Sitedirected mutagenesis of the active site ... - Wiley Online Library

Eur. J. Biochem. 267, 853±860 (2000) q FEBS 2000

Site-directed mutagenesis of the active site serine290 in flavanone 3b-hydroxylase from Petunia hybrida Richard LukacÏin1, Inga GroÈning1, Uwe Pieper2 and Ulrich Matern1 1

Institut fuÈr Pharmazeutische Biologie, Philipps-UniversitaÈt Marburg, Germany; 2Institut fuÈr Biochemie, Justus-Liebig-UniversitaÈt Giessen, Germany

Flavanone 3b-hydroxylase (FHT) catalyzes a pivotal reaction in the formation of flavonoids, catechins, proanthocyanidins and anthocyanidins. In the presence of oxygen and ferrous ions the enzyme couples the oxidative decarboxylation of 2-oxoglutarate, releasing carbon dioxide and succinate, with the oxidation of flavanones to produce dihydroflavonols. The hydroxylase had been cloned from Petunia hybrida and expressed in Escherichia coli, and a rapid isolation method for the highly active, recombinant enzyme had been developed. Sequence alignments of the Petunia hydroxylase with various hydroxylating 2-oxoglutarate-dependent dioxygenases revealed few conserved amino acids, including a strictly conserved serine residue (Ser290). This serine was mutated to threonine, alanine or valine, which represent amino acids found at the corresponding sequence position in other 2-oxoglutarate-dependent enzymes. The mutant enzymes were expressed in E. coli and purified to homogeneity. The catalytic activities of [Thr290]FHT and [Ala290]FHT were still significant, albeit greatly reduced to 20 and 8%, respectively, in comparison to the wild-type enzyme, whereas the activity of [Val290]FHT was negligible (about 1%). Kinetic analyses of purified wild-type and mutant enzymes revealed the functional significance of Ser290 for 2-oxoglutarate-binding. The spatial configurations of the related Fe(II)dependent isopenicillin N and deacetoxycephalosporin C synthases have been reported recently and provide the lead structures for the conformation of other dioxygenases. Circular dichroism spectroscopy was employed to compare the conformation of pure flavanone 3b-hydroxylase with that of isopenicillin N synthase. A double minimum in the far ultraviolet region at 222 nm and 208±210 nm and a maximum at 191±193 nm which are characteristic for a-helical regions were observed, and the spectra of the two dioxygenases fully matched revealing their close structural relationship. Furthermore, the spectrum remained unchanged after addition of either ferrous ions, 2-oxoglutarate or both of these cofactors, ruling out a significant conformational change of the enzyme on cofactor-binding. Keywords: Petunia hybrida; flavonoid biosynthesis; flavanone 3b-hydroxylase; 2-oxoglutarate-binding site; site-directed mutagenesis.

Flavanone 3b-hydroxylase (FHT) catalyzes the 3b-hydroxylation of 2S-flavanones to 2R,3R-dihydroflavonols, which are intermediates in the biosyntheses of flavonols, catechins, proanthocyanidins and anthocyanidins [1]. FHT activity depends on the presence of molecular oxygen, Fe2+, 2-oxoglutarate and ascorbate, and the enzyme classifies as a dioxygenase belonging to the group of nonheme iron(II) enzymes that function in very diverse pathways, e.g. the formation of plant gibberellins or microbial b-lactam antibiotics [2,3]. FHT was described from Petunia hybrida, the cDNA was cloned [4], and the enzyme was expressed in Escherichia coli. Sequence alignments of FHT polypeptides from several plant sources [5] revealed conserved motifs as well as extended homologies with sequences of other intermolecular Correspondence to U. Matern, Institut fuÈr Pharmazeutische Biologie, Philipps-UniversitaÈt Marburg, Deutschhausstrasse 17 A, D-35037 Marburg, Germany. Fax: + 6421 28 26678, Tel.: + 6421 28 22417, E-mail: [email protected] Abbreviations: ACC, 1-aminocyclopropane-1-carboxylic acid; ACV, d-(l-a-aminoadipoyl)-l-cysteinyl-d-valine; DAOCS, deacetoxycephalosporin C synthase; FHT, flavanone 3b-hydroxylase; IPNS, isopenicillin N synthase. (Received 25 October 1999; accepted 1 December 1999)

dioxygenases and related enzymes such as flavonol synthase [6], hyoscyamine 6b-hydroxylase [7], isopenicillin N synthase (IPNS) [8] or deacetoxycephalosporin C synthase (DAOCS) [9]. Among these nonheme iron(II) enzymes, the molecular architecture of IPNS and DAOCS has been studied most extensively and is being considered as a precedent. Both enzymes use ferryl intermediates in catalysis created by reductive splitting of dioxygen, but recruit the required electrons from different sources. While the activity of the FeII ±DAOCS complex is coupled to the oxidative decarboxylation of the cosubstrate 2-oxoglutarate, FeII ±IPNS acts independently of 2-oxoglutarate and generates its ferryl form by the two-electron oxidation (cyclization) of the substrate d-(l-a-aminoadipoyl)l-cysteinyl-d-valine (ACV) reducing dioxygen to two equivalents of water. The metal ion in resting IPNS from Aspergillus nidulans (complexed with divalent manganese) was shown to be octahedrally coordinated by two molecules of water and the residues His214, Asp216, His270, Gln330 of the enzyme polypeptide [8]. In addition, the complex of Aspergillus IPNS with ACV and ferrous iron was crystallized under anaerobic conditions, and an active site model was proposed from X-ray diffraction studies [10]. Accordingly, the substrate ACV replaces Gln330 in the coordination and becomes anchored to the iron center by ligation of its thiolate. Furthermore, substrate

854 R. LukacÏin et al. (Eur. J. Biochem. 267)

q FEBS 2000

Table 1. Synthetic oligonucleotides for the amplification of mutant FHT genes. The exchange of Ser290 by threonine, alanine or valine was accomplished by amplification of the FHT coding DNA as template with the mutated complementary primers listed in the Table. The base triplets encoding Ser290 in wild-type FHT or inducing the mutations are bold printed and the base exchanges are underlined. Enzymes

Partial sequence

Coding change

Wild-type FHT polypeptide Wild-type FHT coding DNA [Thr290]FHT cDNA template [Ala290]FHT cDNA template [Val290]FHT cDNA template

287 293 Ser Arg Leu Ser Ile Ala Thr 5 0 -GCAGCAGGTTATCGATAGCCACG-3 0 3 0 -CGTCGTCCAATTGCTATCGGTGC-5 0 3 0 -CGTCGTCCAATCGCTATCGGTGC-5 0 3 0 -CGTCGTCCAATCACTATCGGTGC-5 0

TCG!ACG TCG!GCG TCG!GTG

binding displaces one molecule of water without distortion of the enzyme core structure to create a pentacoordinated, high spin iron species before the binding of oxygen [10]. Three of the five FeII coordination sites in this complex are still occupied by His214, Asp216 and His270, and the essential function of these strictly conserved amino acids was confirmed by point mutations also for the Petunia FHT [11]. Based on the assumption that the C5-carboxylate of 2-oxoglutarate must be bound to the enzyme during catalysis via a positively charged amino acid, we were able to assign an arginine residue (Arg288) in the FHT, through in vitro mutagenesis and kinetic measurements, that serves as an electrostatic anchor forming a salt bridge with 2-oxoglutarate [11]. This arginine is part of an R-X-S motif conserved in all FHTs, and it was proposed that an equivalent motif is present also in 2-oxoglutarate-dependent dioxygenases of different function [11]. The structural details of the FeII ±DAOCS±oxoglutarate complex published recently from Streptomyces clavuligerus fully supported the conclusions drawn from FHT mutagenesis [9], as an arginine residue corresponding to the Arg288-XaaSer motif in Petunia FHT was confirmed to bind the C5-carboxylate of 2-oxoglutarate. This motif was proposed to be of critical importance also for the activity of the FeII ±IPNS± ACV complex after binding of the ACV valine-carboxylate group to the arginine residue as well as to the proximal serine in the R-X-S motif had been observed [8,10]. The functionality of this conserved serine residue in IPNS for substrate binding was confirmed by recent mutational studies [12], and it is not surprising that X-ray crystallographic studies of the FeII ± DAOCS±oxoglutarate complex also identified the corresponding serine residue as the second ligand for 2-oxoglutarate binding [9]. The DAOCS and IPNS share only 19% amino acid sequence identity, but the modular composition of their secondary structures and the spatial configurations are obviously very similar [9]. Thus the dioxygenases form a family of enzymes that can compensate extensive amino acid substitutions and rely on only few sequence motifs at the appropriate geometric setting. A very similar three-dimensional structure was assumed therefore for all 2-oxoglutarate-dependent dioxygenases, including the FHTs, despite the diversity in their substrate specificities and reaction mechanisms. In order to prove this assumption, we compared the overall configurations of Petunia FHT and Aspergillus IPNS by circular dichroism spectroscopy, and documented further the close relationship of these two enzymes by site-directed mutagenesis of the serine residue in the R-X-S motif that is strictly conserved in FHT and related enzymes.

M AT E R I A L S A N D M E T H O D S Materials Columns of Fractogel EMD BioSEC (S) and Fractogel EMD DEAE 650 (S) were kindly provided by Merck KgaA (Darmstadt, Germany). Biochemicals of analytical grade and E. coli host strains were purchased from Pharmacia (Freiburg, Germany), Sigma (Deisenhofen, Germany), Boehringer-Mannheim (Mannheim, Germany), Roth (Karlsruhe, Germany), Qiagen (Hilden, Germany) or Bio-Rad (MuÈnchen, Germany). [2-14 C]Malonyl-CoA was from Biotrend (KoÈln, Germany). (2S ) [4a,6,8-14 C]Naringenin (2.18 GBq/mmol) was collected from chalcone synthase incubations employing [2-14C]malonylCoA and 4-coumaroyl-CoA [5]. This substrate was stored in ethyl acetate at 220 8C, and radio-TLC analysis (FLA 2000 Bioimager, Fuji, Japan) revealed a purity exceeding 98%. Site-directed mutagenesis Site-directed mutagenesis was accomplished by unique site elimination using the oligonucleotide-directed in vitro mutagenesis technique [11], and oligonucleotides were synthesized (G. Igloi, Institut fuÈr Biologie III, UniversitaÈt Freiburg) as required for the substitution of serine by threonine, alanine and valine, respectively (Table 1). Each individual mutation was verified by DNA sequencing using the single-stranded dideoxynucleotide chain-termination method [13] with adequate sequencing primers. Expression and isolation of wild-type and mutant FHTs The construction of expression vectors was performed with standard cloning techniques [14]. Wild-type or mutated FHT cDNAs were spliced from FHT-pTZ19R constructs with EcoRI, and the isolated FHT inserts were ligated into the vector pAcGP67-B. The orientation of inserts was checked by restriction analysis of each of the cDNA plasmids with PstI and XbaI. The individual FHT cDNAs were isolated after digestion with NcoI and PstI and subcloned into the expression vector pQE6 [11]. E. coli M15, harboring the plasmid pRep4, was transformed with the wild-type and mutant FHT-pQE6 constructs, and the cells were propagated, induced for FHT expression and harvested as described previously [11,15]. A protocol for the rapid purification of FHTs had been devised [16], which was monitored by the concomitant assay of specific

q FEBS 2000

Flavanone 3b-hydroxylase oxoglutarate binding site (Eur. J. Biochem. 267) 855

activity, SDS/PAGE separation [17] and immunoblot crosshybridization [18,19]. Briefly, the cells (2 g pellet) were sonicated for 2 min in 50 mm potassium phosphate buffer at pH 5.5 (20 mL), containing 10 mm EDTA, 5 mm dithiothreitol and 15 mm MgCl2, and the protein was collected from the clear supernatant (30 000 g, 10 min, 5 8C) by ammonium sulfate fractionation (20±55% saturation). The protein was redissolved in 50 mm potassium phosphate buffer at pH 5.5 (2 mL), containing 5 mm dithiothreitol, and the FHTs were purified by successive size exclusion and anion exchange chromatographies on Fractogel EMD BioSEC (S) (1.6 60 cm) and Fractogel EMD DEAE 650 (S) (1 15 cm) in 50 mm Tris/ HCl buffer pH 7.5, in the presence of 5 mm dithiothreitol; the proteins were eluted from the ion exchange column in a linear 0±0.55 m sodium chloride gradient (5.5 column vols) at a flow rate of 1 mL´min21, and homogeneous FHT was collected at 0.21±0.26 m sodium chloride.

with a cylindrical quartz cuvette with a pathlength of 0.05 cm. The temperature of the cell holder was maintained at 5 8C by a circulating water thermostat. The instrument was calibrated with 0.06% ammonium d-10-camphor sulfonate. FHT spectra were recorded in potassium phosphate buffer, pH 6.8, and the protein concentration was adjusted to 0.3 mg´mL21 in all samples. Ferrous iron and 2-oxoglutarate at final concentrations of 20 mm and 100 mm, respectively, were also added individually or in combination, and the CD spectra of these FHT solutions were recorded in order to investigate possible conformational rearrangements of the secondary FHT structures. Reported spectra show the accumulation of 10 scans with 50 nm´min21. The CD spectra of the FHT samples were analyzed for the secondary structure content by the selfconsistent method [20] included in the program package dichroprot v2.4 by Deleage & Geourjon [21].

Sequence analysis

Enzyme assays

Data bank searching and sequence alignments were performed with the programs Entrez and Blast (National Library of Medicine and National Institute of Health, Bethesda, MD, USA).

The standard FHT activity assays were carried out in 0.1 m Bistris/HCl buffer, pH 6.0 (total volume 100 mL), at 37 8C and using (2S)-[4a,6,8-14 C]naringenin as a substrate [5,11]. FHT reaction kinetics comply with a single-substrate reaction mechanism, and the Michaelis constants for the wild-type and mutant FHTs were determined as described previously [11] using 2 mg of pure enzyme protein. Protein concentrations were determined by the Lowry procedure [22] after precipitation of protein with trichloroacetic acid in the presence of deoxycholate [23] and using bovine serum albumin as a standard. The formation of products was examined by chromatography on

Circular dichroism spectroscopy Circular dichroism measurements of homogeneous wild-type and mutant FHTs were done on a Jasco-720 spectropolarimeter (Tokyo, Japan) interfaced to an 486/33 PC and controlled by Jasco software. The Jasco spectropolarimeter was equipped

Fig. 1. Partial alignments of Petunia flavanone b-hydroxylase (PFHT) and Aspergillus IPNS polypeptides. Identical amino acids are bold printed, conservative exchanges are marked by a cross between the two sequences, and the amino acid residues are numbered at the margins of each sequence. Conformational elements identified in the FeII ±IPNS tertiary structure [8] are indicated and numbered below the sequences; b-sheet regions (bx) are designated by open arrows, and a-helical regions (ax) are marked by open arrows on black circles. On the bottom lines, the amino acids ligated to ferrous iron 1, W 2 and W 3 [10], the residue responsible for carboxylate-binding of the ACV substrate in IPNS [10] or of 2-oxoglutarate in PFHT [11] in IPNS are numbered W 4 , and the serine residue participating in oxoglutarate-binding in PFHT is marked W 5. is labeled by W


q FEBS 2000

Table 2. Binding site for 2-oxoglutarate in flavanone 3b-hydroxylase and related enzymes. The conserved motif Arg-Xaa-Ser of the oxoglutaratebinding site was inferred from multiple polypeptide alignments of 52 2-oxoglutarate-dependent and related enzyme sequences in the EMBL library, which included flavanone 3b-hydroxylases, flavonol synthases, anthocyanidin synthases, gibberellin C20 oxidase, gibberellin C2 oxidase, hyoscyamine 6bhydroxylase, deacetoxyvindoline 4-hydroxylase, deacetylcephalosporin C-synthases, isopenicillin N synthase and 1-aminocyclopropane-1-carboxylate oxidases. During catalysis, the basic amino acid residue of this tripeptide (mostly arginine, but lysine in human prolyl 4-hydroxylase) forms a salt bridge with the C5-carboxylate of 2-oxoglutarate. The commonly conserved serine residue is replaced by threonine, alanine, valine or isoleucine in only a few instances as listed. Source

Enzyme

Binding motif

P. hybrida A. thaliana S. clavuligerus Human Chicken

Flavanone 3b-hydroxylase (FHTPH) Gibberellin C20 oxidase (GI20A) Clavaminate synthase (CLASY) Prolyl 4-hydroxylase, a subunit (aPRHH) Lysyl hydroxylase (LYSHY)

Arg-Leu-Ser Arg-Lys-Thr Arg-Lys-Ala Lys-Trp-Val Arg-Tyr-Ile

cellulose thin-layer plates in 15% acetic acid as a solvent, and reference compounds were spotted as described previously [24].

R E S U LT S Choice of mutation A number of highly conserved amino acids and structural motifs had been recognized previously in FHTs from various plant sources through sequence alignment studies with other 2-oxoglutarate-dependent and related nonheme iron(II) enzymes [5]. Among these are two histidines and one aspartate residue which appear to be essential for ferrous iron binding and catalytic activity [11]; these amino acids were identified also as metal-binding ligands in the crystallized FeII ±IPNS complex [8]. Furthermore, in Petunia, FHT Arg288 had been shown to play a crucial role in oxoglutarate-binding, as an exchange for glutamine drastically reduced the affinity of the enzyme to 2-oxoglutarate [11], and this arginine was recognized to be conserved also in one of the b-strands (b14) of IPNS, forming the active site `jelly role' cavity (Fig. 1). Later, it was confirmed that this residue is required for oxoglutarate-binding in the FeII ±DAOCS±oxoglutarate complex [9]. The X-ray diffraction of DAOCS revealed, furthermore, that 2-oxoglutarate is additionally ligated to the serine located next but one to this arginine residue and corresponding to Ser290 in the Arg-XaaSer motif of the Petunia FHT polypeptide (Fig. 1). This motif is conserved in most 2-oxoglutarate-dependent and related enzymes, but a few exceptions are known, e.g. clavaminate synthase from Streptomyces clavuligerus, human prolyl 4-hydroxylase, chicken lysyl hydroxylase or gibberellin C20 oxidase of Arabidopsis thaliana, in which the structurally equivalent threonine or the more hydrophobic alanine, valine and isoleucine, respectively, replace serine in the tripeptide motif (Table 2). Therefore, the relevance of Ser290 for oxoglutaratebinding and catalytic activity of FHT had to be verified, and FHT mutant genes encoding either Ala290, Val290 or Thr290 were generated by site-directed mutagenesis and cloned in the vector pQE6 for expression in E. coli. Expression of wild-type and mutant enzymes Wild-type and mutant FHTs were expressed in E. coli as soluble enzymes, and the level of expression of the mutant enzymes did not differ significantly from that of the wild-type FHT, as estimated by protein staining after SDS/PAGE separation and Western cross-hybridization. The individual enzyme

polypeptides were isolated by extraction of the cells in buffer of pH 5.5 followed by size exclusion and ion exchange chromatographies at pH 7.5. This purification procedure had been elaborated only recently and shown to be rapid enough for the recovery of highly active, homogeneous FHT (32 mkat´kg21 for the wild-type enzyme as compared to 1.9 mkat´kg21 on average from previous purifications [11]) at about 5% yield

Fig. 2. SDS/PAGE separation and immunostaining of purified wildtype and mutant flavanone 3b-hydroxylases. The wild-type enzyme (FHT, lane 1) and the mutants [Thr290]FHT (lane 2), [Ala290]FHT (lane 3) and [Val290]FHT (lane 4) were expressed in E. coli and isolated by a recently devised, rapid procedure [16]. The purified enzymes (2 mg of protein per lane) were separated by SDS/PAGE on a 12.5% polyacrylamide gel and either stained with Coomassie-brilliant-blue (top) or cross-reacted with polyclonal rabbit anti-FHT serum after nitrocellulose blotting (bottom). The molecular mass markers are indicated on the left.

q FEBS 2000


Table 3. Kinetics of wild-type and mutant flavanone 3b-hydroxylases.

Enzyme

Specific activity (mkat´kg21)

Km for 2-oxoglutarate (mmol´L21)

Wild-type FHT [Thr290]FHT [Ala290]FHT [Val290]FHT

31.8 6.6 2.5 0.3

1.9 25.6 80.0 189.6

[16]. The chromatographic elution profiles of the mutant enzymes [Thr290]FHT, [Ala290]FHT and [Val290]FHT did not differ essentially from that of the wild-type FHT, suggesting that the point mutations did not impose major conformational changes on the enzyme polypeptides [11] as had been proposed in another instance [25]. The three mutant FHTs were purified to homogeneity, and the pure enzymes revealed molecular masses equivalent to that of the wild-type FHT upon SDS/ PAGE separation and Western immunostaining (Fig. 2). Kinetic impact of Ser290 Sets of kinetic studies with the purified FHTs revealed that the substrate affinities of the wild-type and mutant enzymes differed greatly (Table 3). Replacement of Ser290 drastically reduced the specific enzyme activity in all instances even in the presence of a large excess of 2-oxoglutarate (up to 0.2 mm). While [Thr290]FHT and [Ala290]FHT retained about 20% and 8%, respectively, of the wild-type activity, [Val290]FHT had

Fig. 3. Circular dichroism spectroscopy of wild-type Petunia flavanone 3b-hydroxylase (FHT wt) and of mutants containing alanine (FHT Ala), threonine (FHT Thr) or valine (FHT Val) in place of Ser290. Spectra of the purified enzymes (0.3 mg´mL21) were recorded in 20 mm potassium phosphate buffer, pH 6.8, scanning through the far ultraviolet range (190±250 nm). The spectra are superimposed for comparison, and the molar ellipticities are expressed as millidegrees´cm22´dmol21.

nearly lost all the activity, falling to a level of less than 1% (Table 3). The apparent affinities of the wild-type and mutant enzymes to 2-oxoglutarate were compared by kinetic enzyme assays as described earlier [11] but using 0.1 m Bistris/HCl buffer of pH 6. A considerable drop in affinity was observed for all mutant FHTs in comparison to the wild-type enzyme (Table 3). While the Km value recorded for [Thr290]FHT was more than ninefold higher than that of the wild-type FHT, factors of more than 42-fold and about 100-fold, respectively, were determined for the [Ala290]FHT and [Val290]FHT mutants. The replacements of Ser290 in FHT by either Thr, Ala or Val must be regarded as conservative exchanges, in particular as such replacements occur naturally in some intermolecular dioxygenases (Table 2). Thus subtle differences in the primary sequence affect the functionality of these dioxygenases; and the pronounced reduction in the overall activity and 2-oxoglutaratebinding of the mutants strongly suggests that Ser290 serves as a ligand for 2-oxoglutarate in Petunia FHT as had been reported for the microbial FeII ±DAOCS±oxoglutarate complex [9]. Configuration of FHT The X-ray diffraction studies on IPNS and DAOCS documented that the tertiary structure of the dioxygenases is of critical importance for their functionality. These enzymes share only 19% sequence identity but fold spatially into very similar structures that include the catalytic `jelly roll' cavity [9]. It is conceivable that similar structural constraints apply to most of the 2-oxoglutarate-dependent enzymes, including FHT, as well as other nonheme iron(II) enzymes. The homology of the


q FEBS 2000

Fig. 4. Circular dichroism spectroscopy of Petunia FHT in the absence (1) and in the presence, of ferrous iron (2), 2-oxoglutarate (3), or a combination of both the cofactors (4). The purified enzyme (0.3 mg´mL21) was dissolved in 20 mm potassium phosphate buffer, pH 6.8, and superimposed spectra were recorded in the far ultraviolet range (190±250 nm) before and after the addition of Fe(NH4)SO4, 2-oxoglutarate (0.1 mm) or a combination of these cofactors. The molar ellipticities are expressed as millidegrees´cm22´dmol21.

polypeptide sequences of FHT and IPNS is in the order of 30% and appears to support such a proposal [5]. Moreover, the sequence identity and conservative exchanges in FHT are clustered mostly in those regions corresponding to motifs of highly conserved tertiary structure and functionality in IPNS (Fig. 1). In order to substantiate further the structural relationship of FHT and IPNS, CD spectroscopy was applied which allows the estimation of the relative percentages of a-helix and b-sheet conformations. This method yields characteristic patterns for defined protein structures in solution and has been used frequently to study conformational changes of enzymes [26,27]. CD spectra are available for IPNS [28,29], and therefore the purified wild-type and mutant FHTs were also subjected to CD spectroscopy in the far ultraviolet region. The CD profile of FHT displayed a characteristic double minimum at 222 nm and 208±210 nm and a maximum at 191±193 nm (Fig. 3), which revealed the presence of extended a-helical regions [26]. Furthermore, the superimposed spectra of wildtype and mutant FHTs fully matched (Fig. 3) and corroborated the fact that the mutations had not affected the catalytic fold of the enzyme [27]. Most importantly, however, the spectra closely resembled the CD spectroscopic pattern recorded for IPNS from S. clavuligerus and Streptomyces jumonjinensis [28,29], and overall the data predict very similar spatial structures for Petunia FHT and the microbial IPNS or DAOCS. The stabilization of FHT activity during purification of the enzyme from petals of red-flowering P. hybrida required the supplementation of buffers with 2-oxoglutarate and ferrous iron [24]; conformational modifications of the enzyme polypeptide upon cofactor binding were assumed to be responsible for this effect (L. Britsch, Merck KgaA, Damstadt, Germany; personal

communication). As significant structural rearrangements of polypeptides in solution can be distinguished by CD spectroscopy [26], the CD spectra of the purified and highly active Petunia FHT were recorded for comparison in the absence and presence of cofactors. The addition of ferrous iron, or 2oxoglutarate, or combinations of these cofactors, did not affect the CD spectra of the recombinant FHT (Fig. 4). These data strongly suggested that binding of the cofactors does not induce a major conformational change on the enzyme protein.

DISCUSSION More than 50 FeII/2-oxoglutarate-dependent dioxygenase enzymes have been characterized from microbial and plant as well as mammalian sources, and most of these enzymes are involved in biosynthetic pathways, e.g. the formation of cephalosporins [30], flavonoids [31±33], alkaloids such as scopolamine [34] and vindoline [35] or gibberellins [36]. All of the enzymes use the four electrons gained from decarboxylation of oxoglutarate to split dioxygen and create the reactive enzyme-ferryl species, but, despite their common cofactor requirements, formally, rather different reactions are being catalyzed. As an example, DAOCS catalyzes the expansion of the penicillin thiazolidine ring to the dihydrothiazine ring in cephalosporin biosynthesis, and molecular insight into the active site and cofactor binding of this class of enzyme has been obtained by structural analysis of DAOCS from S. clavuligerus [9]. Nevertheless, dioxygenases which act independently of 2-oxoglutarate and catalyze hydroxylation or desaturation reactions through one-electron and two-electron reduction of enzyme-bound oxygen, are known to be closely related; several

q FEBS 2000


of such enzymes have also been cloned. IPNS is by far the best characterized example (Fig. 1), and extensive spectroscopic and structural studies of IPNS and of the FeII ±IPNS complex had been reported [8]. DAOCS and IPNS therefore together represent the lead structures for intermolecular dioxygenases, and their scrupulous comparison revealed extended a-helical regions and few strictly conserved motifs in the overall marginally conserved primary sequences (19% identity). Petunia FHT shares a higher degree of sequence identity (Fig. 1) with IPNS (about 30%), and the CD spectra of these two enzymes, which fully matched [28,29] (Fig. 2), suggested an analogous organization in mostly a-helical conformation, as proteins in b-sheet conformation would be characterized by a minimum near 216±218 nm and a maximum near the region between 195 and 200 nm [26]. Furthermore, the CD spectroscopy revealed that binding of cofactors to Petunia FHT (Fig. 3) did not affect the enzyme conformation, although the crystallographic studies had suggested such minor effects for IPNS [8,10]. It is conceivable therefore that the overall tertiary structure and active site motifs of Petunia FHT are arranged just like those of IPNS, DAOCS and probably also other dioxygenases, irrespective of the major discrepancies in the primary enzyme structures. Intermolecular dioxygenases like IPNS or 1-aminocyclopropane-1-carboxylic acid (ACC) oxidase do not require 2-oxoacids for catalysis. For oxygen reduction, IPNS draws two electrons from its substrate ACV during cyclization [37], whereas for the same purpose ACC oxidase uses ascorbate, which is converted to dehydroascorbate [38]. Nevertheless, both enzymes contain the tripeptide motif Arg-Xaa-Ser that was assigned to oxoglutarate binding in Petunia FHT. The arginine of this motif was proposed to ligate oxoglutarate to FHT [11] and the substrate ACV to IPNS [10] for catalysis via salt bridges; the functionality of this residue in ACC oxidase has not yet been confirmed but is probably analogous in binding ascorbate. The recent report by ValegaÊrd et al. [9] on the FeII ±DAOCS±oxoglutarate complex presented the first crystal structure of an oxoglutarate-dependent dioxygenase and provided direct proof for the oxoglutarate-arginine binding. Surprisingly, the serine of the conserved Arg-Xaa-Ser motif was also shown to bind oxoglutarate [9], and the same function was suggested for this serine residue in other 2-oxoglutaratedependent enzymes. It remains nevertheless obvious that the presence of an Arg-Xaa-Ser motif does not always guarantee oxoglutarate-dependency. On the other hand, threonine, alanine, valine or isoleucine replace this residue in some other 2-oxoglutarate-dependent dioxygenases (Table 2), and the role of serine in the conserved tripeptide has been put into question. It must be emphasized, however, that human prolyl hydroxylase and chicken lysyl hydroxylase (Table 2) show rather low affinities to 2-oxoglutarate in comparison to Petunia FHT. The corresponding mutations of Petunia FHT presented in this report clearly demonstrated that Ser290 is absolutely required for high FHT activity. This does not rule out that the impact of the conserved serine residue on catalytic turnover might be modulated by the flanking sequences in other dioxygenases. In summary, Petunia FHT possesses a high degree of similarity with other 2-oxoglutarate-dependent dioxygenases and related nonheme iron(II) enzymes. DAOCS, the only fully characterized oxoglutarate-dependent dioxygenase so far, catalyzes the ring-expansion of penicillin to cephalosporin [9], while FHT hydroxylates its substrates, and the mechanism of FHT action requires further structural investigations. The revised short purification protocol for highly active Petunia FHT has been a major milestone on the way to crystallize the enzyme.

ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft and Fonds der Chemischen Industrie. The authors wish to thank Dr HaÈnisch, Biozentrum Basel, for the initial CD measurements and Dr Urbanke, Medizinische Hochschule Hannover, for valuable discussions.

REFERENCES 1. Heller, W. & Forkmann, G. (1993) Biosynthesis of flavonoids. In The Flavonoids, Advances in Research Since 1986 (Harborne, J.B., eds), pp. 399±425. Chapman & Hall, London. 2. Prescott, A.G.J. (1993) A dilemma of dioxygenases (or where biochemistry and molecular biology fail to meet). J. Exp. Bot. 44, 849±861. 3. De Carolis, E. & Luca, V. (1994) 2-Oxoglutarate-dependent dioxygenase and related enzymes: biochemical characterization. Phytochemistry 36, 1093±1107. 4. Britsch, L., Ruhnau-Brich, B. & Forkmann, G. (1992) Molecular cloning, sequence analysis, and in vitro expression of flavanone 3b-hydroxylase from Petunia hybrida. J. Biol. Chem. 267, 5380±5387. 5. Britsch, L., Dedio, J., Saedler, H. & Forkmann, G. (1993) Molecular characterization of flavanone 3b-hydroxylase: consensus sequence, comparison with related enzymes and the role of conserved histidines. Eur. J. Biochem. 217, 745±754. 6. Holton, T.A., Brugliera, F. & Tanaky, Y. (1993) Cloning and expression of flavonol synthase from Petunia hybrida. Plant J. 4, 1003±1010. 7. Suzuki, K., Yun, D.J., Chen, X.Y., Yamada, Y. & Hashimoto, T. (1999) An Atropa belladonna hyoscyamine 6b-hydroxylase gene is differentially expressed in the root pericycle and anthers. Plant. Mol. Biol. 40, 141±152. 8. Roach, P.L., Clifton, I.J., FuÈloÈp, V., Harlos, K., Barton, G.J., Hajdu, J., Andersson, I., Schofield, C.J. & Baldwin, J.E. (1995) Crystal structure of isopenicillin N-synthase is the first from a new structural family of enzymes. Nature 375, 700±704. 9. ValegaÊrd, K., Van Schelting, T., Lloyd, M.D., Hara, T., Ramaswamy, S., Perrakis, A., Thompson, A., Lee, H.-J., Baldwin, J.E., Schofield, C.J., Hajdu, J. & Andersson, I. (1998) Structure of a cephalosporin synthase. Nature 394, 805±809. 10. Roach, P.L., Clifton, I.J., Hensgens, C.M.H., Shibata, N., Schofield, C.J., Hajdu, J. & Baldwin, J.E. (1997) Structure of isopenicillin N synthase complexed with substrate and the mechanism of penicillin formation. Nature 387, 827±830. 11. LukacÏin, R. & Britsch, L. (1997) Identification of strictly conserved histidine and arginine residues as part of the active site in Petunia hybrida flavanone 3b-hydroxylase. Eur. J. Biochem. 249, 748±757. 12. Loke, P. & Sim, T.-S. (1998) Mutational evidence for the role of serine283 in Cephalosporium acremonium isopenicillin N synthase. FEMS Microbiol. Lett. 165, 353±356. 13. Sanger, F., Nicklen, S. & Coulsen, A.R. (1977) DNA sequencing with chain terminating inhibitors. Proc. Natl Acad. Sci. USA 74, 5463±5467. 14. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, USA. 15. LukacÏin, R., Springob, K., Urbanke, C., Ernwein, C., SchroÈder, G., SchroÈder, J. & Matern, U. (1999) Native acridone synthases I and II from Ruta graveolens L. form homodimers. FEBS Lett. 448, 135±140. 16. LukacÏin, R., GroÈning, I., Schiltz, E., Britsch, L. & Matern, U. (2000) Purification of recombinant flavanone 3b-hydroxylase from Petunia hybrida and assignment of primary site of proteolytic degradation. Arch. Biochem. Biophys. In press. 17. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680±685. 18. Towbin, H., Staehelin, T. & Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl Acad. Sci. USA 76, 4350±4354. 19. Britsch, L. (1990) Purification of flavanone 3b-hydroxylase from


20. 21. 22. 23. 24. 25.

26. 27.

28.

29.

Petunia hybrida: antibody preparation and characterization of a chemogenetically defined mutant. Arch. Biochem. Biophys. 276, 348±354. Sreerama, N. & Woody, R.W. (1993) A self-consistent method for the analysis of protein secondary structure from circular dichroism. Anal. Biochem. 209, 32±44. Deleage, G. & Geourjon, C. (1993) An interactive graphic program for calculating the secondary structure content of proteins from circular dichroism spectrum. Comput. Appl. Biosci. 9, 197±199. Sandermann, H. & Strominger, L. (1972) Purification and properties of C55-isoprenoid alcohol phosphokinase from Staphylococcus aureus. J. Biol. Chem. 247, 5123±5131. Bensadoun, A. & Weinstein, B. (1976) Assay of proteins in the presence of interfering material. Anal. Biochem. 70, 241±250. Britsch, L. & Grisebach, H. (1986) Purification and characterization of (2S)-flavanone 3-hydroxylase from Petunia hybrida. Eur. J. Biochem. 156, 569±577. Huang, Y. & Dennis, D.T. (1995) Histidine residues 139, 363 and 500 are essential for catalytic activity of cofactor-independent phosphoglyceromutase from developing endosperm of the castor plant. Eur. J. Biochem. 229, 395±402. Galla, H.-J., Ed. (1988) Spektroskopische Methoden in der Biochemie, pp. 23±30. Georg Thieme-Verlag, Stuttgart. Roberge, M., Dupont, C., Morosoli, R., Shareck, F. & Kluepfel, D. (1997) Asparagine-127 of xylanase A from Streptomyces lividans, a key residue in glycosyl hydrolases of superfamily 4/7: kinetic evidence for its involvement in stabilization of the catalytic intermediate. Protein Eng. 10, 399±403. Borovok, I., Landman, O., Kreisberg-Zakarin, R., Aharonowitz, Y. & Cohen, G. (1996) Ferrous active site of isopenicillin N synthase: genetic and sequence analysis of the endogenous ligands. Biochemistry 35, 1981±1987. Durairaj, M., Leskiw, B.K. & Jensen, S.E. (1996) Genetic and

q FEBS 2000

30. 31.

32. 33.

34. 35.

36. 37. 38.

biochemical analysis of the cysteinyl residues of isopenicillin N synthase from Streptomyces clavuligerus. Can. J. Microbiol. 42, 870±875. Baldwin, J.E. & Abraham, E.P. (1988) The biosynthesis of penicillins and cephalosporins. Nat. Prod. Rep. 5, 129±145. Forkmann, G., Heller, W. & Grisebach, H. (1980) Anthocyanin biosynthesis in flowers of Matthiola incana. Flavanone 3- and flavonoid 3 0 -hydroxylases. Z. Naturforsch., Sect. C Biosci. 35, 691±695. Britsch, L., Heller, W. & Grisebach, H. (1981) Conversion of flavanone to flavone, dihydroflavonol and flavonol with an enzyme system from cell cultures of parsley. Z. Naturforsch. Sect. C Biosci. 36, 742±750. Anzellotti, D. & Ibrahim, R.K. (1999) A novel a-ketoglutaratedependent dioxygenase catalyzes the 6-hydroxylation of partially methylated flavonols in Chrysoplenium americanum (Poster). Proceedings of the Phytochemical Society of North America, Montreal, Canada, pp. 34±35. Hashimoto, T. & Yamada, Y. (1986) Hyoscyamine 6b-hydroxylase, a 2-oxoglutarate-dependent dioxygenase, in alkaloid-producing root cultures. Plant Physiol. 81, 619±625. De Carolis, E., Chan, F., De Balsevich, J. & Luca, V. (1990) Isolation and characterization of a 2-oxoglutarate-dependent dioxygenase involved in the second-to-last step in vindoline biosynthesis. Plant Physiol. 94, 1323±1329. Hedden, P. & Graebe, J.E. (1982) Cofactor requirements for the soluble oxidases in the metabolism of the C20-gibberellins. J. Plant Growth Regul. 1, 105±116. White, R.L., John, E.-M., Baldwin, J.E. & Abraham, E.P. (1982) Stoichiometry of oxygen consumption in the biosynthesis of isopenicillin from a tripeptide. Biochem. J. 203, 791±793. Dong, J.G., Fernandez-Maculet, J.C. & Yang, S.-F. (1992) Purification and characterization of 1-aminocyclopropane-1-carboxylate oxidase from apple fruit. Proc. Natl Acad. Sci. USA 89, 9789±9793.