Deactivation mechanisms of chloroperoxidase during biotransformations

Inactivation mechanisms of chloroperoxidase (CPO) from Caldariomyces fumago have been investigated with the aim of improving the practical utility of CPO for hydrocarbon oxidation. Deactivation studies in the presence of oxidants (i.e., hydrogen peroxide and t-butyl hydroperoxide) indicated that CPO lost oxidation activity toward hydrocarbon substrates during dismutation of hydrogen peroxide. The loss of enzyme activity was accompanied by the apparent destruction of the heme rather than aggregation or denaturation of the apo-protein. The decrease of enzyme activity was significantly retarded by adding the radical scavenger t-butyl alcohol at pH 4.1, or by optimizing the reaction pH. CPO retained greatest oxidation activity at pH 5-6, which may produce a more favorable ionization state of the key amino acid (Glu-183) and thus reduce radical formation. As a result of higher activity at pH 5-6, the total turnover numbers (TTN, defined as the amount of product produced over the catalytic lifetime of the enzyme) for the oxidation of toluene and o-, m-, p-xylenes in substrate/aqueous emulsion systems ranged from ca. 10% to 110% higher at pH 5.5 (20,000 to 45,000 mol product/mol enzyme) compared to pH 4.1. Furthermore, TTNs of CPO increased with increasing turnover frequencies, indicating that higher activity toward reducing substrates reduces radical formation and stabilizes CPO toward inactivation by H(2)O(2). These findings demonstrate the important relationship between CPO stability and activity, and illustrate that large improvements in CPO activity and stability can be achieved through solvent engineering.     

Chloroperoxidase-catalyzed Epoxidation of Styrene in Aqueous and Nonaqueous Media

To overcome poor product yields and stability in aqueous solution, we have examined the chloroperoxidase (CPO from Caldariomyces fumago ) catalyzed oxidation of styrene in organic media using tert -butyl hydroperoxide as external oxidant. CPO's intrinsic catalytic activity in tert -butanol , as reflected in its k cat value, was ca. one-fourth of that in aqueous buffer, indicating that the enzyme remains highly active in the organic solvent. Styrene epoxidation reactions were modeled in both aqueous and nonaqueous media to provide global kinetic information, which dominates non-initial rate conditions and is heavily influenced by continuous deactivation of the CPO. Deactivation studies revealed that the enzyme is deactivated quickly by the combination of the tert -butyl hydroperoxide and styrene, possibly due to the styrenic free radicals generated during the enzymatic reaction. These results may enable catalyst-engineering strategies to be initiated to improve the prospects of using CPO in nonaqueous media for large-scale epoxidation reactions.     

Chloroperoxidase-catalyzed Epoxidation of Styrene in Aqueous and Nonaqueous Media

To overcome poor product yields and stability in aqueous solution, we have examined the chloroperoxidase (CPO from Caldariomyces fumago ) catalyzed oxidation of styrene in organic media using tert -butyl hydroperoxide as external oxidant. CPO's intrinsic catalytic activity in tert -butanol, as reflected in its k cat value, was ca. one-fourth of that in aqueous buffer, indicating that the enzyme remains highly active in the organic solvent. Styrene epoxidation reactions were modeled in both aqueous and nonaqueous media to provide global kinetic information, which dominates non-initial rate conditions and is heavily influenced by continuous deactivation of the CPO. Deactivation studies revealed that the enzyme is deactivated quickly by the combination of the tert -butyl hydroperoxide and styrene, possibly due to the styrenic free radicals generated during the enzymatic reaction. These results may enable catalyst-engineering strategies to be initiated to improve the prospects of using CPO in nonaqueous media for large-scale epoxidation reactions.     

Stabilization of chloroperoxidase by polyethylene glycols in aqueous media: kinetic studies and synthetic applications

Chloroperoxidase (CPO) is one of the most versatile of the heme peroxidase enzymes for synthetic applications. Despite the potential use of CPO, commercial processes have not been developed because of the low water solubility of many organic substrates of synthetic interest and the limited stability due to inactivation by H(2)O(2). CPO catalytic properties have been studied in aqueous solutions in the presence of short-chain poly(ethylene glycol)s (PEGs), and the sulfoxidation of thioanisole, as model substrate, has been investigated. The addition of PEGs allows a better substrate solubilization in the reaction mixture and the enzyme to retain more of its initial activity, with respect to pure buffer. Kinetic studies were performed to optimize the experimental conditions, and complete enantioselective conversion to the (R)-sulfoxide (ee = 99%) was observed in the presence of PEG 200 and tri(ethylene glycol). The relevant stabilization of chloroperoxidase due to the presence of PEGs allows the enzyme to convert the substrate with significant product yields even after 10 days, with a consequent increase in enzyme productivity. This is a promising result in view of industrial application of the enzyme.     

Use of stable emulsion to improve stability, activity, and enantioselectivity of lipase immobilized in a membrane reactor

The enantiocatalytic performance of immobilized lipase in an emulsion membrane reactor using stable emulsion prepared by membrane emulsification technology was studied. The production of optical pure (S)-naproxen from racemic naproxen methyl ester was used as a model reaction system. The O/W emulsion, containing the substrate in the organic phase, was fed to the enzyme membrane reactor from shell-to-lumen. The enzyme was immobilized in the sponge layer (shell side) of capillary polyamide membrane with 50 kDa cut-off. The aqueous phase was able to permeate through the membrane while the microemulsion was retained by the thin selective layer. Therefore, the substrate was kept in the enzyme-loaded membrane while the water-soluble product was continuously removed from the reaction site. The results show that lipase maintained stable activity during the entire operation time (more than 250 h), showing an enantiomeric excess (96 +/- 2%) comparable to the free enzyme (98 +/- 1%) and much higher compared to similar lipase-loaded membrane reactors used in two-separate phase systems (90%). The results demonstrate that immobilized enzymes can achieve high stability as well as high catalytic activity and enantioselectivity.     

Detoxification of sulfur mustard by enzyme-catalyzed oxidation using chloroperoxidase

One of the most interesting methods for the detoxification of sulfur mustard is enzyme-catalyzed oxidation. This study examined the oxidative destruction of a sulfur mustard by the enzyme chloroperoxidase (EC 1.11.1.10). Chloroperoxidase (CPO) belongs to a group of enzymes that catalyze the oxidation of various organic compounds by peroxide in the presence of a halide ion. The enzymatic oxidation reaction is affected by several factors: pH, presence or absence of chloride ion, temperature, the concentrations of hydrogen peroxide and enzyme and aqueous solubility of the substrate. The optimum reaction conditions were determined by analyzing the effects of all factors, and the following conditions were selected: solvent, Britton–Robinson buffer (pH = 3) with tert-butanol (70:30 v/v); CPO concentration, 16 U/mL; hydrogen peroxide concentration, 40 mmol/L; sodium chloride concentration, 20 mmol/L. Under these reaction conditions, the rate constant for the reaction is 0.006 s⁻¹. The Michaelis constant, a measure of the affinity of an enzyme for a particular substrate, is 1.87 × 10⁻³ M for this system. The Michaelis constant for enzymes with a high affinity for their substrate is in the range of 10⁻⁵ to 10⁻⁴ M, so this value indicates that CPO does not have a very high affinity for sulfur mustard.     

A self-stabilized model of the chymotrypsin catalytic pocket. The energy profile of the overall catalytic cycle

A model of the catalytic triad of chymotrypsin is built assuring the arrangement and properties as they are within the complete enzyme. The model contains 18 amino acid residues of chymotrypsin and its substrate. A total of 135 atoms (including 70 heavy atoms) were subjected to full ab initio geometry optimizations through 127 individual steps along the reaction coordinate of the complete catalytic mechanism. It was shown that the described model of the catalytic apparatus forms a self-stabilized molecule ensemble without the rest of the enzyme and substrate. According to the calculations, the formations of the first and second tetrahedral intermediates in the model have 20.3 and 15.7 kcal/mol activation energy barriers, respectively. Removing elements of the catalytic apparatus such as the (1) catalytic aspartate or (2) the anion hole, as well as (3) inserting a water molecule "early" in the catalytic process, or (4) introducing conformational rigidity of the substrate, results in an increase of the above energy barrier of the first catalytic step in the model by 6.4, 13.7, 3.7, and 4.1 kcal/mol, respectively. Based on the calculated process one can conclude that the catalytic reaction in this model is much more similar to the reaction in the enzyme than to the reference reaction. To our knowledge, this is the first model system that mimics the complete catalytic mechanism.     

Role of aspartate 400, arginine 262, and arginine 401 in the catalytic mechanism of human coproporphyrinogen oxidase

Coproporphyrinogen oxidase (CPO) is the sixth enzyme in the heme biosynthetic pathway, catalyzing two sequential oxidative decarboxylations of propionate moieties on coproporphyrinogen-III forming protoporphyrinogen-IX through a monovinyl intermediate, harderoporphyrinogen. Site-directed mutagenesis studies were carried out on three invariant amino acids, aspartate 400, arginine 262, and arginine 401, to determine residue contribution to substrate binding and/or catalysis by human recombinant CPO. Kinetic analyses were performed on mutant enzymes incubated with three substrates, coproporphyrinogen-III, harderoporphyrinogen, or mesoporphyrinogen-VI, in order to determine catalytic ability to perform the first and/or second oxidative decarboxylation. When Asp400 was mutated to alanine no divinyl product was detected, but the production of a small amount of monovinyl product suggested the K(m) value for coproporphyrinogen-III did not change significantly compared to the wild-type enzyme. Upon mutation of Arg262 to alanine, CPO was again a poor catalyst for the production of a divinyl product, with a catalytic efficiency <0.01% compared to wild-type, including a 15-fold higher K(m) for coproporphyrinogen-III. The efficiency of divinyl product formation for mutant enzyme Arg401Ala was approximately 3% compared to wild-type CPO, with a threefold increase in the K(m) value for coproporphyrinogen-III. These data suggest Asp400, Arg262, and Arg401 are active site amino acids critical for substrate binding and/or catalysis. Possible roles for arginine 262 and 401 include coordination of carboxylate groups of coproporphyrinogen-III, while aspartate 400 may initiate deprotonation of substrate, resulting in an oxidative decarboxylation.     

Novel interface-binding chloroperoxidase for interfacial epoxidation of styrene

A unique interface-binding chloroperoxidase (CPO) was developed and examined for interfacial biocatalysis. Native CPO was conjugated with polystyrene (PS) to form a surfactant-like structure that self assembled at oil-water interfaces. While enantioselectivity of the enzyme was maintained, the interfacial assembly of the enzyme improved its overall catalytic efficiency as compared to that observed with the enzyme contained in the bulk aqueous phase. The PS conjugated CPO (PS-CPO) showed a 2.5-fold enhancement of enzyme productivity versus native CPO under batch reaction conditions for the epoxidation of styrene, whereas a 25-fold improvement was realized in a continuous feeding reaction to reach a productivity of 10 micromol h-1 mg protein-1. The interface-binding enzyme also demonstrated several other advantages such as suppressing unwanted side reactions including the hydrolysis of styrene epoxide products, stabilizing the enzyme by limiting its exposure to both the oxidant H2O2 and epoxide products, and alleviating the deactivating effect of interfacial stress on enzymes by functioning as a surfactant.     

Polyoxometalates as peroxidase mimetics and their applications in H2O2 and glucose detection

Polyoxometalates (H(3)PW(12)O(40), H(4)SiW(12)O(40) and H(3)PMo(12)O(40)) have been proven to possess intrinsic peroxidase-like activity for the first time, which can catalyze oxidation of the peroxidase substrate 3,3',5,5'-tetramethylbenzidine (TMB) by H(2)O(2) to form a blue color in aqueous solution. Among them, H(3)PW(12)O(40) (PW(12)) exhibits higher catalytic activity to TMB than natural enzyme HRP and other two POMs. In addition, H(3)PW(12)O(40)/graphene exhibited higher activity than H(3)PW(12)O(40) in this catalytic oxidation reaction due to the effect of graphene in promoting the electron transfer between the substrate and catalyst. POMs/H(2)O(2)/TMB system provides a simple, accurate approach to colorimetric detection for H(2)O(2) or glucose. The colorimetric method based on POMs showed good response toward H(2)O(2) and glucose detection with a linear range from 1.34×10(-7) to 6.7×10(-5) mol/L and 1×10(-7) to 1×10(-4) mol/L, respectively. The results showed that it is a simple, cheap, more convenient, highly selective, sensitive, and easy handling colorimetric assay.     

A novel reactor concept for the enzymatic reduction of poorly soluble ketones

Reductions of poorly soluble ketones often suffer from low total turnover numbers conferring to the coenzyme and large volumes which are needed for the conversion. The novel emulsion membrane reactor overcomes these limitations. From an emulsion consisting of an organic substrate and an aqueous buffer phase, the aqueous phase is separated selectively by using a hydrophilic ultrafiltration membrane and fed to a subsequent enzyme membrane reactor. The product outflow is recirculated to the emulsion stirred vessel and, due to the partition coefficients, the aqueous phase is recharged with substrate while the product is extracted. This new reactor concept will be compared to the classical enzyme membrane reactor. The latter was operated under the same conditions over a period of 4 months at a space-time yield of 21.2 g l⁻¹ day⁻¹. As a model system the enantioselective reduction of 2-octanone to (S)-2-octanol (ee > 99.5%) is used, carried out by a carbonyl reductase from Candida parapsilosis. NADH is regenerated by formate dehydrogenase from Candida boidinii. In comparison to the classical enzyme membrane reactor the total turnover number could be increased by a factor 9 using the novel emulsion membrane reactor.     

Spectrofluorometric determination of lipase activity.

A fluorescent triglyceride, 1(3)-pyrenylbutanoyl-2,3(1,2)-dipalmitoyl-sn-glycerol, was synthesised, characterised by NMR and mass spectrometry, incorporated into a lipid emulsion and used as a fluorescent substrate for pancreatic lipase. It is shown that the product of the reaction, pyrene butyric acid, diffuses into the aqueous phase resulting in a decrease in the excimer fluorescence of the pyrene fluorophore in the emulsion and an increase in its monomer fluorescence. The phenomenon can be used to assay the enzyme thereby cirumventing the need to extract the fatty acid product.     

The action of plasma amine oxidase on beta-haloamines. Evidence for proton abstraction in the oxidative reaction.

The action of plasma amine oxidase upon beta-Br-ethylamine beta-Cl-ethylamine, beta-OH-phenylethylamine, and beta-Cl-phenylethylamine was examined. Beta-Br-ethylamine is a substrate and irreversible inactivator of the enzyme. The enzyme becomes covalently labeled by the inactivator. Approximately 2 mol of inactivator are incorporated per mol of enzyme (MW 170,000). The reduced enzyme is not inactivated. The enzyme catalyzes the elimination of HCl from beta-Cl-phenylethylamine to produce phenylacetaldehyde. The rate of the elimination reaction is comparable to the normal oxidative reaction. We conclude that the occurrence of this elimination reaction establishes the ability of the enzyme to catalyze proton abstraction from C-1 of the substrate and that proton abstraction occurs during the catalytic oxidation normally catalyzed by plasma amine oxidase. Beta-Cl-ethylamine is only oxidized to corresponding aldehyde. Beta-OH-phenylethylamine is neither oxidized, nor does elimination occur. It is a competitive inhibitor in the oxidation of benzylamine and in the elimination of HCl from beta-Cl-phenylethylamine.     

Structures of Michaelis and product complexes of plant cytokinin dehydrogenase: implications for flavoenzyme catalysis

Cytokinins form a diverse class of compounds that are essential for plant growth. Cytokinin dehydrogenase has a major role in the control of the levels of these plant hormones by catalysing their irreversible oxidation. The crystal structure of Zea mays cytokinin dehydrogenase displays the same two-domain topology of the flavoenzymes of the vanillyl-alcohol oxidase family but its active site cannot be related to that of any other family member. The X-ray analysis reveals a bipartite architecture of the catalytic centre, which consists of a funnel-shaped region on the protein surface and an internal cavity lined by the flavin ring. A pore with diameter of about 4A connects the two active-site regions. Snapshots of two critical steps along the reaction cycle were obtained through the structural analysis of the complexes with a slowly reacting substrate and the reaction product, which correspond to the states immediately before (Michaelis complex) and after (product complex) oxidation has taken place. The substrate displays a "plug-into-socket" binding mode that seals the catalytic site and precisely positions the carbon atom undergoing oxidation in close contact with the reactive locus of the flavin. A polarising H-bond between the substrate amine group and an Asp-Glu pair may facilitate oxidation. Substrate to product conversion results in small atomic movements, which lead to a planar conformation of the reaction product allowing double-bond conjugation. These features in the mechanism of amine recognition and oxidation differ from those observed in other flavin-dependent amine oxidases.     

Variable stoichiometry during the laccase-catalyzed oxidation of aqueous phenol

The oxidation of aqueous phenol through the catalytic action of laccase from Trametes versicolor was studied over a wide range of phenol concentrations and enzyme activities. The stoichiometric ratio, which is defined as the molar ratio of phenol transformed to oxygen consumed in the catalytic reaction, was found to increase with phenol concentration in the reaction mixture from a theoretical lower limit of 1 and to approach a theoretical upper limit of 4. A logistic equation was proposed to relate reaction stoichiometry to substrate concentration and was successfully used to relate these parameters over a range of phenol concentrations extending from approximately 0.15 to 8 mM. This expression was incorporated into two kinetic models in order to account for variations in reaction stoichiometry during the reaction and to extend the range over which the models may be accurately applied. The new models demonstrated an improved ability to predict concentrations of phenol and oxygen over time in a closed batch reaction system.     

Isolation and characterization of a novel oxygenase that catalyzes the first step of n-alkane oxidation in Acinetobacter sp. strain M-1.

In the Finnerty pathway for n-alkane, oxidation in Acinetobacter sp., n-alkanes are postulated to be attacked by a dioxygenase and the product, n-alkyl hydroperoxide, is further metabolized to the corresponding aldehyde via the peroxy acid [W. R. Finnerty, P. 184-188, in A. H. Applewhite (ed.), Proceedings of the World Conference on Biotechnology for the Fats and Oil Industry, 1988]. However, no biochemical evidence regarding the first-step reaction is available. In this study, we found a novel n-alkane-oxidizing enzyme that requires only molecular oxygen, i.e., not NAD(P)H, in our isolate, Acinetobacter sp. strain M-1, and purified it to apparent homogeneity by gel electrophoresis. The purified enzyme is a homodimeric protein with a molecular mass of 134 kDa, contains 1 mol of flavin adenine dinucleotide per mol of subunit, and requires CU2+ for its activity. The enzyme uses n-alkanes ranging in length from 10 to 30 carbon atoms and is also active toward n-alkenes (C12 to C20) and some aromatic compounds with substituted alkyl groups but not toward a branched alkane, alcohol, or aldehyde. Transient accumulation of n-alkyl hydroperoxide was detected in the course of the reaction, and no oxygen radical scavengers affected the enzyme activity. From these properties, the enzyme is most probably a dioxygenase that catalyzes the introduction of two atoms of oxygen to the substrate, leading to the formation of the corresponding n-alkyl hydroperoxide. The enzymatic evidence strongly supports the existence of an n-alkane oxidation pathway, which is initiated by a dioxygenase reaction, in Acinetobacter spp.     

Normal and abnormal heme biosynthesis. Part 7. Synthesis and metabolism of coproporphyrinogen-III analogues with acetate or butyrate side chains on rings C and D. Development of a modified model for the active site of coproporphyrinogen oxidase

Analogues of coproporphyrinogen-III have been prepared with acetate or butyrate groups attached to the C and D pyrrolic subunits. The corresponding porphyrin methyl esters were synthesized by first generating a,c-biladienes by reacting a dipyrrylmethane with pyrrole aldehydes in the presence of HBr. Cyclization with copper(II) chloride in DMF, followed by demetalation with 15% H(2)SO(4)-TFA and reesterification, gave the required porphyrins in excellent yields. Hydrolysis with 25% hydrochloric acid and reduction with sodium-amalgam gave novel diacetate and dibutyrate porphyrinogens 9. Diacetate 9a was incubated with chicken red cell hemolysates (CRH), but gave complex results due to the combined action of two of the enzymes present in these preparations. Separation of uroporphyrinogen decarboxylase (URO-D) from coproporphyrinogen oxidase (CPO) allowed the effects of both enzymes on the diacetate substrate to be assessed. Porphyrinogen 9a proved to be a relatively poor substrate for CPO compared to the natural substrate coproporphyrinogen-III, and only the A ring propionate moiety was processed to a significant extent. Similar results were obtained for incubations of 9a with purified human recombinant CPO. Diacetate 9a was also a substrate for URO-D and a porphyrinogen monoacetate was the major product in this case; however, some conversion of a second acetate unit was also evident. The dibutyrate porphyrinogen 9b was only recognized by the enzyme CPO, but proved to be a modest substrate for incubations with CRH. However, 9b was an excellent substrate for purified human recombinant CPO. The major product for these incubations was a monovinylporphyrinogen, but some divinyl product was also generated in incubations using purified recombinant human CPO. The incubation products were converted into the corresponding porphyrin methyl esters, and these were characterized by proton NMR spectroscopy and mass spectrometry. The results extend our understanding of substrate recognition and catalysis for this intriguing enzyme and have allowed us to extend the active site model for CPO. In addition, the competitive action of both URO-D and CPO on the same diacetate porphyrinogen substrate provides additional perspectives on the potential existence of abnormal pathways for heme biosynthesis.     

P450cam biocatalysis in surfactant-stabilized two-phase emulsions

Cytochrome P450 monooxygenases (P450s) are powerful biocatalysts that have the ability to oxidize a broad range of substrates, often at non-reactive carbon centers. However, incorporation of P450s into synthetic schemes has so far been limited to a few whole-cell transformations. P450 substrates are often hydrophobic and have low water solubility, limiting the amount of product that can be produced. To help overcome this limitation, we have examined P450cam activity in two-phase hexane/water emulsions with and without the anionic surfactant, bis(2-ethylhexyl) sulfosuccinate sodium salt (AOT). Hydroxylation of camphor to hydroxycamphor by the three- component P450cam system was chosen as the model reaction, and regeneration of NADH was accomplished with yeast alcohol dehydrogenase (YADH). P450cam was activated in the surfactant-free emulsions, and addition of AOT improved the activity even further, at least over the range of camphor concentrations for which initial rates were readily measurable in all media. The largest observed rate enhancement was 4.5-fold. Nearly 50-times more product was formed in the surfactant-stabilized emulsions than was achieved in aqueous buffer, with total turnover numbers reaching 28,900 for P450cam and 11,800 for YADH. In the absence of surfactant, the two-phase reaction appeared to be mass-transfer limited, while inclusion of AOT alleviated transport limitations and/or afforded a larger interfacial area for P450 activation. The oxidation of hydroxycamphor to 2,5-diketocamphane was also observed, owing to the large concentration of hydroxycamphor relative to camphor in the aqueous phase of the two-phase emulsion. This competing reaction was accompanied by the uncoupled oxidation of NADH (i.e., NADH oxidation without formation of 2,5-diketocamphane), which reduced the availability of NADH for camphor oxidation and further limited the yield of hydroxycamphor in the two-phase emulsions. These results indicate that a surfactant-stabilized two-phase emulsion is a promising reaction medium for practical P450 biocatalysis, although its effectiveness for a given P450/substrate combination can depend on several factors, including competitive or sequential reactions, product inhibition, and NAD(P)H uncoupling.     

Computational modeling of the catalytic reaction in triosephosphate isomerase

We present a comprehensive analysis of the catalytic cycle of the enzyme triosephosphate isomerase (TIM), including both the reactive chemistry and the catalytic loop and side-chain motions. Combining accurate mixed quantum mechanics/molecular mechanics (QM/MM) and protein structure prediction methods, we have modeled both the structural and chemical aspects of the reversible isomerization of dihydroxyacetone phosphate (DHAP) to d-glyceraldehyde 3-phosphate (GAP), for which there is a wealth of experimental data. The conjunction of this novel computational approach with the use of the recent near-atomic resolution TIM-DHAP Michaelis complex PDB structure, 1NEY.pdb, has enabled us to obtain robust qualitative and, where available, quantitative agreement with a wide range of experimental data. Among the principal conclusions that we are able to draw are the importance of the monoanionic (as opposed to dianioic) form of the substrate phosphate group in the catalytic cycle, detailed positioning and energetics of the key catalytic residues in the active-site, the flexible nature of Glu165, which favors its direct involvement in the formation of the enediol intermediate, energetics of the open and closed form of the catalytic loop region in the presence and absence of substrate, and quantitative reproduction of various experimentally measured reaction rates, typically to within approximately 1 kcal/mol. Our results are consistent with the available experimental data, and provide an initial picture as to why loop opening when GAP is the product has a higher barrier than when DHAP is the product.     

Mathematical analysis of two-phase mass transfer in a batch reactor for the chemical transformation of a steroid

A reactor is described for the conversion of the slightly water-soluble steroid testosterone (T) to 4-androstene-3, 17-dione (4-AD) by enzyme in the presence of excess cofactor. Since the enzyme is subject to substrate inhibition, reaction rates are strong functions of aqueous substrate concentration. High concentrations of the substrate, testosterone, per unit reactor volume are maintained within poly(dimethylsiloxane) beads that are suspended in the aqueous enzyme solution. Mass transfer (controlled by bead size, polymer to water volume ratio, enzyme loading) is used to control the degree and rate of conversion. The reactor dynamics are predicted over a wide range of reaction conditions. The product steroid is recovered in the polymeric beads from the enzyme solution.