Effect of molecular mobility on coupled enzymatic reactions involving cofactor regeneration using nanoparticle-attached enzymes

Journal of Biotechnology 154 (2011) 274–280

Contents lists available at ScienceDirect

Journal of Biotechnology journal homepage: www.elsevier.com/locate/jbiotec

Effect of molecular mobility on coupled enzymatic reactions involving cofactor regeneration using nanoparticle-attached enzymes Muqing Zheng a,b , Songping Zhang a , Guanghui Ma a , Ping Wang c,∗ a b c

National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China Graduate University, Chinese Academy of Sciences, Beijing 100190, China Department of Bioproducts and Biosystems Engineering and Biotechnology Institute, University of Minnesota, St. Paul, MN 55108, USA

a r t i c l e

i n f o

Article history: Received 10 January 2011 Received in revised form 6 April 2011 Accepted 18 April 2011 Available online 12 June 2011 Keywords: Enzyme Superparamagnetic nanoparticles Cofactor regeneration Biotransformation Biosynthesis

a b s t r a c t Cofactor-dependent multi-step enzymatic reactions generally require dynamic interactions among cofactor, enzyme and substrate molecules. Maintaining such molecular interactions can be quite challenging especially when the catalysts are tethered to solid state supports for heterogeneous catalysis for either biosynthesis or biosensing. The current work examines the effects of the pattern of immobilization, which presumably impacts molecular interactions on the surface of solid supports, on the reaction kinetics of a multienzymic system including glutamate dehydrogenase, glucose dehydrogenase and cofactor NAD(H). Interestingly, particle collision due to Brownian motion of nanoparticles successfully enabled the coupled reactions involving a regeneration cycle of NAD(H) even when the enzymes and cofactor were immobilized separately onto superparamagnetic nanoparticles (124 nm). The impact of particle motion and collision was evident in that the overall reaction rate was increased by over 100% by applying a moderate alternating magnetic field (500 Hz, 17 Gs), or using additional spacers, both of which could improve the mobility of the immobilized catalysts. We further observed that integrated immobilization, which allowed the cofactor to be placed in the molecular vicinity of enzymes on the same nanoparticles, could enhance the reaction rate by 1.8 fold. These results demonstrated the feasibility in manipulating molecular interactions among immobilized catalyst components by using nanoscale fabrication for efficient multienzymic biosynthesis. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Conventional enzyme immobilization technologies have been mostly developed for single enzyme reaction systems (Krajewska, ´ 2004; Kramer et al., 2010; Schmid et al., 2001; Szymanska et al., 2009; Wiseman, 1993; Yen et al., 2010; Wang, 2006). One important recent advance in this area is the emergence of nanoscale biocatalysts which promise significantly improved catalytic efficacy, with lifetimes approaching years instead of days or less (Cerdobbel et al., 2010; Kim et al., 2006a,b, 2010; Wang, 2006; Wang et al., 2009). A broad spectrum of nanoscale materials have been examined for biocatalysis, and nanoparticles constitute probably the most extensively studied group of materials (Jia et al., 2003; Ke et al., 2010; Panigrahi et al., 2007; Wang, 2006; Wu et al., 2009). Nanoparticles are attractive supports for biocatalysts in that they offer high surface area-to-volume ratios, easy preparation, versatile surface chemistry and tunable solution and assembling behaviors. Enzyme immobilization with nanoparticles has been

∗ Corresponding author. Tel.: +1 612 625 3064; fax: +1 612 625 6286. E-mail address: [email protected] (P. Wang). 0168-1656/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jbiotec.2011.04.013

mostly achieved through covalent binding with enzyme loadings approaching monolayer surface coverage (Jia et al., 2003). In addition, once dispersed in solution, nanoparticles possess Brownian motion mobility and thus demonstrate enzyme activities comparable to those of free enzymes (Horák et al., 1999; Jia et al., 2003). In fact, it was proposed that Brownian motion-driven nanoparticle biocatalysis may represent a transitional region between the homogeneous catalysis with free enzymes and heterogeneous catalysis with immobilized enzymes (Jia et al., 2003). One important trend in today’s research on biocatalysis is to extend the scope of nanoscale biocatalysis from single enzyme reactions to multistep biocatalysis capitalizing on the unique features of nanomaterials (Wang, 2006). Reaction systems using multiple enzyme and cofactor components can enable predetermined reaction pathways for efficient synthetic and biosensing applications. In particular, cofactor-dependant oxidoreductases can catalyze the synthesis of a wide range of unnatural chemicals, especially chiral acids, alcohols and ketones (Hummel, 1999). However, cofactors are generally too expensive to be consumed stoichiometrically for the production of commodity chemicals. One common way to achieve regeneration and reuse of cofactors is to apply a secondary substrate-driven enzymatic reaction along

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

with the primary synthetic reaction. A particular challenge faced in conducting such multienzymic catalysis is how to facilitate efficient interactions between enzyme and cofactor molecules under immobilized environment (Liu and Wang, 2007). The traditional way to retain enzymes and cofactors in a continuously fed reactor is the use of microcapsules or semi permeable membranes which allow only small molecules such as substrate and product to pass through (Johannes et al., 2006; Lütz et al., 2006; Lin et al., 1999; Stengelin and Patel, 2000). Apparently the space-time productivity of such membrane-contained reaction systems may suffer limitations from mass transfer resistance of the capsules and membranes. Alternatively, immobilization of multiple enzymes and cofactor using materials such as agarose, cellulose, and polyacrylamide gel were also examined; but the results were not satisfying, as the immobilized cofactor could not react well with the immobilized enzymes (Wykes et al., 1975; Yamazaki and Maeda, 1982). Apparently, maintaining dynamic cofactor–enzyme interactions is the key to achieving efficient multistep biocatalysis with immobilized enzymes. Toward that, nanoscale structures provide unique opportunities. One recent work demonstrated that the integration of a multienzymic system within the pores of nanoporous silica glass could facilitate molecular interactions between the cofactor and enzymes through thermal vibration of the flexibly tethered cofactor, thus realizing the coupled reactions with lactate and glucose dehydrogenases (El-Zahab et al., 2004). It was also demonstrated that particle-carried enzymes and cofactor realized multistep reactions through enzyme–cofactor interactions through a mechanism of particle collision (El-Zahab et al., 2008; Liu et al., 2009). Imaginably, many microenvironmental factors can impact enzyme–cofactor interactions, and this work was aimed to investigate parameters that control the reaction kinetics of the nanoparticle-tethered multienzymic catalyst systems. A model reaction system involving glutamate dehydrogenase, glucose dehydrogenase and NAD(H) was applied, as both the enzymes have been studied extensively with catalytic behaviors well known (ElZahab et al., 2004; Liu et al., 2009). The focus of this work is on the effects of immobilization pattern on enzyme–cofactor interactions in terms of observed reaction kinetics. To further probe the role of freedom and mobility of the tethered components, we applied superparamagnetic nanoparticles and spacers for enzyme attachment. Superparamagnetic nanoparticles have attracted a great deal of attention for enzyme immobilization because they can be recovered and reused easily by exerting an external magnetic field (Gardimalla et al., 2005; Huang et al., 2003; Konwarh et al., 2009). For this research, such particles may allow us to tune particle mobility and collision frequency by applying an alternating magnetic field. Earlier work on single enzyme reaction systems has also shown that long spacers for enzyme attachment onto nanoparticles could increase the catalytic efficiency of the immobilized catalysts, presumably by improving the motion freedom and flexibility of attached enzymes (Arica et al., 2009; Cao, 2005; Lee et al., 2009). El-Zahab et al. (2004) also showed that the use of spacers could improve the shuttling frequency of the cofactor between two enzymes co-immobilized within nanopores (El-Zahab et al., 2004). The effect of spacer on particle-tethered multienzymic biocatalysts was therefore also examined as an important factor controlling enzyme–cofactor interactions in the current work.

2. Materials and methods 2.1. Materials Glutamate dehydrogenase (GluDH, EC1.4.1.3, lyophilized powder, 80% protein) was purchased from Fluka (Switzerland). Glucose dehydrogenase (GDH, EC1.1.1.47, lyophilized powder, 60% protein)

275

was obtained from the Amano (Nagoya, Japan). N-(2-aminoethyl)3-aminopropyltrimethoxysilane (AEAPS), poly (ethylene glycol diglycidal ether) (PEGDE) (Mn = 526), and Nicotinamide adenine dinucleotide (NADH) were obtained from Sigma–Aldrich Chemical Co (Saint Louis, USA). Ferric triacetylacetonate (Fe(acec)3 ) was purchased from Acros Organics (Geel, Belgium). Standard glucose test kit was purchased from Beijing Leadman Biochemistry Technology (Beijing, China). All other materials including bovine serum albumin (BSA), ␣-ketoglutarate, ␥-glycidoxypropyl trimethoxysilane (GPS), d-glucose, ethanol, triethanolamine, tetraethyl orthosilicate, triethylamine, hydrochloric acid, sodium dihydrogen phosphate, disodium hydrogen phosphate, and ammonium acetate, were of analytical grade and purchased from Beijing Chemical Reagent Company (Beijing, China). 2.2. Preparation and activation of silica-coated superparamagnetic nanoparticles with -glycidoxypropyl trimethoxysilane (GPS) The silica-coated superparamagnetic nanoparticles were synthesized following a procedure modified from those published previously (Ge et al., 2007; Lu et al., 2002; Wan et al., 2007). Briefly, magnetic Fe3 O4 nanoparticles with an average diameter of 12 nm were synthesized via thermal decomposition of Fe(acec)3 (Fe(C5 H7 O2 )3 ). The resulted magnetite colloidal nanoparticles were then coated with silica through a sol–gel approach based on the hydrolysis of tetraethyl orthosilicate. Size distribution of silica-coated magnetic nanoparticles was determined by scanning electronic microscope (SEM, JSM-6700F, JEOL, Japan) and quasielastic light scattering (QLS, with particle solution refractive index measured as 1.330) with a Coulter Model N4SD particle analyzer (Coulter Electronics, Luton, UK). The silica-coated magnetic nanoparticles were further activated with GPS (Chu et al., 1997; Daniels and Francis, 1998; Lin et al., 2001). Typically, 0.2 g nanoparticles were diluted into a 50-mL ethanol–water solution (v:v = 1:1). The suspension was sonicated for 5 min during which 0.15 mL of triethylamine was added slowly (within 1 min). After transferring the mixture to a 100-mL 3-neck flask equipped with a mechanical stirrer, 0.5 mL GPS was added and the activation reaction was conducted at 60 ◦ C for 3 h with mechanical stirring (400 rpm). The resulted GPS-treated magnetic nanoparticles were collected by magnetic separation using a permanent magnet (with a surface magnetic field as 210 mT), and then washed twice with ethanol and three times with deionized (DI) water. The particles were stored in DI water before being used. 2.3. Activation of silica-coated magnetic nanoparticles with amino groups and PEG–BSA–PEG spacer To introduce spacers of different sizes, nanoparticles were modified first with AEAPS, then with PEG or PEG–BSA–PEG as spacer, following a similar method as reported previously by El-Zahab et al. (2004). The activation with AEAPS followed a similar procedure as that described above for GPS activation, except that 1 mL AEAPS was applied over 5 min (replacing 0.15 mL GPS) and the mixture was incubated at 90 ◦ C under stirring (400 rpm) for 5 h. To apply spacers for enzyme or cofactor attachment, 50 mg PEGDE (Mw 526) was mixed with 50 mg AEAPS-modified magnetic silica nanoparticles in 5 mL phosphate buffer (0.1 M, pH 7.0). The reaction was allowed to last 24 h at room temperature under stirring followed by washing with fresh phosphate buffer for three times. The PEG-attached particles were then applied for enzyme/cofactor immobilization. To generate large spacers, BSA was attached to PEG-activated particles by contacting 50 mg particles with a 5-mL solution containing 10 mg BSA in phosphate buffer (0.1 M, pH 7.0) for 24 h at room temperature under stir-

276

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

Fig. 1. SEM image (A) and hydrodynamic size distribution (B) of the silica-coated superparamagnetic nanoparticles. The data points in (B) were obtained using quasielastic light scattering with the curve as a result of data fitting with Gaussian function.

ring. After washing the particles with fresh phosphate buffer till no protein was detected in the washing solution using Bradford assay (Bradford, 1976), the BSA-attached particles were then re-activated with PEGDE under the same conditions as that developed for the first layer PEG activation. 2.4. Immobilization of individual enzyme (GluDH, or GDH) and cofactor NADH Enzymes and cofactor were covalently attached to particles by reacting with the epoxy groups of PEGDE exposed to the outer surface of the particles. Typically, 10 mg modified magnetic silica nanoparticles (either GPS modified, PEG modified or PEG–BSA–PEG modified) were dispersed in 1 mL phosphate buffer (0.1 M, pH 6.0) containing 0.5 mg enzyme (either GluDH or GDH). The reaction was conducted at 15 ◦ C for 24 h. The immobilized enzyme was then collected by magnetic separation using a permanent magnet and was washed several times with fresh buffer solution. The amount of attached enzyme was determined by subtracting the amount of enzyme found in residue immobilization and washing solutions (Bradford protein assay) from the total quantity initially added for immobilization. NADH was attached through the same procedure except 2 mg NADH was applied to contact with 10 mg activated magnetic nanoparticles in 1 mL triethanolamine–hydrochloric acid buffer (0.1 M, pH 8.0). The reaction was conducted at room temperature for 72 h. The concentration of NADH in aqueous solutions was measured by UV absorbance at 340 nm. It was found that the spacers did not significantly affect the final enzyme and cofactor loadings, which were determined in the range of 38–43 mg/g-nanoparticles for GluDH, 35–40 mg/g-nanoparticles for GDH, and around 20 mg/g-nanoparticles for NADH. 2.5. Coimmobilization of GluDH, GDH and NADH Two forms of coimmobilization, one with two enzymes and the other with both enzymes and cofactor, were prepared. For enzyme coimmobilization, the procedure was the same as that for the immobilization of individual enzymes, except that 10 mg nanoparticles were contacted with 0.35 mg GluDH and 0.35 mg GDH in 1 mL phosphate buffer. The total protein loading was generally about 40 mg/g-nanoparticles. For enzyme–cofactor coimmobilization, a two-step approach was used in which the cofactor was attached first following the same procedure as that described above for cofactor immobilization, followed by the attachment of both enzymes through the same procedure as that developed for enzyme coimmobilization. The loadings were around 20 mg NADH

and 40 mg protein (including GluDH and GDH) for each gram of nanoparticles. 2.6. Activity assay of individual enzymes Activity of GDH was determined by measuring the initial oxidation rate of glucose at 25 ◦ C. Activity unit of GDH was defined as the amount of enzyme needed to oxidize 1 ␮mol d-glucose to d-glucono-␦-lactone in 1 min at pH 8.0 and 25 ◦ C. Generally, the reactant mixture was 990 ␮l phosphate buffer (0.1 M, pH 8.0) containing 0.1 mM NAD+ and 10 mM glucose, to which 10 ␮l solution containing 5 ␮g free or immobilized GDH was added to initiate the reaction. The reaction was monitored by measuring A340 (␧NADH = 6.22 mM−1 cm−1 ) continuously for 5 min using a Unico 2800 spectrophotometer. The activity of GluDH was determined using a similar procedure by monitoring the disappearance of NADH in the reaction mixture. One unit of GluDH was defined as the amount of enzyme needed to reduce 1 ␮mol ␣-ketoglutarate to l-glutamate in 1 min at pH 8.0 and 25 ◦ C. Typically 5 ␮g free or immobilized GluDH was applied for 1 mL phosphate buffer (0.1 M, pH 8.0) containing 0.1 mM NADH, 10 mM ␣-ketoglutarate and 40 mM ammonium acetate at 25 ◦ C. 2.7. Reaction kinetics of coupled reactions Coupled enzymatic reactions were conducted in phosphate buffer (0.1 M, pH 8.0) at 25 ◦ C, and under mechanical stirring (150 rpm) unless stated otherwise. Typically, the reaction solution contained 0.1 U/mL GluDH, 0.1 U/mL GDH, 0.05 mM NADH (either free or immobilized), 10 mM ␣-ketoglutarate, 40 mM ammonium acetate and 10 mM d-glucose. The reaction rate of the coupled reactions was monitored by measuring changes in the concentration of glucose. Aliquots of the reaction solutions were taken periodically for analysis after the removal of biocatalyst particles using a permanent magnet. Glucose concentration was measured using a standard glucose test kit which was based on the PGO enzyme system (peroxidase and glucose oxidase). Standard procedures included mixing 20 ␮L sample solution with 1 mL glucose testing solution at 37 ◦ C for 20 min, followed by absorbance measurement at 520 nm. All experiments were carried out with at least three duplicates, and the results were expressed as average values. Reactions with particle-attached enzymes and cofactor were also conducted with an alternating magnetic field of 500 Hz and 17 Gs through a lab-made device to examine the effect of particle mobility on the reaction kinetics of the coupled reaction system.

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

277

O

O

O

O

GLuDH E1

E2

OH

HO

E2 Cofactor

Cofactor

Cofactor E1

E2

NH2

-ketoglutaric acid

glutamate NADH

E1

NAD

OH O

a

HO

Scheme 1. Different immobilization patterns for enzymes and cofactor with magnetic nanoparticles. (a) E-E-C; (b) EE-C; (c) EEC, each of the immobilization pattern involves different molecular interaction scenarios for cofactor regeneration, with (a) and (b) rely on particle collision, while (c) mainly through enzyme–cofactor affinity interaction on the same particle.

Control reactions were also conducted without external magnetic field, yet with other conditions identical. 3. Results and discussion 3.1. Impact of immobilization on activities of enzymes and cofactor Fig. 1(A) illustrates the SEM image of the silica-coated superparamagnetic nanoparticles prepared for this study. The SEM image showed particles with very smooth surface (within the resolution scale) and an average diameter of 105 nm, which was lower than the average hydrodynamic diameter in water, 124 nm, as measured with a quasielastic light scattering (QLS) particle size analyzer. This discrepancy can be largely attributed to the shrinkage of the dry particles taken for SEM. Nevertheless, the near agreement in particle sizes indicated that a good dispersion of the particles in aqueous solutions was achieved, and that the particles prepared had a very narrow size distribution (Fig. 1(B)). All of these are desired features for the immobilization and kinetic studies on the effect of particle-motion to be probed in the current work. The enzymes and cofactor were immobilized onto the particles through three different patterns as showed in Scheme 1, i.e., separately immobilized enzymes and cofactor (we may label this pattern as E-E-C), coimmobilized enzymes with separately immobilized cofactor (EE-C), and coimmobilized enzymes and cofactor (EEC). As shown in Table 1, the loadings and individual activities of enzymes changed only slightly with different immobilization patterns. For E-E-C immobilization, the loadings of GluDH and GDH were in the range of 42 and 39 mg/g-nanoparticles, respectively. The immobilization procedure was found efficient in that over 80% of the added enzyme was found on the particles, with activities (determined with free cofactor) measured as 0.086 and 0.137 U/mg-nanoparticle for GluDH and GDH, respectively (Table 1). For EE-C immobilization, the total enzyme loading of GluDH and GDH was 43 mg/g-nanoparticle, the activities (with free cofactor) for GluDH and GDH were 0.060, and 0.078 U/mgnanoparticle, respectively. Similar results were obtained for EEC immobilization, the total enzyme loading of GluDH and GDH was 44 mg/g-nanoparticle, the activities (with free cofactor) for GluDH and GDH were 0.056 and 0.084 U/mg-nanoparticle, respectively.

+

OH O

O

c

b

OH

HO

O

HO

OH

GDH

OH

D-glucono-1,5-lactone

OH

OH OH

D-glucose

Scheme 2. Multienzymatic catalytic system with in situ cofactor regeneration. Abbreviations: GluDH—glutamate dehydrogenase; GDH—glucose dehydrogenase; NADH—nicotinamide adenine dinucleotide.

For EEC pattern, the cofactor was attached before the immobilization of enzymes, therefore, either NADH was immobilized separately or collectively, the loading of NADH was the same, 20 mg/g-nanoparticle. It was believed that the attachment of both enzymes and cofactor was achieved through the amino groups on enzyme or adenosine moiety of cofactor. The samples have been washed thoroughly for several times to remove physically absorbed enzymes and cofactor till washing solutions showed no detectable NADH (as detected by the absorbance at 340 nm) or enzymes (using Bradford protein assay). During reaction kinetic studies, we also observed that the reactions stopped immediately upon the removal of the nanoparticle-attached enzymes or cofactor. We further examined the possibility of enzyme leaking from the particles by incubating 10 mg EEC nanoparticles in 1 mL phosphate buffer (0.1 M, pH 8.0) solution with 2 mM NADH for 24 h, and found no protein detectable in the solution. The activities of free GluDH and GDH under the same reaction conditions were measured to be 3.8 and 4.9 U/mg-protein, respectively. From the above individually measured apparent activities of immobilized enzymes, the specific activities of separately immobilized GluDH and GDH were 2.0 and 3.5 U/mg-protein, respectively, indicating activity retention as high as 53 and 72% was achieved for GluDH and GDH. For EE-C and EEC pattern immobilization, the situation was more complicated as individual enzyme loading of GluDH or GDH could not be determined exactly with only total enzyme loadings were known; however, if we assume the same specific activities for both immobilized enzymes, the activities data indicated that the mass ratio between GluDH and GDH were 76:100 and 67:100 for EE-C and EEC immobilization, respectively. 3.2. Effect of enzymes and cofactor immobilization on the multienzymatic reaction kinetics with E-E-C immobilization The multienzymic reaction system involving cofactor regeneration studied in this work was shown in Scheme 2. The translocation of the cofactor between two enzymes was essential to the coupled reactions. NADH was required to facilitate the production of glutamate from ␣-ketoglutarate and got oxidized; it was then regenerated from NAD+ via the oxidation of glucose catalyzed by

Table 1 Enzyme loadings and activities for different immobilization patterns. Immobilization pattern

E-E-C EE-C EEC

Enzyme loading (mg/g-nanoparticle)

Enzyme activity (U/mg-nanoparticle)

GluDH

GDH

GluDH

GDH

42.4 ± 1.5

39.1 ± 1.2

0.086 ± 0.009 0.060 ± 0.012 0.056 ± 0.011

0.137 ± 0.019 0.078 ± 0.016 0.084 ± 0.014

43.2 ± 2.8 44.3 ± 2.5

Glucose concentration (mM)

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

10 8 6 4 2 0

0

10

20

30 40 50 Time (h)

60

70

Fig. 2. Effect of immobilization on activities of enzymes and cofactor. () Free enzymes–free NADH; (♦) immobilized enzymes–free NADH; () free enzymes–immobilized NADH; (䊉) immobilized enzymes–immobilized NADH.

GDH. As the reduced form of the cofactor (NADH) was applied at the beginning of the reaction, the oxidation of glucose was only possible after the production of glutamate. Accordingly glucose concentration was chosen as an overall index of reaction velocity of the coupled reactions. For E-E-C immobilization, it is much easier to handle in that the amounts of enzymes and cofactor can be adjusted at will by controlling the amounts of different particles added to the reaction solutions. Therefore, we started with E-E-C pattern immobilization to study the effect of immobilization on the kinetics of the coupled reaction. Fig. 2 demonstrates the time course of the reactions with four different enzyme/cofactor combinations. As expected, the free enzymes–free NADH combination showed the highest reaction rate (∼42 ␮M/min), and the immobilized enzymes–immobilized NADH system catalyzed the slowest reaction (Fig. 2). It appeared that the immobilization of cofactor affected the overall reaction much more profoundly than the immobilization of enzymes. When immobilized enzymes were applied with free cofactor, the reaction rate accounted for 45% of free enzyme/cofactor system (Fig. 2). However, when immobilized cofactor was applied either with free or immobilized enzymes, the reaction velocity decreased to only ∼2% of the free catalyst/cofactor system. One consideration for that is the activity change of NADH introduced by chemical modification. However, according to a previous research on NADH modification using PEG modifiers (Bückmann et al., 1981), the activity of NADH modified with PEG (Mn , 10,000) was comparable with that of native NADH. The strong effect of particle attachment on cofactor activity observed in the current work may be attributed to the difficulty in cofactor–enzyme interaction. The particles are much larger than the enzymes, both steric and orientation effects may become important factors impairing enzyme–cofactor interactions. 3.3. Effect of immobilization patterns on kinetics of coupled reactions Fig. 3 shows time courses of the coupled reactions with catalysts of different immobilization patterns. As mentioned earlier that the reactions were initiated in the presence of NADH, the observation of changes in glucose concentration is a direct indication of cofactor’s regeneration. Apparently, the cofactor was regenerated even with E-E-C immobilization. From the data shown in Fig. 3, an initial reaction rate of 0.70 ␮M/min could be obtained for E-E-C immobilization. A similar reaction rate with a value of 0.65 ␮M/min was determined for EE-C immobilization which also included the cofactor immobilized separated from the enzymes. For both E-E-C and

10 9 8 7 6 5

0

15

30

45

60

75

Time (h) Fig. 3. Time course of reaction with enzymes and NADH immobilized in different patterns. () EEC; () free enzyme-immobilized cofactor; (♦) E-E-C; (䊉) EE-C.

EE-C immobilization, we may assume that cofactor regeneration cycles can only be realized through particle collision and that particle collision frequency determines the upper limit of reaction rates (Scheme 1). On the other hand, EEC immobilization demonstrated the fastest reaction with an initial reaction rate measured as 1.28 ␮M/min. Interestingly that was even ∼45% higher than what observed for free enzymes with immobilized NADH. We tend to believe that the enzymes and cofactor share dynamic interactions once they are integrated into molecular vicinity on the same particle surface, and such enhanced molecular interactions facilitated the coupled reactions (Scheme 1). 3.4. Effect of alternating magnetic field on the multienzymatic reaction kinetics To further evaluate the role of particle collision in controlling the reaction kinetics for the current catalytic system, we examined the effect of external alternating magnetic field which can mobilize the superparamagnetic nanoparticles. Fig. 4 shows the initial reaction rates for E-E-C immobilization under different conditions. The application of an external magnetic field (500 Hz, 17 Gs) led to an initial reaction rate of 0.77 ␮M/min, which was about 2.3 times of that observed for the control reaction without the magnetic field (Fig. 4). That was also a slight increase in comparison with that achieved with mechanical stirring (150 rpm, 0.70 ␮M/min). This observation further supported the assumption that particle mobility was a controlling factor for the kinetics of such a reaction system.

Reaction rate (µM/min)

Glucose concentration (mM)

278

0.8 0.6 0.4 0.2 0.0 Without mechanical With magnetic field stirring or magnetic field of 500 Hz 17 Gauss

With mechanical stirring of 150 rpm

Fig. 4. Effect of external magnetic field on coupled reactions.

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

1.6

100 PEG-BSA-PEGG PEG-BSA-PE

PEG-BSA-PE A-PEG G PEG-BS PEG-BSA-PEGG GPSPEG-BSA-PE GPS PEG PEG

1.2 PEG GPSPEG GPS

0.8

PEG GPSPEG GPS

0.4

80 60 40 20 0

0.0 Free enzymes+immobilized cofactor system

Residual activity (%)

Reaction rate (µM/min)

2.0

279

E-E-C immobilized system

EEC immobilized system

Fig. 5. Effect of spacer on catalytic efficiency for different immobilized multienzymic systems. Free enzymes + immobilized cofactor; E-E-C; EEC.

3.5. Effect of spacers on activity of immobilized enzymes and cofactor The high reaction activities of EEC immobilization implied that surface interaction and flexibility of tethered catalyst components also impacted the multi-step reaction kinetics greatly. Improved enzyme activities have been reported for immobilized enzymes prepared with appropriate spacers of different lengths (Humayun, 2007; Wang and King, 1979). In this study, the effect of three spacers was examined: GPS, PEG526 , and PEG–BSA–PEG. The use of such spacers did not impact the yield of enzyme immobilization, with effective loadings in the range of 39–44 mg/gnanoparticles, which were similar to those achieved without the spacers. The spacer did not significantly affect the individual activities of immobilized GluDH and GDH either when they were tested separately with free cofactor: 0.078–0.09 U/mg-nanoparticles for GluDH and 0.114–0.137 U/mg-nanoparticles for GDH for E-E-C immobilization; while activities in the ranges of 0.050–0.061 U/mgnanoparticles (GluDH) and 0.065–0.084 U/mg-nanoparticles (GDH) were observed for EEC immobilization. The loading of cofactor NADH maintained in the range of 19.5–20 mg/g-nanoparticles. It appeared that the activities of NADH immobilized with GPS and PEG526 were roughly the same for the coupled reactions (with reaction rates in the order of 0.88 ␮M/min); however, a much larger spacer PEG–BSA–PEG demonstrated a much higher cofactor activity (1.46 ␮M/min) when tested with free enzymes for the coupled reactions (Fig. 5). Apparently larger spacers may provide the cofactor much improved flexibility, facilitating the translocation of the cofactor between the enzymes and thus leading to faster reactions. Imaginably, larger spacers may also raise the cofactor away from particle surface, thus improving the availability of cofactor and reducing steric hindrance effect for enzyme–cofactor interactions. When tested with E-E-C immobilized enzymes, the reaction rate increased to 1.41 ␮M/min with the BSA-based spacer, over 70% improvement from what observed for the smaller spacers (<0.8 ␮M/min). This observation is probably a simple echo of the changes in the activity of the immobilized cofactor NADH as discussed above. The use of spacers did not impact the activity of the EECimmobilized system, with all reaction rates being high and similar to those observed without using spacers. The lack of dependence of cofactor activity on spacer size found with EEC system may reflect a different enzyme–cofactor interaction mechanism from the systems applying separately immobilized cofactor, i.e., the translocation of cofactor may mainly take place in neighboring molecules on the surface of particles via the molecular thermal

0

2 4 6 8 Number of reusing cycles

10

Fig. 6. Reusability of immobilized multienzymic catalysts. () E-E-C; (䊉) EEC.

vibration, which was similar to the multienzymic system as that was integrated inside the nanopores of silica glass as reported earlier (El-Zahab et al., 2004). When longer spacer was applied on the concave surface of nanoporous glass, the distance for cofactor shuttling between neighboring two enzymes would be reduced, thus leading to a higher cofactor shuttling frequency and a higher activity. On the contrary, the longer spacer on nanoparticles may bring the tethered components away from the convex surface and increase the distance between enzymes and cofactor. The improved flexibility of the tethered components may be counter-balanced by the increased distance between them. 3.6. Effect of immobilization on the reusability of multienzymic system One important purpose of immobilizing the catalyst systems including the cofactor is to facilitate catalyst recycling and reuse. The use of magnetic nanoparticles makes such a task easily realized by applying magnetic field for particle recovery. We demonstrate the reusability of the current catalytic system by examining activity loss during reusing cycles with two different patterns of immobilization: E-E-C and EEC. The reactions were allowed to last about one day (23 h) before they were stopped by retrieving the immobilized catalysts magnetically for next round activity test after being washed under the same reaction conditions. The results are illustrated in Fig. 6. For E-E-C immobilization, over 80% residual activity retained after 10 cycles of reuse; in the case of EEC, the reaction activity retained even higher with over 90% residual activity after 10 cycles of reuse. This slight stability difference may indicate that the affinity interaction between neighboring cofactor and enzyme molecules in the case of EEC immobilization may help the protection of enzymes from potential conformational changes introduced by the shear from particle–particle friction during reaction and recycling cycles. Nevertheless, both immobilization patterns showed the capability of serving the purpose of reusing very well. 4. Conclusions In conclusion we demonstrated that particle collision enabled coupled reactions catalyzed by glutamate dehydrogenase, glucose dehydrogenase and NADH even when they were immobilized separately on superparamagnetic nanoparticles. It appeared that the mobility of the immobilized component was critical to the reaction kinetics of the reaction system, and enhanced reaction rates were obtained when an alternating magnetic field was applied.

280

M. Zheng et al. / Journal of Biotechnology 154 (2011) 274–280

It was further demonstrated that the use of spacers for cofactor immobilization or co-immobilizing the cofactor with enzymes both enhanced enzyme–cofactor interactions and substantially improved the reaction rates of cofactor regeneration. These results demonstrated a new and promising approach to preparation of easily reusable immobilized multienzymic catalysts capable of carrying out predetermined reaction pathways for multi-step biosynthetic applications. Acknowledgements The authors thank support from the National Natural Science Foundation of China (Grant Nos. 20728607, 20706054), Chinese Academy of Sciences (KSCX2-YW-G-019), 973 Program (2009CB724705) and 863 Project (2008AA10Z302). The authors thank Ravi Narayanan from Wang lab at UMN for helping with the tests with alternating magnetic field. References Arica, M., Altintas, B., Bayramoglu, G., 2009. Immobilization of laccase onto spacerarm attached non-porous poly (GMA/EGDMA) beads: application for textile dye degradation. Bioresour. Technol. 100 (2), 665–669. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem. 72 (1–2), 248–254. Bückmann, A., Kula, M., Wichmann, R., Wandrey, C., 1981. An efficient synthesis of high-molecular-weight NAD (H) derivatives suitable for continuous operation with coenzyme-dependent enzyme systems. J. Appl. Biochem. 3, 3301– 3315. Cao, L., 2005. Immobilised enzymes: science or art? Curr. Opin. Chem. Biol. 9 (2), 217–226. Cerdobbel, A., Desmet, T., De Winter, K., Maertens, J., Soetaert, W., 2010. Increasing the thermostability of sucrose phosphorylase by multipoint covalent immobilization. J. Biotechnol. 150 (1), 125–130. Chu, L., Daniels, M., Francis, L., 1997. Use of (glycidoxypropyl) trimethoxysilane as a binder in colloidal silica coatings. Chem. Mater. 9 (11), 2577–2582. Daniels, M., Francis, L., 1998. Silane adsorption behavior, microstructure, and properties of glycidoxypropyltrimethoxysilane-modified colloidal silica coatings. J. Colloid Interface Sci. 205 (1), 191–200. El-Zahab, B., Donnelly, D., Wang, P., 2008. Particle-tethered NADH for production of methanol from CO2 catalyzed by coimmobilized enzymes. Biotechnol. Bioeng. 99 (3), 508–514. El-Zahab, B., Jia, H., Wang, P., 2004. Enabling multienzyme biocatalysis using nanoporous materials. Biotechnol. Bioeng. 87 (2), 178–183. Gardimalla, H., Mandal, D., Stevens, P., Yen, M., Gao, Y., 2005. Superparamagnetic nanoparticle-supported enzymatic resolution of racemic carboxylates. Chem. Commun. 2005 (35), 4432–4434. Ge, J., Hu, Y., Biasini, M., Dong, C., Guo, J., Beyermann, W., Yin, Y., 2007. One-step synthesis of highly water-soluble magnetite colloidal nanocrystals. Chem. Eur. J. 13 (25), 7153–7161. Horák, D., Karpí ek, M., Turková, J., Bene, M., 1999. Hydrazide-functionalized poly (2-hydroxyethyl methacrylate) microspheres for immobilization of horseradish peroxidase. Biotechnol. Progr. 15 (2), 208–215. Huang, S., Liao, M., Chen, D., 2003. Direct binding and characterization of lipase onto magnetic nanoparticles. Biotechnol. Progr. 19 (3), 1095–1100. Humayun, M., 2007. Artificial Sight Basic Research, Biomedical Engineering, and Clinical Advances. Hummel, W., 1999. Large-scale applications of NAD (P)-dependent oxidoreductases: recent developments. Trends Biotechnol. 17 (12), 487–492. Jia, H., Zhu, G., Wang, P., 2003. Catalytic behaviors of enzymes attached to nanoparticles: the effect of particle mobility. Biotechnol. Bioeng. 84 (4), 406–414. Johannes, T., Woodyer, R., Zhao, H., 2006. Efficient regeneration of NADPH using an engineered phosphite dehydrogenase. Biotechnol. Bioeng. 96 (1), 18–26.

Ke, R., Yang, W., Xia, X., Xu, Y., Li, Q., 2010. Tandem conjugation of enzyme and antibody on silica nanoparticle for enzyme immunoassay. Anal. Biochem. 406 (1), 8–13. Kim, J., Grate, J., Wang, P., 2006a. Nanostructures for enzyme stabilization. Chem. Eng. Technol. 61 (3), 1017–1026. Kim, J., Jia, H., Wang, P., 2006b. Challenges in biocatalysis for enzyme-based biofuel cells. Biotechnol. Adv. 24 (3), 296–308. Kim, J., Kim, B., Lopez-Ferrer, D., Petritis, K., Smith, R., 2010. Nanobiocatalysis for protein digestion in proteomic analysis. Proteomics. 10 (4), 687–699. Konwarh, R., Karak, N., Rai, S., Mukherjee, A., 2009. Polymer-assisted iron oxide magnetic nanoparticle immobilized keratinase. Nanotechnology 20, 225107. Krajewska, B., 2004. Application of chitin-and chitosan-based materials for enzyme immobilizations: a review. Enzyme Microb. Technol. 35 (2–3), 126–139. Kramer, M., Cruz, J., Pfromm, P., Rezac, M., Czermak, P., 2010. Enantioselective transesterification by Candida antarctica Lipase B immobilized on fumed silica. J. Biotechnol. 150 (1), 80–86. Lütz, S., Rao, N., Wandrey, C., 2006. Membranes in biotechnology. Chem. Eng. Technol. 29 (12), 1404–1415. Lee, D., Ponvel, K., Kim, M., Hwang, S., Ahn, I., Lee, C., 2009. Immobilization of lipase on hydrophobic nano-sized magnetite particles. J. Mol. Catal. B: Enzym. 57 (1–4), 62–66. Lin, J., Siddiqui, J., Ottenbrite, R., 2001. Surface modification of inorganic oxide particles with silane coupling agent and organic dyes. Polym. Adv. Technol. 12 (5), 285–292. Lin, S., Miyawaki, O., Nakamura, K., 1999. Continuous production of l-carnitine with NADH regeneration by a nanofiltration membrane reactor with coimmobilized l-carnitine dehydrogenase and glucose dehydrogenase. J. Biosci. Bioeng. 87 (3), 361–364. Liu, W., Wang, P., 2007. Cofactor regeneration for sustainable enzymatic biosynthesis. Biotechnol. Adv. 25 (4), 369–384. Liu, W., Zhang, S., Wang, P., 2009. Nanoparticle-supported multi-enzyme biocatalysis with in situ cofactor regeneration. J. Biotechnol. 139 (1), 102–107. Lu, Y., Yin, Y., Mayers, B., Xia, Y., 2002. Modifying the surface properties of superparamagnetic iron oxide nanoparticles through a sol–gel approach. Nano Lett. 2 (3), 183–186. Panigrahi, S., Basu, S., Praharaj, S., Pande, S., Jana, S., Pal, A., Ghosh, S., Pal, T., 2007. Synthesis and size-selective catalysis by supported gold nanoparticles: study on heterogeneous and homogeneous catalytic process. J. Phys. Chem. C 111 (12), 4596–4605. Schmid, A., Dordick, J., Hauer, B., Kiener, A., Wubbolts, M., Witholt, B., 2001. Industrial biocatalysis today and tomorrow. Nature 409 (6817), 258–268. Stengelin, M., Patel, R., 2000. Phenylalanine dehydrogenase catalyzed reductive amination of 6-(1 ,3 -dioxolan-2 -yl)-2-keto-hexanoic acid to 6-(1 ,3 -dioxolan2 -yl)-2s-aminohexanoic acid with nadh regeneration and enzyme and cofactor retention. Biocatal. Biotransform. 18 (5), 373–400. ´ Szymanska, K., Bryjak, J., Jarz bski, A., 2009. Immobilization of invertase on mesoporous silicas to obtain hyper active biocatalysts. Top. Catal. 52 (8), 1030–1036. Wan, J., Cai, W., Meng, X., Liu, E., 2007. Monodisperse water-soluble magnetite nanoparticles prepared by polyol process for high-performance magnetic resonance imaging. Chem. Commun. 2007 (47), 5004–5006. Wang, P., 2006. Nanoscale biocatalyst systems. Curr. Opin. Biotechnol. 17 (6), 574–579. Wang, S., King, C., 1979. The use of coenzymes in biochemical reactors. Adv. Biochem. Eng. 12, 119–146. Wang, W., Xu, Y., Wang, D., Li, Z., 2009. Recyclable nanobiocatalyst for enantioselective sulfoxidation: facile fabrication and high performance of chloroperoxidase-coated magnetic nanoparticles with iron oxide core and polymer shell. J. Am. Chem. Soc. 131 (36), 12892–12893. Wiseman, A., 1993. Designer enzyme and cell applications in industry and in environmental monitoring. J. Chem. Technol. Biotechnol. 56 (1), 3–13. Wu, Z., Zhang, B., Yan, B., 2009. Regulation of enzyme activity through interactions with nanoparticles. Int. J. Mol. Sci. 10 (10), 4198–4209. Wykes, J., Dunnill, P., Lilly, M., 1975. Cofactor recycling in an enzyme reactor. A comparison using free and immobilized dehydrogenases with free and immobilized NAD. Biotechnol. Bioeng. 17 (1), 51–68. Yamazaki, Y., Maeda, H., 1982. The co-immobilization of NAD and dehydrogenases and its application to bioreactors for synthesis and analysis. Agric. Biol. Chem. 46 (6), 1571–1581. Yen, M., Hsu, W., Lin, S., 2010. Synthesis of l-homophenylalanine with immobilized enzymes. Process. Biochem. 45 (5), 667–674.