Mixed Polyplex Micelles with Thermoresponsive and Lysine-Based Zwitterionic Shells Derived from Two Poly(vinyl amine)-Based Block Copolymers.

Two series of poly(vinyl amine) (PVAm)-based block copolymers with zwitterionic and thermoresponsive segments were synthesized by the reversible addition-fragmentation chain transfer polymerization. A mixture of the two copolymers, poly(N-acryloyl-l-lysine) (PALysOH) and poly(N-isopropylacrylamide) (PNIPAM), which have the same cationic PVAm chain but different shell-forming segments, were used to prepare mixed polyplex micelles with DNA. Both PVAm-b-PALysOH and PVAm-b-PNIPAM showed low cytotoxicity, with characteristic assembled structures and stimuli-responsive properties. The cationic PVAm segment in both block copolymers showed site-specific interactions with DNA, which were evaluated by dynamic light scattering, zeta potential, circular dichroism, agarose gel electrophoresis, atomic force microscopy, and transmission electron microscopy measurements. The PVAm-b-PNIPAM/DNA polyplexes showed the characteristic temperature-induced formation of assembled structures in which the polyplex size, surface charge, chiroptical property of DNA, and polymer-DNA binding were governed by the nitrogen/phosphate (N/P) ratio. The DNA binding strength and colloidal stability of the PVAm-b-PALysOH/DNA polyplexes could be tuned by introducing an appropriate amount of zwitterionic PALysOH functionality, while maintaining the polyplex size, surface charge, and chiroptical property, regardless of the N/P ratio. The mixed polyplex micelles showed temperature-induced stability originating from the hydrophobic (dehydrated) PNIPAM chains upon heating, and remarkable stability under salty conditions owing to the presence of the zwitterionic PALysOH chain on the polyplex surface.

Stoichiometric Assembly of the Cellulosome Generates Maximum Synergy for the Degradation of Crystalline Cellulose, as Revealed by In Vitro Reconstitution of the Clostridium thermocellum Cellulosome.

The cellulosome is a supramolecular multienzyme complex formed by species-specific interactions between the cohesin modules of scaffoldin proteins and the dockerin modules of a wide variety of polysaccharide-degrading enzymes. Cellulosomal enzymes bound to the scaffoldin protein act synergistically to degrade crystalline cellulose. However, there have been few attempts to reconstitute intact cellulosomes due to the difficulty of heterologously expressing full-length scaffoldin proteins. We describe the synthesis of a full-length scaffoldin protein containing nine cohesin modules, CipA; its deletion derivative containing two cohesin modules, ΔCipA; and three major cellulosomal cellulases, Cel48S, Cel8A, and Cel9K, of the Clostridium thermocellum cellulosome. The proteins were synthesized using a wheat germ cell-free protein synthesis system, and the purified proteins were used to reconstitute cellulosomes. Analysis of the cellulosome assembly using size exclusion chromatography suggested that the dockerin module of the enzymes stoichiometrically bound to the cohesin modules of the scaffoldin protein. The activity profile of the reconstituted cellulosomes indicated that cellulosomes assembled at a CipA/enzyme molar ratio of 1/9 (cohesin/dockerin = 1/1) and showed maximum synergy (4-fold synergy) for the degradation of crystalline substrate and ∼2.4-fold-higher synergy for its degradation than minicellulosomes assembled at a ΔCipA/enzyme molar ratio of 1/2 (cohesin/dockerin = 1/1). These results suggest that the binding of more enzyme molecules on a single scaffoldin protein results in higher synergy for the degradation of crystalline cellulose and that the stoichiometric assembly of the cellulosome, without excess or insufficient enzyme, is crucial for generating maximum synergy for the degradation of crystalline cellulose.

TG1 integrase-based system for site-specific gene integration into bacterial genomes.

Serine-type phage integrases catalyze unidirectional site-specific recombination between the attachment sites, attP and attB, in the phage and host bacterial genomes, respectively; these integrases and DNA target sites function efficiently when transferred into heterologous cells. We previously developed an in vivo site-specific genomic integration system based on actinophage TG1 integrase that introduces ∼2-kbp DNA into an att site inserted into a heterologous Escherichia coli genome. Here, we analyzed the TG1 integrase-mediated integrations of att site-containing ∼10-kbp DNA into the corresponding att site pre-inserted into various genomic locations; moreover, we developed a system that introduces ∼10-kbp DNA into the genome with an efficiency of ∼10(4) transformants/µg DNA. Integrations of attB-containing DNA into an attP-containing genome were more efficient than integrations of attP-containing DNA into an attB-containing genome, and integrations targeting attP inserted near the replication origin, oriC, and the E. coli "centromere" analogue, migS, were more efficient than those targeting attP within other regions of the genome. Because the genomic region proximal to the oriC and migS sites is located at the extreme poles of the cell during chromosomal segregation, the oriC-migS region may be more exposed to the cytosol than are other regions of the E. coli chromosome. Thus, accessibility of pre-inserted attP to attB-containing incoming DNA may be crucial for the integration efficiency by serine-type integrases in heterologous cells. These results may be beneficial to the development of serine-type integrases-based genomic integration systems for various bacterial species.

Enhancement of the enzymatic activity of Escherichia coli acetyl esterase by a double mutation obtained by random mutagenesis.

A double mutant of Escherichia coli acetyl esterase (EcAE) with enhanced enzymatic activity was obtained by random mutagenesis using error-prone PCR and screening for enzymatic activity by observing halo formation on a tributyrin plate. The mutant contained Leu97Phe (L97F) and Leu209Phe (L209F) mutations. Single mutants L97F and L209F were also constructed and analyzed for kinetic parameters, as well as double mutant L97F/L209F. Kinetic analysis using p-nitrophenyl butyrate as substrate indicated that the k(cat) values of L97F and L97F/L209F were larger than that of the wild-type enzyme, by 8.3-fold and 12-fold respectively, whereas no significant change was observed in the k(cat) value of L209F. The K(m) values of L209F and L97F/L209F were smaller than that of the wild-type enzyme, by 2.9-fold and 2.4-fold respectively, whereas no significant change was observed in the K(m) value of L97F. These results indicate that a combination of an increase in k(cat) values due to the L97F mutation and a decrease in K(m) value due to the L209F mutation renders the k(cat)/K(m) value of the double mutant enzyme 29-fold higher than that of the wild-type enzyme.

Site-specific recombinases as tools for heterologous gene integration.

Site-specific recombinases are the enzymes that catalyze site-specific recombination between two specific DNA sequences to mediate DNA integration, excision, resolution, or inversion and that play a pivotal role in the life cycles of many microorganisms including bacteria and bacteriophages. These enzymes are classified as tyrosine-type or serine-type recombinases based on whether a tyrosine or serine residue mediates catalysis. All known tyrosine-type recombinases catalyze the formation of a Holliday junction intermediate, whereas the catalytic mechanism of all known serine-type recombinases includes the 180° rotation and rejoining of cleaved substrate DNAs. Both recombinase families are further subdivided into two families; the tyrosine-type recombinases are subdivided by the recombination directionality, and the serine-type recombinases are subdivided by the protein size. Over more than two decades, many different site-specific recombinases have been applied to in vivo genome engineering, and some of them have been used successfully to mediate integration, deletion, or inversion in a wide variety of heterologous genomes, including those from bacteria to higher eukaryotes. Here, we review the recombination mechanisms of the best characterized recombinases in each site-specific recombinase family and recent advances in the application of these recombinases to genomic manipulation, especially manipulations involving site-specific gene integration into heterologous genomes.

Site-specific recombination system based on actinophage TG1 integrase for gene integration into bacterial genomes.

Phage integrases are enzymes that catalyze unidirectional site-specific recombination between the attachment sites of phage and host bacteria, attP and attB, respectively. We recently developed an in vivo intra-molecular site-specific recombination system based on actinophage TG1 serine-type integrase that efficiently acts between attP and attB on a single plasmid DNA in heterologous Escherichia coli cells. Here, we developed an in vivo inter-molecular site-specific recombination system that efficiently acted between the att site on exogenous non-replicative plasmid DNA and the corresponding att site on endogenous plasmid or genomic DNA in E. coli cells, and the recombination efficiencies increased by a factor of ~10(1-3) in cells expressing TG1 integrase over those without. Moreover, integration of attB-containing incoming plasmid DNA into attP-inserted E. coli genome was more efficient than that of the reverse substrate configuration. Together with our previous result that purified TG1 integrase functions efficiently without auxiliary host factors in vitro, these in vivo results indicate that TG1 integrase may be able to introduce attB-containing circular DNAs efficiently into attP-inserted genomes of many bacterial species in a site-specific and unidirectional manner. This system thus may be beneficial to genome engineering for a wide variety of bacterial species.

In vitro assembly and cellulolytic activity of a ß-glucosidase-integrated cellulosome complex.

The cellulosome is a supramolecular multi-enzyme complex formed by protein interactions between the cohesin modules of scaffoldin proteins and the dockerin module of various polysaccharide-degrading enzymes. In general, the cellulosome exhibits no detectable ß-glucosidase activity to catalyze the conversion of cellobiose to glucose. Because ß-glucosidase prevents product inhibition of cellobiohydrolase by cellobiose, addition of ß-glucosidase to the cellulosome greatly enhances the saccharification of crystalline cellulose and plant biomass. Here, we report the in vitro assembly and cellulolytic activity of a ß-glucosidase-coupled cellulosome complex comprising the three major cellulosomal cellulases and full-length scaffoldin protein of Clostridium (Ruminiclostridium) thermocellum, and Thermoanaerobacter brockii ß-glucosidase fused to the type-I dockerin module of C. thermocellum. We show that the cellulosome complex composed of nearly equal numbers of cellulase and ß-glucosidase molecules exhibits maximum activity toward crystalline cellulose, and saccharification activity decreases as the enzymatic ratio of ß-glucosidase increases. Moreover, ß-glucosidase-coupled and ß-glucosidase-supplemented cellulosome complexes similarly exhibit maximum activity toward crystalline cellulose (i.e. 1.7-fold higher than that of the ß-glucosidase-free cellulosome complex). These results suggest that the enzymatic ratio of cellulase and ß-glucosidase in the assembled complex is crucial for the efficient saccharification of crystalline cellulose by the ß-glucosidase-integrated cellulosome complex.

Comparative Biochemical Analysis of Cellulosomes Isolated from Clostridium clariflavum DSM 19732 and Clostridium thermocellum ATCC 27405 Grown on Plant Biomass.

The cellulosome is a supramolecular multienzyme complex formed via species-specific interactions between the cohesin modules of scaffoldin proteins and the dockerin modules of a wide variety of polysaccharide-degrading enzymes. Here, we report a comparative analysis of cellulosomes prepared from the thermophilic anaerobic bacteria Clostridium (Ruminiclostridium) clariflavum DSM 19732 and Clostridium (Ruminiclostridium) thermocellum ATCC 27405 grown on delignified rice straw. The results indicate that the isolated C. clariflavum cellulosome exhibits lower activity for insoluble cellulosic substrates and higher activity for hemicellulosic substrates, especially for xylan, compared to the isolated C. thermocellum cellulosome. The C. clariflavum cellulosome was separated into large and small complexes by size exclusion chromatography, and the high xylanase activity of the intact complex is mainly attributed to the small complex. Furthermore, both C. clariflavum and C. thermocellum cellulosomes efficiently converted delignified rice straw into soluble sugars with different compositions, whereas a mixture of these cellulosomes exhibited essentially no synergy for the saccharification of delignified rice straw. This is the first study to report that isolated C. clariflavum cellulosomes exhibit greater xylanase activity than isolated C. thermocellum cellulosomes. We also report the effect of a combination of intact cellulosome complexes isolated from different species on the saccharification of plant biomass.

Structural and thermodynamic analyses of Escherichia coli RNase HI variant with quintuple thermostabilizing mutations.

A combination of five thermostabilizing mutations, Gly23-->Ala, His62-->Pro, Val74-->Leu, Lys95-->Gly, and Asp134-->His, has been shown to additively enhance the thermostability of Escherichia coli RNase HI [Akasako A, Haruki M, Oobatake M & Kanaya S (1995) Biochemistry34, 8115-8122]. In this study, we determined the crystal structure of the protein with these mutations (5H-RNase HI) to analyze the effects of the mutations on the structure in detail. The structures of the mutation sites were almost identical to those of the mutant proteins to which the mutations were individually introduced, except for G23A, for which the structure of the single mutant protein is not available. Moreover, only slight changes in the backbone conformation of the protein were observed, and the interactions of the side chains were almost conserved. These results indicate that these mutations almost independently affect the protein structure, and are consistent with the fact that the thermostabiling effects of the mutations are cumulative. We also determined the protein stability curve describing the temperature dependence of the free energy of unfolding of 5H-RNase HI to elucidate the thermostabilization mechanism. The maximal stability for 5H-RNase HI was as high as that for the cysteine-free variant of Thermus thermophilus RNase HI. In contrast, the heat capacity of unfolding for 5H-RNase H was similar to that for E. coli RNase HI, which is considerably higher than that for T. thermophilus RNase HI. These results suggest that 5H-RNase HI is stabilized, in part, by the thermostabilization mechanism adopted by T. thermophilus RNase HI.

Molecularly imprinted polymer-assisted refolding of lysozyme.

For production of active proteins using heterologous expression systems, refolding of proteins from inclusion bodies often creates a bottleneck due to its poor yield. In this study, we show that molecularly imprinted polymer (MIP) toward native lysozyme promotes the folding of chemically denatured lysozyme. The MIP, which was prepared with 1 M acrylamide, 1 M methacrylic acid, 1 M 2-(dimethylamino)ethyl methacrylate, and 5 mg/mL lysozyme, successfully promoted the refolding of lysozyme, whereas the non-imprinted polymer did not. The refolding yield of 90% was achieved when 15 mg of the MIP was added to 0.3 mg of the unfolded lysozyme. The parallel relationship between the refolding yield and the binding capacity of the MIP suggests that MIP promotes refolding through shifting the folding equilibrium toward the native form by binding the refolded protein.

Stabilization of E. coli Ribonuclease HI by the 'stability profile of mutant protein' (SPMP)-inspired random and non-random mutagenesis.

The change in the structural stability of Escherichia coli ribonuclease HI (RNase HI) due to single amino acid substitutions has been estimated computationally by the stability profile of mutant protein (SPMP) [Ota, M., Kanaya, S. Nishikawa, K., 1995. Desk-top analysis of the structural stability of various point mutations introduced into ribonuclease H. J. Mol. Biol. 248, 733-738]. As well, an effective strategy using random mutagenesis and genetic selection has been developed to obtain E. coli RNase HI mutants with enhanced thermostability [Haruki, M., Noguchi, E., Akasako, A., Oobatake, M., Itaya, M., Kanaya, S., 1994. A novel strategy for stabilization of Escherichia coli ribonuclease HI involving a screen for an intragenic suppressor of carboxyl-terminal deletions. J. Biol. Chem. 269, 26904-26911]. In this study, both methods were combined: random mutations were individually introduced to Lys99-Val101 on the N-terminus of the alpha-helix IV and the preceding beta-turn, where substitutions of other amino acid residues were expected to significantly increase the stability from SPMP, and then followed by genetic selection. Val101 to Ala, Gln, and Arg mutations were selected by genetic selection. The Val101-->Ala mutation increased the thermal stability of E. coli RNase HI by 2.0 degrees C in Tm at pH 5.5, whereas the Val101-->Gln and Val101-->Arg mutations decreased the thermostability. Separately, the Lys99-->Pro and Asn100-->Gly mutations were also introduced directly. The Lys99-->Pro mutation increased the thermostability of E. coli RNase HI by 1.8 degrees C in Tm at pH 5.5, whereas the Asn100-->Gly mutation decreased the thermostability by 17 degrees C. In addition, the Lys99-->Pro mutation altered the dependence of the enzymatic activity on divalent metal ions.

Enzymatic diversity of the Clostridium thermocellum cellulosome is crucial for the degradation of crystalline cellulose and plant biomass.

The cellulosome is a supramolecular multienzyme complex comprised of a wide variety of polysaccharide-degrading enzymes and scaffold proteins. The cellulosomal enzymes that bind to the scaffold proteins synergistically degrade crystalline cellulose. Here, we report in vitro reconstitution of the Clostridium thermocellum cellulosome from 40 cellulosomal components and the full-length scaffoldin protein that binds to nine enzyme molecules. These components were each synthesized using a wheat germ cell-free protein synthesis system and purified. Cellulosome complexes were reconstituted from 3, 12, 30, and 40 components based on their contents in the native cellulosome. The activity of the enzyme-saturated complex indicated that greater enzymatic variety generated more synergy for the degradation of crystalline cellulose and delignified rice straw. Surprisingly, a less complete enzyme complex displaying fewer than nine enzyme molecules was more efficient for the degradation of delignified rice straw than the enzyme-saturated complex, despite the fact that the enzyme-saturated complex exhibited maximum synergy for the degradation of crystalline cellulose. These results suggest that greater enzymatic diversity of the cellulosome is crucial for the degradation of crystalline cellulose and plant biomass, and that efficient degradation of different substrates by the cellulosome requires not only a different enzymatic composition, but also different cellulosome structures.

Cloning and characterization of the gene cluster encoding arthrofactin synthetase from Pseudomonas sp. MIS38.

Arthrofactin is a potent cyclic lipopeptide-type biosurfactant produced by Pseudomonas sp. MIS38. In this work, an arthrofactin synthetase gene cluster (arf) spanning 38.7 kb was cloned and characterized. Three genes termed arfA, arfB, and arfC encode ArfA, ArfB, and ArfC, containing two, four, and five functional modules, respectively. Each module bears condensation, adenylation, and thiolation domains, like other nonribosomal peptide synthetases. However, unlike most of them, none of the 11 modules possess the epimerization domain responsible for the conversion of amino acid residues from L to D form. Possible L- and D-Leu adenylation domains specifically recognized only L-Leu. Moreover, two thioesterase domains are tandemly located at the C-terminal end of ArfC. These results suggest that ArfA, ArfB, and ArfC assemble to form a unique structure. Gene disruption of arfB impaired arthrofactin production, reduced swarming activity, and enhanced biofilm formation.

Isolation and characterization of Rhodococcus sp. strains TMP2 and T12 that degrade 2,6,10,14-tetramethylpentadecane (pristane) at moderately low temperatures.

Branched alkanes including 2,6,10,14-tetramethylpentadecane (pristane) are more resistant to biological degradation than straight-chain alkanes especially under low-temperature conditions, such as 10 degrees C. Two bacterial strains, TMP2 and T12, that are capable of degrading pristane at 10 degrees C were isolated and characterized. Both strains grew optimally at 30 degrees C and were identified as Rhodococcus sp. based on the 16S rRNA gene sequences. Strain T12 degraded comparable amounts of pristane in a range of temperatures from 10 to 30 degrees C and strain TMP2 degraded pristane similarly at 10 and 20 degrees C but did not degrade it at 30 degrees C. These data suggest that the strains have adapted their pristane degradation system to moderately low-temperature conditions.

Protein stabilization utilizing a redefined codon.

Recent advances have fundamentally changed the ways in which synthetic amino acids are incorporated into proteins, enabling their efficient and multiple-site incorporation, in addition to the 20 canonical amino acids. This development provides opportunities for fresh approaches toward addressing fundamental problems in bioengineering. In the present study, we showed that the structural stability of proteins can be enhanced by integrating bulky halogenated amino acids at multiple selected sites. Glutathione S-transferase was thus stabilized significantly (by 5.2 and 5.6 kcal/mol) with 3-chloro- and 3-bromo-l-tyrosines, respectively, incorporated at seven selected sites. X-ray crystallographic analyses revealed that the bulky halogen moieties filled internal spaces within the molecules, and formed non-canonical stabilizing interactions with the neighboring residues. This new mechanism for protein stabilization is quite simple and applicable to a wide range of proteins, as demonstrated by the rapid stabilization of the industrially relevant azoreductase.

Cleavage of a DNA-RNA-DNA/DNA chimeric substrate containing a single ribonucleotide at the DNA-RNA junction with prokaryotic RNases HII.

We have analyzed the cleavage specificities of various prokaryotic Type 2 ribonucleases H (RNases H) on chimeric DNA-RNA-DNA/DNA substrates containing one to four ribonucleotides. RNases HII from Bacillus subtilis and Thermococcus kodakaraensis cleaved all of these substrates to produce a DNA segment with a 5'-monoribonucleotide. Consequently, these enzymes cleaved even the chimeric substrate containing a single ribonucleotide at the DNA-RNA junction (5'-side of the single ribonucleotide). In contrast, Escherichia coli RNase HI and B. subtilis RNase HIII did not cleave the chimeric substrate containing a single ribonucleotide. These results suggest that bacterial and archaeal RNases HII are involved in excision of a single ribonucleotide misincorporated into DNA.

Role of repetitive nine-residue sequence motifs in secretion, enzymatic activity, and protein conformation of a family I.3 lipase.

A family I.3 lipase from Pseudomonas sp. MIS38 (PML) contains 12 repeats of a nine-residue sequence motif in the C-terminal region. To elucidate the role of these repetitive sequences, mutant proteins PML5, PML4, PML1, and PML0, in which 7, 8, 11, and all 12 of the repetitive sequences are deleted, and PMLdelta19, in which 19 C-terminal residues are truncated, were constructed. Escherichia coli DH5 cells carrying the Serratia marcescens Lip system permitted the secretion of the wild-type and all of the mutant proteins except for PMLdelta19, although they were partially accumulated in the cells in an insoluble form as well. Both the secretion level and cellular content of the proteins decreased in the order PML > PML5 > PML4 > PML1 > PML0, indicating that repetitive sequences are not required for secretion of PML but are important for its stability in the cells. All the mutant proteins were purified in a refolded form and their biochemical properties were characterized. CD spectra, the Ca2+ contents, and susceptibility to chymotryptic digestion strongly suggested that the five repetitive sequences remaining in PML5 are sufficient to form a beta-roll structure, whereas the four in PML4 are not. PML5 and PMLdelta19 showed both lipase and esterase activities, whereas PML4, PML1, and PML0 were inactive. These results suggest that the enzymatic activity of PML is not seriously affected by a deletion or truncation at the C-terminal region as long as a succession of repetitive sequences can build a beta-roll structure.

Importance of an N-terminal extension in ribonuclease HII from Bacillus stearothermophilus for substrate binding.

The gene encoding ribonuclease HII from Bacillus stearothermophilus was cloned and expressed in Escherichia coli. The overproduced protein, Bst-RNase HII, was purified and biochemically characterized. Bst-RNase HII, which consists of 259 amino acid residues, showed the highest amino acid sequence identity (50.2%) to Bacillus subtilis RNase HII. Like B. subtilis RNase HII, it exhibited Mn2+-dependent RNase H activity. It was, however, more thermostable than B. subtilis RNase HII. When the Bst-RNase HII amino acid sequence is compared with that of Thermococcus kodakaraensis RNase HII, to which it shows 29.8% identity, 30 residues are observed to be truncated from the C-terminus and there is an extension of 71 residues at the N-terminus. The C-terminal truncation results in the loss of the alpha9 helix, which is rich in basic amino acid residues and is therefore important for substrate binding. A truncated protein, Delta59-Bst-RNase HII, in which most of the N-terminal extension was removed, completely lost its RNase H activity. Surface plasmon resonance analysis indicated that this truncated protein did not bind to the substrate. These results suggest that the N-terminal extension of Bst-RNase HII is important for substrate binding. Because B. subtilis RNase HII has an N-terminal extension of the same length and these extensions contain a region in which basic amino acid residues are clustered, the Bacillus enzymes may represent a novel type of RNase H which possesses a substrate-binding domain at the N-terminus.

Cell-free protein synthesis and substrate specificity of full-length endoglucanase CelJ (Cel9D-Cel44A), the largest multi-enzyme subunit of the Clostridium thermocellum cellulosome.

Endoglucanase CelJ (Cel9D-Cel44A) is the largest multi-enzyme subunit of the Clostridium thermocellum cellulosome and is composed of glycoside hydrolase (GH) families 9 and 44 (GH9 and GH44) and carbohydrate-binding module (CBM) families 30 and 44 (CBM30 and CBM44). The study of CelJ has been hampered by the inability to isolate full-length CelJ from recombinant Escherichia coli cells. Here, full-length CelJ and its N- and C-terminal segments, CBM30-GH9 (Cel9D) and GH44-CBM44 (Cel44A), were synthesized using a wheat germ cell-free protein synthesis system and then were purified to homogeneity. Analysis of the substrate specificities of CelJ and its derivatives demonstrated that the fusion of Cel9D and Cel44A results in threefold synergy for the degradation of xyloglucan, one of the major structural polysaccharides of plant cell walls. Because CelJ displayed broad substrate specificity including significant carboxymethylcellulase (CMCase) and xylanase activities in addition to high xyloglucanase activity, CelJ may play an important role in the degradation of plant cell walls, which are composed of highly heterogeneous polysaccharides. Furthermore, because Cel9D, but not Cel44A, acts as a semi-processive endoglucanase, the different modes of action between Cel9D and Cel44A may be responsible for the observed synergistic effect on the activity of CelJ (Cel9D-Cel44A).