Genes encoding acyl-CoA dehydrogenase (AcdH) homologues from Streptomyces coelicolor and Streptomyces avermitilis provide insights into the metabolism of small branched-chain fatty acids and macrolide antibiotic production

Ying-Xin Zhang1, Claudio D. Denoya3, Deborah D. Skinner3, Ronald W. Fedechko3, Hamish A. I. McArthur3, Margaret R. Morgenstern3, Richard A. Davies3, Sandra Lobo4, Kevin A. Reynolds4 and C. Richard Hutchinson1,2

School of Pharmacy1 and Department of Bacteriology2, University of Wisconsin, 425 N. Charter St, Madison, WI 53706, USA
Bioprocess Research, Central Research Division, Pfizer Inc., Groton, CT 06340, USA3
Department of Medicinal Chemistry, School of Pharmacy and Institute for Structural Biology and Drug Discovery, Virginia Commonwealth University, Richmond, VA 23219, USA4

Author for correspondence: C. Richard Hutchinson. Tel: +1 608 262 7582. Fax: +1 608 262 3134. e-mail: crhutchi{at}facstaff.wisc.edu


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The cloning, using a PCR approach, of genes from both Streptomyces coelicolor and Streptomyces avermitilis encoding an acyl-CoA dehydrogenase (AcdH), putatively involved in the catabolism of branched-chain amino acids, is reported. The deduced amino acid sequences of both genes have a high similarity to prokaryotic and eukaryotic short-chain acyl-CoA dehydrogenases. When the S. coelicolor and S. avermitilis acyl-CoA dehydrogenase genes (acdH) were expressed in Escherichia coli, each of the AcdH flavoproteins was able to oxidize the branched-chain acyl-CoA derivatives isobutyryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA, as well as the short straight-chain acyl-CoAs n-butyryl-CoA and n-valeryl-CoA in vitro. NMR spectral data confirmed that the oxidized product of isobutyryl-CoA is methacrylyl-CoA, which is the expected product at the acyl-CoA dehydrogenase step in the catabolism of valine in streptomycetes. Disruption of the S. avermitilis acdH produced a mutant unable to grow on solid minimal medium containing valine, isoleucine or leucine as sole carbon sources. Feeding studies with 13C triple-labelled isobutyrate revealed a significant decrease in the incorporation of label into the methylmalonyl-CoA-derived positions of avermectin in the acdH mutant. In contrast the mutation did not affect incorporation into the malonyl-CoA-derived positions of avermectin. These results are consistent with the acdH gene encoding an acyl-CoA dehydrogenase with a broad substrate specificity that has a role in the catabolism of branched-chain amino acids in S. coelicolor and S. avermitilis.

Keywords: avermectin, branched-chain amino acids, isotope labelling

Abbreviations: AcdH, the specific acyl-CoA dehydrogenases produced by S. avermitilis and S. coelicolor; AD, any other acyl-CoA dehydrogenase; PMS, phenazine methosulfate

The GenBank accession numbers for the sequences described in this paper are AF142581 (Streptomyces coelicolor) and AF143210 (Streptomyces avermitilis).


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We are interested in elucidating the pathways of isoleucine, leucine and valine catabolism in streptomycetes and characterizing the genes involved in the catabolic route leading to methylmalonyl-CoA and propionyl-CoA (Fig. 1), because our previous work (Tang et al., 1994 ; Denoya et al., 1995 ; Skinner et al., 1995 ; Zhang et al., 1996 ) and several other studies (Omura et al., 1983 ; Sherman et al., 1986 ; Reynolds et al., 1988 ; Sood et al., 1988 ; Lebrihi et al., 1992 ) have shown that valine and isoleucine catabolism plays an important role in supplying fatty acid precursors for macrolide and polyether antibiotic formation under certain growth conditions. The msdA gene encoding methylmalonic acid semialdehyde dehydrogenase, an enzyme involved in the later stage of valine catabolism which catalyses the direct conversion of methylmalonic acid semialdehyde to propionyl-CoA, has been cloned and characterized in streptomycetes (Zhang et al., 1996 ), pseudomonads (Steele et al., 1992 ) and mammals (Goodwin et al., 1989 ). Genes encoding the valine (branched-chain amino acid) dehydrogenase (vdh) and branched-chain 2-keto acid dehydrogenase (bkd), which comprise a common pathway catalysing the conversion of valine, isoleucine and leucine to their respective acyl-CoA derivatives, have been cloned and characterized in Streptomyces spp. (Tang et al., 1994 ; Denoya et al., 1995 ; Skinner et al., 1995 ; Tang & Hutchinson, 1993 ) as well as in other microbes and higher animals (Massey et al., 1976 ; Wolf & Akers, 1986 ; Sykes et al., 1987 ; Wexler et al., 1991 ). The acyl-CoA metabolites formed by the common pathway in pseudomonads are catabolized by three separate series of enzymes, one specific for each branched-chain acyl-CoA (Massey et al., 1976 ). The third reaction in the oxidation of branched-chain amino acids is catalysed by an acyl-CoA dehydrogenase (AD), which constitutes the first enzyme after the common pathway. In eukaryotes, two ADs act at this step of the catabolic route. These are isovaleryl-CoA dehydrogenase in leucine metabolism and 2-methyl-branched-chain acyl-CoA dehydrogenase in isoleucine/valine metabolism (Ikeda et al., 1983 ). Although suggested by some data, there has been insufficient evidence to make a categorical assignment of ADs to the common catabolic pathway in bacteria (Massey et al., 1976 ). ADs all catalyse {alpha},ß-dehydrogenation of acyl-CoA esters and transfer electrons to an electron transfer flavoprotein via the same mechanism, but the length and configuration of the hydrocarbon chain of the substrate are distinctly different for each enzyme (Ikeda et al., 1983 ).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 1. Pathways for catabolism of branched-chain amino acids in bacteria. The pathways may diverge along routes A, B or C, after formation of acyl-CoAs by the Bkd enzymes. The symbols on isobutyryl-CoA and its catabolites indicate the predicted relationships between carbons retained or lost during the degradation.

 
As an extension of our earlier work on the vdh (Tang & Hutchinson, 1993 ; Tang et al., 1994 ), bkd (Denoya et al., 1995 ; Skinner et al., 1995 ) and msdA (Zhang et al., 1996 ) genes, a PCR method was used to look for an AD gene in Streptomyces coelicolor and Streptomyces avermitilis. Here, we report the cloning and nucleotide sequencing of acyl-CoA dehydrogenase genes (acdH) from these species. In addition, the genes were expressed in Escherichia coli. The partially purified AcdH proteins were able to oxidize isobutyryl-CoA to methacrylyl-CoA (see Fig. 1), and oxidize n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA, most likely to their respective 2,3-unsaturated derivatives, in the presence of FAD and phenazine methosulfate (PMS). We also describe an S. avermitilis acdH mutant constructed by insertional inactivation. The disruptant is unable to grow on solid minimal medium containing valine, isoleucine or leucine as a sole carbon source and appears from incorporation experiments using labelled isobutyrate to have a decreased ability to process isobutyryl-CoA via methacrylyl-CoA to methylmalonyl-CoA.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
General.
Yeast S-acetyl-CoA synthetase, yeast CoA, methacrylic acid, isobutyryl-CoA, n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA, PMS, FAD and ATP were purchased from Sigma. Cyclohexylcarbonyl-CoA and 1-cyclohexenylcarbonyl-CoA were prepared as described previously (Reynolds et al., 1992 ). Sodium [2,3,3'-13C3]isobutyrate was prepared from [2-13C]diethylmalonate and [13C]iodomethane (Sigma) following a published protocol (Reynolds et al., 1988 ). A Waters HPLC instrument with a model 501 pump system and a Waters 484 variable-wavelength absorbance detector was used to purify the methacrylyl-CoA. Sephadex G-10 from Pharmacia was used for desalting. Proton NMR spectra were recorded on a Bruker AM300 spectrometer in 2H2O. Carbon NMR spectra were recorded on a Bruker AVANCE 500 MHz NMR spectrometer in CDCl3.

Escherichia coli DH5{alpha} (Sambrook et al., 1989 ) was obtained from D. MacNeil (Merck & Co. Research Laboratories), S. coelicolor J802 (dagA1 agaA7) (Hodgson & Chater, 1981 ) from David Hodgson (University of Warwick, UK) and Streptomyces lividans TK64 (Hopwood et al., 1985 ) from David Hopwood (John Innes Centre, Norwich, UK). Plasmid pT7-7 (Tabor, 1990 ) was obtained from Stan Tabor (Harvard University, USA). E. coli DH5{alpha} and BL21(DE3) (Novagen) were grown at 37 °C on Luria–Bertani (LB) medium (Sambrook et al., 1989 ). S. coelicolor J802 was grown on R2YE agar (Hopwood et al., 1985 ) at 30 °C for general use. S. avermitilis ATCC 31272 (Burg et al., 1979 ) was grown as described previously (Denoya et al., 1994 , 1996 ). Solid minimal medium was as described by Hopwood et al. (1985 ), except that SeaKem ME agarose (FMC BioProducts) replaced agar, and the glucose in the medium was replaced with one or more carbon sources at a final concentration of 2·5 g l-1 each.

Oligodeoxynucleotides for PCR and sequencing primers were synthesized with a Perkin-Elmer ABI model 3948 DNA synthesizer/purifier, or purchased from Genosys. EasyStart 100 (PCR Mix-in-a-Tube, Molecular Bio-Products) was used for PCR reactions. Taq polymerase was purchased from Sigma. PCR reactions were carried out in a Perkin-Elmer Cetus model 480 DNA thermal cycler. DNA sequencing was carried out in a Perkin-Elmer ABI Prism 377 DNA Sequencer. All primer syntheses and DNA sequencing related to cloning of the S. coelicolor acdH were done at the Biotechnology Center, University of Wisconsin-Madison; S. avermitilis DNA was sequenced by Lark Technologies.

Preparation of methacrylyl-CoA.
Methacrylyl-CoA was synthesized using acetyl-CoA synthetase as described by Patel & Walt (1987 ). The reaction mixture was modified as follows: CoA (6·5 mmol), ATP (25 mmol), MgCl2 (100 mmol), acetyl-CoA synthetase (3 units) and methacrylic acid (pH 7·2, 500 mmol) were mixed and the reaction was incubated at 37 °C for 2 h.

Methacrylyl-CoA was purified by HPLC on a Nova-Pak C18 column using 20% methanol in 0·05 M phosphate buffer, pH 5·7, at a flow rate of 2·0 ml min-1 with UV detection at 254 nm; the peak with a retention time of 6·7 min was collected. A Sephadex G-10 column with deionized H2O as the mobile phase was used for desalting; the peak absorbing at 254 nm was collected and vacuum-dried. Methacrylyl-CoA was characterized by proton NMR in 2H2O, using the NMR spectrum of isobutyryl-CoA for comparison. NMR spectra of methacrylic acid and the isobutyric acid reference standard were also measured in 2H2O. The NMR data showed two vinylic protons ({delta} 5·7 p.p.m. and {delta} 6·05 p.p.m.) and one C(CH3)=CH2 group ({delta} 1·9 p.p.m.) in the product; in contrast, the spectrum of isobutyryl-CoA had one methine proton ({delta} 2·85 p.p.m.) and a CH(CH3)2 group ({delta} 1·15 p.p.m.).

DNA preparation and amplification.
Small-scale preparations of E. coli plasmid DNA were made as described by Denoya et al. (1985 ) and Morelle (1988 ). Individual DNA fragments and PCR products were separated on agarose gels and purified with a Qiagen QIAEX II gel extraction kit according to the manufacturer’s instructions. Streptomycete genomic DNA was isolated by the SDS method of Hopwood et al. (1985 ). Oligodeoxynucleotides for PCR primers were purified according to the manufacturer’s protocols.

Several AD sequences available from the databases were aligned to identify conserved regions for the design of PCR primers targeting ADs related to branched-chain amino acid catabolism. The leftwards primers described below (P2 and P134) were based on a C-terminal region conserved in a number of short-chain/branched-chain ADs (Willard et al., 1996 ), starting with a Glu residue that has been postulated to serve as the catalytic residue (Rozen et al., 1994 ). Thus, the three nucleotides located at the 3' end of these primers, which have been described as the most important nucleotides for a successful PCR priming in homology cloning (Sommer & Tautz, 1989 ), represent the codon corresponding to Glu381 of the mature human short/branched-chain AD. The sequences of the two primers used to clone the acdH of S. coelicolor were: P1, 5'-AAGAATTCAACGGCACCAAGGCCTGGATCACCAAC-3'; and P2, 5'-AATCTAGAGCGCTGGATCTCGGAGGTGCCCTC-3'. The primers used to clone the acdH from S. avermitilis were: P131, 5'-AAGAATTCATATGAACGGCACCAAGGCCTGGATCACCAAC-3'; and P134, 5'-AAGGATCCTCTAGAGCGCTGGATCTCGGAGGTGCCCTC-3'.

PCR reactions were carried out with S. coelicolor DNA according to the EasyStart 100 protocol. The reaction was boiled for 5 min, cooled to 70 °C and 5 U Taq polymerase was added. Amplification was achieved with 25 cycles of denaturation at 97 °C for 50 s followed by annealing at 60–70 °C for 30 s followed by extension at 70 °C for 1/2–2 min. S. avermitilis genomic DNA was amplified essentially as described previously (Skinner & Denoya, 1993 ). The resulting PCR products were purified as described above and analysed by restriction enzyme digestion and/or DNA sequencing.

DNA hybridization, cloning and sequencing.
Southern blot hybridization and colony hybridization were performed with Hybond-N membranes (Amersham) by standard techniques (Sambrook et al., 1989 ). Digoxigenin-AP labelling, hybridization and detection were done with the Genius kit (Boehringer Mannheim), according to the manufacturer’s protocols. Hybridization was performed at 42 °C overnight, and the blot was washed twice with 1xSSC (0·15 M NaCl and 0·015 M sodium citrate)/0·1% SDS (Sambrook et al., 1989 ) for 5 min at room temperature and twice in 0·1xSSC/0·1% SDS for 15 min at room temperature.

Minilibraries of S. coelicolor J802 genomic DNA were constructed for cloning acdH. Genomic DNA was digested with BamHI and size-fractionated (6–12 kb) by electrophoresis in a 1% agarose gel. Fragments excised from the gel were cloned in pUC18 (Yanisch-Perron et al., 1985 ). The genomic DNA minilibrary was screened by colony hybridization using the PCR product (Fig. 2a) of acdH as a probe to obtain clone pWHM1301, which was shown to contain an approximately 10 kb BamHI DNA fragment. A 3·5 kb PstI fragment that contained the entire acdH gene was subcloned from pWHM1301 into the PstI site of pGEM-3Zf(-) (Promega) to give pWHM1302 (Fig. 2a). A cosmid library of S. avermitilis chromosomal DNA (Skinner et al., 1995 ) was screened by colony hybridization as described previously (Denoya et al., 1994 ) with the S. avermitilis acdH-specific 32P-labelled PCR product. Two positively hybridizing cosmid clones (of >2000 screened) contained overlapping sequences by restriction mapping and Southern blot hybridizations. Genomic DNA restriction fragments carrying acdH were separated and identified by PCR using a sterile toothpick to lift DNA by poking into each gel band as described previously (Kadokami & Lewis, 1994 ). The following genomic fragments were subcloned from the cosmid clones into pGEM-3Zf(-): 2·2 kb BamHI (pCD1106), 7·5 kb BamHI (pCD1119) and 4·5 kb SphI (pCD1468) (Fig. 2b).



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 2. Cloning of Streptomyces acdH genes. (a) The approximately 10 kb BamHI S. coelicolor J802 genomic fragment initially cloned (pWHM1301), and the 3·5 kb PstI fragment containing the entire acdH gene subcloned in the PstI site of pGEM-3Zf(-) as pWHM1302. (b) Restriction map of the S. avermitilis acdH genomic region initially cloned, and location of the fragments subcloned in pGEM-3Z.

 
DNA sequencing reactions were performed with the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction kit (Perkin-Elmer), according to the manufacturer’s protocols. Double-stranded plasmid DNA was used as a template. The Genetics Computer Group (GCG) software (Devereux et al., 1984 ) version 8.0 was used for sequence analysis. Nucleotide-sequence-deduced amino acid sequence data were compared with the National Center for Biotechnology Information (NCBI) database using the Basic blast search (Altschul et al., 1997 ).

Expression of S. coelicolor and S. avermitilis acdH in E. coli.
An NdeI site was introduced into the S. coelicolor acdH translational start codon. The primers used for PCR were: OP1, 5'-AACCATATGGACCACAAACTTTCCCCGGAACTGGAAGAA-3'; and OP2, 5'-GGGACCCGGACTCGGCG-3'. The PCR was carried out as described above with pWHM1302 as a template. The entire acdH gene was assembled by a three-piece ligation of the 0·68 kb NdeI–AatII fragment of the PCR product, the 0·55 kb AatII–BglII fragment of pWHM1302 and the NdeI–BamHI fragment of pT7-7 to give pWHM1303. pWHM1302 was prepared by subcloning the 3·5 kb PstI fragment containing the acdH from pWHM1301 into the PstI site of pGEM-3Zf(-). Similarly, an NdeI site was created at the S. avermitilis acdH translational start codon by PCR using pCD1468 (see above) as template. DNA primers were: P239, 5'-AAGAATTCATATGGACCACCGTCTCACCCCCGA-3' (rightward); and P238, 5' - AAAAGCTTCTGCAGGCCGGACAAGTTCGAAGGTAGGT-3' (leftward). The PCR-amplified DNA fragment was electroeluted from a 0·8% agarose gel, digested with NdeI and PstI, and ligated into pT7-7 to give pCD1513. The new constructs were confirmed by restriction enzyme digestion and the PCR fragment regions were sequenced.

Plasmids pWHM1303 and pCD1513 were introduced by transformation into E. coli BL21(DE3) for acdH expression. The E. coli host was also transformed with pT7-7 to generate a negative control culture. Transformants were grown at 37 °C in LB broth until the OD600 reached 0·6, then induced by adding IPTG to a final concentration of 0·4 mM. Incubation was continued for another 3 h, and whole cells were analysed for protein expression with 10% SDS-PAGE Ready Gels (Bio-Rad). Overexpressed proteins from the acdH and pT7-7 control strains were extracted by sonication; the cell extracts were fractionated by adding ammonium sulfate to 35–55% saturation. The fractionated proteins were analysed by 10% SDS-PAGE Ready Gels and used for the isobutyryl-CoA, n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA oxidation reactions. The protein concentration was determined by the method of Bradford (1976 ).

Characterization of the acdH gene product.
Reaction conditions for the oxidation of isobutyryl-CoA, n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexenylcarbonyl-CoA described by Ikeda et al. (1983 ) were modified as follows. The reaction mixture contained 100 mM phosphate buffer (pH 8·0), 6 mM PMS, 0·4 mM FAD, 1 mM substrate and 40 µg protein. The total volume was 0·5 ml. The mixture was incubated at 37 °C for 30 min (when the methacrylyl-CoA product was isolated, the mixture was incubated for 18 h). Proteins used in the reaction were fractionated by ammonium sulfate precipitation as described above. For better resolution in HPLC analysis, PMS, FAD and FADH2 were removed from the reaction mixture using a Waters Accell Plus QMA cartridge conditioned with 10 ml deionized H2O. Reaction mixture (0·5 ml) was loaded and the cartridges were washed with 2 ml 100 mM phosphate buffer (pH 8·0); substrate and product were eluted with 2 ml 500 mM phosphate buffer (pH 8·0) and collected as 0·25 ml fractions. Fractions that contained substrate (usually in the 3rd and 4th fractions) were analysed by HPLC using the conditions described above for methacrylyl-CoA purification. When isobutyryl-CoA was the substrate, a peak at retention time 6·7 min was collected and characterized by NMR as described above.

Gene replacement in S. avermitilis.
Plasmid pIJ4026, obtained from M. J. Bibb (John Innes Centre, Norwich, UK), was used as a source of the Saccharopolyspora erythraea 1·6 kb BglII fragment carrying ermE (which confers resistance to erythromycin; Bibb et al., 1994 ). Shuttle vector pCD262 is a chimeric construct between pGEM-3Z and pMT660 (Birch & Cullum, 1985 ). When a culture of S. avermitilis transformed with this vector is subjected to stress, such as high temperature or protoplast formation and regeneration in the absence of antibiotic selection, numerous plasmid-free colonies can be recovered (Denoya et al., 1995 ). We have taken advantage of this to use pCD262 as a vector for gene replacements in S. avermitilis. The gene replacement vector pCD1137, a derivative of pCD262, was constructed (see Fig. 6); it was used to transform S. lividans TK64 and reisolated for use in transforming S. avermitilis protoplasts. The latter were prepared and transformed as described previously (Denoya et al., 1995 ).



View larger version (20K):
[in this window]
[in a new window]
 
Fig. 6. Disruption of acdH in S. avermitilis ATCC 31272. Sizes of relevant restriction fragments are indicated. amp, ampicillin-resistance gene; tsr, thiostrepton-resistance gene; erm, erythromycin-resistance gene. Vectors are not drawn to scale.

 
Avermectin incorporation studies.
Seed cultures of the bkd mutant (CD1018) and the bkd acdH double mutant (CD1173) of S. avermitilis were grown for 3 d at 29 °C in 1x6 inch tubes containing 8 ml preform medium and two 5 mm glass beads, using an Innova 4900 shaker (New Brunswick Scientific) at 215 r.p.m. and a 17 ° angle. Preform medium contained: soluble starch [either thin-boiled starch or KOSO (Japan Corn Starch)], 20 g l-1; Pharmamedia (Traders Protein), 15 g l-1; Ardamine pH (Red Star), 5 g l-1; CaCO3, 2 g l-1; 2xBCFA (‘BCFA’ refers to branched-chain fatty acids) containing a final concentration in the medium of 50 p.p.m. 2-(±)-methylbutyric acid, 60 p.p.m. isobutyric acid and 20 p.p.m. isovaleric acid. The pH was adjusted to 7·2. A 2 ml portion of the seed culture was used to inoculate a 300 ml Erlenmeyer flask containing 25 ml production medium, which contained: starch (either thin-boiled starch or KOSO), 160 g l-1; Nutrisoy (Archer Daniels Midland), 10 g l-1; Ardamine pH, 10 g l-1; K2HPO4, 2 g l-1; MgSO4 . 4H2O, 2 g l-1; FeSO4 . 7H2O, 0·02 g l-1; MnCl2, 0·002 g l-1; ZnSO4 . 7H2O, 0·002 g l-1; CaCO3, 14 g l-1; 2xBCFA (as above); and the pH was adjusted to 6·9. After inoculation, the flask was incubated at 29 °C and 200 r.p.m. After 3 d, the fermentation was treated with a mixture of sodium [2,3,3'-13C3]isobutyrate (90 mg) and unlabelled sodium isobutyrate (280 mg) to a final concentration of 4·5 mM, and the fermentation was continued for an additional 10 d. At the end of the fermentation avermectin B1b was isolated and purified by standard HPLC protocols (Burg et al., 1979 ; Denoya et al., 1995 ) as follows. The broth was extracted twice with an equal volume of ethyl acetate, and the ethyl acetate extracts were evaporated to give a brown, oily substance containing the avermectins. This was solubilized in a minimal volume of ethyl acetate (~5 ml) and applied to a 40M KP-SIL flash chromatography column (Biotage, Division of Dyax) equilibrated with ethyl acetate. The avermectins were eluted with ethyl acetate and analysed for avermectin content using a Waters HPLC instrument with a model 600 controller and a Waters 996 photodiode array absorbance detector. Separation was achieved using a Waters C18 symmetry column (3·9x150 mm; elution with 65% acetonitrile/35% water at 1 ml min-1). Fractions containing avermectins were pooled and evaporated to dryness.

Final purification was achieved with a preparative Waters HPLC instrument as above. The material eluted from the 40M KP-SIL column was solubilized in a minimal volume of methanol (~3 ml) and the avermectins were separated on a Waters C18 symmetry column (19x300 mm; elution with 70% acetonitrile/30% water at 15 ml min-1). Fractions containing avermectin B1b were collected, the acetonitrile was removed under vacuum, and the aqueous solution was extracted with three equal volumes of ethyl acetate. The ethyl acetate was then removed under vacuum, yielding 22·4 mg B1b from culture CD1018 and 17·5 mg from CD1173. Each sample was solubilized in 0·5 ml deuterated chloroform for analytical NMR.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Cloning and sequencing of the acdH from S. coelicolor and S. avermitilis
Based on the high degree of sequence homology between all the ADs described in the literature, a homology cloning approach using PCR was investigated. The acdH genes were amplified from S. coelicolor J802 or S. avermitilis ATCC 31272 genomic DNAs using PCR primers designed from analysis of a multiple alignment of published AD peptide sequences (Matsubara et al., 1989 ). A DNA product of approximately 0·67 kb was amplified from each template, cloned and sequenced. The deduced protein products showed high similarity to each other (more than 90% identity) and to products of numerous AD genes in the databases. These PCR products were subsequently used to probe their corresponding genomic libraries. The nucleotide sequence of a 1·7 kb DNA sequence of the S. coelicolor acdH genomic region is available from GenBank (accession no. AF14258). codon preference and gap analyses (Devereux et al., 1984 ) showed that there is only one complete ORF in this region with the expected codon usage pattern for a Streptomyces gene (Bibb et al., 1984 ). The 1161 nt acdH gene is predicted to begin with an ATG at position 265, terminate in a TGA at position 1425, and have a putative RBS (AGGAG) spaced 11 nt from the ATG start codon; it should encode a 386-amino-acid protein with a calculated molecular mass of 41726 Da. The S. avermitilis acdH gene (GenBank accession no. AF143210) also has an ATG start codon with an AGGAG RBS positioned 11 nt upstream, and encodes a primary translation product of 386 amino acids calculated to have a molecular mass of 41737 Da. The amino acid sequences of the AcdH proteins from S. coelicolor and S. avermitilis have high similarity (93% identity overall by gap analysis) and the same length. In the blast search of the NCBI database, both exhibited high similarity to a variety of prokaryotic and eukaryotic ADs: 70%, 50% and 42% overall identity to putative ADs from Mycobacterium tuberculosis (Philipp et al., 1996 ), Bacillus subtilis (YngJ; Tosato et al., 1997 ) and B. subtilis (MmgC; Bryan et al., 1996 ), respectively, 43% identity to a butyryl-CoA dehydrogenase from Clostridium acetobutylicum (Boynton et al., 1996 ), and approximately 42% identity to several mammalian short-chain ADs (Matsubara et al., 1989 ; Naito et al., 1989 ; Kelly et al., 1993 ). In addition, the deduced S. coelicolor and S. avermitilis acdH products contain the consensus signature patterns found in all ADs (Fig. 3), and an amino acid [Glu(Asp)368] that is predicted to serve as the catalytic residue (Rozen et al., 1994 ).



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 3. Acyl-CoA dehydrogenase signature patterns. The S. coelicolor and S. avermitilis AcdH protein sequences (AD-Scoe, S. coelicolor AcdH; AD-Save, S. avermitilis AcdH) were checked for sequence patterns using the Motifs program of the GCG package (Devereux et al., 1984 ). Significant matches with motifs in the PROSITE database are shown. The first amino acid in each sequence is numbered in brackets. The sources are: AD from Bacillus subtilis (AD-Bsub) (Tosato et al., 1997 ), butyryl-CoA dehydrogenase from Clostridium acetobutylicum (BD-Cace) (Boynton et al., 1996 ), short-chain AD from rat (SCAD-Rat) (Matsubara et al., 1989 ) and short-chain AD from human (SCAD-Human) (Naito et al., 1989 ). Shown in bold are the residues conserved in all sequences.

 
DNA sequencing showed that immediately upstream of the S. avermitilis acdH ORF there is an ORF (ORF2) that appears from the overlap between its stop codon and the most likely ATG initiation codon of the acdH ORF (data not shown) to be co-expressed with acdH. Database searching showed that the deduced ORF2 protein has a high degree of similarity to a large number of 3-hydroxy-3-methylglutaryl (HMG)-CoA lyases. The strongest match was to Pseudomonas mevalonii HMG-CoA lyase (more than 40% identity over a 200-amino-acid overlap; Anderson & Rodwell, 1989 ). Interestingly, HMG-CoA lyase is the last enzyme in the leucine catabolic pathway and produces acetyl-CoA and acetoacetate. A homologue of ORF2 was also found immediately upstream of the S. coelicolor acdH, and a truncated ORF (ORF1) was detected immediately upstream of this homologue. The deduced product of the truncated ORF1 has approximately 70% identity (overlap of more than 170 amino acids) to M. tuberculosis biotin carboxyl carrier protein (BCCP), a peptide subunit of enzymes catalysing carboxyl transfer (Norman et al., 1994 ). We speculate that this BCCP is a subunit of the methylcrotonyl-CoA carboxylase that governs step 4 in the leucine catabolic pathway.

Overexpression of S. coelicolor and S. avermitilis acdH in E. coli
For characterization of the S. coelicolor acdH product, we constructed a pT7-7 derivative, pWHM1303, containing the complete acdH coding region under the control of an IPTG-inducible gene for T7 RNA polymerase. E. coli BL21(DE3) containing pWHM1303 was analysed for acdH expression by SDS-PAGE. A protein of 42 kDa, the approximate mass of the predicted S. coelicolor acdH product (41·7 kDa), was detected in the IPTG-induced extracts from E. coli BL21(DE3)(pWHM1303), but not from the pT7-7 control strain (Fig. 4a). Proteins extracted from these strains were fractionated by addition of ammonium sulfate (Fig. 4b) to give enzymically active samples with protein concentrations of 4·6 mg ml-1 and 6·0 mg ml-1 for the pWHM1303 overexpression and pT7-7 control strains, respectively. These protein extracts were used for the oxidation of isobutyryl-CoA, n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA in vitro. Similarly, upon heat induction of the S. avermitilis acdH in E. coli BL21(DE3)(pCD1513), a band of about 42 kDa, which is the approximate mass of the predicted S. avermitilis AcdH protein, was detected in protein extracts.



View larger version (87K):
[in this window]
[in a new window]
 
Fig. 4. PAGE of proteins expressed from the acdH of S. coelicolor in E. coli. (a) Lanes 2, 3, 4 and 5, E. coli BL21(DE3) containing pWHM1303 induced with IPTG for 0, 1, 2 and 3 h; lanes 8 and 9, pT7-7 induced with IPTG for 0 and 3 h; and lanes 6 and 7, no plasmid induced with IPTG for 0 and 3 h. (b) Cell extract from E. coli BL21(DE3) containing pWHM1303. Lane 2 is the total soluble protein; lanes 3, 4 and 5 are from the 0–35%, 35–55% and 55–80% saturation of cell extract with ammonium sulfate. Lane 6 is the same as lane 9 in (a). SDS-PAGE (10% polyacrylamide gel) was used. Lanes 1 in (a) and (b) contain molecular mass standards.

 
Characterization of the acdH product
Dye–bleach assays, described by Hall (1978 ) and Engel (1981 ), are often used for AD purification and characterization, but such an assay cannot be used with cell extracts because of interference by nonspecific reductants. Therefore, we directly looked for product formation by using HPLC analysis to analyse enzyme reaction mixtures. The results with the S. coelicolor AcdH protein and isobutyryl-CoA as substrate are shown in Fig. 5. The expected product appeared only when the overexpressed protein was included in the reaction mixture (Fig. 5d). The product had the same retention time and UV spectrum (data not shown) as a biosynthesized methacrylyl-CoA standard and its proton NMR spectrum (data not shown) was identical to that of methacrylyl-CoA. These results strongly suggest that acdH encodes an enzyme that catalyses the conversion of isobutyryl-CoA to methacrylyl-CoA in S. coelicolor. Most of the isobutyryl-CoA was also consumed in reaction mixtures containing the protein from the control strain, but the product was not methacrylyl-CoA (Fig. 5c).



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 5. HPLC analysis of the oxidation of isobutyryl-CoA to methacrylyl-CoA. (a) A mixture of methacrylyl-CoA (MA-CoA) and isobutyryl-CoA (IB-CoA); (b) 30 µl reaction mixture without protein; (c) 100 µl reaction mixture with protein from the E. coli strain containing the pT7-7 vector; (d) 30 µl reaction mixture with protein from the E. coli strain overexpressing AcdH. A new peak (labelled ‘product’) has the same retention time as the methacrylyl-CoA standard.

 
The results of testing n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA as substrates were similar to those obtained with isobutyryl-CoA. They all were oxidized by the AcdH to products with a shorter retention time (data not shown). The product formed from cyclohexylcarbonyl-CoA had the same retention time as authentic 1-cyclohexenylcarbonyl-CoA (Reynolds et al., 1992 ) when co-injected into the HPLC (data not shown). In addition, similar results were obtained when the S. avermitilis AcdH was analysed using crude extracts from the overexpressing recombinant E. coli strain BL21(DE3)(pCD1513).

Disruption of the acdH gene in S. avermitilis
To confirm that acdH encodes a functional AD in S. avermitilis and to define further its role in the catabolism of branched-chain amino acids, the gene was inactivated by insertion of the Sac. erythraea ermE marker that confers resistance to erythromycin (Bibb et al., 1994 ). The replacement vector was constructed in two steps (Fig. 6). First, both the 2·2 kb BamHI S. avermitilis genomic fragment (from pCD1106) carrying the 5'-half of acdH and the 1·6 kb BglII fragment (from pIJ4026) carrying the ermE marker were cloned into the unique BglII site of the E. coli/Streptomyces shuttle vector pCD262 to create pCD1126. Second, pCD1126 was linearized with BglII and ligated to the 7·5 kb BamHI S. avermitilis fragment (from pCD1119) to create pCD1137. This construct, which was expected to produce an insertional inactivation of the 980 nt ORF in the host genome upon recombination, contains 2·2 and 7·5 kb of homologous DNA to the left and right, respectively, of the target region. The ermE marker lies in the opposite orientation to that of the ORF.

Genomic replacement was achieved by following the approaches of Anzai et al. (1988 ) and Stutzman-Engwall et al. (1992 ) as previously described (Denoya et al., 1995 ). Plasmid pCD1137 was introduced into protoplasts of S. avermitilis ATCC 31272 by transformation, and many primary transformants resistant to both thiostrepton (the vector marker) and erythromycin were obtained. One of these was selected for propagation and for protoplast formation to eliminate the plasmid. After the transformant was plated on regeneration medium containing erythromycin, more than 50% of the colonies were resistant only to erythromycin. Chromosomal DNA was isolated from one of the thiostrepton-sensitive, erythromycin-resistant colonies (S. avermitilis CD1156) and compared with DNA from the parental strain (ATCC 31272). Southern blot hybridization confirmed that the ermE marker had been inserted and had disrupted the target ORF through the expected double crossover (data not shown).

Phenotype of the S. avermitilis disruptants
Table 1 compares the growth of a mutant CD1156 and the wild-type strain. The acdH-disrupted strain had lost the ability to grow on solid minimal medium containing valine, isoleucine and/or leucine as the sole carbon sources. In addition, the mutant grew more weakly than the control strain on acetate or butyrate, but as well as the control on either glucose, caproic acid or caprylic acid as the sole carbon source. To confirm the phenotype of the mutant, a second independent experiment repeated disruption of acdH in S. avermitilis ATCC 31272. Numerous disruptants were recovered and analysed, all showing a similar phenotype to the mutant described above. In addition, analysis of 35 disrupted cultures showed that the mutants grew at normal rates in rich medium and were capable of producing avermectins at least as well as the wild-type.


View this table:
[in this window]
[in a new window]
 
Table 1. Growth properties of S. avermitilis CD1156 acdH mutant (m) and the wild-type strain (wt)

 
To confirm that the phenotype of the disruptant was linked to the insertional inactivation of the acdH gene, a second gene replacement was performed in the CD1156 culture to delete the ermE marker and restore the integrity of the AD-encoding ORF. Numerous erythromycin-sensitive isolates were recovered and all of them exhibited a phenotype similar to that found in the wild-type strain (not shown).

Incorporation of [2,3,3'-13C3]isobutyrate into avermectin
To probe the physiological role of acdH further, incorporation of sodium [2,3,3'-13C3]isobutyrate was investigated using a bkd mutant of S. avermitilis that can make avermectin only when supplied with exogenous short-chain fatty acids. Thus high levels of labelling of the isobutyrate-derived starter unit of avermectin directly reflecting the level of labelling of the isobutyrate added were obtained and purification was simplified as the methylbutyrate-derived avermectins were not produced.

13C-NMR analysis of avermectin B1b (22·4 mg) isolated after incorporation of the labelled isobutyrate into a bkd mutant (S. avermitilis CD1018) showed greater than 10-fold enrichment of the signals corresponding to C-26 (multiplet), C-26a (doublet) and C-27 (doublet) (data not shown). These three carbons are derived from C-2, C-3 and C-3' of isobutyrate. The avermectin 13C-NMR spectrum revealed clear enrichment of the carbons derived from acetate (1·25-fold) and propionate (1·6-fold) (Fig. 7). The pattern of enrichment matched that previously determined from incorporation experiments using [1,2-13C]acetate and [2,3-13C2]propionate (data not shown). Thus these results are consistent with the labelled isobutyrate being (a) isomerized in vivo to n-butyryl-CoA (see Fig. 1) and subsequently degraded to generate [1,2-13C]acetate and [2-13C]acetate, and (b) oxidized directly, presumably via methacrylyl-CoA, to [2,3,3'-13C3]methylmalonyl-CoA (see Fig. 1).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 7. Incorporation of [2,3,3-13C3]isobutyrate into avermectin B1b. (a) Bold, nonwedged bonds indicate intact incorporation of dual C-13 labels into the acetate-, propionate- and isobutyrate-derived positions of avermectin B1b. (b, c) Regions of the 13C-NMR of avermectin B1b obtained after [13C]isobutyrate incorporation into bkd (b) and bkd acdH (c) mutants of S. avermitilis. C-24 (31·13 p.p.m.) and C-12 (39·96 p.p.m.) are derived from C-2 of propionate (methylmalonyl-CoA) while C-16 (34·5 p.p.m.), C-18 (36·9 p.p.m.) and C-20 (40·66 p.p.m.) are derived from C-1 of acetate. The 13C doublets with all five of these signals indicate that the 13C-enriched carbon is bonded directly to a second 13C. C-2' (34·28 p.p.m.) and C-2' (34·77 p.p.m.) are from the disaccharide and are not labelled from the isobutyrate.

 
In an analogous experiment using a bkd acdH double mutant of S. avermitilis (CD1173, 17·5 mg B1b isolated), no change in 13C-enrichments for positions derived from isobutyrate (>10-fold) and acetate (1·35-fold) were detected. However, the enrichment (1·3-fold) in positions derived from C-2 and C-3 of propionate decreased by over 50% (Fig. 7) from that seen using the bkd mutant. These results indicate that the acdH product does not play a significant role in the oxidation of n-butyryl-CoA (despite an ability of the corresponding enzyme to catalyse this reaction in vitro) but does contribute to the oxidation of isobutyryl-CoA to methacrylyl-CoA (see Fig. 1).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We chose S. coelicolor J802 to study the acdH gene because information about the genetics of valine catabolism can be integrated with other knowledge about this genetically well-characterized organism and serve as a model for other Streptomyces spp. In addition, we chose S. avermitilis because the role of the branched-chain amino acid catabolic pathway as a provider of fatty acid precursors for antibiotic biosynthesis has been well-established in this industrially important organism. The enzymes expressed by the acdH cloned from S. coelicolor J802 and S. avermitilis ATCC 31272 were able to oxidize the branched-chain acyl-CoA derivatives isobutyryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA, as well as the short straight-chain acyl-CoAs n-butyryl-CoA and n-valeryl-CoA. AcdH belongs to the AD family of related enzymes that catalyse the oxidation of saturated acyl-CoA thioesters to give 2,3-unsaturated products. Electrons from the reduced ADs are transferred to the flavoprotein and ultimately to the electron transport chain of the cell.

At least six distinct members of this enzyme family have been identified in mammalian cells. Four of them catalyse the first step of the ß-oxidation cycles for straight-chain fatty acids with various chain lengths. These are short (C4–C6 CoA), medium (C4–C16 CoA), long (C6–C22 CoA) and very long chain ADs (Wolf & Akers, 1986 ; Matsubara et al., 1989 ; Kelly et al., 1993 ; Aoyama et al., 1994 ). Two others catalyse the third reaction in the oxidation of branched-chain amino acids. Isovaleryl-CoA dehydrogenase is involved in leucine metabolism and 2-methyl-branched-chain AD is involved in isoleucine/valine metabolism. Recently, a rat short/branched-chain AD able to oxidize 2-methylbutyryl-CoA, isobutyryl-CoA and n-butyryl-CoA efficiently has also been characterized (Willard et al., 1996 ). At least in eukaryotic systems, five of the ADs are tetrameric flavoproteins with a subunit size of 40–45 kDa. Very long chain AD appears to be a dimer of 71 kDa subunits. All ADs share similar structural and functional features, including a consensus signature pattern in their amino acid sequences.

The blast search showed the closest identity (up to 46%) between the deduced AcdH amino acid sequence and a sub-group of ADs consisting mostly of short-chain specific enzymes. As we expected, the AcdH enzyme not only catalysed conversion of isobutyryl-CoA to methacrylyl-CoA, but also oxidized n-butyryl-CoA, n-valeryl-CoA, isovaleryl-CoA and cyclohexylcarbonyl-CoA. Therefore, acdH appears to encode a short/branched-chain AD. An acyl-CoA oxidase purchased from Sigma did not oxidize isobutyryl-CoA to methacrylyl-CoA under conditions that allowed enzymic oxidation of palmitoyl-CoA to its 2,3-unsaturated ester, and clearly has different functions from the AcdHs we describe here.

The properties of the products encoded by acdH in S. avermitilis and S. coelicolor, taken together with the sequence data showing that two upstream ORFs encode products putatively involved in leucine catabolism, support our belief that the acdH product is involved in degradation of branched-chain amino acids. In mammals, two ADs catalyse the third reaction in the oxidation of branched-chain amino acids. These are isovaleryl-CoA dehydrogenase in leucine metabolism and 2-methyl-branched-chain acyl-CoA dehydrogenase in isoleucine/valine metabolism. Both mammalian enzymes have a narrow substrate specificity (Ikeda et al., 1983 ). Since the S. avermitilis acdH-disrupted mutant is unable to grow on valine, isoleucine and/or leucine as sole carbon source(s) one might speculate that Streptomyces spp. have a single branched-chain AD with broader substrate specificity than the mammalian counterparts and a role in the catabolism of all three branched-chain amino acids. However, the [13C]isobutyrate incorporation experiments showed that at least the valine catabolic pathway is still operating in the acdH mutant, though a greater than 50% decrease in 13C at methylmalonate-derived positions of avermectin suggest the catabolic rate is significantly reduced. The latter observation could be explained by assuming that two or more ADs with overlapping specificities contribute to the assimilation of branched-chain amino acids. Inactivation of just one of the overlapping activities, such as the broad substrate range AcdH in the mutant S. avermitilis culture, would be sufficient to prevent growth under the stringent conditions of a minimal medium having branched-chain amino acids as sole carbon source.

The reason why the S. avermitilis acdH mutant grew poorly on acetate and butyrate as a sole carbon source is unclear. It has recently been demonstrated that in addition to the glyoxylate cycle streptomycetes have a novel pathway for growth on acetate as a sole carbon source (Han & Reynolds, 1997 ). Evidence indicates that this pathway passes through a butyryl-CoA intermediate. The inability of the acdH mutant to grow on either acetate or butyrate therefore may implicate the conversion of isobutyryl-CoA to methacrylyl-CoA as a step in the acetate assimilation pathway. Interconversion of butyryl-CoA and isobutyryl-CoA in streptomycetes is well-established (Sherman et al., 1986 ; Reynolds et al., 1988 ).

Several earlier reports have described a relationship between the catabolism of branched-chain amino acids and macrolide antibiotic production (Omura et al., 1983 ; Tang et al., 1994 ), suggesting that under some growth conditions, this catabolic route can provide the n-butyryl-CoA, 2-methylmalonyl-CoA or propionyl-CoA substrates used by the polyketide synthases to build the carbon skeleton of the antibiotic. In the present case the isotopic enrichment of the starter unit and the amount of avermectin produced were not affected when the S. avermitilis acdH mutant was grown in a rich medium, even though about 50% less of the isobutyrate-derived 2-methylmalonate was incorporated into propionate-derived positions of the polyketide. These results confirm that the starter unit for the polyketide portion of avermectin comes from valine and isoleucine catabolism (Denoya et al., 1995 , 1996 ). Because 13C-enrichment of the propionate-derived positions of avermectin B1b was less than 10% of that in the starter unit, they also indicate that in a rich growth medium the amount of methylmalonate and propionate derived from isobutyrate catabolism and used to make the avermectins is very low.


   ACKNOWLEDGEMENTS
 
We thank Stephanie Patton, Virginia Commonwealth University, for preparing the cyclohexylcarbonyl-CoA substrate, and Linda Wisniewski, Pfizer, Inc., for running the NMR samples. This research was supported in part by grants from the National Institutes of Health (GM31925 to C.R.H. and GM50541 to K.A.R.).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped blast and psi-blast: a new generation of protein database search programs. Nucleic Acids Res 25, 3389-3402.[Abstract/Free Full Text]

Anderson, D. H. & Rodwell, V. W. (1989). Nucleotide sequence and expression in Escherichia coli of the 3-hydroxy-3-methylglutaryl coenzyme A lyase gene of Pseudomonas mevalonii. J Bacteriol 171, 6468-6472.[Medline]

Anzai, H., Kumada, Y., Hara, O., Murakami, T., Itoh, R., Takano, E., Imai, S., Satoh, A. & Nagaoka, K. (1988). Replacement of Streptomyces hygroscopicus genomic segments with in vitro altered DNA sequences. J Antibiot 41, 226-233.[Medline]

Aoyama, T., Ueno, I., Kamijo, T. & Hashimoto, T. (1994). Rat very-long-chain acyl-CoA dehydrogenase, a novel mitochondrial acyl-CoA dehydrogenase gene product, is a rate-limiting enzyme in long-chain fatty acid ß-oxidation. cDNA and deduced amino acid sequence and distinct specificities of the cDNA-expressed protein. J Biol Chem 269, 19088-19094.[Abstract/Free Full Text]

Bibb, M. J., Findlay, P. R. & Johnson, M. W. (1984). The relationship between base composition and codon usage in bacterial genes and its use for the simple and reliable identification of protein coding sequences. Gene 30, 157-166.[Medline]

Bibb, M. J., White, J., Ward, J. M. & Janssen, G. R. (1994). The mRNA for the 23S rRNA methylase encoded by the ermE gene of Saccharopolyspora erythraea is translated in the absence of a conventional ribosome-binding site. Mol Microbiol 14, 533-545.[Medline]

Birch, A. W. & Cullum, J. (1985). Temperature-sensitive mutants of the Streptomyces plasmid pIJ702. J Gen Microbiol 131, 1299-1303.[Medline]

Boynton, Z. L., Bennett, G. N. & Rudolph, F. B. (1996). Cloning, sequencing, and expression of clustered genes encoding ß-hydroxybutyryl-coenzyme A (CoA) dehydrogenase, crotonase, and butyryl-CoA dehydrogenase from Clostridium acetobutylicum ATCC 824. J Bacteriol 178, 3015-3024.[Abstract]

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem 72, 248-254.[Medline]

Bryan, E. M., Beall, B. W. & Moran, C. P.Jr (1996). A {sigma}E-dependent operon subject to catabolite repression during sporulation in Bacillus subtilis. J Bacteriol 178, 4778-4786.[Abstract]

Burg, R. W., Miller, B. M., Baker, E. E. & 12 other authors (1979). Avermectins, new family of potent anthelmintic agents: producing organism and fermentation. Antimicrob Agents Chemother 15, 361–367.[Medline]

Denoya, C. D., Trevisan, A. R., Zorzopulos, J., Woloj, M. & Rubeglio, E. (1985). Diversity of plasmid profiles in multiply resistant Klebsiella pneumoniae strains isolated from a single nosocomial environment with a strong antibiotic selection pressure. Microbios Lett 29, 87-93.

Denoya, C. D., Skinner, D. D. & Morgenstern, M. R. (1994). A Streptomyces avermitilis gene encoding a 4-hydroxyphenylpyruvic acid dioxygenase-like protein that directs the production of homogentisic acid and an ochronotic pigment in Escherichia coli. J Bacteriol 176, 5312-5319.[Abstract]

Denoya, C. D., Fedechko, R. W., Hafner, E. W., McArthur, H. A. I., Morgenstern, M. R., Skinner, D. D., Stutzman-Engwall, K., Wax, R. G. & Wernau, W. C. (1995). A second branched-chain {alpha}-keto acid dehydrogenase gene cluster (bkdFGH) from Streptomyces avermitilis: its relationship to avermectin biosynthesis and the construction of a bkdF mutant suitable for the production of novel antiparasitic avermectins. J Bacteriol 177, 3504-3511.[Abstract]

Denoya, C. D., Skinner, D. D., Fedechko, R. W. & McArthur, H. A. I. (1996). Genetic manipulations of the branched-chain fatty acid supply for the biosynthesis of avermectins in Streptomyces avermitilis. Abstracts of The Sixth Conference on the Genetics and Molecular Biology of Industrial Microorganisms (GMBIM), P60. American Society for Microbiology.

Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res 12, 387-395.[Abstract]

Engel, P. C. (1981). Butyryl-CoA dehydrogenase from Megasphaera elsdenii. Methods Enzymol 71, 359-366.

Goodwin, G. W., Rougraff, P. M., Davis, E. J. & Harris, R. A. (1989). Purification and characterization of methylmalonate-semialdehyde dehydrogenase from rat liver: identity to malonate-semialdehyde dehydrogenase. J Biol Chem 264, 14965-14971.[Abstract/Free Full Text]

Hall, C. L. (1978). Acyl-CoA dehydrogenases and electron transferring flavoprotein. Methods Enzymol 53, 502-518.[Medline]

Han, L. & Reynolds, K. A. (1997). A novel alternate anaplerotic pathway to the glyoxylate cycle in streptomycetes. J Bacteriol 179, 5157-5164.[Abstract]

Hodgson, D. A. & Chater, K. F. (1981). A chromosomal locus controlling extracellular agarase production by Streptomyces coelicolor A3(2) and its inactivation by chromosomal integration of plasmid SCP1. J Gen Microbiol 124, 339-348.

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

Ikeda, Y., Dabrowski, C. & Tanaka, K. (1983). Separation and properties of five distinct acyl-CoA dehydrogenases from rat liver mitochondria. J Biol Chem 258, 1066-1076.[Free Full Text]

Kadokami, Y. & Lewis, R. V. (1994). Repeated PCR of a gel band can be used to obtain a single PCR band. BioTechniques 17, 438.[Medline]

Kelly, C. L., Hinsdale, M. E. & Wood, P. A. (1993). Cloning and characterization of the mouse short-chain acyl-CoA dehydrogenase cDNA. Genomics 18, 137-140.[Medline]

Lebrihi, A., Lamsaif, D., Lefebvre, G. & Germain, P. (1992). Effect of ammonium ions on spiramycin biosynthesis in Streptomyces ambofaciens. Appl Microbiol Biotechnol 37, 382-387.[Medline]

Massey, L. M., Sokatch, J. R. & Conrad, R. S. (1976). Branched-chain amino acid catabolism in bacteria. Bacteriol Rev 40, 42-54.[Medline]

Matsubara, Y., Indo, Y., Naito, E., Ozasa, H., Glassberg, R., Vockley, J., Ikeda, Y., Kraus, J. & Tanaka, K. (1989). Molecular cloning and nucleotide sequence of cDNAs encoding the precursors of rat long chain acyl-coenzyme A, short chain acyl-coenzyme A, and isovaleryl-coenzyme A dehydrogenases. Sequence homology of four enzymes of the acyl-CoA dehydrogenase family. J Biol Chem 264, 16321-16331.[Abstract/Free Full Text]

Morelle, G. (1988). A plasmid extraction procedure on a miniprep-scale. Focus 11, 7-8.

Naito, E., Ozasa, H., Ikeda, Y. & Tanaka, K. (1989). Molecular cloning and nucleotide sequence of complementary DNAs encoding human short chain acyl-coenzyme A dehydrogenase and the study of the molecular basis of human short chain acyl-coenzyme A dehydrogenase deficiency. J Clin Investig 83, 1605-1613.[Medline]

Norman, E., De Smet, K. A., Stoker, N. G., Ratledge, C., Wheeler, P. R. & Dale, J. W. (1994). Lipid synthesis in mycobacteria: characterization of the biotin carboxyl carrier protein genes from Mycobacterium leprae and M. tuberculosis. J Bacteriol 176, 2525-2531.[Abstract]

Omura, S., Tsuzuki, K., Tanaka, Y., Sakakibara, H., Aizawa, M. & Lukacs, G. (1983). Valine as a precursor of the n-butyrate unit in the biosynthesis of macrolide aglycones. J Antibiot 36, 614-616.[Medline]

Patel, S. S. & Walt, D. R. (1987). Substrate specificity of acetyl coenzyme A synthetase. J Biol Chem 262, 7132-7134.[Abstract/Free Full Text]

Philipp, W. J., Poulet, S., Eiglmeier, K. & 7 other authors (1996). An integrated map of the genome of the tubercle bacillus, Mycobacterium tuberculosis H37Rv, and comparison with Mycobacterium leprae. Proc Natl Acad Sci USA 93, 3132–3137.

Reynolds, K. A., O’Hagen, D., Gani, D. & Robinson, J. A. (1988). Butyrate metabolism in streptomycetes. Characterization of an intramolecular vicinal interchange rearrangement linking isobutyrate and butyrate in Streptomyces cinnamonensis. J Chem Soc Perkin Trans I 1988, 3195-3208.

Reynolds, K. A., Wang, P., Fox, K. M., Lam, S., Speedie, M. K. & Floss, H. G. (1992). Purification and characterization of a novel enoyl coenzyme A reductase from Streptomyces collinus. J Bacteriol 174, 3850-3854.[Abstract]

ozen, R., Vockley, J., Zhou, L. & 7 other authors (1994). Isolation and expression of a cDNA encoding the precursor for a novel member (ACADSB) of the acyl-CoA dehydrogenase gene family. Genomics 24, 280–287.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sherman, M. M., Yue, S. & Hutchinson, C. R. (1986). Biosynthesis of lasalocid A. Metabolic interrelationships of carboxylic acid precursors and polyether antibiotics. J Antibiot 39, 1135-1143.[Medline]

Skinner, D. D. & Denoya, C. D. (1993). A simple DNA polymerase chain reaction method to locate and define orientation of specific sequences in cloned bacterial genomic fragments. Microbios 75, 125-129.[Medline]

Skinner, D. D., Morgenstern, M. R., Fedechko, R. W. & Denoya, C. D. (1995). Cloning and sequencing of a cluster of genes encoding branched-chain {alpha}-keto acid dehydrogenase from Streptomyces avermitilis and the production of a functional E1[{alpha}ß] component in Escherichia coli. J Bacteriol 177, 183-190.[Abstract]

Sommer, R. & Tautz, D. (1989). Minimal homology requirements for PCR primers. Nucleic Acids Res 17, 6749.[Medline]

Sood, G. R., Ashworth, D. M., Ajaz, A. A. & Robinson, J. A. (1988). Biosynthesis of the polyether antibiotic monensin A. Results from the incorporation of labeled acetate and propionate as a probe of the carbon chain assembly process. J Chem Soc Perkin Trans I 1988, 3183-3194.

Steele, M. I., Lorenz, D., Hatter, K., Park, A. & Sokatch, J. R. (1992). Characterization of the mmsAB operon of Pseudomonas aeruginosa PAO encoding methylmalonate semialdehyde dehydrogenase and 3-hydroxyisobutyrate dehydrogenase. J Biol Chem 267, 13585-13592.[Abstract/Free Full Text]

Stutzman-Engwall, K. J., Otten, S. L. & Hutchinson, C. R. (1992). Regulation of secondary metabolism in Streptomyces spp. and the overproduction of daunorubicin in Streptomyces peucetius. J Bacteriol 174, 144-154.[Abstract]

Sykes, P. J., Burns, G., Menard, J., Hatter, K. & Sokatch, J. R. (1987). Molecular cloning of genes encoding branched-chain keto acid dehydrogenase of Pseudomonas putida. J Bacteriol 169, 1619-1625.[Medline]

Tabor, S. (1990). Expression using the T7 RNA polymerase/promoter system. In Current Protocols in Molecular Biology, pp. 16.2.1–16.2.11. Edited by F. A. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith & K. Struhl. New York: Greene Publishing and Wiley Interscience.

Tang, L. & Hutchinson, C. R. (1993). Sequence, transcriptional and functional analysis of the valine (branched-chain amino acid) dehydrogenase gene of Streptomyces coelicolor. J Bacteriol 175, 4176-4185.[Abstract]

Tang, L., Zhang, Y.-X. & Hutchinson, C. R. (1994). Amino acid catabolism and antibiotic synthesis: valine is a source of precursors for macrolide biosynthesis in Streptomyces ambofaciens and Streptomyces fradiae. J Bacteriol 176, 6107-6119.[Abstract]

Tosato, V., Albertini, A. M., Zotti, M., Sonda, S. & Bruschi, C. V. (1997). Sequence completion, identification and definition of the fengycin operon in Bacillus subtilis 168. Microbiology 143, 3443-3450.[Abstract]

Wexler, I. D., Hemalatha, S. G. & Patel, M. S. (1991). Sequence conservation in the {alpha} and ß subunits of pyruvate dehydrogenase and its similarity to branched-chain {alpha}-keto acid dehydrogenase. FEBS Lett 282, 209-213.[Medline]

Willard, J., Vicanek, C., Battaile, K., Van Veldhoven, P., Fauq, A., Rozen, R. & Vockley, J. (1996). Cloning of a cDNA for short/branched chain acyl-Coenzyme A dehydrogenase from rat and characterization of its tissue expression and substrate specificity. Arch Biochem Biophys 331, 127-133.[Medline]

Wolf, D. A. & Akers, H. A. (1986). Uncertainties remain in the catabolism of valine. Trends Biochem Sci 11, 390-392.

Yanisch-Perron, C., Vieira, J. & Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103-119.[Medline]

Zhang, Y.-X., Tang, L. & Hutchinson, C. R. (1996). Cloning and characterization of a homolog of the mmsA gene encoding methylmalonate-semialdehyde dehydrogenase from Streptomyces coelicolor. J Bacteriol 178, 490-495.[Abstract]

Received 31 December 1999; revised 22 April 1999; accepted 27 April 1999.