Structural and functional relationships in two families of beta-1,4-glycanases

Changes in the molecular-size distribution of insoluble celluloses by the action of recombinant Cellulomonas fimi cellulases

Specific patterns of attacks of cotton, bacterial cellulose and bacterial microcrystalline cellulose (BMCC) by recombinant cellulases of Cellulomonas fimi were investigated. Molecular-size distributions of the celluloses were determined by high-performance size-exclusion chromatography. Chromatography of cotton and bacterial celluloses revealed single major peaks centered over progressively lower molecular-mass positions during attack by endoglucanase CenA. In advanced stages, a second peak appeared at very low average size (approx. 11 glucosyl units); ultimate weight losses were approximately 30%. The isolated catalytic domain of CenA, p30, gave results very similar to those with complete CenA. CenA did not effectively depolymerize or solubilize BMCC significantly. Molecular-size distributions of cotton and bacterial cellulose incubated with endoglucanases CenB or CenD exhibited one major peak regardless of incubation time; low-molecular-mass fragments did not accumulate. Weight losses were 40 and 35% respectively. The single peak shifted to lower-molecular-mass positions as incubation continued, but high-molecular-mass material persisted. CenB and CenD readily attacked and solubilized BMCC (approx. 70%). We conclude that CenA attacks cellulose by preferentially cleaving completely through the cellulose microfibrils at the amorphous sites, and much more slowly by degrading the crystalline surfaces. Conversely, CenB and CenD cleave the amorphous regions much less efficiently while vigorously degrading the surfaces of the crystalline regions of the microfibrils.

Cellobiohydrolase A (CbhA) from the cellulolytic bacterium Cellulomonas fimi is a beta-1,4-exocellobiohydrolase analogous to Trichoderma reesei CBH II

The gene cbhA from the cellulolytic bacterium Cellulomonas fimi encodes a protein of 872 amino acids designated cellobiohydrolase A (CbhA). Mature CbhA contains 832 amino acid residues and has a predicted molecular mass of 85,349 Da. It is composed of five domains: an N-terminal catalytic domain, three repeated sequences of 95 amino acids, and a C-terminal cellulose-binding domain typical of other C. fimi glycanases. The structure and enzymatic activities of the CbhA catalytic domain are closely related to those of CBH II, an exocellobiohydrolase in the glycosyl hydrolase family B from the fungus Trichoderma reesei. CbhA is the first such enzyme to be characterized in bacteria. The data support the proposal that extended loops around the active site distinguish exohydrolases from endohydrolases in this enzyme family.

Alteration of the cleavage mode and of the transglycosylation reactions of the xylanase A of Streptomyces lividans 1326 by site-directed mutagenesis of the Asn173 residue

The amino acid replacement of Asn173 by Asp in the xylanase A (Xln A) of Streptomyces lividans significantly altered its enzymic properties. A time-course hydrolysis of xylan showed that the altered xylanase ([N173D] Xln A) initially produced larger amounts of xylose (X1), xylobiose (X2) and xylotriose (X3) than Xln A, but less xylotetraose (X4). The bond-cleavage frequencies were determined for both enzymes using xylopentaose (X5), xylotetraose (X4) and xylotriose (X3) labelled at the reducing end of the molecule. Xln A hydrolysed X5, yielding 56% X2 and 44% X3, while [N173D]Xln A liberated 90% X2 and only 10% X3. Both enzymes hydrolysed X4 into 100% X2 and X3 into 100% X1. Transglycosylation reactions were detected in HPLC hydrolysis patterns using high substrate concentrations, where larger products than the starting substrates were formed. Their subsequent degradation also affected the yield of hydrolysis products. Using X5 as substrate, products from xylohexaose (X6) up to xylooligosides larger than xylooctaose (X8) were synthesized by Xln A, while [N173D]Xln A produced only a small amount of xyloheptaose (X7) and X8. Xln A hydrolysed X5 into an equivalent amount of X4 and X2 and 1.5-fold more X3. However, [N173D]Xln A yielded the same amount of X3 and X2 but half as much X4. With X4 as substrate, Xln A synthesized twofold more X7 and X6 than [N173D]Xln A. Xln A liberated 1.4-fold more X3 than X2, while [N173D]Xln A yielded twofold more X2 than X3. Xln A liberated almost fourfold more X2 than X1 from X3, while [N173D]Xln A produced only twofold more X2 than X1. These results indicated that the negative charge introduced by the mutation greatly affected the transglycosylation reaction catalysed by this xylanase.

The biological degradation of cellulose

Cellulolytic microorganisms play an important role in the biosphere by recycling cellulose, the most abundant carbohydrate produced by plants. Cellulose is a simple polymer, but it forms insoluble, crystalline microfibrils, which are highly resistant to enzymatic hydrolysis. All organisms known to degrade cellulose efficiently produce a battery of enzymes with different specificities, which act together in synergism. The study of cellulolytic enzymes at the molecular level has revealed some of the features that contribute to their activity. In spite of a considerable diversity, sequence comparisons show that the catalytic cores of cellulases belong to a restricted number of families. Within each family, available data suggest that the various enzymes share a common folding pattern, the same catalytic residues, and the same reaction mechanism, i.e. either single substitution with inversion of configuration or double substitution resulting in retention of the beta-configuration at the anomeric carbon. An increasing number of three-dimensional structures is becoming available for cellulases and xylanases belonging to different families, which will provide paradigms for molecular modeling of related enzymes. In addition to catalytic domains, many cellulolytic enzymes contain domains not involved in catalysis, but participating in substrate binding, multi-enzyme complex formation, or possibly attachment to the cell surface. Presumably, these domains assist in the degradation of crystalline cellulose by preventing the enzymes from being washed off from the surface of the substrate, by focusing hydrolysis on restricted areas in which the substrate is synergistically destabilized by multiple cutting events, and by facilitating recovery of the soluble degradation products by the cellulolytic organism. In most cellulolytic organisms, cellulase synthesis is repressed in the presence of easily metabolized, soluble carbon sources and induced in the presence of cellulose. Induction of cellulases appears to be effected by soluble products generated from cellulose by cellulolytic enzymes synthesized constitutively at a low level. These products are presumably converted into true inducers by transglycosylation reactions. Several applications of cellulases or hemicellulases are being developed for textile, food, and paper pulp processing. These applications are based on the modification of cellulose and hemicellulose by partial hydrolysis. Total hydrolysis of cellulose into glucose, which could be fermented into ethanol, isopropanol or butanol, is not yet economically feasible. However, the need to reduce emissions of greenhouse gases provides an added incentive for the development of processes generating fuels from cellulose, a major renewable carbon source.

Visualization of the adsorption of a bacterial endo-beta-1,4-glucanase and its isolated cellulose-binding domain to crystalline cellulose

Endo-beta-1,4-glucanase A (CenA), a cellulase from the bacterium Cellulomonas fimi, is composed of two domains: a catalytic domain and a cellulose-binding domain. Adsorption of CenA and its isolated cellulose-binding domain (CBD.PTCenA) to Valonia cellulose microcrystals was examined by transmission electron microscopy using an antibody sandwich technique (CenA/CBD.PTCenA-alpha CenA IgG-protein A-gold conjugate). Adsorption of both CenA and CBD.PTCenA occurred along the lengths of the microcrystals, with an apparent preference for certain crystal faces or edges. CenA or CBD.PTCenA, but not the isolated catalytic domain, were shown to prevent the flocculation of microcrystalline bacterial cellulose. The cellulose-binding domain may assist crystalline cellulose hydrolysis in vitro by promoting substrate dispersion.

New families in the classification of glycosyl hydrolases based on amino acid sequence similarities

301 glycosyl hydrolases and related enzymes corresponding to 39 EC entries of the I.U.B. classification system have been classified into 35 families on the basis of amino-acid-sequence similarities [Henrissat (1991) Biochem. J. 280, 309-316]. Approximately half of the families were found to be monospecific (containing only one EC number), whereas the other half were found to be polyspecific (containing at least two EC numbers). A > 60% increase in sequence data for glycosyl hydrolases (181 additional enzymes or enzyme domains sequences have since become available) allowed us to update the classification not only by the addition of more members to already identified families, but also by the finding of ten new families. On the basis of a comparison of 482 sequences corresponding to 52 EC entries, 45 families, out of which 22 are polyspecific, can now be defined. This classification has been implemented in the SWISS-PROT protein sequence data bank.

Cellulose-binding polypeptides from Cellulomonas fimi: endoglucanase D (CenD), a family A beta-1,4-glucanase

Five cellulose-binding polypeptides were detected in Cellulomonas fimi culture supernatants. Two of them are CenA and CenB, endo-beta-1,4-glucanases which have been characterized previously; the other three were previously uncharacterized polypeptides with apparent molecular masses of 120, 95, and 75 kDa. The 75-kDa cellulose-binding protein was designated endoglucanase D (CenD). The cenD gene was cloned and sequenced. It encodes a polypeptide of 747 amino acids. Mature CenD is 708 amino acids long and has a predicted molecular mass of 74,982 Da. Analysis of the predicted amino acid sequence of CenD shows that the enzyme comprises four domains which are separated by short linker polypeptides: an N-terminal catalytic domain of 405 amino acids, two repeated sequences of 95 amino acids each, and a C-terminal domain of 105 amino acids which is > 50% identical to the sequences of cellulose-binding domains in Cex, CenA, and CenB from C. fimi. Amino acid sequence comparison placed the catalytic domain of CenD in family A, subtype 1, of beta-1,4-glycanases. The repeated sequences are more than 40% identical to the sequences of three repeats in CenB and are related to the repeats of fibronectin type III. CenD hydrolyzed the beta-1,4-glucosidic bond with retention of anomeric configuration. The activities of CenD towards various cellulosic substrates were quite different from those of CenA and CenB.

Hidden domains and active site residues in beta-glycanase-encoding gene sequences?

Reading-frame corrective shifts in the nucleotide sequence upstream, within, or downstream from the putative coding region of several beta-glycanase-encoding genes reported in the literature reveal hidden active-site residues or even additional domains, including a cellulose-binding domain on a beta-mannanase-encoding gene. These findings also help in assigning, to cellulase family A, two enzymes previously found to lack sequence similarity with known cellulase families.

Endoglucanase CasA from alkalophilic Streptomyces strain KSM-9 is a typical member of family B of beta-1,4-glucanases

CasA is an endo-beta-1,4-glucanase from Streptomyces KSM-9 belonging to family B of beta-1,4-glucanases. A previous analysis of a portion of the corresponding gene (casA) revealed sequencing errors in a region encoding part of the catalytic site. Additional errors in the original sequence were suspected, based on sequence comparison of the C terminus of CasA with other members of its family. Re-sequencing of the remainder of the casA coding region showed that CasA is a typical member of family B.

Crystallization and preliminary X-ray diffraction analysis of the catalytic domain of Cex, an exo-beta-1,4-glucanase and beta-1,4-xylanase from the bacterium Cellulomonas fimi

Single crystals of the catalytic domain of Cex, an exo-beta-1,4-glucanase and beta-1,4-xylanase from the cellulolytic bacterium Cellulomonas fimi, have been grown in the presence of polyethylene glycol 4000 using the vapour diffusion technique. The crystals, which diffract to better than 2.0 A resolution, belong to space group P4(1)2(1)2 or P4(3)2(1)2 and have cell constants: a = b = 88.21 A, c = 81.10 A; alpha = beta = gamma = 90 degrees.

Specificity mapping of cellulolytic enzymes: classification into families of structurally related proteins confirmed by biochemical analysis

The specificities of 15 cellulolytic enzymes have been examined using chromophoric glycosides derived from D-glucose, cellobiose, higher cellooligosaccharides, lactose, D-xylose, and beta-(1,4)-xylobiose. Coinciding with a classification based on hydrophobic cluster analysis of amino acid sequences, six families each showing a characteristic specificity pattern were observed. Furthermore, in these cases where the anomeric forms of reaction products were determined, results seem to indicate conservation of intrinsic reaction mechanism (single or double displacement) within each family. On the other hand, the low molecular weight substrates do not discriminate exo- from endocellulases. This functional differentiation is speculated to originate from the presence, in exoenzymes, of a tunnel-shaped active site formed by extra loops in their structure.

Cloning and nucleotide sequence of celA1, and endo-beta-1,4-glucanase-encoding gene from Streptomyces halstedii JM8

The celA1 gene encoding an endo-beta-1,4-glucanase from a mesophilic actinomycete, strain JM8, identified as Streptomyces halstedii, was cloned and expressed in S. lividans JI66. From the nucleotide sequence of a 1.7-kb DNA fragment we identified an open reading frame of 963 nucleotides encoding a protein of 321 amino acids, starting at TTG (instead of ATG). The Cel1 mature enzyme is a protein of 294 amino acids (after signal peptide cleavage) and can be included in the beta-glycanase family B (N. R. Gilkes, B. Henrissat, D. G. Kilburn, R. C. Miller, Jr., and R. A. J. Warren, Microbiol. Rev. 55:303-315, 1991). The Cel1 enzyme lacks a cellulose-binding domain as predicted by computer analysis of the sequence and confirmed by Avicel binding experiments. The promoter region of celA1 was identified by S1 mapping; the -35 region closely resembles those of housekeeping Streptomyces promoters. Three imperfectly repeated sequences of 15, 15, and 14 nucleotides were found upstream from celA1 [ATTGGGACCGCTTCC-(N85)-ATTGGGACCGCTTCC-(N2)-TGGGAGC GCTCCCA]; The 14-nucleotide sequence has a perfect palindrome identical to that found in several cellulase-encoding genes from Thermomonospora fusca, an alkalophilic Streptomyces strain, and Streptomyces lividans. This sequence has been implicated in the mechanism of induction exerted by cellobiose. Using an internal celA1 probe, we detected similar genes in several other Streptomyces species, most of them cellulase producers.