In Situ Encapsulation of Horseradish Peroxidase in Electrospun

NANO LETTERS

In Situ Encapsulation of Horseradish Peroxidase in Electrospun Porous Silica Fibers for Potential Biosensor Applications

2006 Vol. 6, No. 5 1042-1046

Alpa C. Patel, Shuxi Li, Jian-Min Yuan, and Yen Wei* Department of Chemistry, Drexel UniVersity, Philadelphia, PennsylVania 19104 Received February 27, 2006; Revised Manuscript Received April 6, 2006

ABSTRACT Nanoporous silica nanofibers have been employed as a matrix to encapsulate horseradish peroxide enzymes via a simple electrospinning method. A viscous solution of prehydrolyzed tetramethyl orthosilicate, β-D-glucose, poly(vinyl alcohol), and enzymes were employed as spinning solution to generate porous fibers in the form of nonwoven mats. The silica fiber mats thus produced have a high surface area because of the small diameter (100 to 200 nm) of the fibers as well as the extreme porosity (2 to 4 nm) of individual fibers caused by the glucose template present in them. The high surface area, mechanical flexibility, thermal stability, reusability, and freedom of encapsulating various enzymes make porous silica nanofibers excellent biosensors.

Various biotic substances are used as sensors for a wide range of applications from detecting hazardous chemicals in the atmosphere to measuring glucose in a human body.1 Even with many successes, utilizing free enzymes for biosensor applications often involves problems of poor stability and reusability of protein molecules.2 To counteract these problems, polymer and sol-gel materials have been utilized to immobilize enzymes, which has proven to increase the storage and thermal properties of enzymes moderately.3,4 However, there are still a number of issues, such as timeconsuming multiple steps for encapsulating and drying enzyme-immobilized materials, small surface area, and finally enzyme leakage, which limit the biosensor applications of such materials.5,6 There have been many efforts to make nanofibers that contain enzymes through the electrospinning process.7-10 Some attractive features, which make electrospun nanofibers interesting candidates for enzyme immobilization, include large surface area, reusability of the fiber mats due to easy separation from the reaction mixture, and retention of catalytic activity due to evaporation of harmful solvents from fibers during electrospinning. Surface stabilization is one of the techniques studied for enzyme immobilization via electrospinning.11,12 In this technique, only the external surface of the electrospun fibers is utilized for enzyme immobiliza* Corresponding author. Herman B. Wagner Professor of Chemistry and Director, The Center for Advanced Polymers and Materials Chemistry, Department of Chemistry, Drexel University, Philadelphia, PA 19104. Tel: (O) (215)895-2650; (Lab) -1644, -2979, and -2982. Fax: (215)8951265. E-mail: [email protected]. 10.1021/nl0604560 CCC: $33.50 Published on Web 04/20/2006

© 2006 American Chemical Society

tion, while the internal surface remains as such. The availability of limited enzyme attachment sites on the surface of the fibers causes low enzyme loading. A loss of enzyme activity was noticed because of the limited mobility and direct exposure of the surface-immobilized enzymes to harsh chemicals and high temperatures. Direct electrospinning of enzymes along with various polymers is another method used to encapsulate and protect enzymes inside polymer fibers. This technique causes limited interaction of enzymes with the substrates because of the complete confinement of most of the enzyme molecules inside the nonporous fibers. Some of these fiber mats employed for enzyme encapsulation are water soluble and hence tend to swell and disintegrate when immersed in solvents, causing enzyme leakage, decreased thermal stability, and nonreusability of the fiber mats, thus requiring extra steps to protect the fibers via coating or crosslinking. Attempts made to cross link the enzyme-encapsulated fibers cause reduced porosity (among overlaying fibers in an nonwoven fiber mat) and conformational changes in the enzyme molecules, which in turn cause limited accessibility of substrates to the active sites of the enzymes, thus causing loss in enzyme activity.13-18 Keeping in mind all of the obstacles faced by researchers to date as mentioned above, a novel method is proposed here in which a biosensor based on direct electrospinning of enzyme-encapsulated mesoporous silica fibers (EIMSF) has been accomplished. This biosensor is a unique combination of two nanotechnologies, that is, the direct encapsulation of enzymes in nanoporous silica matrix19 and the electrostatic nanofiber spinning.20 The unique features that differentiate

and make EIMSF a superior candidate for biosensor applications include (1) high surface area due to the nanometerrange control of mesopore formation inside the silica fibers (e.g., 2 to 10 nm or greater) facilitating substrate diffusion, (2) no deformation of the silica fibers with little or no swelling, (3) no enzyme leakage, (4) high thermal stability, (5) increased reusability, and finally (6) the freedom to encapsulate various kinds of enzymes individually or as mixtures in a short period of time. In this work, tetramethyl orthosilicate (TMOS) was used as a silica precursor, poly(vinyl alcohol) (PVA, Mn 30 000) was used as an extender that assists in easy spinning of silica fibers, and β-D-glucose not only functions as a nonsurfactant template for controlling mesoporosity19,21 of the resultant sol-gel silica fibers but also helps in the fiber spinning by increasing viscosity. The enzyme immobilized inside the mesoporous silica fibers was horseradish peroxidase (HRP). The HRP enzyme is robust and resistant to high temperature and can be used as a glucose sensor.21,22 A typical procedure for preparing 40 wt % glucose and 4 wt % PVA-templated EIMSF included two steps. The first was the preparation of the spinning solution. Thus, 0.77 g of TMOS precursor (accounts for 0.3 g of silica) was prehydrolyzed using 0.25 g of H2O and 30 µL of 40 mM HCl catalyst with continuous stirring for 30 min.21 The reaction mixture was then heated to 60 °C for 30 min to evaporate the methanol produced during hydrolysis. The sol was then cooled to 0 °C and 0.2 g of glucose (dissolved in water to make a 50 wt % solution), 0.01 g of PVA (dissolved in water to make a 8 wt % solution), and 1 mg of commercially available HRP in 1 mL of phosphate buffer (100 mM, pH 6.5) were added with constant stirring to yield the spinning solution. The second step was standard electrospinning20,23 of nanofibers from the silicate-glucose-enzyme sol thus prepared. The sol was filled in a syringe and subjected to an electric filed of 20 kV for electrospinning. A grounded aluminum plate was placed at a distance of 15 cm from the tip of the syringe as a fiber collector. Under the electric field a droplet suspended from the tip of the syringe acted as a feed source from which a charged jet of enzyme-immobilized silica sol solution was ejected out, leaving behind approximately 0.5 g of white, flexible, dry fiber mats containing approximately 1 mg of HRP. Three sets of fibers and three control samples are listed in Table 1. Scanning electron microscopy (SEM, Phillips XL30; Figure 1) results represent enzyme-immobilized silica nanofibers containing 0, 20, and 40 wt % glucose (0EIMSF, 20EIMSF, and 40EIMSF), respectively. The diameter of the fibers is seen to decrease with increase in weight percent of glucose. This change in morphology could be due to the entanglement of glucose molecules with each other as well as with PVA and hydrolyzed TMOS, thus increasing the viscosity of the sol above a critical value required for spinning fibers.24,25 The fibers were then washed thoroughly with phosphate buffer solution for 3 h to extract out the glucose template and PVA polymer. The complete removal of the template and PVA was monitored by FTIR (Perkin-Elmer 1600) and Nano Lett., Vol. 6, No. 5, 2006

Table 1. Material and Activity Parameters of Free HRP and EIMSF sample IDa free HRP 0EIMSF 20EIMSF 40EIMSF 40EIMSF2 40EIMSF (noPVA) 40EIMSF (noHRP)

Vmb Kmb Sc Dd FDe (mg-1 min-1) (mmol L-1) (m2 g-1) (Å) (nm) 500 63 163 227 210 220

0.65 0.3 0.56 0.62 0.60 0.59

80 224 500 450 420 490

7 26 38 35 29 35

500 350 100 200 400 110

a 0, 20, and 40EIMSFs are fibers containing 0, 20, and 40 wt % glucose; 96, 76, and 56 wt % silica; a constant amount of PVA (4 wt %); and HRP. Control samples (1) 40EIMSF2 is similar to 40EIMSF except for the average fiber diameters, (2) 40EIMSF (noPVA) is similar to 40EIMSF except for the absence of PVA, and (3) 40EIMSF (noHRP) is similar to 40EIMSF except for the absence of HRP. b Vm and Km were obtained from the Lineweaver-Burk plots. c Specific surface area was determined by nitrogen adsorption-desorption isotherms d Pore diameter of the fibers was determined by nitrogen adsorption-desorption isotherms. e Average diameters of EIMSF were obtained by measuring about 500 fibers of a single sample from SEM micrographs.

thermogravimetric analysis (TA Q50 TGA) (see Supporting Information Figures 2S and 3S). The buffer solution used to extract glucose and PVA was found to have no enzyme activity, indicating that there was no enzyme leakage from the fibers during template extraction. Such template-extracted fibers did not show any surface morphological changes at macroscopic levels (Figure 1). These fibers were then analyzed for surface area and mesoporosity (Micromeritics ASAP 2010). As anticipated,21 a higher concentration of glucose template used (e.g., 0, 20, or 40 wt %) led to a larger mesopore size (e.g., 0.7, 2.6, or 3.8 nm, respectively) and surface area (Table 1). Figure 2 represents transmission electron microscopy (TEM, JEOL 2010F) cross sections of glucose- and PVA-extracted fibers. The samples for TEM studies were prepared by embedding irregular clusters of template-extracted fibers in resin. The resin was then temperature cured to a hard form and microtomed into 100nm-thick cross sections. Because of the different mechanical properties of the fibers with that of resin, the fibers tend to stick together and lose their spherical shape during sample preparation. The TEM micrographs reveal interconnected mesopores and channels (lighter features in the micrograph), similar to those reported previously.19,21,26 The amount of HRP used in the initial synthesis of the silica fibers was 1 mg commercial enzyme per 0.3 g of template-extracted dry mesoporous silica fiber mats. After the immobilization, the amount of enzyme was further determined by enzymatic assay of the nanofiber samples against activity of 1 mg HRP enzyme in free solution using a standard colorimetric method.26 In this method, templateextracted EIMSF samples were first immersed in a colorless reagent mixture [1 mL phosphate buffer (100 mM, pH 2 to 10), 2 mL aqueous phenol (PhOH, 172 mM)-4-aminoantipyrine (4-AAP, 2.46 mM, colorless dye) and 2 mL of H2O2 (2 to 10 mM)]. The HRP enzyme present inside the nanoporous fibers catalyzed the conversion of H2O2 to water, along with the conversion of the colorless dye to a red color (quinoneimine). The rate of color change was recorded by 1043

Figure 1. Scanning electron micrographs of EIMSF containing (a) 0 wt % glucose (scale bar: 10 µm); (b) 20 wt % glucose (scale bar: 10 µm), and (c) 40 wt % glucose (scale bar: 1 µm). (d) 40 wt % glucose-extracted fibers (scale bar: 2 µm). The inset shows the internal surface morphology of the fibers at macroscopic level.

Figure 2. Low-resolution transmission electron micrograph (TEM) of a cluster of template-extracted cross-sectioned 40EIMSF, scale bar: 0.30 µm. High-resolution TEM of the same, scale bar; 40 nm. The light features in the micrograph represent interconnected mesopores and channels.

Figure 3. Enzyme activity assay of template-extracted EIMSF in colorless reagent mixture [phosphate buffer, phenol (PhOH), 4-aminoantipyrine (4-AAP, colorless dye) and H2O2]. The enzyme present inside the porous fibers converts H2O2 to water in the process converting 4-AAP to quinoneimine (red-colored dye).

using a UV-vis absorbance spectroscopy (Perkin-Elmer λ2S) at 510 nm, and the data thus collected was used to calculate the enzyme activity (see Supporting Information page 8). The Michaelis-Menton plots of HRP-catalyzed reactions are shown in Figure 4A. From these plots, the Michaelis constant (Km) and maximal velocity (Vm) values3 (Table 1) were obtained. The results clearly indicate that the reaction rates under the catalysis of HRP enzyme encapsulated in mesoporous fibers (i.e., 20EIMSF and 40EIMSF) are significantly higher than those in microporous fibers (0EIMSF). This is consistent with our hypothesis that large pore diameters in the host matrix can facilitate substrate diffusion 1044

and, therefore, enhance the apparent enzymatic activity. Apparently, the average diameter of the EIMSF and the addition of PVA in the fibers during electrospinning do not affect the catalytic activity drastically (as seen in 40EIMSF, 40EIMSF2, and 40EIMSF/noPVA in Table 1). It is noted that the specific activity (Vi) value in all cases (Figure 4A) drops slowly with further increase in [H2O2] to greater than 2 mM, which is commonly attributed to reagent inhibition effect.27 The Vm values of encapsulated HRP in fibers were smaller than that of free enzyme because of the small loss of enzyme, which usually takes place during sol-gel immobilization technique.28 The effect of pH on the relative Nano Lett., Vol. 6, No. 5, 2006

Figure 4. (A) Effect of [H2O2] on the initial velocity of EIMSF catalysis at a constant pH of 6.5 and (B) effect of pH on the relative activity of free and EIMSF at constant [H2O2] of 1 mM.

nanometer dimensions, silica fiber mats were quite flexible compared to sol-gel silica in the form of small disks, films, or powders, thus contributing further toward biocatalytic applications. Acknowledgment. This work was supported in part by the National Institutes of Health (Grant no. DE09848), the US Army Research Office, and the Nanotechnology Institute of Southeastern Pennsylvania. Figure 5. Thermal stability (remaining activity %) of the EIMSF after subjecting to heat treatment from 25 to 80 °C for 30 min in phosphate buffer (pH 6.5).

activity of free and immobilized enzymes is shown in Figure 4B. The data are normalized by relating the highest data point on every curve to 100%. The optimum activity in all cases was observed around pH 6.5, typical of HRP behavior.29 It is interesting that the immobilized enzyme seems to maintain a higher relative activity than the free enzyme. The thermal stability of HRP enzymes was studied by heating test tubes of free and fiber immobilized enzymes in phosphate buffer (pH 6.5) in the range of 25 to 80 °C for 30 min in a water bath, cooling, and then conducting an enzyme activity assay. The activity of free enzymes decreased rapidly with increase in temperature because of the thermal denaturation.26,30 In contrast, the encapsulated enzymes showed an increase in activity up to 60 °C and then dropped slightly at 80 °C (Figure 5). This improved thermal stability could be caused by the space confinement in the host matrix that prevents protein unfolding.31,32 Another important feature of EIMSF is the reusability of the fibers. After use in one assay, the fiber mats were washed thoroughly with buffer and reused in another assay. There was little drop in activity. The retention of activity was found to be up to 90% for most of the immobilized fiber samples during the fifth cycle. In conclusion, for the first time, mesoporous silica-based nanofibers have been fabricated and employed as a matrix for the immobilization of enzymes via a simple electrospinning method. The drastically improved features and versatility of this technique is a huge step toward biosensor and biocatalytic applications. The fibers with mesoporosity show fourfold greater activity than the conventional nontemplated silica samples (Table 1) and threefold greater activity than HRP immobilized silica powders.33 Finally, because of the Nano Lett., Vol. 6, No. 5, 2006

Supporting Information Available: Complete characterization data, FTIR and TGA graphs, and detailed enzymatic activity results. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Arnold, M. J. Am. Chem. Soc. 1997, 119, 255. (2) Bornscheuer, U. T. Angew. Chem., Int. Ed. 2003, 42, 3336-3337. (3) Avnir, D.; Braun, S.; Lev, O.; Ottolenghi, M. Chem. Mater. 1994, 6, 1605-1614. (4) Dave, B. C.; Dunn, B.; Valentine, J. S.; Zink, J. I. Anal. Chem. 1994, 66, 1120A-1127A. (5) Huang, S.; Jiang, Z.; Wu, H.; Xu, S. Gaofenzi Tongbao 2003, 2835. (6) Dunn, B.; Zink, J. I. Chem. Mater. 1997, 9, 2280-2291. (7) Kenawy, E. R.; Bowlin, G. L.; Mansfield, K.; Layman, J.; Simpson, D. G.; Sanders, E. H.; Wnek, G. E. J. Controlled Release 2002, 81, 57-64. (8) Wang, X. w.; Hu, Z. m.; Pan, W. l.; Liu, Z. f. Donghua Daxue Xuebao, Ziran Kexueban 2005, 31, 115-119. (9) Lee, S. W.; Belcher, A. M. Nano Lett. 2004, 4, 387-390. (10) Li, D.; Wang, Y.; Xia, Y. Nano Lett. 2003, 3, 1167-1171. (11) Jia, H.; Zhu, G.; Vugrinovich, B.; Kataphinan, W.; Reneker, D. H.; Wang, P. Biotechnol. Prog. 2002, 18, 1027-1032. (12) Hsieh, Y. L.; Wang, Y.; Chen, H. Polym. Prepri. (Am. Chem. Soc., DiV. Polym. Chem.) 2003, 44, 565-566. (13) Herricks, T. E.; Kim, S. H.; Kim, J.; Li, D.; Kwak, J. H.; Grate, J. W.; Kim, S. H.; Xia, Y. J. Mater. Chem. 2005, 15, 3241-3245. (14) Sawicka, K.; Gouma, P.; Simon, S. Sens. Actuators, B 2005, B108, 585-588. (15) Wu, L.; Yuan, X.; Sheng, J. J. Membr. Sci. 2005, 250, 167-173. (16) Xie, J.; Hsieh, Y. L. J. Mater. Sci. 2003, 38, 2125-2133. (17) Zeng, J.; Chen, X.; Liang, Q.; Xu, X.; Jing, X. Macromol. Biosci. 2004, 4, 1118-1125. (18) Zeng, J.; Aigner, A.; Czubayko, F.; Kissel, T.; Wendorff, J. H.; Greiner, A. Biomacromolecules 2005, 6, 1484-1488. (19) Wei, Y.; Xu, J.; Feng, Q.; Lin, M.; Dong, H.; Zhang, W. J.; Wang, C. J. Nanosci. Nanotechnol. 2001, 1, 83-93. (20) Doshi, J.; Reneker, D. H. J. Electrost. 1995, 35, 151-160. (21) Wei, Y.; Jin, D. L.; Ding, T. Z.; Shih, W. H.; Liu, X. H.; Cheng, S. Z. D.; Fu, Q. AdV. Mater. 1998, 10, 313-316. (22) Yamanaka, S. A.; Nishida, F.; Ellerby, L. M.; Nishida, C. R.; Dunn, B.; Valentine, J. S.; Zink, J. I. Chem. Mater. 1992, 4, 495-497. 1045

(23) Patel, A. C.; Li, S.; Wei, Y.; Wang, C.; Zhang, W. Polym. Prepr. (Am. Chem. Soc., DiV. Polym. Chem.) 2004, 45, 553-554. (24) McKee, M. G.; Wilkes, G. L.; Colby, R.; Long, T. E. Macromolecules 2004, 37, 1760-1767. (25) McKee, M. G.; Layman, J. M.; Cashion, M. P.; Long, T. E. Science 2006, 311, 353-355. (26) Wei, Y.; Dong, H.; Xu, J.; Feng, Q. ChemPhysChem 2002, 3, 802808. (27) Baynton, K. J.; Bewtra, J. K.; Biswas, N.; Taylor, K. E. Biochim. Biophys. Acta 1994, 1206, 272-278. (28) Choi, Y. J.; Chae, H. J.; Kim, E. Y. J. Biosci. Bioeng. 1999, 88, 368-373.

1046

(29) Chattopadhyay, K.; Mazumdar, S. Biochemistry 2000, 39, 263-270. (30) Waldner, J. C.; Lahr, S. J.; Edgell, M. H.; Pielak, G. J. Biopolymers 1999, 49, 471-479. (31) Ping, G.; Yuan, J. M.; Sun, Z.; Wei, Y. J. Mol. Recognit. 2004, 17, 433-440. (32) Samuni, U.; Navati, M. S.; Juszczak, L. J.; Dantsker, D.; Yang, M.; Friedman, J. M. J. Phys. Chem. B 2000, 104, 10802-10813. (33) Dong, H. PhD. Dissertation, Drexel University, 2002; Chapter 4, pp 104-133.

NL0604560

Nano Lett., Vol. 6, No. 5, 2006