WO2007050100A2 - Immobilized enzymes and processes for preparing and using same - Google Patents

Abstract

A highly loaded immobilized enzyme conjugate is prepared from a nanopore silica gel support. The enzyme may be immobilized in an amount, based on the silica gel support, of at least about 500 mg enzyme per gram of silica gel. The immobilized enzyme conjugate may be prepared by a method which includes the following steps: (a) surface modifying silicic acid by reacting the silicic acid with a basic aminoalkylsilane in solution in an alcoholic medium, wherein the amount of the basic aminoalkylsilane is sufficient to raise the pH of the solution to induce gelation and form a chemically surface-modified silica gel; (b) activating the chemically surface-modified silica gel by reaction with a cross-linking agent; and (c) contacting the activated chemically surface-modified silica gel with an enzyme. Immobilized thermolysin has been used to prepare a low molecular weight protamine with high efficiency.

Description

IMMOBILIZED ENZYMES AND PROCESSES

FOR PREPARING AND USING SAME

CROSS REFERENCE TO RELATED APPLICATION

[0001] This application claims the benefit of U.S. Provisional Application No.

60/630,213, filed November 24, 2004. This application, in its entirety, is incorporated herein by reference.

FIELD OF THE INVENTION

[0002] In one embodiment, this invention relates to immobilized enzyme conjugates. In another embodiment, the invention relates to methods to prepare the immobilized enzyme conjugate. In another embodiment, the invention relates to chemically modified silica composites which contain functional species that have enhanced utility and to methods for producing same. In another embodiment, the invention relates to the use of such immobilized enzymes in a porous silica matrix as bioreactors, such as for enhancement in production and methods for separation of proteins. In another embodiment, the invention provides methods to control the morphology of the composite materials to enhance performance in bioreactors and in chromatography and affinity separations applications. In another embodiment, the invention relates to production of low molecular weight protamine fragments using a bioreactor according to the present invention.

BACKGROUND OF THE INVENTION

[0003] Because of their ability to improve reaction rates and product specificity, enzymes are a significant part of many industrial processes [I]. They are extensively used in the manufacturing of pharmaceutical drugs and fine chemical production, among many other applications. The use of enzymes in manufacturing processes does possess several drawbacks; enzymes are a relatively expensive reaction component and therefore increase production costs. They are also extremely sensitive to environment conditions and can be easily denatured. Any manufacturing process must consider the conservation of the enzymes within the reactor while also maintaining their activity.

[0004] Often, the most costly step in enzyme production is the downstream separation and purification of the desired protein [2]. Therefore, use of these methods for enzyme recovery from a homogeneous catalysis process can significantly increase production costs. One method of lowering these processing requirements is to use a heterogeneous catalyst - an immobilized enzyme [3-5].

[0005] A method that has long been studied, and has been put into some practice, is the immobilization of enzymes on a solid support. There are three basic methods for enzyme immobilization: entrapment, cross-linking and carrier binding [6]. Entrapment, as the name suggests, is basically the trapping of the enzyme within a cage-like, porous network that allows the substrate to diffuse through. The relatively large enzymes are unable to, or very slowly, diffuse out of the solid network and are therefore retained. Cross-linking is achieved through intermolecular bonding between enzyme molecules. Carrier binding involves the formation of interactions between enzyme and a support. This method can be further categorized as ionic binding, physical adsorption or covalent binding depending on the method. [7]

[0006] A variety of immobilization materials are available for use in both laboratories and industrial processes with many of these materials based on organic polymers (e.g. agarose, dextran, etc.) which fall under the category of "soft gels" [7, 8]. The major drawback of soft gels is their low mechanical strength and in some cases low thermal and chemical stability [9]. Low mechanical strength materials can only sustain a small pressure drop and thus they are limited to low solution flow rates when operated as a packed bed reactor (PBR). This can be a major drawback in industrial processes. The mechanical strength can also limit column size and thereby affect the reactor's efficiency.

[0007] The use of silica in enzyme immobilization according to embodiments of the present invention can provide one or more of the following benefits:

• High surface areas (500 - 1000 m²/g) provide the possibility of high enzyme loadings on the matrix. This would increase the rate of reaction, measured as the product produced over time.

• Greater mechanical strength would allow the use of column reactors at higher flow rates without matrix distortion.

• A silica matrix provides surface hydroxyl groups that can be readily used for a variety of surface modifications. A number of functionalities (thiol, amine, etc) may be introduced on the surface at high loadings (e.g., 7.5 mmole/g of silica for thiol modification).

• An hydrophilic silica surface would minimize fouling problems often encountered with hydrophobic "soft gels". Organic molecules may interact with the surface of these organic gels, creating a film on the surface and interfering with enzyme activity [10, H].

• An open pore morphology of silica gels allows substrates to quickly move into the interior regions of the particle.

• Solvents used in the processing of silica materials may be

environmentally benign.

[0008] Although enzymes have been immobilized on silica gels, the vast majority of the prior art have been done on pre-formed gels. Immobilization on pre-formed gels involve a number of steps, typically including, for example, 1) activation of the gel surface, 2) surface modification with a reactive functional group, 3) attachment of a cross-linking agent, and 4) introduction of the enzyme. This method has been

successfully used for enzyme immobilization; however the loadings have been limited. When the activated gel is introduced to the enzyme solution, enzyme immobilization first occurs at the pore openings. At higher loadings, the enzymes can effectively block the pores preventing enzyme diffusion into the interior. A more effective method would be desirable for the immobilization of enzymes.

[0009] The present inventor(s) and/or co-workers for the assignee have been involved in applying nanotechnology for recovering precious and toxic metals from aqueous waste streams. These efforts resulted in a chemically surface modified silica gel (CSMG) having high surface area with extremely high loading of functional groups that increases adsorption efficiency and capacity. The CSMG may be obtained by reacting a freshly prepared silica gel, which contains many silanol groups (Si-OH) on a large surface area (500-1000 m²/g), with a silane coupling reagent ( e.g., 3-mercaptopropyl- trimethoxysilane, Si(OCH₃) ₃-(CH₂) ₃-SH ) and have applied these nanoadsorbent based CSMG materials to achieve selective adsorption of silver from photographic waste. A description of the preparation of CSMG is found in the commonly assigned U.S.

Application Serial No. 601,888, filed August 9, 2000, the disclosure of which is incorporated herein, in its entirety, by reference thereto. In the present invention, CSMG functionalized with amino groups to a high loading is prepared and then activated with an aldehyde and subsequently the enzyme is infiltrated into the wet gel where it is immobilized.

[0010] Often enzymes are relatively expensive compared to their substrates and are lost in the homogenous reaction process. One method of cost reduction is to recycle and reuse the enzyme from the product stream. Separation can be greatly facilitated by immobilizing the enzyme on a solid matrix. Potential users for the product according to the embodiments of the invention vary from research labs, to academia and large pharmaceutical companies. The engineered, silica based material according to embodiments of the present invention may provide property improvements and should be ideal for large-scale processes. Such materials can find use in heterogeneous catalysis, affinity chromatography, membrane reactors, bio-sensors and drug delivery, among many others.

[0011] Embodiments of this invention encompass methods for the efficient immobilization of enzymes on silica. According to one embodiment of the invention significantly improved loadings and/or simplification of the process may be achieved.

[0012] Specific embodiments of the present invention include:

• surface modification of silica gel with amino functional groups to a high loading, e.g., at least about 500 mg enzyme per gram of silica gel; more preferably, at least about 700 mg enzyme per gram of silica gel;

• immobilization of enzymes, e.g., invertase, to optimize immobilization techniques for other enzymes;

• immobilization of thermolysin for Low Molecular Weight Protamine

(LMWP) production; and

• method to control the silica gel morphology. Controlled regular particle shape and size is advantageous [12].

[0013] Embodiments of the present invention provide effective methods for the immobilization of bio-molecules onto silica surfaces. As an example, an immobilization method is described for the enzyme invertase, which hydrolyzes sucrose into glucose and fructose. As a further example, the immobilization of thermolysin, a protease, is used for the production of Low Molecular Weight Protamine (LMWP) from native protamine.

[0014] In another embodiment, a method is provided to control the silica gel morphology.

[0015] Controlled regular particle shape and size is advantageous. For a packed- bed reactor or chromatographic support, the matrix should have a regular shape and be of uniform size. Spherical particles produced according to embodiments of this invention may be preferred because of improved packing; flow characteristics showing a minimum back-pressure; with maximum surface to volume ratio providing a smaller diffusion path. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1: N₂ adsorption isotherms for air dried silica gel samples, wherein

N-CSMG and GA-N-CSMG samples were air dried at room temperature for 72 hours and then degassed under vacuum for a minimum of two hours at 100 ⁰C prior to measurement of N₂ adsorption.

[0017] Figure 2: Nitrogen adsorption isotherms of freeze-dried silica gels, wherein N-CSMG and GA-CSMG suspensions in deionized water were frozen in liquid nitrogen and then lyophilized; samples were degassed under vacuum at 100 °C until adsorbed moisture was removed prior to analysis.

[0018] Figure 3: Elution profile of protamine fragments using a heparin column and NaCl gradient.

[0019] Figure 4: Determination of product purity by heparin affinity

chromatography.

[0020] Figure 5: Relative immobilized activity (■) and immobilization efficiency (• ) variation with invertase solution concentration. The relative activity plateaus with a 10 g/L invertase solution.

[0021] Figure 6: Optimization of immobilization solution volume yielded a maximum immobilization efficiency (• ) of 71% and relative immobilized activity (■ ) of 90% at an enzyme solution volume of 10 mL.

[0022] Figure 7: Effect of temperature on initial activities of free (■ ) and immobilized (• ) invertase.

[0023] Figure 8: Elution profiles obtained from the repeated use of immobilized thermolysin.

[0024] Figure 9: LMWP Bioreactor: Syringe filter loaded with silica particles containing immobilized thermolysin.

[0025] Figure 10: FPLC set-up for digestion of protamine by immobilized thermolysin.

[0026] Figure 11: Elution profile of protamine digested by immobilized thermolysin reactor at 4°C with a flow rate of 20 μL/min.

[0027] Figure 12: LMWP production and purification by an HPLC system.

[0028] Figure 13: Digestion of protamine by immobilized thermolysin at 25°C.

[0029] Figure 14: Protamine digestion by immobilized thermolysin at 25⁰C and [0030] Figure 15: Emulsion scheme for formation of 3-aminoρropyl(triethoxy)- silane modified silica spheres.

[0031] Figure 16: Micrographs of amino-modified silica spheres formed by a water-in-oil emulsion process at mixing rates of a) 690 RPM and b) 1040 RPM.

[0032] Figure 17: Particles size distribution obtained by the emulsion process with stirring rates of 690 (• ), 860 (♦ ), and 1040 (■ ) RPMs.

[0033] Figure 18: Adsorption of Cu(II) by N-CSMG and spherical N-CSMG particles from a 1000 ppm solution initially at pH 4.5.

[0034] Figure 19: Indicating properties of Thymol blue adsorbed on SDS treated

N-CSMG at a) pH 1.5, b) pH 7.0, and c) pH 11.5.

DETAILED DESCRIPTION OF THE INVENTION

[0035] As used herein, the term "enzyme" refers to a protein that catalyzes at least one biochemical reaction. A compound for which a particular enzyme catalyzes a reaction is typically referred to as a "substrate" of the enzyme. Enzymes typically have molecular weights in excess of 5000.

[0036] In general, six classes or types of enzymes (as classified by the type of reaction that is catalyzed) are recognized: Oxidoreductases; Transferases; hydrolases;

Lyases; Isomerases; Lipases. Any of these classes of enzymes may be used in various embodiments of the present invention.

[0037] The enzyme may be either of a single type or a system consisting of two or more types. The enzyme may include at least one lysing enzyme selected from among lysozyme, chitinase, protease, glucose oxidase, glucanase, .beta.-galactosidase, endo-

.beta.-N-acetylgluco-saminidase and endolysin.

[0038] Enzymes of the present invention may be obtained from natural (such as animals, plants, bacteria, yeast, fungi or virus) or synthetic sources (such as peptide synthesizer or expression vector). Natural sources include Candida species (such as

Candida antarctica), Pseudomonas species (such as Pseudomonas fluorescens or Mucor species, such as Mucor miehei).

[0039] In one embodiment, hydrolytic enzymes, such as, lipase, invertase, esterase or protease, may be used as the immobilized enzyme.

[0040] According to one embodiment of the invention, a silica precursor is used under acidic conditions (pH <3). This condition leads to increased hydrolysis, compared to condensation, and weakly branched polymers that contain a large amount of silanol groups. The addition of a basic catalyst increases the condensation rate and within a few minutes results in the formation of a gel. This process has allowed for the significant improvement in the surface modification of silica gels with various functional groups while also minimizing the shrinkage observed after drying. Another variable that can be adjusted to vary gel structure is the reaction temperature. An increase in reaction temperature leads to a greater degree of branching and therefore a less densely packed structure [13]. Room temperature is used for the fabrication of all silica gels in the following example. The water and catalyst content also determines the rates of hydrolysis and condensation and these, in turn, vary the polymer structure from linear to weakly branched [14]. The co-condensation approach requires fewer processing steps and the material obtained has a more uniform, controlled distribution of the functional group [15, 16]. Although both synthesis routes may be used to prepare ordered, mesoporous silica gels, the co-condensation process has generally been shown to yield higher degrees of surface modification [17].

[0041] Compared with other sol-gel processes that start directly from a colloidal silica process using silicic acid starts with a much lower ionic strength. The low ionic charge is critical to the modification reaction because of an improved solubility for the organic ligand components [18, 19]. Freshly prepared silicic acid is composed of silica particles with very small particle size (2 - 10 nm) and, thus, a very large surface area and plenty of active silanol groups. A silane coupling reagent may be used to incorporate ligand groups on the particle surface and the modified silica sol can be gelled quickly with an adjustment of pH. Two general methods which are available for the chemical modification of mesoporous silicas include: (1) post-synthesis grafting [20] and (2) co- condensation [16]. Post-synthesis grafting involves the reaction of organosilanes with silanol groups present on the matrix surface. Generally this method results in an inhomogeneous surface coverage of the functional group with concentration at the outer surface and pore entrances [21].

[0042] In the co-condensation process, organosilanes are mixed with the silica sol prior to gelation and condensation between the sol and organosilanes results in a silica matrix with a chemically active surface. A homogenous distribution of functional groups can be achieved with the co-condensation process [21] with higher loading densities [17, 22]. However, practically all co-condensation processes to date make use of

tetraethoxysilane (TEOS) or tetramethoxysilane (TMOS) as the source of silica. Only recently has a report been published detailing the synthesis of mesoporous organosilicas from sodium silicate in a co-condensation process [23]. Yu et. al. vigorously mixed acidified solutions of sodium silicate and organosilanes and allowed hydrolysis to occur for one hour. The solution temperature was then raised to 50 °C and NaF added to induce gelation. With this method methyl, mercapto, and vinyl surface modifications were introduced to mesoporous silica gels.

[0043] Although there are many reports using TEOS to produce amino- functionalized silica gels [24-26], only one reference has been found for a co- condensation process which uses sodium silicate [27]. They report a maximum of 50 mol% functionalization of available sites with 3-aminopropyltrimethoxysilane.

[0044] While tremendous advances have been made in the fabrication of composites with a high surface area, high functional loading, and controlled morphology, the immobilization of biological specie on these composites has not seen a significant improvement compared to materials of the past. Efficient immobilization of biologies requires not only a matrix with designed properties, but also the proper protocols and processing conditions. Design of such a system requires a proper understanding of the underlying process and an ability to control it.

[0045] According to one aspect of the invention, there is provided an immobilized enzyme conjugate comprising a nanopore silica gel support and immobilized thereon an enzyme in an amount, based on the silica gel support, of at least about 500 mg enzyme per gram of silica gel. In a preferred embodiment of this aspect of the invention, the immobilized enzyme conjugate comprises a nanopore silica gel supporting at least about 700 mg enzyme per gram of silica gel immobilized thereon.

[0046] According to another aspect of the invention, there is provided a method for producing an immobilized enzyme conjugate. According to this aspect of the invention the method comprises the following steps: (a) surface modifying silicic acid by reacting the silicic acid with a basic aminoalkylsilane in solution in an alcoholic medium, wherein the amount of the basic aminoalkylsilane is sufficient to raise the pH of the solution to induce gelation and form a chemically surface-modified silica gel; (b) activating the chemically surface-modified silica gel by reaction with a cross-linking agent; and (c) contacting the activated chemically surface-modified silica gel with an enzyme. In accordance with one embodiment of this aspect of the invention, a solution of sodium silicate is ion-exchanged to produce silicic acid. In a particular embodiment of this aspect of the invention, surface modification may be accomplished, for example, by introducing 3-aminopropyltriethoxysilane (APTES) to the system, preferably at a small molar excess, e.g., at a ratio of about 1.1 mols per mol silicic acid, which is then stirred into the system with, e.g., a magnetic stir bar. APTES, with its basic amino group, acts as a base-catalyst to induce gelation of the system while it also undergoes the hydrolysis and condensation reactions to be introduced into the network of the forming gel. This embodiment produces chemically surface-modified gels (CSMG) with high loadings and uniform distribution of the amino functional group (N-CSMG). In additional

embodiments of this aspect of the invention, the method can be easily scaled-up by simply increasing the amount of reactants and can be used to produce gel monoliths of desired shapes with the use of an appropriate mold.

[0047] As mentioned earlier, a significant issue in working with mesoporous materials is shrinkage and the corresponding pore collapse. With the realization that permanent gel shrinkage occurs due to the condensation of surface silanol groups, it has been proposed that the shrinkage of gels can be minimized through modification of the silica gel while it is in the wet state. In the wet state, the pore size and specific surface areas are at their maximum. Any attempt to dry the gel, even partially, leads to shrinkage. Therefore, in order to take advantage of the fully open pores, all surface modification reactions should be conducted with the wet gel. To test this hypothesis, we produced N- CSMG gels and a portion of the N-CSMG was further reacted with glutaraldehyde, to produce GA-N-CSMG. Samples taken from both N-CSMG and GA-N-CSMG were then air dried at room temperature, freeze dried, or supercritically dried. Each sample was then further dried under vacuum at 100 ⁰C for a minimum of 2 hours until all adsorbed moisture was removed and then analyzed for surface area. The air dried and

supercritically dried samples were analyzed using a NOVA 1200, Version 6.0 High Speed Gas Sorption Analyzer from Quantachrome Corporation (Boynton Beach, USA).

Analysis of freeze dried samples was conducted by Quantachrome Corporation on an Autosorb Automated Gas Sorption System.

[0048] Pore size distribution is determined by the model developed by Barret,

Joyner and Halenda (BJH model) based on the Kelvin equation, corrected for multilayer adsorption [28]. The pore sizes are divided into three categories based on diameter:

micropores (< 2 nm), mesopores (2 - 50 nm), and macropores (> 50 nm). Mesopores of size >30 nm are ideal for materials used in enzyme immobilization due to improved access of the interior surfaces [6].

[0049] Fig. 1 shows the nitrogen adsorption isotherms for the air dried N-CSMG and GA-N-CSMG samples. Because the y-axis contains the term (IAV), materials that adsorb a higher nitrogen weight, W, show a smaller slope value but possesses greater surface areas. From the plot it is clear that a significant difference is seen in the nitrogen adsorption isotherms of the unmodified N-CSMG and GA-N-CSMG upon air drying. Fig. 2 shows a similar difference between N-CSMG and GA-N-CSMG when the samples are freeze dried. When comparing the results of the freeze and air dried samples, it must be noted that the y-axis in Fig. 2 is only half the scale of that shown in Fig. 1.

Table 1: Specific surface area of N-CSMG and GA-N-CSMG after air drying, freeze drying, and supercritical drying as determined by the BET method.

[0050] Surface area analysis of the material before and after modification has shown considerable difference in material properties. The air dried GA-N-CSMG has a surface area five times greater than the corresponding N-CSMG while the freeze dried GA-N-CSMG shows an almost 50-fold increase in surface area compared to N-CSMG. It is also interesting to note that freeze drying only slightly improves the N-CSMG surface area, compared to air drying, while an increase greater than 13 -fold is seen for GA-N- CSMG. An improvement against shrinkage is clearly seen when the N-CSMG is modified in the wet state. It is likely that the glutaraldehyde works similar to the templating agents commonly used for control of pore morphology in sol-gel processes. Since glutaraldehyde is a bi-functional molecule, it is also possible that it binds to the silica matrix on both ends and braces the gel against the collapsing stresses.

[0051] N-CSMG shows a weight loss of approximately 23 percent according to thermogravimetiic (TGA) analysis. The approximately 15% difference in weight loss, compared to unmodified silica, is directly related to the degree of APTES modification of the gel. With the TGA weight loss profiles, the silanol and APTES in the silica gels can be quantified and the loading densities calculated, assuming homogenous distribution with a surface area of 1250 m²/g-Siθ₂ determined from the surface area of supercritical dried GA-N-CSMG (see Fig. 2). The results are summarized in Table 2 with the loading expressed based on the amount of residual SiO₂, sample weight remaining at the end of TGA, to allow for comparisons.

Table 2 Quantity of silanol and amino groups as determined by TGA.

[0052] The loading densities of silanol and APTES indicate that a large fraction of the active surface area has been amino-modified (approximately 80%) with this sol-gel protocol. Also, with an APTES footprint of 50 A² [29], the surface loading of 3.64 mmol/g SiO₂ is almost 88% of the maximum theoretical APTES loading of 4.16 mmol/g SiO₂ (Equation 1).

[0053] The N-CSMG amino loading of 3.6 mmol/g is significantly higher than most published results and should show improved capacity for enzyme immobilization.

[0054] While the TGA data shows an APTES loading of 3.64 mmol/g SiO₂, the glutaraldehyde loading is more than twice this amount at 7.54 mmol/g SiO₂. This higher loading is likely due to the presence of polymerization between glutaraldehyde molecules [30-32]. Dinitrosalicylic acid was used to determine the presence of the aldehyde groups by their reducing power [33]. The measured reducing power of GA-N-CSMG is equivalent to 3.12 ± 0.5 mmol glutaraldehyde per gram of GA-N-CSMG.

[0055] An array of functionally active proteins and other biological entities, including catalytic antibodies, antigens, live cells, and plant spores have been

encapsulated by sol- gel silica and its hybrids. A thorough review paper provides a clear picture of the current status of the bioencapsulation technology [34]. According to the review, significant advantages of silica encapsulation are: excellent optical and mechanical properties, high resistance to (biological, chemical, and thermal) degradation, simple fabrication, enhanced activities, long and stable shelf life, and application versatilities. However, this review also highlights some substantial hurdles for future technology development. These hurdles include:

• Identification of biocompatible precursors/protocols to prevent denaturation during encapsulation;

• Shrinkage and pore collapse during xerogel formation; and

• Porosity and mechanical stability improvement.

[0056] Covalent immobilization of enzymes, including onto silica surfaces, has been done for a number of years [35]. These methods commonly involve the use of prefabricated silica gels that have chemically inactive surfaces [36]. The gel surface must be activated, with strong acids, to produce the reactive surface hydroxyl groups. The activated surface is then modified with silane couple reagents, such as

aminopropyltriethoxy-silane (APTES), which link functional groups (e.g. amino) to the silica surface via siloxane bonds [37]. The cross-linking agent (e.g. glutaraldehyde) is then introduced to react with the modified surface and finally the enzyme is added. In addition to being process intensive, these methods have been limited to relatively low enzyme loadings.

[0057] Since pre-fabricated silica gels are used, they depend upon diffusion for introduction of the surface modifying agent, cross-linking agent, and the enzyme into the gel interior. A problem that arises, particularly when attempting to obtain a high loading, is the blockage of pores by the diffusing species [38, 39]. Naturally, reactions first occur on the gels outer surface and at pore entrances. At high loadings, the buildup of enzymes at pore entrances can effectively clog the pore preventing further immobilization in the interior regions.

[0058] Many enzyme immobilization techniques involve S_N2-type, nucleophilic substitution reactions. A typical enzyme immobilization system can consist of four different parts: the matrix; a surface modifying agent, if the matrix does not possess reactive functional groups; a cross-linking agent, which attaches the enzyme to the matrix; and, the enzyme. An ideal situation would allow immobilization to occur by simply mixing all matrix and reactive precursors together with the enzyme.

Polymerization of the matrix would then allow for enzyme immobilization at high loadings and even distribution. This is not possible in the vast majority of cases because non-specific interactions between the reactive groups can result in the formation of undesirable bonds.

[0059] A commonly used cross-linking agent, which reacts with amino groups, is the homo-bifunctional molecule-glutaraldehyde. The aldehyde reacts with amine groups to produce imines, or Schiff-bases. In the presence of enzyme and an amino-modified surface, glutaraldehyde can immobilize the enzyme by attaching to an enzyme amino group with one functional group, and to the amino surface with the other. However, if each species were in solution prior to matrix formation, the non-specific nature of the aldehyde reaction will create cross-linked enzymes and also the cross-linking of two amino groups on the surface modifying agent. In addition, the interaction of enzymes with small, reactive molecules can lead to enzyme deactivation [6]. Although a "one-pot" process is preferable, a serial processing method, by introducing the reactive groups one at a time, is used because it minimizes these side-reactions.

[0060] Invertase is a highly used enzyme in the food industry for the production of sweeteners used in beverages, jams, and as artificial honey [I]. As an enzyme, invertase catalyzes cleaving of the α-1,4 glycosidic bonds of sucrose to produce glucose and fructose. In this invention, one embodiment shows that a much higher invertase loadings have been achieved in comparison to previously published results.

Immobilization of enzymes onto a matrix requires compatibilization of the enzyme in the presence of all precursors. The invertase-sucrose system was chosen as an ideal model to study the immobilization capacity of the GA-N-CSMG material. The product formation is followed using the 3,5-dinitrosalicylic acid (DNS) method [33]. Unlike sucrose, glucose and fructose are both reducing sugars. The reduction of DNS by glucose and fructose results in a color change which can be measured spectrophotometrically to determine the product concentration.

[0061] Using the N-CSMG protocol as previously discussed with optimized immobilization steps, immobilized invertase activities of more than 246,000 U/g of GA- N-CSMG have been achieved with a 50 g/L substrate concentration (pH 4.5, 45 ⁰C). This is a significant increase in immobilized activity when compared to published results. TGA analysis has confirmed the high loading of invertase at 0.723g invertase per gram GA-N-CSMG. While the weight of invertase immobilized is significantly improved compared to other published results, it is still less than 25 percent of the theoretical maximum based on the enzyme's Stokes radius. Also, the immobilized invertase exhibited a maximum reaction rate, V_max, only 73 percent of free invertase, an increase in substrate affinity, K_m ^'1, compensated for this and no significant difference in catalytic efficiency is seen.

[0062] In a further embodiment, a bioreactor is fabricated, with the

immobilization of the enzyme thermolysin on GA-N-CSMG, for the production of a pharmaceutical product.

[0063] Protamine is commonly used clinically as an antagonist to heparin, an anticoagulant. However, protamine has been shown to induce some adverse reactions in patients, including some fatalities due to cardiac arrest [40-42]. The discovery of specific sequences within larger protein molecules which are accountable for the transduction abilities of proteins could potentially have significant implications for a broad range of biomedical applications. The development of low molecular weight protamine (LMWP) fragments as nontoxic substitutes for protamine in clinical heparin neutralization has been reported in the prior art [43-45]. These LMWP peptides, which display a high arginine fraction in their amino acid compositions, retain the ability to neutralize heparin yet have significantly less toxicity in vivo [46]. LMWP has also been shown to be equally potent, as protamine, in neutralization of low molecular weight heparins (LMWH), but with the benefit of markedly reduced toxicity.

[0064] Production of LMWP is achieved by thermolysin-catalyzed digestion of native protamine. For the production of such bioactive compounds, an enzyme-mediated reaction is often preferred over a chemical synthesis route because of its simplicities, selectivity, and specificity, as well as its moderate reaction conditions. Furthermore, the immobilization of thermolysin onto a nanoporous support will allow for continuous flow production of LMWP.

[0065] The enzymatic digestion of native protamine yields a series of LMWP species. The current preparation method utilizes a heparin affinity column to separate LMWP into five fractions termed TDSPl to TDSP5 according to the ascending order of their elution ionic strength which correlates with an ascending order in arginine units ranging from 5 to 10. Literature evidence indicates a correlation between the number of arginine units in the peptide sequence to cell membrane transduction ability with an optimum value from 6 tol5 arginine units [47]. Indeed, work by Futaki et. al. clearly indicated the effect of the length of arginine chain on the cell internalization [48]. In their work, peptides composed of even number (4-16) residues of arginine were prepared and tested. The Gly-Cys-amide segment was attached to their C termini for fluorescein labeling. Considerable difference was recognized on the translocation efficiency and intracellular localization among these peptides. R4 showed extremely low translocation activity while R6 and R8 exhibited the maximum internalization and accumulation in the nucleus. The degree of internalization decreased as the chain length was further increased. It has been proposed that the uptake of arginine-rich peptide is mediated by the guanidine head group in the arginine through the formation of a hydrogen bond between arginine and lipid phosphate in the membrane [48-50]. The five LMWP groups, TDSP1-5, contain arginine units of 4, 5, 6, 9, and 10, respectively.

[0066] In this embodiment, a continuous flow reactor using thermolysin immobilized on nanoporous silica was developed for the production of TDSP5. Enzyme encapsulation or immobilization has been commonly utilized for easy removal from a reaction, dispersion into insoluble liquid medium, and stabilization under various environments. By using a nanoporous silica support, the pore morphology, surface ligand groups, and the inherent large surface area can be exploited to achieve high enzyme loadings and, therefore, a reactor of high enzymatic activity can be developed.

[0067] In contrast to peptide synthesis that uses the standard (9-fluorenylmethyl) chloroformate solid-phase chemistry, the thermolysin digestion method approach has the common enzyme-substrate specificity, and therefore, can produce consistent peptide sequences with simplicity and of high quality. Prepared peptides sequences can then be separated and purified by routine heparin-affinity chromatography. The most significant advantage of the LMWP system would be the production of peptides in mass quantities unachievable by conventional peptide synthesis methods. Protamine digestion can be accomplished using free thermolysin. A 10 mg/mL protamine solution, buffered at pH 7.5 by 50 mM TRIS/HCl, is incubated at 37 °C with 1% (w/w) thermolysin and 2mM calcium chloride. The digestion is allowed to proceed for 30 minutes and then stopped by addition of a chelating agent, such as ethylenediamine tetraacetic acid (EDTA) and cooling, e.g., in an ice bath. The desired peptide is then obtained by standard purification methods.

[0068] LMWP purification using a 5 mL HiTrap Heparin column containing a total of approximately 50 mg heparin was known. In this known method, columns are placed between two Alltech 526 HPLC Pumps and a Linear UVIS 200 detector equipped with a flow-through cell. A 1 mL sample loop is connected to a two-position, six-port sample injector. The HPLC system is connected to a PeakSimple Chromatography Data System (Model 202) and the data analyzed by PeakSimple software. [0069] LMWP was purified by diluting 0.2 mL of the digested protamine solution with deionized water to 1 mL and loading this solution into the sample loop. Then, the salt gradient and data collection were initiated by switching the sample injector from the "LOAD" to "INJECT" position. Fig. 3 shows the elution profile obtained by this set-up. The peak labeled "TDSP5" is the desired fragment and is obtained at yields of less than 0.4 mg per cycle. The elution profile was obtained at an absorbance wavelength of 215 nm. Solution flow rate was 1.5 mL per minute with a 5 mL heparin column. The sample loading was 1 mL of a 2.2 mg digest/mL. Buffer A was 50 mM TRIS/HCl (pH 7.5) and Buffer B was 50 mM TRIS/HCl/2 M NaCl (pH 7.5).

[0070] Since the peptide of interest is TDSP5, the efficiency of the purification will be determined by the obtained peak resolution between TDSP5 and the closely eluting TDSP4. Purification of LMWP with a HiTrap heparin column and the salt gradient shown in Fig. 3 gives a peak resolution of 0.81 which yields a TDSP5 product of only 95.8% purity at a recovery of 96.7 percent, as determined through integration of the area between peak profiles and the base line. This purity of less than 96 percent is very low considering that the product is intended for pharmaceutical use. Therapeutic drugs are required in many cases to be 99.997% pure [51]. In order to scale-up the production of LMWP, the peak resolution must first be improved to provide higher purity and yield.

[0071] Using the method developed for the immobilization of invertase, thermolysin was immobilized on nanoporous silica. Enzymatic activity was characterized using an azocasein substrate under varying pH and temperature. A small packed-bed reactor was created by packing the immobilized thermolysin into a syringe filter. The immobilized enzyme showed good retention of activity even after repeated exposure to 2M NaCl solutions. This combined with the optimization of the salt gradient for LMWP purification provides approximately a 30-fold improvement in productivity. High product purity is obtained as seen from the heparin affinity elution profile in Fig. 4.

[0072] Ideally, for a packed-bed reactor or chromatographic support, the matrix should have a regular shape and be of uniform size. Spherical particles are preferential because of improved packing; flow characteristics showing a minimum back-pressure; maximum surface to volume ratio providing a smaller diffusion path. Several methods have been described in literature for the fabrication of silica spheres. A well studied method for producing silica spheres, with diameters in the nanoscale, is the Stober process. In the Stober process, a dilute silica alkoxide solution is mixed with ammonia, which serves as a catalyst to induce condensation [52-54]. Particle formation can be tightly controlled by variation of the reactant concentrations [55], catalyst concentration [56], and reaction temperature [57]. Modified Stδber processes have also been developed for increased control of particle size [58, 59].

[0073] In addition to the Stδber process, silica spheres have been produced by an emulsion system. Many of these processes produce silica beads which is surface modified in subsequent steps. Modification of silica using N-2-(aminoethyl)-3- aminopropyl (trimethoxy)silane has been shown to increase the chemical reactivity of silica for adsorption [60]. Spherical particles, with a size of 50 nm, modified with 3- aminopropyl(triethoxy)silane have been produced using TEOS in a microemulsion process [61]. Silica spheres with an amino loading of 2.3 mmol/g have been reported [62]. These particles however are in the nanoscale and not appropriate for bioreactors where a particle size of at least 50 microns is often desired. In addition, all of these published results use TEOS as the silica source.

[0074] It is well known in the prior art that amino functional groups are able to selectively adsorb metal ions, including copper, nickel, and cobalt [63, 64]. Although the outer morphology appears to be closed by SEM imaging, the spherical particles produced according to the present invention are able to adsorb cobalt, copper, and nickel ions. The clear particles display a drastic change in color associated with the adsorption of these metal ions. If used in a water treatment system, the color change could be used as an indicator of when the material must be changed.

EXAMPLES

[0075] Materials. Invertase (Grade VII) from baker's yeast, glutaraldehyde (25 wt. % in water), (3-aminopropyl)triethoxysilane (min. 98%), and sucrose were purchased from Sigma-Aldrich and used as supplied. Sodium silicate ("N" type) was purchased from PQ Corporation, PA. All other reagents were of analytical grade.

Example 1

Synthesis of amino-chemically surface modified gel (N-CSMG).

[0076] Silicic acid was produced from sodium silicate using an ion-exchange process as previously described. The ion-exchange, accomplished with anionic

Amberlite™ IR120 resin, exchanges sodium ions for hydrogen to produce a low ionic strength, silica sol with reactive silanol groups ( - Si-OH). Formation of N-CSMG was done by addition of ethanol and 3-aminopropyltriethoxy-silane (APTES) to the silicic acid. Addition of the basic APTES to silicic acid increases the solution pH and induces the gelation. Gelation occurs through the condensation of silanol groups. Co- condensation between silica sol and APTES produces silica gels with amino surface modification. The monolithic gel was mechanically broken into particles of llμm average diameter (range of 7 - 38 μm measured on a CEDEX) and washed with ethanol. The ethanol washed suspension was vacuum filtered and washed extensively with deionized water. The final filtered gel was stored at 7°C until used for a maximum of 2 weeks.

Example 2

Glutaraldehyde (GA) activation of N-CSMG to produce glutaraldehyde-activated silica gel (Ga-N-CSMG).

[0077] To activate N-CSMG with glutaraldehyde, the gel was suspended in glutaraldehyde solution at a ratio of 20 mL solution per gram N-CSMG. The suspension was magnetically stirred at room temperature for 24 hours. Then, the suspension was vacuum filtered, to remove the GA solution, and washed extensively with deionized water. The final filtered gel was stored at 7°C until use.

Example 3

Immobilization of invertase on Ga-N-CSMG.

[0078] To immobilize invertase, 500 mg of Ga-N-CSMG was weighed into a 50 mL centrifuge vial. Then, 20 mL of invertase, in 50 mM acetate buffer at pH 4.5, was introduced to the vial and the suspension magnetically stirred at 7°C for a minimum of 48 hours. The suspension was then centrifuged at 10,000g for 10 minutes and the supernatant removed. The invertase-immobilized gel was washed three times with 20 mL of 100 mM acetate buffer (pH 4.5) and centrifuged between each wash. Three washes were found to be adequate for removal of non-immobilized enzymes through activity measurement of the supernatants. The washed gel was suspended in 20 mL of 100 mM acetate (pH 4.5) and stored at 7°C until use.

Determination of Enzyme Activity.

[0079] Invertase activity was determined by monitoring the hydrolysis of 30 mL

50 g/L sucrose in 50 mM buffer in a magnetically stirred, thermostated vessel. The activity was measured by the dinitrosalicylic acid (DNS) method [33]. At various times, 500 μL of reaction solution were withdrawn and introduced to 5 mL of 10 g/L DNS solutions which stopped the reaction. The sample vials were capped and heated in boiling water for 30 minutes prior to addition of 1 mL sodium tartrate (40 wt. %) to stabilize the color. Absorbance was measured at 575 nm on a Gilford Response spectrophotometer and the glucose concentration determined by comparison with a standard curve. Glucose concentration is plotted against time and invertase activity determined from the reaction profile as the initial activity. Enzyme activity is expressed in units per gram (U/g) where a unit is defined as the production of 1 μmole glucose product per minute (μmole/minute). All activity tests were conducted in triplicate.

Optimization of glutaraldehyde concentration.

[0080] The optimum glutaraldehyde concentration was determined by adding 20 mL of glutaraldehyde and one gram N-CSMG (wet weight) to a 50 mL centrifuge vial. After allowing the reaction to take place for 24 hours at room temperature, the gel was thoroughly washed with deionized water. To 500 mg of each sample, 20 mL of 2 g/L invertase in 5OmM acetate (pH 4.5) is added and immobilization is allowed to take place at 7 ⁰C for 48 hours. The gel is then washed several times, suspended in 100 mM acetate buffer (pH 4.5), and measured for invertase activity at room temperature. Glutaraldehyde concentration was varied from 0 to 10 weight percent.

Optimization of invertase concentration.

[0081] The optimum concentration for immobilization of invertase was determined by adding 20 mL of buffered, invertase solution to 500 mg of the optimized GA-N-CSMG. Invertase concentration was varied from 0 to 20 g/L.

[0082] As determined by thermogravimetric analysis, the N-CSMG material contains a functional group loading of approximately 3.6 mmol amino per gram of silica. This loading is significantly higher than the 1.0 mmol/g loading of commercially available products (e.g. Sigma-Aldrich product #364258) and allows for much higher loadings of protein. In order to covalently immobilize enzymes on the N-CSMG, the amino-modified silica is first activated with glutaraldehyde and then, after extensive washing, added to a buffered enzyme solution. Each of these immobilization steps were optimized by using invertase, a well studied enzyme that hydrolyzes the disaccharide sucrose into glucose and fructose. [0083] One advantage that nanoporous silica materials provide, compared to other immobilization matrices, is a high surface area. The surface area of the optimized Ga-N- CSMG was determined using BET analysis based on nitrogen adsorption on a

Quantachrome Nova (Model 1200). Silica based materials experience a significant amount of shrinkage upon drying which would also affect the surface area. One method is to use supercritical drying to avoid the liquid/vapor interface which causes the shrinkage. Supercritical CO₂ was used to dry the optimized Ga-N-CSMG material prior to surface area analysis. This treatment provides a truer surface area measurement of the material in its wet state. After outgassing, under vacuum for 2 hours at 100⁰C, the surface area of the optimized Ga-N-CSMG was measured and found to be approximately 650 m²/g.

[0084] A 20 mL volume of the 0.31% (w/w) solution contains approximately

0.625 mmols glutaraldehyde while a gram of wet N-CSMG, with a dry weight of 90 mg, contains approximately 0.35 mmols of amino groups. Even with this low glutaraldehyde concentration, there is approximately a 1.8 molar ratio of glutaraldehyde to surface amino groups. Increasing the glutaraldehyde concentration further had no effect on the immobilized activity because the system was already over saturated. A 2.5 wt.% glutaraldehyde solution was used in further studies for activation of the silica gel.

[0085] N-CSMG untreated with glutaraldehyde yields a relative immobilized activity of approximately 32 percent. This immobilization may occur through the formation of a salt-bridge between the basic, amino-modified surface and acidic side- chains of aspartic or glutamic acid; a well established interaction seen in proteins [65-67].

Example 4

Optimization of Invertase concentration.

[0086] The optimum invertase solution concentration for immobilization was determined by varying it between 0 to 20 g/L. For each concentration, 20 mL of invertase solution (in 50 mM acetate, pH 4.5) was added to 500 mg (wet weight) of glutaraldehyde- activated N-CSMG gel (GA-N-CSMG).

[0087] As shown in Fig. 5, the optimum immobilized activity was achieved with a

10 g/L invertase solution concentration and no further increase in immobilized activity seen with higher enzyme concentrations. However the immobilization efficiency, obtained by a ratio of activity immobilized to the activity in the original enzyme solution, was only 23% when using 20 mL of a 10 g/L invertase solution with 500 mg of wet glutaraldehyde-activated N-CSMG. The immobilization was further optimized by varying the volume of the 10 g/L invertase solution between 1 and 20 mL and the results are shown in Fig. 6. A solution volume of 10 mL yielded the maximum immobilization efficiency of 11% and provided 90% of the maximum immobilized activity, which was obtained with 20 mL of 10 g/L invertase solution. All further immobilization of invertase onto GA-N-CSMG was done using 500 mg of wet, glutaraldehyde-activated gel with 10 mL of a 10 g/L enzyme solution. As seen from Fig. 6, optimization of immobilization solution volume yielded a maximum immobilization efficiency (®) of 71% and relative immobilized activity (■) of 90% at an enzyme solution volume of 10 mL.

[0088] The effect of temperature on both free and immobilized invertase activity was determined by measuring the hydrolysis of sucrose at temperatures ranging from 25- 75°C. A 50 g/L sucrose solution, buffered with 50 mM acetate at pH 4.5, was used for activity determination. Both the immobilized and free enzymes show a very similar temperature profile as seen in Fig. 7. This indicates that immobilization of invertase on GA-N-CSMG does not significantly alter the enzyme's temperature stability.

[0089] The same data can be plotted on an Arrhenius plot which provides a relation between the log of enzyme activity and inverse temperature. At the lower temperatures (T), higher inverse temperatures (T"¹), the plot is linear as enzymatic activity increases with increasing temperature. As the temperature is increased further enzyme denaturization occurs and a corresponding drop in enzymatic activity is observed.

[0090] The immobilized enzyme shows a decreased energy of activation at 25.4 kJ/mol compared to 31.3 kJ/mol for free invertase. Literature studies have indicated that a decrease in activation energy can be correlated to intraparticle diffusion of the substrate [68]. The appearance of activation energy diffusion effects, for immobilization of invertase on GA-N-CSMG, is an indication that the immobilization is occurring within a porous matrix. In the case of a nonporous matrix, activation energies of the free and immobilized enzyme would be in closer agreement with each other [69].

[0091] The Michaelis-Menten kinetic parameters (V_max and K_m )of both free and immobilized invertase were determined by measuring the initial reaction rates at various initial substrate concentrations. For this set of experiments, the initial sucrose concentration was varied between 8 wt % and 50 wt% with the pH fixed at 4.5 and temperature at 45 ⁰C. The immobilized invertase exhibited a higher affinity for sucrose than free invertase while the maximum reaction rate, V_max, is decreased compared to that of the free enzyme. A catalytic efficiency, which is the ratio of V_max to K_m, was calculated to be similar for both the immobilized and free invertase. The slight increase in substrate affinity is able to compensate for a decreased V_max for immobilized invertase, providing a catalytic efficiency equal to that of the free enzyme. It has been shown that as the effectiveness factor, a ratio of immobilized to free enzyme activity, decreases below unity, the activation energy measured approached the arithmetic mean of the activation energies of diffusion and reaction [70]. GA-N-CSMG gives an efficiency factor of 0.73 for the immobilization of invertase. There are several reasons that can explain the difference in behavior of the free and immobilized invertase. First, the immobilized invertase is located in an environment that is quite different from the environment of free enzyme in the bulk solution. The dependence of immobilized invertase on diffusion of substrate into the matrix interior is also quite different from free invertase, which interacts freely with the bulk solution. In addition, the attachment of invertase to GA-N-CSMG likely causes some change in conformation. A conformational change may explain why the K_1n of immobilized invertase is only 67 percent that of the free. In this instance, immobilization of invertase actually increases its substrate affinity. Most cases show a decrease in substrate affinity due to conformation changes [69].

[0092] Infrared spectra were used to verify the immobilization procedure. As expected, C-H stretching vibration frequency is seen at 2936 cm^"1 for all spectra except the silica gel with contributions from the organosilane,- glutaraldehyde, and the enzyme [71, 72]. The corresponding simple bending vibrations occur at 1407 cm^"1 for the GA-N- CSMG and immobilized invertase spectra. Free aldehyde presence is seen on GA-N- CSMG at 1718 cm^"1 and does not appear in any other spectra. The amine-glutaraldehyde reaction produces an imine N=C bond, Schiff-base, seen at 1647 cm^"1 while an ethylenic C=C bond formed by resonance stabilization of the imine appears at 1563 cm^"1 [73]. Associated with Si-O-Si bonds are the bands appearing in the range 1050-1070 cm^"1 [74].

[0093] TGA confirmed that the N-CSMG is able to achieve a very high loading of invertase with a maximum loading of 1.39 grams invertase per gram of SiO₂. This invertase loading corresponds to a loading of 723 mg invertase per gram of GA-N- CSMG.

[0094] The invertase loading obtained with the GA-N-CSMG is significantly greater than other published results. A comparison of the GA-N-CSMG loading with various other materials is shown in Table 3. With a loading of 723 mg/g matrix, GA-N- CSMG is able to immobilize a three-fold greater weight of invertase than the material with the next highest loading - lactam-amide graft polymer with 225 mg/g. The lactam- amide graft polymer exhibited only a 15 percent retention of enzyme activity. GA-N- CSMG, with an immobilized invertase activity of 340.9 ± 8.8 U/mg protein and 345.4 ± 11.6 U/mg protein for free invertase, provides almost 99 percent recovery of enzymatic activity.

Table 3. Comparison of invertase immobilization loading and the retained activity of GA-N-CSMG with other materials.

[0095] Based on the molecular footprint, calculated from the Stokes radius, and the measured surface area, a theoretical maximum loading of 2.93 grams invertase per gram GA-N-CSMG is achievable. The obtained loading of 0.723 g/g GA-N-CSMG is less than 25 percent of the theoretical maximum. Likely reasons for the lower loading efficiency are the inaccessibility to surface areas by the enzyme due to the presence of small, micropores and the blockage of pore entrances.

Example 5

Thermolysin immobilization

[0096] Thermolysin immobilization was accomplished by adding a 2 g/L thermolysin solution, buffered with TRIS/HCl at pH 7, to glutaraldehyde-activated silica gel (example 2). After extensive washings, the pH and temperature dependent profiles of free and immobilized thermolysin were determined.

[0097] Thermolysin activity was measured color-metrically using an azocasein

(Sigma- Aldrich, Catalogue Number A-2765) substrate. This substrate consists of casein modified with a dye, which upon hydrolysis of the protein becomes soluble. Free and immobilized thermolysin were incubated with 4 mL of 5 wt. % azocasein. To determine the extent of reaction, 500 μL of 15% trichloroacetic acid (TCA) was added to a 1 mL aliquot of the reaction solution. Addition of TCA induces the precipitation of undigested azocasein; the substrate becomes insoluble at low pH. The precipitate is removed from solution by centrifugation at 12,00Og for 5 minutes. Enzymatic activity is then proportional to the dye remaining in solution; measured at a wavelength of 440 nm. Both the free and immobilized thermolysin show a maximum activity at pH 7. There is a rapid drop-off in activity as the pH is changed away from neutral. The immobilized

thermolysin follows a similar activity drop-off, as the free enzyme, for pH greater than 7. However at lower pH, the drop in immobilized thermolysin activity is significantly lower than that of the free enzyme.

[0098] The effect of temperature on both free and immobilized thermolysin activity was determined by measuring the hydrolysis of azocasein at temperatures ranging from 25-85 ⁰C. A 0.5 wt. % azocasein solution, buffered with 50 mM TRIS/HCl at pH 7, was used for activity determination. The substrate solution, 4 mL, was incubated in a temperature-controlled, circulating water bath (Precision, Cat. No. 51221039) for 10 minutes before addition of thermolysin. The extent of reaction was followed over time and the enzyme activity determined. Free thermolysin activity appears to increase only slightly from 25 to 35 ⁰C. As the temperature in increased further, the activity increases linearly, at a higher rate, up to 85 ⁰C. The immobilized thermolysin, on the other hand, shows only a slight increase in activity from 25 to 55°C. Above this temperature, immobilized thermolysin activity increases at a greater rate and linearly up to 85 °C. Azocasein was used to show the retention of thermolysin activity after immobilization on nanoporous silica. The immobilized thermolysin showed a similar pH activity profile to that of the free enzyme. Although some difference was seen in the temperature activity profiles, the trends were similar.

Example 6

Digestion of protamine by immobilized thermolysin in batch reactor

[0099] The immobilized thermolysin activity was also measured for the hydrolysis of protamine. Immobilized thermolysin was introduced to a 10 mg/mL protamine solution buffered at pH 7.4 with 50 mM Tris/HCl/2 mM CaCl₂. The reaction was incubated in a 28.5⁰C water bath for specified times; then placed in ice and EDTA added to stop the reaction. After filtration with a 0.22 μm syringe filter, a sample of the protamine solution was injected into an HPLC gradient system for separation of the individual peptides. The HPLC system was fitted with a 5mL Hi-Trap Heparin affinity column.

Example 7

Digestion of protamine by immobilized thermolysin in a plugged flow reactor using a FPLC operating at 4⁰C

[0100] The use of a plugged flow reactor (PBR) for protamine hydrolysis, by thermolysin, has several advantages. A flow reactor achieves a continuous operation requiring minimum labor. The flow reaction could be integrated with the separation by an affinity column to further streamline the separation and purification tasks. In addition, full automation with software controlled processes may be achieved. The optimum design of such a system obviously requires the integration of several other devices with the enzyme reactor and may involve extensive system engineering. The performance of an immobilized thermolysin system in a PBR mode was studied.

[0101] Because of the high enzyme loadings achieved, a relatively small reactor is sufficient to achieve the required conversion. A 0.22 μm syringe filter (Fisherbrand, Cat. No. 09-719A) was used to filter-off the immobilized thermolysin from a buffered suspension; loading approximately 45 mg of silica matrix containing 33 mg of

immobilized thermolysin. The filter and immobilized thermolysin make-up a packed-bed reactor with a small height to diameter ratio, shown in Fig. 9.

[0102] The LMWP Bioreactor was fitted onto a Fast Performance Liquid

Chromatography (FPLC) system for the production and purification of LMWP. Fig. 10 shows the flow profile of the FPLC system when the position valve is switched from the LOAD to INJECT position.

[0103] With the position valve set at INJECT, the buffer flow from the pump is directed into the top of the Superloop column. This forces a moveable, water-tight piston downwards which forces protamine solution out of the column bottom. The protamine solution flows into the bioreactor, where it is hydrolyzed by the immobilized thermolysin, and then injected into a HiTrap Heparin column. The interaction between the protamine fragments and heparin results in the retention of the peptides within the column.

Switching the position valve to the LOAD position then by-passes the Superloop and directs the buffering solution directly into the chromatography column. A salt gradient is used to elute the peptides with the solution flowing into a UV detector for detection of the elution profiles. [0104] With this FPLC set-up, shown in Fig. 10, the immobilized thermolysin reactor comes in contact with only the buffered protamine solution. The elution buffer, with its high salt content, completely bypasses the reactor. However, as will be seen in the next section, the enzymatic activity is retained even after exposure of the bioreactor to a high salt solution.

[0105] The extent of reaction can be controlled through adjustment of the protamine solution flow rate. Several protamine injection flow rates were tested and the elution profiles determined. Even at the low flow rate of 20 μL/min, a significant amount of native protamine still appears in the elution profile as shown in Fig. 11. This is not surprising because of the reduced thermolysin activity at low temperatures. Since the FPLC is located in a cold room, the hydrolysis reaction occurs at 4⁰C.

[0106] The conversion can be increased by further reduction of the flow rate, increasing the catalyst weight, and increasing the reaction rate. Further reduction of the flow rate is not feasible due to the resulting increase in the cycle time. The catalyst weight can be increased with a larger reactor and the reaction rate can be increased with an increase in temperature.

Example 8

Production of LMWP using HPLC at 25°C and 55°C

[0107] A High Performance Liquid Chromatography (HPLC) system was used to test the immobilized thermolysin reactor's performance at room temperature. The system consisted of two Alltech 526 HPLC pumps, a Linear UVIS 200 detector, and a two- position injection valve. Data acquisition was done by a PeakSimple Chromatography Data System, Model 202. A ImL sample loop was connected to the sample injection port.

[0108] Fig. 12 shows the direction of buffer flow for the INJECT and LOAD valve positions. In the INJECT position, the buffer is directed through the sample loop which then carries the protamine solution through the thermolysin reactor. The reactor exit stream then flows into the chromatography column. When the valve is switched to the LOAD position, although the sample loop is bypassed, the reactor is still a part of the flow stream. With this configuration there is the possibility of enzymatic activity loss due to the high salt gradient.

[0109] The HPLC system was used to follow the hydrolysis of protamine by immobilized thermolysin at 25°C. First the sample loop was loaded with 1 mL of 2 mg/mL protamine solution. TRIS/HCl buffer, pH 7.5, was pumped through the loop at different flow rates with the position valve in the INJECT position. This forces the protamine solution through the immobilized thermolysin reactor and into the heparin column. The resulting elution profiles are shown in Fig. 13.

[0110] The elution profiles clearly show the hydrolysis of protamine into smaller peptides by the immobilized thermolysin reactor. Two different flow rates, 0.25 mL/min and 0.15 mL/min, were used for the hydrolysis reaction. As the flow rate is decreased, the conversion of protamine to LMWP increases. Because the immobilized enzyme is only a thin layer in the reactor, the reaction time is relatively short with these flow rates. This serves as further evidence of the relatively high immobilized activity achieved in this work.

[0111] With the expectation that a higher temperature would yield greater enzymatic activity, the hydrolysis of protamine was also determined with varying reaction temperatures. Protamine solution was incubated at 25°C or 55°C before injecting into the HPLC system. The resulting elution profiles in Fig. 14 show the increased digestion of protamine at the higher temperature.

[0112] The immobilized thermolysin reactor showed no apparent loss in activity after 20 cycles. This includes use of the reactor in the HPLC system where the salt gradient, which flows through the reactor, contacts the immobilized thermolysin.

Retention of activity under a 2M NaCl solution shows that the immobilization method is stable and no leaching occurs.

Example 9

Formation of amino-modified silica spheres by an emulsification process

[0113] Spherical N-CSMG beads were produced using sodium silicate, as the initial source of silica. As detailed in Example 1, an ion-exchange process was used to convert sodium silicate into silicic acid. The silicic acid serves as the water phase of the water-in-oil emulsion and 2-ethyl-l-hexanol is the oil phase. Silicic acid is added to the oil phase, with stirring, at a 1:15 volume ratio. Ethanol is added to the system, at a ethanol: silicic acid ratio of 2:1, as a cosolvent to improve the silicic acid/ APTES compatibility. Under stirring, the silicic acid and ethanol solution form droplets within the oil phase. After five minutes of stirring to achieve a stable droplet size, APTES is added to the emulsion; it enters the oil phase and then diffuses into the droplets. Two events occur as 3-aminopropyl(triethoxy)silane enters the aqueous phase: 1) the ethoxy groups are hydrolyzed to form ethanol, leaving hydroxyl groups on the silane; and 2) the basic amino group causes an increase in droplet pH and induces gelation. This scheme is illustrated in Fig. 15.

Control of spherical particle size

[0114] The silica sphere particle size can be controlled by variation of the stirring speed. In Fig. 16 is shown the micrographs obtained by the emulsion process at two different stirring speeds. Size distribution of the spherical particles was determined from the micrographs obtained for three different stirring rates. The particle size distribution is shown in Fig. 17, with the average particle size and standard deviation, for each stirring rate along with a Gaussian distribution curve. From the curves and average particle size, a clear decrease is seen in particle size with increasing stirring rates. However, a narrower size distribution would be more ideal. In addition to the stirring rates, several other factors can be adjusted to obtain the desired size distribution. Composition of the oil and water phases, surfactants, and even impeller shape can significantly affect the distribution [79-81].

Example 10

Adsorption of metal ions and small molecules

[0115] The Cu(II) adsorption capacity of the amino-modified silica spheres was determined colormetrically. A 25 mg sample of the silica particles was introduced to 10 mL of 1000 ppm Cu(II) solution initially at pH 4.5. The adsorption of Cu(II) was followed through measurement of Cu(II) remaining in solution. A 300 μL aliquot of solution was removed and added to 60 μL of ammonia hydroxide solution. Chelation of copper ions by ammonia gives a blue color and the concentration is determined by absorbance measurement at a wavelength of 440 nm. From the time profile of Cu(II) adsorption by N-CSMG and spherical N-CSMG, shown in Fig. 18, it is clear that the spherical particles have an adsorption capacity less than half that of N-CSMG at 1.5 and 3.1 mmols/g, respectively.

[0116] Using a water-in-oil emulsion process, spherical silica beads are prepared with an amino-modified surface. The amino loading of spherical beads is 1.5 mmols/g which, although significantly lower than the N-CSMG loading, is a higher loading than available in commercial products. Surface modification of the spherical beads relies on the diffusion of APTES into a silicic acid droplet. Because the APTES induces gelation of the droplets outer shell, a significant barrier to APTES diffusion is created. This leaves a core region with a relatively low loading of APTES. The loading can be significantly increased if APTES is homogenously mixed with silicic acid prior to gelation of the beads. Although they can be premixed and then introduced to the oil phase, the gelation occurs rapidly and this inhibits the control of particle size and shape. Control of particle size may be achieved with adjustment of stirring rates. To obtain a narrower size distribution, engineering of the mixer may provide a more uniform mixing profile. In another embodiment, the addition of surfactant(s) can be used to tightly control the emulsion droplet size. Surfactants can also be used as a templating agent to improve the outer shell morphology. The N-CSMG material is made with APTES homogenously mixed with the silicic acid; yielding a homogenous distribution of amino groups on the gel surface. However, the emulsion process of making amino-modified spherical particles depends on the diffusion of APTES into the silicic acid droplets. Gelation thus occurs first at the water-oil interface, forming a solid shell, and progresses toward the center. The shell formation creates a greater barrier to APTES diffusion and results in a lower concentration of amino groups towards the particle center. Although the outer shell has a more closed morphology than the particle interior, it is permeable to glutaraldehyde which is a relatively small molecule.

[0117] The N-CSMG material has a hydrophilic nature and shows no adsorption of the pH dye Thymol Blue. However when treated with the surfactant sodium dodecyl sulfate (SDS) it becomes hydrophobic. The anionic head of SDS interacts with the cationic amino groups in N-CSMG creating a surface with the hydrophobic tails pointing outward. After treatment with SDS, N-CSMG shows adsorption of Thymol Blue and exhibits the pH indicating characteristics of the dye as seen in Fig. 19.

REFERENCES

1. Cheetham, P.S.J., The applications of enzymes in industry, in Handbook of enzyme biotechnology. 1995, Ellis: London, p. 420-40.

2. Conder, J.R. and B.O. Hayek, Adsorption kinetics and equilibria of bovine serum albumin on rigid ion-exchange and hydrophobic interaction chromatography matrices in a stirred cell. Biochemical Engineering Journal, 2000. 6(3): p. 215- 223.

3. Beck, F. and H. Suginome, Industrial Electroorganic Synthesis in Europe. Journal of Synthetic Organic Chemistry Japan, 1991. 49(9): p. 798-808. 4. Chitnis, S. R. and M.M. Sharma, Industrial applications of acid-treated clays as catalysts. Reactive & Functional Polymers, 1997. 32(1): p. 93-115.

5. Laszlo, P., Heterogeneous catalysis of organic reactions. Journal of Physical Organic Chemistry, 1998. 11(5): p. 356-361.

6. Tischer, W. and F. Wedekind, Immobilized enzymes: Methods and applications.

Biocatalysis - from Discovery to Application, 1999. 200: p. 95-126.

7. Abourebyeh, H., et al., Carrier Membrane as a Stationary Phase for Affinity- Chromatography and Kinetic-Studies of Membrane-Bound Enzymes. Journal of Chromatography-Biomedical Applications, 1991. 566(2): p. 341-350.

8. Camperi, S.A., et al., Preparation and characterisation of immobilised metal ion hollow-fibre polysulphone membranes. Their application in high-speed pectic enzyme fractionation. Process Biochemistry, 2004. 39(8): p. 1017-1024.

9. Tharakan, J., et al., Physical and Biochemical-Characterization of 5 Commercial Resins for Immunoaffinity Purification of Factor-Ix. Journal of Chromatography, 1992. 595(1-2): p. 103-111.

10. Li, Z. Y., Z.L. Wu, and J. Li, New methods in immobilization: Biocapsulation by epoxy-amine resins and silicone elastomers. Chinese Journal of Organic

Chemistry, 2003. 23(2): p. 150-154.

11. Amanda, A. and S. K. Mallapragada, Comparison of protein fouling on heat- treated polyvinyl alcohol), poly( ether sulfone) and regenerated cellulose membranes using diffuse reflectance infrared Fourier transform spectroscopy. Biotechnology Progress, 2001. 17(5): p. 917-923.

12. Dasari, G., I. Prince, and M.T.W. Hearn, High-Performance Liquid- Chromatography ofAmino-Acids, Peptides and Proteins .124. Physical

Characterization of Fluidized-Bed Behavior of Chromatographic Packing

Materials. Journal of Chromatography, 1993. 631(1-2): p. 115-124.

13. Brinker, CJ. and G. W. Scherer, Sol-gel science : the physics and chemistry of sol- gel processing. 1990, Boston: Academic Press, xiv, 908.

14. Hench, L.L. and J.K. West, The Sol-Gel Process. Chemical Reviews, 1990. 90(1): p. 33-72.

15. Walcarius, A. and C. Delacote, Rate of access to the binding sites in organically modified silicates. 3. Effect of structure and density of functional groups in mesoporous solids obtained by the co -condensation route. Chemistry of Materials, 2003. 15(22): p. 4181-4192. 16. Burkett, S.L., S.D. Sims, and S. Mann, Synthesis of hybrid inorganic-organic mesoporous silica by co-condensation of siloxane and or ganosiloxane precursors. Chemical Communications, 1996(11): p. 1367-1368.

17. Mori, Y. and TJ. Pinnavaia, Optimizing organic functionality in mesostructured silica: direct assembly of mercaptopropyl groups in wormhole framework structures. Chemistry of Materials, 2001. 13(6): p. 2173-2178.

18. Cserhati, T. and Z. Illes, Salting-out and Salting-in Effects in the Reversed-Phase Thin-Layer Chromatography of Dansylated Amino-Acids - Effect of Acids. Journal of Liquid Chromatography, 1991. 14(8): p. 1495-1510.

19. Tasleem, S., et al., Salting-out coefficients and activity coefficients of alkali and alkaline earth metals in aqueous and aqueous organic mixed solvents. Journal of the Chemical Society of Pakistan, 2004. 26(1): p. 39-43.

20. Beck, J.S., et al., A New Family of Mesoporous Molecular-Sieves Prepared with Liquid-Crystal Templates. Journal of the American Chemical Society, 1992.

114(27): p. 10834-10843.

21. Lim, M.H. and A. Stein, Comparative studies of grafting and direct syntheses of inorganic-organic hybrid mesoporous materials. Chemistry of Materials, 1999. 11(11): p. 3285-3295.

22. Kruk, M., et al., Metamorphosis of ordered mesopores to micropores: Periodic silica with unprecedented loading of pendant reactive organic groups transforms to periodic microporous silica with tailorable pore size. Journal of the American Chemical Society, 2002. 124(22): p. 6383-6392.

23. Yu, N.Y., et al., One-pot synthesis of mesoporous organosϊlicas using sodium

silicate as a substitute for tetraalkoxysilane. Microporous and Mesoporous Materials, 2004. 72(1-3): p. 25-32.

24. Bois, L., et al., Functionalized silica for heavy metal ions adsorption. Colloids and Surfaces a-Physicochemical and Engineering Aspects, 2003. 221(1-3): p. 221-230.

25. Yokoi, T., H. Yoshitake, and T. Tatsumi, Synthesis of amino-functionalized

MCM-41 via direct co-condensation and post-synthesis grafting methods using mono-, di- and tri-amino-organoalkoxysilanes. Journal of Materials Chemistry, 2004. 14(6): p. 951-957.

26. Lee, B., et al., Synthesis of functionalized porous silicas via templating method as heavy metal ion adsorbents: the introduction of surface hydrophilicity onto the surface of adsorbents. Microporous and Mesoporous Materials, 2001. 50(1): p. 77-90.

27. Shah, J., S. S. Kim, and TJ. Pinnavaia, A versatile pathway for the direct assembly of organo-functional mesostructures from sodium silicate. Chemical

Communications, 2004(5): p. 572-573.

28. Barrett, E.P., L.G. Joyner, and P.P. Halenda, The Determination of Pore Volume and Area Distributions in Porous Substances .1. Computations from Nitrogen Isotherms. Journal of the American Chemical Society, 1951. 73(1): p. 373-380.

29. Vansant, E.F., P. van der Voort, and K.C. Vrancken, Characterization and

chemical modification of the silica surface. 1995, Amsterdam; New York:

Elsevier.

30. Overberger, C. G., S. Ishida, and H. Ringsdorf, Intra-Intermolecular

Polymerization of Glutaraldehyde. Journal of Polymer Science, 1962. 62(173): p. S1-&.

31. Rasmusse.Ke and Albrecht.J, Glutaraldehyde - Influence of Ph, Temperature, and Buffering on Polymerization Rate. Histochemistry, 1974. 38(1): p. 19-26.

32. Aso, C. and Y. Aito, Studies on the Polymerization of Bifunctional Monomers .2.

Polymerization of Glutaraldehyde. Makromolekulare Chemie, 1962. 58: p. 195- 203.

33. Miller, G.L., Use of Dinitrosάlicylic Acid Reagent for Determination of Reducing Sugar. Analytical Chemistry, 1959. 31(3): p. 426-428.

34. Gill, I. and A. Ballesteros, Bioencapsulation within synthetic polymers (Part 1): sol-gel encapsulated biologicals. Trends in Biotechnology, 2000. 18(7): p. 282- 296.

35. Weetall, H.H., Preparation of Immobilized Proteins Covalently Coupled through Silane Coupling Agents to Inorganic Supports. Applied Biochemistry and Biotechnology, 1993. 41(3): p. 157-188.

36. Hossain, M.M. and D.D. Do, Fundamental-Studies of Glucose-Oxidase

Immobilization on Controlled Pore Glass. Biotechnology and Bioengineering, 1985. 27(6): p. 842-851.

37. Moreno, J.M. and J.V. Sinisterra, Immobilization of Lipase from Candida- Cylindracea on Inorganic Supports. Journal of Molecular Catalysis, 1994. 93(3): p. 357-369. 38. Pizarro, C, et al., Optimization by experimental design of polyacrylamide gel composition as support for enzyme immobilization by entrapment. Biotechnology and Bioengineering, 1997. 53(5): p. 497-506.

39. Baruque, E.A., M.G.A. Baruque, and G.L. Sant'Anna, Determination of the

enzyme reaction rate in a differential fixed-bed reactor: A case study. Brazilian Journal of Chemical Engineering, 2001. 18(1): p. 1-11.

40. Horrow, J.C., Adverse Reactions to Protamine. International Anesthesiology

Clinics, 1985. 23(3): p. 133-144.

41. Horrow, J.C., Protamine - a Review of Its Toxicity. Anesthesia and Analgesia, 1985. 64(3): p. 348-361.

42. Weiss, M.E., et al., Association of Protamine Ige and Igg Antibodies with Life- Threatening Reactions to Intravenous Protamine. New England Journal of Medicine, 1989. 320(14): p. 886-892.

43. Chang, L.C., et al., Low molecular weight protamine (LMWP) as nontoxic

heparin/low molecular weight heparin antidote (I): Preparation and

characterization. Aaps Pharmsci, 2001. 3(3): p. art. no.-17.

44. Chang, L.C., et al., Low molecular weight protamine (LMWP) as nontoxic

heparin/low molecular weight heparin antidote (II): In vitro evaluation of efficacy and toxicity. Aaps Pharmsci, 2001. 3(3): p. art. no.-18.

45. Lee, L.M., et al., Low molecular weight protamine as nontoxic heparin/low

molecular weight heparin antidote (III): Preliminary in vivo evaluation of efficacy and toxicity using a canine model. Aaps Pharmsci, 2001. 3(3): p. art. no. -19.

46. Liang, J.F., et al., A less toxic heparin antagonist - Low molecular weight

protamine. Biochemistry-Moscow, 2003. 68(1): p. 116-120.

47. Henry, CM., Breaching Barriers. Chemical & Engineering News, 2003. 81(34): p. 35-43, 55.

48. Futaki, S., et al., Arginine-rich peptides - An abundant source of membrane- permeable peptides having potential as carriers for intracellular protein delivery. Journal of Biological Chemistry, 2001. 276(8): p. 5836-5840.

49. Calnan, BJ. , et al., Arginine-Mediated Rna Recognition - the Arginine Fork.

Science, 1991. 252(5009): p. 1167-1171.

50. Wender, P.A., et al., The design, synthesis, and evaluation of molecules that

enable or enhance cellular uptake: Peptoid molecular transporters. Proceedings of the National Academy of Sciences of the United States of America, 2000.

97(24): p. 13003-13008.

51. Wheelwright, S. M., Protein Purification: Design and Scale up of Downstream Processing. 1991, New York: Hanser Publishers.

52. Fox, J.R., et al., Steric Stabilization ofStober Silica Dispersions Using

Organosilanes. Journal of Materials Science, 1987. 22(12): p. 4528-4531.

53. Jelinek, L., et al., Study of the Stober Reaction .1. Properties of Colloidal Silica Spheres Prepared ViaAlkoxide Hydrolysis. Langmuir, 1992. 8(9): p. 2152-2164.

54. Stober, W., A. Fink, and E. Bohn, Controlled Growth of Monodisperse Silica

Spheres in Micron Size Range. Journal of Colloid and Interface Science, 1968. 26(1): p. 62-&.

55. Giesche, H., Synthesis of Monodispersed Silica Powders .2. Controlled Growth Reaction and Continuous Production Process. Journal of the European Ceramic Society, 1994. 14(3): p. 205-214.

56. Liu, R.L., et al., Liquid-state Si-29 NMR study on the chemical kinetics ofStober synthesis. Acta Chimica Sinica, 2004. 62(1): p. 22-27.

57. Chen, S.L., et al., Kinetics of formation of monodisperse colloidal silica particles through the hydrolysis and condensation of tetraethylorthosilicate. Industrial & Engineering Chemistry Research, 1996. 35(12): p. 4487-4493.

58. Dong, P., Preparation of monodisperse silica spheres by growth of silica sol

particles. Acta Physico-Chimica Sinica, 1998. 14(2): p. 109-114.

59. Arkhireeva, A. and J.N. Hay, Synthesis ofsub-200 nm silsesquioxane particles using a modified Stober sol-gel route. Journal of Materials Chemistry, 2003.

13(12): p. 3122-3127.

60. Jesionowski, T., Synthesis of organic-inorganic hybrids via adsorption of dye on an amino silane-functionalised silica surface. Dyes and Pigments, 2002. 55(2-3): p. 133-141.

61. Tan, M.Q., et al., Preparation and time-resolved fluorometric application of

luminescent europium nanop articles. Chemistry of Materials, 2004. 16(12): p. 2494-2498.

62. Etienne, M., B. Lebeau, and A. Walcarius, Organically-modified mesoporous silica spheres with MCM-41 architecture. New Journal of Chemistry, 2002. 26(4): p. 384-386. 63. Yoshitake, H., T. Yokoi, and T. Tatsumi, Adsorption behavior of arsenate at transition metal cations captured by amino-functionalized mesoporous silicas. Chemistry of Materials, 2003. 15(8): p. 1713-1721.

64. Tikhomirova, T.I., et al., Sorption of Noble-Metal Ions on Silica with Chemically Bonded Nitrogen-Containing Ligands. Talanta, 1991. 38(3): p. 261 -21 A.

65. King, S.C., CL. Hansen, and T.H. Wilson, The Interaction between Aspartic Acid-237 and Lysine-358 in the Lactose Carrier of Escherichia-Coli. Biochimica Et Biophysica Acta, 1991. 1062(2): p. 177-186.

66. Dunten, R.L., M. Sahintoth, and H.R. Kaback, Role of the Charge Pair Aspartic Acid-237-Lysine-358 in the Lactose Permease of Escherichia-Coli. Biochemistry, 1993. 32(12): p. 3139-3145.

67. Sapse, A.M., et al., The role of salt bridge formation in glucagon: An

experimental and theoretical study of glucagon analogs and peptide fragments of glucagon. Molecular Medicine, 2002. 8(5): p. 251-262.

68. Miyamoto, K., et al., Intraparticle Diffusion in Reaction Catalyzed by

Immobilized Glucoamylase. Journal of Fermentation Technology, 1973. 51(8): p. 566-574.

69. Bahar, T. and A. Tuncel, Immobilization of invertase onto crosslinked poly(p- chloromethylstyrene) beads. Journal of Applied Polymer Science, 2002. 83(6): p. 1268-1279.

70. De Whalley, H.C.S., ICUMSA Methods of Sugar Analysis. 1964, Amsterdam:

Elsevier Publishing Company.

71. Airoldi, C. and O.A.C. Monteiro, Chitosan-organosilane hybrids - Syntheses, characterization, copper adsorption, and enzyme immobilization. Journal of Applied Polymer Science, 2000. 77(4): p. 797-804.

72. Monteiro, O.A.C. and C. Airoldi, Some studies of crosslinking chitosan- glutaraldehyde interaction in a homogeneous system. International Journal of Biological Macromolecules, 1999. 26(2-3): p. 119-128.

73. Airoldi, C. and O.A.C. Monteiro, Copper adsorption and enzyme immobilization on organosilane-glutaraldehyde hybrids as support. Polymer Bulletin, 2003. 50(1-2): p. 61-68.

74. Coates, J., Interpretation of Infrared Spectra, A Practical Approach, in

Encyclopedia of Analytical Chemistry, R.A. Meyers, Editor. 2000, John Wiley & Sons Ltd: Chichester. p. 10815-10837. 75. Mateo, C, et al., Epoxy-amino groups: A new tool for improved immobilization of proteins by the epoxy method. Biomacromolecules, 2003. 4(3): p. 772-777.

76. Gomez, L. and R. Villalonga, Functional stabilization of invertase by covalent modification with pectin. Biotechnology Letters, 2000. 22(14): p. 1191-1195.

77. Mansour, E.H. and F.M. Dawoud, Immobilization of invertase on celite and on polyacrylamide by an absorption procedure. Journal of the Science of Food and Agriculture, 2003. 83(5): p. 446-450.

78. De Queiroz, A.A.A., et al., Lactam-amide graft copolymers as novel support for enzyme immobilization. Journal of Applied Polymer Science, 2002. 84(4): p. 767- 777.

79. Pacek, A.W., P. Ding, and A.W. Nienow, The effect of volume fraction and

impeller speed on the structure and drop size in aqueous/aqueous dispersions. Chemical Engineering Science, 2001. 56(10): p. 3247-3255.

80. Kraume, M., A. Gabler, and K. Schulze, Influence of physical properties on drop size distributions of stirred liquid-liquid dispersions. Chemical Engineering & Technology, 2004. 27(3): p. 330-334.

81. Perez, M., et al., Surfactant-oil-water systems near the affinity inversion. XII.

Emulsion drop size versus formulation and composition. Journal of Dispersion Science and Technology, 2002. 23(1-3): p. 55-63.

Claims

What is claimed is:

1. An immobilized enzyme conjugate comprising a nanopore silica gel support and immobilized thereon an enzyme in an amount, based on the silica gel support, of at least about 500 mg enzyme per gram of silica gel.

2. An immobilized enzyme conjugate according to claim 1, wherein the amount of immobilized enzyme is at least about 700 mg enzyme per gram of silica gel.

3. An immobilized enzyme conjugate according to claim 1, wherein the enzyme is invertase.

4. An immobilized enzyme conjugate according to claim 1, wherein the enzyme is thermolysin.

5. An immobilized enzyme conjugate according to claim 1, wherein the nanopore silica gel is formed in the presence of the enzyme.

6. An immobilized enzyme conjugate according to claim 1, wherein the nanopore silica gel is formed from silicic acid.

7. An immobilized enzyme according to claim 6, wherein the silicic acid comprises silica sol with reactive silanol groups (-Si-OH) formed by ion-exchanging sodium ions from a solution of sodium silicate with hydrogen ions thereby forming a low ionic strength silicic acid.

8. An immobilized enzyme conjugate according to claim 6, wherein the silicic acid is caused to undergo gelation in the presence of a basic compound.

9. An immobilized enzyme conjugate according to claim 8, wherein the basic compound comprises an aminoalkyl silane.

10. An immobilized enzyme conjugate according to claim 8, wherein the basic compound comprises aminopropyltriethoxysilane to thereby introduce an

aminofunctional group to the silica gel support.

11. A method for preparing an immobilized enzyme conjugate which comprises

(a) surface modifying silicic acid by reacting the silicic acid with a basic

aminoalkylsilane in solution in an alcoholic medium, wherein the amount of the basic aminoalkylsilane is sufficient to raise the pH of the solution to induce gelation and form a chemically surface-modified silica gel;

(b) activating the chemically surface-modified silica gel by reaction with a cross-linking agent; and

(c) contacting the activated chemically surface-modified silica gel with an enzyme.

12. A method according to claim 11, wherein the cross-linking agent is a bifunctional aldehyde.

13. A method according to claim 11, wherein the cross-linking agent is gluteraldehyde.

14. An immobilized enzyme conjugate prepared by the method of any one of claims 11 to 13.

15. An immobilized enzyme conjugate according to claim 14, wherein the enzyme is invertase.

16. A bioreactor comprising the immobilized enzyme conjugate according to claim 1 or claim 14, wherein the enzyme comprises thermolysin.

17. A method for the production of low molecular weight protamine fragments which comprises contacting protamine with the bioreactor of claim 16.

18. A method according to claim 17, further comprising resolving the low molecular weight protamine fragments by high presssure liquid chromatography and collecting a low molecular weight protamine fraction comprising about 10 arginine units per fragment.

19. A method according to claim 17, which comprises contacting the protamine with the bioreactor in a plugged flow reactor.

20. A method according to any one of claims 17, 18 or 19, which comprises contacting the protamine with the bioreactor at a temperature in the range of from about 20 ⁰C to about 75 ⁰C.

21. A method according to claim 20, which comprises contacting the protamine with the bioreactor at a temperature of about 55 °C.

22. An immobilized enzyme conjugate according to claim 1, wherein the silica gel support comprises substantially spherical beads having an average particle size in the range of from about 30 to about 100 nm.

23. An immobilized enzyme conjugate according to claim 22, wherein the silica gel support has an average pore size in the range of from about 2 to about 50 nm.

24. An immobilized enzyme conjugate according to claim 22, wherein the silica gel support has an average pore size in the range of from about 30 to about 50 nm.