Advertisement

Identification of a Flavin:NADH Oxidoreductase Involved in the Biosynthesis of Actinorhodin

PURIFICATION AND CHARACTERIZATION OF THE RECOMBINANT ENZYME *
  • Steven G. Kendrew
    Affiliations
    Department of Biochemistry and Cambridge Centre for Molecular Recognition, University of Cambridge, Tennis Court Road, Cambridge CB2 1QW, United Kingdom
    Search for articles by this author
  • Stephen E. Harding
    Affiliations
    Department of Applied Biochemistry and Food Science, University of Nottingham, Sutton Bonington, LE12 5RD, United Kingdom
    Search for articles by this author
  • David A. Hopwood
    Affiliations
    Department of Genetics, John Innes Centre, Norwich Research Park, Colney, Norwich NR4 7UH, United Kingdom
    Search for articles by this author
  • E. Neil G. Marsh
    Correspondence
    To whom correspondence should be addressed. Tel.: 44-223-333622; Fax: 44-223-333345
    Affiliations
    Department of Biochemistry and Cambridge Centre for Molecular Recognition, University of Cambridge, Tennis Court Road, Cambridge CB2 1QW, United Kingdom
    Search for articles by this author
  • Author Footnotes
    * This work was supported in part by grants from the Royal Society and the Wellcome Trust (to E. N. G. M.). Work in the laboratory of D. A. Hopwood was supported by the BBSRC and the John Innes Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:July 21, 1995DOI:https://doi.org/10.1074/jbc.270.29.17339
      The biosynthesis of the polyketide antibiotic actinorhodin by Streptomyces coelicolor involves the oxidative dimerization and hydroxylation of a precursor, most likely dihydrokalafungin, as the final steps in its formation. Mutations in the actVB gene block these last steps, and the mutants secrete kalafungin as a shunt product. To investigate the role of the actVB gene in these transformations, we have overexpressed the gene in Escherichia coli and purified and characterized the recombinant protein. ActVB was shown to catalyze the reduction of FMN by NADH to give NAD and FMNH2, which, unusually, is released into solution. The protein contains no chromogenic cofactors and exhibits no requirements for added metal ions. The reaction obeys simple kinetics and proceeds through the formation of a ternary complex; Km values for FMN and NADH are 1.5 and 7.3 μM, respectively, and kcat is about 5 s−1. FAD and riboflavin are also substrates for the enzyme, although they have much higher Km values. The subunit structure of the enzyme was investigated by analytical ultracentrifugation, which showed the protein to exist in rapid equilibrium between monomer and dimer forms. The possible role of this oxidoreductase in the oxidative chemistry of actinorhodin biosynthesis is discussed.
      Polyketides comprise a large and diverse family of natural products, which are built up by the head-to-tail condensation of short carboxylic acid units, usually acetate or propionate residues(
      • O'Hagan D.
      ). They are produced by plants, animals, marine organisms, fungi, and, most prolifically, by actinomycete bacteria. Their biosynthesis may be divided into two stages: first, the carbon skeleton is assembled by the requisite polyketide synthase; second, a variety of “tailoring” reactions modify the polyketide structure by, for example, hydroxylation, glycosylation, or methylation.
      Much interest has focused on the assembly of the polyketide backbone, which is analogous to the biosynthesis of fatty acids(
      • O'Hagan D.
      ). The cloning and sequencing of the polyketide synthase genes for several polyketides has confirmed the analogy with fatty acid biosynthesis and has been instrumental in advancing our understanding of the principles of polyketide synthase programming(
      • Hopwood D.A.
      • Sherman D.H.
      ,
      • Katz L.
      • Donadio S.
      ,
      • McDaniel R.
      • Ebert-Khosla S.
      • Hopwood D.A.
      • Khosla C.
      ,
      • Shen B.
      • Hutchinson C.R.
      ,
      • Marsden A.F.A.
      • Caffrey P.
      • Aparicio J.C.
      • Loughran M.S.
      • Staunton J.
      • Leadlay P.F.
      ). In contrast, the enzymes that catalyze the tailoring reactions are generally less well understood, even though these later elaborations of the polyketide structure are usually crucial to the compound's biological activity. The biosynthesis of actinorhodin involves, as a late step, an unusual dimerization in which two benzoquinone units are symmetrically joined by a phenolic oxidative coupling (
      • Gorst-Allman C.P.
      • Rudd B.A.M.
      • Chang C.-J.
      • Floss H.G.
      ) (Fig. 1). Structural features arising from phenolic couplings are found in a very wide range of natural products, including the oligopeptide antibiotic vancomycin and opiate alkaloids(
      • Mann J.
      ), but the enzymic chemistry responsible for them remains largely unexplored. The cloning and sequencing of the relevant part of the actinorhodin biosynthetic cluster (
      • Fernández-Morano M.A.
      • Martnez E.
      • Boto L.
      • Hopwood D.A.
      • Malpartida F.
      ) presented an opportunity to study an example of this novel enzyme activity using molecular biology.
      Figure thumbnail gr1
      Figure 1:Oxidative biosynthesis of actinorhodin from dihydrokalafungin. Mutants blocked in actVB secrete kalafungin as a shunt product. The order of the final (C-8) hydroxylation and dimerization at C-10 is not established.
      One class of actinorhodin pathway mutants, actVB, is believed to lack the ability to perform the oxidative coupling step and secrete kalafungin as a shunt product (Fig. 1)(
      • Cole S.P.
      • Rudd B.A.
      • Hopwood D.A.
      • Chang C.-J.
      • Floss H.G.
      ). Kalafungin can be converted to actinorhodin(
      • Cole S.P.
      • Rudd B.A.
      • Hopwood D.A.
      • Chang C.-J.
      • Floss H.G.
      ), which suggests that either dihydrokalafungin or 8-hydroxy-dihydrokalafungin is likely to be the substrate for dimerization. The two available actVB mutations have been mapped to a single open reading frame encoding a protein of 177 amino acid residues (Mr 18,400) (
      • Fernández-Morano M.A.
      • Martnez E.
      • Boto L.
      • Hopwood D.A.
      • Malpartida F.
      ). Here, we describe the expression of the actVB gene in Escherichia coli and the purification and characterization of the recombinant protein, the first enzyme from the actinorhodin biosynthetic pathway to be characterized biochemically. Initially, we had thought that ActVB might catalyze the dimerization of either kalafungin or dihydrokalafungin directly; however, sequence similarity to a component of pristinamycin IIB hydroxylase1
      V. Blanc and D. Thibaut, personal communication.
      1V. Blanc and D. Thibaut, personal communication.
      encoded by snaC (
      • Blanc V.
      • Blanche J.
      • Crouzet N.
      • Jacques P.
      • Lacroix D.
      • Thibaut D.
      • Zagoreg M.
      ) and successful complementation of an actVB mutant by snaC2
      V. Blanc, manuscript in preparation.
      2V. Blanc, manuscript in preparation.
      suggested that the protein might function as a flavin:NADH oxidoreductase. The ActVB protein does indeed catalyze reduction of FMN by NADH and is therefore probably an essential auxillary enzyme supplying reduced FMN to another enzyme directly involved in oxidative chemistry.

      EXPERIMENTAL PROCEDURES

      Materials

      DNA restriction and modifying enzymes were from Promega. Isopropyl-1-thio-β-D-galactopyranoside was from Melford Laboratories (Ipswich, United Kingdom). Q-Sepharose fast flow anion exchange medium and Sephacryl S-300-HR gel filtration medium were purchased from Pharmacia Biotech Inc. Matrex gel Blue A (Cibacron Blue 3GA) was from Amicon. NADH, FMN, FAD, and riboflavin were from Boehringer Mannheim. pT7-7(
      • Tabor S.
      • Richardson C.C.
      ), 3
      S. Tabor, unpublished work.
      E. coli TG1 carrying the recO mutation(
      • Morrison P.T.
      • Lovett S.
      • Gilson L.E.
      • Kolodner R.C.
      ), and E. coli BL21(DE3) (
      • Studier F.W.
      • Moffatt B.A.
      ) were the kind gift of Dr. P. Caffrey (Dept. of Biochemistry, Cambridge University). Kalafungin, nanaomycin, dihydrokalafungin, and actinorhodin were the kind gift of Prof. S. Omura (Kitasato Institute, Tokyo).

      Construction of Expression Vector for ActVB

      The actVB gene had been subcloned as a 3-kb BglII fragment (encompassing restriction sites 19-21 as defined in (
      • Fernández-Morano M.A.
      • Martnez E.
      • Boto L.
      • Hopwood D.A.
      • Malpartida F.
      )) in pBR329 to give pIJ2347.4
      F. Malpartida, personal communication.
      pIJ2347 was digested with SmaI and SalI and a 1.1-kilobase fragment encompassing all of the actVB gene, except the first 39 nucleotides isolated and ligated into SmaI- and SalI-digested and dephosphorylated pT7-7. The ligation mixture was used to transform E. coli TG1 recO 1504::Tn5 by standard methods(
      • Sambrook J.
      • Fritsch E.F.
      • Maniatis T.
      ), and recombinant colonies were selected by plating on 2TY agar containing 100 μg/ml ampicillin. Recombinant plasmids were analyzed by restriction mapping; the pT7-7 derivative containing the actVB gene minus the first 39 nucleotides was called pACTVB-NT.
      To reconstruct the 5′-end of the actVB gene, two oligonucleotides, TATGGCTGCTGACCAGGGTATGCTGCGTGACGCTATGGCT and GAGCCATAGCGTCACGCAGCATACCCTGGTCAGCAGCCA, corresponding to sense and antisense strands, respectively, were synthesized. These oligonucleotides both optimized codon usage for highly expressed E. coli genes (
      • Grosjean H.
      • Fiers W.
      ) and allowed cloning into the NdeI site of the pT7-7 vector. The oligonucleotides were phosphorylated, annealed together, and ligated into pACTVB-NT previously digested with NdeI, EcoRI, and SmaI (EcoRI digestion served to prevent the polylinker fragment religating into pACTVB-NT). Transformants were selected by plating on 2TY agar containing 100 μg/ml ampicillin. To confirm that the actVB gene had been correctly reconstructed, plasmids were isolated from the recombinant bacteria, and the nucleotide sequence of the first 100 base pairs of the 5′-region was determined. The plasmid containing the actVB gene under the control of the T7 promoter was designated pACTVB.

      Overexpression of Recombinant ActVB Protein

      E. coli BL21(DE3) was transformed with pACTVB(
      • Sambrook J.
      • Fritsch E.F.
      • Maniatis T.
      ). A 10-ml sample of an overnight culture of a transformant was used to inoculate each of six 1-liter flasks containing 500 ml of 2TY media supplemented with 50 mM potassium phosphate, pH 6.9, 10% glycerol, 0.4% glucose, and 100 μg/ml ampicillin. Cultures were grown at 37°C until an A600nm of 3.5 was reached. These cells were harvested by centrifugation and resuspended into six 2-liter flasks containing 1 liter of the same medium but without glucose. Expression of ActVB was induced by adding 238 mg/liter isopropyl-1-thio-β-D-galactopyranoside, and the cells were grown for a further 2.5-3 h (A600nm∼4.0). Cells were harvested by centrifugation, washed with 50 mM Tris-HCl, pH 7.4, and stored at −20°C.

      Purification of Cloned ActVB Protein

      All steps were performed on ice or at 4°C. In a typical purification, 26-g cells (damp weight) were thawed on ice and resuspended in 50 ml of buffer A (50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 10% glycerol). Cells were ruptured by passing them twice through a French pressure cell operating at 15,000 p.s.i. internal pressure, and cell debris was removed by centrifugation (25,000 × g for 15 min). Nucleic acids were precipitated by the dropwise addition of 20% streptomycin sulfate solution to a final concentration of 4% and were removed by centrifugation (25,000 × g for 15 min). The supernatant was then brought to 20% saturation by the slow addition of 107 g/liter solid ammonium sulfate. Proteins precipitated by this step were removed by centrifugation (25,000 × g for 15 min), and the supernatant was brought to 60% saturation by the addition of a further 244 g/liter solid ammonium sulfate. The precipitated protein was recovered by centrifugation, and the pellet was resuspended in a minimal volume of buffer A. The protein was dialyzed overnight against 1 liter of the same buffer.
      The dialysis residue was cleared by centrifugation (25,000 × g for 15 min) and loaded onto a 2.5 × 100-cm Sephacryl S-300-HR gel filtration column equilibrated in buffer A. Proteins were eluted at a flow rate of 28 ml/h, and 3.5-ml fractions were collected. Fractions were analyzed for the presence of ActVB by SDS-PAGE5
      The abbreviation used is: PAGE
      polyacrylamide gel electrophoresis.
      and assayed for FMN:NADH oxidoreductase activity as described below. Pooled peak fractions were applied to a 2.5 × 15-cm Q-Sepharose fast flow anion exchange column equilibrated in buffer B (40 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 10% glycerol) at a flow rate of 60 ml/h. The column was washed with 300 ml of buffer B, and protein was eluted with a 600-ml linear gradient of KCl (0-0.3 M) in buffer B. Fractions (3.7 ml) were analyzed by SDS-PAGE and assayed for FMN:NADH oxidoreductase activity. Fractions containing flavin reductase activity were pooled and concentrated by ultrafiltration in a stirred cell fitted with an Amicon PM10 membrane (exclusion limit, 10 kDa). The ActVB protein, ∼80% pure, was stored at −20°C in the presence of 20% glycerol. Protein concentration was approximately 8-10 mg/ml.
      ActVB protein could be further purified on a small scale by dye ligand chromatography. Nine units of ActVB in 0.5 ml of 20 mM potassium phosphate buffer, pH 6.4, were applied to a small (2 ml) column of Amicon Matrix Blue A equilibrated in the same buffer. The protein was allowed to bind to the column for 30 min, and then proteins that remained unbound were removed by washing with 10 ml of phosphate buffer. ActVB was eluted with 1.5 M potassium chloride in 20 mM potassium phosphate, pH 6.4.

      Protein Analysis

      SDS-PAGE was carried out using the buffer system of Laemmli (
      • Laemmli U.K.
      ) in a 20% polyacrylamide gel, and protein bands were visualized by staining with Coomassie Brilliant Blue R250. The Pharmacia Phast gel system was used with an IEF5-8 gel to determine the isoelectric point of the protein. Protein concentrations were determined by the method of Bradford(
      • Bradford M.M.
      ); bovine serum albumin was used to construct standard curves. Recombinant protein identity was confirmed by N-terminal sequencing of polyvinylidene difluoride-blotted protein on an Applied Biosystems 477A protein sequencer. Exact protein molecular weights were determined on a Fission VG Bio Q Electrospray mass spectrometer. 200 pmol of protein in 20 μl of 50% acetonitrile, 0.1% trifluoroacetic acid were used per injection.

      Enzyme Assay

      Flavin:NADH oxidoreductase activity was assayed by monitoring the decrease in absorption of NADH at 340 nm (∊ = 6.22 × 103M−1 cm−1) in 50 mM Tris-HCl, pH 7.4, at 25°C. Steady state kinetic measurements were performed with a 1-cm light path in a final volume of 1 ml. For routine determination of enzyme activity, 100 μM NADH and 40 μM FMN were present, and assays were initiated by the addition of enzyme. To maintain enzyme activity, enzyme dilutions were made into 10% glycerol to give a working stock solution. Solutions containing FMN were kept in the dark to avoid photoreduction.
      Anaerobic assays were performed in gas-tight cuvettes under argon. Solutions were de-oxygenated by bubbling through with argon gas that had been scrubbed free of oxygen.

      Analytical Ultracentrifugation

      Sedimentation velocity measurements were performed on a MSE Centriscan 75 analytical ultracentrifuge, and sedimentation equilibrium measurements were performed on a Beckman XLA machine, using methods previously described (
      • Marsh E.N.
      • Harding S.E.
      ). Measurements were made at 20°C, and sedimentation was monitored at 280 nm using scanning absorption optics. Protein samples were made up at a concentration of 0.5-0.7 mg/ml in TES buffer, pH 7.5.

      RESULTS

      Expression of ActVB

      ActVB was overexpressed in E. coli by placing the gene under the control of strong bacteriophage T7 φ10 promoter(
      • Morrison P.T.
      • Lovett S.
      • Gilson L.E.
      • Kolodner R.C.
      ). The subcloning strategy involved reconstructing the 5′-portion of the gene using synthetic oligonucleotides (Fig. 2), and this allowed the first 13 codons to be changed to those found in highly expressed E. coli genes(
      • Grosjean H.
      • Fiers W.
      ). This has been shown to improve the expression levels of some Streptomyces proteins, which are otherwise poorly expressed in E. coli due to the high GC bias of Streptomyces DNA(
      • Gramajo H.C.
      • White J.
      • Hutchinson C.R.
      • Bibb M.J.
      ). Optimal levels of ActVB expression were achieved by first growing the cells until late exponential phase (A600nm∼3.5-4.0) in the presence of 0.4% glucose, which represses the basal level of T7 RNA polymerase expression. Expression was induced by removing them from the glucose-containing medium by centrifugation and resuspending in fresh medium containing isopropyl-1-thio-β-D-galactopyranoside. The cells were grown for a further 2.5-3 h, by which time they were nearing stationary phase, and expression of ActVB was maximal. The inclusion of 10% glycerol and 50 mM phosphate, pH 6.9, in the medium increased the yield of ActVB and may have increased the proportion of soluble recombinant protein. Typically, 26 g of cells were produced from 6 liters of medium, and in general the overexpressed protein comprised 5-10% of total cell protein as judged by SDS-PAGE.
      Figure thumbnail gr2
      Figure 2:Subcloning strategy for the construction of pACTVB. Inset, the protein sequence, the Streptomyces coelicolor 5′-DNA sequence, and the replacement 5′-oligonucleotide sequence (coding strand) for the N-terminal part of ActVB. Nucleotides underlined represent those changed during oligonucleotide synthesis for optimal codon usage in E. coli.

      Purification of ActVB

      ActVB was produced both in soluble form and as inclusion bodies in approximately equal amounts. The soluble protein, rather than the inclusion bodies, was purified since it was not known whether the protein could be refolded. To effect the purification, it was necessary to include 10% glycerol in all the buffers since aggregation of ActVB proved a considerable problem, and glycerol appeared to help the protein remain soluble. A purification based on size exclusion chromatography on Sephacryl S-300-HR and anion exchange chromatography on Q-Sepharose fast flow medium resulted in ActVB protein that was more than 80% pure. Removal of the remaining proteins proved more difficult. Various chromatographic techniques, including hydrophobic interaction chromatography on phenyl-Sepharose, fast protein liquid chromatography using a Pharmacia Mono Q column and affinity chromatography on FMN-agarose, failed to provide significant further purification. However, nearly pure ActVB protein could be produced on a small scale by binding it to a Cibacron Blue 3GA dye column. Although this step removed nearly all the remaining contaminating proteins, recovery was only 33% on this step, and the purified ActVB actually had a slightly lower specific activity. Attempts to scale up this purification step proved unsuccessful as very substantial losses of protein were encountered.
      After purification by dye column chromatography, ActVB was very nearly pure as judged by SDS-PAGE (Fig. 3). The two major protein bands seen on the gel corresponded to different forms of ActVB (as discussed below), and minor contaminating bands of Mr≈ 30,000 and 35,000 were also still present. The purified protein was stable when stored at −20°C in the presence of 20% glycerol. The yields and specific activities for each step are shown in Table 1.
      Figure thumbnail gr3
      Figure 3:Purification of ActVB. SDS-PAGE of samples taken after each step of the purification (gels stained with Coomassie Brilliant Blue) is shown. Lane1, soluble cell fraction; lane2, ammonium sulfate fractionation; lane3, pooled fractions after gel filtration chromatography; lane4, pooled fractions after ion exchange chromatography; lane5, protein after dye ligand chromatography. The positions and molecular weights of marker proteins are indicated at the sides of the gel.
      ActVB was found to precipitate irreversibly from solution when the pH was reduced below 6.5, which indicated a relatively high isoelectric point and restricted the choice of purification methods. The isoelectric point, calculated from the protein sequence, is 6.61, and the experimentally determined value by isoelectric focusing is 6.45. The low yields encountered in the dye column chromatography step were probably caused by the need to conduct this step at a pH close to the isoelectric point; at higher pH levels, ActVB was not bound by the column. Presumably, the losses encountered when trying to scale up the purification were due to the protein precipitating irreversibly on the column.
      During the expression of ActVB, a protein of lower molecular mass (≈16,000 Da) was also seen (Fig. 3). The smaller protein copurified with ActVB and had the same N-terminal sequence, indicating that it was derived from ActVB by proteolysis near the C terminus. The proportion of proteolyzed protein in the preparation could be reduced by decreasing the length of time that cells were grown after induction and by the use of proteolytic inhibitors throughout the purification.
      The site of proteolysis was investigated by electrospray mass spectrometry. The mass spectrum of a mixture of the two ActVB forms showed two major series of peaks. One of these was due to a species of Mr 18265 ± 2, which corresponds to ActVB from which the N-terminal methionine residue has been cleaved (calculated Mr = 18,262). The other species has an Mr of 16,645 ± 2, which would correspond to ActVB lacking the N-terminal methionine and cleaved after Cys-158 (calculated Mr = 16,642). Since it was not possible to separate the truncated protein from the full-length form, it is not clear whether it is active or not.

      Physical Characterization and Quaternary Structure of ActVB

      Initial estimations of molecular weight made by gel filtration on a calibrated Pharmacia Superose-12 fast protein liquid chromatography column at pH 7.5 yielded a Mr of 32,000 for ActVB, suggesting it to be dimeric. The subunit structure was investigated in more detail by analytical ultracentrifugation. The weight-averaged molecular weight over the whole solute distribution, M°r,w, was determined at several different pH values between 5.0 and 8.5. M°r,w decreased progressively, as the pH was raised, from 31,000 ± 2000 at pH 5.0 to 25,000 ± 1000 at pH 8.5. At pH 7.5, under the conditions in which gel filtration was performed, M°r,w for ActVB was 28,000 ± 2000. Only a single symmetrical boundary was seen during sedimentation velocity, indicating that the protein exists in a rapid equilibrium between monomer and dimer forms, with higher pH promoting dissociation. The point weight average molecular weight extrapolated to zero concentration gave a value of 18,000 ± 5000, which corresponds to the Mr of the monomeric protein, as expected for a system that is in true equilibrium between monomer and dimer. The sedimentation coefficient, s20,w, at pH 7.5, was 2.93 ± 0.08 S, a value typical for globular proteins of this size. At lower pH values, higher molecular weight species were detected at the cell base, again demonstrating the propensity of the protein to aggregate when near its pI.

      Substrate Specificity and Kinetics

      ActVB prepared from E. coli was colorless, and the UV-visible spectrum showed no evidence for any chromogenic cofactors. Oxidoreductase activity depended on both NADH and flavin being added in the assay; no requirement for any other cofactors was apparent. The most effective substrates were NADH and FMN, but FAD and riboflavin could be turned over by the enzyme, although the Km values were much higher. Under aerobic conditions, the oxidation of NADH could be effected with catalytic amounts of FMN, suggesting that FMNH2 was being recycled by reaction with oxygen to form H2O2 and FMN. This was confirmed when the assay was performed anaerobically. NADH now reduced FMN stoichiometrically, and the absorption band at 450 nm corresponding to FMN was completely bleached when excess NADH was added (Fig. 4). When oxygen was admitted to the assay, further oxidation of NADH took place, and the absorption at 450 nm due to FMN returned.
      Figure thumbnail gr4
      Figure 4:Reduction of FMN catalyzed by ActVB. i, UV-visible spectrum before addition of enzyme; ii, after complete reduction by NADH.
      The steady state kinetic properties of the enzyme were investigated; velocities were measured at 25°C in 50 mM Tris-HCl buffer, pH 7.4. The Km for NADH was determined at various concentrations of FMN, and the data were fitted by computer. This yielded values for the true Km for NADH of 7.3 ± 0.6 μM and Kmfor FMN of 1.5 ± 0.1 μM. The Hanes plot of the data (Fig. 5) clearly demonstrates an intersecting pattern characteristic of the formation of a ternary complex between ActVB, NADH, and FMN. The specific activity of ActVB was calculated as 8.2 units/mg, which, assuming that both forms of ActVB are active, gives a value for kcat of 5 s−1. The apparent Km values for FAD and riboflavin, determined at 100 μM NADH, were 11.5 ± 0.6 and 13.5 ± 0.6 μM, respectively.
      Figure thumbnail gr5
      Figure 5:Hanes plot of kinetic data for ActVB. Enzyme activity was measured at fixed concentrations of ActVB and varying concentrations of NADH and FMN. The concentrations of FMN used were 40 μM (▪), 20 μM (□), 10 μM (▴), 5 μM (○), 2 μM (♦), and 1 μM (Δ).
      The stoichiometry of FMN binding to ActVB was investigated using equilibrium ultrafiltration(
      • Wilson K.
      ). A ≈10 μM solution of ActVB protein was incubated in the presence of 50 μM FMN (i.e. a saturating concentration of FMN) for 5 min, and the solution was subjected to ultrafiltration using a membrane with a 10,000-Da cut off. The concentrations of FMN in the filtrate and retentate were then determined by the absorbance at 450 nm. The FMN concentration in the retentate was 58 μM, while that in the filtrate was 43 μM; therefore, the concentration of bound FMN was 15 μM. The protein concentration in the retentate was determined as 17 μM, and hence 88% of the protein bound FMN. Given that the ActVB preparation was approximately 85% pure, we interpret this as indicating that 1 mol of FMN is bound per mol of ActVB.

      DISCUSSION

      The cloning and sequencing of the actinorhodin biosynthetic gene cluster has afforded the opportunity to study in detail the biochemistry of polyketide assembly. ActVB is the first enzyme from this pathway to be overexpressed in E. coli and physically and kinetically characterized. The actVB gene was originally the focus of our investigation because mutations mapping to this gene abolished two chemically interesting steps, the final hydroxylation and dimerization, the enzymology of which is little explored.
      Initially, we thought that ActVB might catalyze either the hydroxylation or dimerization of kalafungin or dihydrokalafungin directly. However, extensive trial assays using a variety of different buffers, pH, metal ions, cofactors such as flavin, and both oxidized and reduced nicotinamide coenzymes failed to produce any evidence that ActVB could catalyze the hydroxylation or dimerization of either potential substrate. The initial suggestion that ActVB might be an oxidoreductase came from sequence similarity observed between ActVB and a component of pristinamycin IIB hydroxylase1 encoded by snaC (
      • Blanc V.
      • Blanche J.
      • Crouzet N.
      • Jacques P.
      • Lacroix D.
      • Thibaut D.
      • Zagoreg M.
      ) and, subsequently, successful complementation of an actVB mutant by snaC.2 Recently, the cloning of a FAD:NADH-dependent 4-hydroxyphenylacetate 3-hydroxylase from E. coli has been reported(
      • Prieto M.A.
      • Garcia J.L.
      ). The enzyme comprises two separable protein components, the smaller of which, HpaC, also has sequence similarity with ActVB. This suggests that HpaC may similarly function as flavin:NADH oxidoreductase. These enzymes may comprise a new class of reduced flavin-dependent oxidases.
      We subsequently confirmed that ActVB is a flavin:NADH oxidoreductase; the preferred substrate is FMN, but FAD and riboflavin are effective substrates at high concentrations. Since it was not possible to purify ActVB to homogeneity, the enzyme activity might possibly result from a contaminating protein. However, we feel this is unlikely because of the sequence similarity to SnaC and the fact that ActVB binds FMN stoichiometrically. Unusually, FMN behaves as a true substrate in this reaction rather than as a tightly bound cofactor. This is confirmed by the absence of any absorption bands attributable to flavin in the UV-visible spectrum of the purified protein, the quantitative reduction of substrate levels of FMN under anaerobic conditions in the presence of excess NADH, and steady state kinetic analysis, which clearly indicates the formation of a ternary complex of FMN and NADH with ActVB, implying direct reduction of FMN by NADH.
      The only well studied example in which free FMNH2 serves as a cosubstrate is the oxidation (by molecular oxygen) of aliphatic aldehydes by bacterial luciferase in luminescent bacteria. FMNH2 for this reaction is provided by FMN:NADH oxidoreductases, of which examples from Vibrio fischeri and Vibrio harveyi have recently been overexpressed in E. coli and characterized(
      • Zenno S.
      • Saigo K.
      • Kanoh H.
      • Inouye S.
      ,
      • Lei B.
      • Liu M.
      • Huang S.
      • Tu S.-C.
      ). ActVB shows no homology to these enzymes and differs in structure and kinetic properties. The Vibrio oxidoreductases are monomers of approximately 25,000 Da, while ActVB is dimeric and substantially smaller, with a subunit mass of 18,400 Da. The Vibrio enzymes contain tightly bound FMN, distinct from FMN undergoing reduction, and exhibit ping-pong kinetics, implying that an intermediate reduced flavoprotein is formed. In contrast, the reduction of FMN by ActVB, as discussed above, appears to proceed via a ternary complex.
      The order of the hydroxylation and dimerization steps that transform dihydrokalafungin to actinorhodin is unknown. On simple chemical grounds, dimerization was favored as the first step since the phenolic hydroxyl at C-11 would direct coupling to the ortho (C-10) position. Accumulation of kalafungin by the actVB mutants was therefore taken as evidence that they were blocked in dimerization itself(
      • Fernández-Morano M.A.
      • Martnez E.
      • Boto L.
      • Hopwood D.A.
      • Malpartida F.
      ,
      • Cole S.P.
      • Rudd B.A.
      • Hopwood D.A.
      • Chang C.-J.
      • Floss H.G.
      ). However, hydroxylation could precede dimerization if the dimerase could distinguish between the two para-substituted hydroxyl groups (at C-8 and C-11) and so direct the regiochemistry of coupling to C-10 rather than C-9. Evidence that one or more genes essential for C-8 hydroxylation are located in the actVA region of the act cluster (far from the actVB gene) came from the finding that this DNA segment could cause Streptomyces sp. AM 7161, the producer of medermycin, which lacks the C-8 hydroxyl, to make mederrhodin A, which is hydroxylated at C-8(
      • Hopwood D.A.
      • Malpartida F.
      • Kieser H.M.
      • Ikeda H.
      • Duncan J.
      • Fujii I.
      • Rudd B.A.M.
      • Floss H.G.
      • Omura S.
      ). Therefore, if ActVB is indeed involved in this hydroxylation step, as suggested by its similarity to HpaC and SnaC, Streptomyces sp AM 7161 must carry an equivalent gene.
      There are many examples in which reduced flavin is used to activate dioxygen toward hydroxylation chemistry, which occurs through the intermediacy of a flavin 4a hydroperoxide(
      • Ghisla S.
      • Massey V.
      ). The chemistry of the dimerization is more speculative. Dihydrogeodin oxidase, one of the few enzymes catalyzing oxidative couplings that have been characterized, requires copper and uses molecular oxygen as the oxidant(
      • Fujii I.
      • Iijma H.
      • Tsukita S.
      • Ebizaka Y.
      • Samaka U.
      ). However, flavin-mediated redox chemistry is sufficiently versatile that reduced FMN might well also play a role in the oxidative dimerization leading to actinorhodin. We are currently investigating genes within the actVA region of the actinorhodin biosynthetic cluster to identify the hydroxylase and dimerase enzymes and to clarify the role of FMNH2 generated by ActVB, the timing of hydroxylation and dimerization, and the chemistry associated with these transformations.

      Acknowledgments

      We thank Francisco Malpartida (Centro Nacional de Biotecnologia, Madrid) (who provided pIJ2347), Maureen Bibb, and Peter Revill (John Innes Centre) for helpful discussions. We also thank Paul Skelton (Dept. of Chemistry, University of Cambridge) for protein analysis by electrospray mass spectrometry and Mike Wheldon of the Protein and Nucleic Acid Facility (Dept. of Biochemistry, University of Cambridge) for the synthesis of oligonucleotides and protein sequence determination.

      REFERENCES

        • O'Hagan D.
        The Polyketide Metabolites. Horwood, Chichester, UK1991
        • Hopwood D.A.
        • Sherman D.H.
        Annu. Rev. Genet. 1990; 24: 37-66
        • Katz L.
        • Donadio S.
        Annu. Rev. Microbiol. 1993; 47: 875-912
        • McDaniel R.
        • Ebert-Khosla S.
        • Hopwood D.A.
        • Khosla C.
        Science. 1993; 262: 1546-1550
        • Shen B.
        • Hutchinson C.R.
        Science. 1993; 262: 1535-1540
        • Marsden A.F.A.
        • Caffrey P.
        • Aparicio J.C.
        • Loughran M.S.
        • Staunton J.
        • Leadlay P.F.
        Science. 1994; 263: 378-380
        • Gorst-Allman C.P.
        • Rudd B.A.M.
        • Chang C.-J.
        • Floss H.G.
        J. Org. Chem. 1981; 46: 455-456
        • Mann J.
        Secondary Metabolism. Clarendon Press, Oxford, UK1987
        • Fernández-Morano M.A.
        • Martnez E.
        • Boto L.
        • Hopwood D.A.
        • Malpartida F.
        J. Biol. Chem. 1992; 267: 19278-19290
        • Cole S.P.
        • Rudd B.A.
        • Hopwood D.A.
        • Chang C.-J.
        • Floss H.G.
        J. Antibiot. (Tokyo). 1987; 40: 340-347
        • Blanc V.
        • Blanche J.
        • Crouzet N.
        • Jacques P.
        • Lacroix D.
        • Thibaut D.
        • Zagoreg M.
        Patent WO 94/08014, France. 1994;
        • Tabor S.
        • Richardson C.C.
        Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 1074-1078
        • Morrison P.T.
        • Lovett S.
        • Gilson L.E.
        • Kolodner R.C.
        J. Bacteriol. 1989; 171: 3641-3649
        • Studier F.W.
        • Moffatt B.A.
        J. Mol. Biol. 1986; 189: 113-130
        • Sambrook J.
        • Fritsch E.F.
        • Maniatis T.
        Molecular Cloning. A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989
        • Grosjean H.
        • Fiers W.
        Gene (Amst.). 1982; 18: 199-209
        • Laemmli U.K.
        Nature. 1970; 227: 680-685
        • Bradford M.M.
        Anal. Biochem. 1976; 72: 248-254
        • Marsh E.N.
        • Harding S.E.
        Biochem. J. 1993; 290: 551-555
        • Gramajo H.C.
        • White J.
        • Hutchinson C.R.
        • Bibb M.J.
        J. Bacteriol. 1991; 173: 6475-6483
        • Wilson K.
        Wilson K. Walker J.M. Principles and Techniques of Practical Biochemistry. Cambridge University Press, Cambridge1994: 223
        • Prieto M.A.
        • Garcia J.L.
        J. Biol. Chem. 1994; 269: 22823-22829
        • Zenno S.
        • Saigo K.
        • Kanoh H.
        • Inouye S.
        J. Bacteriol. 1994; 176: 3536-3543
        • Lei B.
        • Liu M.
        • Huang S.
        • Tu S.-C.
        J. Bacteriol. 1994; 176: 3552-3558
        • Hopwood D.A.
        • Malpartida F.
        • Kieser H.M.
        • Ikeda H.
        • Duncan J.
        • Fujii I.
        • Rudd B.A.M.
        • Floss H.G.
        • Omura S.
        Nature. 1985; 314: 642-644
        • Ghisla S.
        • Massey V.
        Eur. J. Biochem. 1989; 181: 1-17
        • Fujii I.
        • Iijma H.
        • Tsukita S.
        • Ebizaka Y.
        • Samaka U.
        J. Biochem. (Tokyo). 1987; 101: 11-17