Ionic liquid effects on the activity of monooxygenase P450 BM-3

Abstract

P450 monooxygenases oxidize a broad range of substrates including fatty acids, alcohols, aliphatic and aromatic hydrocarbons which are often poorly soluble in water and have K_m values in the millimolar range. Cosolvent and host–guest chemistry have thus been supplemented to increase catalytic efficiencies. Ionic liquids are alternative solvents and cosolvents in biocatalysis and have received an escalating interest due to unique properties such as tunable hydrophobicity and hydrophilicity. Effects of ten ionic liquids on cytochrome P450 BM-3 activity were investigated by evaluating the influence of hydrophobicity and ion pairs on P450 BM-3. Investigated ionic liquids comprised a range of four anions, four 1-alkyl-3-methylimidazolium derivatives differing in hydrophobic chain length, two altered head groups (pyridinium and pyrrolidinium) and an additional functional group in the side chain (nitrile). Results indicate that varying the chain length of the 1-alkyl-3-methylimidazolium has apparent effects on the resistance of P450 BM-3 (based on EC₅₀), with increasing inhibition in the order of EMIM Cl (232 ± 8) > BMIM Cl (148 ± 5) > HMIM Cl (32 ± 2) > OMIM Cl (9 ± 1).

---

Introduction

Ionic liquids have received increasing attention in recent years as alternative solvents in biocatalysis.^1–4 These studies were performed on isolated/immobilized enzymes or whole cell systems, in biphasic or homogenous ionic liquid–water mixtures and pure ionic liquids.^2,5–7 Stability, activity and enantioselectivity were altered through the use of ionic liquids for enzyme classes such as hydrolases (lipase,^5,8 esterase⁶ and protease⁹) and oxidoreductases (formate dehydrogenase,¹⁰ horseradish peroxidase¹¹ and cytochrome c¹²). Research on biocatalysis in ionic liquids is driven by solubility properties, low vapor pressure, and high thermal stability of ionic liquids which lead to lower inhalatory exposure and non-flammability. The latter are important considerations in large-scale industrial processes.

Cytochrome P450s are oxidoreductases and form a ubiquitous enzyme family which plays key roles in biochemical pathways, xenobiotic metabolism and detoxification. Cytochrome P450s can hydroxylate or epoxidize substrates by inserting a single oxygen molecule from oxygen into an organic compound.^13–15 These oxygenation reactions are often difficult to achieve through organic syntheses, especially when regio- or stereoselective oxygenation at ambient temperature, using oxygen as oxidant is required. P450 monooxygenases are attractive industrial biocatalysts due to their selectivity and wide spectrum of substrates accepted (alkanes, alkenes, alcohols, fatty acids, amides, polyaromatic hydrocarbons and heterocyclics).^16–19 However hydrophobic substrates of P450s often have K_m values in the millimolar range thus requiring the use of organic solvents or cosolvents for efficient biotransformations. P450 enzymes often cannot tolerate required amounts of organic solvents or cosolvents.^20,21

P450 BM-3 is one of the most well-studied and industrially important P450 enzymes (structure,^22–24catalytic efficiency^18,25 and substrate specificity^16,26) due to its robustness and design as a single component system (reductase and heme domain being fused on a single polypeptide²⁷). Ionic liquids represent an alternative to solvents and cosolvents for industrial important P450 BM-3 substrates such as fatty acids, alcohols, aliphatic and aromatic hydrocarbons which are known to be totally or partially miscible with ionic liquids.^28–30 The influence of ionic liquids has barely been studied for cytochrome P450s, and only one report for a multi-component system (CYP3A4) has been published.³¹ There have been no reports on the study of ionic liquid effects on enzymatic activity for single component P450 monooxygenases.

In this study, effects of ionic liquids and their hydrophobic side chains on P450 BM-3 activity are quantified. First hypotheses on inhibitory influences are further generated by measuring the OMIM Cl resistance of P450 BM-3 mutant F87A, in which the F87A substitution decreases tolerance of P450 BM-3 towards organic solvents.²¹

Results

The pNCA (p-nitrophenoxycarboxylic acid) assay³² was used to determine the activity of P450 BM-3 under the influence of ionic liquids using p-nitrophenoxydodecanoic acid (12-pNCA) as substrate. The pH remained stable under assay conditions. Concentration range of ten ionic liquids as well as calculated EC₅₀ values are listed in Table 1. The statistical calculations for EC₅₀ were performed using software library drfit for the R language and environment.³³ The tolerance of P450 BM-3 towards the investigated ionic liquids varies between EC₅₀ values of 9 and 232 mM. Among the ten ionic liquids, P450 BM-3 has the highest tolerance towards EMIM Cl (EC₅₀ = 232 ± 8 mM) and the least tolerance towards OMIM Cl (EC₅₀ = 9 ± 1 mM).

Table 1 Compounds, abbreviations, concentrations of ionic liquids used for P450 BM-3 activity measurements and calculated EC₅₀ values of ionic liquids employed in inhibitory studies of P450 BM-3. All parameters were determined from triplicate measurements

Compound	Abbreviation	Concentration range/mM	EC50^a/mM
a EC50 values determined based on triplicate measurements (confidence intervals α = 0.025)
1-Ethyl-3-methylimidazolium chloride	EMIM Cl	0–420	232 ± 8
1-Ethyl-3-methylimidazolium trifluoromethanesulfonate	EMIM CF3SO3	0–180	94 ± 6
1-Ethyl-3-methylimidazolium trifluoroacetate	EMIM CF3COO	0–180	72 ± 3
1-Ethyl-3-methylimidazolium tetrafluoroborate	EMIM BF4	0–100	42 ± 3
1-Butylpyridinium chloride	Bpy Cl	0–264	195 ± 9
1-Cyanomethyl-3-methylimidazolium chloride	1M11CN Cl	0–360	182 ± 8
1-Butyl-1-methylpyrrolidinium chloride	BMPyr Cl	0–300	175 ± 8
1-Butyl-3-methylimidazolium chloride	BMIM Cl	0–240	148 ± 5
1-Hexyl-3-methylimidazolium chloride	HMIM Cl	0–72	32 ± 2
1-Octyl-3-methylimidazolium chloride	OMIM Cl	0–15	9 ± 1
Sodium chloride	NaCl	0–600	202 ± 9
Sodium trifluoromethanesulfonate	NaCF3SO3	0–300	166 ± 8
Sodium trifluoroacetate	NaCF3COO	0–360	162 ± 7
Sodium tetrafluoroborate	NaBF4	0–360	119 ± 8

---

Effect of the anions and EMIM cation on P450 BM-3 activity

The EC₅₀ values of EMIM with different anions (Cl^–, CF₃SO₃^–, CF₃COO^– and BF₄^–) show that P450 BM-3 has the lowest tolerance towards EMIM BF₄. The residual activities for each anion are normalized to the activity when no ionic liquid is added and are shown in Fig. 1. It can be observed from Fig. 1 that EMIM Cl has a visibly lower inhibitory effect compared to the other EMIM anions. Inhibition is also reflected by the absolute decrease of EC₅₀ from EMIM Cl to EMIM CF₃SO₃, which is at least 2.7-fold higher than any difference between the three other investigated EMIM salts (EMIM CF₃SO₃, EMIM CF₃COO and EMIM BF₄). In fact Fig. 1 shows that the order of inhibition for EMIM CF₃SO₃ and EMIM CF₃COO changes around 150 mM suggesting very similar inhibitions by these two ionic liquids.

Fig. 1 Residual activities of P450 BM-3 in presence of EMIM and anions (Cl^–, CF₃SO₃^–. CF₃COO^–, BF₄^–). Residual activity is given as ratio of initial catalytic turnover number in the presence of ionic liquids to that in the absence of ionic liquids. –■–: EMIM Cl; –□–: EMIM CF₃SO₃; –●–: EMIM CF₃COO; –○–: EMIM BF₄.

The Na⁺ salts of the above four anions (Cl^–, CF₃SO₃^–, CF₃COO^– and BF₄^–) show lower inhibitory influences than the corresponding EMIM salts except NaCl (EC₅₀ value 0.9-fold of the EMIM Cl). These results suggest that reduced activity of P450 BM-3 can mainly be attributed to an anion effect. Differences in the extend of inhibition indicate that inhibitory differences between the Na⁺ and EMIM salts are not solely due to an anion effect, but rather depend on EMIM and Na⁺ counter ion effects on corresponding anions.

As the chloride anion has the lowest inhibition among the tested anion, it was used as the counter ion for all subsequent tests on ionic liquids with different cations.

Effect of the alkyl-chain length of EMIM cations on P450 BM-3 activity

Effects of alkyl chain length for the imidazolium cation were investigated by using 1-alkyl-3-methylimidazolium chlorides in which the alkyl-chain at one N-atom was stepwise elongated from ethyl- (C₂) to octyl- (C₈). Fig. 2 shows a decreasing tolerance of P450 BM-3 with stepwise increased alkyl chain length. From the EC₅₀ values in Table 1, the tolerance of P450 BM-3 towards OMIM Cl is almost 20-fold lower than EMIM Cl. Fig. 2 and EC₅₀ values in Table 1 show a clear increase in inhibition in accordance to alkyl-chain elongations from C₂ to C₈.

Fig. 2 Effect of hydrophobic side chain of 1-alkyl-3-methylimidazolium on residual activities of P450 BM-3. Residual activity is given as ratio of initial catalytic turnover number in the presence of ionic liquids to that in the absence of ionic liquids. –■–: EMIM Cl; –□–: BMIM Cl; –●–: HMIM Cl; –○–: OMIM Cl.

Effect of cation aromaticity, size and free electron pair in alkyl-chain of imidazolium cation

Three other cationic structures were investigated: Bpy Cl, BMPyr Cl and IM11CN Cl. The resulted EC₅₀ are in between EMIM Cl and BMIM Cl. Bpy Cl has a pyridinium head group which is larger compared to the imidazolium head group (six- compared to five-membered ring). However this does not seem to have large effects on the inhibition (EC₅₀ ∼1.3-fold of BMIM Cl) and EC₅₀ remains lower than EMIM Cl. BMPyr Cl has a side chain on its cation which is the same as BMIM, but unlike the imidazolium (BMIM) head group, the pyrrolidinium (BMPyrH) head group does not have an aromatic ring system. Aromaticity however has only a minor effect on the inhibition. The BMPyr Cl EC₅₀ value is ∼1.2-fold higher than that of BMIM Cl (Table 1). The side chain carrying the nitrile functional group on IM11CN Cl has a chain length that is in between EMIM and BMIM. Differing from the alkyl imidazolium side chains, a polar nitrile functional group was introduced. Possible electrostatic interactions have however no notable influence on P450 BM-3 inhibition (EC₅₀ for IM11CN Cl lies between EMIM Cl and BMIM Cl).

k _cat and K_m values for P450 BM-3 in ionic liquid–water mixture

As seen in Table 1 and Fig. 2, the effect of the length of side chain has an apparent influence on the inhibition of P450 BM-3. For further insight, the kinetic parameters of the P450 BM-3 hydroxylation of 12-pNCA were determined in the presence of the imidazolium chlorides (Table 2). For each of the 1-alkyl-3-methylimidazolium chlorides (BMIM Cl, HMIM Cl and OMIM Cl), the amount of ionic liquid equivalent to the EC₅₀ concentration was employed. In all three ionic liquids a lower k_cat and a higher K_m was obtained which led to a 4.0–4.6-fold lower catalytic efficiency (k_cat/K_m) when compared to the P450 BM-3 activity in no ionic liquid. In detail, the k_cat was reduced to 0.50–0.62-fold while the K_m increased 2.0–2.6-fold. Kinetic constants (k_cat, K_m, K_i) were calculated by assuming mixed inhibition kinetics of P450 BM-3 (Table 2). Table 2 shows that K_i decreases 18-fold as the side chain length increases from BMIM to OMIM. Changes in K_i suggest that the increasing inhibition as side chain length increases can be attributed to a stronger association of the ionic liquid to P450 BM-3. Increasing from BMIM Cl to HMIM Cl shows largest decrease in K_i (5.6-fold) and EC₅₀ values (4.6-fold) when comparing 1-alkyl-3-methylimidazolium from C₄ to C₈.

Table 2 Effect of 1-alkyl-3-methylimidazolium chlorides on the kinetic parameters (k_cat, K_m, K_i) of P450 BM-3 wildtype and mutant F87A. All parameters were determined from triplicate measurements

Inhibitor	Inhibitor added/mM	k cat/min^–1	K m/µM	K cat/Km	k cat ratio^a	K m ratio^a	K i ^b/mM
Wild type
No inhibitor	0	136.6 ± 2.6	2.8 ± 0.3	48.8	—	—	—
BMIM Cl	148	68.2 ± 0.7	5.6 ± 0.3	12.2	0.50	2.0	49.2
HMIM Cl	32	77.8 ± 1.0	7.4 ± 0.4	10.5	0.57	2.6	8.8
OMIM Cl	9	84.2 ± 2.2	7.4 ± 0.7	11.4	0.62	2.6	2.7

---

Mutant F87A
a k cat and Km ratio is the ratio of kinetic parameters in the presence of ionic liquids to that in the absence of ionic liquids. b K i = dissociation constant of inhibitor to enzyme.
No inhibitor	0	404.8 ± 5.7	7.2 ± 0.4	56.2	—	—	—
OMIM Cl	9	31.4 ± 1.8	20.2 ± 2.8	1.6	0.08	2.8	0.3

---

Role of amino acid position 87 in P450 BM-3 ionic liquid tolerance

The amino acid residue phenylalanine 87 is important for the organic solvent resistance of P450 BM-3. Mutating this residue to alanine resulted in a significant drop in resistance of P450 BM-3 towards DMSO, THF, ethanol and acetonitrile.²¹ The catalytic efficiency of P450 BM-3 F87A mutant in the presence of OMIM Cl decreased 35-fold compared to when no ionic liquid was added, which is significantly larger than the 4.3-fold decrease observed in wildtype P450 BM-3 (see Table 2). The change in K_m is of the same order of magnitude between the wildtype and F87A mutant while the major contributor of this drastic difference in catalytic efficiency is the 12.5-fold decrease in k_cat in F87A compared to the 1.6-fold decrease seen in the wildtype. A lower K_i is also observed suggesting better association of OMIM Cl in F87A mutant compared to the wildtype.

NADPH consumption and P450 spectra using ionic liquids as substrate

The NADPH consumption assay was used to investigate the possibility of the ionic liquids being substrates of P450 BM-3, particularly the long chain 1-alkyl-3-methylimidazolium. Except for IM11CN Cl, no NADPH consumption higher than the negative control was observed. For IM11CN Cl, a NADPH consumption of <12% of employed NADPH was detected after 20 min in absence of any other substrate. All employed ionic liquids except IM11CN Cl are therefore not or very poor substrates of P450 BM-3. NADPH consumption assay also shows that OMIM Cl is not a substrate of the F87A variant. Spectral measurements of P450 BM-3 in the presence of 1-alkyl-3-methylimidazoliums (BMIM Cl, HMIM Cl, OMIM Cl) did not resulted in any spectral changes (absorbance decrease around 420 nm and increase around 390 nm). These spectral changes are characteristic of substrate binding in P450 BM-3 which leads to perturbation of the sixth water ligand from the heme iron center. The iron changes thereby from a low spin to a high spin configuration that subsequently leads to an increase in the heme iron reduction potential, a pre-requisite for the electron transfer from the redox partner. All experimental results suggest a lack of inhibitor coordination to the heme iron as reported for the cosolvent DMSO.²²

Discussion

Principles of structure–function relationships of enzymes and biological systems in ionic liquids have been discovered and quantified.^34–36 There has been so far only one study on the effects of ionic liquids on cytochrome P450 activities³¹ and this report does not evaluate the potential use of ionic liquids for industrial biotransformations. In this manuscript, we have performed a systematic study to quantify ionic liquid effects on P450 BM-3 activity. In detail we studied the influence of four anions and EMIM cation, the effect of alkyl-chain length of 1-alkyl-3-methylimidazolium cations, cation aromaticity and size as well as an additional free electron pair in the alkyl-chain of imidazolium cation on P450 BM-3 activity. These influences of the P450 BM-3 activity were quantified by determining the EC₅₀ values, k_cat, K_m and K_i using the pNCA activity assay.³² Based on these results the EMIM Cl (EC₅₀ = 232 ± 8 mM) is a preferred candidate for biotransformations in ionic liquids.

EC₅₀ values are at least an order of magnitude higher for P450 BM-3 wildtype when compared to studies on acetylcholinesterase.³⁴ When compared to hydrolases P450 BM-3 resistance to ionic liquids is however low. Hydrolases even function in pure ionic liquid solutions.^5,6,37 The Cl^– anion has a significantly lower inhibitory effect than CF₃SO₃^–, CF₃COO^– and BF₄^– in both cations (EMIM, Na⁺; see Table 1 and Fig. 1). For the cations (EMIM, Na⁺), no general anion trends could be observed. The EMIM and Na⁺ cations however have similar inhibitory effects when Cl^– ion is used as counter anion (Table 1 and Fig. 1). One possible explanation of the different anion effects is the Hofmeister series which suggests that kosmotropic anions stabilize enzymes while chaotropic anions destabilize biocatalysts. According to the Hofmeister series, chaotropicity increases in the order of CF₃COO^– > Cl^– > BF₄^–.^38,39 The observed inhibition does show that the most chaotropic anion BF₄^– is most inhibitory for P450 BM-3, while CF₃COO^– and Cl^– behave differently from the Hofmeister series. Such exceptions are also reported for bovine carbonic anhydrase, succinate dehydrogenase and NADH oxidase.³⁸

The most significant activity effect due to changes in the cation was seen when the alkyl chain length of the 1-alkyl-3-methylimidazolium cation was increased, which has also been previously observed in acetylcholinesterase.³⁴ Additionally, a decreased enantioselectivity of a protease from Bacillus licheniformis had been reported with increasing cation side-chain length (EMIM CF₃COO > BMIM CF₃COO > HMIM CF₃COO).⁴⁰ It had been suggested that increased destabilization with increased alkyl chain length follows the Hofmeister series where longer side-chain length increases hydrophobicity and kosmotropicity of cations.³⁹ The Hofmeister series remains the best possible postulation for enzyme destabilization by cations but it is difficult to generalize for all the enzyme systems that have been studied so far (lipase, esterase, dehydrogenase and P450s).³⁸ For example the kinetic resolution of 1-phenylethanol by Candida antarcticalipase A and Candida antarcticalipase B improves when the 1-alkyl-3-methyllimidazolium cation chain length increases from BMIM BF₄ to HMIM BF₄ and OMIM BF₄.⁴¹

Increasing the alkyl side chain from ethyl- to octyl- leads to more than 25-fold decrease in the EC₅₀ value for P450 BM-3 wildtype. Table 2 summarizes the kinetic parameters (k_cat, K_m, K_i) of BMIM Cl, HMIM Cl and OMIM Cl. Analysis of the P450 BM-3 structure shows a substrate binding pocket close to heme that is lined by hydrophobic residues (F42, Y51, L181, M185, L188, A328, A330 and M354).⁴² The increased inhibition can be attributed to the extended hydrophobicity of longer side chains which cause stronger associations with the hydrophobic patches in the P450 BM-3 substrate channel. Consequently the K_m increases as substrate access is adversely affected due to competition from the 1-alkyl-3-methylimidazolium cation. Changing the cation to pyrrolidinium or addition of a nitrile functional group (Table 1 and 2) does not impact significantly on the EC₅₀ values (Table 1 and 2).

As the full crystal structure of P450 BM-3 has not been resolved, one can only base hypotheses on the crystallized heme domain of P450 BM-3. One possible explanation of the lower K_m and k_cat of P450 BM-3 is the binding of 1-alkyl-3-methylimidazolium cation at the entrance of the substrate access channel. To obtain further insight, short molecular dynamics (MD) simulation of BMIM, HMIM and OMIM at the entrance of the substrate channel and inside the channel were performed. For the MD at the substrate channel entrance, the final conformations show the head group pointing towards Glu13, likely due to electrostatic interaction; the hydrophobic tail points into the substrate channel (see Fig. 3). Glu13 lies within the hydrophobic patch (P11, L14, L17, P18, P45 and A191) which is believed to be a site important for initial binding of lipophilic substrates.⁴² The presence of the 1-alkyl-3-methylimidazolium cation may interfere with this initial binding thereby affecting substrate binding and leading to higher K_m values (Table 2). Despite the substrate channel being hydrophobic and having a positively charged R47 which coordinates the carboxylate terminal of the fatty acid substrates, it is known that positively charged substrates can enter and be converted by P450 BM-3.⁴³ MD of the positively charged 1-alkyl-3-methylimidazoliums inside the substrate channel were performed and the positively charged head group moves away from Arg47 and orients towards Glu435 with the hydrophobic tail pointing towards the heme (see Fig. 3). OMIM, which has the longest hydrophobic tail of the investigated ionic liquids, is located ∼5 Å from the Fe iron, 9 Å from Y51 and 13 Å from R47 according to the MD simulations. The 1-alkyl-3-methylimidazoliums in the substrate channel in such a configuration can interfere with the orientation of the substrate with respect to the heme, thereby affecting the terminal hydroxylation of 12-pNCA and lead to the lower k_cat values.

Fig. 3 Final conformations from molecular dynamics docking simulations. Top row (substrate channel entrance; from left to right): BMIM, HMIM, OMIM; Bottom row (inside substrate binding channel; from left to right): BMIM, HMIM, OMIM. Figures on the left are labeled by highlighting important P450 BM-3 positions/regions. The 1-alkyl-3-methylimidazolium is colored in pink.

The mutation F87A is known to influence P450 BM-3's tolerance towards organic solvents.²¹ The OMIM Cl resistance of P450 BM-3 F87A was thus determined. The change in phenylalanine 87 to alanine led to a lower resistance of P450 BM-3 to OMIM Cl as reflected in the 35-fold decrease in k_cat/K_m compared to the 4-fold decrease observed in P450 BM-3 wildtype (Table 2). The change in k_cat/K_m is contributed largely by the 12.5-fold lower k_cat value (Table 2). The presence of phenylalanine 87 at the end of channel can possibly interact further with the p-nitrophenyl group of the 12-pNCA substrate through π–π interaction and contribute to a catalytic “active” orientation of 12-pNCA, especially if the OMIM inside the substrate channel affects substrate orientation. Changing F87 to alanine removes this interaction and leads to a much lower k_cat value. More extensive studies are currently being performed to elucidate the actual mode of inhibition by the 1-alkyl-3-methylimidazoliums.

Conclusion

EMIM Cl is identified as a preferred ionic liquid as cosolvent for biotransformations catalyzed by P450 BM-3. Increasing alkyl side chain from ethyl- to octyl- in the imidazolium cation significantly increases (26-fold) the inhibitory effect of ionic liquids which can likely be attributed to increase hydrophobicity of the alkyl chain. Potential binding sites lie at the hydrophobic mouth of the substrate access channel (OMIM head group pointing towards Glu13) and inside the substrate channel (OMIM head group pointing towards Glu435). P450 BM-3 shows, based on the EC₅₀ values, a lower resistance to ionic liquids compared to previously studied organic solvents such as DMSO (50% residual activity at ∼3 M), acetonitrile (50% residual activity at ∼1 M) and ethanol (50% residual activity at ∼1 M).²¹ However for EMIM Cl, IM11CN Cl, Bpy Cl and BMPyr Cl, higher resistances were observed compared to THF (50% residual activity at ∼0.1 M). Changing of the cation to pyridinium and pyrrolidinium or addition of the nitrile functional group did not generate obvious changes in EC₅₀.

Directed P450 BM-3 evolution studies employing EMIM Cl are currently being performed for discovering principles on how P450 BM-3 and monooxygenases in general can be reengineered to operate efficiently in EMIM Cl.

Experimental

Materials

All chemicals used were of analytical reagent grade or higher quality and purchased from Sigma-Aldrich Chemie (Taufkirchen, Germany), Applichem (Darmstadt, Germany) and Carl Roth (Karlsruhe, Germany). NADPH tetrasodium salt was purchased from Codexis (Redwood city, CA, United States), p-nitrophenoxydodecanoic acid (12-pNCA) was synthesized as previously described³² and the ionic liquids used were from Merck KGaA (Darmstadt, Germany).

P450 BM-3 expression and purification

P450 BM-3 wildtype and mutants were expressed in 1-L Erlenmeyer flask containing 250 mL of terrific broth (100 mg L^–1ampicillin, 250 µL of trace elements as previously described²¹). The cells were harvested by centrifugation (4 °C, 3220 g, 15 min) and the cell pellet was resuspended in K_xH_3–xPO₄ (25 mM, pH 7.5) and lysed by high-pressure homogenization (EmulsiFlex-C3, Avestin Inc, Ottawa, Canada; two cycles of homogenizing pressure 1500–1800 bar). Cell lysate was centrifuged (20000 g, 45 min) and the supernatant was further cleared through a 0.45 µm filter (17765 K, Sartorius, Hannover, Germany). The filtrate was loaded on a DEAE650s anion exchanger column (Tosoh Bioscience, Stuttgart, Germany) and purified as previously described.⁴⁴ P450 BM-3 concentrations were measured by CO-difference spectra as reported by Omura and Sato using ε = 91 mM^–1 cm^–1.⁴⁵

Determination of pH in ionic liquid–water mixture

The pH of the reaction mixture used for activity tests was determined using a laboratory pH meter (Knick, Berlin, Germany). A reaction mixture (2.5 mL) consist of 1.5 mL buffer (0.3 M Tris pH 8.2), 0.3 mL ionic liquid (0.5 M EMIM Cl; 0.18 M EMIM CF₃SO₃ and EMIM CF₃COO; 0.096 M EMIM BF₄; 0.3 M BMPyr Cl; 0.36 M 1M11CNCl; 0.24 M BMIM Cl; 0.072 M HMIM Cl; 0.015 M OMIM Cl), 75 µM 12-pNCA and 0.5% v/v DMSO. The pH was equilibrated (10 min) or until a stable reading was obtained.

Determination of NADPH consumption and P450 spectra using ionic liquids as substrate

For NADPH consumption assay, 600 µL tris buffer (0.3 M, pH 8.2), 0.0975 nmol P450 BM-3 WT and 10 µL of ionic liquids (0.175 M EMIM Cl; 0.075 M EMIM CF₃SO₃ and EMIM CF₃COO; 0.04 M EMIM BF₄; 0.15 M 1M11CN Cl; 0.125 M BMPyr Cl; 0.1 M BMIM Cl; 0.03 M HMIM Cl; 0.0125 M OMIM Cl) were used. The reaction mixture was topped up by supplementing doubly-distilled H₂O (900 µL), incubated (5 min) and initiated by adding NADPH (0.25 mM; final reaction volume 1 mL). NADPH consumption was followed spectrophotometrically (340 nm, 10 min; Specord200, Analytik Jena AG, Jena, Germany) in disposable cuvettes. A positive control was performed by adding 12-pNCA (75 µM) and no ionic liquids to the reaction mixture; a negative control was performed in the absence of 12-pNCA substrates. For spectra measurements, 0.2 nmol of P450 BM-3 were added to phosphate buffer (25 mM, pH 7.5) with ionic liquids (148 mM BMIM Cl; 32 mM HMIM Cl; 9 mM OMIM Cl) and incubated for 5 min. Subsequently spectra measurements between 350 and 700 nm were recorded using a Specord200 (Analytik Jena AG).

Determination of P450 BM-3 activity in ionic liquid–water mixture

Residual activities of P450 BM-3 WT in the presence of ionic liquids were measured in 96-well microtiter plates (Grenier Bio-One GmbH, Frickenhausen, Germany) and ionic liquids were always prepared freshly in doubly-distilled H₂O. The 200 µL reaction mixture finally consists of 0.18 M Tris buffer (pH 8.2), gradient concentrations of ionic liquids, 75 µM 12-pNCA, 0.5% DMSO and 0.0195 nmol P450 BM-3 WT. Employed concentrations of ionic liquids and the Na⁺ salts used for these activity measurements are listed in Table 1. Reaction mixtures were always incubated for 5 min before 12-pNCA conversion was initiated by adding NADPH (25 µL, 2.5 mM). Conversion of the 12-pNCA was monitored using the Tecan Sunrise microtiter plate reader (Tecan, Zurich, Switzerland) at 405 nm for a duration of 6–10 min with an average of 3 s interval between each reading cycle. All P450 BM-3 residual activities in ionic liquid–water mixture were determined with at least triplicate measurements.

Determination of k_cat and K_m values for P450 BM-3 in ionic liquid–water mixture

The kinetic constants were determined using the 12-pNCA assay. The assay was performed in disposable cuvettes as previously described³² except that the tris buffer composition was changed (0.18 M, pH 8.2, final concentration of DMSO is 0.5%). Amount of P450 BM-3 per reaction was always 0.1 nmol. The activity was monitored at 410 nm using Specord 200 (Analytik Jena AG). 12-pNCA concentrations were varied from 2.5–125 µM. All measurements were made in triplicates and the k_cat and K_m values were determined using GraphPad Prism software (GraphPad Software, San Diego, CA, USA) with fitting to a one-site binding hyperbola equation using non-linear regression.

Imidazolium cation docking in the substrate binding site of P450 BM-3

For MD simulations the atomic coordinates of P450 BM-3 wildtype were obtained using the crystal structure file (pdb code 1bu7).²⁴ All crystallographic water and other ligand molecules were removed except the water molecule coordinating the iron in the active site of the enzyme. Polar and aromatic hydrogen atoms were added to the crystal structure. Atomic partial charges for the protein were assigned from the GROMOS96 43b1 force field.⁴⁶ The initial structure of the three ionic liquids were optimized at HF level using 6-31G** basis set using Gaussian03 program (www.gaussian.com). The force field parameters for the ionic liquid were taken from published data.⁴⁷ Initial docked conformations were manually generated using PBD Viewer. The initial docked conformation of 1-alkyl-3-methylimidazolium at the substrate channel entrance was placed between P45 and A191,^48,49 while initial docked conformation inside the channel was selected according to the palmitoleic acid substrate (pdb code 1fag)²³ close to R47 and Y51. Optimization of the docked structures was performed using standard energy minimization followed by a molecular dynamics simulation of at least 30 ps using position restraints on the protein backbone. All the simulation were performed using the program GROMACS (version 3.1) (www.gromacs.org). Final conformations were used to analyse the cation position.

Notes and references

1. S. Park and R. J. Kazlauskas, Curr. Opin. Biotechnol., 2003, 14, 432–437.
2. R. A. Sheldon, R. M. Lau, M. J. Sorgedrager, F. van Rantwijk and K. R. Seddon, Green Chem., 2002, 4, 147–151.
3. F. van Rantwijk, R. M. Lau and R. A. Sheldon, Trends Biotechnol., 2003, 21, 131–138.
4. Z. Yang and W. B. Pan, Enzyme Microb. Technol., 2005, 37, 19–28.
5. J. L. Kaar, A. M. Jesionowski, J. A. Berberich, R. Moulton and A. J. Russell, J. Am. Chem. Soc., 2003, 125, 4125–4131.
6. M. Persson and U. T. Bornscheuer, J. Mol. Catal. B: Enzym., 2003, 22, 21–27.
7. H. Pfruender, R. Jones and D. Weuster-Botz, J. Biotechnol., 2006, 124, 182–190.
8. S. J. Nara, J. R. Harjani and M. M. Salunkhe, Tetrahedron Lett., 2002, 43, 2979–2982.
9. H. Zhao, L. Jackson, Z. Y. Song and A. Olubajo, Tetrahedron: Asymmetry, 2006, 17, 2491–2498.
10. N. Kaftzik, P. Wasserscheid and U. Kragl, Org. Process Res. Dev., 2002, 6, 553–557.
11. S. Sgalla, G. Fabrizi, S. Cacchi, A. Macone, A. Bonamore and A. Boffi, J. Mol. Catal. B: Enzym., 2007, 44, 144–148.
12. J. A. Laszlo and D. L. Compton, J. Mol. Catal. B: Enzym., 2002, 18, 109–120.
13. A. Chefson and K. Auclair, Mol. Biosyst., 2006, 2, 462–469.
14. A. J. Fulco, Ann. Rev. Pharm. Toxicol., 1991, 31, 177–203.
15. D. F. V. Lewis and A. Wiseman, Enzyme Microb. Technol., 2005, 36, 377–384.
16. D. Appel, S. Lutz-Wahl, P. Fischer, U. Schwaneberg and R. D. Schmid, J. Biotechnol., 2001, 88, 167–171.
17. A. B. Carmichael and L. L. Wong, Eur. J. Biochem., 2001, 268, 3117–3125.
18. A. Glieder, E. T. Farinas and F. H. Arnold, Nat. Biotechnol., 2002, 20, 1135–1139.
19. S. Graham-Lorence, G. Truan, J. A. Peterson, J. R. Falck, S. Z. Wei, C. Helvig and J. H. Capdevila, J. Biol. Chem., 1997, 272, 1127–1135.
20. S. Kumar, L. Sun, B. K. Muralidhara, J. R. Halpert and J. R. Halpert, Protein Eng. Des. Sel., 2006, 19, 547–554.
21. T. S. Wong, F. H. Arnold and U. Schwaneberg, Biotechnol. Bioeng., 2004, 85, 351–358.
22. J. Kuper, T. S. Wong, D. Roccatano, M. Wilmanns and U. Schwaneberg, J. Am. Chem. Soc., 2007, 129, 5786–5787.
23. H. Y. Li and T. L. Poulos, Nat. Struct. Biol., 1997, 4, 140–146.
24. I. F. Sevrioukova, H. Y. Li, H. Zhang, J. A. Peterson and T. L. Poulos, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 1863–1868.
25. K. L. Tee and U. Schwaneberg, Angew. Chem., Int. Ed., 2006, 45, 5380–5383.
26. T. S. Wong, N. Wu, D. Roccatano, M. Zacharias and U. Schwaneberg, J. Biomol. Screen., 2005, 10, 246–252.
27. L. O. Narhi and A. J. Fulco, J. Biol. Chem., 1987, 262, 6683–6690.
28. U. Domanska and A. Marciniak, J. Chem. Thermodyn., 2005, 37, 577–585.
29. U. Domanska and A. Marciniak, Fluid Phase Equilib., 2006, 260, 9–18.
30. S. Park, F. Viklund, K. Hult and R. J. Kazlauskas, Green Chem., 2003, 5, 715–719.
31. A. Chefson and K. Auclair, ChemBioChem, 2007, 8, 1189–1197.
32. U. Schwaneberg, C. Schmidt-Dannert, J. Schmitt and R. D. Schmid, Anal. Biochem., 1999, 269, 359–366.
33. J. Ranke, Drfit: Dose–response data evaluation, 2005 (computer program).
34. F. Stock, J. Hoffmann, J. Ranke, R. Stormann, B. Ondruschka and B. Jastorff, Green Chem., 2004, 6, 286–290.
35. S. Stolte, J. Arning, U. Bottin-Weber, M. Matzke, F. Stock, K. Thiele, M. Uerdingen, U. Welz-Biermann, B. Jastorff and J. Ranke, Green Chem., 2006, 8, 621–629.
36. S. Stolte, J. Arning, U. Bottin-Weber, A. Müller, W. R. Pitner, U. Welz-Biermann, B. Jastorff and J. Ranke, Green Chem., 2007, 9, 760–767.
37. R. M. Lau, F. van Rantwijk, K. R. Seddon and R. A. Sheldon, Org. Lett., 2000, 2, 4189–4191.
38. H. Zhao, J. Mol. Catal. B: Enzym., 2005, 37, 16–25.
39. H. Zhao, O. Olubajo, Z. Y. Song, A. L. Sims, T. E. Person, R. A. Lawal and L. A. Holley, Bioorg. Chem., 2006, 34, 15–25.
40. H. Zhao, S. M. Campbell, L. Jackson, Z. Y. Song and O. Olubajo, Tetrahedron: Asymmetry, 2006, 17, 377–383.
41. S. H. Schofer, N. Kaftzik, P. Wasserscheid and U. Kragl, Chem. Commun., 2001, 425–426.
42. K. G. Ravichandran, S. S. Boddupalli, C. A. Hasemann, J. A. Peterson and J. Deisenhofer, Science, 1993, 261, 731–736.
43. C. F. Oliver, S. Modi, W. U. Primrose, L. Y. Lian and G. C. K. Roberts, Biochem. J., 1997, 327, 537–544.
44. U. Schwaneberg, A. Sprauer, C. Schmidt-Dannert and R. D. Schmid, J. Chromatogr., A, 1999, 848, 149–159.
45. T. Omura and R. Sato, J. Biol. Chem., 1964, 239, 2370–2378.
46. W. F. van Gunsteren, S. R. Billeter, A. A. Eising, P. H. Hünenberger, P. Krüger, A. E. Mark, W. R. P. Scott and I. G. Tironi, in Biomolecular Simulation: The GROMOS96 Manual and User Guide, vdf Hochschulverlang, ETH Zürich, Switzerland, 1996.
47. N. M. Micaelo, A. M. Baptista and C. M. Soares, J. Phys. Chem. B, 2006, 110, 14444–14451.
48. M. D. Paulsen and R. L. Ornstein, Proteins: Struct., Funct., Genet., 1995, 21, 237–243.
49. D. Roccatano, T. S. Wong, U. Schwaneberg and M. Zacharias, Biopolymers, 2006, 83, 467–476.

---