Journal of Molecular Biology
Volume 275, Issue 2, 16 January 1998, Pages 309-325
Journal home page for Journal of Molecular Biology

Regular article
High-resolution crystal structures reveal how a cellulose chain is bound in the 50 Å long tunnel of cellobiohydrolase I from Trichoderma reesei1

https://doi.org/10.1006/jmbi.1997.1437Get rights and content

Abstract

Detailed information has been obtained, by means of protein X-ray crystallography, on how a cellulose chain is bound in the cellulose-binding tunnel of cellobiohydrolase I (CBHI), the major cellulase in the hydrolysis of native, crystalline cellulose by the fungus Trichoderma reesei. Three high-resolution crystal structures of different catalytically deficient mutants of CBHI in complex with cellotetraose, cellopentaose and cellohexaose have been refined at 1.9, 1.7 and 1.9 Å resolution, respectively. The observed binding of cellooligomers in the tunnel allowed unambiguous identification of ten well-defined subsites for glucosyl units that span a length of ∼50 Å. All bound oligomers have the same directionality and orientation, and the positions of the glucosyl units in each binding site agree remarkably well between the different complexes. The binding mode observed here corresponds to that expected during productive binding of a cellulose chain. The structures support the hypothesis that hydrolysis by CBHI proceeds from the reducing towards the non-reducing end of a cellulose chain, and they provide a structural explanation for the observed distribution of initial hydrolysis products.

Introduction

Cellulose is a natural, linear polymer composed of 1,4-linked β-d-glucose units. As the predominant component of plant cell walls, native cellulose is the most abundant organic substance in the biosphere. The growing needs for utilisation and recovery of natural resources has prompted extensive scientific efforts towards a better understanding of the basic mechanisms behind cellulose degradation. In nature, cellulose is recycled through the action of versatile, multicomponent microbial enzyme systems (reviewed in Tomme et al., 1995). Among these, the enzymes secreted by the filamentous fungus Trichoderma reesei are known to be highly effective against native, crystalline cellulose.

T. reesei produces four endoglucanases and two cellobiohydrolases. Cellobiohydrolase I and II (CBHI and CBHII) are believed to hydrolyse cellulose chains from their ends, and to be responsible for hydrolysis of the more inaccessible, crystalline regions of cellulose, an event in which CBHI, in particular, appears to play a key role Chanzy et al 1983, Irwin et al 1993. The T. reesei cellulases have a two-domain organisation with a large catalytic domain connected to a small cellulose-binding domain (CBD) via a glycosylated linker peptide (Tomme et al., 1988). In the case of CBHI, both domains have been shown to bind specifically to the cellulose surface Stahlberg et al 1991, Linder et al 1995. Recent experiments have shown that the CBD of CBHI binds to the crystalline cellulose surface reversibly, exhibiting an exchange rate which is compatible with the rate of hydrolysis of the catalytic domain (Linder & Teeri, 1996). The CBD, together with the linker region, thus mediates a dynamic attachment of the catalytic domain on the crystallite’s surface.

CBHI is a retaining enzyme (Knowles et al., 1988), belonging to family 7 of the glycosyl hydrolases (Henrissat 1991, Henrissat and Bairoch 1993; reviewed by Davies & Henrissat, 1995). It hydrolyses β-1,4-linkages from the ends of cellulose chains to liberate β-cellobiose as the main product Schmid and Wandrey 1990, Vrsanska and Biely 1992, Nidetzky et al 1994. We have reported the crystal structure of the catalytic domain of T. reesei CBHI complexed with a glucoside derivative, o-iodobenzyl-1-thio-β-d-glucoside (IBTG; Divne et al., 1994). CBHI folds into a large β-sandwich, similar in topology to those found in the legume-lectin family (concanavalin A; Edelman et al., 1972), and the family 16 bacterial lichenases (Bacillus 1,3-1,4-β-glucanase; Keitel et al., 1993).

A striking feature of the structure is an extensive cellulose-binding tunnel that was initially estimated to contain seven glucosyl-binding sites (Divne et al., 1994). The tunnel structure in CBHI and the location of the active site at the far end of the tunnel suggested that the enzyme could processively split off cellobiose units from the reducing end of the chain, while the glucosyl units on the non-reducing side of the point of cleavage remain attached to the enzyme (Divne et al., 1994). The suggestion that CBHI hydrolyses a cellulose chain from the reducing end, rather than the non-reducing end, is also in agreement with biochemical studies Vrsanska and Biely 1992, Barr et al 1996. The presence of the tunnel, that can provide a large number of interactions with loose cellulose-chain ends on the surface of a crystallite during enzymatic action, provided a plausible explanation for the specific binding of the catalytic domain to the cellulose surface.

The active site in the CBHI tunnel was identified from the position of the IBTG ligand in the wild-type structure, and three acidic residues, Glu212, Asp214 and Glu217, were suggested to be catalytically important (Divne et al., 1994). In particular, the pair Glu212-Glu217 appeared suitably positioned to act as the nucleophile and acid/base catalyst, respectively, in a double-displacement nucleophilic substitution at the anomeric C1 atom Koshland 1953, Sinnott 1990. The role of Asp214 would then be to maintain the correct position and protonation state of Glu212. Our recent results from activity measurements using catalytically impaired mutants of CBHI have lent support to this hypothesis (Ståhlberg et al., 1996). Three isosteric, single mutations were introduced to produce the variants E212Q, D214N and E217Q. The E212Q and E217Q mutants were unable to degrade crystalline cellulose, whereas the D214N mutant displayed high residual activity. On the small chromophoric substrate 2-chloro-4-nitrophenyl-β-lactoside, the kcat values were reduced to 1/2000 (E212Q), 1/85 (D214N) and 1/370 (E217Q) of the wild-type activity, with unaffected KM values. Crystal structures of E212Q and D214N CBHI without bound saccharides showed that the active-site structure remained unaffected by the mutations (Ståhlberg et al., 1996).

The aim of this study has been to arrive at a better understanding of how CBHI is able to hydrolyse crystalline cellulose by studying the detailed binding of a substrate chain inside the substrate tunnel. The residual activity of the variants E212Q and E217Q proved sufficiently low to allow crystal-structure determination of the mutant enzymes in complex with oligomeric fragments of the natural substrate. We report here the structures of E212Q CBHI in complex with cellotetraose and cellopentaose, and of E217Q CBHI with cellohexaose. Based on these structures we present a complete mapping of glucosyl-binding sites in the tunnel of T. reesei CBHI, and discuss the implications of the interactions in the tunnel for hydrolysis of crystalline cellulose.

Section snippets

Quality of the final models and overall structures

Statistics on data, refinement and the final models are given in Table 1. The models E212QG4, E212QG5 and E217QG6 have been refined at 1.9, 1.7 and 1.9 Å resolution to final R (Rfree) values of 0.196 (0.240), 0.190 (0.235) and 0.176 (0.215), respectively. The complexes crystallise in the same space group, I 222. All models contain the complete catalytic domain (residues 1 to 434), one N-acetyl glucosamine residue covalently linked to Asn270, and two Co2+(one of which lies on a crystallographic

Discussion

Our first structural studies on CBHI (Divne et al., 1994) revealed a molecule with a ∼40 Å long active-site tunnel lined with aminoacid side-chains forming a complex set of hydrogen bonds and salt links. We estimated that the tunnel contained seven glucosyl-binding sites, four of which would be formed by tryptophan residues. The initial structure was solved in complex with o-iodobenzyl-1-thio-β-d-glucoside, which allowed the identification of three amino acids (Glu212, Asp214 and Glu217),

Crystallisation and data collection

The mutagenesis, expression and purification of the mutant proteins has been described previously (Ståhlberg et al., 1996). Crystallization experiments were carried out using the hanging-drop vapour-diffusion method (McPherson, 1982). Crystals of E212Q and E217Q were obtained by mixing equal volumes of protein (8 mg/ml) and reservoir solution (100 mM morpholinoethane sulphonic acid (Mes) buffer (pH 6.0), 18% (w/v) mPEG 5000, 10 mM CoCl2 and 0.02% (w/v) azide). The drops were equilibrated at

Acknowledgements

This work has been supported by the European Commission (Biotechnology program DG XII, BIO2 CT943030), NUTEK, the Swedish Natural Science Research Council, Nordisk Industrifond and Alko Ltd. We thank the following people for providing coordinates prior to their official release: Dr Wolfram Saenger for the cellotetraose coordinates, Dr Ivo Tews for the chitobiase complex coordinates and Dr Gideon Davies for the thio-oligosaccharide coordinates. We also thank Dr Gerard Kleywegt for collecting the

References (59)

  • J. Ståhlberg et al.

    Activity studies and crystal structures of catalytically deficient mutants of cellobiohydrolase I from Trichoderma reesei

    J. Mol. Biol.

    (1996)
  • P. Tomme et al.

    Cellulose hydrolysis by bacteria and fungi

    Adv. Microb. Physiol.

    (1995)
  • M. Vrsanská et al.

    The cellobiohydrolase I from Trichoderma reesei QM 9414action on cello-oligosaccharides

    Carbohyd. Res.

    (1992)
  • N.K. Vyas

    Atomic features of protein-carbohydrate interactions

    Curr. Opin. Struct. Biol.

    (1991)
  • B.K. Barr et al.

    Identification of two functionally distinct classes of exocellulases

    Biochemistry

    (1996)
  • P. Biely et al.

    Substrate-binding site of endo-1,4-β-xylanase of the yeast Cryptococcus albidus

    Eur. J. Biochem.

    (1991)
  • A.T. Brünger

    X-PLOR: A System for Crystallography and NMR

    (1990)
  • A.T. Brünger

    Free R valuea novel statistical quantity for assessing the accuracy of crystal structures

    Nature

    (1992)
  • A.T. Brünger et al.

    Crystallographic R factor refinement by molecular dynamics

    Science

    (1987)
  • M. Claeyssens et al.

    Fungal cellulase systems. Comparisons of the specificities of the cellobiohydrolases isolated from Penicillium pinophilum and Trichoderma reesei

    Biochem. J.

    (1989)
  • The CCP4 suiteprograms for protein crystallography

    Acta Crystallog. sect. D

    (1994)
  • G.J. Davies et al.

    Nomenclature for sugar-binding subsites in glycosyl hydrolases

    Biochem. J.

    (1997)
  • C. Divne et al.

    The three-dimensional structure of the catalytic core of cellobiohydrolase I from Trichoderma reesei

    Science

    (1994)
  • G.M. Edelman et al.

    The covalent and three-dimensional structure of concanavalin A

    Proc. Natl Acad. Sci. USA

    (1972)
  • R.A. Engh et al.

    Accurate bond and angle parameters for X-ray protein structure refinement

    Acta Crystallog. sect. A

    (1991)
  • K.H. Gardner et al.

    The structure of native cellulose

    Biopolymers

    (1974)
  • B. Henrissat

    A classification of glycosyl hydrolases based on amino acid sequence similarities

    Biochem. J.

    (1991)
  • B. Henrissat et al.

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

    Biochem. J.

    (1993)
  • A. Hodel et al.

    Model bias in macromolecular crystal structures

    Acta Crystallog. sect. A

    (1992)
  • Cited by (374)

    View all citing articles on Scopus
    1

    Edited by K. Nagai

    View full text