Metallomics

Dynamic Article Links

Cite this: Metallomics, 2011, 3, 410–416

PAPER

www.rsc.org/metallomics

Control of the stereo-selectivity of styrene epoxidation by cytochrome P450 BM3 using structure-based mutagenesiswz Wei-Cheng Huang,a Paul M. Cullis,b Emma Lloyd Ravenb and Gordon C. K. Roberts*a Received 19th November 2010, Accepted 17th December 2010 DOI: 10.1039/c0mt00082e The potential of flavocytochrome P450 BM3 (CYP102A1) from Bacillus megaterium for biocatalysis and biotechnological application is widely acknowledged. The catalytic and structural analysis of the Ala82Phe mutant of P450 BM3 has shown that filling a hydrophobic pocket near the active site improved the binding of small molecules, such as indole (see Huang et al., J. Mol. Biol., 2007, 373, 633) and styrene. In this paper, additional mutations at Thr438 are shown to decrease the binding of and catalytic activity towards laurate, whereas they significantly increased the stereo-specificity of styrene epoxidation. Production of R-styrene oxide with 48% and 64% e.e., respectively, was achieved by the Ala82Phe-Thr438Leu and Ala82Phe-Thr438Phe mutants. These structure-based mutants of P450 BM3 illustrate the promise of rational design of synthetically useful biocatalysts for regio- and stereo- specific mono-oxygenation reactions.

Introduction The cytochromes P450 are a large superfamily of haem-thiolate mono-oxygenase enzymes, found in almost all forms of life, which catalyse the activation of molecular oxygen and the addition of one atom of oxygen to their substrate.2 There is considerable sequence diversity within the superfamily; those members whose structures have been determined share a common overall fold, while differing markedly in their active site architecture, leading to very diverse substrate specificity. The bacterial cytochrome P450 CYP102A1 (P450 BM3) from Bacillus megaterium has been extensively studied structurally and mechanistically.3,4 It catalyses the hydroxylation of C12–C16 saturated fatty acids at the (o-1), (o-2) and (o-3) positions.5,6 P450 BM3 is a 119kDa polypeptide which contains a P450 domain and a diflavin NADPH-cytochrome P450 reductase domain5,7 similar to that in the mammalian drugmetabolising mono-oxygenase system. Unlike most P450s, therefore, which require additional electron transfer partners, P450 BM3 is catalytically self-sufficient. Perhaps because of this, P450 BM3 also has the highest catalytic activity of any a

Henry Wellcome Laboratories of Structural Biology, Department of Biochemistry, University of Leicester, Henry, Wellcome Building, PO Box 138, Lancaster Road, Leicester LE1 9HN, United Kingdom. E-mail: [email protected]; Tel: +44(0)116 229 7100 b Department of Chemistry, University of Leicester, Leicester LE1 7RH, United Kingdom w This article is published as part of a themed issue on Cytochromes, Guest Edited by Norbert Jakubowski and Peter Roos. z Electronic supplementary information (ESI) available. See DOI: 10.1039/c0mt00082e

410

Metallomics, 2011, 3, 410–416

P450 mono-oxygenase identified to date. It has been widely used as the starting point for the construction of mutants with altered specificity for use in chemical synthesis, where the ability of P450s to insert an oxygen atom into an unactivated C–H bond has potentially valuable applications. P450 BM3 can be prepared in large quantities by soluble expression in Escherichia coli3,8 and it can also be expressed on the surface of E. coli,9 making it an ideal industrial biocatalytic enzyme.10 Both structure-based mutant design and directed evolution approaches have been used to produce mutants with markedly altered specificity and/or regiospecificity (e.g. ref. 11–13). Crystallographic studies of wild-type and mutant forms of P450 BM3 have provided evidence for ‘substrate-free’ and ‘substrate-bound’ conformations,1,14,15 which appear to coexist in equilibrium in solution in the absence of substrate. However, in none of the available structures of substrate complexes is the fatty acid bound with the (o-1), (o-2) and/or (o-3) carbons positioned close to the iron in a position for hydroxylation. Instead, the ‘o-end’ of the fatty acid becomes sequestered in a hydrophobic pocket between phenylalanines 81 and 87, created by rotation of the aromatic ring of phenylalanine 87 by B901 and a rearrangement of nearby side-chains (notably isoleucine 263 and leucine 437), with its terminal methyl group in contact with alanine 82. In this position, the o to o-6 carbons of the fatty acid are all between 7.5 A˚ and 10 A˚ from the iron centre, too distant for hydroxylation (Fig. 1). The existence of a hydrophobic pocket within the active site cavity, but relatively distant from the iron, also has implications for attempts to engineer the enzyme’s activity towards novel target substrates. It seems likely that small hydrophobic molecules could preferentially bind in the pocket between phenylalanines This journal is

c

The Royal Society of Chemistry 2011

the catalysis of styrene epoxidation. Styrene oxide is a useful intermediate in the synthesis of a range of pharmaceuticals (e.g. ref. 18), and enzymatic bio-transformations offer attractive alternatives for the production of chiral epoxides.19 The double mutants were found to show increased catalytic efficiency for styrene epoxidation and also, notably, to produce a significant enantiomeric excess of R-styrene oxide, suggesting that they have the potential to be a useful addition to the range of P450 BM3 mutants for biocatalysis.

Experimental procedures Materials

Fig. 1 Substrate-binding pocket in the N-palmitoylglycine-bound crystal structure of the wild-type P450 BM3 (PDB: 1JPZ).15 The van der Waals surface of the substrate-binding pocket is coloured in magenta and a space-filling representation of the substrate is in blue and the haem coloured in red.

81 and 87, resulting in non-productive complexes. There have been many reports of changing the substrate specificity of P450 BM3 by substitution of phenylalanine 87 by a smaller residue such as glycine, alanine or valine (e.g. ref. 8,16,17). Such substitutions would destroy the hydrophobic pocket, but would also increase the active site volume. More recently we have filled the hydrophobic pocket rather than destroying it, by substitution of alanine 82 by the larger hydrophobic residues isoleucine, phenylalanine and tryptophan,1 while subsequently others have combined an F87A or F87V substitution with the replacement of alanine 328 by larger residues to avoid too great an increase in active site volume.11,13 We observed that the mutants A82F and A82 W bound fatty acids orders of magnitude more tightly compared to the wild type, and were effective catalysts of the oxidation of indole, resulting in formation of indigo, suggesting that they may exhibit generally improved activity towards small molecules. In order to build on the increased activity of the A82F mutant towards small substrates with the aim of improving regio- and stereo-selectivity, we have focussed on residue threonine 438, which is located near the o-terminal end of the bound substrate with its side-chain pointing into the substrate binding channel1 and which has not hitherto been studied by mutagenesis. Simple models of A82F-T438 mutants of P450 BM3 (Fig. 2) indicate that larger side chains at position 438 have the potential to constrain the catalytic position of the bound substrate in the ligand binding channel. Side-chains up to the size of phenylalanine could be accommodated in the binding site of the A82F mutant while still leaving room for a substrate, but the larger tyrosine and tryptophan could not; in addition, a tyrosine residue in this position would place an oxygen atom within the hydrophobic pocket, with likely deleterious effects on substrate binding. In the work reported here, we have explored the use of double mutants of P450 BM3 combining the A82F mutation with the substitution of Thr438 by hydrophobic residues (Val, Ile, Leu, and Phe) for This journal is

c

The Royal Society of Chemistry 2011

The QuikChange XL mutagenesis kit was obtained from Stratagene, UK, and oligonucleotide primers from the Protein and Nucleic Acid Chemistry Laboratory, University of Leicester, UK. Restriction enzymes were obtained from New England Biolabs. Chromatography columns and media from were obtained from Amersham Biosciences, Completet Protease inhibitors from Roche. All other chemicals, of analytical grade or higher, were from Sigma Aldrich UK Ltd. Construction, expression and purification of P450 BM3 A82F-T438 mutants Mutants at T438 (Val/Ile/Leu/Phe) along with A82F were constructed by using the QuikChange XL or Multi SiteDirected Mutagenesis kit from Stratagene. P450 BM3 and its mutants were expressed using E. coli JM109 cells harbouring plasmids based on pGLWBM3AW and proteins were expressed and purified as described previously.1 Briefly, the cell pellets were resuspended in Buffer A (50 mM potassium phosphate, pH 7.5, 1 mM EDTA, 1 mM benzamidineHCl, 1 mM DTT) supplemented with Completet Protease inhibitors, lysed by sonication and centrifuged at 50 000 g for 1 h at 4 1C. The soluble fraction was loaded onto a DEAE-Sepharose Fast-Flow column pre-equilibrated with buffer A. The proteins were eluted with a linear gradient of 0–500 mM potassium chloride in buffer A. The fractions exhibiting the highest haem content, with A418/A280 4 0.3, were pooled and then loaded onto a Hi-Load 26/10 Q-Sepharose HP column, and the protein samples were eluted with a linear gradient of 0-500 mM potassium chloride in 20 mM potassium phosphate, pH 7.5, 1 mM EDTA, 1 mM benzamidineHCl, 1 mM DTT. Again fractions with the highest haem content (A418/A280 4 0.5–0.8), were pooled. Finally, to remove any bound substratelike molecules, the protein was treated with 5 molar equivalents of NADPH for 5 min at room temperature; any products, NADP+, or residual NADPH were removed from the protein sample by extensive ultrafiltration against buffer A prior to storage in 50% (v/v) glycerol at 20 1C. All the cytochrome P450 BM3 mutants exhibited typical cytochrome P450 spectra on reduction in the present of carbon monoxide with absorbance maxima at 448 nm and minimal (o5%) formation of inactive ‘P420’ enzyme. Protein concentrations were determined by the CO-difference method using a difference extinction coefficient De (450–490 nm) of 91 mM1 cm1.20 Metallomics, 2011, 3, 410–416

411

S-styrene oxide and R-styrene oxide were both set up at 500 mM in 10 ml 50 mM potassium phosphate buffer and prepared through the same extraction procedure in order to compare and identify the products catalysed from P450 BM3. The petroleum ether phase was analysed by gas chromatography using a chiral Supelco b-DEXTM 120 column. Determination of dissociation constant (Kd) for substrate binding by optical spectroscopy

Fig. 2 Models of the active site of the palmitate complex of P450 BM3. (A) the A82F mutant and (B-D) the mutants with in addition (B) isoleucine, (C) leucine, and (D) phenylalanine in place of Thr438. The binding site models are based on the palmitate-bound crystal structure of the P450 BM3 A82F mutant (PDB: 2UWH; Huang, et al.1). The van der Waals surface of the substrate-binding pocket is coloured in magenta and a space-filling representation of the substrate palmitate is superimposed in blue. The side chains of residue 438 are highlighted in yellow, the F82 residue in brown and the haem in red.

Determination of catalytic activity by NADPH consumption NADPH consumption assays were carried out using reaction mixtures containing 0.1–0.5 mM P450 BM3 and varying concentrations of substrates. Reactions were initiated by the addition of 250 mM NADPH, and the decrease in absorbance at 340 nm monitored (e = 6210 M1 cm1). Initial rates were obtained from the first 20 s of the reaction.

UV/Visible spectroscopy was carried out using a Cary 300 Bio UV/Visible spectrophotometer equipped with a Peltier temperature control unit and Cary WinUV software. All experiments were conducted in buffer of 50 mM potassium phosphate, pH 8.0, at 30 1C. The absorbance change of the haem Soreˆt band was measured on addition of increasing quantities of ligand to a protein solution containing 0.1–0.5 mM P450 BM3 in the sample cuvette, and to buffer solution in the reference cuvette. The data were corrected for any dilution of the protein during the titration of ligand, and the difference in absorbance change between 390 nm and 420 nm, D(A390  A420), was plotted against substrate concentration. The dissociation constants (Kd) were estimated using nonlinear curve fitting in Microcal OriginPro 7.5 software. Computational docking The protein structure used was based on the crystal structure of the palmitate complex of the A82F mutant (PDB code 2UWH;1), from which the bound ligand was removed; appropriate side-chains were altered to simulate the desired mutations and local energy minimization was performed. Protein-ligand docking was performed using the CCDC program GOLD.21 All residues were kept rigid during docking runs, except that the side-chains of F82, F87 and residue-438 were allowed to be flexible using library rotamers of corresponding amino acids. From 104 trials, the lowest-energy 10 docking poses were examined; energies were estimated using ChemScore22 using parameters,23 optimised for ligand-protein docking with P450s and other heme enzymes.

Identification and quantification of laurate hydroxylation products by NMR spectrometry The products of laurate hydroxylation by wild-type and mutants of P450 BM3 were identified from 1 ml reactions initially containing 250 mM laurate and 250 mM NADPH in 50 mM potassium phosphate buffer in 2H2O, pH 8.0. The reaction was started by addition of 0.3 mM P450 BM3 and incubated at 30 1C for 3 h to complete the whole reaction. The NMR spectra of the reaction mixtures were recorded at 298 K using a Bruker 600 MHz instrument. The quantities of each product were estimated by integration of the resonances of the methyl group of the fatty acid. Analysis of styrene epoxidation by gas chromatography Reaction mixtures (final volume 10 ml) contained 500 mM styrene and 200 mM NADPH in 50 mM potassium phosphate buffer, pH 8.0; the reaction was started by addition of 0.3 mM enzyme and incubated at 30 1C for 3 h in order to allow the reaction to go to completion. The reaction mixtures were then extracted with petroleum ether. The reference of authentic 412

Metallomics, 2011, 3, 410–416

Results and discussion Mutant characterisation The proteins of the P450 BM3 double mutants, A82F/T438(V/I/L/F), like the A82F parent mutant,1 were initially purified in a predominately high-spin state, as indicated by their absorbance spectra, suggesting that a substrate-like molecule occupied the active site and displaced the water molecule in the sixth coordination position of the iron. Treatment of the purified proteins with 5 molar equivalents of NADPH, followed by buffer exchange using extensive ultrafiltration to remove reaction products, resulted in conversion to a predominantly ( Z 58%) low-spin form for the A82F-T438(V/I/F) mutants, but not for the A82F/T438L mutant (supplemental Fig. S1). Although the A82F/T438L mutant, like the other mutants, is able to oxidise NADPH and to hydroxylate lauric acid, indole and styrene (see below), its haem remained around 80% in the high-spin state after the treatment with NADPH. It is possible that it is the side-chain This journal is

c

The Royal Society of Chemistry 2011

Table 1

Binding and kinetic parameters for lauric acid with P450 BM3 and A82F/T438(V/I/LF) mutants

Enzyme

Kda (mM)

High spin (%)

KMb (mM)

kcatb (sec1)

Coupling ratioc (%)

Apparent rate of product formationd (sec1)

Wild-typee A82Fe A82F-T438V A82F-T438I A82F-T438L A82F-T438F

270 (14) 0.34 (0.03) 86 (9.3) 35 (4.9) N/A 45.6 (4.1)

60 92 44 63 [79]f 53

265 (19) o20 112 (18) 89 (9) N/A N/A

28 (1) 26 (1) 26 (1) 28 (1) N/A N/A

98 89 77 70 68 25

9.2 9.0 7.4 7.2 3.9 0.1

(2) (8) (6) (8) (12) (3)

a Dissociation constants (Kd) were estimated, using a single-site binding model, from the absorbance changes, D(A390  A420), on addition of 0 to 1 mM of lauric acid to a protein solution containing 0.8 mM enzyme. b KM and kcat were estimated from the initial rates of NADPH consumption at 0 to 1 mM laurate, 250 mM NADPH and 0.2 mM enzyme. c Hydroxylated products were quantified from the 1H NMR spectra. The coupling ratio (%) was defined as 100  (total amount of hydroxylated products)/(amount of NADPH consumed). d Product formation rates were estimated by multiplying the apparent NADPH consumption rate (at 0.3 mM P450 BM3, 250 mM laurate, 250 mM NADPH in 50 mM potassium phosphate, 2H2O) by the coupling ratio. e Parameters for wild-type P450 BM3 and the A82F mutant are from Huang et al.1. f Measured in the absence of substrate.

of the leucine-438 residue in the active site of the A82F/T438L mutant which displaces the water molecule coordinated to the haem iron, resulting in the predominately high spin state. All four A82F/T438(V/I/L/F) mutants were able to hydroxylate lauric acid. As we showed earlier,1 the A82F mutation leads to substantially tighter binding of lauric acid. The additional mutations at position 438 decrease binding relative to P450 BM3 A82F, but all four double mutants still bind lauric acid more tightly than the wild-type enzyme (Table 1). The A82F/T438V and A82F/T438I mutants catalyse hydroxylation of lauric acid almost as efficiently as does P450 BM3 A82F, but for A82F/T438L and, particularly, A82F/T438F the rate of hydroxylation and its coupling to NADPH consumption is significantly lower (Table 1). This presumably reflects clashes between the larger side-chains at position 438 and the bound substrate molecule (Fig. 2). We earlier reported that the introduction of P450 BM3 A82F into E. coli JM109 host cells led to the generation of an insoluble blue dye, which was shown to arise from the hydroxylation of indole by the enzyme and the formation of indigo.1 Similarly, cells containing the A82F/T438L or A82F/T438F mutants of P450 BM3 generated indigo during normal growth, indicating that these mutants can also hydroxylate indole. Interestingly, cells containing the A82F/T438V or A82F/T438I mutants of P450 BM3 did not generate indigo

Table 2

and were thus presumably unable to hydroxylate indole efficiently (Fig. S2, ESIz). The fact that replacement of Thr438 by the similar sized Val438 substantially decreases activity against indole suggests the possibility that hydrogenbonding to this threonine side-chain might be involved in orienting indole in the active site. The larger side-chains at this position in the A82F/T438L and A82F/T438F would substantially decrease the size of the hydrophobic pocket and this might then lead to orientation of the indole molecule without the need for hydrogen-bonding to Thr438. Styrene binding and epoxidation The binding of styrene to wild-type P450 BM3 and the A82F/T438(V/I/L/F) mutants was investigated by monitoring the change of the haem Soreˆt band absorbance on increasing the styrene concentration. For wild-type P450 BM3 styrene binding, like indole binding,1 leads to a ‘Type-I-like’ binding spectrum24 representing a partial shift (B20%) towards the high-spin state. Similar shifts were seen for the A82F/T438V, A82F/T438I and A82F/T438F mutants, whereas the A82F mutant showed a much larger shift and the A82F/T438L mutant showed a modest shift from 79% high-spin in the absence of substrate to 86% at saturating styrene (Fig. S3, ESIz), making it impossible to determine Kd values accurately. The concentration-dependence of the optical changes could be

Styrene binding and epoxidation by A82F/T438(V/I/L/F) mutants

Enzyme

NADPH R-styrene Dissociation constant, Dissociation constant, consumption Styrene oxide Kd2a (mM) High spina (%) rate (s1) formedb (mM) Couplingc (%) oxided (% e.e.) Kd1(a (mM)

Wild-type A82F A82F-T438V A82F-T438I A82F-T438L A82F-T438F

0.17 (0.01) 0.32 (0.02) 0.39 (0.02) 41.87 (5.31) —e o0.05

0.93 (0.06) 0.38 (0.02) 42 410 —e 1.07 (0.08)

22 98 23 40 86 30

0.9 6.0 2.4 3.3 13.9 1.8

0 68 41 27 39 20

(1.9) (1.6) (1) (1.7) (0.5)

0 23 (0.6) 14 (0.5) 9 (0.3) 13 (0.6) 7 (0.2)

0 4 (2) 9 (7) 2 (1) 48 (5) 64 (8)

a The dissociation constants (Kd) were estimated, using a two-binding-site model incorporating a tight-binding correction,25 from the absorbance changes of the haem Soreˆt band D(A390  A420) on addition of increasing quantities of styrene. b The amount of styrene oxide formed from the reaction mixtures of 300 mM NADPH, 800 mM styrene and 1 mM P450 BM3 was estimated by peak area integration of HPLC chromatogram. c The coupling (%) was defined as 100  (amounts of styrene oxide)/(amounts of NADPH supplied). d The enantiomeric excess (e.e.) is defined as: e.e. (%) = 100  |(R-S)|/(R+S),26 in which S and R represent the intensity (%) of the respective peak areas in the GC chromatogram of the two styrene oxide stereo isomers. e Kd values could not be determined accurately due to the very small change in absorbance.

This journal is

c

The Royal Society of Chemistry 2011

Metallomics, 2011, 3, 410–416

413

Fig. 3 Models of the binding of styrene and 3-chlorostyrene to P450 BM3 mutants. The predicted binding of styrene to P450 BM3 A82F (A) and to A82F/T438F (B) shows that the styrene molecule adopts essentially the same mode of binding in both cases, but with a significantly more restricted orientation in the latter case. This orientation is such that the product would be the R-isomer, as shown in (D) with the oxygen atom of R-styrene oxide over the haem iron. (C) shows the predicted mode of binding of 3-chlorostyrene to the F82G mutant. The protein structure used was based on the crystal structure of the palmitate complex of the A82F mutant (PDB code 2UWH; Huang et al.1) and the models were obtained by computational docking as described in the Experimental section.

described by a two-binding-site model, with one dissociation constant in the micromolar range and one in the millimolar range; the latter could not be determined accurately due to the limited aqueous solubility of styrene. Interestingly, indole and styrene show similar binding to the A82F mutant, with Kd1 0.14 mM for indole and 0.32 mM for styrene, and Kd2 0.24 mM for indole and 0.38 mM for styrene (Table 2).1 When Thr438 is replaced by the isosteric but hydrophobic valine, there is no effect on the dissociation constant for the tight binding site for styrene (Kd1 = 0.32 mM for A82F, 0.39 mM for A82F-T438V). However, replacement of this residue by larger hydrophobic side-chains has disparate effects: Kd1 is 130-fold higher in A82F/T438I and 4 6-fold lower for A82F/T438F (Table 2). The formation of styrene oxide was measured in reaction mixtures containing 300 mM NADPH, 800 mM styrene and 1 mM enzyme, allowing the reaction to proceed until all the NADPH had been consumed. As shown in Table 2, undetectable amounts of styrene oxide were formed by the wild-type enzyme. However, as noted earlier for indole,1 the A82F mutation fills the hydrophobic pocket and results in increased activity towards small molecules; for styrene, the coupling of epoxidation to NADPH oxidation was as high as 23%. This mutant produced a racemic mixture of R- and S-styrene oxide in a ratio of around 1 : 1 (o 5% e.e.). The A82F/T438(V/I/L/F) mutants showed somewhat decreased styrene epoxidation activity compared to the A82F mutant, with a coupling to NADPH oxidation of 23% for A82F diminishing to B10% for Thr438 mutants (Table 2). The A82F/T438V and A82F/T438I mutants, like the parent A82F, essentially generated a racemic mixture, with o10% e.e. Strikingly, however, the A82F/T438V and A82F/T438F 414

Metallomics, 2011, 3, 410–416

mutants produced predominantly R-styrene oxide, with an enantiomeric excess of 50–60% for these two mutants (Table 2). It is therefore apparent that the size of side chain at residue 438 is a key determinant of the stereo-selectivity of P450 BM3 in styrene epoxidation. It is interesting to note that the side-chains in positions 82 and 438 have quite distinct effects on styrene epoxidation by P450 BM3. Introduction of a phenylalanine side-chain at position 82 (A82F) markedly increases the activity of the enzyme as a styrene epoxidase, but with no stereo-selectivity. By contrast, introduction of an additional phenylalanine side-chain at position 438 (A82F/T438F) slightly decreases the activity of the enzyme towards styrene but leads to a substantial degree of stereoselectivity. A number of other mutants of P450 BM3, either structure-based or produced by random mutagenesis and selection, have been examined for their activity towards styrene.17,27 While there are significant variations in the reported activity of wild-type P450 BM3, the most striking observation is that removal of the phenylalanine residue at position 87 (F87G) leads to the production of stereochemically almost pure R-3-chlorostyrene oxide.17 In an attempt to rationalise these observations, we have used computational docking methods to investigate the effects of these mutations on the possible modes of styrene and 3-chlorostyrene binding in the P450 BM3 mutants. Docking of styrene into the A82F/T438F mutant revealed a favoured mode of binding in which the styrene molecule is positioned between the three phenylalanine residues at positions 82, 87 and 438, making non-polar interactions with them and locating the C1–C2 atoms of styrene above the iron atom of haem (Fig. 3). It is apparent from Fig. 3 that, in this mode of binding, as the side chain of residue 438 increases in size from threonine to phenylalanine, more space is occupied on one side of the styrene molecule, restricting its possible orientation and explaining the progressive increase in the proportion of R-styrene oxide as the size of the side chain of residue 438 increases. Indeed, the top 10 lowest-energy docked structures for styrene binding to the A82F/T438F mutant show an orientation which would lead to the R-epoxide (Fig. 3). Docking of 3-chlorostyrene into the F87G mutant again shows a single mode of binding in the 10 lowest-energy docked structures, but this is quite different from that seen for the A82F/T438F mutant. The ring of 3-chlorostyrene is situated at a position similar to that formerly occupied by the ring of Phe87 ring (Fig. 3), in an orientation which would lead to the production of the R-stereoisomer, is substantially preferred by this mutant. By contrast, the docking scores indicate no significant orientational preference for styrene itself; in this mutant, therefore, the stereo-selectivity depends strongly on the ligand substitution. Thus the enantiomeric excess achieved in the epoxidation of styrene derivatives by mutants of P450 BM3 is likely to depend on the combination of the substitution pattern of the substrate and the nature of the side-chains defining the hydrophobic pocket. We have shown that decreasing the size of the pocket by the A82F and T438F substitutions leads to a substantial increase in stereospecificity, but this is not yet sufficient for pratical biocatalytic application, at least with unsubstituted This journal is

c

The Royal Society of Chemistry 2011

styrene. A further residue which defines the size and shape of the pocket is Ala328, and substitution of this residue by larger hydrophobic side-chains has been shown to improve activity and regioselectivity in epoxidation of amorpha-4,11-diene11 and of (4 R)-limonene.13 A combination of substitutions at positions 82, 87, 328 and 438 has the potential to yield an enzyme which will catalyse highly stereospecific epoxidation reactions, but it is likely that the particular combination of substitutions required will depend on the substrate of interest.

Acknowledgements This work was supported by the BBSRC (grant E20186) and by an ORSAS grant (to W.-C. H.). We are grateful to Professor Luet Wong for the gift of the pGLWBM3 plasmid and to Dr Andrew C. G. Westlake and Dr Jacqueline Ellis for valuable discussions.

13

Notes and references 1 W.-C. Huang, A. C. G. Westlake, J.-D. Mare´chal, M. G. Joyce, P. C. E. Moody and G. C. K. Roberts, Filling a hole in cytochrome P450 BM3 improves substrate binding and catalytic efficiency, J. Mol. Biol., 2007, 373, 633–651. 2 T. M. Makris, I. Denisov, I. Schlichting and S. G. Sligar, in Cytochrome P450: Structure, Mechanism and Biochemistry, ed. P. R. O. d. Montellano, Kluwer Academic/Plenum Publishers, New York, 2005, pp. 149–182; M. J. Cryle, P. R. Ortiz de Montellano and J. J. De Voss, Cyclopropyl containing fatty acids as mechanistic probes for cytochromes P450, J. Org. Chem., 2005, 70, 2455–2469. 3 A. W. Munro, D. G. Leys, K. J. McLean, K. R. Marshall, T. W. B. Ost, S. Daff, C. S. Miles, S. K. Chapman, D. A. Lysek, C. C. Moser, C. C. Page and P. L. Dutton, P450 BM3: the very model of a modern flavocytochrome, Trends Biochem. Sci., 2002, 27, 250–257. 4 T. L. Poulos, Intermediates in P450 catalysis, Philos. Trans. R. Soc. London, Ser. A, 2005, 363, 793–806; O. Pylypenko and I. Schlichting, Structural aspects of ligand binding to and electron transfer in bacterial and fungal P450s, Annu. Rev. Biochem., 2004, 73, 991–1018. 5 A. J. Fulco, P450BM-3 and other inducible bacterial P450 cytochromes: Biochemistry and regulation, Annu. Rev. Pharmacol., 1991, 31, 177–203. 6 L. O. Narhi and A. J. Fulco, Characterization of a catalytically self-sufficient 119,000- dalton cytochrome P-450 monooxygenase induced by barbiturates n Bacillus megaterium, Journal of Biological Chemistry., 1986, 261, 7160–7169. 7 L. O. Narhi and A. J. Fulco, Identification and characterization of two functional domains in cytochrome P-450BM-3, a catalytically self-sufficient monooxygenase induced by barbiturates in Bacillus megaterium, Journal of Biological Chemistry, 1987, 262, 6683–6690. 8 A. B. Carmichael and L.-L. Wong, Protein engineering of Bacillus megaterium CYP102., Eur. J. Biochem., 2001, 268, 3117–3125. 9 S. K. Yim, D. H. Kim, H. C. Jung, J. G. Pan, H. S. Kang, T. Ahn and C. H. Yun, Surface display of heme- and diflavin-containing cytochrome P450 BM3 in Escherichia coli: a whole cell biocatalyst for oxidation, J. Microbiol. Biotechnol., 2010, 20, 712–7, DOI: JMB020-04-10 [pii]. 10 R. Bernhardt, Cytochromes P450 as versatile biocatalysts, J. Biotechnol., 2006, 124, 128–145. 11 J. A. Dietrich, Y. Yoshikuni, K. J. Fisher, F. X. Woolard, D. Ockey, D. J. McPhee, N. S. Renninger, M. C. Y. Chang, D. Baker and J. D. Keasling, A Novel Semi-biosynthetic Route for Artemisinin Production Using Engineered Substrate-Promiscuous P450(BM3), ACS Chem. Biol., 2009, 4, 261–267, DOI: 10.1021/ cb900006h. 12 J. C. Lewis, S. Bastian, C. S. Bennett, Y. Fu, Y. Mitsuda, M. M. Chen, W. A. Greenberg, C. H. Wong and F. H. Arnold,

This journal is

c

The Royal Society of Chemistry 2011

14

15 16

17

18

Chemoenzymatic elaboration of monosaccharides using engineered cytochrome P450(BM3) demethylases, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 16550–16555, DOI: 10.1073/ pnas.0908954106; A. M. Sawayama, M. M. Y. Chen, P. Kulanthaivel, M. S. Kuo, H. Hemmerle and F. H. Arnold, A Panel of Cytochrome P450 BM3 Variants to Produce Drug Metabolites and Diversify Lead Compounds, Chem.–Eur. J., 2009, 15, 11723–11729, DOI: 10.1002/chem.200900643; M. G. Shapiro, G. G. Westmeyer, P. A. Romero, J. O. Szablowski, B. Kuster, A. Shah, C. R. Otey, R. Langer, F. H. Arnold and A. Jasanoff, Directed evolution of a magnetic resonance imaging contrast agent for noninvasive imaging of dopamine, Nat. Biotechnol., 2010, 28, 264–270, DOI: DOI: 10.1038/nbt.1609; C. J. C. Whitehouse, S. G. Bell and L. L. Wong, Desaturation of Alkylbenzenes by Cytochrome P450(BM3) (CYP102A1), Chem.–Eur. J., 2008, 14, 10905–10908, DOI: 10.1002/chem.200801927; C. J. C. Whitehouse, S. G. Bell, W. Yang, J. A. Yorke, C. F. Blanford, A. J. F. Strong, E. J. Morse, M. Bartlam, Z. H. Rao and L. L. Wong, A Highly Active Single-Mutation Variant of P450(BM3) (CYP102A1), ChemBioChem, 2009, 10, 1654–1656, DOI: 10.1002/cbic.200900279. A. Seifert, S. Vomund, K. Grohmann, S. Kriening, V. B. Urlacher, S. Laschat and J. Pleiss, Rational Design of a Minimal and Highly Enriched CYP102A1 Mutant Library with Improved Regio-, Stereo- and Chemoselectivity, ChemBioChem, 2009, 10, 853–861, DOI: 10.1002/cbic.200800799. D. C. Haines, B. Chen, D. R. Tomchick, M. Bondlela, A. Hegde, M. Machius and J. A. Peterson, Crystal Structure of InhibitorBound P450BM-3 Reveals Open Conformation of Substrate Access Channel, Biochemistry, 2008. e-publish; M. G. Joyce, H. M. Girvan, A. W. Munro and D. Leys, A single mutation in cytochrome P450 BM3 induces the conformational rearrangement seen upon substrate binding in the wild-type enzyme, J. Biol. Chem., 2004, 279, 23287–23293; H. Li and T. L. Poulos, The structure of the cytochrome P450BM-3 haem domain complexed with the fatty acid substrate, palmitoleic acid, Nat. Struct. Biol., 1997, 4, 140–146; K. G. Ravichandran, S. S. Boddupalli, C. A. Hasermann, J. A. Peterson and J. Deisenhofer, Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450’s, Science, 1993, 261, 731–736. D. C. Haines, D. R. Tomchick, M. Machius and J. A. Peterson, Pivotal role of water in the mechanism of P450BM-3, Biochemistry, 2001, 40, 13456–13465. D. Appel, S. Lutz-Wahl, P. Fischer, U. Schwaneberg and R. D. Schmid, A P450 BM-3 mutant hydroxylates alkanes, cycloalkanes, arenes and heteroarenes, J. Biotechnol., 2001, 88, 167–171; P. C. Cirino and F. H. Arnold, Regioselectivity and activity of cytochrome P450 BM-3 and mutant F87A in reactions driven by hydrogen peroxide, Adv. Synth. Catal., 2002, 344, 932–937; L. A. Cowart, J. R. Falck and J. H. Capdevila, Structural determinants of active site binding affinity and metabolism by cytochrome P450 BM-3, Arch. Biochem. Biophys., 2001, 387, 117–124; S. Graham-Lorence, G. Truan, J. A. Peterson, J. R. Falck, S. Wei, C. Helvig and J. H. Capdevila, An active site substitution, F87V, converts cytochrome P450 BM-3 into a regio- and stereoselective (14S,15R)-arachidonic acid epoxygenase, J. Biol. Chem., 1997, 272, 1127–1135; W. T. Sulistyaningdyah, J. Ogawa, Q. S. Li, C. Maeda, Y. Yano, R. D. Schmid and S. Shimizu, Hydroxylation activity of P450 BM-3 mutant F87V towards aromatic compounds and its application to the synthesis of hydroquinone derivatives from phenolic compounds, Appl. Microbiol. Biotechnol., 2005, 67, 556–562. Q. S. Li, J. Ogawa, R. D. Schmid and S. Shimizu, Residue size at position 87 of cytochrome P450 BM-3 determines its stereoselectivity in propylbenzene and 3-chlorostyrene oxidation, FEBS Lett., 2001, 508, 249–252. J. D. Bloom, M. D. Dutia, B. D. Johnson, A. Wissner, M. G. Burns, E. E. Largis, J. A. Dolan and T. H. Claus, Disodium (R,R)-5-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl]-amino] propyl]1,3-benzodioxole-2,2-dicarboxylate (CL 316,243). A potent betaadrenergic agonist virtually specific for beta 3 receptors. A promising antidiabetic and antiobesity agent, J. Med. Chem., 1992, 35, 3081–4; O. Takahashi, J. Umezawa, K. Furuhashi and M. Takagi, Stereocontrol of a tertiary hydroxyl group via microbial epoxidation: A facile synthesis of prostaglandin o-chains, Tetrahedron Lett., 1989, 30, 1583–1584.

Metallomics, 2011, 3, 410–416

415

19 C. A. Martinez and J. D. Stewart, Cytochrome P450’s: potential catalysts for asymmetric olefin epoxidations, Curr. Org. Chem., 2000, 4, 263–282. 20 T. Omura and R. Sato, The carbon monoxide-binding pigment of liver microsomes II. Solubilization, purification, and properties, Journal of Biological Chemistry, 1964, 239, 2379–2385. 21 G. Jones, P. Willett and R. C. Glen, Molecular recognition of receptor sites using a genetic algorithm with a description of desolvation, J. Mol. Biol., 1995, 245, 43–53; G. Jones, P. Willett, R. C. Glen, A. R. Leach and R. Taylor, Development and validation of a genetic algorithm for flexible docking, J. Mol. Biol., 1997, 267, 727–748. 22 M. D. Eldridge, C. W. Murray, T. R. Auton, G. V. Paolini and R. P. Mee, Empirical scoring functions: I. The development of a fast empirical scoring function to estimate the binding affinity of ligands in receptor complexes, J. Comput.-Aided Mol. Des., 1997, 11, 425–445. 23 S. B. Kirton, C. W. Murray, M. L. Verdonk and R. D. Taylor, Prediction of binding modes for ligands in the cytochromes P450 and other heme-containing proteins, Proteins: Struct., Funct., Bioinf., 2005, 58, 836–844. 24 R. E. White and M. J. Coon, Heme ligand replacement reactions of cytochrome P-450. Characterization of the bonding atom of the axial ligand trans to thiolate as oxygen, Journal of Biological Chemistry, 1982, 257, 3073–3083; D. Kim, Z. L. Wu and

416

Metallomics, 2011, 3, 410–416

F. P. Guengerich, Analysis of coumarin 7-hydroxylation activity of cytochrome P450 2A6 using random mutagenesis, J. Biol. Chem., 2005, 280, 40319–40327. 25 S. Modi, M. J. Paine, M. J. Sutcliffe, L. Y. Lian, W. U. Primrose, C. R. Wolf and G. C. K. Roberts, A model for human cytochrome P450 2D6 based on homology modeling and NMR studies of substrate binding, Biochemistry, 1996, 35, 4540–4550; S. He, S. Modi, D. S. Bendall and J. C. Gray, The surface-exposed tyrosine residue Tyr83 of pea plastocyanin is involved in both binding and electron transfer reactions with cytochrome f, EMBO Journal, 1991, 10, 4011–4016. 26 S. Panke, B. Witholt, A. Schmid and M. G. Wubbolts, Towards a biocatalyst for (S)-styrene oxide production: characterization of the styrene degradation pathway of Pseudomonas sp. strain VLB120, Applied and Environmental Microbiology, 1998, 64, 2032–2043. 27 S. Eiben, L. Kaysser, S. Maurer, K. Ku¨hnel, V. B. Urlacher and R. D. Schmid, Preparative use of isolated CYP102 monooxygenasesA critical appraisal, J. Biotechnol., 2006, 124, 662–669; K. L. Tee and U. Schwaneberg, A screening system for the directed evolution of epoxygenases: Importance of position 184 in P450 BM3 for stereoselective styrene epoxidation, Angew. Chem., Int. Ed., 2006, 45, 5380–5383; E. T. Farinas, M. Alcalde and F. Arnold, Alkene epoxidation catalyzed by cytochrome P450 BM-3 139-3, Tetrahedron, 2004, 60, 525–528.

This journal is

c

The Royal Society of Chemistry 2011

Supplementary Material (ESI) for Metallomics This journal is © The Royal Society of Chemistry 2011

Control of the stereo-selectivity of styrene epoxidation by cytochrome P450 BM3 using structure-based mutagenesis Wei-Cheng Huang a, Paul M. Cullis b, Emma Lloyd Raven b, and Gordon C.K. Roberts a* 5

Received (in XXX, XXX) Xth XXXXXXXXX 200X, Accepted Xth XXXXXXXXX 200X First published on the web Xth XXXXXXXXX 200X DOI: 10.1039/b000000x

Supplemental figures 10

15

Figure S1 UV-Visible absorption spectra of purified P450 BM3 A82F-T438(V/I/L/F) mutants of ‘resting’ enzymes (red lines), enzymes reduced by sodium dithionite (green lines) and the reduced form CO-bound spectra (black lines). The CO-difference spectra were shown in insets for each mutant.

20

25

This journal is © The Royal Society of Chemistry [year]

Metallomics, [year], [vol], 00–00 | 1

Supplementary Material (ESI) for Metallomics This journal is © The Royal Society of Chemistry 2011 Styrene epoxidation by cytochrome P450 BM3 mutants

5

Figure S2 Phenotypes of E. coli JM109 cells containing P450 BM3 A82F-T438V, A82F-T438I, A82F-T438L, and A82F-T438F mutant genes streaked on a LB-Amp agar plate. Wild-type, A82F-T438V and A82F-T438I present a normal E. coli phenotype colour; on the other hand, A82F, A82F-T438L and A82F-T438F transform the phenotype of E. coli cell to blue colour, reflecting the formation of indigo. 10

2 | Metallomics, [year], [vol], 00–00

This journal is © The Royal Society of Chemistry [year]

Styrene epoxidation by cytochrome P450 BM3 mutants

Figure S3 Optical titration of styrene binding to (A) wild-type P450 BM3, (B) A82F mutant, and (C) A82F-T438V mutant.

5

10

Figure S4 GC chromatograms of the products of styrene epoxidation by P450 BM3 mutants. Samples were petroleum ether extracts of reaction mixtures containing 800 µM styrene, 300 µM NADPH and 1 µM P450 BM3 A82F-T438L mutant

This journal is © The Royal Society of Chemistry [year]

Metallomics, [year], [vol], 00–00 | 3