oped a special cross-linker based on dextran (a mixture of glucose oligomers). The modified dextran is too large to enter the active site; hence, in those cases where glutaraldehyde caused loss of activity, active CLEAs were obtained using this large molecule as a cross-linking agent.
Because the enzyme molecules are packed together in a small volume as compared with the free protein in solution, one might expect mass-transport limitations with fast reactions. However, on the time scale of the average biocatalytic synthesis such effects generally turned out to be negligible.
One major advantage of CLEAs is their facile separation afterwards. Because their size is in the micrometer range, filtration is easier; brief centrifugation also results in complete recovery of the catalyst. The stability is highly dependent on the reaction conditions used, but typically detrimental conditions—such as high stirring rate or the bubbling of oxygen necessary for many oxidases—pose few problems. Furthermore, no foam formation is observed when using CLEAs.
As noted earlier, another advantage of the CLEA technology is that it provides the possibility of preparing immobilized biocatalysts containing more than one enzyme (combi-CLEAs). We have demonstrated this concept in the synthesis of (5)-mandelic acid, from benzaldehyde, and aqueous cyanide, using a combi-CLEA containing an (5)-oxynitrilase and a nitrilase (19). The use of a combination of an enantioselective oxynitrilase and an aselective nitrilase is that the latter catalyses hydrolysis of the intermediate cyanohydrin, thus driving the equilibrium of the first step to the right.
The number of enzyme molecules and the way they are packed together in an aggregate can be expected to have a crucial influence on the activity of the aggregate as a whole. Hence, knowing the influential factors and being able to control them would pave the way for changing the CLEA from a phenomenon into a mature, well-defined catalytic particle. During aggregation the solubility of the enzyme in the surrounding medium decreases. When this process is slow the enzyme might denature because of the severe force exerted on its structure. If it is able to find neighboring protein molecules to surround it in time, chances are fairly good that it will retain its tertiary structure. As seen from the high-throughput experiments, increasing the speed of aggregation in cases of poor activity recovery always gave a completely active aggregate. Now the question is, to what extent will enzymes aggregate? Most likely, the surface tension of the aggregate will set the diameter. It is similar to water/oil mixtures where small droplets (emulsion) are formed. Cell- or droplet-like structures of protein aggregates, depending on the surface hydrophobicity, can in that case be expected. The interplay between free-energy increases by interfacial surface formation and free-energy decrease in solids formation governs the critical nucleus size according to the classical nucleation theory. The final aggregate size of the primary particles is then governed by the ratio of nucleation and growth but will likely be small.
Scanning electron microscopy showed a very uniform structure of the aggregates. From the different enzyme aggregates we examined, two types emerged. Type 1 aggregate was formed by Candida antarctica lipase B (see Fig. 1). This enzyme is scarcely glycosylated and highly lipophilic. The diameter of the aggregate is about 1 |m with a small deviation. Taking an enzyme size for CaLB of 5 x 5 x 5 nm, a single CLEA particle contains a maximum of 8 x 106 enzyme molecules. Aggregates built from enzymes with a more glycosylated surface or a surface with more hydrophilic residues were found to be smaller, approx 0.1 |m in diameter, called Type 2 aggregates. C. rugosa lipase (see Fig. 2) and Prunus amygdalus ^-oxynitrilase are examples of that type. These enzymes are glycosylated and therefore have a more hydrophilic surface. One CLEA particle of this aggregate contains a maximum of 8 x 103 enzyme molecules. Although Type 1 aggregates can contain a thousand times more enzyme molecules than Type 2, in both cases the enzymes find themselves in an apparently ideal environment where their natural function is thriving. Because the enzyme molecules are packed together in a small volume compared to the free protein in solution, one might expect mass-transport limitations, particularly with fast reactions. If the CLEAs are finely dispersed in solution—as is the case when the cross-link step is completed and subsequently quenched—this effect is actually small.
CLEAs can form larger clusters (see Fig. 3) that do have mass-transport limitations, especially in fast ultraviolet (UV) tests. The size of these clusters can be up to 100 |m, making them visible to the naked eye. The number of CLEAs in a cluster is much less uniform than enzymes in an aggregate: it can vary from a few to a few hundred thousand. The previous findings with laser scattering (see Fig. 4)
which suggested a variable number of enzyme molecules per aggregate is now rationalized as being a variable number of aggregates per cluster.
Centrifugation leads to increased cluster formation. When a dispersed CLEA is assayed for activity, directly after dilution of the cross-link medium, a higher
activity is observed than when the sample is centrifuged and redispersed. This presses the CLEA particles close together elevating mass-transport limitations to a noticeable level. The interaction between particles can be either reversible (as is demonstrated when these clusters are put in aqueous organic solvents, causing them to break up) or irreversible (when active cross-link residues on the CLEA surface form covalent bonds between individual particles). Some differences between enzymes were observed: with CaLB very large and hydrophobic clusters were obtained whereas with P-galactosidase well-dispersible suspensions were common. The most noticeable structural difference between these two enzymes is that P-galactosidase is extensively glycosylated and CaLB is not. Comparing activities for CaLB CLEA, found in the relatively fast hydrolysis of /-nitrophenyl propionate measured via UV/Vis absorption, with the slower hydrolysis of triacetin (monitored by titration) the mass-transport limitation was obvious. Compared with free enzyme the first showed 35% activity recovery and the latter 177%. For P-galactosidase however, activity recovery found in the hydrolysis of /-nitrophenyl-P-d-galactopyranoside and in lactose hydrolysis was the same. These two CLEAs emphasize the important effect of cluster formation on the apparent activity.
One major advantage of CLEAs is their facile separation from aqueous solutions. In contrast to free enzyme, brief centrifugation results in complete recovery of the catalyst. Another way of isolating CLEAs from a reaction mixture is filtration. The aggregate or cluster size determines how successful filtration will be. Certainly the large clusters do not pose any problem for even a large pore glass filter and the single sub-micrometer aggregates can be filtered with low-MW membrane filters. For medium size clusters the widely used micrometer pore size filters can be used. We applied several filters ranging from 0.2 to 5 |m. As expected, the 1 |m size C. antarctica lipase CLEAs passed through filters with pore sizes ranging from 1 to 5 |m. More surprisingly, they also passed through the 0.2 |m filters demonstrating the elasticity of the "enzyme balls." This suggests that the individual enzymes in an aggregate do have some freedom of movement which is undoubtedly beneficial for their activity. Smaller aggregates like the ones formed from C. rugosa lipase elegantly showed the effect of clustering. With their 0.1 |m size they could be filtered with 0.5 |m filters. Although the polypropylene (PPXL) filters were completely blocked, filtration was possible with the Zeta Plus filters which have a graded density resulting in better cake building. Controlling the size of the clusters means controlling the ability to filter the CLEAs.
The CLEA technology has many advantages in the context of industrial applications. The method is exquisitely simple and amenable to rapid optimization, which translates to low costs and short time-to-market. It is applicable to a wide variety of enzymes, including crude preparations, affording stable, recyclable catalysts with high retention of activity.
In contrast to CLECs, there is no need for the enzyme to be available in crystalline form and the technique can be applicable to the preparation of combi CLEAs containing two or more enzymes. Synthesis of CLEAs in the presence of addi tives, such as crown ethers or surfactants, provides the possibility of "locking" the immobilized enzyme in a more favorable conformation, resulting in increased activity and/or (enantio)selectivities.
Currently the technology is being commercialized by CLEA Technologies on the basis of an exclusive licence from Delft University of Technology. We believe that CLEAs will, in the future, be widely applied in industrial biotransformations and other areas requiring immobilized enzymes.
1. Lipases (Novozymes, Bagsvœrd, Denmark).
2. Trypsin (Novozymes).
3. Galactose oxidase (Hercules, Barneveld, The Netherlands).
4. ß-galactosidase and phytase (DSM, Delft, The Netherlands).
5. Laccase, alcohol dehydrogenase, and formate dehydrogenase (Jülich Fine Chemicals, Jülich, Germany).
6. All other enzymes and chemicals were purchased from Sigma.
7. Scanning Electron Microscope (Philips XL 20).
8. Samples of CLEAs were freeze-dried prior to analysis. UV/Vis spectroscopy was performed on a Varian Cary 3 Bio equipped with a Cary temperature controller or a Greiner microplate U-FORM 96K microtiter plate reader at room temperature.
9. 50-mL Centrifuge tube with a magnetic strirrer bar.
10. 25% (v/v) Glutaraldehyde solution.
11. Lipase assay buffer: 100 mMpotassium phosphate, pH 7.4, 0.4 mMp-nitrophenyl propionate at 25°C.
12. Slow reaction substrate for lipases: Triacetin (2%v/v) in 50 mMTris buffer, pH 7.4, at 40°C.
14. Phytase activity buffer: p-nitrophenyl phosphate (0.4 mM) in 0.1 M acetate, pH 4.5.
15. A solution containing 200 mM galactose (see Note 1), o-toluidine (1.75 mM), peroxidase (60 U/mL), and 100 mM potassium phosphate, pH 7.3, was used to detect galactose oxidase activity.
16. A solution containing glucose (200 mM) (see Note 1), o-toluidine (1.75 mM), peroxidase (60 U/mL), and 100 mM potassium phosphate, pH 7.3, was used to detect glucose oxidase activity.
17. ß-Galactosidase activity buffer: 100 mM potassium phosphate, pH 7.3, 0.4 mM p-nitrophenyl-ß-D-galactopyranoside with hydrolysis (24) was monitored at 25°C and 400 nm.
18. Alcohol dehydrogenase activity buffer: 1.5 mMp-chloro-acetophenone in 100 mM potassium phosphate buffer, pH 6.0, and 0.25 mM nicotinamide-adenine dinucle-otide (NADH).
19. Formate dehydrogenase activity buffer: 100 mMpotassium phosphate buffer, pH 7.5, containing 160 mM sodium formate and 0.25 mM NAD.
20. The filter system was supplied by CUNO Benelux. 0.2-5.0 |im PPXL filters and a 0.5 |m Zeta Plus 050 HT filter were used.
3.1. Cross-Linked Enzyme Aggregate Activity (see Note 2)
1. For the cross-link optimization 90 ||L precipitant containing glutaraldehyde in varying concentrations was added to 10 ||L enzyme solution (see Note 3).
2. The samples were incubated at room temperature without shaking. For the volumetric activity optimization 90 | L precipitant containing glutaraldehyde was added to 10 | L solution with a varying protein content and after the specific cross-linking time this mixture was quenched with 900 | L buffer.
1. In a 50-mL centrifuge tube with a magnetic stirrer bar 1-mL enzyme (25 mg) solution in 100 mM buffer (mentioned in the activity assay) was added to 9-mL precipitant at room temperature unless stated otherwise.
2. Glutaraldehyde was then added to the desired end concentration and the suspension was stirred for 2.5 h.
3. The suspension was diluted with 10-mL buffer and centrifuged.
4. The pellet was resuspended in buffer and centrifuged again.
5. This procedure was performed two times in total.
6. Subsequently, the washed CLEA was resuspended in 1-mL buffer and stored at 4°C.
1. The standard activity test was performed with p-nitrophenyl propionate as substrate (7.8 mg in 1 mL ethanol; 10 |L/mL lipase assay buffer).
2. The reaction was monitored at 400 nm (see Note 4).
3. Additionally, (slow) hydrolysis of 2% triacetine (v/v) in 50-mM Tris buffer, pH 7.4, and 40°C was performed by 0.1 M NaOH titration.
4. C. antarctica lipase A [EC 184.108.40.206] (CaLA) contained 125 U/mL.
5. The CLEA was prepared in DME with 100 mM glutaraldehyde and 7 U/mL.
6. C. antarctica lipase B [EC 220.127.116.11] (CaLB) contained 308 units and 10 mg protein/mL.
7. The Cal B CLEA was prepared in DME with 150 mM glutaraldehyde and 15 U/mL.
8. Thermomyceslanuginosuslipase [EC 18.104.22.168] (TlL) contained 66.7 U/mL.
9. The TlL CLEA was prepared in tert-butyl alcohol with 100 mM glutaraldehyde and 5 U/mL.
10. Rhizomucormieheilipase [EC 22.214.171.124] (RmL) contained 44.1 U/mL.
11. The RmL CLEA was prepared in tert-butyl alcohol with 100 mM glutaraldehyde and 3 U/mL.
1. Enzymatic oxidation (20) of ABTS was monitored at 420 nm and 25°C (see Note 5)
2. Coriolus versicolor laccase [EC 126.96.36.199] (Lacc) contained 0.66 U/mg.
3. The CLEA was prepared in polyethylene glycol (PEG) with 100 mM glutaraldehyde and 2 U/mL.
1. Hydrolysis of p-nitrophenyl phosphate was monitored at 400 nm to determine the activities of phytase (21). (See Note 6).
2. Aspergillus nigerphytase [EC 188.8.131.52] (Phyt) contained 47 U/mg.
3. The CLEA was prepared in ethyl lactate with 100 mMglutaraldehyde and 120 U/mL.
3.2.4. Galactose Oxidase
1. The oxidation of o-toluidine (22) was monitored at 425 nm.
2. Dactylium dendroides galactose oxidase [EC 184.108.40.206] (GalOx) contained 3000 U/mL.
3. The CLEA was prepared in tert-butyl alcohol with 100 mM glutaraldehyde and 225 U/mL.
1. The activity (23) was monitored using the same test as for the lipases (hydrolysis of p-nitrophenyl propionate).
2. Porcine Trypsin contained 30% protein/mg.
3. The CLEA was prepared in saturated ammonium sulphate with 100 mM glutaraldehyde and 2 mg/mL.
3.2.6. Glucose Oxidase
1. The oxidation of o-toluidine was monitored at 425 nm.
2. A. niger glucose oxidase [EC 220.127.116.11] (GlcOx) contained 50 U and 20% protein/mg.
3. The CLEA was prepared in saturated ethyl lactate with 100 mM glutaraldehyde and 125 U/mL.
1. The hydrolysis of p-nitrophenyl-ß-D-galactopyranoside (24) is monitored at 25°C and 400 nm.
2. A. oryzae ß-galactosidase [EC 18.104.22.168] (Gal-ase) contained 227 U/mg.
3. The CLEA was prepared in 2-propanol with 100 mMglutaraldehyde and 850 U/mL.
3.2.8. Alcohol Dehydrogenase
1. Enzyme activity (20) was monitored at 30°C and 340 nm (Mol ext. = 6.22 mM-1cm-1).
2. Rhodococcus erythropolis alcohol dehydrogenase [EC 22.214.171.124] (ADH) contained 77 U and 16.1 mg protein/mL.
3. The CLEA was prepared in saturated ammonium sulphate with 8 mM glutaraldehyde at 4°C and 7.7 U/mL.
3.2.9. Formate Dehydrogenase
1. Enzyme activity was detected (26) with an activity buffer at 30°C and 340 nm (Mol ext. = 6.22 mM-1cm-1).
2. C. boidinii formate dehydrogenase [EC 126.96.36.199] (FDH) contained 70 U and 26.3 mg/mL.
3. The CLEA was prepared in saturated ammonium sulphate with 8 mM glutaraldehyde at 4°C and 7 U/mL.
3.3. Scale-Up and Preparative Use of fi-Galactosidase Cross-Linked
1. Add drop-wise 200 mg of fi-galactosidase in 20-mL potassium phosphate buffer, pH 7.3, to 80 mL 2-propanol in a 250-mL flask with magnetic stirrer immersed in a water bath at room temperature.
2. After complete addition, 3.7 mL 25% glutaraldehyde was added and the suspension was stirred for 1 h.
3. CLEA was then centrifuged off, washed twice with 50 mL 50% ammonium sulphate solution, and subsequently stored in it.
4. To 20 mL of a 50 mMlactose solution in 25 mMpotassium phosphate buffer, pH 7.3, either 4 mg of free enzyme was added or an equivalent CLEA suspension (made from 4 mg).
5. After 3 h or overnight incubation with magnetic stirring at room temperature the reaction mixtures were centrifuged and the supernatants concentrated in vacuo. 13C NMR showed 55% conversion in both cases, whereas overnight incubation gave quantitative hydrolysis.
6. Upon assaying the recovered CLEA, it was found that no activity was lost during lactose hydrolysis.
1. An amount of 100 mg CLEA was suspended in 10 mL of demineralized water. The liquid was pressed through the filter with 2 bar nitrogen gas.
2. The filtrate was then assayed for enzyme activity.
1. Solution is prepared 1 d in advance to allow for mutarotation.
2. For the assays, an enzyme activity is necessary that induces a change in absorption (AA) of 0.1 to 0.5/min. Dilutions of the enzymes were made so that the AA per minute never exceeded 0.5. All activities were correlated to the native enzyme, taken as 100% (no absolute activities are given).
3. To the glutaraldehyde stock 1 vol% phosphoric acid was added and the pH was adjusted to 7.3 with diluted sodium hydroxide prior to mixing with the precipitants
4. Blank reaction rate: AA 0.00318/min
5. The assay buffer was saturated with oxygen prior to addition to the cuvet.
6. The blank reaction at pH 4.5 was negligible.
1. D'Souza, S. F. (1999) Immobilized enzymes in bioprocess. Curr. Sci. 77, 69-79.
2. Kragl, U. (1996) Enzyme membrane reactors. In: IndustrialEnzymology, 2nd edition (Godfrey, T., and West, S., eds.), MacMillan, Basingstoke, pp. 274-283.
3. Fernandez-Lorente, G., Terreni, M., Mateo, C., et al. (2001) Modulation of lipase properties in macro-aqueous systems by controlled enzyme immobilization: enantioselective hydrolysis of a chiral ester by immobilized Pseudomonas lipase. Enzyme Microb. Technol. 28, 389-396.
4. Cao, L., van Langen, L. M., and Sheldon, R. A. (2003) Immobilised enzymes: carrier-bound or carrier-free? Curr. Opin. Biotechnol. 14, 387-394.
5. Doscher, M. S. and Richards, F. M. (1963) The actvitiy of an enzyme in the crystalline state: Ribonuclease S. J. Biol. Chem. 238, 2399-2406.
6. Quiocho, F. A. and Richards, F. M. (1964) Intermolecular cross-linking of a protein in the crystalline state: carboxypeptidase A. Proc. Natl. Acad. Sci. USA 114, 7314-7316.
7. St. Clair, N. L. and Navia, M. A. (1992) Cross-linked enzyme crystals as robust biocatalysts. J. Am. Chem. Soc. 114, 7314-7316.
8. Margolin, A. L. and Navia, M. A. (2001) Protein crystals as novel catalytic materials. Angew. Chem. Int. Ed. Engl. 40, 2204-2222.
9. Lalonde, J. (1997) Practical catalysis with enzyme crystals. Chemtech 27 (2), 38-45.
10. Margolin, A. L. (1996) Novel crystalline catalysts. Tibtech 14, 223-230.
11. Haring, D. and Schreier, P. (1999) Cross-linked enzyme crystals. Curr. Opin. Biotechnol. 3, 35-38.
12. Brown, D. L. and Glatz, C. E. (1987) Aggregate breakage in protein precipitation. Chem. Eng. Sci. 42, 1831-1839.
13. Cao, L., van Rantwijk, F., and Sheldon, R. A. (2000) Cross-linked enzyme aggregates: a simple and effective method for the immobilization of penicillin acylase. Org. Lett. 2, 1361-1364.
14. Wegman, M. A., Janssen, M. H. A., van Rantwijk, F., and Sheldon, R. A. (2001) Towards biocatalytic synthesis of P-lactam antibiotics. Adv. Synth. Catal. 343, 559-576.
15. van Langen, L. M., Oosthoek, N. H. P., van Rantwijk, F., and Sheldon, R. A. (2003) Pencillin acylase catalysed synthesis of ampicillin in hydrophilic organic solvents. Adv. Synth. Catal. 345, 797-801.
16. Lopez-Serrano, P., Cao, L., van Rantwijk, F., and Sheldon, R. A. (2002) NL 1017258, to Delft University of Technology.
17. Theil, F. (2000) Enhancement of selectivity and reactivity of lipases by additives. Tetrahedron 56, 2905-2919.
18. Lopez-Serrano, P., Cao, L., van Rantwijk, F., and Sheldon, R. A. (2002) Cross-linked enzyme aggregates with enhanced activity: application to lipases. Biotechnol. Lett. 24, 1379-1383.
19. Mateo, C., Chmura, A., Rustler, S., van Rantwijk, F., Stolz, A., and Sheldon, R. A., manuscript in preparation.
20. Bourbonnais, R. and Paice, M. G. (1990) Oxidation of nonphenolic substrates -an expanded role for laccase in lignin biodegradation. FEBS Lett, 267, 99-102.
21. Dvorakova, J., Volfova, O., and Kopecky, J. (1997) Characterization of phytase produced by Aspergillus niger. Folia Microbiologica 42, 349-352.
22. Avigad, G., Asensio, C., Horecker, B. L., and Amaral, D. (1962) d-Galactose oxidase of Polyporys circinatus. J. Biol. Chem. 237, 2736-2740.
23. Asaad, N. and Engberts, J. F. B. N. (2003) Cytosol-mimetic chemistry: Kinetics of the trypsin-catalyzed hydrolysis of p-nitrophenyl acetate upon addition of polyethylene glycol and N-tert-butyl acetoacetamide. J. Am. Chem. Soc. 125, 6874-6875.
24. Kim, C. S., Ji, E. S., and Oh, D. K. (2003) Expression and characterization of Kluyveromyces lactis P-galactosidase in Escherichia coli. Biotechnol. Lett. 25, 1769-1774.
25. Rella, R., Raia, C. A., Pensa, M., et al. (1987) A novel archaebacterial NAD+-dependent alcohol dehydrogenase. Purification and properties. Eur. J. Biochem. 167, 475-479.
26. Gröger, H., Hummel, W., Rollmann, C., et al. (2004) Preparative asymmetric reduction of ketones in a biphasic medium with an (5)-alcohol dehydrogenase under in siiu-cofactor-recycling with a formate dehydrogenase. Tetrahedron 60, 633-640.
Immobilization-Stabilization of Enzymes by Multipoint Covalent Attachment on Supports Activated With Epoxy Groups
Cesar Mateo, Olga Abian, Gloria Fernández-Lorente, Benevides C. C. Pessela, Valeria Grazu, Jose M. Guisan, and Roberto Fernandez-Lafuente
Commercial epoxy supports may be very useful tools to stabilize proteins via multipoint covalent attachment if the immobilization is properly designed. In this chapter, a protocol to take full advantage of the support's possibilities is described. The basics of the protocol are as follows: (1) the enzymes are hydrophobically adsorbed on the supports at high ionic strength. (2) There is an "intermolecular" covalent reaction between the adsorbed protein and the supports. (3) The immobilized protein is incubated at alkaline pH to increase the multipoint covalent attachment, thereby stabilizing the enzyme. (4) The hydrophobic surface of the support is hydrophylized by reaction of the remaining groups with amino acids in order to reduce the unfavorable enzyme-support hydrophobic interactions. This strategy has produced a significant increase in the stability of penicillin G acylase compared with the stability achieved using conventional protocols.
Key Words: Multipoint covalent attachment; hydrophobic interactions; hydro-phylization; enzyme stabilization.
Many protocols for covalent immobilization of proteins have been reported. Many are very useful, at least to immobilize and solve the problem of the immobilization at laboratory scale. In these cases, the researcher has to be an expert in both support activation and protein immobilization methods (1-9). In fact, enzyme immobilization is already considered a very well-developed technique. However, most protocols have some drawbacks when used to quantitatively immobilize under mild experimental conditions. One is the need for large amounts of protein per milliliter of support throughout long-term handling of the activated supports when the immobilization is carried out at an industrial level. There are no
From: Methods in Biotechnology: Immobilization of Enzymes and Cells, Second Edition Edited by: J. M. Guisan © Humana Press Inc., Totowa, NJ
current methodologies that may be used for immobilization of large amounts of enzyme in a solid under mild experimental conditions (1-9). Keeping in mind these considerations, epoxy-activated supports seem to be almost ideal systems with which to develop easy protocols for enzyme immobilization. Epoxy groups are very stable at neutral pH values so commercial supports can therefore be stored for long periods of time, meaning that they can be prepared in advance of the enzyme immobilization. Moreover, these epoxy supports may react with very different chemical groups present in the amino acids such as thiol, amine, or aromatic alcohols.
While other popular immobilization protocols may promote great alterations in the physical properties of the protein surface or in some instances yield labile enzyme-support attachments (e.g., when using BrCN-activated supports) (10,11), epoxy supports are able to react with different nucleophilic groups of the protein surface to form extremely strong linkages (secondary amino, ether, or thioether bonds) with a minimal chemical modification of the protein because pK values of the new secondary amino groups are very similar to those of the pre-existing primary amino ones. Thus, the linkages are established through short spacer arms. This is advantageous because they may be used to stabilize enzymes through multipoint covalent attachment via control of the enzyme-support interactions (12). In this way, as it is reported, the residues of proteins involved in immobilization retain their relative positions almost completely and are largely unaffected during any conformational change promoted by heat, pH, organic solvents, or any other distorting agents (13-15). Thus, such multipoint, covalently immobilized enzymes should remain more stable than both their soluble counterparts and other randomly immobilized preparations when exposed to any distorting reagent.
At the end of the immobilization process, epoxy groups can be easily blocked by reaction with very different thiol or amine compounds under mild conditions (16); this prevents further uncontrolled reaction between the support and the enzyme that might decrease its stability.
Negligible immobilizations are found when crude Escherichia coli protein preparations are incubated in the presence of epoxy-agarose supports at both high-and low-ionic strengths. Similar immobilization yields are found when other hydrophobic commercial supports are used at low ionic strength conditions. These results point out that the previous hydrophobic adsorption of proteins on these supports at high-ionic strength (described as necessary for the protein immobilization on these supports [12,17,18]) is a consequence of the hydrophobicity of the support's core and not a result of the presence of the epoxy groups covering the support surface.
Thus, to immobilize proteins onto epoxy supports, the use of a hydrophobic support in the presence of high-ionic strength is necessary. In these conditions the previous adsorption of the protein at neutral pH (where the soluble enzymes are normally more stable) is promoted, and in a second step the covalent intramolecular reaction between the nucleophylic groups and the epoxide of the support is produced (see Fig. 1). In these conditions it is possible to immobilize 70% of the proteins contained in a crude strain of E. coli.
1.2. Optimization of Protein-Support Multipoint Covalent Attachment
As just mentioned, an intense multipoint covalent enzyme-support attachment can stabilize the protein structure, therefore highly stabilized enzyme preparations could be obtained.
The reactivity of the nucleophilic groups present in the protein surface becomes a key parameter in establishing a multipoint covalent attachment. These nucleophilic groups are poorly reactive at neutral pHs (see Fig. 2). Thus, multipoint covalent attachment requires long reaction times at alkaline pHs. Moreover, support and enzyme surfaces are not complementary; therefore, multipoint interaction may require long interaction times to permit the correct alignment between the reactive groups (see Fig. 3).
The blockage of the immobilized support fulfills a double objective: (1) eliminate the reactivity of the support and (2) alter the physical properties of the support (important when considering their hydrophobic nature). A proper blockage may promote the hydrophylization of the support surface and avoid the hydrophobic interactions with the protein hydrophobic pockets that could interact with the partially distorted structure of the protein and promote an apparent "destabilization" of the enzyme (see Fig. 4).
In fact, most of the commercial supports recommend the use of 2-mercaptoethanol as a blocking agent. However, there are numerous compounds that are able to react with the epoxy groups—such as other thiols, amino acids, and other amines—that can yield hydrophilic surfaces.
1. Epoxy Sepabeads® EC-EP (Resindion SRL, Milan, Italy) (see Note 1).
2. Eupergit® C and 250L (Degussa, Dusseldorf, Germany) (see Note 2).
3. Immobilization buffer: 1 Msodium phosphate buffer, pH 7.0, adjusted with 5 M NaOH (see Note 3).
4. Incubation buffer: 100 mM sodium phosphate adjusted with 5 MNaOH to pH 10.0. Some additives may be used (see Note 4).
5. Blocking buffer: 3 Mglycine, pH 8.5.
Fig. 3. Enzyme immobilization on epoxy supports at pH 10.0.
3.1. Preparation of Epoxy Supports and Enzyme Solution
1. Wash the support 10 times with 5 vol of distilled water, pH 7.0, at 4°C using a Büchner flask with a glass-sintered funnel connected to a vacuum line for filtration.
2. Dissolve the proteins in 1 M sodium phosphate, pH 7.0 (see Subheading 2.). Take a sample as a reference blank and test the enzyme activity (see Note 5).
3.2. Immobilization of Proteins on Amino-Epoxy Supports
1. Add the support to the enzyme solution.
2. Keep the suspension (enzyme and gel) under mild stirring at 25°C. (see Note 6)
3. Periodic samples of the supernatant and suspension were taken for assay of enzyme activity. Supernatant was achieved either by using pipet filter or by centrifugation of the suspension (see Note 7).
Was this article helpful?