Eur. J. Biochem. 268, 1712±1721 (2001) q FEBS 2001
Cyclic dipeptide oxidase from Streptomyces noursei Isolation, purification and partial characterization of a novel, amino acyl a,b-dehydrogenase Muriel Gondry, Sylvie Lautru, Guillaume Fusai, Gilles Meunier, Andre MeÂnez and Roger Genet CEA /Saclay, DeÂpartement d'IngeÂnierie et d'Etudes des ProteÂines, Gif-sur-Yvette, France
Cyclic dipeptide oxidase is a novel enzyme that specifically catalyzes the formation of a,b-dehydro-Phe (DPhe) and a,b-dehydro-Leu (DLeu) residues during the biosynthesis of albonoursin, cyclo(DPhe-DLeu), an antibiotic produced by Streptomyces noursei. It was purified 600-fold with a 30% overall recovery, and consists of the association of a single type of subunit with a relative molecular mass of 21 066 resulting in a large homopolymer of relative molecular mass over 2 000 000. The enzyme exhibits a typical flavoprotein spectrum with maxima at 343.5 and 447.5 nm, the flavin prosthetic group being covalently bound to the protein. The catalytic reaction of the natural substrate cyclo(l-Phe-l-Leu) occurs in a two-step sequential reaction leading first to cyclo(a,b-dehydro-Phe-l-Leu)
and finally to albonoursin. Kinetic parameters for the first step were determined (Km 53 mm; k 0.69 s21). The enzyme was shown to catalyze the conversion of a variety of cyclo(dipeptides) and can be reoxidized at the expense of molecular oxygen by producing H2O2. This reaction mechanism, which differs from those already described for the formation of a,b-dehydro-amino acids, might consist of the transient formation of an intermediate imine followed by its rearrangement into an a,b-dehydro-residue.
Peptide secondary metabolites, as produced by prokaryotes and lower eukaryotes, exhibit great structural diversity. These bioactive products include antibiotics, plant and animal toxins, immunosuppressants, and enzyme inhibitors. Such metabolites are derived from peptides synthesized by the conventional ribosomal approach or produced in a nonribosomal way by peptide synthetases [1]. A number of these peptides may contain nonproteinogenic residues, including dehydro-, hydroxy-, methyl- or d-residues, which can be either formed and then incorporated during the peptide elongation by modular peptide synthetases [2,3] or produced by enzymatic processing from peptide or pseudopeptide metabolites by specifically dedicated enzymes [4±7]. Among these unconventional residues, a,b-dehydroamino acids are widely distributed in microbial metabolites [8±11]. Dehydro-metabolites exhibit a variety of biological activities and an increased resistance to proteolytic degradation, which can be related to conformational constraints and restricted orientation of the a,b-dehydro-amino acyl side chain [12]. In some cases, a,b-dehydro-amino acids
can also behave as key intermediates in biosynthetic processes owing to the electrophilicity of the double bond which allows further addition reactions. Such reactions occur for instance during the biosynthesis of microbial lantibiotics, from a,b-dehydro-alanyl or a,b-dehydrocysteinyl residues [6,13]. In another case, it was postulated that a,b-dehydro-tryptophan might be the precursor in the biosynthesis of violacein, a blue pigment produced by Chromobacterium violaceum [14]. The large occurrence and variety of dehydro-metabolites suggest that a large number of specific enzymes might be involved in the formation of a,b-dehydro-amino acids. Our goal was thus to document the diversity of these enzyme systems to provide a better understanding of the role and function of a,b-dehydro-amino acids in biological processes. In the present paper, we provide new insights into the mechanism of formation of a,b-dehydro-amino acids, in particular in the case of cyclic dipeptide metabolites, the biosynthesis pathways of which remain poorly understood. We describe the isolation, purification and partial characterization of a new enzyme system involved in the nonribosomal biosynthesis of albonoursin in Streptomyces noursei. Albonoursin, a bis-dehydro-diketopiperazine of structure cyclo(DPhe-DLeu), exhibits an antibiotic activity towards some Bacillus spp. and a weak antitumour activity [15]. We demonstrate that the new enzyme is directly responsible for the sequential conversion of cyclo(l-Phe-l-Leu) to cyclo(DPhe-l-Leu) and then cyclo(DPhe-DLeu) (Scheme 1). The enzyme is a homopolymeric flavoenzyme, with novel structural and mechanistic characteristics. In addition, we show it transforms various cyclo(dipeptides) into cyclo(a,b-dehydro-dipeptides)
Correspondence to R. Genet or M. Gondry, CEA /Saclay, DeÂpartement d'IngeÂnierie et d'Etudes des ProteÂines, F91191 Gif-sur-Yvette Cedex, France. Fax: 1 33 169089071, Tel.: 1 33 169082576, E-mail:
[email protected] or
[email protected] Abbreviations: D, a,b-dehydro-; DPhe, a,b-dehydro-Phenylalanine; DLeu, a,b-dehydro-Leucine. Enzyme: cyclo(dipeptide):oxygen oxidoreductase (EC 1.3.3.x). Note: a web site is available at http://www-dsv.cea.fr (Received 24 November 2000, accepted 19 January 2001)
Keywords: a, b-dehydro-amino acid; albonoursin; amino acyl a, b-dehydrogenase; diketopiperazine; Streptomyces noursei.
q FEBS 2001
Cyclic dipeptide oxidase from S. noursei (Eur. J. Biochem. 268) 1713
O NH HN
at 200 r.p.m. (CNRS/Laboratoire de fermentation, Gif-surYvette, France). The culture was stopped after 28 h, at the end of the exponential phase of bacterial growth. The cells were harvested by centrifugation and then stored frozen at 280 8C. Total yield was about 4150 g of fresh bacteria including remaining agar from a 200-L culture.
O cyclo(L-Phe-L-Leu) Detection of enzyme activity in vivo
O NH HN O cyclo( ∆ Phe-L-Leu) O NH HN O
cyclo(∆ Phe-∆Leu) Albonoursin
Scheme 1. Sequential metabolization of cyclo(l l-Phe-l l-Leu) to albonoursin in S. noursei cultures.
that can adopt both E and Z configurations. We named this new enzyme cyclic dipeptide oxidase or cyclo(dipeptide): oxygen oxidoreductase.
E X P E R I M E N TA L P R O C E D U R E S Materials The S. noursei strain (ATCC 11455) was obtained from the American Type Culture Collection (ATCC). N-Acetyl-lPhenylalanyl-l-Leucine, N-acetyl-l-Leucyl-l-Phenylalanine, cyclo(l-Phe-l-Leu), cyclo(l-Trp-l-Trp), cyclo(l-Phe-Gly), cyclo(l-Ser-Gly) and cyclo(l-Glu-Gly) were from Bachem AG. Protein standards for gel filtration chromatography were from Bio-Rad. Other protein standards for native or SDS-polyacrylamide gel electrophoreses and the ampholytes for isoelectric focusing or agarose-EF were from Amersham Pharmacia Biotech. All other products were from Sigma-Aldrich. Microorganism and cultivation S. noursei was grown at 28 8C with shaking in ATCC medium 5 supplemented with 0.1% agar or in RPR liquid medium containing 30 g´L21 bio-trypcase (bioMerieux), 1 g´L21 KH2PO4, 1 g´L21 agar. For enzyme purification, S. noursei cells were cultivated in 100 mL of medium 5 for 24 h. The resulting culture was inoculated for 24 h into a F 020 Chemap fermenter containing 17 L of medium 5 and then transferred to a F 300B Chemap fermenter containing 200 L of the same medium, aerated by mechanical stirring
Enzyme activity was detected throughout the bacterial growth by following the conversion of tritiated cyclo(lPhe-l-Leu) to tritiated dehydro-products. [3H]Cyclo(l-Phel-Leu) was prepared according to Evans et al. [16] by isotopic exchange of tritium gas on the aromatic ring of the phenylalanyl residue, in the presence of palladium oxide as a catalyst. [3H]Cyclo(l-Phe-l-Leu) was obtained with a specific radioactivity of 19.2 GBq/mmole. 3.7 MBq of [3H]cyclo(l-Phe-l-Leu) were added to 5 mL of RPR medium at the initiation of the bacterial culture. 150 mL fractions of culture filtrate, i.e. 110 kBq, were taken at increasing incubation times and filtered using Ultrafree-MC (Millipore) with a 10-kDa cut-off. The resultant samples were analyzed by reverse-phase chromatography (C18 column 4.6 250 mm, Vydac) using a linear gradient from 0% to 55% acetonitrile in 0.1% trifluoroacetic acid for 55 min (flow rate, 1 mL´min21). Conversion products were identified by comparison with reference products. Enzyme purification All procedures were carried out at 4 8C in the presence of 5 mm EDTA and 2 mm phenylmethanesulfonyl fluoride unless specified otherwise. Centrifugation at 39 200 g for 30 min was performed throughout the purification procedure. Protein concentrations were determined using a Bio-Rad protein assay kit with bovine serum albumin as standard [17]. Cell disruption and extraction. Frozen S. noursei cells (130 g) were suspended in 130 mL of 100 mm Tris/HCl buffer, pH 7, containing 5% glycerol. The bacterial suspension was passed twice through a Manton Gaulin homogenizer. The extract was then centrifuged. Ammonium sulfate precipitation. The supernatant was brought to 30% ammonium sulfate saturation. After centrifugation, the enzyme was precipitated by bringing the ammonium sulfate saturation to 60%. The resulting precipitate was then slowly redissolved and equilibrated in 50 mm Tris/HCl buffer, pH 7, containing 0.2 m NaCl. Thermal denaturation. The enzyme solution was incubated at 50 8C for 30 min and centrifuged to remove the precipitate. Chromatographic steps. A three-step chromatographic purification was applied: (a) ion exchange chromatography using a Q-Sepharose Fast Flow column (50 300 mm, Pharmacia); (b) hydrophobic chromatography using an EMD propyl column (20 50 mm, Merck); and, finally (c), gel filtration on a Superose 6 HR column (10 300 mm, Pharmacia). Details are given in the figure
1714 M. Gondry et al. (Eur. J. Biochem. 268)
legends. Enzymatic fractions were pooled before being stored frozen at 280 8C. Enzyme assays Cyclo(l-Phe-l-His) was used throughout this work as a model substrate. The standard assay mixture comprised 1 mm cyclo(l-Phe-l-His) in 100 mm Tris/HCl buffer, pH 8. The reaction was initiated by adding the enzyme, and the increase in absorbance at 297 nm due to the formation of cyclo(DPhe-l-His) was monitored at 30 8C using a Uvikon 943 spectrophotometer. An average value of molar absorption coefficient of 19 mm21´cm21 at 297 nm, estimated from the values given by Shimohigashi et al. [18] for a series of DPhe derivatives, was used throughout this study. One unit of the enzyme was defined as the amount catalyzing the formation of 1 mmol of cyclo(DPhe-lHis) per min under standard assay conditions. Specific activity was expressed as enzyme units per mg of protein. Molar activity can be defined as 1 mol of cyclo(DPhe-lHis) produced per s per mol of enzyme monomer, expressed in s21. Anaerobic assays were performed as above, except that an oxygen-scavenging system was added to the reaction mixture [19]. In brief, the standard assay mixture was supplemented with 0.4 mm protocatechuate and deaerated for 10 min with argon. 0.5 U protocatechuate dioxygenase was then added and, after a 10-min incubation at 30 8C, the enzyme was in turn added to the solution. Production of H2O2 was determined by measuring the oxidation of 4-hydroxyphenylacetic acid into a fluorescent product (lexc 318 nm; lem 405 nm) in a horseradish peroxidase-coupled assay [20]. The reaction mixture comprised 1 mm cyclo(l-Phe-l-His) in 100 mm Tris/HCl buffer, pH 8, 1 mm 4-hydroxyphenylacetic acid and 1 U horseradish peroxidase. The reaction was initiated by adding the enzyme, and the increase in fluorescence at 405 nm was monitored at 30 8C using a Jasco FP-750 spectrofluorometer. Fluorescence intensity was correlated with H2O2 production through the use of standard curves established using known amounts of H2O2. Kinetic analysis Kinetics under standard conditions were monitored at 30 8C using a Uvikon 943 spectrophotometer. Initial velocities were evaluated during the first minute of the reactions at various substrate concentrations (enzyme concentration, 2.11 mg´mL21). Km and k parameters were determined using kaleidagraph (version 3.0.8) software (Abelbeck Software) by nonlinear regression from the direct plot. Molecular mass determination of the enzyme The relative molecular mass of the native enzyme was determined by high performance gel permeation chromatography from its mobility relative to the mobility of protein standards under the conditions described above. The mass spectrum of the purified enzyme was determined by Atheris Laboratories (Switzerland) in the linear and positive mode on a MALDI-TOF mass spectrometer (Voyager Elite, PerSeptive Biosystems). External calibration of the mass scale was performed with a solution of horse myoglobin. The sample was mixed with a saturated solution
q FEBS 2001
(10 mg´mL21) of sinapinic acid in 35% acetonitrile and 0.1% aqueous trifluoroacetic acid as a matrix and then analyzed. Polyacrylamide gel electrophoresis SDS/PAGE was performed in 5-mm thick precast polyacrylamide gradient gels (8±18%) (ExcelGel-SDS; Amersham Pharmacia Biotech) with a 2117-Multiphor II unit. The relative molecular mass of the monomer was determined as compared with those of protein standards. Native PAGE was carried out under standard conditions using the CleanGel 25S and buffer kit native pH 8.9 (Amersham Pharmacia Biotech) in 0.5-mm thick precast 10% polyacrylamide gels. The resultant gels were either subjected to silver staining [21] or submitted to a seconddimensional separation by SDS/PAGE as described above, or else cut into 2-mm slices for enzyme activity determination by incubating slices under standard assay conditions for 24 h. The incubation mixtures were filtered using Ultrafree-MC (Millipore) with a cut-off of 10 kDa and then injected into a reverse-phase C18 column, which was developed as previously described. Isoelectric focusing The isoelectric point of the native enzyme was determined on agarose gel electrophoresis [22], from its mobility relative to the mobility of protein standards. A 1-mm thick agarose-EF gel (1%) containing 7.2% preblended ampholines, pH 3.5±9.5, was used. The enzyme activity was assayed from 2-mm slices (see above). Mass spectrometry analysis of reaction products Purified dehydro-products, or crude reaction mixtures obtained after incubating dipeptides in the presence of cyclic dipeptide oxidase, were both analyzed by mass spectrometry. Analyses were carried out in positive ionization mode on a Quattro II analysis mass spectrometer fitted with an electrospray ion source, under control of the Mass Lynx data system (Micromass). The lyophilized samples were dissolved in water/acetonitrile (50 : 50) containing 0.2% formic acid, and portions of 8 mL were introduced into the mass spectrometer at a flow rate of 10 mL´min21. Assuming that the presence of the double bond does not affect the ionization process for a specific cyclodipeptide and its derived dehydro-products, quantification was estimated on the basis of their relative signal intensity, proportional to their concentration. This was further confirmed by using cyclo(Gly-l-His) as an internal standard.
R E S U LT S Detection of the enzyme activity in vivo [3H]Cyclo(l-Phe-l-Leu) was added at the initiation stage of the S. noursei culture and its metabolization was followed throughout bacterial growth. The exogenous substrate was modified by the cells into two main tritiated products which were identified by HPLC-chromatography as [3H]cyclo(DPhe-l-Leu) and [3H]cyclo(DPhe-DLeu), using
q FEBS 2001
Cyclic dipeptide oxidase from S. noursei (Eur. J. Biochem. 268) 1715
Dehydrogenation Yield (%)
6 5
O 3H
NH
4
HN O cyclo(∆Phe-∆Leu)
3 O
2
3
H
NH HN
1
O cyclo(∆Phe-L-Leu)
0 0
50
100
150
200
Time (h) Fig. 1. Time-course conversion of exogenous [3H]cyclo(l l-Phe-l ll-Leu) then [3H]cyclo(DPhe-DLeu) by Leu) into [3H]cyclo(DPhe-l S. noursei cells. [3H]Cyclo(l-Phe-l-Leu) (3.7 MBq) was added to 5 mL of RPR medium at the initiation of the S. noursei cultivation. Aliquots of the culture filtrate were taken at various time intervals and then subjected to reverse-phase chromatography. The yield of conversion into [3H]cyclo(DPhe-l-Leu) (X) and [3H]cyclo(DPheDLeu) (W) was calculated by comparing the radioactivity associated with each compound to the total injected radioactivity.
reference compounds. Figure 1 shows that, in the first 10 h of bacterial growth, [3H]-cyclo(l-Phe-l-Leu) is converted exclusively into [3H]-cyclo(DPhe-l-Leu) to reach a maximal yield of conversion of about 3.5%. Then, the level of [3H]cyclo(DPhe-l-Leu) progressively decreases with increasing concentration of [3H]cyclo(DPhe-DLeu), to completely vanish after a 100-h incubation period. The level of [3H]cyclo(DPhe-DLeu) levels off at 5.5%, and then remains stable up to 200 h (Fig. 1). These results clearly show that albonoursin is formed in a sequential process, the cyclo(DPhe-l-Leu) being an intermediate in the biosynthesis of the final antibiotic (Scheme 1). Expression and distribution of cyclo(dipeptide) dehydrogenase activity Cyclo(l-Phe-l-Leu) dehydrogenation occurs essentially during the exponential growth phase of S. noursei culture concomitantly with the biosynthesis of the antibiotic albonoursin (data not shown). Dehydrogenase activity was retained in extracts of disrupted cells; the whole activity was recovered in the supernatant after centrifugation for 30 min at 39 200 g, suggesting that the enzyme is located in the cytoplasm of the bacteria. Purification of cyclo(dipeptide) dehydrogenase activity The enzyme was purified to homogeneity by a three-step chromatographic procedure (Fig. 2). Purification yields and enzymatic activities were determined under standard conditions (Table 1). To follow enzyme enrichment throughout our purification protocol, we used cyclo(l-Phe-l-His) as a model substrate because of higher solubility in aqueous solution as compared to that of the natural substrate, cyclo(l-Phe-l-Leu). The overall purification yielded 1.9 mg of enzyme isolated from 130 g of frozen bacteria, with a 600-fold increase in specific activity and a 30%
Fig. 2. Three-step chromatographic procedure for the purification of cyclic dipeptide oxidase. The UV absorption at 280 nm is expressed as the relative absorbance and the enzyme activity was measured under standard kinetic conditions. (A) Ion exchange chromatography. About 280 mg of protein resulting from the thermal denaturation step were loaded at 5 mL´min21 onto a Q Sepharose Fast Flow column (50 300 mm, Pharmacia) equilibrated with 50 mm Tris/HCl buffer, pH 7, containing 0.2 m NaCl. The enzyme was eluted at 10 mL´min21 with a step gradient from 0.2 m NaCl to 1 m NaCl. Enzymatically active fractions were pooled and adjusted to 1 m ammonium sulfate. (B) Hydrophobic chromatography. The purified fraction (0.2 mg´mL21 protein) was applied to an EMD propyl column (20 50 mm, Merck) equilibrated with 50 mm Tris/HCl buffer, pH 8, containing 1 m ammonium sulfate. Elution was carried out at 25 8C at 1.5 mL´min21 with a step gradient from 1 m to 0 m ammonium sulfate. Enzymatically active fractions were concentrated by precipitation at 40% ammonium sulfate saturation and then equilibrated with 50 mm Tris/HCl buffer, pH 8, containing 100 mm NaCl. (C) High performance gel permeation chromatography. 256 mg of protein were subjected to gel filtration on a Superose 6 HR column (10 300 mm, Pharmacia). The column was developed at 25 8C at 0.5 mL´min21, with 50 mm Tris/HCl buffer, pH 8, containing 100 mm NaCl. The spectrum of enzymatically active fractions (about 130 mg´mL21 protein) in 100 mm Tris/HCl buffer, pH 8, is shown in the inset.
recovery. The specific activity of the purified enzyme was estimated to be 1.84 ^ 0.05 U´mg21. The purified enzyme was stable in the absence of any additive for 24 h at room temperature and for several months upon storage at 280 8C.
1716 M. Gondry et al. (Eur. J. Biochem. 268)
q FEBS 2001
Table 1. Purification of cyclic dipeptide oxidase from S. noursei. Enzyme purification from 130 g of frozen bacterial cells, wet weight.
Step
Total protein (mg)
Total activity (U)
Specific activity (1023 U´mg21)
Yield (%)
Purification (-fold)
Crude extract 30±60%-saturated ammonium sulfate Thermal denaturation at 50 8C Q-Sepharose EMD Propyl Superose 6 after 40%-saturated ammonium sulfate
2940 414 280.5 76 4.8 1.9
9.2 11.8 10.1 8.8 4.4 3.5
3.1 28.6 36 115 920 1842
± 100 85.6 74.6 37.3 29.7
1.0 9.2 11.6 37.1 297.8 594.2
The purified enzyme exhibits a yellow colour related with a typical absorption spectrum of a flavin-containing protein, as inferred from the presence of absorption peaks at (lmax) 343.5 and 447.5 nm (Fig. 2C). The flavin cofactor was not released from the protein during thermal or acidic denaturation, or in the presence of 6 m guanidinium chloride, showing a covalent association to the protein. Molecular mass and oligomeric structure At each step of the purification procedure, a progressive enrichment of a single protein band was detected by SDS/PAGE and its relative molecular mass (Mr) was estimated to be approximately 23 000 (Fig. 3). This estimation was refined by mass spectrometry analysis and a molecular ion peak was detected at m/z 21 067 ([M 1 H]1) (not shown). When subjected to gel-filtration chromatography, the native purified enzyme showed an apparent hydrodynamic volume large enough to be excluded on a Superose 6 column (Fig. 2C), suggesting an apparent relative molecular
Fig. 3. SDS/PAGE of enzyme fractions throughout the purification. The gel consisted of a 0.5-mm thick precast linear polyacrylamide gradient gel (8±18%) with silver staining. Lane T, molecular mass standards: phosphorylase b (94 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor (20.1 kDa), a-lactalbumin (14.4 kDa); lane 1, thermal denaturation (< 1.5 mg proteins); lane 2, ion exchange-purified enzymatically active fraction (< 1 mg); lane 3, hydrophobic interaction-purified active fraction (< 0.3 mg); lane 4, gel permeationpurified active fraction (< 0.3 mg).
mass over 2000 000. The same result was obtained from the extraction step despite the presence of detergents (1% Nonidet-P40 or 1% Chaps), sulfobetaines (1 m SB 201 or 1 m SB 195) [23] or reducing agents (2 mm Tris(2-carboxyethyl)phosphine hydrochloride or 5 mm dithiothreitol). Isoelectric focusing on a thin-layer 1% agarose gel revealed a single band, with a pI of approximately 3.8, which was found to be enzymatically active. These results, together with SDS/PAGE analysis (Fig. 3), suggested a homopolymeric structure of the enzyme. The purified enzyme was then submitted to 10% polyacrylamide native gel electrophoreses and the resultant gels were either subjected to silver staining and enzyme assay, or submitted to a second-dimensional separation by SDS/PAGE as described under Experimental procedures (Fig. 4). Protein bands were detected all along the gel during the first-dimensional separation (Fig. 4A), each band exhibiting an enzymatic activity which appears closely related to the protein concentration (Fig. 4B). This dissociation pattern strongly suggests that the enzyme underwent an ordered disruption of subunit association. The
Fig. 4. Native PAGE analysis of purified enzyme. The general experimental conditions are given under Experimental procedures. (A) Native PAGE of purified enzyme. The gel consisted of a 5-mm thick precast 10% polyacrylamide. Left-hand lane, molecular mass standards: thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), lactate dehydrogenase (140 kDa), albumin (66 kDa); right-hand lane, purified enzyme (1.5 mg). (B) Enzymatic activity determination of protein bands from native PAGE, which were incubated under standard conditions for 24 h.
q FEBS 2001
Cyclic dipeptide oxidase from S. noursei (Eur. J. Biochem. 268) 1717 Table 2. Conversion of various dipeptides to mono- and bisdehydro-products after incubation with cyclic dipeptide oxidase. The dipeptides were incubated for 18 h at 30 8C at a concentration of 2.4 mm, with the exception of cyclo(l-Phe-l-Leu) and cyclo(l-Trp-lTrp) which were, respectively, used at 0.3 mm and 1 mm, with 5% dimethylsulfoxide.
Fig. 5. Analysis of the enzymatic conversion of cyclo(l l-Phe-l l-His is). 3 mm cyclo(l-Phe-l-His) was incubated in 100 mm Tris/HCl buffer, pH 8, in the presence of 1.6 U, for 72 h at 30 8C. (A) Chromatogram at 338 nm of the reaction mixture. The components were separated by reverse-phase chromatography as described under Experimental procedures. (B) UV spectra of the products standardized according to their maximal intensity.
second-dimensional separation of the S. noursei enzyme by SDS/PAGE clearly showed that each previously detected band comprised exclusively one type of subunit, with a relative molecular mass of around 23 000 (not shown) and confirmed the polymeric structure of the enzyme. Determination of the N-terminal residue of the peptide sequence by the phenylisothiocyanate degradation procedure (Edman degradation) gave no phenylthiohydantoin derivative, indicating that the N-terminus of the protein is probably blocked. Dissociation experiments of the native enzyme in relatively mild conditions, such as 0.8 m KSCN for 12 h, led to a flavoprotein component of about 23 000 Da (as calibrated by gel filtration) with a residual enzyme activity. Further experiments are in progress to better characterize the quaternary structure of the polymeric enzyme. Enzyme activity and substrate specificity The purified enzyme was incubated in the presence of the model substrate cyclo(l-Phe-l-His) for 72 h at 30 8C. Reverse-phase chromatography of the reaction showed six resolved peaks (Fig. 5A) and the corresponding products were tentatively identified on the basis of both their UV spectra and mass spectrometry. Peak 2 (retention time, 19.9 min), the major component observed in the 220-nm chromatogram (not shown), was characterized by an absorption band at (lmax) 297 nm and a molecular ion peak at m/z 283 ([M 1 H]1), and identified as cyclo(DPhel-His). Peak 1 (retention time, 18.8 min) coeluted with the nonconverted substrate; once purified, it displayed a UV spectrum with an absorption band at (lmax) 297 nm and was thus assumed to be the second stereoisomer (Z or E isomer) of the cyclo(DPhe-l-His). Peaks 3 and 6 (retention times, 23.4 and 28.5 min), characterized by the same absorption band at (lmax) 338 nm (Fig. 5B) and a similar molecular ion peak at m/z 281 ([M 1 H]1), were identified as two stereoisomers of the cyclo(DPhe-DHis). Finally, peaks 4 and 5 (retention times, 27.2 and 27.5 min), both displayed UV spectra with an absorption band at (lmax) 350 nm (Fig. 5B). Though it was not possible to clearly
Dipeptides
Mono-dehydro-products Bis-dehydro-products (%) (%)
Cyclo(l-Phe-l-Leu) Cyclo(l-Phe-l-His) Cyclo(l-Trp-l-Trp) Cyclo(l-Leu-l-Ala) Cyclo(l-Phe-Gly) Cyclo(l-Leu-Gly) Cyclo(l-Ser-Gly) Cyclo(l-Glu-Gly) N-Ac-l-Phe-l-Leu N-Ac-l-Leu-l-Phe
24 23 74 13 36 9 11 3 0 0
67 ,1 14 4 ± ± ± ± 0 0
identify these two minor products due to their low concentration, we rationally assumed they correspond to two minor stereoisomers of the cyclo(DPhe-DHis) as referred to their UV spectra (Fig. 5B). All these results, together with those from kinetic experiments, showed that the enzymatic conversion of cyclo(l-Phe-l-His) led to the formation of first DPhe and then DHis residues which could adopt both E and Z configurations. Though it was not possible to obtain sufficient quantities of each dehydroproduct for subsequent characterization, similar results were observed in the presence of cyclo(l-Phe-l-Leu). However, as naturally occurring albonoursin purified from S. noursei cultures was shown to display a Z,Z geometry [24], it is tempting to conclude that the in vivo enzymatic reaction could predominantly lead to dehydro-products under Z configuration, the most stable from a thermodynamic point of view. Substrate specificity was investigated by following the dehydrogenation of a variety of dipeptides incubated in the presence of purified enzyme, in 50 mm ammonium carbonate for 18 h at 30 8C. Table 2 shows that linear dipeptides (N-Ac-l-Phe-l-Leu and N-Ac-l-Leu-l-Phe) do not behave as substrates, whereas all the tested cyclic dipeptides undergo an a,b-dehydrogenation, suggesting that the diketopiperazine ring is essential for the enzymatic reaction. In particular, the conversion of cyclo(l-Phe-lLeu) leads to mono- and bis-dehydroproducts, i.e. the albonoursin cyclo(DPhe-DLeu). Nevertheless, dehydrogenation yields depend on the nature of the targeted side chains. To a first approximation, cyclodipeptides constituted of aromatic as well as hydrophobic residues appear to be the best substrates. Also, the nature of the second side chain seems to affect the rate of dehydrogenation of the first one. Catalytic properties The steady-state kinetic experiments with cyclo(l-Phe-lHis) and cyclo(l-Phe-l-Leu) were performed under standard conditions. One may note that under these conditions and considering the short time of data acquisition, a,b-dehydrogenation only occurred at the Phe residue,
Relative Activity (%)
1718 M. Gondry et al. (Eur. J. Biochem. 268)
100
A
q FEBS 2001
C
B
80 60 40 20 0 5
7
9
pH
11
20
40
60
Temperature (˚C)
0
2
4
6
[NaCl] (M)
allowing us to characterize the first step of the reaction; a conventional hyperbolic curve was observed in both cases (data not shown). From the nonlinear regression analysis of the reaction with cyclo(l-Phe-l-His), we calculated an apparent Km value of 67 ^ 6 mm and a kcat value of 0.453 ^ 0.007 s21. A similar experiment performed with cyclo(l-Phe-l-Leu) as a substrate led to a Km value of 53 ^ 3 mm and a kcat value of 0.69 ^ 0.01 s21. The catalytic efficiency, as inferred from the kcat /Km ratio, was almost twice as high as the natural substrate as compared to that of the model substrate. The Hill coefficient as determined from the Hill plot was equal to 1.1 ^ 0.1 and 0.9 ^ 0.1 for the substrates cyclo(l-Phe-l-His) and cyclo(l-Phe-l-His), respectively (not shown). It was therefore concluded that the purified enzyme behave as a Michaelian enzyme, with no apparent cooperative effects between the monomers. The dehydrogenation of cyclo(l-Phe-l-His) was not observed under anaerobic conditions (see Experimental procedures), showing that molecular oxygen is required for the enzymatic reaction by acting as an electron acceptor. Formation of H2O2 was assayed by measuring the oxidation of 4-hydroxyphenylacetic acid to a fluorescent product in a horseradish peroxidase-coupled assay [20]. Steady-state kinetic measurements showed that the rate of formation of the fluorescent product, detected at 405 nm, was proportional to the concentration of the purified enzyme. The molar activity was estimated to be 75% as compared to that obtained with cyclo(l-Phe-l-His) under standard conditions. This difference, which was due to instability of the fluorescent product, was weak enough to allow us to assume that the production of H2O2 was stoichiometric during the enzymatic reaction. These results indicate that the enzyme can catalyze the formation of an a,bunsaturation at the expense of molecular O2 by producing one molecule of H2O2. Optimal conditions for the enzymatic reaction Enzyme activity was assayed with pH values ranging from 5.5 to 10.5. All other parameters were maintained constant as defined in the standard assay; the ionic strength was held constant at 100 mm using two buffer systems (formate/ Mes/ Tris buffer: 50 mm formate, 50 mm Mes, 100 mm Tris/HCl; Mes/ Tris/ethanolamine buffer: 100 mm Mes,
Fig. 6. Effects of pH, temperature, and ionic strength on cyclic dipeptide oxidase activity. Residual activity (%) was measured under standard assay conditions with the following exceptions. (A) Assays were performed at various pH values using buffer allowing a constant ionic strength at 100 mm: formate/Mes/ Tris buffer, containing 50 mm Formate, 50 mm Mes, 100 mm Tris/HCl (X) and Mes/ Tris/ethanolamine buffer, containing 100 mm Mes, 50 mm Tris/HCl, 50 mm ethanolamine (W). (B) Standard assays were performed at various temperatures. (C) Standard assays were performed in formate/Mes/ Tris buffer, pH 7, at various NaCl concentrations.
50 mm Tris/HCl, 50 mm ethanolamine) [25]. Under these conditions, the molar activity reached a maximum value around pH 8, irrespective of the nature of the buffer used during the experiment (Fig. 6A). When assayed between 16 and 60 8C, the molar activity increased quite linearly with temperature and was nearly twice as high at 60 8C as compared to 30 8C (Fig. 6B). Above 60 8C, the enzyme activity was dramatically decreased, probably due to thermal inactivation. Finally, the enzyme activity was also tested under standard conditions according to ionic strength values. The rate of the reaction remained maximal in Tris/HCl buffer ranging from 10 mm to 100 mm. Then, molar activity diminished with increasing ionic strength to reach 50% at 1 m NaCl and 3.5% at 6 m NaCl (Fig. 6C). Enzyme stability Effect of enzyme concentration. The purified enzyme was diluted in 100 mm Tris/HCl buffer at pH 8 and incubated at 30 8C for 20 min and 2 h. Under these conditions, the enzyme activity was not affected by a change in enzyme concentration over a range varying from 1 to 410 mg´mL21. Effect of pH. The purified enzyme was diluted in appropriate buffers with a pH range between 6.0 and 8.7 and incubated at 30 8C for 6.5 h. The enzyme activity was quite stable at neutral and basic pH values, whereas it decreased at acidic pH. The residual activity was 75% after 6.5 h at pH 6.0. Thermal stability. The purified enzyme was heated for a given time between 60 and 75 8C, in 100 mm Tris/HCl buffer and at pH 8. We observed an inactivation process that follows first-order kinetics (Fig. 7), suggesting that thermal unfolding may be primarily responsible for inactivation. The activation energy associated with the observed thermal inactivation was calculated from the equation: Ea S´R
1
where Ea is the activation energy, S is the slope of the Arrhenius plot (Fig. 7), and R is the universal gas constant. The resulting Ea value was equal to 290 kJ´mol21, indicating that the enzyme is rather stable, in agreement with our data showing an increase in enzyme activity with increasing temperature (Fig. 6B). According to Voordouw
q FEBS 2001
Cyclic dipeptide oxidase from S. noursei (Eur. J. Biochem. 268) 1719
Residual activity (%)
100
60 ˚C -3 -4
ln k
73 ˚C
10
-5
75 ˚C
67 ˚C
-6
2.90 2.95 3.00 3
70 ˚C
-1
1 0 /T (K )
1
0
200
400
600
Heating time (min) Fig. 7. Thermal inactivation of cyclic dipeptide oxidase. The enzyme (25 mg´mL21) was incubated in 100 mm Tris/HCl buffer, pH 8, at the appropriate temperature. Aliquots were assayed under standard conditions in order to determine their remaining activity. Residual activity is expressed relative to that of an unheated sample. The Arrhenius plot is shown in the inset.
et al. [26], an enzyme can be classified as thermostable when its free energy of inactivation (DG³) is at least equal to 104.6 kJ´mol21, at 70 8C, under conditions that maximize the thermal stability. The enzyme purified from S. noursei is characterized by a DG³ value of 106.5 kJ´mol21, which makes it a thermostable enzyme.
DISCUSSION In a preliminary report, we mentioned that an enzymatic activity isolated from S. noursei is likely to be responsible for the a,b-dehydrogenation of phenylalanyl and leucyl residues during the biosynthesis of the antibiotic albonoursin, cyclo(DPhe-DLeu) [27]. Recently, a similar enzymatic activity was detected in Streptomyces sp. KO-23 later identified as Streptomyces albulus [28,29]. We now show that cyclo(l-Phe-l-Leu), probably arising from nonribosomal peptide synthesis [30±33], is an intermediate in the biosynthetic pathway of albonoursin. By a two-step sequential dehydrogenation, this intermediate is first converted to cyclo(DPhe-l-Leu) and finally to albonoursin (Scheme 1). The dehydrogenase activity was purified from S. noursei. It is a new flavoenzyme, which transforms various cyclo(dipeptides) into cyclo(a,b-dehydro-dipeptides). The enzyme can be reoxidized at the expense of molecular oxygen by producing one molecule of H2O2. Therefore, we propose that this novel enzyme purified from S. noursei and catalyzing the a,b-dehydrogenation of cyclo(dipeptides) with oxygen as a final acceptor be named cyclic dipeptide oxidase or cyclo(dipeptide):oxygen oxidoreductase (EC 1.3.3.x). Cyclic dipeptide oxidase is highly stable, being unaffected by a change in its molar concentration, or by neutral and basic pH values. It can be recognized as a thermostable enzyme as defined by Voordouw et al. [26]. The enzyme's activity is maximal from pH 7 to 9 and is not altered by the presence of reducing agents. As judged from its characteristic UV spectrum, cyclic dipeptide oxidase is a novel member of the flavoprotein
family. The flavin prosthetic group, covalently linked to the enzyme, is involved in the enzyme activity as indicated by the observations that flavin reduction depends on substrate concentration, and flavin enrichment parallels the enzyme purification. The enzyme also possesses an original quaternary structure. While it consists of a single type of subunit, with a relative molecular mass of 21 066, it was obtained in vitro in a form that is large enough to be excluded on a Superose 6 column (Mr 2 000 000). This large molecular mass seems to be an intrinsic property of the native enzyme as it was observed irrespective of the extraction process, the presence of additives, and enzyme concentration. Therefore, the cyclic dipeptide oxidase is a large homopolymer comprising over 100 monomers. Such quaternary structure organization is not very common, although a few examples have been reported in the literature. Thus, phosphoribosylpyrophosphate synthetase [34±36], acetyl-CoA carboxylase [37±39] and phosphofructokinase [40], all consist of one type of subunit that associates in vitro in large homopolymeric forms. Moreover, there is some evidence for the existence in vivo of homopolymeric states of acetyl-CoA carboxylase [37] and phosphoribosylpyrophosphate synthetase [36]. In fact, their states of subunit association allow a modulation of enzymatic activities, the large homopolymeric forms being the active ones. The actual state of polymerization of the cyclic dipeptide oxidase in intact bacterial cells remains to be established. However, when subjected to native PAGE, the enzyme undergoes dissociation by successive loss of one or several subunits, each of the resulting polymeric forms clearly retaining enzymatic activity. Further experiments are now in progress to explore the mode of association of the monomers, as well as to determine the nature of the flavin cofactor and its mode of covalent linkage to the enzyme. Cloning of the gene and overexpression of the cyclic dipeptide oxidase being under way (S. Lautru, M. Gondry, R. Genet & J. L. Pemodet, unpublished work), sufficient amount of the protein will be soon available to refine our understanding of the enzyme system. Cyclic dipeptide oxidase from S. noursei appears today clearly distinct from other enzyme systems which have been reported to be involved in the formation of a,b-dehydro-amino acids. l-Tryptophan 2 0 ,3 0 -oxidase from C. violaceum, a hemoprotein isolated in our laboratory, was shown to catalyze the formation of a,b-dehydrotryptophan as a precursor in the biosynthesis of the blue pigment violacein [14]. This enzyme is a hetero-oligomer consisting of two subunits, a (Mr 14 000) and b (Mr 74 000), organized in an octameric structure of type (ab)8. It was postulated to proceed via the direct a,b-dehydrogenation of the amino acid side chain (Scheme 2, route a) [41]. This enzyme system can be compared to three other ones mentioned in the literature. First, cyclopeptine dehydrogenase from Penicillium cyclopium which catalyzes the formation of a dehydro-phenylalanyl residue at an intermediate stage of the biosynthesis of various alkaloids [4]. No information is available regarding its structural organization except that a FAD cofactor is associated with the protein [4]. However, it was shown that this NAD-dependent enzyme proceeds via a direct a,b-dehydrogenation mechanism (Scheme 2, route a) [42]. Second, Epi D from Staphylococcus epidermidis which is involved in the biosynthesis of the lantibiotic epidermin. It
1720 M. Gondry et al. (Eur. J. Biochem. 268)
q FEBS 2001
R O N H H
(b) HO O
O
(c)
R
R O
in microbial metabolites [44], is here highlighted for the first time. Therefore, introduction of an a,b-unsaturation in a peptide can be achieved by a variety of structurally unrelated enzymes. In addition, there is a diversity of reaction mechanisms available to reach that unsaturation. As a result, we wish to suggest that the converging presence of an a,b-unsaturation in peptidyl-like secondary metabolites may be of great physiological importance for the microorganisms, being related to the existence of a large family of enzymes we propose to name `amino acyl a,b-dehydrogenase' with regard to the catalyzed reaction.
(a) N H H
ACKNOWLEDGMENT
N O
(b')
O R O N H
(c')
We acknowledge G. assistance, and J.-L. R. StoÈcklin (Atheris (CEA /DCC, France) discussions.
Potier and C. Laguri for excellent technical Tarride for peptide tritiation. We also thank Laboratories, Switzerland) and H. Virelizier for mass spectrometry analyses and helpful
REFERENCES O
Scheme 2. Possible pathways for enzymatic synthesis of a,b-dehydro-amino acyl residues. Route a, direct a,b-dehydrogenation; route b-b 0 , b-hydroxylation followed by dehydration; route c-c 0 , formation of an imine followed by a rearrangement to enamine.
consists of the association of a single type of subunit with a relative molecular mass of about 18 000 resulting in a homododecamer [43]. This FMN-containing protein catalyzes the removal of two reducing equivalents from the C-terminal cysteinyl side chain [13]. Third, pristinamycin synthase from Streptomyces pristinaespiralis which catalyses the conversion of pristinamycin IIB into pristinamycin IIA, differing from each other in the presence of an a,b-dehydro-prolyl residue. It is an (a,b) heterodimer, its subunits having relative molecular masses of about 35 000 and 50,000, respectively. This monooxygenase uses the oxidized-FMN provided by a NADH:FMN reductase to catalyze the transient b-hydroxylation of a d-prolyl residue, immediately followed by a dehydration reaction leading to the unsaturated residue (Scheme 2, route b-b 0 ) [5]. Although they both use flavin cofactors, these three enzyme systems exhibit quite different mechanisms, substrate specificity and structural features, the first one being involved in a post-translational process whereas the two others are associated with peptide synthetases. The new isolated enzyme, cyclic dipeptide oxidase, is clearly distinct from those described above by both its structural properties and dehydrogenation mechanism. Based on the formation of the a,b-dehydro-products under both Z and E configurations, we postulated a twostep mechanism involving the transient formation of an intermediate imine followed by its rearrangement into an a,b-dehydro-residue (Scheme 2, route c-c 0 ). The formation of the two stereoisomers might arise from a free rotation around the Cû carbon before the formation of the double bond between Ca and Cû carbons. Such a mechanism, originally hypothesized by Bycroft to explain the common occurrence of a,b-dehydro-and d-amino acids
1. von Dohren, H., Dieckmann, R. & Pavela-Vrancic, M. (1999) The nonribosomal code. Chem. Biol. 6, R273±R279. 2. Marahiel, M.A., Stachelhaus, T. & Mootz, H.D. (1997) Modular peptide synthetases involved in nonribosomal peptide synthesis. Chem. Rev. 97, 2651±2673. 3. von DoÈhren, H., Keller, U., Vater, J. & Zocher, R. (1997) Multifunctional peptide synthetases. Chem. Rev. 97, 2675±2705. 4. Aboutabl, E.S.A. & Luckner, M. (1975) Cyclopeptine dehydrogenase in Penicillium cyclopium. Phytochemistry 14, 2573±2577. 5. Thibaut, D., Ratet, N., Bisch, D., Faucher, D., Debussche, L. & Blanche, F. (1995) Purification of the two-enzyme system catalyzing the oxidation of the d-proline residue of pristinamycin IIB during the last step of pristinamycin IIA biosynthesis. J. Bacteriol. 177, 5199±5205. 6. Sahl, H.G., Jack, R.W. & Bierbaum, G. (1995) Biosynthesis and biological activities of lantibiotics with unique post-translational modifications. Eur. J. Biochem. 230, 827±853. 7. Heck, S.D., Faraci, W.S., Kelbaugh, P.R., Saccomano, N.A., Thadeio, P.F. & Volkmann, R.A. (1996) Posttranslational amino acid epimerization: enzyme-catalyzed isomerization of amino acid residues in peptide chains. Proc. Natl Acad. Sci. USA 93, 4036±4039. 8. Gross, E. (1976) a,b-Unsaturated amino acids. In Handbook of Biochemistry and Molecular Biology (Fasman, G.D., ed.), pp. 111±176. CRC Press Inc., Boca Raton, FL, USA. È hler, E. & Poisel, H. (1979) 9. Schmidt, U., HaÈusler, J., O Dehydroamino acids, a-hydroxy-a-amino acids and a-mercaptoa-amino acids. In Progress in the Chemistry of Organic Natural Products (Herz, W., ed.), pp. 251±327. Springer Verlag, New York. 10. Stammer, C.H. (1982) Dehydroamino acids and peptides. In Chemistry and Biochemistry of Amino Acids, Peptides and Proteins (Weinstein, B., ed.), pp. 33±74. Marcel Dekker Inc., New York. 11. Noda, K., Shimohigashi, Y. & Izumiya, N. (1983) a,b-Dehydroamino acids and peptides. In The Peptides: Analysis, Synthesis, Biology (Gross, E. & Meienhofer, J., eds), pp. 285±339. Academic Press, New York. 12. Shimohigashi, Y. & Stammer, C.H. (1983) Dehydro-enkephalins. Part 7. A potent dehydroleucine-enkephalin resistant to carboxypeptidase. J. Chem. Soc. Perkin Trans. 1, 803±808. 13. Kupke, T., Stevanovic, S., Sahl, H.G. & Gotz, F. (1992) Purification and characterization of EpiD, a flavoprotein involved
q FEBS 2001
14.
15. 16.
17.
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28.
Cyclic dipeptide oxidase from S. noursei (Eur. J. Biochem. 268) 1721
in the biosynthesis of the lantibiotic epidermin. J. Bacteriol. 174, 5354±5361. Genet, R., Denoyelle, C. & Menez, A. (1994) Purification and partial characterization of an amino acid alpha,beta-dehydrogenase, l-tryptophan 2 0 ,3 0 -oxidase from Chromobacterium violaceum. J. Biol. Chem. 269, 18177±18184. Fukushima, K., Yazawa, K. & Arai, T. (1973) Biological activities of albonoursin. J. Antibiot. (Tokyo) 26, 175±176. Evans, E.A., Sheppard, H.C., Turner, J.C. & David, C.W. (1974) A new approach to specific labelling of organic compounds with tritium: catalysed exchange in solution with tritium gas. J. Label. Compounds 10, 569±587. 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. Shimohigashi, Y., Dunning, J.W., Grim, M.D. & Stammer, C.H. (1981) Ultraviolet spectra of dehydropeptides by double beam measurement. J. Chem. Soc. Perkin Trans. 2, 1171±1175. Cross, A.R., Yarchover, J.L. & Curnutte, J.T. (1994) The superoxide-generating system of human neutrophils possesses a novel diaphorase activity. Evidence for distinct regulation of electron flow within NADPH oxidase by p67-phox and p47-phox. J. Biol. Chem. 269, 21448±21454. Poosch, M.S. & Yamazaki, R.K. (1986) Determination of peroxisomal fatty acyl-CoA oxidase activity using a lauroylCoA-based fluorometric assay. Biochim. Biophys. Acta 884, 585±593. Blum, H., Beier, H. & Gross, H.J. (1987) Improved silver staining of plant proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 8, 93±99. Saravis, C.A. & Zamcheck, N. (1979) Isoelectric focusing in agarose. J. Immunol. Methods 29, 91±96. Vuillard, L., Braun-Breton, C. & Rabilloud, T. (1995) Nondetergent sulphobetaines: a new class of mild solubilization agents for protein purification. Biochem. J. 305, 337±343. Shin, C.-G., Hayakawa, M., Mikami, K. & Yoshimura, J. (1977) Syntheses and configurational assignments of albonoursin and its three geometric isomers. Tetrahedron Lett. 10, 863±836. Ellis, K.J. & Morrison, J.F. (1982) Buffers of constant ionic strength for studying pH-dependent processes. Methods Enzymol. 87, 405±426. Voordouw, G., Milo, C. & Roche, R.S. (1976) Role of bound calcium ions in thermostable, proteolytic enzymes. Separation of intrinsic and calcium ion contributions to the kinetic thermal stability. Biochemistry 15, 3716±3724. Genet, R., Hammadi, A., Fusai, G., Gondry, M., Meunier, G. & MeÂnez, A. (1998) Enzymatic a,b-dehydrogenation of amino acid side chains as a precursor step for protein labelling. In Peptides 1996, Proceedings of the Twenty-fourth European Peptide Symposium, 8±13 September, Edinburgh, UK (Ramage, R. & Epton, R., eds), pp. 413±414. Mayflower Scientific Ltd., Kingswinford, UK Kanzaki, H., Imura, D., Sashida, R., Kobayashi, A. & Kawazu, K. (1999) Effective production of dehydro cyclic dipeptide
29. 30. 31.
32. 33. 34. 35.
36. 37. 38. 39.
40.
41.
42.
43.
44.
albonoursin exhibiting pronuclear fusion inhibitory activity. I. Taxonomy and fermentation. J. Antibiot. (Tokyo) 52, 1017±1022. Kanzaki, H., Imura, D., Nitoda, T. & Kawazu, K. (1999) Enzymatic dehydrogenation of cyclo (L-Phe-L-Leu) to a bioactive derivative, albonoursin, J. Mol. Catal. B: Enzymatic 6, 265±270. Gerlach, M., Schwelle, N., Lerbs, W. & Luckner, M. (1985) Enzymatic synthesis of cyclopeptine intermediates in Penicillium cyclopium. Phytochemistry 24, 1935±1939. Stachelhaus, T., Mootz, H.D., Bergendahl, V. & Marahiel, M.A. (1998) Peptide bond formation in nonribosomal peptide biosynthesis. Catalytic role of the condensation domain. J. Biol. Chem. 273, 22773±22781. Belshaw, P.J., Walsh, C.T. & Stachelhaus, T. (1999) AminoacylCoAs as probes of condensation domain selectivity in nonribosomal peptide synthesis. Science 284, 486±489. Mootz, H.D., Schwarzer, D. & Marahiel, M.A. (2000) Construction of hybrid peptide synthetases by module and domain fusions. Proc. Natl Acad. Sci. USA 97, 5848±5853. Fox, I.H. & Kelley, W.N. (1971) Human phosphoribosylpyrophosphate synthetase. Distribution, purification, and properties. J. Biol. Chem. 246, 5739±5748. Becker, M.A., Meyer, L.J., Huisman, W.H., Lazar, C. & Adams, W.B. (1977) Human erythrocyte phosphoribosylpyrophosphate synthetase. Subunit analysis and states of subunit association. J. Biol. Chem. 252, 3911±3918. Meyer, L.J. & Becker, M.A. (1977) Human erythrocyte phosphoribosylpyrophosphate synthetase. Dependence of activity on state of subunit association. J. Biol. Chem. 252, 3919±3925. Meredith, M.J. & Lane, M.D. (1978) Acetyl-CoA carboxylase. Evidence for polymeric filament to protomer transition in the intact avian liver cell. J. Biol. Chem. 253, 3381±3383. Beaty, N.B. & Lane, M.D. (1983) The polymerization of acetylCoA carboxylase. J. Biol. Chem. 258, 13051±13055. Thampy, K.G. & Wakil, S.J. (1988) Regulation of acetylcoenzyme A carboxylase. I. Purification and properties of two forms of acetyl-coenzyme A carboxylase from rat liver. J. Biol. Chem. 263, 6447±6453. Johnson, C.S. & Deal, W.C. Jr (1982) High concentration active enzyme centrifugation studies with pig kidney phosphofructokinase. Detection of 9.8 S, 25 S, and 53 S active polymeric forms. J. Biol. Chem. 257, 913±916. Genet, R., Benetti, P.H., Hammadi, A. & Menez, A. (1995) l-Tryptophan 2 0 ,3 0 -oxidase from Chromobacterium violaceum. Substrate specificity and mechanistic implications. J. Biol. Chem. 270, 23540±23545. Aboutabl, E.S.A., El Azzouny, A., Winter, K. & Luckner, M. (1976) Stereochemical aspects of the conversion of cyclopeptine into dehydrocyclopeptine by cyclopeptine dehydrogenase from Penicillium cyclopium. Phytochemistry 15, 1925±1928. Kupke, T., Uebele, M., Schmid, D., Jung, G., Blaesse, M. & Steinbacher, S. (2000) Molecular characterization of lantibioticsynthesizing enzyme EpiD reveals a function for bacterial dfp proteins in coenzyme A biosynthesis. J. Biol. Chem. 275, 31838±31846. Bycroft, B.W. (1969) Structural relationships in microbial peptides. Nature 224, 595±597.
