Activity of glucose oxidase functionalized onto magnetic nanoparticles

Background Magnetic nanoparticles have been significantly used for coupling with biomolecules, due to their unique properties. Methods Magnetic nanoparticles were synthesized by thermal co-precipitation of ferric and ferrous chloride using two different base solutions. Glucose oxidase was bound to the particles by direct attachment via carbodiimide activation or by thiophene acetylation of magnetic nanoparticles. Transmission electron microscopy was used to characterize the size and structure of the particles while the binding of glucose oxidase to the particles was confirmed using Fourier transform infrared spectroscopy. Results The direct binding of glucose oxidase via carbodiimide activity was found to be more effective, resulting in bound enzyme efficiencies between 94–100% while thiophene acetylation was 66–72% efficient. Kinetic and stability studies showed that the enzyme activity was more preserved upon binding onto the nanoparticles when subjected to thermal and various pH conditions. The overall activity of glucose oxidase was improved when bound to magnetic nanoparticles Conclusion Binding of enzyme onto magnetic nanoparticles via carbodiimide activation is a very efficient method for developing bioconjugates for biological applications


Background
The immobilization of biomolecules onto insoluble supports is an important tool for the fabrication of a diverse range of functional materials or devices [1]. Enzyme immobilization for example, is a desired biological procedure because of the possible application of immobilized enzymes in continuous operations, product purification, and catalyst recycling [2]. Furthermore, immobilization provides many advantages such as enhanced stability, easy separation from reaction mixture, possible modula-tion of the catalytic properties, and easier prevention of microbial growth [3].
In the last decade, nanosize materials have been widely used as support for this purpose. Among these materials, magnetic nanoparticles are very popular when used in conjunction with biological materials including proteins, peptides, enzymes [4][5][6][7][8][9] antibodies and nucleic acids [8], because of their unique properties [4][5][6][7][8][9]. The ability to tract magnetically labeled entities or target organelles using magnetic force offer the opportunity to conduct biological operations with increased specificity.
Magnetite nanoparticles have been used as support material for binding of enzymes including yeast alcohol dehydrogenase [4] and lipase [5] directly via carbodiimide activation. This method brought about considerable promise because of its simplicity and efficiency. Recently, γ-Fe 2 O 3 magnetic nanoparticles were used for binding Candida rugosa lipase after acetylation of thiophene functionalized nanoparticles, or through nitroso-derivative formed on the surface of the particles by reacting nitroso tetrafluoroborate in methylene chloride. Both methods and more effectively lipase immobilized on acetylated nanoparticles exhibited long term stability. Glucose oxidase (GOX, β-D-glucose oxygen 1-oxidoreductase, EC 1.1.3.4) is a homodimer flavoprotein containing two active sites per molecule [12,13]. It catalyses the oxidation of β-D-glucose to gluconic acid, concomitant with the reduction of oxygen to hydrogen peroxide. Glucose oxidase has been used to test various types of enzyme immobilization, and is the most commonly studied in the construction of biosensors for glucose assay development [12,14,15]. A more recent study [16] examined the activity of cholesterol oxidase activity using carbodiimide activation.
Here, we report the stability and enzymatic activity of glucose oxidase immobilized onto Fe 3 O 4 magnetic nanoparticles using two binding methods, the direct binding via carbodiimide activation of amino functionalized particles and binding to thiophene-functionalized acetylated nanoparticles. A comparison of the stability and activity of glucose oxidase immobilized to magnetite using different protocols will lay the foundation for magnetic immunoassays. The size and structure of the nanoparticles were characterized using Transmission electron microscopy (TEM) and Fourier Transform Infrared (FTIR) spectroscopy, respectively. The stability, activity, and kinetic behavior of bound glucose oxidase were also examined. A second functionalization protocol using a modification of the strategy adopted to immobilize Candida rugosa lipase on the γ-Fe 2 O 3 [12] was implemented (Magnetic nanoparticles II). Briefly, 1.5 g Fe 3 O 4 was added to 5 g of 11-bromoundecanoic acid dissolved in 15 mL of ethanol. 11-bromoundecanoic acid was covalently linked to the nanoparticles surfaces by heating the mixture with microwave irradiation for 10 min. Functionalization of the particle was achieved through nucleophilic substitution with the 2-thiophene thiolate. In practice, 2-thiophene thiolate (7 mL) was added to the mixture containing the particles and heated in microwave for 5 minutes. The mixture was washed with ethanol and transferred in a round bottom flask. Acetic anhydride (4 mL) and 34.6 mL of iodine (0.01N) were successively added to the particles and agitated. The mixture was heated for 1 h under reflux condition [17]. The particles were washed several times with water, once with 10% sodium carbonate solution and finally with ethanol. The acetylated particles were reacted directly with the enzyme covalently linked to the particles via C = N bond [11].

Materials and methods
For the attachment of glucose to nanoparticles, 2 mL of the GOX solution (1000 units/mL) was added to 50 mg of functionalized magnetic nanoparticles in a test tube and sonicated at 15°C for 3 h. The supernatant containing unbound enzymes was separated from the magnetic nanoparticles using the magnetic separator, and the enzymes bound to magnetic nanoparticles were then used for activity determination. A schematic of the procedures used for both attachments are presented in Figure 1. The amount of protein in the supernatant was determined by a colorimetric method at 595 nm with the Biorad Protein Assay Reagent Concentrate using bovine serum albumin (BSA) as the protein standard. The amount of bound enzymes was calculated from: Where A is the amount of bound enzyme, C i and C s are the concentration of the enzyme initially added for attachment and in the supernatant, respectively (mg/mL) and V is the volume of the reaction medium (mL). The size of The samples for TEM analysis were prepared as follows: a drop of magnetic nanoparticles was dispersed in nanopure water. The resulting solution was sonicated for 4 min to obtain better particle dispersion. A drop of the dispersed solution was then deposited onto a copper grid and dried overnight at room temperature. The binding of GOX onto the magnetic nanoparticles was investigated using FTIR spectroscopy. Samples for FTIR analysis were prepared in phosphate buffer pH 7.4. The activity of bound GOX was determined by measuring the initial rate of formation of hydrogen peroxide at a given temperature following the formation of a red quinoneimine dye. The principle of enzymatic determination of the activity of glucose oxidase is described as follows: Glucose is oxidized by glucose oxidase to gluconate and hydrogen peroxide. Phenol + 4-AAP, in the presence of peroxidase (POX), produces a quinoneimine dye that is measured at 500 nm using a Beckman Du Spectrometer to provide an absorbance that is proportional to the concentration of glucose in the sample. The reaction is described as follows: The activity of glucose oxidase was measured as follows. An assay mixture was prepared by mixing 500 U of horseradish peroxidase, 0.015 mmol of 4-aminoantipyrine Br(CH 2 ) 10 CO 2 H -0 2 C(CH 2 ) 10 Br concentration of hydrogen peroxide following the procedure by Trinder [18]. The activity of the enzyme can be calculated using the following equation: Where ABS sample denotes the absorbance of the sample, ABS Std is the absorbance of the standard solution, and C the concentration of glucose in the sample.
The effect of temperature on the free GOX, GOX-Fe 3 O 4 I and GOX-Fe 3 O 4 II was estimated by determining the concentration of glucose in the sample at various temperatures. A solution of the assay mixture was added to the various centrifuge test tubes containing bound or free enzymes. The samples were stored in a water bath at specific temperatures (37, 50, 60, 70 and 80°C) and the absorbance was monitored at fixed time intervals to determine the glucose content. The effect of pH on GOX was monitored by measuring the initial rate of glucose oxidation by glucose oxidase in different phosphate and carbonate buffer solutions of pH (5-10) at 25°C.
The thermal stability of free GOX, GOX-Fe 3 O 4 I, and GOX-Fe 3 O 4 II were determined by measuring the residual activity of the enzyme at 25°C, after being exposed to different temperatures (37-80°C) in phosphate buffer (0.05 M, pH 7.4) for 30 min. Aliquots of the reacting solutions were taken at periodic intervals (every 30 min for 6 h) and assayed for enzymatic activity as described above. The first-order inactivation rate constant, k was calculated from the equation: where A 0 is the initial activity, A is the activity after time t (min) and k is the reaction constant.
The storage stability was examined by measuring the change in the concentration of glucose at room temperature at different time intervals (4 days

Results and discussion
Synthesis of magnetic nanoparticles with equivalent amounts of ferric and ferrous chlorides resulted in 66% of magnetic nanoparticles I and 73% of magnetic nanoparticles II. These yields suggested that the synthesis using NaOH is more advantageous for large scale production of magnetic nanoparticles. "Bare" Fe 3 O 4 I and Fe 3 O 4 II, and their GOX bound counterparts shown in the TEM micrographs in Figures (2A, B, C and 2D), respectively reveal that the particles are fine and spherical. The sizes of the particles of each sample were evaluated from 2 different TEM images. The diameter was in the range between 9 and 13 nm. The coefficient of variation between different measurements was less than 7%. There was no significant change in the size of the 'bare" particles and GOX bound particles. However, signs of agglomeration of the particles were visible in the samples. The agglomerates were not considered in the examination of the size distribution of the magnetic particles because of the assumption that agglomerated particles do not describe the original size of the particles. Figure 3 shows the distribution of the particles sizes.  Figure 4 indicating enzyme attachment onto the particles. The amino groups on the surface of the particles resulted from the use of concentrated ammonia solution during the coprecipitation of Fe 2+ and Fe 3+ as demonstrated by [5] and [19].

A B C D
The most important aspect of this study is related to the retention of biocatalytical activity of GOX after binding to magnetic nanoparticles. The amount of bound GOX was estimated using the UV-vis spectrophotometer and the catalytic activity of free and bound GOX was compared.
Kinetic parameters (V m and K m ) were estimated from the double reciprocal plots of the initial rates of glucose oxidation by GOX. The double reciprocal plots are presented in Figure 5. The Michaelis-Menten constants V max and K m for GOX are shown in Table 2. V max of free GOX, GOX-Fe 3 O 4 I and GOX-Fe 3 O 4 II were 0.731, 0.803, and 0766 µmol/min mL and, the corresponding K m values were, 0.383, 0.208, and 0.237 mM, respectively. The V max value of GOX immobilized on magnetic nanoparticles (I and II) was higher than that of the free enzyme. The highest V max was obtained with GOX-nanoparticles I. Since a low K m indicates a high degree of affinity of the enzyme to substrate [5], calculations showed that the affinity of the enzyme to the substrate increased in the order of free GOX, GOX-Fe 3 O 4 II and GOD-Fe 3 O 4 I, respectively. The high affinity of the enzyme to the substrate may be explained by a favorable change in the structural organization of the enzyme due to the immobilization procedure [20]. Consequently, the active sites of the enzymes    The effect of pH on the activities of the free and bound GOX was investigated in the pH range of 5-10 at 25°C ( Figure 6). Each data point was the average of two measurements. The coefficient of variation between measurements was between 2 and 5%. In the pH range between 6 and 7.4, the enzyme activity increased in all the systems. However, the enzyme activity was higher in GOX-Fe 3 O 4 I than in GOX-Fe 3 O 4 II and free GOX. The activity reached 100% at pH 7.4, and decreased above this pH value to 43% for GOX-Fe 3 O 4 I, and to 26% for GOX-Fe 3 O 4 II and 9% for the free GOX at pH 10. The free GOX experienced a more severe loss in activity, while GOX-Fe 3 O 4 I retained greater activity as the pH increased. In this system, the binding process occurred directly upon activation of the particle surface using carbodiimide, while the binding of GOX-Fe 3 O 4 II involved a C = N bond formed on the acetylated thiophene. It can be argued that the direct binding between the protein and the amino bond in the former case exhibited a greater resistance to a medium with higher alkalinity. This medium appeared even more constraining to the free enzyme and placed the enzyme in an electrostatic state that might affect the activity.
The effect of temperature on the activity of free GOX was examined by measuring its relative activity when stored at various temperatures. Figure 7a and 7b shows the effect of temperature on GOX-Fe 3 O 4 I and GOX-Fe 3 O 4 II at various temperatures. Each data point represents the average of duplicate measurements (coefficient of variation was less than 6 %). It can be observed that at 37°C, the enzyme retained its activity for about 80 minutes before showing a slight decrease. At 50, 60, 70 and 80°C the activity decreased as the temperature increased in both systems. In GOX-Fe 3 O 4 , the remaining activity was 23% at 50°C and 15% at 60°C after 270 min. For this time period and duration, the remaining activities were 9% and 0%, respectively for GOX-    Residual activity (%) pH mol for free-GOX, GOX-Fe 3 O 4 I, and GOX-Fe 3 O 4 II, respectively. These results show that the unbound enzyme has the highest activation energy, while GOX-Fe 3 O 4 I had the lowest. The low activation energy of GOX associated with binding to magnetic nanoparticles suggests that the energy requirement on the surfaces of the nanoparticles for enzymatic activity is reduced. Table 3 shows the inactivation rates constants (k) at 50, 60 70, and 80°C. The rate constants increased with increasing temperature in the order GOX-Fe 3 O 4 I, GOX-Fe 3 O 4 II and free GOX. Here again, binding GOX to magnetic nanoparticles minimized structural denaturation due to heat treatment. Covalent binding was expected to provide the enzyme with the protection against structural denaturation due to the unfavorable solvent-protein interactions, and thus result in activation effect [21], a possible reason for a better activity of the bound enzyme compared with the free enzyme after heat treatment. GOX-Fe 3 O 4 II had a lower stability at higher temperatures compared to GOX-Fe 3 O 4 I. The reason for this difference could reside in the stability of the binding, since the binding methods are so far the major difference between these systems. Indeed with GOX-Fe 3 O 4 I, the binding of the enzyme occurred through the amino groups on the surfaces of the particles and the carboxylic groups of proteins in the enzymes [5] which is a natural way for protein binding, while in the case of GOX-Fe 3 O 4 II the N atom to which the enzyme is attached shared a double bond with the carbon atom which is less stable than the amide bond.
Loss of storage stability is a major concern in enzyme preservation. The storage stability of the enzyme was examined for 33 days. Figure 9 shows the storage stabilities of free GOX, GOX-Fe 3

1/Tx10 3 (1/K)
measurements (coefficient of variation of the measurements was between 1 and 5%). The activity decreased with time in all the systems. Total loss of activity was observed after 20 days for the free GOX and 28 days for GOX-Fe 3 O 4 II while GOX-Fe 3 O 4 I retained 26% activity after 33 days of storage under identical conditions. The stability of the enzyme was found to improve upon binding to the magnetic nanoparticles but the most significant improvement in stability was observed for the GOX-Fe 3 O 4 I nanoparticle complex. The fixation on the surface of the magnetic nanoparticles has been a tangible argument supporting the prevention of auto-digestion of the enzyme and lysis, and the subsequent conservation of its activity [4]. This argument supports our results and justifies the long term stability of GOX-nanoparticles I and GOX-nanoparticles II over the free GOX. The efficiency of binding of GOX via carbodiimide activation over the binding by thiophene acetylation may be attributed to the potential of carbodiimide to activate the carboxylic acid side chains partially buried at the surface or in active sites of the enzyme, as well as the amino groups on the nanoparticles, favoring the formation and the stability of the amide bond [22]. This may explain why the amount of bound enzymes is higher in the binding via carbodiimide than with the thiophene acetylation. Furthermore, carbodiimide might cause cross-linking of the enzyme providing a better stability to its quaternary structure [23] and a subsequent improvement in stability.

Conclusion
Magnetic nanoparticles were synthesized by thermal coprecipitation of ferric and ferrous chlorides using two different base solutions. GOX was bound to the particles by direct attachment via carbodiimide activation and chemically via covalent attachment onto thiophene acetylated magnetic nanoparticles. Confirmation of the binding was demonstrated by FTIR spectroscopy and the sizes of the particles were characterized by TEM. The direct binding of GOX via carbodiimide activation was more effective and resulted in binding efficiency in the range between 94-100% while the binding efficiency was only between 66-72% for the GOX-Fe 3 O 4 II complex. Kinetic and stability studies showed that the enzyme activity was more preserved upon binding onto the nanoparticles when the complex was subjected to thermal and pH variations. This study shows that binding onto magnetic nanoparticles can allow the enzyme to acquire the conformational and structural arrangement for a better activity and stability, and suggests that binding of enzyme onto magnetic nanoparticles via carbodiimide activation was efficient for creating bioconjugates for a variety of applications in health and food safety.

Authors' contributions
Drs. Gilles Kouassi and Joseph Irudayaraj were the primary authors. They were responsible for the concept and experimental plan of the article. Dr Gregory MacCarty was the secondary author and contributed to the overall effort.