UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The expression of cellulomomas fimi cellulase genes in Brevibacterium lactofermentum and characterization… Paradis, François William François William 1990

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

Item Metadata

Download

Media
831-UBC_1990_A1 P37.pdf [ 10.57MB ]
Metadata
JSON: 831-1.0098751.json
JSON-LD: 831-1.0098751-ld.json
RDF/XML (Pretty): 831-1.0098751-rdf.xml
RDF/JSON: 831-1.0098751-rdf.json
Turtle: 831-1.0098751-turtle.txt
N-Triples: 831-1.0098751-rdf-ntriples.txt
Original Record: 831-1.0098751-source.json
Full Text
831-1.0098751-fulltext.txt
Citation
831-1.0098751.ris

Full Text

THE EXPRESSION OF CELLULOMONAS FIMICELLULASE GENES IN BREVIBACTERIUM LACTOFERMENTUM AND CHARACTERIZATION OF RECOMBINANT C. FIMI 3-GLUCOSIDASE A FROM E. COLI. By FRANCOIS WILLIAM PARADIS B.Sc, Universite LAVAL, 1983 M.Sc, Universite LAVAL, 1987 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) (Genetics program) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April, 1990 ©Frangois William Paradis, 1990 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 The University of British Columbia Vancouver, Canada DE-6 (2/88) i i ABSTRACT In the first part of this thesis, I describe the expression of C. fimi cellulase genes in the closely related Brevibacterium lactofermentum by generating a shuttle vector able to replicate selectively in the latter and carrying full length cellulase-encoding genes. The expression of those genes apparently originated from some unpredicted regulatory sequences, possibly located within the vector itself. The enzymatic activity was mostly found in the culture medium in B. lactofermentum indicating that the organism secreted the enzymes. The putative C. fimi promoter sequences did not function in B. lactofermentum, making difficult the analysis of their roles in expression of C. fimi cellulase genes. In the second part of this thesis, I describe the characterization of a recombinant C. fimi exo-13-1,4-glucosidase (CbgA) expressed in E. coli. The purified enzyme had a M r of 183 kDa and hydrolysed various 6-glucosides with a preference for cello-oligosaccharides in the order C5>C4>C3>C2. The intact CbgA polypeptide was not required for enzymatic activity since removal of about 700 residues from the amino terminus did not reduce activity. The purified enzyme was used to raise polyclonal antibodies which in turn were used to identify the corresponding enzyme in C. fimi. During the fractionation of C. fimi 6-glucosidases, several enzymes hydrolyzing various (3-glucosides were isolated together with the native CbgA, which was present in the culture medium as part of a protein aggregate. Part of the nucleotide sequence of the 7.2 kb insert was determined. Alignments of the N-terminal amino acid sequences of the purified CbgA and truncated polypeptides with the partial nucleotide sequence of the cloned C. fimi DNA showed that precise excision was responsible for the appearance of a truncated form of CbgA. Alignment of the amino-terminal sequence of a C b g A : C e x Q B D fusion peptide indicated that the pre-mature CbgA starts with a putative leader sequence of 49 amino acids which is followed by a region rich in Pro and Ala residues. Two GTG translational initiation codons followed by sequences resembling i i i prokaryotic ribosome binding sites and separated by a large open reading frame were identified from data obtained after in vitro site-directed mutagenesis of the most upstream initiation codon suggesting that internal re-initiation may occur and that upstream regulatory sequences had not been isolated. i v TABLE OF CONTENTS page ABSTRACT ii TABLE OF CONTENTS . . iv LIST OF TABLES vii LIST OF FIGURES viii LIST OF ABBREVIATIONS x ACKNOWLEDGEMENTS xiii 1. INTRODUCTION 1.1. Background 1 1.2. The enzymatic hydrolysis of crystalline cellulose 3 1.3. The cellulases of Cellulomonas fimi and their genes 5 1.3.1. Molecular cloning and characteristics 6 1.3.2. The cex gene and its product 8 1.3.3. The cenA gene and its product 10 1.3.4. The cenB gene and its product 10 1.3.5. The cenC gene and its product 10 1.3.6. The cbgA gene and its product 11 1.3.7. Transcriptional analysis of C. fimi cellulase genes . . . 11 1.4. Heterologous expression of C.fimi cellulase genes . 11 1.4.1. Expression in E. coli 11 1.4.2. Expression in other Gram-negative bacteria 13 1.4.3. Expression in the yeast Saccharomyces cerevisiae . . 13 1.4.4. Expression in Gram-positive bacteria 14 1.5. Cloning systems in Coryneform bacteria 15 1.6. The first objective of this thesis 16 1.7. Characteristics of 13-glucosidases . . 18 1.7.1. Substrate specificity and measurement of activity . . . 19 1.7.2. Classification of G-glucosidases 21 1.8. The preliminary characterization of C. fimi B-glucosidases . . . 24 1.9. The second objective of this thesis 25 V 2. MATERIALS AND METHODS 2.1. Bacterial strains, plasmids, phage and growth conditions . . 26 2.2. Enzymes and reagents 26 2.3. Transformation of bacteria by plasmid DNA 26 2.4. Enzyme assays and protein determination 28 2.5. Fractionation of bacterial cells 29 2.5.1. Fractionation of E. coli cells 29 2.5.2. Fractionation of C. fimi cells . . . . . . . . . . . . . 29 2.6. The purification of B-glucosidases 30 2.6.1. The purification of recombinant CbgA 30 2.6.2. The purification of C. fimi 3-glucosidases 31 2.7. Characterization of purified recombinant CbgA 31 2.7.1. The effects of pH and temperature on CbgA activity 31 2.7.2. Determination of the kinetic parameters for CbgA . . . 31 2.8. Electrophoretic analysis of proteins 32 2.9. Production of antiserum and Western blot analysis 33 2.10. Determination of the NH2-terminal amino acid sequences and amino acid compositions of recombinant CbgA polypeptides 35 2.11. DNA and RNA methodology 35 2.11.1. Isolation of plasmid DNA and restriction digest analysis 35 2.11.2. Isolation of RNA and RNase-free work 36 2.11.3. Northern blot analysis of mRNA 37 2.11.4. Preparation of 3 2 P labelled DNA hybridization probes . 38 2.11.5. Hybrid protection analysis 39 2.11.6. DNA sequencing 39 2.11.7. Oligonucleotide directed mutagenesis . . . . . . . . 41 3. RESULTS AND DISCUSSIONS PART 1. The expression of Cellulomonas fimi cellulase genes in Brevibacterium lactofermentum 3.1. Construction of the shuttle vector and subcloning of the cellulase genes 42 3.2. Transformation . 42 3.3. Gene expression and detection of cellulolytic activities . . . . 45 v i 3.4. Detection of cellulase gene specific transcripts 49 3.5. Mapping of the cenA mRNA 5' ends 51 PART 2. The characterization of recombinant C. fimi B-glucosidase A 3.6. The expression of recombinant CbgA in E. coli 54 3.6.1. Cloning, subcloning and specific activities 54 3.6.2. The location of recombinant CbgA in E. coli 57 3.7. Characterization of recombinant CbgA from E. co//7pUC13:62A31 60 3.7.1. The purification of CbgA from E. coli 60 3.7.2. The effects of pH and temperature on CbgA activity . . 67 3.7.3. The kinetic parameters for recombinant CbgA 67 3.7.4. The amino acid composition of CbgA and various enzyme 71 3.8. The analyses of C. fimi B-glucosidases . . 73 3.8.1. The effect of carbon source on B-glucosidase activity . 73 3.8.2. The fractionation of C. fimi B-glucosidases . . . . . . . 73 3.8.3. The detection of B-glucosidase activity after gel electrophoresis 76 3.8.4. Summary of C. fimi B-glucosidases 82 3.9. Antiserum and Western blot analysis 85 3.9.1. Generation of rabbit polyclonal antibodies 85 3.9.2. Western blot analysis of the activity gel 86 3.9.3. Western blot analysis of denatured protein samples . . 86 3.9.4. Western blot analysis of proteins from various clones 91 3.10. DNA and amino acid sequence determination 94 3.10.1. DNA sequence analysis of the 5'end of cbgh\ 94 3.10.2. N-terminal amino acid sequences of polypeptides related to CbgA 99 3.10.3. Mutation of the CbgA translational start site 102 3.10.4. Cellulose affinity chromatography of a CbgA'.CexQgD fusion peptide . 105 3.11. General discussion 107 REFERENCES 109 APPENDIX 120 v i i LIST OF TABLES I. C. fimi cellulases and their genes 9 II. Km's for cellodextrin hydrolysis by various exo-(3-1,4-glucosidase from cellulolytic fungi 23 III. Bacterial strains, plasmids and phage 27 IV. Sequencing primers 40 V. Cellulolytic activities of bacterial strains . 47 VI. Localisation of B-glucosidase activity in E.co//7pUC13:62A31 . . 58 VII. Purification of recombinant CbgA from £. co///pUC13:62A31 . . 66 VIII. Kinetic parameters for recombinant CbgA on various 6-glucosides 69 IX. Amino acid compositions of CbgA and related enzymes . . . . 72 X. Characteristics of C. fimi 6-glucosidases 83 XI. Comparison of leader peptides from C. fimi cellulase 98 XII. N-terminal amino acid sequences of CbgA-CexQ B D, p183, p137 and p60 100 v i i i LIST OF FIGURES 1. Structure of cellulose 2 2. Possible routes of glucose formation in saccharification of cellulose . . 4 3. Structures of CenA and Cex 7 4. Structures of various (3-glucosides 20 5. Titration of antiserum by ELISA on (A) E. coli and (B) C. fimi proteins 34 6. Construction of the shuttle vector and its derivatives . 43 7. Screening of cellulase activities in transformants 46 8. Detection of specific transcripts by Northern blot analysis 50 9. S1 nuclease protection analysis of cenA transcripts 52 10. Diagram of pEC62 and derivatives 55 11. Scheme for the generation of a translational frameshift in pEC62 . . 56 12. Profile of E. co//7pUC13:62A31 membrane fraction separated on a sucrose density gradient 59 13. Fractionation of E. co///pUC13.62A31 cytoplasmic proteins on a DEAE-sephadex A-50-120 column 61 14. Fractionation of Pool I on Mono Q 63 15. Fractionation of stored Pool I on Mono Q 64 16. 7% SDS-PAGE analysis of protein fractions from the ion-exchange purification of recombinant CbgA 65 17. Effect of pH (A) and temperature (B) on the hydrolysis of pNPG by recombinant CbgA 68 18. Thermostability study of purified CbgA . 68 19. Total G-glucosidase (pNPGase) activity in C. fimi cultures grown on various carbon sources 74 20. Fractionation of C. fimi cytoplasmic proteins on a DEAE-sephadex A-50-120 column 75 21 . Fractionation of Pool II on Mono Q 77 22. Fractionation of Pool III on Mono Q 78 23. Zymogram of partially denatured protein samples using MUG as a substrate 80 24. Western blot analysis of the zymogram 87 25. Western blot analysis of the recombinant and native CbgA 88 ix 26. Western blot analysis of E. coli and C. fimi proteins 90 27. Western blot analysis of protein extracts from (3-glucosidase deletion clones expressed in E. coli 92 28. Restriction digest analysis of various deletion clones 93 29. Nucleotide sequence of the 5' end of cbgA . 95 30. Mutation of the putative GTG start site of CbgA 103 3 1 . Western blot analysis of the polypeptides produced by the CbgA translational frameshift mutant 104 32. Purification of the CbgA:CexQBD f u s ' o n peptide 106 X LIST OF ABBREVIATIONS A.A. ; amino acid abg ; Agrobacterium sp.gene encoding B-glucosidase (Abg) Ap ; ampicillin ATCC ; American Type Culture Collection ATP ; adenosine triphosphate B-gal ; 6-galactosidase Bgl ; B-glucosidase bp base pair (s) BSA bovine serum albumine CBD cellulose binding domain cbgA C. fimi gene encoding B-glucosidase A (CbgA) CbgB Cellulomonas fimi 6-glucosidase B CbgC Cellulomonas fimi B-glucosidase C CbgD Cellulomonas fimi B-glucosidase D cenA C. fimi gene encoding endoglucanase A (CenA) cenB C. fimi gene encoding endoglucanase B (CenB) cenC ; C. fimi gene encoding endoglucanase C (CenC) cex ; C. fimi gene encoding exoglucanase (Cex) CIAP ; calf intestine alkaline phosphatase CMC ; carboxymethyl cellulose DEAE ; diethylaminoethyl DMSO ; dimethylsulfoxide DNA ; deoxyribonucleic acid DNAse 1 ; deoxyribonuclease 1 DNS ; dinitrosalicylic acid dNTP ; deoxynucleoside triphosphate (dATP, dCTP, dGTP, dTTP) ds ; double stranded DTT ; dithiothreitol EDTA ; ethylenediaminetetraacetic acid ELISA ; enzyme-linked-immunoadsorbant assay Eng ; endo-B-1,4-glucanase Exg ; exo-B-1,4-glucanase x i FPLC fast protein liquid chromatography HEPES N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid Kb ; 1000 base pair (s) kDa lOOODaltons Km kanamycin Lac Z' , the first 78 amino acid of E. coli 6-galactosidase lac Zpo lac promoter operator LB(ls) Luria broth (low salt) LPS lipopolysaccharide MIC minimal inhibitory concentration MOPS morpholinepropanesulfonic acid Mr relative molecular mass mRNA messenger RNA MUG 4-methylumbelliferyl-6-D-glucoside MUC 4-methylumbelliferyl-l3-D-cellobioside MUGase enzyme which hydrolyses MUG NAD nicotinamide adenine dinucleotide NADP nicotinamide adenine dinucleotide phosphate nt nucleotide (s) O.D. optical density ORF open reading frame PAGE polyacrylamide gel electrophoresis PEG , polyethylene glycol PMSF ; phenylmethylsulfonylfluoride PNK ; polynucleotide kinase pNPC ; p-nitrophenyl-6-D-cellobioside pNPCase ; enzyme which hydrolyses pNPC pNPG ; p-nitrophenyl-8-D-glucoside pNPGase ; enzyme which hydrolyses pNPG PSI ; pounds per square inch RBS ; ribosome binding site RNA ; ribonucleic acid RNase A ; ribonuclease A RPM ; revolution per minute RS ; modified R2-medium x i i SDS sodium dodecyl sulfate SET Sucrose/Tris/EDTA s/n supernatant SS single stranded SSC standard sodium citrate TBE Tris/Borate/EDTA buffer TE Tris/EDTA buffer T c R tetracycline resistance determinant TSA trypticase soy agar TSB trypticase soy broth x i i i ACKNOWLEDGEMENTS I would like to thank those whom I have trusted, for their guidance and support throughout the course of this work: My supervisor Dr. R.C. Miller, Jr., my committee members Drs. R.A.J. Warren, D.G. Kilburn, S. Withers and C. Douglas, and special thanks to my friend and acting member, Dr. J.T. Beatty (chance favours the prepared mind, L. Pasteur). I wish to thank Dr. N.R. Gilkes for the oligo-saccharides and discussions, Ms. S. Keilland for the amino acid sequences and Mr. T. Atkinson for the synthesis of oligonucleotides. To my fellow graduate students, staff and faculty of the department of Microbiology-many thanks for the unforgettable memories. Finallement, je tiens a remercier les membres de ma famille pour la patience et le support moral donnes tout au long de ce travail. Cette these fut financee en partie par mon pere, le Dr. Desmond Paradis. Je dedie cette these a celui qui a su eveiller mon esprit scientifique, le Dr. V. Smirnoff, et a celle a qui de loin me protege, ma grand-mere G. Biron. 1 1. INTRODUCTION 1.1. Background Plant biomass is the most abundant source of carbohydrate available on earth. This nearly unlimited amount of stored solar energy is extensively used by most species, including human, effecting its primary benefits, mainly the production of oxygen and being the first element in the food chain. The scientific community is now concerned about the long term depletion of cellulosic material and methods to improve its utilization by recycling of industrial wastes has become a major concern. Because of its abundance, research has been focussed on the conversion of biomass into more suitable products such as ethanol and acetone, amino acids and nucleotides, even single-cell protein and the production of recombinant material. In nature, cellulose molecules composed of 100 to 10,000 glucose subunits linked by 6-1,4-glucosidic bonds are associated in a highly resistant matrix with lignin and hemicellulose (Fig. 1). This association forms a major structural barrier to the availability of cellulose, and efficient pretreatment methods which promote the complete hydrolysis of cellulose are practically non-existent. As a result of strong inter- and intramolecular hydrogen bonding, cellulose is insoluble in water and in most organic solvents. Nevertheless, a number of cellulose solvents exist, namely, strong mineral acids, quaternary ammonium bases, non-aqueous organic solvents like DMSO or even metal complex solvents like iron-tartaric acid-alkali. Cellulose can be enzymatically degraded by a set of enzymes called cellulases. Enzymatic degradation is slow because of the structure of the substrate but seems to be a preferred choice because of the absence of corrosion and environmental pollution. A variety of organisms can efficiently hydrolyze crystalline cellulose using a process which requires the synergistic action of several enzymes with various substrate specificities. All cellulolytic enzymes are found to have multiple forms which differ in their relative activities on a broad range of substrate. The complex mechanism of enzymatic hydrolysis of cellulose has been reviewed recently (Coughlan, 1985; Beguin etal, 1987). 2 Fig.1. Structure of cellulose, (a) Stereochemical representation of the cellulose molecule. Arrows A and B represent the 6-1,4-linkages within the cellulose fibril, (b) Organisation of cellulose molecules in the elementary fibril. Highly ordered or relative disarray are represented by the crystalline and amorphous regions, respectivelly. (c) Cross-section of a wood fiber. The elementary fibrils are surrounded by hemicellulose and lignin, reducing the accessibility of cellulose to enzymatic degradation (adapted from Fan etai, 1980). CH 2OH CH 2OH CH 2OH "ST CELLOBIOSE b) GLUCOSE CRYSTALLINE REGION AMORPHOUS REGION CRYSTALLINE REGION FIBRIL 30A CELLULOSE ELEMENTARY FIBRIL HEMICELLULOSE 3 1.2. The enzymatic hydrolysis of crystalline cellulose Studies on the extracellular enzymes from a number of cellulolytic fungi, particularly Trichoderma species, have provided us with much of our basic information on the mechanism of cellulase action (Ericksson and Wood, 1985). However, other mechanisms exist in bacteria where the optimal conditions of pH and temperature may be different. The overall kinetic scheme can be represented by this rather simplified form: where S is the original substrate, G n are cello-oligosaccharides with degree of polymerization greater than 2, G2 is cellobiose and G is glucose. As can be seen from this scheme, glucose can be formed either from higher oligosaccharides or from cellobiose. The rate of glucose formation essentially depends on the physical state of the substrate. The rate of hydrolysis decreases with solubility of the substrate in the order CM-cellulose (soluble)- filter paper- microcrystaline cellulose (insoluble). Cellulose hydrolysis requires the synergistic action of various enzymes (Fig.2). In cellulase systems able to degrade crystalline cellulose, three main types of enzymes are found (Shewale, 1982). Endo-1,4-6-D-glucanases (EC 3.2.1.4, 1,4-B-D-glucan glucanohydrolase) hydrolyze the 1,4-B-glucosidic bonds of cellulose molecules randomly generating new ends, and decreasing the degree of polymerization rapidly. The prolonged action of endoglucanases on long molecules of cellulose results in soluble material, mainly cellodextrins and some cellobiose. The rate of attack by endoglucanase which is the first enzyme in the multienzyme chain, decreases progressively with the extent of conversion of cellulose as a result of the exhaustion of more accessible regions of the substrate. The exo-CELLULOSE G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-G-CELLOBIOSE tf CELLODEHTRINS fi-l,4-Glucan g lucanohyd ro lase ff fl-1,4-Glucan ce l l ob iohyd ro lase R -1,4 -glucan glucanohydrolase fl-1,4-Glucan c e l l o b i o h y d r o l a s e G-G, G-G-G, G _ G f e-G-G, G-G-G-G, G-G-G-G, G-G-G-G-G, G-G-G-G-G, G-G-G-G-G-G, G-G-G-G-G-G, ti-glucosidase I J3-g|uco$ldo$e ( E H 0 - f l - 1 , 4 - g l u c a n I ( E H 0 - f l - 1 , 4 - g l u c a n g lucohyd ro lase ) I g l u c o h y d r o l a s e ) ? G-G, G-G, fi-Glucosidase E K O - B -1,4-Glucan g l u c o -h y d r o l a s e GLUCOSE G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, G, Fig 2. Possible r o u t e s o f g lucose f o r m a t i o n in s a c c h a r i f i c a t i o n o f ce l lu lose (Sheuia le , 1982) 5 cellobiohydrolases (EC 3.2.1.91, 1,4-8-D-glucan cellobiohydrolase) act by removing cellobiose from the non-reducing ends of cellulose chains. Exoglucanases when present are major components in some cellulase systems but are usually absent from most of them. Finally, B-glucosidases (EC 3.2.1.74, exo-1,4-6-glucan glucohydrolase and EC 3.2.1.21, B-D-glucoside glucohydrolase) are very important components of these systems because they complete the hydrolysis of soluble cellodextrins into glucose. The B-glucosidases remove the inhibitory effect of cellobiose on endoglucanases and exoglucanases and could represent the rate limiting step in the saccharification of cellulose (Coughlan, 1985). It is assumed that glucose is formed almost exclusively by the hydrolysis of cellobiose by cellobiases. In Trichoderma longibrachiatum, T. reesei, T. lignorum and Aspergillus feotidus, the selective inhibition of cellobiase activity by gluconolactone has shown that in the initial stages of hydrolysis of cellulose, exoglucosidase plays the main role in the formation of glucose from soluble and insoluble polymeric substrates (Sinitsyn and Klesov, 1981). Due to the fact that until that time there was no suitable method for determining exoglucosidase activity, exoglucosidases are among the least studied components of the cellulase complex. The properties of B-glucosidases previously reviewed by Shewale (1982) and Woodward and Wiseman (1982) will be presented since this thesis is largely concerned with the characterization of recombinant C. fimi B-glucosidase A (CbgA). 1.3 The cellulases of Cellulomonas fimi and their genes. C. fimi is a Gram-positive coryneform facultative anaerobic bacterium with DNA of 72 mole % G+C (Stackebrant and Kandler, 1979). By secreting into the culture medium a complex mixture of cellulolytic enzymes (Langsford era/., 1984), C. fimi can grow on microcrystalline cellulose as the sole source of carbon and energy. It is one of the most intensively studied cellulolytic bacteria at both the biochemical and genetic levels (Beguin et al., 1987). The initial biochemical analysis of C. fimi cellulolytic enzymes has identified up to 10 different cellulolytic components in culture supematants, some of which could result from proteolysis or change in glycosylation (Langsford et al., 1984). More recently our studies of cellulose hydrolysis 6 by C. fimi extracellular enzymes have identified at least four different cellulases. The major components of this cellulolytic system are CenA and Cex, endoglucanase and exoglucanase respectively. The two other enzymes, CenB and CenC, are endoglucanases which are minor components of the system. The multiplicity of cellulolytic enzymes may be related to the broad array of substrates generated during the hydrolysis of microcrystalline cellulose. Our ultimate goal has been the characterization of all C. fimi components that are secreted and involved in cellulose hydrolysis. 1.3.1. Molecular cloning and characteristics The molecular cloning of a C. fimi cellulase gene was first reported by Whittle et ai. (Whittle et ai, 1982). So far, five cellulase encoding genes have been isolated from this organism and characterized in E. coli (Gilkes etal., 1984; Bates, 1987; Moserefa/. , 1989). Biochemical analyses of the native and recombinant forms of Cex and CenA has revealed unusual characteristics for these enzymes. On SDS-PAGE, the native endoglucanase A has an M r of 53.0 kDa, whereas its recombinant counterpart has an M r of 48.7 kDa. Native exoglucanase has an M r of 49.3 kDa but the corresponding recombinant enzyme has an M r of 47.1 to 47.3 kDa (Gilkes etal., 1988). The discrepancies observed in migration between the native and recombinant enzymes reflects glycosylation of the native enzymes (Langsford etal., 1987). C. fimigCex and gCenA are believed to have O-linked sugars which may act in protecting substrate bound forms against proteolytic cleavage. CenA and Cex contain three distinct regions: an irregular, hydroxyamino acid rich, low charge density region and an ordered, high charge-density region, separated by a short sequence of about 20 amino acids containing only Pro and Thr residues (Fig. 3). Both enzymes hydrolyze CM-cellulose but only the exoglucanase hydrolyses pNPC and MUC (O'Neill, 1986). CenA and Cex in the presence of salt bind strongly to the insoluble substrate Avicel, a microcrystalline cellulose of degree of polymerization between 100 and 250. The cellulose binding domain of CenA (CBD-CenA) is located at the N-terminal part of the enzyme whereas the cellulose binding domain of Cex (CBD-Cex) is located at its C-terminus. 7 Endoglucanase A H N 2 H N 2 I "JK>K JK>K X A COOH Exoglucanase « x -A X -J I COOH Fig 3. Structures of CenA and Cex. H denotes the Pro-Thr box E^3 denotes the active site El denotes the cellulose binding domain fO denotes the ordered charged region 8 Recently, Ong et al. (1989) have reported the construction of a fusion between the 5'-end of the abg gene of an Agrobacterium sp. encoding a 6-glucosidase and the 3'-end of the cex gene encoding the C. fimi CBD-Cex. The hybrid protein was immobilized on Whatman CF1 cellulose column and could retain 48% of its activity. The binding properties were identical to those of Cex in that the bound proteins could be eluted from the cellulose column with water. The non-covalent binding of CBD-Cex to cellulose is strong and stable enough to be used in enzyme immobilization. The CBD of CenA was also used to bind enzymes to cellulose when the CBD was fused at the N-termini (Greenwood etal., 1989). Therefore, fusions of the CBD's at either ends of a molecule can be achieved depending on which C. fimi gene is more convenient. The other C. fimi enzymes, CenB and CenC, are endoglucanases which hydrolyze CM-cellulose. CenB binds weakly to autoclaved Avicel (Owalabi et al., 1988) whereas CenC does not seem to be substrate associated (Moser, 1988). Apparently, these two enzymes are not glycosylated. A summary of the characteristics of the C. fimi c loned genes and the corresponding enzymes is shown in Table I. 1.3.2. The cex gene and its product The cex structural gene initiating with an ATG codon is 1452 bp in length and encodes a mature polypeptide of 484 amino acids (O'Neil et al., 1986a). The DNA sequence predicts a 41 amino acid leader peptide the function of which is to export the Exg to the periplasmic space of E. coli (O'Neil et al., 1986a). A more detailed analysis of the recombinant Exg has demonstrated the necessity of its leader sequence for export (O'Neil ef al., 1986b). By removing all the amino acids of the leader sequence and fusing the mature enzyme to six amino acids of the N-terminus of B-gal, most of the Exg activity remained in the cytoplasm of E. co//'(0'Neil etal., 1986b). Both secreted native and recombinant enzymes have been purified to homogeneity (Gilkes etal., 1988). Transcription of the cex gene in E. coli is not driven by the endogenous C. fimi DNA 5'-flanking sequence as shown by the requirement of an external promoter for its expression in the Gram-negative bacterium (Greenberg, 1988). 9 TABLE I. C. //m/cellulases and their genes Gene Start ORF stop Protein #A.A. Specific ref. transcript codon(bps) codon Leader Mature CMC Gly Glu cex ATG 1452 TGA Cex 41 484 ++ +/- - 1 cenk ATG 1347 TGA CenA 31 449 ++ - - 2 cenB ATG 3135 TGA CenB 33 1011 ++ + + 3 cenC GTG 3306 TGA CenC 32 1180 ++ - 4 ORF; open reading frame #A.A.; number of amino acids per polypeptide ++, +, +/-, -; indicates the relative amount ofspecific mRNA detected from cells grown on various carbon sources (CM-cellulose, glycerol, glucose). 1. O'Neill, 1986 2. Wong, 1986 3. Owalabi, 1988; Meinke, personal communication 4. Moser, 1988; Coutinho, personal communication 10 1.3.3. The cenA gene and its product The cenA structural gene initiating with an ATG codon is 1347 bps in length and encodes a mature polypeptide of 449 amino acids. The nucleotide sequence predicts a leader peptide of 31 amino acids, responsible for the export of EngA into the periplasm of E. coli (Wong et al., 1986). The location of the native enzyme is extracellular and both native and recombinant forms have been purified to homogeneity (Gilkes et al., 1984a; Wong et al., 1986). The putative cenA 5'-flanking ribosome binding site appears to be functional in E. coli but the expression of the cenA gene in E. coli is also dependent on the presence of an external promoter (Wong et ai, 1986). 1.3.4. The cenB gene and its product The cenB gene encodes a polypeptide of M r 110 kDa (Owalabi et al., 1988). From the size of this peptide, one could predict that the structural gene would require at least 3 Kb of coding sequence. Some attempts to purify this minor component of the C. fimi cellulolytic system have been unsuccessful, but the recombinant enzyme was purified by affinity chromatography on Avicel (Owalabi era/. , 1988). The coding sequence starts with an ATG codon. The predicted amino acid sequence includes a 33 amino acid long leader peptide which allows export of the recombinant EngB to the periplasm of E. coli (Owalabi era/., 1988). The transcription of cenB in E coli also requires the participation of an external promoter. The putative cenB ribosome-binding site appears to be functional in E. coli (Owalabi et al., 1988). 1.3.5. The cenC gene and its product The DNA sequence of cenC encodes two related polypeptides of M r 130 and 120 kDa. Both recombinant and native polypeptides share the same N-termini and kinetic parameters (Moser et al., 1989). These minor components of the C. fimi cellulolytic system, which are apparently not substrate associated, have been purified to homogeneity and require up to 3306 bp of DNA coding sequence (J.B. Coutinho, personal communication). The nucleotide sequence predicts putative 32 amino acid long leader 1 1 peptide initiating at a GTG codon. 1.3.6. The cbgA gene and its product The recombinant C. fimi G-glucosidase A (CbgA) is a phosphate-independent aryl-B-D-glucosidase which preferentially hydrolyses pNPG, but has some activity towards cellobiose (Bates, 1987). The level of expression of cbgA in E. coli cells carrying the recombinant plasmid pUC13:62 was low, so expression was increased considerably by deleting the 5' end of the insert. The recombinant enzyme may correspond to a secreted enzyme detected in the culture medium of C. fimi grown on CMC. 1.3.7. Transcriptional analysis of C. fimi cellulase genes Transcriptional analyses of cenA, cex, cenB, cenC and clg (or cex-linked gene) by S1 nuclease mapping identified several 5'-mRNA putative start sites (Greenberg etal, 1987a,b; Moser et al., 1989). The analysis of putative promoter sequences has revealed a region of 50 bp which display up to 64% homology that could be involved in expression of the C. fimi genes (Greenberg, 1988). The closest homologies were found with promoters from Streptomyces spp., a Gram-positive bacterium with high % G+C in its DNA. When RNA was isolated from E. coli cells carrying recombinant plasmids with putative C. fimi promoter sequences, no transcripts were found which initiated within the inserts (Greenberg et al., 1987b). However, there is no definitive evidence for the presence of the putative C. fimi promoter sequences on the cloned DNA fragments since no in vivo or in vitro transcriptional systems are available for this organism. 1.4. Heterologous expression of C. fimi cellulase genes. 1.4.1. Expression in E. coli The first report in 1982 (Whittle etal., 1982) on the heterologous expression of a C. fimi cellulase gene was in the host E. coli C600 carrying the recombinant plasmid pDW1 (a pBR322 derivative carrying a 20.2 Kb-BamH\ insert) expressing Exg activity. Soon after, three clones (namely pEC1, 2 and 3 encoding Cex, CenA and CenB, respectively) expressing various cellulase activities were isolated. Significant fractions of the activities were exported to the periplasm of E. coli (Gilkes etal., 1984). Surprisingly, pEC2 encoded a fusion polypeptide in which CenA, lacking 12 its leader peptide and the first 45 amino acids of the mature protein, was fused to the T c R determinant (Wong era/, 1986). Cloning of the entire gene led to 50% of the CenA activity being in the periplasm, but there was little activity in the culture supernatant (Wong et ai, 1986). An E. coli strain leaking up to 40% of total cellulase activity as well as other enzymes into the culture medium was isolated following mutagenesis with nitrosoguanidine (Gilkes era/., 1984b). The cex gene was over expressed by fusing it to a synthetic ribosome-binding site and an initiating ATG which were under the control of the lambda P\_ promoter contained on the runaway replication vector pCP3. Up to 20% of total E. coli protein was Cex (O'Neil era/., 1986b). The overproduced enzyme formed insoluble aggregates which could be recovered by centrifugation and solubilized in 6M urea or 5M guanidine hydrochloride. However, active enzyme was not recovered. Subcloning of cenB from pBR322 to pUC19 increased its level of expression several fold (Owolabi et ai, 1988). A total of seven polypeptides of M r 110 kDa and lower were produced in E. coli carrying pJB301 plasmid, a pUC19:cenB construct. The putative CenB leader peptide allowed transport of CenB into the periplasm of E. coli, even when the N-terminal 17 amino acids were replaced with part of the lacZ alpha-peptide (Owolabi et al., 1988) suggesting that other regions of the protein may be involved in its secretion. CenC was detected when E. coli cells carrying the recombinant plasmid pTZP-cenC were incubated at 30°C but not at 37°C. Apparently, expression of this gene reduces cell viability (Moser, 1988). Electron microscopic analysis demonstrated the aggregation of the recombinant material around the bacterial chromatin (Moser, 1988). Under noninduced condition, E. coli JM101 carrying pTZP-cenC transported to its periplasm up to 58% of the total CenC enzymatic activity. The pre-enzyme was processed correctly by E. coli leader peptidase (Moser et ai, 1989). Furthermore, a difference of about 10 kDa in size between the cytoplasmic form and the periplasmic form of the recombinant CenC suggested a proteolytic processing of more than the CenC leader peptide can account for suggesting a processing at the C-terminus of this enzyme (Moser, 1988). 13 The lack of affinity of CenC for crystalline cellulose could indicate its preference for soluble intermediates generated during cellulose hydrolysis. 1.4.2. Expression in other Gram-negative bacteria Since the native substrate containing cellulose is often characterized by a high ratio of carbon-.nitrogen, the growth of microorganisms that hydrolyze cellulose could become more efficient when supplemented with an additional source of nitrogen (Bisaria and Ghose, 1981). In an attempt to combine cellulose hydrolysis and nitrogen fixation in a single free living organism, the cex, cenA and cenB genes were subcloned individually into a broad-host-range vector and transferred by conjugation into the photosynthetic bacterium Rhodobacter capsulatus (Johnson etal., 1986). A shuttle vector compatible with Rhodobacter capsulatus DNA replicative and conjugative mechanisms allowed expression of the cellulase genes under the control of the R. capsulatus oxygen-regulated p t / f o p e r o n promoter formally known as the rxcA promoter. Significant cellulase activity was detectable in cell extracts of E. coli and the expression of the cellulase genes was dependent on the presence of the external promoter (Johnson etal., 1986). With the objective of growing non-cellulolytic bacteria on cellulose, the cex and cenA genes were placed together under the control of the divergent tet promoters of transposon Tn10 and transferred into the broad-host-range plasmid, pJRD215 (Din, 1989). Both genes were expressed in all tested bacteria (E. coli C600 MM294, E. coli K12 W4860 (Cel+), R. capsulatus B10 and K. pneumoniae M5a1). Nevertheless, none of them secreted cellulase activity into the culture medium and transformants were unable to grow on cellulose. 1.4.3. Expression in the yeast Saccharomyces cerevisiae The secretion of cellulases is required for their action on cellulose. In contrast to Gram-negative bacteria, Gram-positive bacteria and even the eucaryotic yeasts are better hosts for expression and secretion of C. fimi cellulases because of their ability to secrete various proteins into the culture medium. Expression vectors were constructed carrying the cenA or the cex gene of C. fimi and transferred into the yeast Saccharomyces cerevisiae. The coding sequences for the mature cellulases were fused to 14 leader sequences, ribosome-binding sites and promoter sequences of proteins secreted by yeast (Skipper ef al., 1985; Curry et al., 1988). S. cerevisiae did produce low levels of extracellular glycosylated cellulases but required appropriate yeast leader peptide sequences for secretion of the enzymes. Subsequently, both cellulase genes were inserted in tandem on the expression vector pMV2Adel (Wong era/., 1988). Both enzymes were produced and secreted by Saccharomyces cerevisiae using the alpha-Gal signal peptide but the cells were unable to grow on cellulose suggesting that perhaps both enzymes are not sufficient to sustain growth on this substrate. 1.4.4. Expression in Gram-positive bacteria The Gram-positive bacterium Bacillus subtilis secretes a variety of polypeptides, and is well characterized physiologically and genetically. The cex gene was subcloned into several plasmid vectors which replicate in B. subtilis including an E. coli/ B. subtilis shuttle vector. The cex gene was not maintained in B. subtilis, being excised from all the plasmids tested (D.J. Whittle and V. Tai, personal communication). The instability of the gene may have resulted from the marked difference in the moles per cent G+C contents of B. subtilis DNA (41%) and that of C. fimi DNA (72%). 15 1.5. Cloning systems in Coryneform bacteria Coryneform bacteria are part of the family of Gram-positive, aerobic, non-sporulating irregular rod shaped bacteria with a high genomic % G+C content (54% to 73%) that are widely distributed in nature (Komagata e r a / . , 1969; Bousfield and Callely, 1978). These include Corynebacterium, Brevibacterium, Arthrobacter, Microbacterium and Cellulomonas genera. Some are pathogens of plants and animals, including man, others are used in industrial amino acid fermentations (Yamada e r a / . , 1972). Despite their physical similarities, DNA hybridization studies as well as 16S and 5S rRNA sequence analyses suggest a clear heterogeneity amongst the species included in these genera (Martin et al., 1987). Most genetic studies on coryneform strains have focused on the isolation and characterization of amino acid biosynthetic mutants for the production of high levels of selected amino acids. More recently, recombinant DNA techniques through the development of appropriate cloning vectors and transformation systems have become available in several Corynebacteria (Santamaria et al., 1984; Katsumata et al., 1984; Santamaria e r a / . , 1985; Yoshihama et al., 1985; Yeh et al., 1986; Singer, 1986; Shaw and Harley, 1988). A large number of endogenous plasmids has now been reported in Corynebacteria and used as starting material for vector construction (Martin era/. , 1987). Since most of the plasmids are cryptic, this involves the introduction of selectable markers into endogenous plasmids. The selectable markers used so far have been mostly antibiotic resistance genes isolated from different bacteria. Nevertheless, a need for additional markers is evident since only a few are expressed enough to allow direct selection: the aph II gene from Tn5, the genes conferring streptomycin and spectinomycin resistance from pCG4, the chloramphenicol and hygromycin resistance genes from Streptomyces and perhaps a few more markers that are still poorly characterized (Martin etal., 1987). Problems related to the construction of cloning vectors in Corynebacteria is the lack of knowledge of the indispensable 16 sequences involved in plasmid establishment and instability of large hybrid plasmids likely due to recombination proficiency of the bacterial host. Furthermore, all published transformation procedures for Corynebacteria are based on the treatment of protoplasts with DNA and PEG. The removal of the corynomycolic acid-containing cell wall is difficult to achieve without affecting cell viability, thereby lowering transformation efficiency. Only one in 10,000 protoplasts is transformed under the best experimental conditions (Martin etal., 1987). Other factors, like the secretion of nucleases into the medium and the existence of a restriction-modification system in the host cell leading to low frequencies of transformation have been reported (Katsumata et al., 1984; Smith et al., 1986; Paradis et al., 1987). More recently, transformation of bacterial cells by electroporation techniques has been achieved in Corynebacteria and seems promising in terms of better efficiencies (Chassy et al., 1988). 1.6. The first objective of this thesis. From the perspective of studying heterologous promoter functionality and secretion of cellulases, two C. fimi cellulase genes were transferred into the closely related strain Brevibacterium lactofermentum (Paradis etal., 1987; this thesis). The use of this industrial strain was believed to allow secretion of the enzymes in the culture medium because of the properties of the leader peptides present on those enzymes. Furthermore, the presence of putative C. fimi promoter sequences on the cloned fragments may allow expression of their corresponding genes or display some promoter functionality in B. lactofermentum allowing the analysis of those involved in gene expression. B. lactofermentum is a coryheform bacterium whose genomic DNA contains 64 mol % G+C (Bousfield and Callely, 1978). It and several derivatives are used industrially for the production of L-glutamic acid, L-phenylalanine, L-leucine, L-threonine and L-lysine (Yamada et al., 1972a; Nakamori era/., 1987). It contains the 4.4 Kb cryptic plasmid pBLI (Santamaria etal., 1984) that has been used widely in the 17 construction of cloning vectors (Santamaria et al., 1984; Miwa ef al., 1985; Yeh etal., 1986; Martin et al., 1987), one of which was used to express the B. subtilis alpha-amylase gene in B. lactofermentum f rom its own promoter (Smith etal., 1986). As well as being a coryneform bacterium with high % G+C genomic content, B. lactofermentum possesses a well established transformation procedure (Santamaria ef ai, 1984). Expression of cloned C. fimi genes in the related organism Brevibacterium lactofermentum might allow the role of the individual enzymes to be determined in the absence of other contaminating cellulases. Furthermore, because the gene products might be secreted, production and purification of enzymes could be simplified. 18 1.7. Characteristics of B-glucosidases (3-Glucosidases (B-D-glucoside glucohydrolase ; EC 3.2.1.21) are enzymes which hydrolyze various compounds with 6-D-glucosidic linkages. They are widely distributed in nature and have been isolated from bacteria, fungi, plants and animals (Shewale, 1982; Woodward and Wiseman, 1982; Dinur et al., 1986). This discussion considers only those involved in the hydrolysis of cellulose to glucose as a component of a cellulase complex. B-Glucosidases are not strictly speaking cellulases but they are very important components of cellulase systems. B-Glucosidases can hydrolyze cellobiose, remove glucose from the non-reducing ends of cello- oligosaccharides, transglycosylate and hydrolyze various artificial B-D-glucosides (Coughlan, 1985). Their transglycosylation activities may generate inducers of the synthesis of cellulase complexes (Coughlan, 1985). Low levels of B-glucosidase activity can result in poor saccharif ication of cellulose because of the inhibition of endoglucanases and cellobiohydrolases by cellobiose (Shewale, 1982). Adequate levels of B-glucosidase activity are therefore essential to ensure maximal conversion of cellulose to glucose. Before the hydrolysis of cellulose becomes a commercial process, several problems concerning B-glucosidases must be addressed. Possible solutions to these problems have already been suggested (Woodward and Wiseman, 1982). The inhibitory effect of glucose could be removed by either chemical modification of the glucose binding site with protection of the substrate binding site or conversion of glucose to fructose by added invertase. The lack of secretion of the enzymes could be overcome by using an appropriate secretion system. The thermo-instability of the enzymes during a reaction could be decreased by addition of Pro residues within the enzymes where the stability of a protein could be increased by changing specific residues by site-directed mutagenesis leading to a decrease in configurational entropy of unfolding (Matthews etal., 1987). The susceptibility of the enzymes to proteolytic cleavages could be changed by the addition of covalently 19 bound carbohydrates and protease inhibitors. 6-Glucosidases are intracellular, cell-bound or extracellular in different organisms, or in the same organism depending on the type of inducer. B-Glucosidases are usually secreted into the culture medium by cellulolytic fungi (Moldoveanu and Kluepfel, 1983). Bacterial B-glucosidases are often intracellular or cell associated (Gong and Tsao, 1979). Nevertheless, extracellular bacterial B-glucosidases have been reported in Clostridium stercorarium (Bronnenmeier and Standenbauer, 1988), Bacteroides succinogenes (Groleau and Forsberg, 1981), from an hybrid of Cellulomonas species and Bacillus subtilis (Gokhale et al., 1984; Gokhale and Deobakar, 1989) and also from the bacterium Acetivibrio cellulolyticus (Saddler and Khan, 1981). The M r for fungal B-glucosidases, which may contain 0 to 90% (w/w) carbohydrates, ranges from 35000 for an extracellular enzyme of T. reesei up to 480000 for the G1 enzyme of Monilia sp. The B-glucosidase of A. fumigatus is possibly a dimer with an M r of 340000 and the Botryodiplodia theobromae enzyme with an M r of about 350000 could be constituted of 8 subunits, each composed of 4 identical non-catalytic polypeptides (Coughlan, 1985). The M r for bacterial B-glucosidases may range from 50000 for the enzyme of C. thermocellum up to 122000 for the enzyme of E. herbicola (Woodward and Wiseman, 1982). The Agrobacterium sp. enzyme has an M r of 100000 and could be a dimer (Day and Withers, 1986). 1.7.1. Substrate specificity and measurement of activity The B-glucosidases recognize a wide variety of aglycones provided the glucosyl moiety has a B-configuration (Fig.4). The rate of hydrolysis depends on the nature of the aglycone moiety. The enzymes are sometimes specific for the hydroxyl group at position C 4 . The determination of K m and Vmax at different pH values during the hydrolysis of cellobiose by B-glucosidases could indicate the involvement of a carboxylate group and a protonated nitrogen atom of an imidazole group in catalysis. The first step involves splitting of the 20 FIG. 4. STRUCTURE OF VARIOUS B-GLUCOSIDES. R1, 1-0-METHYL-B-D-GLUCOSIDE; R2, CELLOBIOSE; R3, 4-NITROPHENYL-6-D-GLUCOSIDE(PNPG); R4, 4-METHYLUMBELLIFERYL-B-D-GLUCOSIDE (MUG); 5 21 aglycone moiety with simultaneous formation of a glucosyl-enzyme complex which then reacts with water, yielding glucose. Reactions catalysed by B-glucosidases may be assayed by measuring the glucose produced. The glucose is then measured spectrophotometrically using either a coupled glucose oxidase system or a coupled glucose-6-phosphate dehydrogenase and NADP system, or chemically. These assay methods are accurate and sensitive but laborious and time-consuming. An alternative method which allows the reaction to be followed continuously involves following the production of glucose at 340 nm by making use of ATP, hexokinase, glucose-6-phosphate dehydrogenase and NADP. The latter is converted into its reduced form NADPH which absorbs strongly at 340 nm (Hsuanya and Laidler, 1984). More directly, B-glucosidase activity can be monitored in one step by measuring spectrophotometrically either the release of p-nitrophenol (pNP) from p-NP-B-D-glucoside at 410 nm (Stoppock et al., 1982) or of 4-methylumbelliferol (4-UM) from 4-MU-B-D-glucoside at 340 nm (Sprey and Lambert, 1983). Cellobiose cleaving activity should not be determined by using aryl-B-glucosides or by using assays not optimized for phosphorolytic cleavage, as previously mentioned by Schimzefa/. , (1983). More recently, an automated high-pressure liquid chromatography (HPLC) system allows the quasicontinuous analysis of all initial, intermediary and end products that appear during the hydrolysis of oligosaccharides (Schmid and Wandrey, 1989). Additional analysis can be performed by anion-exchange chromatography of the respective carbohydrate borate complexes with an automated sugar analyser (Schmid and Wandrey, 1989). 1.7.2. Classification of B-glucosidases In general, B-glucosidases are separable into two groups: those in the first group have high affinities for aryl- and alkyl-B-D-glucosides but low affinity for cellobiose and are known as aryl-B-D-glucosidases; those in the second group are highly active on cellobiose and higher 22 cellodextrins and are known as true cellobiases. Nevertheless, the distinction between these two groups is not clear cut because many 6-glucosidases hydrolyze both aryl-6-D-glucosides and cellobiose efficiently. 8-Glucosidases hydrolyzing preferentially aryl- and alkyl-(3-glucosides have been reported in Clostridium thermocellum (Ait etal., 1979), Flavobacterium (Sano etal., 1975), Kluyveromyces fragilis (Raynal and Guerineau, 1984), Aspergillus niger (Pentilla etal., 1984 ) , Cellulomonas uda CB4 (Nakamura etal., 1986), Cellulomonas fimi (Wakarchuk et al., 1984) and Escherichia coli (Schaefler, 1967). Even though the bgl operon of E. coli encodes all functions necessary for the regulated uptake and utilization of aryl-8-glucosides, the operon is cryptic and requires mutational activation. B-Glucosidases which preferentially hydrolyze cellobiose (or true cellobiases) have been reported in Clostridium thermocellum (Alexander, 1968), Cellulomonas fimi strain R2 (Sato and Takahashi, 1967), Cellulomonas sp., Cellulomonas flavigena, Cellulomonas cartalyticum, Cellulomonas uda ( Schimz et al., 1983), Escherichia adecarboxylata (Armentrout and Brown, 1981), Agrobacterium faecalis (Han and Srinivasan, 1969) and E. coli (Kricker and Hall, 1984). In E. coli, the eel operon for cellobiose catabolism is not expressed in wild-type strains but as for the bgl operon, it requires activation by mutational event to sustain bacterial growth on this substrate. Several pathways of cellobiose utilization occur in nature. It can be degraded by hydrolysis, by oxidation to cellobionic acid followed by hydrolysis, by ATP-dependent phosphorylation of the C-6-OH at the non-reducing moiety of cellobiose followed by hydrolysis or by inorganic phosphate dependent phosphoro lys is leading to equimolar concentration of glucose-1-phosphate and D-glucose (Schimz et al., 1983). The last two mechanisms occur frequently in bacteria which do not secrete the enzymes into the culture medium. In cellulolytic fungi, the occurrence of two kinds of B-glucosidases, 23 Table II. K m s for cellodextrin hydrolysis by various exo-B-1,4-gluco-sidases from cellulolytic fungi K m (mM) Enzyme C2 C3 C4 C5 C6 Ref. Aspergillus niger Enzyme 1 1.1 0.49 0.13 0.26 0.36 1 Enzyme 2 1.64 0.54 0.087 0.31 0.49 Enzyme 3 0.91 0.52 0.085 0.63 0.72 A. phoenicis 0.75 NA 0.36 NA NA 2 Pyricularia orysae GB-1 0.91 0.34 0.26 NA NA 3 Sclerotium rolfsii BG-1 3.6 1.0 0.5 0.49 0.62 4 BG-2 3.07 1.23 0.85 0.40 0.37 BG-3 5.84 1.98 0.83 0.76 0.37 BG-4 4.15 0.70 0.50 0.55 0.67 Trichoderma koningii 0.84 0.41 0.16 0.095 0.22 5 T. viride 1.5 NA 0.30 NA NA 6 T. reesei 0.54 0.08 0.15 0.16 0.18 7 NA: not available 1. King and Smibert, 1963; 2. Sternberg et al., 1977; 3. Hirayama et al., 1978; 4. Shewale and Sadana, 1981; 5. Halliwell and Vincent, 1977; 6. Berghem and Pettersson, 1974; 7. Schmid and Wandrey, 1989; 24 one with aryl-6-glucosidase and one with a broad activity towards different glucosides, is also a common feature. A decrease of K m with increase in the chain length of the cello-oligosaccharides (at least up to cellopentaose) has been reported for several 6-glucosidases isolated from various fungi (Table II). The exoglucosidase activities of these enzymes were demonstrated by chromatographic identification and quantitative determination of the products of cellodextrin hydrolysis. The enzymes are classif ied as exo-6-1,4-glucan glucohydrolase (or exoglucosidases) because of their sequential action on cellodextrins. Finally, it seems that 6-glucosidase in cellulolytic fungi is a multifunctional enzyme in the sense that it is not merely a cellobiase but rather an exo-6-1,4-glucan glucohydrolase. 1.8. The preliminary characterization of C. fimi 8-glucosidases The first report on C. fimi 6-glucosidases described the presence in the cell extracts of at least two enzymes, one an aryl-8-D-glucosidase and one a 6-D-glucoside glucohydrolase. Both activities are inducible by growth on cellulosic substrates (Wakarchuk et ai., 1984). A 6-glucosidase-encoding clone reacting with antiserum to C. fimi culture supernatant, was isolated together with the other cellulase genes (Gilkes etal., 1984) and partially characterized ( Bates, 1987). The recombinant enzyme, Cbg, did not require phosphate for the hydrolysis of pNPG or cellobiose. Gene expression in E. coli requires an E. coli promoter (Bates, 1987). Zymograms using MUG as a substrate identify comigrating bands of activity in C. fimi culture supernatants and in extracts from E. coli cells expressing the gene. This suggested that the recombinant enzyme corresponds to a 8-glucosidase secreted by C. fimi. This is unusual because extracellular 6-glucosidase activity was not detected previously in C. fimi culture (Wakarchuk etal., 1984). Furthermore, a series of 5'-end deletions of the cloned C. fimi DNA fragment gave a mutant called pUC13:62A31, with increased activity. Apparently, the increase in expression resulted from a more optimal spacing between the Plac promoter and the cbgA RBS (Bates, 1987). More recently, Kim and Pack have reported the molecular cloning in E. 25 coli of two C. fimi B-glucosidase encoding genes (Kim and Pack, 1989). For the expression of 6-glucosidase activity in E. coli, both genes require the presence of an E. coli promoter. One enzyme hydrolyses both pNPG and cellobiose, the other only pNPG (Kim and Pack, 1989). The corresponding native enzymes were not identified in C. fimi; the recombinant enzymes themselves must be characterized further. 1.9. The second objective of this thesis The general goal of our group is to characterize every secreted component of the C. fimi cellulase complex involved in the hydrolysis of cellulose to glucose. So far, five genes encoding cellulolytic enzymes have been isolated and characterized both at the protein and DNA levels. In C. fimi very little is known about its B-glucosidases. During the characterization of the recombinant CbgA and identification of the corresponding native enzyme, several other enzymes with 6-glucosidase activity were identified. Apparently, up to three different enzymes, based on their catalytic properties, were detected along with the CbgA enzyme which is the only one found secreted in the culture medium. The second part of this thesis describes the characterization of the recombinant enzyme CbgA both at the genetic and biochemical levels. 2 6 2. MATERIALS AND METHODS 2.1. Bacterial strains, plasmids, phage and growth conditions Table III list the bacterial strains, plasmids and phage used. C. fimi strain ATCC 484 was grown at 30°C with agitation in low salt-LB medium (Bacto-tryptone, 10 g; Bacto-yeast extract, 5 g ; NaCl, 0.5 g; pH 7.0) or in basal medium (Stewart and Leatherwood, 1976) supplemented with 1 % (w/v) of an appropriate source of carbon (glucose, glycerol, cellobiose or CMC (Sigma; low viscosity)). Brevibacterium lactofermentum strain ATCC 21798 was grown at 30°C with agitation in TSB or MMYC media (Santamaria etal., 1984). E. coli strains were grown at 37°C with agitation in LB, TYP or M9CA media (Maniatis et al., 1982). Solid media contained 1.5% (w/v) agar (Difco). If the media contained CMC, the agar was reduced to 1.1% (w/v). Ap and Km sulfate (Sigma Co., St. Louis, MO) were used at a final concentration of 5 to 100u.g/mL, as required. 2.2. Enzymes and reagents Restriction endonucleases, DNA and RNA modifying enzymes were purchased from various sources (Bethesda Research Lab., Pharmacia, New England Biolabs and Boehringer-Mannheim Inc.) and used according to the supplier. Chemicals for electrophoresis were supplied by Bio-Rad laboratories. Nitrocellulose membranes BA85 were from Schleicher & Schuell Inc. Radionucleotides were from New England Nuclear (Dupont-NEN) and Amersham Inc. Yeast tRNA and redistilled phenol were from Bethesda Research Laboratories. Ribonuclease A, pNPG, pNPC, MUG and MUC were from Sigma Chemical Co. 2.3. Transformation of bacteria by plasmid DNA E. coli cells were transformed by a modification of the calcium chloride procedure (Mandel and Higa, 1970). B. lactofermentum cells were transformed by a protoplast fusion technique according to Santamaria etal., (1984) with some minor modifications. Briefly, 27 Table III. Bacterial strains, plasmids and phage Bacterial strains Relevant characteristics references E. coli JM83 ara A(lac-pro AB) rpsL 080 lacZ AM15 1 E. coli JM101 supE thi A(lac-proAB)[F tra A36 proAB 1 lacl-ZAM15 E. coli RZ1032 Hfr KL16 PO/45[lysA(61-62)] dut1 ung1 2 thi l relA1 zbd-279:.1n10 supE44 B. lactofermentum aecR ATCC21798 C. fimi cellulose hydrolysis ATCC484 Plasmids Relevant characteristics pUC8 to 19 ApR l a c Z 3 pUC8:cenA ApR lac Z' CenA 4 pUC12:A25 ApR lac Z' Cex 5 pUC4K ApR lac Z K m R 6 pBR322 ApR TcR 7 pBL1 cryptic 8 pBK10 ApR KmR 4 pFWPIO ApR KmR Cex 4 pFWP30 A p R KmR CenA 4 pFWP1030 ApR KmR Cex CenA 4 pEC62 ApR CbgA 9 pTZ19R ApR lac Z' orifl 10 Phage Relevant characteristics M13K07 KmR 10 1. Yannish-Perron etal., 1985; 2. Kunkel etal., 1987: 3. Messing, 1983; 4. Paradis et al., 1987; 5. O'Neil, 1986; 6. Oka etal., 1981; 7. Boyere fa / . , 1977; 8. Santamaria et al., 1984; 9. Bates, 1987; 10. Viera, 1985. 2 8 protoplast formation was monitored by microscopy and by increase in osmotic sensitivity after hypotonic dilution. Protoplasts were fused and 50 uJL aliquots of the fusion mixture were plated on RS regeneration medium. After 24 h regeneration at 30°C, the agar was overlayed with TSB medium precooled to 37°C and containing 0.4% low-melting agarose and 50 ug Kanamycin sulfate/mL Putative transformants were transferred to TSA media after five to ten days of incubation at 30°C and tested for cellulolytic activity. Clones expressing cex were detected by fluorescence on agar containing MUC (final concentration 500 u.M). Those expressing cenA were detected on agar containing 1 % CMC, using Congo red to visualize the zones of hydrolysis (Teather and Wood, 1982). 2.4. Enzyme assays and protein determination Cell extracts were prepared by concentrating the cells 10 to 50-fold by centrifugation and resuspending the cells in 50 mM K-phosphate buffer (pH 7.0). The suspensions were either sonicated using a Bronson sonifier with a microprobe set at an intensity of 2 for 3 bursts of 30 sec or ruptured with a French press for larger volumes. The exoglucanase activity of cell extracts was determined with pNPC as a substrate (Gilkes etal., 1984a). The endoglucanase activity of cell extracts was determined by DNS assay (Miller, 1959) using CM-Cellulose as substrate (Gilkes etal., 1984a). 6-Glucosidase activity of cell extracts was determined spectrophotometrically by following the release of p-nitrophenol from pNPG at 410 nm (Stoppok ef al., 1982; Berghem and Pettersson, 1973). Glucose-6-phosphate dehydrogenase activity, a cytoplasmic enzyme marker, was measured by following the reduction of NADP at 30°C and 340 nm according to Worthington (1977). B-Lactamase activity, a periplasmic space marker, was measured by following the hydrolysis of the chromogenic substrate nitrocefine at 37°C and 486 nm according to O'Callaghan etal. (1972). NADH oxidase activity, an inner-membrane marker, was measured by the method of Osborn etal. (1972). 2-Keto-3-deoxyoctonate content, a sugar 29 exclusively found in the LPS of outer membranes, was estimated by the modified method of Osborn etal. (Darveau and Hancock, 1983). Protein concentration was determined according to Bradford (1976) using BSA as standard. One unit (U) of enzyme is defined as the amount of enzyme required to release 1 umol of product per min. 2.5. Fractionation of bacterial cells E. coli cells were fractionated into periplasmic, cytoplasmic and total membrane fractions, and C. fimi cultures were fractionated into cytoplasmic, total membrane and culture supernatant fractions according to Gerhardt et al., (1981) with minor modifications. 2.5.1. Fractionation of E. coli cells The location of recombinant CbgA in E. coli JM83 carry ing pUC13:62A31 was established by fractionating the cells and measuring B-glucosidase activity. Cells were grown in 200 mL of Luria broth at 37°C with agitation. Exponential and stationary phase cultures (O.D.soOnm °f °-8 and 2.3, respectively) were chilled on ice for 5 min and harvested at 8000 X g for 15 min at 4°C in a JA 20 rotor (Beckman). Periplasmic fractions were first prepared by a standard osmotic shock procedure (Nossal and Heppel, 1966). The shocked cells were recovered by centrifugation and re-suspended in 1/20 vol ice-cold 10 mM HEPES buffer, pH 7.4. The cells were ruptured with a French press. Cell debris was removed by centrifugation at 8000 X g for 10 min at 4°C and membranes were sedimented by ultracentrifugation at 110 000 X g for 2 hrs at 7°C. The supernatants were decanted and kept as cytoplasmic fractions. The membranes were resuspended in 10 mM HEPES buffer (pH 7.4) containing 5 mM-EDTA and further analysed by isopycnic sucrose density gradient centrifugation (Osborn and Munson, 1974). 2.5.2. Fractionation of C. fimi cells C. fimi cells were grown to late log phase (O.D.goOnm o f 1 -7) in 1 L 30 of 1 % CM-cellulose containing basal medium, chilled on ice for 5 min prior to collection at 8000 X gr for 15 min at 4°C in a JA20 rotor (Beckman). Proteins in the culture medium were concentrated by ultrafiltration through a PM10 membrane (Amicon) at 4°C for 12 to 16 hrs. Protein aggregates containing a substantial amount of the total pNPGase activity detected in the culture medium were obtained from the concentrated material by ultracentrifugation. Cytoplasmic proteins and crude envelope fractions were obtained as described above for E. coli. 2.6. The purification of 8-glucosidases 2.6.1. The purification of recombinant CbgA E. co//7pUC13:62A31 was grown for 8 hrs at 30°C in 40 L of Luria broth in a LH Scientific 5000 series Fermenter (air pressure at 1 bar, air flow at 20 L/min, agitation at 200 rpm). Cells were harvested at an O.D. 600nm °f 0 7 using an air-driven Sharpies continuous flow centrifuge (Laboratory Presurtite centrifuge), set at 50 000 RPM and 20 PSI, at a flow rate of 700-850 mL/min. Turbidity of the culture supernatant was observed periodically to follow the efficiency of centrifugation. The cell paste was resuspended in 400 mL of ice cold buffer A (50 mM Tris-CI (pH7.0); 5 mM EDTA; 1 mM PMSF; 1 0 ' 7 M Pepstatin A). Cytoplasmic proteins were prepared by rupturing 100 mL of the cell suspension with a French press. Cell debris was removed by low speed centrifugation and the supernatant was further clarified by ultracentrifugation at 110 000 X g for 80 min at 7°C. To the clarified extract was added 1.5% (w/v) streptomycin sulfate and the precipitated nucleic acids removed by centrifugation. Two ammonium sulfate cuts (the first cut 28%, the second cut 45%) were performed on the supernatant fraction, where precipitates were removed by centrifugation. The 45% cut precipitate was dissolved in buffer B and dialysed overnight at 4°C against buffer B (Tris-CI 10 mM-EDTA 1 mM (pH7.0)). The dialysate was fractionated at 22° by ion-exchange chromatography on DEAE-Sephadex A-50-120. Bound proteins were eluted at 44 ml_/hr by stepwise increase in the NaCl salt concentration from 0 to 1 M. Fractions of 2.5 mL each were collected and 31 the active fractions from two peaks (I and II) containing 6-glucosidase activity were pooled separately (Pool I and Pool II). Proteins in each pool were concentrated by ultrafiltration through a PM10 membrane (Amicon), salt concentration was decreased 10 fold by dilution with buffer B and samples were further concentrated to a final volume of 5 mL by ultrafiltration. The final purification was achieved by anion-exchange chromatography of 1.0 mL of desalted Pool I on Mono Q resin (Pharmacia). The speed of elution was set at 1 mL/min and bound proteins were eluted using a 0 to 1 M NaCl salt gradient. 2.6.2. The purification of C. fimi 6-glucosidases. 10 L of 1%-CM-cellulose basal medium were inoculated with a preculture containing C. fimi cells and the culture grown at 30°C to an O.D. 600 nm °* The culture was chilled on ice, then centrifuged at 4400 X g for 10 min at 4°C. The cells were suspended in 100 mL of ice cold buffer A. The suspension was passed 10X through a French press. The resulting extract was fractionated as described above. Active fractions with pNPGase/cellobiase, pNPGase and mainly cellobiase activities were grouped as Pool I, Pool II and Pool III, respectively. The 6-glucosidases present in Pool II and III were further purified by anion-exchange chromatography on Mono Q resin as described above. 2.7. Characterization of purified recombinant CbgA 2.7.1. The effects of pH and temperature on CbgA activity Reaction mixtures contained either 1.4 ug of purified CbgA m L - 1 or 140 ug of a crude cell extract m L - 1 , 100 mM phosphate of the required pH, and 5 mM pNPG in a total volume of 1 mL. After incubation at the appropriate temperature for 60 min, the reactions were stopped by the addition of 1/2 vol 1 M Na2C03 and the absorbance measured at 410 nm. 2.7.2. Determination of the kinetic parameters for CbgA The kinetic parameters of the purified recombinant enzyme CbgA 32 were determined from Lineweaver-Burke plots for each 6-glucosidic substrates used in this study (see appendix). All reactions were done in 1 mL final volume in a thermostatted cuvette chamber at 37°C and followed using a spectrophotometer (Varian DMS 100 UV Visible) connected to a Screen-Plotter (SANYO) and laboratory strip chart recorder. Each assay contained 1 u.g of the purified recombinant enzyme from E. coli JM83/pUC13:62A31 strain, 1 X reaction buffer (10 X stock ; 200 mM KPO4 (pH 7.0)-50 \ig BSA/mL-1 mM DTT) and various concentrations of substrates (0.04-8.0 mM pNPG, 0.025-5.0 mM pNPC, 0.5-50 mM cellobiose, 0.25-10 mM cellotr iose, 0.25-10 mM cellotetraose, 0.04-5.0 mM cellopentaose). The hydrolysis of p-nitrophenyl derivatives (pNPG and pNPC) was followed at 410 nm and the molar extinction coefficient for p-nitrophenol at pH 7.0 was estimated from standard solution (Sigma). The hydrolysis of cellodextrins (cellobiose from Sigma; cellotriose, cellotetraose and cellopentaose kindly provided by N.R. Gilkes from U.B.C.) to glucose was followed by making use of a glucose coupled assay (Sigma kit 15-10) containing hexokinase and glucose-6-phosphate dehydrogenase and following the reduction of NADP at 340 nm (Hsuanyu and Laidler, 1984). The hydrolysis of all 6-glucosides was initiated by adding the recombinant CbgA to the prewarmed reaction mixture. Hydrolysis was not detected in the absence of enzyme with all substrates tested. Glucose inhibition (0-11 mM) was measured on pNPG (0.1, 0.3 mM and 1.0 mM) as substrate (Cornish-Bowden, 1979). 2.8. Electrophoretic analysis of proteins Proteins were analysed by electrophoresis in 7.5 or 10% polyacrylamide gels in the presence of 0 . 1 % sodium dodecyl sulfate (SDS) (Laemmli, 1970) and gels were stained with 0.03% Coomassie brillant blue (R-250) dissolved in 10% acetic acid and 50% methanol. Excess stain was removed by soaking the gel in 10% acetic acid-50% methanol until the background was negligible. When necessary, protein samples were denatured by boiling for 2 min. Calibration standards 33 used were myosin (205 kDa) ; 6 -ga lac tos idase (116 kDa) ; phosphorylase B (97.4 kDa); bovine serum albumin (66 kDa); catalase (57 kDa); glutamate dehydrogenase (53 kDa); ovalbumin (45 kDa); alcohol dehydrogenase (43 kDa); glyceraldehyde-3-phosphate dehydrogenase (36 kDa) and carbonic anhydrase (29 kDa). For zymograms, unheated samples were appl ied to 7% polyacrylamide gels in standard sample loading buffer containing SDS and B-mercaptoethanol. After electrophoresis, bands of P3-glucosidase activity were visualized by soaking the gel for 30 min at 30°C in a solution containing 1 mM 4-methylumbelliferyl- B-D-glucoside (MUG)-0.1 M potassium phosphate (pH 6.8)-1% Triton X-100 solution (Lacks and Springhorn, 1980). Fluorescent bands were visualized under shortwave UV light and the gel was photographed using type 57 film (Kodak). 2.9. Production of antiserum and Western blot analysis A total of 150 (ig of purified recombinant CbgA was mixed with Freund's complete and incomplete adjuvant (1:1). At intervals of two weeks, 1/3 of the mixture was injected intramuscularly into a New Zealand white rabbit. One week after the third injection, the rabbit was bled, the serum collected and tested against E. coli and C. fimi proteins in conjunction with the rabbit preimmune serum using an enzyme-linked-immunoadsorbant assay (ELISA, figure 5A and 5B) as previously described (Voller etal., 1976). The antiserum was made 0.02% in sodium azide and aliquots were kept at 4°C and -20°C. Polypept ides from polyacrylamide gels were blotted onto nitrocellulose membranes by electrotransfer and detected by immunoblotting as previously described (Towbin etal., 1979) using the alkaline phosphatase/5-bromo- 4-chloro-3- indolyl-phosphate (X-phosphate) detection system (Blake etal., 1984). Prestained calibration standards used were myosin (rabbit skeletal muscle, 210 kDa); phosphorylase B (rabbit muscle, 110 kDa); bovine serum albumin (75 kDa); ovalbumin (hen egg white, 45 kDa); soybean trypsin inhibitor (29 kDa); B-lactoglobulin (18.4) and lysozyme (hen egg white, 14.3 kDa) (BRL). 34 Fig. 5. Titration of antiserum by ELISA on (A) E. coli and (B) C. fimi proteins. Test antigens were coupled in triplicate to microtiter plates using the following amounts of protein per well: (A) 50 ng of purified CbgA, 10 ug of E. co//7pUC13:62A31 total cell extract and 10 ug of E. co///pUC13 total cell extract; (B) 10 ug of C. fimi cytoplasmic fraction, 10 ug of C. fimi membrane fraction and 2 ug of C. fimi culture supernatant protein aggregate fraction. The pre-immune (open symbols) and immune (black symbols) rabbit sera were tested. The absorbances at 405 nm reflect the mean values of the activity of alkaline phosphatase conjugated to goat IgG antibodies against rabbit IgG bound to the test antigens. o o r-~ CO i o I E c in o 2, Q b 2.0 r CbgA (0.05|ig/0.2U) A31 (10ng/0.8U) pUC13 (10jig) CbgA (0.05ng/0.2U) A31 (10|ig/0.8U) pUC13 (10ng) dilution B o o r-co i o CM in o Q 6 1.0 r 0.5 o.o Cyto. (10ug/0.15U) Memb.(10ug/0.3U) S/N (2ng/0.15U) Cyto. (10ag/0.15U) Memb.(10ug/0.3U) S/N (2ug/0.15U) dilution 35 2.10. Determination of the NH2-terminal amino acid sequences and amino acid compositions of recombinant CbgA polypeptides The NH2-terminal amino acid sequences were determined by automated Edman degradation using an Applied Biosystems model 470A gas sequenator utilizing the resident sequencing program. These analyses were kindly performed by Sandy Kielland at the University of Victoria sequencing facility. Amino acid composition was determined at the same facility. 2.11. DNA and RNA methodology 2.11.1. Isolation of plasmid DNA and restriction digest analysis Plasmid DNA was isolated from E. coli strains by a modification of the alkaline-lysis procedure (Birnboim and Doly, 1979). When required, the DNA was purified further by centrifugation to equilibrium in CsCl density gradients containing ethidium bromide (Maniatis etal., 1982). Plasmid DNA was isolated from B. lactofermentum by a modification of the procedure of Santamaria etal., (1984). B. lactofermentum cells were grown overnight in TSB medium. The cells were collected from a 1.5 mL sample by centrifugation at 12,000 X grand resuspended in 200 uL of lysis buffer containing lysozyme at 3 mg/mL final concentration. The mixture was incubated at 37°C for 4 hrs with occasional mixing. Protoplasts were lysed by adding 400 uL of a 1 % SDS-0.2 N NaOH solution prewarmed at 65°C and mixing immediately by inversion. After 5 min incubation at room temperature, 300 u.L of ice cold 3 M sodium acetate, pH 5.2, were added and mixed in for 10 sec with a vortex mixer. Tubes were kept on ice for 15 min, then the turbid lysates were spun at 12,000 X g tor 10 min at 4°C. The decanted supernatants were extracted once with 1 vol of phenol and once with 1 vol of chloroform-isoamyl alcohol (24:1). The nucleic acids were precipitated by adding 1 vol of cold isopropanol and the mixtures were stored at -80°C for 30 min. The tubes were centrifuged in a microfuge (Fisher) for 10 min at 4°C. The pellets were rinsed with 70% ethanol, dried and then incubated at 36 37°C for 30 min in 25 p± of TE buffer containing RNase A at 50(j.g/mL final concentration. Single stranded pTZ DNA was isolated by a modification of the procedure of Kristensen etal. (1987). A 2.0 mL overnight culture of M13K07 infected JM101 was centrifuged. 1.5 mL of the supernatant were mixed with 200 u l of 25% PEG-3.5 M NaCl and incubated on ice for 30 min. After centrifugation at 12,000 X g for 15 min, the phage pellet was treated with 1.0 mL of 4 M NaCIC>4-10 mM Tris pH 7.5-1 mM EDTA. The DNA released was adsorbed to a glass fiber filter (GF/C from Whatman International Ltd) of 7 mm diameter. The filters were washed 3 X with 500 \iL of 70% ethanol. The DNA was eluted with water. DNA restriction f ragments were resolved by agarose gel electrophoresis (0.6 to 1.5% (w/v) agarose). Restriction fragments from lambda DNA digested with Hind\\\ or /-//Vidlll-EcoRI served as size markers. 2.11.2. Isolation of RNA and RNase-free work All chemicals and reagents used during RNA work were purchased only for this purpose and kept separate from regular laboratory supplies to avoid RNase contamination. The glassware used was either baked at 300°C for 3 hrs or bought as disposable labware. When required, solutions were treated with 0.2% (v/v) diethylpyrocarbonate (DEPC) as described elsewhere (Ehrenberg etal., 1976; Maniatis etal., 1982). Plastic pipette tips and microfuge tubes were sterilized by autoclaving only. RNA was isolated from all bacterial strains by a modification of published procedures (Greenberg etal., 1987a). Briefly, 100 mL cultures were rapidly cooled on ice and transferred to pre-chilled centrifuge tubes (-20°C). Cells were harvested by centrifugation for 5 min at 6000 X g and resuspended in 1/10 volume of 50 mM Tris-HCl (pH 6.8)-2 mM EDTA-1.0% SDS. The suspension was transferred to a clean tube and boiled for 2 min in a water bath. The lysate was cooled on ice 37 for 5 min and 1/2 vol of ice cold NaCl was added and mixed in using a Vortex mixer. The sample was kept on ice for 5 min, then centrifuged for 10 min at 10,000 X g. The clear supernatant fluid was decanted into a 30 mL Corex glass tube. The nucleic acids were precipitated by addition of 2.5 volumes of 95% ethanol, stored at -20°C for 12 to 16 hrs and recovered by centrifugation for 20 min at 10,000 X g. The pellet was then washed with 70% ethanol at -20°C and redissolved in 2.0 mL of 10 mM Tris-HCl (pH 7.5)-40 mM NaCI-5 mM MgCl2 . DNA was removed by treatment with 5 U of RQ1 DNasel (Promega) for 15 min at 37°C. EDTA was added to 5 mM and the mixture was extracted twice with phenol-chloroform (1:1) and once with chloroform. RNA was recovered by precipitation with 2.5 volumes of 95% ethanol followed by centrifugation for 10 min at 10,000 X g. The RNA pellet was washed with 70% ethanol and finally dissolved in RNA storage buffer (20 mM NaPC>4 (pH 6.5)-1 mM EDTA). Samples were divided into aliquots and stored at -20°C. The RNA preparations were analysed by agarose gel electrophoresis after staining with ethidium bromide. The size markers used were ssDNA HaeUl digested M13 mp11 where the 525-base fragment arised from partial digestion (Greenberg etal., 1987a,b) and a 0.24-9.5 Kb RNA ladder purchased from BRL. RNA concentrations were determined by A26O n m using a value of 37 ug per O.D^gn-2.11.3. Northern blot analysis of mRNA For Northern blot analysis of specific transcripts, up to 30 ug of bacterial RNA were precipitated with ethanol, redissolved in 10 uL of 20 mM MOPS-1 mM EDTA-5 mM sodium acetate (running buffer [pH 7.0]) with 50% formamide and 2.2 M formaldehyde, heated for 5 min at 68°C, and cooled briefly on ice. Loading dye was added (to give 3%[w/v] Ficol (Pharmacia) and 0.02% (w/v) bromophenol blue and xylene cyanol). The samples were loaded onto a 1.0% agarose-6.6% formaldehyde gel alongside 3 2 p . | a D e | | e q molecular weight markers. Electrophoresis was at 40 mA with recirculation of running buffer. RNA was blotted onto 3 8 BioTrans membranes (Pall, Inc.) in 20X SSC (3 M NaCI-3 M sodium citrate) by electrotransfer for 12 to 16 hrs (Southern, 1975). The membrane was allowed to dry then baked at 80°C for 2 hrs. Prehybridization and hybridization were performed according to the membrane supplier. The 3 2 P - l a b e l l e d probes (cex specific: 5' GTGGCCGGGTGCGGGCGTGGTCCTAGGCAT 3'; and cenA specific: 5' CAGCGCTGCGGCGGTTCTGCGGGTGGACAT 3') were synthesized chemically by Tom Atkinson essentially as described (Atkinson and Smith, 1984) and labelled by N. Greenberg at U.B.C. The membranes were first washed at room temperature with 6X SSC buffer followed by gradual increase of temperature and salt stringency. Filters were wrapped in S a r a n - W r a p ™ and exposed to X-ray film (Kodak) with intensifying screens at -70°C for 16 hrs to 3 days. 2.11.4. Preparation of 3 2 P labelled DNA hybridization probes Plasmid DNA was digested conveniently by restriction enzymes under the appropriate conditions suggested by the manufacturer. The mixture was extracted with phenol-chloroform (1:1) and the DNA was precipitated with 95% ethanol. The 5' phosphate was removed from the DNA ends by treatment with calf intestinal alkaline phosphatase (CIAP) and the ends were labelled with [ g a m m a - 3 2 P ] - A T P (3,000-7000 Ci/mmol) and T4 polynucleotide kinase (PNK) as described previously (Maniatis etal., 1982). Unincorporated label was removed using a Sephadex G-50 column (Maniatis etal., 1982) and the reaction was monitored by liquid scintillation spectrometry in an ISOCAP-300 (Nuclear Chicago). If required, restriction digests were performed on the labelled DNA to liberate fragments uniquely labeled at one end. The probes were purified by electrophoresis in 8% polyacrylamide gels from which they were electroeluted and further precipitated by addition of 2 vol of 95% ethanol and 0.1 vol of 3M sodium acetate. Pellets were rinsed with 70% ethanol, dried in a spin vac and redissolved in water. The efficiency of the kinasing reaction was monitored by liquid scintillation counting. 39 2.11.5. Hybrid protection analysis The 5' ends of mRNA were identified with labeled DNA probes essentially as described previously (Favoloro etal., 1980). Up to 30 ug of RNA and end-labelled DNA probe (0.01 to 0.03 pmol) were precipitated together by ethanol, redissolved in 30 uL of hybridization buffer (0.4 M NaCI-0.04 M sodium phosphate [pH 6.5]-0.4 mM EDTA-80% formamide), heated for 15 min at 85°C and quickly transferred to 60°C water bath for 3 hrs. Samples were diluted with 300 uL ice-cold S1 buffer (30 mM sodium acetate [pH 4.5]-28 mM NaCI-4.5 mM ZnSC»4) containing about 1000 U of S1 nuclease, and incubated for 30 min at 37°C. The reactions were terminated by addition of 75 uL stop buffer (2.5 mM ammonium acetate-50 mM EDTA) and 20 ug yeast tRNA as carrier. Nucleic acids were precipitated with 400 uL isopropanol and recovered by centrifugation. Pellets were dissolved in sequencing buffer (90% formamide, 0.02% [w/v] bromophenol blue and xylene cyanol) (Maniatis et al., 1982) and heated at 90°C for 2 min. Nucleic acid fragments were resolved by PAGE alongside a DNA sequencing ladder and the dried gels were exposed to X-ray film (Eastman Kodak) at -70°C with intensifying screens. 2.11.6. DNA sequencing DNA was sequenced by modifications of the standard chemical (Maxam and Gilbert, 1980) and enzymatic chain termination (Sanger ef al., 1977) methods. For chemical sequencing, DNA fragments were 5'end-labelled using PNK, digested with restriction enzymes and purified by PAGE as described above. The chemical modification and cleavage reactions were performed as described (Maniatis et al., 1982). For enzymatic sequencing, several methods were used. For sequencing of double stranded template, DNA was isolated according to Kraft et al., (1988) and the reactions were performed using the Pharmacia Sequencing Kit, T7 DNA polymerase (Pharmacia) and [a lpha- 3 5 S]dATP 40 Table IV. Sequencing primers M13/pUC 17mer-reverse S'-dCAGGAAACAGCTATGAC-S' M13/pUC 17mer-universal 5'-dGTAAAACGACGGCCAGT-3' Micro 1 17mer-94nt 5'-P0 4-dACAGGCACCGACCAGGC-3' Micro 2 17mer-380nr 5 ,-P04-dGCCCAGCGGGTCGCCGG-3' 4 1 (500 Ci/mmol) with various primers (Table IV). DNA compressions were resolved by making use of deoxy-7-deaza-guanosine triphosphate (Mizusawa et ai, 1986) in place of dGTP and by using Taq polymerase (Pharmacia). 2.11.7. Oligonucleotide directed mutagenesis The creation of a single base pair frameshift at position 34 of the cbgA putative GTG start site was obtained by primer elongation (Atkinson and Smith, 1984). Cells carrying pTZ19R-2 were grown and single stranded DNA was isolated as described previously (Kunkel et al., 1987). In the hybridization mixture, 20 pmoles of single stranded pTZ19R-2 DNA and 200 pmoles of primer FP21 (5' PO4 -GGAGTGA CCGCTGCTGCGTGC-OH-3') were mixed in 200 uL of HincW buffer (BRL, insertion of a £ destroys the CbgA putative GTG start site). The mixture was heated to 90°C for 5 min and slowly cooled to 25°C. The hybridization mix was then kept at 4°C for 12-16 hrs. Primer extension was initiated by the addition of 15 U of Klenow fragment of E. coli DNA polymerase I (BRL) plus 500 uM-dNTPs, 1 mM-DTT, 1 mM ATP and 15 U of T4 DNA ligase (NEB). After 20 min incubation at 37°C, an additional 15 U of Klenow enzyme were added and incubation continued for another 20 min at 37°C. The sample was extracted once with phenol-chloroform (1:1) and the DNA was recovered by ethanol precipitation followed by resuspension in 50 uL TE buffer. An aliquot of 5 uL was used to transform E. coli JM101 and some single stranded DNA was isolated from transformants as described (see above). DNA sequencing was performed to identify clones with an extra C at position 34 giving rise to pTZ19R-2FP21. By subcloning the wild-type Stu\-EcoR\ fragment in pTZ19R-2FP21, pTZ19R-7 was finally isolated and screened for the extra C base by sequencing and further assayed for B-glucosidase activity and Western blot antigenic peptides. 42 3. RESULTS AND DISCUSSIONS P A R T 1 . The expression of Cellulomonas fimi cellulase genes in Brevibacterium lactofermentum. 3.1. Construction of the shuttle vector and subcloning of the cellulase genes The cryptic plasmid pBL1 was iolated from B. lactofermentum; it was joined to pBR322 by their unique HindiU sites to give pRB3. The 1.3 Kb kanamycin resistance cartridge of pUC4K (Oka etal., 1981), was inserted into the unique BamHl site of pRB3 to give pBK10 (Fig. 6). The MIC values for kanamycin sulfate in wild type bacterial strains were 0.8 u.g/mL for B. lactofermentum, 12.5 u.g/mL for E. coli and 350 jig/mL for C. fimi. The values for the strains transformed with pBK10 were 800 uxj/mL for B. lactofermentum and over 1 mg/mL for E. coli. Three derivatives of pBK10 were constructed (Fig. 6). pFWPIO carries the entire coding sequence of cex together with its putative regulatory region (Greenberg era/., 1987a). pFWP30 carries the entire coding sequence of cenA together with its putative regulatory region (Greenberg etal., 1987a). pFWP1030 carries both of these genes, separated by more than 1 Kb of C. fimi DNA. From previous studies (Stuber and Bujard, 1981), there does not appear to be a putative promoter in the pBR322 sequence upstream of these genes that could be held responsible for the cellulase genes expression. 3.2. Transformation Transformation efficiency by heterologous plasmid DNA was lower in B. lactofermentum (<1 transformant per iig of DNA) than previously reported (Santamaria et al., 1984). This may have been due to the presence of DNA-degrading enzymes which were detected by DNase test agar (Difco) (not shown). The transformation efficiency may also be ( reduced by restriction because the efficiency of transformation was increased 100-fold when the plasmid DNA was isolated from B. lactofermentum rather than from E. coli (data not given; see also Santamaria ef al., 1985). pBK10 and its derivatives were stable in both 43 Fig. 6. Construction of the shuttle vector and its derivatives. Restriction enzymes are: B, fiamHI; E, EcoRI and H, HindlU. Partial digest are designated by prefix p (e.g., pH/ndlll, etc.). Cryptic plasmid pBL1 was isolated from B. lactofermentum (ATCC21798). Plasmid pBL1 and pBR322 were digested at their unique HindlU sites and joined together to form pRB3. pRB3 was digested with SamHi, and the K m R cartridge of pUC4K was inserted into it to form the shuttle vector pBK10. This was digested with EcoRI and partially digested with HindlU to allow forced cloning of two individual EcoRI-H/ndlll fragments. The fragment coding for Cex gave rise to pFWPIO, that coding for CenA gave rise to pFWP30. Finally, pFWPIO was digested with EcoRI and partially digested with BamHI to allow forced cloning of an EcoRI-SamHI fragment encoding CenA to give pFWP1030. 45 E . coli and B. lactofermentum. The cryptic plasmid originally present in the wild-type strain of B. lactofermentum was incompatible with pBK10 and its derivatives under selective pressure with Km (not shown). 3.3. Gene expression and detection of cellulolytic activities The cellulase genes were expressed in both E. coli and B. lactofermentum transformants carrying pFWPIO, pFWP30 or pFWP1030 (Fig. 7). Only weak Exg activity was detected in E . coli cells carrying pFWPIO or pFWP1030, even after incubation for five days on L agar-MUC plates at 30°C (not shown). On the other hand, strong Exg activity was detected in B. lactofermentum carrying pFWPIO (Fig. 7). EngA activity was detected in E . coli cells carrying pFWP30 or pFWP1030 (Fig. 7). EngA activity was also detected in B. lactofermentum carrying pFWP30 or pFWP1030 (Fig. 7). Surprisingly, Exg activity was not detected by plate assay in B. lactofermentum carrying pFWP1030, even though the plasmid could be recovered intact from the cells. In broth, pFWP30 or pFWP1030 expressed five to seven times more EngA in B. lactofermentum (691 units/20 mL and 509 units/20 mL, respectively) than in E . coli (97 units/20 mL and 108 units/20 mL, respectively) and pFWP1030 expressed over seven times more Exg in B. lactofermentum (5.8 units/20 mL) than in E. coli (0.8 units/20 mL) (Table V). There was a striking difference in the location of the enzyme in the two organisms: more than 95% of the activity in B. lactofermentum was in the culture supernatant, compared with only 30 to 50% for E . coli. However, the level of expression in E . coli was very low. Similar results were obtained for cultures grown in minimal medium (Table V). Cellulase activities detected in E . coli supernatants probably results from cell lysis since most of the activity is usually found in the periplasm of exponentially growing cells. In E . coli carrying pFWP1030, the amount of Exg produced was almost identical to that produced by E . coli carrying pFWPIO (Table V). Again, Exg activity in B. lactofermentum carrying pFWP1030 was about ten-fold less than that in B. lactofermentum carrying pFWPIO (Table V). This difference was not seen with the corresponding E . coli strains. The reason for this is unknown. 46 Fig. 7. Screening of cellulase activities in transformants. Transformants were streaked on L agar containing 50 jig Km/mL; 500 (iM MUC; and 1% CMC. The plate was photographed under long-wave UV light to detect Exg activity (plate 1), then stained with Congo red (Teather and Wood, 1982) before photographing under white light to detect EngA activity (plate 2). E. coli was transformed with (A) pBK10, (B) pFWPIO, (C) pFWP30, (D) pFWP1030; B. lactofermentum was transformed with (E) pBK10, (F) pFWPIO, (G)pFWP30, (H)pFWP1030. 47 Table. V. Cellulolytic activities of bacterial strains. Each strain was inoculated from a single colony into 20 mL of medium containing Km sulfate at 50 ug/mL. The cultures were incubated at their optimal growth temperature, B. lactofermentum at 30°C and E. coli at 37°C. The cells were collected by centrifugation at the end of the exponential phase of growth and ruptured by sonication. Cell extracts (ex.) and supernatants (s/n) were used to measure cellulase activities at 37°C. In the Exg assay, one unit of enzyme released 1 nmol of pNP per min per mL of culture. In the EngA assay, one unit of enzyme released 1 nmol of glucose equivalents per min per mL of culture. Specific activity is units per mg of protein. Symbol "<" is less then 0.1 unit/mg of protein for Exg and less then 1.0 unit/mg of protein for EngA. The total units (Tot. U.) are given for 20 mL of culture. 48 Brevibacterium lactofermentum Escherichia coli pNPC ONS pNPC DNS SD.ACt. Tot.U. SD.Act. Tot.U. SD.Act. Tot.U. SD.Act. Tot.U. rich media plasmids ex. < < < < pBKIO 0 0 0 0 s/n < < < < C.x. 0.14 2.7 0.10 2.1 pFWPIO 5.8 71 0.8 14 S/n 3.24 75 0.61 87 ex. < 20 < 20 PFV/P30 0 691 0 97 s/n < 829 < 33 ex. < 12 0.12 9.4 PFWP1030 0.5 509 0.8 108 s/n 0.46 1118 0.43 65 minimal  media plasmids c.x. < < < < pBKIO 0 0 0 0 s/n < < < < ex. 0.34 2.7 0.12 3.4 PFWPIO 3.5 109 0.3 12 S/n 1.7 589 0.48 115 ex. < 20 < 16 PFWP30 0 1408 0 169 s/n < 1064 < 154 ex. < 45 < 23 PFWP1030 0.3 868 0.4 131 S/n 0.2 751 0.54 51 49 Restriction analysis of plasmid DNA indicated that the structural gene and the regulatory region were both still present. However, it is possible that some promoter located in the pBR322 plasmid is used by B. lactofermentum and is responsible for the expression to the cellulase genes. Since the bla gene is not expressed in B. lactofermentum(San-tamaria era/. , 1984). it is possible that the P4 promoter identified by Stuber and Bujard (1981) (located between the origin of replication of pBR322 and the bla gene and oriented toward the bla gene) could be responsible for the cellulase gene expression. 3.4. Detection of cellulase gene specific transcripts The positive control RNA species isolated from C. fimi and detected by the cenA probe was approximately 1400 bases in length (Fig. 8, lane A) as previously observed (Greenberg et al., 1987a). The RNA species isolated from B. lactofermentum carrying pFWP30 plasmid and detected by the cenA probe are approximately 1700, 1400 and 300 bases in length (Fig. 8, lane B and C). The positive control RNA species isolated from C. fimi and detected by the cex probe was approximately 1500 bases in length (Fig. 8, lane F) as previously observed (Greenberg et al., 1987a). The RNA species isolated from B. lactofermentum carrying pFWPIO plasmid and detected by the cex probe are approximately 2300 and 1300 bases in length (Fig. 8, lane D and E). The stringencies of probe hybridizations were based on the specificity of the positive controls. Nevertheless, a negative control like mRNA from B. lactofermentum carrying the shuttle vector pBK10 would have been appropriate to justify the specificity of the detected hybrids in that bacterium. Assuming that the RNA polymerase putative stop sites for the cellulase genes are recognized by B. lactofermentum transcriptional apparatus and considering the additional 5' end of non-coding C. fimi DNA present in pFWPIO and pFWP30 plasmids, the largest cenA and cex transcripts (1700 and 2300 bases in length, respectively) would originate from a common region in pBR322 just upstream from the EcoRI site. In the cex lanes, none of the transcript (lane D and E) detected from B. lactofermentum carrying pFWPIO plasmid was similar in length with the cex positive control C. fimi mRNA (lane F). This suggests that 5 0 Fig. 8. Detection of specific transcripts by Northern blot analysis. RNA was isolated from C. fimi culture grown in basal medium supplemented with 1% CMC and B. lactofermentum carrying pFWPIO or pFWP30 plasmids grown in rich medium containing 50 u.g Km/mL Total RNA was denatured with formaldehyde, fractionated on a formaldehyde gel containing 1% (w/v) agarose, and blotted onto a Biotrans membrane (Pall, Inc.). Hybridizations were done with 32P-labelled 30-mers oligos complementary to cenA mRNA (lane A, B, C) or cex mRNA (Lane D, E, F). The filters were exposed to X-Ray films for 18 hrs in the case of cenA and 3 days in the case of cex. Lane A and F, 20 u,g of C. fimi RNA; B and C, 20 and 30 u.g of B. /acfofera7enft/m/pFWP30 RNA, respectively; lane D and E, 30 and 20 |j.g of B. lactofermentum/pF\NP'\0 RNA, respectively. Size markers (in nucleotides) are rVaelll digested single-stranded M13mp11 32p_|aDe||ec| DNA fragments. The expected size of cenA mRNA (1438 nt) and cex mRNA (1564 nt) are indicated. 51 initiation of transcription of the cex gene originated upstream from the C. fimi regulatory sequences. In contrast, in the cenA lanes, one of the transcripts detected from B. lactofermentum carrying pFWP30 plasmid is similar in length (about 1400 bases) with the cenA positive control C. fimi mRNA (Lane A) which suggests that initiation of transcription of the cenA gene may occur at the expected site. Nevertheless, Northern blot analysis does not identify transcriptional initiation sites but the approximate size of a transcript which is prone to processing or nuclease degradation. Since only cenA gave a comigrating mRNA band, the initiation of transcription of the cenA mRNA in B. lactofermentum carrying plasmid pFWP30 was further investigated by fine resolution S1 nuclease mapping. 3.5. Mapping the cenA mRNA 5' ends Transcripts produced in vivo by C. fimi cells grown on CMC containing basal medium and by B. lactofermentum carrying pFWP30 plasmid were analysed by high resolution S1 nuclease mapping essentially as described (Greenberg etal., 1987a). A 315 bp Sma\-Sa/I DNA restriction fragment isolated from the plasmid pNG101 (Greenberg et al., 1987a) was used as the hybridisation probe. The probe uniquely labelled at the 5' Sail site, was made single stranded by denaturation and hybridised with total RNA isolated from C. fimi and B. lactofermentum carrying pFWP30 plasmid. The mixtures favouring RNA-DNA hybrids were degraded with S1 nuclease. The length of the protected probe species were determined by gel electrophoresis under denaturing conditions. A sequencing ladder of the Sma\-Sal\ probe created by chemical cleavage served as control to measure the length of the residual probe and identify the 5' end of the CenA mRNA. The autoradiograph (Fig. 9) revealed a major mRNA 5' end in the C. fimi control lane (Lane 1) with a protected fragment of 162 bases in length. When these sequencing data and the weaker exposure of the ladder were compared with the previously published data, it was estimated that the mapped 5' end corresponded to a C base located 52 bases upstream of the ATG as previously observed as a -1 start site (Greenberg etal., 1987a). The conditions of hybridization must have 52 Fig. 9. S1 nuclease protection analysis of cenA transcripts. After hybridisation with RNA and treatment with S1 nuclease, the remaining Sma\-Sal\ cenA-specific labeled DNA probe was analysed in a 8% polyacrylamide-7 M urea sequencing gel alongside probe sequenced by the base-specific chemical cleavage method (Maxam and Gilbert, 1980). The chemical sequencing ladders of the probe is shown. Protection of the probe by: lane 1, RNA from C. fimi ; lane 2, RNA from B. lactofermentumlpF\NPZO; lane 3, by yeast tRNA. Lane 4 represent the probe alone. Sizes are in nucleotides. G G+AT+C C 1 2 3 4 53 been optimal since a unique band was detected (Lane 1) where the usual self DNA-DNA probe hybrid (Lane 4) was not. Interestingly, the full length of the probe was protected by B. lactofermentum mRNA (lane 2) carrying pFWP30 compared with the untreated probe alone (Lane 4) indicating that the 5' end of the cenA mRNA in that strain is located upstream of the Sma\ site. Furthermore, this observation suggests that the cenA putative regulatory sequence is not used or recognized by the B. lactofermentum transcriptional apparatus and that an unreported promoter located somewhere upstream is used by B. lactofermentum for expression of the cloned genes. This is a likely explanation of the lack of exoglucanase activity in B. lactofermentum carrying the pFWP1030 construct. The labelled probe was not protected in mapping studies with control yeast tRNA (lane 3). With RNA isolated from B. lactofermentum as well as E. coli cells carrying recombinant plasmids with putative C. fimi promoter sequences, transcripts were never found which initiated within the inserts (Greenberg ef al., 1987b; this thesis). Several reasons could be suggested for these observations: (1) E. coli and B. lactofermentum RNA polymerases may not have been able to recognize the C. fimi promoters due to the absence of an appropriate sigma factor, (2) the RNA polymerases may have recognized the promoters but were incapable of initiating or elongating the transcripts, or (3) the resulting "hybrid" transcripts may have been intrinsically unstable. It is proposed that the expression of the cex and cenA genes in B. lactofermentum originated from a similar promoter sequence located within the vector itself and that conclusive evidence for that promoter would require more stringent studies beyond the scope of this study. In conclusion, even though the cellulase gene putative promoter sequences did not function as expected in the closely related bacterium B. lactofermentum, the cellulase genes were expressed and their products were secreted into the culture supernatant. The B. lactofermentum system could allow studies of the implicated sequences and factors required for secretion of the cellulase enzymes in this bacterium. 54 PART 2. The characterization of recombinant C. fimi 6-glucosidase A from E. coli. 3.6. The expression of recombinant CbgA in E. coli 3.6.1. Cloning, subcloning and specific activities The original clone cbgA was first isolated from a family of immunopositive E. coli transformants generated from a pBR322-C. fimi genomic library (Gilkes etal., 1984). The recombinant plasmid pEC62 carries a 7.2 Kb-SamHI insert (Fig. 10) and expresses in E. coli an enzyme with 6-glucosidase activity (CbgA). Expression of the cloned gene, cbgA, in E. coli requires a heterologous promoter (Bates, 1987; this thesis). To test for the presence of a CbgA translational initiation codon within the 7.2 Kb insert, a translational frameshift was created at the 5'-end SamHI site of pEC62 (Fig. 11). A loss of the enzymatic activity would have resulted if the recombinant 6-glucosidase originated from a translational fusion between CbgA and the T c R determinant of pBR322, whereas, the detection of 6-glucosidase activity would have confirmed the presence of a CbgA translational start site within the insert, and perhaps the presence of the entire CbgA structural gene as well. As an initial step, the problem of having an additional SamHI site at the 3'-end of the insert was circumvented by deleting the 3' end-1.5 Kb BglU-BamH\ fragment followed by religation of the compatible ends giving rise to pEC62.1. No loss in specific activities was observed from either the pEC62.1 or pEC62.2 genetic constructs when expression was compared to that of the original plasmid pEC62 (Fig. 10). These data confirmed the presence of a CbgA translational start site within the 7.2 Kb SamHI insert and suggests that 5.6 Kb of insert DNA is sufficient to produce an active peptide. To increase the expression of cbgA in E. coli, the 7.2Kb SamHI insert was transferred from pEC62 to the high copy number plasmid pUC13 giving rise to pUC13:62 (Bates, 1987). No increase of the specific activity was observed from this construct (Fig. 10). Deletions were generated at the 5' end of the 7.2Kb insert with exonuclease Sa/31 and fragments of various sizes were isolated and subcloned into pUC13. E. coli transformants were screened by the MUGase plate assay for 55 Fig. 10. Diagram of pEC62 and derivatives. The circular plasmids are shown in a linear fashion. The open bars represent pBR322 DNA; the dashed bars represent C. fimi DNA; the solid bars represent pUC13 DNA; The single lines represent deleted DNA regions. The arrow indicates the functional orientation for the T c R or Lac promoters as well as the orientation of the cbgA gene in the insert. The scale of the drawing is given at the top. The length of each plasmid is indicated. The deletion in pUC13:62A31 extends to the 5' proximal Xhol site. The specific B-glucosidase activity for each plasmid expressed in E. coli is given in units (1 U releasing 1 nmol of pNP per min at 37°C) per mg of protein of total cell extracts. Restriction sites : B, BamHl; Bg, BglW; K, Kpn\; X, Xhol 1 Kb B X K XX Bg/XK B Size plasmid Sp.Act. (Kb) (U/mg) 11.6 pEC62 2.0 10.1 pEC62.1 1.9 10.1 pEC62.2 2.0 9.90 pUC13:62 2.3 9.50 pUC13:62A27 3.0 9.45 pUC13:62Al7 14 9.36 pUC13:62A28 <0.1 9.22 pUC13:62A24 <0.1 9.20 pUC13:62A31 120 8.76 pUC13:62A21-4 27 8.13 pUC13:62A21-15 26 56 Fig. 11 . Scheme for the generation of a translational frameshift in pEC62. Plasmids are shown in a linear fashion. The original recombinant plasmid pEC62 carrying a 7.2 Kb SamHI fragment of C. fimi DNA (dash boxes) in pBR322 (open boxes) was digested to completion with Bglll and partially digested with SamHI. Ligation of compatible cohesive ends of the remaining 10.1 Kb fragment eliminated the 3' proximal SamHI site creating pEC62.1 containing a unique 5' proximal SamHI site. Finally, pEC62.1 was digested to completion with SamHI and the ends were filled with dNTP's by Klenow enzyme followed by religation of the newly generated blunt ends creating pEC62.2. This construct contains the translational frameshift at the SamHI T c R promoter proximal region. Relevant restriction sites are indicated. The location of the cbgA gene and orientation of transcription is given. Putative and newly generated DNA sequence is shown in lower diagram of pEC62.2. BamHI BamHI Bglll BamHI pEC62., deletion of 1.5 Kb 3* Bglll-BamHI fragment -Bglll/BamHI partial -ligation Bglll/BamHI BamHI pEC62.1. Bglll/BamHI M l creation of translational frameshift -BamHI -Klenow & dNTP's -ligation Bglll/BamHI pEC62.2 GGATCGATCC CCTAGCTAGG — c b g A | Bglll/BamHI 57 increase in 6-glucosidase activity (Bates, 1987). The size of the deletions was estimated by restriction digest analysis. The plasmid giving the highest expression in the E. coli (pUC13:62A31) has a deletion of about 700 bp from the 5' end of the insert and gives a 60 fold increase in specific activity compared with the original clone pEC62 (Fig. 10). The shortest deletion (pUC13:62A28) which results in the loss of 13-glucosidase activity is about 540 bp. This suggests that the CbgA translational start site is located within this 540 bp region. It is appropriate to mention at this point that the A24, A27 and A28 deletions are out of frame fusions with the 8-Gal alpha peptide, whereas, the other deletions are in frame fusions (see section 3.10) and they will be discussed later on. The expression of 6-glucosidase activity detectable in E. coli cells carrying pUC13:62A21-15 or pEC62.1 plasmids suggests that less than 3.8 Kb of DNA is sufficient to produce an active peptide. The high expressing clone pUC13:62A31 was used to determine the location of recombinant enzyme in E. coli. 3.6.2. The location of recombinant CbgA in E. coli The location of CbgA in E. coli cells carrying pUC13:62A31 was determined by fractionation of the cellular compartments into cytoplasm, periplasm and total membrane fractions. Cells grown to midlog and stationary phases were fractionated and each fraction was tested for the presence of enzymes and sugar markers that are specific for and representative of E. coli compartments (Table VI). In either logarithmic or stationary phase cells, almost all of the 6-glucosidase activity was cell associated, mostly found in the cytoplasmic fraction and partially in the membrane fractions of E. coli cells (Table VI). Little 8-glucosidase activity was detectable in the periplasmic fractions of E. coli cel ls carrying either pUC13:62 (not shown) or its A31 derivative (Table V). No activity was detectable in culture supernatants of either cultures (data not shown). This is not surprising since E. coli rarely secretes proteins into the culture medium. Since 6-glucosidases are sometimes membrane associated (Hwang and Suzuki, 1976; Umile and Kubicek, 1986), the total membrane fraction of the stationary phase culture was analysed further by isopycnic 58 Table VI. Localisation of B-glucosidase activity in E. co//JM83/pUC13:62A31 Fractions Protein B-glucosidase G-6-PDH B-lactamase KDO NADHox (mg) (total mU) (total mU) (total mU) (|ig) (total mU) (LOG PHASE) SET wash 0.3 9 n.d. 12 n.d. <1 cytoplasm 2.5 415 513 2 <1 1220 membranes 1.3 30 <1 <1 63 1459 periplasm 0.3 7 <1 28 <1 <1 Total cells 5.0 496 586 37 100 5200 (STATIONARY PHASE) SET wash 0.9 <1 n.d. , 15 n.d. <1 cytoplasm 22.4 3219 698 147 <1 255 membranes 10.0 1374 14 16 87 1894 periplasm 0.9 1 <1 53 18 <1 Total cells 61.6 4468 635 235 224 2080 Protein fractions in this table were obtained as described in section 2.5.1. Enzymatic assays, sugar content and protein determination in this table are as described in section 2.4. SET : 40% sucrose- 33 mM Tris- 0.1 mM EDTA (pH 7.0) < : less then One unit of enzyme in this table releases one fimol of product per min under specified conditions. 59 Fig. 12. Profile of E. coli JM83/pUC13:62A31 membrane fraction separated on a sucrose density gradient. Approximately 950 u.g of proteins from the membrane fraction of stationary phase cells was loaded onto a 30% to 55% sucrose step gradient containing 5 mM-EDTA and 10 mM-HEPES buffer (pH 7.4). Fractions of 350 uL were collected dropwise, absorbance at 280 nm, 6-glucosidase and NADH oxidase activities were measured. The bottom and top of the gradient are shown. Sample 0 corresponds to the pellet at the bottom of the tube. 60 sucrose density gradient centrifugation (Fig. 12). The location of 6-glucosidase activity was compared with that of NADH oxidase which is membrane associated in E. coli (Osborn etal., 1972). The small amount of sediment at the bottom of the tube (sample 0) was kept for analysis. One has to consider the fact that the use of M g 2 + during the osmotic shock treatment or at any point in the isolation procedure results in extremely poor separation of inner and outer membranes (Osborn and Munson, 1974). Also, membranes from cells grown on very rich media appear to be more difficult to separate because of a possible increase in the number of zones of adhesion between the outer and cytoplasmic membranes in rapidly growing cells (Gerhardt et al., 1981). Therefore, virtually all the membrane material (Fig. 12) was recovered at the intermediate position "M" (Osborn and Munson, 1974). About 94% of the recovered NADH oxidase activity was found at buoyant density of about 1.19 g/cc indicating the location of the bacterial inner membranes. About 67% of the recovered B-glucosidase activity was located on the top of the gradient. These data suggest that the recombinant enzyme CbgA expressed from pUC13:62A31 plasmid is not bound to the bacterial inner membrane like the E. coli NADH oxidase enzyme but rather seems to co-sediment with the membranes during ultracentrifugation. This suggests that CbgA may form dense particles perhaps by multimeric association with itself or other proteins present in the sample. 3.7. Characterization of recombinant CbgA from E. co//7pUC13:62A31 3.7.1. Purification of CbgA from E. coli. The high expressing clone pUC13:62A31 was used to produce the recombinant B-glucosidase in E. coli JM83 for purification. The elution profile of E. coli cytoplasmic fraction from DEAE-sephadex A-50-120 column (Fig. 13) shows two major peaks of B-glucosidase activity eluting at 0.35 M NaCl and 0.55 M NaCl respectively. The fractions under the first peak of activity which contained about 50% of the total B-glucosidase activity (fraction 40 to 60) were pooled (Pool I) and processed further. The fractions under the second peak of activity which contained about 25% of the total B-glucosidase activity (fraction 76 to 84) were also pooled (Pool II) and proteins were concentrated by ultrafiltration. 61 Fig. 13. Fractionation of E. co//JM83/pUC13:62A31 cytoplasmic proteins on a DEAE-sephadex A-50-120 column. The cytoplasmic fraction of E. coli carrying pUC13:62A31 was applied to a DEAE-sephadex A-50-120 (2.5 cm x 11 cm) column. After washing the column at a flow rate of 44 ml_/hr with 50 mM Tris-5 mM EDTA (pH 7.0), the bound proteins were eluted by stepwise increases in the NaCl concentration as shown. Fractions of 2.5 mL were collected and the absorbance at 280 nm, conductivity and pNPGase activity were determined. Fractions 40 to 60 (Pool I) and 76 to 84 (Pool II) were combined separately. 0 10 20 30 40 50 60 70 80 90 100110 Fraction number 62 The proteins in Pool I were separated further by ion exchange on Mono Q column (Fig. 14 and 15). Several fractions obtained during purification of the recombinant enzyme were analysed by SDS-PAGE. The purification scheme gave a fraction (Fig. 14, fraction 24) containing a major polypeptide, as visualized by Coomassie blue staining of SDS-PAGE (Fig. 16-lane 7). The purified recombinant enzyme CbgA from Mono Q resin has a molecular weight of approximately 183 000 (p183) and a specific activity of 12957 mU/mg of protein with pNPG as substrate (Table Vll). The recombinant enzyme was purified about 100-fold to homogeneity, as visualized from SDS-PAGE. A yield of 2 . 1 % was obtained from the purification scheme and was estimated from the total recovery after Mono Q column (Fig. 14). The presence of more than one peak of activity eluting from DEAE-sephadex A-50-120 is not because of the presence on the 7.2 Kb insert of two separate B-glucosidase encoding genes. Western blotting analysis of fully denatured proteins from Pool II has indicated the presence of a major polypeptide of 183 kDa as well as much larger components that could originate from aggregation of CbgA with itself or with other peptides (data not shown). The storage at 4°C of Pool I from DEAE-Sephadex A-50-120 gave rise to additional products (Fig. 16-Lane 10 and 11) which coeluted with the p183 peptide during a second purification on Mono Q column (Fig. 15) . Based on their apparent molecular weights and associativity, these additional peptides (p137 and p60) could originate from proteolysis of the p183 peptide (see section 3.10). Furthermore, the specific activities of fractions containing p183-p137 or p183-p137-p60 peptides suggests that p60 does not have B-glucosidase activity. Further evidence that the p137 polypeptide is a cleavage product of the p183 polypeptide comes from the observation that p137 reacted with the anti-CbgA antibodies on Western blots (see section 3.9). DNA sequence and N-terminal amino acid sequence analysis of p137 brings about conclusive evidence of this relatedness. However, I have no evidence to support the origin of the p60 polypeptide (see section 3.10). Some cleavage products are also observed in C. fimi, and in both cases, they seem to interact in an aggregating fashion. The function of the cleavage products and the nature of the aggregation of CbgA subunits remains unknown. 63 Fig. 14. Fractionation of Pool I on Mono Q. One fifth of the Pool I fraction from the DEAE-sephadex column was applied to a Mono Q anion-exchange column. After washing the column at a flow rate of 1 mL/min with 50 mM Tris-5 mM EDTA (pH 7.0), bound proteins were eluted by stepwise increases in the NaCl concentration as shown. Eluting fractions were monitored by absorbance at 280 nm and analysed for pNPGase activity. 0 10 20 30 40 50 Fraction number 64 Fig. 15. Fractionation of stored Pool I on Mono Q. One fifth of the Pool I fraction from the DEAE-sephadex column that was stored at 4°C for several months was applied to a Mono Q anion-exchange column. After washing the column at a flow rate of 1 mL/min with 50 mM Tris-5 mM EDTA (pH 7.0), bound proteins were eluted by stepwise increases in the NaCl concentration as shown. Eluting fractions were monitored by absorbance at 280 nm and analysed for pNPGase activity. Fractions 22 to 27 and 28 to 31 were pooled separately. 65 Fig. 16. 7% SDS-PAGE analysis of protein fractions from the ion-exchange purification of recombinant CbgA. Lane 1, 26 ug (10 U) of French press total crude extract; lane 2, 35 ug (10 U) of the cytosol fraction; lane 3, 58 ug (10 U) of the membrane fraction; lane 4, 28 ug (10 U) of the streptomycin sulfate fraction; lane 5, 15 ug (10 U) of the ammonium sulfate fraction; lane 6, 3 ug (10 U) of Pool I fraction; lane 7, 1.4 ug (7U) of fraction 24 from first Mono Q column; lane 8, 1.4 ug (6U) of fraction 25 from first Mono Q column; lane 9, 1.4 ug (5U) of fraction 26 from first Mono Q column; lane 10, 6.9 ug (27 U) of fraction 28 to 31 from second Mono Q column; lane 11, 6.3 ug (56 U) of fraction 22 to 27 from second Mono Q column. Proteins from lane 10 and 11 were isolated from the stored Pool I fraction and ran on a separate gel where Rf's were adjusted accordingly. M r standards (kDa) are shown on the left. Polypeptides relevant to this thesis are identified on the right. 66 Table Vll. Purification of recombinant CbgA from E. coli JM83/ pUC13:62A31 Fraction Tot. protein Tot. units Specific activity Purification Y ie ld (mg) (U/mg of prot.) factor (%) Crude extract 6300 9.5x10 s 151 Cytoplasm 5900 7.1 x 1 0 5 120 1 100 Amm. S04 cut 352 2.3x10 5 653 5.4 42 DEAE-Sephadex 44 4 .9x10 4 1114 9.3 35 Mono Q-fraction 24 0.14 1.8x10 3 12957 108 2.1 The values in this table are from 10 L of cell culture with various fractions obtained during the purification of recombinant CbgA. The units in this table are defined as nmol of pNP released per min at 37°C. 67 3.7.2. The effects of pH and temperature on CbgA activity The 6-glucosidase activity of the crude extract from E. coli carrying pUC13:62A31 and the purified CbgA was tested using 5 mM pNPG as substrate at 37°C at various pH values (Fig.17-A). Both the crude extract and purified enzyme showed maximal activity at about pH 5.5. The crude extract shows higher activity than the purified enzyme below and above pH 5.5 perhaps because of a protein-protein interaction protecting CbgA from acid/alkaly inactivation or activation at the active site. The 6-glucosidase activity of the crude CbgA enzyme extract from E. coli carrying pUC13:62A31 and the purified enzyme was tested after 60 min incubation at various temperatures in the presence of pNPG (5 mM) (Fig.17-B). Both samples showed identical behavior with highest activity at a temperature of about 42°C. Nevertheless, at that temperature, a significant denaturation of the enzyme occurs (Fig. 18). The thermal inactivation of the purified enzyme CbgA was monitored by measuring the remaining activity against pNPG at 37°C. The enzyme was incubated at various temperatures in phosphate buffer (pH 7.0) without substrate. The residual activity presented under the form of an Arrhenius plot shows the effect of temperature on the purified enzyme. 3.7.3. The kinetic parameters of recombinant CbgA The kinetic parameters of the purified recombinant CbgA were determined from Lineweaver-Burke plots for each of the substrates tested (Table Vlll). An example of such plots is given in the Appendix (1-A to D). The recombinant enzyme CbgA can hydrolyze a wide variety of aglycones. It acts preferentially on longer cellodextrins (C5>C4>C3>C2) with K m values comparable to those previously reported for 6-glucosidases from fungi (see Table II). The rate of hydrolysis depends on the nature of the aglycone moiety. The relationship between the V m a x / K m values and the number of glucose molecules per cello-oligosaccharide shows an increase in substrate specificity with the increase in length of the substrate. Thus, in the hydrolysis of cellulose, CbgA could be involved more in the hydrolysis of cellodextrins than of cellobiose, assuming equal concentrations of both substrates. This strongly suggests that CbgA is an exo-8-1,4-glucosidase and must be 68 Fig. 17-A. Effect of pH on the hydrolysis of p N P G by recombinant CbgA 1 0 " 4.0 5.0 6.0 7.0 8.0 p H Fig. 17-B. Effect of temperature on the hydrolysis of p N P G by recombinant CbgA 0 1 0 2 0 3 0 4 0 5 0 Temperature ( ° C ) Fig. 18. Thermostabi l i ty study of purified C b g A 1 . 8 0 r 1 . 4 0 ' ' ' 1 ' 3 . 1 0 3 . 2 0 3 . 3 0 1/T"(absoluteX1000) 6 9 Table Vlll. Kinetic parameters of recombinant CbgA on various 6-glucosides Substrate K m V m a x V m a x / K m (mM) (umol prod/min/mg) (umol prod/min/mg) mM substrate pNPG 0.13 4.4 33.8 Cellobiose 2.50 31.2 13 Cellotriose 0.27 8.8 33 Cellotetraose 0.33 20.0 61 Cellopentaose 0.08 9.2 115 Glucose (Ki) 2.00 - -The kinetic parameters from this table were determined from Lineweaver-Burk plots as described in section 2.7.2 (also Appendix 1). Proteins were determine according to Bradford (1976) using BSA as a standard. 70 transported outside the cell to be able to act on the cellodextrins released from hydrolysis of cellulose by C. fimi endoglucanases. If this is the case, CbgA should possess a leader peptide like sequence for secretion as for the Exg and EngA, B, and C from the same organism. Furthermore, one may speculate that other B-glucosidases may be present in C. fimi to hydrolyze cellobiose and other B-glucosides from plants. The Ki value of 2 mM glucose was measured using pNPG as substrate and is in the same range as those for bacterial and fungal B-glucosidases ranging from 0.66 to 135 mM glucose. Product inhibition for B-glucosidases has been reported to be either competitive or non-competitive depending on the source of the enzyme (Coughlan, 1985). In this case, the inhibition of CbgA activity on pNPG by glucose is of the competitive type (see appendix 1-E) meaning that the product glucose competes with the substrate during the formation of enzyme-substrate complexes. An interesting behavior was observed for the hydrolysis of pNPC (see appendix 1-F) which was also observed and reported elsewhere (Day and Withers, 1986). At high substrate concentration, very little nitrophenol is initially detected at 410 nm whereas at low concentration, a faster release of nitrophenol occurs. This suggests that the enzyme initially hydrolyses the bond between the two sugar residues to produce glucose and pNPG with no change in the absorbance at 410 nm. Subsequently, pNPG will compete with the remaining pNPC and soon enough, adsorbance at 410 nm will increase notably. The recombinant B-glucosidase would therefore be better classified as an exo-glucosidase enzyme. Interestingly, the recombinant enzyme is also specific for the hydroxyl group at position C2 since it does not hydrolyze the substrate p-nitrophenyl-B-mannoside (pNP-man) which differs from pNPG by having the C2 hydroxyl group in the opposite configuration. The recombinant enzyme shows weak activity on pNP-Acetylglucosamine which has a larger group and right configuration at this position (data not shown). This result could suggest a role for the C2-OH group of the sugar during enzyme activity or may simply reflect poor binding of the pNP-man with its axial hydroxyl. No change in absorbance was measured on pNP-7 1 Galactoside, pNP-Lactoside or pNP-Xyloside (not shown). Interestingly, only at high cellopentaose substrate concentration does the Lineweaver-Burk plot show an upward deflection of the curve (see Appendix 1-D). The upward deflection in Lineweaver-Burk plots for cellopentaose hydrolysis at high substrate concentration could be due to the excess of substrate causing enzyme inhibition. In this case, the substrate may bind at a different position away from the catalytic site with no reaction occuring preventing proper binding of another substrate molecule. Another explanat ion could be because of the transglycosylation reaction that could take place under those conditions as previously observed (Gusakov etal., 1984) where cellopentaose could become the acceptor molecule generating C6 and C4. Interestingly, the synthesis of inducer molecules are believed to require transglycosylation activity by enzymes that are associated with the membranes generating molecules which are capable of inducing for example, the lactose operon (Jobe and Bourgeois, 1972) and cellulase complexes (Nisizawa et al., 1971; Vaheri etal., 1979). 3.7.4. The amino acid composition of CbgA and various enzymes The amino acid composition of the purified enzyme CbgA and related enzymes is given in Table IX. No major differences are evident. 72 Table IX. Amino acid composition of various enzymes Molar rat io 3 A.A. CbgA A b g 1 C e x 2 C e n A 2 C e n B 3 C e n C 4 asx 0.73 1.00 0.83 0.86 0.71 0.68 glx 0.76 0.74 0.62 0.67 0.53 0.94 ser 0.71 0.35 0.50 0.59 0.66 0.60 giy 1.36 0.89 0.70 1.10 0.84 1.00 his 0.11 0.31 0.08 0.06 0.11 0.13 arg 0.33 0.43 0.27 0.35 0.19 0.22 thr 0.53 0.35 0.72 0.90 1.03 0.59 ala 1.00 1.00 1.00 1.00 1.00 1.00 pro 0.16 0.54 0.47 0.67 0.60 0.65 tyr 0.12 0.39 0.17 0.24 0.39 0.33 val 0.60 0.59 0.57 0.51 0.65 0.67 met 0.04 0.28 0.08 0.04 0.02 0.05 cys 0.11 0.11 0.10 0.12 0.06 -ile 0.48 0.26 0.13 0.20 0.11 0.12 leu 0.13 0.67 0.37 0.49 0.52 0.68 phe 0.11 0.39 0.35 0.18 0.23 0.22 lys 0.17 0.35 0.28 0.27 0.35 0.15 trp - 0.24 0.20 0.33 0.19 -#A.A. 1634 458 443 418 1065 1161 M r (kDa) 183 51 49 52 110 120 1, Wakarchuk, 1987; 2, Langsford, 1988; 3, Owolabi, 1988; 4, Moser, 1988. a, The amino acid composition is given on a Mol % basis relative to alanine residues equal to 1.00. (-) , Not available #A.A., Number of residues per enzyme, approximate for CbgA 73 3.8. The identification of the nativeC. fimi B-glucosidase A. To be able to compare the size, cellular location and enzymatic activity of the recombinant enzyme with the corresponding native enzyme, C. fimi B-glucosidases were part ia l ly pur i f ied and characterized. Proteins from a C. //m/CM-Cellulose grown culture were fractionated into cytoplasmic, membrane associated and secreted protein fractions. 3.8.1. The effect of carbon source on B-glucosidase activity During the study of cellulose hydrolysis by C. fimi, several enzymes with B-glucosidase activity have been detected. The enzyme activity is always observed independently of the carbon source suggesting that in C. fimi, B-glucosidase activity is constitutive. Nevertheless, the total pNPGase activity detected in C. fimi cultures varies with the carbon source used (Fig. 19). An increase of 2.4 fold in total B-glucosidase activity is observed when cells are grown in rich medium and a 3.7 fold increase is observed in minimal medium containing CM-cellulose. 6-glucosidase activity was found mostly in the cytosol fraction of C. fimi (from 65% up to 100% depending on the carbon source) but some activity was detected in the membrane fractions and in the culture supernatants, which suggested a multiple location of B-glucosidases in this organism (data not shown). 3.8.2. The fractionation of C. fimi B-glucosidases About 80% of the pNPGase activity in a culture grown on basal medium containing 1 % CMC was cell associated. About 66% of the cell-associated activity was in the cytoplasm and about 13% was membrane-associated. About 16% of the activity in cultures was in the culture supernatant and 2 1 % of the unrecovered activity was attributed to experimental loss, most of it still present in the discarded pellet of cell debris which was not processed. The B-glucosidases in the cytoplasm of CMC-grown cells were first fractionated by anion-exchange chromatography on a DEAE-Sephadex A-50-120 column (Fig. 20). B-Glucosidase activities were separated into 3 pools. Pool I (fractions 27 to 32) contained both pNPGase and 74 Fig. 19. Total B-glucosidase (pNPGase) activity in C. fimi cultures grown on various carbon sources. C. fimi cells were grown at 30°C in 100 mL of media containing 1 % of the various carbon sources to late log phase. Each Culture was tested for total B-glucosidase activity using pNPG as substrate. Values are given in total miliunits (mU) of enzyme per culture per O.D.6oonm where one unit releases one umol of pNP per min at 37°C. Carbon source 75 Fig. 20. Fractionation of C. fimi cytoplasmic proteins on a DEAE-Sephadex A-50-120 column. C. fimi cytoplasmic protein fraction obtained from cells grown in basal medium containing 1%-CM-cellulose was applied to a DEAE-Sephadex A-50-120 column (2.5 cm x 11 cm). After washing with 50 mM Tris-5 mM EDTA (pH 7.0), bound proteins were eluted with a linear salt gradient (0 - 0.85 M NaCl). Fractions were tested for protein concentration, pNPGase and cellobiase activities (legend on axis). Fractions 27 to 32 (Pool I), 44 to 57 (Pool II) and 58 to 69 (Pool III) were pooled separately. The proteins in Pool II and III were fractionated further on a Mono Q column. Pool III 0 10 20 30 40 50 60 70 80 90 100 Fraction number 76 cellobiase activities; Pool II (fractions 44 to 57) contained mostly pNPGase activity; and Pool III (fractions 58 to 69) contained mostly cellobiase activity with a small shoulder of pNPGase activity. The proteins concentrated from Pools II and III were purified further on a Mono Q column (Fig. 21 and 22) because it was beleived that one of those fractions may contain the native CbgA enzyme and that Pool I with its low amount of activity was not to be considered important. A single peak of pNPGase activity was obtained from Pool II (Fig 21-fraction 14 and 15) and fractions 14 and 15 were combined. Three major peaks of pNPGase activity were obtained from Pool III (Fig. 22- fraction 19 and 20, 32 and 33 and 39 to 41) and these fractions were combined, respectively. There was also a major peak with activity on cellobiose but not on pNPG (Fig. 22-fraction 34 to 37) and these active fractions were also combined. The combined active fractions 14-15, 32-33, 34-37 and 39-41 were then analysed further by gel electrophoresis for their content. 3.8.3. The detection of 6-glucosidase activity after gel electrophoresis The proteins from the various active fractions were resuspended in SDS-PAGE loading buffer containing 6-mercaptoethanol and were kept at 25°C prior to loading on gels. The proteins in those samples are beleived to have migrated in the polyacrylamide gel according to their globular size since all samples were not fully denatured by heat treatment. The rationale for not fully denaturing the proteins prior to the electrophoresis comes from an early observation that showed the absence of migration of CbgA during non-denaturing gel electrophoresis where SDS is omitted (not shown). 6-Glucosidase activity from those cells had previously been detected at the junction of the stacking and separating gels but was not discussed further because active bands were detected in the separat ing gels (Wakarchuk, personal communication). The usual explanation for absence of appropriate migration of a protein under non-denaturing conditions is often attributed to its isoelectric point which could be located above 7.0 preventing a net negative charge to the peptide (Jovin etal., 1964; Sigma, technical bulletin no. NKR-137). In the case of CbgA, aggregation or multime-rization of the enzyme could have precluded its migration in the gel. 77 Fig. 21. Fractionation of Pool II on Mono Q. Pool II from the DEAE-sephadex column was applied to a Mono Q anion-exchange column. After washing the column with 50 mM Tris-5 mM EDTA (pH 7.0), bound proteins were eluted by a linear increase in the NaCl concentration as shown. Fractions were analysed for protein and pNPGase activity. Fractions 14 and 15 were pooled. Fraction number 78 Fig. 22. Fractionation of Pool III on Mono Q. Pool III fraction from the DEAE-sephadex column was applied to a Mono Q anion-exchange column. After washing the column with 50 mM Tris-5 mM EDTA (pH 7.0), bound proteins were eluted by linear increase in the NaCl concentration as shown. Fractions were analysed for protein, and for cellobiase and pNPGase activities. Fractions 19 and 20, 32 and 33, 34 to 37 and 39 to 41 were pooled seperately. 0 10 20 30 40 50 Fraction number 79 Furthermore, heat denaturation of CbgA and other proteins (see Lacks and Springhorn, 1980) sometimes completely inactivates the enzyme. Activity can not be restored after removal of SDS (data not shown). The negatively charged SDS present in the gel and sample buffer which may have bound partially to the protein allowed CbgA to enter the stacking and separating gels and MUGase activity to be observed after removal of the SDS. Various protein fractions from E. coli and C. fimi cultures were analysed by zymogram (Fig. 23). Total cell extract from E. co//7pUC13 lacked MUGase activity (lane 1), that from E. co//7pUC13:62A31 gave a major active band at position B (lane 2) which comigrated with the major active component in the purified enzyme (lane 3). The purified enzyme contained several active components with a fast migrating band at position D. All these active components were found to react with the CbgA-directed antiserum (Fig. 24-lane 3). SDS-PAGE of the fully denatured purified enzyme, however, gave a unique band stained with Coomassie blue or reacting with the antibodies (Fig 16, lane 7; Fig 25, lane 1). The C. fimi enzyme(s) behave similarly (Fig. 23, lane 4 to 13). The C. fimi culture supernatant contains an apparent aggregate (lane 4) which also contains multiple components. A major active band (A) was located at the junction of the stacking and separating gels. The C. fimi membranes (lane 5) also gave an active band (A) located at the top of the separating gel. Due to the low concentration of B-glucosidase activity present in the C. fimi cytoplasm (lane 6), no MUGase activity was detectable. Nevertheless, C. fimi cytoplasm which was further fractionated into more concentrated and active fractions (see above), is shown in lanes 7 to 13 (Fig. 23). The Pool I fraction (lane 7) gave an active band at position D which comigrated with the fastest migrating band of the purified recombinant enzyme. The band at position D from Pool I also reacted with the antiserum (Fig. 24-lane 7). The native enzyme present in Pool I could correspond to the cloned enzyme and was named CbgA. I wish to emphasize that at this point, it is an early suggestion to relate CbgA to the enzyme in Pool I as more evidence will be shown later in the thesis. 80 Fig. 23. Zymogram of partially denatured protein samples using MUG as a substrate. By mixing various protein samples with loading buffer, non-fully denatured protein samples from E. coli (lanes 1 to 3) and C. fimi (lanes 4 to 13) were analysed by 10% SDS-PAGE. Active bands were visualised by soaking the gel in 1 mM MUG -1%(v/v) Triton X-100 -100 mM phosphate buffer (pH 6.6). Lane 1, 40 ug of E. coli JM83/pUC13 total cell extract; lane 2, 40 ug of E. coli JM83/pUC13:62A31 total cell extract; lane 3, 188 ng of purified recombinant CbgA. Lane 4, 1 9 u g o f C. fimi culture supernatant protein aggregate fraction; lane 5, 30 ug of C. fimi membrane fraction; lane 6, 40 ug of C. fimi cytosol fraction. Lane 7, 10 ug of C. fimi Pool I fraction; lane 8, 37 ug of C. fimi Pool II fraction; lane 9, 10 ug of C. fimi fraction 14 and 15 from Mono Q of Pool II; lane 10, 40 ug of C. fimi Pool III fraction; lane 11, 21 ug of C. fimi fraction 32 and 33 from Mono Q of Pool III; lane 12, 11 ug of C. fimi fraction 34 to 37 from Mono Q of Pool III; lane 13, 7.7 ug of C. fimi fraction 39 to 41 from Mono Q of Pool III. After 30 min incubation at 30°C, active bands were visualized by U.V. illumination and picture was taken. Letters A, B, C, D, E1 and E2 on the left indicate the locations of active bands. 81 The Pool II fraction (lane 8) gave very weak MUGase activity with a band located at position E1. The Pool II (lane 8), fraction 14-15 (lane 9) and fraction 19-20 (not shown) seem to lack MUGase activity since appreciable amount of pNPGase activity was present in those lanes. This suggests that MUGase activity from Pool II may originate from contaminating activity from Pool III (see Fig. 20) and that fractions 14-15 and 19-20 could contain identical enzymes. The Pool III fraction (lane 10) contained multiple components with 3 major active bands detected at position A, E1, and E2. Similar components are detected in the lane containing fraction 32-33 (lane 11). The fraction 39-41 (lane 13) showed active bands at position E1 and uniquely at position C. The enzyme from Pool II (Fig. 21) called CbgB1 is responsible for the major peak of pNPGase activity from C. fimi cytoplasm (Fig. 20) and is anticipated to have contaminating amount of activity in Pool III. The contaminating activity could correspond to fraction 19 to 20 from Pool III (Fig. 22) and this enzyme was called CbgB2. This suggests that CbgB1 and CbgB2 enzymes originating from different Pools (Pool II and III, respectively), may contain identical enzymes and could be named CbgB. The major active bands detected at the junction of the gels (A) could originate from aggregation of proteins. The other enzymes from fractions 32 to 33 and 39 to 41 of Pool III (Fig. 22) could also^be identical enzymes that are isomers and were called CbgD1 and CbgD2, respectively. Nevertheless, this could not be confirmed by the data obtained so far. The fraction 34-37 with strong cellobiase activity did not hydrolyze MUG efficiently (lane 12) and was called CbgC. The weak MUGase activity from that sample which was detected above position E2 is also detected in Pool III (lane 10) and in sample containing CbgD1 (lane 11). This suggests that CbgC is not responsible for the appearance of this MUGase active band and, consequently, CbgC may not have strong affinity for MUG or for pNPG. To be able to compare protein migration patterns and determine the sizes of the active peptides, various samples were analysed by SDS-PAGE after heat denaturation then stained with Coomassie blue in parallel with the corresponding non heat-denatured samples analysed by zymogram and Coomassie blue (data not shown). The CbgB1 and 82 CbgB2 samples stained with Coomassie blue contained several peptides and had identical protein migration patterns. These samples contained a major peptide with an Mr of 57 kDa. Nevertheless, the CbgB enzymes could not be identified with certainty because these enzymes were only partially purified and could not hydrolyze MUG when analysed by zymogram. The sample containing CbgC which lacked MUGase activity was stained with Coomassie blue and showed a unique band with an Mr of 88 kDa. From identical migration patterns between denatured and non-denatured samples containing either CbgD1 or CbgD2 enzymes, Mr were assigned to the active peptides. The CbgD1 enzyme has an Mr of 35 kDa. The CbgD2 enzyme showed two active bands comigrating with denatured peptides with Mr of 45 and 47 kDa. The Pool I fractions containing CbgA enzyme was not analyzed under those conditions. The Cbg enzymes were quantitatively analysed for their ability to hydrolyse pNPG and cellobiose (Table X). The CbgA enzyme could hydrolyse pNPG. The CbgB enzymes could hydrolyze pNPG and cellobiose. The CbgC enzyme could only hydrolyze cellobiose. The CbgD enzymes could hydrolyze pNPG but only CbgD1 had some activity on cellobiose. 3.8.4. Summary of C. fimi B-glucosidases A summary of the characteristics obtained so far on C. fimi B-glucosidases indicates the various differences between those enzymes and the relationship between the recombinant CbgA and its corresponding form in C. fimi. The C. fimi cytoplasmic fraction shows heterogeneity in its 8-glucosidases and so far, 4 different enzymes have been partially purified by ion exchange chromatography. Based on their enzymatic activities, a summary of the properties of these enzymes is given in table X. The CbgB enzymes have the highest specific activity on pNPG whereas the CbgC enzyme had the highest specific activity on cellobiose. Consequently, most of the pNPGase activity of C. fimi seems to be due to CbgB, whereas, most of the cellobiase activity seems to be due to CbgC. The CbgD enzymes were able to hydrolyse 83 Table X. Characteristics of C. fimi 13-glucosidases Fractions Enzyme Specific activity (U/mg prot.) pNPGase 1 Cel lobiase 2 MUGase M r (kDa) Pool 1 CbgA 47 0.3 + 183(a) Pool II 250 n.d. -14 and 15 CbgB1 450 5.5 - 57(b) Pool III 50 n.d. + 19 and 20 CbgB2 146 2.1 - 57(b) 32 and 33 CbgD1 52 5.3 + 35 34 to 37 CbgC <1 218 +/- 88 39 to 41 CbgD2 122 <0.1 + 45-47 The fractions in this table are from C. fimi culture grown on CMC basal medium and were obtained as described in section 2.6.2. Pool I, II and III fractions are from DEAE-sephadex column (Fig. 20). Fraction 14 and 15 is from Mono Q column of Pool II (Fig. 21). Fractions 19 and 20, 32 and 33, 34 to 37 and 39 to 41 are from Mono Q column of Pool III (Fig. 22). 1. One unit of enzyme releases 1 urnol of pNP per minute at 37°C. 2. One unit of enzyme releases 1 urnol of glucose per minute at 37°C. +, -, presence or absence of MUGase activity from zymogram (a) , M r of larger peptide from Western blot analysis (b) , M r after SDS-PAGE of major peptide in that sample n.d. not determined 84 pNPG but not as efficiently as the CbgB enzymes. The activity of CbgD1 on cellobiose could originate from contaminating CbgC activity since CbgD1 originates from fraction 32 to 33 and CbgC from fraction 34 to 37 (Fig. 22), fraction 33 containing appreciable amount of cellobiase activity. Eventhough a definite statement should not be made until further purification is achieved, the Cbg enzymes appear to be distinctive from each other in their relative substrate specificity. The presence in bacteria of more than one B-glucosidase gene has been reported antecedently (Gokhale and Deobagkar, 1989). Multiple B-glucosidase components are common in microorganisms and may possibly result from (i) complex formation between enzymes and polysaccharides; (ii) dissociation of subunits during experimentation; (iii) proteolytic cleavage of the peptide; (iv) multiple enzymes with different functions required for growth of the organism (Shewale, 1982). A preliminary study has shown that C. fimi contains at least two distinct 6-glucosidases (Wakarchuk etal., 1984). More recently, Kim and Pack (1989) have reported the molecular cloning in E. coli of two B-glucosidase genes from C. fimi. One of them encodes an aryl-6-glucosidase that hydrolyzes only pNPG but not cellobiose and may well correspond to the CbgD enzymes. The other gene encodes a "true" cellobiase that hydrolyzes both cellobiose and pNPG, and may well correspond to the CbgB enzymes. In C. fimi, CbgB is possibly active in cleaving phenolic glucosides of plant origin that may be liberated by plants following the attack by this cellulolytic bacterium. In addition, a third enzyme, called CbgC, has been isolated and was shown to have high activity on cellobiose. It is noteworthy to mention that the C. fimi genes were isolated by screening a genomic bank with pNPG as a substrate making unlikely the molecular cloning of the true cellobiase, CbgC, which does not hydrolyze this substrate. It appears that CbgC is the C. fimi B-glucoside glucohydrolase or "true" cellobiase that is involved in the hydrolysis of cellobiose released from the hydrolysis of cellulose by C. fimi cellulases. Finally, more studies are required to identify the corresponding native CbgA that could be present in the cytoplasmic fraction Pool I and/or in the secreted fraction as part of the protein aggregate. 85 3.9. Antiserum and Western blot analysis Various protein samples were obtained from E. coli and C. fimi where the relationship between the recombinant enzyme CbgA and the native enzyme remained to be established on another basis than simply enzymatic activity. A highly specific antiserum to CbgA was necessary for the detection of the recombinant forms of the enzyme within a mixture of bacterial proteins. Also, the detection of the native forms of CbgA would give some interesting information regarding its cellular location. Finally, a stronger relation could be established between the previously characterized recombinant CbgA and its corresponding forms in C. fimi. 3.9.1. Generation of rabbit polyclonal antibodies. For the detection of small quantities of antigenic peptides through the powerful technique of Western blot analysis, polyclonal antibodies to purified recombinant enzyme CbgA were raised in New Zealand white rabbit (section 2.8). An ELISA on protein fractions from E. co///pUC13:62 gave a titre of 1 0 " 4 for the antiserum with virtually no background for the E . co//7pUC13 cell extracts (Fig. 5A). The pre-immune rabbit serum also shows zero cross reactivity with the protein extracts from E . co//7pUC13:62 or with the purified enzyme CbgA. This indicates that the antiserum is highly specific for CbgA from E . coli and that it can be used at the appropriate dilution of 10~ 4 to detect antigenic peptides from SDS-PAGE. An ELISA on protein fractions from C. fimi culture supernatant (Fig. 5B) also gave a titre of 10~ 4 for the antiserum when compared to the rabbit pre-immune serum. Interestingly, both the C. fimi culture supernatant fraction, containing the protein aggregate, and also the membrane fraction from ultracentrifugation of crude protein extract gave a positive ELISA response. This result suggests that CbgA in C. fimi could be secreted in the culture medium. It isn't clear if the native CbgA from the C. fimi membrane fraction is physically attached to the membranes or, as for the recombinant enzyme detected in the E . coli membrane fraction, is present as aggregates which sediment at high speed centrifugation. The rabbit antiserum was used to analyse various protein fractions by Western blot. 3.9.2. Western blot analysis of the activity gel. The proteins present in the MUGase activity gel were blotted onto NC membrane and further analysed by Western blot (Fig. 24). The negative control containing E. co//7pUC13 cell extract (lane 1) shows no reactivity with the antiserum whereas all the enzymatically active components of E. co//7pUC13:62A31 cell extract (lane 2) or of the purified recombinant enzyme (lane 3) reacted with the antibodies. It is appropriate to mention that once boiled at 100°C for 2 min, the purified enzyme shows only a single reactive band (Fig. 25-lane 1). This suggests that the ladder of active bands in fig. 23 lane 3 could represent a set of partially denatured CbgA or aggregating fragments forming multimeric forms. This multiple banding phenomenon was also observed with the proteins in the C. fimi culture supernatant aggregate (Fig. 23-lane 4) which could also contain various denatured or aggregating forms of CbgA. All the MUGase active components in that fraction reacted with the antibodies. It is also appropriate to mention that once boiled at 1 0 0 ° C f o r 2 min, the C. fimi protein aggregate shows only 3 reactive bands (Fig. 25-lane 8). The nature of the aggregation is unknown but 6-glucosidase enzymes have been reported to form multimers in other organisms (Coughlan, 1985) and could be responsible for these observations . The active band in Pool I reacts with the antibodies and the kinetic characteristics of the Pool I enzyme are very similar to those previously reported for the recombinant CbgA enzyme. This indicates that the purified recombinant enzyme CbgA corresponds to the native enzyme present in Pool I from C. fimi cytoplasm. In the other lanes (Fig. 24-lane 8 to 13) none of the MUGase active components reacted with the antibodies indicating a clear difference between CbgA and the other 6-glucosidases characterized so far. 3.9.3. Western blot analysis of denatured protein samples. Western blot analysis using antiserum raised against the recombinant enzyme CbgA was used to detect specific peptides in fully denatured fractions separated by SDS-PAGE (Fig. 25). The purified recombinant CbgA (lane 1) shows a single band of about 183 kDa. The negative 87 Fig. 24. Western blot analysis of the zymogram. Proteins separated by electrophoresis were blotted onto NC filter and analysed by Western blot using anti-serum raised against purified recombinant CbgA. L a n e l , 40 ug of E. co//JM83/pUC13 total cell extract; lane 2, 40 ug of E. coli JM83/pUC13:62A31 total cell extract; lane 3, 188 ng of purified recombinant CbgA. Lane 4, 19 ug of C. fimi culture supernatant protein aggregate fraction; lane 5, 30 ug of C. fimi membrane fraction; lane 6, 40 ug of C. fimicytosol fraction. Lane 7, 10 ug of C. fimi Pool I fraction; lane 8, 37 ug of C. fimi Pool II fraction; lane 9, 10 ug of C. fimi fraction 19 and 20 from Mono Q of Pool II; lane 10, 40 ug of C. fimi Pool III fraction; lane 11, 21 ug of C. fimi fraction 32 and 33 from Mono Q of Pool III; lane 12, 11 ug of C. fimi fraction 34 to 37 from Mono Q of Pool III; lane 13, 7.7 ug of C. fimi fraction 39 to 41 from Mono Q of Pool III. 88 Fig. 25. Western blot analysis of the recombinant and native enzyme CbgA. Various protein samples were fractionated by 7% SDS-PAGE after denaturation at 100°C for 2 min. Samples were blotted onto NC filter prior to detection of proteins with antiserum raised against the purified recombinant CbgA. Lanes 1 to 4 contain E. coli proteins and lanes 5 to 9 contain C. fimi proteins. Lane 1, 50 ng of purified recombinant CbgA; lane 2, 30 u.g of E. coli JM83/pUC 13 total cell extract; lane 3, 30 u,g of E. co//JM83/pUC13:62 total cell extract; lane 4, 30 u.g of E. coli JM83/pUC13:62A31 total cell extract. Lane 5, 30 u.g of C. //m/'cytosol fraction; lane 6, 5 u.g of C. fimi Pool I fraction; lane 7, 30 u.g of C. fimi membrane fraction; lane 8, 9.5 u.g of C. fimi culture supernatant protein aggregate fraction; lane 9, 50 u.g of C. fimi cytosol (100X) from LBIs grown culture. Numbers on the left indicate Mr markers (kDa). 89 control E. co//7pUC13 (lane 2) shows no reactivity with the antibodies whereas proteins from E. co//7pUC13:62 (lane 3) contains a p183 band as well as very weak bands of about 197 and 210 kDa in size. When CbgA is over expressed in E. coli cells carrying pUC13:62A31 plasmid proteolysis of the recombinant enzyme is evident (lane 4). The major peptide in that lane comigrates with the purified enzyme in lane 1. One would expect a larger peptide from E. coli cells carrying pUC13:62 plasmid which contains the entire 7.2 Kb fragment (lane 3). A number of products, including a processed p183 product, are present. This suggests that the size of the major peptide expressed from pUC13:62A31 roughly correlates with the size of the processed product produced by E. coli carrying pUC13:62 plasmid. The intensity of the immunogenic protein bands is comparable to the specific activities of the samples. A more accurate study of the proteins made from the various deletions will be discussed later. Specific bands could be detected in fractions of C. fimi cells grown on CMC (Fig. 25). In the cytoplasmic fraction of CMC-grown cells (lane 5) and LBIs-CMC grown cells (lane 9), two bands of low intensity with molecular weights of 197 and 183 kDa were only visible on the original filter. Partially purified C. fimi cytosol fraction (lane 6) contains p183 and p137 peptides and a weakly reacting p60 peptide (see below). The C. fimi fraction enriched for membranes by ultracentrifugation of cleared broken cell lysate shows the presence of two polypeptides with M r of 197 and 183 kDa (lane 7). Finally, the C. fimi protein aggregate isolated from its culture supernatant shows the presence of several polypeptides (lane 8) with M r of 197, 170 and 137 kDa in sizes. The p60 polypeptide visible as a weak band on the original Western blot (and not visible on the photograph Fig. 25) is clearly visible in Fig. 26 (lane 2). Lane 1 (Fig. 26) which corresponds to a FPLC purified fraction of E. coli proteins (also Fig. 16-lane 10) contains a p183 and a p137 polypeptide but does not contain the p60 polypeptide. This p60 polypeptide is evident in the C. fimi pool I fraction (lane 2). For comparisons, this figure shows the p183 and p137 polypeptides from different origins and suggests that specific proteolysis of CbgA occurs in a similar way in both organisms. Nevertheless, only in the lane 9 0 Fig. 26. Western blot analysis of E coli and C. fimi proteins. Proteins from the fractions containing the purified p183, p137 and p60 polypeptides from E. coli (Fig. 15-fraction 28 to 31) and C. fimi Pool I (Fig. 20-fraction 27 to 32) were fully denatured, fractionated by 7% SDS-PAGE and blotted onto a NC filter. These fractions were then analysed by Western blotting using CbgA anti-serum. Lane 1, 13 ug (114 U) of protein from E. coli p183, p137 and p60 peptides. Lane 2, 5 ug (1 U) of protein from C. fimi Pool I fraction. The numbers on the left indicate Mr markers (kDa). 91 containing C. fimi proteins is the p60 polypeptide detectable with the antibodies. The absence of a p60 reacting polypeptide in lane 1 suggests that in E. coli (Fig. 16-lane 10), the FPLC purified p60 polypeptide may not correspond to the p60 polypeptide detected from C. fimi. The native enzyme in C. fimi has been identified, and like the recombinant enzyme CbgA, it could be cleaved into smaller polypeptides presumably by proteolysis. The presence of the native enzyme in the culture supernatant of CMC-grown C. fimi cells suggests again that CbgA is probably secreted by this organism into the culture medium to facilitate hydrolysis of cello-oligosaccharides to glucose. 3.9.4. Western blot analysis of proteins from various clones. Various B-glucosidase expressing clones obtained from deletions of 5' end DNA regions were analysed by Western blot (Fig. 27). Plasmid DNA was digested with Psfl restriction enzyme and restriction fragments were fractionated by agarose gel electrophoresis. The decrease in size of the major peptides detected by Western blot analysis (Fig. 27) correlates roughly with the decrease in size of the DNA in A17, A31, A21-4 and A21-15 clones (Fig. 28). With the exception of clone A27, all peptides show a gradual decrease in size with decrease in length of the DNA coding sequences. A peptide was detected from C. fimi cytoplasmic proteins purified by affinity chromatography on an antibody column (lane 11) and could comigrate with a peptide from A27 clone (Fig. 27-lane 3). Proteolysis is likely responsible for the appearance of lower-Mr species. A peptide of approximately 137 kDa in size is detected in most samples suggesting that proteolysis of CbgA in E. coli could occur at a specific location downstream from the A21-15 fusion point. This peptide may correspond to the previously isolated p137 polypeptide originating from processing of CbgA by E. coli. Therefore, up to 2 Kb of DNA can be deleted and still produce an active polypeptide suggesting that the CbgA catalytic domain is located in the carboxy terminal two third of the protein requiring less than 3.4 Kb of DNA coding sequence. 92 Fig. 27. Western blot analysis of protein extracts from B-glucosidase deletion clones expressed in E. coli. Various denatured protein samples from E. coli cells carrying plasmids encoding B-glucosidase were fractionated by 7% SDS-PAGE and analysed by Western blotting using CbgA anti-serum. Lanes 1 to 7, 30 jig of proteins from E. coli total cells extracts carrying : 1, pUC13; 2, pUC13:62; 3, pUC13:62A27; 4, pUC13:62A17; 5, pUC13:62A31; 6, pUC13:62A21-4; 7, pUC13:62A21-15 plasmids. Lane 8, 0.7 jig (2.6 U) of purified CbgA; lane 9, 2.5 ng (23 U) of p183-p137 polypeptides; lane 10, 5.5 ng (21 U) of p183-p137-p60 polypeptides. Lane 11 contains 0.7 (ig of protein from an antibody affinity purified fraction of the cytosol of LBIs-CMC grown C. fimi cells. Prestained Mr markers (kDa) are shown and the location of p183 and p137 is indicated. 93 Fig. 28. Restriction digest analysis of various deletion clones. Plasmid DNA isolated from various clones were digested with Pst\ and fractionated by 1 % agarose gel electrophoresis in 1 X TBE buffer. Lane 1, H/ndll l-EcoRI lambda DNA markers; lane 2, pUC13:62; lane 3, pUC13:62A27; lane 4, pUC13:62A17; lane 5, pUC13:62A31; lane 6, pUC13:62A21-4; lane 7, pUC13:62A21-15 plasmids. Sizes of M r standards are shown (Kb). 1 2 3 4 5 6 7 94 3.10. DNA and amino acid sequence determination DNA sequencing was initiated at the region surrounding the A31 fusion point and going in both direction. Part of the nucleotide sequence of the 7.2 Kb insert was determined, the region covered starts from the BamHI site at the 5' end of the insert up to approximatly 2200 bp downstream (Fig. 29) but not all regions were sequenced on both strands. Nevertheless, some important conclusions could be drawn. 3.10.1. DNA sequence analysis of the 5' end of cbgA. A unique ATG located at position 636 was found to be in frame with the CbgA coding sequence of the A31 clone. Nevertheless, the upstream region did not indicate the presence of a typical prokaryotic RBS which would allow initiation of translation of CbgA at this ATG. To verify whether the region located upstream of this ATG does promote expression of cbgA, S1 nuclease analysis of total RNA from C. fimi CMC-grown cells was performed. Initially, a 214 bp 5'labelled probe which covers this region was generated by labelling the 5* end of the Xho\ site at position 705 followed by digestion with Stu\ at position 493. The uniquely labelled fragment released was isolated and used to detect the presence of a specific mRNA from C. fimi CMC-grown cells by Northern blot analysis (Appendix 2). The autoradiogram was put in Appendix because of the poor quality and low intensity of the Northern blot. This may have resulted from a low transfer efficiency or degradation of large mRNA's. Nevertheless, a specific mRNA band of approximatly 5400 nucleotides in length was detected as well as a weaker band just under the 2.37 nt marker. This RNA preparation was allowed to hydridize to the heat denatured probe which was then followed by S1 nuclease analysis of the protected DNA-RNA hybrids (Appendix 3). At the lowest concentration of S1 nuclease used (lane 1), several 5' ends were detected and were mapped at the positions 593, 637 and 675 nt relative to the DNA sequence (Fig. 29). These detected fragments were 112, 68 and 30 bases in length, respectively. As more sequencing data will become available and in the context of those identified 5' ends, the importance of those ends remains to be established since the A28 fusion located upstream of those mapped ends is out of frame with the CbgA 95 Fig. 29. Nucleotide sequence of the 5' end of cbgA. The 2189 nucleotides shown start at the eamHI site at the 5' end and extend past the Kpn\ site in the 7.2 Kb insert. RBS indicates the putative ribosome-binding sites preceding the GTG initiation codons. The proposed signal sequence is indicated in the boxed area. The N-terminal amino acid sequences of CbgA:CexcBD a n d polypeptides p183 and p137 are underlined. Relevant restriction sites are indicated. The fusion points of various deletion clones are shown ( ^ ). Numbers refered as #bp/#residue below. 1 BamHI •1 RBS gga tec ggt cgt ggt gtc acg aag gag tga cc 33 / 1 63 / 11 gtg ctg cgt gee cgt ccg acc ctg etc cgc cgc acc cgc acc acc gec .39.?, age ccg ccg |met leu arg ala arg pro thr leu leu arg arg thr arg thr thr ala gly ser pro prol 93 / 21 123 / 31 gac agg cac cga cca ggc gtc gec gcg ctg acg gcg etc gcg etc acg gtc ccc etc gcg |asp arg his arg pro gly val ala ala leu thr ala leu ala leu thr val pro leu alal 153 / 41 183 / 51 etc gee gcg gee gca ccc gec gcg gee teg gee acg gcg acc etc ccg gag gca ccc gca lieu ala| ala ala ala pro ala ala ala ser ala thr ala thr leu pro glu ala pro ala 213 / 61 243 / 71 ccg gcg a eg gcg teg tec gec ccg gca gcg gca ccg gaa gee gca ccg gcg gee gcg ggc pro ala thr ala ser ser ala pro ala ala ala pro glu ala ala pro ala ala ala gly 273 / 81 303 / 91 gac ctg gec egg acc ggg acg gcg acg gee teg cag cac cag gcg gac ggc gac ggc acg asp leu ala arg thr gly thr ala thr ala ser gin his gin a l a asp gly asp gly thr 333 / 101 363 / 111 t t c C C C ccg gac gcg gcg ate gac ggc gac ccg gcg acc cgc tgg gcg age ggc aac ggc phe pro pro asp ala ala l i e asp gly asp pro ala thr arg t r p ala ser gly asn gly 393 / 121 27 423 / 131 17 ccg gac gcg gac gtc gag t t c acg gec tgg etc cag gtc gac etc ggt gcg acg gcg teg pro asp ala asp val glu phe thr ala trp leu gin val asp leu gly ala thr ala ser 453 / 141 tgg 483 / 151RBSStul tac gtc gac egg gtg gcg etc gcg gag gcg gcg tac gcg aag gee egg gtg cag gtc val asp arg val ala leu ala trp glu ala ala tyr ala lys ala tyr arg val gin val 513 / 161 28 543 / 171 gcg a eg gee gee ccg cag gac ccg gcg t^g tgg acg acg gtg cac acc gag acg gcg ggc ala thr ala ala pro gin asp pro ala ser t r p thr thr val his thr glu thr ala gly 573 / 181 603 / 191 gac gga ggg acg gac gac gtc acg etc ccg acg ccg gee gac gcg cgc tac gtg cgc ate asp gly gly thr asp asp val thr leu pro thr pro ala asp ala arg tyr v a l arg i l e 633 / 201 663 / 211 24 cag atg gac gcg cgc acg teg t t c gac tgg gac gee ccg acg ctg cac ^ g tac ggc tac gin met asp ala arg thr ser phe asp trp asp ala pro thr leu his trp tyr gly tyr 693 / 221 723 / 231 teg ctg t t c gcg etc gag gtc tac ggc acg ccg ggc gcg gtg gcg acg gcg t t c ggg acg ser leu phe ala leu glu val tyr gly thr pro gly ala val ala thr ala phe gly thr 753 / 241 783 / 251 age cgt gtc egg gtg ccg gcg ggc cag acc gcg cag gtg ccg gtc gtc etc gcg get ccg ser giy val arg val pro ala gly gin thr ala gin val pro val val leu ala ala pro 813 / 261 gtg 843 / 271 Smal gtg gcg cag gac acg acc egg gtc gcg teg acg ggc ggc acg gcg gtg ccc ggg acg val ala gin asp thr thr val arg val ala ser thr gly gly thr ala val pro gly thr 96 873 / 281 903 / 291 gac t t c acc gco etc gac gag acg etc acg t t c ccg gcg ggc gcg acg acg gec acg gtc asp phe thr ala val asp glu thr leu thr phe pro ala gly ala thr thr ala thr val 933 / 301 963 / 311 gac gtg gtg acg acg gac cac ggc ccg ctg gec ccg gtc egg acg gtc gtg ctg gag ctg asp val val thr thr asp his gly pro leu ala pro val arg thr val val leu glu leu 993 / 321 1023 / 331 acg gag ccg ggc gac ggc etc gtc ctg ggc ggc cgc acg acg gcg acg gtg acg ate acg thr glu pro gly asp gly leu v a l leu gly gly arg thr thr ala thr val thr i l e thr 1053 / 341 1083 / 351 ccg cac egg ccg ctg ccg gac gtc ggc gcg gtg acg gtg etc gac gac tac gag gac ggc pro his arg pro leu pro asp val gly ala val thr val leu asp asp tyr glu asp gly 1113 / 361 21-4 1143 / 371 gtg ccg gcg ggc tac acg acg tgg ggg age ggc gca ccg gtg acg ccg gtg ctg age acg val pro ala gly tyr thr thr t r p gly ser gly ala pro val thr pro val leu ser thr 1173 / 381 1203 / 391 acg acc acg gac cga ccg ggt gcg ccg gcg ggc age cac gcg ctg gtc ggc acc gtg ggc thr thr thr asp arg pro gly ala pro ala gly ser his ala leu val gly thr val gly 1233 / 401 1263 / 411 ggc ccg gcg gga ccc ggt gac tgg t t c ggg etc acg cac gac ctg ccg ccg acg gac tgg gly pro ala gly pro gly asp t r p phe gly leu thr his asp leu pro pro thr asp trp 1293 / 421 1323 / 431 teg cac cac gac ggg t t c acg t t c tgg t t c etc ggc acg ggc ggc ggc ggg ctg ctg egg ser asp his asp gly phe thr phe trp phe leu gly thr gly gly gly gly leu leu arg 1353 / 441 1383 / 451 tac gag etc aag age ggc ggg cag ctg t t c gag acg teg gtc gtg gac gac acg gcg ggc tyr glu leu lys ser gly gly gin leu phe glu thr ser val val asp asp thr ala gly 1413 / 461 1443 / 471 Smal tgg cgc egg gtc aac gtc gcg t t c ggc gcg cct gcg ctg aag aac gac ccg ggc age gac trp arg arg val asn val ala phe gly ala pro ala leu lys asn asp pro gly ser asp 1473 / 481 1503 / 491 gcg egg t t c gac ccg acg gcg teg acg ggc tgg gcg ate acg ctg acc gac ctg ggc gcg ala arg phe asp pro thr ala ser thr gly t r p ala i l e thr leu thr asp leu gly ala 1533 / 501 1563 / 511 gcg tgg cag ctg gac gac etc ggc ctg tac gac cgc gtg acg acg gtc gag gac gcg gag ala trp gin leu asp asp leu gly leu tyr asp arg val thr thr val glu asp ala glu 1593 / 521 1623 / 531 ggc gac gtc ccc get cgc gag ccg ggc age acg gtc ggc ctg t t c acg tgg ggc teg teg gly asp val pro ala arg glu pro gly ser thr val gly leu phe thr trp gly ser ser 1653 / 541 .1683 / 551 ggc get cag gtg teg etc ggc gtg acg cag cag gac cgc gag ggc ggt ccg gcg gac aac gly ala gin val ser leu gly v a l thr gin gin asp arg glu gly gly pro ala asp asn 1713 / 561 1743 / 571 21-15 cac gtg etc teg ggg cgc eta cct ggt ccg teg ggc ggc tgg ggc ggg t t c aqc cag aac his val leu ser gly arg leu pro gly pro ser gly gly trp gly gly phe ser gin asn 1773 / 581 1803 / 591 etc gec gcg ccg cag gac tgg age teg t t c cgc ggc ate egg ctg etc tgg tac gcg teg leu ala ala pro gin asp trp ser ser phe arg gly i l e arg leu leu trp tyr ala ser 1833 / 601 1863 / 611 cag gac acg cgc ccc gcg teg ccg acg gec ggt gac gac ate aag gtc gag etc aag gac gin asp thr arg pro ala ser pro thr ala gly asp asp i l e lys val glu leu lys asp 1893 / 621 ' 1923 / 631 ggc ggc ccg gac ggc gag cac teg gag ctg tgg gcg acg acg t t c aag gac aac tgg teg gly gly pro asp gly glu his ser glu leu t r p ala thr thr phe lys asp asn trp ser 1953 / 641 1983 / 651 ccc gac ggc age cgc tgg aag etc gtc gag ctg ccg t t c gac cag t t c acg ctg ggc ggg pro asp gly ser arg trp lys leu val glu leu pro phe asp gin phe thr leu gly gly 2013 / 661 2043 / 671 tac cag ccg ggt gac gcg cag acc cgc aac ggc acg etc gac etc acg teg gcg tgg ggg tyr gin pro gly asp ala gin thr arg asn gly thr leu asp leu thr ser ala trp gly Kpnl / 6 8 1 2103 / 691 tac gee ctg acg t t c gtg ccg ggc acc gcg aac ccg gtg cgc tgg gcg gtc gac gac gtg tyr ala leu thr phe val pro gly thr ala asn pro val arg trp ala val asp asp val 2133 / 701 2163 /'• 711 cag ctg tac ggc teg geg gtg ccc gcg ccg acg gec gag gtc get ccg gec acc gac gin leu tyr gly ser ala val pro ala pro thr ala glu val ala pro ala thr asp 97 reading frame and lacks expression of enzymatic activity. Also, the probe is fully protected in lanes 1 and 2 (Appendix 3) which suggests that an additional 5' end could be located upstream of the sequence covered by the probe used. As more sequencing data became available, the legitimate transcriptional start site of the cbgA gene should be located upstream of the Stu\ site. The deletions generated previously at the 5' end of the cbgA structural gene were localized by DNA sequencing of the fusion region. The sizes of the remaining insert DNA in clones A17, A31, A21-4 and A21-15 (Fig. 29) correlated roughly with the sizes of the major peptides detected by Western blot analysis (Fig. 27), showing that CbgA initiated several hundred bp upstream of the A17 fusion point. A putative GTG initiation codon, preceded by the typical Shine-Dalgarno sequence GAAGGAG, was located 33 bp downstream of the SamHI site. The presence of a GTG codon for initiation of translation is uncommun in E. coli but is to be expected more frequently in microorganisms with high G+C genomic content like C. fimi and its cenC gene (Table 1) and Streptomyces (Bibb etal., 1985). This GTG may correspond to the CbgA initiation codon. Furthermore, a typical prokaryotic purine rich RBS sequence GAAGGAG which is complementary to the 3'-OH end sequence CUCCUUA-OH of E. coli 16S rRNA subunit is located upstream of this GTG and has a spacing of about 5 bases which is within average distance for RBS (Hawley and McClure, 1983). Consequently, this putative translational initiation site would most certainly explain the detection of 8-glucosidase activity after translational frameshift of the SamHI site at the 5' end of the insert. It is likely that upstream regulatory sequences have not been isolated within this insert. The deduced amino acid sequence following the putative GTG start site shows striking similarities with the prokaryotic signal peptides of Cex and CenA, B and C (Table XI). The presence of CbgA in C. fimi culture supernatants could require some kind of secretory signal. By comparison with those C. fimi enzymes, it is suggested that a signal peptidase processing site may be present between alanine residues 42 and 43, whereas, a prediction of prokaryotic secretory signal sequence from PC GENE computer program has suggested a processing site 98 Table XI. Comparison of leader peptides from C. fimi cellulases Protein Residues CbgA Cex CenA CenB CenC CbgA Cex CenA CenB CenC ML-RARPT LLRRTRTT-AG-SPPDRHRPGVAAPLT MP -RTTP AP GHP ARG ART AL RTT : RRR AATL -MSTR-R-T A-AA—LL- AA AA ML-R QVP RTLVAGGS A-LA MVSR-R-SSQ—ARG AL T-A WATLA -ALA—LTV-P-L ALA AAAPAAA*£M: V-VGATV-VLP AQA*A1T_ VAVGG-VTAL-T-TTAAQA*ARG_ VAVGV-L-VAP-LATGAAA*A£T_ LALA—L-AG-S-GT-ALA*A£E Amino acid residues are indicated in single letter code. Space (-) indicates a gap left to improve the alignment. Underlined letters represent the first residues of the mature enzymes. Star (*) indicate the location of the leader peptidase cleavage site. 99 between residues 51 and 52. The following region located between amino acids 43 and 79 is enriched in Pro and Ala residues representing about 78% hydrophobic residues. Regions enriched in Ala, Pro, Thr or Ser have been reported for various cellulolytic enzymes where the specific function of such sequences remains to be determined (Warren etal., 1986; Hall era/ . , 1989: Nakai etal., 1988). The carboxy proximal segment of the endo-type semi-alkaline cellulase CasA leader sequence from the alkalophilic Streptomyces strain KSM-9 contains an Ala-Pro repeat region which shows a striking homology with the CbgA Pro-Ala rich region (Nakai et al., 1988). The casA gene product apparently initiates with a GTG codon which is followed by an unusual long leader sequence of 70 A.A. This Pro-Ala rich region could be an hinge region involved in secretion of CbgA by C. fimi. The elevated expression of CbgA from E. coli c a r r y i n g pUC13:62A17 and A31 plasmids could be the result of deletion of DNA encoding the Pro-Ala rich region from CbgA as E. coli would become Pro-Ala depleted if required to produce high levels of proteins rich in specific residues as previously mentioned (Hall et al., 1989). It is well established that amino acid starvation shuts down synthesis of the translational machinery in E. coli. There is little explanation other than the large size of the CbgA protein for the lack of secretion of CbgA into the periplasm of E. coli carrying pUC13:62 plasmid. Nevertheless, some evidence that the CbgA leader peptide can promote export of a protein in the periplasmic space of E coli comes from Western blot analysis data of osmotic shockate of E. co/ /7pUC18:cbgA-cexQQQ expressing a CbgA:CexQgQ fusion peptide (section 3.10.4). This fusion peptide which is composed of part of the N-terminal region of CbgA is about 7 times smaller than CbgA and was detected in the periplasmic space of those cells. 3.10.2. N-terminal amino acid sequences of polypeptides related to CbgA The N-terminal amino acid sequences of the p183, p137 and p60 polypeptides were determined (Table XII) because it was critical to 1 0 0 Table XII. N-terminal amino acid sequences of CbgA:CexQgp, p183, p137 and p60 Polypeptide N-terminal amino acid sequence C b g A : C e x C B D SATATLPEAPAPATA p183* GTPGAVATAF (B-Gal-CbgA) S_VYGTPGAVA p137 GSAVPAPTA-VAPATD p60 AAKDVJ<FG.NDA-VKMLGVNVLADAVKVTL (continue) GPKG-NVVLDKSFGAPTITKDGVS-A-QIL-D The amino acid residues are indicated in single letter code. * Two residues were obtained for each cycle, indicating ragged ends as shown. A gap (-) indicates the absence of detectable phenylthiohydantion in that cycle. Data were provided by Ms. S.Keilland from the University of Victoria. The DNA probes were synthesized according to the underlined sequence of the p60 polypeptide. 1 0 1 establish a relation between the DNA data and the protein work. The N-terminal amino acid sequencing data of the p183 peptide gave a sequence mainly composed of 2 residues per cycle. The location of the amino acid sequences obtained by Edman degradation of the p183 polypeptide which is a fusion peptide with 7 residues belonging to (3-gal polypeptide, were resolved by comparison of the corresponding amino acid sequence of the DNA sequencing data of the pUC13:62A31 fusion region with the amino acid sequencing data. By starting with the Gly residue of the first cycle, a first sequence was located following the predicted amino acid sequence and matching one residue at a time from each cycle. The remaining amino acid sequence of each cycle was then compared with the predicted sequence and was found to correspond to the sequence encoded in part by the fused 3-gal alpha-peptide, the Ser residue was located at the fusion point and was part of that peptide. At the fourth cycle, twice the amount of Gly residues were detected originating from the simultanous appearance of residues 229 and 232 of CbgA. These data indicate the presence of one polypeptide with different N-terminal sequences apparently processed by proteolytic cleavages, between residues Arg-6 and Ser-7 of the B-gal alpha-peptide (not shown) and between residues Tyr-228 and Gly-229 of CbgA (Fig. 29) near the fusion point of A31. Furthermore, the p137 amino acid sequence (Table XII) was located at the end of the given sequence and was also processed between identical residues, Tyr-703 and Gly-704. A rather specific processing event was also observed in E. coli expressing Cex from C. fimi (Gilkes etal., 1988). In an attempt to locate the p60 amino acid sequence data to the DNA, a set of DNA probes correspon-ding to part of the p60 amino acid sequence was used to screen the entire 7.2 Kb insert by Southern blot analysis. The probes were a mixture of 17-mers with the possible sequences AA(G/A)GA(C/T)GTNAA (G/A)TT(C/j)GG corresponding to residues K-D-V-K-F-G located at the beginning of p60 amino acid sequence (Table XII). This region was chosen because of the low number of variants for the synthesis of oligo-probes with multiple codons. No specific hybridization was obtained, therefore, the location of the amino acid sequence of the p60 polypeptide and its relatedness to CbgA could not be established. The only 102 interesting feature was an 83% homology between residues 18 to 56 of p60 and residues 39 to 77 of the 65 kDa cell wall protein antigen of Mycobacterium tuberculosis (Shinnick, 1987). Proteolysis is most likely to be responsible for the apperearance of lower-Mr species from the various deletions analysed so far as well as from C. fimi where a peptide of approximatly 137 kDa in size was detected in most samples. This peptide could correspond to the previously isolated p137 peptide originating from processing between residues Tyr-703 and Gly-704 of CbgA isolated from E. coli carrying pUC13:62A31 plasmid. The reasons for the removal of more than 700 residues from the N-terminus of CbgA is not clear. These amino acids are not essential for activity since 1765 bp of DNA were deleted in the A21-15 clone without a strong decrease in the specific activity. The presence of long amino acid sequences between the signal sequences and the N-termini of the mature processed proteins have been reported for the secreted alkaline and neutral proteases of B. amyloliquefaciens (Vasantha era/.,1984). Similarly, proteolytic cleavage also occurs after a Tyr residue in the case of the apr[BamP] gene product (Vasantha et al., 1984). A possible role of the pro-sequences of these proteases could be in secretion of these enzymes. The function of the pro-sequence from CbgA is unclear at present. 3.10.3. Mutation of the CbgA translational start site To confirm the putative GTG start site of translation, a single base frameshift mutation was generated within the codon (Fig. 30). The frameshift did not reduce the expression of CbgA in E. coli, suggesting that translation was initiated at another codon further downstream. The polypeptide produced by the frameshift mutant was smaller than that produced by the wild-type (Fig. 31). This suggested the presence of an internal reinitiation site downstream from the putative GTG start site. Examination of DNA sequence revealed a putative Shine-Dalgarno sequence (CGAAGG) followed by a GTG codon located at position 504. The A27 deletion mutant which is out of frame with the CbgA coding sequence and also with the Plac reading frame has a similar specific activity to the wild-type but produces a number of peptides, all of which 103 Fig. 30. Mutation of the putative GTG start site of CbgA. The putative CbgA GTG start site (underlined) was mutated by primer extension. The 2 Kb fragment from the C. fimi DNA (hatched bars) subcloned into pUC19 giving rise to pUC19:1.9 plasmid was transferred into the vector pTZ19R. Consequently, pTZ19R-2 ssDNA was isolated and allowed to hybridize with the FP21 oligo prior to synthesis of dsDNA. Double stranded pTZ19R-2 plasmid containing the extra C base (pTZ19R-2FP21) was identified by DNA sequencing. The wild type 6.7 Kb Stu\-EcoR\ fragment was subcloned into pTZ19R-2FP21 creating pTZ19R-7. The DNA sequence of the beginning of the insert is shown, the location of the putative RBS is indicated and star (*) shows the location of the C base added to the wild type DNA sequence. Vector DNA (solid bars) and convenient restriction sites are shown. Scale of drawing is indicated above. 104 Fig. 31. Western blot analysis of the polypeptides produced by the cbgA translational frameshift mutant. Lane 1, 30 u.g of E. co//7pTZ17R-7 total cell extract (or mutant); lane 2, 30ng of E. co//7pUC13:62 total cell extract (or wild-type); lane 3, 30 ng of E. co//7pTZ19R total cell extract (control); lane 4, 30 u.g of E. co//7pUC13 total cell extract (control). Prestained Mr standards (kDa) are shown on the right. Antigenic polypeptides were detected using CbgA antiserum (see section 2.9). Arrows indicate the location of the relevant polypeptides. 1 2 3 4 - 2 1 0 - 1 1 0 -76 1° 5 are smaller than the wild-type polypeptide. The end-points of the deletions in A27 and A28 are nucleotides 408 and 540, respectively. This supports the suggestion that the GTG at nucleotide 504 functions as an internal translation initiation site. The A28 deletion is also out of frame but it does not express B-glucosidase activity. Internal reinitiation was also observed in the x/nZ gene of C. thermocellum (Grepinet et al., 1988) and in the x/nA gene of P. fluorescens subspecies cellulosa (Hall ef al., 1989). In the case of cbgA, the initiation of translation from the upstream sequence adds 157 residues to CbgA and could allow secretion of the enzyme whereas initiation from the downstream GTG deletes the putative leader sequence and the Pro-Ala rich region and may force the enzyme produced to remain in the cytosol. CbgA was indeed detected in C. fimi cytosol and culture medium. 3.10.4. Cellulose affinity chromatography of a C b g A : C e x Q B D fusion peptide To facilitate the determination of the N-terminal amino acid sequence of CbgA, a fusion protein was constructed comprising the N-terminal part of CbgA fused to the cellulose binding domain of Cex (CexQgo)- The fusion product was purified by cellulose affinity column chromatography essentially as described (Gilkes etal., 1988). Initially, a gene fusion was constructed using in-frame Stu\ restriction sites within CbgA and Cex coding sequences. The fusion peptide CbgA:CexQgQ, expressed in E. coli carrying the pUC18:cdgrA-cexQBQ plasmid, contained the first 154 residues of CbgA with its putative leader sequence and Pro-Ala rich region (Fig. 29) fusedto the last 133 residues of Cex comprising the Pro-Thr box and the cellulose binding domain. The CbgA:CexQgD fusion protein reacted with antiserum to Cex and was detected in E. coli periplasmic space but its Mr of 37 kDa (Fig. 32) was 8.5 kDa larger than the expected value. The discrepancy between the predicted and observed Mr values could be attributed to the presence of the Pro-Thr box. Peptides derived from CenA and containing the Pro-Thr box migrate slower on SDS-PAGE (Gilkes etal., 1989). The amino acid sequence of the N-terminal region of the fusion peptide matched perfectly well with residues 50 to 64 of the predicted sequence of CbgA 106 Fig. 32. Purification of the CbgA:CexcBD fusion peptide. The CbgA:CexcBD fusion peptide was purified from E. coli/p\JC'\8:cbgA-cexQBQ total cell extract by cellulose affinity chromatography on CF1 cellulose. The water eluate was concentrated by ultrafiltration and proteins were analyzed by (A) Coomassie blue staining and (B) Western blot of a 10% SDS-PAGE using antiserum to Cex protein. Lane 1, Mr standards (kDa); lane 2, 23 ug of protein; lane 3, prestained Mr standards; lane 4, 23 ug of protein. The location of the antigenic band corresponding to the CbgA:CexQgp fusion peptide is shown. 107 (Fig. 29). The observed Mr (37 kDa) of the fusion polypeptide is 13.5 kDa greater than predicted suggesting that the Pro-Ala rich region of CbgA may also affect migration of peptides on SDS-PAGE. 3.11. General discussion The recombinant C. fimi B-glucosidase A (CbgA) was characterized in E. coli. The gene was over expressed and the enzyme produced was purified. The kinetic parameters indicated that CbgA is an exo-B-1,4-glucan glucohydrolase (EC3.2.1.74) because of its linear action and preference for soluble cellodextrins released during the hydrolysis of cellulose. Similar enzymes of fungal and bacterial origins acting on oligomeric substrates, such as cellodextrin glucohydrolase, have been reported elsewhere (see Schmid and Wandrey, 1989). So far, 13 different fungal exo-B-1,4-glucosidases have been characterized and shown to hydrolyze cello-oligosaccharides more efficiently than cellobiose (Schmid and Wandrey, 1989). Only one similar bacterial exo-B-1,4-glucosidase has been characterized but this enzyme hydrolyses pNPG more efficiently than the oligo-saccharides and is found cell-associated (Ait era/., 1982). The B-glucosidase expressed in E. coli carrying the pUC13:62 plasmid is apparently a very large protein ( M r of 210 kDa). E. coli expressing cbgA does not appear to be affected by the production of this heterologous peptide suggesting that it can produce proteins of unusually high molecular weights. Only a few proteins with molecular masses greater than 100 kDa are found in E. coli (Tsung et al., 1989). The lack of secretion by E. coli of a normally secreted protein is often a problem that can be circumvented by the use of an alternative host. Apparently, CbgA can be processed in a specific and perhaps identical manner in E. co//and C. fimi suggesting that the structural features of the protein may dictate the way processing occurs. There is some evidence that CbgA forms aggregates either with itself, as do other B-glucosidases, or with other proteins. The nature of the aggregation remains unknown at present. Interestingly, the aggregation of several different peptides that have various cellulolytic activities occurs in nature. These complex structures are known as "cellulosomes". The 108 corresponding C. fimi CbgA enzyme has been identified within a group of isoenzymes hydrolyzing various B-glucosides. CbgA is secreted by C. fimi which is unusual for a bacterial B-glucosidase. The analysis of the 5' end of cbgA was found to be very informative regarding the mode of expression and structure of part of the protein indicating : (1) the presence of a putative leader sequence for secretion of the enzyme, composition of which was similar to those found in the other C. fimi cellulase leader peptides, (2) an internal translational initiation site presumably involved in the production of a peptide lacking its secretory signal peptide. This could preclude the secretion of that peptide into the culture medium. A need for the enzyme within the cell and also secreted into the medium can be rationalized if, during growth on cellulose, large cellodextrins must be hydrolysed in the culture medium whereas smaller ones enter the cells. One could also speculate that the expression of cbgA in C. fimi is regulated by a set of two different promoters. The first one catabolite repressed and inducible by cellulosic derivatives responsible for expression of cbgA from the most upstream region allowing secretion of CbgA in the culture medium where its activity is required. The second one constitutive and located within the structural gene, may prevent secretion of CbgA which could be involved in the hydrolysis of other B-glucosides that may enter the cells or the generation of inducer molecules from transglycosylation activity. Interestingly, a sequence of 9 bp inverted repeats is located near a -35 region upstream from the second translational initiation site and may be some kind of regulatory sequence. The analysis of this rather speculative mode of expression of cbgA in C. fimi would require cloning of more of the region upstream from the gene followed by transcriptional analysis of induced and non-induced states of the system. Mutants lacking CbgA activity are essential to define the precise function of CbgA in the hydrolysis of cellulose by Cellulomonas fimi. Finally, CbgA could be used in conjunction with an endo-glucanase to produce glucose directly from cellulose. C. fimi 6-glucosidase A may play an important role in the hydrolysis of cellulose. Its size may be indicative of the complexity of this role. 109 REFERENCES Ait, N., Creuset, N. and Cattaneo, J. (1979) Characterization and purification of thermostable B-glucosidase from Clostridium thermocellum. Biochem. Biophys. Res. Comm. Q0_: 537-546. Ait, N., Creuzetm, N. and Cattaneo, J. (1982) Properties of B-glucosidase purified from Clostridium thermocellum. J. Gen. Microbiol. 128 : 569-577. Alexander, J.K. (1968) Purification and specificity of cellobiose phosphorylase from Clostridium thermocellum. J. Biol. Chem. 242 : 2899-2904. Armentrout, R.W. and Brown, R.D. (1981) Molecular cloning of genes for cellobiose utilization and their expression in Escherichia coli. Appl. Environ. Microbiol. 41:1355-1362. Atkinson, T. and Smith. M. (1984) In; Oligonucleotide synthesis: a practical approach. Gait, N.J.(ed), IRL Press, Washington. Bates, N.C. (1987) Characterization of cbg : a cloned gene encoding an extracellular B-glucosidase from Cellulomonas fimi. M. Sc. Thesis, the University of British-Columbia. Beguin, P., Gilkes, N.R., Kilburn, D.G., Miller, R.C.Jr., O'Neil, G.P. and Warren, R.A.J. (1987) Cloning of cellulase genes. CRC Critical Reviews in Biotechnology, vol 6 (2), pp.129-162. Berghem, L.E.R. and Pettersson, L.G. )1973) The machanism of enzymatic cellulose degradation. Purification of a cellulolytic enzyme from Trichoderma viride active on highly ordered cellulose. Eur. J. Biochem. 37 :21 -30 . Birnboim, H.L and Doly, J. (1979) A rapid alkaline extraction procedure for screening of recombinant plasmid DNA. Nucl. Ac. Res. 7 :1513-1523. Bisaria, V.S. and Ghose, T.K. (1981) Biodegradation of cellulosic materials: substrates, microorganisms, enzymes and products. Enzyme Microb. Technol. 3 :90-104. Blake, M.S., Johnston, K.H., Russel-Jones, G.J. and Gotschlich, E.C. (1984) A rapid, sensitive method for detection of alkaline phosphatase-conjugate anti-antibody on western blots. Analytical Biochem. 136 : 175-179. Bousfield, I.J. and Callely, A.G. (eds) (1978) Coryneform bacteria. Academic Press. London. Boyer, H.W., Betlach, M., Bolivar, F., Rodriguez, R.L, Heyneker, H.L, Shine, J. and Goodman, H.M. (1977) The construction of molecular cloning 110 vehicles, In Beers, R.F. and Basset,-E.G. (eds.), Recombinant DNA molecules, 1j0_, Miles Symposium Series. Raven Press, New York, pp.9-19. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 7_£ : 248-254. Bronnenmeier, K. and Standenbauer, W.N. (1988) Purification and properties of an extracellular B-glucosidase from the celllolytic thermophile Clostridium stercorarium. Appl. Microbiol. Biotechnol. 2£ : 380-386. Chassy, B.M., Mercenier, A. and Flickinger, J. (1988) Transformation of bacteria by electroporation. Elsevier Publication, Cambridge, Tibtech 6 : 303-310. Cornish-Bowden, A. (1981) Fundamentals of enzyme kinetices. Butterworth & Co (Publishers) Ltd. UK. Coughlan, M.P. (1985) The properties of fungal and bacterial cellulases with comment on their production and application. Biotechnology and Genetic Engineering Reviews, vol 3, pp.53-109. Curry, C , Gilkes, N., O'Neill, G., Miller, R.C.Jr. and Skipper, N. (1988) Expression and secretion of Cellulomonas fimi exoglucanase in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 54 : 476-484. Darveau, R.P. and Hancock, R.E.W. (1983) Procedure for isolation of bacterial l ipopolysaccharide from both smooth and rough Pseudomonas aeruginosa and Salmonella typhumurium strains. J. Bact. 155 : 831-838. Day. A.G. and Withers, S.G. (1985) The purification and characterization of a B-glucosidase from Alcaligenes feacalis. Biochem. Cell. Biol. 64 : 914-922. Dayhoff, M.O., Dayhoff, R.E. and Hunt, L.T. (1976) In "Atlas of protein sequence and structure". Dayhoff, M.O., Ed., National Biomedical Research Fondation : Washington, DC, 5_ : 301-310. Din, N., Beck, C.F., Miller, R.C.Jr., Kilburn, D.G. and Warren, R.A.J. (1990) Expression of the Cellulomonas fimi cellulase genes cex and cenA from the divergent tet promoters of transposon Tn10. Arch. Microbiol. 153 : 129-133. Dinur, T., Osiecki, K.M., Legler, G., Gatt, S., Desnick, R.J. and Grabowski, G.A. (1986) Human acid B-glucosidase: Isolation and amino acid sequence of a peptide containing the catalytic site. Proc. Natl. Acad. Sc. USA. fi2 : 1660-1664. i n Ehrenberg, L , Fedorcsak, I. and Lolymosy, F. (1976) Diethylpyrocarbonate in nucleic acid research. J. Mol. Biol. 16_ : 189-262. Ericksson, K.E. and Wood, T.M. (1985) in "Biosynthesis and Biodegradation of Wood components". (T. Higuchi, ed) Academic Press, New York, p.469. Fan, L.T., Gharpuray, M.M. and Lee, Y.-H. (1987) Cellulose hydrolysis. Biotechnology Monographs, vol. 3, Favoloro, J. , Triesman, R. and Kamen, R. (1980) Transcription maps of polyoma virus-specific RNA: Analysis by two-dimentional nuclease S1 gel mapping. Methods Enzymol. £5 : 718-749. Gerhardt, P., Murray, R.G.E., Costilow, R.N., Nester, E.W., Wood, W.A., Krieg, N.R. and Phillips, G.B. (1981) Manual of methods for general bacteriology., A.S.M. Washington, D.C. 20006. Gilkes, N.R., Kilburn, D.G., Langsford, M.L, Miller, R.C.Jr., Wakarchuk, W.W., Warren, R.A.J., Whittle, D.J. and Wong, W.K.R. (1984) Isolation and characterization of Escherichia coli clones expressing cellulase genes from Cellulomonas fimi. J. Gen. Microbiol. 130 :1377-1384. Gilkes, N.R'!, Langsford, M.L, Kilburn, D.G., Miller, R.C.Jr. and Warren, R.A.J. (1984a) Mode of action and substrate specificities of cellulases from cloned bacterial genes. J. Biol. Chem. £52 : 10455-10459. Gilkes, N.R., Kilburn, D.G., Miller, R.C.Jr. and Warren, R.A.J. (1984b) A mutant of Escherichia coli that leaks cellulase activity encoded by cloned cellulase genes from Cellulomonas fimi. Bio/Technology 2 : 259-263. Gilkes, N.R., Warren, R.A.J., Miller, R.C.Jr. and Kilburn, D.G. (1988) Precise excision of the cellulose binding domains from two Cellulomonas fimi cellulases by a homologous protease and the effect on catalysis. J. Biol. Chem. 2£3_: 10401-10407. Gilkes, N.R., Kilburn, D.G., Miller, R.C.Jr. and Warren, R.A.J. (1989) Structural and functional analysis of a bacterial cellulase by proteolysis. J. Biol. Chem. 2§4 : 17802-17808. Gokhale, D.V., Soo-Han, E., Srinivan, V.R. and Dileep, N.D. (1984) Transfer of DNA coding for cellulases from Cellulomonas species to Bacillus subtilis by protoplast fusion. Biotechnology Letters £ : 627-632. Gokhale, D.V. and Deobagkar, D.N. (1989) Differential expression of xylanases and endoglucanases in the hybrid derived from intergeneric protoplast fusion between a Cellulomonas sp. and Bacillus subtilis. Appl. Environ. Microbiol. §5 : 2675-2680. Gong, C.-S. and Tsao, G.T. (1979) Cellulase and biosynthesis regulation. Annual Reports on Fermentation Processes. 3.: 111-140. 112 Greenberg, N.M., Warren, R.A.J., Kilburn, D.J. and Miller, R.C.Jr. (1987a) Regulation, initiation and termination of the cen A and cex transcripts of Cellulomonas fimi. J. Bact. 169 : 646-653. Greenberg, N.M., Warren, R.A.J., Kilburn, D.J. and Miller, R.C.Jr. (1987b) Regulation and initiation of cenB transcripts of Cellulomonas fimi. J . Bact. 1£9_ : 4674-4677. Greenberg, N.M. (1988) Cellulase gene transcription in Cellulomonas fimi and an Agrobacterium. Ph. D. Thesis., the University of British-Columbia. Greenwood, J.M., Gilkes, N.R., Kilburn, D.J., Miller, R.C.Jr. and Warren, R.A.J. (1989) Fusion to an endoglucanase allows alkaline phosphatase to bind to cellulose. FEBS letters 244 :127-131. Grepinet, O., Chebrou, M.-C. and Beguin, P. (1988) Nucleotide sequence and deletion analysis of the xylanase {xynZ) of Clostridium thermocellum. J. Bact. 17Q : 4582-4588. Groleau, D. and Forsberg, C.W. (1981) Cellulolytic activity of rumen bacterium Bacteroided succinogenes. Can. J. Microbiol. 27 :517-530. Gusakov, A.V., Sinitsyn, A.P., Goldstein, G.H. and Klyosov, A.A. (1984) Kinetics and mathematical model ot hydrolysis and transglycosylation catalysed by cellobiase. Enzyme Microbial Technology 6 : 275-282. Hall, J., Hazlewood, G.P., Huskisson, N.S., Durrant, A.J. and Gilbert, H.J. (1989) Conserved serine-rich sequences in xylanase and cellulase from Pseudomonas fluorescens subspecies cellulosa: internal signal sequence and unusual protein processing. Molec. Microbiol. 3_ 1211-1219. Han, Y.W. and Srinivasan, V.R. (1969) Purification and characterization of 6-glucosidase of Alcaligenes faecalis. J. Bact. 100 :1353-1363. Hsuanya, Y. and Laidler, K.J. (1984) Kinetics of the hydrolysis of cellobiose catalysed by B-glucosidase. Can. J. Biochem. Cell. Biol. 63 :167-175. Hwang, J.-T. and Suzuki, H. (1976) Intracellular distribution and some properties of B-glucosidases of a cellulolytic pseudomonad. Agr. Biol. Chem. 4J2 : 2169-2175. Jobe, A. and Bourgeois, S. (1972) Lac repressor-operator interaction. VI. The natural inducer of the operon. J. Mol. Biol. 6J£: 397-408. Johnson, J.A., Wong, W.K.R. and Beatty, J.T. (1986) Expression of cellulase genes in Rhodobacter capsulatus by use of plasmid expression vectors. J. Bact. 1 £ Z : 604-610. 113 Jovin, T., Chrambach, A. and Naughton, M.A. (1964) An apparatus for temperature-regulated polycarylamide gel electrophoresis. Anal. Biochem. 9_: 351-369. Katsumata, R., Ozaki, A., Oka, t. and Furuya, A. (1984) Protoplasts transformation of glutamate-producing bacteria with plasmid DNA. J. Bact. 159 : 306-311. Kim, H.K. and Pack, M.Y. (1989) Cloning and expression of Cellulomonas fimi B-glucosidase genes in Escherichia coli. Enz. Microb. Technol. 11 : 313-316. Komagata, K., Yamada, K. and Ogawa, H. (1969) Taxonomic studies on coryneform bacteria. I. Division of bacterial cells. J. Gen. Appl. Microbiol. 15: 243-259. Kraft, R., Tardiff, J., Kranter, K.S. and Leinwand, L.A. (1988) Using mini-prep plasmid DNA for sequencing double stranded templates with sequenasetm. Bio/Techniques £ : 544-547. Kricker, M. and Hall, B.G. (1984) Direct evolution of cellobiose utilization in Escherichia coli K12. Mol. Biol. Evol. 1:171 -182. Kristensen, T., Hartmut, V. and Wilhelm, A. (1987) A simple and rapid preparation of M13 sequencing templates for manual and automated dideoxy sequencing. Nucl. Ac. Res. 15 : 5507-5516. Kunkel, T.A., Roberts, J.D. and Zakour, R.A. (1987) Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154 : 367-382. Lacks, S.A. and Springhorn, S.S. (1980) Renaturation of enzymes after polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. J. Biol. Chem. 255 : 7467-7473. Langsford, M.L, Gilkes, N.R., Wakarchuk, W.W., Kilburn, D.G., Miller, R.C.Jr. and Warren, R.A.J. (1984) The cellulase system of Cellulomonas fimi. J. Gen. Microbiol. 12Q_: 1367-1376. Langsford, M.L, Gilkes, N.R., Singh, B., Moser, B., Miller, R.C.Jr., Warren, R.A.J, and Kilburn, D.G. (1987) Glycosylation of bacterial cellulases prevents proteolytic cleavage between functional domains. FEBS letters 225:163-167. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 221 • 680-685. Mandel, M. and Higa, A. (1970) Calcium dependent bacteriophage DNA infection. J. Mol. Biol. 52 :154-1 1 4 Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular cloning. Cold Spring Harbour Laboratory, Cold Spring Harbour, New York. Martin, J.F., Santamaria, R., Sandoval, H., Real, G.D., Mateos, L.M., Gil, J.A. and Aguilar, A. (1987) Cloning systems in amino acid-producing corynebacteria. Bio/Technology Review 5_ :137-146. Matthews, B.W.,Nicholson, H. and Becktel, W.J. (1987) Enhanced protein thermostability from site-directed mutations that decrease the entropy of unfolding. Proc. Natl. Acad. Sci. USA fi4 : 6663-6667. Maxam, A.M. and Gilbert, W. (1980) Sequencing end-labelled DNA with base-specific chemical cleavages. Methods Enzymol. 65.: 499-560. Messing, J. (1983) New M13 vectors for cloning. Methods Enzymol. 101 : 20-78. Miller, G.L. (1959) Use of dinitrosalicylic acid reagent for determination of reducing sugar. Analyt. Chem. 3J_: 426-428. Miwa, K., Matsui, K., Terabe, M., Ito, K., Ishida, M., Tagaki, H., Nakamori, S. and Sano, K. (1985) Construction of novel shuttle vectors and a cosmid vector for the glutamic acid-producing bacteria Brevibacterium lactofermentum and Corynebacterium glutamicum. Gene 3_9_: 281 -286. Mizusaqa, S., Nishimura, S. and Seela, F. (1986) Improvement of the dideoxy chain terminator method of DNA sequencing by use of deoxy-7-deaza guanosine triphosphate in place of dGTP. Nucleic Acid Research 1.4 : 1319-1324. Moldoveanu, N. and Kluepfel, D. (1983) Comparison of 8-glucosidase activities in different Streptomyces strains. Appl. Environ. Microbiol. 46 : 17-21. Moser, B. (1988) Molecular cloning, characterization and expression of the endoglucanase C gene of Cellulomonas fimi and properties of the native and recombinant gene products. Ph. D. Thesis. The University of British-columbia. Moser, B., Gilkes, N.R., Kilburn, D.J., Warren, R.A.J, and Miller, R.C.Jr. (1989) Purification and characterization of endoglucanase C of Cellulomonas fimi, cloning of the gene and analysis of in vivo transcripts of the gene. Appl. Environ. Microbiol. 5_5_: 2480-2487. Nakai, R., Horinouchi, S. arid Beppu, T. (1988) Cloning and nucleotide sequence of a cellulase gene, casA, from an alkalophilic Streptomyces strain. Gene £5 : 229-238. Nakamori, S., Ishida, M., Takagi, H., Ito, K., Miwa, K. and Sano, K. (1987) 115 Improved L-threonine production by the amplification of the gene encoding homoserine dehydrogenase in Brevibacterium lactofermentum. Agricultural and Biological Chemistry 5J_: 87-91. Nakamura, K., Misawa, N. and Kitamura, K. (1986) Cellulase genes of Cellulomonas uda CB4. I. Cloning and expression of 8-glucosidase genes in Escherichia coli. J. Biotechnol. 2 : 239-246. Nisizawa, T., Suzuki, H., Nakayama, M. and Nisizawa, K. (1971) Inductive formation of cellulase by sophorose in Trichoderma viride. J. Biochem. 7_0_ : 375-385. Nossal, N.G. and Heppel, L.A. (1966) The release of enzymes by osmotic shock from Escherichia coli in exponential phase. J. Biol. Chem. 241 : 3055-3062. O'Callaghan, C.H., Morris, A., Kirby, S.M. and Shingler, A.H. (1972d) Novel method for detection of B-lactamase by using a chromogenic cephalosporin substrate. Antimicrob. Ag. Chemother., 1: 283-288. Oka, A., Sugisaki, H. and Takanami, M. (1981) Nucleotide sequence of the kanamycin resistance transposon Tn903. J. Mol. Biol. 147 : 217-226. O'Neil, G.P. (1986) Structure of the gene encoding the exoglucanase of Cellulomonas fimi and its expression and secretion in Escherichia coli. Ph. D. Thesis, the University of British-Columbia. O'Neil, G.P., Goh, S.H., Warren, R.A.J., Kilburn, D.J. and Miller, R.C.Jr. (1986) Structure of the gene encoding the exoglucanase of Cellulomonas fimi. Gene 44 : 325-330. O'Neil, G.P., Kilburn, D.J., Warren, R.A.J, and Miller, R:C.Jr. (1986a) Overproduction from a cellulase gene with a high guanosine-plus-cytosine content in Escherichia coli. Appl. Environ. Microbiol. 52 : 737-7'43. O'Neil, G.P., Warren, R.A.J., Kilburn, D.J. and Miller, R.C.Jr. (1986b) Secretion of Cellulomonas fimi exoglucanase by Escherichia coli. Gene 44 : 331-336. Ong, E., Gilkes, N.R., Kilburn, D.J., Miller, R.C.Jr. and Warren, R.A.J. (1989) Enzyme immobilization using the cellulose-binding domain of a Cellulomonas fimi exoglucanase. Bio/Technology 7 : 604-607. Osborn, M.J., Gander, J.E., Parisi, E. and Carson, J. (1972) Mechanism of assembly of the outer membrane of Salmonella typhimurium. Isolation and characterization of cytoplasmic and outer membranes. J. Biol. Chem. 247 : 3962-3972. Osborn, M.J. and Munson, R. (1974) Separation of the inner (cytoplasmic) and 116 outer membranes of Gram-negative bacteria. Methods Enzymol. 3J_: 642-667. Owolabi, J.B. (1988) Characterization and expression of Cellulomonas fimi endoglucanase B gene and properties of the gene product from Escherichia coli. Ph. D. Thesis. The University of British-Columbia. Owolabi, J.B., Beguin, P., Kilburn, D.J., Miller, R.C.Jr. and Warren, R.A.J. (1988) Expression in Escherichia coli of the Cellulomonas fimi structural gene for endoglucanase B. Appl. Environ. Microbiol. 54 : 518-523. Pack, Y.-H., Hori, H., Suzuki, K.-l., Osawa, S. and komagata, K. (1987) Phylogenetic analysis of the coryneform bacteria by 5S rRNA sequences. J. Bact. 1801-1806. Paradis, F.W., Warren, R.A.J., Kilburn, D.J. and Miller, R.C.Jr. (1987) The expression of Cellulomonas fimi cellulase genes in Brevibacterium lactofermentum. Gene 6J_: 199-206. Pentilla, M.E., Nevalainen, K.M.H., Raynal, A. and Knowles, J.C.K. (1984) Cloning of Aspergillus niger genes in yeast. Expression of the gene coding Aspergillus B-glucosidase. Mol. Gen. Genet. 194 : 494-499. Raynald, A. and Guerineau, M. (1984) Cloning and expression of the structural gene for B-glucosidase of Kluyveromyces fragilis i n Escherichia coli and Saccharomyces cerevisiae. Mol. Gen. Genet. 195 : 108-115. Saddler, J.N. and Khan. A.W. (1981) Cellulolytic enzyme system of Acetivibrio cellulolyticus. Can. J. Microbiol. 27 : 288-294. Sanger, F., Nicklen, S. and Coulson, A.R. (1977) DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA JA : 5463-5467. Sano, K., Amemura, A. and Harada, T. (1975) Purification and properties of a B-1,6-glucosidase from Flavobacterium. Biochemica et biophysica acta. 377 : 410-420. Santamaria, R., Gil, J.A., Mesas, J.M. and Martin, J. (1984) Characterization of an endogenous plasmid and development of cloning vectors and a transformation system in Brevibacterium lactofermentum. J. Gen. Microbiol. 1^ 0 : 2237-2246. Santamaria, R.I., Gil, J.A. and Martin, J.F. (1985) High-frequency transformation of Brevibacterium lactofermentum protoplasts by plasmid DNA. J. Bact. 1£2 : 463-467. Sato, M. and Takahashi, H. (1967) Fermentation of C'4-labelled cellobiose by Cellulomonas fimi. Agricult. Biologic. Chem. 31: 470-474. 117 Schaefler, S. (1967) Inducible system for the utilization of B-glucosides in Escherichia coli. J. Bact. 22 : 254-263. Schimz, K.-L, Broil, B. and John, B. (1983) Cellobiose phosphorylase (EC 2.4.1.20) of Cellulomonas: occurence, induction, and its role in cellobiose metabolism. Arch. Microbiol. 135 : 241 -249. Schmid, G. and Wandrey, C. (1989) Characterization of a cellodextrin glucohydrolase with soluble oligomeric substrates: Experimental results and modeling of concentration-time-course data. Biotech. Bioengineer. 22 : 1445-1460. Shaw, P-.C. and Hartley, B.S. (1988) A host-vector system for an Arthrobacter species . J. Gen. Microbiol. 124 : 903-911. Shewale, J.G. and Sadana, J.C. (1981) Purification, characterization and properties of 6-glucosidase enzymes from Sclerotium solfsii. Archs Biochem. Biophys. 202 :185-196. Shewale, J.G. (1982) B-glucosidase : its role in cellulase synthesis and hydrolysis of cellulose. Can. J. Biochem. 14 : 435-443. Shinnick, T.M. (1987) The 65-kilodalton antigen of Mycobacterium tuberculosis. J. Bact. 1S2 :1080-1088. Singer, M.E. (1986) Plasmid transformation in Arthrobacter sp. H13A. In: Abstracts of the A.S.M. Annual Meeting, March 23-28, Washington, D.C. Sinitsyn, A.P. and Klesov, A.A. (1981) Comparison of roles of exo-1,4-8-glucosidase and cellobiase in enzymatic hydrolysis of celluloase. Biochem. 4£ : 165-174. Skipper, N., Sutherland, M., Davies, R.W., Kilburn, D.J., Miller, R.C.Jr., Warren, R.A.J, and Wong, R. (1985) Secretion of a bacterial cellulase by yeast. Science 22Q : 958-960. Smith, M.D., Flickinger, J.L, Lineberger, D.W. and Schmidt, B. (1986) Protoplast transformation in coryneform bacteria and introduction of an alpha-amylase gene from Bacillus amuloliquefaciens into Brevibacterium lactofermentum. Appl. Environ. Microbiol. 5J_: 634-639. Southern, E.M. (1975) Detection of specific sequences amoung DNA fragments separated by gel electrophoresis. J. Mol. Biol. 28 : 503-517. Sprey, B. and Lambert, C. (1983) Titration curves of cellulases from Trichoderma reesei : Demonstration of a cellulase-xylanase-B-glucosidase-containing complex. FEMS Microbiol. Letters 12 : 217-222. Stackebrandt, E. and Kandler, O. (1979) Taxonomy of the genus 118 Cellulomonas, based on phenotypic characters and DNA-DNA homology and proposal of seven neotype strains. Int. J. Syst. Bacteriol. 29 : 263-282. Stuber, D. and Bujard, H. (1981) Organization of transcriptional signals in plasmids pBR322 and pACYC184. Proc. Natl. Acad. Sci. USA 7_8_: 167-171. Stoppock, W., Rapp.P. and Wagner, F. (1982) Formation, localization and regulation of endo-1,4-8-glucanases and 8-glucosidases from Cellulomonas uda. Appl. Environ. Microbiol. 44 : 43-53. Stewart, B.J. and Leatherwood, J.M. (1976) Derepressed synthesis of cellulase by Cellulomonas. J. Bact. 122 : 609-615. Teather, R.M. and Wood, P.J. (1982) Use of Congo red-polysaccharide interactions in enumeration and characterization of cellulolytic bacteria from bovine rumen. Appl. Environ. Microbiol. 42 : 777-780. Towbin, H., Staehelin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76 : 4350-4354. Tsung, K., Inouye, S. and Inouye, M. (1989) Factors affecting the efficiency of protein synthesis in Escherichia coli. J. Biol. Chem. 2M : 4428-4433. Umile, C. and Kubicek, CP. (1986) A constitutive plasma-membrane bound 6-glucosidase in Trichoderma reesei. FEMS Microbiol. Letters 34 : 291-295. Vaheri, M., Leisola, M. and Kauppinen, V. (1979) Transglycosylation products of cellulase system of Trichoderma reesie. Bio/Technology Lett 1: 41-46. Vansatha, N., Thompson, L.D., Rhodes, C, Banner, C, Nagle, C, Nagle, J. and Filpula, D. (1984) Genes for alkaline protease and neutral protease from Bacillus amyloliquefaciens contain a large open reading frame between the regions coding for signal sequence and mature protein. J. Bact. 159.: 811-819. Vieira, J. (1985) United States Biochemical Company technical literature accompaning the vectors pTZ 18/19. Voller, A., Bidwell, D. and Barlett, A. (1976) Microplate immunoassays for the immunodiagnosis of virus infections, pp.506-512. In: N. Rose and H. Freedman (eds.), Manual of clinical immunology, first ed. A.S.M., Washington, D.C. Wakarchuk, W.W., Kilburn, D.G., Miller, R.C. and Warren, R.A.J. (1984) The preliminary characterization of the B-glucosidases of Cellulomonas fimi. J. Gen. Microbiol. 120. : 1385-1389. 119 Wakarchuk, W.W. (1987) The molecular cloning and characterization of a 6-glucosidase gene from an Agrobacterium. Ph. D. Thesis. The University of British-Columbia. Warren, R.A.J., Beck, C.F., Gilkes, N.R., Kilburn, D.G., Langsford, M.L, Miller, R.C. Jr., O'Neill, G.P., Scheufeus, M. and Wong, W.K.R. (1986) Sequence conservation and region shuffling in an endoglucanase and an exoglucanase from Cellulomonas fimi. Proteins 1 : 335-341. Whittle, D.J., Kilburn, D.G., Warren, R.A.J, and Miller, R.C.Jr. (1982) Molecular cloning of a Cellulomonas fimi cellulase gene in Escherichia coli. Gene 17 : 139-145. Woodward, J. and Wiseman, A. (1982) Fungal and other B-D-glucosidases-their properties and applications. Enzyme Microb. Technol., 4 : 73-79. Wong, R.W.K. (1986) The cloning and characterization of an endoglucanase gene of Cellulomonas fimi. Ph. D. Thesis, The University of British-Columbia. Wong, R.W.K., Gerhard, B., Guo, Z-.M., Kilburn, D.G., Warren, R.A.J, and Miller, R.C.Jr. (1986) Characterization and structure of an endoglucanase gene cenA of Cellulomonas fimi. Gene 44 : 315-324. Wong, R.W.K., Curry, C, Davies, R.W. and Kilburn, D.G. (1988) Wood hydrolysis by Cellulomonas fimi endoglucanase and exoglucanase coexpressed as secreted enzymes in Saccharomyces cerevisiae. Bio/Technology §.: 713-719. Worthington, CC. (1988) Worthington Enzyme Manual, Worthington Biochemical Corporation, Freehold, New Jersey, USA 07728. Yamada, K. and Komagata, K. (1972) Taxonomic studies on coryneform bacteria. V. Classification of coryneform bacteria. J. Gen. Appl. Microbiol. 12:417-431. Yamada, K., Kinoshita, S., Tsunoda, T. and Aida, K. (1972) The microbial production of amino acids. Wiley, New York. Yannish-Perron, C, Vieira, J. and Messing, J. (1985) Improved M13 phage cloning vectors and host strains: Nucleotide sequences of M13 mp and pUC vectors. Gene 22 :103-119. Yeh, P., Oreglia, J., Prevost, F. and sicard, A.M. (1986) A shuttle vector system for Brevibacterium lactofermentum. Gene 47 : 301-306. Yoshihama, M., Higashiro, K., Rao, E.A., Akedo, M., Shanabruch, W.G., Follettie, M.T., Walker, G.C. and Sinskey, A.J. (1985) Cloning vector system for Corynebacterium glutamicum . J. Bact. 162 : 591-597. 120 Appendix 1. The double-reciprocal plots of reciprocal rates (Wo) against reciprocal substrate concentration (1/[S]) are given for the hydrolysis of cellobiose (A) and higher cellodextrins (B, C, D) by the purified CbgA as described in section 2.7.2. The substrates are shown in shadowed legends. The intercept on the ordinate (1/Vo) axis corresponds to 1/Vmax and the intercept on the abcissa (1/[S]) axis corresponds to -1/Km. The inhibition of CbgA activity by glucose on the hydrolysis of pNPG at various substrate concentrations is shown by plotting the reciprocal rate (1A/o) against glucose inhibitor concentration ([i]) (E). The crossing point of all three lines reflected on the abcissa is equal to -Ki. The hydrolysis of p-nitrophenyl derivatives (pNPG and pNPC) by CbgA is shown by plotting the continuous increase in optical density (410 nm) in function of time (F). The substrate and various concentrations are indicated in the legend. s.o 4.0 ol/min/mL)) 3.0 - —Q— cellobiose 1 • E .A 2.0 -1 • 1.0 • f -1.0 0.0 1.0 2.0 M[S] (1/mM) E 2.0 r 1/[S] (1/mM) 1 2 1 [I] (mM) 122 Time (min) 123 Appendix 2. Detection of specific transcript by Northern blot analysis. RNA isolated from C. fimi CMC grown cells was denatured with formaldehyde, fractionated on a formaldehyde gel containing 1% (w/v) agarose and electrotransfered to a Biotrans membrane (Pall, Inc.). Hybridization was done with 32P-labelled Xho\-Stu\ probe (9.2 X 104 cpm) to : lane 1, 28.6 ug of RNA; lane 2, 14.3 ug of RNA. Mr markers are shown (nt). The membrane was exposed for 3 days at -70°C with intensifying screens. The location of the detected transcripts is shown by arrows. 0.24 124 Appendix 3. S1 nuclease protection analysis of cbgA transcript. After hybridisation of the Stu\-Xho\ cbgA specific labeled DNA probe with RNA from C. fimi CMC-grown cells, the DNA-RNA hybrid was digested with various amount of S1 nuclease and analysed in a 8% polyacrylamide-7M urea containing gel alongside probe sequenced by the base-specific chemical cleavage method. The chemical sequencing ladders of the probe is shown. Protection of the probe by 30 u.g of RNA digested with ; lane 1, 50 U; lane 2, 150 U; lane 3, 300 U of S1 nuclease. Lane 4, 30 u.g of yeast tRNA digested with 300 U of S1 nuclease, lane 5, probe alone. Qfl C C 1 2 34 5 

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.831.1-0098751/manifest

Comment

Related Items