Synthesis of 3-tert-Butylcatechol by anEngineered Monooxygenase

Andreas Meyer,1 Martin Held,1 Andreas Schmid,1 Hans-Peter E. Kohler,2

Bernard Witholt1

1Institute of Biotechnology, ETHZ, Swiss Federal Institute of Technology,ETH Honggerberg, HPT, CH-8093, Zurich, Switzerland; telephone: +41 16333286; fax: +41 1 6331051; e-mail: bw@biotech.biol.ethz.ch2Environmental Microbiology and Molecular Ecotoxicology, EAWAG,Swiss Federal Institute of Environmental Sciences and Technology,CH-8600 Dubendorf, Switzerland

Received 30 April 2002; accepted 15 July 2002

DOI: 10.1002/bit.10487

Abstract: Recombinant Escherichia coli JM101 was usedfor the in vivo biocatalytic synthesis of 3-tert-butyl-catechol. The bacterial strain synthesized the laboratory-evolved variant HbpAT2 of 2-hydroxybiphenyl 3-mono-oxygenase (HbpA, EC 1.14.13.44) from Pseudomonas az-elaica HBP1. The mutant enzyme HbpAT2 is able tohydroxylate 2-tert-butylphenol to the corresponding cat-echol, a reaction that is not catalyzed by the wild-typeenzyme. The biotransformation was performed in a 3-Lbioreactor for 24 h. To mitigate the toxicity of the 2-tert-butylphenol starting material, we applied a limited sub-strate feed. Continuous in situ product removal with thehydrophobic resin Amberlite� XAD-4 was used to sepa-rate the product from culture broth. In addition, bindingto the resin stabilized the product, which was importantbecause 3-tert-butylcatechol is very labile in aqueous so-lution. The productivity of the process was 63 mg L−1 h−1

so that after 24 h, 3.0 g of 3-tert-butylcatechol were iso-lated. Down-stream processing consisted of two steps.First, bound 2-tert-butylphenol and 3-tert-butylcatecholwere eluted from Amberlite� XAD-4 with methanol. Sec-ond, the two compounds were separated over neutralaluminum oxide, which selectively binds the producedcatechol but not the phenol substrate. The final purity of3-tert-butylcatechol was greater than 98%. © 2003 WileyPeriodicals, Inc. Biotechnol Bioeng 81: 518–524, 2003.Keywords: engineered monooxygenase; in situ productremoval; 2-hydroxybiphenyl 3-monooxygenase (EC1.14.13.44); 3-tert-butylcatechol

INTRODUCTION

Alkylated catechols are used as raw materials for the syn-thesis of pharmaceuticals, agricultural chemicals, dye de-velopers, and polymerization inhibitors (Yoo et al., 1999).The industrial route to these compounds proceeds by elec-trophilic substitution, where the metal halides AlCl3, FeCl3,and ZnCl2 serve as Lewis acid catalysts (Wade, 1995).Problems of safety and waste disposal associated with these

processes have stimulated the development of alternativeroutes. New approaches include the use of heterogeneouscatalysts such as silica gel or acidic zeolites (Kamitori et al.,1984; Yoo et al., 1999). However, these chemical reactionspreferentially yield the 4-substituted or polyalkylated iso-mers; 3-alkylcatechols are notoriously difficult to obtain(Held et al., 1998). An effective alternate route to 3-n-alkylcatechols proceeds by intramolecular cyclization of2-alkanoyl-2,5-dimethoxytetrahydrofurans with aqueousacid, but this route is not applicable to the synthesis of3-sec- or 3-tert-alkylcatechols (Miyakoshi and Togashi,1990).

Due to their high selectivity, biological catalysts offer apotent and environmentally friendly alternative to thechemical production of 3-substituted catechols (Held et al.,1999; Robinson et al., 1992). The corresponding benzenesor 2-substituted phenols, which are readily available, serveas starting materials and are transformed to the desiredproducts by regioselective oxidation with a mono- or dioxy-genase. Although this approach has been successful for thesynthesis of different alkylated, halogenated, and aryl-substituted catechols (Held et al., 1998; Johnston and Ren-ganathan, 1987; Robinson et al., 1992; Schmid et al., 1998),it has not been useful for the synthesis of 3-tert-butyl-catechol. The reason for this is the lack of available en-zymes with corresponding activities.

Tuning the activity of a biocatalyst to a selected activityis a major topic in molecular biology. Both rational designby site-directed mutagenesis and random approaches, suchas DNA shuffling or error-prone PCR followed by screen-ing for appropriate modifications (Moore and Arnold, 1996;Stemmer, 1994), have been used to change and tune thesubstrate specificity of biocatalysts (Graf et al., 1987;Schmidt-Dannert and Arnold, 1998; Shao and Arnold,1996; van den Heuvel et al., 2000; Zhao and Arnold, 1999).

Since we have been interested in developing biocatalystsfor the production of various interesting and syntheticallychallenging target catechols by biooxidation of the corre-

Correspondence to: Bernard WitholtContract grant sponsor: Swiss National Science FoundationContract grant number: 5002-046098

© 2003 Wiley Periodicals, Inc.

sponding phenols, we have used directed enzyme evolution(Meyer et al., 2002) to modify 2-hydroxybiphenyl 3-mono-oxygenase (HbpA, EC 1.14.13.44) of Pseudomonasazelaica HBP1 (Kohler et al., 1988). HbpA is a homo-tetrameric, NADH-dependent flavoprotein aromatic hy-droxylase (Suske et al., 1997). We used in vitro manganesemutagenesis to generate HbpA variants and screened forimproved monooxygenase activity on various 2-substitutedphenols by observing the formation of colors indicative ofthe autooxidation of the reaction products. We characterizedseveral such mutants (Meyer et al., 2002). One of these,which we denoted HbpAT2, contained the two amino acidsubstitutions Val368Ala and Leu417Phe. HbpAT2 has a kcat

of 0.5 s−1 with 2-tert-butylphenol as the substrate, while kcat

< 0.1 s−1 for the wild-type protein (Meyer et al., 2002). Herewe report the application of HbpAT2 for the synthesis of3-tert-butylcatechol using a recombinant whole-cell bio-catalyst. Due to the bactericidal properties of both substrateand product, an integrated process with limited starting ma-terial feed and in situ product removal was used (Held et al.,1998, 1999; Lye and Woodley, 1999).

MATERIALS AND METHODS

Chemicals and Media Components

2-tert-Butylphenol, Amberlite� XAD-4, aluminum oxide,and all other chemicals were purchased from Fluka AG(Buchs, Switzerland). Solvents were obtained from Bio-solve Ltd (Valkenswaard, The Netherlands), and compo-nents for complex media were obtained from Difco Labo-ratories (Detroit, MI).

Strains and Plasmids

Escherichia coli JM101 [F�, traD36, lacIq, �(lacZ)M15,proAB/supE, thi, �(lac-proAB)] (Yanish-Perron et al.,1985) containing plasmid pAMT2 was used as a biocatalyst.pAMT2 is a pUC18 (Yanish-Perron et al., 1985) derivativeharboring the hbpAT2 gene as a SalI/NsiI fragment undercontrol of the lacZ promoter (Meyer et al., 2002).

Media and Growth Conditions

E. coli JM101 or recombinants thereof were usually storedas stab cultures and initially grown as 5-mL LB preculture(Sambrook et al., 1989). For assays or biotransformations,such cultures were inoculated 1:100 in M9 medium (Sam-brook et al., 1989), supplemented with 0.001% (w/v) thia-min, 0.1% (v/v) MT trace element solution (Lageveen et al.,1988), 0.1 mM CaCl2, 2 mM MgSO4, and 0.5% (w/v) glu-cose or 0.5% (v/v) glycerol as carbon source. For recombi-nants 150 mg L−1 ampicillin was added to ensure pAMT2maintenance.

Cultivations were carried out at 30°C, and biomass con-centration was determined at 450 nm (Witholt, 1972).

Analytical Methods

2-tert-Butylphenol and 3-tert-butylcatechol were quantifiedwith reverse-phase HPLC (Hewlett-Packard HP 1050 TiHPLC equipped with a DAD 1040 M diode array detector)using a Hypersil ODS column (5 �m, 4.5 × 125 mm), asdescribed elsewhere (Held et al., 1999). Samples, whichwere freed from solids by centrifugation, were diluted 1:1with methanol that was acidified with 0.1% (v/v) phospho-ric acid. Elution was carried out with a flow rate of 1.5 mLmin−1 under isocratic conditions with a methanol-to-waterratio of 40:60, also containing 0.1% (v/v) H3PO4.

Biotransformation

The biotransformation experiment was performed in a 3-Lbioreactor from Bioengineering AG (Wald, Switzerland)connected to an external loop for in situ product extraction(Fig. 1). Cell growth and biotransformation were time-separated to optimize cultivation and production indepen-dently (Held et al., 1999).

Cell Cultivation

The preculture was grown overnight in M9 medium (seeGrowth Conditions Section) with 0.5% (w/v) glucose ascarbon source. Thereafter, the bioreactor containing 2 L ofM9 medium and 1% (w/v) glucose was inoculated with 50mL of preculture, and the cells were grown overnight witha stirrer speed of 1,500 rpm and aeration by pressurized air(1 L min−1). Two hours prior to biotransformation experi-ments, the culture was pulsed with 0.5% (v/v) glycerol and150 mg L−1 ampicillin and the stirrer speed was increased to3,000 rpm.

Preparation of Product Extraction Module

Due to the bactericidal properties of substrate and productand the instability of the formed 3-tert-butylcatechol, the

Figure 1. Schematic representation of the experimental set-up. A productrecovery loop was filled with Amberlite� XAD-4 and connected to a 3-Lbioreactor. Substrate was added by a high-precision pump (P1), and thefermentation broth was circulated (P2) through the extraction module.

MEYER ET AL.: BIOCATALYTIC SYNTHESIS OF 3-TERT-BUTYLCATECHOL 519

biotransformation was carried out with limited substratefeed and in situ product removal (Held et al., 1999). Theextraction module consisted of a cylindrical external loop(diameter 8 cm, length 18 cm) filled with 390 g of Amber-lite� XAD-4, a polystyrene-based hydrophobic resin. Theloop and the tubing were sterilized by flushing with 70%ethanol for 24 h. After being washed with sterile water, themodule was filled with sterile M9 medium and connected tothe reactor.

Biotransformation

One hour before starting the biotransformation the contentof the bioreactor was circulated (Fig. 1, P2) through theexternal loop with a flux of 400 mL min−1. Thereafter,2-tert-butylphenol was fed (Fig. 1, P1) at a rate of 0.15 gL−1 h−1 from a 2 M stock solution in methanol. When thedissolved oxygen concentration in the bioreactor increasedabove 80% DOT, indicating carbon source depletion, 50 mLof 30% (v/v) glycerol were added.

Product Purification

Elution from Amberlite� XAD-4

After biotransformation, the extraction loop was removedfrom the reactor and the Amberlite� XAD-4 washed withdistilled water. Subsequently, the resin was transferred to aglass column and bound compounds were eluted withmethanol (0.1% HCl) and collected in 20 fractions of 100mL. Acidification to pH 2 was necessary to stabilize theformed catechol (Held et al., 1999). After HPLC analysis,fractions containing product were pooled and filteredthrough a 2-�m nitrocellulose filter. Finally, the methanolwas removed in a rotary evaporator.

Reverse-Phase Chromatography

Reverse-phase chromatography was performed by prepara-tive HPLC (Hewlett-Packard HP 1050Ti) with C8-silica assolid phase and 40% MeOH (0.1% trifluoroacetic acid) asmobile phase. Eluted compounds were detected with a di-ode array detector (Hewlett-Packard HP DAD 1040M).

Aluminum Oxide Column

The fact that aluminum oxide efficiently binds catechols butnot phenols (Pras et al., 1990; Schmid et al., 2001) was usedto separate 3-tert-butylcatechol and 2-tert-butylphenol. Thebinding capacity was evaluated with two model compounds.Per gram of aluminum oxide, 91 �mol of catechol (Mw 110)and 107 �mol, 2,3-dihydroxybiphenyl (Mw 186) werebound. The capacity for 3-tert-butylcatechol was thereforeassumed to be approximately 100 �mol g−1.

A glass column was filled with activated aluminum oxide(type 507C neutral, 100–125 mesh) and washed with hex-

ane. An aliquot of the compounds eluted from XAD-4 wasloaded onto the column bed, and unbound substances werewashed out with methanol. 3-tert-Butylcatechol was elutedwith 10% 5 M HCl in methanol. Fractions of 15 mL werecollected and analyzed by reverse-phase HPLC.

RESULTS

Growth Inhibition of E. coli JM101by 2-tert-Butylphenol

Substituted phenols generally have antimicrobial properties(Davidson and Brandon, 1981). The effect of 2-tert-butyl-phenol on the growth of E. coli JM101 (pUC18) was de-termined in 250-mL shaking flasks filled with 50 mL of M9medium and 0.5% (w/v) glucose as carbon source. The cellswere inoculated to a biomass concentration of 20 mg L−1.After 6 h the cultures had reached the early exponentialphase and 2-tert-butylphenol was added in concentrationsranging from 0 to 1.5 mM. Significant growth inhibitionwas observed at all concentrations above 0.5 mM (Fig. 2).When the added 2-tert-butylphenol exceeded 1 mM, partialcell lysis could be observed and substantially lower cellconcentrations were reached after incubation overnight. Atconcentrations above 1.5 mM, no further growth was ob-served.

HbpAT2 Activity in the Presenceof 2-tert-Butylphenol

The formation of 3-tert-butylcatechol by E. coli JM101(pAMT2) was determined in 250-mL shaking flasks with 50mL of M9 medium and 0.5% (w/v) glucose as carbonsource. HbpAT2 synthesis was induced in the early expo-nential phase by the addition of 0.2 mM IPTG. Two hourslater 2-tert-butylphenol was added in concentrations of0–1.5 mM, and the cultures were incubated overnight at

Figure 2. Growth of E. coli JM101 (pUC18) in the presence of differentconcentrations of 2-tert-butylphenol. Biomass concentration was deter-mined spectrophotometrically at 450 nm (Witholt, 1972). The arrow indi-cates the addition of 2-tert-butylphenol (2TBP) in concentrations of 0 (�),0.1 (�), 0.5 (�), 1 (�), and 1.5 mM (�).

520 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 81, NO. 5, MARCH 5, 2003

30°C. The instability of the formed 3-tert-butylcatechol,which forms a dark polymer, was used to identify active cellcultures. Product was formed in cultures with substrate con-centrations up to 1 mM (Fig. 3).

Biotransformation of 2-tert-Butylphenolto 3-tert-Butylcatechol

E. coli JM101 (pAMT2) was grown overnight on M9 me-dium with 1% glucose as carbon source as described in theMaterials and Methods section. This led to a final biomassconcentration of 3.2 g CDW L−1 and no significant hbpAT2

expression. After being pulsed with 0.5% glycerol, the dis-solved oxygen tension (DOT) of the culture decreased im-mediately, indicating a viable and growing cell culture. Inaddition, changing the carbon source induced hbpAT2 ex-pression, indicating that catabolite repression of the lacZpromoter was reduced or eliminated. The content of thereactor was pumped through the extraction module, and 1 hlater the substrate feed was switched on. The amount of2-tert-butylphenol and 3-tert-butylcatechol in the reactionvessel was analyzed by HPLC and never exceeded 6 mg L−1

for the substrate and 3 mg L−1 for the product. Neithersubstrate nor product was detected in the reflux from theexternal loop, indicating complete extraction of these com-pounds. After 24 h, a total of 7.5 g of 2-tert-butylphenol hadbeen added and the substrate feed was switched off. Theculture broth was circulated through the extraction loop foranother 10 min. At the end of the biotransformation thebiomass concentration was 7.6 g CDW L−1.

3-tert-Butylcatechol Recovery From XAD-4

Amberlite� XAD-4 has no effect on the growth of recom-binant E. coli JM101 or on HbpAT2 activity and can effi-ciently be used for the adsorption of hydrophobic com-pounds from culture broth (Held, 2000). After biotransfor-mation, the extraction loop was disconnected from thereactor, the resin was washed with water, and the boundcompounds were eluted with 2 L of methanol (0.1% HCl).In total, 4.6 g of 2-tert-butylphenol and 3.0 g of 3-tert-butylcatechol were collected, which corresponds to a con-version yield of 36.8% and a product space-time yield of

0.063 g L−1 h−1. Substrate and product were present insimilar ratios in all fractions (Fig. 4). The mass balancerevealed a molar recovery of 97% (substrate plus product),which indicated that no significant product polymerizationtook place. After filtration and methanol removal, 45 mL ofbrownish oil remained.

Separation of 3-tert-ButylcatecholFrom 2-tert-Butylphenol

To separate remaining substrate and formed product, twodifferent approaches were used: reverse-phase chromatog-raphy and separation on neutral aluminum oxide (pH 7).

Reverse-phase Chromatography

A 1-mL sample containing 40 g L−1 3-tert-butylcatecholand 61 g L−1 2-tert-butylphenol was loaded and eluted un-der isocratic conditions. Retention times were 18–25 minfor the catechol and more than 28 min for the phenol. All the3-tert-butylcatechol was collected in a volume of 35 mL.

Separation on Aluminum Oxide

A 2-mL sample containing 80 g L−1 3-tert-butylcatecholand 122 g L−1 2-tert-butylphenol was loaded onto neutralaluminum oxide. The column was flushed with methanol,and fractions of 15 mL were collected. 2-tert-Butylphenolwas washed out and was present in fractions 2–5 (Fig. 5).When no more phenol was detected in the effluent, thesolvent was changed to 10% 5 M HCl in methanol. The3-tert-butylcatechol that subsequently came off the columnwas collected in fractions 7–11 (Fig. 5). The mass balancerevealed a 3-tert-butylcatechol yield of 84%. Purity waschecked with reverse-phase HPLC and determined to begreater than 98%.

Figure 3. Cultures of E. coli JM101 (pAMT2) grown in the presence ofdifferent concentrations of 2-tert-butylphenol. The dark color derives fromthe polymerization of 3-tert-butylcatechol, which is formed by active cul-tures.

Figure 4. Elution of 2-tert-butylphenol and 3-tert-butylcatechol fromXAD-4. After biotransformation the formed 3-tert-butylcatechol and theremaining substrate were eluted from the hydrophobic resin AmberLite�

XAD-4 (390 g) and collected in 20 fractions of 100 mL. 2TBP, 2-tert-butylphenol; 3TBC, 3-tert-butylcatechol.

MEYER ET AL.: BIOCATALYTIC SYNTHESIS OF 3-TERT-BUTYLCATECHOL 521

3-tert-Butylcatechol Authentication

For verification of the authenticity of 3-tert-butylcatechol(Fig. 6) 1H- and 13C-NMR spectra were recorded. For themeasurements a sample of the HPLC-purified product wastaken. 1H-NMR (CDCl3): 6.81 (d, Cf-H); 6.98 (dd, Ce-H);6.82 (d, Cd-H); 1.52 (s, Ch-H); 5.01 (br, 1 OH); 5.70 (br, 1OH). 13C-NMR (CDCl3): 137.1 (Ca); 143.2 (Cb); 143.8(Cc); 113.4 (Cd); 119.5 (Ce); 119.8 (Cf); 35.0 (Cg); 29.9(Ch). The measured spectra are in agreement with earlierdata (Kamitori et al., 1984; Yoo et al., 1999).

DISCUSSION

Advances in protein engineering, especially in randomapproaches such as directed enzyme evolution, significantlyenlarge the scope of biocatalytic synthesis. In this study, wedescribed the gram-scale production of 3-tert-butylcatecholby a laboratory-evolved monooxygenase.

Experimental Set-Up

Phenolic compounds permeabilize bacterial membranes(Davidson and Brandon, 1981). This explains the antimi-

crobial properties of such compounds. The addition of2-tert-butylphenol in concentrations of 1.5 mM was toxicfor E. coli JM101 cells and led to complete cell lysis. There-fore, a limited substrate feed was necessary for bioconver-sion of this substrate. An alternative process, based on theapplication of cell-free preparations of the mutant monoox-ygenase with an enzymatic or electrochemical regenerationsystem for NADH could have been developed (Hollmann etal., 2001; Kragl et al., 1996; Kula and Wandrey, 1987).Such approaches have been used successfully with dehy-drogenases in enzyme membrane reactors. However, theapplication of oxygenases, which constitute a more complexclass of enzymes, has led to only a few practical examplesdue to the limited stability of the enzymes under in vitroprocess conditions (Duetz et al., 2001). In contrast, recom-binant E. coli JM101 synthesizing HbpAT2 showed in vivoactivity for up to 24 h without any loss. The instability of theformed 3-tert-butylcatechol at neutral pH required the in-stant removal of the product from culture broth. This wasachieved with Amberlite� XAD-4, which has already beenused for in situ product recovery of several substituted cat-echols (Held et al., 1998, 1999).

3-tert-Butylcatechol Synthesis

The productivity of the process with the mutant 2-hydroxy-biphenyl 3-monooxygenase was only 6-fold lower than thatof the corresponding optimized biotransformation for theproduction of 2,3-dihydroxybiphenyl with the wild-type en-zyme (Held et al., 1999). Considering that the specific ac-tivity of purified HbpAT2 with 2-tert-butylphenol as thesubstrate is only 4% of the value of HbpA with the naturalsubstrate, the obtained productivity is unexpectedly highand significantly higher than for chemical catalysis (Kami-tori et al., 1984; Yoo et al., 1999). We expect that biocata-lyst engineering and process optimization could increase theproductivity further. The yield for conversion of 2-tert-butylphenol was determined to be 36.8% and could easilybe raised by reducing the substrate feed rate. For synthesisof 2,3-dihydroxybiphenyl, yields of up to 97% could bereached in this way. Yields for chemical catalysis wereconsiderably lower. Synthesis with silica gel as the catalystresulted in conversions of 8% (Kamitori et al., 1984), whilein the best reaction set-up with acidic zeolites as the cata-lyst, yields of 3.1% were obtained (Yoo et al., 1999). Areason for the low yields of chemical synthesis is the for-mation of side products. The hydroxyl groups of catecholpolarize the benzene ring, resulting in a higher electrondensity at carbon C3/6 compared to C4/5. As a consequencethe main product of chemical catechol alkylation is the4-substituted isomer (Kamitori et al., 1984; Yoo et al.,1999).

Side-product formation was not observed for the biocata-lytic production of 3-tert-butylcatechol. As is true for thewild-type enzyme, HbpAT2 hydroxylates 2-substituted phe-nols in an absolutely regioselective manner. This fact wasalso useful for the development of easy downstream pro-

Figure 5. Separation of 2-tert-butylphenol and 3-tert-butylcatechol overneutral aluminum oxide. 2TBP, 2-tert-butylphenol; 3TBC, 3-tert-butylcatechol.

Figure 6. 3-tert-Butylcatechol. Chemical structure of 3-tert-butyl-catechol with assigned nomenclature for 1H- and 13C-NMR spectra.

522 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 81, NO. 5, MARCH 5, 2003

cessing based on the catechol functionality of the synthe-sized product.

Product Purification

Adsorption of the formed 3-tert-butylcatechol on the hydro-phobic resin Amberlite� XAD-4 during the biotransforma-tion was the first step of the downstream processing. Bothseparation from culture broth and stabilization wereachieved and after elution more than 94% of the 3-tert-butylcatechol was recovered in 1 liter of methanol. In com-parison to a two-liquid-phase process of the same scale, inwhich a second organic phase (e.g., octane) serves as theextraction medium, the product-containing solvents to beprocessed have similar volumes. However, stabilization ofthe produced catechol occurred only when the product wasimmobilized on a solid matrix.

The separation of 3-tert-butylcatechol from 2-tert-butyl-phenol by reverse-phase HPLC led to a highly pure productpreparation, as could be seen from the 1H- and 13C-NMRspectra. However, HPLC methods are of limited capacityand therefore difficult to apply on large scales. A betteroption is the separation of the product on aluminum oxide,which selectively binds catechols but not phenols. The ca-pacity of the matrix is determined by catechol functional-ities rather than by molecular weight as shown by immobi-lization experiments with model compounds. Because theamount absorbed at pH 7 is about 4 times higher than at pH4 (Held, 2000) the separation was carried out at neutral pH.As a consequence, part of the 3-tert-butylcatechol polymer-ized (as indicated by brownish color formation on the col-umn) and was not eluted. This resulted in a final yield of84% for product purification. The small amount of poly-merization was deemed acceptable compared to the benefitsof the reduced column volume.

In this article we have shown that the application of amodified 2-hydroxybiphenyl 3-monooxygenase (HbpAT2),in combination with in situ product recovery and productpurification over aluminum oxide, allows the efficient syn-thesis of 3-tert-butylcatechol. The approach described hereprovides a general procedure for the production of a broadvariety of 3-substituted catechols by engineered HbpA.

We thank Carsten Boehler (ETH Zurich) for recording the NMRspectra.

References

Davidson PM, Brandon AL. 1981. Antimicrobial activity of non-halogenated phenolic compounds. J Food Prot 44:623–634.

Duetz WA, van Beilen JB, Witholt B. 2001. Using proteins in their naturalenvironment: potential and limitations of microbial whole-cell hydrox-ylations in applied catalysis. Curr Opin Biotechnol 12:419–425.

Graf L, Craik CS, Patthy A, Roczniak S, Fletterick RJ, Rutter WJ. 1987.Selective alteration of substrate specificity by replacement of asparticacid-189 with lysine in the binding pocket of trypsin. Biochemistry26:2616–2623.

Held M. 2000. Biocatalytic synthesis of 3-substituted catechols. Ph.D.thesis. Zurich, Switzerland: ETH Zurich.

Held M, Schmid A, Kohler H-PE, Suske W, Witholt B, Wubbolts MG.1999. An integrated process for the production of toxic catechols fromtoxic phenols based on a designer biocatalyst. Biotechnol Bioeng 62:641–648.

Held M, Suske W, Schmid A, Engesser K-H, Kohler H-PE, Witholt B,Wubbolts MG. 1998. Preparative scale production of 3-substitutedcatechols using a novel monooxygenase from Pseudomonas azelaicaHBP1. J Mol Catal B 5:87–93.

Hollmann F, Schmid A, Steckhan E. 2001. The first synthetic applicationof a monooxygenase employing indirect electrochemical NADH re-generation. Angew Chem Int Ed 40:169–171.

Johnston JB, Renganathan V. 1987. Production of substituted catecholsfrom substituted benzenes by Pseudomonas sp. Enzyme Microb Tech-nol 9:706–708.

Kamitori Y, Hojo M, Masuda R, Izumi T, Tsukamoto S. 1984. Silica gelas an effective catalyst for the alkylation of phenols and some hetero-cyclic aromatic compounds. J Org Chem 49:4161–4165.

Kohler H-PE, Kohler-Staub D, Focht DD. 1988. Degradation of 2-hy-droxybiphenyl and 2,2�-dihydroxybiphenyl by Pseudomonas sp. strainHBP1. Appl Environ Microbiol 54:2683–2688.

Kragl U, Kruse W, Hummel W, Wandrey C. 1996. Enzyme engineeringaspects of biocatalysis: cofactor regeneration as example. BiotechnolBioeng 52:309–319.

Kula M-R, Wandrey C. 1987. Continuous enzymatic transformation in anenzyme-membrane reaction with simultaneous NADH regeneration.Methods Enzymol 136:9–21.

Lageveen RG, Huisman GW, Preusting H, Ketelaar P, Eggink G, WitholtB. 1988. Formation of polyesters by Pseudomonas oleovorans: effectof substrates on formation and composition of poly-(R)-3-hydroxyal-kanoates and poly-(R)-3-hydroxyalkenoates. Appl Environ Microbiol54:2924–2932.

Lye GJ, Woodley JM. 1999. Application of in situ product-removal tech-niques to biocatalytic processes. Trends Biotechnol 17:395–402.

Meyer A, Schmid A, Held M, Westphal AH, Rothlisberger M, KohlerH-PE, van Berkel WJH, Witholt B. 2002. Changing the substratereactivity of 2-hydroxybiphenyl 3-monooxygenase from Pseudomonasazelaica HBP1 by directed evolution. J Biol Chem 277:5575–5582.

Miyakoshi T, Togashi H. 1990. Synthesis of 3-alkylcatechols via intramo-lecular cyclization. Synthesis 5:407–410.

Moore JC, Arnold FH. 1996. Directed evolution of a para-nitrobenzylesterase for aqueous–organic solvents. Nat Biotechnol 14:458–467.

Pras N, Batterman S, Dijkstra D, Horn AS, Malingre TM. 1990. Continu-ous production of the pharmaceutical 7,8-dihydroxy N-di-n-propyl2-aminotetralin using a phenoloxidase from cell cultures of Mucunapruriens. Plant Cell Tissue Organ Cult 23:209–215.

Robinson GK, Stephens GM, Dalton H, Geary PJ. 1992. The production ofcatechols from benzenes and toluene by Pseudomonas putida in glu-cose fed-batch culture. Biocatalysis 6:81–100.

Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a laboratorymanual, 2nd edition. Plainview, NY: Cold Spring Harbor LaboratoryPress.

Schmid A, Kohler H-PE, Engesser K-H. 1998. E. coli JM109 pHBP461, arecombinant biocatalyst for the regioselective monohydroxylation ofortho-substituted phenols to their corresponding 3-substituted cat-echols. J Mol Catal B 5:311–316.

Schmid A, Vereyken I, Held M, Witholt B. 2001. Preparative regio- andchemoselective functionalization of hydrocarbons catalyzed by cellfree preparations of 2-hydroxybiphenyl 3-monooxygenase. J MolCatal B 11:455–462.

Schmidt-Dannert C, Arnold F-H. 1998. Directed evolution of industrialenzymes. Trends Biotechnol 17:135–136.

Shao Z, Arnold FH. 1996. Engineering new functions and altering existingfunctions. Curr Opin Struct Biol 6:513–518.

Stemmer WPC. 1994. Rapid evolution of a protein in vitro by DNA shuf-fling. Nature 370:389–391.

Suske WA, Held M, Schmid A, Fleischmann T, Wubbolts MG, Kohler

MEYER ET AL.: BIOCATALYTIC SYNTHESIS OF 3-TERT-BUTYLCATECHOL 523

H-P. 1997. Purification and characterization of 2-hydroxybiphenyl3-monooxygenase, a novel NADH-dependent, FAD-containing aro-matic hydroxylase from Pseudomonas azelaica HBP1. J Biol Chem272:24257–24265.

van den Huevel RHH, Fraaije MW, Ferrer M, Mattevi A, van Berkel WJH.2000. Inversion of stereospecificity of vanillyl-alcohol oxidase. ProcNatl Acad Sci USA 97:9455–9460.

Wade LG. 1995. Organic chemistry, 3rd edition. Englewood Cliffs, NJ:Prentice Hall.

Witholt B. 1972. Method for isolating mutants overproducing NAD and itsprecursors. J Bacteriol 109:350–364.

Yanish-Perron C, Vieira J, Messing J. 1985. Improved M13 phage cloningvectors and host strains: nucleotide sequences of the M13mp18 andpUC19 vectors. Gene 33:103–109.

Yoo JW, Lee CW, Park S-E, Ko J. 1999. Alkylation of catechol witht-butyl alcohol over acidic zeolites. Appl Catal A 187:225–232.

Zhao H, Arnold FH. 1999. Directed evolution converts subtilisin E into afunctional equivalent of thermitase. Protein Eng 12:47–53.

524 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 81, NO. 5, MARCH 5, 2003