Appl Microbiol Biotechnol (2003) 60:708712 DOI 10.

1007/s00253-002-1181-7

SHORT CONTRIBUTION

R. Pilz E. Hammer F. Schauer U. Kragl

Laccase-catalysed synthesis of coupling products of phenolic substrates in different reactors

Received: 10 July 2002 / Revised: 15 October 2002 / Accepted: 18 October 2002 / Published online: 18 December 2002  Springer-Verlag 2002

Abstract Substrate oxidation of aromatic substances by the enzyme laccase followed by a heteromolecular coupling with a co-substrate is a promising possibility for the synthesis of new compounds. To find a suitable reactor for the effective production of new compounds, the laccase-catalysed coupling of 3-(3,4-dihydroxyphenyl)propionic acid with 4-aminobenzoic acid was investigated as a model system. Based on the kinetic parameters, a mathematical model was used to predict the reaction yield and oxygen demand in a discontinuously stirred tank reactor and a continuously operated stirred tank reactor. Membrane processes were used for bubblefree aeration of the system and to recover the soluble enzyme.

Introduction
Laccases are enzymes of considerable biotechnological interest, since they can be used for numerous applications including delignification or decolouration of paper pulp, detoxification of environmental pollutants, textile dyebleaching or diagnostic assays (Xu 1999). All of these applications are based on the ability of the laccase to oxidise a wide range of aromatic substances including phenolic substrates. During the laccase-catalysed reaction, O2 is reduced to water (Reinhammar and Malmstrm 1981). The copper-containing enzyme laccase (E.C. 1.10.3.2) has been detected in various plant species (e.g. mango, mung bean, peach), some bacteria (e.g. Azospirillum lipoferum), various insects and in more than 40 different
R. Pilz U. Kragl ()) Rostock University, Department of Chemistry, 18051 Rostock, Germany e-mail: udo.kragl@chemie.uni-rostock.de Tel.: +49-381-4986450 Fax: +49-381-4986452 E. Hammer F. Schauer Ernst Moritz Arndt University, Institute of Microbiology, 17487 Greifswald, Germany

fungi, which produce it as an extracellular enzyme (Gianfreda et al. 1999). Laccase generally contains four copper atoms per monomer bound to three redox sites (type 13) and differ from each other in their electronic properties. It is believed that the mechanism of laccase catalysis involves three steps: (1) the reduction of copper type 1 by reducing a substrate, (2) internal electron transfer between the different copper types, and (3) the reduction of O2 to water at the type 2 and the type 3 copper site (Messerschmidt 1994; Thurston 1994; Yaropolov et al. 1994). The substrate oxidation catalysed by laccase is a oneelectron reaction, in which a free (cation) radical is generated (Thurston 1994). The unstable radical may undergo further non-enzymatic reactions, e.g. hydration, disproportionation or polymerisation. In addition, crosscoupling with different substances has been observed (Tatsumi et al. 1992, 1994). Using this reaction, it is possible to produce novel compounds, e.g. with antibiotic properties (Schauer et al. 2001). For technical purposes, effective production of such new compounds by laccase catalysis should be carried out in a bioreactor. To find a suitable reactor type we have determined the kinetic parameters maximum reaction velocity (Vmax), MichaelisMenten constant (Km), and the product inhibition (Ki) of a model reaction. These parameters were then used for the mathematical simulation of several reactor types. The model reaction used was the laccase-catalysed crosscoupling of 3-(3,4-dihydoxyphenyl)propionic acid and 4aminobenzoic acid (Fig. 1).

Materials and methods

Chemicals 3-(3,4-Dihydroxyphenyl)propionic acid, 2,2-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)diammonium salt (ABTS) and phosphoric acid (crystallised) were obtained from Fluka (Deisenhofen, Germany). p-Aminobenzoic acid, methanol, sodium acetate, and acetic acid were purchased from Baker (Deventer, The Netherlands).

709 The concentrations of the reactants and the cross-coupling product were determined using an HPLC equipped with a UV detector (Knauer, Berlin, Germany) operating at 280 nm. The column was a 1252 mm reversed phase column (Nucleosil C181003, Machery-Nagel, Dren, Germany). The reaction mixture was separated at a flow rate of 0.3 ml min1 using an initial mobile phase of 20:80 (v/v) methanol:0.1% phosphoric acid, changing to 80:20 (v/v) methanol:0.1% phosphoric acid run over a period of 10 min and held for an additional 1 min. To calibrate the HPLC system, the cross-coupling product was isolated from the reaction mixture by solid phase extraction. The C18-column type Bond elut (VARIAN, Harbor City, Calif.) filled with 0.5 g material was activated with 2 ml each of methanol and water and then charged with reaction mixture containing the product. Water containing 1030% (v/v) methanol was used to wash out the remaining reactants. After drying the column with argon, the cross-coupling product was eluted with acetonitrile and dried under vacuum. The cross-coupling product was silylated with Silyl-991 (BSTFA (N,O-bis-trimethylsilyl-trifluoracetamide)-TMCS (trimethyl-chlorsilane) 99:1, Macherey-Nagel) and identified by gas chromatography/mass spectrometry (Mikolasch et al. 2002).

Fig. 1 Cross-coupling reaction catalysed by laccase from Pycnoporus cinnabarinus Preparation of laccase Laccase was prepared from cultures of the white rot fungus Pycnoporus cinnabarinus (DSM 15225) isolated from an oak tree in northern Germany. Cultures were prepared by inoculating a nitrogen-rich medium (Braun-Lllemann et al. 1997) according to Jonas et al. (2000). After 6 days of incubation, the cultures were harvested and the supernatant separated from mycelia by filtration through glass fibre filters (GF6, Schleicher & Schuell, Dassel, Germany). The extracellular enzyme was bound on DEAESephacel (Sigma, Steinheim, Germany) as described by Jonas et al. (2000), eluted by a high salt buffer (20 mM sodium acetate, 700 mM sodium chloride, pH 6.0) and desalted over a HiPrep 26/10 column (Pharmacia, Freiburg, Germany). Discontinuously stirred tank reactor To determine the mass balance and the kinetic parameters in a discontinuously stirred tank reactor (STR), dark glass vials with a volume of 5 ml were used. The vials were filled with 1 ml reaction mixture and were vigorously shaken over the reaction period (400 min1). Continuously operated enzyme-membrane reactor Continuous enzyme reactions were performed in a continuously operated enzyme-membrane reactor (EMR). The reaction chamber consisted of acrylic glass and had a volume of 44 ml. It had connections to substrate inlet and product outlet, temperature control fluid (inlet/outlet) and two sampling ports. The sampling ports were located above the membrane, which retained the enzyme. The membrane used in the reactor was a NADIR C010F (NADIR Filtration, Wiesbaden, Germany) with a cut off of 10 kDa. The acetate buffer (0.02 M, pH 5.0), which contained the reactants, was fed into the EMR by a membrane pump. The solution in the reaction chamber was mixed by magnetic stirring. The product and the residual reactants permeated through the membrane and left the reactor. Different residence times were achieved by varying the flow rate of the inlet. All reactions were carried out at 25C. In the case of experiments with bubble-free aeration, the two sampling ports of the reaction chamber were connected to a silicon tube (2 m, i.d. 4 mm, wall thickness 0.3 mm; BIW Isolierstoffe, Ennepetal, Germany) in the form of a closed loop. For this type of tube an oxygen transfer rate of 72 g day1 m2 was determined (Rissom 1999). The reaction solution was pumped through the aeration tube by a peristaltic pump (16 ml min1). The volume of the tube (0.025 l) was added to the volume of the reaction chamber. Analyses The laccase activity was determined at pH 5.0 by monitoring the oxidation of ABTS at 420 nm (e420=36,000 M1 cm1, Bourbonnais and Paice 1990). A 100-l sample was added to 110 l 5 mM ABTS solution and to 790 l 0.02 M sodium acetate buffer (pH 5.0) and the change in absorbance monitored over 1 min (Jonas et al. 1998). One unit of enzyme was defined as 1 mol of substrate oxidised per minute.

Results
The reaction Incubation of 3-(3,4-dihydoxyphenyl)propionic acid and 4-aminobenzoic acid (Fig. 1) with a laccase preparation from Pycnoporus cinnabarinus revealed one major product 4-[2-(2-carboxyethyl)-4,5-dihydoxyphenylamino]benzoic acid. The transformation was accompanied by a change in colour of the fluid to dark red. Incubation of each substrate alone with laccase showed no enzyme activity in the case of 4-aminobenzoic acid, whereas 3-(3,4-dihydoxyphenyl)propionic acid was converted to several products. Therefore 3-(3,4-dihydoxyphenyl)propionic acid must be the substrate that is bound to the enzyme. To analyse the mass balance of the reaction (Fig. 1), the concentrations of the two reactants and of the crosscoupling product were measured with respect to time. For this experiment we used an STR with contact to the atmosphere; 81.1% of the 3-(3,4-dihydoxyphenyl)propionic acid was converted into the cross-coupling product (Fig. 2). After 120 min in the STR no further transfor-

Fig. 2 Mass balance of the cross-coupling reaction, c0=0.001 M (reactants equimolar), cE=0.049 U ml1

710 Fig. 3 Experimental and simulated data for the production of 4-[2-(2-carboxyethyl)-4,5-dihydroxyphenylamino]benzoic acid in a stirred tank reactor (STR), c0=0.001 M equimolar, cE= 0.049 U ml1 (A) and in the continuously operated stirred tank reactor (CSTR) c0=0.001 M equimolar, cE= 0.170 U ml1 (B)

mation could be detected although residues of 3-(3,4dihydroxyphenyl)propionic acid as well as 4-aminobenzoic acid were detected. HPLC analysis showed the presence of small amounts (about 9%) of at least two byproducts, which have not been identified so far. The concentration of the cross-coupling product decreased after 200 min, suggesting that the cross-coupling product itself might be transformed further by chemical or enzymatic reaction. Determination of Vmax, Km and Ki Experiments were carried out in the STR, to determine the kinetic parameters Vmax, Km and Ki from initial rate measurements. These parameters were required for the mathematical simulation of the several reactor types. The relationship between reaction velocity and substrate concentration was obtained by varying the substrate concentration at a constant concentration (0.01 M) of 4-aminobenzoic acid and an enzyme activity of 0.049 U ml1 based on the standard assay. The resulting parameters were: Vmax =(50.11.58) 106 mol l1 min1 and Km =(10.41.38) 104 mol l1. Substrate inhibition was not observed under these conditions. Ki was determined by adding different amounts of isolated product to an equimolar (0.004 M) reaction mixture before starting the coupling reaction, from which a competitive Ki =(25.83.1) 105 mol l1 was calculated. Modelling Computer aided modelling of the STR was carried out using the software Scientist (Micromath Scientific Software, Salt Lake City, Utah) and the data of Vmax, Km and Ki. The experimental data for the STR and the simulated data for substrate consumption and product formation (Fig. 3A) show a good correlation, showing this simple model is sufficient to describe the reaction system. This suggests such simulations can provide an estimate of the yields of the reaction that could be reached in other reactor types.

Comparatively stable enzymes can be used more effectively in a continuously operated reactor (CSTR) designed as an EMR than in a STR (Kragl et al. 1996). In the EMR, an ultrafiltration membrane is used to retain the enzyme whereas the reactants and the cross-coupling product pass through the membrane. In the following experiments an EMR was used with a membrane with a cut-off of 10 kDa. In contrast to the results obtained for the STR, the experimental and simulated data did not correlate well (Fig. 3B). A correlation was seen only for low yields corresponding to 1020 min residence time. At longer residence times synthesis of the coupling product slowed and yield was lower than expected from the simulation. For instance, a residence time of 70 min gave a yield of only 40%, whereas the simulation predicted a yield greater than 70%. This difference in yield was assumed to be caused by a limited oxygen supply in the reactor. The EMR is a sealed system, preventing access of air to the reaction mixture. The acetate buffer (pH 5.0) used for the reaction contained 0.16 mmol l1 dissolved oxygen. However, total conversion of 1 mmol l1 substrate consumes 0.25 mmol l1 O2. Therefore, additional oxygen is necessary for complete conversion. The standard method of feeding oxygen by bubbling air into the reaction mixture might cause deactivation of the enzyme. For this reason, we tried an integrated bubblefree aeration system (Rissom et al. 1997) using a thinwalled silicon tube to supply oxygen. While the reaction mixture is pumped through the tube, the oxygen diffuses through the tube surface into the reaction mixture. An aeration tube 2 m in length was sufficient to avoid oxygen limitation. Under these conditions experimental and calculated data were in good agreement (Fig. 3B). Using integrated bubble-free aeration a yield of 80% was reached, similar to that in the STR. In order to reach total conversion of substrate at initial concentrations higher than 0.001 M, a greater tube surface for the diffusion of oxygen might be necessary. This can be achieved by increasing the length of the tube. In total, the reactor was operated for 4 days. The average space-time yield was 1.3 g l1 day1.

711

Discussion
The phenol-removing activity of laccase is going to be used not only for environmental applications but also in bleaching processes in the textile industry or in the treatment of beverages (reviewed by Gianfreda et al. 1999). In contrast, the synthesis potential of laccasecatalysed reactions was considered as moderate (Schmid et al. 2002). The coupling activity of the polyphenol oxidase laccase with a view to the synthesis of novel compounds (Anyanwutaku et al. 1994; Bhalerao et al. 1994; Lindequist and Schauer 2002; Mikolasch et al. 2002) and materials (Httermann et al. 2001) has been studied in detail in only a few cases. Nevertheless, the conversion of phenolic substrates with laccase by coupling of small molecules is a promising possibility for the synthesis of new compounds. It might be a way to generate new therapeutic compounds for the treatment of microbial infections or cancer, or to modify already known ones. To use such reactions effectively, it is necessary to generate data to scale up the process and identify possible bottlenecks. Immobilisation of the laccase can reduce the amount of enzyme required for use in the synthesis process. Several methods have been shown to be applicable for this purpose (Palmieri et al. 1994; Crecchio et al. 1995). However, in some cases the products synthesised may interact with the immobilisation matrix, resulting in decreased yields. Therefore, we favoured the use of a membrane reactor to retain the enzyme in a defined reaction volume. On basis of the kinetic experiments we were able to determine the basic kinetic parameters, Vmax, Km and Ki, necessary to carry out a mathematical simulation of several reactor types. The estimation of the Km value with 1 mmol l1 point to a relatively low affinity of the enzyme for the substrate 3-(3,4-dihydroxyphenyl) propionic acid. The Km value for other laccase substrates like syringaldazine, catechol or 2,6-dimethylphenol were described to be in the range 0.0010.26 mmol l1 (Xu 1999). In case of the STR, the simulated and the experimental data of the laccase-catalysed reaction were in good agreement. Therefore, a mathematical calculation of the yield should be possible for any reactor type, thus avoiding scale up optimisation. However, the EMR experiments showed that an optimal laccase-catalysed synthesis strongly depends upon a sufficient supply of oxygen. Without additional aeration the experimental product yield was much lower than the calculated one. Because of this, some type of aeration device will be a necessary component of the reactor. Diffusion of oxygen into the input is one possibility, air bubbling is another. We plan to investigate these processes in the future to optimise the stability of the enzyme, further improve the reactor design and increase substrate concentrations in order to scale up the process. This will also allow the productivity of the process to be increased. These investigations will be of special importance if the reaction will be used for synthetic purposes on a multi-gram scale.

Several examples of reaction systems of commercial interest are described in Mikolasch et al. (2002).
Acknowledgements This work was financially supported by the European Funds for Regional Development and by the Ministry for Science and Education of Mecklenburg-Vorpommern. We thank R. Jack (Greifswald, Germany) for revising the manuscript.

References
Anyanwutaku IO, Petroski RJ, Rosazza JPN (1994) Oxidative coupling of mithramicin and hydroquinone catalyzed by copper oxidases and benzoquinone. Implications for the mechanism of action of aureolic acid antibiotics. Bioorg Med Chem 2:543 551 Bhalerao UT, Muralikrishna C, Rani BR (1994) Laccase enzyme catalysed efficient synthesis of 3-substituted 1,2,4-triazolo(4,3b)(4,1,2)benzothiazidine-8-ones. Tetrahedron 50:40194024 Bourbonnais R, Paice MG (1990) Oxidation of non-phenolic substrates: an expanded role for laccase in lignin biodegradation. FEBS Lett 267:99102 Braun-Lllemann A, Majcherezyk A, Httermann A (1997) Degradation of styrene by white rot fungi. Appl Microbiol Biotechnol 47:150155 Crecchio C, Ruggiero P, Pizzigallo MDR (1995) Polyphenoloxidases immobilized in organic gels: properties and applications in the detoxification of aromatic compounds. Biotechnol Bioeng 48:585591 Gianfreda L, Xu F, Bollag JM (1999) Laccases: a useful group of oxidoreductive enzymes. Bioremed J 3:125 Httermann A, Mai C, Kharazipour A (2001) Modification of lignin for the production of new compounded materials. Appl Microbiol Biotechnol 55:387394 Jonas U, Hammer E, Schauer F, Bollag JM (1998) Transformation of 2-hydroxydibenzofuran by laccases of the white rot fungi Trametes versicolor and Pycnoporus cinnabarinus and characterization of oligomerization products. Biodegradation 8:321 328 Jonas U, Hammer E, Haupt ETK, Schauer F (2000) Characterisation of coupling products formed by biotransformation of biphenyl and diphenyl ether by the white rot fungus Pycnoporus cinnabarinus. Arch Microbiol 174:393398 Kragl U, Kruse W, Hummel W, Wandrey C (1996) Enzyme engineering aspects of biocatalysis: cofactor regeneration as example. Biotechnol Bioeng 52:309319 Lindequist U, Schauer F (2002) Bioactive natural compounds. New possibilities for their derivatization. Screening 3:4849 Messerschmidt A (1994) Blue copper oxidases. Adv Inorg Chem 40:121185 Mikolasch A, Hammer E, Jonas U, Popowski K, Stielow A, Schauer F (2002) Synthesis of 3-(3,4-dihydroxy-phenyl)-propionic acid derivatives by N-coupling of amines using laccase. Tetrahedron 58:75897593 Palmieri G, Giardina P, Desiderio B, Marzullo L, Giamberini M, Sannia G (1994) A new enzyme immobilization procedure using copper alginate gel: application to a fungal phenol oxidase. Enzyme Microb Technol 16:151158 Reinhammar B, Malmstrm BG (1981) Blue copper-containing oxidases. In: Spiro TG (ed) Copper proteins. Wiley, New York, pp 109149 Rissom S (1999) Membranverfahren fr Redoxenzyme, Gasversorgung-Reaktion-Produktextraktion. Dissertation Rheinische Friedrich-Wilhelms-Universitt Bonn Rissom S, Schwarz-Linek U, Vogel M, Tishkov VI, Kragl U (1997) Synthesis of chiral e-lactones in a two-enzyme system of cyclohexanone mono-oxygenase and formate dehydrogenase with integrated bubble-free aeration. Tetrahedron Asymmetry 8:25232526

712 Schauer F, Lindequist U, Hammer E, Jlich WD, Schfer A, Jonas U (2001) Biotransformation von biologisch aktiven Verbindungen aus verschiedenen chemischen Stoffklassen mittels der Enzyme Laccase und Manganperoxidase. Patentschrift PCP/EP 01/07152 Schmid A, Hollmann F, Bhler B (2002) Oxidation of phenols. In: Drauz K, Waldmann H (eds) Enzyme catalysis in organic synthesis. Wiley-VCH, Weinheim, p 1174 Tatsumi K, Wada S, Ichikawa H, Lin SY, Bollag JM (1992) Crosscoupling of a chloroaniline and phenolic acids catalyzed by a fungal enzyme. Water Sci Technol 26:21572160 Tatsumi K, Freyer A, Minard R, Bollag JM (1994) Enzymatic coupling of chloroanilines with syringic acid, vanillic acid and protocatechuic acid. Soil Biol Biochem 26:735742 Thurston F (1994) The structure and function of fungal laccases. Microbiology 140:1926 Xu F (1999) Laccase. In: Flickinger MC, Drew SW (eds) Encyclopedia of bioprocessing technology: fermentation, biocatalysis and bioseparation, Wiley, New York, pp 15451554 Yaropolov AI, Skorobogatko OV, Vartanov SS, Varfolomeyev SV (1994) Laccase properties, catalytic mechanism and applicability. Appl Biochem Biotechnol 49:257280