Structure of a Bacillus subtilis endo-beta-1,4-glucanase gene

Cloning and Characterization of an Endoglucanase Gene from Actinomyces sp. Korean Native Goat 40

A gene from Actinomyces sp. Korean native goat (KNG) 40 that encodes an endo-β-1,4-glucanase, EG1, was cloned and expressed in Escherichia coli (E. coli) DH5α. Recombinant plasmid DNA from a positive clone with a 3.2 kb insert hydrolyzing carboxyl methyl-cellulose (CMC) was designated as pDS3. The entire nucleotide sequence was determined, and an open-reading frame (ORF) was deduced. The ORF encodes a polypeptide of 684 amino acids. The recombinant EG1 produced in E. coli DH5α harboring pDS3 was purified in one step using affinity chromatography on crystalline cellulose and characterized. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis/zymogram analysis of the purified enzyme revealed two protein bands of 57.1 and 54.1 kDa. The amino terminal sequences of these two bands matched those of the deduced ones, starting from residue 166 and 208, respectively. Putative signal sequences, a Shine–Dalgarno-type ribosomal binding site, and promoter sequences related to the consensus sequences were deduced. EG1 has a typical tripartite structure of cellulase, a catalytic domain, a serine-rich linker region, and a cellulose-binding domain. The optimal temperature for the activity of the purified enzyme was 55°C, but it retained over 90% of maximum activity in a broad temperature range (40°C to 60°C). The optimal pH for the enzyme activity was 6.0. Kinetic parameters, K_m and V_max of rEG1 were 0.39% CMC and 143 U/mg, respectively.

Isolation, characterization and manipulation of cellulase genes

The complete hydrolysis of cellulose requires a number of different enzymes including endoglucanase, exoglucanase and beta-glucosidase. These enzymes function in concert as part of a 'cellulase'complex called a cellulosome. In order (i) to develop a better understanding of the biochemical nature of the cellulase complex as well as the genetic regulation of its integral components and (ii) to utilize cellulases either as purified enzymes or as part of an engineered organism for a variety of purposes, researchers have, as a first step, used recombinant DNA technology to isolate the genes for these enzymes from a variety of organisms. This review provides some perspective on the current status of the isolation, characterization and manipulation of cellulase genes and specifically discusses (i) strategies for the isolation of endoglucanase, exoglucanase and beta-glucosidase genes; (ii) DNA sequence characterization of the cellulase genes and their accompanying regulatory elements; (iii) the expression of cellulase genes in heterologous host organisms and (iv) some of the proposed uses for isolated cellulase genes.

Regulation of carbon catabolism in Bacillus species

The gram-positive bacterium Bacillus subtilisis capable of using numerous carbohydrates as single sources of carbon and energy. In this review, we discuss the mechanisms of carbon catabolism and its regulation. Like many other bacteria, B. subtilis uses glucose as the most preferred source of carbon and energy. Expression of genes involved in catabolism of many other substrates depends on their presence (induction) and the absence of carbon sources that can be well metabolized (catabolite repression). Induction is achieved by different mechanisms, with antitermination apparently more common in B. subtilis than in other bacteria. Catabolite repression is regulated in a completely different way than in enteric bacteria. The components mediating carbon catabolite repression in B. subtilis are also found in many other gram-positive bacteria of low GC content.

A novel thermostable multidomain 1,4-beta-xylanase from 'Caldibacillus cellulovorans' and effect of its xylan-binding domain on enzyme activity

The nucleotide sequence of the complete xynA gene, encoding a novel multidomain xylanase XynA of 'Caldibacillus cellulovorans', was determined by genomic-walking PCR. The putative XynA comprises an N-terminal domain (D1), recently identified as a xylan-binding domain (XBD), homologous to non-catalytic thermostabilizing domains from other xylanases. D1 is followed by a xylanase catalytic domain (D2) homologous to family 10 glycosyl hydrolases. Downstream of this domain two cellulose-binding domains (CBD), D3 and D4, were found linked via proline-threonine (PT)-rich peptides. Both CBDs showed sequence similarity to family IIIb CBDs. Upstream of xynA an incomplete open reading frame was identified, encoding a putative C-terminal CBD homologous to family IIIb CBDs. Two expression plasmids encoding the N-terminal XBD plus the catalytic domain (XynAd1/2) and the xylanase catalytic domain alone (XynAd2) were constructed and the biochemical properties of the recombinant enzymes compared. The absence of the XBD resulted in a decrease in thermostability of the catalytic domain from 70 degrees C (XynAd1/2) to 60 degrees C (XynAd2). Substrate-specificity experiments and analysis of the main products released from xylan hydrolysis indicate that both recombinant enzymes act as endo-1, 4-beta-xylanases, but differ in their ability to cleave small xylooligosaccharides.

Purification and characterization of cellulases produced by two Bacillus strains

Cellulases produced by two Bacillus strains, CH43 and HR68, isolated from hot springs in Zimbabwe, were purified to homogeneity from culture supernatants. Both enzymes had molecular mass of 40 kDa and isoelectric point of 5.4. The enzymes also resembled each other in N-terminal amino acid sequence which was Ala-Gly-Thr-Lys-Thr-Pro-Val-Ala-Lys-Asn-Gly-Gln, showing 100% homology with that of endoglucanases from Bacillus subtilis belonging to glycoside hydrolase family five. The cellulases were optimally active in the pH range of 5-6.5. The optimum temperature was 65 and 70 degrees C for the endoglucanase of CH43 and HR68, respectively. The CH43 enzyme was stable at 50 degrees C in a pH range of 6-10, and HR68 at pH 6-8. Both the enzymes retained complete activity for at least 24 h at 50 degrees C. The enzymes showed highest activity with beta-glucan as substrate followed by carboxymethylcellulose. Significant activity was also observed with crystalline forms of cellulose such as filter paper and Avicel, particularly for HR68 cellulase. For carboxymethycellulose, the CH43 and HR68 cellulases had a Km of 1.5 and 1.7 mg ml(-1), respectively, and Vmax of 0.93 and 1.70 mmol glucose min(-1) mg protein(-1) respectively. The activity of the enzymes was not influenced by most metal ions at 1 mM concentration, but was increased by about 38% by Co2+. The inhibition by Hg2+ and Mn2+ was higher for CH43 than for HR68 enzyme. Ag+ inhibited the CH43 activity but stimulated the HR68 activity. The CH43 cellulase was inhibited by N-bromosuccinimide and iodoacetamide while HR68 was unaffected.

A gene encoding a novel multidomain beta-1,4-mannanase from Caldibacillus cellulovorans and action of the recombinant enzyme on kraft pulp

Genomic walking PCR was used to obtained a 4,567-bp nucleotide sequence from Caldibacillus cellulovorans. Analysis of this sequence revealed that there were three open reading frames, designated ORF1, ORF2, and ORF3. Incomplete ORF1 encoded a putative C-terminal cellulose-binding domain (CBD) homologous to members of CBD family IIIb, while putative ORF3 encoded a protein of unknown function. The putative ManA protein encoded by complete manA ORF2 was an enzyme with a novel multidomain structure and was composed of four domains in the following order: a putative N-terminal domain (D1) of unknown function, an internal CBD (D2), a beta-mannanase catalytic domain (D3), and a C-terminal CBD (D4). All four domains were linked via proline-threonine-rich peptides. Both of the CBDs exhibited sequence similarity to family IIIb CBDs, while the mannanase catalytic domain exhibited homology to the family 5 glycosyl hydrolases. The purified recombinant enzyme ManAd3 expressed from the cloned catalytic domain (D3) exhibited optimum activity at 85 degrees C and pH 6.0 and was extremely thermostable at 70 degrees C. This enzyme exhibited high specificity with the substituted galactomannan locust bean gum, while more substituted galacto- and glucomannans were poorly hydrolyzed. Preliminary studies to determine the effect of the recombinant ManAd3 and a recombinant thermostable beta-xylanase on oxygen-delignified Pinus radiata kraft pulp revealed that there was an increase in the brightness of the bleached pulp.

Expression and processing of a bacterial endoglucanase in transgenic mice

The C6.5 endoglucanase from Bacillus subtilis catalyzes the hydrolyses of beta-glucans. This enzyme, which is also produced by many ruminant microbes, is not part of the normal digestive repertoire of monogastric animals. We have generated transgenic mice which express the C6.5 endoglucanase gene specifically in the pancreas with secretion of the enzyme into the small intestine. The secreted enzyme has a molecular mass of 55 kDa which is reduced by protease digestion to the principal forms of 37 and 35 kDa. These truncated forms are resistant to further protease degradation and exhibit enhanced specific activity compared to the native enzyme. These results encourage further investigation of the utility of this transgene for enhancing the digestive capability of monogastric animals.

Gene cloning, DNA sequencing, and expression of thermostable beta-mannanase from Bacillus stearothermophilus

A DNA genomic library constructed from Bacillus stearothermophilus, a gram-positive, facultative thermophilic aerobe that secretes a thermostable beta-mannanase, was screened for mannan hydrolytic activity. Recombinant beta-mannanase activity was detected on the basis of the clearing of halos around Escherichia coli colonies grown on a dye-labelled substrate, Remazol brilliant blue-locust bean gum. The nucleotide sequence of the mannanase gene, manF, corresponded to an open reading frame of 2,085 bp that codes for a 32-amino-acid signal peptide and a mature protein with a molecular mass of 76,089 Da. From sequence analysis, ManF belongs to glycosyl hydrolase family 5 and exhibits higher similarity to eukaryotic than to bacterial mannanases. The manF coding sequence was subcloned into the pH6EX3 expression plasmid and expressed in E. coli as a recombinant fusion protein containing a hexahistidine N-terminal sequence. The fusion protein has thermostability similar to the native enzyme and was purified by Ni2+ affinity chromatography. The values for the kinetic parameters Vmax and Km were 384 U/mg and 2.4 mg/ml, respectively, for the recombinant mannanase and were comparable to those of the native enzyme.

Expression of a bacterial endo (1-4)-beta-glucanase gene in mammalian cells and post translational modification of the gene product

An endo (1-4)-beta-glucanase gene C6.5 from Bacillus subtilis has been expressed in Chinese hamster ovary (CHO) cells and pancreatic 266-6 cells. The fusion gene, stably transfected into CHO cells consisted of the mouse Amy-2.2 signal peptide coding sequence and the endoglucanase gene C6.5 transcribed from the early SV40 promoter/enhancer, using the dihydrofolate reductase gene as a selective marker. The gene construct transfected into pancreatic 266-6 cells consisted of the mouse Amy-2.2 promoter/enhancer and signal peptide coding sequence and the same C6.5 sequences using the xanthine-guanine phosphoribosyl transferase gene (gpt) as the selective marker. The stably transfected CHO cells synthesized endoglucanase at 1.1 U/mg cell protein in a 72 h culture, with 89% of the activity secreted into the culture fluid in a glycosylated form of 66 kDa as compared with the unglycosylated 53 kDa form expressed in E. coli. Glycosylation did not change the specific activity, protease resistance, or cellulose binding of the endoglucanase as compared to the unglycosylated form of the enzyme from E. coli. The level of expression in the stably transfected pancreatic cells was substantially lower at 0.3 mU/mg cell protein with all detectable activity present in the culture fluid. The secreted enzyme from pancreatic cells was glycosylated with a mass similar to that secreted from CHO cells.

Compilation and analysis of Bacillus subtilis sigma A-dependent promoter sequences: evidence for extended contact between RNA polymerase and upstream promoter DNA

Sequence analysis of 236 promoters recognized by the Bacillus subtilis sigma A-RNA polymerase reveals an extended promoter structure. The most highly conserved bases include the -35 and -10 hexanucleotide core elements and a TG dinucleotide at position -15, -14. In addition, several weakly conserved A and T residues are present upstream of the -35 region. Analysis of dinucleotide composition reveals A2- and T2-rich sequences in the upstream promoter region (-36 to -70) which are phased with the DNA helix: An tracts are common near -43, -54 and -65; Tn tracts predominate at the intervening positions. When compared with larger regions of the genome, upstream promoter regions have an excess of An and Tn sequences for n > 4. These data indicate that an RNA polymerase binding site affects DNA sequence as far upstream as -70. This sequence conservation is discussed in light of recent evidence that the alpha subunits of the polymerase core bind DNA and that the promoter may wrap around RNA polymerase.

celA, another gene coding for a multidomain cellulase from the extreme thermophile Caldocellum saccharolyticum

Caldocellum saccharolyticum is an extremely thermophilic anaerobic bacterium capable of growth on cellulose and hemicellulose as sole carbon sources. Cellulase and hemicellulase genes have been found clustered together on its genome. The gene for one of the cellulases (celA) was isolated on a lambda genomic library clone, sequenced and found to comprise a large open-reading frame of 5253 base pairs that could be translated into a peptide of 1751 amino acids. To date, it is the largest cellulase gene sequenced. The translated product is a multidomain structure composed of two catalytic domains and two cellulose-binding domains linked by proline-threonine-rich regions (PT linkers). The N-terminal domain of celA encodes for an endoglucanase activity on carboxymethylcellulose, consistent with its high homology to the sequences of several other endo-1, 4-beta-D-glucanases. The carboxyterminal domain shows sequence homology with a cellulase from Clostridium thermocellum (CelS), which is known to act synergistically with a second component to hydrolyze crystalline cellulose. In the absence of a Caldocellum homologue for this second protein, we can detect no activity from this domain.

Mutation analysis of the cellulose-binding domain of the Clostridium cellulovorans cellulose-binding protein A

Cellulose-binding protein A (CbpA) has been previously shown to mediate the interaction between crystalline cellulose substrates and the cellulase enzyme complex of Clostridium cellulovorans. CbpA contains a family III cellulose-binding domain (CBD) which, when expressed independently, binds specifically to crystalline cellulose. A series of N- and C-terminal deletions and a series of small internal deletions of the CBD were created to determine whether the entire region previously described as a CBD is required for the cellulose-binding function. The N- and C-terminal deletions reduced binding affinity by 10- to 100-fold. Small internal deletions of the CBD resulted in substantial reduction of CBD function. Some, but not all, point mutations throughout the sequence had significant disruptive effects on the binding ability of the CBD. Thus, mutations in any region of the CBD had effects on the binding of the fragment to cellulose. The results indicate that the entire 163-amino-acid region of the CBD is required for maximal binding to crystalline cellulose.

Nucleotide sequence of an endo-beta-1,4-glucanase gene from Bacillus subtilis CK-2

The gene encoding endo-beta-1,4-glucanase in Bacillus subtilis CK-2 was cloned into Escherichia coli DH5 alpha, and the nucleotide sequence determined. The 1500 bp gene encodes a protein of 499 amino-acid residues with a calculated molecular mass of 55,261, and is equipped with a typical B. subtilis signal peptide. Nucleotide sequence comparison revealed only 2 basepairs deviation between this gene and the endo-beta-1,4-glucanase gene of B. subtilis PAP115, and 93% to 95% homology was found between the amino acid sequences of these enzymes and other B. subtilis endo-beta-1,4-glucanases. Regions of similarity were also observed between the carboxy-terminal part of these enzymes and the part of the B. lautus PL236 celA enzyme constituting the cellulose-binding domain.

Molecular analysis of the major cellulase (CelV) of Erwinia carotovora: evidence for an evolutionary "mix-and-match" of enzyme domains

The structural gene for the major cellulase of Erwinia carotovora subspecies carotovora (Ecc) was isolated and expressed in Escherichia coli. Sequencing of the gene (celV) revealed a typical signal sequence and two functional domains in the enzyme; a catalytic domain linked by a short proline/threonine-rich linker to a cellulose-binding domain (CBD). The deduced amino acid sequence of the catalytic domain showed homology with cellulases of Family A, including enzymes from Bacillus spp. and Erwinia chrysanthemi CelZ, whereas the CBD showed homology with cellulases from several diverse families, supporting a "mix-and-match" hypothesis for evolution of this domain. Analysis of the substrate specificity of CelV showed it to be an endoglucanase with some exoglucanase activity. The pH optimum is about 7.0 and the temperature optimum about 42 degrees C. CelV is secreted by Ecc and by the taxonomically related Erwinia carotovora subspecies atroseptica (Eca) but not by E. coli. Overproduction of the enzyme from multicopy plasmids in Ecc appears to overload the secretory mechanism.

Sequencing of a Clostridium thermocellum gene (cipA) encoding the cellulosomal SL-protein reveals an unusual degree of internal homology

It is known that two proteins of the cellulosomal complex of Clostridium thermocellum (SL and SS) together degrade crystalline cellulose. SL is a glycoprotein of 210,000 Da which enhances the binding to cellulose and the activity of SS, an endoglucanase of 83,000 Da. We have previously reported the cloning of a DNA fragment encoding the N-terminal end of the SL protein using antibodies raised against the native protein. A chromosomal walking approach using an EcoRI and a Bam HI-Sau3A gene library allowed us to isolate the C-terminal end of the gene. Sequencing of both fragments revealed the existence of a leader peptide as has been found in cellulases of the same organism. This leader sequence is followed by a stretch of 14 amino acids that is identical to the N-terminal amino acid sequence of the native secreted protein. The open reading frame (ORF) of this gene encodes a protein of 196,800 Da and is followed by a hairpin loop that could be involved in transcription termination. Within the open reading frame (ORF), we found nine internal repeated elements (IREs) of about 500 nucleotides each. Seven of these sequences displayed 98-100% homology and were located adjacent to each other within the structural gene without intervening regions. The remaining two, located on the N-terminal end of the gene, showed a significantly lower homology. Bearing in mind the inherent instability of reiterated regions, we confirmed the authenticity of our clones by Southern blot analysis using chromosomal C. thermocellum DNA and ruled out the possibility of rearrangements during the cloning and sequencing process. The sequenced gene is designated cipA and the encoded SL protein CipA.

Protein secretion in Bacillus species

Bacilli secrete numerous proteins into the environment. Many of the secretory proteins, their export signals, and their processing steps during secretion have been characterized in detail. In contrast, the molecular mechanisms of protein secretion have been relatively poorly characterized. However, several components of the protein secretion machinery have been identified and cloned recently, which is likely to lead to rapid expansion of the knowledge of the protein secretion mechanism in Bacillus species. Comparison of the presently known export components of Bacillus species with those of Escherichia coli suggests that the mechanism of protein translocation across the cytoplasmic membrane is conserved among gram-negative and gram-positive bacteria differences are found in steps preceding and following the translocation process. Many of the secretory proteins of bacilli are produced industrially, but several problems have been encountered in the production of Bacillus heterologous secretory proteins. In the final section we discuss these problems and point out some possibilities to overcome them.

The beta-mannanase from "Caldocellum saccharolyticum" is part of a multidomain enzyme

The complete sequence of a beta-mannanase gene from an anaerobic extreme thermophile was determined, and it shows that the expressed protein consists of two catalytic domains and two binding domains separated by spacer regions rich in proline and threonine residues. The amino-terminal catalytic domain has beta-mannanase activity, and the carboxy-terminal domain acts as an endoglucanase. Neither domain shows homology with any other cellulase or hemicellulase sequence at the nucleic acid or protein level.

celA from Bacillus lautus PL236 encodes a novel cellulose-binding endo-beta-1,4-glucanase

celA from the cellulolytic bacterium Bacillus lautus PL236 encodes EG-A, an endo-beta-1,4-glucanase. An open reading frame of 2,100 bp preceded by a ribosome-binding site encodes a protein with a molecular mass of 76,863 Da with a typical signal sequence. The NH2-terminal active domain of EG-A is not homologous to any reported cellulase or xylanase and may represent a new family of such enzymes. A 150-amino-acid COOH-terminal peptide is homologous to noncatalytic domains in several other cellulases (A. Meinke, N.R. Gilkes, D.G. Kilburn, R.C. Miller, Jr., and R.A.J. Warren, J. Bacteriol. 173:7126-7135, 1991). Upstream of celA, a partial open reading frame encodes a 145-amino-acid peptide which also belongs to the family mentioned. Zymogram analysis of extracts from Escherichia coli and supernatants of Bacillus subtilis and B. megaterium, including protease-deficient mutants thereof, which express celA, revealed two active proteins, EG-A-L and EG-A-S, with Mrs of 74,000 and 57,000, respectively. The proportion of EG-A-L to EG-A-S depends on the extracellular proteolytic activity of the host organism, indicating that EG-A-S arises from posttranslational proteolytic modification of EG-A-L. Since EG-A-S has an NH2 terminus corresponding to the predicted NH2-terminal sequence of EG-A, processing appears to take place between the catalytic and noncatalytic domains described. EG-A-L and EG-A-S were purified to homogeneity and shown to have almost identical characteristics with respect to activity against soluble substrates and pH and temperature dependency. EG-A-L binds strongly to cellulose, in contrast to EG-A-S, and has higher activity against insoluble substrates than the latter. We conclude that the COOH-terminal 17,000-Mr peptide of EG-A-L constitutes a cellulose-binding domain.

The binding of Cellulomonas fimi endoglucanase C (CenC) to cellulose and Sephadex is mediated by the N-terminal repeats

Endoglucanase C (CenC) from Cellulomonas fimi binds to cellulose and to Sephadex. The enzyme has two contiguous 150-amino-acid repeats (N1 and N2) at its N-terminus and two unrelated contiguous 100-amino-acid repeats (C1 and C2) at its C-terminus. Polypeptides corresponding to N1, N1N2, C1, and C1C2 were produced by expression of appropriate cenC gene fragments in Escherichia coli. N1N2, but not N1 alone, binds to Sephadex; both polypeptides bind to Avicel, (a heterogeneous cellulose preparation containing both crystalline and non-crystalline components). Neither C1 nor C1C2 binds to Avicel or Sephadex. N1N2 and N1 bind to regenerated ('amorphous') cellulose but not to bacterial crystalline cellulose; the cellulose-binding domain of C. fimi exoglucanase Cex binds to both of these forms of cellulose. Amino acid sequence comparison reveals that N1 and N2 are distantly related to the cellulose-binding domains of Cex and C. fimi endoglucanases A and B.

Primary sequence analysis of Clostridium cellulovorans cellulose binding protein A

The cbpA gene for the Clostridium cellulovorans cellulose binding protein (CbpA), which is part of the multisubunit cellulase complex, has been cloned and sequenced. When cbpA was expressed in Escherichia coli, proteins capable of binding to crystalline cellulose and of interacting with anti-CbpA were observed. The cbpA gene consists of 5544 base pairs and encodes a protein containing 1848 amino acids with a molecular mass of 189,036 Da. The open reading frame is preceded by a Gram-positive-type ribosome binding site. A signal peptide sequence of 28 amino acids is present at its N terminus. The encoded protein is highly hydrophobic with extremely high levels of threonine and valine residues. There are two types of putative cellulose binding domains of approximately 100 amino acids that are slightly hydrophilic and eight conserved, highly hydrophobic beta-sheet regions of approximately 140 amino acids. These latter hydrophobic regions may be the CbpA domains that interact with the different enzymatic subunits of the cellulase complex.

Nucleotide sequence of the cellulase gene celF of Clostridium thermocellum

The nucleotide sequence of the celF gene of Clostridium thermocellum was determined. The open reading frame extended over 2217 bp. The encoded 739-aa polypeptide, CelF, with a Mw = 82,015, was an endoglucanase with activity against carboxymethylcellulose. The N terminus showed a typical signal peptide, and a cleavage site after Ala-27 was predicted. From residues 28 to 470, the sequence of CelF was related to the catalytic domains of type E2 endoglucanases, with a strong homology to the endoglucanases CelZ of Clostridium stercorarium and CenB of Cellulomonas fimi. The catalytic region was followed by a 134-aa segment also present in C. stercorarium CelZ and in C. fimi CenB, and belonging to the family of non-catalytic, presumably cellulose-binding domains first identified in Bacillus subtilis endoglucanase. A 21-aa segment rich in Pro/Thr/Ser residues separated the putative cellulose-binding region from the COOH-terminal region, which contained two conserved stretches of 24 amino acids closely similar to those previously described in endoglucanases CelA, CelB, CelD, CelE, CelH and CelX, and xylanase XynZ of C. thermocellum.

Multiple domains in endoglucanase B (CenB) from Cellulomonas fimi: functions and relatedness to domains in other polypeptides

Endoglucanase B (CenB) from the bacterium Cellulomonas fimi is divided into five discrete domains by linker sequences rich in proline and hydroxyamino acids (A. Meinke, C. Braun, N. R. Gilkes, D. G. Kilburn, R. C. Miller, Jr., and R. A. J. Warren, J. Bacteriol. 173:308-314, 1991). The catalytic domain of 608 amino acids is at the N terminus. The sequence of the first 477 amino acids in the catalytic domain is related to the sequences of cellulases in family E, which includes procaryotic and eucaryotic enzymes. The sequence of the last 131 amino acids of the catalytic domain is related to sequences present in a number of cellulases from different families. The catalytic domain alone can bind to cellulose, and this binding is mediated at least in part by the C-terminal 131 amino acids. Deletion of these 131 amino acids reduces but does not eliminate activity. The catalytic domain is followed by three domains which are repeats of a 98-amino-acid sequence. The repeats are approximately 50% identical to two repeats of 95 amino acids in a chitinase from Bacillus circulans which are related to fibronectin type III repeats (T. Watanabe, K. Suzuki, K. Oyanagi, K. Ohnishi, and H. Tanaka, J. Biol. Chem. 265:15659-15665, 1990). The C-terminal domain of 101 amino acids is related to sequences, present in a number of bacterial cellulases and xylanases from different families, which form cellulose-binding domains (CBDs). It functions as a CBD when fused to a heterologous polypeptide. Cells of Escherichia coli expressing the wild-type cenB gene accumulate both native CenB and a stable proteolytic fragment of 41 kDa comprising the three repeats and the C-terminal CBD. The 41-kDa polypeptide binds to cellulose but lacks enzymatic activity.

Domains in microbial beta-1, 4-glycanases: sequence conservation, function, and enzyme families

Several types of domain occur in beta-1, 4-glycanases. The best characterized of these are the catalytic domains and the cellulose-binding domains. The domains may be joined by linker sequences rich in proline or hydroxyamino acids or both. Some of the enzymes contain repeated sequences up to 150 amino acids in length. The enzymes can be grouped into families on the basis of sequence similarities between the catalytic domains. There are sequence similarities between the cellulose-binding domains, of which two types have been identified, and also between some domains of unknown function. The beta-1, 4-glycanases appear to have arisen by the shuffling of a relatively small number of progenitor sequences.

Nucleotide sequence of the endoglucanase C gene (cenC) of Cellulomonas fimi, its high-level expression in Escherichia coli, and characterization of its products

The cenC gene of Cellulomonas fimi, encoding endoglucanase CenC, has an open reading frame of 1101 codons closely followed by a 9 bp inverted repeat. The predicted amino acid sequence of mature CenC, which is 1069 amino acids long, is very unusual in that it has a 150-amino-acid tandem repeat at the N-terminus and an unrelated 100-amino-acid tandem repeat at the C-terminus. CenC belongs to subfamily E1 of the beta-1,4-glycanases. High-level expression in Escherichia coli of cenC from a 3.6 kbp fragment of C. fimi DNA leads to levels of CenC which exceed 10% of total cell protein. Most of the CenC is in the cytoplasm in an inactive form. About 60% of the active fraction of CenC is in the periplasm. The catalytic properties of the active CenC are indistinguishable from those of native CenC from C. fimi. The Mr of CenC from E. coli and C. fimi is approximately 130 kDa. E. coli and C. fimi also produce an endoglucanase, CenC', of approximate Mr 120kDa and with the same N-terminal amino acid sequence and catalytic properties as CenC. CenC' appears to be a proteolytic product of CenC. CenC and CenC' can bind to cellulose and to Sephadex. CenC is the most active component of the C. fimi cellulase system isolated to date.

Purification and properties of an acid endo-1,4-beta-glucanase from Bacillus sp. KSM-330

A novel acid cellulase (endo-1,4-beta-glucanase, EC 3.2.1.4) was found in a culture of Bacillus sp. KSM-330 isolated from soil. One-step chromatography on a column of CM-Bio-Gel A yielded a homogeneous enzyme, as determined by silver staining of both sodium dodecyl sulphate (SDS) and nondenaturing gels. The enzyme had a molecular mass of 42 kDa, as determined by SDS-polyacrylamide gel electrophoresis. The isoelectric point was higher than pH 10. The N-terminal amino acid sequence of the enzyme was Val-Ala-Lys-Glu-Met-Lys-Pro-Phe-Pro-Gln-Gln-Val-Asn-Tyr-Ser-Gly-Ile-Leu- Lys-Pro . This enzyme had an optimum pH for activity of 5.2, being active over an extremely narrow range of pH values, from 4.2 to 6.9; below and above these pH values no activity was detectable. The optimum temperature at pH 5.2 was around 45 degrees C. The enzyme efficiently hydrolysed carboxymethylcellulose (CMC) and lichenan, but more crystalline forms of cellulose, curdlan, laminarin, 4-nitrophenyl-beta-D-glucopyranoside and 4-nitrophenyl-beta-D-cellobioside were barely hydrolysed. The enzymic activity was inhibited by Hg2+ but was not affected by other inhibitors of thiol enzymes, such as 4-chloromercuribenzoate. N-ethylmaleimide and monoiodoacetate. N-Bromosuccinimide abolished the enzymic activity, and CMC protected the enzyme from inactivation by this tryptophan-specific oxidant. It is suggested that a tryptophan residue(s) is involved in the mechanism of action of the Bacillus cellulase and that the inhibition of enzymic activity by Hg2+ is ascribable to interactions with the tryptophan residue(s) rather than with thiol group(s).

Recent advances in rumen microbial ecology and metabolism: potential impact on nutrient output

Feedstuffs consumed by ruminants are all initially exposed to fermentative activity in the rumen prior to gastric and intestinal digestion. The extent and type of transformation of feedstuffs thus determines the productive performance of the host. Research on rumen microbial ecology and metabolism is essentially a study of the interactions between the host, microorganisms present, substrates available, and end products of digestion. Furthermore, the interactions of the normal microbial flora with the host can be manipulated to improve the efficiency of nutrient utilization in ruminant animals. Three important areas of ruminal fermentation will be reviewed, N metabolism, fiber degradation, and biotransformation of toxic compounds. The extent of protein degradation and the rate of uptake of resultant peptides and ammonia are extremely important factors in determining the efficiency of N utilization by rumen bacteria and, therefore, the relative amounts of microbial or bypass protein available to the host. Strategies aimed at identifying and characterizing rate-limiting enzymes of cellulolytic bacteria are essential in elucidating mechanisms involved in ruminal fiber degradation. Results obtained with ruminococci will be described. The detoxification of phytotoxins by passage through the gastrointestinal tract of ruminants is a process deserving special attention and several examples will be presented. Opportunities for manipulation of rumen fermentation are good. However, successful manipulation and full exploitation depend on a through understanding of the mechanisms involved.

celB, a gene coding for a bifunctional cellulase from the extreme thermophile "Caldocellum saccharolyticum"

"Caldocellum saccharolyticum" is an obligatory anaerobic thermophilic bacterium. A gene from this organism, designated celB, has been cloned in Escherichia coli as part of a bacteriophage lambda gene library. This gene produces a thermostable cellulase that shows both endoglucanase and exoglucanase activities on test substrates and is able to degrade crystalline cellulose to glucose. The sequence of celB has homology with both exo- and endoglucanases described by others. It appears to have a central domain without enzymatic activity which is joined to the enzymatic domains by runs of amino acids rich in proline and threonine (PT boxes). Deletion analysis shows that the exoglucanase activity is located in the amino-terminal domain of the enzyme and that endoglucanase activity is located in the carboxy-terminal domain. There are internal transcriptional and translational start sites within the gene. The intact gene has been cloned into a temperature-inducible expression vector, pJLA602, and overexpressed in E. coli. Polyacrylamide gel electrophoresis showed that celB produced a protein with a molecular weight of 118,000 to 120,000. A number of smaller proteins with activity against carboxymethyl cellulose and 4-methyl umbelliferyl-beta-D-cellobioside were also produced. These are believed to be the result of alternative translational start sites and/or proteolytic degradation products of the translated gene product.

Sequence analysis of the Clostridium stercorarium celZ gene encoding a thermoactive cellulase (Avicelase I): identification of catalytic and cellulose-binding domains

The nucleotide sequence of the celZ gene coding for a thermostable endo-beta-1,4-glucanase (Avicelase I) of Clostridium stercorarium was determined. The structural gene consists of an open reading frame of 2958 bp which encodes a preprotein of 986 amino acids with an Mr of 109,000. The signal peptide cleavage site was identified by comparison with the N-terminal amino acid sequence of Avicelase I purified from C. stercorarium culture supernatants. The recombinant protein expressed in Escherichia coli is proteolytically cleaved into catalytic and cellulose-binding fragments of about 50 kDa each. Sequence comparison revealed that the N-terminal half of Avicelase I is closely related to avocado (Persea americana) cellulase. Homology is also observed with Clostridium thermocellum endoglucanase D and Pseudomonas fluorescens cellulase. The cellulose-binding region was located in the C-terminal half of Avicelase I. It consists of a reiterated domain of 88 amino acids flanked by a repeated sequence about 140 amino acids in length. The C-terminal flanking sequence is highly homologous to the non-catalytic domain of Bacillus subtilis endoglucanase and Caldocellum saccharolyticum endoglucanase B. It is proposed that the enhanced cellulolytic activity of Avicelase I is due to the presence of multiple cellulose-binding sites.

Multiple endo-beta-1,4-glucanase-encoding genes from Bacillus lautus PL236 and characterization of the celB gene

A Bacillus lautus strain was isolated from compost by its ability to degrade microcrystalline Avicel cellulose and acid-swollen cellulose. Three DNA fragments cloned in Escherichia coli encoded at least four endo-beta-1,4-glucanases (EG), of which at least two were contained on one DNA fragment. Another fragment, of 2.5 kb and carrying celB, was cloned in the shuttle-vector plasmid, pJKK3-1, and expressed in E. coli and Bacillus subtilis. The fragment was sequenced and shown to encode a 62-kDa protein, which was found as a 56-kDa mature and active EG in extracts of E. coli and in the supernatant of B. subtilis. The deduced amino acid (aa) sequence has a homology of 37% identical aa on a stretch of 295 aa to EG-E of Clostridium thermocellum. A low level of homology is detected with the Bacillus-type EG.

Cloning and sequencing of a Bacteroides ruminicola B(1)4 endoglucanase gene

Bacteroides ruminicola B(1)4, a noncellulolytic rumen bacterium, produces an endoglucanase (carboxymethylcellulase [CMCase]) that is excreted into the culture supernatant. Cultures grown on glucose, fructose, maltose, mannose, and cellobiose had high specific activities of CMCase (greater than 3 mmol of reducing sugar per mg of protein per min), but its synthesis was repressed by sucrose. B. rumincola did not grow on either ball-milled or acid-swollen cellulose even though the CMCase could hydrolyze swollen cellulose. The CMCase gene was cloned into Escherichia coli, and its nucleotide sequence contained a single open reading frame coding for a protein of 40,481 daltons. The enzyme was overproduced in E. coli under the control of the tac promoter and purified to homogeneity. The N-terminal sequence, amino acid composition, and molecular weight of the purified enzyme were similar to the values predicted from the open reading frame of the DNA sequence. However, the CMCase present in B. ruminicola was found to have a monomer molecular weight of 88,000 by Western immunoblotting. This discrepancy appeared to have resulted from our having cloned only part of the CMCase gene into E. coli. The amino acid sequence of the CMCase showed homology to sequences of beta-glucanases from Ruminococcus albus and Clostridium thermocellum.

Structure of the beta-1,3-1,4-glucanase gene of Bacillus macerans: homologies to other beta-glucanases

The nucleotide sequence of an 852 base pair (bp) DNA fragment containing the entire gene coding for thermostable beta-1,3-1,4-glucanase of Bacillus macerans has been determined. The bglM gene comprises an open reading frame (ORF) of 711 bp (237 codons) starting with ATG at position 93 and extending to the translational stop codon TAA at position 804. The deduced amino acid sequence of the mature protein shows 70% homology to published sequences of mesophilic beta-1,3-1,4-glucanases from B. subtilis and B. amyloliquefaciens. The sequence coding for mature beta-glucanase is preceded by a putative signal peptide of 25 amino acid residues, and a sequence resembling a ribosome-binding site (GGAGG) before the initiation codon. By contrast with the processed protein, the N-terminal amino acid sequence constituting the putative leader peptide bears no or only weak homology to signal peptides of mesophilic Bacillus endo-beta-glucanases. The B. macerans signal peptide appears to be functional in exporting the enzyme to the periplasm in E. coli. More than 50% of the whole glucanase activity was localized in the periplasmic space and in the supernatant. Whereas homology to endo-1,4-beta-glucanases is completely lacking, a weak amino acid homology between the sequence surrounding the active site of phage T4 lysozyme and a sequence spanning residues 126 through 161 of B. macerans endo-beta-glucanase could be identified.

The Glu residue in the conserved Asn-Glu-Pro sequence of two highly divergent endo-beta-1,4-glucanases is essential for enzymatic activity

We initially aligned 28 different cellulase sequences in pairwise fashion and found half of them have the sequence -Asn-Glu-Pro- located in a region flanked by hydrophobic-rich amino acids. Based on lysozyme as a model, the glutamate residue could be essential for enzyme function. We tested this possibility by site-directed mutagenesis of the genes coding Bacillus polymyxa and Bacillus subtilis endo-beta-1,4-glucanases. The genes and amino acid sequences of these two enzymes show very little similarity. Change of Glu-194 and Glu-169 to the isosteric glutamine form in these respective enzymes resulted in a dramatic loss of CMCase activity which could be restored by reverse mutation. Similar mutations to less-conserved residues, Glu-72 and Glu-147, of the B. subtilis enzyme did not cause any loss of activity.

Cloning, sequence analysis, and expression of genes encoding xylan-degrading enzymes from the thermophile "Caldocellum saccharolyticum"

A lambda recombinant bacteriophage coding for xylanase and beta-xylosidase activity has been isolated from a genomic library of the extremely thermophilic anaerobe "Caldocellum saccharolyticum." Partial Sau3AI fragments of the lambda recombinant DNA were ligated into pBR322. A recombinant plasmid with an insertion of ca. 7 kilobases of thermophilic DNA expressing both enzymatic activities was isolated. The location of the genes has been established by analyzing deletion derivatives, and the DNA sequence of 6.067 kilobases of the insert has been determined. Five open reading frames (ORFs) were found, one of which (ORF1; Mr 40,455) appears to code for a xylanase (XynA) which also acts on o-nitrophenyl-beta-D-xylopyranoside. Another, ORF5 (Mr 56,365), codes for a beta-xylosidase (XynB). The xynA gene product shows significant homology with the xylanases from the alkalophilic Bacillus sp. strain C125 and Clostridium thermocellum.

Molecular cloning, expression, and characterization of endo-beta-1,4-glucanase genes from Bacillus polymyxa and Bacillus circulans

Endo-beta-1,4-glucanase genes from Bacillus circulans and from B. polymyxa were cloned by direct expression by using bacteriophage M13mp9 as the vector. The enzymatic activity of the gene products was detected by using either the Congo red assay or hydroxyethyl cellulose dyed with Ostazin Brilliant Red H-3B. The B. circulans and B. subtilis PAP115 endo-beta-1,4-glucanase genes were shown to be homologous by the use of restriction endonuclease site mapping, DNA-DNA hybridization, S1 nuclease digestion after heteroduplex formation, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the protein products. Analysis of the nucleotide sequence of 3.1 kilobase pairs of cloned B. polymyxa DNA revealed two convergently transcribed open reading frames (ORFs) consisting of 398 codons (endoglucanase) and 187 codons (ORF2) and separated by 374 nucleotides. The coding region of the B. polymyxa endoglucanase gene would theoretically produce a 44-kilodalton preprotein. Expression of the B. polymyxa endoglucanase in Escherichia coli was due to a fusion of the endoglucanase gene at codon 30 with codon 9 of the lacZ alpha-peptide gene. The B. polymyxa endoglucanase has 34% amino acid similarity to the Clostridium thermocellum celB endoglucanase sequence but very little similarity to endoglucanases from other Bacillus species. ORF2 has 28% amino acid similarity to the NH2-terminal half of the E. coli lac repressor protein, which is responsible for DNA binding.

Sequence analysis of the Clostridium cellulolyticum endoglucanase-A-encoding gene, celCCA

The nucleotide sequence of a Clostridium cellulolyticum endo-beta-1,4- glucanase (EGCCA)-encoding gene (celCCA) and its flanking regions, was determined. An open reading frame (ORF) of 1425 bp was found, encoding a protein of 475 amino acids (aa). This ORF began with an ATG start codon and ended with a TAA ochre stop codon. The N-terminal region of the EGCCA protein resembled a typical signal sequence of a Gram-positive bacterial extracellular protein. A putative signal peptidase cleavage site was determined. EGCCA, without a signal peptide, was found to be composed of more than 35% hydrophobic aa and to have an Mr of 50715. Comparison of the encoded sequence with other known cellulase sequences showed the existence of various kinds of aa sequence homologies. First, a strong homology was found between the C-terminal region of EGCCA, containing a reiterated stretch of 24 aa, and the conserved reiterated region previously found to exist in four Clostridium thermocellum endoglucanases and one xylanase from the same organism. This region was suspected of playing a role in organizing the cellulosome complex. Second, an extensive homology was found between EGCCA and the N-terminal region of the large endoglucanase, EGE, from C. thermocellum, which suggests that they may have a common ancestral gene. Third, a region, which extended for 21 aa residues beginning at aa + 127, was found to be homologous with regions of cellulases belonging to Bacilli, Clostridia and Erwinia chrysanthemi.

Structure of a Ruminococcus albus endo-1,4-beta-glucanase gene

A chromosomal DNA fragment encoding an endo-1,4-beta-glucanase I (Eg I) gene from Ruminococcus albus cloned and expressed in Escherichia coli with pUC18 was fully sequenced by the dideoxy-chain termination method. The sequence contained a consensus promoter sequence and a structural amino acid sequence. The initial 43 amino acids of the protein were deduced to be a signal sequence, since they are missing in the mature protein (Eg I). High homology was found when the amino acid sequence of the Eg I was compared with that of endoglucanase E from Clostridium thermocellum. Codon usage of the gene was not biased. These results suggested that the properties of the Eg I gene from R. albus was specified from the known beta-glucanase genes of the other organisms.

Cloning and sequencing of an endoglucanase (end1) gene from Butyrivibrio fibrisolvens H17c

The nucleotide sequence of a 2.8 kb DNA segment containing an endoglucanase gene (end1) from Butyrivibrio fibrisolvens H17c was determined. The B. fibrisolvens H17c gene was expressed from its own regulatory region in Escherichia coli and three putative consensus promoter sequences were identified upstream of a ribosome binding site and an ATG start codon. The complete amino acid sequence (547 residues) was deduced and homology with the Clostridium thermocellum celE gene product (EGE) was demonstrated. The endoglucanase contained a typical amino-terminal signal sequence and five repeated sequences (PDPTPVD) between amino acids 412-447. The endoglucanase showed relatively high endoglucanase activity against endoglucanase-specific substrates with beta 1-4 linkages but low activity against xylan and an exoglucanase-specific substrate, p-nitrophenyl-beta-D-cellobioside.

Purification and characterization of endoglucanase C of Cellulomonas fimi, cloning of the gene, and analysis of in vivo transcripts of the gene

Two nonglycosylated endoglucanases which bind to Sephadex were purified from culture supernatants of Cellulomonas fimi grown on microcrystalline cellulose. Their Mrs were 120,000 and 130,000. The N-terminal amino acid sequences of the enzymes were identical, suggesting that the enzymes were related. A DNA fragment encoding this N-terminal sequence was cloned in Escherichia coli. The nucleotide sequence corresponding to the N-terminal amino acid sequence was preceded by a sequence encoding a typical leader peptide. Transcripts hybridizing to the cloned fragment were detected in total RNA isolated from C. fimi cells grown on carboxymethyl cellulose but not from cells grown on glycerol or glucose. Transcription started at a cluster of sites 53 to 59 nucleotides upstream of a GUG translation initiation codon and terminated at either of two closely spaced C residues immediately downstream of a region of potential secondary structure. The size of the transcript was approximately 3.5 kilobases, sufficient to encode a polypeptide of 130 kilodaltons. The 130-kilodalton polypeptide is designated endoglucanase C (CenC), and the gene encoding it is designated cenC.

Nucleotide sequence and analysis of the lethal factor gene (lef) from Bacillus anthracis

The nucleotide sequence of the Bacillus anthracis lethal factor (LF) gene (lef) has been determined. LF is part of the tripartite protein exotoxin of B. anthracis along with protective antigen (PA) and edema factor (EF). The apparent ATG start codon, which is located immediately upstream from codons which specify the first 16 amino acids (aa) of the mature secreted LF, is preceded by an AAAGGAG sequence, which is its probable ribosome-binding site. This ATG codon begins a continuous 2427-bp open reading frame which encodes the 809-aa LF-precursor protein with an Mr of 93,798. The mature secreted protein (776 aa; Mr 90,237) was preceded by a 33-aa signal peptide which has characteristics in common with leader peptides for other secreted proteins of the Bacillus species. The codon usage of the LF gene reflects its high (70%) A + T content. The N-terminus of LF (first 300 aa) shared extensive homology with the N-terminus of the anthrax EF protein. Since LF and EF each bind PA at the same site, these homologous regions probably represent their common PA-binding domains.

Aspergillus oryzae has two nearly identical Taka-amylase genes, each containing eight introns

cDNA and genomic DNA for two nearly identical genes, AmyI and AmyII, coding for the enzyme Taka-amylase A (TA-A) of the fungus Aspergillus oryzae have been cloned and characterized. These genes are apparently unlinked, differing by only 3 nucleotides (nt) out of the 2720 nt that span the coding regions. The 617-nt 5'-flanking regions differ only at nt -372 (T or A) from the putative ATG start codon and contain four sets of short, inverted repeats (IR) upstream from the putative TATAAA box at nt -100 and the transcription start point at nt -69. The coding regions consist of 499 codons disrupted by eight intervening sequences. The putative proenzymes differ by only two amino acids (aa) and consist of the 478-aa extracellular enzyme plus a 21-aa hydrophobic leader sequence. Except for the replacement site changes in codons 35 (Arg----Gln) and 151 (Phe----Leu), the identity of the two genes continues downstream for 58 nt past the TGA stop codon before diverging. Exon 9 codes for 94 of the 98 aa of the domain B of mature TA-A. Little conservation of TA-A exons was found when these exons were aligned with those of human amylase. The genes are flanked by at least 6 to 10 kb of unrelated chromosomal nucleotide sequence. The Amy genes are co-expressed, since mRNA (cDNA) specific to the 3'-UTR of both genes was recovered from mycelia grown on starch, a known inducer of TA-A biosynthesis. The 3'-UTRs of cDNAs related to AmyI are shorter (128 nt and 145 nt) than those of AmyII (179 nt and 297 nt). The AmyI specific 3'-UTR is characterized by the absence of IR sequences and the presence of a putative 'AATAAA' polyadenylation signal.

Hyperexpression of a Bacillus circulans xylanase gene in Escherichia coli and characterization of the gene product

A 4.0-kilobase (kb) fragment of Bacillus circulans genomic DNA inserted into pUC19 and encoding endoxylanase activity was subjected to a series of subclonings. A 1.0-kb HindIII-HincII subfragment was found to code for xylanase activity. Maximum expression levels were observed with a subclone that contained an additional 0.3-kb sequence upstream from the coding region. Enhancer sequences in the upstream region are thought to be responsible for these high expression levels. Southern hybridization analyses revealed that the cloned gene hybridized with genomic DNA from Bacillus subtilis and Bacillus polymyxa. Xylanase activity expressed by Escherichia coli harboring the cloned gene was located primarily in the intracellular fraction. Levels of up to 7 U/ml or 35 mg/liter were obtained. The protein product was purified by ion exchange and gel permeation chromatography. The xylanase had a molecular weight of 20,500 and an isoelectric point of 9.0.

The third cellulase of alkalophilic Bacillus sp. strain N-4: evolutionary relationships within the cel gene family

The third cellulase gene (celC) of Bacillus sp. strain N-4 was cloned in plasmid pBR322 and was located within a 5.5-kb HindIII fragment. The cellulase encoded by this fragment had an Mr of about 100,000 and showed optimum activity around pH 9. These properties were different from those of the enzymes encoded by the celA and celB genes of the same organism. The amino acid sequence deduced from the nucleotide sequence was found to be highly homologous to the CEL-F enzyme from Bacillus sp. strain No. 1139 [Fukumori et al., J. Gen. Microbiol. 132 (1986) 2329-2335]. An evolutionary relationship observed among the four cellulases of alkalophilic Bacillus strains and that of Bacillus subtilis endoglucanase suggested that ancestral genes for alkaline and neutral cellulases diverged early in the evolution of these enzymes.

Species-specific variation in signal peptide design. Implications for protein secretion in foreign hosts

Secretory signal peptides from individual prokaryotic and eukaryotic species have been analyzed, and the lengths and amino acid compositions of the positively charged amino-terminal region, the central hydrophobic region, and the carboxy-terminal cleavage-region have been compared. We find distinct differences between species in all three regions. Implications for protein secretion in foreign hosts are discussed.