OUP user menu

Family 6 carbohydrate-binding modules display multiple β1,3-linked glucan-specific binding interfaces

Márcia A.S. Correia , Virgínia M.R. Pires , Harry J. Gilbert , David N. Bolam , Vânia O. Fernandes , Victor D. Alves , José A.M. Prates , Luís M.A. Ferreira , Carlos M.G.A. Fontes
DOI: http://dx.doi.org/10.1111/j.1574-6968.2009.01764.x 48-57 First published online: 1 November 2009

Abstract

Noncatalytic carbohydrate-binding modules (CBMs), which are found in a variety of carbohydrate-degrading enzymes, have been grouped into sequence-based families. CBMs, by recruiting their appended enzymes onto the surface of the target substrate, potentiate catalysis particularly against insoluble substrates. Family 6 CBMs (CBM6s) display unusual properties in that they present two potential ligand-binding sites termed clefts A and B, respectively. Cleft B is located on the concave surface of the β-sandwich fold while cleft A, the more common binding site, is formed by the loops that connect the inner and the outer β-sheets. Here, we report the biochemical properties of CBM6-1 from Cellvibrio mixtus CmCel5A. The data reveal that CBM6-1 specifically recognizes β1,3-glucans through residues located both in cleft A and in cleft B. In contrast, a previous report showed that a CBM6 derived from a Bacillus halodurans laminarinase binds to β1,3-glucans only in cleft A. These studies reveal a different mechanism by which a highly conserved protein platform can recognize β1,3-glucans.

Keywords
  • protein : carbohydrate interactions
  • carbohydrate-binding module (CBM)
  • glycoside hydrolase (GH)
  • Cellvibrio

Introduction

Enzyme–substrate targeting is critical to the efficient hydrolysis of plant cell wall polysaccharides, a key process for carbon recycling within the biosphere. Plant cell walls are intricate macromolecules composed of a variety of polysaccharides that establish complex interactions, thus limiting substrate accessibility to enzymatic attack (Brett & Waldren, 1996). Plant cell wall hydrolases, therefore, play a crucial role in numerous biological processes, while also being of considerable industrial importance, particularly in the exploitation of lignocellulose as an environmentally sustainable substrate for biofuel production (Boudet et al., 2003; Schell et al., 2004; Ragauskas et al., 2006). A generic feature of enzymes that hydrolyse complex carbohydrates is their modular structure; plant cell wall hydrolases contain catalytic and noncatalytic modules linked by flexible linker sequences. The most common noncatalytic modules found in these enzymes are carbohydrate-binding modules (CBMs), which are grouped into sequence-based families (Coutinho & Henrissat, 1999). The general function of CBMs is to promote and maintain the interaction of the enzyme with the target substrate, thereby increasing the efficiency of catalysis (Hall et al., 1995; Bolam et al., 1998; Fernandes et al., 1999; Boraston et al., 2003a). Based on the topology of the carbohydrate-binding site, CBMs have been classified into three types (Boraston et al., 2004). The binding site of type A CBMs, which recognize crystalline polysaccharides, consists of three planar aromatic residues, a topology that is consistent with the planar conformation of the ligand. In contrast, type B and type C CBMs recognize single carbohydrate chains either internally or at the termini, respectively, and present a ligand specificity that reflects the substrate specificity of the appended catalytic domain (Boraston et al., 2004). Structural studies revealed that type B and C CBMs accommodate their target ligands in clefts and pockets, respectively (Boraston et al., 2004).

Family 6 CBMs (CBM6s) consist of a large family that display a variety of ligand specificities and adopt a canonical β-jelly roll fold. The surface of CBM6 proteins contain two potential ligand-binding sites termed cleft A and cleft B (Czjzek et al., 2001; Boraston et al., 2003b; Henshaw et al., 2004; Pires et al., 2004). One of these sites, cleft B, is located on the concave surface, while the second site, cleft A, is on the edge of the protein and is formed by the loops that connect the inner and the outer β-sheets. Although cleft B displays a topology that is typical of type B binding sites, cleft A generally, but not exclusively, resembles a type C carbohydrate-binding region, which recognizes single sugars or the termini of polymeric carbohydrates (Van Bueren et al., 2005). The variation in ligand specificity in CBM6 reflects, in part, the functionality of these two binding sites. Thus, CBM6s that bind xylan (Czjzek et al., 2001; Boraston et al., 2003b), laminarin (Van Bueren et al., 2005) and the nonreducing glucose of β-glucans (Henshaw et al., 2004; Pires et al., 2004) have been described, with ligand binding occurring in cleft A. Although cleft A forms a deep cleft in CBM6s that bind to the internal regions of xylan, in the majority of proteins in this family, cleft A forms a pocket. This pocket-like topology is exemplified by a CBM6 from a Bacillus halodurans laminarinase; the protein module binds to the terminal sugar of the helical glucan in cleft A, but the ligand extends across the surface of the protein (Van Bueren et al., 2005). In addition, the functional cleft B binding site of CBM6-2 from CmCel5B (Henshaw et al., 2004; Pires et al., 2004), which presents a concave surface, recognizes the internal regions of β1,4- or mixed linked β1,3-β1,4-glucans, while cleft A binds to the nonreducing end of β-linked gluco- and xylo-configured polymers. Functional cleft B carbohydrate-binding sites were also identified in CBM6s appended to β-agarases (Henshaw et al., 2006) and α-agarases (Flament et al., 2007; Michel et al., 2009).

A previously described cellulase from Cellvibrio mixtus, CmCel5B, presents a modular architecture containing an N-terminal family 5 glycoside hydrolase (GH) catalytic module, followed by an internal CBM6, termed CBM6-1, and a C-terminal CBM6, termed CBM6-2 (Fontes et al., 1998). Cellvibrio mixtus is a saprophytic mesophilic bacterium that colonizes the soil and efficiently degrades a variety of plant cell wall polysaccharides. The structure of CBM6-2 revealed the presence of two functional ligand-binding sites, cleft A and cleft B, respectively (Henshaw et al., 2004; Pires et al., 2004). Here, we present the biochemical properties of C. mixtus CBM6-1. The data revealed that CBM6-1 binds exclusively to β1,3-glucans through residues located both at clefts A and B. Therefore, β1,3-glucans can interact with CBM6 modules at different binding sites, highlighting the plasticity in carbohydrate recognition displayed by this protein fold.

Materials and methods

Protein expression and purification

Genes encoding the truncated derivatives of CmCel5B, CBM6-1, CBM6-2 and CBM6-1/2 were amplified by PCR from C. mixtus genomic DNA using the primers listed in Table 1 using the thermostable DNA polymerase NZYPremium (NZYTech Ltd). Forward and reverse primers incorporated, respectively, EcoRI and XhoI restriction sites, which were used for the subsequent cloning into pET32a (Table 1). Amplified DNA was directly cloned into pNZY28 (NZYTech Ltd) and sequenced to ensure that no mutations were accumulated during the amplification. The genes were subsequently subcloned into EcoRI/XhoI-restricted pET32a (Novagen), generating pCBM6-1, pCBM6-2, and pCBM6-1/2, respectively (Sambrook et al., 1989). All recombinant derivatives contained an internal His6-tag and consisted of recombinant thioredoxin fusion proteins. To engineer a gene encoding two copies of CBM6-2 organized in tandem, DNA encoding CBM6-2 and the linker sequence separating CBM6-1 and CBM6-2, was amplified from C. mixtus genomic DNA with the primers listed in Table 1. The forward primer contained an engineered SalI restriction site. The resulting PCR fragment was subcloned from pNZY28 (NZYTech Ltd), on a SalI–XhoI restriction fragment, into the XhoI site of pCBM6-2. The stop codon that separated the two genes was subsequently removed using the mutagenesis primers listed in Table 1, allowing the production of a thioredoxin fusion protein containing two CBM6-2 copies organized in tandem. Genes encoding CmCel5B catalytic domain, GH5, and the full-length enzyme, GH5-CBM6-1/2, were amplified from C. mixtus genomic DNA as described above and using the primers listed in Table 1. The resulting PCR products were cloned into pNZY28 (NZYTech Ltd) and sequenced to ensure that no mutations had occurred during the amplification. The recombinant plasmids obtained were digested with NheI and XhoI and the excised DNA fragment was cloned into the similarly restricted expression vector pET21a, to generate pGH5 and pGH5-CBM6-1/2. The two recombinant proteins contained a C-terminal His6-tag. The molecular architectures of all the proteins generated in this study are presented in Fig. 1.

View this table:
1

Primers used to obtain the genes encoding CmCel5B derivatives used in this work and for the mutagenesis of CBM6-1

CloneSequence (5′→3′)Direction
CBM6-1CTC GAA TTC ACC TGC ACC AAA GCC AATFOR
CAC CTC GAG TTA TTC TAC TTT CAG CCA GTTREV
CBM6-2CTC GAA TTC GTA ATC GCG ACT ATT CAGFOR
CAC CTC GAG TTA ATG TGT CTT GTT GREV
CBM6-1/2CTC GAA TTC ACC TGC ACC AAA GCC AATFOR
CAC CTC GAG TTA ATG TGT CTT GTT GREV
CBM6-2/2CTC GTC GAC AGC ACT GGC GGT GAC AACFOR
CAC CTC GAG TTA ATG TGT CTT GTT GREV
CBM6-2/2-FLCAT CAA CAA GAC ACA TGA AAG CAC TGG CGG TGA CFOR
GTC ACC GCC AGT GCT TTC ATG TGT CTT GTT GAT GREV
GH5CTC GCT AGC GTA CCC GCG TTA CAG GTGFOR
CAC CTC GAG ATC GAT GGG TTT ACGREV
GH5-CBM6-1/2CTC GCT AGC GTA CCC GCG TTA CAG GTGFOR
CAC CTC GAG ATG TGT CTT GTT GAT GAGREV
CBM6-1 Y33AGGT TTG AAT GTC GGC GCT ATC GAT GGC GGC GACFOR
GTC GCC GCC ATC GAT AGC GCC GAT ATT CAA ACCREV
CBM6-1 W92ACCC GCG ACG GGC GGC GCG CAA AAC TGG CAG ACCFOR
GGT CTG CCA GTT TTG CGC GCC GCC CGT CGC GGGREV
CBM6-1 E73AGGA CAG CTC CAA TTG GCA AAA GCC GGT GGC AGCFOR
CT GCC ACC GGC TTT TGC CAA TTG GAG CTG TCCREV
CBM6-2 E73ACAG CTT GAC ATT TGC AGA AGC AGG CGGFOR
CCG CCT GCT TCT GCA AAT GTC AGG CTGREV
  • Engineered restriction sites and mutation points are depicted in bold.

1

Molecular architecture of truncated derivatives of the CmCel5B used in this study.

Escherichia coli Origami DE3 cells harbouring all the recombinant expression vectors, except pGH5 and pGH5-CBM6-1/2, were cultured in Luria–Bertani broth at 37 °C to the mid-exponential phase (A600 nm 0.6), and recombinant protein expression was induced by the addition of 1 mM isopropyl 1-thio-β-d-galactopyranoside and incubation for a further 16 h at 19 °C. BL21 DE3 E. coli were used to express the catalytically active CmCel5B derivatives GH5 and GH5-CBM6-1/2, encoded by pGH5 and pGH5-CBM6-1/2, respectively, following the same procedure described above. Soluble recombinant proteins were purified by immobilized metal ion affinity chromatography as described previously (Charnock et al., 2000; Carvalho et al., 2003; Dias et al., 2004) and buffer exchanged into 20 mM Tris-HCl buffer, pH 7.5, containing 100 mM NaCl and 5 mM CaCl2 (Buffer A). Sodium dodecyl sulphate polyacrylamide gel electrophoresis showed that all the recombinant proteins were >95% pure.

Source of sugars used

All soluble polysaccharides were purchased from Megazyme International (Bray, County Wicklow, Ireland), except oat spelt xylan, laminarin and hydroxyethylcellulose, which were obtained from Sigma. Avicel (PH101) was obtained from Serva while acid-swollen cellulose was prepared as described previously (Najmudin et al., 2006).

Mutagenesis

Site-directed mutagenesis was carried out using the PCR-based NZYMutagenesis site-directed mutagenesis kit (NZYTech Ltd) according to the manufacturer's instructions, using plasmids pCBM6-1 and pCBM6-1/2 as templates. The sequences of the primers used to generate these mutants are displayed in Table 1. The mutated DNA sequences were sequenced to ensure that only the appropriate mutations had been incorporated into the nucleic acid.

Affinity gel electrophoresis (AGE)

The affinity of CBM6-1 and other CmCel5B derivatives for a range of soluble polysaccharides was determined by AGE. The method used was essentially that described by Tomme et al. (2000) using the polysaccharide ligands at a concentration of 0.1% (w/v) unless stated otherwise. Electrophoresis was carried out for 4 h at room temperature in native polyacrylamide gels containing 10% (w/v) acrylamide. The nonbinding negative control protein was bovine serum albumin. Quantitative assessment of binding was carried out as described previously (Takeo, 1984), using polysaccharide concentrations ranging from 0.001% to 0.5% (w/v).

Isothermal titration calorimetry (ITC)

The binding of the CBMs to carbohydrates was quantified through ITC. ITC was carried essentially as described previously (Xie et al., 2001; Carvalho et al., 2004; Flint et al., 2004), except that proteins were dialyzed into 50 mM NaHepes, pH 7.5, containing 2 mM CaCl2, unless otherwise specified, at 25 °C. The concentration of the ligands in the syringe was ∼3 mM for the oligosaccharides and 0.5% (w/v) for the polysaccharides and the CBMs in the reaction cell were at 80–100 μM. Integrated heat effects, after correction for heats of dilution, were analyzed by nonlinear regression using a single site-binding model (microcal origin, ver. 5.0, Microcal Software, Northampton, MA). The molar concentration of CBM binding sites present in the polysaccharide ligands was determined as described previously (Bolam et al., 2004). The fitted data yield the association constant (Ka) and the enthalpy of binding (ΔH). Other thermodynamic parameters were calculated using the standard thermodynamic equation: Embedded Image

All data show the average and SD of three independent titrations.

Enzyme assays

Enzyme activity was determined at 37 °C in 20 mM Tris-HCl, pH 7.5, 20 mM NaCl, 5 mM CaCl2 (Buffer A), using various polysaccharides as substrates and the rate of reducing sugar released was measured with the method described by Miller (1959). The assay mixture (600 μL) contained 20 μL of enzyme (with appropriate dilution) and concentrations of substrate ranging from 0.15% to 2.00% (w/v). The mixture was incubated at 37 °C for 5, 10 and 15 min to confirm linearity of the reaction. All reported results are the mean of three separate experiments.

Results and discussion

Ligand specificity of CBM6-1

CmCel5B is a modular cellulase containing an N-terminal family 5 glycoside hydrolase (GH) catalytic domain followed by CBM6-1 and CBM6-2 (Fontes et al., 1998). The structure of CBM6-2 revealed the presence of two ligand-binding sites in the β-sandwich structure, which were termed cleft A and cleft B (Henshaw et al., 2004; Pires et al., 2004). Cleft A was shown to bind to terminal glucose and xylose residues, while cleft B recognizes cellulose and β1,4-1,3-mixed linked glucans (Henshaw et al., 2004). The ligand specificity of CmCel5B CBM6-1 remains unknown and it is possible that both CBM6-1 and CBM6-2 act in concert during carbohydrate recognition. Alignment of CBM6-1, CBM6-2 and the laminarin-specific CBM6 of B. halodurans suggests that CBM6-1 cleft A is indeed functional, because all residues involved in ligand recognition in CBM6-2 are conserved in CBM6-1 (Fig. 2). In contrast, there are significant differences in the residues presented on the surface of cleft B in the two CmCel5B CBM6s. Significantly, major changes are observed in subsite 1 (CBM6-2 cleft B contains four sugar-binding subsites), where Gly44, present in CBM6-2, is absent in CBM6-1 and CBM6-2 Gln110 is replaced by leucine in CBM6-1. More significantly, in subsite 4, Lys114, which makes several direct hydrogen bonds with the sugar, is replaced by a serine in CBM6-1. These amino acid differences suggest that CBM6-1 cleft B should display a much restricted binding interface.

2

(a) Structural alignment of CBM6-1 and CBM6-2 with the laminarin-specific CBM6 from Bacillus halodurans (BhCBM6) and the xylan-specific CBM6 from Clostridium stercorarium (CsCBM6). Residues belonging to cleft A are boxed in orange and those belonging to cleft B are boxed in green. Arrows above the alignment represent the secondary structure elements of Cellvibrio mixtus CBM6-2. Residues that were subjected to mutagenesis are highlighted with an arrow below the sequence. The protein alignment was prepared with clustal w (Thompson et al., 1994). (b) Three-dimensional structure of C. mixtus CBM6-2 shown in color ramped format in complex with two cellobiose molecules bound to clefts A and B, respectively (PDB ID: 1UYX). Amino acids dominating carbohydrate recognition at the two binding sites are indicated.

In order to clarify the functional implications of these amino acid substitutions in the ligand affinity of CBM6-1, the CmCel5B noncatalytic domain was purified to electrophoretic homogeneity and its ligand specificity was evaluated by ITC and AGE. The data, presented in Figs 3 and 4, revealed that CBM6-1 binds to laminarin, with an affinity constant Ka of 4 × 103 M−1. Thus, CBM6-1 displays a relatively modest affinity for laminarin and determination of the thermodynamic parameters could not be performed with accuracy. However, in general, binding of CBMs to soluble saccharides is enthalpy-driven, with entropy making an unfavourable contribution to ligand binding (Notenboom et al., 2001; Boraston et al., 2003c; Bueren et al., 2005). Significantly, AGE did not reveal any binding of CBM6-1 to laminarin, which may reflect the low degree of polymerization of the polysaccharide (Nisizawa et al., 1963), and the stoichiometric binding of the polysaccharide to the ligand, which, collectively, would result in little electrophoretic retardation of the protein. A ligand specificity screen using AGE revealed that CBM6-1 displays no significant affinity for xyloglucan, the β1,4-β1,3-mixed glucans barley β-glucan and lichenan, the β1,4-glucan hydroxyethylcellulose, konjac glucomannan, oat spelt xylan, the β1,3-glucans laminarin and curdlan, carob galactomannan, potato galactan, pullulan or pustulan (data not shown). ITC confirmed that, without β1,3-glucans, CBM6-1 does not interact with β1,3-β1,4-mixed linked glucans or the other polysaccharides tested. In addition, CBM6-1 lacks the capacity to interact with insoluble cellulose preparations of Avicel and acid-swollen cellulose (data not shown).

3

Interaction of CmCel5B derivatives with lichenan analysed by AGE. Recombinant proteins were purified through IMAC (a) and subjected to AGE as described in Materials and methods (b). The proteins under analysis were CBM6-1/2 (lane 2), CBM6-1/2 with a mutation in cleft B of CBM6-1 (lane 3), CBM6-1/2 with a mutation in cleft B of CBM6-2 (lane 4), CBM6-1/2, with mutations in cleft B of both CBM6s (lane 5), CBM6-2 (lane 6) and CBM6-2/2 (lane 7). Lanes 1 and 8 contain protein molecular mass markers (a) and bovine serum albumin (b). The affinity of CmCel5B truncated derivatives from (a) and (b) for lichenan was determined through AGE (c). ac, % (w/v) ligand concentration.

4

ITC of wild type (CBM6-1) and mutant derivatives of CBM6-1 with laminarin. Upper parts of each panel show the raw heats of binding and the lower parts are the integrated heats.

CBM6-1 and CBM6-2 do not act cooperatively to bind polysaccharides

Although CBM6-2 displays specificity for β1,4- or mixed linked β1,4-β1,3-glucans, reflecting the substrate specificity of the appended CmCel5B GH5 catalytic module, while CBM6-1 recognizes β1,3-glucans, it is possible that the two modules display cooperativity particularly for polysaccharides that contain both β1,4- and β1,3-gluco linkages. To assess this possibility, the capacity of CBM6-1 fused to CBM6-2 (CBM6-1/2) to bind to β-glucans was evaluated through AGE. The data, examples of which are presented in Fig. 3b and the Ka values reported in Fig. 3c, suggest that the affinity of CBM6-1/2 for barley β-glucan is similar to CBM6-2, indicating that the two modules do not act cooperatively. To confirm that CBM6-2 can participate in avidity-mediated increases in affinity, two CBM6-2 modules were fused in tandem and the capacity of the bimodular protein to interact with mixed linked β-glucan was evaluated. The data (Fig. 3) confirm that a significant increase in affinity is observed when two copies of CBM6-2 are fused together. This phenomenon has been described for a variety of CBMs that display the same ligand specificities and are located in the same enzyme (Bolam et al., 2001; Boraston et al., 2002; Gilbert et al., 2002). To further confirm the inability of CBM6-1 to affect the interaction of CBM6-2 with mixed linked β-glucan, the pivotal residues involved in ligand binding were mutated separately and together in the cleft B regions of the two modules within CBM6-1/2 and the affinity of the resulting proteins for mixed-linked β-glucan was evaluated (cleft B exclusively interacts with this polysaccharide in CBM6-2). The data (Fig. 3) confirm that apparently only cleft B from CBM6-2 has the capacity to interact with mixed linked glucans. Therefore, taken together, the data presented here suggest that CBM6-1 does not act cooperatively with CBM6-2 to bind polysaccharides.

Mapping the ligand-binding site by mutagenesis

To evaluate as to which CBM6-1 binding cleft interacts with laminarin, the affinity of various mutant derivatives was evaluated by ITC. By analogy with CBM6-2, Tyr33 and Trp92 are key residues involved in ligand recognition in cleft A, while Glu73 plays a major role in carbohydrate recognition in cleft B. Thus, the following CBM6-1 mutants were constructed: Y33A (cleft A), W92A (cleft A), Y33A/W92A (cleft A), E73A (cleft B) and Y33A/E73A/W92A (clefts A and B) – CBM6-2 numbering used for simplification. The data presented in Fig. 4 revealed that all CBM6-1 mutant derivatives analysed display no significant binding affinity for laminarin, suggesting that both clefts contribute to laminarin recognition in CBM6-1. In addition, wild-type CBM6-1 is unable to recognize both laminarihexose and cellohexaose (data not shown). Lack of affinity for laminarihexose suggests that CBM6-1 laminarin-binding cleft accommodates carbohydrates with a degree of polymerization >6. Taken together, these results indicate that CBM6-1 binds to laminarin through residues located both in clefts A and B. Therefore, CBM6s have evolved the capacity to interact with laminarin in at least three different binding sites. Thus, cleft A of CmCel5B CBM6-2 binds to the nonreducing end of all β-linked glucans or xylans assessed. In contrast, cleft A of B. halodurans CBM6 binds specifically to the nonreducing end of laminarin and the carbohydrate establishes several productive interactions along its chain with the surface of the protein that is formed by the loops connecting the two β-sheets, which display a convex topology (Fig. 5). The residues mediating ligand recognition in BhCBM6, which are not located in cleft B, are not conserved in CBM6-1. In this work, CmCel5B CBM6-1 is shown to accommodate laminarin through residues located in clefts A and B, revealing a third location for laminarin recognition within the CBM6 family. The discrete, and topologically distinct, laminarin-binding sites may be intimately related to the capacity of CBM6 members to display an extensive range of ligand specificities and is in sharp contrast to other CBM families where variation in carbohydrate recognition reflects differences in the topology of a single binding site.

5

The three-dimensional structure of BhCBM6 (PDB ID: 1W9W). The structure of BhCBM6 is shown in color ramped format in complex with laminohexaose, which is depicted as grey carbons and red oxygens. Residues making contact with the ligand at cleft A are highlighted. The figure demonstrates that laminarin does not interact with cleft B, whose putative interacting residues are also depicted. The structure also shows that only a β1,3-glucan considerably longer than laminohexaose would be able to interact with both clefts A and B.

Role of CBM6 modules in the function of CmCel5B

To evaluate the role of the tandem CBM6s in the function of the glycoside hydrolase CmCel5B, derivatives of the enzyme, comprising the GH5 catalytic module alone (GH5) or fused to CBM6-1 and CBM6-2 (GH5-CBM6-1/2), were produced in the recombinant form. The temperature optimum of GH5 and GH5-CBM6-1/2 is ∼37 °C, with maximal activity at pH 7.5 (data not shown). The biochemical properties of both recombinant derivatives indicate that CmCel5B hydrolyses cellulosic substrates such as carboxymethyl cellulose, mixed linked β1,4-β1,3-glucans and glucomannans (Table 2). The enzyme did not hydrolyse laminarin, xyloglucan, mannan, galactomannan, carob-galactomannan, arabinogalactan, lupin galactan, arabinan, rhamnogalacturan, pectin galactan lupin, pectic galactan potato, polygalactutronic acid, rhamnogalacturan I, pustulan, pululan, pectin from apples and pectin from citrus. Comparison of the activities of GH5 and GH5-CBM6-1/2 revealed that the tandem CBM6s appear to mediate a threefold increase in the activity of CmCel5B against the insoluble substrate Avicel, but not significantly, against purified soluble polysaccharides (Table 2). This feature is common to many cellulose-binding CBMs (Hall et al., 1995; Bolam et al., 1998; Fernandes et al., 1999) and is believed to result from the increased concentration of the enzyme on the surface of the recalcitrant substrate. As soluble purified polysaccharides are highly accessible to enzyme attack, there is no requirement to sequester these biocatalysts on the surface of these isolated glycan chains.

View this table:
2

Specific activities of GH5 and GH5-CBM6-1/2 for a range of plant cell wall polysaccharides and targeting effect of CBM6-1/2

SubstrateEnzyme activity (kat)
GH5GH5-CBM6-1/2
Specific activity
Avicel1030
β-Glucan19 80919 694
Lichenan18 87618 575
Carboxymethylcellulose14 91214 779
Glucomannan12 26712 273
Xyloglucan (XG)00
  • * Moles of product formed per mole of enzyme per minute.

  • Assays performed at 37°C with 0.15% (w/v) of substrate except for Avicel where 2% (w/v) substrate was used.

  • The enzymes do not present measurable reducing sugar activity against laminarin, xyloglucan, mannan, galactomannan, carob-galactomannan, arabinogalactan, lupin galactan, arabinan, rhamnogalacturan, pectin galactan lupin, pectic galactan potato, polygalactutronic acid, rhamnogalacturan I, pustulan, pululan, pectin from apples and pectin from citrus.

Conclusion: biological rational for β1,3-glucan recognition by CmCel5B

In this report, CBM6-1 from C. mixtus CmCel5B was shown to bind β1,3-glucans through residues located both in clefts A and B and the ligand-binding interface is suggested to accommodate more than six β1,3-linked glucose residues. The biological significance of this specificity is intriguing because β1,3-glucans are usually not present in the cell walls of terrestrial plants. However, it is well known that β1,3 polysaccharides are important components of fungal cell walls (Cid et al., 1995). In addition, although β1,3-glucans are not known to be part of the cell walls of green, red or brown algae (Kloareg & Quatrano, 1988), laminarin that is localized in the vacuole is a storage polysaccharide in brown algae (Read et al., 1996). Cellvibrio mixtus is a saprophytic bacterium that colonizes the soil and, therefore, it is unlikely that algae polysaccharides are widely used as energy and carbon sources by this bacterium. However, it is possible that C. mixtus hydrolyses fungal cell wall polysaccharides. As fungi are the initial colonizers of plant biomass, it is possible that saprophytic bacteria could have developed fungal lytic mechanisms allowing them to compete with fungal species that also use plant cell wall carbohydrates as an energy source. Intriguingly, the catalytic module appended to CBM6-1 consists of a typical cellulase, arguing against this possibility. The fungal-targeting hypothesis may, however, be relevant to other C. mixtus glycoside hydrolases, such as CmCel16A, which is encoded by a gene that clusters with the CmCel5B locus (Centeno et al., 2006). CmCel16A presents a GH16 catalytic domain, which typically hydrolyses β1,3-glucans, and a family 32 CBM that specifically recognizes β1,3-glucans (C. Fontes, unpublished data). In contrast, CBM6-1 may target CmCel5B to plant cell wall locations that are being actively degraded by fungi, where a variety of complex carbohydrates are accessible to enzyme attack. When the enzyme is located in the vicinity of the fungal cell wall, the action of the microbial eukaryote may expose plant cell wall β1,3-1,4-glucans, which are targeted by the C-terminal CBM6-2 of the glucanase, enabling the catalytic module to degrade its target substrate. Therefore, the cooperative action of CBM6-1 and CBM6-2 allows CmCel5B to be directed to the region of the cell wall that is being degraded actively by fungi, through CBM6-1, where the catalytic module, in harness with CBM6-2, will hydrolyse plant cell wall β1,3-1,4-glucans.

In addition to the above-described hypothesis, CBM6-1 may also target the appended catalytic domain to the cell wall of oomycetes, a group of filamentous, unicellular heterokonts, physically resembling fungi, which may be plant pathogens (Bartnicki-Garcia, 1968). Nevertheless, this possibility seems unlikely because C. mixtus itself is not a plant pathogen. However, the CBM6-1 of CmCel5B may direct the enzyme to callose, a β1,3-glucan molecule produced by plants in response to a fungal attack (Verma & Hong, 2001). This targeting mechanism would allow the enzyme to be in close proximity with the plant cell wall mixed linked glucans, which are the substrates for the enzyme catalytic domain. Finally, it is also possible that the primary substrates targeted by CmCel5B are the β1,3-1,4-glucans located in the fungal cell wall itself. It has been recently demonstrated that the cell wall of Aspergillus fumigatus contains linear β1,3-1,4-glucans, which corresponds to 10% of the total β-glucan content (Fontaine et al., 2002). Chitin, galactomannan and the linear β1,3-1,4-glucans were found to be covalently linked to the nonreducing end of β1,3-glucans, explaining the specificity of CBM6-1 (Fontaine et al., 2002). The prevalence of β1,3-1,4-glucans in the cell wall of cellulolytic and hemicellulolytic fungi, however, is currently unknown.

Acknowledgements

This work was supported by grant PTDC/BIAPRO/69732/2006 from the Fundação para a Ciência e a Tecnologia, Portugal, and by the individual grants SFRH/BD/23784/2005 (to M.A.S.C.) and SFRH/BD/12562/2003 (to V.M.R.P.) from the same institution.

Footnotes

  • Editor: Dieter Jahn

References

View Abstract