IE85134B1 - Use of Crosslinked Crystals as a Novel Form of Enzyme Immobilization - Google Patents

Abstract

ABSTRACT A method of immobilizing an enzyme by forming crystals of the enzyme and, generally, also cross-linking the resulting crystals through the use of a bifunclional reagent; crosslinked, immobilized enzyme crystals (CLECS) made by this method; lyophilization of CLECs made by this method; lyophilised crosslinked immobilized CLECs and a method of making a desired product by means of a reaction catalysed by a CLEC or set of CLECs.

Description

PATENTS ACT, 1992 2004/0113 USE OF CROSSLINKED CRYSTALS AS A NOVEL FORM OF ENZYME TMMOBTLIZATION VERTEX PHARMACEUTICALS INCORPORATED Enzymes are used as industrial catalysts for the large and laboratory scale economical production of fine and specialty chemicals (Jones, J.B., Tetrahedron 42: 3351-3403 (1986)), for the production of foodstuffs (zaks et. al., Trends in Biotechnology 6: 272-275 (1988)), and as tools for the synthesis of organic compounds (Wong, C.-H., Science 244: 1145-1152 (1989); CHEMTRACTS-org. Chem. 3: 91-111 (1990); Klibanov, A.M., Acc. Chem. Res. 23: 114-120 (1990)).

Enzyme-based manufacturing can significantly reduce the environmental pollution burden implicit in the large scale manufacturing of otherwise unusable chemical intermediates, as shown in the large scale production of acrylamide using the enzyme, nitrile hydratase (Nagasawa, T. and Yamada, H., Trends in Biotechnology 7: 153-158 (1989)).

Enzymes are also used in biosensor applications to detect various substances of clinical, industrial and other interest (Hall, E., ”Biosensors“, open University Press (1990)). In the clinical area, enzymes may be used in extracorporeal therapy, such as hemodialysis and hemofiltration, where the enzymes selectively remove waste and toxic materials from blood (Klein, H.

E85134 and Langer, R., Trends in Biotechnology 4: 179-185 (1986)). Enzymes are used in these areas because they function efficiently as catalysts for a broad range of reaction types, at modest temperatures, and with substrate specificity and stereoselectivity.

Nonetheless, there are disadvantages associated with the use of soluble enzyme catalysts which have limited their use in industrial and laboratory chemical processes (Akiyama et. al., CHEMTECH 627-634 (1988)).

Enzymes are expensive and relatively unstable compared to most industrial and laboratory catalysts, even when they are used in aqueous media where enzymes normally function. Many of the more economically interesting chemical reactions carried out in common practice are incompatible with aqueous media, where, for example, substrates and products are often insoluble or unstable, and where hydrolysis can compete significantly. In addition, the recovery of soluble enzyme catalyst from product and unreacted substrate in the feedstock often requires the application of compli- cated and expensive separation technology. Finally, enzymes are difficult to store in a manner that retains their activity and functional integrity, for commer- cially reasonable periods of time (months to years) without having to resort to refrigeration (4°C to -80°C to liquid N, temperatures), or to maintenance in aqueous solvents of suitable ionic strength, pH, etc.

Enzyme immobilization methods have, in many instances, circumvented these disadvantages.

Immobilization can improve the stability of enzyme {catalysts and protect their functional integrity in the harsh solvent environments and extreme temperatures characteristic of industrial and laboratory chemical processes (Hartmeier, W., Trends in Biotechnology 3: 149-153 (1985)). Continuous flow processes may be operated with immobilized enzyme particles in columns, for example, where the soluble feedstock passes over the particles and is gradually converted into product.

As used herein, the term enzyme immobilization refers to the insolubilization of enzyme catalyst by attachment to, encapsulation of, or by aggregation into macroscopic (10” mm) particles.

A number of useful reviews of enzyme immob- ilization methods have appeared in the literature (Maugh, T.H., Science 223: 474-476 (1984); Tramper, J., Trends in Biotechnology 3: 45-50 (1985)). Maugh describes five general approaches to the immobilization of enzymes. These include: adsorption on solid supports (such as ion-exchange resins); covalent attachments to supports (such as ion-exchange resins, porous ceramics or glass beads); entrapment in poly- meric gels; encapsulation; and the precipitation of soluble proteins by cross-linking them with bifunc- tional reagents in a random and undefined manner.

EPfA-367302 (soejima 3; _l.) describes a water-soluble cross-linked Achromobacter protease I for use in a process for semi-synthesizing human insulin. EP-A-341503 (visuri) describes a water- insoluble glucose isomerase which is formed by a crystalline enzyme converted to solid form by cross- linking with dialdehyde (glutaraldehyde) in the presence of a compound containing an amino group. In addition, one can immobilize whole cells (usually dead and made permeable) which have expressed the desired enzyme activity at high levels (e.g., Nagasawa, T. and Yamada, H., Trends in Biotechnology 7: 153-158 (1989)).

Each of these immobilization procedures has its own advantages and limitations and none can be considered optimal or dominating. In most of them, the enzyme catalyst ultimately represents only a small fraction of the total volume of material present in the chemical reactor. As such, the bulk of the immobilized medium is made up of inert, but often costly carrier material. In all of them, the immobilizing interactions of the enzyme catalyst molecules with each other and/or with the carrier material tend to be random and undefined. As a result, although these interactions confer some enhanced stability to the enzyme catalyst molecules, their relative non-specificity and irregularity makes that stabilization sub-optimal and irregular. In most cases, access to the active site of the enzyme catalyst remains ill-defined. In addition, the immobilization methods described above fail to deal with problems associated with storage and refrigeration. Nor can conventionally immobilized enzymes generally be manipulated, as in being exchanged into one or another solvent of choice, without risk to the structural and functional integrity of the enzyme. In practical terms, except for the attached tether to the carrier particle, conventionally immobilized enzymes bear close resemblance to soluble enzymes, and share with them a susceptibility to denaturation and loss of function in harsh environments.

In general, immobilization methods lead to a reduction of observed enzyme-catalyzed reaction rates relative to those obtained in solution. This is mostly a consequence of the limits of inward diffusion of substrate and outward diffusion of product within the immobilized enzyme particle (Quiocho, F.A., and Richards, F.M., Biochemistry 5: 4062-4076 (1967)).

The necessary presence of inert carrier in the immobilized enzyme particles increases the mean free path between the solvent exterior of the immobilized enzyme particle and the active site of the enzyme catalyst and thus exacerbates these diffusion problems. when dealing with immobilized cells, the diffusion problem is particularly severe, even if cell walls and membranes are made permeable to substrate and product in some way. one would further be concerned with the multitude of contaminating enzymatic activities, metabolites, and toxins contained in cells, and with the stability of cells in harsh solvents or extreme temperature operating environments. An improved immobilization technique which avoids the limitations of the presently available methods would be helpful in promoting the use of enzymes as industrial catalysts, particularly if it were shown to be useful on a large scale (Daniels, M.J., Methods in Bnzymology 136: 371-379 (1987)).

The present invention relates to cross-linked immobilized enzyme crystals (referred to as CLECS or CLIECS).

In the method disclosed herein, small (lO*mm) protein crystals are grown from aqueous solutions, or aqueous solutions containing organic solvents, in which the enzyme catalyst is structurally and functionally stable. In a preferred embodiment, crystals are then cross-linked with a bifunctional reagent, such as glutaraldehyde. This cross-linking results in the stabilization of the crystal lattice contacts between the individual enzyme catalyst molecules constituting the crystal. As a result of this added stabilization, the cross-linked immobilized enzyme crystals can function at elevated temperatures, extremes of pH and in harsh aqueous, organic, or near-anhydrous media, including mixtures of these.

That is, a CLEC of the present invention can function in environments incompatible with the functional integrity of the corresponding uncrystallized, uncrosslinked, native enzyme or conventionally immobilized enzyme catalysts.

In addition, CLECs made by this method can be subjected to lyophilization, producing a lyophilized CLEC which can be stored in this lyophilized form at non-refrigerated (room) temperatures for extended periods of time, and which can be easily reconstituted in aqueous, organic, or mixed aqueous-organic solvents of choice, without the formation of amorphous suspensions and with minimal risk of denaturation.

The present invention relates to CLECS produced by this method and to their use in laboratory and large scale industrial production of selected materials, such as chiral organic molecules, peptides, carbohydrates, lipids, or other chemical species. Presently, these are typically prepared by conventional chemical methods, which may require harsh conditions (e.g. aqueous, organic or near-anhydrous solvents, mixed aqueous/organic solvents or elevated temperatures) that are incompatible with the functional integrity of uncrystallized, uncrosslinked, native enzyme catalyst. other macromolecules with catalytic activity can also be incorporated into the proposed CLEC technology. These might include catalytic antibodies (Lerner, R.A., Benkovic, S.J., and Schultz, P.G., Science 252:659-667 (1991)) and catalytic polynucleotides (Cech, T.R., Cell 64:667-669 (1991); Celander, D.W., and Cech, T.R. Science, 2S1:401-407 (1991)).

The present specification also describes a method of making a selected product by means of a reaction catalyzed by a CLEC of the present invention.

In an example of this method and practice, the enzyme thermolysin, a zinc metalloprotease, was used to synthesize a chiral precursor of the dipeptidyl artificial sweetener, aspartame. The enzyme thermolysin was crystallized from a starting aqueous solution of 45% dimethyl sulfoxide, and 55% 1.4M calcium acetate, 0.05M sodium cacodylate, pH 6.5. The resulting crystals were cross-linked with glutaraldehyde to form a thermolysin CLEC. The thermolysin CLEC was then transferred from the aqueous crystallization solution in which it was made, into a solution of ethyl acetate containing the substrates, N-(benzyloxycarbonyl)-L-aspartic acid (Z-L-Asp) and L-phenylalanine methyl ester (L-Phe-one).

The thermolysin CLEC was then used to catalyze a condensation reaction of the two substrates to synthesize N-(benzyloxycarbonyl)-L-aspartyl- L-phenylalanine methyl ester (Z-L-Asp-L-Phe-oMe), which is the dipeptidyl precursor of the artificial sweetener aspartame. Using any one of many known techniques (see, e.g. Lindeberg, G., J. Chem Ed. 64: 1062-1064 (1987)) the L-aspartic acid in the synthesized dipeptidyl precursor can be deprotected by the removal of the benzyloxycarbonyl (Z-) group to produce aspartame (L-Asp-L-Phe-OMe).

In a second example of this method and practice, the enzyme thermolysin was used to produce thermolysin CLECs. The activity and stability of thermolysin CLECs were compared to that of soluble thermolysin under optimum conditions and conditions of extreme pH and termperature, following incubation in the presence of organic solvents and following incubation in the presence of exogenous protease.

The enzyme thermolysin was crystallized from a solution of 1.2M calcium acetate and 30% dimethyl -.9- sulfoxide pH 8.0. The resulting crystals were crosslinked with glutaraldehyde at a concentration of 12.5% to form a thermolysin CLEO.) The thermclysin CLEC was then lyophilized by a standard procedure (Cooper, T.G., o o em’ , pp. 379-380 (John Wiley and sons, NY (1977)) to form a lyophilized enzyme CLEC of thermolysin. This lyophilized CLEC was then transformed directly into the different aqueous, organic, and mixed aqueous/organic solvents of choice without an intervening solvent exchange procedure, without formation of amorphous suspensions, and with minimal risk of denaturation. These solvents included acetonitrile, dioxane, acetone, and tetrahydrofuran, but not to the exclusion of others. Following incubation, activity was assayed spectrophotometrically by cleavage of the dipeptide substrate FAGLA (furylacryloyl-glycyl-L-leucine amide).

In a third example of this method and practice, the enzyme elastase (porcine pancreatic) was crystallized from an aqueous solution of 5.5 mg/ml protein in 0.1 M sodium acetate at pH 5.0 at room temperature (Sawyer, L. et al., J. Mol. Biol. 1l8:137-208). The resulting crystals were crosslinked with glutaraldehyde at a concentration of 5% to form an elastase CLEC. The elastase CLEC was lyophilized as described in Example 2.

In a fourth example of this method and practice, and as disclosed here, the enzyme esterase (porcine liver) was crystallized from an aqueous solution of 15 mg/ml protein in 0.25 M calcium acetate at pH 5.6 at room temperature. The resulting crystals were crosslinked with glutaraldehyde at a concentration of 12.5% to form an esterase CLEC.

The esterase CLEC was lyophilized as described in Example 2.

In a fifth example of this method and practice, and as disclosed here, the enzyme lipase (ceotrichum gandidgm) was crystallized from an aqueous solution of 20 mg/ml protein in 50 mM Tris at pH 7 at room temperature. The resulting crystals were crosslinked with glutaraldehyde at a concentration of 12.5% to form a lipase CLEC. The ’ lipase CLEC was lyophilized as described in Example 2.

In a sixth example of this method and practice, the enzyme lysozyme (hen egg white) was crystallized from an aqueous solution of 40 mg/ml protein in 40 mM sodium acetate buffer containing % sodium chloride at pH 7.4 at room temperature (Blake, C.C.F. et al., Nature, 196:1l73 (1962)). The resulting crystals were crosslinked with glutaraldehyde at a concentration of 20% to form a lysozyme CLEC. The lysozyme CLEC was lyophilized as described in Example 2.

In a seventh example of this method and practice, the enzyme asparaginase (Escherichia 9911) was crystallized from an aqueous solution of 25 mg/ml protein in 50 mM sodium acetate and 33% ethanol at pH 5.0 at 4°C. The crystallization is a modification of the procedure described by Grabner et al. [U.S. Patent 3,664,926 (1972)]. As disclosed here, the resulting crystals were crosslinked with glutaraldehyde at a concentration of 7.5% to form an asparaginase CLEC. The asparaginase CLEC was lyophilized as described in Example 2. other enzymes which can be immobilized in a similar manner and used to catalyze an appropriate reaction include luciferase and urease. other enzymes, such as those listed in Tables 1-5, can also be crystallized and crosslinked using this method, to produce a desired CLEC which can, in turn, be used to catalyze a reaction which results in production of a selected product or to catalyze a reaction which is an intermediate step (i.e. one in a series of reactions) in the production of a selected product.

CLECs have several key characteristics that confer significant advantages over conventional enzyme immobilization methods presently in use. CLECs dispense with the need for a separate, inert support structure. Lack of an inert support will improve substrate and product diffusion properties within CLECs and provides enzyme concentrations within the crystal that are close to the theoretical packing limit for molecules of such size. High enzyme concentrations can lead to significant operational economies through the increased effective activity of a given volume of catalyst, reduction in substrate contact time with enzyme and overall reduction in plant size and capital costs (Daniels,M.J., Methods in Enzymol. 136: 371-379 (1987)). The uniformity across crystal volume and enhanced stability of the constituent enzyme in CLECs creates novel opportunities for the use of enzyme catalysis in harsh conditions, such as elevated temperature, and aqueous, organic or near-anhydrous solvents, as well as mixtures of these. In addition, the restricted solvent access and regular protein environment implicit in a crystal lattice should lead to improved metal ion and co-factor retention for CLECs vs. conventional immobilized enzyme systems. grief Qescription of the Drawings Figure 1 is a graphic representation of results of assessment of enzymatic activity of soluble and thermolysin CLEC.

Figure 2 is a graphic representation of results of a comparison of pH dependencies of thermolysin CLEC and soluble thermolysin.

Figure 3 is a graphic representation of measurement of the activity of soluble and crystalline thermolysin after incubation at 65°C.

Figure 4 is a graphic representation of results of assessment of resistance of soluble and thermolysin CLEC to exogenous proteolytic degradation.

Figure 5 is a graphic representation of results of the assessment of enzymatic activity for soluble elastase and the corresponding elastase CLEC.

Figure 6 is a graphic representation of the resistance of soluble elastase and the corresponding elastase CLEC to exogenous proteolytic degradation.

Figure 7 is a graphic representation of results of the assessment of enzymatic activity for soluble esterase and the corresponding esterase CLEC.

Figure 8 is a graphic representation of the resistance of soluble esterase and the corresponding esterase,CLEC to exogenous proteolytic degradation.

Figure 9 is a graphic representation of results of the assessment of enzymatic activity for soluble lipase and the corresponding lipase CLEC.

Figure 10 is a graphic representation of results of the assessment of enzymatic activity for soluble lysozyme and the corresponding lysozyme CLEC.

Figure 11 is a graphic representation of results of the assessment of enzymatic activity for soluble asparaginase and the corresponding asparaginase CLEC. -l3a- Detailed Description Of The Invention The present invention relates to . An enzyme crystal crosslinked with a multifunctional crosslinking agent, said crosslinked enzyme crystal having resistance to exogenous Pronasem proteolysis and being selected from the group consisting of: b) C) pyridoxal phosphate enzymes, metalloenzymes, CoA-requiring enzymes and sulfurases (PAPS); thermolysin, nitrilase, aminocyclase and hydantoinase; elastase, aminopeptidase, asparaginase, glutaminase, arginase, urease, alkaline phosphatase, lipase, preferably phospholipase, B-lactamase, ACL hydrolase, L-ACT hydrolase, arylsulfatase, glycosidase, preferably [3-galactosidase and B-fructosidase; amidase, protease, peptidase, lactase, amino acid‘ esterase, '-phosphodiesterase, nuclease and creatine deiminase; dehydrogenase, amino acid dehydrogenase, formate dehydrogenase, lactate dehydrogenase, glutamate dehydrogenase, hydroxysteroid dehydrogenase, succinate glucosepymvate dehydrogenase, dehydrogenase, glucose dehydrogenase, alcohol dehydrogenase, catalase, superoxide dismutase, peroxidase, tyrosinase, isomerase, flavoenzyme, nitrate/nitrite reductase, oxidoreductase (NAD/P), oxidase and lyase; or kinase (ATP), preferably creatine kinase, transaminase, transferase, glycosyl transferase and methyl transferase (SAM) for use in a medicament.

. The enzyme crystal according to . The enzyme crystal according to The enzyme crystal according to , wherein said enzyme is thermolysin. , wherein said enzyme is elastase. , wherein said enzyme is asparaginase. — 13b- The enzyme crystal according to 1, wherein said enzyme is lipase.

The enzyme crystal according to 1, wherein said oxidase is alcohol oxidase, glucose oxidase, bilirubin oxidase, lactate oxidase, cholesterol oxidase or L-amino oxidase, and/or said lyase is aldolase, L-aspartate 4-decarboxylase, phenylalanine ammonium lyase, carbonic anhydrase, nitrile hydratase, L-aspartase, fumarase or heparinase.

The enzyme crystal according to 1, selected from the group consisting of: L- tryptophan synthetase and strictodine synthetase.

The enzyme crystal according to 1, wherein said enzyme is as defined in c) of claim 1 and wherein said crosslinked enzyme crystal retains at least 91% of its initial activity after incubation for three hours in the presence of a concentration of Pronasem that causes the soluble uncrosslinked form of the enzyme that is‘ crystallized to form said enzyme crystal that is crosslinked to retain at most 6% of its initial activity under the same conditions, and said enzyme being esterase, and The enzyme crystal according to any one of 1 to 8, wherein the crosslinker is glutaraldehyde.

A simple, general procedure that assures stability and function for a given enzyme or set of enzymes under conditions which are of interest to the synthetic chemist and which are too harsh for use with enzymes using presently available methods, would be very useful. Cross-linked immobilized enzyme crystals (referred to as CLECS or CLIECS) as described here can be used for this purpose. stabilization of the crystal lattice and of the constituent enzyme catalysts in the crystal by the cross-linking reaction permits the use of CLECs in environments, including aqueous, organic or near-anhydrous solvents, mixtures of these solvents, extremes of pH and elevated temperatures, which are incompatible with enzyme function using presently available methods. In addition, the stabilization of the crystal lattice in CLECs makes possible the lyophilization of CLECs by standard methods.

Lyophilized CLECs can be stored for commercially attractive periods of time (months to years) in the absence of refrigeration, and facilitate the rapid and uncomplicated utilization of cLECs in industrial and laboratory scale processes by the simple addition of solvents of choice, without need for intervening solvent exchange processes. CLBCS are also highly resistant to digestion by exogenous proteases. The method of the present invention facilitates the use of versatile enzyme catalysts in mainstream industrial chemical processes, as well as in the laboratory synthesis of novel compounds for research.

The method described includes cross-linking of a crystal enzyme and is described in detail in the following sections.

The regular interactions between the constituent enzyme molecules in the crystal lattice of a CLEC result in well defined pores of limited size leading to the enzyme molecules within the body of a CLEC. As a result, substrates larger than the available pore size will not penetrate the body of the CLBC particle.

As a consequence of the limited pore size, many enzymatic reactions of commercial and academic interest involving substrates larger than the pore size of the cLECs would be beyond the scope of the present invention. This would include most reactions involving large polymers, such as proteins, polynucleotides, polysaccharides, and other organic polymers, where the number of polymeric subunits would be such as to make the polymer larger than the crystal pore size in CLECs.

In such instances, however, catalysis can still take place on the CLEC surface.

The specification discloses a method of immobilizing a selected protein, particularly an enzyme, by crystallizing and crosslinking the protein, resulting in production of a crosslinked immobilized enzyme crystal (CLEC) which can be used to catalyze production of a selected product, such as a peptide, carbohydrate, lipid or chiral organic molecule. The present invention relates to such CLECS and discloses a method of making a selected product by means of a CLEC-catalyzed reaction or CLEC-catalyzed step in a series of reactions. In one example, the dipeptidyl precursor of aspartame has been produced in a condensation reaction catalyzed by cross-linked immobilized thermolysin made by the present method. In another embodiment of this invention, the indicator substrate, FAGLA, has been cleaved to produce a colorimetric product, whose presence is indicative of enzyme activity in a thermolysin CLEC. FAGLA hydrolysis has been used as a model reaction to indicate the robustness of the thermolysin CLEC in a number of environments that would be normally incompatible with that enzyme's activity.

In other examples, the enzymes elastase, esterase, lipase, asparaginase, and lysozyme have been used to cleave various indicated substances, such as p-nitrophenyl acetate (esterase and lipase), succinyl- (ala)3-p-nitroanilide (elastase), 4- methylumbelliferyl N-acetyl-chitrioside (lysozyme) and NADH (asparaginase).

By this method, one of ordinary skill in the art can adapt a protocol for making a desired product by means of a reaction catalyzed by an immobilized enzyme. The enzyme of interest, when crystallized from an appropriate solution, can be cross-linked with glutaraldehyde or other suitable bifunctional reagent in the crystallization solution to produce a CLEC of that enzyme. Subsequently, the CLEC of the enzyme of choice can be lyophilized as described in Example 2.

There are several advantages which the use of a CLEC offers over presently—available enzyme-catalyzed methods. For example, the cross-linked crystal matrix in a CLEC provides its own support. Expensive carrier beads, glasses, gels, or films are not required in order to tie down the enzyme catalyst, as they are in presently-available immobilization methods. As a result, the concentration of enzyme in a CLEC is close to the theoretical packing limit that can be achieved for molecules of a given size, greatly exceeding densities achievable even in concentrated solutions.

The entire CLEC consists of active enzyme (and not inactive carrier), and thus, the diffusion-related reduction of enzyme reaction rates usually observed with conventionally immobilized enzymes relative to enzymes in solution should be minimized, since the mean free path for substrate and product between active enzyme and free solvent will be greatly shortened for CLECs (compared to a conventional immobilized enzyme carrier particles). These high protein densities will be particularly useful in biosensor, analytical and other applications requiring large amounts of protein in small volumes. In industrial processes, the superior performance and compactness of CLECs results in significant operating economies, by increasing the effective activity of a given volume of catalyst, thereby allowing reductions in plant size, as well as capital costs (Daniels, M.J., Methods in Enzymol. 136: 371-379 (1987)). CLECS are relatively monodisperse, with a macroscopic size and shape reflecting natural crystal growth characteristics of the individual enzyme catalysts. Replacement of existing carrier-immobilized enzyme media with CLECs should not be difficult, since both systems are comparable in size and shape, and both can be similarly recovered from feedstock by any number of simple methods, including basic economical operations such as filtration, centrifugation, decantation of solvent, and others.

In addition, the use of lyophilized CLECs permits routine handling and storage of these materials prior to use (dry storage at room temperature without refrigeration, for extended periods of time).

Lyophilized CLECS also allow for routine formulation by direct addition of solvents and substrates of interest, without lengthy solvent exchange processes, or the formation of amorphous suspensions. The lyophilized CLEC form extends the general utility of the enzymes as catalysts to a broader spectrum of enzymes and functional conditions.

A second advantage of a CLEC is that cross-linking of the crystallized enzyme stabilizes and strengthens the crystal lattice and the constituent enzyme molecules, both mechanically and chemically. As a result, a CLEC may be the only means of achieving significant concentrations of active enzyme catalyst in harsh aqueous, organic, near-anhydrous solvents, or in aqueous-organic solvent mixtures. The use of enzymes as catalysts in organic syntheses has been hampered by their tendency to denature in the presence of non—aqueous solvents, and particularly, in mixtures of aqueous and non—aqueous solvents (Klibanov, A.M., Trends in Biochemical Sciences, 14:141-144 (1989)). In CLECS, the restriction of conformational mobility that leads to stability is provided by the inter- molecular contacts and cross-links between the constituent enzyme molecules making up the crystal lattice, rather than by the near-absence of water in the medium. As a result, intermediate water concentrations can be tolerated by enzymes when formulated as CLECs, as has previously not been possible (see Table 12). In commercial applications, aqueous-organic solvent mixtures allow manipulation of product formation by taking advantage of relative solubilities of products and substrates.

Even in aqueous media, enzyme catalysts, immobilized or soluble, are subject to mechanical forces within a chemical reactor that can lead to denaturation and a shortened half-life. The chemical cross-links within the CLEC provide the necessary mechanical strength (Quiocho and Richards, Proc. Natl. Acad. Sci. (USA) 52: 833-839 (1964)) that results in increased reactor life for the enzyme catalyst.

A third advantage of a CLEC is that as a result of its crystalline nature, a CLEC can achieve uniformity across the entire cross-linked crystal volume.

Crystalline enzymes as described herein are grown and cross-linked in an aqueous environment and, therefore, the arrangement of molecules within the crystal lattice remains uniform and regular. This uniformity is maintained by the intermolecular contacts and chemical cross-links between the enzyme molecules constituting the crystal lattice, even when exchanged into other aqueous, organic or near-anhydrous media, or mixed aqueous/organic solvents. In all of these solvents, the enzyme molecules maintain a uniform distance from each other, forming well-defined stable pores within the CLECs that facilitate access of substrate to the enzyme catalysts, as well as removal of product.

Uniformity of enzyme activity is critical in industrial, medical and analytical applications where reproducibility and consistency are paramount.

A fourth advantage of using a CLEC is that it should exhibit an increased operational and storage half-life. Lattice interactions, even in the absence of cross-linking, are known to stabilize proteins, due in part to restrictions of the conformational degrees of freedom needed for protein denaturation. In CLECs, the lattice interactions, when fixed by chemical cross-links, are particularly important in preventing denaturation, especially in mixtures of aqueous and non-aqueous solvents (Klibanov, A.M., Trends in Biochemical Sciences 14: 141-144 (1989)). Enzymes that have been in the crystalline state for months or years routinely retain a high percentage of their catalytic activity. Cross-linked immobilized enzyme crystals stored in anhydrous solvents will be even further protected from microbial contamination and damage, which is a serious problem in storing large quantities of protein in a nutrient rich, aqueous environment.

In the case of a lyophilized CLEC, the immobilized enzyme is stored in the absence of solvent. That, and the stabilization achieved by cross-linking allows for the storage in the absence of refrigeration for long periods of time.

A fifth advantage of using a CLEC is that it should exhibit enhanced temperature stability as a consequence of the cross-links stabilizing the crystal lattice. Carrying out reactions at a higher temperature than that used with conventional methods would increase reaction rates for the chemical reactions of interest, both thermodynamically, and by enhancing the diffusion rate into and out of the crystal lattice of CLECs. These combined effects would represent a major improvement in reaction efficiency, because they would maximize the productivity of a given quantity of enzyme catalyst, which is generally the most expensive component of the reaction process (Daniels, M.J., Methods in Enzymol. 136: 371-379 (1987)). The temperature stability exhibited by CLECs is remarkable because most enzyme systems require mild reaction conditions. CLECs would also be stabilized against denaturation by transient high temperatures during storage.

A final advantage of use of a CLEC is that pores of regular size and shape are created between individual enzyme molecules in the underlying crystal lattice. This restricted solvent accessibility greatly enhances the metal ion or cofactor retention characteristics of CLEC as compared to conventionally immobilized enzymes and enzymes in solution. This property of CLEC will permit the use of economically superior continuous-flow processes in situations (see e.g. oyama et. al. Methods in Enzymol. 136 503-516 (1987)) where enzyme would otherwise be inactivated by metal ion or cofactor leaching. For example, in the thermolysin-mediated synthesis of the dipeptidyl aspartame precursor, Z-L-Asp-L-Phe-One, conventionally immobilized enzyme is known to lose catalytic activity in continuous-flow column processes, in part through the leaching of calcium ions essential for thermolysin activity. In practice, leaching of calcium ions has forced the use of less efficient batch processes (Nakanishi et. a1., Biotechnology 3: 459-464 (1985)).

Leaching occurs when calcium ion complexes are formed with substrate Z-L-Asp, in competition with the natural calcium binding sites on the surface of the enzyme, resulting in the loss of catalytic activity. The high density of enzyme, and the correspondingly limited volume accessible to solvent in the interstices of the CLECs, discourages the formation of the competing L-Asp-Ca** complexes responsible for metal ion leaching.

Preparation of CLECs - enzyme crystallization By this method, a cross-linked immobilized enzyme crystal (or CLEC) is prepared as follows: Enzyme crystals are grown by the controlled precipitation of protein out of aqueous solution, or aqueous solution containing organic solvents.

Conditions to be controlled include, for example, the rate of evaporation of solvent, the presence of appropriate co-solutes and buffers, and the pH and temperature. A comprehensive review of the various factors affecting the crystallization of proteins has been published by McPherson (Methods Enzymol. 114: 112 (1985)). In addition, both McPherson and Gilliland (J.

Crystal Growth 90: 51-59 (1988)) have compiled comprehensive lists of all proteins and nucleic acids that have been reported as crystallized, as well as the conditions that lead to their crystallization. A compendium of crystals and crystallization recipes, as well as a repository of coordinates of solved protein and nucleic acid crystal structures, is maintained by the Protein Data Bank (Bernstein et. al. J. Mol. Biol. 112: 535-542 (1977)) at the Brookhaven National Laboratory. Such references can be used to determine the conditions necessary for the crystallization of a given protein or enzyme previously crystallized, as a prelude to the formation of an appropriate CLEC, and can guide the formulation of a crystallization strategy for proteins that have not. Alternatively, an intelligent trial and error search strategy (see eg., Carter, C.W. Jr. and Carter, C.W., J. Biol. Chem. 254: 12219-12223 (1979)) can, in most instances, produce suitable crystallization conditions for most proteins, including, but not limited to, those discussed above, provided that an acceptable level of purity can been achieved for these. The level of purity required can vary widely from protein to protein. In the case of lysozyme, for example, the enzyme has been crystallized directly from its unpurified source, the hen egg-white (Gilliland, G.L., J. Crystal Growth 90: 51-59 (1933)).

For use as CLECS, the large single crystals which are needed for X-ray diffraction analysis are not required, and may, in fact, be undesirable because of diffusion problems related to crystal size. Microcrystalline showers (ie. crystals of order 10* mm in size/cross section) are suitable for CLECs and are often observed, though seldom reported in the X-ray crystallographic literature. Micro-crystals are very useful in the method of this invention to minimize problems with diffusion (see eg., Quiocho, F.A., and Richards, F.M., Biochemistry 5: 4062-4076 (1967)).

In general, crystals are produced by combining the protein to be crystallized with an appropriate aqueous solvent or aqueous solvent containing appropriate precipitating agents, such as salts or organics. The solvent is combined with the protein at a temperature determined experimentally to be appropriate for the induction of crystallization and acceptable for the maintenance of protein stability and activity. The solvent can optionally include co-solutes, such as divalent cations, co-factors or chaotropes, as well as buffer species to control pH. The need for co-solutes and their concentrations are determined experimentally to facilitate crystallization. In an industrial scale process, the controlled precipitation leading to crystallization can best be carried out by the simple combination of protein, precipitant, co-solutes, and optionally buffers in a batch process. Alternative laboratory crystallization methods, such as dialysis, or vapor diffusion can also be adapted. McPherson (Methods Enzymol. 114: 112 (1985)), and Gilliland (J.

Crystal Growth 90: 51-59 (1988)) include a comprehensive list of suitable conditions in their reviews of the crystallization literature.

Occasionally, incompatibility between the cross-linking reagent and the crystallization medium might require exchanging the crystals into a more suitable solvent system.

Many of the proteins for which crystallization conditions have already been described in the literature, have considerable potential as practical enzyme catalysts in industrial and laboratory chemical processes, and are directly subject to formulation as CLECs within the method of this invention. Table 1 is a sampling of enzymes that have already been crystallized. Note that the conditions reported in most of these references have been optimized for the growth of large, diffraction quality crystals, often at great effort. Some degree of adjustment of conditions for the smaller crystals used in making CLECS might be necessary in some cases.

Table 1 Enzyme Microbial or biological source References (including those cited therein) - alcohol dehydrogenase horse liver Eklund er aI.,J.Mol.Biol. 146: 561- 587 (I981) - alcohol oxidase Pichia pa.SIOri$ Boys et aI.,J.MoI.Bio1. 208: 211-212 (1989) Tykarska er a!.. J.Prorein Chem. 9: 83-86 (1990) - aldolase (fructose-bisphosphate) rabbit muscle calf muscle human muscle Eagles el al., J.M01.Biol. 45: 533-544 (1969) Heidmer et al.. Science 171: 677-680 (1971) Goryunov et aI.,Biofizika 14: 1116- 1117 (1969) Millar er a1., Trarzs.R0y.S0c.L0nd.

B293: 209-214 (1981) Drosophila rnelariogaster Brenner et aI.. J..Biol.Chem. 257: 11747-11749 (1982) - aldolase (PKDG) Pseudamonax putida Vandlen et al., J.Bi0t’.Chem. 248: ‘ 2251-2253 (1973) - alkaline phosphatase Escherichia coli Sowadski er a{.. J.MoI.BioI. 150: 245-272 (1981) -asparaginase Erwinia carotova North et al., Nature 224: 594-595 (1969) Escherichia coil’ Epp et £11.. Eur.J.Bi0chem. 20: 432- 437 (1971) Escherichia coli Yonei e1 al.,J.Mol.BioI. 110: 179- Proteus vulgarix (1977) Tetsuya er al.,J.Bio1.Chem. 248: 7620-762] (1972) - carbonic anhydrase human erythrocyte (C) Kannan er a1.. J.Mo1.Bio1. 12: 740- 760(l965) human erythrocyte (B) Kannan er aI..l.Mo1.BioI. 63: 601- 604 (1972) bovine erythrocyte Carlsson et al. }.Mo1.Bio1. 80: 373- 375 (1973) - catalzme horse erythrocyte Glauser ex al.,Ac1a Cryst. 21: 175- 177 (1966) MiCr0COCC1LV[uieuK Marie et aI..J.Mo1.Biol. 129: 675- Penicillium vilale (1979) Vainshtein er a1.,Acta Cryst. A37: C29(1981) bovine liver Eventoff et. aI..J.M01.BioI. 103: 799- 801 (1976) . creatine kjnase bovine heart Gillilancl er aI..J.MoI.Biol. 170: 791- 793 (1983) rabbit muscle McPherson, J.Mol.Biol. 81: 83-86 (1973) - gluiarninase Actenobacter glutanimaszflcans Pseudomonas 7A Wlodawer el aI.,.,’.MOI.Biol. 99: 295- 299(1975) Wlodawer et aI.,J.MoI.BioI. 112: 515-519 (1977) Table 1 continued: Enzyme Microbial or biological References source (including those cited therein) - glucose oxidase Aspergillur niger Kalisz at 014g J.M0l.BioI. 213: 207- 209 (1990) -B-Iactamases Staphylococcusaureus Moult et a1.. Biochem J. 225: 167- 176 (1985) Bacillus cereu: Sutton et a1.. Biochem J. 248: 181- 188 (1987) - lacate dehydrogenase porcine Hacker: er al., J.Mo1.Biol. 78: 665- 673 (1973) chicken Pickles er al.. !.Mo1,Biol. 9: 598-600 (1964) doglish Adams et al.,J.Mol.Biol. 41: 159-188 (1969) Bacillus stearolhermophilus Schar er a1..J.Mol.Bio1. 154: 349-353 (1982) - lipase Gcorric/mm candidum Haza er al.. J.Biochem. 86: 1821-1827 (1979) horse pancreatic Lombardo er al.,.I.M0l.Bi0I. 205: 259- 261 (1989) Mucor meihei Brady el al.J\/azure 343: 767-770 (1990) human pancreatic Winkler er al.. Nature 343: 771-774 (1990) - luciferase Firefly Green,A.A.,et al., Biochem. Biophys.

Ac1a.20: 170 (1956) - luciferase Vibrio harveyii Swanson et al .J.Bz'o1.Chem. 260: 1287-1289 (1985) - nitrile hydratase Brevibacterium R312 Nagasawa er al, Biochem.Biophy:.Re5.

P. chlororaphis B23 Commun. 139: 1305-1312 (1986) Nagasawa e1.al., Eur.J.Bi0c/zem. 162: -698 (1987) - peroxidase horseradish Braithwaite e1al.,J.Mol.Biol. 106: 229-230 (1976) horseradish roots (T ype E4) Aibara er aI..J.8iochem. 90: 489-496 (1981) Japanese radish Morita. Acza Crvrz. A28: S52 (1979) - peroxidase (chloride) Caldaromycesfumago Rubin ez al., J Biol .Chem. 257: 7768- 7769 (1982) v peroxidase (cytochrome) Sarchomyces ccrevzxae Poulos ez al.,1.Biol.C-‘rem. 253: 3730-3735 (1978) - peroxidase (glutaihionc) bovine eryihrocyie Ladenstein er al.J.Mol.BioI. 104: 877-882 (1979) - subtilisin Bacillus subzilis (Novo) Drgenlh at al., J.Mol.Biol. 28: 543-544 (1 67) Bacillus amyloliquefaciens Wright e1al.. Nature 221: 235-242 (BPN') (1969) Bacillus subtiIir(Car1sberg) Petsko ez aI..J.Mol.Bi0I. 106: 453- (1976) Table 1 continued: Arthrobacler B3728 Srrepzomyces olivochromogenes Sireptomyces violaceoniger Aclinoplanes missouriensis Enzyme Microbial or biological References source (including those cited therein) - superoxide dismutase bovine Richardson er al.. J.BioI.Chem. 247: 6368-6369 (1972) spinach Morita et aI., J.Mo1.Bio1. 86: 685~686 (1974) Saccharomyces cerevisiae, Beam et al.,J.Mol.BioI. 105: 327-332 Escherichia coli (1976) Baciilus szearolhermophillus Bridgen e1 01...! .Mo(.Biol. 105: 333- 335 (1976) Pseudomona: ovalis Yamakura ex 01., !.Bio1.C hem. 251: 4792-4793 (1976) - Lhermolysin Bacillus zhermoprareolyricus Matthews er al., Nature New Bio!. 238: 37-41 (1972) ~ urease jack bean Sumner,J.B., J.BioI. Chem. 69: 435 (1926) - xylose isomerase Sirepzomyces rubiginosus Carrel) el al., J.Bi0l.Chem. 259: 3230-3236 (1984) Akins et aI., Bioc/iym.Biophy5 Acid 874: 375-377 (1986) Father et al., Protein Engineering 1: 459-466 (1987) Glasfeld at a!.,J.Bio1.Chemi 263: 14612-146130988) Rey er al.. Fr0tetn5:SIruC.Func.Genet. 4: 165-172 (1988) Ereparatiog of CLECs - cross-linkigg reaction Once crystals are grown in a suitable medium, they can be cross-linked. Cross-linking results in stabilization of the crystal lattice by introducing covalent links between the constituent enzyme molecules in the crystal. This makes possible the transfer of enzyme into an alternate reaction environment that might otherwise be incompatible with the existence of the crystal lattice, or even with the existence of intact undenatured protein. Cross-linking can be achieved by a wide variety of bifunctional reagents, although in practice, simple, inexpensive glutaraldehyde has become the reagent of choice. (For a representative listing of other available cross-linking reagents, one can consult, for example, the 1990 catalog of the Pierce Chemical Company).

Cross-linking with glutaraldehyde forms strong covalent bonds between primarily lysine amino acid residues within and between the enzyme molecules in the crystal lattice that constitute the crystal. The cross-linking interactions prevent the constituent enzyme molecules in the crystal from going back into solution, effectively insolubilizing or immobilizing the enzyme molecules into microcrystalline (ideally 10” mm) particles. The macroscopic, immobilized, insolubilized crystals can then be readily separated from the feedstock containing product and unreacted substrate by simple procedures such as filtration, decantation, and others. They can also be used in CLEC packed columns in continuous flow processes, where they exhibit enhanced cofactor and metal ion retention properties.

By the method CLECS are obtained for use as enzyme catalysts in existing and novel environments. The enhanced stability of the cLEcs, which results from the cross-linking reaction, makes it possible to transfer the CLEC into a solvent (e.g., aqueous, organic or near-anhydrous solvents, or a mixture of these), in which it would otherwise be incompatible, and to carry out chemical reactor operation at elevated temperatures of extremes of pH.

The macroscopic CLEC catalyst particles can also be readily manipulated, allowing recovery from feedstock by simple methods, such as filtration, centrifugation, or decantation of solvent. In addition, these can be used in packed columns in continuous flow processes.

Egepagation of CLECS - lyophilization A suspension of one volume of cross-linked thermolysin crystals in ten volumes of demineralized water at pH 7.0 was lyophilized overnight using a VirTis Model #24 lyophilizer. Lyophilized crystals were stored at room temperature or at 4°C prior to reconstitution, which was accomplished by adding ten volumes of the solvent of choice directly onto crystals taken from storage. Re-hydrated crystals were reconstituted in 1omM calcium acetate buffer at pH 7.0 for the FAGLA cleavage experiments. Reconstituted lyophilized CLECs were routinely stored at room temperature. In contrast, soluble enzyme required -31.. storage at -70°C to maintain specific activity longer than a week. This protocol was used for all the enzymes described in the exemplification included here.

Synthesis Qf gspartame precursor with thggmolysin CLEC The method of the present invention, by which cross-linked crystal enzymes are produced, is described below and exemplified by the production of cross-linked immobilized enzyme crystals of thermolysin for use in the production of the dipeptidyl precursor of aspartame, in ethyl acetate, which is a near-anhydrous organic, solvent. Thermolysin, a protein which has been crystallized and whose structure has been solved at l.6A resolution (Holmes and Matthews, J. Mol. Biol. 160: 623-639 (1982)), is one example of an enzyme which can be used as a CLEC in the present method.

Thermolysin is used in the manufacture of the artificial sweetener aspartame (Isowa et. al. U.S.

Patent #4,436,92S (1984); Lindeberg, J. Chem. Ed. 64: 1062-1064 (1987); Nakanishi et. al., Biotechnology 3: 459-464 (1985); Oyama, et. al., Methods in Enzymol. 136: 503-516 (1987)). At the present time, most aspartame appears to be produced by a conventional synthetic chemistry approach, although use of conventionally immobilized thermolysin in near-anhydrous media has produced encouraging results (Oyama et. al., J. Org. Chem. 46: 5242-5244 (1981); Nakanishi et. al., Biotechnology 3: 459-464 (1985)).

Improvement in the enzymatic approach to aspartame production, such as is possible through use of the present method, would make it competitive with the presently-used method, both in terms of convenience and cost (Oyama, et. a1., uethods in gnzymol. ;3 : 503-516 (1937)). ssessmen e in C The method has also been used to produce thermolysin CLECs which have been assessed as to their pH dependence and stability, stability at elevated temperature, resistance to exogenous proteolysis and stability in the presence of an organic solvent. Thermolysin CLECs were compared to soluble thermolysin, as described in detail in Example 2 and Figures 1-4. Results of the assessment showed the following: . As to pH dependence and stability, both forms demonstrate maximum activity at pH 7 and demonstrate similar activity in the acidic range. In the alkaline pH range, the CLEC maintains maximum activity to pH 10; the soluble thermolysin has 75% activity at pH .5, only 25% activity at pH 9 and is completely inactive at pH 9.5.

. The additional stabilization achieved in CLECs results in enzymatic activity at higher temperatures than is possible with soluble thermolysin. Enhanced stability of CLBC thermolysin at lower temperatures makes storage simpler than it is for the soluble enzyme. Thermal stability and resistance to autolysis was also demonstrated for thermolysin CLECs, which retained maximum activity after five days of incubation at 65°C. In contrast, soluble thermolysin lost % of its initial activity after two hours incubation and demonstrated negligible activity after 24 hours incubation at 65°C.

. Enzymatic activity of thermolysin CLECs was unaffected by four days‘ incubation in the presence of the powerful streptococcal protease, Pronase®. In contrast, soluble thermolysin was rapidly degraded and lost all activity after 90 minutes incubation.

. Thermolysin CLECs and soluble thermolysin exhibited markedly different stability in the presence of organic solvents, as shown in Table 12. The thermolysin CLECs retained greater than 95% maximum activity following incubation with all organic solvents assessed.

These features of thermolysin CLECS and other enzyme CLECs make them particularly useful, since they are easier to store, more stable and less easily inactivated or degraded than corresponding soluble enzymes.

Assessment of Elastase CLECs The method has also been used to produce elastase CLECS which have been assessed as to their activity and resistance to exogenous proteolysis. Elastase CLECs were compared to soluble elastase, as described in detail in Example 3 and Figures 5 and 6. Results of the assessment demonstrated the following: . Elastase CLECs retain approximately 50% activity compared to soluble enzyme. . soluble elastase was rapidly degraded by protease. Activity of soluble elastase was reduced to 50% of initial activity following ten minutes incubation in the presence of protease. After one hour incubation the soluble enzyme had lost more than 90% of its initial activity. In contrast the enzymatic activity of the elastase CLEC was unaffected by incubation with protease.

Assessment of Egterase CLECs The method has also been used to produce esterase CLECs which have been assessed as to their activity and resistance to exogenous proteolysis. Esterase CLECs were compared to soluble esterase, as described in detail in Example 4 and Figures 7 and 8. Results of the assessment demonstrated the following: . Esterase CLECS retain approximately 50% activity compared to soluble enzyme. . soluble esterase was highly susceptible to proteolytic degradation. Activity of soluble esterase was reduced to 50% of initial activity following ten minutes incubation in the presence of protease. After one hour incubation the soluble enzyme had lost more than 90% of its initial activity. In contrast the enzymatic activity of the esterase CLEC was unaffected by incubation with protease. ssess t of i a e CL Cs The method has also been used to produce lipase CLECs which have been assessed as to their activity. Lipase cLECs were compared to soluble lipase, as described in detail in Example 5 and Figure 9. Results of the assessment demonstrated that lipase CLECs retain approximately 90% activity compared to soluble enzyme.

Assessment Qf Lysozyme CLECs The method has also been used to produce lysozyme CLECs which have been assessed as to their activity and resistance to exogenous proteolysis. Lysozyme CLECs were compared to soluble lysozyme as described in detail in Example 6 and Figure . Results of the assessment demonstrated that lysozyme CLECs retain approximately 50% activity compared to soluble enzyme.

Assessment of Asparaginase CL£Cs The method has also been used to produce asparaginase CLBCs which have been assessed as to their activity. Asparaginase CLECs were compared to soluble asparaginase, as described in detail in Example 7 and Figure 11. Results of the assessment demonstrated the following asparaginase CLECs retain approximately 77% activity compared to soluble enzyme.

Genera ‘c bi ’ of C Ecs As disclosed here, CLECs represent a novel technology with broad use in many areas, including, but not limited to, industrial scale syntheses, laboratory tools, biosensors, and medical applications. Examples of various systems using conventionally immobilized enzyme methods in their execution are given in Tables 2-5 below. One skilled in the art should be able to adapt these, and similar systems, to the CLBC technology disclosed in this application. To illustrate this, specific examples are discussed in more detail from each of the categories listed.

Table 2 below lists examples which use conventionally immobilized enzymes in an industrial process, which examples can be readily adapted to the CLEC technology disclosed here. -37.. glycine Table 2 Enzyme Production or Substrates References application (including those cited) - thermolysin 0 aspartame precursor Z-Asp, L-Phe-OMe Oyama et al.. J. Org_ Chem 46: 5242-5244 (1931) Nakanishi et aI., Trends in Biotechnology 3: 459-464 (1985) - subulisin - aspaname L-Asp-L-Phe, OMe Davino,A.A., US Patent 4293648 (1981) - lipase - cocoa fat palm oils Harwood, J .. Trends in substitutes Biochemical Sciences 14: 125- (I989) Macrae, A.R., J. Am. Oil Chem Soc. 60: 291-294 (1983) - niu'ile hydratase, - acrylamide acrylonitrile Nagasawa, T. and Yarnada, I-L, nitrilases, amidase Trends in Biotechnology 7: 153-158 (1989) - amino acylase (fungal) - amino acid N-acyl-D,L amino acids Schmidt-Kasuier,G. & Egerer, - amino acid esterase, resolution esters of D,L amino P. in Biotechnology vol 63: - subtilisin acids 387~421 (1984) and references - amidases amides of DJ. amino therein. - hydamionases acids - specific dehydrogenases hydanloins Fusee,M.C.. Methods in - amino peptjdase a-hydroxycarboxylic Enzymology 136: 463 (1987) - transaminase acids Fusee,M.C., Methods in Enzymology 136: 463 (1987) - amino acid - amino acid keto or hydroxy acids dehydrogenase + fomiate production: genera] Rozzell,J .D.. Methods in dehydrogenase Enzymology 136: 479 (1987) - amino acid Enzymes in Industry; Ed production: specific Gerhartz.W.,VCH Press 1990 - L-aspanase L aspartic acid fumamte/fumaric acid - L-aspartate 4- L-alanine L-aspanic acid. decarboxylase ammonium fumarate aspanase + L aspartate 4 decar-boxylase I.-lysine D,L-a amino c- - ACL hydrolasc caprolactam (ACL) - L-ACT hydrolasc 1.-cysteine DL-2amino2 thiazoline carbox ylic acid L-isoleucine L-methionine v lyase L-phenylalanine cinnamatc ~ L-tryptophan synthetase L-lryplophan indole. L-serine L-valine - fumarase L-malic acid fumarate - hydamoinase D n carbamoyl p- Sp-hydroxy hydantoin hydroxy- phenyl Table 2 continued: Enzyme Production or Substrates References application (including those cited) - lipases. esterases - resolution of synthetic chemistry Jones, J .B., Tetrahedron 42: racemates by 3351-3403 (1988) stercoselective Butt. S. and Roberts, S.M., synthesis Natural Produc1Reports 489- (1986). and references cited therein for a more comprehensive review of this area - fumarase - L—malic acid fumaric acid Chibata et aI.. Methods in Enzymology 136: 455 (1987) - lactase, B-galactosidases - disaccharide lactose & N—acety1 Larsson er al., Methods in synthesis eg galactosamine Enzymo/ogy 136: 230 (1987) galactosyl-N acetyl galactosamine - lipase, esterase - L-menthol 4 isomer mix Fukui,S., Tanaka,A.. Methods in Enzymology 136: 293 (I987) - amidascs - D-valine DJ. amino acid amide Schmidt-Kastner,G. & Egerer, (intennediate for P. in Biotechnology vol 62: pyrethroid 387-421 (1984) and references insecticide fluvinate) therein. lipase (Candida 0 R(+)2 phenoxy~ 2 chloro propionic acids Biocatalysts in Organic cylindricea) propionic acids Syntheses eds Tramper. van de (herbicides) Plas & Linko; Proceedings of International Symposium in Netherlands 1985 - lipases, esterases, - organic syntheses Jones, J.B., Tetrahedron 42: arnidases, aldolases rnonoglycerides 3351-3403 (1988) peptides Butt. S. and Roberts. S.M., - proteases, peptidases - yeast lipase - 2(p-chlorophenoxy) resolution of racemic Natural ProduclRep0r15 489- 503 (1986), and references cited therein for a more propionic acid: ester comprehensive review of this herbicide area - strictodine synthetase - alkaloid production Pfitzner et aI., Method: in eg strictosidine Enzvmology 136: 342 (1987) - penicillin acylase - 6-amino penicillin G or V En: Eng 6: 291 (1982) - penicillin amidase penicillinanic acid Enz Eng 8: 155 and 7-ADCA o hydroxysteroid ~ steroid Carrea er a1. Methods in dehydrogenases transfonnations Enzymology 136: 150 (1987) - Sphosphodiesterase, - 5'-ribonucleotides Keller et al. Methods in nucleases Ertzymology 136: 517 (1987) - esteme - B-lactarn precursor corresponding diesters Japanese patent application: (chiral mono esters 82-159, 493 (1931) Biseibutsu eg Bamino glutaric Company acid monoalkyl ester) - lipases ‘ B-blockers Kloosterman,M er aI.,Trends in Biotechnology 6: 251-256 (1988) Production of acrvlamide using CLEC technology The following is a description of one use of the method: the adaptation of acrylamide production from immobilized cells which overproduce nitrile hydratase enzyme (Nagasawa, T. and Yamada, H., Trends in Biotechnology 7: 153-158 (1989)) to the CLEC technology previously disclosed herein.

Industrial scale production of acrylamide, an important commodity chemical, has been described by Yamada and collaborators (Nagasawa, T. and Yamada, H., Trends in Biotechnology 7: 153-158 (1989)). Kilotons of acrylamide per year are produced in chemical reactors loaded with entrapped cells selected as overproducers of the enzyme nitrile hydratase. Nitrile hydratase has also been reported as purified and crystallized from two sources, Brevibaggerium R312 (Nagasawa et. al., Biochem. Biophys. Res. Commun. 139: 1305-1312 (1986) and B. chlororaphis B23 (Nagasawa et. al., Eur. J. Biochem. 162: 691-698 (1987). As disclosed here, these crystalline enzymes can each be immobilized by crosslinking with glutaraldehyde or other suitable crosslinking reagent to produce a CLEC.

The nitrile hydratase CLECs can then be used in a conventional reactor, in place of the entrapped cells currently used. Adapting this process to CLEC technology leads to immediate advantages. These include; reduced plant size and improved throughput resulting from the enhanced activity per unit volume implicit in the higher enzyme concentration in the CLECS, and improved substrate and product diffusion rates; reduction in undesired contamination and side reactions, resulting from the higher purity of CLECs; and reduced sensitivity to microbial contamination in the absence of cells. In addition, there are other benefits available only to a CLEC-based method. These benefits include: higher temperature operation to improve reaction rates; the ability to operate in aqueous, organic and near-anhydrous solvents, allowing optimization of the acrylamide production reaction; and enhanced half-life in operation and storage, resulting from the higher chemical and mechanical stability of CLECs, particularly in unconventional solvents. -40a- Medical Applications of CLEC technology — extracorporeal treatment The present invention provides for . An enzyme crystal crosslinked with a multifunctional crosslinking agent, said crosslinked enzyme crystal having resistance to exogenous Pronasem proteolysis and being selected from the group consisting of: b) G) pyridoxal phosphate enzymes, metalloenzymes, CoA-requiring enzymes and sulfurases (PAPS); thermolysin, nitrilase, arninocyclase and hydantoinase; elastase, aminopeptidase, asparaginase, glutaminase, arginase, urease, alkaline phosphatase, lipase, preferably phospholipase, B-lactamase, ACL hydrolase, L-ACT hydrolase, arylsulfatase, glycosidase, preferably B—galactosidase and B-fructosidase,‘ amidase, protease, peptidase, lactase, amino acid‘ esterase, ‘-phosphodiestcrase, nuclease and creatine deiminase; dehydrogenase, amino acid dehydrogenase, formate dehydrogenase, lactate dehydrogenase, glutamate dehydrogenase, hydroxysteroid dehydrogenase, succinate dehydrogenase, glucose glucosepyruvate dehydrogenase, dehydrogenase, alcohol dehydrogenase, catalase, superoxide dismutase, peroxidase, tyrosinase, isomerase, flavoenzyme, nitrate/nitrite reductase, oxidoreductase (NAD/P), oxidase and lyase; or kinase (ATP), preferably crcatine kinase, transaminase, transferase, glycosyl transferase and methyl transferase (SAM) for use in a medicament.

. The enzyme crystal according to 3. The enzyme crystal according to . The enzyme crystal according to , wherein said enzyme is thermolysin. , wherein said enzyme is elastase. , wherein said enzyme is asparaginase. —40b- The enzyme crystal according to 1, wherein said enzyme is lipase.

The enzyme crystal according to 1, wherein said oxidase is alcohol oxidase, glucose oxidase, bilirubin oxidase, lactate oxidase, cholesterol oxidase or L-amino oxidase. and/or said lyase is aldolase, L-aspartate 4-decarboxylase, phenylalanine ammonium lyase, carbonic anhydrase, nitrile hydratase, L-aspartase, fumarase or heparinase.

The enzyme crystal according to 1, selected from the group consisting of: L- tryptophan synthetase and strictodine synthetase.

The enzyme crystal according to 1, wherein said enzyme is as defined in c) of claim 1 and wherein said crosslinked enzyme crystal retains at least 91% of its initial activity after incubation for three hours in the presence of a concentration of Pronasem that causes the soluble uncrosslinked form of the enzyme that is crystallized to form said enzyme crystal that is crosslinked to retain at most 6% of its initial activity under the same conditions, and said enzyme being esterase, and The enzyme crystal according to any one of 1 to 8 , wherein the crosslinker is glutaralclehyde. - 40c — A CLEC or a set of CLEcs can be used, for example, to remove a component of a fluid, such as blood, generally by altering the component, and thus, converting it to a substance not detrimental to an individual or which can be removed by normal body processes (e.g., via detoxification, or degradation in the liver, excretion via the kidneys).

In this application, an appropriately-selected CLEC or set of CLECs is brought into contact with body fluid, which contains the component to be altered, or a reactant (product or substrate) of a reaction in which the component participates, upon which the enzyme in __* —-«-w»~-~~ » the CLEC acts. As a result, the enzyme is able to act upon the component to be altered or with another substance which is a product of a reaction in which the component to be altered participates. The activity of the enzyme results in direct alteration of the component to be removed or in alteration of the product of the reaction in which the component participates (thus making continuation of the reaction impossible).

This can be carried out through the use of an extracorporeal device which includes an appropriately—selected CLBC or set of CLECs and a retaining means which is made of a material; such as a porous material on which a CLEC is retained or a tube in which a CLEC is present, which allows contact between the component itself or the substance in the fluid which is a product of a reaction in which the component to be altered participates.

This might also be achieved by the insertion of an appropriate CLBC into a suitable body compartment, such as the peritoneum or a lymph node, where the CLEC would have access to bodily fluids. This insertion might be done surgically, or by injection of the CLBC slurry.

Direct injection of CLEC into the blood stream would not be appropriate, given the high risk of embolism.

The use of appropriate CLECs in this area might serve as an alternative to genetic methods in enzyme replacement therapy to correct a natural deficiency, such as, e.g. phenylketonuria.

Table 3 illustrates some of the medical applications in which CLECs could be used. For the majority of these cases, the extra-corporeal treatment is still in the research phase, but the benefits that CLECs offer could provide novel treatments in areas in which there was previously no alternative treatment.

W... _..._.._._..._.......__.___._.......s._.._........./. .

TABLE 3 Entylne Removal of:- Disease/patients treated References em ployed - asparaginase - asparagine - leukemia Klein.M.. Langer,R.. Trend: in (Removal of asparagine.an Biotechnology 4: l'79~l85 important cancer nuuient, harms (1986) and references therein leukemic cells which cannot manufacture the essential amino Chang,T.M.S.. Methods in acid - asparagine; normal cells Enzyrnology 137: 444-457 can manufacture asparaginc and so (1987) & references therein are unaffected by this treatment) - heparinase ~ heparin - deheparinization for Langer.R.,et al.. Science 217: hemoperfusion patients eg kidney 261-263 (1982) dialysis ~ bilirubin oxidase - bilirubin - neonatal jaundice Lavin,A.,et a!., Science 230: -545 (1985) - carboxypeptidase - methotrexate ' chemotherapy patients Pitt,A.M.,et al., Appl.

Bt'ochem.Bt'o:echnoI.8: 55-68 (1983) - tyrosinase - aromatic amino - liver failure exhibiting Chang.'l‘.M.S., Sem.Liver acids pathological elevations of amino Dis.$er. 6: 148 (1986) acids - phenylalanine - phenylalanine - phenylketonuria and liver failure Ambrus.C.M., et aI., ammonium lyase J .Pharm.& Exp.Ther. 224: -602 (I983) - multi-enzyme ° urea (convened - detoxification for chronic renal Chang.T.M.S., Methods in system including: into glutamic and failure patients Enzymology 137: 444-457 urease. glutamate other amino acids, (1987) & references therein dehydro-genase, via ammonia) glucose, Chang,T.M.S.. Enzyme Eng dehydrogenase & a S: 225 (1980) transaminase - arginase - arginine - familial hyperargininaemia KanalasJJ. er al.. Biochem.

Med. 27: 46-55 (1982) - glutamate s ammonia - liver failure Maugh,T.l-1., Science 223: dehydrogenase & 474-476 (1984) ammonia A particular application of the present method is the heparin lyase system for blood deheparinization (Bernstein et al., Methods in Enzymology 137: 515-529 (1987)), which is discussed below.

All extracorporeal devices perfused with blood, such as kidney dialysis, continuous arteriovenous hemofiltration or extracorporeal membrane oxygenators, require heparinization of the patient to avoid blood clotting. However, heparinization of the patient leads to hemorrhagic complications and remains a threat to human safety. These problems increase as the perfusion times increase, for example with the membrane oxygenator, and can lead to serious bleeding. After extracorporeal therapy, heparin may be removed from blood by employing a heparinase device at the effluent of the extracorporeal device which eliminates all heparin from the blood returned to the patient and thus avoids the current problems of heparinization.

Published research (Langer et al. Science 217: 261-263 (1982)); Bernstein et al., Methods in Bnzymology 137: 515-529 (1987)), details the problems presented by conventionally immobilized enzymes used in extracorporeal devices. The principal problem is that conventional immobilization results in a low retention of enzyme activity per unit volume, thus requiring a large volume of immobilized enzyme to perform the necessary heparinization. This volume is too large to be of practical use in humans. However, the high retention of activity per unit volume in CLEC, due to the lack of inert support circumvents this problem and __________,___,_______,____.i___.__._..____.i.__.,_._ .-___ offers a practical solution to deheparinization of humans. The enhanced stability of CLEC will reduce the disassociation of enzyme from the crosslinked crystal.

This is superior to the less stable, conventionally immobilized enzymes, because immune responses resulting from enzyme leakage will be reduced. CLEC temperature stability prevents denaturation of the enzyme due to high transient temperatures during storage; it is likely that CLEC may retain high activity, even when stored at room temperature. In addition, CLEC will be cheaper and more convenient to use than their conventionally immobilized counterparts because of their longer operational and storage lifetimes.

Additional applications of CLEC technology: biosensors A CLEC or a set of CLECs can be used as a component of a sensor, referred to as a biosensor, useful for detecting and/or quantitating an analyte of interest in a fluid, such as body fluid (e.g., blood, urine), chemical and laboratory reaction media, organic media, water, culture medium and beverages. In some instances, the fluid in question can be a gas, as in an alcohol breath analyzer (Barzana, E., Klibanov, A., and Karell, M., NASA Tech Briefs l3:104 (1989)). In this application an appropriately-selected CLEC or set of cLEcs is brought into contact with a fluid to be analyzed for the analyte of interest. The analyte of interest can be measured directly (e.g., blood glucose level) or indirectly (e.g., by detecting or quantitating a substance which is a reactant (product or substrate) in a reaction in which the analyte of interest participates). In either case, the CLEC is able to act upon the analyte or the substance which is a reactant in a reaction in which the analyte also participates. The activity of the enzyme results in a detectable change (e.g., change in pH, production of light, heat, change in electrical potential) which is detected and/or quantitated by an appropriate detecting means (e.g., pH electrode, light or heat sensing device, means for measuring electrical change) (Janata, J., et al., Anal. Chem. 62: 33R-44R (1990)). Any means useful for detecting the change resulting from the enzyme—catalyzed method can be used. A biosensor of the present invention includes a CLEC or set of CLECs and a retaining means for the CLEC which allows contact between the CLBC(s) and the analyte of interest or the substance in the fluid which is a reactant in the reaction in which the analyte of interest participates.

Table 4 illustrates some of the biosensor applications in which CLECs could be used. Currently immobilized enzymes are used in these applications, but suffer from low stability, low enzyme density, short lifetimes and lack of reproducibility. These examples can be readily adapted to the CLEC technology disclosed here. __-.e -._.. .. .

Table 4 Detection oi‘:- ~ glucose l Enzyme Employed - glucose oxidase Application vdiabeucs Reference J ~ Daniles,B.,Mossbach,K..

Methods in Enzymology 137: 4-7 (1987) - Hall,E "Biosensors " Open University Press (1990) ~ 'l'aylor,R., Proceed.

Biotechnology Conference 1989; 275-287 - Anthony ez al., "Biosensorr, Fundamentals and Applications". Oxford Universityfless (1987) l - creatinine deiminase - creatinine °urease ‘Urea - kidney function - Tabata,M. et aI., Anal.

Biochem.134:44 (1983) - kidney function ° Hsuie,G.l-I et al., PoIym.Mater.Sci.Eng. 57: 825-829 (1987) ° Kobos.e1 al. AnaI.Chcm. 60: 1996-1998 (1988) - lactate oxidase & - lactate dehydnogenzse ‘->g1uc0se—6-pyruvate dehydmgenase sucrose and ATP - Glucosephosphate. - clinical applications - Blaedel, W,J. & .lenkins,R.

A., Anal. Chem. 48(8): 1240 (1976) - Sagaguchi. Y.,et al., J.

Appl. Biochem 3: 32 0981) - diabetics and other lflcdical ~ ibid as glucose oxidase - alcohol dehydrogenase, - ethanol & other alcohol oxidase acids alcohols; acetic, formic - breathalysers and industrial applications - Romette,J.1..et aI..

Methods in Enzymology 137: 217-225 (1987) ~ Ho,M.1-1.. Methods in Enzymology 137: 271-288 (1987) - B-fructosidase ~ sucrose - industrial applications - Romette.J_L.et aI., Methods in Enzymology 137: 217-225 (1987) - cholesterol oxidasc - cholesterol - cholesterol testing - Satoh,l.Method5 in Enzymo1ogyl37: 217-225 (1987) - catalase - uric acid, cholesterol - atherosclerotic and - Sa1oh,1.Methods in - carboxy peptidase \ - methotrexate other medical Enzymology 137: 217-225 (1987) - cancer - ibid as glucose oxidase - B-lactamase penicillinase ~ alkaline hoshatase ° ho hale v carbonic anhydrase 0 carbon dioxide - industrial. laboratory & ~ ibid as glucose oxidase environmental _ applications - L-amino acid oxidase - amino acids - medical and industrial - ibid as glucose oxidase - penicillin - medical 0 Anzai et al., Bull.Chem. - metabolite monitoring SocJprt. 60: 4133-4137 (1988) - ibid as glucose oxidase ' nitrate/nitrite reductase - nitrates & nitrites v metabolite and food monitoring - ibid as glucose oxidase Table 4 continued: photon release Enzyme Emploved Detection of:- Application Reference J - arylsulfatase - sulfate ~ metabolite monitoring - ibid as glucose oxidase - succinate - succinate - industrial - ibid as glucose oxidase dehydrogenase - bacterial luciferase FMNH2 and coupled . dclwjon of1()'18 v WannlundJ .,er al., reactions molar quantities of "Luminescen1 3SS3)'53 FMNH2 through Perspectives in measmemem of photon endocrinology and clinical mime chemistry"; Eds Serio,M. and Pazzagli,M. 1: 125 (1982) - Kurkijarvi er al., Methods in Enzymology 137: 171- 18] (1987) - ftrefly luciferase ATP and coupled . daecuon of 10-12 o Kurkijarvi et a1.. Methods reactions molar quantities of ATP in Enzymology 137: )7]- tluough measurement of 131 (1937) - Murachi er al., Methods in Enzymology 137: 260-271 (1988) -49..

In the method as it may be carried out for the analysis of samples in a biosensor, it is particularly desirable to produce the largest possible detectable signal from the smallest possible quantity of substrate and catalyst. In this regard, the CLEC technology disclosed here is particularly attractive, since it achieves the highest possible concentration of enzyme catalyst in a given volume.

Often, considerable efforts are made to couple an ultimate enzymatic reaction of interest, either directly, or through suitable intermediates, to the production of light by enzymes like luciferase (Kurkijarvi et al., Methods in Enzymol. 137: 171~181 (1988)). This is done in order to take advantage of the unparalleled sensitivity and efficiency of photon detection equipment, which allows for the detection of femtomolar concentrations of enzyme reaction products under appropriate conditions. Following this principle, biosensor systems have been designed, using conventionally immobilized enzymes, to detect various substrates of clinical and other interest. Light producing reactions have been coupled to assay reactions detecting substrates like, D-glucose, L-lactate, L-glutamate and ethanol, among others, at extremely low concentration. with regard to this application, the luciferase enzyme from yibrig harveyii has been reported as crystallized (Swanson et al., J. Biol. chem. 260: 1287-1289 (1985)). Crystals of this luciferase can be crosslinked with glutaraldehyde or other suitable reagent to form a CLEC of luciferase. For biosensor and analytical uses, a CLEC of luciferase offers many advantages over conventionally immobilized enzyme. In a CLEC, the entire volume of the luciferase CLEC would consist of light emitting enzyme. In a conventionally immobilized enzyme system, however, as much as 95% of the total volume is taken up by "inert" carrier material, which more likely functions as an absorber of the light emitted by enzyme. In addition, the enhanced stability of CLECs should facilitate storage at room temperature, and also makes possible novel sensing applications in harsh environments and elevated temperatures.

Additional applications of CLEC technoloqv - laboratory reactions CLECs may be used as laboratory reagents in small columns or in batch processes, which can be used to carry out laboratory reactions. Some of the broad categories of reactions are in Table 5. -51..

Table 5 Enzyme Employed Reaction type catalvzed Reference - lipases, phospholipases - Stereoselective synthesis: including - Zaks,A.& K1ibanov,A.M. esterification, transesterification, Proc.Nat.Acad.Sci.U$A. 82: 3192- aminolysis, lactonizations. 3196 (1985) polycondensations, acylation, o KIibanov.A.M. Acc.Chem.Res. 23: oxirnolysis and resolution of racemic 114-120 (1990) and references therein mixtures - Wong.C.l-I.,ChemIracts-Organic Chemistry 3:91-ll] (1990) and references therein ~ esterases - stereoseleetive synthesis and - Kobayashi et al.. Tetrahedron Letters resolution Vol 25, #24: 2557-2560 (1984) - Schneider et al.. Agnew. Chem. Int.

Ed.Eng1.23 (#1): 64-68 (1984) - tyrosinase ~ oxidation of phenols to produce - l(azandjian.R.Z and Klibanov, A.M. quinones J.Am.ChemiS0c 110: 584-539 (1986) - proteases, eg subtilisin - stereoselective acylation of ~ Riva et al. J.Am.Chem.Soc.1l0: S84- carbohydrates S89 (1988) - oxidztses - selective oxidation of hydrocarbons - Klibanov,A.M. Acc.Chem.Re.r. 23: -120 (1990) and references therein - other enzymes not - Stereoselective synthesis: - Wong,C.l-l..Chemrracts-Organic requiring co-factors: Chemistry 3:91-111 (1990) and isomerases, lyases. references therein aldolases, glycosyl transferases. glvcosidases - other enzymes not - Stereoselective synthesis: - Wong.C.I-l.,Chenurac1s-Organic requiring added co-factors: Chemistry 3:91 -1 11 (1990) and flavoenzymes, pyridoxal references therein phosphate enzymes. metallocnzymes - enzymes requiring co~ - Stereoselective synthesis: - Wong,C.H.,Chenurac1s-Organic factors: kinases (ATP), Chemistry 3:91-111 (1990) and oxidoreductatses (NAD/P), references therein methyl transfer-ases (SAM), COA-requuing enzymes, sulfurases Schneider et al. (Agnew. Chem. Int. Ed. Engl. 23 (No.1): 64-68 (1984)) illustrates how enzymes may be used in organic syntheses. Pig liver esterase was employed in the meso-ester transformation into a chiral mono-ester in an aqueous phosphate buffer.

The advantages of CLEC catalyzed reactions for laboratory use are threefold. First, CLEC retain high activity in harsh environments (eg. aqueous, organic, near-anhydrous solvents and mixtures of these, and at high temperatures) that are typical of laboratory chemical synthesis experiments. Second, CLEC exhibit high operational and storage stability, which is appropriate for intermittent laboratory experiments.

Third, their high activity per unit volume will allow shorter reaction times and require smaller volumes of enzyme (per unit of activity). Thus, the advantages that CLECs offer over free or immobilized enzymes, provide organic chemists with an alternative, highly selective, synthetic tool.

In all of these instances described above, but not limited to these, the method described can be adapted by one of ordinary skill in the art, to convert a process using a conventionally immobilized enzyme catalyst to the use of a CLEC of the appropriate enzyme. CLECs can not only replace conventional immobilized enzymes, but also be used in cell mediated transformations. The present invention will now be illustrated by the following Examples.

-S3- EXAMPLE ; Crystallization and crosslinkinq of thermolysin for synthesis of the aspartame precursor. z-Asp—Ehe-OMe. sta ' ‘o 250 mg of thermolysin from fiagillus thermopgoteolyticus was purchased from Boehringer-Mannheim GmbH, and dissolved in 4 ml of 45% dimethyl sulfoxide (DMSO) and 55% 1.40 M calcium acetate, 0.50 M sodium cacodylate at pH 6.5. These starting conditions are similar to those described by Matthews et. al. for the production of diffraction quality thermolysin crystals (see, eg., Holmes and Matthews, J. Mol. Biol. 160: 623-639 (1982)). The protein solution was then concentrated to 1 ml in a Centricon 10 micro-concentrator. A good yield of microcrystals was obtained by a process of flash crystallization, now disclosed here, in which 1 ml of water, or 1.40 M calcium acetate, 0.50 M sodium cacodylate at pH 6.5, was rapidly injected into either of the thermolysin-DMSO solutions described above. A shower of hexagonal micro-crystals of approximately uniform dimensions (approx. 10“ mm in length) results from this process. crossligking of thgrmolysin migrocrystals.

The protocol used in this specific example of the method of this invention is an adaptation of that described by Nakanishi et. al. (Biotechnology 3: 459-464 (1985)), in which protocol, thermolysin was first adsorbed onto a carrier bead composed of the ion-exchange resin Amberlite XAD-7, and subsequently immobilized by cross-linking with glutaraldehyde (Quicho and Richards, Proc. Natl. Acad. Sci. (USA) 522833-839 (1964)). In this exemplification, the microcrystals of thermolysin obtained above were centrifuged and pelleted, and the supernatant was discarded. 5 ml of 17.5% technical grade glutaraldehyde, in 2.5 % DMSO, 0.05M calcium acetate, and 0.02SM sodium cacodylate at pH 6.5, were then added to the microcrystals. The mixture was incubated with gentle agitation at 37°C for 4 hours. The crosslinking reaction was stopped by repeated washing of the crystals with 10 ml aliquots of water to remove the glutaraldehyde solution. The washed cross-linked thermolysin crystals constitute the thermolysin CLEC used below as a catalyst. gynthesis gf z-Asp-Ehe-OMe in an agueous solution ml of a thermolysin CLEC suspension were added to a continuous stirred-batch reactor incubated at 37°C. After centrifugation and decantation of the supernatant, an aqueous reaction mixture was added to the CLECs. This solution was prepared by mixing 80 mg of Z-L-Asp and 80 mg of L-Phe-OMe-Hcl in 1 ml of water, with acetic acid added to obtain a pH of 7.0. samples were taken for analysis by HPLC. Table 6 below shows the HPLC peak height of the Z-L—Asp substrate peak after the indicated time of reaction, normalized to 1 at time t=0. Since Z-L-Asp is rate limiting in this reaction, measuring its depletion is equivalent to measuring the appearance of product Z-L-Asp-L-Phe-OMe (Nakanishi et. al. Biotechnology 3: 459-464 (1985)).

Table 6 also includes the normalized peak height of limiting Z-L-Asp substrate remaining, and an estimate of the degree of completion of the reaction. It is clear that the reaction proceeded to about 20% completion within the first 30 seconds and plateaued there. These results are consistent with the observations of Nakanishi et al. (Biotechnology 3: 459-464 (1985)) when using conventionally immobilized thermolysin in an aqueous reaction mixture as above, and are attributable to the sparing solubility of the Z-L-Asp-L-Phe-OMe product in water.

TABLE 6 Reaction Peak Height Percent Time (sec) (Normalized) Completion .000 .727 27.3% 60 0.857 14.3% 120 0.940 6.0% 180 0.797 20.3% Synthesis of Z-Asp-Phe-0Me in a near-anhvdrous solution ml of a thermolysin CLEC suspension were added to a continuous stirred-batch reactor incubated at 37°C. After centrifugation and decantation of the supernatant, a near-anhydrous organic reaction mixture was added to the CLECs. This solution was prepared by mixing 80 mg of Z-L-Asp and 240 mg of L-Phe—OMe in 1 ml of 99% ethyl acetate and 1% water. Samples were taken for analysis by HPLC. Table 7 below shows the HPLC peak height of the Z-L-Asp substrate peak after the indicated time of reaction, normalized to 1 at time t=o. Since Z-L-Asp is rate limiting in this reaction, measuring its depletion is equivalent to measuring the appearance of product Z-L-Asp-L-Phe-OMe (Nakanishi et. al. Biotechnology 3: 459~464 (1985)). Table #7 also includes the normalized peak height of limiting Z-L-Asp substrate remaining, and an estimate of the degree of completion of the reaction. In this case, the reaction proceeded to about 70% completion within the first 30 seconds and plateaued there. These results are also consistent with the observations of Nakanishi et al.

(Biotechnology 3: 459-464 (1985)) with conventionally immobilized thermolysin in a near-anhydrous reaction mixture, and are attributable to product inhibition of the enzyme.

TABLE 7 Reaction Peak Height Percent Time (sec) (Normalized) Completion 0 1.000 .323 67.7% 60 0.314 68.6% 120 0.305 69.5% .272 72.8% -S7...

BXAMPLB z Crystallization, cross-linking and lyophilization of thermal s n and assessme t a f gesulting product Crygtalligation of thermglygin Thermolysin (Diawa Kasei K.K., Japan) was dissolved in 10 mM calcium acetate (Sigma), pH 10.0, to a concentration of 10% (w/v). The pH of the solution was maintained at 10.0 by titration with 2 M NaOH.

Following complete solubilization, the protein solution was titrated to pH 8.0 with 2 M Hcl. Solid calcium acetate was added to 1.2 M. Dimethyl sulfoxide (sigma) was then added to 30%. The protein was concentrated to 100 mg/ml by ultrafiltration in an Amicon stir cell (10,000 MWCO membrane). Concentrated enzyme was aliquoted and stored at -70°C. Thermolysin was crystallized by the addition of 9 volumes demineralized water to 1 volume concentrated (100 mg/ml) protein solution. The solution was briefly vortexed and allowed to stand overnight at room temperature.

Crystals were washed with 10 volumes of 10 Mm calcium acetate pH 7.0 and recovered by low speed centrifuga- tion (10 min at 1500 x G, Beckman GPR centrifuge).

The rapid addition of water to a concentrated (100 mg/ml) solution of thermolysin induces the formation of crystals which become visible under low—power magnification within ten minutes. Crystal size is reproducibly dependent on the final protein concentra- tion. Three volumes of water to one volume of thermolysin concentrate (100 mg/ml) will produce 0.5mm long, X-ray diffraction quality hexagonal rods that correspond to the crystals described earlier by Colman et al. (Colman, P.M., Jansonius, J.N. and Matthews, B.W., J. Mol. Biol. 70: 701-724 (1972)). as confirmed by us by diffraction analysis. Adding ten volumes of water to one of protein concentrate reduces the length of the resulting crystals to 0.05mm. These micro- crystals are preferred in CLEC applications, since they tend to minimize diffusion problems related to crystal size (see eg., Quiocho, F.A. and Richards, F.M.

Biochemistry 5: 4062-4076 (1967)). Within a given batch of protein, crystal size was consistently uniform. (Crystals 0.05 - 0.10 mm in length were used in this study to facilitate accurate pipetting of crystalline suspensions.) Densitometer scans of SD8- PAGE showed a six-fold purification of the enzyme on crystallization, significantly increasing the specific activity of the CLECs. Crystallization resulted in a % decrease in the total activity of the CLEC protein compared to soluble thermolysin, when assayed by spectrophotometric cleavage of the dipeptide substrate furylacryloyl-glycyl-L-leucine-amide (FAGLA), as described below. ggosslinking of thermolysin crystals Thermolysin crystals were crosslinked for 3 hours at room temperature in a solution of 12.5% glutaraldehyde (Sigma), 5% DMSO and 50 mM Tris pH 6.5.

The crosslinked crystals were washed 3 times in demineralized water and recovered by low speed centrifugation, as described in respect to crystal- lization of thermolysin. Chemical cross-linking of enzyme crystals stabilizes the crystal lattice and the constituent enzyme molecules in the crystal suf- ficiently so as to permit the practical use of CLECs in environments that are otherwise incompatible with enzyme function. There was no measurable difference in enzymatic activity between the crosslinked and un- crosslinked crystals when assayed (spectrophoto- metrically) by monitoring cleavage of the dipeptide substrate FAGLA (described below). Moreover, cross- linking stabilizes CLECs to the point that they can be lyophilized, with retention of full enzymatic activity upon reconstitution in aqueous, organic, and mixed aqueous-organic solvents as shown in Figure 1 and Table 8. Although crystallization resulted in a 30% decrease in the specific activity of the CLEC protein compared to soluble thermolysin, crosslinking and lyophilization of the CLBCS did not further diminish specific activity.

TABLE 8 - Thermolysin Activity Absorbance 345 nm Time (min) CLEC Soluble Enzyme 1 0.0 0.314 0.315 2 1.0 0.272 0.271 3 3.0 0.235 0.218 4 5.0 0.204 0.198 10.0 0.184 0.185 6 15.0 0.183 0.1 Enz a ’ ‘v of solub e a C EC t ermol sin The catalytic activity of soluble and CLEC thermolysin was assayed (Feder, J. and Schuck, J.M., Biochemistry 9: 2784-2791 (1970)) by hydrolysis of the blocked dipeptide substrate furylacry1oy1-glycy1-L- leucine-amide (FAGLA) (schweizerhall). cleavage of the amide bond was measured spectrophotometrically by a decrease in absorbance at 345 nm. Initial enzyme concentration was 10”M by Bradford protein deter- mination and densitometer scanning (Pharmacia LKB Ultroscan XL) of Coomassie stained SDS-PAGE gels. CLEC enzyme is defined as reconstituted lyophilized cross- linked thermolysin crystals. Soluble enzyme is defined as thermolysin concentrated to 100 mg/ml. Enzyme was added to a 5 ml reaction volume containing substrate.

Aliquots of the reaction mix were removed at the indicated times, and absorbance at 345nm was measured.

CLEC thermolysin was separated from the reaction mix by brief centrifugation (Beckman, microcentrifuge E) before reading absorbance. Absorbance was fitted to a pseudo first order rate equation and kcat/Km was calculated by dividing the fitted value by enzyme concentration (Multifit 2.0 Curve Fitting for the Apple Macintosh Computer, Day Computing P.O. Box 327, Milton, Cambridge CB4 6WL, U.K. (1990)). pH Dependence and Stability The pH optimum and stability of the soluble enzyme were compared to that of thermolysin CLECs by cleavage of the dipeptide substrate FAGLA. Results are shown in Figure 2 and Table 9. Both soluble and crystalline enzyme forms demonstrate maximum activity at pH 7.

CLECs and soluble thermolysin also demonstrated similar activity in the acidic range and the bell shaped pH profile generated by the soluble enzyme was in good agreement with published data (Peder, J. and Schuck, J.M., Biochemistry 9: 2784-2791 (1970)). In the alkaline pH range, however, the crystalline enzyme maintains maximum activity, to pH 10, while the soluble enzyme has 75% activity at pH 8.5, and only 25% activity at pH 9. At pH 9.5, the soluble enzyme is completely inactive. ‘ TABLE 9 - Thermolysin pH Curve % Maximum Activity pH CLEC Soluble Enzyme 1 5.0 10.250 5.170 2 5.5 9.750 6.070 3 6.0 52.500 39.100 4 6.5 85.000 74.610 7.0 97.500 100.000 6 7.5 100.000 98.650 7 8.0 97.500 82.920 8 8.5 95.000 71.910 9 9.0 96.250 24.720 9.5 95.000 0.000 11 10.0 90.000 0.000 Stability at elevated temperature one can achieve higher reaction rates and lower diffusion times for substrates and products by operating a given chemical process at higher temperature, where one is usually limited by the temperature stability of substrates and products. In enzyme-based catalysis, however, it is often the loss of enzymatic activity that sets the practical limit on the temperature that a process can be run. The additional stabilization achieved in CLECs allows for enzymatic activity at much higher temperatures than is possible for soluble enzyme.

The enhanced stability at lower temperatures simplifies the routine long term storage of the CLEC catalysts. For example, it was necessary to store concentrated (>Somg/ml) solutions of soluble thermolysin at -80°C to retain maximum specific activity. At room temperature, activity was usually lost within one day. In contrast, rehydrated thermolysin CLECs could be routinely stored for months at room temperature with no apparent loss of activity.

Unreconstituted lyophilized CLECs of thermolysin appear to be viable indefinitely.

Thermal stability and resistance to autolysis were demonstrated in thermolysin CLECS following incubation at 65°C for five consecutive days (Figure 3 and Table ). Thermolysin CLECs retained maximum activity after five days incubation at elevated temperature. In contrast, the soluble thermolysin lost 50% of its initial activity after only two hours incubation and demonstrated negligible activity after 24 hours incubation at 65°C.

TABLE 10 - Thermolysin Thermal stability at 65°C 8 Maximum Activity Time (days) CLEC soluble Enzyme 1 0.000 100.000 100.000 2 0.041 70.000 3 0.083 96.000 50.000 4 0.164 32.000 0.246 17.000 6 0.410 97.0 10.000 7 1.000 101.0 2.000 8 2.000 97.0 9 3.000 94.0 4.000 96.0 11 5.000 92.0 The activity of soluble and CLEC thermolysin was measured following incubation at 65°C. Soluble thermolysin was incubated in 10 mM calcium acetate, 50 mM Tris pH 7.0 in a 65°C water bath. The reaction volume was Soopl. Final protein concentration was 10 mg/ml. Aliquots were removed at times 0, 1, 2, 4, 6, , and 18 hours. The samples were assayed by SDS-PAGE and FAGLA cleavage at room temperature as described above. For the thermolysin CLECs, a 250 pl crystal suspension in 10mM calcium acetate and 50 mM Tris was also incubated in a 65°C water bath. Activity was assayed at times 0, 1, 6, 24, 48, 72, 96, and 120 hours by FAGLA cleavage. gesistagcg tg exogenous proteolysis Assessment of the resistance of the thermolysin CLEC to the action of an exogenous protease was also carried out. SDS-PAGE (Sodium dodecyl sulfate poly acrylamide gel electrophoresis) analysis suggests that commercial enzymes can contain a substantial percentage of contaminants, some of which might have proteolytic activity against the principal soluble enzyme species.

Given the packing of enzyme molecules in a crystal lattice one might assume that the interior enzyme molecules in a CLEC would be protected from proteolysis. To test this possibility, thermolysin CLECs and a soluble enzyme preparation were incubated in the presence of the streptococcal protease, Pronase®, a nonspecific protease capable of digesting most proteins to free amino acids (Calbiochem 1990 Catalog; LaJolla, CA).

Soluble and CLEC thermolysin were incubated in 50mM Tris, pH 7.5, at 40°C in the presence of the protease Pronase® (Calbiochem). The Pronase® to thermolysin ratio was 1/40. To inhibit thermolysin autolysis and prevent the proteolytic destruction of pronase by the thermolysin, EDTA was added to the soluble enzyme reaction to a final concentration of 100 mM (EDTA inhibits thermolysin activity but not Pronase®). At the times indicated aliquots were removed from the reaction mix and activity was assayed spectrophotometrically by cleavage of the dipeptide substrates FAGLA. To offset thermolysin inhibition due to the presence of EDTA, the spectrophotometric assay of soluble enzyme activity was performed in 0.5 M calcium acetate buffer pH 7.0 and enzyme concentration was increased two fold. Crosslinked crystalline enzyme was assayed as described above.

As can be seen in Figure 4 and Table 11, the soluble thermolysin was rapidly degraded and lost all activity after 90 minutes incubation. In contrast, the activity of the thermolysin CLEC was unaffected by four days incubation in the presence of protease. This near imperviousness to proteolysis is of particular interest in diagnostic biosensor applications where a suitable CLEC might be called upon to act in the presence of an unknown cocktail of naturally occurring proteolytic enzymes.

LE 1 Protease Resistance Time % Maximum Activity Time (days) CLEC soluble Enzyme (min) 1 0.000 100.0 100.0 0.000 2 0.003 25.0 5.000 3 0.010 17.5 15.000 4 0.021 9.5 30.000 0.042 98.0 3.0 60.000 6 0.063 1.0 90.000 7 0.084 101.0 0.0 8 1.000 97.0 9 2.000 99.0 3.000 98.0 11 4.000 96.0 Stability in the presence of organic solvent In order for enzymes to gain ideal acceptance as viable industrial catalysts, they must be able to function without excessive intervention in the practical environment of manufacturing processes. In particular, this would include the use of aqueous, polar and non-polar organic solvents, and mixtures of these. In commercial applications, aqueous-organic solvent mixtures allow manipulation of product formation by taking advantage of relative solubilities of products and substrates.

Soluble thermolysin and thermolysin CLECs exhibited markedly different stability in the presence of organic solvents. (Table 12). Soluble enzyme concentrations which could be incubated in organic solvent were limited to a maximum of 10 mg/ml.

Concentrations greater than this value resulted in the instantaneous precipitation of thermolysin upon addition of organic solvent. In contrast, thermolysin CLEC concentrations were limited only by the volume occupied by the crystals. Soluble thermolysin retained the greatest activity (75%) following incubation in acetone, and the least (36%) in tetrahydrofuran.

Following a one hour incubation in the presence of acetonitrile or dioxane the soluble enzyme lost approximately 50% of its initial activity. The CLEC thermolysin retained greater than 95% maximum activity following incubation with all organics assayed.

TABLE 12 ” % Maximum Activity Soluble CLEO Enzyme Acetonitrile 42 102 Dioxane 66 97 Acetone ' 75 99 war‘ as 95 t Totrahydro Puran Stab’ it o an’ so ve ts Thermolysin CLECS or soluble thermolysin preparations were incubated in 5o%'(v/v) solutions of the indicated organic solvents. A 100 pl slurry of thermolysin CLECs (10 mg/ml) in 10mM Tris pH 7 was placed in a 1/2 dram glass vial. An equal volume of the indicated organic solvent was added and the mixture was briefly vortexed. Twenty pl of soluble thermolysin (100 mg/ml) was diluted in 80 pl of 0.01SM Tris buffer pH 7.0 in a 1/2 dram glass vial. A 100 pl volume of organic solvent was then added to the protein solution and briefly vortexed. CLEC and soluble enzyme were incubated in the presence of organic solvent for one hour at 40°C. Following incubation, enzyme activity was assayed by cleavage of the dipeptide substrate FAGLA as described.

Low water concentration is thought to disfavor unfolding to intermediate states on the path to enzyme denaturation. In CLECs, this restriction of conformational mobility is provided by the inter- molecular contacts and cross-links between the constituent enzyme molecules making up the crystal lattice, rather than by the near-absence of water in the medium. As a result, intermediate water-organic solvent concentrations are readily tolerated by enzymes when formulated as CLECs, something previously unobserved with enzymes (see Table 12). This discovery opens up whole new areas of synthetic chemistry to exploitation using enzyme catalysis.

Even in near-anhydrous organic solvents, however, the routine use of enzymes has been hampered by their tendency to form ill-defined suspensions that are subject to clumping and other aggregation problems.

This property makes these preparations inherently unattractive for large scale industrial processes. In contrast, CLECs and the constituent enzymes within the crystal lattice, remain mono-disperse in all these solvents.

Comparison with other immobilization methods A number of useful reviews of enzyme immobilization methods have appeared in the literature (Maugh, T.H., Science, 223: 474-476 (1984)); Tramper, J., Trends in Biotechnology 3: 45-50 (1985)) . In these, the enzyme always represents a small fraction of the total volume of the immobilized particle, the bulk of it being inert carrier material. The carrier increases the mean free path between the solvent exterior of the immobilized enzyme particle and the enzyme active sites, exacerbating diffusion problems (Quiocho, F.A. and Richards, F.M., Biochemistry 5: 4062-4076 (1967)).

In a CLEC, the cross-linked crystal matrix provides its own support, eliminating the need for a carrier. As a result, the concentration of enzyme in a CLEC is close to the theoretical packing limit that can be achieved for molecules of a given size, greatly exceeding densities achievable even in concentrated solutions. The entire CLEC consists of active enzyme, and thus, the diffusion-related reduction of enzyme reaction rates usually observed with conventionally immobilized enzymes relative to enzymes in solution are minimized (see Figure 1), since the mean free path for substrate and product between active enzyme and free solvent will be greatly shortened for CLECs (compared to a conventional immobilized enzyme carrier particles). Importantly, the constituent enzyme in CLECs is intrinsically mono-disperse, and can be recovered by simple manipulations of the CLEC particles, such as filtration, centrifugation or decantation of solvent.

EXAMPLE 3 Crystallization. gross—linking and lygphilization of elastase and assessment of characteristics of the resulting pggguct.

Crystallization Q; elastase Lyophilized porcine pancreatic elastase (serva) was dissolved in 0.1 M sodium acetate pH 5.0 to a concentration of 5 mg/ml (w/v) at room temperature.

Rod shaped elastase crystals were visible within one minute of the complete solvation of the protein. The crystallization solution was transferred to 4°C and crystallization was completed overnight. Crystals were recovered by centrifugation as previously described.

Cross-linking of elastase crystals A 200 ul volume of elastase crystals was added to a 1.3 ml solution of 5.77% glutaraldehyde and 1.5 M sodium acetate pH 5.0. The crystals were crosslinked for one hour with mild agitation (stir plate).

Following cross-linking the crystals were washed with three 15 ml volumes of 0.2 M Tris pH 8.0. The elastase CLEC was lyophilized as described in Example 2.

Enz atic ctiv't of soluble a elastase The catalytic activity of soluble and CLEC elastase was assayed spectrophotometrically by measuring hydrolysis of the substrate succinyl - (Ala), p-nitroanilide (Bachem) [Bieth, et a1. Biochem. Med 11:350-357 (1974)) (Table 13 Figure 5). Cleavage was monitored by increasing absorbance at 410 nm. Initial substrate concentration was 2 X 10‘. Enzyme concentration was 2.8 X 10”M. CLEC or soluble enzyme was added to a 5 ml reaction volume containing substrate in 0.2 M Tris pH 8.0. As described previously, CLEC enzyme was removed from the reaction mix prior to measuring absorbance.

TABLE 13 - Blastase Activity Absorbance 400 nm Time (min) CLEC Soluble Enzyme 1 0.0 0.000 0.000 2 0.5 0.103 0.205 3 1.0 0.195 0.390 4 2.0 0.366 0.672 3.0 0.523 0.923 6 4.0 0.657 1.098 7 5.0 0.780 1.227 8 6.0 0.888 1.326 9 7.0 0.974 1.393 10.0 1.170 1.512 11 15.0 1.365 1.586 Resistance to exoqenous Droteolvgis Assessment of the resistance of the elastase CLEC to the action of protease was also performed under identical conditions as described for thermolysin (Example 2). Activity of the soluble and CLEC enzyme, following incubation with protease, was assayed by hydrolysis of the nitroanilide substrate as described above (Table 14 and Figure 6).

TABLE 14 - Blastase Resistance to Proteolysis % Maximum Activity Time CLEO soluble Enzyme .0 100.0 100.0 2 10.0 53.0 3 20.0 32.0 4 30.0 101.0 18.0 45.0 11.0 6 60.0 102.0 8.0 7 120.0 101.0 3.0 8 180.0 103.0 2.0 Example 5 Crystallization, cross-linking and lyophilization of gstergsg and assessment 0: cgaracteristigs of tne resulting product.

Crystallization of esterase As disclosed here, 30 mg/ml ammonium sulfate suspension of pig liver esterase (Fluka) was dissolved in 0.25 M calcium acetate pH 5.6 at room temperature.

Esterase crystals were visible within several minutes following addition of the calcium acetate solution.

The crystallization solution was allowed to stand at room temperature and crystallization was completed overnight. Crystals were recovered by centrifugation as previously described in Example 2.

Cross-linkina of esterase crystals As disclosed here, a 300 pl volume of esterase crystals were added to a 5 ml solution of 12.5% glutaraldehyde and 0.5 M sodium acetate pH 5.5. The crystals were crosslinked for one hour with mild agitation (stir plate). Following cross-linking the crystals were washed with three 15 ml volumes of 0.5 M calcium acetate pH 6.3. The esterase CLEC was lyophilized as previously described in Example 2.

Enzymatic activity of soluble and CLEC esterase The catalytic activity of soluble and CLEC esterase was assayed spectrophotometrically by monitoring hydrolysis of the substrate p-nitrophenyl acetate (Fluka) (Table 15 and Figure 7). Cleavage was monitored by increasing absorbance at 400 nm. Initial substrate concentration was 0.001%. Enzyme concentration was 1 X 1o‘M. CLEC or soluble enzyme was added to a 5 ml reaction volume containing substrate in 0.25 M calcium acetate pH 6.3. As described previously in example 2, CLEC enzyme was removed from the reaction mix by centrifugation prior to measuring absorbance.

TABLB 15 - Estorase Activity Absorbance 400 an Time (min) CLBC Soluble Enzyme 1 0.0 0.000 0.000 2 0.5 0.770 0.252 3 1.0 0.128 0.297 4 2.0 0.208 0.337 3.0 0.260 0.346 6 5.0 0.324 0.353 7 7.0 0.353 0.359 8 10.0 0.369 0.368 Resistance to exogenous groteolysis Assessment of the resistance of the esterase CLEC to the action of protease was also performed under identical conditions as described for thermolysin (Example 2). Activity of the soluble and CLEC enzyme, following incubation with protease, was assayed by hydrolysis of the substrate p-nitrophenyl acetate as described above (Table 16 and Figure B).

TABLE 16 - Bsterase Resistance to Proteolysis % Maximum Activity Time (min) CLEO soluble Enzymes .0 100.0 100.0 2 10.0 68.0 3 20.0 47.0 4 30.0 99.0 25.0 45.0 20.0 6 60.0 97.0 ' 16.0 7 120.0 94.0 10.0 8 180.0 91.0 6.0 gxample 5 Crystallization, cross-linking and lyophilization of lipase and assessment of gnagacteristics of the resulting product.

Crystallization of lipase As disclosed here, the enzyme lipase (Geotrichum candidum) was crystallized by vapor diffusion from an aqueous solution of 20 mg/ml protein in 50 mM Tris pH 7 containing 8% ammonium sulfate. Bipyrimidal crystals were visible after 20 to 30 days incubation at room temperature. Crystals were recovered by centrifugation, as previously described in Example 2.

C s - ’ ' f i ase c stals As disclosed here, lipase crystals were added to a solution of 12.5% glutaraldehyde and 50 mM Tris pH 5.6.

The crystals were crosslinked for one hour. Following cross-linking the crystals were washed with three 15 ml volumes of 50 mM Tris pH 7.0. The lipase CLEC was lyophilized, as previously described in Example 2. n2 ati ac v't o soluble and LEC ' ase The catalytic activity of soluble and CLEC lipase was assayed spectrophotometrically by monitoring hydrolysis of the substrate p-nitrophenyl acetate (Table 17 and Figure 9). Cleavage was monitored by increasing absorbance at 400 nm. Initial substrate concentration was 0.005%. Enzyme concentration was 1.5 X 10*M. CLEC or soluble enzyme was added to a 5 ml reaction volume containing substrate in 0.2 M Tris pH 7.0 at room temperature. As described previously in Example 2, CLEC enzyme was removed from the reaction mix by centrifugation prior to measuring absorbance.

TABLE 17 - Lipase Activity Absorbance 400 nm Time (min) CLBC soluble Enzyme .0 0.000 0.000 2 1.0 0.013 0.021 3 5.0 0.094 0.116 4 10.0 0.164 0.186 15.0 0.248 0.258 6 30.0 0.346 0.357 7 45.0 0.407 0.420 8 60.0 0.461 0.459 9 90.0 0.497 0.502 Example 6 Crystallization, cross-linking and lyoghiligation of lysozyme and assessment of characteristics of the resulting prgduct Crystallization of lysogyme Following the method of Blake, C.C.F. et al., Nature 196:1173 (1962) 200 mg of lyophilized hen egg white lysozyme (Boehringer Mannheim) was dissolved in 2.5 ml of 0.04 M sodium acetate buffer pH 4.7 at room temperature. Following solvation of the protein, 2.5 ml of 10% sodium chloride were added to the lysozyme solution dropwise with stirring. The crystallization solution was allowed to stand overnight at room temperature, and crystallization was completed in forty-eight hours. Crystals were recovered by centrifugation, as previously described in Example 2. os - ' o c sta As described here, a 500 ul volume of lysozyme crystals was added to 10 ml of 24% glutaraldehyde and 50 mM Tris pH 5.6 containing 20% sodium chloride. The crystals were crosslinked for 20 minutes with mild agitation (stir plate). Following cross-linking the crystals were washed with three 50 ml volumes of 20 mM calcium acetate and 50 mM potassium chloride pH 5.3.

The lysozyme CLEC was lyophilized, as previously described in Example 2.

Enzymatic activitv of soluble and CLEC lvsozvme The catalytic activity of soluble and CLEC lysozyme was assayed by measuring the rate of hydrolysis of the substrate 4-methylumbelliferyl N- acetyl-chitrioside (Fluka) (Yang, Y. and Hamaguchi, K.

J. Biochem. 8:1003-1014 (1980) (Table 18 and Figure ). The release of 4-methylumbelliferone was followed fluorimetrically (Perkin Elmer Model LS-50). Initial substrate concentration was 1.42 X 10*. Enzyme concentration was 3 X 10”. CLEC or soluble enzyme was added to a 2 ml reaction volume containing substrate in mM calcium acetate and 50 mM potassium chloride pH .3 at 42°C. The amount of 4-methylumelliferone was determined fluorimetrically by measuring fluorescence intensities at 450 nm with excitation at 360 nm. slit width for both excitation and emission was 10 mm. As described previously in Example 2, CLEC enzyme was removed from the reaction mix by centrifugation prior to measuring fluorescence.

TABLE 18 - Lysozyno Activity rluorescenco Time (min) CLEO soluble Enzyme 1 0.000 0.000 0.000 2 10.000 4.400 18.900 3 30.000 10.500 29.400 4 60.000 27.500 44.800 90.000 33.800 51.700 6 120.000 45.900 59.800 Example 7 grystallization. cross-linking and lyophiligation gf asparaginase and assessment og characteristics of the resultinq product.

Crystallization of asgaraginase As a modification of the procedure described by Grabner et al. [U.S. Patent 3,664,926 (1972)] 25 mg of lyophilized asparaginase (Worthington) were dissolved in 500 pl of 50 mM sodium phosphate buffer pH 7.2. The solution was cooled to 4°C and the pH adjusted to 5.0 with 1 M acetic acid. Cold (-20°C) ethanol was then added dropwise to the asparaginase solution to a final concentration of 33%. The solution was incubated at 4°C. Crystallization was completed in forty-eight hours. Crystals were recovered by centrifugation as previously described.

Cross-linking gf aspagaginase crystals As disclosed here, asparaginase crystals were crosslinked in a solution of 7.5% glutaraldehyde in 50 mM sodium phosphate buffer pH 5.6. Following crosslinking the crystals were washed with five 15 ml volumes of 50 mM tris pH 7;o. The asparaginase CLECs were lyophilized, as previously described in Example 2. t‘ a 'v't ub e nd ' ase The catalytic activity of soluble and CLEC asparaginase was assayed spectrophotometrically by measuring evolution of ammonium ion in the coupled enzymatic reaction described below (all reagents were purchased from Boehringer Mannheim) (Table 19 and Figure 11). asparaginase L-asparagine -------------- --> asparate + NH;+ glutamate dehydrogenase Nfln-+ NADH + a ketoglutarate --------------------- --> glutamic acid + NAD+ oxidation of NADH was measured by decreasing absorbance at 340 nm. Initial NADH concentration was 1.4 mg/ml.

Asparagine concentration was 103M. Alpha ketoglutarate concentration was 10‘M. Glutamate dehydrogenase concentration was 10”M. Asparaginase concentration was 2.3 x 10*M. As described previously in Example 2, CLEC enzyme was removed from the reaction mix by centrifugation prior to measuring absorbance.

TABLE 19 - Asparaginaso Activity Absorbanco 340 um Time (min) CLEC soluble Enzyme 1 0.0 1.000 1.000 2 1.0 0.867 0.825 3 3.0 0.739 0.684 4 5.0 0.603 0.538 10.0 0.502 0.406 6 15.0 0.449 0.338 7 30.0 0.328 0.199 8 45.0 0.211 0.187

Claims (1)

1. CLAIMS An enzyme crystal crosslinked with a multifunctional crosslinlcing agent, said crosslinked enzyme crystal having resistance to exogenous Pronasem proteolysis and being selected from the group consisting of: a) pyridoxal phosphate enzymes, metalloenzymes, CoA-requiring enzymes and sulfurases (PAPS); b) thermolysin, nitrilase, aminocyclase and hydantoinase; c) elastase, aminopeptidase, asparaginase, glutaminase, arginase, urease, alkaline phosphatase, lipase, preferably phospholipase, B-lactamase, ACL hydrolase, L-ACT hydrolase, arylsulfatase, glycosidase, preferably [3-galactosidase and B-fructosidase, ’ amidase, protease, peptidase, lactase, amino acid ' esterase, 5'-phosphodiesterase, nuclease and creatine deiminase; d) dehydrogenase, amino acid dehydrogenase, formate dehydrogenase, lactate dehydrogenase, glutamate dehydrogenase, hydroxysteroid dehydrogenase, glucosepyruvate dehydrogenase, succinate dehydrogenase, glucose dehydrogenase, alcohol dehydrogenase, catalase, superoxide dismutase, peroxidase, tyrosinase, isomerase, flavoenzyme, nitrate/nitrite reductase, oxidoreductase (NAD/P), oxidase and lyase; or e) kinase (ATP), preferably creatine kinase, transaminase, transferase, glycosyl transferase and methyl transferase (SAM) for use in a medicament. The enzyme crystal according to claim 1, wherein said enzyme is thermolysin. The enzyme crystal according to claim 1, wherein said enzyme is elastase. The enzyme crystal according to claim 1, wherein said enzyme is asparaginase. The enzyme crystal according to claim 1, wherein said enzyme is lipase. The enzyme crystal according to claim 1, wherein said oxidase is alcohol oxidase, glucose oxidase, bilirubin oxidase, lactate oxidase, cholesterol oxidase or L-amino oxidase, and/or said lyase is aldolase, L-aspartate 4-decarboxylase, phenylalanine ammonium lyase, carbonic anhydrase, nitrile hydratase, L-aspartase, fumarase or heparinase. The enzyme crystal according to claim 1, selected from the group consisting of: L- tryptophan synthetase and strictodine synthetase. The enzyme crystal according to claim 1, wherein said enzyme is as defined in c) of claim 1 and wherein said crosslinked enzyme crystal retains at least 91% of its initial activity after incubation for three hours in the presence of a concentration of Pronasem that causes the soluble uncrosslinked fomi of the enzyme that is crystallized to form said enzyme crystal that is crosslinlced to retain at most 6% of its initial activity under the same conditions, and said enzyme being esterase. The enzyme crystal according to any one of claims 1 to 8, wherein the crosslinker is glutaraldehyde. F. R. KELLY & CO., AGENTS FOR THE APPLICANTS