Cloning, overexpression, and purification of novobiocic acid synthetase from Streptomyces spheroides NCIMB 11891.

Novobiocic acid synthetase, a key enzyme in the biosynthesis of the antibiotic novobiocin, was cloned from the novobiocin producer Streptomyces spheroides NCIMB 11891. The enzyme is encoded by the gene novL, which codes for a protein of 527 amino acids with a calculated mass of 56,885 Da. The protein was overexpressed as a His(6) fusion protein in Escherichia coli and purified to apparent homogeneity by affinity chromatography and gel chromatography. The purified enzyme catalyzed the formation of an amide bond between 3-dimethylallyl-4-hydroxybenzoic acid (ring A of novobiocin) and 3-amino-4,7-dihydroxy-8-methyl coumarin (ring B of novobiocin) in an ATP-dependent reaction. NovL shows homology to the superfamily of adenylate-forming enzymes, and indeed the formation of an acyl adenylate from ring A and ATP was demonstrated by an ATP-PP(i) exchange assay. The purified enzyme exhibited both activation and transferase activity, i.e. it catalyzed both the activation of ring A as acyl adenylate and the subsequent transfer of the acyl group to the amino group of ring B. It is active as a monomer as determined by gel filtration chromatography. The reaction was specific for ATP as nucleotide triphosphate and dependent on the presence of Mg(2+) or Mn(2+). Apparent K(m) values for ring A and ring B were determined as 19 and 131 micrometer respectively. Of several analogues of ring A, only 3-geranyl-4-hydroxybenzoate and to a lesser extent 3-methyl-4-aminobenzoate were accepted as substrates.

are derived from tyrosine, ring C is derived from glucose, and the prenyl group of ring A is formed via the nonmevalonate pathway (1)(2)(3). The antimicrobial activity of novobiocin results from its interaction with bacterial DNA gyrase, which has been investigated by x-ray crystallographic studies (4 -7). The detailed knowledge available about the structural elements of novobiocin involved in its binding to the biological target may permit rational approaches in the search for new aminocoumarin derivatives. Recently, the development of new, synthetic aminocoumarin compounds with gyrase-inhibiting activity has been reported (8 -10). Like novobiocin itself (Albamycin®, Pharmacia & Upjohn), such new aminocoumarins may serve as antibiotics for the treatment of infections with multi-resistant Gram-positive bacteria such as Staphylococcus aureus or Staphylococcus epidermidis (11)(12)(13).
Genetic engineering and combinatorial biosynthesis in bacteria provide an important new tool for drug discovery. Besides polyketide synthetases, peptide synthetases especially have been successfully used for such approaches (14 -17). Knowledge of the sequence and function of the genes involved in the biosynthesis of natural products is a prerequisite for such research. We have recently cloned and sequenced the biosynthetic gene cluster for novobiocin from S. spheroides NCIMB 11891 and have assigned functions to the biosynthetic genes by comparison with GenBank TM entries and by gene inactivation experiments (18). A key step in the biosynthesis of novobiocin is the formation of the amide bond between ring A and ring B ( Fig. 1) in an ATP-dependent reaction; this enzymatic reaction, termed novobiocic acid synthetase, has been demonstrated previously in crude extracts from a novobiocin-producing strain (19). A detailed investigation of this reaction is of particular interest for the development of new aminocoumarin antibiotics; whereas ring B and ring C are essential for the binding of novobiocin to gyrase, the structure of ring A can be varied without loss of antibiotic activity (10,20). It has been suggested that the structure of ring A influences the uptake of the antibiotic through the bacterial membrane (20,21). Cloning of the gene(s) for novobiocic acid synthetase and investigation of the substrate specificity of this reaction may therefore assist in the development of novobiocin derivatives with a modified ring A. We now report the cloning, overexpression, purification, and characterization of novobiocic acid synthetase from S. spheroides NCIMB 11891, encoded by the gene novL. [ 32 P]Tetrasodium pyrophosphate (126.4 GBq/mmol) was obtained from NEN Life Science Products. Ring B and novobiocic acid were kindly provided by Pharmacia & Upjohn, Inc. (Kalamazoo, MI). 3-Cyclohexyl-4-hydroxybenzoic acid was a gift from L. Wessjohann (Amsterdam, Netherlands). Plicatin B was kindly provided by R. Bates (Bangkok, Thailand). Ring A was obtained by hydrolysis of novobiocin as described previously (19). 3-Dimethylallyl-4-hydroxycinnamic acid * This work was supported by a grant from the Deutsche Forschungsgemeinschaft. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Chemicals and Radiochemicals
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF170880.
was synthesized by hydrolysis of plicatin B as described in Bates et al. (22); EI-MS analysis on a TSQ70 spectrometer (Finnigan, Bremen, Germany) using methanol as solvent confirmed the identity of the product (observed molecular weight, 232.2; theoretical molecular weight of C 14 H 16 O 3 , 232.3). 3-Geranyl-4-hydroxybenzoic acid (GBA) was synthesized enzymatically from 4-hydroxybenzoic acid and geranyl diphosphate with 4-hydroxybenzoate polyprenyltransferase of Escherichia coli (23); the incubation mixture (3 ml) contained 0.4 mM 4-hydroxybenzoic acid, 2 mM geranyl diphosphate, 50 mM MgCl 2 , 1 mM KF, 50 mM Tris-HCl buffer, pH 8.0, and 0.2 mg/ml membrane protein fraction and was incubated for 60 min at 37°C. The reaction was stopped by addition of 90 l of concentrated formic acid. GBA was extracted with 30 ml of n-hexan, the organic phase was evaporated, and the residue was dissolved in 50 mM Tris-HCl buffer, pH 8.0.

Bacterial Strains, Cloning Vectors, and DNA Manipulations
The bacterial strains and plasmids used in this study are listed in Table I. Plasmid pUWL201 was kindly provided by A. Bechthold (Tü bingen, Germany) and originally obtained from U. Wehmeier (Wuppertal, Germany). Cloning experiments were performed in E. coli XL1 Blue MRFЈ by standard procedures (24). Heterologous expression experiments with Streptomyces lividans TK24 were carried out as described previously (18). Enzyme activity was determined after cultivation in CDM medium (20 g/ml thiostrepton) for 3-4 days at 28°C and 170 rpm in baffled shake flasks. DNA manipulations and standard genetic techniques in E. coli and Streptomyces species were carried out as described in Sambrook et al. (24) and Hopwood et al. (25).

Construction of pMS65, pMS77, and pMS79 for Expression in S. lividans TK24
The streptomyces expression vector pUWL201, containing the ermE up promoter, was used for the construction of the expression plasmids. For the expression of novL, a 2.1-kb ApaI fragment from plasmid p9 -6GE9 containing novL was cloned into the same site of pGEM-11Zf(ϩ) to give pMS76. Restriction analysis confirmed that the EcoRI site of pGEM-11Zf(ϩ) was located upstream of novL. The insert of pMS76 was excised with EcoRI and XbaI and ligated into the same sites of pUWL201 to give pMS77. pMS65, containing novH as a 1.95-kb EcoRI-BglII fragment in expression vector pUWL201, has been described previously (18). pMS79, containing both novH and novL in pUWL201, was prepared by ligation of the same 1.95-kb EcoRI-BglII fragment carrying novH into the EcoRI-BamHI-sites of pMS76. Both genes in the resulting plasmid pMS78 were oriented in the same direction. The 4.08-kb insert of pMS78 was excised with EcoRI and XbaI and cloned into the same sites of pUWL201 to yield pMS79. Expression constructs were transformed into S. lividans TK24 and examined for novobiocic acid synthetase activity as described previously (18).

Construction of pMS80 for Expression of NovL as Fusion Protein
with a C-terminal His 6 Tag novL was amplified by PCR using pMS76 DNA as template. An SphI site was introduced at the place of the natural start codon, using primer novL-1 (5Ј-TAGCCACGCATGCCGAACAAGGATCAC-3Ј; bold letters represent the SphI site). At the C terminus, a BamHI site was introduced before and an EcoRI site behind the stop codon, using primer novL-3 (5Ј-CATCGAATTCTCAGGATCCCCTGTCCACCA-3Ј; bold letters represent introduced restriction sites). The PCR mixture (100 l) contained 100 ng of pMS76 template, 22 pmol of each primer, 0.2 mM dNTPs (Stratagene), Pfu DNA polymerase reaction buffer, and 5% (v/v) Me 2 SO. 2.5 units of cloned Pfu DNA polymerase (Stratagene) were added after an initial denaturation for 5 min at 96°C, followed by 27 cycles (95°C for 90 s, 72°C for 45 s, and 72°C for 4 min). The PCR product was digested with SphI and BamHI before ligation into the same sites of the expression vector pQE70, resulting in a C-terminal in-frame fusion with the His 6 tag of pQE70. The resulting plasmid was designated as pMS80.

Construction of pMS82 for Expression of NovL as Fusion Protein
with an N-terminal His 6 Tag novL was again amplified by PCR. A BglII site was introduced at the N-terminal side for in-frame ligation to the His 6 tag of pRSet B using primer novL-4 (5Ј-CGAAAGATCTCCACATATGGCGAACAAGG-3Ј). The GTG start codon of novL was changed to an ATG codon by introduction of a NdeI site, which offers the possibility to remove the sequence for the N-terminal His tag by NdeI digestion and religation. Primer novL-3 (see above) was again used as C-terminal primer. PCR was carried out as described above, with an annealing temperature of 65°C. The BglII-EcoRI-digested PCR product was ligated into the same sites of pRSet B to create pMS82. The use of primer novL-3 resulted in the extension of the C-terminal end of the encoded protein from the original -VDR to -VDRGS. Expression experiments with pMS82 were performed in E. coli BL21(DE3)pLysS cultivated in NZYCM broth (24) supplemented with 50 g/ml carbenicillin and 34 g/ml chloramphenicol at 37°C.

Expression of pMS80 and Purification of Novobiocic Acid Synthetase
E. coli XL1 Blue MRFЈ was used as host for the expression of pMS80. Cells were cultured at 30°C in LB medium (24) supplemented with 50 g/ml carbenicillin until an A 600 of 0.7 was reached. 0.5 mM isopropyl-␤-D-thiogalactoside (IPTG) was added, and after further growth for 3 h at 30°C, cells were harvested by centrifugation and washed with 50 mM Tris-HCl, pH 8.0. All subsequent steps were carried out at 4°C. Cells (3 g) were suspended in 3 ml of lysis buffer (50 mM NaH 2 PO 4 , pH 8.0, 300 mM NaCl, 10 mM imidazole, 1 mg/ml lysozyme). After incubation on ice for 30 min, the cell suspension was sonicated for 1 min (Branson Sonifier 250). 10 g/ml RNase A and 5 g/ml DNase I were added, and the mixture was incubated on ice for further 10 min. After removal of cellular debris by centrifugation (17,500 ϫ g for 30 min), 1 ml of Ni-NTA-agarose slurry (50% (w/v) nickel-nitrilotriacetic acid agarose resin suspension in 30% (v/v) ethanol, precharged with Ni 2ϩ ) (Qiagen) were added and mixed gently by shaking for 60 min. The lysate-Ni-NTA-agarose mixture was loaded into a column. Unbound proteins were removed by washing with 8 ml of wash buffer (50 mM NaH 2 PO 4 , pH 8.0, 300 mM NaCl, 20 mM imidazole), and the NovL fusion protein was eluted with 2 ml of elution buffer (50 mM NaH 2 PO 4 , pH 8.0, 300 mM NaCl, 250 mM imidazole). The Ni-NTA-agarose eluate was applied to a HiLoad 26/60 Superdex 200 column (Amersham Pharmacia Biotech) that had been equilibrated with 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5 mM dithiothreitol, and 50 M phenylmethylsulfonyl fluoride. Chroma- tography was carried out with the same buffer at a flow rate of 1.5 ml/min. Fractions were assayed for novobiocic acid synthetase activity.

Protein Analysis
Protein concentrations were determined by the Bradford method (26) using bovine serum albumin as a standard. SDS-PAGE was carried out according to the method of Laemmli (27), and protein bands were stained with Coomassie Brilliant Blue R-250. The molecular weight of native NovL was determined by gel filtration on a HiLoad 26/60 Superdex 200 column using the buffer described above. The column was calibrated with blue dextran 2000, aldolase (molecular weight, 158,000), albumin (molecular weight, 67,000), ovalbumin (molecular weight, 43,000), and ribonuclease A (molecular weight, 13,700) (Amersham Pharmacia Biotech).

Enzyme Assays
Novobiocic Acid Synthetase Assay-The novobiocic acid synthetase assay contained 1 mM ring A, 1 mM ring B, 5 mM ATP, 5 mM MnCl 2 , and 50 mM Tris-HCl, pH 8.0, in a final volume of 100 l. To assay the activity of crude extracts, 20 -100 g of protein and an incubation time of 20 min were used. To assay the activity of purified novobiocic acid synthetase, a maximum of 5 g of enzyme and an incubation time of 7 min were used to ensure linearity of the product formation. The reaction was carried out at 30°C and stopped by addition of 5 l of 1.5 M trichloroacetic acid. The reaction mixture was extracted with 1 ml of ethyl acetate, the organic phase was evaporated, and the residue was dissolved in H 2 O/methanol (50:50, v/v). HPLC analysis was carried out using a Multosphere RP18 -5 column (250 ϫ 4 mm, 5 m; C ϩ S Chromatographie Service, Dü ren, Germany) with a linear gradient from 60 to 100% methanol in 1% aqueous formic acid and detection at 305 nm. Authentic novobiocic acid was used as a standard.
ATP-PP i Exchange Assay-The reaction mixture of the ATP-PP i exchange assay (28) contained (in a final volume of 100 l) 50 mM Tris-HCl, pH 8.0, 5 mM MnCl 2 , 5 mM ATP, 0.1 mM tetrasodium pyrophosphate, 34.4 kBq [ 32 P]tetrasodium pyrophosphate, 4 g of purified novobiocic acid synthetase, and 1 mM of the tested substrates ring A or ring B. After incubation for 20 min at 30°C, the reaction was stopped by adding 1 ml of a mixture containing 1.2% (w/v) activated charcoal, 0.1 M tetrasodium pyrophosphate, and 3% (v/v) perchloric acid. The charcoal was pelleted by centrifugation (14,000 rpm for 5 min), washed twice with 1 ml of water, and finally resuspended in 0.5 ml of water. The charcoal-bound radioactivity was measured using a Tri-Carb 2100TR scintillation analyzer (Canberra-Packard) after addition of 9 ml of liquid scintillation fluid (Rotiszint ® Eco Plus, Roth).

Preparative Isolation and EI-MS Analysis of Enzymatic Products
To confirm the identity of the product of the novobiocic acid synthetase reaction, 2 ml (22.5 mg of protein) of a crude extract from S. lividans TK24 transformed with the expression construct pMS77 were passed through a Sephadex G-25 column and incubated in 50 mM Tris-HCl, pH 8.0, with 1 mM ring A, 1 mM ring B, 5 mM ATP, and 5 mM MnCl 2 in a final volume of 15 ml for 60 min at 30°C. After addition of 750 l of 1.5 M trichloroacetic acid, the incubation mixture was extracted for three times with 20 ml of ethyl acetate. The organic phases were combined, evaporated, and dissolved in H 2 O/methanol (50:50, v/v), and the reaction product was purified by HPLC as described for the novobiocic acid synthetase assay (see above). EI-MS was carried out as described under "Chemicals and Radiochemicals" using dichloromethane as a solvent. A molecular weight of 395.2 was observed (for novobiocic acid, C 22 H 21 NO 6 , the molecular weight was 395.4).
To confirm the identity of the product formed from the ring A analogue GBA, 50 g of purified NovL, 0.5 mM GBA, 1 mM Ring B, 5 mM ATP, and 5 mM MnCl 2 were incubated in a final volume of 1 ml in 50 mM Tris-HCl, pH 8.0 for 20 min at 30°C. The reaction was stopped by addition of 50 l of 1.5 M trichloroacetic acid. After extraction with 3 ϫ 1 ml of ethyl acetate and evaporation of the organic phase, the residue was dissolved in in H 2 O/methanol (50:50 v/v), and the reaction product was purified by HPLC as described for the novobiocic acid synthetase assay (see above). EI-MS analysis was performed as described under "Chemicals and Radiochemicals," and a molecular weight of 463.3 was observed (theoretical result: molecular weight of C 27 H 29 NO 6 was 463.5).

Identification of the Novobiocic Acid Synthetase
Gene-Sequence analysis of the novobiocin biosynthetic gene cluster revealed two genes for which homology searches suggested a possible involvement in the novobiocic acid synthetase reaction (18); the deduced protein sequence of novH showed similarity to nonribosomal peptide synthetases, and novL showed homology to acyl-CoA synthetases and 4-coumarate:CoA ligases. Novobiocic acid synthetase activity could be demonstrated (18) upon heterologous expression of a 6.96 kb DNA fragment comprising novH, novL, and three further complete open reading frames (Fig. 2, pMS73). To identify the genes involved in this reaction, we now prepared additional constructs containing these genes, expressed them in S. lividans TK24, and examined the resulting novobiocic acid synthetase activity. All constructs were derived from pUWL201 and contained the ermE up promoter for foreign gene expression. Whereas expression of novH (Fig. 2, pMS65) yielded no activity, novobiocic acid synthetase was clearly detected upon expression of novL (pMS77). The identity of the enzymatic product was confirmed by HPLC in comparison with authentic substance and by preparative isolation of the product followed by mass spectroscopy (EI-MS; see "Experimental Procedures"). Simultaneous expression of novH and novL (pMS79) did not increase activity in comparison to the expression of novL alone (Fig. 2), and likewise a mixture of the two enzyme extracts obtained after separate expression of genes novH and novL (pMS65 and pMS77), respectively, did not show higher activity than the extract obtained from the expression of novL alone (data not shown). This demonstrates that the novobiocic acid synthetase reaction is catalyzed by NovL alone and that NovH is not required for this activity. Within the novobiocin biosynthetic gene cluster (GenBank TM accession number AF170880), novL spans positions 12457-14040. It comprises 1584 base pairs and encodes a protein of 527 amino acids (calculated mass, 56,885 Da). The coding region has an overall G ϩ C content of 70.1%. Upstream of the GTG initiation codon, a putative ribosomal binding site (AGG-TAG) was identified. Fig. 3 shows that NovL contains several conserved motifs supposed to be involved in common steps of adenylate formation, i.e. nucleotide binding, PP i release, and adenylation of the carboxylate moiety of the substrate. These motifs include Box I (SSGTTGXPKGV) and a sequence similar to the Box II motif (usually GEICIRG) of 4-coumarate:CoA ligases (29), as well as motifs A8 and A10 in the C-terminal domain (30). As confirmed by mutational analysis, both the conserved Lys of the Box I motif and the conserved Arg of motif A8 cooperate in coordinating the pyrophosphate release during adenylate formation (29). The conserved Lys of motif A10 interacts with the carboxyl group of the substrate as well as with the ribose oxygens O-4Ј and O-5Ј (30). Becker-André et al. (31) suggested a participation of the central cysteine within Box II motif of 4-coumarate:CoA ligases in thiolester formation, but this hypothesis was recently disproven by mutational analysis (29). A distinction between coenzyme A ligases and enzymes that merely form acyl adenylates is therefore not possible from sequence data. In contrast to nonribosomal peptide synthetases, NovL did not show a 4Ј-phosphopantetheinyl attachment site. Overexpression of the novL Gene in E. coli-For further characterization, NovL was expressed in E. coli in form of different fusion proteins with a His 6 residue for metal affinity chromatography (see "Experimental Procedures"). Fusion of the His tag to the N terminus of NovL resulted in the expression of an unsoluble protein, and only minimal novobiocic acid synthetase activity (0.74 pkat/mg protein) was detectable in the soluble fraction. Fusion of the His tag to the C terminus of NovL gave approximately 7-fold higher activities (5.1 pkat/mg). SDS-PAGE analysis showed, after induction with IPTG, the formation of a 60.6-kDa protein (calculated mass, 58.1 kDa) (Fig. 4), most of which was still in the insoluble fraction. The amount of soluble protein could be increased, however, by decreasing the growth temperature to 30°C, reducing IPTG concentration to 0.5 mM, induction at a later stage of the growth phase (A 600 ϭ 0.7), and shortening of the induction period to 3 h. Under these conditions, soluble novobiocic acid synthetase was obtained in an activity of 461 pkat/mg protein.
Purification of Novobiocic Acid Synthetase-The fusion protein of NovL with the C-terminal His 6 tag was purified by metal affinity chromatography. Because SDS-PAGE of the eluate (Fig.  4) showed some impurities, gel chromatography on Superdex 200 was used as additional purification step and yielded a protein of apparent homogeneity (Fig. 4). The gel chromatography step also served for the removal of imidazole; imidazole was necessary for elution from the metal affinity column but greatly reduced the stability of the enzyme. After removal of imidazole and in the presence of 50 M phenylmethylsulfonyl fluoride and 5 mM dithiothreitol, the purified enzyme could be stored at 4°C for 2 days with only 22% loss of activity. Overall, the enzyme was purified 115-fold with an overall yield of 55% (Table II).
Molecular Weight Determination and Characterization of NovL-By SDS-PAGE, the molecular mass of the His-tagged protein was determined as 60 kDa (Fig. 4). By gel filtration, the native molecular mass of the enzyme resulted as 57-58 kDa (calculated mass, 58.1 kDa), showing that the protein is active as a monomer. The purified NovL exhibited a specific activity of 50.6 nkat/mg protein, corresponding to a turnover rate of 2.9 mol of substrate/mol enzyme/s. The reaction was strictly dependent on the presence of native enzyme, ring A, ring B, ATP, and divalent cations such as Mg 2ϩ and Mn 2ϩ . EDTA (0.5 mM) reduced the enzyme activity to 1.4% of the maximal activity. Product formation in the novobiocic acid synthetase assay showed a linear dependence on the protein amount up to 5 g of purified protein in the assay and on incubation time up to 10 min. The pH optimum was 8.0 in Tris-HCl buffer and 7.0 in phosphate buffer.
Substrate Specificity-Substrate specificity was tested with several analogues of ring A (Table III). Replacement of the dimethylallyl side chain of ring A by a geranyl side chain still allowed a substantial product formation (26% of the value obtained with the natural substrate). The geranylated reaction product was identified by EI-MS after preparative isolation. Some product formation was also detectable with 3-methyl-4aminobenzoic acid. In contrast, 3-dimethylallyl-4-hydroxycinnamic acid was not accepted as a substrate, and neither were 4-hydroxybenzoic acid derivatives with bulky or polar substituents (Table III). No product formation was observed after incubation with 7-amino-4-methyl coumarin instead of ring B. Specificity for the nucleotide triphosphate was tested with 5 mM ATP, GTP, CTP, and TTP. Novobiocic acid synthetase activity was only detectable in the presence of ATP (100%) and to a much lesser extent with GTP (2.4%). The other nucleotides were completely ineffective. The reaction followed Michaelis-Menten kinetics for the substrates ring A and ring B in the presence of 5 mM ATP. Apparent K m values of 19 M for ring A and 131 M for ring B were determined by the Lineweaver-Burk method.
Mechanism of Activation-Peptide synthetase and acyl-CoA ligase reactions usually proceed via acyl adenylates (32). Because the reaction ATP ϩ carboxylic acid º acyl adenylate ϩ PP i is an equilibrium reaction, the formation of [ 32 P]ATP can be observed when [ 32 P]PP i is added to the reaction mixture. This ATP-PP i exchange assay can be used to monitor the activity of adenylate forming enzymes. As shown in Table IV, the purified enzyme exhibited adenylation activity in this assay when incubated with ring A and ATP. No adenylation activity was observed when ring A was replaced by ring B. This strongly suggests that the reaction proceeds via the formation of a ring A-AMP intermediate followed by the transfer of the acyl group of ring A onto the amino group of ring B. An involvement of CoA seems unlikely, because the addition of CoA (0.5 mM) to the novobiocic acid synthetase assay did not increase novobiocic acid formation. DISCUSSION The antibiotic novobiocin contains two aromatic rings: ring A and ring B (Fig. 1). The linkage of those two rings by an ATP-dependent formation of an amide bond, i.e. the novobiocic acid synthetase reaction, has been demonstrated previously in crude enzyme extracts (19). Cloning of the novobiocin biosyn-  thetic gene cluster (18) has now permitted investigation of this reaction on the molecular genetic level. Sequence analysis of the gene cluster showed two genes, novH and novL, which showed distinct homology to the superfamily of adenylateforming enzymes and which might therefore be involved in the novobiocic acid synthetase reaction. Heterologous expression of a fragment comprising both genes indeed resulted in the formation of acitve novobiocic acid synthetase (18). The superfamily of adenylate-forming enzymes is divided into two groups (32): group I enzymes, including the nonribosomal peptide synthetases, contain a highly conserved 4Ј-phosphopantetheinyl attachment site. After activation of their acyl substrate in form of an acyl adenylate, they bind the acyl substrate covalently as thioester to their phosphopantetheinyl cofactor. In contrast, group II enzymes, including 4-coumarate: CoA-ligases, luciferases, acetyl-CoA synthetases, and long chain fatty acid CoA synthetases do not contain the 4Ј-phosphopantetheinyl attachment site; consequently they activate their acyl substrates as adenylates but do not thereafter form covalent bonds between the substrate and the enzyme. Nonribosomal peptide synthesis has been investigated in detail in the biosynthesis of many antibiotics (30,33). It requires activation of each amino acid (or amino acid derivative) as aminoacyl adenylate, followed by covalent thioester linkage to the 4Ј-phosphopantetheinyl cofactor and transfer of the activated carboxyl group to the amino group of the next acyl intermediate. For each amino acid (or derivative), these functions are assembled on a different peptide synthetase module. In anal-ogy, a possible hypothesis for the biosynthesis of novobiocic acid was an activation and covalent enzyme binding of both ring A and ring B (or of precursors thereof) by separate peptide synthetase modules and a transfer of the acyl group of ring A to the amino group of activated ring B.
Our present study, however, proves a different mechanism. The presence of the 4Ј-phosphopantetheinyl attachment site within the NovH sequence suggested its classification as an enzyme of group I, but the novobiocic acid synthetase NovL did not contain this motif and was classified as group II enzyme. NovL alone catalyzed both the activation of ring A in form of an adenylate as well as its transfer to the amino group of ring B. No indication was found for an involvement of NovH in this reaction, and the overexpression and homogenous purification of enzymatically acitve NovL ruled out the participation of any other enzyme. Therefore, the mechanism of the novobiocic acid synthesis is different from the nonribosomal peptide synthesis described in other antibiotic-producing organisms.
Consistent with the sequence homology of NovL to adenylate-forming enzymes, biochemical evidence for an adenylating activity of NovL was provided by the ATP-PP i exchange assay (Table IV). As expected, an activation of ring A but no activation of ring B was observed. Therefore ring B itself, rather than an activated intermediate thereof, is the acyl group acceptor in this reaction. Coenzyme A was not necessary for novobiocic acid formation.
These findings leave the question for the role of novH, which shows significant homologies to peptide synthetases. A likely hypothesis is that NovH may be required for the activation of tyrosine or an intermediate derived from tyrosine, during the biosynthesis of ring B or ring A (Fig. 1).
The biochemical properties of the novobiocic acid synthetase (molecular mass, K m values, and turnover rate) were within the range of corresponding data reported for other group II adenylate forming enzymes, e.g. acyl CoA synthetases (34,35) or 4-coumarate:CoA-ligases (36,37). Our investigation indicated a rather strict specificity of the enzyme for ATP as nucleotide. In contrast, there was some flexibility for the structure of the acyl substrate (ring A), e.g. GBA was accepted as substrate at a substantial rate (Table III). Therefore, inactivation of genes required for ring A biosynthesis and supplementation of culture broth of the resulting mutant with suitable synthetic analogues of ring A may open an easy and efficient access to novobiocin derivatives. Likewise, identification and mutation of individual amino acids affecting the substrate specifity of NovL could offer possibilities for the production of new aminocoumarin antibiotics. a Product formation with purified NovL was determined as described for the novobiocic acid synthetase assay under "Experimental Procedures," with 1 mM of the respective substrates. Data are the mean values of duplicate measurements. b 3-Geranyl-4-hydroxybenzoic acid was synthesized as described under "Experimental Procedures."