UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structure-function relationships of endoglucanase C (CenC) from Cellulomonas fimi Coutinho , John B. 1992

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

Item Metadata


831-ubc_1992_fall_coutinho_john_b.pdf [ 3.1MB ]
JSON: 831-1.0086543.json
JSON-LD: 831-1.0086543-ld.json
RDF/XML (Pretty): 831-1.0086543-rdf.xml
RDF/JSON: 831-1.0086543-rdf.json
Turtle: 831-1.0086543-turtle.txt
N-Triples: 831-1.0086543-rdf-ntriples.txt
Original Record: 831-1.0086543-source.json
Full Text

Full Text

STRUCTURE-FUNCTION RELATIONSHIPS OFENDOGLUCANASE C (CENC) FROM CELLULOMONAS FIMIbyJOHN B. COUTINHOB.Sc., Queen’s University, 1982M.Sc., Queen’s University, 1986A ThESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of MicrobiologyWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1992© John B. Coutinho, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.______________Department ofThe University of British ColumbiaVancouver, CanadaDate SF4’,— ‘VJDE-6 (2/88)UABSTRACTThe cenC gene of Ceilulomonasfimi, encoding endoglucanase CenC had an open readingframe of 1101 codons closely followed by a 9 bp inverted repeat. The amino acid sequence ofmature CenC, which was 1069 amino acids long, is very unusual in that it had a 150 amino acid-long tandem repeat (N1N2) at the N-terminus and an unrelated 100 amino acid-long tandem repeat(C1C2) at the C-terminus. Similarity of the central domain of CenC to the catalytic domains ofother endoglucanases placed CenC in subfamily El of the 3-l,4-g1ycanases. CenC could beaffinity purified on cellulose or Sephadex. The catalytic properties of recombinant CenC from E.coil, for the substrate carboxymethylcellulose were indistinguishable from those of native CenCfrom C. fimi.In order to determine which of the repeats N1N2 or C1C2 bind to cellulose or toSephadex, both repeats were cloned separately. N1N2 mediated binding to both cellulose andSephadex. Nl or N2 alone did not bind Sephadex but did bind cellulose. The C-terminal repeats,alone or in combination, did not mediate binding to cellulose or to Sephadex. N1N2 bound toregenerated cellulose (phosphoric acid swollen cellulose) but had negligible affinity for bacterialmicrocrystalline cellulose.To show that the N-terminal repeats could be used as an affinity tag for a polypeptide otherthan CenC, N1N2 was fused to the catalytic domain of CenA (another endoglucanase from C.fimi). The resulting fusion polypeptide C’ ‘A could be affinity purified on cellulose or Sephadexand retained catalytic activity. C’ ‘A was also used to study the influence of the binding domain onthe hydrolysis of cellulosic substrates by comparison to CenA. C’ ‘A had higher activity on theamorphous substrate cellulose azure when compared to CenA. C’ ‘A and CenA had similaractivities on regenerated cellulose with the most striking difference being the poor activity of C’ ‘Aon crystalline cellulose. The ability (or lack thereof) of the binding domain to adsorb to crystallinecellulose correlated well with the ability of the enzyme to hydrolyze the substrate.illTABLE OF CONTENTSPageABSTRACT iiTABLE OF CONTENTS iiiLIST OF TABLES viLIST OF FIGURES viiLIST OF ABBREVIATIONS ixACKNOWLEDGEMENTS xi1. Introduction 11.1. Structure of cellulose I1.2. Enzymatic hydrolysis of cellulose 21.3. Measurement of cellulase activities 41.4. Motifs of cellulolytic enzymes 51.5. Cellulases from Cellulomonas fin2i 81.6. Objectives of this thesis 102. Materials and Methods 132. 1. Bacterial strains, plasmids and phage vectors 132.2. Media and growth conditions 132.3. Enzymes and reagents 152.4. DNA preparation and sequence determination 162.5. Cloning the 3’ end of the cenC gene 172.6. Construction of deletion mutants of cenC 172.6.1 N-terminal deletions of CenC 172.6.2 C-terminal deletions of CenC 18iv2.7. Cloning of the N-tenninal repeats N1N2 and Ni and theC-terminal repeats C1C2 and Ci 182.8. Cloning of N2 using Nuclease Ba131 192.9. Deletion analysis of Ni 192.10. Substituting N1N2 for the CBD of endogiucanase CenA 192.11. Purification of CenC, CenCA.1 and CenCA2 202.12. Expression of C1C2 and Cl 212.13. Purification of N1N2, Ni and N2 212.14. Purification of C’ ‘A 222.15. Measurements of CenC activity 222.16. Measurements of C’ ‘A activity 232.17. Oligosaccharide products of CenC 232.18. Determination of protein concentration 232.19. Amino-acid sequence determination 242.20. Protein electrophoresis 242.21. Immunoblotting 242.22. Adsorption of N1N2 or Ni to Avicel and Sephadex 252.23. Adsorption of N1N2, Ni or CBDCexto BMCC and PASC 252.24. Adsorption of N1N2, Ni and N2 to PASC 262.25. Rapid small scale screening procedure for functionalbinding activity 262.26. Adsorption of CenA and C’ ‘A to BMCC and PASC 262.27. Hydrolysis of CenA and C’ ‘A by C.Jirni protease 272.28. Amino-acid sequence comparisons 27V3. Results3. 1. Nucleotide and amino acid sequence of cenC 283.2. Expression of CenC in E. coli 423.3. 5’ and 3’ deletions of cenC 513.4. Kinetic parameters for native and recombinant CenC 593.5. Expression of fragments of the cenC gene 653.6. Binding of the N- and C-terminal repeats of CenCto Avicel and Sephadex 723.7. Sequence relatedness between N1N2, Niand other CBDs 753.8. Expression of N2 803.9. Adsorption of N1N2, Ni and N2 to (PASC) 803.10. Consequences of substituting N1N2 for theCBD of endoglucanase CenA 874. Discussion 965. Summary 1066. References 107viLIST OF TABLESTable Page1. Plasmids encoding cenC and its derivatives 142. Kinetic parameters for CenC 603. Partition coefficients of N1N2, Ni and CBDCex forPASC and BMCC 744. Families of cellulose binding domains 765. Adsorption of N1N2, Ni and N2 to PASC 856. Adsorption of CenA and C’ ‘A to BMCCand PASC 917. Activity of CenA and C’ ‘A oncellulosic substrates 92vuLIST OF FIGURESFigure Page1. The structure of cellulose 22. Organization of catalytic and non-catalytic domains incellulases from C. find 103. Nuclease Ba131 deletions of eeoC 294. Mini-preparation of plasmid DNA from clones obtainedafter Nuclease Ba131 deletions 305. Identification of a DNA fragment containing the 3’ endof cenC 316. Identification of clones containing the3’ end of cenC 327. Plasmids used in cloning the entire ceiiC gene 338. Nucleotide sequence of eeoC and its flanking regionsand the deduced amino acid sequence of CenC 359. Repeated amino acid sequences in CenC 3810. Sequence relatedness in the catalytic domains off3-1,4-glycanases of subfamily El 3911. Consensus sequences around the DAGD and H-R peptidesin subfamilies El and E2 4312. Family E of 13-1,4-glycosidases 4513. Expression of CenC in E. coil 4714. Expression of CenC in E. coil under differentphysiological conditions 5015. N-terminal deletions of CenC 5216. C-terminal deletions of CenC 5317. Purification of CenC, CenC CAl and CenC CA2by affinity chromatography 54v1u18. Purification of CenC by anion exchange chromatography 5519. Purification of CenC by gel-filtration 5720. SDS-PAGE analysis of purified CenC and its two C-terminaldeletions 5821. CenC’s oligosaccharide cleavage pattern 6122. Temperature and pH optima for CenC 6323. Plasmids encoding polypeptides for the N-andC-terminal repeats 6624. Western blot of polypeptides N1N2 and Ni synthesizedin E. co/i 6725. CiC2 and Cl production in E. co/i 6826. Purification of N1N2 and NI by affinity chromatography 6927. Adsorption isotherms for N1N2, Ni and CBDCex toPASC or BMCC 7328. Dendogram of Family II CBDs 7829. Amino acid sequence similarities between Ni, N2and other CBDs from C. Jinii 7930. Plasmids encoding polypeptides N1N2, Ni and N2 8131. Purification of N2, Ni and Ni N2 by affinity chromatography 8232. SDS-PAGE analysis of the adsorption of N1N2, Niand N2 to regenerated cellulose (PASC) 8333. Deletion mutants of Ni 8634. Substitution of N1N2 for the CBD of CenA 8835. Adsorption of CenA and C’ ‘A to BMCC and PASC 8936. Organization of CenA and C’ ‘A 9437. Hydrolysis of CenA and C’ ‘A with C. JImi protease 9538. Sequence similarities between lectins and C1C2 9839. Potential applications of CBDs with different specificities 102xLIST OF ABBREVIATIONSInterposon omega[BI Bound polypeptideBMCC Bacterial microcrystalline celluloseC1C2 C-terminal repeats of CenCC’ ‘A CenC CBD fused to CenA catalytic domainCBD Cellulose binding domainCeID Endoglucanase D from Clostridium thermocellumCenA Endoglucanase A from CellulomonasfimiCenC Endoglucanase C from Cellulomonasfin1iCenC Proteolytic derivative of CenCCMC CarboxymethylcelluloseCNPC ChioronitrophenolcellobioseDNS Dinitrosalicyclic acid[F] Free polypeptideG3 CellotetraoseHBAH Hydrobenzoic acid hydrazideIPTG Isopropyl-13-D-thiogalactosideKcat Enzyme turnover numberkDa KilodaltonsKm Michaelis-Menten constantMUC Methylumbelliferyl cellobiosideN1N2 N-terminal repeats of CenCp30 CenA proteolytic fragmentPASC Phosphoric acid swollen cellulose! regenerated cellulosepG2 2-chloro-4-nitrophenol cellobiosep1 Isoelectric pointPMSF Phenylmethylsulfonyl fluoridepNPC p-nitrophenyl-13-cellobiosideSDS-PAGE Sodium dodecyl sulfate-polyaciylarnide gel electrophoresisVmax Maximum rate of enzyme reactionxxiACKNOWLEDGEMENTSI would like to thank three very special persons, my supervisors Dr. D.G. Kilburn, Dr.R.C.Miller Jr. and Dr. R.A.J. Warren for giving me the opportunity to work on cellulases. Mythanks to my committee members Dr. Speert and Dr. Hancock for their guidance and their usefulcriticisms and recommendations not only with the project but also in helping me improve mypublic speaking. My friends who made my stay in Vancouver a very pleasant one, they includeRudy Vlyasevi, Nicky Ahmed, the rookies Dave Nordquist and Howard Damude, the scottish ladpit yer specs on Alasdair MacLeod, the veterans Francois Paradis, Roger Graham, Celia Ramirez,Zahra Assouline and the classmates that I started with Neena Din and Jeffrey Greenwood. Thanksto Norm Greenberg who helped me learn molecular biology techniques and to all persons whocontributed in a very big way to this thesis including Bernard Moser, Shen Hua, Emily Kwan,Tara Young, Andreas Meinke, Lando Robillo, Gary Lesnicki, Helen Smith, Donald Trimbur andNeil Gilkes. The persons who made the biggest contributions need a very special thank you andthey are Edgar Ong, Peter Tomme and Pat Miller. My thanks also to a very special and caringfriend Brigitte Massot who has taught me many things. Finally my very special appreciation formy mother and father and for my brother and sister in law, for their love and encouragement,without whom this surely would not have come to pass.This work was supported by grants fromthe Natural Sciences and Engineering Research Council of Canada.11. Introduction1.1. Structure of celluloseCellulose is a homo-polymer of D-glucose. The linear, unbranched chains of glucoseresidues are linked to each other by f3 -1,4- glucosidic bonds. The basic recurringdisaccharide is cellobiose. The f3- linkage (Figure la) in cellulose assumes an extendedconformation with parallel cellulose chains being held together by cross-links of hydrogenbonds between the microfibrils (Figure lb). The intramolecular bonds thus help tomaintain the rigidity of the cellulose chain.There are two types of cellulose, cellulose I and cellulose II (Sharrock, 1988).Cellulose I, also commonly referred to as ‘native’ celluloses is the form producednaturally by plants, bacteria and algae and is crystalline in nature. Until recently there wasgeneral agreement on the structure of native cellulose, however data from 13C-NMRspectroscopy (Atalla and van der Hart, 1984) support the existence of two distinctcrystalline forms of native cellulose, cellulose Io and 1f3 based on differences in resonancepatterns. Cellulose lix is predominant in bacterial (Acetobacter) and algal (Valonia)celluloses whereas form I3 is predominant in celluloses from higher plants. Cellulose IIon the other hand is a relatively non-crystalline form of cellulose produced by treatment ofnative cellulose with sodium hydroxide.Within cellulose fibres there are areas of complete order, i.e. crystalline areas, andalso less ordered or amorphous regions. The degree of crystallinity within fibres varieswith the source of the cellulose and the treatment to which it has been subjected. In thenative state, cellulose fibres are associated with other polymers such as hemicellulose,pectin and lignin. The ciystallinity of the native material and its association with lignin arethe major factors that inhibit the enzymatic hydrolysis of cellulose (Sharon, 1975;Lehninger, 1982; Coughlan, 1985).2—oI I(b)Figure 1. The structure of cellulose a) Cellulose chain; the Dglucose units are in 13-1,4-linkage. b) Schematic drawing showing how parallel cellulose chains are held together bya crosslinking of hydrogen bonds (Lehninger, 1982).1.2. Enzymatic hydrolysis of celluloseCellulose is a renewable energy resource with approximately 4 x 1010 tonssynthesized annually by photosynthesis. By catalyzing the decay of forest and agriculturalwaste, the cellulases in combination With hemicellulases and ligninases recycle nutrientswhich would otherwise remain as inert waste for a considerably longer period of time.Equally as important, is the ability to control the activities of cellulolytic organisms in thetextile, paper and lumber industries. In the lumber indusy alone, millions of dollars arespent annually on preservatives to protect lumber (Coughlan, 1985). Cellulosic materialscan be pretreated using physicochemical processes to enhance their enzymatic hydrolysis(Millet eta!., 1976; Brownell and Saddler, 1987). Strains yielding high levels ofcellulases (Montenecourt, 1983) as well as mutant strains that are resistant to end-productinhibition have been isolated (Stewart and Leatherwood, 1976; Choudhury eta!., 1980).(a3Cellulose can be hydrolyzed and utilized by many bacteria and fungi: Aerobicbacteria such as Cellulonionas species, gliding bacteria (Cytophaga aid Sporocytophaga)and Streptomycetes; anaerobes such as rumen organisms Bacteroides succinogenes,Butyrvibrio fibrisolvens) and Clostridium thermocellurn; and various fungi of whichTrichoderma reesei is most prominent (Gottschalk, 1986).Cellulolytic enzymes have been traditionally divided into three classes,endoglucanases (E.C.3 .2.1.4), exoglucanases or cellobiohydrolases (E.C.3.2. 1.91) and13-glucosidases (E.C. Cellulases all cleave the same 13-1,4-glycosidic bond, butthere must be variation in the microenvironment of these bonds in natural substrates sincecellulolytic enzymes are generally induced as multienzyme systems with more than oneenyzme representing each class and the different classes of enzymes working in synergyfor cellulose hydrolysis (Coughlan and Ljungdahl, 1988; Béguin et al., 1990; Gilkes etat., 1991a). It is not clear what contributes to the differences in the microenvironment ofthe bond, but accessibility to the substrate, preferences for certain sizes of cellodextrinsand the inhibition of a particular enzyme by a certain type of product may play a role.Synergism between endo- arid exo-glucanases has been shown for a number of enzymes(Wood, 1975; Eriksson, 1975; Ryu eta!., 1984; Tomme eta!., 1990). The snynergisticaction is most marked on highly crystalline substrates, low with amorphous cellulose andnegligible with soluble derivatives (Wood and McCrae, 1979). The first model for thebiological degradation of cellulose was put forward by Reese (Reese et al., 1950) inwhich he proposed a Cl component necessary for the degradation of crystalline cellulose.This contribution stimulated research on the biochemical basis for cellulose hydrolysis andthe discovery of different types of cellulases (Nunirni etal., 1981; Klesov eta!., 1983)helped formulate the basis for present day models for the enzymatic hydrolysis ofcellulose (Béguin ci’ al.,1987; Enari and Niku-Paavola, 1987; Stahlberg eta!., 1991). Thegenerally accepted model that has emerged involves the internal splitting of the cellulosemolecule in the amorphous region by endoglucanases. The non-reducing ends generated4by this reaction become the substrate for exoglucanases which processively degradecellulose molecules in stepwise fashion from a non-reducing end liberating cellobiose as aproduct. Cellobiose and other low molecular weight cellodextrins are then split by 13-glucosidases to generate glucose thus preventing the buildup of cellobiose which inhibitsexoglucanase and endoglucanase activity (Beguin et at., 1987).1.3. Measurement of cellulase activitiesSeveral methods are available for the detection and measurement of the activities ofcomponents of the cellulase system (Mullings, 1985; Sharrock, 1988). A simplequalitative assay relies on the staining of carboxymethylcellulose in solid growth media orpolyacrylamide gels (Béguin, 1983) with Congo red. Hydrolysis of this substratefollowed by Congo red staining provides a zone of clearing (Teather and Wood, 1982).Reducing sugars may be determined by the dinitrosalicylic acid assay (Miller, 1959) or theNelson-Somogyi assay (Nelson, 1952). The increase in fluidity accompanyinghydrolysis of carboxymethylcellulose may be assayed viscometrically (Almin et a!.,1975). A plot of increase in relative fluidity versus time indicates the mode of attack onthe substrate and helps to differentiate endoglucanases from exoglucanases. Plots ofincrease in relative fluidity versus reducing equivalents have been used to compareindividual endoglucanases (Gilkes ci’ al., 1984). Exoglucanase activity has beendetermined by measuring the release of p-nitrophenol from p-nitrophenyl-13-cellobioside(pNPC) (Deshpande eta!., 1984) or release of methylumbelliferone frommethylumbelliferyl cellobioside (MUC) (van Tilbeurgh ci’ a!., 1982). 13-glucosidaseactivity is determined by measuring the release of p-nitrophenol from p-nitrophenyl-13-glucoside or by the release of glucose from cellobiose (Mullings, 1985).51.4. Motifs of cellulolytic enzymesThe cloning of cellulase genes, the determination of their nucleotide sequences, andthe analysis of the amino acid sequences predicted from the nucleotide sequences have ledto a greatly increased understanding of these enzymes. The amino acid sequences of morethan sixty 13- 1,4-glycanases have been deduced from the nucleotide sequences of theirgenes. Characterization of the enzymes and/or analysis of their amino acid sequences hasshown that many of them comprise two or more functional domains (Knowles et a!.,1987; Aubert eta!., 1988; Béguin, 1990; Gilkes eta!., 1991a). In some, if not all, suchenzymes, the domains function independently, and retain their functions when separatedby proteolysis (Calza eta!., 1985; Van Tilbeurgh eta!., 1986; Langsford eta!., 1987;Tomme eta!., 1988; Gilkes eta!., 1988; Ghangas and Wilson, 1988; Stahlberg eta!.1988; McGavin and Forsberg, 1989). A frequent arrangement is a catalytic domain linkedto a cellulose-binding domain (CBD) by a linker sequence rich in proline and/orhydroxyamino acids (Gilbert et al., 1990; Béguin, 1990; Gilkes eta!., 1991a).Catalytic domains: A given microorganism will produce a number of cellulaseswhich generally differ in overall amino acid sequence but which may share shortconserved sequences. Cellulases can be grouped into families of related enzymes on thebasis of amino acid sequence relatedness in their catalytic domains (Henrissat et a!.,1989). Sequence identity in the catalytic domains of cellulases and xylanases has beenreviewed recently and the known sequences can be grouped into nine families A-I whichare quite distinct from one another (Henrissat et a!., 1989; Gilkes et a!., 1991a). Althoughcellulases cleave the same 13-1 ,4-glycosidic bond, it is postulated that the differences lie inthe microenvironment of the cleavage site.Cellulose-binding domains (CBDs): Many cellulases bind to cellulose but themechanism and significance of this interaction is not clear. Reese’s group was the first topropose a non-enzymatic factor termed Cl that helped with the hydrolysis of crystallinecellulose (Reese ci’ a!., 1950). Klesov and his colleagues working with crude enzyme6fractions drew a correlation between the binding strength of an enzyme and the ability todegrade crystalline cellulose (Rabinovitch et a!., 1982; Klesov eta!., 1983; Chernoglazovet a!., 1983), however, the existence of an essential non-enzymatic factor wasdemonstrated by Ljungdahl when he isolated the “yellow affinity substance” from culturefiltrates of C. thermocellunz which promoted the binding of endoglucanases to cellulosefibres (Ljungdahl eta!., 1983). One year later Griffin isolated a factor from T. reeseifiltrates, a factor, that generated microfibrils from filter paper without hydrolysis (Griffinet a!., 1984). Since then many enzymes have been cloned that have a discrete cellulosebinding domain but till the present time, only the CBDs of the exoglucanase/xylanase Cexand the endoglucanase CenA of C.Jmnii and the cellobiohydrolases CBH1 and CBHII ofT. reesei have been characterized in any detail. The CBDs have been reviewed recently(Gilkes eta!., 1991a) and have been grouped into five families I-V based on amino acidsequence similarities (Coutinho eta!., 1992). Most of the CBDs vary in size from 35-150 amino acids. Although not essential for catalytic activity, the CBDs do influence theactivities of the catalytic domains (van Tilbeurgh eta!., 1986; Gilkes eta!., 1988; Tommeeta!., 1988; Ghangas and Wilson, 1988). The binding of CBDs to cellulose has beenanalyzed with the native enzymes or with the isolated CBDs (Tomme et a!., 1990;Stahlberg eta!., 1991; Din eta!., 1991; Coutinho eta!., 1992, Gilkes eta!., 1992). Itseems likely that the catalytic domains also influence the binding of the CBDs to celluloseTomnie eta!., 1990; Gilkes eta!,, 1992). The solution of the three dimensional structureof the C-terminal cellulose binding domain of cellobiohydrolase I from T. reesei bynuclear magnetic resonance has been determined (Kraulis eta!., 1989). The protein has awedgelike shape with an amphiphilic character, one face being predominantly hydrophilicand the other mainly hydrophobic.Linkers: These are short regions of amino acids rich in proline and/or hydroxyaniinoacids. Most range in size from 6-40 amino acids and are thought to connect differentfunctional domains (Gilkes et a!., 1991 a). Linker sequences have been observed in many7bacterial and fungal cellulases and have been termed a ‘hinge’ (Warren et a!., 1986;Langsford eta!., 1987; Béguin, 1990; Gough eta!., 1990, Gilkes eta!., 1991a) primarilybecause of their similarity to immunoglobulin molecules (Burton et a!., 1989) where theirpresence may serve as a hinge to provide flexibility to the two arms. This has allowedspeculation that the Pro-Thr like regions may also act as a hinge and thus allowconformational mobility for interactions between the binding domain, the catalytic domainand the substrate (Ferreira et al., 1990; Shen et al., 1991). Linker regions because of theirextended hinge like structure are often sensitive to proteolysis. Cleavage sites immediatelyadjacent to linker regions have been defined for CenA and Cex (Gifices eta!., 1988) andCenB from C.jmnii (Meinke eta!., 199 la,b) and for other cellulolytic enzymes (Tomme eta!., 1988; Gough et a!., 1990). Replacement of the hinge region of CenA with the hingeof immunoglobulin A (IgA) allowed hydrolysis of CenA with an IgA protease (Miller eta!., 1992). Linker regions are also extensively 0-glycosylated (Tomme eta!., 1988;Jentoft, 1990; Williamson eta!., 1992) and this post-transalational modification has beenshown to afford protection from proteases (Langsford et at., 1987). Another potentialapproach to afford protection of enzymes from proteolytic cleavage is by linkerreplacement.Repeated sequences: Most -1,4-glycanases are in the range 30 kDa to 70 kDa(Aubert eta!., 1988; Béguin, 1990; Gilkes eta!., 1991a). Some f3-1,4- glycanases arelarger than this and may contain other domains in addition to a catalytic domain and aCBD. For example, repeated sequences 20 to 150 amino acids long occur in several I1,4-glycanases of widely different molecular weights (Hall eta!., 1988, Béguin, 1990;Gilkes et a!., 1991 a). The functions of the repeated sequences are mostly unknown, butin endoglucanase CeIZ from Clostridiuni stercorarium they bind the enzyme to cellulose(Jauris eta!., 1990). The C-terminal repeats of endoglucanase Ce1D from C!ostridiumthermoce!!uni have some sequence similarities to a number of calcium-binding8polypeptides, but their removal does not affect calcium binding by the enzyme (Chauvauxeta!., 1990).1.5. Cellulases from Cellulonionas flmiCellulase systems tend to be very complex, and their components can be difficult topurify by biochemical methods. Gene cloning and expression in an appropriate host whichis free of cellulase activity facilitates the isolation of a particular enzyme from even traceamounts of the other components. Our group has focused its efforts on gene cloningtechnology to express genes encoding individual enzymes of the multienzyme cellulasecomplex of Cellulomonasfirni in Escherichiu coli.Cellulomonasfirni is a Gram-positive, aerobic, mesophilic, rod-shaped bacterium.Béguin ci’ at. (1977, 1978) studied the cellulases from Cellulomonas species and reportedtwo classes of enzymes. One class comprised enzymes which were tightly bound tocellulose and the other enzymes which were found free in the culture supematant.Langsford eta!. (1984) reported the presence of 10 components with cellulase activity insupernatants from cultures of C.fImi grown on Avicel. Four cellulolytic enzymes fromC.firni have been characterized (Gilkes ci’ at., 1991b). Two of them, CenA (418 aminoacids) an endoglucanase (Wong et at., 1986) and Cex (443 amino acids) an exoglucanase/xylanase (O’Neill ci’ at., 1986) are major extracellular enzymes produced when C.Jlmiis grown on microcrystalline cellulose. Comparison of the amino acid sequences of Cexand CenA show that each enzyme has three distinct regions: a short sequence of about 20amino acid residues composed only of proline and threonine (Pro-Thr linker), which isconserved almost perfectly in the two enzymes; a region rich in hydroxyamino acids but oflow charge density which is 50% conserved (cellulose-binding domain); and a region,comprising about 70% of the polypeptide, the catalytic domain, which is not conserved.The order of the regions is reversed in the two enzymes (Fig. 2) with the Pro-Thr linkerseparating the catalytic domain from the binding domain in both enzymes (Warren et a!.,91986). Domain shuffling of the catalytic domains of CenA and Cex (Warren et at., 1987)as well as proteolysis with C. firni protease of both these enzymes have confirmed theindependent binding ability of the binding domain in the absence of the catalytic domain(Gillces eta!., 1988; 1989).More recently the genes for two other cellulolytic enzymes of C.Jlmi CenB(Owolabi eta!., 1988) and CenC (Moser et al., 1989) have been cloned and expressed inE. coil. CenB, an endoglucanase, has a molecular mass of 110 kDa and sharessimilarities with CenA and Cex in that it too can bind microcrystalline cellulose. It has anunusual sequence organization consisting of five domains: a catalytic domain at the N-terminus with an internal cellulose-binding domain, followed by three repeats related tofibronectin type III repeats and lastly a C-terminal cellulose-binding domain (Fig. 2)Meinke eta!., 1991a, b).CenC is an endoglucanase and has an apparent molecular mass of 130 kDa. It bindsboth to Sephadex and to cellulose. It was purified from C.Jirni culture supematants byvirtue of its Sephadex-binding capacity (Moser et at., 1989), and may correspond toenzyme CI, an endoglucanase of Mr 118 purified from the culture supernatant ofCeIlu!omonas sp. IIbc by Sephadex affinity chromatography (Beguin and Eisen, 1978).In C.fimi and in E. coil a truncated form (120 kDa) of the enzyme sharing the sameamino-terminus, CenC’, is produced by proteolysis of CenC (Moser et a!., 1989). CenCseems to be a minor component on a percentage weight basis of the cellulase system of C.fimi (Gilkes et al., 1984; Moser eta!., 1989).10CfCenACf CexJLCfCe ri Bs’\\.,,,..‘‘‘.%——,,.. .‘•1 1012Figure 2. Organization of catalytic and non-catalytic domains in cellulases from C.fimi. CfCenA, B and CfCex are the endo-B-1,4-glucanases andexo-13-1,4-glucanase/xylanase of C.firni. The primary structures are drawnapproximately to scale and are numbered from the amino terminus of the matureprotein. Catalytic domain ii:i, Linker region , Binding domain (Family II),Binding domain (Family III), Repeats E+3.1.6. Objectives of this thesisThe goal of this study was to correlate the structure of endoglucanase CenC with itsfunction. In terms of structure CenA (45 kDa) is about half the size of CenC yet bothappear to have the same function which is the cleavage of f3-1,4-glycosidic bonds. Theprimary objective of this thesis was to determine the sequence of CenC which might helpexplain why CenC is so large. Sequencing of cenC revealed that the gene was incompleteand was terminating in an adjacent sequence of the vector DNA in which it was cloned.11The missing sequence of the gene was cloned as described in the section on materials andmethods and this was followed by identifying putative domains by sequence comparison.CenC can bind to Sephadex and to cellulose. It was of interest to determine whichpart of the polypeptide was required for this function. Identification of the amino acidsequence(s) involved in binding is most easily accomplished by examining the isolateddomains. Proteolytic separation of domains has been used to dissect the structural andfunctional organization of endoglucanase CenA (44 kDa) and exoglucanase Cex (47kDa)from C.fImi (Gilkes eta!. , 1988, 1989) and of other 3-1,4-glycanases (Tomme eta!.,1988; Ghangas and Wilson, 1988). For large, multidomain polypeptides such as CenC,however, proteolytic separation of domains is difficult. It is easier to dissect the gene andto express independently the sequences encoding each domain. This approach was usedto correlate structure with binding activity either by deletion analysis and/or by expressingthe domains alone or in tandem using molecular genetic techniques.To fully understand the mechanisms by which cellulases degrade cellulose, it isnecessary to determine the roles of the CBDs. Exoglucanase and endoglucanase from T.reesei require a CBD for efficient hydrolysis of crystalline cellulose (Van Tilbeurgh et a!.,1986; Tomme eta!., 1988) while in C. therinocelluni a CBD increase the affinity ofendoglucanase E (EGE) for its substrate (Durrant ci’ a!., 1991). Although the CBDs lackhydrolytic activity, at least one of them, the CBD of CenA, can disrupt cellulose fibres(Din eta!., 1991). One way to compare the different CBDs is to fuse them to the samecatalytic domain. This study describes the construction of such a gene fusion whichencodes a mutant form of recombinant CenA in which the CBD of CenA is replaced withthe CBD of CenC. The influence of the CBD of CenC on the catalytic activity of themutant enzyme for a number of cellulosic substrates was compared to CenA.Recombinant CenA (non-glycosylated CenA produced in E. coli) is cleaved by C.Jimi protease adjacent to its linker sequence to generate a p30 fragment that has catalyticactivity but lacks the binding domain (Gilkes eta!., 1988; 1989; Shen eta!., 1991). Themutant form of CenA described above was designed in a manner that the 23 amino acidPT linker of CenA was replaced by the linker lying adjacent to the CenC binding domain.The resulting polypeptide was tested to see if it was more resistant to C. fimi protease thanCenA.Using a number of approaches which included studying intact CenC, its deletionmutants and by interchanging domains between CenC and CenA the objective of thisthesis is to postulate the special role CenC may play in the cellulase complex ofCellulomonasjmnii.12132. Materials and Methods2.1. Bacterial strains plasmids and phage vectorsCellulomonasfimi ATCC 484 was used for the work described in this thesis. TheEscherichia coli strain JM1O1 has been described previously (Yannish-Perron eta!.,1985). Site-directed mutagenesis was optimized by using the E. co/i host strain RZ1032(dut, ung-) (Kunkel et al., 1987). The plasmids pTZ-18R and 19R and helper phageM13K07 for the preparation of single-stranded DNA were from Pharmacia (Dorval,Quebec). Plasmid pUC18-1.6 cenA encoding CenA (Guo eta!., 1988; Wong eta!., 1986)and pUC18-1.6 cenA L\Pro-Thr (Shen et al., 1991) were described previously. Plasmidsencoding cenC and its derivatives are presented in Table. 1.2.2. Media and growth conditionsC.fimi ATCC 484 was grown at 30 C in Leatherwood’s basal medium (Stewartand Leatherwood, 1976). The E. CO/i strains were grown in Luria-Bertani (LB) medium(Sambrook et al., 1989) supplemented with 100 jig ampicillin ml at 37 C, for theproduction of recombinant protein unless stated otherwise or in 2 x YT medium (Messing,1983) for the production of single-stranded DNA for sequencing. Cultures were inducedwhen the A600 reached 1.0, with 0.1 mM final concentration of isopropyl-13-D-thiogalactoside (IPTG). Solid medium contained 1.5% Difco agar. Cells expressinginterposon 2 were selected on LB containing 75 jig ampicillin, 25 jig streptomycin and 25jig spectinomycin ml-’. Endoglucanase activity was detected on solid medium containing0.1% carboxyniethylcellulose by staining with Congo red (Teather and Wood, 1982). Theinteraction of the dye Congo red with intact B-glucans provides the basis of this rapid andsensitive assay for endoglucanases. Potential colonies carrying recombinant plasmidsexpressing endoglucanase activity were plated on a master plate and a duplicate plate. Theplates were incubated at 3T C, overnight. The colonies are then scraped of the plate, theplates were flooded with Congo red 2 mglml and left to rock gently on a platform for 1514minutes. The Congo red solution was poured off and the plates were flooded with IMsodium chloride solution to enhance the destaining process. The zones of hydrolysisappeared as a yellow halo against a red background.Table 1. Plasmids encoding cenC and its derivatives.Plasmid Description* Polypeptide ReferencepTZ-18R-8/5 Plasmid encoding cenC, CenC’ Moser et a!., 1989missing 3’ end: aa 1023-1069 missing.pTZP-CenC High-expression plasmid CenC CM Moser Ph.D. thesis, 1988encoding cenC, missing 3’end: aa 1023-1069 missing.pTZ-JC1 Plasmid encoding the 3’ ‘CenC Coutinho eta!., 1991end of cenC: aa1023-1069present.pTZ-JC2 High-expression plasmid CenC Coutinho et a!., 1991encoding ceiiC : CenC LPA+ aa 1-1069.pTZ-JC2 AKpnI Internal deletion of C1C2: CenC CA2 This studymissing aa 954-1004.pTZ-JC3 Plasmid encoding both N- N1N2 Coutinho et a!., 1992terminal repeats N1N2:CenC LP + aa 1-299.pTZ-JC6 Plasmid encoding Ni: Ni Coutinho et a!., 1992CenC LP + aa 1-150.pTZ-J7 Plasmid encoding both C- C1C2 Coutinho et a!., 1992terminal repeats C1C2:CenCLP+aa 1-11 fused toamino acids 860-1069.* Numbers refer to aa (amino acids) of the mature (processed) polypeptide.A LP refers to the (32 amino acid) leader peptide of CenC.15Table 1. Plasmids encoding cenC and its derivatives (Contd).Plasmid Description* Polypeptide ReferencepTZ-JC13 Plasmid encoding the C’ ‘A This studyfusion of CBD cenC to thecatalytic domain of cenACenC LPA + aa 1-299 ofCenC fused to aa 135-4 18of CenA.pTZ-JC14 Plasmid encoding Cl: Cl Coutinho et al., 1992CenC LP + aa 1-11 fused toamino acids 860-993.pTZ-JC34 Plasmid encoding N2: N2 This studyCenC LP + aa 1-38 fused toaa 167-299.* Numbers refer to aa (amino acids) of the mature (processed) polypeptide.A LP refers to the (32 amino acid) leader peptide of CenC.2.3. Enzymes and reagentsLow-viscosity carboxymethyl cellulose (CMC), dinitrosalicyclic acid (DNS),hydrobenzoic acid hydrazide (HBAH), IPTG and phenylmethylsulfonyl fluoride (PMSF)were from Sigma. The degree of polymerization of the CMC was 400 and the degree ofsubstitution 0.7. Restriction endonucleases, T7 DNA polyrnerase, T4 DNA ligase anddeoxyribonucleotides were from Pharmacia (Dorval, Quebec), New England Biolabs Inc.(Beverly, MA, U.S.A.) or Bethesda Research Labs., Inc. (Burlington, Ontario).Radioactive nucleotides were purchase from New England Nuclear Corp. (Dupont CanadaInc. Missisauga, Ontario). Oligodeoxyribonucleotide primers were synthesized on anApplied Biosysterns 380A automated DNA synthesizer (Oligonucleotide synthesis lab,University of British Columbia) and further purified by chromatography on a Sep-Pak C-18 reverse phase cartridge (Millipore/Waters Associates, Milford, MA, U.S.A.) (Atkinsonand Smith, 1984). Nitrocellulose membranes BA85 were from Schleicher and Schuell16Inc. (Keene, NH, U.S.A.). Avicel (microcrystalline cellulose) was from FMCInternational, Ireland. Sephadex 0-50 was from Pharmacia. Bacterial crystalline cellulose(BMCC) was from the cellulosic pellicle synthesized by Acetobacter xylinum (ATCC23769) (Hestrin, 1963) and phosphoric acid swollen cellulose (PASC) was prepared byprecipitation of a solution of Avicel PH1O1 in 70% (w/v) phosphoric acid (Wood, 1988).Ultra-filtration membranes, PM1O and PM3O were from Amicon (Danvers, MA, U.S.A.).Other chemicals and reagents were from BDH (Vancouver, B.C.) or Sigma Chemical Co.(St. Louis, MO, U.S.A.).2.4. DNA preparation and sequence determinationThe following procedures were described previously: isolation of total DNA fromC.fimi by a lysozyme-SDS-pronase procedure (Whittle et a!., 1982); isolation of plasmidDNA (Birnboim and Doly, 1979); equilibrium density gradient centrifugation in CsC1-ethidiurn bromide (Sambrook eta!., 1989); isolation of single-stranded DNA forsequencing (Sambrook et a!., 1989). The DNA sequence of cenC was determined byNuclease Ba131 deletions of the EcoRJ-HindJII fragment in plasmid pTZ-l 8R-8/5. 100 igof CsCI gradient, purified DNA, cut with the appropriate restriction enzyme, was treatedwith 5U of Nuclease Ba13 1 in the buffer recommended by the supplier for 60 minutes at37° C and in a final volume of 100 jtl. 5 tl samples were removed at appropriate intervalsand the reaction stopped by doing a phenol-chloroform extraction followed by ethanolprecipitation (Sarnbrook et a!., 1989). The extent of the deletions was monitored byagarose gel electrophoresis. Conditions were modified as required by either increasing ordecreasing the amount of nuclease. Samples with the appropriate level of digestion werepooled and the DNA excised and subcloned into pTZ-18R (Pharmacia). The second strandwas sequentially deleted and sequenced in the same way by cloning the fragment in pTZ19R. Single-stranded DNA was sequenced by the enzymatic method (S anger eta!., 1977)using T7 DNA polymerase and35S-cc-dATP (Tabor and Richardson, 1987) and replacing17dGTP with 7-deaza-dGTP (Mizusawa et at., 1986). The nucleotide mixes were adjustedfor the high G+C content of C.fimi DNA. The synthetic oligonucleotide P-816 waslabelled with 32P using [ct-32P]-ATP and T4 polynucleotide kinase, then used to screentransformants by hybridization (Sambrook et a!., 1989).2.5. Cloning the 3’ end of the cenC geneA KpnI site was located close to the end of C.finii DNA in pTZ-18R-8/5. Anoligonucleotide (P-8/6) was synthesized which was complementary to the 30 bases at the3’ end of C.fimi DNA in pTZ-18R-815. After labelling with 32P as described above, theoligonucleotide was used to probe C. Jimi DNA which had been digested with KpnI andPstI. A 1.8 kb fragment hybridized with the probe. DNA was recovered from this area ofthe agarose gel, ligated into pTZ-18R which had been digested with KpnI and PstI, thenused to transform E. coti JM1O1(Sambrook ci’ at., 1989). Clones containing potentialinserts were transferred onto nitrocellulose filters, lysed with alkali and the filters probedwith the oligonucleotide. Positive clones were sequenced. A clone designated pT’Z-JCl,was obtained which contained the 3’ end of cenC. pTZ-JC1 was digested with KpnI andHindill and the 1.8 kb fragment recovered and ligated to the large fragment released frompTZP-CenC cut with KpnI and Hindu! to give pTZ-JC2.2.6. Construction of deletion mutants of cenC2.6.1. N-terminal deletions of CenCConstruction of deletion mutants using Nuclease Ba13 1: N-terminal internaldeletion mutants of CenC were created by linearizing plasmid pTZ-JC2 at the uniqueBamHI restriction site followed by treatment with Nuclease Bal3l. Conditions were asdescribed previously but in this case 20 jig of DNA was digested with 1U of nuclease in atotal volume of 20 j.il for 20 minutes. Samples were removed at intervals of 1 minute18and analyzed by agarose gel electrophoresis. Samples containing appropriately digestedDNA were phenol-chloroform extracted and ethanol precipitated. The DNA was ligatedwith T4 DNA ligase and transformed into E. coil JM1O1. Ampicillin-resistant colonieswere screened for carboxyrnethylcellulase activity. Individual colonies were subjected tomini-preparation of DNA and the DNA cut with restriction endonucleases to detect theextent of deleted DNA. Suitable clones were sequenced to map the precise extent of thedeletion.2.6.2. C-terminal deletions of CenCPlasmid pTZP-CenC (CenC CM), missing the 3’ end of cenC, wasobtained in the initial cloning of cenC and has been described.Construction of pTZ-JC2 AKpnI (CenC CA2) a C-terminal internaldeletion: pTZ-JC2 carrying the entire coding sequence for cenC has two in-frame KpnIsites separated by 150 base pairs (amino acid 954-1004). Plasmid DNA was digested withKpnI and the fragment recovered after agarose gel electrophoresis of the digest. Thefragment was ligated then used to transform E. coil JM1O1. Ampicillin resistant colonieswere screened ([Feather and Wood, 1982) for carboxymethylcellulase (CMCase) activity.The plasmid DNAs from positive clones were sequenced across the junction between theKpnI sites. A plasmid with the correct sequence was designated CenC CA2.CenC Cz3 was made by inserting the 2 fragment at a HpaI site,(nucleotides 2837-2842) 11 codons downstream from the start of the region encoding Cl.2.7. Cloning of the N-terminal repeats N1N2 and Ni and the C-terminal repeats C1C2and ClpTZ-JC2 and the interposon were digested with Sn,aI, then ligated together togive pTZ-JC3. pTZ-JC2 was digested with Sad, then with mung bean nuclease andligated to SmaI-digested 2 to give pTZ-JC6. A second Apal site was introduced into pTZ19JC2 by site-directed loop-out mutagenesis (Zhou et at., 1990). The nucleotides encodingamino acids 859 and 860 of mature CenC were changed from GAGCTC to GGGCCC(Coutinho eta!., 1991). The mutant DNA was digested with Apal and the larger fragmentligated to give pTZ-JC7. pTZ-JC7 was digested with Mitt!; then the ends filled in usingthe Kienow fragment of DNA polymerasel and the blunt-ended DNA ligated to SmaJdigested 2 to give pTZ-JC14.2.8. Cloning of N2 using Nuclease Ba131To obtain N2, an N-terminal internal deletion mutant created in Section 2.6.1. wastruncated by introducing the 2 fragment at the Snial site (nucleotides 1050-1055) asdescribed before for generating N1N2.2.9. Deletion analysis of NiNi internal deletions (Ni 3 and A4) were obtained by the same method asdescribed in Section 2.6.1. except that the 2 fragment was introduced at the Sad site(nucleotides 606-6 11), following treatment with mung bean nuclease, as described forgenerating Ni. C-terminal truncations of Ni (Ni M and A2) were obtained by insertingthe Q interposon at sites MtuI (nucleotides 559-564) and BstEJI (nucleotides 466-47 1)respectively.2.10. Substituting N1N2 for the CBD of endoglucanase CenApTZ-JC2 encoding cenC has a Snial site between the sequence encoding its CBDand the catalytic domain. pUCi8-i.6 cenAAproT1w encodes a deletion mutant of thecenA gene in which the sequence encoding the linker connecting the CBD and the catalyticdomain is replaced with a HpaI recognition sequence. pTZ-JC2 was digested with Smaland HindIII and the 3.9 kb fragment was recovered after agarose gel electrophoresis of thedigest. pUCi8-i.6 cenAApi•oTIr was digested with HpaI and HindIll and the 1.1 kb20fragment recovered as above. The fragments were ligated, then used to transform E. coliJM1O1. Ampicillin resistant colonies were screened for carboxymethylcellulase activity.The plasmid DNAs from positive colonies were sequenced across the junction betweenthe cenA and cenC fragments. A plasmid with the correct sequence was designated pTZJC13 and the fusion polypeptide it encoded CBDCe11- CatCenA (C” A).2.11. Purification of CenC, CenC CAl and CenC CA2Cells were grown in LB medium supplemented with 100 .ig ampicillin m1’ at 30°C. After overnight growth, the cells were isolated with a Sharples centrifuge at 33,000g.The cells were ruptured by two passages through a French pressure cell in the presence of1 mM PMSF and 0.01 mM pepstatin A. Streptomycin sulfate was added to a finalconcentration of 1.5% (wlv). After overnight precipitation of nucleic acids at 4°C, theextract was clarified by centrifugation twice at 35,000g in a Beckman JA2O rotor. Avicelwas added to a final concentration of 1% (w/v). The suspension was stirred at 4° C for 1hour. The Avicel was recovered by filtration through a Whatman GF-A filter and washedwith 1M NaC1-50 mM potassium phosphate, pH 7.0 - 0.02% sodium azide to removenon-specifically adsorbed proteins and then with 50 mM potassium phosphate, pH 7.0 -0.02% sodium azide till the A280 nm of the filtrate was <0.01. The adsorbed proteinswere eluted with water until the A280 nm of the filtrate was again <0.01. The solution ofdesorbed proteins was concentrated by ultrafiltration through an Amicon PM3Omembrane, and the water exchanged with 20 mM potassium phosphate, pH 5.8 - 0.02%sodium azide. This fraction was then subjected to anion-exchange chromatography withthe aid of an FPLC using Mono-Q as the matrix (Pharmacia). Bound material was elutedusing a linear salt gradient of 0 - 1.0 M NaCI in 20 mM potassium phosphate, pH 5.8 -0.02% sodium azide over a period of 200 minutes at a flow rate of 1 ml mm4.50 111 ofthe peak fractions were assayed for activity on a Titertek plate using 50 j.tl CNPC as thecolorimetric substrate and then analyzed by SDS-PAGE. The active fraction was21concentrated on a PM3O membrane and the buffer exchanged for 50 mM potassiumphosphate, pH 7.0 - 0.02% sodium azide. The concentrated material was made up to afinal concentration of glycerol of 2% (v/v) and then applied to a P-150, Bio-Rad gel-filtration column, mesh size 100-200 at a flow rate of 0.2 ml min1.Active fractions weretested as before, analyzed by SDS-PAGE and concentrated in an Amicon ultrafiltrationunit.2.12. Expression of C1C2 and ClCultures of JM1OI/pTZ-JC7 or JM1O1/pTZ-JC14 were grown as above andpolypeptides of the expected sizes corresponding to C1C2 and Cl were observed in theculture supernatants. The amino acid sequences of the N-tennini were confirmed by aminoacid sequencing.2.13. Purification of N1N2, Ni and N2Cultures of JM1O1/pTZ-JC3, JM1O1/pTZ-JC6 or JM1O1/pTZ-JC34 were grownat 370 C to A600 of 1.0, then induced with IPTG. After growth overnight, the cells wereremoved by centrifugation. Avicel or Sephadex G-50 was added to the supematant (lg100 m11). The suspension was kept at 40 C for 3 hours with occasional stirring. TheAvicel/Sephadex was recovered by filtration through a Whatman GF-A filter and washedas described previously for the purification intact CenC. Adsorbed proteins wererecovered by washing the Avicel/Sephadex with water. The water extract wasconcentrated by ultrafiltration through an Arnicon PM1O membrane and the waterexchanged with 50 mM potassium phosphate - 0.02% sodium azide. The N-termini of allthree polypeptides were confirmed by amino acid sequencing. N2 was purified in thesame way as N1N2 and Ni.222.14. Purification of C” AC’ A (60.9 kDa) was purified from culture supernatants of E. coli JM1O1/pTZ-JC13 in the same way as described for N1N2, Ni or N2. The purity of C’ A wasmonitored by SDS-PAGE and the N-terminal sequence confirmed by amino acidsequencing.2.15. Measurements of CenC activityThe activity of CenC on CMC by the HBAH method was measured as describedpreviously (Moser eta!., 1988; Coutinho et at., 1991). The assay conditions for theHBAH method were as follows: 0.1 ml of appropriately diluted enzyme solution wasmixed with 0.4 ml of 0.5% carboxmethylcellulose in 50 mM sodium citrate buffer, pH6.8. After incubation at 300 C for 30 minutes, the reaction was stopped with 1.0 mlHBAH reagent (Lever, 1973). The tubes were steamed for 12 minutes and the absorbancewas read at 420 nm. The kinetic parameters of CenC using chloronitrophenylcellobioside(CNPC) as the substrate were measured and compared to Ce1D (endogiucanase D from C.thermocellurn) another endoglucanase from Family El (Tomme eta!., 1991). Assays weredone in triplicate. Briefly a stock concentration of the enzyme 3.85 tM was diluted 1:10and 20 j.il, added to 80 tl of 50 mM buffer (Na2HPO4 / citric acid buffer) with 100-1000.jM CNPC as the substrate. The release of 2-chloro-4-nitrophenol was monitoredcontinuously (Hitachi U-2000 spectrophotmeter) at 400 nm. Temperature optimummeasurements were done in the same way except 2 mM CNPC was used as the substrate.Temperatures were kept constant by using a circulating water bath (Forma Scientific)directly connected to the spectrophotometer chamber. Buffers for the determination of thepH optimum of CenC were 50 mM: acetate buffer (pH 4.0), citrate-phosphate buffer (pH5.0), phosphate buffer (pH 6.0, 7.0) and Tris-hydrochioride buffer pH 8.0) (Perrin andDempsey, 1974).232.16. Measurements of C’ ‘A activityThe activity of the purified enzyme on CMC, Avicel, BMCC, cellulose azure andPASC was measured by the HBAH method (Gilkes et at., 1988; Shen et a!., 1991).Assays were done in triplicate. Initial rates of hydrolysis were measured by assaying forthe amount of reducing sugar released with the HBAH reagent (Lever, 1973). Theabsorbance of the colored reagent was read at 420 nm against blanks containing equivalentamounts of enzyme added after mixing substrate with HBAH reagent.2.17. Oligosaccharide products of CenCOligosaccharides G3-G6 were from Seikagaku (Tokyo, Japan). Cleavage productswere separated on a Waters, Dextro-pak column (Millipore) using an HPLC (Waters) andmonitored by a Waters 410 differential refractometer. The flow rate was 1 mi/mm at roomtemperature with water as the solvent. The column was capable of separating the a fromthe f3- anomer for oligosaccharides G3 and G4, with the a anomer having a slightlylonger retention time. Reaction mixtures contained substrate concentrations in the range0.5 - 2.0 mM and were incubated with 5U of enzyme at 37°C in a final volume of 40 jilfor times ranging from 2 minutes to 240 minutes. 30 .tl of sample was injected onto thecolumn for analysis.Cleavage, of 2-chloro-4-nitrophenyl glycosides (derived from cellodextrins), at Clwas followed spectrophotometrically at an absorbance of 400 nm, following the release ofphenol (Tomme ci’ a!., 1991).2.18. Deteririination of protein concentrationProtein concentration was routinely measured by dye binding (Bradford, 1976)using the Bio-Rad protein assay kit. Purified protein concentration was determined byabsorbance at 280 nm using the extinction coefficient obtained for the polypeptide(Scopes, 1974).242.19. Amino-acid sequence determinationPolypeptides were prepared for sequencing by electrotransfer from SDS-PAGEgels (Laemmli, 1970) on to a polyvinylidine difluoride membrane (PVDF membrane,Millipore) (Matsudaira, 1987, 1990). The N-tenninal amino acid sequences of the excisedbands were determined by automated Edman degradation with an Applied Biosystems470A gas phase sequenator (Protein Sequencing Lab., University of Victoria, B.C.).2.20. Protein electrophoresisProteins were analyzed by SDS-PAGE (Laeimrili, 1970) and detected byCoommassie blue staining. Calibration standards were myosin, rabbit muscle (212 kDa);13-galactosidase, E. coli (130 kDa); phosphorylase B, rabbit muscle (97.4 kDa); albumin,bovine (68 kDa); glutamate dehydrogenase (53 kDa); ovalbumin (43 kDa);glyceraldehyde-3-phosphate dehydrogenase, rabbit muscle (36 kDa) and carbonicanhydrase (29 kDa). Low range calibration standards were from Bethesda ResearchLabs.: ovalbumin (43 kDa); carbonic anhydrase (29 kDa); 13-lactoglobulin (18 kDa);lysozyme (14 kDa); bovine trypsin inhibitor (6 kDa) and insulin (3 kDa).2.21. ImmunoblottingCenC or its derivatives were detected by immunoblotting (Harlow and Lane,1988). Bovine serum albumin (3 %) was used for blocking residual sites on thenitrocellulose membrane and washes included 0.05 % Tween 20 to get rid of non-specificadsorption. Western blotting was done using polyclonal sera against CenC (Moser, Ph.D.thesis) and goat anti-rabbit IgG-alkaline phosphatase conjugate as the secondary antibody.252.22. Adsorption of N1N2 or Ni to Avicel and SephadexThe binding to Avicel or Sephadex G-50 of polypeptides in cell extracts or culturesupematants was evaluated by allowing the polypeptides to adsorb as described for thepurification of Ni and N1N2. The washed adsorbent with adsorbed polypeptides washeated in gel loading buffer (Laemmli, 1970) prior to analysis by SDS-PAGE.2.23. Adsorption of N1N2, Ni or CBDCex to BMCC and PASCBMCC was extracted from cultures of Acetobacter xylinum (ATCC 23769) grownon peptone, yeast extract, glucose medium (Hestrin, 1963). PASC was prepared bytreating Avicel with phosphoric acid (Wood, 1988).CBDCex was used as a comparison and purified polypeptide was provided by Dr.E. Ong (E. Ong, Ph.D. thesis). N1N2, Ni and CBDCex were stable for 36 hours underthe test conditions. The optimal times for equilibration were determined by incubatingsaturating levels of purified polypeptide with cellulose (1 mg of PASC or 1 mg of BMCC)in a total volume of 1 ml 50 mM potassium phosphate buffer pH 7.0 - 0.02% sodiumazide in a 1.5 ml Eppendorf tube at 30° C with gentle agitation in a rotary mixer.Equilibration was reached between 30 minutes and 1 hour after adding the polypeptide; theamount of unbound polypeptide in the supernatant changed negligibly between 1 hour and36 hours after addition. For convenience, mixtures were allowed to equilibrate for 24hours at 30° C in all subsequent experiments. Tubes containing substrate and nopolypeptide were used as controls. After equilibration, the tubes were centrifuged at15,000g for 10 minutes. The supernatant liquid was transferred to another tube andcentrifuged again for 10 minutes. The absorbance of the supernatant fluid was measured at280 nm. Protein concentrations were determined from the extinction coefficients at thiswavelength. Adsorption was expressed as jamoles protein bound g1 cellulose. All assayswere done in triplicate.262.24. Adsorption of N1N2, Ni and N2 to PASCN1N2, Ni and N2 were incubated with or without 1 mg of PASC atconcentrations of 25, 100 and 250 ig m1’ in a total volume of 1 ml of 50 mM potassiumphosphate buffer, pH 7.0 - 0.02% sodium azide. To allow for equilibration the sampleswere incubated at room temperature in a rotary shaker for a period of 24 hours in 1.5 mlEppendorf tubes. The tubes were centrifuged at 15,000g for 10 mm. Samples wereanalyzed by SDS-PAGE (Laemmli, 1970). Gels were scanned by a Bio-Image (Millipore)computer assisted densitometer to get a relative estimate of the percent of free polypeptidepresent in solution.2.25. Rapid small scale screening procedure for functional binding activityCultures carrying individual deletion mutants were grown in 500 ml of culturemedium at 37° C to A600 of 1.0, then induced with IPTG. The cells were removed bycentrifugation and the culture supernatant was mixed with 10 g of Avicel. The Avicel waswashed and the proteins desorbed with 100 ml of water. The eluted material was thenconcentrated and exchanged with 50 mM potassium phosphate - 0.02% sodium azide byultrafiltration through an Amicon PM 10 membrane and analyzed on a SDS-PAGE gel.2.26. Adsorption of CenA and C’’ A to BMCC and PASCIncreasing amounts (25, 100 and 250 j.ig) of purified CenA and C” A were mixedwith 1 mg of BMCC in 1 ml of 50 mM potassium phosphate, pH 7.0 - 0.02% sodiumazide in 1.5 ml Eppendorf tubes. Controls did not contain BMCC. The tubes wereincubated at room temperature on a rotary mixer. Both polypeptides were stable for at least24 hours under these conditions. After 8 hours, the tubes were centrifuged at 15, 000g for10 mm. The supernatants were transferred to fresh tubes and centrifuged again at 15,000g for 10 mm. Then a 500 jil sample of each supernatant was transferred to anEppendorf tube containing 500 .ig PASC. The second set of tubes, together with those27containing the remainder of the supematants from the first set of control tubes, wereincubated as before for 8 hours. The supernatants were recovered after two centrifugationsas before. Equal quantities of all supernatants, including the residual supematants from thefirst set of tubes, were analyzed by SDS-PAGE. Gels were scanned by a Bio-ImageMillipore) computer assisted densitometer to get a relative estimate of the percent of freepolypeptide present in solution.2.27. Hydrolysis of CenA and C’’ A by C.fimi proteaseC.firni protease was prepared from culture supernatant as described previously(Gilkes eta!., 1988). The activity of the preparation was measured by incubating sampleswith 10 mg of hide powder azure in 1.5 ml of 50 mM phosphate buffer pH 7.0 at 370 C.The A585 nm of the supernatant was measured at intervals. One unit of activity gave anincrease in A585 nm of 1.0 Fr1. 80 .tg samples of CenA and C’’ A were incubated withC.Jirni protease for 3 hours at 30° C in 50 iI 50 mM potassium phosphate, pH 7.0 -0.02% azide. After incubation, samples from each tube were analyzed by SDS-PAGE.2.28. Amino acid sequence comparisonsAnalysis were done on a PC-AT compatible microcomputer, using programs inPC-Gene (Intelligenetics Inc., Mountain View, California). PCLUSTAL was used formultiple sequence alignments and generation of dendogranis.283. Results3.1. Nucleotide and amino acid sequence of cenCCenC was cloned initially into 2gt1 1, using a partial Saullia digest of C.fimiDNA. Screening of the library was done using a restricted pooi of oligonucleotidescorresponding to the codons for amino acids 7-11 of the N-terminal of CenC. Thefragment encoding CenC was then subcloned into pTZ18R to give pTZ18R-8/5 (Moser etat., 1989). Sequencing of the C. fIrni DNA after Nuclease Ba13 1 deletions of pTZ1 8R-8/5 (Fig. 3,4) showed cenC to be incomplete.The 18 amino acids, sequence RLPFTSCAVCLQDSMRRR, at the 3’ end of theopen reading frame established by the N-terminal amino acid sequence of CenC (Moser etat., 1989), corresponded to codons from ?.gt1 1.The missing fragment of the gene was cloned as described in Materials andMethods. Briefly, a 30 mer synthetic oligonucleotide P-8/6 corresponding to thenucleotide sequence at the 3’end of the truncated gene was used to probe C.fimi genomicDNA cut with restriction enzymes KpnI and PstI. A 1.8 kb KpnI/PstI fragment wasobtained after digestion of C.fin2i genomic DNA (encoding the 3’ end of the gene) (Fig.5)was recovered and cloned into pTZ- 1 8R. Colony hybridizations of potential clones gavefour colonies which carried the correct insert (Fig.6) and the plasmid encoding the 3’ endof the gene was designated pTZ-JC1(Fig. 7). Sequencing of the C.firni DNA in pTZ-JC1showed that 46 codons were missing from the cenC sequence in pTZ18R-8/5. PlasmidpTZ-JC2 (Fig. 7) encodes complete CenC and was generated by subcloning the KpnIHindlil fragment from pTZ-JC1 into pTZP-CenC (which carries a portable transalationinitiation site for improved expression).Codon usage in cenC was very similar to that in cex, cenA and cenB Meinke eta!., 1991a; data not shown), reflecting the 71 moles% G+C in C.fimi DNA.29Figure 3. Nuclease Ba13 I deletion of plasmid p1Z— I XR—X/5 to generate deletion clones forthe sequencin of cenC. The DNA molecular v. eight markers:X/I 1 and X/E. H refer to‘ambda DNA cut with Hi,ullhI and I-/inS/li ILcuRI respectively. Time course digestion ofpTZ— I 8R—/5. in buffer recommended b the supplier, from 0—60 minutes at 37 C using 5U of Nuclease Bal3 I . I 00u of DNA. in a total volume of 100 p.1.pTZ 18R_815kb HEIH0 Bal3123.1.3.5030Figure 4. Mini-preparation of plasmid DNA from clones generated after Nuclease Ba13 1deletions and cut with BarnHI. The DNA molecular weight markers: HIE and P referto lambda DNA cut with Hind!!!! EcoR! and PstI respectively.kb A)’HIE P23 12 190 56DELETEDCLONES“N”aaa31Figure 5. Southern blot of C.Jirni DNA cut with K (KpnI), P (PstI), K/P (KpnI and Pstf)and probed with the synthetic oligonucleotide P-8/6 labelled with [cc-32P]-ATP to locatethe fragment encoding the 3’ end of cenC. Uncut plasmid pTZ-18R-8/5 was used as acontrol. A unique band corresponding to a size of 1.8 kb was excised and cloned intopTZ-18R also cut with KpnI and PstI, and then transformed into E. coli JM1O1.C.timi I8R8/5kb K P KIP —r -—723.81.8j32Figure 6. Colony hybridization of potential clones carrying the 1.8 kb KpnIIPstI fragmentencoding the 3’ end of cenC with oligonucleotide P-816. Nitrocellulose filters were probedin duplicate to minimize the possibility of picking non-specific background spots aspotential clones. C refers to JM1O1/pTZ18R-815 as control.•( :‘VI .•133Figure 7. Plasmids used for the cloning of complete cenC and for its high-expression. C.fimi DNA is indicated by the double line, with the cenC coding sequence in black. Thearrow shows the direction of transcription of cenC. Plasmids: pTZ18R-8/5, encodes cenCbut missing the 3’ end of the gene. pTZP-CenC is the same clone engineered for high-expression with a portable transalation initiation site. pTZ-JC1 encodes the 3’ end of cenCand pTZ-JC2 encodes complete cenC.Blunt!BamHI.HiIIII-lindlllBluntBamHlHindill34Twenty-one codons were not used in cenC, and of these, twenty were not used in any ofthe genes.Endoglucanase C from Cellulornonasfirni is one of the largest 3-1,4-g1ycanasescharacterized to date. The mature polypeptide is 1069 amino acids long, preceded by aleader peptide of 32 amino acids (Fig. 8). The p1 of CenC calculated from its predictedamino acid sequence is 4.1. This may contribute to the discrepancy between the molecularmass of 112, 969 calculated from the predicted amino acid sequence and the molecularmass of 130 kDa observed by SDS-PAGE. Acidic proteins can give Mrs by SDS-PAGEwhich are significantly greater than their actual molecular masses (Kaufman eta!.,1984).The promoter and ribosome binding site of CenC were identified previously (Fig.8; Moser et at., 1989). A 9 bp inverted repeat 7 nucleotides downstream of the putativetranslational stop codon could be a transcription termination signal (Fig. 8), which wouldgive a mRNA of —3.37 kb. This was in good agreement with the size of 3.5 kbdetermined for CenC rnRNA from C.JInii (Moser ci’ at., 1989).The most striking features of the predicted amino acid sequence of CenC were thelong tandem repeats at both ends. Two contiguous repeats of —150 amino acids at the N-terminus (Ni, N2) were 50% similar, and two unrelated, contiguous repeats of —100amino acids at the C-terminus (Cl,C2) were 60% similar (Fig. 9) (Coutinho eta!., 1991).It is rare amongst 13-1 ,4-glycanases for the catalytic domain to be flanked at both ends byextended sequences of amino acids (Beguin, 1990; Gilkes et a!. , 1991a).Amino acids 300-880 of CenC shared greater than 30% similarity with the catalyticdomains of f3-1,4-glycanases of sub-family El (Fig.8, 10; Henrissat eta!. , 1989;Béguin, 1990; Gilkes eta!., 1991a). There was much less similarity with the catalyticdomains of members of sub-family E2 (data not shown). Interestingly, the tetrapeptideDAGD (amino acids 47 1-474 in CenC) occurs 305-329 amino acids from the tripeptide HR (amino acids 799-80 1 in CenC) in all of the enzymes in family E. The consensussequences around these motifs in subfamilies El and E2 share some similarity35Figure 8. Nucleotide sequence of cenC and its flanking regions, and the deduced aminoacid sequence of CenC. The nucleotide sequence is numbered with +1 as thetranscriptional start site. The amino acid sequence is numbered with +1 as the N-terminalamino acid of mature CenC. -10 and -35 are the promoter sequences, and S.D. is theribosome binding site identified previously (Moser et at., 1989). The leader peptide, 32amino acids, precedes A, the leader peptide processing site. The inverted repeatsdownstream of the translational stop codon are underlined with arrows. The repeats in theamino acid sequence are boxed. AV and BV are the junctions between the N-terminaland C-terminal repeats, respectively. The KpnI site near the 3’ end of the codingsequence is boxed. P-8/6 indicates the sequence of the oligonucleotide probe used toclone the 3’ terminus of the cenC coding sequence. The underlined amino acid sequenceTi 15 matches completely, the sequence of the tryptic peptide determined from thedigestion of CenC previously, and also used for generating an oligonucleotide probe in theinitial cloning of cenC (Moser et at., 1989). This sequence appears in theEMBL/GenBankIDDBJ Nucleotide Sequence Data Libraries under the accession numberX57858.-62-35P-10+1SD.___________59_____________________179MVSRRSSQARGALTAVVxrLALALAasGTALAIAsPIGZGTI8A299F’00Gp5GWVAYGT1)GPL0TSTGALCVAVPA0SAQY0V0VVI48 419ILNGVAIREGTTYTLRYTATASTDV?VRALVGQNGAPYGTVI88 539IL0TSPALTSSPRQVTSPPPASAPYPAPPAA00PS0QIAPI128659ILCCFSA0AWTFCLDOVALDBIlEVSLLPHP3FAKSL0PWSL168779ICTSRPVFADCRF4CVDLPG0QONPWDAGLVYNGVPVGKGSS209999VLSFPASATPDHPVRVLVOKG0GAYRPAFEQGSAPLTGI2481019IPATREYAPPSNLPPPP00DAP0QVAFHI.0KA0AYKPCISQI288______________________1139IVSLTTSATIPPP0YEPDTOPRVRVNQVGYLPPGPKRATLVT3291259NAASPVAWSLRDA00VVVAN0PSSPR0VKPSAAQAVHVLN3681379FS0VTTQCACYTLVA00KPSRPF0ID00LYQQLRY0ALNY4081499FYLARSCTSISA0VVCKKVARKACHVCVAPNQC0P0VPCI4481619CPRDYV0CWTC0VRL0VS00WV0A00II0KVVVN00IAVCQ4881739LLQTVSRALHAGPA0ALA00PLDVPBHCNNVP0VL0BARw5281859ELZWHLSHIVPEGKYACHVHHKVItDKCWTCLPLLPA00PQ568Ti151979ARSLHRPSTAATLNLSAVAAQCARLLKPY0PQLAQPLLKA6082099ARPTWAAAQHPALYAPCKACA0CCCAYNNSQVA0KPYWA6482219AAKLYLTTCE0AFATAVPP5PLHPA0VFPADCFCWCSVAA6882339LCRL0LAPVPNKLPCL0AVQSSVVKCAQKYLAAQACQCPC7282459S1.‘1SPPCCEVVWCSSSQVANNLVVVAPAV0LP0DKRFRAA7682579TLECL0VLPCRNALNQSVVP0WCKVASHQQHSRWFAHQLN8082699PSLPSPPPCSLACCPNSQAAPW0PPPKAAPP00CAPSACV848_________2819V0KIQAWSTNELPVNWNSALSWVASUVA0QCSAKPVPITAI8882939VTRQPV0ATVALCA0APFTAKASCVPAPPVRWQVRACR]92830591wK0VAGATCTPLTVRATARTNCTRYRAVFTNAAGSVKSAV9683179IVRLTVKRIIAAPVVTQHPA0VRARVCPRAVFRAAA0CVPTPCj1008P-ale3299LVVWQVRWCCCSU.RPIPWAPSPPLSVPVPVLAACPKIRAVF10483419ITNAVGTAATEPAELAVQRPRSI*>-<1069353638Figure 9. Repeated amino acid sequences in CenC.A N-terminal repeats.. C-terminal repeats.:, identical amino acids; , conservative changes;-, gaps left to improve the alignment.Amino acids said to be ‘similar’ are: A,S,T; D,E; N,Q; R,K; I,L,M,V; F,Y,W1 60* . . * * * • *.. *. • * * .*. ..• *149 EvErLpHTSF?ESLGPWSLYGTSEPVF2D-GRMCVDLPGGQ 4DGLVYNGVPVGEXESY 20961 TLRrrATAST VRALVGGAPYG1VLDT-SPALTSEPRQVTErFTASATYPATPAADDP 121• *••*.* .* * *. .* *.*. .210 VLSFrASATPEt1PVRVLVGGGAYRTFEQS?PLTGEPATREYAFrSNLTFP--PD3DAP 269122 BQIAFQLGGFS7DAWTFCI.DDVALDS 148•.*.. .. .* •.* .*. *270 -GQVFHI1KAG---AYE’CISQVSLTTSAT 296886 TAP rEQ-P VRAGX-WVAGATGrTLTVRAT 945* *•* *.976 AAPVVr-QHPADVRA’GTPAVAAADY?TPC\M VGGSWRPIP3ATSTTLSV?%1T 1036946 ?IDTRYRAV?rNAAGSVESAVVRLTVER 975*.:““:* * :*:1037 VLAAGTEYRAWrNAVGTAATEPAEUVQRPRS 1069Figure 10. Sequence similarity in the catalytic domains of 3-1,4-g1ycanases of subfamilyEl. Symbols for identity, conservative changes and gaps, as for Fig. 9. CfCenC,endoglucanase C of C. flnii; PfEndA, endoglucanase A of Pseudomonasfluorescenssubsp. cellulosa (Hall and Gilbert, 1988); CtCe1D, endoglucanase D of Clostridiumthermocellum (Joliff et at., 1986a); BfCedl, cellodextrinase of Butyrivibriofibrisolvens(Berger et a!., 1990). The sequences are numbered from the N-terminal amino acids ofthe mature enzymes, except where marked A In these cases the numbering is from thestart of the signal sequence because the sites of processing have not been determined.3940CfCenC: 300 GYEPUrG-PRVRVNQVGYLPFGPKRATLVTDAAEPVAWELRflJDGVVVAD3TSEPR*: ::: ::*::*: :*:* *: :* :*PfidA: 2 9** :* *:*.: * * :* * :*** *: *CtCe1D: 44’ ITBYQRIRLNSIGFIPNHSKKA- --TIA?\NCSTFYVVKEDTIVYTGTATSM*BfCed: 38 AFtX3KVEHFGT----CfCenC: GVEPSAAQAVHVLDFSDTQGAG’X7LVAETSRPFDIDLYQQLRYD.LNYFY• * .* *. *. • ..* .*.*..* .**. •. . **PfEndA: GSDASSGLZIHIDLSSVTATGSGFrLIVGGDSSYPFSISSAFYDALKYF1: * * * : ::: :* * : : :**CCCe1D:* ** ... . * . .. . .* .*BfCed: --DEISGED1’YVADFSLTEEGKYKI-VADGQESVLFSISNDAYDKI21KDICKcFYCfCenC: LARSGI’EIEADVVGEE---YA REAGHVGVAPNQGLYIWPCIGPRDYYD3-* . .. . ...*PfEndA: HSGIAIE1’flI’GGGGSYASMSPWSRPAGHLNQGANKGDMNVPCWSGTcNYS--:*** : :*CtCe1D: ILRCGTSVSA--TYNG-IHY SM GPCHDAYLt71I**. * • • • . •BfCed: YLRCGDALSX--EFAG-EYY- H KPCHTI’EA1V--YCfCenC: WrCDYRLDVSGYDAGDHGKYVVNGGIAVGQLLQrYERALHAGTAD-LAIXTLD*. ..* • *••.**PfEndA: LNVTKGDAGDHGKYVVNGGISVPLt2’IYERAQHITGNLAAVA3SMN*. ** *. • * **CtCe1D: NGQHTKKDSTKDAGDYNKYVVNAGITVGSMFLAWE---H--FKEQ-LEPVALE*. **. * .. .* . *BfCed: GEDVEPVDGGWHDGDYGRYSTAGAVAVANILYGVR---F- -FKGL-LD-VHYDCfCenC: VP---DVPDVLDEARWELThLSMIVPEG-EYAGN*XVFDEGWTGLPLLP*.*..*. .. . *PfEndA: IP---ESGNGVADILDEARWQMANQVPQGQAKAGMAHHICIIWVGW1’GLPLAP** ****•*. . . . *. • .* * • *CtCe1D: IP---ESIPDFLDELKYEIEThTMQYPIDG-- -SGRWHKVSTRNFGGF-IMP.*• •*•** .. . • ...*. . . *•.BfCed: IPK/AGDKLPEILAEVKVELDFL1QRG---S4--H?ITFNHAPF-LMPC fCeriC: ADDPQARSLPSTAATLNLSAVAAARLLEPYDPQLTLLEAARPTWAAQEH* * * . . . .. *.PfEnA: HDPQQRALVPPSTAATULAATAAQAARIWKDID1.GFAALCLTAAERAAAQAN:: * ** . *• • • **CtCe1D: EDERFF’JWSSAATADFV7TTAMAARIFRPYDPQYAEKCINAAKVSYEFUN* : :* :*: ..*: * * **. *:BfCed: DDREELFLFSVSSLATADIAVFALAYTTYDAEYADKU1QKSLLAYKWLLtJ41CfCenC: PAL-YAPGEAGA3GGAYNDSQVADEAELYLTI’GEDAFATA---VflSPLH• . • *..•* * * *PfEridA: PND-IYSGGGGG-YGDRFDEFWAAAELYITrGDSRYLFr---INNYTLE* : : ::* :::::** :: *CtCe1D:• * • • .•• **•* • .BfCed: PDELLFRPGQ-YDFDISRFACALYEATSrCzYSDAQEL}RLEECfCenC: TADVFI’AD GFGGSVAALGRLDLTVPNELPGLDAVQSSVVEGAQEYL** *PfEndA: RTDFGWPD TELL MSLAWPATHTNSLRIAAPHWIASTHLTS* • ** * *.*: *CtCe1D: FSKKIEAD FVNIJWfYLLSERPGKNPALVQSIKDSLLSTAD* •. • . • • **BfCed: FDKNAQ1GYQGECLAEVAGLGSLSLLUR--EEN-ALCSLARNSFVEDCfCenC: AAQAGQGFGSLYSPPG-GEYW3SSSQV?NNLVVVATAYDLTGDERFRAATLEGLD** . . . *. •PfEndA: ASGYPAPLSSLE YSNSVINKLVLNGLYDFSGNQNFALGVSKGIN* * : * *** *: • :* *CtCe1D::*: *: * :: :*: *: *: *BiCed: RLVVSGFGLGDFILUC1LAIRIPEYKLALEGLDCfCenC:* *.*.PfdA: YLFGSNVLSTSFITGLGVAQPHPFtZAGALNSNYPWAPPGALSGGPN---AGL• .** *.. . *..CtCe1D: HVFGBNYRSYVrGrJ3flJPPt?HDRR-SGtX--IWEPWPWfLVGG* • . ... . * • *BfCed: YILXNSMDISYNGEKAPHRP-TAVDD--IEEV1PGLVSGGPNSGLFUEC fCenC: 881* * .* ..* .**.** *.*.•.* •*•. * • **PfdA: EDSLSASPLSGCTSRPATCWDSIWSEITflAPLWVLGFYNDFAAT 604”* * • * * **CtCe1D: GWPGPVDIQDSYQTNEIAINZAALIYALAGFVNYNSA 577”*. ..* .**..*• *BfCed: RAQTLRGK- ---GLPPMKCYIDHIDLYSLNEITIYSPLVFALSGILE 547”42(Fig. 11,12) (Coutinho etal., 1991). Similar considerations apply to three recentadditions to family E, EndAFS from Fibrobacter succinogenes AR1(Cavicchioli eta!.,1991) and enzymes El and E4 from Thernionosporafusca (Wilson, D.B., 1992).Domains in Cex, CenA and CenB are separated by linkers, short sequences rich in prolineand/or hydroxyamino acids (O’Neill ci’ al., 1986; Wong et a!., 1986; Meinke eta!.,1991a). In CenC, the sequences SLTTSATPPP (amino acids 292-299) and PVPTAP(amino acids 983-988) linked the N-terminal and C-terminal repeats, respectively, to theremainder of the polypeptide (Fig. 8). Amino acids 809-818, with the sequencePSLPSPPPGS, may delineate a domain of -60 amino acids adjacent to the C-terminalrepeats (Fig. 8).3.2. Expression of CenC in E. coliThe fragment of C.fimi DNA in pTZ18R-8/5 was manipulated to increase thelevel of expression of cenC before it was learned that the gene lacked its 3’ end (Moser,Ph.D. Thesis, 1988; Coutinho eta!., 1991). In pTZ18R-8/5, the Plac on the vector wasseparated from the 5-end of cenC by some 600 bp of C.Jirni DNA containing severalinverted repeats, which could interfere with transcription and/or translation in E. co/i.Furthermore, C.firni promoters for other cellulase genes function poorly in E. co/i(O’Neill et a!., 1986; Wong eta!., 1986; Owolabi et al., 1988). Therefore, thepresumed coding sequence of cenC, except for the GTG translational start codon, wasfused to an ATG codon at the 3’-end of a portable translation initiation site (PTIS), and thefusion put under the control of Plac on pTZ-1 8R to give pTZP-cenC (Moser,, 1988;Coutinho eta!., 1991). This reduced the size of the C.finii DNA insert to about 3.6 kbp.Subsequently, the 3’ end of the gene was subcloned from pTZ-JC1 into pTZP-cenC togive pTZ-JC2 (Fig. 7, Coutinho et at., 1991).Fig. 13 (top panel) shows the expression of pTZ- 1 8R as the control (with noinsert), pTZP-CenC and pTZ-JC2 respectively. While the level of expression of CenC asFigure 11. Consensus sequences around the DAGD and H-R peptides in subfamilies Eland E2. Symbols as for Fig. 9. 0 and B in the amino acid sequences denote (aromaticamino acids and histidine) and basic amino acids, respectively. Enzymes names are as inFig. 10.4344SUBFAMILY El:CfCenC: 457 WrCDYRLDVSGGWYDAGDffGK’xVVNGGIAV 486PfEndA: 186” GTCNYSLNVTKGWYDAGDf-{GKYWNGG ISV 215”CCCe1D: 184” NGQHKK1DSTKGWHTYGDYNKYVVNGITV 213”BfCed: 132” GEDVEPVDVTGGWHDGDYGRYSTAGAVAV 161”El CONSENSUS CWODAGEO-BY VSUBFAMILY E2:CfCenB: 42 ADVGDLTGGWDGDHVKFGFPMAFSA 71* **PaCe1: 65” 1XSSYNVDLVGG1fDAGLLKFGLPMATr 94”* . ...*_.. * **D270 -6: 68” KNGIYNLSGGYFD?GDVKFGLPMAYSM 97”CtCe1Z: 60” LX3ADVGLDLTGGWYDGDHVKFNLFMAYSQ 96”E2 CONSENSUS L-GG000AGD--KF--PMAO--E1/E2 CONSENSUS GOODaGO-SUBFAMILY El:CfCenC: 778 GAIQSYGEVASHQQHSIFAHQLDPSLPSPPPGSLGG 822*PfEndA: 502” C STFITGtJ3TAQP ALl SNYPPPGALSGG 546”CtCe1D: 495” GRNTfNRS TGLGINPPMNPHDRR-SGAtX--IWEPWPGYLVGG 536”• . •.. . * . *BfCed: 453” GCNSMDISY ONGEKAFPHLRP-TAVDD-- IPWPGLVSGG 494”El CONSENSUS G-N----SO-TG-G H-R PG---GGSUBFAMILY E2:CfCenB: 356 G4PRSSSYVVGFCANPPTAPNHRTAHGSWLDSITPAQSRHVLY 400• . . . .*. • • • . .. . * *PaCel: 390” GQNPA SYMVGFGERYPQHVHHRGSSLPSVQVHPNSIPQAGFQ 434”• .. •** ... * * * . *t270-6 369” C PNQQSFVGMGPNYPIPH AAHS1TNDI1NWNLYLLK 413”*.. • * * .CtCe1Z 379” ALSSGRSYVVGFGVNPPUPHHRTAHSSWADSMSVPDYHPHVLI 422”E2 CONSSUS SO-VG-G---P---HHP.E1/E2 CONSENSUS SO--G-G H-R45Figure 12. Family E of 13-1 ,4-glycosidases. The numbers refer to the positions ofamino acids. i’’’’ catalytic domains of subfamily El. “‘‘i catalytic domains ofsubfamily E2; sequences common to both types of catalytic domain;repeated sequences; linker sequences; ‘ cellulose-binding domains.CfCenC, PfEndA, CtCe1D and BfCedl as in Fig. 8. CfCenB, endoglucanase B of C.fimi (Meinke et a!., 1991a). CsCe1Z, endoglucanase Z of Clostridium stercorarium(Jauris et al., 1990). Dd270-6, spore germination specific protein of Dictyosteliumdiscoideum (Giorda et at., 1990). PaCel, endoglucanase of Persea americana (Tuckereta!., 1987).CfCenC1503004908109801069PEndAI•IH150935CICoID150820BfCod[105520F4W///A1Ir45380CfConBCsCeIZDd270-6PaCel1012LW////AIi40956ireeezsj.40680I rA%%.I3546547Figure 13. Expression of CenC in E. coli. Top panel. Arrows indicate the location ofCenC in JM1O1/pTZP-CenC and JM1O1/pT’Z-JC2 on a Comassie stained SDS-PAGE gel.JM1O1/pTZ-18R (no insert) was used as control. Cultures were grown in the absence ofinducer IPTG at 300 C. Cells were harvested and fractionated as described previously(Moser, 1988). and the fractions were analyzed by SDS-PAGE. pf: periplasmic fractionobtained by osmotic shock (Nossal and Heppel, 1966); cf: cytoplasmic fraction (obtainedby French press of osmotically shocked cells) and Ce: whole cell extract (obtained byFrench press of whole cells). Bottom panel. Western blot of identical gel probed with antiCenC serum showing a large number of degradation products, specific to CenC, as thesewere not present in the fractions of the control culture JM1O1 carrying the plasmid pTZ18R (Pharmacia) with no insert.48kDa205116WESTERN BLOTIAntI CenC—__— -m — — .- ———— —18Rpf ci ce p1pTZP JC2ci ce pt ci ce66362949observed on SDS-PAGE gels of crude cell extracts, was high, most of the CenCaccumulated in the cytoplasm in an inactive form (Moser, 1988). Furthermore thepolypeptide was very unstable as depicted by the Western blot of an identical gel (lowerpanel) probed with anti-CenC serum.In order to improve the expression a number of physiological parameters weretested which included: a) altering the growth temperature Fig. 14 (lanes 1, 2 and 3represent CenC purified from cultures grown at 25, 30 and 37° C respectively. b) theabsence of the inducer TPTG or its addition at A600 of 1.0 Fig. 14 (lanes 4 and 5respectively) and c) the age of the inoculum (lane 6 represents the initial starter culturegrown for less than 3 hours and A600 less than 0.4 and lane 7 represents the initial starterculture grown overnight). Lanes 8 and 9 represent CenC and CenC purified fromCfimi.Despite the difficulty in obtaining reproducible results for the expression of CenC,there were three physiological parameters that were ascertained over the course of theseexperiments that favoured the expression of CenC (i) Growth of the culture at 30° C (ii)The absence of the inducer IPTG and (iii) The use of a fresh starter inoculum. Theseconditions were used for all subsequent purifications. Two additional problems in thepurification of CenC were that some of the degradation products of CenC also hadcatalytic activity (much of the activity remained in the cell extracts even after Avicelbinding) and furthermore despite the amount of intact protein seen in crude cell extractsand the ability of the antibody to recognize the polypeptide, most of this material did notbind Avicel. The yields of purified CenC from a 10 1 culture varied from 20 j.ig to 100 .tgand this agreed well with previous results obtained in the lab (Moser Ph.D. thesis, 1988).50Figure 14. Production and purification of CenC from E. coli grown under differentphysiological conditions. Comassie stained, SDS-PAGE gel of the fraction from wholecell extracts, remaining bound to Avicel and desorbed with water. The location of fulllength CenC is indicated by an arrow. The lower bands are degradation products of CenCwhich retain the capacity to bind Avicel. Lanes 1, 2 and 3: Effect of growth temperature,25, 30 and 37°C respectively; Lanes 4 and 5: Effect of the absence or presence of theinducer IPTG, Lane 4, - IPTG and 5, + IPTG; Lanes 6 and 7: Effect of the age of thestarter culture, Lane 6, starter culture grown for 3 hours, Lane 7, starter culture grownovernight. Lanes 8 and 9: C3.2 and C3.1 are native forms of CenC purified from C.Jlmi..-c_-2 3 4 5 6 789-.--kDa205116661I.513.3. 5’ and 3’ deletions of cenCSequence comparisons with other cellulases suggested that the central domain(amino acids 300-880) was responsible for catalytic activity. Attempts were made to tryand see how much of the the N-terminal and C-terminal repeats could be deleted withoutloss of catalytic activity as assayed by CMCase activity on a Congo red plate. Fig. 15depicts the results obtained for the N-terminal deletions. Fig. 16 depicts the resultsobtained for deletions made in the C-terminal repeats of cenC.The properties of the C-terminal deletion mutants (Fig. 16) revealed deletions ofamino acids 954-1004 and deletion of amino acids 1022-1069 did not result in a loss ofCMCase activity. In order to test if the deletion mutants could still bind to cellulose andSephadex, the two constructs, CenC CAl and CenC CA2 were chosen for further study.A three step purification protocol employing cellulose affinity, anion-exchange and gel-filtration chromatography was used to purify whole CenC and the two C-terminaldeletions from their breakdown products.Fig. 17 shows the polypeptides obtained after affinity chromatography (CB:cellulose binding). The lower bands are breakdown products of CenC which still have thecapacity to bind cellulose. Figures 18 and 19 show the active fractions obtained, afterFPLC assisted ion-exchange chromatography on a Mono-Q anion-exchange columnfollowed by gel-filtration on a bio-gel matrix with a exclusion of 150 kDa, for CenC.Similar results using this purification protocol were obtained for the two C-terminaldeletions as well (Data not shown). Lanes 1-8 show (Fig.18) active fractions eluted fromthe Mono-Q column using a linear increasing salt gradient. A further purification step52Figure 15. N-teririinal deletions of CenC. CMC activity: Carboxyrnethylcellulase activitywas detected qualitatively, by Congo red staining, and then looking for the absence orpresence of halos around the colony. Dotted lines refer to internal deletions obtained byNuclease Bal. 31 deletions. Numbers refer to the amino acid sequence of mature(processed) CenC.CenC N11 38138 3001 28 167 300CMCActivity+CenC Nz3 [:ZZZZ__________CenCNi N2 Cl C2886 1069CenC NA2_+886 1069886 10691 11 498 886 106953Figure 16. C-terminal deletions of CenC. Captions as in Fig. 15. Deletion I was obtainedas a result of the initial cloning of CenC and deletion 3 was constructed by insertion of theomega interposon. The internal deletion, deletion 2, was constructed by excision of aKpnI-KpnJ fragment as described in materials and methods.CenC CA3CMCActivity+Ni N2CenCCenC CAlCenC CA2Ci C21 300 886 1069—%1 300 886 10221 300 886 / \ 1069954 1004++300 89754Figure 17. Cellulose-affinity chromatography of CenC, CenC CAl and CenC CA2,analyzed by SDS-PAGE. CB: cellulose bound. The location of CenC and its deletions areshown by an arrow. M: molecular mass marker. Lanes I and 2: CenC and CenC’ purifiedfrom C.fimi. Lanes 3,4 and 5 are fractions of CenC, CenC CAl and CenC CA2 obtainedafter binding cell extract to Avicel followed by washing of bound material with high saltand desorption with water.kDa M I 2 3 4 5205116 U29CB55Figure 18. SDS-PAGE of CenC purified by FPLC assisted anion exchangechromatography on Mono-Q. The location of CenC is shown by an arrow. Lanes 1-8 arefractions of active CenC eluted by an increasing linear salt gradient. Mono-Q: Anionexchange matrix (Pharnacia).3 4 5 6 7 8::- — —kDa 1 22051166641 -R—-MONO..Q56involving gel—filtration allowed the separation ot the top band (Fig. 19. lanes 1. 2 and 3)from other contaminatin degradation products (Fig. 19. lanes 4 — ). Fig. 20 shows thetinal products obtained for all three polypeptides. Polvpeptides encoded by both C—terminal deletion mutants retained the ability to hind both Avicel and Sephadex indicatingthat the presence of much of the C—terminal repeats (either amino acids 954—1004 or aminoacids 1022—1069) are not essential br bindin to these substrates.57Figure 19. SDS-PAGE of CenC after gel-filtration on P-150. Lanes 1-8 are fractions ofactive CenC eluted from the gel-filtration column. P-150: Polyacrylarnide bio-gel,150,000 mol. mass. size exclusion matrix (Biorad). The location of CenC is shown by anarrow.kDal 2345 678205 —116 — —. —664129P. 15058Ml 2 3Figure 20. SDS-PAGE analysis of intact CenC (Lane 1) and its two C-terminal deletionsCenC CM (Lane 2) and CenC Cz\2 (Lane 3). M: Molecular mass marker. The location ofCenC and the two deletion derivatives are indicated by an arrow.kDa20511666593.4. Kinetic parameters for native and recombinant CenCThe kinetic parameters for hydrolysis of CMC by native and recombinant CenCare presented in Table 2A.and compared to those of CenA and CenB. CenC is the mostactive component of the C.fImi cellulase system isolated to date (Moser et a!., 1989;Coutinho eta!., 1991). The hydrolysis of 2-chloro-4-nitrophenyl cellobioside by nativeCenC is compared to Ce1D (endoglucanase from Clostridiurn therrnoce!lum ) and arepresented in Table 2B.The oligosaccharide cleavage patterns for CenC on the paranitrophenol derivatives:2-chloro-4-nitrophenol lactoside; 2-chloro-4-nitrophenol cellobioside (pG2), trioside(pG3) and tetraoside (pG4) and on the cellodextrins G3-G6 (detected by HPLC) areshown in Fig. 21. The lower panel shows a sample chromatograrn for G6 cleavage inwhich the sample was injected onto the HPLC column, immediately after enzyme andsubstrate were mixed together. pG2 and the 2-chloro-4-nitrophenol lactoside were cleavedat Gi but pG3 and p04 were not. The addition of cellobiose to the reaction mixtureinhibited the cleavage at Cl of pG2. CenC gave several products from cellohexaose andcellopentaose.The generation and interconversion of the anomeric forms of G3 indicated thatCenC cleaved the 13-1 ,4-glucosidic bond with inversion of the anomeric configuration.The theoretical cx-OH: 13-OH ratio at G3 after hydrolysis of cellohexaose or cellopentaoseshould be greater than 1 shortly after cleavage (before the two anomeric forms reachequilibrium) if it is an inverting enzyme and less than 1 if it is a retaining enzyme. Themeasured ratio was 2.2 indicating the change in configuration from f3 to cx. Racemizationis complete within about 40 minutes and the measurements were made at intervals between5 and 20 minutes (by integrating the area under the peak). CenC cleaved cellotetraosepreferentially to yield cellobiose. G3 was hydrolyzed to G2 and 01 but the preferred pointof hydrolysis (Cl or C3) could not be deten’nined by HPLC. The temperature and pHoptima of CenC on CNPC were 45CC and pH 5.0 respectively (Fig.22).60TABLE 2A. Kinetic parameters for hydrolysis of CMC by native andrecombinant endoglucanases.Enzyme Source *KmCenC” C.fimi 0.15 ± 0.02 12.33 ± 0.21CenC’ C.fimi 0.10 ± 0.02 11.39 ± 0.31CenC E. coli 0.13 ± 0.03 12.69 ± 0.44CenC’ E. coli 0.30 ± 0.03 14.86 ± 0.30±CenA E. coli 0.17 ± 0.10 2.76 ± 0.050CenB E. coil 0.51 ± 0.05 2.87 ± 0.12Initial velocities for the production of reducing sugars at 30°C were determined withHBAH over a substrate range of 0.8 to 13.0 Km. Values were derived by weighted linearregression analyses of Lineweaver-Burke plots.*Km values: mgCMC/ml+Vm values: imoles reducing sugar produced per nmole protein.“The author acknowledges the help of B. Moser for these measurements.± From Langsford eta!, 1984.o J.B. Owolabi, personal communication.TABLE 2B. Kinetic parameters for hydrolysis of CNPC by CenCand Ce1DEnzyme Km (riM) Kcat (s1) KcatlKm (s’M’)CeLD 98 7.9 80,000CenC 479 60.6 126,600Activity measurements were performed at 2YC, pH 5.0 with CNPC (100-1000 pM) asthe substrate. Kinetic parameters were derived from (S/V) vs (S) plots (Hanes, 1932).The values for Ce1D from Clostridiuni therniocellum were reported previously Tomme etal., 1991).Figure 21. Oligosaccharide cleavage patterns of CenC for various cellodextrins. Cleavageof the p-nitrophenol conjugated cellodextrins were followed by the absence or presence ofcolor. Cleavage sites for the non-conjugated cellodextrins was monitored by separation ofthe cleavage products by HPLC and the separated products monitored by a diffractometer.A sample chromatograrn showing the separation of the cleavage products, including theseparation of the alpha from the beta anorner for cellodextrins G3 and G4, on the HPLCis presented in the lower panel.6162EEEFLEJEEIL-L-L-L1• p-nitrophenolglucose residueA galactose residuef3-1 ,4-glucosidic bondcleavage site6.0005.00G3xG44.003.00 -______________ _______0.50 1.00 1.50 .Ofl 2.5u:<11jLFigure 22. Temperature and pH optima for the hydrolysis of chioronitrophenylcellobioseCNPC at Cl by native CenC. The reaction, monitoring the release of phenol, wasfollowed spectrophotometrically at an absorbance of 400 nrn. The enzyme had atemperature optimum of 45CC and a pH optimum of 5.5636430 -204.0 5.0EE>C>100—3.0160 —1207.06.0pH8.0 9.080 -400-20 30 40 50 60Temp (C)653.5. Expression of fragments of the cenC gene.The N-terminal repeats of CenC were designated Ni (amino acids 1-148) and N2(amino acids 149-296); the C-terminal repeats were designated Cl (amino acids 886-975)and C2 (amino acids 976-1069). Amino acid sequence comparison revealed that the N-terminal repeats (N1N2) were distantly related to the cellulose-binding domains of Cex,CenA and CenB (Fig. 29). Sequence comparisons of the C-terminal repeats (C1C2) didnot reveal any obvious siniilarites to any other cellulase sequences reported to date.Deletion mutants of the cenC gene were constructed which encoded the leader peptide pluseither Ni (amino acids 1-150), N1N2 (amino acids 1-299), Cl (amino acids 860-993) orC1C2 (amino acids 860-1069) (Fig. 23).Polypeptides corresponding to Ni and N1N2 were detected on Western blots offractions from cultures of E. coil JM1O1 carrying plasmids pTZ-JC6 and pTZ-JC3,respectively, using anti-CenC serum (Fig. 24). The low molecular mass band seen in cellextracts of E. coil JMIO1/pTZ-JC3 was also seen in the E. coil JM1O1 pTZ-18R controlcell extract and was considered to represent the non-specific interaction of the antiserumwith an E. coil protein. Polypeptides Cl and CIC2 reacted poorly with anti CenC’ serum(data not shown); however they were detected as unique bands by SDS-PAGE inappropriate E. coli extracts following induction with IPTG (Fig. 25). The extinctioncoefficients for N1N2 and Ni as determined by UV absorbance (Scopes, 1974) were1.06 and 1.10 respectively, in good agreement with the values predicted from thetryptophan and tyrosine content (Cantor and Schimmel, 1980) of the proteins (1.21 and1.20 respectively). The apparent molecular masses of N1N2 and Ni (20.0 and 40.0 kDa,respectively) were significantly greater than those predicted from their correspondingDNA sequences (15.4 kDa and 30.4 kDa) (Fig. 26). The predicted pT’s of Ni and N1N2(3.54 and 3.48, respectively) showed that both are highly acidic polypeptides and thismay account for the discrepancies between their predicted and apparent molecular masses.Similar considerations apply to intact CenC (Coutinho ci’ at., 1991).66Figure 23. Polypeptides encoded by the various plasmids. The plasmid encoding CenCor the CenC polypeptide is shown in parenthesis. The locations are indicated of therestriction endonuclease sites used to construct the various plasmids. , N-terminalrepeats; , C-terminal repeats. The numbers refer to the amino acid residues,starting from the N-terminus of mature CenC (Coutinho et al., 1991). Thetranscriptional and transalatiorial stops were from interposon omega and the promoterused was the lac promoter.CenC (pTZ-JC2)Apal SadSmal//1 148 300Apal’ MM\/886 975 1069N1N2 (pTZ-JC3)Ni (pTZ-JC6)1 150Cl C2 (pTZ-JC7)860 975 1069860 993Cl (pTZ-JC14)67Figure 24. Intracellular and extracellular distribution of polypeptides N1N2 and Nisynthesized in E. coli JM 101 by Western blotting. Polypeptides from cell extracts (ce) andculture supernatants (cs) were analyzed by SDS-PAGE followed by Western blotting anddetection with polyclonal anti-CenC’ serum. Plasmids pTZ-JC3 and pTZ-JC6 encodeN1N2 and Ni, respectively; pTZ-18R is the control plasrnid, without cenC DNA. Ni wasdetected wealdy by the antibody. The low molecular weight band seen in cell extracts ofE. coil JMiO1/pTZ-JC3 was also seen in theE. coil JM1O1/pTZ-18R and wasconsidered to represent the non-specific interaction of the antiserum with an E. coilprotein.JC3 JC6 18Rcecscecs ceFigure 25. Intracellular and extracellular distribution of polypeptides C1C2 and Clsynthesized in E. coli JM1O1. Polypeptides from cell extracts (ce) and culture supernatants(cs) were analyzed by SDS-PAGE, 6h after induction with 0.1 mM IPTG. The left lane ofeach pair corresponds to induction when the A600 of the culture supematant reached 0.4;the right corresponds to induction when the A600 reached 1.0 Polypeptides C1C2 andCl, indicated by arrows, are encoded by pTZ-JC7 and pTZ-JC14 respectively. Molecularmass markers are shown in the far left lane.JC7 JC14ce Cs ce- —I•“68CskDa116663629 4’Figure 26. Purified N1N2 and Ni obtained by affinity chromatography. A. Polypeptidesfrom E. coli JM1O1 cell extracts (ce) and culture supernatants (cs), and the purifiedpolypeptides (pp) analyzed by SDS-PAGE. Plasmids pTZ-JC3 and pTZ-JC6 encodeN1N2 and Ni respectively; pTZ-18R is the control plasmid, without cenC DNA. N1N2and Ni were purified by affinity chromatography on Sephadex G-50 and Avicel PHiO1,respectively, as described in Materials and Methods. B. SDS-PAGE analysis of 10, 20and 40 .tg of Ni purified from E. coli JM1O1 (pTZ-JC6), and 10 and 40 jig of N1N2purified from E. coli JM1O1 (pTZ-JC3), to evaluate levels of contamination with otherpolypeptides. The far left lane of each panel shows the molecular mass markers.6970ppJC618R JC3kDa ce Cs pp cs97664129AJC3kDa JC69766 —. *.4129 — j-—.BThe apparent molecular masses of Ci and C1C2 (16.0 and 28.0 kDa, respectively)agreed well with those predicted from their corresponding DNA sequences (16.2 and 26.4kDa, respectively); their predicted p1’s are 5.07 and 8.76, respectively. N-terminalsequence analysis revealed that Ni, N1N2, Ci and C1C2 all had the same N-terminus(ASPIG), as expected (Fig. 23). Ni, N1N2, Ci and C1C2 were devoid of detectablecarboxymethylcellulase activity (Data not shown).71723.6. Binding of the N- and C-terminal repeats to cellulose and to Sephadex.Polypeptides N1N2 and Ni (Fig. 26 A) both bound to Avicel (a heterogenousmicrocrystalline cellulose preparation containing both crystalline and non-crystallineregions), but only N1N2 bound to Sephadex G-50. Neither C1C2 nor Ci bound to Avicelor to Sephadex G-50 (data not shown). N1N2 and Ni were purified to virtualhomogeneity by affinity chromatography on Avicel or Sephadex G-50 (Fig. 26B).The binding specificities of N1N2 and Ni for two different cellulose allomorphswere investigated by determination of their adsorption to PASC and to BMCC. Theadsorption of the CBD of Cex (CBDCex) was also included for comparison. CBDCex(11.1 kDa) had an extinction coefficient of 2.30, in agreement with its predicted value (E.Ong, Ph.D. thesis). Adsorption of the three polypeptides to the two forms of cellulose isdescribed by isotherms bound [B] vs free [F]) shown in Fig. 27. The partition coefficient(the initial slope of the adsorption isotherm) was determined as a measure of the relativeaffinities of Ni, N1N2 and CBDCex for a particular cellulose allomorph (Table 3). Allthree polypeptides adsorbed to regenerated cellulose (Fig. 27).The partition coefficients for N1N2 and NI for regenerated cellulose (Table 3) weresignificantly different (approximately three fold lower) than the corresponding values forCBDCex with a level of confidence of> 95% (the difference in the partition coefficientsbetween the two means was greater than 2.5 standard deviations). Saturation ofregenerated cellulose was not obtained with Ni, N1N2 or CBDCex at the highestpolypeptide concentration tested. In contrast to the adsorption of N1N2 and Ni toregenerated cellulose, there was no apparent adsorption of either polypeptide to bacterialcrystalline cellulose; however, CBDCex adsorbed well to this substrate (Fig. 27, Table 3).There was no detectable binding of the C-terminal repeats (Ci C2 and Ci) to eithersubstrate confiririing what was observed earlier, in that C-tenriinal deletions of CenC,retained both catalytic activity and binding to cellulose and to Sephadex.Figure 27. Adsorption isotherms for N1N2, Ni and CBDCex. Adsorption of the variouspolypeptides to PASC (Panels A,B,C (i) ) or BMCC (Panels A,B,C (ii)) was determinedas described in Materials and Methods. [F] is the free polypeptide concentration (tiM) and[B] is the bound polypeptide concentration (jimol per g cellulose) at equilibrium. PASCrefers to phosphoric acid swollen cellulose and BMCC refers to bacterial microcrystallinecellulose.73A(i)CBDc0-PASC...•.••I60504030201000 50 100 150 200 250 300B(i)N1N2-PASC.... ....•••6050403020100A(ii)q)0E0aS0ECSa-CB0X -BMCC0 0 00 0fO0 20 40 60 80 100B(ii)NIN2-BMCC20151050201510500C(ii)20 40 60 80 10020150E:i 100CSa- 500 20 40 60 80 100C(i)N1-PASC..•••I0 50 100 150 200N1-BMcc201510500 50 100 150 200[P] iM[P] jiM74Table 3. Partition coefficients of N1N2, Ni and CBDCex for PASC and BMCC. PASCrefers to phosphoric acid swollen cellulose and BMCC refers to bacterial microcrystallinecellulose.Partition Coefficient(L.g1)1lnitial slope of adsorption isotherm plot([P1 ad VS. [P1).PolypeptidePASC BMCCCBDCex 0.598 0.967N1N2 0.203 0.001Ni 0.236 0.001753.7. Sequence relatedness between Ni, N2 and other CBDsf3-1,4-glycanases can be grouped into families according to amino acid sequencerelatedness within their catalytic domains (Henrissat et at., 1989; Beguin, 1990; Gilkes eta!., 1991a). As more amino acid sequences are deduced from the nucleotide sequences ofgenes, it is becoming clear that there are also families of CBDs (Gilkes eta!., 199 1a. Atpresent, however, some of the families contain only one or two members. The CBDswithin a family are quite uniform in size and amino-acid sequence, but there areconsiderable differences between the families (Table 4). Since the catalytic domainfamilies are designated by letters, a different designation should be used for the CBDfamilies; Roman numerals are used in this study (Coutinho et a!., 1992).Ni and N2 of CenC, and the T. reesei EglIl sequence, could be in a sub-family offamily II (Table 4, Fig. 28). Although there is little overall sequence identity between Ni,N2 and the other CBDs, all of them contain two cysteines —iOO amino acids apart, witharomatic, aliphatic and hydroxyamino acids at the same or similar locations between them(Fig. 29). However, the other CBDs are —l06 amino acids long, whereas Ni and N2 are—150 amino acids long. The dendogram also suggests that the D. discoideurn 270-ilsequences a, b are in a third sub-family (Fig. 28).Table4Familiesofcellulose-bindingdomainsFamily-OrganismEnzymeAminoacids*TerminusBinding#ReferenceI IIHurnicoingrisenvar.thermoideaCbhl36C1PhanerochnetechrysosporiumCbhI35C2TrichodermareeseiCbhI36C+3T.reeseiCbhII36N+4T.reeseiEglI36C5T.reeseiEglIl36N+6TrichodermavirideCbhI36C7ButyrivibriofibrisolvensEnd!95C8CellulomonasfirniCenA106N+9C.firniCenB103C+10C.firniCex106C+11C.fimiCenCNI148N+12•CenCN2148ICellulo,nonnsfinvigenaC1fX106C13Dictyosteliumdiscoideum270-Ha98C14270-lIb106IMicrobispornbispornCeIA100C15PseudornonasfluoresceusEgIA100C16subsp.cellulosnP.fluorescenssubsp.cellulosaEgIB102N+17P.fluorescenssubsp.cellulosaXynA101N+18P.fluorcsccnssubsp.ccllulosaXynB/C99N+19TherruornonospornfuscaE286C+20T.fuscnE5103N20T.reeseiEglIl90I6,21Table4Familiesofcellulose-bindingdomains(cont’d)FamilyOrganismEnzymeAminoacids*TerminusBinding#ReferenceIIIBacillussubtilisDLGEnd132C22B.subtilisN-24End132C23B.subtilisPAP115End132C24CaldocellurnsaccharblyticumCeIB136I25,26C.fimiCenB131I+27Clostridiurnstercorariurn+Ce1ZC144I+28Ce1ZC133CIVClostridiumtherrnocel!urn°CelE2401+29,30VErwiniachrysanthemiEgZ63C+31*Nljmber ofaminoacidsinbindingdomain;±CandCareina50kDaIragmentelCeIZwhichbindscellulosehutiscatalyticallyinactive;#bindingofthedomaintocellulosehasactuaHybeendemonstrated;Othebindingdomainissomewherewithinasequenceof240aminoacids.N:bindingdomainisattheN-terminus;C:bindingdomainisattheC-terminus;I:bindingdomainisinternal.References:(1)Azvedociat.,1990(2)Simsciat.,1988(3)Shoemakerciat.,1983(4)Chenciat.,1987(5)Penttilactat.,1986(6)Teenetat.,1987(7)Chengciat.,1990(8)l3ergerci(it.1989(9)‘vVongciat.,1986(10) Meinkeetal.,1991a(11)O’Neillcial.,1986(12) Coutinhoca!.,1991(13)AlTawheed,1988(14) Giordaeta!.,1990(15)Yablonskyetat.,1988(16)1-lallandGilbert,1988(17)Gilbert cia!.1990(18)Hall etal.,1989(19)Kellettetal.,1990(20)Laoeta!.,1991(21)Meinkecia!.,1992(22)RobsonandChambliss,1986(23)Nakamuracia!.,1987(24)Mackaycia!.,1986(25)SauletnI.,1989(26)Saulciat.,1990(27) Meinkeetat.1991b(28)Jaurisetal,,1990(29)Hall ciat.,1988(30)Durranteta!.,1991, (31)Pycia!.,1991.78Figure 28. Dendogram of Family II CBDs. Complete sequence data are contained in thereferences listed in Table 4 (Bfi, B.Jlbrisolvens: Cfi, C.fimi; Cfl, C.fiavigena; Ddi, D.discoideum; M bi, M. bispora; Pfl, P. fluorescens su bsp. cellulosa; Tfu, T. fusca; Tre, T.reesei.).Cfi CenC NiCfi CenC N2Tre Egi IICfi CenAMbi CeIACfl CenBCfl CIfXCfi CexTfuE2Tfu E5Ff1 EgIAPfl EgIBPfl XynAPfI XynBFf1 XynCSf1 End IDdi 270-1 laDdi 270-1 lb79Figure 29. Amino acid sequence similarities between Ni, N2 and other CBDs from C.Jimi. Boxes enclose three or more identical amino acids;• indicates similar amino acids.The numbers at the ends of the sequences refer to the positions of those amino acids in thesequences of the mature enzymes. Amino acids said to be ‘similar’ are: A,S,T; D,E;N,Q; R,K; I,L,M,V; F,Y,W.CeriCNi 33 C-V--APAGSAQ-YGVG-VLN_GVAIEEGTTY• • • ••CenCN2 181 C-V--DLPGGQGNPWDAGL-VYN_GVPVGEGESYV• • F-i •CenA 4 CRVDYAVTNQWPGGFGANVTITNLGD_PVS_SWK•• • • •Cex 341 C Q V L W G V - N Q W N T G F T A N V TI!IK N T S S A P V D - - C W T• • •CenB 913 CTVVYS-TNSWNVGFTGSVKITNTTTPLT___WT•• • • •CenC Ni L R Y T A T A S T DV R A L V G Q N G A P Y G T V L D - P 5 P A L• • • • • ••CenC N2 L S F T A S A T P D M P V R V L V C E C G C A Y R T A F EG S A P L• • • •• • •CenA L D W T Y T A G Q R I Q Q L W N G T A S T N G G Q V• •• • •Cex LTFSFPSGQQVT QAWSSTVTQSQS V•• • • • • ••CenB LGFAFPSGQQVT QGWSATWSQTGT V•• • • • • ••CenC Ni T S E P R Q V T E P F - - - - T A S A T Y P A T P A A D D P E C Q I A• • •CenCN2 TGEPATREYAF____TSNLT[P__PDGDAPGQ ACenASVTSLPWNCSIPTGCTASFFFNGSWACSNPT_ PASCex TVRNAPWNGS IPACGTAQFGFNGSHTGTNAA- PT•• • • •CenB TATGLSWNATLQPCQSTDIGFNGSHPGTNTN_ PAS•• •• • ••CenCN1 FQLGGFSADAWFC 140CenCN2 FHL-GK-AGAYEFC 285CenA FSLNG TTC 103•Cex FSLNG TPC 440•CenB FTVNG EVC 1011••803.8. Expression of N2A deletion mutant encoding N2 was constructed by linearizing the fragmentencoding N1N2 at the unique BamHI site followed by digestion with NucleaseBal3l(Fig.30). The polypeptide could be purified by affinity adsorption to Avicel (Fig. 31) butnot to Sephadex G-50 (Data not shown). N-terminal sequence analysis revealed that N2had the expected N-terminus (ASPIG) (Fig. 30). The polypeptide (17.3 kDa) had anextinction coefficient of 1.09 as determined by UV absorbance (Scopes, 1974) and ingood agreement with the value predicted from the tryptophan and tyrosine contents(Cantor and Schimmel, 1980) of the polypeptide (1.18).3.9. Adsorption of N1N2, Ni and N2 to PASCFig. 32, Panel A shows the relative adsorption (pg/rng) of N1N2, Ni and N2 onPASC. The capacity of NIN2 and N2 to adsorb to PASC was greater than Ni. Panel Bshows the free polypeptide in solution when N1N2 and N2 are compared with each otherwith N1N2 having a higher capacity to adsorb to PASC. Panel C shows competitionbetween N2 and Ni with N2 having a slightly higher capacity of adsorption over Ni.The relative adsorption capacities of the three polypeptides in the presence of one another(based on Jig/mg of PASC) are N1N2 > N2 > Ni. Table 5 summarizes the results, usinga densitometer, to scan the SDS-PAGE gels for the relative amount of free polypeptide leftin solution post-binding. It must be understood however that on a molar basis, becauseN1N2 is about twice the size of Ni or N2, it would have half the number of molecules.The assumption in the above experiment that the number of binding equivalents is thesame for all three polypeptides, with N1N2 having 2 and Nl or N2 having 1, does nottake into account the fact that steric hindrance between the binding sites in N1N2 mightaffect the adsorption process. The experiment should be redone using equivalent molaramounts of the three polypeptides which would make up for the difference in the numberof molecules. None of the Ni deletion derivatives bound to Avicel (Fig. 33).81Figure 30. Polypeptides encoded by the various plasmids. The plasmid encoding eachpolypeptide appears in parentheses. The locations of the restriction endonuclease sitesused to construct the plasmids are indicated. The stippled regions indicate the N-terminalrepeats. The dotted lines indicate the extent of deletion using Nuclease Ba131. Thenumbers refer to the amino acid residues of CenC.N1N2 (pTZ-JC3)Ni (pTZ-JC6)I138ISON2 (pTZ-JC34)167 299Figure 31. SDS-PAGE analysis ol purilied Ni NI and Xl N2 . Polypeplides from E.uuii Thl 101 culture supernatants (cs) were puriIc1 b affinity chromatography onAvicel.The far left Jane shuw the molecular mass marlers.kDa N2 Ni N1N243•29181483Figure 32. SDS-PAGE analysis of the adsorption of N1N2, Ni and N2 to 1 mg of(PASC). From left to right, increasing concentrations of polypeptide in paired lanes(Materials and Methods): Open circles are control samples incubated without cellulose andrun on a gel to determine the amount of starting material and closed circles are sampleswhich were run after adsorption to cellulose to ascertain the amount of polypeptideremaining in supernatants after binding. Panel A: N1N2 vs Ni vs N2. Panel B: N1N2 vsN2. Panel C: N2 vs Ni. PASC: Phosphoric acid swollen cellulose. The far left laneshows the molecular mass markers. An estimate of the relative amount of free polypeptideremaining in solution, by densitornetric scanning of the SDS-PAGE gel, before and afterbinding, appears in Table 5.Ao• o•o•kDaBC43—___— N1N229———— —N2— — NiO•O•O•k Da———— N1N229%-—— —N2o•o•o•kOa29%—— —N 2•__-- NIPASC85Table 5: Adsorption of N1N2, Ni and N2 to PASC. Free polypeptide (% of startingmaterial) in supernatant after adsorption.Polypeptide concentration[25] [100] [250]Substrate % free polypeptide, after adsorptionb PolypeptidePanel APASCd N.D.C 20 60 N1N2N.D. 45 90 N240 85 90 NiPanel BPASC N.D. 10 40 N1N2N.D. 40 85 N2Panel CPASC N.D. 30 75 N225 65 90 Nia [Concentration] jig/mib Amount in supernatant before adsorption was taken as 100%c None detectedd Phosphoric acid swollen cellulose (regenerated cellulose)86Figure 33. Deletions of Ni. Restriction endonuclease sites used to generate the deletionsare included. Binding: As detected by adsorption to Avicel PH 101. Dotted lines refer tothe extent of each deletion. NI Al and NI A2 were created by truncation at Miul andBstEll respectively using the 2 fragment. NI A3 and Ni A4 were created by digestionwith BaniHi followed by treatment with Nuclease Ba131. Numbers refer to the amino acidresidues of mature (processed) CenC.BindingNi (pTZ-.JC6) +N1MN1A21 102 150N1A3N1A4— — —1 133 150j3I1501 34 64 150873.10. Consequences of substituting N1N2 for the CBD of endoglucanase CenAFig. 34 shows the cloning strategy for fusing the sequence encoding the catalyticdomain of cenA with the sequence encoding the cellulose binding domain of cenCresulting in plasmid pTZ-JC13 which encodes CBDCenC- CatCenA (C’ ‘A). The mutantenzyme retained both the binding properties of the CenC binding domain as well thecatalytic activity borne by the CenA catalytic domain. N-terminal sequencing of thepolypeptide confirmed the expected N-terminal sequence (ASPIG). The extinctioncoefficient (1 cm, 1 mg i1, 280 nm) for the polypeptide was determined by far UVabsorbance. The value of 2.12 obtained was in good agreement with its predicted value of1.81 (Scopes, 1974).Results with N1N2 showed that the CenC CBD had negligible affinity for BMCC.To test if the same held true for C’ ‘A, a mixture containing CenA (non-glycosylatedCenA produced in E. coil) and C’ ‘A were tested for binding to BMCC and then toPASC. Fig. 35A shows the removal of CenA from a mixture of CenA and C’ ‘A whenBMCC is used as the adsorbent followed in panel B by the removal of both CenA and C’‘A when PASC is used as the affinity adsorbent. These results confirmed that the bindingdomain of CenC retained its specificity (negligible affinity for BMCC) in the fusionpolypeptide C’ ‘A. The relative amount of free polypeptide remaining in solution(determined by scanning the SDS-PAGE gel) before and after binding is given in Table 6.The activities of CenA on a range of substrates were compared to those of C’ ‘A(Table 7). CenA produced in E. co/i is sensitive to hydrolysis in its linker region and isconverted into a 30 kDa fragment (p30) which has catalytic activity but lacks the bindingdomain (Gilkes et a!., 1988). The activity of p30 against the soluble substrate CMC washigher than that of either C’ ‘A or CenA. Higher activities for p30 were also observedagainst PASC and cellulose azure. The activity of p30 against Avicel was lower than thatof C’ ‘A or CenA. The activity of C’ ‘A on Avicel was significantly higher than that of88Figure 34. Construction scheme for domain swapping of CBDCenC (NIN2) withCBDCenA. Plasmid pUC18-1.6 cenA encoding polypeptide CenA was digested withHpaI and Hind!!!. Plasmid pTZ-JC2 encoding polypeptide CenC was digested with Sma!and Hind!!!. The fragments were isolated, then ligated and the plasmid encodingpolypeptide C ‘A, designated pTZ-JC13. PT 23 refers to the 23 amino acid PT linker inCenA and PPP refers to the triproline linker in CenC.SmalHind Iflsoata large fragmentHpalHind IllIsolate small fragmectT4 DNA Ligase.PPP rer•TGA89Figure 35. Adsorption of CenA and C’ ‘A to BMCC and PASC. Panel A. Increasingamounts (paired lanes) of purified CenA and C’ ‘A were incubated without BMCC (-) (ascontrols, to determine the amount of starting material) or mixed with 1 mg of BMCC (+)to ascertain the amount of free polypeptide present in solution, post-binding. CenA waspreferentially adsorbed by BMCC from a solution bearing both CenA and C’ ‘A. PanelB. Samples after the adsorption experiment to BMCC were rebound to PASC. Pairedlanes (-) and (+) refer to incubations without or with PASC. Both CenA and C ‘Aadsorbed to PASC. The far left lane in each panel shows the molecular mass markers.BMCC: Bacterial microcrystalline cellulose. PASC: Phosphoric acid swollen cellulose(Regenerated cellulose).90A BMCCkDa_ - + - + - +11666C’ ‘ACe n A43 — — —36B PASCkDa - + - + - +11666c”ACen A3691Table 6: Adsorption of CenA and C’ ‘A to BMCC and PASC. Free polypeptide (% ofstarting material) in supernatant after adsorption.Polypeptide concentration[25] [1001 [250]Substrate % free polypeptide, after adsorptionb PolypeptideBMCCcI100 100 100 C’ ‘AN.D.C 15 30 CenAPASC10 50 80 C’AN.D. N.D. 5 CenAa [Concentration] jig/mib Amount in supernatant before adsorption was taken as 100%C None detectedd Bacterial microcrystalline cellulosee Phosphoric acid swollen cellulose (regenerated cellulose)Table7:EnzymeactivityofCenAandC’‘Aagainst arangeofcellulosicsubstratesEnzymeActivityonCMCaCelluloseazurePASCbAvicelBMCCCkatal/moldunits/molekatal/molmillikatal/molmillikatalJrnolngCenA22.5±0.1145.5±13.6±0.239.1±0.44.5±0.20C’‘A26.1±0.4327.8±33.7±0.231.2±0.21.2±0.05p3033.2±0.4586.7±44.6±0.219.3±0.61.0±0.04aCai.boxyietlylcelft1losebpljosphoric acidswollencellulose.CBactei.jalmicrocrystallinecellulose.dkatal/mol, activityof Imolproduct/s/mo]enzyme.eUnits/nmiol,productionofA595nm=1/h/mmolenzyme.93p30 alone indicating that the CenCCBD did influence the activity of the fusionpolypeptide.The results for CenA and p30 were similar to those obtained previously(Gilkes eta!., 1988; Shen eta!., 1991).The activities of C’ ‘A against the amorphous substrates PASC and cellulose azurewere compared to those of CenA. C’ ‘A had identical activity on PASC however it hadgreater than twice the activity on cellulose azure. Another striking difference was thepoorer ability of C’ ‘A to hydrolyze crystalline substrates. The activity of C’ ‘A was 20%and 70% lower on Avicel and BMCC respectively as compared to CenA. CenA had thehighest activities on crystalline cellulose, C’ ‘A had intermediate activity on crystallinecellulose and p30 had highest activity on soluble and amorphous substrates.In the cloning strategy in Fig. 34 the 23 amino-acid PT linker separating the CenAcatalytic domain from its cellulose binding domain (which is sensitive to cleavage by C.fimi protease to generate p30) was removed and replaced by a linker (PPP) directlyfollowing the N-terminal repeats (Fig.36).Fig. 37 shows the effect of increasing amounts of C.Jmnii protease on CenA andC’ ‘A. Removal of the protease site from C’ ‘A resulted in enhanced stability of thepolypeptide to C.JIrni protease as compared to CenA.94Figure 36. Bar diagram indicating features of CenA and C’ ‘A. V indicates thelocation of the 23 amino acid PT linker in CenA and the PPP linker in C’ ‘A. TheC.fimi protease sensitive site in CenA is located adjacent to the PT linker. Cleavage ofCenA by C.fimi protease results in a P30 proteolytic fragment. renA bindingdomain; E%1 CenC binding domain; linker region; I 1 CenA catalyticdomain.PT linkerV C. Ini pco(ease seCenA P301 112 134 418PPP linkerC ‘A1 150 300 584Ni N295Figure 37. Hydrolysis of CenA and C’ ‘A with C.fInii protease. Both polypeptides weretreated with 0 (Control: Starting material, untreated with protease), 0.1, 0.5 and 1.0 unitsof C.fimi protease and the products of hydrolysis analyzed by SDS-PAGE. P30 is aproteolytic fragment derived from the hydrolysis of CenA. The far left lane shows themolecular mass markers.C”A Cen Ak Da1166643 — —29 p301.0 0.5 0.1 0 1.0 0.5 0.1 0C. fimi protease (U)964. DiscussionIdentities within the amino acid sequences of the catalytic domains of 13-1,4-glycanases allow grouping of the enzymes into families (Knowles et a!., 1987; Henrissateta!. 1989; Béguin, 1990; Gilkes et al., 1991a). The enzymes of family E can begrouped into sub-families El and E2 (Béguin, 1990). Amino acids 300-880 of CenCcontain the catalytic domain; this sequence places CenC in sub-family El (Fig. 10).Although the similarities between the sub-families are much less than those betweenenzymes within a sub-family, the sequence DAGD occurs in all members of the family,and may be part of the active site because carboxyl groups are involved in glycosidic bondcleavage (Sinnott, 1987; Chauvaux et at., 1992). In enzymes from sub-family El it isabout 150 amino acids from the N-terminus of the catalytic domain, for those of E2 it isabout 40 amino acids from this terminus (Joliff ci’ at., 1986a; Tucker eta!. , 1987; Halland Gilbert, 1988; Berger ci’ a!., 1990; Giorda ci’ at. , 1990; Meinke et a!., 1991a;Coutinho eta!., 1991). The common sequence H-R, —320 amino acids away, may alsobe important for catalytic activity. Residues near its C-terminus are required for thecatalytic activity of Ce1D of Ctostridiuni thermocellun? (Béguin, 1990; Tomme et a!.,1991). In CbhII of Trichodernia reesei, for which the three-dimensional structure isknown, the substrate binding site of the catalytic domain involves amino acids 135-428(Rouvinen et al., 1990). Aspartic acid residues 175 and 221 are involved in glycosidicbond cleavage. Asp 175 occurs in the sequence V-Y--P-RDC, which is conserved in allmembers of its family; Asp221, however, is not conserved in all members of the family(Rouvinen ci’ al., 1990). Interestingly, the sequence Y-D-GH--W is conserved in allmembers of the family. In CbhII this histidine is His266. It is conceivable, therefore,that the DAGD and H-R sequences, although separated by >300 amino acids in theenzymes of family E, are close to each other in the three-dimensional structures and formparts of their active sites. The crystallization of CelD, a member of family El, whose three97dimensional structure is known confirms the location of DAGD and the H-R sequences inthe same vicinity (Juy et al., 1992).Many proteins accumulate in the cytoplasm as granules when the genes encodingthem are overexpressed in E. coli (Kane and Hartley, 1988; Wilcox and Studnicka, 1988),including Cex (O’Neill et al. , 1986) and CenA (Z.M. Guo, personal communication).The proteins in these inclusion bodies can be inactive, but granule formation alone maynot cause inactivation. CenC unlike Ce1D from Clostridium thermocellum (Joliff et a!.,1986b), does not form granules, but most of it is inactive and appears as a soluble form inthe cytoplasm (Coutinho ci’ al., 1991). As with Cex, CenA and CenB, E. coli exportssome CenC to the periplasm.CenC is very susceptible to proteolysis in E. coli. Deletions into both the N-terminal or C-terminal repeats did not result in improved stability of the enzyme when itwas produced in E. coli. C-terminal deletions revealed that most of C1C2 is not essentialfor catalytic activity.The functions of the C-terminal repeats in CenC are not known at present. NeitherC1C2 nor Cl could be affinity purified on Avicel or Sephadex. A search of the entireSWISS PROT data-bank of protein sequences with Cl or C2 (approximately 100 amino-acids in size each) revealed no obvious similarities but there was one interesting aspect ofthe alignments that allows speculation as to what the repeats may be doing. There is aperfectly conserved region of 7 amino acids in both Cl and C2 (YRAVFTNA). Thisregion was observed to be very similar to a region in rat and human hepatic lectins whichtoo are well conserved in this region (Patthy, 1987). The alignments are shown in Fig.38. This region in the rat and human hepatic lectins comprises part of the carbohydraterecognition domain (CRD) of these proteins and is similar to a number of other lectinssuch as mannose-binding protein (Drickamer, 1989). It is interesting to note that despitethe small size of 10 amino-acids of this conserved region, a peptide isolated by proteolytic98digestion from Cytisus sessilifolius anti-H(O) lectin (DTYFGKTYNPW) of 11 amino-acids retains the ability to bind carbohydrate (Konarni et al., 1992)The component enzymes in some cellulase systems form multienzyme complexes(Lamed and Bayer, 1988). Although each of the C.JInii cellulases examined to date has adiscrete CBD (O’Neill et at. , 1986; Wong et at., 1986; Meinke et at., 1991a; Coutinho eta!., 1992) there may be interaction between the enzymes on the substrate. The C-terminalrepeats might be involved in such interactions. One possible function of CenC’s C-terminal repeats may be the recognition and attachment to other glycosylated enzymeswhich might improve synergy in the hydrolysis of cellulose. Another possibility is thatCenC via its C-terminal repeats may allow binding to as yet unidentified substrates.Figure 38. Similarity of the conserved region within the C-terminal repeats to aconserved region withing the carbohydrate recognition domains (CRDs) of lectins.B, charged residue; 0, aromatic residue; : , identical residue; * , conserved residueHHL1,2 and RHL1,2: human hepatic lectins and rat hepatic lectins respectively.The sequences are numbered from the N-terminal amino acid.HHL1 CR0 220 GTDYETGFKN 229*..HHL2 CR0 243 GTDYRHNYKN 252*..RHL1 CR0 226 GTDYETGFKN 235.*. *RHL2 CR0 192 GTEYRSNFKN 201CenC Cl 950 GTRYRAVFTN 959CenC C2 1041 GTEYRAVFTN 1050CONSENSUS GTBYB- -0-N99CenC is the most active endoglucanase isolated from C.fImi to date, but it seemsto be a minor component of the cellulase system of this organism (Moser et at. , 1989).Endoglucanase D, whose catalytic domain is also in Family El like CenC, is also the mostactive endoglucanase from C. thermocetlum (Joliff eta!., 1986b; Tomme et at., 1991;Table 2). Both enzymes have similar cleavage specificities (Fig.17).Cex, CenA and CenB of C.firni have a discrete CBD of about 110 amino acids, atthe N-terminus in CenA, at the C-tenriini in Cex and CenB (Gilkes et at., 1988; Gilkes eta!. , 1991a). The sequences of the CBDs of the three enzymes are about 50% similar(Meinke et al. , 1991a), with four conserved tryptophans, two conserved cysteines and anumber of asparagine and glycine residues at several other sites (Gilkes eta!., 1991a).The N-terminal repeats of CenC are 150 amino acids in length. Sequence comparison ofthe N-terminal repeats of CenC with other C.JInii CBDs revealed that the cysteineresidues in both Ni and N2 were conserved and were also separated by about 100 aminoacids as was seen for the other CBDs. Interestingly, most of the tryptophans in thebinding domains of CenA, CenB and Cex were substituted for tyrosines in the CenCCBD. There were a number of other residues that were also conserved suggesting thatboth Ni and N2 were cellulose-binding domains which were similar to but distantlyrelated to the cellulose-binding domains of Cex and C.fimi endoglucanases A and B.CenC can bind both to cellulose and to Sephadex. In order to confirm that the N-terminalrepeats could bind independently to cellulose or Sephadex, polypeptides corresponding toNi, N2 and N1N2 were affinity purified on Avicel and/or Sephadex, proving that the N-terminal repeats of CenC constitute a cellulose- and a Sephadex-binding domain. It isnoteworthy that both repeats are necessary for binding to Sephadex whereas a singlerepeat mediated binding to cellulose. Cellulose is a polymer of -D-glucopyranosy1 unitsjoined by 1 ,4-glycosidic bonds. The backbone of the polymer has a fully extended, flatconformation (Gardner and Blackwell, 1974). Sephadex is prepared by cross-linking a100bacterial dextran with epichiorohydrin. The native dextran is a polymer of cL-Dglucopyranosyl units. The units in the backbone of the polymer are linked by a,1—6bonds, with sidechains attached to the backbone by CL,1—>3 bonds (Gasciolli, eta!.,1991). The 1,6 linkages give the backbone a loosely jointed conformation because theglucosyl units are separated by three rather than two bonds. This conformation may allowbinding by the paired but not the single repeats. Whatever the explanation, the ability ofthe paired N-terminal repeats of CenC to bind both CL-linked and 13-linked glucosepolymers is striking. CenC had no detectable activity on Sephadex or raw starch. Bindingto Sephadex has also been observed for CenA (N.R. Gilkes, personal communication),concavalin A (Agarwal and Goldstein, 1967) and pea lectin (Stubbs eta!., 1986).Adsorption of N1N2, Ni or N2 to regenerated cellulose, but not to bacterialcrystalline cellulose indicates a specificity of these domains for cellulose II and/oramorphous cellulose. This is the first reported example of a CBD showing specificity for aparticular structural form of cellulose. Avicel is a heterogeneous form of celluloseprepared by partial acid hydrolysis of wood fibres and contains both crystalline andamorphous components (Ooshima eta!., 1983; Marshall and Sixsmith, 1974).Presumably, adsorption of N1N2, Ni or N2 to Avicel involves binding to the non-crystalline regions of this substrate. It is possible that adsorption of CenC in vivo isdependent on prior disruption of crystalline cellulose by other cellulolytic components; forexample the CBD of CenA has been shown to disrupt the surface structure of cottonfibres, apparently by a non-catalytic process (Din eta!., 1991). Although it is not knownwhether such disruption enhances the adsorption of CenC, mechanisms of this kind maycontribute to the observed synergism between some cellulolytic enzymes.The fusion of CBDCenC (N1N2) to the catalytic domain of CenA was stable whenexpressed and could be purified from the culture supernatants of E. co!i using Avicel asthe affinity adsorbent. The endoglucanase activity borne by the catalytic domain of CenAand the lack of affinity of CBDCenC for BMCC were retained by the fusion polypeptide101indicating the tri-proline linker of CenC allowed the separate and independent folding ofthe catalytic and cellulose binding domains.The use of CBD’s bearing different specificities as affinity tags has potentialpractical application in biotechnology for the selective binding of different CBD fusionproteins. This was demonstrated by the selective removal of CenA from a mixture bearingboth CenA and C’ ‘A using BMCC as the affinity adsorbent.The introduction of a specific protease site such as IEGR for Factor X between theCBD and the tagged polypeptide facilitates cleavage of the purified fusion polypeptide(Ong eta!, 1989; Sassenfeld, 1990). In such situations it is preferable not only to recoverthe CBD after cleavage of the fusion polypeptide by the protease (thus leaving thepolypeptide of interest in solution) but it is also advantageous to recover the protease usedfor the digestion. To recover the protease itself in as pure a form as possible it would bepreferable to tag the protease with a CBD having a different specificity to that of the CBDused for fusion to the polypeptide of interest (Fig. 39).The use of CBD’s bearing different specificities may also have applications inindustry for example the brewing industry, where it would be advantageous not only torecover enzymes but also to allow addition or removal of enzymes at different times bothduring the fermentation process and in the downstream processing of the brew prior tobottling. The preferential tagging of polypeptides may also have applications in theautomation of diagnostic tests and techniques in molecular biology techniques, forexample, the sequential treatment of DNA by restriction and modification enzymes incloning experiments.This is the first time that by using separate cellulose binding domains and using apreferred adsorbent for one of the two CBD’s that each can be differentially removedfrom a single solution bearing both enzymes.102Protease siteVContaminating proteinPurified poIeptide PAmorphouscelluloseFigure. 39 Purification scheme for fusion polypeptides bearing CBDs with differentaCBD A Polwep(ide P4 aMicrocrystallinecellulose/WashL rAProtease CBD NI N2///iFspecificities.103The activity of C’ A was tested on a variety of cellulosic substrates and comparedto CenA: BMCC (predominantly form I, highly crystalline cellulose) (Henrissat andChanzy, 1986; Ross etal., 1991); Avicel (heterogeneous cellulose containing bothcrystalline and amorphous components); PASC (regenerated cellulose, thought to bepredominantly form II) (Sarko, 1986) although other workers have concluded that it isamorphous (Lee eta!., 1982; Ooshima ci’ at., 1983); Cellulose azure (dye-linkedamorphous cellulose) (Fernley, 1962) and CMC (methyl substituted soluble cellulose).C’ ‘A had higher activity on the amorphous substrate cellulose azure but a similar activityon PASC as compared to CenA. The isolated catalytic domain of CenA, p30, had thehighest activity on both soluble and amorphous substrates. C’ ‘A was more active onAvicel than p30 alone but the most striking difference was the poor activity of C’ ‘A oncrystalline cellulose as compared to CenA. The results indicate that CBDs have a markedinfluence on catalytic activity only when the substrate is of an insoluble relativelycrystalline nature such as Avicel and BMCC. It is remarkable that Klesov and hiscolleagues, with great insight, were able to predict the correlation between the ability ofenzymes to adsorb to crystalline cellulose and the hydrolyis of crystalline cellulose(Rabinovitch eta!., 1982; Klesov eta!., 1983; Chernoglazov eta!., 1983). Almost adecade later and with a better understanding of the molecular basis of attachment betweenenzyme and substrate, it is much easier to appreciate their contribution in understandingthe hydrolysis of cellulose.The ability of the catalytic domain to function on different substrates based on thespecificity of the binding domain permits the engineering of enzymes suited for thehydrolysis of a particular substrate. The study of such engineered enzymes along withnatural isolates should lead to an even greater understanding of cellulose hydrolysis.Cellulases may consist of two or more domains linked to each other by linkersequences rich in Ser, Thr and Pro residues (Gilkes etal., 1991a). Low angle X-raystudies (Schmuck et at., 1986; Abuja eta!., 1988) suggest that these regions are flexible.104A recent search of natural linker sequences found at domain interfaces of proteins in theBrookhaven protein structural databank revealed that so far oniy relatively short linkers offive or six amino acids have been identified as general candidates to link protein moleculesor domains through gene fusion (Argos, 1990), therefore detailed information concerningthe structures of flexible linker regions long enough to be useful for the design andconstruction of novel fusion proteins is not yet available (Takkinen et aL, 1991).Sequences composed of 3 or more consecutive proline residues have beenobserved in the linker regions of cellulases (Gilkes ci’ a!., 1991a). The use of proline inlinker oligopeptides was considered controversial as linkers maintain an extendedconformation (Jentoft, 1990; Williamson et at., 1992) and proline can be consideredhydrophobic, tends to bend and kink the main chain and is rigid in dihedral conformation.Nonetheless it was a preferred residue amongst the natural linkers and may thus beconsidered desirable (Argos, 1990). A conserved triproline sequence has been postulatedto generate a sharp bend in the polypeptide chain of myelin basic protein (Brostoff andEylar, 1971). Subsequent nuclear magnetic resonance experiments failed to confirm thepresence of cis-proline bonds indicating instead that the proline residues had an all-transconformation in aqueous solution (Fraser and Deber, 1985) inconsistent with a sharp bendwith chain reversal (Brostoff and Eylar, 1971). NMR work done on a separate triprolinepeptide also indicated the absence of any sharp bend at the PPP sequence and indicatedthat the segments of the peptide on opposite sides of PPP were distinctly separated fromeach other (Nygaard eta!., 1984). These results suggest that the triproline sequence mayact as a rigid spacer that separates parts of the protein. This may help explain why theconformation of the polypeptide C’ ‘A in this region is more resistant to C.flmi protease.The high activity of CenC on soluble substrate, its inability to bind to highlycrystalline cellulose and the modified activity of the catalytic domain of CenA influencedby the CBD of CenC suggest that the role of CenC in the C.Jmnii cellulase system maybefor the hydrolysis of noncrystalline and or amorphous cellulose. Further improvements in105the stable expression of CenC would allow for some very interesting experiments on thesynergy of CenC with enzymes that prefer crystalline cellulose. Furthermore unlike CenAand Cex, which were recovered from the residual cellulose in cultures of C.fimi grownwith Avicel (Langsford ci’ at. 1984), CenC and CenC’ remained in the supernatant ofsuch cultures and were isolated by virtue of their capacities for binding to Sephadex(Moser eta!., 1989). It was not clear at that time why CenC and CenC’ were notrecovered on the residual Avicel. The present study suggests that Cex and CenA, whichare major components of the C.finii cellulase system, may saturate the available bindingsites and this may have prevented the recovery of CenC under such conditions.This thesis has attempted to correlate amino-acid sequence with function for theendoglucanase CenC. The large size of CenC is contributed to by the presence ofrepeating sequences at the N and C-terminal of the catalytic domain. It is postulated thatthe hydrolysis of crystalline cellulose requires the presence of endoglucanase A and B aswell as exoglucanase/xylanase Cex. As digestion proceeds, more amorphous regions areexposed and soluble cellodextrins are produced; it is suggested that CenC may play agreater role at this stage of the digestion process. The function of the C-terminal repeats ofCenC are not known at present but they may have a role in protein-protein interactions.1065. SummaryThe cenC gene was sequenced.The missing Y end was identified by Southern blotting of C.fimi genomic DNA.Sequence comparison to other cellulolytic enzymes revealed three regions, N-terminal repeats, C-terminal repeats and the central region comprising thecatalytic domain.The N-terminals repeats were cloned and expressed separately and shown to havebinding independent of the catalytic domain.The binding domain of CenC is unusual, the first described to have affinity foramorphous cellulose but negligible affinity for bacterial crystallinecellulose. It can be produced in large amounts (>20 mg/I) in E. co/i.The industrial advantage of using the CenC CBD as an affinity tag is that it can beeluted from cellulose with water unlike most other CBD fusions whichrequire much harsher and more expensive conditions for elution usingguanadinium hydrochloride.Binding domains with different specificities allowed the preferential adsorption ofpolypeptides from a mixture by use of the appropriate affinity adsorbent.Although the function of the C-terminal repeats is not known at the present, it wasshown that most of the C-terminal repeats could be deleted without anyobservable effect on catalytic or binding activity.Replacement of CBDCenA with CBDCenC in CenA allowed the role of theCBDs in cellulose hydrolysis to be examined. The CBDs have a significantinfluence on catalytic activity especially on insoluble substrates.Replacement of the PT linker of CenA (which is sensitive to C.fimi protease)with a PPP linker in C’ ‘A increased the stability of the enzyme toproteolysis byC.fmnii protease.1076. ReferencesAbuja, P.M., M, Schmuck, I. Pilz, P. Tomme, M. Claeyssens and H. Esterbauer 1988.Structural and functional domains of cellobiohydrolase I from Trichoderma reesei.A small angle X-ray scattering study of the intact enzyme and its core. Eur. J.Biophys. 15:339-342.Agrawal, B.B.L. and I.J. Goldstein 1967. Protein-Carbohydrate interaction VI. Isolationof concavalin A by specific adsorption on cross-linked dextran gels. Biochim etBiophysica Acta 147:262-27 1.A1-Tawheed, A.R. 1988. Molecular characterization of cellulase genes from Cellulomonasflavigena. M.Sc. Thesis. Trinity College, Dublin, Ireland.Almin, K.E., K.E. Eriksson and L.G. Petterson 1975. Extracellular enzymesystem utilized by the fungus Sporotrichiun? pulverulentum (Chryosporiumlignorurn) for the breakdown of cellulose. 2. Activities of the five endo-13-1,4-glucanases towards carboxymethylcellulose. Eur. J. Biochern. 51:207-211.Argos, P. 1990. An investigation of oligopeptides linking domains in protein tertiarystructures and possible candidates for general gene fusion. J. Mo!. Biol. 211:943-958.Atalla, R.H. and D.L. van der Hart 1984. Native cellulose: A composite of two distinctcrystalline forms.Science 223:283-284.108Atkinson, T. and M. Smith. 1984. Solid phase synthesis of oligodeoxyribonucleotides bythe phosphite triester method. In N.J. Gait (ed.) Oligonucleotide Synthesis: APractical Approach. IRL Press, Oxford, pp.35-81.Aubert, 3.-P., P. Béguin and J. Millet (ed.). 1988. Biochemistry and genetics of cellulosedegradation. FEMS Symp. 43:1-428.Azevedo, M. de 0., M.S.S. Felipe, S. Astolfi-Filho and A. Radford 1990. Cloning,sequencing and homologies of the cbh-I (exoglucanase) gene of Humicola griseavar. thermoidea. J. Gen. Microbiol. 136: 2569-2576.Béguin, P., H. Eisen and A.Roupas 1977. Free and cellulose bound cellulases fromCellulonionas sp. J. Gen. Microbiol. 101:191-196.Béguin, P. and H. Eisen 1978. Purification and partial characterization of threeextracellular cellulases from Cellulornonas sp. Eur. J. Biochem. 87: 525-53 1.Béguin, P. 1983. Detection of cellulase activity in polyacrylarnide gels using Congo red-stained agar replicas. Anal. Biocheni. 131:333-336.Béguin, P., N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr., G.P. O’Neill and R.A.J.Warren 1987. Cloning of cellulase genes. CRC Critical Rev. Biotechnol. 6:129-162.Béguin, P. 1990. Molecular biology of cellulose degradation. Annu. Rev. Microbiol.44: 219-248.109Berger, E., W. A. Jones, D. T. Jones and D.R. Woods 1989. Cloning and sequencingof an endoglucanase (endi) gene from Bulyrivibriofibrisolvens H17c. Mol. Gen.Genet. 219:193-198.Berger, E., W. A. Jones, D. T. Jones and D.R. Woods 1990. Sequencing andexpression of a cellodextrinase (cedl) gene from Butyrivibriofibrisolvens H17ccloned in Escherichia co/i. Mo/cc. Gen. Genet. 223: 310-318.Birnboim, H.L. and J. Doly 1979. A rapid alkaline extraction procedure for screening ofrecombinant plasmid DNA. Nucleic Acids Res. 7:15513-15523.Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgramquantities of proteins utilizing the principle of protein-dye binding. Anal.Biochem. 72:248-254.Brostoff, S. and E.H. Eylar 1971. Localization of inethylated arginine in the Al proteinfrom myelin. Proc. Nat. Acad. Sci. 68:765-769.Brownell, H.H. and J. N. Saddler 1987. Steam pretreatment of lignocellulosic materialfor enzymatic hydrolysis. Biotechno/. Bioeng. 29:228-235.Burton, J., S.G. Wood, A. Pedyczak and 1.Z. Siemion 1989. Conformationalpreferences of sequential of the hinge region of human IgAl immunoglobulinmolecule. II. Biophys. Chem. 33:39-45.110Caiza, R.E., D.C. Irwin and D.B. Wilson 1985. Purification and characterization of twof3-1 ,4-endoglucanases from ThermomonosporaJusca. Biochemistry 24:7797-7804.Cantor, C.R. and P.R. Schimmel 1980. Biophysical Chemistry Part II. Techniques forthe Study of Biological Structure and Function. Freeman, San Francisco, p.380.Cavicchioli, R., P.D. East and K. Watson 1991. endAFS, a novel family Eendoglucanase gene from Fibrobacter succinogenes AR1.J. Bacteriol. 173:3265-3268.Chauvaux, S., P. Béguin, J. -P. Aubert, K.M. Bhat, L.A. Gow, T.M. Wood and A.Bairoch 1990. Calcium-binding affinity and calcium-enhanced activity ofClostridiuni rhermocellum endoglucanase D. Biocheni J 265: 261-265.Chauvaux, S, P. Béguin and J. -P. Aubert 1992. Site-directed mutagenesis of essentialcarboxylic residues in Clostridiuni thermocellum endoglucanase Ce1D. J. Biol.Chem. 267:4472-4478.Chen, C.M., M. Gritzali and D.W. Stafford 1987. Nucleotide sequence and deducedprimary structure of cellobiohydrolase II from Trichoderma reesei. BiolTechnol1: 691-696.Cheng, C., N. Tsukagoshi and S. Udaka 1990. Nucleotide sequence of thecellobiohydrolase gene from Trichoderma viride. Nuci Acids Res 18: 5559.111Chernoglazov, V.M., A.A. Klesov and O.V. Ermolova 1983. Effect of isoenzymes ofendoglucanase from Trichoderma Iongibrachiatum differing in strength ofadsorption on cellulose on soluble, amorphous, and crystalline substrates: varyingrole of adsorption efficiency. Transalated from Biokhimiya, Vol. 48, No. 10, pp.1617-1623. Biochemistry 48:1391-1396.Choudhury, N., P.P. Gary and N.W. Dunn 1978. Isolation of a cotton wool degradingstrain of Cellulomonas: mutants with altered ability to degrade cotton wool. Aust.J. Biol. Sci. 31:553-564.Coughian, M. P. 1985. The properties of fungal and bacterial cellulases with commentson their production and application. Biotech. Genetic Engineering Rev. 3:39-109.Coughlan, M. P. and L.G. Ljungdahl 1988. Comparative biochemistry of fungal andbacterial cellulolytic systems. FEMS Symp. 43:11-30.Coutinho, J.B., B. Moser, D.G. Kilburn, R.A.J. and R.C. Miller, Jr. 1991. Nucleotidesequence of the endoglucanase C gene (cenC) of Cellulomonasfimi, its high-levelexpression in Escherichia co/i, and characterization of its products. Mol Microbiol5: 1221-1233.Coutinho, J.B., N.R. Gilkes, R.A.J. Warren, D.G.Kilburn and R.C.Miller Jr. 1992.The binding of Cellulomonasfinii endoglucanase C (CenC) to cellulose andSephadex is mediated by the N-terminal repeats. Mol Microbiol 6:1243-1252.112Deshpande, M.V., K.E. Eriksson and B. Pettersson 1984. An assay for selectivedetermination of exo-B-1,4-glucanases in a mixture of cellulolytic enzymes. Anal.Biochem. 138:481-487.Din, N., N.R. Gilkes, B. Tekant, R.C. Miller Jr., R.A.J. Warren and D.G. Kilbum.1991. Non-hydrolytic disruption of cellulose fibres by the binding domain of abacterial cellulase. Biotechnology 9: 1096-1099.Drickamer, K. 1989. Multiple subfamilies of carbohydrate recognition domains in animallectins. In Carbohydrate recognition in cellularfunction. Wiley, Chichester (CibaFoundation Symposium 145, pp.45-61.Durrant, A.J., J. Hall, G.P. Hazelwood and H.J. Gilbert 1991. The non-catalytic C-terminal region of endoglucanase E from Clostridium thermocellum contains acellulose-binding domain. Biochem 1 273: 289-293.Enari, T. M. and M.L. Niku-Paavola 1987. Enzymatic hydrolysis of cellulose: is thecurrent theory of mechanisms of hydrolysis valid? RC Critical Rev.Biotechnol. 5:67-87.Eriksson, K.E. 1975. Enzyme mechanisms involved in the degradation of woodcomponents, pp. 263-280. In M. Bailey, T.M. Enari and M. Linko (eds.),Symposium on enzymatic hydrolysis of cellulose. SITRA, Helsinki.Femley, H. N. 1962. The use of reactive dyestuffs in enzyrnology: New substrates forcellulolytic enzymes. Biocheni 1. 87: 90-95.113Ferreira, L.M.A., A.J. Durrant, J. Hall, G.P. Hazeiwood and H.J. Gilbert 1990. Spatialseparation of protein domains is not necessary for catalytic activity or substratebinding in a xylanase. Biocheni J. 269: 26 1-264.Ferreira, L.M.A, G.P. Hazeiwood, P.J. Barker and H.J. Gilbert 1991. Thecellodextrinase from Pseudornonasfluorescens subsp. cellulosa consists ofmultiple functional domains. Biochem J. 279:793-799.Fraser, P.E. and C.M. Deber 1985. Structure and function of the proline-rich region ofmyelin basic protein. Biochernistiy 24:4593-4598.Gardner, K.H. and J. Blackwell 1974. The structure of native cellulose. Biopolymers13: 1975-2001.Gasciolli, V., L. Choplin, F, Paul and P. Monsan 1991. Viscous properties andmolecular characterization of enzyi’natically size-controlled oligodextrans inaqueous solution. JBiotechnol 19: 192-202.Ghangas, G.S. and D.B. Wilson 1988. Cloning of the Thermomonospora fuscaendoglucanase E2 gene in Streptoniyces lividans: affinity purification andfunctional domains of the cloned gene product. App. Environ. Microbiol.54:2521-2526.Gilbert, H.J., 3. Hall, G.P. Hazeiwood and L.M.A. Ferreira 1990. The N-terminalregion of an endoglucanase from Pseudomonasjluorescens subspecies cellulosaconstitutes a cellulose-binding domain that is distinct from the catalytic centre.Mo! Microbiol 4: 759-767.114Gilkes, N.R., M.L. Langsford, D.G. Kilburn, R.C. Miller, Jr., and R.A.J. Warren1984. Mode of action and substrate specificities of cellulases from cloned bacterialgenes. J. Biol. Cheni. 259:10455-10459.Gilkes, N.R., R.A.J. Warren, R.C. Miller, Jr., D.G. Kilburn. 1988. Precise excisionof the cellulose binding domains from two CellulonionasfImi cellulases by ahomologous protease and the effect on catalysis. J. Biol. Chen2. 263: 1040 1-10407.Gilkes, N.R., D.G. Kilburn, R.C. Miller Jr. and R.A.J. Warren 1989. Structural andfunctional analysis of a bacterial cellulase by proteolysis. J. Biol. Chem.264:17802-17 808.Gilkes, N.R., B. Henrissat, D.G. Kilburn, R.C. Miller, Jr. and R.A.J. Warren 1991a.Domains in microbial f3-1,4-glycanases: sequence conservation, function, andenzyme families. MicrobiolRev. 55: 305-3 15.Gilkes, N.R., A, Meinke, J.B. Coutinho, E. Ong, J,M. Greenwood, R.C. Miller, Jr.,D.G. Kilburn and R.A.J. Warren 1991b. Cellulomonasfimi 8-1,4-glucanases. InEnzymes in biomass conversion. Leatharn, G.F. and M.E. Himmel (eds.)American Chemical Society, Series 460, Washington, pp. 349-360.Gilkes, N.R., E. Jervis, B. Henrissat, B. Tekant, R.C. Miller Jr., R.A.J. Warren andD.G. Kilburn 1992. The adsorption of a bacterial cellulase and its two isolateddomains to crystalline cellulose. J. Biol. Chem. 267:6743-6749.115Giorda, R., T. Ohmachi, D.R. Shaw and H.L. Ennis 1990. A shared internal threonineglutairiic acid-threonine-proline repeat defines a family of Dictyosteliumdiscoideum spore germination specific proteins. Biochemistry 29: 7264-7269.Gottschalk, 0. 1986. Degradation of polymers by exoenzymes. In BacterialMetabolism. Springer-Verlag, New York, pp. 142-143.Gough, C.L., J.M. Dow, J. Keen, B. Henrissat and M. J. Daniels 1990. Nucleotidesequence of the gene encoding the major endoglucanase of Xanthomonascampestris pv. campestris. Gene 89:53-59.Griffin, H., F.R. Dintzis, Krull, L. and EL. Baker 1984. A microfibril generating factorfrom the enzyme complex of Trichoderma reesei. Biotechnol. Bioeng. 26:296-300.Guo, Z.M., N. Arfman, E. Ong, N.R. Gilkes, D.G. Kilburn, R.A.J. Wan-en and R.C.Miller Jr. 1988. Leakage of Cellulomonasfinii cellulases from Escherichia coli.FEMS Microbiol. Left. 49:279-283.Hall, J. and H.J. Gilbert 1988. The nucleotide sequence of a carboxymethylcellulasegene from Pseudomonasfluorescens subsp. cellulosa. Mo! Gen Genet 213: 112-117.Hall, J., G.P. Hazeiwood, P.J. Barker and H.J. Gilbert 1988. Conserved reiterateddomains in Clostridium therniocellum endoglucanases are not essential for catalyticactivity. Gene 69:29-38.116Hall, J., G.P. Hazlewood, N.S. Huskisson, A.J. Durrant and H.J. Gilbert 1989.Conserved serine-rich sequences in xylanase and cellulase from Pseudomonasfluorescens subsp. cellulose: internal signal sequence and unusual proteinprocessing. Mol Microbiol 3: 1211-1219.Harlow, E and D. Lane 1988. Antibodies: A Laboratoiy Manual. Cold Spring HarborLaboratory Press, New York.Henrissat, B., and H. Chanzy 1986. Enzymatic breakdown of cellulose crystals. In R.A.Young and R.M. Rowell (ed.) Cellulose: Structure, Modification and Hydrolysis.Wiley, N.Y., pp. 337-47.Henrissat, B., M. Claeyssens, P. Tomme, L. Lemesle and J.P. Mornon 1989.Cellulase families revealed by hydrophobic cluster analysis. Gene 81: 83-95.Hestrin, S. 1963. Bacterial cellulose. In Methods in Carbohydrate Chemistry. Vol.3.Whistler, R.L. (ed.). New York: Academic Press, pp. 4-9.Jauris, S., K.P. RUcknagel, W.H. Schwarz, P. Kratzsch, K. Bronnenmeier and W.L.Staudenbauer 1990. Sequence analysis of the Clostridium stercorarium celZ geneencoding a thermoactive cellulase (Avicelase I): identification of catalytic andcellulose-binding domains. Mo!. Gen. Genet. 233: 258-267.Jentoft, N. 1990. Why are proteins 0-glycosylated? Trends Biol. Sci. 15:29 1-294.117Joliff, G., P. Béguin and J.-P. Aubert 1986a. Nucleotide sequence of the cellulase genecelD encoding endoglucanase D of Clostridium thermocellum. Nucleic Acids Res.14: 8605-8613.buff, G, P. Béguin, M. Juy, J. Millet, A. Ryter, R. Poijak and 1.-P. Aubert 1986b.Isolation, crystallization and properties of a new cellulase of Clostridiumtherniocellum overproduced in Escherichia co/i. Biotechnology 4:896-900.Juy, M., A.G. Amit, P.M. Aizari, R.J. Poijak, M. Claeyssens, P. Béguin and J.-P.Aubert 1992. Three-dimensional structure of a thermostable bacterial cellulase.Nature 357:89-9 1.Kane, J.F. and D.L. Hartley, D.L. 1988. Formation of recombinant protein inclusionbodies in Escherichia co/i. Trends Biotechnol. 6:95-10 1.Kaufman, E., N. Geisler, N. and K. Weber 1984. SDS-PAGE strongly overestimatesthe molecular masses of the neurofilament proteins. FEBS Lett. 170: 8 1-86.Kellett, L.E., D.M. Poole, L.M.A. Ferreira, A.J. Durrant, G.P. Haziewood and H.J.Gilbert 1990. Xylanase B and an arabinofuranosidase from Pseudomonasfluorescens subsp. cel/u/osa contain identical cellulose-binding domains and areencoded by adjacent genes. Biochern J 272: 369-376.118Klesov, A.A., V.M. Chernoglazov, M.L. Rabionovich, M.V. Glazov and M.D.Adamenkova 1983. Ability of cellulases to degrade crystalline cellulose as a resultof their effective adsorption on the substrate: experimental confirmation andtheoretical interpretation. Transalated from Biokhimiya, Vol. 48, No. 9, pp. 1411-1420. Biochemistry 48:1207-1215Knowles, J., P. Lehtovaara and T. Teen 1987. Cellulase families and their genes.Trends Biotechnol. 5:255-26 1.Konami, Y., K. Yamarnoto, T. Osawa and T. Irimura 1992. Correlation betweencarbohydrate-binding specificity and amino-acid sequence of carbohydrate-bindingregions of Cytisus-type anti-H(O) lectins. FEBS Lett. 304:129-35.Kraulis, P.J., G.M. Clore, M. Nilges, T. Aiwyn Jones, G. Pettersson, J. Knowles andA.M. Gronenborn 1989. Determination of the Three-dimensional solutionstructure of the C-terminal domain of Cellobiohydrolase I from Trichodermareesei A study using nuclear magnetic resonance and hybrid distance geometry-dynamical simulated annealing. Biochemistiy 28:7241-7257.Kunkel, T.A., J.D. Roberts and R.A. Zakour 1987. Rapid and efficient site-specificmutagenesis without phenotypic selection. Methods Enzymol. 154:367-382.Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature 277: 680-685.Lamed, R. and E.A. Bayer 1988. The cellulosome of Clostridiuni thermocellum. AdvAppi Microbiol 33:1-46.119Langsford, M.L., N.R. Gilkes, W.W. Wakarchuk, D.G. Kilburn, R.C. Miller, Jr. andR.A.J. Warren 1984. The cellulase system of Cellulomonasfimi. J. Gen.Microbiol. 130:1367-1376.Langsford, M.L., N.R. Gilkes, B. Singh, B. Moser, R.C. Miller, Jr., R.A.J. Warrenand D.G. Kilburn 1987. Glycosylation of bacterial cellulases preventsproteolytic cleavage between functional domains. FEBSLett. 225:163-167.Lao, G., G.S. Ghangas, E.D. Jung and D.B. Wilson 1991. DNA sequences of three 1-1,4-endoglucanase genes from Thermomonosporafusca. J Bacteriol 173: 3397-3407.Lee, S.B., H.S. Shin, D.D.Y. Ryu and M. Mandels 1982. Adsorption of cellulase oncellulose: effects of physicochernical properties of cellulose on adsorption and rateof hydrolysis. Biotechnol. Bioeng. 24:2137-53.Lehninger, A.L. 1982. Carbohydrates: Structure and biological function. In Principles ofbiochemistiy. Worth Publishers Inc., New York, pp.289-292.Lever, M. 1973. Colorimetric and fluorornetric carbohydrate determination with phydrobenzoic acid hydrazide. Biocheni. Med. 7:274-28 1.Ljungdahl, L., B. Pettersson, K.E. Eriksson and J. Wiegel 1983. A yellow affinitysubstance involved in the cellulolytic system of Clostridiuni thermocellum.Current Microbiol. 9:195-200.120Mackay, R.M., A. Lo, G. Willick, M. Zuker, S. Baird, M. Dove, F. Moranelli and V.Seligy 1986. Structure of a Bacillus subtilis endo-3-1,4-glucanase gene. NuciAcids Res 14: 9159-9170.Marshall, K., and D. Sixsmith 1974. Some physical characteristics of microcrystallinecellulose. Drug Devel Commun 1:51-71.Matsudaira, p. 1987. Sequence from picomole quantities of proteins electroblotted ontopolyvinylidene difluoride membrane. J. Biol. Chern. 262:10035-38.Matsudaira, p. 1990. Limited N-terminal sequence analysis. Methods Enzymol. 182:602-613.McGavin, M. and C.W. Forsberg 1989. Catalytic and substrate binding domains ofendoglucanase 2 from Bacteroides succinogenes. J. Bacteriol. 171:3310-3315.Messing, J. 1983. New M13 vectors for cloning. Methods Enzymol. 101:20-78.Meinke, A., C. Braun, N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr. and R.A.J. Warren1991a. Unusual sequence organization in CenB, an inverting endoglucanase fromCellulonionasjimi. JBacteriol 173: 308-314.Meinke, A., N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr. and R.A.J. Warren 1991b.Multiple domains in Endoglucanase B (CenB) from Cellulomonasfimi: Functionsand relatedness to domains in other polypeptides. J. Bacteriol 173: 7126-7 135.121Meinke, A., N.R. Gilkes, D.G. Kilburn, R.C. Miller Jr. and R.A.J. Warren 1992.Bacterial cellulose binding domain-like sequences in eucaryotic polypeptides.Protein Seq Data Anal. In Press.Miller, G.L. 1959. Use of dinitrosalicylic acid reagent for determination of reducingsugars. Anal. Chem. 31:426-428.Miller, P.B., H. Shen, N.R. Gilkes, D.G. Kilburn, R.C. Miller Jr., A. G. Plaut andR.A.J. Warren 1992. Endoglucanase A from Cellulomonasfimi in which thehinge sequence of human IgAl is substituted for the linker connecting its twodomains is hydrolyzed by IgA proteases from Neisseria gonorrhoeae. FEMSMicrobiol. Lett. 92:199-204.Millett, M.A., A.J. Baker and L.D. Satter 1976. Physical and chemical pretreatments forenhancing cellulose saccharification. Biotechnol. Bioeng. Symp. 6:125-153.Mizusawa, S., S. Nishimura and F. Seela 1986. Improvement of the dideoxy chaintermination method of DNA sequencing by use of deoxy-7-deazaguanosinetriphosphate in place of dGTP. Nuci. Acids Res. 14:1319-1324.Montenecourt, B.S. 1983. Trichoderma reesei cellulases. Trends Biotechnol. 1:156-161.Moser, B. 1988. Molecular cloning, characterization and expression of theEndoglucanase C gene of Cellulomonasfinii and properties of the native andrecombinant gene product. Ph.D. thesis. Univ. of British Columbia, Vancouver,B.C., Canada.122Moser, B., N.R. Gilkes, D.G. Kilburn, R.A.J. Warren and R.C. Miller Jr. 1989.Purification and characterization of endoglucanase C of Cellulomonasfimi,cloning of the gene, and analysis of in vivo transcripts of the gene. App!.Environ. Microbiol. 55: 2480-2487.Mullings, R. 1985. Measurement of saccharification by cellulases. Enzy. Microb.Technol. 7:586-591.Nakamura, A., T. Uozurni and T. Beppu 1987. Nucleotide sequence of a cellulase geneof Bacillus subtilis. Eur J Bioclieni 164: 3 17-320.Nelson, N. 1952. A photometric adaptation of the Somogyi method for the determinationof glucose. J. Biol. Cheni. 153:376-380.Nossal, N.G. and L.A. Heppel 1966. The release of enzymes by osmotic shock fromEscherichia co/i in exponential phase. J. Biol. Chem. 241 :3055-3062.Nummi, M., M. Niku-Paavola, T. Enari and V. Raunio 1981. Isolation of cellulases bymeans of biospecific sorption on amorphous cellulose. Anal. Biochem. 116:137-141.Nygaard, E., G.L. Mendz, W.J. Moore and R.E. Martenson 1984. NMR of a pepticpeptide spanning the triprolyl sequence in myelin basic protein. Biochemistry23:4003-4010.123O’Neill, G.P., S.H. Gob, R.A.J. Warren, D.G. Kilburn and R.C.Miller, Jr.1986.Structure of the gene encoding the exoglucanase of Cellulomonasfimi.Gene 44 :325-330.Ong, E., J.M. Greenwood, N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr. and R.A.J.Warren 1989. The cellulose-binding domains of cellulases: tools forbiotechnology. Trends Biotechnol 7:329-43.Ong, E. 1992. Enzyme immobilization using the cellulose binding domain of theCellulomonasJlmi exoglucanase. Ph. D. thesis. Univ. of British Columbia,Vancouver, B.C., Canada.Ooshima, H., M. Sakata and Y. Harano 1983. Adsorption of cellulase from Trichodermaviride on cellulose. Biotechnol Bioeng 25:3103-3114.Owolabi, J.B., P. Béguin, D.G. Kilburn, R.C. Miller, Jr., and R.A.J. Warren. 1988.Expression in Escherichia coli of the Cellulomonasfinii structural gene forendoglucanase B. Appi. Environ. Microbiol. 54:518-523.Patthy, L. 1987. Is lung surfactant protein a lectin-collagen hybrid? Nature 325:490.Penttila, M., P. Lehtovaara, H. Nevalainen, R. Bhikhabhai and J. Knowles 1986.Homology between cellulase genes of Trichodernia reesei: complete nucleotidesequence of the endoglucanase I gene. Gene 45: 253-263.Perrin, D.D. and B. Dempsey 1974. Buffers for pH and metal ion control. Publishers,Chapman and Hall, New York.124Py, B., I. Bortoli-German, J. Haiech, M. Chippaux and F. Barras 1991. Cellulase EGZof Erwinia chysanthemi: structural organization and importance of His98 andGlu 133 residues for catalysis. Protein Enging 4:325-333.Rabinovitch, M.L., Nguen Van Viet and A.A. Klesov 1982. Adsorption of cellulolyticenzymes on cellulose and kinetics of the action of adsorbed enzymes. Two typesof interactions of the enzymes with an insoluble substrate. Transalated fromBiokhimiya, Vol. 47, No. 3, pp.465-477.Biochemistry 47:394-403.Reese, E.T., R.G.H. Siu and H.S. Levinson 1950. The biological degradation of solublecellulose derivatives and its relationship to the mechanism of cellulose hydrolysis.J. Bacteriol. 59:485-497.Robson, L.M. and G. H. Chambliss 1986. Cloning of the Bacillus subtilis DLG 13-1,4-glucanase gene and its expression in Escherichia coli and Bacillus subtilis. JBacteriol 165: 612-619.Ross, P., R. Mayer and M. Benziman 1991. Cellulose biosynthesis and function inbacteria. Microbiol. Rev. 55:35-58.Rouvinen, J., T. Bergfors, T. Teen, J.K.C. Knowles, T.A. Jones 1990. Three-dimensional structure of cellobiohydrolase II from Trichoderma reesei. Science249: 380-385.125Ryu, D.D.Y., C. Kim and M. Mandels 1984. Competitive adsorption of cellulasecomponents and its signficance in a synergistic mechanism. Biotechnol. Bioeng.26:488-496.Sambrook, J., E.F. Fritsch and T. Maniatis 1989. Molecular Cloning: A LaboratoryManual, 2nd ed. Cold Spring Harbor Laboratory Press, New York.Sanger, F., S. Nicklen and A.R. Coulson, A.R. 1977. DNA sequencing with chaintenninating inhibitors. Proc Natl Acad Sd USA 74: 5463-5467.Sarko, A. 1986. Recent X-ray crystallographic studies of celluloses. In Cellulose:Structure, Modification and Hydrolysis. Young R.A., and Rowell, R.M. (eds).New York: Wiley lnterscience. pp.29-66.Sassenfeld, H.M. 1990. Engineering proteins for purification. Trends Biotechnol 8:88-93.Saul, D.J., L.C. Williams, D.R. Love, L.W. Chamley and P.L. Bergquist 1989.Nucleotide sequence of a gene from Caldocelluin saccharolyticum coding forexocellulase and endocellulase activity. NucI Acids Res 17: 439.Scopes, R.K. 1974. Measurement of protein by spectrophotometry at 205nm. Anal.Biocheni. 59:277-282.Schmuck, M., I. Pilz, M. Hayn and H. Esterbauer 1986. Investigation ofcellobiohydrolase from Trichoderma reesei by small-angle X-ray scattering.Biotechnol. Lett. 8:397-402.126Sharon, N. 1975. Introduction. In Complex carbohydrates: Their chernistiy,biosynthesis, and functions. Addison-Wesley Publishing Co. Mass., pA..Sharrock, K.R. 1988. Cellulase assay methods: a review. Journal ofBiochem. Biophys.Methods. 17: 8 1-106.Shen, H., M. Schmuck, I. Pilz, N.R. Gilkes, D.G. Kilburn, R.C. Miller Jr. and R.A.J.Warren 1991. Deletions of the linker connecting the catalytic and cellulose-bindingdomains of Endoglucanase A (CenA) of Cellulomonasfimi alters its conformationand catalytic activity. J. Biol. Chem. 266: 11335-11340.Shoemaker, S., V. Schweickart, M. Ladner, D. Gelfand, W. Kwock, K. Myambo andM. Innis 1983. Molecular cloning of exo-cellobiohydrolase I derived fromTrichoderma reesei strain L27. BiolTechnol 1:691-696.Sims, P.F.G., C. James and P. Broda 1988. The identification, molecular cloning andcharacterization of a gene from Phanerochaete chtysosporium that shows stronghomology to the exo-cellobiohydrolase I gene from Trichoderma reesei. Gene74:411-422.Sinnott, M.L. 1987. Glycosyl group transfer. In Enzyme Mechanisms. Royal Societyof Chemistry, London. pp. 25 9-297.Stahlberg, J., G. Johansson and G. Pettersson 1988. A binding-site deficient,catalytically active, core protein of endoglucanase III from the culture filtrate ofTrichoderma reesei. Eur. J. Biochem. 173:179-183.127Stahlberg, J., G. Johansson and G. Pettersson 1991. A new model for enzymatichydrolysis of cellulose based on the two-domain structure of cellobiohydrolase I.Biotechnology 9:286-290.Stewart, B.J. and J.M. Leatherwood 1976. Derepressed synthesis of cellulase byCellulomonas. J. Bacteriol. 128:609-615.Stubbs, M.E., J.P. Carver and R.J. Dunn 1986. Production of pea lectin in Escherichiacoli. J. Biol. Cheni. 261:6141-6144.Tabor, S. and C.C. Richardson 1987. DNA sequence analysis with a modifiedbacteriophage T7 DNA polymerase. Proc. Nat!. Acad. Sc. USA. 84:4767-7 1.Takkinen, K., M. Laukkanen, D. Sizmann, K. Alfthan, T. Immonen, L. Vanne, M.Kaartinen, J.K.C. Knowles and T. Teen 1991. An active single-chain antibodycontaining a cellulase linker domain is secreted by Escherichia coli. Protein Eng.4:837-84 1.Teather, R.M., and P.J.Wood 1982. Use of Congo-red-polysaccharide interactions inenumeration and characterization of cellulolytic bacteria from bovine rumen. App!.Environ. Microbiol. 43:777-780.Teen, T.T., P. Lehtovaara, S. Kauppinen, I. Salovuoni and J. Knowles 1987.Homologous domains in Trichodernia reesei cellulolytic enzymes: gene sequenceand expression of cellobiohydrolase II. Gene 51: 43-52.128Tomme, P., H. Van Tilbeurgh, G. Pettersson, J. Van Damme, J. Vandekerckhove, J.Knowles, T. Teen and M. Claeyssens 1988. Studies of the cellulolytic system ofTrichodernia reesei QM9414. Analysis of domain function in twocellobiohydrolases by limited proteolysis. Eur. J. Biocheni. 170: 575-58 1.Tomme, P., V. Heriban and M. Claeyssens 1990. Adsorption of two cellobiohydrolasesfrom Trichoderma reesei to avicel: evidence for “exo-exo” synergism and possible“loose complex” formation. Biotechnology Lett. 12:525-530.Tomme, P., S. Chavaux, P. Béguin , J. Millet, J.-P. Aubert and M. Claeyssens 1991.Identification of a histidine residue in the active center of endoglucanase D fromClostridiurn thermocellum J. Biol. Cheni. 266: 103 13-10318.Tucker, M.L., M.L. Durbin, M.T. Clegg and L.N. Lewis 1987. Avocado cellulase:nucleotide sequence of a putative full-length eDNA clone and evidence for a smallgene family. Plant Molec. Biol. 9: 197-203.van Tilbeurgh, H., M. Claeyssens and C. K. de Bruyne 1982. The use of 4-methylumbelliferyl and other chromophoric glycosides in the study of cellulolyticenzymes. FEBS Lett. 149:152-156.van Tilbeurgh, H., P. Tomme, M. Claeyssens, R. Bhikhabhai and G. Pettersson 1986.Limited proteolysis of cellobiohydrolase I from Trichoderma reesei. Separation offunctional domains. FEBS Len 204: 223-227.129Warren, R.A.J., C.F. Beck, N.R. Gilkes, D.G. Kilburn, M.L. Langsford, R.C. Miller,Jr., G.P. O’Neill, M. Scheufens and W.K.R.Wong. 1986. Sequenceconservation and region shuffling in an endoglucanase and an exoglucanase fromCellulon1onas flnii. Proteins 1:335-341.Warren, R.A.J, B. Gerhard, N.R. Gilkes, J.B. Owolabi, D.G. Kilbum and R.C. Miller,Jr. 1987. A bifunctional exoglucanase-endoglucanase fusion protein. Gene61:421-427.Whittle, D.J., D.G. Kilburn, R,A.J. Warren and R.C.Miller Jr. 1982. Molecularcloning of a Cellulomonasjmnzi cellulase gene in Escherichia coli. Gene 17: 139-145.Wilcox, G. and G.M. Studnicka 1988. Expression of foreign proteins inmicroorganisms. Biotechnol. Appi. Biochern. 10:500-509.Williamson, G, N.J. Beishaw and M.P. Williamson 1992. 0-glycosylation in Aspergillusglucoamylase. Biocheni J. 282:423-428.Wilson, D.B. 1992. Biochemistry and genetics of Actinomycete cellulases. CriticalReviews in Biotechnology 12:45-63.Wong, R.W.K., B. Gerhard, Z.M. Guo, D.G. Kilburn, R.A.J. Warren and R.C. Miller,Jr. 1986. Characterization and structure of an endoglucanase gene cenA ofCellulornonas fimi. Gene 44:315-324.130Wood, T.M. 1975. Properties and mode of action of cellulases. Biotechnol. Bioeng.Synip. 5:111-137.Wood, T.M. 1988. Preparation of crystalline, amorphous and dyed cellulase substrates.Meth Enzynwl 160:19-25.Wood, T.M. and S.I. McCrae 1979. Synergism between enzymes involved in thesolubilization of native cellulose. pp. 18 1-209. In R.D. Brown, and L. Jurasek(eds.), Advances in Chemical Series vol. 181. American Chemical Society,Washington, D.C.Yablonsky, M.D., T. Bartley, K.O. Elliston, S.K. Kahrs, Z.P. Shalita and D.E. Eveleigh1988. Characterization and cloning of the cellulase complex of Microbisporabispora. In Biochemistiy and Genetics of Cellulose Degradation. Aubert, J.-P.,Béguin, P. and Millet, J. (Editors). FEMS Symposium. 43. Academic Press,London pp. 249-266.Yanisch-Perron, C., C. Vieiera and I. Messing 1985. Improved M13 phage cloningvectors and host strains: nucleotide sequences of the M13mp18 and pUC19vectors. Gene 33: 103-119.Zhou, C., L. Abigair and A.Y.Jong 1990. A protocol for using T7 DNA polymerase inoligonucleotide site-directed mutagenesi s. BioTechniques 8: 503.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items