Structural stability of glucose oxidase encapsulated in liposomes to inhibition by hydrogen peroxide produced during glucose oxidation

Biochemical Engineering Journal 30 (2006) 158–163

Structural stability of glucose oxidase encapsulated in liposomes to inhibition by hydrogen peroxide produced during glucose oxidation Makoto Yoshimoto ∗ , Mitsunobu Sato, Shaoqing Wang, Kimitoshi Fukunaga, Katsumi Nakao Department of Applied Chemistry and Chemical Engineering, Faculty of Engineering, Yamaguchi University, 2-16-1 Tokiwadai, Ube 755-8611, Japan Received 18 October 2005; received in revised form 6 February 2006; accepted 10 March 2006

Abstract The glucose oxidase (GO) consists of two identical subunits each of which contains noncovalently bound flavin adenine dinucleotide (FAD) cofactor. GO is known to be inactivated due to hydrogen peroxide (H2 O2 ) produced in the oxidation of glucose. In our previous paper, the liposomal GO showed a much higher stability to H2 O2 than the free enzyme. In this work, to deduce the structure and state of the liposomal GO, the fluorescence properties of the tryptophan residue and FAD cofactor in free GO during the glucose oxidation were measured for its tertiary structure and redox state, respectively. The tryptophan fluorescence data revealed that the initial glucose concentration lower than 0.6 mM resulted in almost no alteration in the tertiary structure, while the higher concentration did in a remarkable change in the structure due to the increase in catalytic turnover. On the other hand, the FAD fluorescence data showed that the reduced FAD was accumulated in the initial stage of the reaction. When glucose was completely consumed, the FAD restored the initial oxidized form for the initial glucose concentrations lower than 0.6 mM, whereas for the higher concentrations the reduced FAD tended to form an inactive complex with H2 O2 leading to the deactivated enzyme. In the case of the liposomal GO at even such a high initial glucose concentration as 10 mM, the glucose concentration inside liposome was previously estimated to be lower than 0.2 mM due to its low permeability to glucose. Consequently, the formation of the inactive complex was proved to be effectively depressed in the liposomal GO reaction. © 2006 Elsevier B.V. All rights reserved. Keywords: Liposomal glucose oxidase; Tryptophan fluorescence; FAD fluorescence; Enzyme tertiary structure; Enzyme inactivation; Glucose permeability

1. Introduction Enzyme molecules encapsulated in the phospholipid vesicles, i.e., liposomes are shown to be modified in their reactivity and stabilized compared to free enzyme [1]. We reported the highly stable liposomal glucose oxidase in catalyzing a prolonged oxidation of glucose producing gluconic acid and hydrogen peroxide (H2 O2 ) [2]. In the active glucose oxidase (GO) molecule, two tightly bound flavin adenine dinucleotide (FAD) molecules are stabilized through their interactions with the particular aromatic amino acid residues in GO [3]. H2 O2 is known to competitively inhibit the activity of GO through forming the inactive complex between the reduced form of FAD cofactor and H2 O2 produced in the catalytic turnover of GO [4]. Recently, we reported the stable liposomal GO in catalyzing the oxidation of glucose [5]. The mechanism for the highly stabilized liposomal GO was

∗

Corresponding author. Tel.: +81 836 85 9271; fax: +81 836 85 9201. E-mail address: [email protected] (M. Yoshimoto).

1369-703X/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2006.03.004

revealed mainly focusing on the role of the catalase contained in the commercially available GO. The catalase was stabilized by encapsulating in liposomes and catalyzed the decomposition of H2 O2 produced continuously as opposed to free CA. It was also reported that, when the rate of the liposomal GO reaction was increased through enhancing the glucose permeability across the liposome membrane, the inhibition of the GO activity became pronounced due to the increased accumulation of H2 O2 . The mechanistic details on the above deactivation of GO has been remained unknown at the enzyme molecular level. Since the liposomal aqueous phase is very small in volume and surrounded by fragile lipid bilayer membranes, the structure of enzyme present in liposomes are difficult to directly observe. One of the effective approaches for analyzing the glucose oxidation catalyzed by liposomal enzyme would be to examine the free enzyme reaction which proceeds under a similar condition to the liposomal system. For the liposomal GO-catalyzed oxidation of glucose, the lower glucose concentration inside liposomes compared to that outside liposomes was suggested to be the

M. Yoshimoto et al. / Biochemical Engineering Journal 30 (2006) 158–163

additional critical feature in analyzing the structural stability of the GO molecule [6]. In the present work, the fluorescence measurements [3,7] were performed to evaluate the redox state of FAD cofactor as well as the local tertiary structure around the FAD molecules in GO molecules during the free GO-catalyzed glucose oxidation. In particular, the effects of the glucose concentration on the fluorescence properties of the enzyme were dynamically examined in order to elucidate the high stability of the liposomal GO to H2 O2 on a GO molecular basis. 2. Materials and methods 2.1. Materials Glucose oxidase (GO) from Aspergillus niger (EC 1.1.3.4) was purchased from Toyobo Co. Ltd. (Osaka, Japan). POPC (1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) was obtained from Avanti Polar Lipids Inc. (Alabaster, AL, USA). ␤-dGlucose was obtained from Wako Pure Chemical Industries Ltd. (Osaka, Japan). 2.2. Preparation of glucose oxidase-containing liposomes (GOL) The GOL was prepared by the extrusion technique with the 50 mM Tris–HCl/100 mM NaCl buffer at pH 7.4 (denoted as Tris buffer) as reported previously [2,5]. The enzyme-free liposomes were also prepared. 2.3. Oxidation of glucose catalyzed by glucose oxidase (GO) The oxidation of glucose was initiated at 27 ◦ C in the Tris buffer of pH 7.4 by adding glucose solution to the GO solution (1.3 mg GO/mL solution) or GOL suspension ([lipid] = 1.0 mM, 1.1 mg GO/mL suspension) in a test tube to give the initial glucose concentration of 10 mM. The total reaction volume was 1.5 mL. The time course of the glucose conversion and the remaining GO activity were then measured. For the liposomal GO reaction, the glucose conversion was calculated from the remaining glucose concentration measured with the enzymatic method. This method was applicable to the sample containing the negligible amount of H2 O2 because it measures the H2 O2 produced from glucose [5]. The assay samples from the liposomal GO reaction were reported to contain almost no H2 O2 due to the liposome-stabilized catalase activity in the commercially available GO [5]. For the free GO reaction accumulating H2 O2 , in which the moles of glucose consumed were almost equal to moles of H2 O2 accumulated, the conversion was calculated from the H2 O2 concentration measured with the titanium sulfate method [4]. For the remaining GO activity measurements, the free GO or GOL was separated from the reaction solution containing glucose as well as the products H2 O2 and gluconic acid by using the GPC. The GO activity was measured in the same way as previously described [2,5,8]. The intrinsic GO activity in the GOL was measured after solubi-

159

lization of liposomes with an excess amount of cholate. The remaining GO activity was determined as the intrinsic activity at the certain reaction time relative to that before initiating the reaction. 2.4. Measurement of tryptophan fluorescence The intrinsic tryptophan fluorescence was measured for the free GO during catalyzing the glucose oxidation in the Tris buffer of pH 7.4 at 27 ◦ C. The emission spectra of 1 mL of the GO reaction solution above at the fixed GO concentration of 1.3 mg/mL were periodically recorded between 300 and 400 nm at the excitation wavelength of 280 nm [9] using a spectrofluorometer (JASCO FP-750). Time course of the tryptophan fluorescence spectra during the GO-catalyzed glucose oxidation was observed for the different initial glucose concentrations ranging from 0.05 to 20 mM. All the measurements were carried out in the cuvette thermostatted at 27 ◦ C using the perche-type temperature controller (JASCO ETC-272T). The amount of catalase contained in the commercially available GO was much smaller than that of GO so that catalase had negligible effect on the fluorescence measurements of GO. 2.5. Measurement of flavin adenine dinucleotide (FAD) fluorescence A GO molecule contains two FAD molecules of which oxidized state is essential for the catalytic activity of GO. To estimate the redox state of the FAD within GO molecules during the glucose oxidation, the time course of the emission fluorescence spectra of FAD was measured between 480 and 560 nm at the excitation wavelength of 365 nm [9] at the fixed GO concentration of 1.3 mg/mL for the different initial glucose concentration in the cuvette. The intact FAD fluorescence spectrum of the GO was measured at 24 h after incubation of GO in the Tris buffer of pH 7.4 at 27 ◦ C in the absence of glucose. The reaction, i.e., measurement temperature was controlled at 27 ◦ C throughout the experiment in the same way as in the tryptophan fluorescence measurement above. 3. Results and discussion 3.1. Activity of free and liposomal glucose oxidase during catalyzing glucose oxidation Fig. 1(a) and (b) shows the time courses of the remaining GO activity and glucose conversion during the prolonged oxidation of glucose catalyzed by the free GO (1.3 GO mg/mL solution) and by the liposomal GO ([lipid] = 1.0 mM, 1.1 mg GO/mL suspension), respectively, at the initial glucose concentration (G0 ) of 10 mM. It is seen in the figure that more than 90% of free GO is deactivated within 24 h, while more than 90% of the activity of liposomal GO is maintained during the reaction period of 10 days. The glucose conversion is 85% at 24 h for the free GO reaction and 83% at 10 days for the liposomal GO reaction. It was reported that the free GO deactivation was caused by the accumulation of an inactive complex formed between the

160

M. Yoshimoto et al. / Biochemical Engineering Journal 30 (2006) 158–163

Fig. 1. Time course of remaining glucose oxidase (GO) activity (closed squares) and glucose conversion (open squares) in oxidation of glucose catalyzed by (a) free GO (1.3 mg GO/mL solution) and (b) liposomal GO ([lipid] = 1.0 mM, 1.1 mg GO/mL liposome). Initial glucose concentration, pH and reaction temperature were 10 mM, 7.4 and 27 ◦ C, respectively, in both reactions.

reduced state of GO and H2 O2 [4], which were both formed in the initial stage of the reaction. The results shown in Fig. 1 suggest that a much larger initial reaction rate in the case of free GO-catalyzed reaction causes the higher deactivation of GO due to the H2 O2 accumulated more rapidly. However, the mechanism for the deactivation of free GO at its molecular level has not fully been understood. Recently, we have revealed that the liposomal GO is highly stable in its activity even at the later stage of the reaction partly because of the presence of a small amount of catalase contained in the commercially available GO [5]. The catalase was remarkably stabilized in the liposomes and could decompose H2 O2 formed continuously, keeping the enzyme in the liposome almost free of the inhibition. To further increase the reactivity and stability of the liposomal GO, it is required to clarify the effect of the higher reaction rate on the stability of the liposomal GO. In the following, the redox state of FAD cofactor as well as the tertiary structure around the FAD in GO are examined at the various G0 values, which correspond to the different reaction rates in the liposomes with the various membrane permeation rates of glucose.

Fig. 2. Time course of tryptophan fluorescence spectra for free GO in catalyzing oxidation of glucose in the Tris buffer of pH 7.4 at 27 ◦ C at the initial glucose concentration G0 of (a) 0.5 mM and (b) 20 mM. The concentration of GO was 1.3 mg/mL.

3.2. Relation of tertiary structure of glucose oxidase and redox state of FAD within the enzyme to glucose concentration Fig. 2(a) and (b) shows the time courses of the tryptophan fluorescence spectra measured at the G0 values of 0.5 and 20 mM, respectively. It has been reported that four aromatic residues from tryptophan and tyrosine are located near the FAD within the GO molecules [3]. It is seen in Fig. 2(b) that, at G0 = 20 mM, the fluorescence intensity (ITrp ) increases with time in the initial stage of the reaction accompanied by a red shift in the wavelength of maximum fluorescence (λmax ). The increase in the ITrp value is generally caused by the conformational change of GO molecules, and the red shift of λmax means the increased water accessibility of the tryptophan residue. As shown in Fig. 2(b), the spectrum at 24 h is similar in both ITrp and λmax to the spectrum observed in the absence of glucose. At 24 h, most of the initially charged glucose was consumed. These results suggest that the extent of perturbation in the tertiary GO structure around FAD was dependent on the glucose concentration in the reaction solution. In a separate experiment, no practical change in the tryptophan fluorescence spectra was observed throughout a

M. Yoshimoto et al. / Biochemical Engineering Journal 30 (2006) 158–163

161

earlier stage of the reaction, the formation and accumulation of the reduced FAD (FAD·H2 ) during the catalytic process are suggested to cause the decrease in the IFAD value. Since the IFAD value remains decreased at 24 h as shown in Fig. 3(b), the reaction-induced alteration in the IFAD value for free GO is proved to occur irreversibly at G0 = 20 mM. Almost all of the GO molecules are considered to be deactivated based on the result shown in Fig. 1. Thus, the IFAD values obtained in the 24-h reaction that are much smaller than the IFAD value of the oxidized FAD suggest that most of the GO molecules are present as the inactive form of GO. At G0 = 0.5 mM (Fig. 3(a)), IFAD was initially decreased followed by being increased to the level of the oxidized FAD. This means that little deactivation of GO was induced in the almost whole reaction process initiated at G0 = 0.5 mM. The structure of the interface region between the GO subunits was shown to be related to the redox state of the FAD cofactor [7]. The tryptophan and FAD fluorescence measurements (Figs. 2 and 3) quantitatively show the interrelation between the tertiary structure of GO and the redox state of FAD during catalyzing the oxidation of glucose. 3.3. Effect of initial glucose concentration G0 on change in structure and redox state of free glucose oxidase during catalyzing oxidation of glucose

Fig. 3. Time course of FAD fluorescence spectra for free GO in catalyzing oxidation of glucose in the Tris buffer of pH 7.4 at 27 ◦ C at the initial glucose concentration G0 of (a) 0.5 mM and (b) 20 mM. The concentration of GO was 1.3 mg/mL.

24 h-incubation of GO in the presence of 10 mM H2 O2 without glucose, although about 40% of the GO molecules were deactivated at 24 h in the H2 O2 -containing solution. This result is consistent with the result shown in Fig. 2(b) where the tryptophan fluorescence spectrum of GO at 24 h was almost restored to that obtained without glucose in spite of the presence of H2 O2 as in the former case. In contrast, no significant change in the tryptophan fluorescence spectra is seen in the reaction initiated at G0 = 0.5 mM (Fig. 2(a)). This result confirms that the tertiary structure of GO is dominated by the glucose concentration. At such a low G0 value, free GO is seen to catalyze the oxidation of glucose keeping its stable conformation. Fig. 3(a) and (b) shows the time courses of the FAD fluorescence spectra measured at the G0 values of 0.5 and 20 mM, respectively. As seen in Fig. 3(b) (G0 = 20 mM), the intensity of FAD fluorescence (IFAD ) is significantly decreased at the initial stage of the reaction. The oxidized FAD is known to be the active form and a redox change in the FAD molecules to occur in the oxidation of glucose. Considering the increased catalytic turnover of GO due to the higher glucose concentration in the

Fig. 4(a) shows an effect of G0 on the maximum value of ITrp , ITrp,max of free GO obtained from the ITrp data for the different values of G0 (0.05–20 mM) at 5 min and 24 h after initiating the respective glucose oxidations. It is found in the figure that there is a critical value of G0 (see vertical dotted line) at which the intramolecular GO interaction for maintaining its tertiary structure is significantly altered at the reaction time of 5 min. The structures of GO for G0 higher than 0.6 mM are seen to be almost the same as the structure for G0 = 20 mM (Fig. 4(a)). A small decrease in the ITrp value (Fig. 4(a)) as well as a small blue shift of λmax (data not shown) were observed at 24 h with increasing G0 from 5 to 20 mM. These suggest that the tryptophan residues moved into a little hydrophobic environment, which may be caused by their intermolecular association in the presence of a more highly concentrated glucose. Notably, at G0 ≤ 0.5 mM, both the ITrp,max (Fig. 4(a)) and λmax values (data not shown) are almost unchanged with varying G0 regardless of the reaction time, meaning that GO molecules can keep their compact and stable conformation throughout the reaction period in such a low G0 range. Analogously to the above figure, Fig. 4(b) shows the effect of G0 on the maximum IFAD value, IFAD,max at 5 min and 24 h after initiating the reaction. The IFAD,max value is seen to decrease from the IFAD value (57 a.u.) for the fully oxidized FAD within the reaction time of 5 min regardless of G0 due to the reactioninduced formation of the reduced FAD. The IFAD,max value is almost restored to the level of the original oxidized FAD (57 a.u.) at 24 h in the case of G0 ≤ 0.5 mM. On the other hand, in the case of the higher G0 , the IFAD,max values at 24 h are still lower than the above value of the oxidized FAD. The persistent formation of the inactive FAD at 24 h (Fig. 4(b)) and the significant change in

162

M. Yoshimoto et al. / Biochemical Engineering Journal 30 (2006) 158–163

glucose concentration inside liposomes was lower than 0.2 mM throughout the reaction period. This was due to the permeability barrier of the liposome membrane to glucose, even at the initial glucose concentration of 10 mM. Therefore, it follows that the liposomal GO undergoes little shift of the tertiary structure (Fig. 4(a)) and redox state (Fig. 4(b)) to the unfavorable structure and state, respectively. In addition, as previously reported [5]. the liposomal catalase is effective to depress the accumulation of H2 O2 produced during the prolonged liposomal GO reaction initiated at high glucose concentrations. The liposomal reaction system is thus suitable for stabilizing the tertiary structure and oxidized FAD of GO and for effectively reducing the inhibitory effect due to H2 O2 accumulated in the prolonged oxidation of glucose. The result shown in Fig. 4 would be utilized as a criterion to obtain the highest possible reactivity of the liposomal GO by increasing the liposome membrane permeability to glucose keeping the highest possible stability of the enzyme simultaneously.

4. Conclusions The results obtained in the present work are summarized as follows:

Fig. 4. Effect of initial glucose concentration on maximum (a) tryptophan and (b) FAD fluorescence intensities, measured at 5 min (open squares) and 24 h (closed squares) after initiating free GO-catalyzed reaction in the Tris buffer of pH 7.4 at 27 ◦ C. The concentration of GO was 1.3 mg/mL. The vertical dotted lines represent the critical value of G0 (0.6 mM) at which the maximum tryptophan and FAD fluorescence intensities are changed significantly.

the GO tertiary structure at 5 min (Fig. 4(a)) are seen to become pronounced at almost the same critical value of G0 (0.6 mM). This means that the local conformational alteration around the FAD cofactor within GO is closely linked to the formation of the inactive complex. The reason for the decrease in the IFAD,max values of GO at 24 h with increasing G0 from 5 to 20 mM is probably that the association of the partially denatured enzyme with H2 O2 is more pronounced at the higher G0 . This is also indicated in Fig. 4(a) where the ITrp,max at 24 h is decreased with increasing G0 from 5 to 20 mM. These observations elucidate that the structure and activity of free GO are effectively maintained by regulating the glucose concentration appropriately, for example, in such a way that G0 < 0.6 mM as in the present experiment. It should be noted that free GO was deactivated during the glucose oxidation in the enzyme-free liposome suspension ([POPC] = 1.0 mM) containing glucose added at the initial concentration of 10 mM. Therefore, it is not the interaction between the lipid membranes and GO molecules but the encapsulation of GO in liposomes that is responsible for the high stability of the liposomal GO. Our kinetic model developed previously for the liposomal GO-catalyzed reaction [6] showed that the

(1) Free GO was readily inactivated during catalyzing the oxidation of 10 mM glucose, suggesting the formation and accumulation of the inactive complex formed between the reduced FAD and H2 O2 in the initial stage of reaction. In the liposomal GO system, in marked contrast, little deactivation of the GO in liposomes was observed throughout the prolonged reaction period. (2) The tertiary structure of GO was dependent on the glucose concentration as revealed by the tryptophan fluorescence analysis. The structure of GO was perturbed significantly at the high initial glucose concentration G0 , followed by restoring the original GO structure with the consumption of glucose. The redox state of FAD cofactor within GO was predominantly changed into its reduced form without any activity in such a high G0 range from 5 to 20 mM. (3) An appreciable decrease in the fluorescence intensity of FAD within free GO, indicating the accumulation of an inactive complex of GO, was observed at the G0 value higher than ca. 0.6 mM. At the lower G0 , the reactive form of the oxidized FAD was restored as glucose was converted. The formation of the inactive complex of GO, causing the enzyme inactivation, was suggested to be closely related to the local conformational perturbation around the FAD cofactor as revealed by the tryptophan fluorescence analysis for free GO. It was concluded that GO was stabilized in the liposomal reaction system because the glucose concentration in liposomes was kept low enough irrespective of the concentration in the liquid bulk due to the permeability barrier of the liposome membrane to glucose.

M. Yoshimoto et al. / Biochemical Engineering Journal 30 (2006) 158–163

Acknowledgment This work was supported in part by the Japan Society of the Promotion of Science (No. 17760625).

[5]

References

[6]

[1] P. Walde, S. Ichikawa, Enzymes inside lipid vesicles: preparation, reactivity and applications, Biomol. Eng. 18 (2001) 143–177. [2] S. Wang, M. Yoshimoto, K. Fukunaga, K. Nakao, Optimal covalent immobilization of glucose oxidase-containing liposomes for highly stale biocatalyst in bioreactor, Biotechnol. Bioeng. 83 (2003) 444–453. [3] D. Zhong, A.H. Zewail, Femtosecond dynamics of flavoproteins: charge separation and recombination in riboflavine (vitamin B2)-binding protein and in glucose oxidase enzyme, Proc. Nat. Acad. Sci. U.S.A. 98 (2001) 11867–11872. [4] J. Bao, K. Furumoto, M. Yoshimoto, K. Fukunaga, K. Nakao, Competitive inhibition by hydrogen peroxide produced in glucose oxida-

[7] [8]

[9]

163

tion catalyzed by glucose oxidase, Biochem. Eng. J. 13 (2003) 69– 72. M. Yoshimoto, Y. Miyazaki, M. Sato, K. Fukunaga, R. Kuboi, K. Nakao, Mechanism for high stability of liposomal glucose oxidase to inhibitor hydrogen peroxide produced in prolonged glucose oxidation, Bioconjugate Chem. 15 (2004) 1055–1061. M. Yoshimoto, S. Wang, Y. Arimatsu, K. Furumoto, K. Fukunaga, K. Nakao, A kinetic model for glucose oxidation catalyzed by immobilized glucose oxidase-containing liposomes in a mini-scale external loop airlift bubble column, J. Chem. Eng. Jpn. 37 (2004) 1012– 1018. A. Haouz, C. Twist, C. Zentz, P. Tauc, B. Alpert, Dynamic and structural properties of glucose oxidase enzyme, Eur. Biophys. J. 27 (1998) 19–25. A.S.T. Huggett, D.A. Nixon, Use of glucose oxidase, peroxidase, and o-dianisidine in determination of blood and urinary glucose, Lancet 273 (1957) 368–370. S.K. Haq, Md.F. Ahmad, R.H. Khan, The acid-induced state of glucose oxidase exists as a compact folded intermediate, Biochem. Biophys. Res. Commun. 303 (2003) 685–692.