Historically, the first enzyme immobilization was accomplished in 1916 by Nelson and Griffin, using Al(OH)3 and charcoal for immobilization of invertase from yeast (Nelson and Griffin, 1916). However, in the subsequent period organic supports were preferred, due to the higher reactivity and easier modification or activation of their functional groups, making them suitable for coupling with enzymes. Nowadays, the inorganic carriers, particularly the silica-based materials, are back in the attention. Considering the possible industrial applications, the inorganic supports present several advantages, as high mechanical resistance, temperature and pH stability, resistance to organic solvents and microbial attack.
Numerous inorganic matrixes have been tested as supports for immobilization of enzymes, Part of them were cheap minerals or other materials of natural origin, but synthetic products have numerous advantages and are now used in most applications. An overview of selected inorganic carriers is presented in Table 1.2. Inorganic supports have been used for immobilization by adsorption, entrapment and covalent binding techniques.
Table 1.2. Inorganic supports used for immobilization of enzymes
Pumice cellulase adsorption Pazarlioğlu et al.,
2005
Kaolinite lipase adsorption Iso et al., 2001
Clay minerals cellulase adsorption Sinegani et al.,
2005
Montmorillonite lipase adsorbtion Scherer et al., 2011
Bentonite lipase adsorption Yeşiloğlu, 2005
Synthetic
Alumina pancreatin adsorption Silvestre et al.,
2009
Activated carbon pancreatin adsorption Silvestre et al., 2009
Pumice, ZrOCl2 activated
cellulase adsorption Pazarlioğlu et al., 2005
Mesoporous molecular sieves
horeseradish peroxidase, trypsin
adsorption Deere et al., 2003 Ceramics, derivatized lipase
glucoamylase
adsorption Kamori et al., 2002 Mesoporous silica peroxidase
covalent binding Huckel et al., 1996
Controlled pore
β-D-galactosidase covalent bindig, glutaraldehyde activation
Saito et al., 1994
Magnetic Fe3O4 glucose oxidase, chymotrypsin
peroxidase adsorption Ma et al., 2003
Silica,
entrapment Lei et al., 2002
Adsorption is a simple method that usually results in preservation of catalytic activity, as the active centre is not involved in the immobilization process. Immobilization on natural inorganic supports was used mainly on empirical basis, based on the observation that a certain support has been proved efficient for immobilization of several enzymes. It is the simplest technique and probably the cheapest method to obtain immobilized biocatalysts, but also results in the highest degree of protein desorption during the biocatalytic process.
Consequently, several studies were initiated to tailor the best protein-carrier combinations, based on the physicochemical properties of the support and the surface potential of the biomolecules (Torres-Salas et al., 2011).
Among natural inorganic carriers, porous kaolinite particles were considered appropriate for immobilization of Pseudomonas fluorescens lipase (Iso et al., 2001). When used for production of biodiesels, the activity of immobilized lipase was highly increased in comparison with the free enzyme. The immobilized enzyme was used repeatedly, without troublesome method of separation and significant decrease of its activity. The authors considered that the active sites of the enzyme became more effective following immobilization, as the enzyme was dispersed on the surface of carrier particles.
A very simple natural carrier, CaCO3 has been efficiently used for immobilization of Pseudomonas lipase, to obtain a biocatalyst for transesterification of docosahexaenoic acid ethyl ester with glycerol (Roşu et al., 1998). CaCO3 particles played the role of surfactant and enzyme transporter to the glycerol-ethyl ester interface. Microscopic analysis of the emulsified reaction mixture revealed that CaCO3 particles stabilized the substrate droplets by adsorption on their surface. The reaction rate was five times higher with lipase adsorbed on fine CaCO3 powder compared to the dissolved free lipase, due to a larger interfacial area available for the enzyme action.
Although they are cheap and easily available, natural inorganic materials are not commonly employed as immobilization carriers, since their characteristics are generally not constant. Synthetic materials are more appropriate, as their properties: surface area, surface morphology, chemical composition, porosity, pore distribution, can be assessed for a specific purpose. Moreover, such supports are suitable for structural functionalization.
Numerous synthetic inorganic materials have been investigated as carriers: metals, metal oxides, porous glass, and ceramics, as results from Table 1.2.
Silvestre et al. investigated the immobilization of pancreatin on alumina and activated carbon by adsorption. Activated carbon proved to be a more efficient support, considering
both immobilization yield of the enzyme (100%, compared to 37% on alumina) and catalytic efficiency of the adsorbed biocatalyst. Pancreatin adsorbed on activated carbon was reused 5 times for hydrolysis of whey without any loose of activity, while in case of alumina only 2 hydrolysis cycles were performed at 100% activity. However, at longer utilization (up to 20 reaction cycles) the performance of the two support materials was similar, resulting in 65% and 67% residual activity, respectively (Silvestre et al., 2009).
Since first described in 1992, mesoporous molecular sieves were considered attractive candidates for a wide range of applications in catalysis, having large surface areas (up to 1000 m2g-1), highly ordered pore structures, and tight pore size distributions (Deere et al., 2003). Different adsorption techniques were investigated in connection with these types of carriers, concluding that functionalized mesoporous molecular sieves present enhanced interactions within the protein surface and the surface of the support material. Deere et al.
used two MCM-41 type mesoporous molecular sieve materials, with average pore diameters of 28 and 45Å, for adsorption immobilization of several oxidative enzymes. The proteins have penetrated the mesoporosity, but only partially travelled down through the pore channels. At high protein loadings, a substantial amount was no more accessible to the substrate and the activities declined, due to packing of protein molecules into the mesoporosity (Deere et al., 2003).
As porous silica or glass materials have disadvantages of solubility at pH above 8.0 and hydrolysis of siloxane bonds at strong acidic pH values, for enzymes requiring extreme pH values utilization of transition metals as carriers could be an alternative. Utilization of chelating properties of transition metals was investigated by several research groups.
Particularly, titanium and zirconium oxides have been demonstrated to be useful for activation of different supports, like cellulose, porous glass, or silica gel (Kennedy and Cabral, 1985). A different approach, proposed by Huckel et al., employed porous zirconia, obtained by plasma spray technology, as support material for immobilization. The advantages of zirconia-based materials are high density and extremely high chemical, thermal and pH stabilities. The surface of zirconia was hydrothermally treated and activated with 3-isothiocyanato-propyltriethoxy silane, followed by covalent attachment of enzymes to this activated support. Several proteolytic enzymes were immobilized in this way, obtaining protein coupling efficiencies in the range of 6-80 mg/g. The highest specific activity was obtained for zirconia-chymotrypsin biocatalyst, reaching 61% of activity of the free enzyme (Huckel et al., 1996).
Controlled pore glass and ceramic materials were considered for a long period among the most important inorganic supports. Controlled-pore glass has wide solvent compatibility, can be obtained in 4.5-400 nm range with narrow pore size distribution, but is unstable at pH>8, limiting its applicability (Kennedy and Cabral, 1987). Due to relative chemical inertness of inorganic support materials, which have mainly hydroxyl groups on their surface, derivatization with a silanization agent holding an organic functional group was necessary, to generate an activated support (Fig. 1.1).
O
R - etoxy; X - organic functional group
Fig. 1.1. Silanization of the surface of porous glass (Kennedy and Cabral, 1987)
γ-Aminopropylsilane derivatives of porous glass were the most popular activated supports, being commercialized since the 1970’s by Corning Inc., USA, as carriers for covalent enzyme immobilization (Chibata, 1978). The organosilane derivatives of porous glass were in some cases reacted with the enzyme using a cross-linker. In other cases, the grafted organic functional groups were transformed in more reactive derivatives, able to form covalent linkages with enzymes in mild conditions.
Azevedo et al. reported immobilization of alcohol oxidase on γ-propylamino derivatized controlled pore glass by covalent attachment, using glutaraldehyde as cross-linker. The best results were obtained with porous glass beads having 100 µm diameter and 500Ǻ pore size. Using this biocatalyst in a continuous bioreactor, the initial performance was preserved for more than 14 h operation at 32°C (Azevedo et al., 2004).
Other porous inorganic carriers, like as porous ceramics or porous silica, are more suitable for industrial purposes than controlled-pore glass, on condition to have been obtained with desired pore size and activated in a proper way. Saito et al. used porous ceramic with 37.6 nm mean pore diameter and 77.7 m2/g specific surface for immobilization of a thermostable β-D-galactosidase from Escherichia coli. From several activating organosilanes, 3-[2-(2-aminoethyleminoethylamino)propyl]trimethoxysilane led to the
highest residual activity when β-D-galactosidase was immobilized on the derivatized support by glutaraldehyde cross-linking. The amount of covalently immobilized β-D -galactosidase was double, compared to the enzyme immobilized by physical adsorption on the same carrier. The specific activity of the immobilized enzyme was about 50%, related to the free enzyme (Saito et al., 1994). A different approach to immobilize enzymes on a ceramic carrier was proposed by Kamori et al. They carried out functionalization of the ceramic material by reaction with organosilanes, obtaining methacryloyloxy and amino functional groups on the porous surface. Further on, the immobilization was accomplished by adsorption on these organo-functionalized supports (Fig. 1.2). If the pore size of support and the organic functional group were appropriately selected, the activated support was able to selectively concentrate the enzyme protein from a crude enzyme preparate. Both lipase and glucoamylase immobilized by this method demonstrated high catalytic efficiency (Kamori et al., 2002).
Fig. 1.2. Ceramic support for immobilization of enzymes, non-functionalized (a), derivatized with trimethoxysilyl(propyl) methacrylate (b), and derivatized with
3-(triethoxysilyl)propylamine (Kamori et al., 2002)
Mesoporous silica was proved particularly efficient for immobilization, allowing to develop an immobilization strategy based on structural and functional characteristics of the enzyme (Lee et al., 2009). Ordered mesoporous silicas provide excellent opportunities for immobilization of enzymes via covalent binding, due to the availability of well defined silanol groups. These groups provide reactive sites for functionalization and offer tunable surface properties, allowing to control the position and density of the immobilized catalyst precisely. Particularly, the sol-gel technique, developed in the last decades, was able to produce a new generation of robust and efficient silica-based biocatalysts, with excellent catalytic performances (Pierre, 2004). Since most matrixes obtained by the sol-gel method and utilized for enzyme entrapment are hybrid inorganic-organic materials, they will be discussed in a separate section.
Functionalized magnetic materials
Utilization of magnetic materials instead of conventional water-insoluble supports was first reported in the 1970’s (Zaborsky, 1976). The main advantages of these carriers were the ease of separation and process control. The immobilized enzyme can be simply separated from the reaction mixture, even from viscous solutions or suspensions (Brady and Jordaan, 2009). Moreover, such biocatalysts are particularly suited for fluidized bed operations (Bahar and Celebi, 2000). Most magnetic micro- or nanoparticles used for immobilization were of core-shell type, having the biomolecules connected to the magnetic core through an organic or polymeric shell (Ma et al., 2003). They are basically non-porous supports, but may present some advantages compared to the use of porous supports, as absence of external diffusion problems. The most commonly used magnetic material was magnetite, Fe3O4, a superparamagnetic material. Various immobilization methodologies have been investigated, like as: (i) functionalization of magnetite followed by covalent binding (Koneracká et al., 2002); (ii) adsorption of enzyme on activated magnetite followed by cross-linking with glutaraldehyde (Van Leemputten and Horisberger, 1974); (iii) coating of magnetic particles with a polymeric material by polymerization (Sureshkumar and Lee, 2011) or copolymerization (Kondoi and Fukuda, 1997) in presence of magnetite, followed by covalent attachment of the enzyme. Several types of magnetic composite materials will be discussed in Section 1.3.1.2.
Covalent binding was mainly used for immobilization of enzymes on modified magnetic nanoparticles, by formation of covalent linkages between functional groups on the activated particle surface and appropriate functional groups of enzymes (particularly amino groups). This method often resulted in decrease of enzyme activity, but could achieve strong binding of enzyme on the magnetic particle. Magnetic particles were synthesized as macro ions by Koneracká et al. by the co-precipitation method, and used for immobilization of several enzymes, e.g. glucose oxidase and chymotrypsin. Covalent coupling of enzyme was accomplished with 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride as cross-linking agent. The presence of free hydroxyl groups on the surface of particles was responsible for binding of the protein. The maximum activity of the immobilized enzyme was reached by performing the coupling reaction at pH 4.5 (Koneracká et al., 2002). Another carbodiimide derivative, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide, was studied as activating agent for covalent
immobilization of lipase on Fe3O4 magnetic nanoparticles. The obtained biocatalyst was successfully used for production of biodiesel, as the immobilized lipase retained about 75% of the initial activity and was suitable for recycling (Xie and Ma, 2010). In a different approach, magnetic Fe3O4 nanoparticles were prepared by the chemical coprecipitation method and subsequently coated with 3-aminopropyltriethoxysilane, to achieve amino-functionalized magnetic nanoparticles. The coupling agent for covalent attachment of lipase to the surface of activated nanoparticle was glutaraldehyde, as presented in Fig. 1.3.
The activity recovery yield of Serratia marcescens lipase was up to 62% (Hu et al., 2009).
Fig. 1.3. Reaction scheme of covalent immobilization of lipase on amino-functionalized Fe3O4 nanoparticles (Hu et al., 2009)
Although covalent binding was obviously the most popular method for immobilization on coated or composite magnetic particles, it must be pointed out that adsorption interactions between the organic coatings covering the surface of particles and the appropriate functional groups of the enzymes can not be excluded. Moreover, it was reported the attachment of horseradish peroxidase on such a derivatized magnetic support only by adsorption. 3-Aminopropyl triethoxysilane was used to coat the surface of the Fe3O4 nanoparticles with a near monolayer of amino silane (67% coverage ratio), and the resulting amino active groups were able to adsorb biomolecules (Ma et al., 2003).
Magnetic nanoparticles
Nanotechnology-inspired biocatalyst systems were introduced only in the recent years, the first reports dating from the late 1980’s (Wang, 2006). Enzyme molecules are in order of nanometers, making them possible to benefit from the high surface area-to-volume ratio of nanomaterials. Carbon nanotubes, superparamagnetic nanoparticles, and mesoporous materials represent the most important classes of nano-sized matrices. However,
nanoparticles as such have two disadvantages when used as carriers of enzymes: (i) they often form aggregates and ultrasonication has to be carried out for temporary dispersion;
(ii) because their size, separation by either centrifugation or membranes is not simple.
Superparamagnetism can solve both these problems. Superparamagnetic materials become magnetic only in the presence of a magnetic field. Fortunately, magnetic particles with size less than 30 nm show superparamagnetism, meaning that they disperse easily in solution and can be recovered by use of a simple magnet (Gupta et al., 2011).
Most studies with nanoparticles have been dedicated to improvement of enzyme activity and loading, rather than to enzyme stabilization. However, good enzyme stabilization was reported with covalently attached lipase on magnetic γ-Fe2O3 nanoparticles. The surface of nanoparticles was functionalized by Dyal et al. with 11-bromoundecanoic acid, and the resulted intermediate was reacted with 2-thiophene thiolate, to give thiophene-functionalized nanoparticles. These have been acetylated with acetic anhydride, or were reacted with nitrosonium tetrafluoroborate, to produce the appropriate nitroso derivative.
The acetylated nanoparticles were reacted directly with the enzyme, which was chemically bonded to the nanoparticle surface via a C=N bond. The nitroso-functionalized nanoparticles were reduced with SnCl2, and the enzyme was chemically bonded to the resulted amine-functionalized nanoparticles by glutaraldehyde. It must be noticed that coating with 11-bromoundecanoic acid decreased the magnetization of γ-Fe2O3 only by 12%. The enzyme immobilized on acetylated nanoparticles demonstrated long-term stability, with only 15% decrease of activity over one month (Dyal et al., 2003).