Selective oxidative demethylation of veratric acid to vanillic acid by CYP199A4 from Rhodopseudomonas palustris HaA2

Abstract

CYP199A4 (RPB3613) from Rhodopseudomonas palustris HaA2 is a hememonooxygenase that catalyzes the hydroxylation of para-substituted benzoic acids. Monooxygenase activity of CYP199A4 can be reconstituted in a Class I electron transfer chain with an associated [2Fe–2S] ferredoxin, HaPux, (RPB3614) and the flavin-dependent reductase, HaPuR, (RPB3656) that is not associated with a CYPgene. CYP199A4 and the ferredoxin HaPux are produced in greater quantities using recombinant Escherichia coli expression systems when compared to the equivalent proteins in the closely related CYP199A2–Pux–PuR Class I system from R. palustris CGA009. HaPuR and HaPux can also replace PuR and Pux in supporting the CYP199A2 enzyme turnover with high activity. Whole-cell in vivo substrate oxidation systems for CYP199A4 and CYP199A2 with HaPux and HaPuR as the electron transfer proteins have been constructed. These E. coli systems were capable of selectively demethylating veratric acid at the para position to produce vanillic acid at rates of up to 15.3 μM (g-cdw)⁻¹ min⁻¹ and yields of up to 1.2 g L⁻¹.

---

Introduction

Cytochrome P450 (CYP) enzymes are heme-dependent monooxygenases with a wide range of physiological roles, including hormone and secondary metabolitebiosynthesis, as well as xenobiotic metabolism.^1,2 P450 enzymes primarily catalyze the insertion of an oxygen atom from atmospheric dioxygen into carbon–hydrogen bonds of organic molecules but can also catalyze other reactions such as ring coupling, ring formation and contraction, and reductive reactions.³ Bacterial CYP enzymes have a wide range of functionality and their C–H bond oxidation activity may have applications in chemical synthesis.

Rhodopseudomonas palustris is a family of purple photosynthetic bacteria that are isolated in temperate, aquatic sediments and which possess extraordinary metabolic versatility. Bacteria classified as Rhodopseudomonas spp. can grow with or without light, or oxygen, fix nitrogen and have a wide range of biodegradation abilities.⁴ Studies have shown that different Rhodopseudomonas species share many characteristics but that each isolate has a unique set of genes for functions that take advantage of the microenvironment in which they are found.^5–7 The sequenced genome of R. palustris CGA009 revealed seven potential CYP enzyme genes.⁴ We recently reported the gene cloning, production and purification of the complete CYP199A2 Class I system P450 enzyme from R. palustris CGA009. CYP199A2 (RPA1871) was found to catalyze the oxidation of para-substituted benzoic acids⁸ using an electron transfer chain consisting of a [2Fe–2S] ferredoxin, palustrisredoxin (Pux, RPA1872), and a flavin-dependent ferredoxin reductase, palustrisredoxin reductase (PuR, RPA3782) with NADH as the source of electrons.^9,10 However the potential for using the CYP199A2 system in biotechnological processes is hampered by the relatively low heterologous production of the CYP199A2 and the Pux enzymes in E. coli.

Other R. palustris species contain similar genes to those of the CYP199A2 system in their genome . However, while the majority of the R. palustris strains isolated and analyzed so far have a gene that encodes a protein similar to the ferredoxin reductase PuR (e.g. strains TIE-1, BisB5, HaA2, BisB18 and BisA53),⁵ only certain strains contained genes which were analogous to CYP199A2 and the associated ferredoxin Pux (strains TIE-1 and HaA2). We have isolated the equivalent CYP enzyme, ferredoxin and ferredoxin reductase (CYP199A4, HaPux, HaPuR) from R. palustris HaA2. This Rhodopseudomonas species was isolated from uncontaminated freshwater marsh sediment at a site in Haren (Netherlands). The gene inventory of R. palustris HaA2 predicts that it has adapted for growth viaaerobic or microaerobic respiration in light rather than by anaerobic heterotrophic growth.^5,7,11,12 All three proteins of the CYP199A4 system were readily produced and isolated using E. coli as a heterologous host. Substrate binding, turnover activity and product formation studies were carried out with para-substituted benzoic acids. The electron transfer proteins HaPuR and HaPux could support high activity to CYP199A4 which was found to oxidatively demethylate veratric acid exclusively at the para position to produce vanillic acid. We have also constructed catalytically competent in vivo substrate oxidation systems for CYP199A4 and CYP199A2 enzymes using HaPux and HaPuR as the electron transfer chain. In shake flask experiments the CYP199A4–HaPux–HaPuR E. coli system was capable of oxidizing substrates efficiently at rates of 15.3 μM (g-cell dry weight)⁻¹ min⁻¹ and could produce up to 1.2 g L⁻¹ of vanillic acid from veratric acid in 6 hours.

Results and discussion

Heterologous production and purification of CYP199A4, HaPux and HaPuR

BLAST searches of the R. palustris HaA2 genome using the amino acid sequences of CYP199A2, Pux and PuR from the Class I CYP199A2 system from R. palustris CGA009 identified three equivalent genes (see ESI† , Table S1 and Fig. S1–S3). These genes encoded the proteins CYP199A4 (RPB3613), HaPux (RPB3614) and HaPuR (RPB3656). The CYP199A4, HaPux and HaPuRgenes were amplified by PCR, cloned and the encoded proteins produced in E. coli in their soluble forms (see Experimental).

Ferredoxin–NAD(P) reductases (FNRs) can be grouped into two families: plant-type and glutathione reductase-type.¹³Gene analysis revealed that HaPuR belongs to the oxygenase-coupled NADH-dependent ferredoxin reductase (ONFR) class of the glutathione reductase-type FNRs. The gene sequence of HaPuR possesses conserved motifs responsible for the binding of FAD (GXGXA, residues 9–13) and the binding of NADH (GXGXXG, residues 150–155, Fig. S1, ESI† ). The HaPuR protein was over-produced in soluble form in E. coli and purified in good yields (>30 mg L⁻¹ of shake flask culture). The observed mass of HaPuR was 43 543 ± 2.0 Da (calculated 43 532.1 Da) consistent with a protein of 404 amino acid residues (Fig. S4a, ESI† ). A second mass was observed at 43 674.0 ± 2.0 Da which corresponds to incomplete cleavage of the N-terminal methionine (calculated 43 663.2 Da, Fig. S4, ESI† ). The HaPuR protein had absorption maxima at 272, 383, 455 and 482 nm (Fig. 1a). On denaturation and removal of the protein a spectrum consistent with that of an FADcofactor was observed. The FAD content and spectral properties of HaPuR, PuR and PdR were found to be almost identical (PuR had maxima at 272, 383, 455 and 482 nm and PdR had maxima at 272, 380, 454 and 485 nm). The HaPuR concentration was calculated using ε₄₅₅ = 10.0 mM⁻¹ cm⁻¹ (Table S2, ESI† ).

Fig. 1 (a) Electronic spectra of HaPux and HaPuR. (b) Electronic spectra of CYP199A4; substrate free and the reduced CO form, (c) electronic spectra of CYP199A4; substrate free, 4-methoxybenzoic acid and veratric acid bound forms.

The HaPux gene contains sequence motifs that indicate it is a [2Fe–2S] ferredoxin. Such ferredoxins can be classified into three distinct classes (plant-type, bacterial-type, and vertebrate-type) based on the amino acid sequence relationships, the arrangement of the conserved cysteine residues and geometry of the [2Fe–2S] cluster. HaPux can be identified as belonging to the vertebrate family from a three-element fingerprint: motif 1 includes the region with the three conserved cysteines, which are ligands of the [2Fe–2S] center (residues Cys39-His48); motif 2 contains a cluster of negatively charged residues (Ala58-Asp72) and motif 3 includes the sequence around the fourth conserved Cys ligand of the [2Fe–2S] cluster (Ser81-Ile89, Fig. S2, ESI† ).^14,15

The HaPux protein was produced in soluble form in E. coli and purified in good yields (>10 mg L⁻¹ of shake flask culture). The electrospray mass spectrum of native HaPux revealed a relative molecular mass of 11 298.0 ± 2.0 Da (calculated 11 296.5 Da, Fig. S4b, ESI† ), consistent with a mature protein with 105 amino acid residues. The absorption spectra of HaPux had maxima at 280, 330, 416 and 457 nm (Fig. 1a) which are similar to the [2Fe–2S] ferredoxins Pux and Pdx. The iron content of the ferredoxin was determined by acid denaturation of the protein, followed by assaying the concentration of the iron–bathophenanthroline disulfonate complex. The extinction coefficients for native HaPux were calculated to be ε₃₃₀ = 14.1 mM⁻¹ cm⁻¹, ε₄₁₆ = 11.2 mM⁻¹ cm⁻¹ and ε₄₅₇ = 9.1 mM⁻¹ cm⁻¹ (Table S2, ESI† ).

The CYP199A4 protein was produced in soluble form and purified in good yields (>20 mg L⁻¹ of shake flask culture). The observed mass of CYP199A4 was 44 404.0 ± 2.0 Da (calculated 44 405.8 Da) consistent with a protein of 409 amino acids (Fig. S4c, ESI† ). CYP199A4 was isolated in the low-spin ferric form with a Soret maximum at 418 nm and showed the characteristic 449 nm absorption for the Fe(II)CO complex with no evidence of the inactive P420 form (Fig. 1b). The extinction coefficients were calculated to be ε₄₁₉ = 103.1 mM⁻¹ cm⁻¹ for the ferric resting state (Table S2, ESI† ). The addition of 4-methoxybenzoic acid resulted in a >95% change in the Soret band from 419 nm to 394 nm (Fig. 1c).

Overall the three proteins isolated from R. palustris HaA2 could be a Class I P450 electron transfer system. The HaPux and the CYP199A4 enzymes from R. palustris HaA2 were produced in higher yields in E. coli than Pux and CYP199A2 from R. palustris CGA009. The isolated yield of HaPux was higher (>5-fold) than that of Pux using identical conditions and the same vector–host combination.

Electron transfer activity of the CYP199A4 system

The small molecule electron acceptors2,6-dichloroindophenol (DCIP) and Fe(CN)₆³⁻ were used to examine the oxidation of NADH and NADPH by HaPuR. Under steady-state turnover conditions the rates of NADH oxidation were 40–110 times higher than those with NADPH (Table 1). The preference of HaPuR for NADH over NADPH mirrored that found for PuR.

Table 1 The activity of the ferredoxin reductases from R. palustris HaA2 (HaPuR) and R. palustris CGA009 (PuR) in mediating the oxidation of NADH and NADPH by ferricyanide and dichloroindophenol (DCIP) in 50 mM Tris, pH 7.4. Rates are in nmol (nmol FdR)⁻¹ s⁻¹. The data are given as mean ± S.D. with n ≥ 3

Ferredoxin reductase	Electron acceptor
	Fe(CN)6^3−		DCIP
	NADH	NADPH	NADH	NADPH
HaPuR	1110 ± 18	25 ± 0.5	228 ± 7.6	1.7 ± 0.2
PuR	1400 ± 20	32 ± 3.6	199 ± 9.3	2.6 ± 0.2

---

The kinetic parameters of electron transfer from the ferredoxin reductase (FdR) to the ferredoxin (Fdx) were investigated using cyt c as the terminal electron acceptor. Electron transfer from Fdx to cyt c is fast while FdR reduces cyt c slowly.^16,17 HaPuR was also found to transfer electrons slowly to cyt c, while cyt c reduction occurred within the time of mixing reduced HaPux with oxidized cyt c and measuring the absorbance at 550 nm (data not shown). Hence by titrating increasing concentrations of the ferredoxin, K_m and k_cat values were determined for the reduction of HaPux by HaPuR and PuR (Table 2, Fig. 2). HaPuR–HaPux binding was tighter than PuR–Pux (K_m: 2.7 ± 0.12 vs. 4.2 ± 0.3 μM). PuR could also support HaPux mediated cyt c reduction with a similar K_m and k_cat to HaPuR. The rate of HaPuR reduction of HaPux (k_cat: 243 ± 3.3 s⁻¹, Fig. 2) is similar to that of PuR reduction of Pux (k_cat: 262 ± 6.0 s⁻¹).

Fig. 2 Michaelis–Menten analysis of the kinetics of electron transfer (a) from ferredoxins to cytochrome P450, and (b) from ferredoxin reductases to ferredoxins using equine cytochrome c as the terminal electron acceptor. PuR: palustrisredoxin reductase R. palustris CGA009, Pux: palustrisredoxin R. palustris CGA009, HaPuR: palustrisredoxin reductase R. palustris HaA2, HaPux: palustrisredoxin R. palustris HaA2.

Table 2 Michaelis–Menten parameters (50 mM Tris, pH 7.4) for (a) reduction of ferredoxins, HaPux and Pux, by the ferredoxin reductases, HaPuR and PuR, from R. palustris HaA2 and R. palustris CGA009 using cytochrome c as the terminal electron acceptor, and (b) oxidation of ferredoxins, HaPux and Pux, by CYP199A2 and CYP199A4. 4-Methoxybenzoic acid was used as substrate under conditions where the first electron transfer from Fdx to CYP is the slow step. All data are given as mean ± S.D. with n ≥ 3

Electron transfer chain	50 mM Tris, pH 7.4
	k cat/s^−1	K m/μM
(a) FdR–Fdx–Cyt c
HaPuR–HaPux	243 ± 3.3	2.7 ± 0.12
PuR–HaPux	226 ± 3.9	2.0 ± 0.13
PuR–Pux	262 ± 6.0	4.2 ± 0.3
(b) FdR–Fdx–CYP
HaPuR–HaPux–CYP199A4	42.6 ± 0.80	1.41 ± 0.10
HaPuR–HaPux–CYP199A2	48.4 ± 1.9	0.53 ± 0.10
PuR–Pux–CYP199A2	37.9 ± 0.80	0.45 ± 0.04

---

The first electron transfer from HaPux to CYP199A4 was studied under conditions where this step is the slowest in the catalytic cycle, using 4-methoxybenzoic acid as substrate. Electron transfer from HaPux to CYP199A4 was faster (k_cat: 42.6 s⁻¹) than from Pux to CYP199A2 (k_cat: 37.9 s⁻¹) but the proteins were less tightly bound (K_m: 1.41 μM vs. K_m: 0.45 μM). HaPux binding to CYP199A2 (K_m: 0.53 μM) was stronger than to CYP199A4 and similar to Pux binding to CYP199A2. Electron transfer from HaPux to CYP199A2 (k_cat: 48.4 ± 1.9 s⁻¹) was faster than that for HaPux to CYP199A4.

The steady-state assays using a ratio of 1 : 10 : 1 of FdR–Fdx–CYP showed that the CYP199A4 system was more active by ∼50% than the CYP199A2 system (Table 3). The NADH turnover rate observed under these conditions, with 4-methoxybenzoic acid as the substrate, was similar to the k_cat observed under saturating ferredoxin conditions (36.3 nmol (nmol (nmol CYP)⁻¹ s⁻¹).

Table 3 Substrate binding and steady-state catalytic turnover parameters for 4-methoxybenzoic acid, 4-ethylbenzoic acid and veratric acid, with the CYP199A4 system from R. palustris HaA2 and the CYP199A2 system from R. palustris CGA009. The reaction mixtures (50 mM Tris, pH 7.4) contained 0.5 μM P450, 5 μM Fdx and 0.5 μM FdR. Coupling is the percentage of NADH consumed in the reaction that led to the formation of products. All data are given as mean ± S.D. with n ≥ 3. Rates are given in nmol (nmol CYP)⁻¹ min⁻¹. The NADH turnover rates of HaPuR–HaPux–CYP199A2 and HaPuR–Pux–CYP199A2 are 1670 ± 45 and 2250 ± 10 nmol (nmol CYP)⁻¹ min⁻¹, respectively

P450 system	Spin-state shift	K d/μM	NADH turnover rate	Product formation rate	Coupling (%)
HaPuR–HaPux–CYP199A4
4-Methoxybenzoic acid	≥95%	0.034 ± 0.003	2176 ± 57	1971 ± 47	91 ± 2
4-Ethylbenzoic acid	≥95%	0.336 ± 0.016	1177 ± 34	1026 ± 52	88 ± 2
Veratric acid	70%	29.5 ± 3.1	1101 ± 59	1061 ± 88	96 ± 3
PuR–Pux–CYP199A2
4-Methoxybenzoic acid	≥95%	0.085 ± 0.005	1438 ± 49	1375 ± 55	97 ± 2
4-Ethylbenzoic acid	≥95%	0.524 ± 0.043	763 ± 11	620 ± 34	81 ± 2
Veratric acid	70%	36.2 ± 2.3	n.d.	n.d.	n.d.

---

The HaPuR enzyme could also support electron transfer to Pux–CYP199A2 with high activity. HaPux could also replace Pux in the CYP199A2 system, suggesting that the molecular interactions involved in electron transfer from NADH to the [2Fe–2S] ferredoxin and subsequently to the P450 enzyme in these two systems are similar.

Substrate binding and product formation

Both 4-methoxy- and 4-ethylbenzoic acid induced a ≥95% spin-state shift with CYP199A4 (Table 3). The dissociation constants, K_d, were obtained by measuring the maximum difference in absorbance against concentration of substrate and fitting to a hyperobolic equation or the tight binding quadratic equation (Fig. 3).

Fig. 3 Example of (a) quadratic fit for tight binding to CYP199A4 (4-methoxybenzoic acid) and (b) hyperbolic fit for weaker binding (veratric acid). The enzyme concentration was 0.2 μM in 50 mM Tris, pH 7.4, at 30 °C. Each point is in triplicate and error bars show the standard deviations.

The binding constant of 4-methoxybenzoic acid was almost 10-fold tighter than that of 4-ethylbenzoic acid (e.g. 0.034 ± 0.003 vs. 0.336 ± 0.016 μM for CYP199A4). Veratric acid, which is related to 4-methoxybenzoic acid but has an additional methoxy group at the 3-position, gave a spin-state shift of 70% (Fig. 1c) and the binding was considerably weaker (29.5 ± 3.1 μM, Fig. 3b). This agrees with previous data showing that 3-substituted benzoic acids bind less tightly than 4-substituted analogues.⁸ This could arise from steric clashes between the substituent in the 3-position and the amino acids which line the protein active site. Substrate binding was tighter for CYP199A4 than for CYP199A2 for all substrates measured (e.g. 0.034 ± 0.003 vs. 0.085 ± 0.005 μM for 4-methoxybenzoic acid).

CYP199A4 oxidation of 4-methoxybenzoic acid resulted in a single product, 4-hydroxybenzoic acid, presumably viaoxidation of a methyl C–H bond followed by spontaneous decomposition of the hemi-acetal (Scheme 1).⁹ Calibration of the GC detector response showed that the enzyme is highly efficient in coupling NADH consumption to product generation (96%), with a product formation rate of 1971 ± 47 nmol (nmol CYP)⁻¹ min⁻¹ (Table 3). The two major products from 4-ethylbenzoic acidoxidation were 4-vinylbenzoic acid (38%) formed by C_α–C_β bond desaturation, and 4-(1-hydroxyethyl)-benzoic acid (51%) from hydroxylation of the C_α benzylic carbon. This is similar to the product distribution reported previously with CYP199A2.^9,10 The minor products, 4-epoxyethylbenzoic acid (4%) and 4-acetylbenzoic acid (4%), most likely arose from further oxidation of the two major products and one assigned as 4-(2-hydroxyethyl)-benzoic acid (3%) was also observed. The total product formation rate was 1026 ± 52 nmol (nmol CYP)⁻¹ min⁻¹ with high efficiency (88%) of NADH utilization for product formation (Scheme 1). Veratric acid was efficiently demethylated solely at the 4-position to yield vanillic acid by both CYP199A4 and CYP199A2 (CYP199A4 product formation rate 1061 ± 88 nmol (nmol CYP)⁻¹ min⁻¹; Table 3). CYP199A4 has no significant activity with vanillic acid when compared to veratric acid (Scheme 1). This is advantageous as demethylase enzymes are known to remove both the methyl groups from veratric acid to form protocatechuic acid.^18,19 These CYP199 systems could potentially provide an alternative route to vanillic acid which can in turn be reduced to vanillin in high yield by enzymatic and whole-cell biotransformations .^20,21

Scheme 1 Oxidation of para-substituted benzoic acids by CYP199A4. (1) 4-Methoxybenzoic acid, (2) 4-ethylbenzoic acid, (3) veratric acid, (4) 4-hydroxybenzoic acid, (5) 4-(1-hydroxyethyl)-benzoic acid, (6) 4-vinyl benzoic acid, (7) vanillic acid.

Whole-cell substrate oxidation systems

The feasibility of using P450 enzymes for biotransformations depends on maximizing the catalytic lifetime whilst removing the need to use expensive biological cofactors. In vitrocofactor regeneration or replacement of the natural cofactors using chemical or electrochemical methods have all been attempted with a variety of P450 enzymes, with different degrees of success.^22–24 Alternatively all the proteins in a P450 electron transfer chain can be expressed in a single heterologous host to form a catalytically competent in vivo substrate oxidation system.^25,26 To this end we constructed HaPux–HaPux–CYP199A2 and HaPux–HaPux–CYP199A4 expression systems using commercially available vectors. We used the HaPuR and HaPux electron transfer proteins for the CYP199A2 enzyme as they were capable of supporting efficient activity with this enzyme and the HaPux ferredoxin was produced in much higher quantities in E. coli than was Pux. Two vectors (pRSFDuet-1 and pETDuet-1) were used to allow the expression of four genes (HaPuR, HaPux on pETDuet-1; HaPux and the CYP199A4 or CYP199A2 genes on pRSFDuet-1). Two copies of the HaPux gene were used so as to give an excess of the ferredoxin in the E. coli cells in order to maximize the CYP activity (Fig. S5, ESI† ). The pETDuet-HaPux-HaPuR and pRSFDuet-HaPux-CYP199A4(2) plasmids were transformed into an E. coli host and the cells were grown and proteins produced as described in the experimental section. Both CYP199A4 and CYP199A2 systems produced 5–6 g L⁻¹ of a dark red/brown cell pellet indicating successful production of the ferredoxin and P450 components of the system. The P450 concentration was quantitated by measuring the ferrous-CO spectra of the cell lysate. CYP199A4 was produced in greater quantities compared to CYP199A2 (31.8 mg L⁻¹vs. 9.8 mg L⁻¹). The expression levels of the HaPuR and HaPux components are more difficult to measure but using pETDuet-HaPux-HaPuR, the pRSFDuet-HaPux plasmids, an estimate of the HaPux concentration of ∼3 μM was obtained which is more than double the concentration of the P450 component in the cell (∼0.7 μM). The harvested cell pellets were resuspended in an equal volume of E. coli minimal media (EMM) and up to 4 mM substrate (4-methoxybenzoic acid and veratric acid) were added. Samples were removed periodically and monitored for substrate consumption and product formation. When all the substrate had been consumed additional 4 mM aliquots of substrate were added and the reaction mixtures were left for up to 20 h.

The whole-cell CYP199A4–HaPux–HaPuR system was found to be approximately five times more active than the CYP199A2–HaPux–HaPuR system under identical conditions. Both were significantly more active (>10 times) than whole-cell oxidation systems constructed by replacing HaPux with Pux (data not shown). This was presumably due to the lower levels of production of CYP199A2 and Pux in E. coli compared to CYP199A4 and HaPux. A 500 mL suspension of 2.5–3 g wet weight of cells containing the CYP199A4–HaPux–HaPuR system was found to consume 4 mM of 4-methoxybenzoic acid and veratric acid completely within 2 h. Time-course analysis by GC showed that the rates of 4-hydroybenzoic acid and vanillic acid formation were 15.2 and 15.4 μM (g-cdw)⁻¹ min⁻¹, respectively (Fig. 4). The addition of further aliquots of 4 mM substrate every 2 h resulted in the formation of up to 7.3 mM 4-hydroxybenzoic acid and 7.4 mM vanillic acid after 5 h (3 experiments carried out for each substrate). This corresponds to 10⁴ total turnovers and ∼1.2 g of vanillic acid and ∼1 g of 4-hydroxybenzoic acid per litre of culture. When the whole-cell oxidation growths were left overnight with an extra 20 mM substrate added there was no significant increase in the amount of product formation suggesting that substrate or product toxicity had terminated the reactions under these conditions.²⁷

Fig. 4 Time-course analysis of veratric acidoxidation to vanillic acid using the pRSFDuet-HaPux-CYP199A4 and pETDuet-HaPux-HaPuR whole-cell oxidation system. The quantity of the veratric acid substrate increases as further aliquots are added at 60, 120 and 240 min.

These whole-cell oxidation systems can oxidatively demethylate both 4-methoxybenzoic acid and veratric acid (exclusively at the para position) to efficiently produce 4-hydroxybenzoic acid and vanillic acid, respectively. Both products are formed at very similar rates despite the faster in vitro conversion of 4-methoxybenzoic acid with the CYP199A4 system. This suggests that the activity of the enzyme may not be the rate-determining step in the in vivo turnovers. The transport of the substrate into the E. coli cell across the membrane or the regeneration of cellular NADH may be possible limiting factors in rate of the reaction.^28,29 The rapid and selective conversion of veratric acid to vanillic acid in this system under non-optimized conditions is encouraging and may provide an alternative route to vanillin.

Experimental

General

Enzymes for molecular biology were from New England Biolabs, UK. KOD polymerase (Merck Biosciences, UK) was used for the PCR steps. General reagents were from Sigma-Aldrich or Merck, UK. NADH was from Roche Diagnostics, UK, and isopropyl-β-D-thiogalactopyranoside (IPTG), dithiothreitol (DTT), buffer components and growth media were from Melford Laboratories, UK. K₃Fe(CN)₆, 2,6-dichloroindophenol (DCIP), 4-ethylbenzoic acid, 4-methoxybenzoic acid, veratric acid, vanillic acid, N,O-bis-(trimethylsilyl)trifluoroacetamide with trimethylchlorosilane (BSTFA + TMCS, 99 : 1), and equine cytochrome c were from Sigma-Aldrich. The genomic DNA of R. palustris HaA2 was obtained from Prof. Caroline Harwood (University of Washington, Seattle).

General DNA and microbiological experiments were carried out by standard methods.³⁰Proteins were stored at −20 °C in 50 mM Tris, pH 7.4, containing 50% v/v glycerol. Glycerol was removed immediately before use by gel filtration on a 5 mL PD-10 column (GE Healthcare, UK) by eluting with 50 mM Tris, pH 7.4. UV/Vis spectra and spectroscopic activity assays were recorded at 30 ± 0.5 °C on a Varian CARY-50 or 1E spectrophotometer. Gas chromatography analysis was performed on a ThermoFinnegan TRACE instrument equipped with a CP-SIL 8CB fused silica column (15 m × 0.32 mm; Varian) using helium as the carrier gas and flame ionization detection. Both the injector and detector were held at 250 °C. Electrospray proteinmass spectrometry (ES-MS) was carried out on a Micromass Platform II instrument.

Enzymes and molecular biology

The genes encoding HaPuR and CYP199A4 were amplified by PCR methods via 25 cycles of strand separation at 95 °C for 1 min followed by annealing at 50 °C and extension at 68 °C for 1 min 40 s + 2 s per cycle. The gene encoding HaPux was amplified via a similar method with an extension time of 1 min + 1 s per cycle. The oligonucleotides used were: CYP199A4 5′-ttaattccatggtcagcaatagctccgcggag-3′ (Nco I) and 5′- ttaattaagcttttattaggcaggagtcagcttgacc-3′ (Hind III); HaPuR 5′-ttaattcatatgaacgacacggtcttgattg-3′ (Nde I) and 5′-ttaattaagcttttattacgccatcgccttctttaggtcg-3′ (Hind III); HaPux 5′-ttaattcatatgccgagtatcaccttcatc-3′ (Nde I) and 5′-ttaattaagcttttattaggtctgtcgttcgggcagg-3′. The restriction sites are in italics and a modified base (a → g) in the N-terminal primer of CYP199A4 to allow the incorporation of the Nco I restriction site is highlighted in bold and italics. This changes the N-terminal residue from isoleucine to valine (atc → gtc).

The amplified genes were cloned into the expression vectors pET26a, pET28a (Merck Biosciences) or pCWOri+ using the Nde I or Nco I and Hind III restriction sites. All the amplified genes were fully sequenced on an ABI 377XL Prism DNA sequencer by the Geneservice DNA sequencing facility at the Department of Biochemistry, University of Oxford.

Recombinant protein production and purification

CYP199A4. The RPB3613gene was over-expressed and the encoded CYP199A4 protein over-produced using the pET28a vector in E. coli BL21(DE3). A single colony was inoculated into 200 mL Luria–Bertani broth containing 30 mg L⁻¹kanamycin (LB_kan) and grown at 37 °C and 200 rpm overnight. This culture was then inoculated at a ratio of 10 mL L⁻¹ into 2 × 1 L of 2YT_kan medium and grown at 37 °C and 200 rpm for 8 h to OD₆₀₀ ≈ 1.0. Recombinant protein production was induced by the addition of 0.1 mM IPTG (from a 1 M stock in water). The incubator temperature was lowered to 25 °C and the shaker speed reduced to 120 rpm and the culture grown for a further 16 h. The red cell pellet harvested from the culture by centrifugation was resuspended in 200 mL buffer T (50 mM Tris, pH 7.4, 1 mM DTT) and the cells were lysed by sonication. Cell debris was removed by centrifugation at 37 000 g for 30 min at 4 °C and the protein was fractionated from the supernatant using ammonium sulfateprecipitation (25%–50% fraction). The protein was redissolved in buffer T and desalted using a Sephadex G-25 column (200 mm × 40 mm) that had been pre-equilibrated with the same buffer. The protein was loaded onto a Fast Flow DEAE Sepharose column (40 mm × 200 mm; GE Healthcare) and eluted using a linear salt gradient of KCl (0–250 mM) in buffer T developed over 8 column bed-volumes at a flow rate of 8 mL min⁻¹. The red colored fractions containing CYP199A4 were pooled and concentrated using ultrafiltration. The protein was then loaded onto a Source Q column (120 mm × 26 mm; GE Healthcare), and eluted using a linear gradient of 0–100 mM KCl in buffer T. All fractions with A₄₁₉/A₂₈₀ > 2.0 were combined and concentrated by ultrafiltration. Protein concentrations were estimated using ε_448–491 = 91.0 mM⁻¹ cm⁻¹ for the CO difference spectra.³¹ Using these values ε₄₁₉ was determined to be 103.1 mM⁻¹ cm⁻¹ for the ferric resting state.

HaPuR. The HaPuR protein was produced using E. coli DH5α transformed with a pCWori+ plasmid containing the RPB3656gene and purified using methods similar to those used for PdR and PuR.^32,33 A single colony was inoculated into 200 mL Luria–Bertani broth (LB) containing 100 μg mL⁻¹carbenicillin (LB_carb) and grown at 37 °C and 200 rpm overnight. 15 mL of this culture were then inoculated into 6 × 1 L of 2YT_carb medium and grown at 37 °C and 200 rpm until OD₆₀₀ ≈ 1. Recombinant protein production was induced by the addition of 1 mM IPTG (from a 1 M stock in water) and the culture was then grown for another 6 h. The yellow cell pellet harvested by centrifugation (5000 g, 5 min, 4 °C) was resuspended in 200 mL buffer T and the cells were lysed by sonication. Cell debris was removed by centrifugation at 37 000 g for 30 min at 4 °C. The supernatant was loaded onto a Fast Flow DEAE Sepharose column equilibrated with buffer T. The protein was eluted using a linear salt gradient of KCl (50–200 mM) in buffer T developed over 8 column bed-volumes at a flow rate of 8 mL min⁻¹. The yellow protein-containing fractions were collected and desalted by successive concentration and dilution using ultrafiltration. The protein was then loaded onto a Source Q column and eluted using a linear gradient of 50–200 mM KCl in buffer T. Purified fractions (A₂₈₀/A₄₅₄ > 6.0) were combined and concentrated by ultrafiltration.

The flavin content was determined by boiling HaPuR for 10 min in the dark. The denatured protein was removed by centrifugation (21 000 g for 3 min). The electronic spectrum of the released flavincofactor was recorded and the concentration calculated using an extinction coefficient of 9.2 mM⁻¹ cm⁻¹ at 473 nm.³⁴ The method was validated using PdR as a control.

HaPux. HaPux was produced in E. coli BL21(DE3) harboring a pET26a plasmid containing the RPB3614gene. The protein was purified using methods similar to that used for Pux and Pux-B from R. palustris CGA009.^8,9 A single colony was inoculated into 200 mL LB_kan containing 30 mg L⁻¹kanamycin and grown at 37 °C and 200 rpm overnight. This culture was then inoculated at a ratio of 25 mL L⁻¹ into 6 × 1 L of 2YT_kan medium, and grown at 37 °C and 200 rpm for 4–6 h to OD₆₀₀ ≈ 1. Recombinant protein production was induced by the addition of 0.1 mM IPTG (from a 1 M stock in water). The incubator temperature was lowered to 25 °C and the shaker speed reduced to 120 rpm and the culture was then grown for at least another 6 h. The brown cell pellet harvested from the culture by centrifugation was resuspended in 200 mL buffer T2 (10 mM Tris, pH 7.4, 20% v/v glycerol, 1% v/v Triton X-100, 1 mM DTT and 1% v/v β-mercaptoethanol) and the cells were lysed by sonication. Cell debris was removed by centrifugation at 37 000 g for 30 min at 4 °C. The supernatant was loaded onto a DEAE Fastflow Sepharose column (200 mm × 50 mm; GE Healthcare) and the protein eluted using a linear salt gradient of KCl (100–400 mM) in buffer T developed over eight column bed-volumes at a flow rate of 8 mL min⁻¹. The brown protein-containing fractions were collected and concentrated by ultrafiltration and then desalted using a Sephadex G-25 column (200 mm × 40 mm) that had been pre-equilibrated with buffer T. The final purification step was anion-exchange chromatography using a Source-Q column (120 mm × 26 mm; GE Healthcare), and the HaPux protein was eluted using a linear gradient of 75–300 mM KCl in buffer T developed over 20 column bed-volumes at a flow rate of 8 mL min⁻¹. Purified fractions (A₃₂₅/A₂₈₀ > 0.65) were combined and concentrated by ultrafiltration. An equal volume of glycerol was added, the solution filtered through a 0.22 μm sterile syringe filter and stored at −20 °C. Aliquots of this frozen solution were further purified viagel filtration on a G50 Sephadex column (1000 mm × 10 mm) using Buffer T.

The ferredoxin iron content and hence extinction coefficients were determined by denaturing the proteins with 0.3 volumes of 12 M HCl and heating at 100 °C for 15 minutes. Distilled water (400 μL) was then added to the sample. Precipitated material was removed by centrifugation at 21 000 g at ambient temperature for 4 min. Aliquots (50–200 μL) of the supernatant were then removed and each was made up to 1.5 mL with 0.5 M Tris buffer (pH 8.5). The iron was reduced by addition of 100 μL of sodium ascorbate solution (5% w/v in H₂O) and finally 400 μL of a 0.1% w/v bathophenanthroline disulfonate solution in H₂O were added and the mixture was left for at least 1 h. The iron content was assayed by UV/Vis spectroscopy against a blank containing no iron, and the iron was quantitated using ε₅₃₅ = 22.14 mM⁻¹ cm⁻¹.^35,36 The iron content was obtained from a standard curve using FeSO₄ solutions and the method was validated using native Pdx as a control. The HaPux concentration was calculated using ε₄₁₆ = 11.2 mM⁻¹ cm⁻¹.

The expression and purification of CYP199A2, Pux, and PuR have been described elsewhere.^8,9

Substrate binding titrations

The protein was diluted to 0.2 μM using 50 mM Tris, pH 7.4, in 2.5 mL and 0.5–2 μL of substrate were added using a Hamilton syringe from either a 1, 10 or 100 mM stock solution in ethanol. The sample was mixed and the peak-to-trough difference in absorption recorded on the UV spectrophotometer between 700 nm and 250 nm. Further aliquots of substrate were added until the peak-to-trough difference did not shift further. The dissociation constants, K_d, were obtained by fitting the difference in absorbance against concentration of substrate to the hyperbolic equation (eqn (1)):where ΔA is the peak-to-trough absorbance difference, ΔA_max is the maximum absorbance difference, [S] is the substrate concentration. Several substrates exhibited extremely tight binding of K_d < 1 μM. In these instances the data were fitted to the tight binding quadratic equation (eqn (2)):³⁷where ΔA is the peak-to-trough absorbance difference, ΔA_max is the maximum absorbance difference, [S] is the substrate concentration and [E] is the enzyme concentration.

Activity assays

The ability of HaPuR to mediate NADH and NADPHoxidation was compared by assaying electron transfer to the small molecule electron acceptorsFe(CN)₆³⁻ and DCIP. Incubation mixtures (1.2 mL) contained 50 mM Tris, pH 7.4, 10 nM HaPuR and 1 mM Fe(CN)₆³⁻ or 100 μM DCIP. NADH or NADPH (300–500 μM) was added and the absorbance was monitored at 420 nm for Fe(CN)₆³⁻ (ε₄₂₀ = 1.04 mM⁻¹ cm⁻¹) and 600 nm for DCIP (ε₆₀₀ = 21.0 mM⁻¹ cm⁻¹).

NADH turnover rate assays were performed with mixtures (1.2 mL) containing 50 mM Tris, pH 7.4, 0.5 μM CYP, 5 μM ferredoxin (Fdx), 0.5 μM ferredoxin reductase (FdR) and 100 μg mL⁻¹ bovine liver catalase (a 1 : 10 : 1 ratio of CYP : Fdx : FdR which should be in a reigime close to k_cat). The mixtures were oxygenated and then equilibrated at 30 °C for 2 min. Substrates were added as a 100 mM stock solution in ethanol to a final concentration of 1 mM. NADH was added to ca. 320 μM (A₃₄₀ = 2.00) and the absorbance at 340 nm was monitored. The rate of NADH turnover was calculated using ε₃₄₀ = 6.22 mM⁻¹ cm⁻¹ over the linear part of the data (Fig. S6, ESI† ). To determine the K_m and k_cat for electron transfer from ferredoxins to CYP enzymes the assays were carried out as above but with a lower CYP concentration (0.1 μM) and the Fdx concentration was varied from 1 to 40 μM.

Steady-state cytochrome c (cyt c) reduction assays to determine the kinetic parameters for electron transfer reduction of Fdx by FdR were performed with mixtures (1.2 mL final assay volume) containing 50 mM Tris, pH 7.4, 1 nM FdR and a range of Fdx concentrations from 1 to 40 μM. The mixtures were equilibrated at 30 °C for 2 min. Cyt c was added as a 20 mg mL⁻¹ stock solution in 50 mM Tris, pH 7.4, to a final concentration of 30 μM. NADH was added to ca. 100 μM and the absorbance at 550 nm was monitored. The initial rate of cyt c reduction was calculated using ε₅₅₀ = 22.1 mM⁻¹ cm⁻¹.

K _m and k_cat values for the FdR–Fdx and Fdx–CYP interactions were obtained by fitting the initial rate of cyt c reduction or NADH oxidation, respectively, against the Fdx concentration to a hyperbolic function using the Origin 8 software (Origin Labs).

Analysis of metabolites

After the NADH had been consumed in substrate oxidation incubations, 990 μL of the reaction mixture were mixed with 10 μL of internal standard solution (25 mM 9-hydroxyfluorene in ethanol) and 2 μL of concentrated HCl. The mixture was extracted three times with 400 μL of ethyl acetate and the organic extracts were combined and dried over MgSO₄. Solvent was evaporated under a stream of dinitrogen and the sample dissolved in 200 μL acetonitrile. Excess (25 μL) BSTFA + TMCS (99 : 1) was added and the mixture left for at least 120 min to produce the trimethylsilyl ester of the carboxylic acid group and trimethylsilyl ether of the alcohol, if formed. The reaction mixtures were used directly for GC analysis. The oven temperature was held at 100 °C for 1 min and then increased at 15 °C min⁻¹ up to 220 °C. The retention times for the trimethylsilyl derivatives were 5.55 min for 4-ethylbenzoic acid, 5.70 min for 4-vinylbenzoic acid, 6.18 min for 4-methoxybenzoic acid, 7.00 min for 4-hydroxybenzoic acid, 7.37 min for 4-(1-hydroxyethyl)-benzoic acid, 7.57 min for veratric acid and 7.96 min for vanillic acid. The internal standard eluted at 8.25 min. Products were calibrated against authentic samples of 4-vinylbenzoic acid, 4-hydroxybenzoic acid and vanillic acid.⁹

Construction of the in vivo systems

The HaPuxgene was re-amplified using the same method as above using a different C-terminal oligonucleotide to incorporate a Kpn I restriction site (5′-ttaattggtaccctattaggtctgtcgttcgggcaggcg-3′; Kpn I site italicized). The amplified gene was cloned into the pETDuet-1 and pRSFDuet-1 vectors using the Nde I and Kpn I restriction sites to yield the pETDuet-HaPux and pRSFDuet-HaPux plasmids, respectively.

The HaPuRgene was amplified using the same protocol as above but an Nco I restriction site was introduced at the 5′ end using the following oligonucleotide; 5′- ttaattccatggccgacacggtcttgattgctggag-3′. The three bases of the first codon have been converted from aac → gcc to allow the incorporation of the Nco I site at the N-terminal this changes the first amino acid from Asn → Ala. The amplified HaPuRgene was incorporated into the pET-Duet-HaPux vector using the Nco I and Hind III restriction sites to yield the plasmid pETDuet-HaPux-HaPuR. The CYP199A4 and CYP199A2genes were excised from pET28 vector using the Nco I and Hind III restriction enzymes introduced previously⁹ and cloned into the pRSF-HaPux vector to yield the plasmids pRSFDuet-HaPux-CYP199A4 and pRSFDuet-HaPux-CYP199A2.

Growth and expression of the in vivo system

The plasmids pETDuet-HaPux-HaPuR and pRSFDuet-HaPux-CYP199A4 were both transformed into competent BL21(DE3) cells and grown on LB plates containing kanamycin (30 μg mL⁻¹) and carbenicillin (100 μg mL⁻¹) (LB_kan–carb). A single colony was inoculated into 500 mL 2YT broth (2YT_kan–carb) in a 2 L flask and grown at 37 °C overnight. Protein expression was induced by the addition of 60 μM IPTG (from a 0.4 M stock in H₂O) and the temperature was reduced to 20 °C and the shaker speed reduced to 110 rpm. The growths were allowed to continue for another 24 h before the culture was harvested by centrifugation and the wet weight of the cell pellet was recorded. The pellet was washed in EMM media²⁵ and resuspended in an equal volume of EMM_kan–carb. The substrates 4-methoxybenzoic acid and veratric acid (4 mM from a 100 mM stock in EtOH) were added and a 1 mL aliquot was removed for analysis. The whole-cell reaction mixtures were then shaken at 220 rpm and 20 °C. Further 1 mL aliquots were then taken (at 20, 40, 60, 80, 100, 120, 180 and 240 min and 20 h) and the samples were extracted and analyzed by GC. Further additions of substrate were made when it was found to be exhausted and also after 6 hours when a larger aliquot of 4-methoxybenzoic acid or veratric acid was added and analyzed again at 21 h.

Cell dry weights (cdw) were calculated from the cell wet weight by incubating a known amount of the wet cell pellet at 150 °C until a constant mass was achieved. The activities of the whole-cell systems are reported as μM (g-cdw)⁻¹ min⁻¹.

Conclusion

In summary here we report the cloning, expression and characterization of a complete Class I electron transfer system from R. palustris HaA2 consisting of CYP199A4, a ferredoxin reductase (HaPuR) and a [2Fe–2S] ferredoxin (HaPux). The catalytic activity, product formation and substrate binding of the CYP199A4–HaPux–HaPuR system have been studied. CYP199A4 was found to catalyze efficient oxidative demethylation of 4-methoxybenzoic and veratric acids at the 4-position. Further studies on electron transfer proteins in the CYP199A4 and CYP199A2 systems could identify residues that are involved in protein recognition and electron transfer between the ferredoxin reductase, ferredoxin and CYP enzymes of these systems. We have also constructed whole-cell oxidation systems enabling in vivo preparative scale reactions for product formation including the rapid conversion of veratric acid to vanillic acid with 100% selectivity. The CYP199A4 system has the advantage over the CYP199A2 system in that the enzyme components, particularly the ferredoxin (HaPux), are produced in greater quantities in E. coli and consequently the whole-cell CYP199A4 system is considerably more active.

Acknowledgements

SGB would like to thank Prof. Caroline Harwood and Dr Yasuhiro Oda (University of Washington, Seattle) for supplying the genomic DNA of R. palustris HaA2. This work was supported by the Higher Education Funding Council for England, the Engineering and Physical Sciences Research Council, and Biotechnology and Biological Sciences Research Council (EP-D048559-1), UK.

References

1. F. P. Guengerich, J. Biol. Chem., 1991, 266, 10019–10022.
2. Cytochrome P450: Structure, Mechanism, and Biochemistry, ed. P. R. Ortiz de Montellano, Kluwer Academic/Plenum Press, New York, 3rd edn, 2005.
3. E. M. Isin and F. P. Guengerich, Biochim. Biophys. Acta, 2007, 1770, 314–329.
4. F. W. Larimer, P. Chain, L. Hauser, J. Lamerdin, S. Malfatti, L. Do, M. L. Land, D. A. Pelletier, J. T. Beatty, A. S. Lang, F. R. Tabita, J. L. Gibson, T. E. Hanson, C. Bobst, J. L. Torres y Torres, C. Peres, F. H. Harrison, J. Gibson and C. S. Harwood, Nat. Biotechnol., 2004, 22, 55–61.
5. Y. Oda, F. W. Larimer, P. S. Chain, S. Malfatti, M. V. Shin, L. M. Vergez, L. Hauser, M. L. Land, S. Braatsch, J. T. Beatty, D. A. Pelletier, A. L. Schaefer and C. S. Harwood, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 18543–18548.
6. Y. Oda, W. G. Meijer, J. L. Gibson, J. C. Gottschal and L. J. Forney, Microb. Ecol., 2004, 47, 68–79.
7. Y. Oda, W. Wanders, L. A. Huisman, W. G. Meijer, J. C. Gottschal and L. J. Forney, Appl. Environ. Microbiol., 2002, 68, 3467–3477.
8. S. G. Bell, N. Hoskins, F. Xu, D. Caprotti, Z. Rao and L. L. Wong, Biochem. Biophys. Res. Commun., 2006, 342, 191–196.
9. S. G. Bell, F. Xu, I. Forward, M. Bartlam, Z. Rao and L. L. Wong, J. Mol. Biol., 2008, 383, 561–574.
10. F. Xu, S. G. Bell, Y. Peng, E. O. D. Johnson, M. Bartlam, Z. Rao and L.-L. Wong, Proteins: Struct., Funct., Bioinf., 2009 DOI:10.1002/prot.22510.
11. S. J. Bent, C. L. Gucker, Y. Oda and L. J. Forney, Appl. Environ. Microbiol., 2003, 69, 5192–5197.
12. Y. Oda, B. Star, L. A. Huisman, J. C. Gottschal and L. J. Forney, Appl. Environ. Microbiol., 2003, 69, 5186–5191.
13. A. Aliverti, V. Pandini, A. Pennati, M. de Rosa and G. Zanetti, Arch. Biochem. Biophys., 2008, 474, 283–291.
14. A. V. Grinberg, F. Hannemann, B. Schiffler, J. Muller, U. Heinemann and R. Bernhardt, Proteins: Struct., Funct., Bioinf., 2000, 40, 590–612.
15. J. J. Muller, A. Muller, M. Rottmann, R. Bernhardt and U. Heinemann, J. Mol. Biol., 1999, 294, 501–513.
16. J. D. Lambeth and S. Kriengsiri, J. Biol. Chem., 1985, 260, 8810–8816.
17. P. W. Roome and J. A. Peterson, Arch. Biochem. Biophys., 1988, 266, 32–40.
18. M. Nishimura, D. Ishiyama and J. Davies, Biosci., Biotechnol., Biochem., 2006, 70, 2316–2319.
19. V. Venturi, F. Zennaro, G. Degrassi, B. C. Okeke and C. V. Bruschi, Microbiology, 1998, 144, 965–973.
20. P. Venkitasubramanian, L. Daniels and J. P. Rosazza, J. Biol. Chem., 2007, 282, 478–485.
21. R. H. van den Heuvel, W. A. van den Berg, S. Rovida and W. J. van Berkel, J. Biol. Chem., 2004, 279, 33492–33500.
22. R. Bernhardt, J. Biotechnol., 2006, 124, 128–145.
23. A. Chefson and K. Auclair, Mol. BioSyst., 2006, 2, 462–469.
24. P. Hlavica, Biotechnol. Adv., 2009, 27, 103–121.
25. S. G. Bell, C. F. Harford-Cross and L.-L. Wong, Protein Eng., 2001, 14, 797–802.
26. M. K. Julsing, S. Cornelissen, B. Buhler and A. Schmid, Curr. Opin. Chem. Biol., 2008, 12, 177–186.
27. M. Chamkha, E. Record, J. L. Garcia, M. Asther and M. Labat, Curr. Microbiol., 2002, 44, 341–349.
28. J. M. Woodley, Adv. Appl. Microbiol., 2006, 60, 1–15.
29. P. Y. Kim, D. J. Pollard and J. M. Woodley, Biotechnol. Prog., 2007, 23, 74–82.
30. J. Sambrook, E. F. Fritsch and T. Maniatis, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, 2nd edn, 1989.
31. T. Omura and R. Sato, J. Biol. Chem., 1964, 239, 2370–2378.
32. Y. Peng, F. Xu, S. G. Bell, L. L. Wong and Z. Rao, Acta Crystallogr., Sect. F: Struct. Biol. Cryst. Commun., 2007, 63, 422–425.
33. J. A. Peterson, M. C. Lorence and B. Armaneh, J. Biol. Chem., 1990, 265, 6066–6073.
34. A. Aliverti, B. Curti and M. A. Vanoni, Methods Mol. Biol. (Totowa, N. J.), 1999, 131, 9–23.
35. D. E. Blair and H. Diehl, Anal. Chem., 1961, 33, 867–870.
36. J. M. Moulis and J. Meyer, Biochemistry, 1982, 21, 4762–4771.
37. J. W. Williams and J. F. Morrison, Methods Enzymol., 1979, 63, 437–467.

---

Footnote

† Electronic supplementary information (ESI) available: Sequence alignments of the CYP199A4, CYP199A2 and CYP101A1 Class I electron transfer systems and mass spectra of the purified proteins are provided. See DOI: 10.1039/b913487e

---