Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Enzyme immobilization using the cellulose-binding domain of the Cellulomonas Fimi exoglucanase Ong, Edgar 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1992_spring_ong_edgar.pdf [ 9.51MB ]
Metadata
JSON: 831-1.0098964.json
JSON-LD: 831-1.0098964-ld.json
RDF/XML (Pretty): 831-1.0098964-rdf.xml
RDF/JSON: 831-1.0098964-rdf.json
Turtle: 831-1.0098964-turtle.txt
N-Triples: 831-1.0098964-rdf-ntriples.txt
Original Record: 831-1.0098964-source.json
Full Text
831-1.0098964-fulltext.txt
Citation
831-1.0098964.ris

Full Text

ENZYME IMMOBILIZATION USING THE CELLULOSE-BINDING DOMAIN OF THE CELLULOMONAS FIMI EXOGLUCANASE by EDGAR ONG B.S., University of Santo Tomas, 1983 M.Sc, Katholieke Universiteit Leuven, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Microbiology THE UNIVERSITY OF BRITISH COLUMBIA February 1992 © Edgar Ong, 1992 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of MICROBIOLOGY The University of British Columbia Vancouver, Canada 26 March 1992 Date DE-6 (2/88) ii A B S T R A C T A new strategy employing molecular genetic techniques to produce fusion polypeptides was used for enzyme immobilization. The cellulose-binding domain (CBD) of the Cellulomonas fimi exoglucanase (Cex) was used as an affinity "tag" by fusing the sequence encoding CBDcex to the 3'-ends of the genes encoding two 6-glucosidases, Abg and Cbg, from a mesophilic Agrobacterium sp. and a thermophilic bacterium, Caldocellum saccharolyticum, respectively. The resulting fusion polypeptides, Abg-CBDcex and Cbg-CBDcex, were purified to >95% homogeneity in a single step by affinity chromatography on cellulose using water as the desorbing agent. Protein recovery varied from 58-70%. The matrix, pH and temperature affected the reversibility of binding. These parameters can be selected to facilitate either purification or stable immobilization of the polypeptide. Fusion of the affinity "tag" appeared not to affect the conformation of the fusion partners significantly, since their specific activities were not altered appreciably. Kinetic analyses revealed that the increase in apparent Km of the immobilized Abg-CBDcex was not directly related to the fusion, but rather, to external mass transfer resistance. The percentage of activity retained by the immobilized enzyme was inversely related to the amount of fusion polypeptide adsorbed to cellulose. This was affected by the presence of an extra 57 amino acid residues between the fusion partners. Abg-CBDcex was stably adsorbed to cellulose between 4° to 50°; pH 3 to 8.5; and in the presence of up to 1 M NaCl. The immobilized Abg-CBDcex column exhibited long-term operational stability (at least 15 days at 37°) and gave 50-70% substrate conversion at the flow rates employed. At 50° Abg-CBDcex lost activity but remained bound to cellulose. Immobilized Cbg-CBDcex was stable at 70° for at least 3 days with no apparent iii desorption of fusion polypeptide from the column. The use of a more crystalline cellulosic support improved the stability of binding of both fusion polypeptides. A plasmid that expressed the gene encoding CBDcex alone was also constructed to facilitate biochemical analysis of the binding mechanism. CBDcex was purified from culture supernatant to > 98% homogeneity in a single step by batch affinity chromatography on cellulose. Unlike the fusion polypeptides, it was not desorbed with water. The presence of xylan in the CBDcex preparation affected its susceptibility to cleavage by pepsin. CBDcex bound stably to both amorphous and crystalline cellulose, and under a wide range of temperatures, pHs and detergent concentrations. T A B L E OF C O N T E N T S Page ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES vii LIST OF FIGURES viii LIST OF ABBREVIATIONS xi ACKNOWLEDGMENTS xiii 1. INTRODUCTION 1 1.1 Enzyme immobilization 1 1.2 Fusion polypeptides for enzyme immobilization 4 1.3 Cellulose-binding domains 5 1.4 Cellulose structure 8 1.5 Objectives of the study 10 2 . MATERIALS AND METHODS 11 2.1 Chemicals, media components, buffers and enzymes 11 2.2 Bacteria, plasmids, growth media and conditions 11 2.3 Recombinant DNA work 14 2.4 Screening for gene expression 15 2.5 Production of polypeptides 15 2.5.1 Small-scale production of polypeptides 15 2.5.2 Large-scale production of fusion polypeptides 16 2.5.3 Large-scale production of CBDcex from culture supernatant 17 2.6 Determination of protein and carbohydrate concentrations... 18 2.7 Gel electrophoresis - staining for protein 18 2.8 Gel electrophoresis - in situ detection of enzyme activity 19 2.9 Amino acid sequence determinations 19 2.10 Protease digestion of CBDcex 20 2.11 Determination of disulfide bonds in CBDcex 20 2.12 Binding analysis 20 2.13 Enzyme kinetics 21 2.14 Immobilization of Abg-CBDcex and Cbg-CBDcex 22 2.15 Stability of CBDCex, Abg-CBDcex and Cbg-CBDcex 23 2.15.1 pH 23 2.15.2 Ionic strength 24 2.15.3 Temperature 25 2.15.4 Time 25 2.15.5 Detergents 26 2.16 Performance of Abg-CBDcex and Cbg-CBDcex immobilized enzyme columns 27 RESULTS 30 3.1 The cellulose-binding domain of C.fimi exoglucanase Cex (CBDcex) 30 3.1.1 Construction of plasmids expressing CBDcex 30 3.1.2 Production of CBDcex 38 3.1.3 Properties of CBDcex 38 3.1.4 Binding of CBDce x 42 3.1.5 CBDcex contains a disulfide bridge 50 3.1.6 Sensitivity of CBDcex to proteases 50 3.2 A fusion polypeptide comprising CBDcex fused to the C-terminus of a 6-glucosidase (Abg) from an Agrobacterium sp. (Abg-CBDcex) 57 3.2.1 Construction of plasmids expressing abg-CBDCex 57 3.2.2 Improved yield of Abg-CBDcexl in E. coli CAG456/pE01 57 3.2.3 Screening of Abg-CBDcex 64 3.2.4 Large-scale production of fusion polypeptides 67 3.2.5 Properties of the fusion polypeptides 67 3.2.6 Binding of Abg-CBDcexl to cellulose 72 3.2.7 Stability of fusion polypeptides 73 3.2.8 Enzyme kinetics 73 3.2.9 Immobilization yields for the fusion polypeptides.... 80 3.2.10 Stabilities of the immobilized enzymes 82 3.2.11 Cellulosic matrices 86 3.3 A fusion polypeptide comprising CBDcex fused to the C-terminus of a 6-glucosidase (Cbg) from Caldocellum sac char olyticum (Cbg-CBDcex) 90 3.3.1 Construction of a plasmid expressing cbg-CBDcex— 90 3.3.2 Production of Cbg-CBDcex 90 3.3.3 Properties of Cbg-CBDcex 93 3.3.4 Cbg-CBDcex as an immobilized enzyme 96 DISCUSSION 99 REFERENCES 110 vii L I S T OF T A B L E S Table Page 1.1 Immobilization of carbohydrases 3 1.2 Immobilization of fusion polypeptides 5 2.1 Escherichia coli strains 12 2.2 Plasmids 13 3.1 Purification of Abg-CBDcexl by affinity chromatography on CF1™ cellulose 70 3.2 Purification of fusion polypeptides 72 3.3 Catalytic activities of Abg and the fusion polypeptides 77 3.4 Activity of immobilized Abg-CBDcexl at different temperatures 89 L I S T O F F I G U R E S Figure Page 1.1 Domain arrangement in CenA and Cex 6 1.2 Structures of cellulose, cellulose I and cellulose II 9 2.1 Schematic diagram of the system to test performance of Abg-CBDcex and Cbg-CBDcex immobilized enzyme columns 28 3.1 Block diagrams of CBDcex and fusion polypeptides 31 3.2 Construction of pTZE07 32 3.3 Construction of pTZE07 (PTIS) 34 3.4 Construction of pTZE04 36 3.5 CBDcex production in E. coli JM101/pTZEO7 39 3.6 CBDcex production by E. coli JM101/pTZEO7 and JM101/ pTZE07 (PTIS); Purity of CBDCex 40 3.7 Removal of xylan from purified CBDcex 41 3.8 Adsorption of CBDcex to cellulose and a-chitin 45 3.9 Relative affinities for the binding of CBDcex to cellulose and a-chitin 46 3.10 Adsorption of CBDcex to BMCC at different temperatures... 47 3.11 Adsorption of CBDcex to BMCC at different pHs 48 3.12 The influence of detergents on the binding of CBDcex to cellulose 49 3.13 The influence of 6-mercaptoethanol on the mobility of CBDcex 51 3.14 Trypsin sensitivity of CBDcex 52 3.15 Digestion of CBDcex with pepsin 55 3.16 Construction of pEOl 58 ix 3.17 Construction of pTZEOl and pTZE02 60 3.18 Construction of pTZE03 62 3.19 Abg-CBDcexl production by various strains of E. coli 65 3.20 Integrity of Abg-CBDcex produced by various strains of E. coli 66 3.21 Purification of Abg-CBDcexl by affinity chromatography on CF1™ cellulose 68 3.22 The influence of cellulose matrix on the recovery of Abg-CBDcexl 69 3.23 Column output of Abg-CBDCexl purified from CF1™ cellulose 71 3.24 Heterogeneity of purified Abg-CBDcexl 74 3.25 Detection of heterodimers in purified Abg-CBDcex 75 3.26 Stability of Abg-CBDcexl adsorbed to Avicel™ 76 3.27 Column flow characteristics of pNP produced from an immobilized Abg-CBDcexl membrane column 79 3.28 Adsorption equilibria and immobilization yields for Abg-CBDcexl & 3 adsorbed to Avicel™ 81 3.29 Stability of immobilized Abg-CBDcexl as a function of pH... 83 3.30 Desorption of Abg-CBDcexl from cellulose with a pH gradient 84 3.31 The influence of ionic strength on the stability of Abg-CBDcexl 85 3.32 Performance of Abg-CBDcexl immobilized columns: effect of cellulose structure 87 3.33 Performance of Abg-CBDcexl immobilized columns: temperature stability 88 3.34 Construction of pTZEOlO 91 X 3.35 Purification of Cbg-CBDcex 94 3.36 Purification of Cbg-CBDcex 95 3.37 Binding of Cbg-CBDcex to cellulose 97 3.38 Temperature stability of Cbg-CBDcex immobilized enzyme column 98 L I S T OF A B B R E V I A T I O N S xi 2F-DNPG [N0] [P] [Po] a Abg Abg-CBDcex BMCC BME BSA CBD CBDcenA CBDcex Cbg Cbg-CBDcex CD CenA Cex dH20 DNP DTNB DTT £280nm FPLC GdmCl IPTG 2\4'-dinitrophenyl-2-deoxy-2-fluoro-B-D-glucopyranoside Concentration of binding sites in the absence of protein Protein concentration Initial protein concentration Number of lattice units occupied by a protein Agrobacterium sp. 6-glucosidase Fusion polypeptide between Abg and CBDcex Bacterial microcrystalline cellulose 6-mercaptoethanol Bovine serum albumin Cellulose-binding domain CBD of CenA CBD of Cex Caldocellum saccharolyticum 6-glucosidase Fusion polypeptide between Cbg and CBDcex Catalytic domain Cellulomonas fimi endoglucanase A Cellulomonas fimi exoglucanase Distilled water Dinitrophenolate 5,5'-dithionitrobenzoic acid Dithiothreitol Extinction coefficient at 280nm Fast protein liquid chromatography Guanidinium chloride Isopropyl-6-D-thiogalactoside Ka kbp Kcat kDa Km LB LP MUG Pi PMSF pNP pNPG PTH PTIS SDS-PAGE TPCK Vmax X-glu Equilibrium association constant kilobase pairs Enzyme turnover number kilodaltons Michaelis-Menten constant Luria-Bertani Leader peptide Methylumbelliferyl-6-D-glucoside Isoelectric point phenylmethylsulfonylfluoride p-nitrophenolate /7-nitrophenyl-6-D-glucopyranoside Phenylthiohydantoin Portable translation initiation site sodium dodecyl sulfate-polyacrylamide gel electrophoresis iV-tosyl-L-phenylalanine chloromethyl ketone Maximum rate of enzyme reaction 5-bromo-4-chloro-3-indolyl-6-D-glucoside A C K N O W L E D G M E N T S This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada. Many people helped me along the way. I would like to express my sincere appreciation particularly to the following: Doug Kilburn, my supervisor, for his enthusiasm, patience and excellent supervision, and for believing in my ability to accomplish things; Tony Warren, Bob Miller and Neil Gilkes for their constant support, creative ideas and direction, and helpful discussion; Emily Kwan for her technical help and friendship; Helen Smith for providing every little thing that I needed to complete my experiments; Don Trimbur for the use of TS Graph (version 2); Gary Lesnicki and Lando Robillo for growing up cells in the fermentor; The Wesbrook (now NiCE) Deputies for the friendship, Friday night's out, and informal, but substantial discussion (sciencewise or otherwise); Everyone else in the Cellulase Labs, the Media Room and the Department for the laughter, care and camaraderie. I lovingly dedicate this thesis to the memory of my mother, Mrs. Rose Ong (1931-1990), for her love and sacrifice up to the very end. 1 . I N T R O D U C T I O N 1 1.1 Enzyme immobilization Enzyme immobilization is an important aspect of biotechnology. It is a process by which movement of enzymes is restricted physically or chemically. It is a part of the much broader area of immobilization technology wherein proteins, other than enzymes (e.g., antibodies, receptors, etc.), and cells are fixed to or entrapped within a matrix. Enzymes can be immobilized by either binding or physical retention (Hartmeier, 1988). Binding to a support can be accomplished either directly by adsorptive, ionic or covalent attachment, or by cross-linking enzymes to the support using glutaraldehyde and similar agents. Enzymes can be retained physically either by matrix entrapment or by membrane enclosure. In most enzyme immobilization processes, the mechanisms overlap somewhat and several could function concurrently. Enzymes are immobilized for the following reasons: (1) substrate conversion can proceed continuously rather than batchwise; (2) the products of bioconversion are free of enzymes; (3) the product yield per unit volume and time is increased; (4) the potential problem of substrate and/or product inhibition is substantially reduced especially in a column technique; (5) the immobilized enzyme often shows improved pH and temperature stability, and catalytic efficiency (Gianfreda et ah, 1985; Webster et al., 1985; Baker et ah, 1988; Kimura et al., 1989; Mozhaev, 1990; Reiken et ah, 1990; Mattiasson & Kaul, 1991). Efficiency and stability are the two most important parameters for a successful immobilized enzyme process (Monsan & Combes, 1988). The immobilized enzyme should maintain its native conformation, or, if changed, should retain a significant proportion of the catalytic activity of the soluble enzyme. The immobilized enzyme should remain bound and maintain its 2 catalytic activity for prolonged periods in continuous operation without the need for enzyme replenishment. In 1916, Nelson and Griffin reported the first immobilization process in which invertase from yeast was bound to activated charcoal and still maintained catalytic activity. Significant advances in enzyme immobilization did not begin until the 1960s. In 1969, the first large-scale industrial application of an immobilized enzyme was demonstrated with the production of L-amino acids using bound L-aminoacylase (Chibata, 1978). Since the 1970s, immobilization technology has advanced beyond the simple one-enzyme type of hydrolytic/ isomeric reactions to include many novel immobilization processes utilizing a great variety of natural and chemically-modified supports. Recent examples include the use of immobilized cells/proteins for multi-step enzymatic reactions (Karkare, 1991; Kimura etal., 1990a); for biochemical synthesis (Jayakumari & Pillai, 1990; Yoshida et al, 1990; Kimura et al., 1990b); for clinical diagnosis (Roda et al., 1991; Chien et al, 1991) and analytical biosensors (Guilbault, 1984; Lowe, 1984; Guilbault, 1988; Guilbault etal, 1991; Huang etal, 1991; Taylor, 1991); for detoxification of pesticides (Caldwell & Raushel, 1991); and for purification of target proteins (Baneyx & Georgiou, 1989; Greenwood et al, 1989; Chen et al, 1991). Several excellent reviews and monographs deal with the various aspects of immobilization technology (Chibata, 1978; Buchholz, 1979; Bernath & Venkatasubramanian, 1987; Hartmeier, 1988; Taylor, 1991). Carbohydrases have been immobilized on a large number of supports by various modes with varying success in terms of the amount of activity retained after immobilization, i.e. the immobilization yield (Table 1.1). Various supports have been used for immobilizing proteins. The ideal support should be (1) stable, insoluble and chemically inert, (2) porous and permeable, yet rigid and resistant to compression, and (3) cheap and reusable. Natural cellulose satisfies most of these requirements, and is readily available in Table 1.1 Immobilization of carbohydrases Enzyme 6-amylase 6-galactosidase 6-galactosidase 6-galactosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-glucosidase 6-xylosidase Cellulases Glucoamylase Glucose isomerase Glucose isomerase Invertase Maltohexao-hydrolase Support polystyrene/Al3+ alumina hydrophobic cotton cloth alginate Sepharose ConA-Sepharose alginate hydrophobic cotton cloth PHSA4/PHEMA5 alginate acrylamide Sepharose ConA-Sephadex alumina/TiCU ConA-Sepharose DEAE-cellulose porous glass beads DEAE-cellulose poly(ethylene-vinyl alcohol) acrylamide Mode ionic covalent/GA2 adsorption covalent/CD3 covalent adsorption covalent adsorption entrapment covalent/GP6 entrapment covalent adsorption/GA covalent adsorption/ covalent-CNBr ionic covalent adsorption/ ionic ionic copolymeri-zation/radiation IYl 7 25 50 42 44 93 30 50 50/26 38 51 33 95/71 70 9 61 47 70 13 75 Reference Roy & Hegde, 1987 Nakanishi et al., 1983 Sharma & Yamazaki, 1984 Dominquez et al, 1988 Kierstan et al, 1982 Lee & Woodward, 1983 Hahn-Hagerdal, 1984 Sharma & Yamazaki, 1984 AXfzmetal, 1987 Fujikawa et al., 1988 Roy etai, 1989 Roy etal, 1989 Husain & Saleemuddin, 1989 Oguntimein & Reilly, 1980 Woodward & Zachry, 1982 Tomar & Prabhu, 1985 Strandberg & Smiley, 1972 Chen et al., 1981 Imai etal., 1986 Nakakxte et al., 1983 1 Immobilization yield (percentage 2 Glutaraldehyde; 3 Carbodiimide; hydroxyethylfnathacrylate of activity retained by immobilized enzyme); 4 Genipin; 5 poly-human serum albumin; 6 poly-4 a variety of different forms (fibres, powders and papers). Reactive sites in cellulose may be generated for the covalent attachment of proteins (Chen et al., 1981; Cannon et al, 1984; Tomar & Prabhu, 1985; Jin & Toda, 1988; Dumitriu et al., 1989). The principal disadvantages of this approach involving covalent attachment are the extra expense of the derivatized support, and the potential for conformational changes in the protein as a consequence of immobilization, which ultimately result in decreased catalytic activity. 1.2 Fusion polypeptides for enzyme immobilization This thesis demonstrates the use of a new strategy for enzyme immobilization on cellulose in which a cellulose-binding domain (CBD) from a bacterial exoglucanase is fused to a heterologous enzyme by molecular genetic techniques (Ong et al., 1989a). The hybrid polypeptide produced by this process binds to cellulose and retains the catalytic activity of the fused enzyme. The fusion of genes using molecular genetic techniques allows the production of hybrid polypeptides comprised of two or more separate functional domains (Biilow & Mosbach, 1991). In this way one can custom-design any desired protein, whose gene is available, fused to an affinity "tag". This affinity "tag" can be located at either the N- or C-terminus of the fusion partner, depending on whether fusion at one end inactivates the partner (Greenwood et al., 1989; Ong et al., 1989a). There are numerous examples in the literature describing the use of various affinity "tags" for purification of proteins (Sassenfeld, 1990). The design is basically similar for all of these hybrid polypeptides: the polypeptide of interest is fused to a peptide with high affinity for an appropriate ligand. A protease recognition site can be incorporated between the two partners allowing for the precise cleavage of the fusion protein to release the protein of interest. Under appropriate conditions, the hybrid polypeptide can also be left adsorbed to the matrix for use in an immobilized enzyme reactor (Table 1.2). Table 1.2 Immobilization of fusion polypeptides Enzyme B-glucosidase B-lactamase GlcDHcys44* Bacterial luciferase Affinity tag CBD2 Protein A Cysteine +LDH4 Protein A Support Cellulose IgG on Sepharose 4B Thiopropyl-Sepharose Hollow fibre (< 15 K) IgG on Sepharose 4B Mode adsorption adsorption adsorption entrapment adsorption Stability1 70% > 15 days NT>3 ND 56% at 2.5 days ND Reference Ong et al., 1989a; 1991 Baneyx et al., 1990 Persson et al., 1991 Persson et al., 1991 Lindbladh et al., 1991 1 continuous mode; 2 cellulose-binding domain; 3 not determined; 4 lactate dehydrogenase; * glucose dehydrogenase with Asp44Cys 1.3 Cellulose-binding domains Cellulomonas fimi is a Gram-positive, aerobic, mesophilic, rod-shaped bacterium that produces a cellulase system composed of multiple enzyme components -some of which bind tightly to cellulose. The genes for the major cellulase components in C.fimi have been cloned, sequenced, characterized and expressed in E. coli (O'Neill et al, 1986; Wong et al, 1986; Owolabi et al, 1988; Coutinho et al., 1991; Meinke et al., 1991). Two of the most well-characterized cellulases of C.fimi are endoglucanase A (CenA) and an exoglucanase (Cex) (Figure 1.1). Both enzymes are glycoproteins with molecular masses of 53 kDa and 49.3 kDa, respectively. When produced in E. coli their molecular masses, determined by SDS-PAGE analysis, are 48.7 kDa and 47.3 kDa, respectively, NHo CenA + 112 134 Cex NH-COOH 418 ^H 316 335 COOH 443 Figure 1.1 Domain arrangement in CenA and Cex. The catalytic domain (unshaded area) is separated from the cellulose-binding domain (shaded area) by a Pro-Thr linker (striped area). The numbers refer to the amino acid residues of the mature protein. The arrows refer to cleavage site by a C. fimi serine protease. reflecting the lack of glycosylation in the recombinant forms. CenA and Cex are both composed of two domains, the catalytic domain (CD) and the cellulose-binding domain (CBD), separated from each other by a short linker composed principally of prolyl and threonyl residues, the Pro-Thr linker. The CD of Cex is found at the N-terminus and the CBD at the C-terminus. This arrangement is reversed in CenA. Proteolysis of recombinant CenA and Cex with a C. fimi serine protease releases a core peptide that retains enzymatic activity but no longer binds to cellulose (Figure 1.1; Gilkes et al., 1988). The shorter peptide bearing the CBD could be selectively removed by adsorption to Avicel™ (Gilkes et al., 1988). CenA and Cex are therefore enzymes containing two functionally independent domains separated by a linker sequence. 7 The CBDs of Cex and CenA are 108 and 111 amino acids long, respectively, and their sequences share more than 50% identity (O'Neill et al., 1986; Wong et al., 1986). Similar sequences are found in many cellulases and xylanases (Gilkes et al., 1991b). The amino acid sequences of the C.fimi-type. bacterial CBDs are highly conserved, with low numbers of pharged amino acids, high contents of hydroxy amino acids, and conserved tryptophan, asparagine and glycine residues (Ong et al., 1989b; Gilkes et al, 1991b). Two conserved cysteine residues, which participate in disulfide bond formation in CenA and Cex, are found at the N- and C-termini in all CBDs but one (endoglucanasel of Butyrivibrio fibrisolvens). Tryptophan residues are a feature of other proteins which interact with polysaccharides: the starch-binding domain of a glucoamylase from Aspergillus niger (Svensson et al., 1989); the pilus-associated adhesion proteins in E. coli (Lund et al., 1988), and an animal lectin carbohydrate-recognition domain (Drickamer, 1988). At present the only cellulase catalytic domain for which the three-dimensional structure has been solved is that of cellobiohydrolase II (Cbhll) from Trichoderma reesei (Rouvinen et al., 1990), but this and other fungal CBDs are structurally distinct from the C. fimi type. Small-angle X-ray scattering analyses of CenA and Cex show that the molecules are tadpole shaped. The catalytic domain forms an ellipsoidal head region and the Pro-Thr linker and CBD forms an extended tail region (Pilz et al., 1990; Shen et al., 1991; Gilkes, unpublished results). The CBD of Cbhl was synthesized chemically and shown by two-dimensional nuclear magnetic resonance spectroscopy to be wedge-shaped with two disulfide bridges (Kraulis etal., 1989). The mechanism and significance of binding of a CBD to cellulose are poorly understood. In a number of sugar-binding proteins found in the periplasm of Gram-negative bacteria, charged and aromatic amino acids are the 8 principal residues participating in the formation of an extensive hydrogen bond network and van der Waals interaction between the protein and the carbohydrate substrate (Quiocho, 1986; Johnson et ah, 1988). Tryptophan in particular is thought to play a role in substrate specificity by restricting the binding of sugar epimers because of steric hindrance or polarity of the hydroxyl groups (Quiocho, 1988). Some water molecules also participate in hydrogen bond formation. Other amino acid residues, especially the hydroxy amino acids, form a secondary shell around the principal amino acids, thereby allowing more hydrogen bond formation (Spurlino et al., 1991). 1.4 Cellulose structure Cellulose is the principal polysaccharide found in the cell walls of plants. It is a linear polymer of 6-D-glucopyranosyl units linked by 6-1,4-glucosidic bonds. Cellulose molecules have up to 10,000 glucopyranosyl units linked to form long chains of molecular weight up to 1.62 million. Native cellulose, or cellulose I, has a very ordered and regular structure. It forms fibrils composed of long, threadlike bundles of molecules that are stabilized laterally by hydrogen bonds between hydroxyl groups. Cellulose I exhibits a parallel-chain structure, with each repeating cellobiosyl unit forming two intramolecular hydrogen bonds and linked by intermolecular bonds to its neighbors (Figure 1.2). When native cellulose is swollen with acid or alkali, or regenerated from solution by precipitation, cellulose I is converted to cellulose II in which the molecules are in an anti-parallel arrangement. This affords increased intermolecular hydrogen bonding resulting in a more stable and lower-energy structure than cellulose I. There is still a controversy as to the nature of the transformation between cellulose I and II (Blackwell, 1981; Sarko, 1986). Crystalline cellulose resists the entry of water or enzyme molecules, making it 9 A 1 f r ? T r l ' r St™CtUreS ° f C e U u l ° S e ( A ) ' C e U u l o s e J W- m d «*>totae H £ L AH, ? H ' / Df' f ial P l a n e P r ° J e C , i 0 n S : d 0 t t e d l ine- intennolecular H-bond. Adapted from Blackwell (1981). resistant to biological hydrolysis. However, interspersed with the regular crystalline regions are paracrystalline or amorphous regions of a less ordered nature which are more susceptible to enzymatic hydrolysis. Treatment of cellulose with acid or alkali can reduce the degree of crystallinity, thereby facilitating enzymatic hydrolysis. 1.5 Objectives of the study The general objective of this study was to explore the feasibility of using the cellulose-binding domain (CBD) of the C. fimi exoglucanase Cex for enzyme immobilization. The specific objectives included: (1) the construction of fusion polypeptides to serve as model systems for enzyme immobilization on cellulose; (2) the functional characterization of the isolated CBDcex and fusion polypeptides, especially with respect to their pH and temperature stability, and binding to various types of cellulose; (3) the operation of immobilized enzyme columns in a continuous mode. 2 . MATERIALS AND METHODS 2.1 Chemicals, media components, buffers and enzymes All chemicals used were of analytical or HPLC grade. Media components were from Difco. Buffers were prepared as described previously (Gomori, 1955; Stoll & Blanchard, 1990). Restriction and modifying enzymes were used according to manufacturers' recommended procedures. Bacterial microcrystalline cellulose (BMCC) was prepared from cultures of Acetobacter xylinum (Hestrin, 1963). Phosphoric acid swollen cellulose or regenerated cellulose was prepared as described previously (Wood, 1988). Avicel™ PH-101, ot-chitin, Cellufine™, and CF1™ cellulose were purchased from FMC International (Ireland), Sigma, Grace Co. (Amicon), and Whatman, respectively. 2.2 Bacteria, plasmids, growth media and conditions Escherichia coli was the host bacterium for recombinant DNA work and protein production. The E. coli strains plus their relevant genotypes, and the plasmids plus their associated genetic characters used in the study are listed in Tables 2.1 and 2.2, respectively. The medium used for growing E. coli strains carrying recombinant DNA was Luria-Bertani (LB) medium (tryptone, 10 g.L-1; yeast extract, 5 g.L-1; NaCl, 10 g.L-1) supplemented with 100 jug ampicillin (Sigma). mL_1 (Sambrook et al., 1989). All recombinant E. coli strains were grown under inducing conditions (0.1 mM final concentration of isopropyl-B-D-thiogalactoside [IPTG; Sigma]) at 37° for approximately 16-18 h unless stated otherwise. Table 2.1 Escherichia coli strains Strain Genotypes Reference CAG440 lacZam trpam phoam supCts mal Baker et al., 1984 rpsL phe rel CAG456 CAG440, rpoUl65 (/tf/?R165) Baker et al, 1984 JM101 F fraD36 lacW A{lacZ)Ml5 Yannish-Perron et proAB/supE thi A(lac-proAB) al., 1985 JM109 F traD36 lacW A{lacZ)M\5 Yannish-Perron et proAB/recAl end Al gyrA96 al., 1985 (Nalr) thi hsdRll (rK-mK+) supE44 rel Al A(lac-proAB) PM191 dra drm thr leu thi lacY recA56 Meacock & Cohen, supE 1980 RZ1032 HfrKL16 PO/45 [lysA(6l-62)] Kunkel et al., 1987 dutl ungl thil rel Al Zbd-279::Tn70 supEU Table 2.2 Plasmids Plasmid Genetic characters Reference pABG5 pEOl pNZ1070 pTZ18R pTZlSR-cbg pTZEOl pTZE02 pTZE03 pTZE04 pTZE06 pTZE07 pTZE07 (PTIS) pTZE08 pTZEOlO pUC12 pVCll-l.lcex bla plac abg bla plac abg-CBDcex bla plac flori cbg bla plac flori bla plac flori cbg bla plac flori abg-CBDCexl bla plac flori abg-CBDCex2 bla plac flori abg-CBDCex3 bla plac flori CBDcex bla plac flori cex bla plac flori CBDcex bla plac flori CBDCex bla plac flori cbg bla plac flori cbg-CBDCex bla plac bla plac cex Wakarchuk et al., 1986 Ong et al, 1989 Love etal, 1988 Pharmacia D. Trimbur, unpublished work This study This study This study This study This study This study This study This study This study Vieira & Messing, 1982 O'Neill et al, 1986 2.3 Recombinant DNA work All protocols for recombinant DNA work were described previously (Sambrook et al., 1989) unless stated otherwise. Double-stranded plasmid DNA was prepared by the alkaline-lysis method. It was purified by cesium chloride/ ethidium bromide ultracentrifugation if required. Plasmid DNA for double-stranded sequencing was prepared as described previously (Kraft et al., 1988). Single-stranded Ml 3 phage DNA was prepared by extraction with phenol-chloroform. Uridine-containing single-stranded DNA was prepared as described previously (McClary et al., 1989). After restriction endonuclease digestion of the desired plasmid, DNA fragments were separated by agarose gel electrophoresis and purified by the GeneClean™ method (BiolOl, La Jolla, CA). Following ligation of the desired insert(s) and vector with T4 DNA ligase, the ligated mixture was transformed into competent E. coli cells (Hanahan, 1982). Transformed cells were selected on LB-ampicillin plates. Oligodeoxyribonucleotides for mutagenesis (Kunkel et al., 1987) and DNA sequencing (Sanger etal., 1977) were synthesized using an Applied Biosystems automated DNA synthesizer model 380A (Oligonucleotide Synthesis Facility, University of British Columbia, Vancouver, B.C.). They were purified by polyacrylamide gel electrophoresis and reversed-phase chromatography on Sep-Pak columns (Millipore)(Atkinson & Smith, 1984). Single- or double-stranded DNA was sequenced by the dideoxyribo-nucleotide chain-terminating method (Sanger et al., 1977) using modified T7 DNA polymerase and 35S-cc-dATP (Tabor & Richardson, 1987). 2.4 Screening for gene expression Transformants were screened by plating on LB plates supplemented with ampicillin, IPTG and either 100 |iM methylumbelliferyl-B-D-glucoside (MUG; Sigma) or 100 JLIM 5-bromo-4-chloro-3-indolyl-6-D-glucoside (X-glu; Sigma). After incubation overnight at 37°, 6-glucosidase activity was detected by the production of either fluorescent (MUG) colonies when observed under long-wavelength UV light, or blue (X-glu) colonies. Transformants were also screened by colony immunoblotting (Hanahan & Meselson, 1983) and Western blot analysis (Harlow & Lane, 1988). Primary polyclonal antisera against Agrobacterium 6-glucosidase (Abg) and C. find exoglucanase (Cex) were raised in rabbits. Goat anti-rabbit IgG-alkaline phosphatase conjugate (BRL) was used as secondary antibody. The detection reagents were 5-bromo-4-chloro-3-indolyl-phosphate (Sigma) and nitroblue tetrazolium dye (Sigma). 2.5 Production of polypeptides 2.5.1 Small-scale production of polypeptides E. coli cells were grown in 25 mL LB supplemented with ampicillin and IPTG in a 125-mL flask. The culture was grown overnight at 37° (30° for E. coli CAG456), 150 rpm (New Brunswick Scientific G-25 incubator shaker). Cells were harvested by centrifugation at 17,400g, 4° for 10 min, washed in 5 mL of 50 mM potassium 50 mM potassium phosphate buffer, pH 7, pH 7 and recentrifuged. The cell pellet was resuspended in 3 mL 50 mM potassium phosphate buffer, pH 7 supplemented with 3 mM EDTA. Cells were homogenized using a 5-mL French press cell (Aminco). The cell extract was clarified by centrifugation at 17,400g, 4° for 20 min. Cell extract was stored at 4° in the presence of 0.02% NaN3, 1 mM phenylmethylsulfonyl-fluoride (PMSF; BDH), and 1 JIM pepstatin A (Sigma). 2.5.2 Large-scale production of fusion polypeptides E. coli cells were grown in 60 L LB supplemented with ampicillin and IPTG in a 110-L fermentor (L.H. Fermentation). Growth conditions were the same as in 2.5.1. Cells were separated from the culture medium by centrifugation at 31,000g (Sharpies Inc.). Cells were washed in 500 mL of 50 mM potassium phosphate buffer, pH 7, centrifuged at 4,000g, 4° for 10 min and resuspended in 800 mL of 50 mM potassium phosphate buffer, pH 7 supplemented with 3 mM EDTA. Cells were homogenized in 40 mL batches using a 50-mL French press cell. PMSF and pepstatin A were added immediately to prevent proteolysis. Cell debris was removed by centrifugation at 17,400g, 4° for 25 min. Nucleic acids were precipitated from the supernatant by addition of streptomycin sulfate (Sigma) to a final concentration of 1.5% (wt/v) and incubating the mixture overnight. The precipitate was removed by centrifugation at 17,400g, 4° for 20 min. The supernatant was then centrifuged at 150,000g, 4° for 40 min, using a Beckman Ti 50.2 rotor, so as to remove membrane-associated debris and loaded onto cellulose columns as described below. CF1™ cellulose was washed extensively with distilled H2O (dH20) to remove fines and then packed into a column in 50 mM potassium phosphate buffer, pH 7. Two column sizes were routinely used: a 30 cm x 5 cm diameter column (~ 400 mL bed volume; Pharmacia XK50/30) and a 60 cm x 5 cm diameter column (~ 800 mL bed volume; Pharmacia XK50/60). After equilibration of the column with 50 mM potassium phosphate buffer, pH 7 at 4°, the sample was loaded into the column at a flow rate of 1 mL.mhr1. The column was washed with 5 bed-volumes of 1 M NaCl in 50 mM potassium phosphate buffer, pH 7 followed by 2 bed-volumes of 50 mM potassium phosphate buffer, pH 7. Desorption was effected using a concave descending gradient of 50 mL 50 mM potassium phosphate buffer, pH 7 and 2 L of dH20. For Cbg-CBDcex further desorption was achieved using a 0-8 M gradient of guanidinium chloride (GdmCl) in 50 mM potassium phosphate buffer, pH 7. The column run was controlled and monitored with an FPLC™ (fast protein liquid chromatography) system (Pharmacia). Peak fractions were identified by on-line absorbance readings at 280 nm, pooled and concentrated by ultrafiltration through a 10 kDa-cutoff membrane (Amicon PM-10). If necessary, the polypeptide sample was further purified by MonoQ anion-exchange chromatography using the FPLC™ system. Enzyme preparations were stored at 4° in the presence of 0.02% NaN3. 2.5.3 Large-scale production of CBDcex from culture supernatant E. coli cells were grown as in 2.5.2. The cells were removed by centrifugation at 31,000g (Sharpies) and the supernatant was transferred to 3 x 20-L jugs (Pyrex) - each containing 330 g of Avicel™ or CFl™ cellulose. The suspension was stirred periodically for 3 to 5 h at room temperature to allow binding of protein to cellulose. After the cellulose settled, the supernatant was aspirated. Each 330 g of cellulose was resuspended in 1 L of 1 M NaCl in 50 mM potassium phosphate buffer, pH 7, stirred for 30 min at room temperature and filtered through a GF/C glass fibre filter (Whatman). The NaCl wash was repeated, followed by two washes with 1 L of 50 mM potassium phosphate buffer, pH 7. Bound proteins were desorbed from the cellulose with 500 mL of 8 M GdmCl in 50 mM potassium phosphate buffer, pH 7. The eluate was concentrated and exchanged with dH20 by ultrafiltration using a 2 kDa-cutoff membrane (Amicon YM-2). The polypeptide was either lyophilized or stored in solution at 4° in the presence of 0.02% NaN3. The material purified from Avicel™ was contaminated with xylan (see 3.1.2) and was used for all subsequent experiments except when noted. Contaminating xylan (see pp. 37, 39 & 48) was removed from purified CBDcex by cation-exchange chromatography on MacroprepS beads (Bio-Rad) using a pH gradient consisting of 20 mM formic acid/trimethylamine buffer, pH 4 and 20 mM trimethylamine/formic acid, pH 7. 2.6 Determination of protein and carbohydrate concentrations Protein concentrations in cell extract and partially purified preparations were determined by dye binding (Bradford, 1976). Bovine serum albumin (BSA) was used as the standard. Protein concentrations in purified preparations were determined by absorbance at 280 nm using the extinction coefficient obtained for the purified polypeptides (Scopes, 1974). The carbohydrate contents of purified polypeptides were determined by the phenol-sulfuric acid method (Chaplin, 1986). D-Glucose was used as the standard. 2.7 Gel electrophoresis - staining for protein Polypeptides were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (Laemmli; 1970; Schagger & von Jagow, 1987). Non-denaturing PAGE was performed in the same manner except for the omission of SDS, 6-mercaptoethanol (BME) and heat treatment; the pH of the system was 8.8; the stacking gel was 3%, the resolving gel 7.5%; the running buffer was Tris-Tricine without SDS (Schagger and von Jagow, 1987). Isoelectric focusing PAGE was performed using a Phast gel system (Pharmacia). Protein bands were visualized by either Coomassie blue staining (Merril, 1990) or silver staining (Merril, 1990). Quantification of protein bands was done using a densitometer (Molecular Dynamics ImageQuant version 3.0, Sunnyville, CA). The following proteins were used as molecular weight markers: myosin (205-212 kDa); 8-galactosidase (116-130 kDa); phosphorylase B (97.4 kDa); bovine serum albumin (66-68 kDa); catalase (57.5 kDa); glutamate dehydrogenase (53 kDa); alcohol dehydrogenase (43-45 kDa); ovalbumin (41 kDa); glucose-3-phosphate dehydrogenase (36 kDa); carbonic anhydrase (29 kDa); soybean trypsin inhibitor (20.1 kDa); lysozyme (14 kDa); and cytochrome C (12.3 kDa). 2.8 Gel electrophoresis - in situ detection of enzymatic activity Polypeptides were resolved by non-denaturing PAGE. The gel was then incubated in 50 mM potassium phosphate buffer, pH 7 for 15 min at room temp. 6-glucosidase activity was detected by immersing the gel in 2 mM MUG for 2-5 min at room temp. Fluorescent protein bands were detected under long-wavelength UV light. 2.9 Amino acid sequence determinations Polypeptides were resolved by SDS-PAGE, then electroblotted (0.5 amperes, 30 min) onto a poly(vinylidine difluoride) membrane (Immobilon™; Millipore) (Matsudaira, 1987; 1990). Stained bands corresponding to the peptide of interest were excised from the blots and sequenced directly by automated Edman degradation using an Applied Biosystems 470A gas-phase sequenator with on-line PTH analyzer and 900A system controller and data analyzer (Protein Sequencing Facility, University of Victoria, Victoria, B.C.). 2.10 Protease digestion of CBDcex The following proteases were used: endoproteinase Lys-C (Boehringer-Mannheim); endoproteinase Asp-N (Sigma); TPCK-treated trypsin (Cooper); papain (Sigma); pepsin (Sigma) and a C.fimi serine protease (Gilkes etal., 1988). Each of the proteases was used at its optimum pH and temperature in either the presence or absence of urea and/or dithiothreitol (Aitken et al., 1989; Matsudaira, 1989). The protease to CBDcex weight ratio varied from 1:5 to 1:30. Digestion was monitored by SDS-PAGE. 2.11 Determination of disulfide bonds in CBDcex Purified CBDcex was analyzed by SDS-PAGE in the absence of BME. Reaction of 5,5'-dithionitrobenzoic acid (DTNB) with free thiol groups in CBDcex was determined as described previously (Creighton, 1989). 2.12 Binding analysis Purified polypeptides (0.15-90 jxM) were mixed with either 5 mg of Avicel™, 1 mg of BMCC, 1 mg a-chitin or 1 mg regenerated cellulose equilibrated in 1 mL of 50 mM potassium phosphate buffer, pH 7 contained in 1.7 mL micro-centrifuge tubes. Tube contents were mixed continually by rotation at 2 rpm on a Labquake Shaker™ (Labindustries, Inc., Berkeley, CA). After equilibration at the desired temperature for 3 to 24 h, substrate and bound polypeptides were removed by centrifugation at 14,800g for 10 min. The supernatant was removed and assayed either for 6-glucosidase activity (Abg-CBDcex) or by absorbance at 280 nm (CBDCex). 2.13 Enzyme kinetics Purified Abg-CBDcex or Cbg-CBDcex was diluted in 50 mM potassium phosphate buffer, pH 7 containing 0.2 mg BSA.mL-1 as a stabilizer. 6-glucosidase activity was estimated by measuring the initial rate of p-nitrophenolate (pNP) formation from /?-nitrophenyl-B-D-glucopyranoside (pNPG; Sigma). One unit of 6-glucosidase or pNPGase activity releases one |imole of pNP per min from pNPG in either 50 mM potassium phosphate buffer, pH 7 at 37° (Abg-CBDcex) or phosphate-citrate buffer, pH 6.2 at 70° (Cbg-CBDcex) (standard conditions). The Michaelis-Menten kinetic parameters of Abg and soluble Abg-CBDcex were determined using pNPG concentrations from approximately 0.1 x Km to 20 x Km. The Vmax for Abg-CBDcex bound to cellulose and resuspended in a stirred cell was determined by active site titration using 2',4'-dinitrophenyl-2-deoxy-2-fluoro-B-D-glucopyranoside (2F-DNPG) as a substrate analogue (Withers & Street, 1988). 2F-DNPG inactivated the bound, active 6-glucosidase fusion polypeptide by the rapid formation of a covalent glycosyl-enzyme intermediate at the active site with concomitant release of an equimolar quantity of dinitrophenolate (DNP) (Withers & Street, 1988). The amount of DNP produced was directly proportional to the number of active sites of the bound fusion polypeptide. The phenol product was measured at a wavelength of 400 nm. The column flow characteristics and the apparent Km of immobilized Abg-CBDcex m a continuous column reactor were determined as described previously (Lilly et ah, 1966; Ford et aL, 1972). The filter holder (see 2.14) containing fusion polypeptide adsorbed to cellulose acetate membranes was equilibrated with 50 mM potassium phosphate buffer, pH 7 at room temperature for 30 min. pNPG was perfused through the column and the appearance of pNP in the column effluent was measured by absorbance at 405 nm. The absorbance was plotted against the column effluent volume to obtain a time distribution of pNP in the column. The apparent Km was determined using a similar procedure and a pNPG concentration range of 150-10,000 uM. Reaction was carried out at 37°, 60 mL.h"1 flow rate. 2.14 Immobilization of Abg-CBDcex and Cbg-CBDcex The following cellulosic adsorbents were used: dewaxed, absorbent cotton (Fisher); CF1™ cellulose; Avicel™, Cellufine™, and cellulose acetate membranes (Memtek). Cellulose was first resuspended in d t^O and centrifuged or filtered under vacuum. This procedure was repeated once with dH20 and twice with 50 mM potassium phosphate buffer, pH 7. Finally, the cellulose was resuspended to a fixed volume in 50 mM potassium phosphate buffer, pH 7. Sufficient Abg-CBDcex was then added to saturate the cellulose. After brief mixing, the suspension was incubated at 4° for 3 h with gentle agitation, then centrifuged. The supernatant was assayed for 6-glucosidase activity and the activity of the enzyme bound to cellulose was determined from the difference between this value and the original activity in the preparation. The pellet was resuspended in 50 mM potassium phosphate buffer, pH 7, the mixture centrifuged and the supernatant discarded. The moist pellet was then kept at 4° until required. For Cbg-CBDcex five cellulose acetate membranes (0.2 mm x 25 mm diameter per membrane) were stacked and installed in a filter holder (Memtek). The membrane was equilibrated by pumping 50 mM potassium phosphate buffer, pH 7 through the holder at a given temperature. A given amount of fusion protein (4 nmoles) was then passed through the holder at a flow rate of 1 mL.mhr1. The amount of fusion protein bound to cellulose was determined from the difference between the 6-glucosidase activity found in the flow through and that added initially. 2.15 Stability of CBDcex, Abg-CBDcex and Cbg-CBDcex adsorbed to cellulose 2.15.1 pH 2.15.1.1 CBDcex The buffers used were 25 mM citrate, pH 3 and 5; 25 mM phosphate, pH 7; and 25 mM carbonate, pH 9 and 11. Increasing concentrations of purified CBDcex (0.9-90 |iM) were adsorbed to 1 mg of BMCC (2 mg.mL-1 made up in the appropriate buffer) in a total volume of 1 mL contained in 1.7 mL micro-centrifuge tubes. After equilibration at the desired pH, at 22° with constant agitation for 18-24 h, the cellulose was pelleted by centrifugation at 17,400g, 22° for 10 min. The concentrations of polypeptides in the supernatant were measured by absorbance at 280 nm. 2.15.1.2 Abg-CBDcexl The buffers (50 mM) used were: KC1-HC1 (pH 2); glycine-HCl (pH 3); sodium acetate-acetic acid (pH 4 and 5); citrate-phosphate (pH 6 and 6.5); phosphate (mono- and dibasic) (pH 7, 7.5 and 8); glycine-NaOH (pH 9 and 10) and phosphate-NaOH (pH 11). Cellufine™ (3 g) was saturated with Abg-CBDcexl (124.1 U). The Cellufine™ was filtered on a glass fibre filter. A 200 mg sample of the Cellufine™ was placed in a vial and 4.8 mL of buffer of the appropriate pH was added to it. The Cellufine™ was resuspended with a vortex mixer. The suspension was incubated in a shaker bath at 25°, 150 rpm for 3 days. A 0.5 mL sample of the suspension was centrifuged; the Cellufine™ was washed in 50 mM potassium phosphate buffer, pH 7, resuspended in 0.5 mL of 50 mM potassium phosphate buffer, pH 7 and assayed for 6-glucosidase activity under standard conditions. A control sample of Abg-CBDcexl in solution (0.5 |aM) was incubated and assayed as above. Bound fusion polypeptide was also analyzed by SDS-PAGE. A uniform amount of Cellufine™ (2 mg), after incubation at the indicated pH, was extracted with SDS loading buffer. The solution obtained was then loaded onto a 10%T SDS-polyacrylamide gel. The binding stability was also tested by eluting a column (2.5 cm x 0.5 cm diameter; Pharmacia HR 5/2, containing 0.5 mL of Cellufine™ suspension, 50% v/v) containing 0.03 umoles of bound Abg-CBDcexl with a pH gradient of the appropriate constant ionic strength buffers (50 mM sodium acetate, pH 6.9 and 0.2 M HC1 + 50 mM KC1, pH 1.2; 50 mM triethanolamine, pH 7.1 and 0.2 M NaOH + 50 mM KC1, pH 12.6) (Elving et al, 1956). Protein desorption was monitored by absorption at 280 nm. Fractions (1 mL) were collected and the pH of each measured at 25°. 2.15.2 Ionic strength Stability of soluble and bound Abg-CBDcexl as a function of ionic strength was determined as follows: a 500 mg sample of Abg-CBDcexl -Cellufine™ suspension was placed in a vial to which 4.5 mL of dH20 or potassium phosphate buffer (pH 7) of the appropriate ionic strength were added. The cellulose was resuspended and incubated at 25°, 150 rpm for 24 h. Subsequently, a 0.5 mL sample of the suspension was taken, washed and analyzed for 6-glucosidase activity as in the pH stability study. A control sample of Abg-CBDcexl in solution, 1.2 |J,M, was prepared in the same manner. Residual 6-glucosidase activities were determined under standard conditions. The binding stability of the fusion polypeptide at varying ionic strength was also tested using a column flow-through technique. A column (2.5 cm x 0.5 cm diameter; Pharmacia HR5/2) containing 0.5 mL of Cellufine™ suspension (50% v/v) with 0.03 jimoles of bound Abg-CBDcexl was eluted with a linear salt gradient. The buffers used for the linear gradient were from 0-1 M NaCl in 50 mM potassium phosphate buffer, pH 7 at a flow rate of 0.1 mL.mur1. Protein desorption was monitored by absorption at 280 nm. 2.15.3 Temperature The buffer used was 25 mM potassium phosphate buffer, pH 7. Increasing concentrations of purified CBDcex (0.9-90 |J,M) were adsorbed to 1 mg of BMCC (2 mg.mL-1 in 50 mM potassium phosphate buffer, pH 7) in a total volume of 1 mL contained in 1.7 mL micro-centrifuge tubes. After equilibration at the desired temperature with constant agitation for 18-24 h , the cellulose was pelleted by centrifugation at 17,400g, 22° for 10 min. The protein concentrations in the supernatant were measured by absorbance at 280 nm. 2.15.4 Time Stability of the fusion polypeptide during storage was determined as follows: Abg-CBDcexl adsorbed to Avicel™ (0.012 umoles enzyme.g Avicel™-1) was incubated in a final volume of 10 mL (containing 1 g Avicel™) of 50 mM potassium phosphate buffer, pH 7 for 7 weeks at 4° and 37°. A 0.5 mL sample of the suspension was taken weekly and centrifuged to pellet the Avicel™. The Avicel™ was washed once in 50 mM potassium phosphate buffer, pH 7, then resuspended with 0.5 mL of 50 mM potassium phosphate buffer, pH 7. The suspensions were then assayed for 6-glucosidase activity under standard conditions to determine residual activity. As a control, 4.8 (imoles of Abg-CBDcexl were incubated in 10 mL of 50 mM potassium phosphate buffer, pH 7, and assayed in the same manner as the immobilized enzyme. In addition, all samples were analyzed by SDS-PAGE. Samples of the soluble Abg-CBDcexl were first concentrated by lyophilization. After addition of SDS loading buffer into each sample, volumes equivalent to equal original weights of the test protein were loaded onto a 10%T SDS-poly aery lamide gel. 2.15.5 Detergents Triton X-100, a non-ionic detergent, and SDS, an ionic detergent, were tested at concentrations from 0.002% to 2% (v/v). Nine nmoles CBDcex were bound to 5 mg Cellufine™. The Cellufine™ was washed and resuspended in a given concentration of detergent. The mixture was incubated at 22° with constant agitation for 4 h, centrifuged and the supernatant discarded. Polypeptides remaining bound to the cellulose was extracted with SDS loading dye and analyzed by SDS-PAGE. An analogous procedure was used to test whether binding could occur in the presence of detergent. In this case binding of CBDcex to cellulose was done in the presence of detergent for 2 h at 22° with constant agitation. 2.16 Performance of Abg-CBDcex and Cbg-CBDcex immobilized enzyme columns A schematic diagram of the system used to test the performance of immobilized enzymes is shown in Figure 2.1. Cotton fibres were cut to lengths of 0.85 mm or less with a Wiley cutting mill (A. Thomas Co.). One to two grams of the cut cotton were washed twice with dH20 and once with 50 mM potassium phosphate buffer, pH 7 by filtration and resuspension. The fibres were mixed with Abg-CBDcexl in 50 mL 50 mM potassium phosphate buffer, pH 7, incubated at 4° with shaking at 200 rpm for 3 h, then centrifuged at 17,400 g, 4° for 15 min. The supernatant was removed and assayed for fi-glucosidase activity. The fibres were resuspended in 50 mM potassium phosphate buffer, pH 7 and packed into a jacketed column (20 cm x 1.6 cm diameter; ~35 mL bed volume; Pharmacia XK16/20). The column was connected to a circulating water bath (Forma) and equilibrated at the desired temperature. Unbound fusion polypeptide was desorbed with 50 mM potassium phosphate buffer, pH 7. Substrate (3 mM pNPG in 50 mM potassium phosphate buffer, pH 7) was then perfused through the column at a constant flow rate using a peristaltic pump (Pharmacia P3). The outflow from the column was collected in 7.5 mL fractions to which an equal volume of 1 M Na2C03 was added by the same peristaltic pump to inactivate any enzyme that might have desorbed from the support. A sample of each fraction was transferred to a microtitre plate and the pNP produced was measured by absorbance at 405 nm. The same protocol was followed when CF1™ cellulose was used as a support instead of cotton except that it was not pretreated mechanically. Essentially the same protocol was followed for Cbg-CBDcex except that the enzyme was immobilized on 5 cellulose acetate membranes (0.2 mm x 25 mm CELLULOSE FUSION PROTEIN 1 M Na2COa n MIXTURE WASH PACK j j f j f i j nr V - - 4 pNPG in buffer Alternative immobilization arrangement using cellulose acetate membranes • <« [ 70 #R] rt*i _^y - • A 405 nm w Figure 2.1 Schematic diagram of the system to test performance of Abg-CBDcex and Cbg-CBDcex immobilized enzyme columns 29 diameter per membrane; Memtek) stacked in series in a filter holder. The buffer used was phosphate-citrate buffer, pH 6.2 optimal for Cbg. 3 . R E S U L T S 3.1 The cellulose-binding domain of C. fimi exoglucanase Cex (CBDCex; Figure 3.1) 3.1.1 Construction of plasmids expressing CBDcex A C. fimi serine protease cleaves Cex into two peptides - a core peptide that retains enzymatic activity against />-nitrophenyl-B-D-cellobioside, and a shorter peptide that still adsorbs to cellulose (Gilkes et al., 1988). N-terminal amino acid sequence analysis of this shorter peptide indicated a sequence corresponding to the amino acids immediately after the Pro-Thr linker (Figure 1.1; Gilkes et al., 1988). In this study the cellulose-binding domain of Cex (CBDcex) corresponded to the same sequence starting at the serine residue immediately after the Pro-Thr linker and extending to the C-terminus of Cex. Two extra amino acids (A & S) were introduced at the N-terminus of CBDcex corresponding to an Nhel site created to facilitate DNA manipulation. Three plasmids were constructed to express CBDcex: pTZE07, pTZE07 (PTIS) and pTZE04 (Figures 3.2, 3.3 & 3.4). The coding sequence for the Cex leader peptide was directly fused to the sequence encoding CBDcex in both pTZE07 and pTZE07 (PTIS). This allowed the export of the polypeptide to the periplasm of the cells. In the latter plasmid a portable translation initiation site (PTIS) corresponding to the consensus Shine-Dalgarno sequence was introduced upstream of the ATG. The plasmid pTZE04 was a cloning vector containing the multiple cloning sites of pTZ18R introduced upstream of the CBDcex coding sequence. This vector was made to facilitate cloning of heterologous genes to the CBDcex coding sequence. In all three plasmids expression was controlled by the c B D c « 11 kDa Abg-CBD^ I 68 kDa 110 ASSGP 459 496 516 624 M/TDPN Abg-CBDCex2 66 kDa 459 498 606 M/TDPN 459 569 Abg-CBDCex3 62 kDa M/TDPN Cbg-CBD^x 64 kDa 452 562 ASFPK Figure 3.1 Block diagrams of CBDcex and fusion polypeptides. Abg: Agrobacterium 6-glucosidase. Cbg: Caldocellum saccharolyticum 6-glucosidase. Sizes in kDa are based on the predicted amino acid sequences. Numbers refer to amino acid residues. The letters give the N-terminal amino acid sequences determined experimentally. Q : Abg or Cbg. U : Cex catalytic domain. • : Pro-Thr linker. • : CBDcex-Figure 3.2 Construction of pTZE07. p\JC12-l.lcex (PTIS) (O'Neill et aL, 1986) (see Figure 3.3) was digested completely with BamHI and Hindlll. The cex insert was isolated and ligated to pTZ18R, which had been digested completely with BamHI and Hindin, to give pTZl.lcex. An Nhel site was introduced between the Cex leader peptide (LP) coding sequence and the coding sequence for the Cex catalytic domain by oligonucleotide-directed in vitro mutagenesis to give pTZE06. pTZE06 was digested completely with Seal, Nhel and Sspl and the 1.9 kbp fragment containing the Cex LP coding sequence was isolated. The plasmid pTZE02 (see 3.2.1) was digested completely with Seal and Nhel, and the 1.8 kbp fragment containing the CBDcex coding sequence was isolated. The 1.8 kbp and the 1.9 kbp fragments were ligated to give pTZE07. 33 Nhel Seal Nhel Seal Seal Nhel Seal Nhel Sspl 1800 bp NheKcex CBD)-Scal 1891 bp Scal-(cex LP)-Nhel T4 DNA Ligase TGA Smal ^ P s t l Sphl Hhdlll Seal Figure 3.3 Construction of pTZE07 (PTIS). pUC12-l.lc<?x (PTIS) was completely digested with Seal, BamHI and PstI, and the 1.7 kbp fragment containing the PTIS sequence was isolated. pTZE07 was digested completely with Seal, BamHI and PstI, and the 1 kbp and 0.85 kbp fragments were isolated. The 1.7, 1 and 0.85 kbp fragments were ligated to give pTZE07 (PTIS). 35 PstJ BamHI Pstl BamHI Seal Seal Seal BamHI Pstl 1700 bp Scal-(PTIS)-BamHI Seal BamHI Pstl 1000 bp Scal-(f1 or/)-Pstl 850 bp Pstl-(cex CBD)-BamHI DNA Ligase. Sad BamHI Ps« Sphl Hindlll Seal Figure 3.4 Construction of pTZE04. pTZE02 was digested completely with Nhel. The fragments were made blunt-ended using kPolI, then digested completely with PstI. The 0.7 kbp fragment containing CBDCex was isolated. pTZ18R was digested with Xbal, made blunt-ended as above, digested with PstI, and ligated to the 0.7 kbp fragment from pTZE02 to give pTZE04. 37 Nhel Xbal kPoll Pstl Isolate 2.8 kbp fragment Nhel kPoll Pstl Isolate 0.7 kbp fragment -T4 DNA Ligase Lac p romote r -SGal I Cex CBD sequence Gly Asn Ser Ser Ser Val Pro Gly Asp Pro Leu Ala Ser .GGG AAT TCG AGC TCG GTA CCC GGG GAT CCT CTA GCT AGC EcoRI I SacI I Kpnl I I BamHI I I Nhel Ser Gly Pro Ala TCC GGT CCG GCC. Pstl SpW Hindlll lac promoter. The correctness of the constructs was confirmed by restriction endonuclease analysis and DNA sequencing. 3.1.2 Production of CBDcex The production of CBDcex in E. coli JM101/pTZEO7 was detected by Western blot analysis (Figure 3.5). Approximately 30% of the CBDcex produced by the cell was found in the culture medium. The introduction of the consensus Shine-Dalgarno sequence upstream of the ATG of the coding sequence for the Cex leader peptide increased the production of CBDcex by 4-fold (Figure 3.6 A). CBDcex was purified by batch affinity chromatography on Avicel™. The typical yield was between 10-15 mg CBDcex-L"1 culture medium. Recovery of CBDcex bound to either Avicel™ or CF1™ cellulose using GdmCl was greater than 90%. The material obtained was >98% pure (Figure 3.6 B). The material purified from Avicel™ was contaminated with xylan (up to 30 K mol. wt.) as determined by proton nuclear magnetic resonance spectroscopy and fast-atom bombardment mass spectrometry (J. Carver, personal communication). The xylan was removed by cation-exchange chromatography (Figure 3.7). The A28O nm peaks 2 and 3 corresponded to CBDcex- The reason for two CBDcex peaks was not clear. Subsequent proteolytic digestion by pepsin (see 3.1.6) involved peak 3 only. 3.1.3 Properties of CBDcex The mol. wt. of CBDceX, deduced from the DNA-derived protein sequence (O'Neill et al., 1986), was 11,081. The Mr of CBDcex as determined by SDS-PAGE was U K . The isoelectric point (pi) for the polypeptide was approximately 8.6. The predicted pi was 8.3 (Skoog & Wichman, 1986). The A 1 2 3 4 5 6 7 8 11 kDa-1 2 3 4 5 6 7 8 11kDa-Figure 3.5 CBDcex production in E. coli JM101/pTZEO7. Crude ceU extracts were prepared from two transformants. A sample (50 jig total protein) was bound to 1.25 mg of Avicel™. The mixture was incubated for 3 h on ice with periodic agitation. After centrifugation the cellulose pellet was washed with 2 x 100 JLIL of 1M NaCl in phosphate buffer followed by 2 x 100 |jL phosphate buffer. Following centrifugation the pellet was resuspended in 20 \\L of 2x SDS loading dye, boiled for 2.5 min and centrifuged. Equivalent amounts of proteins were analyzed by SDS-PAGE. Lane 1, cell extract of E. coli JM101/pTZ18R. Lane 2, Cex cleaved with a C. firm serine protease. The upper band, p36, is the Cex catalytic domain with the Pro-Thr linker. The lower band is CBDcex-Lanes 3 and 6, crude cell extracts from two transformants. Lanes 4 and 7, unbound proteins. Lanes 5 and 8, proteins which bound to Avicel™. (A) Coomassie blue-stained gel. (B) Western blot. The primary antiserum used was rabbit anti-Cex. A PTZE07 PTZE07/PTIS B kDa 1 2 3 4 Figure 3.6 A. CBDcex production by E. coli JM101/pTZEO7 and JM101/pTZEO7 (PTIS). Culture medium (500 uL) and crude cell extract (50 fig) were bound separately to 5 mg Avicel™. The mixtures were treated as in the legend to Figure 3.5. The polypeptides were analyzed by SDS-PAGE. Lane 1, molecular weight markers. Lane 2, purified CBDcex- Lanes 3 and 6, Avicel™-bound proteins from culture medium. Lanes 4 and 7, Avicel™-bound proteins from crude cell extract. Lanes 5 and 8, crude cell extracts corresponding to l/8th of the amounts loaded in lanes 4 and 7. B. Purity of CBDcex- CBDcex was purified by batch affinity chromatography on CF1™ cellulose, then analyzed by SDS-PAGE. Lane 1, molecular weight markers. Lanes 2-4 are loaded with 12, 60 and 120 jig CBDcex, respectively. CD CO O o CD "5 E * 1500 1000 500 n 1 -j j .LajOQ>HfH / 2 „ ^n, 71 11 < ^ 3 JL-KobA-r • --• - 100 - 80 c CD 60 2 CD X Q. H40 0 20 40 60 80 Fraction Number - 20 0 100 0.8 0.6 E c o CO CM C CD 0.4 | 0.2 J 0 . 0 * Figure 3.7 Removal of xylan from purified CBDcex- A 12.5 mg sample of CBDcex in 20 mM trimethylamine/formic acid, pH 4 was loaded onto a MacroprepS Econocolumn (Bio-Rad). Sensitivity of the UV monitor was set at 2.0 absorbance units. The gradient was formed with 20 mM trimethylamine/ formic acid, pH 4 and pH 7. (O) carbohydrate detected by the phenol-sulfuric acid assay. ( • ) absorbance at 280 nm. (—) pH gradient. £280nm for CBDcex was 2.3 mL.mg-i.cnr1. The predicted 8280nm (Cantor & Schimmel, 1980) was 2.6 mL.mg-1.cnr1. The N-terminal amino acid sequence was ASSGP, in agreement with the predicted sequence (O'Neill et al., 1986). The polypeptide reacted with anti-Cex antiserum (Figure 3.5). 3.1.4 Binding of CBDcex An adsorption model describing the interaction of a large ligand (protein) with a lattice of overlapping potential binding sites (cellobiosyl units) was used to analyze binding of CBDcex to cellulose (Gilkes et al., 1992). The dimensions of CBDcenA (Pilz et al., 1990; Shen et al., 1991) are greater than the dimensions of the repeating cellobiosyl unit on the cellulose surface (Henrissat et al., 1988). This results in the CBD occupying several lattice units at any given time. If a binding site is larger than one lattice unit, the surface can be considered as an array of overlapping potential binding sites. The Langmuir adsorption isotherm models a single ligand interacting with only one receptor (Stuart & Ristroph, 1985; Steiner et al, 1988). This model is not valid for the CBD. A probability function must be included to find the concentration of available binding sites (Gilkes et al., 1992). The probability depends on both the concentration and configuration of proteins bound on the cellulose surface. This complication can be avoided if one only considers low concentrations of bound protein where the spacing is such that any two neighboring CBD molecules do not exclude the binding of a third CBD. The modified Langmuir equation (1) can be "linearized" by rearranging it in a double reciprocal form (2) which emphasizes data for the lower concentration range. bound [P] 1 bound [P] [NQ] • K a • free [P] 1 + a • K a * free[P] 1 1 + Ka- [N0] free[P] a [N0] (1) (2) where bound [P] is the concentration of bound protein (jamole.g cellulose-1), free [P] is the concentration of free protein (umolar), [N0] is the concentration of binding sites in the absence of protein (|nmole.g cellulose-1), Ka is the equilibrium association constant (L.|i.mole-1) and a is the number of lattice units occupied by a single protein molecule (jimole lattice unit.jxmole protein-1). Bound protein was determined from the difference between the initial protein concentration ([P0], JiM) and the equilibrium concentration of free protein. The adsorption isotherm was derived by plotting bound [P] vs. free [P]. The saturation value was estimated at the highest total protein concentrations. The absolute value of Ka cannot be determined unless [NQ] is known (Gilkes et al., 1992). However, a relative affinity value [(Ka • [No])"1; L.g cellulose-1] can be used to compare the affinities of various proteins for a given preparation of cellulose (i.e. when [N0] is constant). The relative affinity was determined from the initial slope of the double reciprocal plot (1/bound [P] vs. 1/free [P]), i.e. slope = (Ka • [N0])-1, at low protein concentrations. The binding of CBDcex to Avicel™, BMCC, a-chitin and regenerated cellulose was determined. Regenerated cellulose had the highest capacity for binding CBDcex (Figure 3.8). It becomes predominantly amorphous in nature after acid treatment. BMCC bound 5-fold more CBDcex than did Avicel™, suggesting that BMCC had significantly more surface area for adsorption. CBDcex bound most strongly to BMCC and least strongly to regenerated cellulose (Figure 3.9), indicating a tighter adsorption of the polypeptide to cellulose with a higher crystallinity. The relative affinity and saturation level of CBDcex adsorbed to a-chitin was about twice that of Avicel™. Adsorption to a-chitin suggests that CBDcex recognizes a common structural pattern in both cellulose and chitin. Two operating parameters of importance for immobilized enzyme columns are temperature and pH. CBDcex bound to BMCC over wide ranges of temperature and pH (Figures 3.10; 3.11). The relative affinities were not markedly influenced by temperature within the range tested, although a lower value was observed at 50°. The relative affinity increased with increasing pH. The highest value was observed at pH 9, around the isoelectric point of CBDcex-The saturation levels remained essentially the same. Detergents might block or destabilize the binding of a CBD fusion polypeptide. This would preclude the use of detergents in wash solutions or reaction mixtures. CBDcex adsorbed to and remained bound to cellulose in the presence of up to 2% Triton X-100, a non-ionic detergent (Figure 3.12). It adsorbed to and remained bound to cellulose in the presence of 0.2% SDS, an ionic detergent (Figure 3.12). Higher concentrations of SDS destabilized the binding. Avicel BMCC o> T> C "5 E TJ C =J £ 40 9 30h E =$. s - 20 10 - * — • — ^ 0 50 100 150 200 Free [P] U.M Chitin 10 20 Free [P] uM 30 4U £ 30 E a. 5T 2 0 1 1<> CQ g • q> "o E TJ C 3 0 10 20 30 40 Free[P] u.M Regenerated cellulose (RC) 100 200 Free[P] uM Substrate Avicel BMCC Chitin RC Saturation level (nmol.g-1) >3 >15 >7 >40 Figure 3.8 Adsorption of CBDcex to cellulose and a-chitin. Avicel™ (5 mg), BMCC (1 mg), chitin (1 mg) or regenerated cellulose (1 mg) was mixed with varying initial concentrations of CBDcex ([Pol = 0.9-180 uM for Avicel™ and [P0] = 1.8-300 JJM for BMCC, chitin and regenerated cellulose). Samples were treated as in 2.12. 46 Avicel BMCC 1/Free [P] nM" Chitin 10 1/Free [P] jiM 20 30 -1 1/Free [P] uM 1 Regenerated cellulose (RC) 0.20 0.00 0.00 0.05 0.10 1/Free [P] uM" 0.15 Substrate Avicel BMCC Chitin RC Relative affinity (L.g~1) 8 58 18 1 Figure 3.9 Relative affinities for the binding of CBDcex to cellulose and a-chitin. Details are given in the legend to Figure 3.8. o E a. CL 10 1 a, •o c 3 O m 10 20 30 40 50 Free[P] nM &^  10 10 20 30 40 50 Free[P] (iM 10 20 30 40 Free[P] (iM I -o c 3 O CD I •o c 3 o CD I •o c 3 o CD 10 1/Free[P] \M 20 30 1 2 4 1/Free[P] nM 6 8 10 -1 Temperature (°C) 22 40 50 Relative affinity (L.g-1) 58 58 18 Saturation level Oimol.g"1) >15 >20 >15 Figure 3.10 Adsorption of CBDcex to BMCC (1 mg) at 22°, 40° and 50°, pH 7. [P0] = 1.8-90 |iM. Details are given in 2.12. pH 3 CD P 3 -0. -o c 3 £0 'q> O fc 3 . t n. • o c 3 £ 1 q> 3 . C 3 zu 10 ' • 0 10 pH 7 10 n 0 10 pH 9 10 > / £ • 0 10 20 30 Free[P] jtM " • 20 30 Free[P] nM —— 5 20 30 Free[P] nM • 40 • 40 • 40 50 50 50 T -o E 3 . CO E" T3 C 3 o m '-,-o E a. CO Q. TJ C 3 O m ,-o E 3 . CO •o c 3 o m 0.50 0.00 1/Free [P] \iM 2 4 1/Free [P] nM 6 8 -1 2 4 1/Free [P] pM 6 8 1 10 10 PH 3 7 9 Relative affinity (L.g"1) 20 45 108 Saturation level (nmol.g"1) >10 >15 >15 Figure 3.11 Adsorption of CBDcex to BMCC (1 mg) at pH 3, 7 and 9, 22°. [P0] = 1.8-90 nM Details are given in 2.12. 49 A 1 2 3 4 5 6 7 8 9 10 © 100 CD CD Q o 2 o •*—* CD > CO CD OC 80 60 40 20 0 8 9 10 Lane Number Figure 3.12 The influence of detergents on the binding of CBDcex to cellulose. Details are given in 2.15.4. The percentage of gel used was 16%T. Protein loading per lane was approximately 1.8 nmoles. The detergent was either added after binding of CBDcex to cellulose (A, Q ) or present during binding of CBDcex to cellulose (B, S3). C, densitometric scan of protein bands. Lanes 1-5, CBDcex treated with Triton X-100 (0, 0.002%, 0.02%, 0.2% and 2%, respectively). Lanes 6-10, CBDcex treated with SDS (0, 0.002%, 0.02%, 0.2% and 2%, respectively). 3.1.5 CBDcex contains a disulfide bridge An indirect way of showing the presence of disulfide bonds in a protein is to analyze the migration of the protein in an SDS-polyacrylamide gel in the presence or absence of a reducing agent such as 6-mercaptoethanol (BME) or 1,4-dithiothreitol (DTT). Proteins that are not reduced, i.e. that have disulfide bridges, have a different hydrodynamic behavior and as a result appear to be smaller and therefore migrate further in an SDS-polyacrylamide gel (Creighton, 1989). There are 3 disulfide bridges in Cex, one of which is in the CBD (Gilkes et al., 1991a). CBDcex migrated more slowly in the presence than in the absence of BME (Figure 3.13), showing that the two cysteine residues in CBDcex formed a disulfide bridge. The failure of CBDcex to react with DTNB (Creighton, 1989) confirmed this (data not shown). This is consistent with the finding of Gilkes et al. (1991a). 3.1.6 Sensitivity of CBDcex to proteases Digestion of a protein with proteases of different specificities often provides information regarding its conformation (Price & Johnson, 1989). Proteolysis pinpoints specific residues that are exposed and susceptible to digestion. Denaturants and reducing agents can be added to monitor changes in protease susceptibility of the protein as a result of unfolding or the reduction of disulfide bridge(s). CBDcex purified by affinity chromatography on Avicel™ was resistant to cleavage by papain, endoproteinase Lys-C, endoproteinase Asp-N, trypsin, and a C. fimi serine protease. Trypsin hydrolyzed CBDcex, but only slowly (Figure 3.14). There are two possible trypsin cleavage sites in CBDcex : K28 and R68. Complete digestion should give fragments of 2.9, 4 and 4.1 kDa; partial digestion 51 kDa 1 2 3 4 5 6 7 -12 — Figure 3.13 The influence of B-mercaptoethanol (BME) on the mobility of CBDcex- The percentage of gel used was 15%T. Protein loading per lane was approximately 0.7 nmoles. Lane 1 is molecular weight markers. CBDcex was electrophoresed either in the presence (lanes 3-6) or absence (lanes 2, 7-9) of BME with no heat treatment (lanes 2 & 3) or incubated for 3 min at 37° (lanes 4 & 7), 60° (lanes 5 & 8) and 100° (lanes 6 & 9). Figure 3.14 Trypsin sensitivity of CBDcex A. Single time-point trypsin digestion of CBDcex- CBDcex (554 jxg; 50 nmoles) was dissolved in 123 \ih of 8 M urea in 0.4 M NaHC03 and 15 pL of 0.5 M DTT, and incubated at 50° for 15 min. The sample was diluted to 1000 HL with dH20 supplemented with 20 mM CaCl2. TPCK-treated trypsin (10 mg.mL-1 stock in 1 mM HC1) was added to a final weight ratio of 1:30 (trypsin:CBDcex). The solution was incubated at 37° for 2 h. The reaction was stopped by adding PMSF (10 mg.mL-1 stock in 100% isopropanol) to a final concentration of 100 jig.mL-1. Buffer (0.4 M NaHC03) replaced either urea or DTT, or both in the other tubes. Samples (each equivalent to 11 jig; 1 nmole) were analyzed by SDS-PAGE. Lanes 1-6 are + CBDcex Lane 1 is + urea, - DTT, + trypsin. Lane 2 is + urea, - DTT, - trypsin. Lane 3 is - urea, + DTT, + trypsin. Lane 4 is - urea, + DTT, - trypsin. Lane 5 is + urea, + DTT, + trypsin. Lane 6 is + urea, + DTT, - trypsin. Lane 7 is - CBDcex, + urea, + DTT, + trypsin. Lane 8 is + CBDcex, - urea, - DTT, - trypsin. Lane 9 is trypsin (10 jag). B . Time course digestion of CBDcex with trypsin. CBDcex was digested as above in the presence of urea and DTT. Numbers refer to the incubation time in min. A 1 2 3 4 5 6 7 8 9 B O T 1 0 1 5 2 0 3 0 4 5 BO 9 0 1 2 0 -53 could give fragments of 8.1, 6.9, 4.1, and 2.9 kDa. Since CBDcex was only cleaved in the presence of both urea and DTT, denaturation appears essential to expose the amino acid residue(s) recognized by trypsin. Analysis of CBDcex prepared for two-dimensional nuclear magnetic resonance spectroscopy revealed that the material purified on Avicel™ contained xylan (J. Carver, personal communication). The xylan associated with CBDcex (see 3.1.2) affected its sensitivity to pepsin. The xylan-free material was resistant to pepsin, whereas the contaminated material was rapidly degraded (Figure 3.15A & B). When the xylan-free material was readsorbed to Avicel™, desorbed with GdmCl and the GdmCl removed, it became susceptible to pepsin (Figure 3.15C). Apparently, the xylan contaminant picked up by CBDcex from Avicel™ alters the conformation of the polypeptide, making it susceptible to pepsin. Figure 3.15 Digestion of CBDcex with pepsin A. Avicel™-purified CBDcex (200 ng; 18 nmoles) was diluted to 200 \\L with 0.4% acetic acid, pH 3. Pepsin (5 mg.mL-1 stock in 1 mM HCl) was added to a final weight ratio of 1:5. The digest was incubated at 37°. Samples were removed at intervals and added to an equal volume of 1 M Tris.Cl buffer, pH 8 to stop the reaction. 10 \ig (1 nmole) from each sample were analyzed by SDS-PAGE. Pepsin was replaced with acetic acid buffer in the control digest. Lane 1, molecular weight markers; lanes 2 and 8, no pepsin control at 0 and 1 h, respectively; lanes 3-7 are pepsin digestion of CBDcex at 0, 5,15, 30 and 60 min, respectively. B . Details are as above except CBDcex was purified further by cation-exchange chromatography on MacroprepS beads (Bio-Rad). C. CBDcex from B was bound to Avicel™, desorbed with 8 M GdmCl and exchanged with dH20. The sample was treated as in A. Lane 1, molecular weight markers; lanes 2-4, 5-7 and 8-10, CBDcex in B, CBDcex in C, and CBDcex in A, respectively. Digestion for each sample was 0, 5 and 60 min, respectively. 56 A kOa 1 2 3 4 5 6 7 8 20— 14— C 1 2 3 4 5 6 7 8 9 10 3.2 A fusion polypeptide comprising CBDcex fused to the C-terminus of a B-glucosidase (Abg) from an Agrobacterium sp. (Abg-CBDcex; Figure 3.1) 3.2.1 Construction of plasmids expressing abg-CBDcex The model enzyme chosen for immobilization work comprised the 6-glucosidase Abg (Han & Srinivasan, 1969) fused at its C-terminus to the N-terminus of CBDcex- It was designated Abg-CBDcex- The first plasmid construct (pEOl) expressed the fusion protein Abg-CBDcexl- In this polypeptide 37 amino acids of the C-terminus of Cex catalytic domain and the Pro-Thr linker separated the fusion partners (Figure 3.16). Oligonucleotide-directed in vitro "loop out" mutagenesis was used to remove the linker, giving pTZE02 (Figure 3.17). In a third construct, pTZE03, the C-terminus of Abg was fused precisely to the N-terminus of CBDcex (Figure 3.18). Gene expression in all three plasmids was controlled by the lac promoter. The correctness of the constructs was confirmed by restriction endonuclease analysis and DNA sequencing. The fusion polypeptides expressed by pEOl, pTZE02 and pTZE03 were designated Abg-CBDcexl, Abg-CBDcex2 and Abg-CBDcex3, respectively. 3.2.2 Improved yield of Abg-CBDcexl in E. coli CAG456/pE01 E. coli CAG456 has an amber mutation in the htpR locus and a temperature-sensitive suppressor t-RNA; E. coli CAG440 is an htpR+ isogenic variant of CAG456 (Baker et al., 1984). The htpR locus encodes the sigma factor required for transcription of heat shock and other SOS genes (Grossman et al., 1984). Figure 3.16 Construction of pEOl. pUCl2-1.1 cex (PTIS) was digested completely with Seal and Ndel, and the 1.1 kbp fragment was isolated containing the CBDcex coding sequence. pABG5 (Wakarchuk et al, 1986) was first digested completely with Ndel then partially with Ncol. A 3.8 kbp fragment containing the whole vector sequence plus the sequence encoding all but the last six amino acids of Abg was isolated. The 1.1 kbp and 3.8 kbp fragments were ligated using an adapter which encoded the last six amino acids of Abg, to give pEOl. 59 BamHI EcoRI, sad Ndel Seal I Seal Ndel Isolate 1.1 kbp Scal-(cex CBD)-Ndel Ncol Ncol I Ndel Partial Ncol Isolate 3.8 kbp Ndel-(a£»g)-Ncol Ncol CAT GGG GTT GCC AAG GGG CC CAA CGG TTC CCC T4 DNA Ligase abg Hindlll Psti Figure 3.17 Construction of pTZEOl and pTZE02. pEOl was digested completely with Sad and PstI. Two fragments were isolated from the digest: a 1.7 kbp fragment containing abg, cex catalytic domain (CD), cex PT and part of CBDcex, and a 0.6 kbp fragment containing remainder of the CBDcex and cex 3' non-coding region. pTZ18R was digested with Sad and PstI, then ligated to the two fragments from pEOl to give pTZEOl. pTZE02 was made by replacing the sequence encoding the Pro-Thr linker in pTZEOl with an Nhel site using oligonucleotide-directed in vitro "loop-out" mutagenesis. pEOI + pTZ1 8R Sacl Pstl T4 DNA Ligase PT linker replaced by Nhel site using oligonucleotide-directed in vitro mutagenesis Figure 3.18 Construction of pTZE03. pTZE02 was digested completely with Nhel. The Nhel sites were filled-in using kPolI. The blunt-ended DNA was digested partially with Ncol. A 5.1 kbp fragment containing all of pTZE02 except the sequence encoding the last six amino acids of abg was isolated. An adapter (see Figure 3.16) was ligated to the 5.1 kbp fragment to give pTZE03. The Nhel site in pTZE02 was not restored in pTZE03. 63 Seal Nhel kPoll Partial Ncol Ncol i Ncol abg f 1 ori 7 + Ncol CAT GGG GTT GCC AAG GGG T CC CAA CGG TTC CCC.A H G V A K G abg T4 DNA Ligase cex CBD TCT AGC S S TCC GGT. S G Overproduction of foreign proteins by E. coli may cause stress to which the cells respond by producing proteases (e.g. Lon protease) which may degrade the overproduced proteins (Goldberg & Goff, 1986). Some proteins are unstable when produced in a wild type strain, but exhibit significantly longer half-lives when produced in an htpR strain (Baker et al., 1984). There was a 3-fold increase in total enzyme activity (to 298 pNPGase units) and a 4-fold increase in the specific activity of crude extracts (to 96 units.|4.moH) of Abg-CBDcexl when it was produced in CAG456 compared to JM109. The specific activities of Abg-CBDcexl produced in CAG440 and JM109 were comparable. The CAG456 cells also contained more of the intact fusion protein (Figure 3.19). The low molecular weight polypeptides (<30 K) in the cell extracts (4, 5 and 6 in Figure 3.19) represented non-specific antibody binding. The lower molecular weight polypeptides in Abg-CBDcexl (<55 K; 3 in Figure 3.19) were degradation products formed during storage of the enzyme (Ong et al., 1989a). 3.2.3 Screening of Abg-CBDcex The intact fusion polypeptides reacted with both anti-Abg and anti-Cex antisera (Figure 3.20). The degradation product(s) reacted only with anti-Abg antiserum indicating that these corresponded to Abg. The degradation products which were larger than native Abg were probably Abg with some amino acid residues from either the Cex catalytic domain or the CBDcex attached to its C-terminus, which did not react with the anti-Cex antiserum. 65 1 2 3 4 5 Figure 3.19 Abg-CBDcexl production by various strains of E. coli. Polypeptides were resolved by SDS-PAGE and detected by Western blotting. The primary antiserum used was rabbit anti-Cex. Triangle indicates Abg-CBDcexl- Lanes 2 and 3, cellulose-purified Cex and Abg-CBDcexl, respectively; lanes 1, 4, 5 and 6, equivalent amounts of protein from JM109, JM109/pEOl, CAG440/pEOl, and CAG456/pE01, respectively. 3 4 5 B 7 kDa —aio - l a o IIQfftli&qffa. —|^^gy^ jifcMBiMllfc AjjA^^ta^. ^^^^^V^ _ s22S 12 5F ^ ^p W i f • • * te« —sa — 5 3 « | 4 1 f? f"f 3 a 5 B 1 2 3 a 5 Figure 3.20 Integrity of Abg-CBDcex produced by various E. coli strains. Lane 7, molecular weight markers. Lanes 5 and 6, purified Abg-CBDcexl and Cex, respectively; lanes 1, 2, 3 and 4, equivalent amounts of cell extracts from JM101/pTZ18R, JM101/pTZEO3, JM101/pTZEO2, and JMlOl/pTZEOl, respectively. Primary antisera used were rabbit anti-Abg (B) and anti-Cex (C). 3.2.4 Large-scale purification of fusion polypeptides CBDcex fusion polypeptides could be purified virtually to homogeneity in a single step by adsorption to and desorption from cellulose. After removal of the nucleic acids from the crude cell extract, the fusion polypeptide was adsorbed to cellulose. Non-specifically adsorbed proteins were desorbed with 1 M NaCl in phosphate buffer. The fusion polypeptide was then desorbed with dH20. The rate of desorption became significant when the conductivity of the column effluent decreased to < 1 jamho (Figure 3.21). The nature of the cellulose influenced the stability of adsorption. Only 20% of the bound fusion polypeptide was desorbed from Cellufine™ compared to 70% from CF1™ cellulose (Figure 3.22). This phenomenon is discussed in 3.2.11 and section 4. A typical purification scheme using CF1™ cellulose is summarized in Table 3.1. Column elution is shown in Figure 3.23. Comparable results were obtained using this protocol for Abg-CBDcexl> 2 and 3 (Table 3.2). The desorption yield was between 58-70%. A 100- to 260-fold purification of the fusion polypeptides was achieved using affinity chromatography on cellulose. The specific activity of the purified Abg-CBDcexl was comparable to Abg that was purified using a combination of gel filtration and ion-exchange chromatography (K. Rupitz, personal communication). Large-scale production of CBDcex and Cbg-CBDcex is presented in 3.1.2 and 3.3.2, respectively. 3.2.5 Properties of the fusion polypeptides The Mrs (68, 66 and 62 K) for the 3 fusion polypeptides as determined by SDS-PAGE were very similar to the predicted molecular weights (O'Neill et al., 1986; Wakarchuk et al., 1988). The N-terminal amino acid sequences of intact Abg-CBDcexl and its degradation products were identical to that of mature Abg • E E o ZJ T3 C o O 300 200 100 S o E > 1 CD (A CO 55 O O _3 O) I 50 100 150 Fraction Number J 0 Figure 3.21 Purification of Abg-CBDcexl by affinity chromatography on CFl cellulose. Details are given in 2.5.2. vz, CF1 Cellulose I I 0.0 0.2 0.4 0.6 0.8 6-glucosidase activity 1.0 Figure 3.22 The influence of cellulose matrix on the recovery of Abg-CBDcexl- ( E3 ), recovered bound activity; ( U ), unrecovered bound activity; ( I )> unbound activity. Table 3.1 Purification of Abg-CBDcexl by affinity chromatography on CFl cellulose Purifi-cation stage* Volume (mL) Activity (Units. mL"1) Total activity (Units) [Protein] (mg.mL"1) Total [Protein] (mg) Specific activity (Units.mg"1) Overall yield (%) Desorption yield (%) Fold-purification A B C D E F 867 848 885 1,750 22 21.9 17.4 0.12 0.98 411.1 18,987 14,755 106 1,719 12,950 9,045 21.3 14.3 0.78 6.39 3.44 18,467 12,126 690 11,179 257 76 1.02 1.22 50.4 119.6 100 78 68 48 100 70 1.0 1.2 49 117.2 *A, French press cell extract; B, Streptomycin sulfate-treated and GF/C-filtered cell extract; C, Column flow through; D, High and low salt buffer washes; E, Cellulose-adsorbed polypeptide; F, d.H20-desorbed polypeptide after ultrafiltration 71 so aa 98 io2 io6 -no I T * n a 3 a kDa Figure 3.23 Column output of Abg-CBDcexl purified from CFl™ cellulose. Equal volume of each fraction (90-120) was analyzed by SDS-PAGE. 1, intact fusion polypeptide; 2, degradation product; 3, Abg-CBDcexl from a previous purification; 4, molecular weight markers. Table 3.2 Purification of fusion polypeptides Fusion Total Total Specific Fold- Yield polypeptide Step* activity protein activity purifi- (%) (Units) (mg) (U.mg-1) cation Abg-CBDcexl 1 12,947 10,701 1.21 1 100 2 9,045 76 119 98 70 Abg-CBDCex2 1 3,301 7,709 0.43 1 100 2 1,911 29 66 153 58 Abg-CBDCex3 1 7,256 15,773 0.46 1 100 2 4,590 38 121 262 63 "^Clarified cell extract prepared by rupture in a French pressure cell and precipitation of nucleic acids with streptomycin sulfate. 2Water-desorbed polypeptides using affinity chromatography on CF1™ cellulose and concentrated by ultrafiltration with an Amicon cell. (M/TDPN). A significant proportion retained the N-terminal methionine residue as found previously for the recombinant Abg (Wakarchuk et ah, 1988). The predicted pi for Abg-CBDcexl was 5.1 (Skoog & Wichman, 1986). The fusion polypeptides reacted with both anti-Abg and anti-Cex antisera (Figure 3.19). 3.2.6 Binding of Abg-CBDcexl to cellulose Abg is a dimeric protein (Day & Withers, 1986). Abg-CBDcex, like CenA and Cex (Gilkes et al., 1988; 1989), is susceptible to proteolysis between the two fusion partners. Analysis of purified Abg-CBDcexl by non-denaturing PAGE revealed three distinct species, only two of which bound to cellulose (Figure 3.24). All three species hydrolyzed MUG (Figure 3.25). One of the species binding to cellulose contained only the 68 kDa polypeptide, the other contained both the 68 and the 51 kDa polypeptides (Figure 3.25). The species unable to bind to cellulose contained only the 51 kDa polypeptide. This was probably formed by subunit exchange between the two types of cellulose-binding dimers during and after affinity purification on cellulose. Abg-CBDcexl> like Abg itself, forms dimers. Both homodimers (Abg-CBDcexl/Abg-CBDcexl) and heterodimers (Abg-CBDcexl/Abg) bind to cellulose. The observation that dimeric polypeptides can be adsorbed to cellulose if one or both of the monomers carry a CBD suggests that CBD fusion polypeptides might be used to purify the components of protein complexes by linking a constituent polypeptide to cellulose through a fused CBD. 3.2.7 Stability of fusion polypeptides Increased stability is one benefit of immobilizing an enzyme on a solid matrix (Hartmeier, 1988). Abg-CBDcexl adsorbed to Avicel™ was stable and remained bound at 4° and 37° for up to 7 weeks. In solution, in contrast, the fusion polypeptide was stable at 4° but not at 37° (Figure 3.26). Clearly, binding to cellulose protects the fusion polypeptide from inactivation. 3.2.8 Enzyme kinetics Fusion of an enzyme to a CBD might affect its catalytic activity. The specific activities (jxmol pNP.mur1.|j,mol protein-1) for the purified fusion polypeptides were 8,296 for Abg-CBDCexl; 4,356 for Abg-CBDCex2 and 7,502 for Abg-CBDcex3 compared to 7,344 for recombinant Abg. Therefore, fusion of Abg to c-Figure 3.24 Heterogeneity of purified Abg-CBDcexl- Purified Abg-CBDcexl and the fraction of it binding to Avicel™ were analyzed by non-denaturing PAGE using a 7.5%T gel. Lanes: 1, purified Abg-CBDcexl; 2, fraction binding to Avicel™. A, intact fusion homodimer; B, heterodimer; C, Abg homodimer. 1 2 3 4 kDa 5 6 7 8 u H 68 -5 1 -Figure 3.25 Detection of heterodimers in purified Abg-CBDcex- Abg and Abg-CBDcexl were subjected to non-denaturing PAGE (lanes 1 and 2). Enzymatically active bands were detected by applying MUG to the gel, then examining it under long-wavelength UV light (lanes 3 and 4). Active bands were cut out and incubated ovemight at 4° in 20 u,L phosphate buffer. Then 45 jxL 2x SDS-loading dye was added and the sample agitated on a vortex mixer. After centrifugation at 14,800g for 10 min at room temperature, the supematants were analyzed by SDS-PAGE (lanes 5-8). Lanes 1, 3 and 5, purified Abg; 2 and 4, Abg-CBDcexl; 6, lane 4 upper band; 7, lane 4 middle band; 8, lane 4 lower band. A B Immobilized fusion Soluble fusion 0.0 0.0 2.0 4.0 6.0 Time (weeks) 8.0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Figure 3.26 Stability of Abg-CBDcexl adsorbed to Avicel. A. Residual 6-glucosidase activity of bound Abg-CBDcexl expressed relative to initial activity at week 0. B. SDS-PAGE analysis of adsorbed Abg-CBDcexl. Numbers refer to time in weeks. Letters refer to different sets of samples: A, bound fusion at 4°; B, soluble fusion at 4°; C, bound fusion at 37°; D, soluble fusion at 37°. Experimental details are given in 2.15.3. 77 part of the Cex catalytic domain, Pro-Thr linker and the CBD (Abg-CBDcexl) or to the CBD directly (Abg-CBDcex3) did not alter significantly the activity of Abg. However, when the Pro-Thr linker was removed leaving Abg joined to CBDcex by 37 amino acids of the C-terminus of the catalytic domain of Cex (Abg-CBDcex2), the specific activity was decreased by almost 40%. Presumably this fusion polypeptide was unable to form a fully functional conformation. Abg-CBDcex2 was also unstable, being completely degraded to Abg and CBDcex during 6 months storage at 4° in the presence of 0.02% NaN3 and protease inhibitors (data not shown). It was not studied further. The KCat values for the fusion polypeptides (Table 3.3) were very similar; thus the catalytic activity of Abg on pNPG was not affected significantly by the fusion of CBDcex to its C-terminus. The Km values for Abg and Abg-CBDcexl were almost identical, whereas the Km for Abg-CBDcex3 increased by about 3-fold. Table 3.3 Catalytic activities of Abg and the fusion polypeptides Enzyme Km Kcat KcatlKm (HM pNPG) (s-1) (nM-l.s-1) Abg 75 125 1.67 Abg-CBDcexl 70 143 2.04 Abg-CBDCex3 202 145 0.72 Immobilization of an enzyme may alter its catalytic activity (Hartmeier, 1988). The catalytic activity of Abg-CBDcexl adsorbed to cellulose was examined. The specific mechanism-based inhibitor 2F-DNPG has been used to titrate the active sites of retaining glycosidases such as Abg and Cex (Withers et al, 1987; Withers et al, 1988; Tull et al, 1991). The method is particularly useful for immobilized enzymes since it allows determination of Vmax of the active, bound enzymes. Kcat for immobilized Abg-CBDcexl was 113 s_1. As this value was similar to the Kcat values of the enzymes in solution (Table 3.3), immobilization appeared not to affect catalytic activity. Km for immobilized Abg-CBDcexl in stirred suspension was 296 |LiM pNPG. This was 4-fold higher than Km of the soluble enzyme suggesting that the transport of substrate to the active site of the immobilized enzyme was rate limiting, presumably due to external mass transfer resistance. Ideally, the operation of an immobilized column should follow a plug-flow velocity profile wherein every element in the solution moves through the column at the same velocity (Lilly et al, 1966). Under appropriate conditions, a column operated in a plug-flow would maximize catalytic conversion. The flow pattern of a reactor can be characterized from a plot of column product against effluent volume or residence time (Kohlwey & Cheryan, 1981). Plug flow is characterized by an abrupt change in the effluent after a given time equal to the residence time of the solution in the column. A deviation from this, like a gradual change of effluent over time, indicates significant relative fluid motion and mixing (Lilly et al, 1966). The time distribution plot showed that the membrane column of immobilized Abg-CBDcexl did not follow a plug-flow profile (Figure 3.27). This indicates that there was considerable variation in the perfusion velocity resulting in the differences in time distribution and the accessibility of substrates to or transport of products from the active site of the immobilized fusion polypeptide. E c LO o •<fr < CD > "•£< CD DC 1.0 0.8 0.6 0.4 0.2 0.0 -— -• ~ i ! • • • • • • • • • • • • • • • • • • • » • • 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Time (min) Figure 3.27 Column flow characteristics of pNP ( • ) produced from an immobilized Abg-CBDcexl membrane column. Enzyme loading was 10 nmoles for 5 cellulose acetate membranes. Row rate was 60 mL.rr1. Temperature was 22°. (—), ideal plug-flow reactor. The apparent Km of Abg-CBDcexl adsorbed to cellulose was determined in a continuous column reactor. An integrated Michaelis-Menten equation was used to express the total reaction occurring in the column (Lilly et al., 1966; Ford et al, 1972; Patwardhan & Karanth, 1982). The apparent Km was 7-fold higher than that observed in a stirred suspension (data not shown), as would be expected with mass transfer limitation. 3.2.9 Immobilization yields for the fusion polypeptides Immobilization yield is the percentage of the actual activity retained by the enzyme when immobilized. At near saturation of the cellulose the immobilization yield was about 30% for Abg-CBDcexl> and about 10% for Abg-CBDcex3. These values were higher when the Avicel™ was not saturated with enzyme (Figure 3.28). Abg did not adsorb to cellulose (data not shown). At least in this case the immobilization yield was improved by the presence of an extra 57 amino acids between the affinity "tag" and the fusion partner, presumably by extending the Abg portion of the fusion polypeptide into the solution. The decrease in immobilization yield as the amount of bound enzyme increased may be a result of steric hindrance at higher levels of immobilized polypeptide. However, the lower immobilization yield of Abg-CBDcex3 (lacking the 57 extra amino acids) at the same enzyme loading level as Abg-CBDcexl (Figure 3.28) suggests that the proximity of the Abg to the cellulose surface may be significant. The saturation levels were >2,000 U.g Avicel™"1 (0.24 umoles protein.g Avicel™-!) for Abg-CBDCexl, and >3,500 U.g Avicel™-! (0.49 |imoles protein.g Avicel™-!) for Abg-CBDcex3 (Figure 3.28). The level of actual 4000 £ 'a £ o 8 -2 2000 •o c o D ) O 0.40 10 20 30 40 50 Free activity U.mL"' 0.00 0.0 O Abg-CBDCex1 • Abg-CBDCex3 1.0 2.0 3.0 Free[P] |iM 2 © c o N j Q O E E 100 Free activity U.mL" 1.0 2.0 3.0 Free[P] uM Figure 3.28 Adsorption equilibria (top) and immobilization yields (bottom) for Abg-CBDcexl & 3 adsorbed to Avicel. Initial enzyme activity = 6-86 U.mL-l (Abg-CBDcexl) and 6-81 U.mL-1 (Abg-CBDcex3). [P0] = 0.7-10.3 jiM (Abg-CBDcexl) and 0.8-11.3 JIM (Abg-CBDcex3). The bound fusion protein was assayed for 6-glucosidase activity under standard conditions (M). The theoretical bound activity (T) was the difference between the activity added initially and the activity in the supernatant. Immobilization yield was (M/T)100. bound 6-glucosidase activity exhibits an inverse relationship to the size of the bound polypeptide. 3.2.10 Stabilities of the immobilized enzymes There are two aspects to the stability of an immobilized enzyme: its intrinsic stability and loss from the matrix. The latter was of particular interest for Abg-CBDcexl because it was immobilized by simple adsorption to the cellulose surface. Abg-CBDcexl was inactivated by prolonged incubation at or below pH 5.0, either in solution or adsorbed to cellulose (Figure 3.29A), but desorption was insignificant (Figure 3.29B). However, the enzyme was desorbed at pH 8-9 (Figure 3.29B). In solution it remained active at pH 7.5-9, but it was inactivated above pH 9. The pH stability of Abg-CBDcexl in solution was similar to that reported for Abg (Han and Srinivasan, 1969). As expected, Abg-CBDcexl was desorbed from a cellulose column by a pH gradient from 7-12, but not by one from 7-2 (Figure 3.30). The enzyme desorbed at >pH 9 was inactive. Adsorbed Abg-CBDcexl remained active and bound to cellulose during prolonged incubation at pH 6 to 7. Trichoderma reesei cellulase also desorbed at alkali pH (Otter et at., 1989). Abg-CBDcexl in solution was stable for 24 h at 25°, pH 7.0 and ionic strengths from 0 to 1 M (Figure 3.31A). Immobilized Abg-CBDcexl was also stable under these conditions but was largely desorbed from the cellulose in dH20 (Figure 3.31 A). Desorption of a column containing 70% of fusion polypeptide bound to cellulose with a linear NaCl gradient (0-1 M) confirmed this (Figure 3.3IB). No desorption of the fusion polypeptide was observed indicating that the binding of Abg-CBDcexl was stable up to 1 M NaCl. A >% +-» > o < CD W cti • a w o o =3 D ) CC CD > - 1 — ' CO CD cc 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1 1 I I O Soluble fusion E3 Immobilized fusion 6.5 7.5 8 SL 10 11 PH PH 4 5 6 7 8 9 10 11 ; • • I B • • • • : — 2 l O — 1 2 D — 9 7 - B 8 — 5 8 — 5 3 kDa Figure 3.29 Stability of immobilized Abg-CBDcexl as a function of pH. A. Activity. Residual 6-glucosidase activity of soluble Abg-CBDcexl and adsorbed Abg-CBDcexl assayed under standard conditions and expressed relative to the value obtained at pH 7. B. Binding. SDS-PAGE analysis of adsorbed Abg-CBDcexl after incubation at different pH values. Arrow refers to Abg-CBDcexl- Last lane is molecular weight markers. Experimental details are given in 2.15.1. 12.0 -10.0 -IE 2.0 -20 40 60 Fraction Number o o x o CO CM < Figure 3.30 Desorption of Abg-CBDcexl from cellulose with a pH gradient. Open circles indicate absorption at 280 nm. Closed circles indicate pH. Experimental details are given in 2.15.1. 85 > •*—• o < CD .> • I — • CD c CD o td E23 Immobilized • Soluble 100 80 60 40 20 0 i I I r i i i i M I i I 1 10 20 50 100 200 Ionic Strength (mM) 500 1000 20 30 40 50 Fraction Number 1 o o ' »o.oo 60 70 Figure 3.31 The influence of ionic strength on the stability of Abg-CBDcexl. A. Activity. Residual 6-glucosidase activity of soluble Abg-CBDcexl and adsorbed Abg-CBDcexl assayed under standard conditions and expressed relative to the value obtained at 50 mM NaCl in phosphate buffer. B. Binding. Stability of Abg-CBDcexl in a salt gradient. Closed circles refer to A280nm readings. Line is the salt gradient (0-1 M NaCl). Experimental details are given in 2.15.2. 3.2.11 Cellulose matrices Since CBDcex adsorbed more strongly to crystalline than to amorphous cellulose (section 3.1.4), three types of cellulose were tested as matrices for Abg-CBDcex immobilization: CFl™ cellulose, cotton and cellulose acetate membrane. Abg-CBDcexl was adsorbed to CFl™ cellulose, which was then packed into a column and equilibrated at 4°. A solution of pNPG was perfused through the column, and the amount of pNP released in the effluent was monitored. The conversion of pNPG to pNP dropped from 70% initially to 25% during the first 13 days of operation, but remained nearly constant during the following 8 days (Figure 3.32). When adsorbed to cotton fibres, enzyme activity decreased only 12% in the 15 day period (Figure 3.32). This suggested that, in the case of Abg-CBDcexl adsorbed on the CFl™ cellulose, the enzyme bound to the amorphous regions of the cellulose was released gradually during the first 13 days of operation of the column. The enzyme tightly bound to the crystalline regions of the CFl™ cellulose was not released and accounted for the stably bound activity. When Abg-CBDcexl was adsorbed to stacked cellulose acetate membranes, activity remained fairly constant throughout 14 days of continuous operation of the column (Figure 3.32). The substrate conversion rate was only 30% compared to 70% for cotton. This could be explained by the shorter residence time (4.8 min) in the cellulose acetate membrane column compared to 40 min in the column packed with either cotton or CFl™ cellulose. Abg-CBDcexl adsorbed to cotton was stable for at least 10 days continuous operation at 4°, 10° and 37° (Figure 3.33). It was inactivated completely in 3 days at 50°, but protein was not released from the column. Abg-CBDcexl in solution was inactivated after 3 days at 50°. The activity of adsorbed Abg-CBDcexl was highest at temperatures between 20° and 30° (Table 100 CF1 cellulose * - 100 o 80 60 -40 -20 -100 a. o a. o c g > c o O o c o .e a. o Q. o •— e g "w | 0 * c 0 o o • Cotton o o - o I • Cellulose acetate 8 10 12 14 16 18 20 Time (days) Figure 3.32 Performance of Abg-CBDcexl immobilized enzyme columns: effect of cellulose structure. Amount of cellulose used was 1 g (CFl™ cellulose or cotton) or 5 cellulose acetate membranes (0.2 mm x 25 mm diameter). Enzyme loading was from 0.9-2.3 nmoles protein.g"1 cellulose or 3 nmoles protein.membrane"1. Temperature of column run was 4° (CFl™ or cotton) or 22° (membrane). Row rate was 7.8 mL.rr1. Phosphate buffer only was passed through the CFl™ cellulose column from days 6 to 12. Further details are given in 2.16. >» • > •G cm CD .> CD DC 1.0 0.5 0.0 » '*9l%V^Bl,^%?*%I^^0«>000*>oooo0 tf "OOOOodPOooOoOOOOOOOlo CO 1.0 0.5 0.0 1.0 0.5 0.0 1.0 0.5 -0.0 A A 10* &V^%w^ 37c o o •o»%-^ O O ro o > a. • 50° 4 6 8 Time (days) 10 12 Figure 3.33 Performance of Abg-CBDcexl immobilized enzyme columns: temperature stability. Amount of cotton used was 2 g. Enzyme loading was from 3.7-22 nmoles protein.g-1 cellulose. Row rate was 7.8 mL.lT1. Values are expressed relative to the initial, highest p-nitrophenol produced at that temperature. Further details are given in 2.16. 3.4). Conversion of pNPG to pNP was approximately 50% even after 20 h of operation. Although Abg-CBDcexl remained bound at 50°, the Abg domain was inactivated. Table 3.4 Activity of immobilized Abg-CBDcexl at different temperatures Temperature Percent conversion to pNP °C 5h 20h 4 30 30 10 38 38 20 50 50 30 48 48 40 45 25 Amount of cotton used was 2 g. Enzyme loading was 0.013 |imoles protein.g"1 cotton. Flow rate was 7.8 mL.h"1. 3.3 A fusion polypeptide comprising CBDcex fused to the C-terminus of a fi-glucosidase (Cbg) from C. saccharolyticum (Cbg-CBDcex; Figure 3.1) 3.3.1 Construction of a plasmid expressing cbg-CBDCex C. saccharolyticum is a thermophilic, cellulolytic bacterium (Love & Streiff, 1987). It produces a fi-glucosidase (Cbg) with a temperature optimum for hydrolysis of 70° (Love & Streiff, 1987). Furthermore, it shares 39% sequence identity with Abg and is virtually the same size. These properties emphasized the validity of comparing the two enzymes as fusion polypeptides with CBDcex-Cbg was fused at its C-terminus to CBDcex to examine its efficiency as an immobilized enzyme at temperatures at which Abg was inactivated. Construction of the plasmid expressing cbg-CBDcex involved the introduction of two Nhel sites at the sequences corresponding to the N- and C-termini of the mature Cbg. The 5' end of the cbg was ligated to the sequence encoding the Cex leader peptide. The sequence encoding CBDcex (without the Pro-Thr linker sequence) was ligated to the 3' end of cbg just before the stop codon. The gene fusion cbg-CBDcex was expressed from the lac promoter with a consensus Shine-Dalgarno sequence located just upstream of the ATG of the coding sequence for the Cex leader peptide (Figure 3.34). The correctness of the construct was confirmed by restriction endonuclease analysis and DNA sequencing. 3.3.2 Purification of Cbg-CBDcex The binding of Abg-CBDcexl to cellulose was stable at 50° (section 3.2.11) and Cbg was known to be stable from 50°-70° (Love & Streiff, 1987). This made it likely that heat treatment could be used as an initial step in the purification of Figure 3.34 Construction of pTZEO 10. The Hindin fragment of pNZ1070 (Love et al., 1988), containing the cbg coding sequence, was cloned into the Hindlll site of pTZ18R. Nhel sites were introduced just after the ATG and just before the TAA codons of cbg by in vitro mutagenesis. The resulting plasmid, pTZlSR-cbg (Nhel), was digested completely with Nhel and a 1.4 kbp fragment containing the cbg sequence was isolated. pTZE07 (PTIS) was digested completely with Nhel and ligated to the 1.4 kbp cbg fragment to give pTZEOlO. 92 Nhel sites introduced by in vitro mutagenesis Nhel Nhel cbg 1.4 kbp N A S AATGCTAGC Nhal Nhel Linearized plasmid T 4 DNA Ligase BamHI Cbg-CBDcex- Much of the protein in a crude cell extract was removed by heating at 50° for 30 min (Figure 3.35), without loss of Cbg-CBDcex activity. Cbg-CBDcex was recovered from the treated, clarified extract by adsorption to cellulose (Figure 3.35). Like Abg-CBDcex> Cbg-CBDcex was sensitive to proteolysis. The polypeptide reacting with anti-Cex antiserum had a similar Mr to that of Abg-CBDcex3, about 62 K. Cbg-CBDcex could be desorbed from CF1™ cellulose either with dH20 or 8 M GdmCl. However, when dH20 was used, further Cbg-CBDcex could be released with GdmCl (Figure 3.36). The polypeptide desorbed with GdmCl comprised only the intact fusion polypeptide, whereas that desorbed by dH20 also included degradation products (Figure 3.36). This suggested that adsorption of the intact homodimers (assuming that Cbg by analogy to Abg is also a dimer) was stronger than that of the heterodimers. 3.3.3 Properties of Cbg-CBDcex The Mr of the intact fusion polypeptide was 64 K, in good agreement with that calculated from the predicted protein sequence (O'Neill et ah, 1986; Love et ah, 1988). The predicted pi was 5.6 (Skoog & Wichman, 1986). The N-terminal amino acid sequences of the Cbg-CBDcex desorbed with dH20 or GdmCl, and of the degradation product were identical to that of Cbg (ASFPK). The intact fusion protein (both dH20- and GdmCl-desorbed) was recognized by the anti-Cex antiserum, whereas the degradation product was not (Figure 3.35). Thus Cbg-CBDcex appeared to be sensitive to proteolysis between the two domains. As with Abg-CBDcexl/Abg heterodimers, the resulting Cbg was purified as a heterodimer with Cbg-CBDcex-94 A 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 Figure 3.35 Purification of Cbg-CBDcex- Crude cell extracts were prepared as in Figure 3.5. Cell extracts were then heated to 50° or 70° for 30 min followed by centrifugation. The treated cell extract was allowed to bind to CF1™ cellulose for 2 h at 4°. Samples were treated as in Figure 3.5. Equivalent volumes were analyzed by SDS-PAGE (A). Lane 1, purified Abg-CBDcex3. Lane 2, JM101/pTZ18R cell extract. Lane 3, culture medium. Lanes 4, 6 & 8, unbound cell extracts incubated at 4°, 50° and 70°, respectively. Untreated (4°) and heat-treated (50° & 70°) CF1™ cellulose-bound proteins, lanes 5, 7 & 9, respectively. Arrow refers to the intact Cbg-CBDcex- B, Western blot of A. Primary antiserum used was rabbit anti-Cex. 95 Figure 3.36 Purification of Cbg-CBDcex- Lane 1, molecular weight markers; lane 2, crude cell extract after rupture of cells with a French press; lane 3, crude cell extract after treatment with streptomycin sulfate and after ultracentrifugation; lanes 4 & 5, high and low salt buffer washes, respectively; lanes 6 & 7, dH20-desorbed before and after heat treatment at 50° for 30 min, respectively; lane 8, GdmCl-desorbed fusion polypeptide. The fusion polypeptide was most active at 70° (192 pNPGase U.mL-1) compared to 49.9 pNPGase U.mL-1 at 50°. The specific activities of the dH20-and GdmCl-desorbed fusion polypeptides were 7,785 U.|amoH and 13,323 U.|imoH, respectively. The almost 2-fold difference in specific activities could be the result of inactivation of a significant proportion of the dH20-desorbed preparation. Cbg-CBDcex bound stably to CF1™ cellulose at 70° (Figure 3.37). Densitometric scan of the adsorbed fusion polypeptide incubated overnight at 70° showed a decrease of only 8% compared to the 4° control. Attempts to desorb the fusion polypeptide from CF1™ cellulose with dH20 or high pH after incubation at 70° for 24 h were unsuccessful (data not shown). Most of the activity remained in the cellulose. Only a partial recovery of activity was obtained when GdmCl was used. This suggests that the CBD part of the fusion polypeptide may have become "fixed" on the cellulose surface as a result of its thermal denaturation. 3.3.4 Cbg-CBDcex as an immobilized enzyme Cbg-CBDcex adsorbed to cellulose acetate membranes was stable during continuous operation at 70° for more than 70 h (Figure 3.38). 97 1 2 3 Figure 3.37 Binding of Cbg-CBDCex to cellulose. Cbg-CBDCex (0.32 nmoles) was added to 1 mg BMCC and allowed to bind overnight at either 4° or 70°. The samples were treated as in Figure 3.5. Lane 1, Cbg-CBDcex before binding to BMCC. Lanes 2 & 3, Cbg-CBDcex left in the cellulose pellet after incubation at 4° and 70°, respectively. 98 1.0 | > "ft 0.8 -< CD O) 0.4 -• CO CD 1 0.2 -CD DC 0 20 40 60 80 Time (h) Figure 3.38 Temperature stability of Cbg-CBDcex immobilized enzyme column. Cellulose used was 5 stacked cellulose acetate membranes. Enzyme loading was 3.9 nmoles for 5 membranes. Flow rate was 8.8 mL.lr1. Temperature of column run was 70°. Further details are given in 2.16. 4 . D I S C U S S I O N The CBD of Cex was used to produce hybrid polypeptides for the purpose of enzyme immobilization. The high affinity of CBDcex for both native and modified cellulose allowed the strong and specific adsorption of fusion polypeptides without detectable leaching in an immobilized enzyme column. This work also demonstrated that CBDcex can be used as a "tag" for affinity chromatography on cellulose, thereby facilitating the purification of fusion polypeptides. 6-glucosidase was chosen as a reporter enzyme for enzyme immobilization for the following reasons: (1) the genes for a mesophilic 6-glucosidase (Abg) from an Agrobacterium sp. and a thermophilic 6-glucosidase (Cbg) from C. saccharolyticum were available, and both had been sequenced and expressed in E. coli (Wakarchuk et al., 1986; Wakarchuk et al, 1988; Love et al, 1988); (2) the enzymes had been characterized (Han & Srinivasan, 1969; Day & Withers, 1986; Love & Streiff, 1987); (3) a simple spectrophotometric assay using /7-nitrophenol-6-D-glucopyranoside as substrate was available to monitor activity. Recombinant Abg does not have a leader peptide recognized by the E. coli leader peptidase, therefore Abg was found predominantly in the cytoplasm of the cells (Wakarchuk et al., 1986). As with Abg, Abg-CBDcex was found in the cytoplasm of E. coli. Both CBDcex and Cbg-CBDcex had the Cex leader peptide at their N-termini; this allowed export of the polypeptide to the periplasm of E. coli. The production yield of CBDcex in E. coli was > 6% total cell protein, resulting in its accumulation in the periplasm and subsequent leakage into the culture supernatant. Non-specific protein leakage into the culture supernatant in E. coli is not uncommon (Gilkes et al., 1984; Guo et al., 1988; Shen et al., 1991), but the exact mechanism is not clear. Lysis of cells was ruled out since cytoplasmic enzyme markers (e.g., glucose-6-phosphate dehydrogenase) were not detected in the culture supernatant. However, significant levels of the periplasmic enzyme marker 6-lactamase were detected in the culture supernatant. It appears that the accumulation of polypeptides in the periplasm alters the integrity of the outer membrane, leading to non-specific leakage of both recombinant polypeptides. Unlike CBDcex, Cbg-CBDcex did not leak into the culture supernatant, probably because of its low synthesis level (< 0.05% of total cell protein). Abg-CBDcex and Cbg-CBDcex were therefore isolated from whole cell extracts. Affinity chromatography is one of the most powerful procedures for purifying proteins (Ostrove, 1990). This procedure exploits the specific binding properties of the target protein. Unlike other chromatographic or filtration techniques, affinity chromatography relies on the highly specific interaction of biomolecules. Properties such as shape, conformational changes or recognition of certain regions in the molecule have all been used to effect separation. Different eluents (e.g. substrate analogs, chaotropic agents, and changes in pH, temperature or salt concentration) have been used to desorb the desired molecule from the matrix, but in some cases, irreversible denaturation of the target molecule results in poor yield. Affinity chromatography usually requires the use of ligands coupled chemically to matrices. As with similar supports used for immobilizing enzymes, the principal disadvantage is the expense of preparing such specialized matrices. The strategy employed for purifying polypeptides in this study utilized inexpensive and readily available cellulosic matrices, and water for desorption of the fusion polypeptides. Like the intact exoglucanase Cex (Gilkes et al., 1988), the fusion polypeptides were desorbed from cellulose with water with a recovery varying from 58-70%. CBDcex, however, was not desorbed with water but required 8 M GdmCl, with a recovery >90%. Surprisingly, in view of these differences in conditions for desorption, the relative affinities of Abg-CBDcexl (7-8 L.g-1) and CBDcex (7.5 L.g-1) for Avicel™ were similar. This suggests that the size of the polypeptide may influence the packing on the cellulose surface. The smaller CBDcex could pack more tightly, hindering the access of water molecules for desorption. Whereas for the larger hybrid polypeptides, the fusion partner might sterically block binding, resulting in lower and less dense surface packing (e.g. > 0.6 jimoles Abg-CBDcexl.g Avicel™-1 compared to > 3 (xmoles CBDcex-g Avicel™-1). This in turn might facilitate the entry of water molecules to desorb the polypeptide much more readily. Cex is a monomer (Gilkes, unpublished results), whereas Abg (Day & Withers, 1986) and presumably Cbg are both dimers. The inability of water to completely desorb the fusion polypeptides from cellulose can also be attributed to the dimeric nature of the fusion partner. At least for Cbg-CBDcex the fusion polypeptides desorbed with water consisted of heterodimers with one partner bearing a CBD, whereas the unrecovered fusion polypeptides consisted of homodimers in which each partner had a CBD (Figure 3.36). It seems, therefore, that the homodimers bind more strongly to cellulose than the heterodimers because of the two-fold increase in the number of hydrogen bonds and van der Waals interaction between the homodimers and the cellulose matrix. The different saturation levels of CBDcex to a number of substrates (Figure 3.8) relates at least in part to the difference in available surface area. Heterogeneity of the substrate structure may also play a role since the various cellulosic substrates employed differ in the extent and nature of their crystallinity. Regenerated cellulose is thought to be predominantly cellulose II (Figure 1.2; Blackwell, 1981; Sarko, 1986), although other workers have concluded that it is amorphous (Lee et al., 1982; Ooshima et al., 1983). The regeneration also results in a reduction in particle size and consequent increase in surface area (Lee et al., 1982; Ooshima et al, 1983). BMCC consists of uniform bundles of microfibrils, 20-50 nm wide and several |im long (Henrissat & Chanzy, 1986; Ross et al, 1991). The microfibrils are crystalline and consist of cellulose I. Avicel™, on the other hand, is a relatively heterogeneous cellulose preparation made from partially acid hydrolyzed wood fibres containing both crystalline and amorphous components (Ooshima et al., 1983). It is a mixture of rod- and irregularly-shaped particles/aggregates with varying intraparticle pores (1-10 nm) and interparticle voids (> 5 um) (Marshall & Sixsmith, 1974). CBDcex was also shown to adsorb to a-chitin. Chitin is a polymer of N-acetylglucosamine residues linked by 6-1,4-glycosidic linkages much like cellulose. Regenerated chitin (or a-chitin) forms an anti-parallel structure similar to cellulose II (Blackwell, 1981). The adsorption of CBDcex to chitin suggests a common recognition motif in both types of substrates. Adsorption is obviously not blocked by the 2-acetamido substitution of the glucosyl residues. The CBDcex fusion polypeptides used for enzyme immobilization also bound to other cellulosic matrices (sections 3.2.4, 3.2.11, and 3.3.4). Cotton is one of the purest naturally occurring celluloses. It has a high crystallinity of 70-84% (Lee et al, 1982; Wood, 1988). CF1™ ceUulose is fibrous in nature with an average fibre length of 200 |im and diameter of 13.5 |J.m [E. Heilweil (Whatman, Inc.), personal communication]. It is made from highly purified cotton and has an oc-cellulose content of at least 98%. CF1™ cellulose is therefore free of hemicelluloses and other polysaccharides. Its manufacture does not involve either acid or alkali swelling and regeneration, and is therefore characterized as cellulose I. The crystallinity of CF1™ cellulose is estimated at 75-85%. It is the preferred support for chromatographic application because of its good packing and flow rate characteristics. Cellulose acetate is made from wood pulp prepared with acids or organic solvents followed by acetylation and partial deacetylation. The pulping conditions generally hydrolyze amorphous cellulose, resulting in a high cc-cellulose content (> 90%) (Klausmeier, 1986). Depending on the manufacturing process, cellulose acetate contains regions of varying crystallinity. Cellufine™ consists of microcrystalline particles of 45-105 Jim diameter, and is produced from cellulose triacetate after removal of the ester groups [D. Boyd (Amicon), personal communication]. It is evident that the nature of the substrate affects the reversibility of adsorption. The reasons for this are not immediately apparent but may be related to the crystallinity of, and available surface area on, the cellulose. For example, CF1™ cellulose allowed good recovery of CBDcex fusion polypeptides, and regenerated cellulose showed a very high capacity for binding CBDcex- By contrast, the use of a more crystalline cellulose (e.g. cotton; Cellufine™; BMCC; cellulose acetate) resulted in stable immobilization of the CBDcex fusion polypeptides. Protein purification by affinity chromatography requires reversible binding to the matrix, whereas for enzyme immobilization, a tight and irreversible binding is usually preferred. It is an advantage to have a choice of cellulosic supports with different affinities for CBDcex for particular biotechnological applications. The saturation levels of CBDcex bound to BMCC (Figures 3.10) did not vary appreciably over the temperatures tested. The relative affinity values however did show a marked decrease at 50° (Figure 3.11), suggesting that a portion of the polypeptides could be denatured and therefore did not bind to BMCC. The portion of polypeptides that did bind probably adsorbed during the first few seconds of incubation before the temperature reached 50°, and remained bound subsequently. This is supported by the finding that CBDcenA adsorbed to BMCC almost immediately at 30° (Gilkes et al, 1992). More importantly, this also demonstrates that CBDcex could still be used for enzyme immobilization under operating conditions extending well beyond those normally encountered. Although Abg was inactivated at 50°, the fusion polypeptide remained bound to cellulose. Moreover, the stable and continuous operation of a Cbg-CBDcex immobilized column confirmed unequivocably that CBDcex continued to bind at operating temperatures up to 70° with no apparent leaching of the fusion polypeptide (Figures 3.37 & 3.38). A possible explanation for the binding stability at elevated temperature is that the CBDcex part of the thermophilic fusion polypeptide was denatured in situ and "fixed" on the surface of the cellulose, with no obvious detrimental effect on the conformation of Cbg. Irreversible denaturation of the adsorbed CBD at high temperature would not be detrimental for enzyme immobilization as long as the enzyme retained catalytic activity. The adsorption of CBDcex to cellulose at high pH can be explained partly by the pi of the polypeptide. At its pi, a protein exhibits the least intra- and intermolecular electrostatic repulsions, resulting in a more compact structure and greater surface packing. Glucoamylase adsorption to starch showed maximum saturation at its pi of 3.5, and decreased with increasing pH (Dalmia & Nikolov, 1991). The same trend was observed for CBDcex although the differences in the saturation levels were not large (Figure 3.11). The higher relative affinity of CBDcex at pH 9 could also be explained by the decreased electrostatic repulsion among polypeptides at this pH. The lower relative affinity of CBDcex at pH 3 could have resulted from an increased charge repulsion effect and/or changes in the conformation of the polypeptide. The desorption of the fusion polypeptide at a pH (> 9) higher than its pi (~ 5) may have resulted from the electrostatic repulsions among fusion polypeptides on the cellulose surface. It is clear that no generalization can be made from the effect of temperatures and pHs on the binding of CBDcex to cellulose. Until more is known about the structure-function relations of CBD and its substrate, each fusion polypeptide constructed should be tested individually. The immobilization yields for the two fusion polypeptides, Abg-CBDcexl and 3 (section 3.2.9) were comparable to those reported for 6-glucosidases immobilized by other methods (Table 1.1). An exception is the immobilization of B-glucosidase through sugar moieties to concanavalin A-Sepharose. The activity retained was 90%. This difference in the performance of two analogous immobilization techniques might be due to denser packing of fusion polypeptides on the cellulose surface, resulting in steric hindrance of the bound molecules. Alternatively, the activity of the Abg fusion polypeptides might be reduced by the interaction of the active site of the fusion polypeptide with a cellobiosyl unit in the cellulose, to the exclusion of pNPG, i.e. cellulose could act as a competitive inhibitor of Abg-CBDcex under these conditions. In CenA, the Pro-Thr linker forms an extended, rigid segment connecting the ellipsoidal catalytic domain to the CBD (Pilz et al, 1990; Shen et al., 1991). It is not known if the Pro-Thr linker performs the same function in Cex. There is no model as yet for Cex, but the Cex paired distance distribution function from small-angle X-ray scattering analyses shows that the overall shape is similar to that of CenA (Schmuck & Gilkes, unpublished results). Moreover, the Pro-Thr linker-CBD amino acid sequence is very similar to that in CenA. The observation that Abg-CBDcex3 (lacking an intervening 57 amino acids) showed less activity than Abg-CBDcexl is consistent with a role for the linker in reducing inhibitory interaction of Abg with the cellulose surface. The immobilization yield of Abg-CBDcex could be improved by reducing the enzyme loading. The enzyme loading and operational stability of the ConA-adsorbed B-glucosidase were not reported, thus comparisons with the performance of Abg-CBDcex adsorbed to cellulose cannot be made. Cex could be cleaved by a C.fimi serine protease (Gilkes et al., 1988) and papain (Gilkes et al., 1991a) into a stable core peptide (p33) corresponding to the catalytic domain and a shorter peptide (p8) corresponding to the CBD. p8 was resistant to further cleavage. CBDcex was resistant to cleavage by papain and the C. fimi protease (data not shown). Native CenA could be cleaved by trypsin at sites inside the CBD producing p45 and p38 fragments, and, to a lesser extent, at the junction of the Pro-Thr linker and the catalytic domain producing a p30 fragment (Gilkes et al., 1989). It was concluded that CBDcenA had a rather open native conformation relative to the catalytic domain which resisted protease cleavage. Trypsin cleaved CBDcex m the presence of urea and DTT, albeit slowly. This suggests that CBDcex in solution adopts a more tightly folded conformation than CBDcenA affording protection from proteolytic attack. Pepsin is a carboxyl protease that hydrolyzes peptides with hydrophobic residues on either side of the scissile bond (Fersht, 1985). It is therefore a broad-spectrum protease capable of digesting polypeptides into very small fragments. It also requires a very acidic (pH 1-5) environment to function. CBDcex contaminated with xylan was rapidly digested by pepsin, whereas the xylan-free material was completely resistant (Figure 3.15). Xylan thus alters the sensitivity of CBDcex to pepsin presumably by changing the conformation of the polypeptide. It is possible that, under the conditions used for pepsin digestion (pH 4), the xylan associated with CBDcex was dissociated since xylan was removed during cation-exchange chromatography of CBDcex at pH 4 (Figure 3.7). However, the fact that pepsin digestion of the xylan-contaminated CBDcex* but not the purified material, still occurred at pH 4 argues against this possibility. This paradox could be explained by arguing that the low pH did not cause the dissociation of the xylan material from CBDcex- It did however allow the displacement of the xylan contaminant during cation-exchange chromatography favouring ionic interaction over hydrogen bonding. The strength of most ionic bonds is ~ 500 kJ.moH, which makes them stronger than either hydrogen bonds ( ~25 kJ.moH) or van der Waals interaction (~ 1 kJ.moH) (Israelachvili, 1992). Charged residues (like lysine, arginine, glutamic and aspartic acids) have all been shown to participate in hydrogen bond formation in the arabinose-binding (Quiocho, 1988) and maltose-binding (Spurlino et al., 1991) proteins. Further confirmation of adsorption at acid pH comes from the observation that CBDcex bound stably to cellulose at pH 3 (Figure 3.11). Detergents are a class of molecules with amphiphilic structure (Neugebauer, 1990). Each molecule has both hydrophobic and hydrophilic moieties and is capable of forming micelles with other detergent molecules. Non-ionic detergents (e.g. Triton X-100) act by disrupting hydrophobic interactions within and between molecules. In addition, ionic detergents (e.g. SDS) also bind to proteins causing gross structural changes and denaturation. The stable adsorption of CBDcex to cellulose in Triton X-100 (up to 2%; section 3.1.4) and SDS (up to 0.2%) suggests the potential use of CBDcex for fusion to polypeptides (e.g. membrane-associated lipoproteins) that require the presence of detergents for their solubilization and purification. More fundamentally, the ability of CBDcex to bind to cellulose in the presence of detergents suggests that the interaction between the CBD and cellulose is not mediated by hydrophobic interaction. However, hydrophobic interaction may well be involved in the structural stability of CBDcex because of the high aliphatic (24%) and aromatic (10%) amino acid contents of the polypeptide. The adsorption plots of CBDcex (1/bound [P] vs. 1/free [P]; figures 3.9, 3.10 & 3.11) and Abg-CBDcexl (bound [P]/free [P] vs. bound [P] (Scatchard); Ong et al., 1991) were non-linear. Non-linearity is often indicative of binding site heterogeneity (i.e. the presence of two or more classes of binding sites) or cooperativity (i.e. ligand-ligand interaction) (Dahlquist, 1978). Classical Scatchard analysis is valid for small ligands interacting with independent binding but not for large biomolecules which can potentially occupy several binding sites at any given time (e.g. DNA-protein interaction [McGhee & von Hippel, 1974]). Since the dimensions of the CBD (Pilz et al, 1990; Shen et al, 1991) greatly exceed those of the repeating cellobiosyl units, the adsorption process may involve productive (binding) or non-productive (masking) or both interactions with more than one lattice unit (Gilkes et al, 1992). By extension of the DNA-protein interaction model (McGhee & von Hippel, 1974) the cellulose surface is viewed as a two-dimensional lattice with an array of overlapping potential sites (Gilkes et al, 1992). Despite these considerations many previous studies on the adsorption of cellulases used a simple one- or two-site model based on the Langmuir adsorption isotherm (Lee etal., 1982; Ooshima et al., 1983; Stuart & Ristroph, 1985; Steiner et al., 1988; Woodward et al, 1988; Stahlberg et al, 1991). These analyses are further confounded by the use of very heterogeneous substrates like Avicel™. It is obvious that analysis of the adsorption process is inherently complex, and requires a better understanding of the structure of the CBD and the topology of cellulose. The exact mechanism of CBD adsorption to cellulose is not known. A more thorough understanding of the adsorption process would allow the operating conditions and the choice of cellulosic matrix for purification and/or immobilization to be defined more precisely. Published results (Reese, E.T., 1982; Gilkes et al, 1988; Owolabi et al, 1988; Ong et al, 1989b) strongly suggest that hydrogen bonding, together with van der Waals interactions, between the CBD and glucosyl units are the principal forces involved in the adsorption process. Binding does not appear to be mediated by salt linkages because native cellulose is uncharged and adsorbed CBDcex fusion polypeptides were stable at NaCl concentrations up to 1 M (section 3.2.4). Water molecules would compete with the hydrogen bond network between substrate and polypeptide, resulting in desorption of the polypeptide. The size and packing of the polypeptide on the cellulose surface, and the degree of crystallinity of the cellulose used also affect desorption, presumably by blocking access of water molecules or allowing tighter binding of the polypeptide to cellulose or both. A definitive explanation of the binding mechanism may be possible when the structure of the CBD, alone or bound to its substrate, is solved to a resolution comparable to that of the maltose-binding protein (Spurlino et al., 1991). X-ray crystallographic and two-dimensional nuclear magnetic resonance spectroscopic analyses of CBDcex are now underway. In summary, the potential of fusion polypeptides containing CBDcex or related CBDs for affinity purification and/or enzyme immobilization on a commercial scale has been demonstrated. Simultaneous purification and immobilization would result in simplified downstream processing, improved recovery of proteins and significantly reduced purification costs. The simple and stable coupling of an enzyme-CBD fusion to a readily available, inexpensive and stable cellulosic support opens the possibility of using this technique as a generic method for immobilizing a complete range of proteins. Although this generic method for making fusion proteins necessitates the fusion partner tolerating the affinity "tag", the availability of other C.fimi-type CBDs (e.g. CBDcenA) for fusion to the N-terminus of the fusion partner (Greenwood et al., 1989) increases the flexibility of the method. This is especially important if fusion at the C-terminus of the protein is detrimental to its function. 5 . R E F E R E N C E S Aitken, A., M.J. Geisow, J.B.C. Findlay, C. Holmes & A. Yarwood. 1989. Peptide preparation and characterization. In J.B.C. Findlay & M.J. Geisow (ed.) Protein Sequencing: A Practical Approach. IRL Press, Oxford, pp. 43-68. Alfani, F., L. Cantarella, A. Gallifuoco, L. Pezzullo, V. Scardi & M. Cantarella. 1987. Characterization of the 6-glucosidase activity associated with immobilized cellulase of Aspergillus niger. Annals New York Acad. Sci. 501:503-07. Atkinson, T. & M. Smith. 1984. Solid phase synthesis of oligodeoxyribo-nucleotides by the phosphite triester method. In N.J. Gait (ed.) Oligonucleotide Synthesis: A Practical Approach. IRL Press, Oxford, pp. 35-81. Baker, J.O., K.K. Oh, K. Grohmann & M.E. Himmel. 1988. Thermal stabilization of fungal 8-glucosidase through glutaraldehyde crosslinking. Biotechnol. Letters 10:325-330. Baker, T.A., A.D. Grossman & C.A. Gross. 1984. A gene regulating the heat shock response in Escherichia coli also affects proteolysis. Proc. Natl. Acad. Sci. USA 81:6779-6783. Baneyx, F. & G. Georgiou. 1989. Expression, purification, and enzymatic characterization of a protein A-6-lactamase hybrid protein. Enzyme Microb. Technol. 11:559-67. Baneyx, F., C. Schmidt & G. Georgiou. 1990. Affinity immobilization of a genetically engineered bifunctional hybrid protein. Enzyme Microb. Technol. 12:337-42. Bernath, F.R. & K. Venkatasubramanian. 1987. Methods of enzyme immobilization. In A. Demain & N. Solomon (ed.) Manual of Industrial Microbiology and Biotechnology. American Society for Microbiology, Washington, D.C., pp. 230-47. Blackwell, J. 1981. The structures of cellulose and chitin. In R. Srinivasan (ed.) Biomolecular Structure, Conformation, Function and Evolution. Vol. I: Diffraction and Related Studies. Pergamon, New York, pp. 523-35. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of proteins utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. Buchholz, K. (ed.) 1979. Characterization of Immobilized Biocatalysts. Verlag Chemie, Weinheim. Biilow, L. & K. Mosbach. 1991. Multienzyme systems obtained by gene fusion. Trends Biotechnol. 9:226-31. Caldwell, S.R. & F.M. Raushel. 1991. Detoxification of organophosphate pesticides using an immobilized phosphotriesterase from Pseudomonas diminuta. Biotechnol. Bioeng. 37:103-09. Cannon, J.J., L.F. Chen, M.C. Flickinger & G.T. Tsao. 1984. The development of an immobilized lactate oxidase system for lactic acid analysis. Biotechnol. Bioeng. 26:167-73. Cantor, C.R. & P.R. Schimmel. 1980. Biophysical Chemistry Part II. Techniques for the Study of Biological Structure and Function. Freeman, San Francisco, p. 380. Chaplin, M.F. 1986. Monosaccharides. In M.F. Chaplin & J.F. Kennedy (ed.) Carbohydrate Analysis: A Practical Approach. IRL Press, Oxford, p. 2. Chen, L., C. Ford, A. Kusnadi & Z.L. Nikolov. 1991. Improved adsorption to starch of a 6-galactosidase fusion protein containing the starch-binding domain from Aspergillus glucoamylase. Biotechnol. Prog. 7:225-29. Chen, L.F., C.S. Gong & G.T. Tsao. 1981. Immobilized glucose isomerase on DEAE cellulose beads. Starch 33:58-63. Chibata, I. (ed.) 1978. Immobilized Enzymes: Research and Applications Wiley, New York. Chien, C.H., Y.H. Wei & J.F. Shaw. 1991. Immobilized enzyme system for determination of sialic acid in serum or urine. Enzyme Microb. Technol. 13:45-52. Coutinho, J.B., B. Moser, D.G. Kilburn, R.AJ. Warren & R.C. Miller, Jr. 1991. Nucleotide sequence of the endoglucanase C gene (cenC) of Cellulomonas fimi, its high-level expression in Escherichia coli, and characterization of its products. Mol. Microbiol. 5:1221-33. Creighton, T.E. 1989. Disulphide bonds between cysteine residues. In T.E. Creighton (ed.) Protein Structure: A Practical Approach. IRL Press, Oxford. Dahlquist, F.W. 1978. The meaning of Scatchard and Hill plots. Methods Enzymol. 48:270-99. Dalmia, B.K. & Z.L. Nikolov. 1991. Characterization of glucoamylase adsorption to raw starch. Enzyme Microb. Technol. 13:982-990. Day, A.G. & S.G. Withers. 1986. The purification and characterization of a 6-glucosidase from Alcaligenes faecalis. Biochem. Cell Biol. 64:914-22. Dominquez, E., M. Nilsson & B. Hahn-Hagerdal. 1988. Carbodiimide coupling of 6-galactosidase from Aspergillus oryzae to alginate. Enzyme Microb. Technol. 10:606-10. Drickamer, K. 1988. Two distinct classes of carbohydrate-recognition domains in animal lectins. / . Biol. Chem. 263:9557-60. Dumitriu, S., M. Popa, V. Artenie & F. Dan. 1989. Bioactive polymers. 56: urease immobilization on carboxymethylcellulose. Biotechnol. Bioeng. 34: 283-290. Elving, P.J., J.M. Markowitz & I. Rosenthal. 1956. Preparation of buffer systems of constant ionic strength. Anal. Chem. 28:1179-80. Fersht, A. 1985. Enzyme Structure and Mechanism, 2nd ed. Freeman, New York. Ford, J.R., A.H. Lambert, W. Cohen & R.P. Chambers. 1972. Recirculation reactor system for kinetic studies of immobilized enzymes. Biotechnol. Bioeng. Symp. 3:267-84. Fujikawa, S., T. Yokota & K. Koga. 1988. Immobilization of B-glucosidase in calcium alginate gel using genipin as a new type of cross-linking reagent of natural origin. Appl. Microbiol. Biotechnol. 28:440-41. Gianfreda, L., M. Modaferri & G. Greco, Jr. 1985. Enzyme stabilization towards thermal, chemical and proteolytic deactivation. Enzyme Microb. Technol. 7:78-82. Gilkes, N.R., D.G. Kilburn, R.C. Miller, Jr. & R.A.J. Warren. 1984. A mutant of Escherichia coli that leaks cellulase activity encoded by cellulase genes for Cellulomonas fimi. Bio/Technology 2:259. Gilkes, N.R., D.G. Kilburn, R.C. Miller, Jr. & R.A.J. Warren. 1989. Structural and functional analysis of a bacterial cellulase by proteolysis. / . Biol. Chem. 264:17802-08. Gilkes, N.R., E. Jervis, B. Henrissat, B. Tekant, R.C. Miller, Jr., R.AJ. Warren & D.G. Kilburn. 1992. The adsorption of a bacterial cellulase and its two isolated domains to crystalline cellulose. / . Biol. Chem.., in press. Gilkes, N.R., M. Claeyssens, R. Aebersold, B. Henrissat, A. Meinke, H.D. Morrison, D.G. Kilburn, R.A.J. Warren & R.C. Miller, Jr. 1991a. Structural and functional relationships in two families of 6-1,4-glycanases. Eur. J. Biochem. 202:367-77. Gilkes, N.R., B. Henrissat, D.G. Kilburn, R.C. Miller, Jr. & R.A.J. Warren. 1991b. Domains in microbial 6-1,4-glycanases: sequence conservation, function, and enzyme families. Microbiol. Rev. 55:303-15. Gilkes, N.R., R.A.J. Warren, R.C. Miller, Jr. & D.G. Kilburn. 1988. Precise excision of the cellulose-binding domains from two Cellulomonas fimi cellulases by a homologous protease and the effect on catalysis. / . Biol. Chem. 263:10401-07. Goldberg, A.L. & S.A. Goff. 1986. The selective degradation of abnormal proteins in bacteria. In W. Reznikoff and L. Gold (ed.) Maximizing Gene Expression Butterworths, Boston, pp. 287-313 Gomori, G. 1955. Preparation of buffers for use in enzyme studies. Methods Enzymol. 1:138-46. Greenwood, J.M., N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr. & R.AJ. Warren. 1989. Fusion to an endoglucanase allows alkaline phosphatase to bind to cellulose. FEBS Letters 244:127-31. Grossman, A.D., J.W. Erickson & C.A. Gross. 1984. The htpR gene product of E. coli is a sigma factor for heat-shock promoters. Cell 38:383-90. Guilbault, G.G. 1984. Analytical Uses of Immobilized Enzymes. Dekker, New York. Guilbault, G.G. 1988. Enzyme electrode probes. Methods Enzymol. 137:14-29. Guilbault, G.G., J.M. Kauffman & G.J. Patriarche. 1991. Immobilized enzyme electrodes as biosensors. In R.F. Taylor (ed.) Protein Immobilization. Dekker, New York, pp. 209-62. Guo, Z.M., N. Arfman, E. Ong, N.R. Gilkes, D.G. Kilburn, R.AJ. Warren & R.C. Miller, Jr. 1988. Leakage of Cellulomonas fimi cellulases from Escherichia coli. FEMS Microbiol. Lett. 49:279-83. Hahn-Hagerdal, B. 1984. An enzyme coimmobilized with a microorganism: the conversion of cellobiose to ethanol using 6-glucosidase and Saccharomyces cerevisiae in calcium alginate gels. Biotechnol. Bioeng. 26:111-1 A. Han, Y.W. & V.R. Srinivasan, V.R. 1969. Purification and characterization of 6-glucosidase of Alcaligenes faecalis. J. Bacteriol. 100:1355-63. Hanahan, D. & M. Meselson. 1983. Plasmid screening at high colony density. Gene 10:63. Hanahan, D. 1982. Studies on transformation of Escherichia coli with plasmids. J.Mol.Biol. 166:557. Harlow, E. & D. Lane. 1988. Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York. Hartmeier, W. 1988. Immobilized Biocatalysts. Springer-Verlag, Berlin. Henrissat, B. & H. Chanzy. 1986. Enzymatic breakdown of cellulose crystals. In R.A. Young & R.M. Rowell (ed.) Cellulose: Structure, Modification and Hydrolysis. Wiley, N.Y., pp.337-47. Henrissat, B., B. Vigny, A. Buleon & S. Perez. 1988. Possible adsorption sites of cellulases on crystalline cellulose. FEBS Letters 231:177-82. Hestrin, S. 1963. Bacterial cellulose. Methods Carbohyd. Chem. 3:4-9. Huang, Y.L., S.Y. Li, B.A.A. Dremel, U. Bilitewski & R.D. Schmid. 1991. On-line determination of glucose concentration throughout animal cell cultures based on chemiluminescent detection of hydrogen peroxide coupled with flow-injection analysis. J. Biotechnol. 18:161-72. Husain, Q. & M. Saleemuddin. 1989. An inexpensive procedure for the immobilization of glycoenzymes on Sephadex G-50 using crude concanavalin A. Biotechnol. Appl. Biochem. 11:508-12. Imai, K., T. Shiomi, K. Uchida & M. Miya. 1986. Immobilization of enzyme onto poly(ethylene-vinyl alcohol) membrane. Biotechnol. Bioeng. 28:198-203. Israelachvili, J. 1992. Intermodular and Surface Forces, 2nd ed. Academic Press, London, p. 125. Jayakumari, V.G. & N.R. Pillai. 1990. Peptide synthesis in a triphasic medium catalyzed by papain immobilized on a crosslinked polystyrene support. Indian J. Chem. 29:514-17. Jin, F. & K. Toda. 1988. Preparation of immobilized papain covalently bound on natural cellulose for treatment of beer. Biotechnol. Letters 10:221-23. Johnson, L.N., J. Cheetham, P.J. McLaughlin, K.R. Acharya, D. Barford & D.C. Phillips. 1988. Protein-oligosaccharide interactions: lysozyme, phosphorylase, amylases. Current Topics Microbiol. Immunol. 139:81-134. Karkare, S.B. 1991. Immobilized microbial and animal cells as enzyme reactors. In R.F. Taylor (ed.) Protein Immobilization. Dekker, New York, pp. 319-37. Kierstan, M , A. McHale & M.P. Coughlan. 1982. The production of ethanol from cellobiose using immobilized 6-glucosidase coentrapped with yeast in alginate gels. Biotechnol. Bioeng. 24:1461-63. Kimura, T., M. Ogata, H. Kobayashi, M. Yoshida, K. Oishi & T. Nakakuki. 1990a. Continuous production of maltotetraose using a dual immobilized enzyme system of maltotetraose-forming amylase and pullulanase. Biotechnol. Bioeng. 36:790-96. Kimura, T., M. Ogata, M. Yoshida & T. Nakakuki. 1989. Stability of immobilized maltotetraose-forming amylase from Pseudomonas stutzeri. Biotechnol. Bioeng. 33:845-55. Kimura, Y., T. Yoshida, K. Muraya, K. Nakanishi & R. Matsuno. 1990b. Continuous synthesis of a tripeptide by successive condensation and transesterification catalyzed by two immobilized proteinases in organic solvent. Agri. Biol. Chem. 54:1433-40. Klausmeier, W.H. 1986. Economic analysis of an innovative process for cellulose acetate production. In R.A. Young & R.M. Rowell (ed.) Cellulose: Structure, Modification and Hydrolysis. Wiley, N.Y., pp. 187-201. Kohlwey, D.E. & M. Cheryan. 1981. Performance of a 6-D-galactosidase hollow fibre reactor. Enzyme Microb. Technol. 3:64-68. Kraft, R., J. Tardiff, K.S. Krauter & L.A. Leinwand. 1988. Using mini-prep plasmid DNA for sequencing double-stranded template with Sequenase. BioTechniques 6:544-48. Kraulis, P.M., M.G. Clore, M. Nilges, T.A. Jones, G. Pettersson, J. Knowles & A.M. Gronenborn. 1989. Determination of the three-dimensional solution structure of the C-terminal domain of cellobiohydrolase I from Trichoderma reesei. A study using nuclear magnetic resonance and hybrid distance geometry-dynamical simulated annealing. Biochemistry 28:7241-57. Kunkel, T.A., J.D. Roberts & R.A. Zakour. 1987. Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154:367-82. Laemmli, K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-85. Lee, J.M. & J. Woodward. 1983. Properties and application of immobilized 6-D-glucosidase coentrapped with Zymomonas mobilis in calcium alginate. Biotechnol. Bioeng. 25:2441-51. Lee, S.B., H.S. Shin, D.D.Y. Ryu & M. Mandels. 1982. Adsorption of cellulase on cellulose: effects of physicochemical properties of cellulose on adsorption and rate of hydrolysis. Biotechnol. Bioeng. 24:2137-53. Lilly, M.D., W.E. Hornby & E.M. Crook. 1966. The kinetics of carboxymethylcellulose-ficin in packed beds. Biochem. J. 100:718-23. Love, D.R. & M.B. Streiff. 1987. Molecular cloning of a 6-glucosidase gene from an extremely thermophilic anaerobe in E. coli and B. subtilis. Bio/Technology 5:385-87. Love, D.R., R. Fisher & P.L. Bergquist. 1988. Sequence, structure and expression of a cloned fi-glucosidase gene from an extreme thermophile. Mol. Gen. Genet. 213:84-92. Lowe, C.R. 1984. Biosensors. Trends Biotechnol. 2:59-65. Lund, B., F. Lindberg & S. Normark. 1988. Structure and antigenic properties of the tip-located P pilus proteins of uropathogenic Escherichia coli. J. Bacteriol. 170:1887-94. Marshall, K. & D. Sixsmith. 1974. Some physical characteristics of microcrystalline cellulose 1. powders for pharmaceutical use. Drug Devel. Commun. 1:51-71. Matsudaira, P. (ed.) 1989. A Practical Guide to Protein and Peptide Purification for Microsequencing. Academic Press, San Diego, CA. Matsudaira, P. 1987. Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membrane. / . Biol. Chem. 262:10035-38. Matsudaira, P. 1990. Limited N-terminal sequence analysis. Methods Enzymol. 182:602-13. Mattiasson, B. & R. Kaul. 1991. Determination of coupling yields and handling of labile proteins in immobilization technology. In R.F. Taylor (ed.) Protein Immobilization. Dekker, New York, pp. 161-79. McClary, J.A., F. Witney & J. Geisselsoder. 1989. Efficient site-directed in vitro mutagenesis using phagemid vectors. BioTechniques 7:282-89. McGhee, J.D. & P.H. von Hippel. 1974. Theoretical aspects of DNA-protein interactions: co-operative and non-co-operative binding of large ligands to a one-dimensional homogeneous lattice. / . Mol. Biol. 86:469-89. Meacock, P.A. & S.N. Cohen. 1980. Partitioning of bacterial plasmids during cell division, a ds-acting locus that accomplishes stable plasmid inheritance. Cell 20:529-42. Meinke, A., C. Braun, N.R. Gilkes, D.G. Kilburn, R.C. MiUer, Jr. & R.AJ. Warren. 1991. Unusual sequence organization in CenB, an inverting endoglucanase from Cellulomonas fimi. J. Bacteriol. 173:308-14. Merril, C.R. 1990. Gel staining techniques. Methods Enzymol. 182:477-88. Monsan, P. & D. Combes. 1988. Enzyme stabilization by immobilization. Methods Enzymol. 137:584-98. Mozhaev, V.V. 1990. Immobilization provides additional stabilization of an initially stable protease from Thermoactinomyces vulgaris (Thermitase). Biotechnol. Techniques 4:255-56. Nakakuki, T., T. Hayashi, M. Monma, K. Kawashiwa & K. Kainuma. 1983. Immobilization of the exo-maltohexaohydrolase by the irradiation method. Biotechnol. Bioeng. 25:1095-107. Nakanishi, K., R. Matsuno, K. Torii, K. Yamamoto & T. Kamikubo. 1983. Properties of immobilized 6-D-galactosidase from Bacillus circulans. Enzyme Microb. Technol. 5:115-20. Nelson, J.M. & E.G. Griffin. 1916. Adsorption of invertase. / . Am. Chem. Soc. 38:1109-15. Neugebauer, J.M. 1990. Detergents: an overview. Methods Enzymol. 182:239-53. O'NeiU, G., S.H. Goh, R.A.J. Warren, D.G. Kilbum & R.C. Miller, Jr. 1986. Structure of the gene encoding the exoglucanase of Cellulomonas fimi. Gene 44:325-30. Oguntimein, G.B. & P.J. Reilly. 1980. Purification and immobilization of Aspergillus niger 6-xylosidase. Biotechnol. Bioeng. 22:1127-42. Ong, E., N.R. Gilkes, R.A.J. Warren, R.C. Miller, Jr. & D.G. Kilbum. 1989a. Enzyme immobilization using the cellulose-binding domain of a Cellulomonas fimi exoglucanase. Bio/Technology 7:604-07. Ong, E., J.M. Greenwood, N.R. Gilkes, D.G. Kilbum, R.C. Miller, Jr. & R.A.J. Warren. 1989b. The cellulose-binding domains of cellulases: tools for biotechnology. Trends Biotechnol. 7:239-43. Ong, E., N.R. Gilkes, R.C. Miller, Jr., R.A.J. Warren & D.G. Kilbum. 1991. Enzyme immobilization using a cellulose-binding domain: properties of a 6-glucosidase fusion protein. Enzyme Microb. Technol. 13:59-65. Ooshima, H., M. Sakata & Y. Harano. 1983. Adsorption of cellulase from Trichoderma viride on cellulose. Biotechnol. Bioeng. 24:3103-14. Ostrove, S. 1990. Affinity chromatography: general methods. Methods Enzymol. 182:357-71. Otter, D.E., P.A. Munro, G.K. Scott & R. Geddes. 1989. Desorption of Trichoderma reesei cellulase from cellulose by a range of desorbents. Biotechnol. Bioeng. 34:291-98. Owolabi, J.B., P. Beguin, D.G. Kilburn, R.C. Miller, Jr. & R.A.J. Warren. 1988. Expression in Escherichia coli of the Cellulomonas fimi structural gene for endoglucanase B. Applied Environ. Microbiol. 54:518-23. Patwardhan, V.S. & N.G. Karanth. 1982. Film diffusional influences on the kinetic parameters in packed-bed immobilized enzyme reactors. Biotechnol. Bioeng. 24:763-80. Peitersen, N., J. Medeiros & M. Mandels. 1977. Adsorption of Trichoderma cellulase on cellulose. Biotechnol. Bioeng. 19:1091-94. Pilz, I., E. Schwarz, D.G. Kilburn, R.C. Miller, Jr., R.A.J. Warren & N.R. Gilkes. 1990. The tertiary structure of a bacterial cellulase determined by small-angle x-ray scattering analysis. Biochem. J. 271:277-80. Price, N.C. & CM. Johnson. 1989. Proteinases as probes of conformation of soluble proteins. In R.J. Beynon & J.S. Bond (ed.) Proteolytic Enzymes: A Practical Approach. IRL Press, Oxford. Quiocho, F.A. & N.K. Vyas. 1984. Novel stereospecificity of the L-arabinose-binding protein. Nature 310:381-86. Quiocho, F.A. 1986. Carbohydrate-binding proteins: tertiary structures and protein-sugar interactions. Ann. Rev. Biochem. 55:287-315. Quiocho, F.A. 1988. Molecular features and basic understanding of protein-carbohydrate interactions: the arabinose-binding protein-sugar complex. Current Topics Microbiol. Immunol. 139:135-48. Reiken, S.R., R.J. Knob & D.M. Briedis. 1990. Evaluation of intrinsic kinetics in hollow fiber reactor systems. Enzyme Microb. Technol. 12:736-42. Roda, A., S. Girotti, B. Grigolo, S. Ghini, G. Carrea, R. Bovara, I. Zini & R. Grimaldi. 1991. Microdialysis and luminescent probe: analytical and clinical aspects. Biosensors & Bioelectronics 6:21-30. Ross, P., R. Mayer & M. Benziman. 1991. Cellulose biosynthesis and function in bacteria. Microbiol. Rev. 55:35-58. Rouvinen, J., T. Bergfors, T. Teeri, J.K.C. Knowles & T.A. Jones. 1990. Three-dimensional structure of cellobiohydrolase II from Trichoderma reesei. Science 249:380-86. Roy, F. & M.V. Hegde. 1987. Immobilization of 6-amylase on polystyrene cation exchange resin equilibrated with Al3+ ions (IR-120 Al3+). Enzyme Microb. Technol. 9:550-52. Roy, S.K., S.K. Raha, S.K. Dey & S.L. Chakrabarty. 1989. Immobilization of 6-glucosidase from Myceliophthora thermophila D-14. Enzyme Microb. Technol. 11:431-35. Sambrook, J., E.F. Fritsch & T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, New York. Sanger, F., S. Nicklen & A.R. Coulson. 1977. DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-67. Sarko, A. 1986. Recent x-ray crystallographic studies of celluloses. In R.A. Young & R.M. Rowell (ed.) Cellulose: Structure, Modification and Hydrolysis. Wiley, New York, pp. 29-49. Sassenfeld, H.M. 1990. Engineering proteins for purification. Trends Biotechnol. 8:88-93. Schagger, H. & G. von Jagow. 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166:368-79. Scopes, R.K. 1974. Measurement of protein by spectrophotometry at 205nm. Anal Biochem. 59:277-82. Sharma, S. & H. Yamazaki. 1984. Preparation of hydrophobic cotton cloth. Biotechnol. Lett. 6:301-06. Shen, H., M. Schmuck, I. Pilz, N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr. & R.A.J. Warren. 1991. Deletion of the linker connecting the catalytic and cellulose-binding domains of endoglucanase A (CenA) of Cellulomonas fimi alters its conformation and catalytic activity. / . Biol. Chem. 266:11335-40. Skoog, B. & A. Wichman. 1986. Calculation of the isoelectric points of polypeptides from the amino acid composition. Trends Anal. Chem. 5:82-83. Spurlino, J.C., G.Y. Lu & F.A. Quiocho. 1991. The 2.3-A resolution structure of the maltose- or maltodextrin-binding protein, a primary receptor of bacterial active transport and chemotaxis. / . Biol. Chem. 266:5202-19. Stahlberg, J., G. Johansson & G. Pettersson. 1991. A new model for enzymatic hydrolysis of cellulose based on the two-domain structure of cellobiohydrolase I. Bio/Technology 9:286-90. Steiner, W., W. Sattler & H. Esterbauer. 1988. Adsorption of Trichoderma reesei cellulase on cellulose: experimental data and their analysis by different equations. Biotechnol. Bioeng. 32:853-65. Stoll, V.S. & J.S. Blanchard. 1990. Buffers: principles and practice. Methods Enzymol. 182:24-38. Strandberg, G.W. & K.L. Smiley. 1972. Glucose isomerase covalently bound to porous glass beads. Biotechnol. Bioeng. 14:509-13. Stuart, J.Y. & D.L. Ristroph. 1985. Analysis of cellulose-cellulase adsorption data: a fundamental approach. Biotechnol. Bioeng. 27:1056-59. Svensson, B., H. Jespersen, M.R. Sierks & E.A. MacGregor. 1989. Sequence homology between putative raw-starch binding domains from different starch-degrading enzymes. Biochem. J. 264:309-11. Tabor, S. & C.C. Richardson. 1987. DNA sequence analysis with a modified bacteriophage T7 DNA polymerase. Proc. Natl. Acad. Sc. USA. 84:4767-71. Taylor, R.F. (ed.) 1991. Protein Immobilization. Dekker, New York. Taylor, R.F. 1991. Immobilized antibody-and receptor-based biosensors. In R.F. Taylor (ed.) Protein Immobilization. Dekker, New York, pp.263-318. Tomar, M. & K.A. Prabhu. 1985. Immobilization of glucoamylase on DEAE-cellulose activated with chloride compounds. Enzyme Microb. Technol. 7:557-60. Tull, D., S.G. Withers, N.R. Gilkes, D.G. Kilburn, R.AJ. Warren & R. Aebersold. 1991. Glutamic acid 274 is the nucleophile in the active site of a "retaining" exoglucanase from Cellulomonas fimi. J. Biol. Chem. 266:15621-25. Vieira, J. & J. Messing. 1982. The pUC plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19:259-68. Wakarchuk, W.W., D.G. Kilburn, R.C. Miller, Jr. & R.AJ. Warren. 1986. The molecular cloning and expression of a cellobiase gene from an Agrobacterium in Escherichia coli. Mol. Gen. Genet. 205:146-52. Wakarchuk, W.W., N.M. Greenberg, D.G. Kilburn, R.C. Miller, Jr. & R.AJ. Warren. 1988. Structure and transcription analysis of the gene encoding a cellobiase from Agrobacterium sp. strain ATCC 21400. / . Bacteriol. 170:301-07. Webster, LA., C.E. Schwier & F.S. Bates. 1985. Using the rotational masking concept to enhance substrate inhibited reaction rates: controlled pore supports for enzyme immobilization. Enzyme Microb. Technol. 7:266-7r4. Withers, S.G. & LP. Street. 1988. Identification of a covalent a-D-glucopyranosyl enzyme intermediate formed on a 8-glucosidase. / . Am. Chem. Soc. 110:8551-53. Withers, S.G., LP. Street & D.H. Dolphin. 1987. 2-deoxy-2-fluoroglucosides: a novel class of mechanism-based glucosidase inhibitors. J. Amer. Chem. Soc. 109:7530-31. Withers, S.G., K. Rupitz & LP. Street. 1988. 2-deoxy-2-fluoro-D-glycosyl fluorides: a new class of specific mechanism-based glycosidase inhibitors. / . Biol. Chem. 263:7929-7932. Wong, W.K.R., B. Gerhard, Z.M. Guo, D.G. Kilburn, R.AJ. Warren & R.C. Miller, Jr. 1986. Characterization and structure of an endoglucanase gene cenA of Cellulomonas flmi. Gene 44:315-24. Wood, T.M. 1988. Preparation of crystalline, amorphous and dyed cellulase susbstrates. Methods Enzymol. 160:19-25. Woodward, J. & G.S. Zachry. 1982. Immobilization of cellulase through its carbohydrate side chains - a rationale for its recovery and reuse. Enzyme Microb. Technol. 4:245-48. Woodward, J., M.K. Hayes & N.E. Lee. 1988. Hydrolysis of cellulose by saturating and non-saturating concentrations of cellulase: implications for synergism. Bio/Technology 6:301-04. Yannish-Perron, C , J. Vieira & J. Messing. 1985. Improved ml3 phage cloning vectors and host strains: nucleotide sequence of the ml3mpl8 and pUC19 vectors. Gene 33:103-19. Yoshida, Y., M. Kawase & T. Shiraishi. 1990. Synthesis of branched cyclodextrin with immobilized enzyme bioreactor. Hakkokogaku Kaishi 68:197-204. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0098964/manifest

Comment

Related Items