna Kuczek1

Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Weigla 12, 53-114 Wroclaw, Poland1
Department of Biochemistry, University of Leicester, Leicester LE1 7RH, UK2
Department of Molecular Microbiology, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK3

Author for correspondence: Katarzyna Kuczek. Tel: +48 71 3732274. Fax: +48 71 373 2587. e-mail: kuczek{at}immuno.iitd.pan.wroc.pl


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Type I polyketide synthases (PKSs) are complexes of large, multimodular enzymes that catalyse biosynthesis of polyketide compounds via repetitive reaction sequences, during which each step is catalysed by a separate enzymic domain. Many type I PKSs, and also non-ribosomal peptide synthetase clusters, contain additional thioesterase genes located adjacent to PKS genes. These are discrete proteins called type II thioesterases (TE IIs) to distinguish them from chain-terminating thioesterase (TE I) domains that are usually fused to the terminal PKS module. A gene of a new TE II, scoT, associated with the cluster of putative type I PKS genes from Streptomyces coelicolor A3(2), was found. The deduced amino acid sequence of the gene product shows extensive similarity to other authentic thioesterase enzymes, including conservation of characteristic motifs and residues involved in catalysis. When expressed in the heterologous host Streptomyces fradiae, scoT successfully complemented the resident TE II gene (tylO), and, by restoring a significant level of macrolide production, proved to be catalytically equivalent to the TylO protein. S1 nuclease mapping of scoT revealed a single potential transcription start point with expression being switched on for a short period of time during a transition phase of growth.

Keywords: thioesterase type II, Streptomyces fradiae, disruption mutant complementation, S1 nuclease mapping

Abbreviations: PKS, polyketide synthase; TE, thioesterase

The GenBank accession number for the sequence reported in this paper is AF109727.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Polyketides are a large and structurally a diverse group of compounds, many of which exhibit biological activity as antibiotics, immunosuppressants or anticancer drugs. Polyketides are synthesized by a common mode of condensation and reduction reactions similar to that of fatty acid biosynthesis. Nascent polyketides are processed by large multienzyme complexes, polyketide synthases (PKSs). In the type I PKSs, involved in production of macrolide antibiotics such as erythromycin, reactions of the biosynthetic cycle are catalysed sequentially by separate enzymic domains housed in large multifunctional polypeptides. Each complete cycle of condensation and reduction reactions is catalysed by a module, a functional unit of the PKS. The substrate acyl chains which undergo successive reactions are tethered as thioesters by acyl carrier domains of the PKS polypeptides. A terminal thioesterase domain (TE) catalyses release and cyclization of the full-length (fully processed) polyketide chain (Katz & Donadio, 1993 ).

Many type I PKSs, and also non-ribosomal peptide synthetase clusters, contain additional TE genes located adjacent to the PKS genes within the cluster of antibiotic biosynthetic genes (Weissman et al., 1998 ; Schneider & Marahiel, 1998 ; Shaw-Reid et al., 1999 ; August et al., 1998 ; Xue et al., 1998 ; Heathcote et al., 2001 ). The products of such genes are discrete proteins called type II thioesterases (TE IIs) to distinguish them from chain-terminating thioesterase (TE I) domains (Gokhale et al., 1999 ). In fatty acid synthase complexes, TE IIs are alternative chain-terminating enzymes exhibiting hydrolase activity towards medium-chain-length acyl thioesters (Smith, 1994 ). There is no obvious role for the TE IIs associated with multienzyme PKS complexes, nor is the mechanism of their action known.

The function of TE IIs is predicted from gene-disruption analysis, complementation studies, and determination of their substrate specificities (Weissman et al., 1998 ; Butler et al., 1999 ; Heathcote et al., 2001 ). Polyketide production is drastically reduced, by 90% or more, in strains with a deleted TE II gene (Xue et al., 1998 ; Butler et al., 1999 ; Doi-Katayama et al., 2000 ), indicating an important function, proposed to involve editing of aberrant intermediates (Butler et al., 1999 ) during the course of polyketide biosynthesis. More recently, this hypothesis has been confirmed and the mechanism clarified. Thus, the TylO protein displayed hydrolytic activity in vitro towards short-chain acyl-CoAs, indicating that the enzyme could remove aberrantly decarboxylated (and, therefore, non-reactive) extender acyl chains from the PKS during polyketide biosynthesis (Heathcote et al., 2001 ). By hydrolytic release of such aberrant acyl groups, TE II was proposed to unblock PKS modules and restore overall efficiency of the complex enzyme.

The proposed common role for TE IIs in PKS multienzyme systems raises the question of whether specific TE IIs might be replaceable by other TE IIs normally associated with other PKS complexes. Based on the TE II substrate-specificity studies published to date (Weissman et al., 1998 ; Heathcote et al., 2001 ), such enzymes do not seem to select the structure of the acyl substrates.

Genetic engineering studies allow assembly of novel polyketide chains following fusion, swapping or repositioning of catalytic domains, modules or whole peptides within PKS polypeptides (Hutchinson & Fujii, 1995 ; Ranganathan et al., 1999 ; Tang et al., 2000 ). In engineered PKSs, co-expression of TE IIs in addition to other PKS proteins might help in achieving elevated levels of the polyketide products.

We were studying a gene cluster for the new PKS type I in Streptomyces coelicolor A3(2) (Kuczek et al., 1997 ; K. Pawlik, M. Kotowska & K. Kuczek, unpublished data; GenBank accession numbers U88833, AF109727 and AF 202898) located in a previously unmapped region of the chromosome, between cosmids 2H4 and 10H5 (Redenbach et al., 1996 ), now covered by cosmids 2C4, 1G7, BAC8D1 and IF3 (http://www.sanger.ac.uk/Projects/S_coelicolor/). During our studies, we found a new gene, encoding a putative TE II (ScoT). The gene is located within the putative polyketide biosynthetic gene cluster. In this paper, we describe this gene. When expressed in the heterologous host, Streptomyces fradiae, scoT functionally complemented disruption of the native TE II gene, tylO.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, cosmid clones, plasmids and growth media.
The Streptomyces strains used are listed in Table 1. S. fradiae and its mutants were maintained and propagated at 37 °C on AS1 agar (Wilson & Cundliffe, 1998 ) or at 30 °C in Difco tryptic soy broth. S. coelicolor A3(2) was maintained and propagated at 30 °C on 79 agar medium [IMET Catalogue of Strains (Jena) 1987] or SMM agar (Kieser et al., 2000 ). Escherichia coli SURE strain [e14-(McrA-) {Delta}(mcrCB–hsdSMR–mrr)171 endA1 supE44 thi-1 gyrA96 relA1 lac recB recJ sbcC umuC::Tn5 (Kanr) uvrC (F' proAB lacIqZ {Delta}M15 Tn10 (Tetr)] was a host for cosmid clone 1G7 of S. coelicolor A3(2) genomic DNA in SuperCos vector (Stratagene), which was kindly provided by H. M. Kieser (John Innes Institute, Norwich, UK). pBluescript SK(+) and SK(-) phagemid vectors (Stratagene) were used for cloning of restriction fragments from the cosmid and for their sequencing. E. coli K-12 strain DH5 {alpha} was used for routine plasmid manipulation and was grown in standard media (Sambrook et al., 1989 ). DNA was introduced into S. fradiae via conjugal transfer from E. coli S17-1 (Simon et al., 1983 ).


View this table:
[in this window]
[in a new window]
 
Table 1. Streptomyces strains used in this work

 
DNA and RNA isolation, hybridization and sequencing.
Plasmids propagated in E. coli were isolated according to standard procedures (Sambrook et al., 1989 ). Cosmid DNA was purified by the procedure for high-molecular-mass plasmid DNA preparation and was used for restriction analysis and cloning of scoT by standard genetic procedures (Sambrook et al., 1989 ). The hybridization probe was random-prime labelled with DIG-11-dUTP. Labelling, hybridization (at 68 °C) and detection of hybrids were carried out according to the manufacturer’s recommended procedure (The DIG System User’s Guide for Filter Hybridization; Boehringer) with CSPD (Tropix) as a substrate for the luminescent reaction. The sequence of the S. coelicolor DNA was determined both manually (Kuczek et al., 1998 ), by using the chain-termination method with the Sequenase version 2.0 sequencing kit of Amersham, and by automated sequencing (performed by Qiagen Sequencing Services, Hilden, Germany). The sequence was determined on both strands and was submitted to GenBank (accession no. AF109727). Comparisons of the nucleotide and amino acid sequences with the databases were performed with the BLAST and CLUSTAL W programs. RNA was isolated as described by Strauch et al. (1991 ).

Construction of a recombinant plasmid for complementation analysis.
An S. fradiae tylO-disruption strain was complemented with cloned S. coelicolor A3(2) DNA containing scoT, using the conjugative vector pLST9828 (Butler et al., 1999 ). This is a derivative of pSET152 (Bierman et al., 1992 ) that contains the powerful constitutive promoter ermEp* to drive expression of inserted DNA fragments following integration into the chromosomal {phi}C31 attB site. A DNA fragment containing scoT was ligated into pLST9828 in a two-step process. First, the DNA containing the C-terminal part of ScoT was PCR-amplified from the pBSK(-) plasmid clone template (Fig. 1c) using M13 reverse primer and the TE-Rev primer with an engineered XbaI site (underlined): 5'-TTTTCTAGATGTCGTACGTACACGGA-3'. The PCR product was purified using the Qiaex II DNA purification kit, digested, and ligated into pLST9828 using the BamHI and XbaI sites. Then, a second PCR product containing the N-terminal part of ScoT was obtained from the template of another pBSK(+) construct (Fig. 1b) using the T7 universal primer and the TE-Fw primer with an engineered BamHI site (underlined): 5'-TTTTTTGGATCCGATGGGAAGTGACTGGTT-3'. The 50 µl reaction mixture contained 5 µl 10x PCR DyNAzyme buffer (Finnzymes), 1 µl 10 mM deoxynucleoside triphosphate mixture, 50 pmol each oligonucleotide, about 10 ng template DNA and 1 µl DyNAzyme II DNA polymerase (Finnzymes). Cycling was as follows: a hot start at 96 °C for 6 min, 1 min at 80 °C (adding of the enzyme), 31 cycles with denaturation at 95 °C for 1 min, annealing at 63–65 °C for 1 min and extension at 72 °C for 1·5 min, followed by a final extension at 72 °C for 5 min. The product was digested with BamHI and ligated into the pLST9828 derivative obtained in the first step of the cloning procedure. Gentamicin (15 µg ml-1) was used for the selection of E. coli DH5{alpha} transformants. The authenticity and orientation of the cloned fragments were confirmed by automated sequence analysis. The gene cloned in pLST9828 was introduced into S. fradiae C373.1, a tylosin-producing strain, by transconjugation from E. coli S17-1 as described elsewhere (Butler et al., 1999 ).



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 1. Cloning strategy of scoT and its fragments in different vectors. (a) Restriction map of the T3-terminal fragment of cosmid 1G7 containing the scoT gene; scbR2 denotes another ORF identified upstream of scoT. (b, c) Clones containing fragments of scoT encoding the N- and C-terminal parts of TE II, respectively. Arrows indicate primers used for PCR amplification of the respective fragments for cloning into the pLST9828 vector. (d) The recombinant plasmid pLST9828 containing the entire sequence of scoT for complementation study.

 
Fermentation and analysis of its products.
Strains of S. fradiae were fermented in media described elsewhere, and fermentation products were extracted and analysed by reverse-phase HPLC, with absorbance measurement at 282 nm, also as described elsewhere (Butler et al., 1999 ). Desmycosin was used as an internal standard to identify fermentation products in tylO-disrupted strains.

S1 nuclease mapping.
For each S1 nuclease reaction, 30 or 40 µg RNA was hybridized in NaTCA buffer [Murray, 1986 ; solid NaTCA (Aldrich) dissolved to 3 M final concentration in 50 mM PIPES, 5 mM EDTA, pH 7·0) to about 0·002 pmol (approx. 104 Cerenkov c.p.m.] of the probe. The oligonucleotide 5'-GTCGAACTCGGGCGTCAGCTC-3' was uniquely labelled at its 5' end with [32P]ATP using T4 polynucleotide kinase, and was used in the PCR with the unlabelled oligonucleotide 5'-CCTCGGCGGCGGAGAGAAT-3', which anneals upstream of the scoT promoter, to generate a 430 bp probe. The PCR used M145 total DNA as a template. Subsequent steps were as described by Strauch et al. (1991 ).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
TE II gene scoT
We investigated cosmid 1G7 from the library of chromosomal DNA of S. coelicolor A3(2). The cosmid is located between cosmids 2C4 and BAC8D1 (http://www.sanger.ac.uk/Projects/S_coelicolor/). DNA of cosmid 1G7, digested with DraI and BamHI, was screened with a probe having a sequence characteristic of acyltransferase active sites (Kuczek et al., 1997 ). Under non-stringent conditions, six fragments were found to hybridize with the probe. These hybridizing fragments were cloned for sequence determination which, with the aid of CODONPREFERENCE analysis, revealed an ORF of 807 bp, located on two fragments and ending 265 bp from the T3 end of the 1G7 cosmid (Fig. 1a–c). The ORF, designated scoT, was deduced to encode a protein of 268 amino acid residues (molecular mass 28686 Da; pI 6·17), of which about 53% were predicted to be hydrophobic.

Comparison of the scoT sequence (GenBank AF109727) with others in the databases revealed extensive similarities with TEs from various actinomycetes and other bacteria, and also with rat S-acyl fatty acid synthase complex. The greatest similarity was found with TE IIs (Pfam00975), i.e. 43% identity with a TE II (AF040570) from the rifamycin biosynthetic gene cluster of Amycolatopsis mediterranei, 43% identity with a TE II (X60379) associated with 6-deoxyerythronolide B synthase from Saccharopolyspora erythraea, and 40% identity with TylO (U08223), the TE II involved in tylosin biosynthesis in Streptomyces fradiae. ScoT is predicted to belong to the well-known alpha/beta hydrolase family (Pfam00975).

The ATG start codon of the scoT gene is preceded by a potential RBS sequence (AAGGGG) ending 8 bp before the start (nucleotides complementary to the 3' end of the 16S rRNA from Streptomyces lividans are underlined) (Strohl, 1992 ) (see Fig. 4b). The ORF ends with a TGA stop codon. The amino acid motif GxSxG (x=any amino acid) characteristic of acyltransferases and TEs, with Ser-90 as the active-site residue, is present within the deduced sequence of the ScoT protein. A second conserved amino acid which might also be involved in catalysis is His-224 (Fig. 2). Usage of AGC as the codon for the active-site serine is typical of TE IIs [AGY for TE II and TCN for TE I, where Y=(C, T), N=(A, C, G, T)] (Smith, 1994 ).



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 2. Alignment of the amino acid sequence of the active-site region of the putative thioesterase II from S. coelicolor A3(2) (‘scoT’) with another TE: ‘amy’, TE II from the rifamycin biosynthetic gene cluster (AF040571) of A. mediterranei; ‘sac’, TE (X60379) from Saccharopolyspora erythraea; ‘sfr’, TE II (U08223) from the tylosin biosynthetic gene cluster of S. fradiae. Amino acids identical to those encoded by scoT are shown in bold. The conservative motif of the TE catalytic centre and the conservative histidine approx. 100 amino acids downstream are underlined; the active-site serine is indicated by an asterisk.

 
Biological activity of TE II
We attempted to determine whether the product of scoT has TE II activity. This was examined by studying whether it could functionally replace the native TE II (i.e. TylO) in the tylosin producer, S. fradiae. ScoT and TylO show extensive amino acid sequence similarity, and the complementation system for the TE II gene-disrupted S. fradiae strain was constructed (Butler et al., 1999 ). Therefore, scoT was cloned in a conjugative expression vector, pLST9828 (Fig. 1d), and integrated into the chromosome, at the attB site, of a tylO-disrupted strain of S. fradiae under the control of the strong, constitutive promoter ermEp* (Butler et al., 1999 ). This was done via conjugal transfer from E. coli, and three resultant transconjugant strains were subjected to fermentation analysis using quadruplicate cultures of each strain. Since TylO is co-expressed with the downstream gene, tylCVI (which encodes an enzyme involved in synthesis of the deoxyhexose sugar mycarose), disruption of tylO results in a polar effect on tylCVI expression, such that the tylO-disrupted strain produces demycarosyl-tylosin (desmycosin), as do strains in which the disruption is successfully complemented by cloned DNA (Butler et al., 1999 ).

Thus, the tylO disruptant was complemented by scoT, as measured by the yield of desmycosin in fermentations analysed by HPLC with desmycosin as the internal standard (Fig. 3a–c). Desmycosin production was restored to up to 48% of the level of macrolide produced by the wild-type strain (Fig. 3c). Because of the close similarity in the molar absorbance coefficients of tylosin and desmycosin, the amount of desmycosin produced as a fermentation product by a mutant strain was directly compared with the amount of tylosin produced by the wild-type strain. Control fermentation of the non-complemented, tylO-disrupted strain yielded only minimal amounts of desmycosin (Fig. 3b). These results showed that the TE II gene, scoT, from S. coelicolor A3(2) could, by complementation, restore macrolide production to a significant level in the tylO-disrupted strain of S. fradiae.



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 3. HPLC analysis of fermentation products of S. fradiae. (a) Wild-type; (b) tylO-disrupted strain; (c) tylO-disrupted strain, complemented with scoT.

 
S1 nuclease protection experiments with scoT transcription
We examined if scoT undergoes transcription in the cells of its native host, S. coelicolor A3(2). Expression of scoT was analysed at different growth stages, using S1 nuclease protection analysis of RNA extracted from S. coelicolor A3(2) M145 cultures grown in SMM liquid medium. Transcription start sites were identified 77–78 nt upstream of the translation start site of scoT (Fig. 4a). The apparent promoter -35 and -10 hexamer sequences are TGTCCA and CCTGGT, respectively, separated by a spacer region of 17 nt. The -10 hexamer sequence is separated from the +1 region by 6 nt. There is little similarity between the -35 and -10 regions of scoT and well-defined promoter regions with consensus sequences TTGAC(Pu) for the -35 region and TAG(Pu)(Pu)T for the -10 region (Strohl, 1992 ) (Fig. 4b).



View larger version (31K):
[in this window]
[in a new window]
 
Fig. 4. S1 nuclease transcript mapping of scoT. (a) S1 nuclease mapping of the transcriptional start sites of scoT. A 430 bp PCR product (nt) labelled uniquely at the 5' end position was used as a probe for scoT mRNAs. Asterisks indicate the probable transcription start points; the sequences shown are those of the template strand. Lanes T, G, C and A are sequence ladders derived from the same labelled primer that was used to generate the PCR product. (b) The promoter region of scoT. Possible -10- and -35-region sequences are underlined. A possible RBS is shown shaded. Transcriptional start points are indicated by asterisks, and the direction of transcription is indicated by an arrow. The translation start codon is shown boxed. The numbers are nucleotide positions as denoted in GenBank accession no. AF109727. (c) Analysis of the scoT transcript, using RNA isolated from a liquid SMM-grown culture of S. coelicolor A3(2) strain M145. ‘EXP’, ‘TRANSITION’ and ‘STAT’ indicate the exponential, transition and stationary phases of growth, respectively. The shaded box labelled ‘RED’ denotes the presence of undecylprodigiosin in the mycelium.

 
Transcription of scoT was hardly detectable during exponential growth and early transition phase. The expression was induced at the 19th hour of growth, i.e. in late transition phase, and was quickly shut down in stationary phase (Fig. 4c).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
ScoT is TE II protein
The translated sequence of scoT displayed extensive similarity to the deduced sequences of TE IIs associated with type I PKS complexes. Such similarity extended to several characteristics: (1) the overall amino acid sequence similarity; (2) the conservation of characteristic amino acid motifs; (3) the molecular mass of the peptide; (4) the distance from the amino terminus of the protein to the active-site serine (usually about 100 residues); (5) the position of the conserved histidine residue; (6) active-site serine codon AGC (Smith, 1994 ).

The amino acid sequences of all the known TE IIs are strongly conserved, especially in implicated catalytically active sites. The sequence data suggest that scoT encodes a TE II which, presumably, is catalytically active. However, in the absence of data relating to its substrate specificity, the probability that ScoT might be capable of functioning in association with a heterologous PKS system could not be predicted.

ScoT complements the natural function of TE II in S. fradiae
ScoT was able to replace TylO, the TE II of the tylosin pathway, by restoring the efficiency of macrolide biosynthesis in the tylO-disrupted strain. The two enzymes appear, therefore, to be equivalent in their catalytic function. Interestingly, in their natural context, these enzymes are apparently associated with PKS enzymes that differ in module number and in the nature of the extender acyl units incorporated by the respective modules. The primary product of the tylosin PKS in S. fradiae, a linear polyketide derived from condensation of five propionate units, two acetate units and one butyrate unit (Baltz & Seno, 1988 ), undergoes cyclization into a 16-atom lactone ring, tylactone. This means that in tylactone biosynthesis, malonyl-CoA, methylmalonyl-CoA and ethylmalonyl-CoA are used as extender units. Although the polyketide product of the modular PKS with which ScoT putatively associates has not yet been determined, the sequences of the acyltransferase domains in each of the five extension modules of the PKS (Kuczek et al., 1997 ; K. Pawlik, M. Kotowska & K. Kuczek, unpublished data) suggest that all five should use malonyl-CoA extender units according to the consensus sequence motifs correlated with the substrate specificity of the acyltransferase domains (Haydock et al., 1995 ).

In any event, if ScoT is active as an editing enzyme with both the tylosin PKS and its natural PKS partner, it must be able to hydrolyse thioester bonds irrespective of the length of the extender units employed or the size of the nascent polyketide intermediates, since ScoT can apparently cooperate in the biosynthesis of a product longer than that of its native PKS partner.

We suggest that a mechanism other than substrate selectivity, probably based on differences in kinetic rates of hydrolysis by ScoT and polyketide condensation, is involved in competition between these two reactions. To clarify this, however, further studies which involve kinetic measurements of the enzyme activity are needed.

Transcription of scoT occurs in a growth-phase-dependent manner
S1 nuclease mapping of the scoT promoter region in S. coelicolor A3(2) showed that transcription was growth-phase dependent. It was detectable only during late transition phase, indicating the operation of a regulatory system that prevents expression of the gene throughout most of the growth cycle. Since production of secondary metabolites by Streptomyces is commonly initiated during transition phase, onset of scoT expression might be correlated with expression of the gene cluster encoding type I PKS, located close to scoT. Our results on this will be published in due course.


   ACKNOWLEDGEMENTS
 
The authors are grateful to Professor D. A. Hopwood and Dr H. Kieser for providing cosmid clone 1G7 and S. coelicolor A3(2) strain M145. This work was partly supported by The Committee of Scientific Research of Poland (under project 6 P04B 025 16) and by The Royal Society Short Term Visit Program. E.T. was supported by a grant from the Human Frontiers Science Program (grant RG0330/1998-M).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
August, P. R., Tang, L., Yoon, Y. J. & 9 other authors. (1998). Biosynthesis of the ansamycin antibiotic rifamycin: deductions from the molecular analysis of the rif biosynthetic gene cluster of Amycolatopsis mediterranei S699. Chem Biol 5, 69–79.[Medline]

Baltz, R. & Seno, E. T. (1988). Genetics of Streptomyces fradiae and tylosin biosynthesis. Annu Rev Microbiol 42, 547-574.[Medline]

Bierman, M., Logan, R., O’Brien, K., Seno, E. T., Rao, R. N. & Schoner, B. E. (1992). Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces spp. Gene 116, 43-49.[Medline]

Butler, A. R., Bate, N. & Cundliffe, E. (1999). Impact of thioesterase activity on tylosin biosynthesis in Streptomyces fradiae. Chem Biol 6, 287-292.[Medline]

Doi-Katayama, Y., Yoon, Y. J., Choi, C.-Y., Yu, T-W., Floss, H. G. & Hutchinson, R. (2000). Thioesterases and the premature termination of polyketide chain elongation in rifamycin B biosynthesis by Amycolatopsis mediteranei S699. J Antibiot 53, 484-495.[Medline]

Gokhale, R. S., Hunziker, D., Cane, D. E. & Khosla, C. (1999). Mechanism and specificity of the terminal thioesterase domain from the erythromycin polyketide synthase. Chem Biol 6, 117-125.[Medline]

Haydock, S. F., Aparicio, J. F., Molnar, I. & 7 other authors (1995). Divergent sequence motifs correlated with the substrate specificity of (methyl)malonyl-CoA: acyl carrier protein transacylase domains in modular polyketide synthases. FEBS Lett 374, 246–248.[Medline]

Heathcote, M. L., Staunton, J. & Leadlay, P. F. (2001). Role of type II thioesterases: evidence from removal of short acyl chains produced by aberrant decarboxylation of chain extender units. Chem Biol 8, 207-220.[Medline]

Hopwood, D. A., Bibb, M. J., Chater, K. F. & 7 other authors (1985). Genetic Manipulation of Streptomyces: a Laboratory Manual. Norwich: John Innes Foundation.

Hutchinson, C. R. & Fujii, I. (1995). Polyketide synthase gene manipulation: a structure-function approach in engineering novel antibiotics. Annu Rev Microbiol 49, 201-238.[Medline]

Katz, L. & Donadio, S. (1993). Polyketide synthesis: prospects for hybrid antibiotics. Annu Rev Microbiol 47, 875-912.[Medline]

Kieser, T., Bibb, M. J., Buttner, M. J., Chater, K. F. & Hopwood, D. A. (2000). Practical Streptomyces genetics. Norwich: John Innes Foundation.

Kuczek, K., Pawlik, K., Kotowska, M. & Mordarski, M. (1997). Streptomyces coelicolor DNA homologous with acyltransferase domains of type I polyketide synthase gene complex. FEMS Microbiol Lett 157, 195-200.[Medline]

Kuczek, K., Kotowska, M., Wiernik, D. & Mordarski, M. (1998). Single-stranded DNA production from phagemids containing GC-rich DNA fragments. BioTechniques 24, 214-215.[Medline]

Murray, M. G. (1986). Use of sodium trichloroacetate and mung bean nuclease to increase sensitivity and precision during transcript mapping. Anal Biochem 158, 165-170.[Medline]

Ranganathan, A., Timoney, M., Bycroft, M. & 8 other authors (1999). Knowledge-based design of bimodular and trimodular polyketide synthases based on domain and module swaps: a route to simple statin analogues. Chem Biol 6, 731–741.[Medline]

Redenbach, M., Kieser, H. M., Denapaite, D., Eichner, A., Cullum, J., Kinashi, H. & Hopwood, D. A. (1996). A set of ordered cosmids and a detailed genetic and physical map for the 8 Mb Streptomyces coelicolor A3(2) chromosome. Mol Microbiol 21, 77-96.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schneider, A. & Marahiel, M. A. (1998). Genetic evidence for a role of thioesterase domains, integrated in or associated with peptide synthetases, in non-ribosomal peptide biosynthesis in Bacillus subtilis. Arch Microbiol 169, 404-410.[Medline]

Seno, E. T. & Baltz, R. H. (1982). S-Adenosyl-L-methionine:macrocin O-methyl-transferase activities in a series of Streptomyces fradiae mutants that produce different levels of the macrolide antibiotic tylosin. Antimicrob Agents Chemother 21, 758-763.[Medline]

Shaw-Reid, C. A., Kelleher, N. L., Losey, H. C., Gehring, A. M., Berg, C. & Walsh, C. T. (1999). Assembly line enzymology by multimodulator nonribosomal peptide synthetases: the thioesterase domain of E. coli EntF catalyses both elongation and cyclolactonization. Chem Biol 6, 385-400.[Medline]

Simon, R., Priefer, U. & Pühler, A. (1983). A broad host range mobilisation system for in vivo genetic engineering: transposon mutagenesis in gram-negative bacteria. Biotechnology 1, 784-791.

Smith, S. (1994). The animal fatty acid synthase: one gene, one polypeptide, seven enzymes. FASEB J 8, 1248-1259.[Abstract/Free Full Text]

Strauch, E., Takano, E., Baylis, H. A. & Bibb, M. J. (1991). The stringent response in Streptomyces coelicolor A3(2). Mol Microbiol 5, 289-298.[Medline]

Strohl, R. (1992). Compilation and analysis of DNA sequences associated with apparent streptomycete promoters. Nucleic Acids Res 20, 961-974.[Abstract]

Tang, L., Fu, H. & McDaniel, R. (2000). Formation of functional heterologous complexes using subunits from the picromycin, erythromycin and oleandomycin polyketide synthases. Chem Biol 7, 77-84.[Medline]

Weissman, K. J., Cameron, J. S., Hanefeld, U., Aggarwal, R., Bycroft, M., Staunton, J. & Leadlay, P. F. (1998). The thioesterase of the erythromycin-producing polyketide synthase: influence of acyl chain structure on the mode of release of substrate analogues from the acyl enzyme intermediates. Chem Int Ed 37, 1437-1440.

Wilson, V. T. & Cundliffe, E. (1998). Characterization and targeted disruption of a glycosyltransferase gene in the tylosin producer, Streptomyces fradiae. Gene 214, 95-100.[Medline]

Xue, Y., Zhao, L., Liu, H. W. & Sherman, D. H. (1998). A gene cluster for the macrolide antibiotic biosynthesis in Streptomyces venezuelae: architecture of metabolic diversity. Proc Natl Acad Sci USA 95, 12111-12116.[Abstract/Free Full Text]

Received 9 November 2001; revised 28 January 2002; accepted 11 February 2002.