Cellulose-binding domains

Cellulose-binding domains

(Parte 1 de 6)

Review article

Cellulose-binding domains Biotechnological applications

Ilan Levy, Oded Shoseyov*

The Institute of Plant Science and Genetics in Agriculture and The Otto Warburg Centre for Agricultural

Biotechnology, The Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, PO Box 12, Rehovot 76100, Israel

Accepted 6 January 2002


Many researchers have acknowledged the fact that there exists an immense potential for the application of the cellulose-binding domains (CBDs) in the field of biotechnology. This becomes apparent when the phrase ‘‘cellulose-binding domain’’ is used as the key word for a computerized patent search; more then 150 hits are retrieved. Cellulose is an ideal matrix for large-scale affinity purification procedures. This chemically inert matrix has excellent physical properties as well as low affinity for nonspecific protein binding. It is available in a diverse range of forms and sizes, is pharmaceutically safe, and relatively inexpensive. Present studies into the application of CBDs in industry have established that they can be applied in the modification of physical and chemical properties of composite materials and the development of modified materials with improved properties. In agro-biotechnology, CBDs can be used to modify polysaccharide materials both in vivo and in vitro. The CBDs exert nonhydrolytic fiber disruption on cellulose-containing materials. The potential applications of ‘‘CBD technology’’ range from modulating the architecture of individual cells to the modification of an entire organism. Expressing these genes under specific promoters and using appropriate trafficking signals, can be used to alter the nutritional value and texture of agricultural crops and their final products. D 2002 Elsevier Science Inc. All rights reserved.

Keywords: Cellulose; Cell wall; Cellulose-binding domain (CBD); Pulp; Paper; Biotechnology

* Corresponding author. The Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew

University of Jerusalem, PO Box 12, Rehovot 76100, Israel. Tel.: +972-8-9481084; fax: +972-8-9462283. E-mail address: shoseyov@agri.huji.ac.il (O. Shoseyov).

w.elsevier.com/locate/biotechadv Biotechnology Advances 20 (2002) 191–213

1. Introduction

It was proposed in late 1940s, that the initial stage in the enzymatic degradation of crystalline cellulose involves the action of an unknown nonhydrolytic component termed C1. This component was thought to be responsible for destabilization (nonhydrolytic disruption) of the cellulose structure, making the substrate accessible to the enzyme, Cx component (Reese et al., 1950). The cellulose-binding domain (CBD) was first demonstrated in the fungus Trichoderma reesei and the bacterium Cellulomonas fimi (Van Tilbeurgh et al., 1986; Gilkes et al., 1988). The connecting linker between the CBD moiety and the enzyme proved to be susceptible to proteolysis, thus allowing for isolation of the individual domain by limited proteolysis. Forty years after the C1–CX model was proposed, the first C1 component was cloned from Clostridium cellulovorans and C. fimi (Shoseyov et al., 1990; Shoseyov and

Doi, 1990; Din et al., 1991; Goldstein et al., 1993). This achievement gave researchers an opportunity to study the C1–CX hypothesis. To date, domain structures and biochemical functions of many CBDs have been deciphered (for review, see Gilkes et al., 1991; Davis,

In earlier studies of CBD–cellulose interactions, the presence of a CBD was shown to increase the effective concentration of enzyme on insoluble cellulose substrates, thereby assisting the enzyme through the phase transfer from soluble fraction (the enzyme) to insoluble fraction (the substrate) (Shoseyov and Doi, 1990; Beguin and Aubert, 1994; Din et al., 1994; Linder et al., 1995; Tomme et al., 1995a; Bolam et al., 1998; Suurnakki et al., 2000).

CBDs have been found in hydrolytic and nonhydrolytic proteins. In proteins that possess hydrolytic activity (cellulases, xylenases), the CBD is a discrete domain that concentrates the catalytic domains on the surface of the insoluble cellulose substrate (Gilkes et al., 1991; Tomme et al., 1995a,b, 1998; Linder et al., 1997; Teeri et al., 1998). The CBDs present in proteins that do not have hydrolytic activity compose part of a scaffolding subunit that organizes the catalytic subunits into a cohesive multienzyme complex known as a cellulosome. The enzymatic complex was found to function more efficiently in the degradation of cellulosic substrates (Woodward et al., 1988; Shoseyov and Doi, 1990; Doi et al., 1994; Beguin and Alzari, 1998; Bayer et al., 1998a,b). Removal of the CBD from the cellulase molecule or from the scaffolding in cellulosomes dramatically decreased enzymatic activity (Van Tilbeurgh et al., 1986; Tomme et al., 1988; Hefford et al., 1992; Goldstein et al., 1993; Coutinho et al., 1993; Carrard and Linder, 1999).

CBDs have also been found in several polysaccharide-degrading enzymes. In T. reesei,

CBD has been identified in hemicellulase, endo-mannanase and acetyl-xylanesterase (Stalbrand et al., 1995; Margolles-Clark et al., 1996). CBDs have been recognized in xylanase originating from Clostridium thermocellum (Kulkarni et al., 1999; Kim et al., 2000), esterase from Penicillium funiculosum (Kroon et al., 2000), and pectate lyase in Pseudomonas cellulosa (Brown et al., 2001). In addition, there exists the intriguing presence of such a domain in b-glucosidase located in Phanerochaete chrysosporium (Lymar et al., 1995). The presence of putative CBDs in plant endoglucanases has also been reported (Catala and Bennett, 1998; Trainotti et al., 1999). Expansins, that are believed to play a role in nonhydrolytic cell wall expansion, are homologues to CBDs and possess cellulose binding capabilities in vitro (Cosgrove, 2000).

Today, more than 200 putative sequences, in over 40 different species, have been identified. The binding domains are classified into 14 different families based on amino acid sequence, binding specificity, and structure (Gilkes et al., 1991; Tomme et al., 1995a,b, 1998; Bayer et al., 1998a). Families V and VIII contain only one member each, while Families I, I, and II consist of 40 or more members (Tomme et al., 1998). The CBDs can contain 30–180 amino acids, and exist as a single, double, or triple domain in one protein. Their location within the parental protein can be either C- or N-terminal and occasionally centrally positioned in the polypeptide chain. The affinity and specificity towards different cellulose allomorphs can vary (for an extended review on CBDs, see Gilkes et al., 1991; Henrissat, 1994; Tomme et al., 1995b, 1998; Bayer et al., 1998a,b. Extensive data and classification can de found in the Carbohydrate-Binding Module Family Server at http:// afmb.cnrs-mrs.fr/~pedro/CAZY/cbm.html). Three-dimensional structures of representative members of CBD Families I, IIa, I, IV, V, VI, IX, and XV have been resolved by crystallography and NMR (Xu et al., 1995; Tormo et al., 1996; Johnson et al., 1996; Brun et al., 1997; Sakon et al., 1997; Mattinen et al., 1998; Notenboom et al., 2001; Szabo et al., 2001; Czjzek et al., 2001). Data from these structures indicate that CBDs from different families are structurally similar and that their cellulose binding capacity can be attributed, at least in part, to several aromatic amino acids that compose their hydrophobic surface. The positions and angles between these aromatic amino acids differ between various CBD members. CBDCex, from Family IIa, contains a b-barrel-type backbone that displays aromatic amino acids on a relatively flat surface (Din et al., 1994; Tormo et al., 1996;

Nagy et al., 1998). On the other hand, CBDN1 from Family V, displays its aromatic amino acids in a narrow groove (Johnson et al., 1996; Tomme et al., 1996). CBDs from the same organism can differ in their binding specificity (Carrard and Linder, 1999) and, occasionally, two CBDs located on the same enzyme can also exhibit this distinction (Brun et al., 2000). Biochemical studies have shown that the course of events leading to the binding of CBD to cellulose is directed by several driving forces. In the case of CBDCex, which binds to crystalline cellulose, the process is entropically driven. The decrease in entropy can be attributed to a net loss in conformational freedom of the polysaccharide and protein side chains. Water hydration upon binding may be another factor leading to lower entropy (Creagh et al., 1996). On the other hand, binding of CBDN1 to amorphous cellulose is driven by enthalpy. This force can be attributed to heat release, which occurs upon complex formation that transpires through hydrogen and van der Waals bonding between the equatorial hydroxyl of the glucopyranosyl ring and the polar amino acids (Brun et al., 2000). Although the interaction of the CBD with the cellulose is occasionally irreversible, contact with the cellulose surface is dynamic. Jervis et al. (1997) demonstrated by using fluorescence recovery techniques, that CBDCex is mobile on the surface of crystalline cellulose when it appears in isolated form or as a module in xylanase. Furthermore, it was hypothesized that the binding of

Family IIa CBD from C. fimi to cellulose occurs either along or across the chain (McLean et al., 2000).

CBDs, constituting isolated modules, are utilized in many different applications. This article will review the potential applications of CBDs in diverse fields of biotechnology.

2. Bioprocessing

Large-scale recovery and purification of biologically active molecules continues to be a challenge for many biotechnology companies. Various purification procedures have been developed, of which biospecific affinity purification (affinity chromatography) has become one of the most rapidly developing divisions of immobilized affinity ligand technology. To date, several affinity tags have been developed that vary in size from several amino acids to a complete protein. Each individual affinity-based purification system embodies specific advantages. Cellulose, when compared with most immobilization systems, is an economical support-matrix for large-scale protein purification (for review, see Harakas, 1994; Wilchek and Chaiken, 2000; Lowe, 2001). The wide use of CBD as an affinity tag in expression and purification is illustrated in Fig. 1. This subject has been extensively reviewed (Ong et al., 1989; Greenwood et al., 1992; Bayer et al., 1994; Tomme et al., 1998; Saleemuddin, 1999) and described in several patents (Shoseyov et al., 1997, 1998a,b; Kilburn et al., 1999a,b;

Fig. 1. CBD-based expression and purification of recombinant proteins. Protein expression and purification via CBD involves several steps. (1) Gene fusion between cbd and a gene of interest. (2) Transformation of ligated plasmid vector into a prokaryotic or eukaryotic expression system. (3) Overexpression of the recombinant protein. (4) Purification by immobilization of CBD-tagged protein on cellulose. (5) Reconstitution of the target protein by (A) gentle elution of target protein from cellulosic matrix, (B) addition of a ligand or a substrate, or (C) proteolytic cleavage of the engineered sequence located between the CBD and the target protein.

Meade, 2000; Meade et al., 2001); therefore, this section will review only current developments.

Recent reports relating to multifunctional studies have further affirmed the feasibility of employing CBD as an affinity tag. Atrazin dechlorinating enzyme (Kauffmann et al., 2000), a-amylase (Bjornvad et al., 1998), lipase B (Rotticci-Mulder et al., 2001), glucoamylase (Jiang and Radford, 2000), and organophosphorus hydrolase (Richins et al., 2000) have been expressed as CBD-fused enzymes while retaining their high specific activity. In addition, human T cell connective tissue-activator peptide-I (CTAP-I) (Rechter et al., 1999), human hsp60 epitope (Shpigel et al., 1998a,b), protein A (Shpigel et al., 2000), and murine stem-cell factor (SCF) (Doheny et al., 1999; Boraston et al., 2001) were also expressed as CBD-fused proteins. All these studies established the fact, that CBDs can be employed as high-capacity purification tags for the isolation of biologically active target peptides, at relatively low cost.

Matrix-assisted refolding of recombinant proteins is one of the approaches taken in order to prevent the aggregation of protein during the course of renaturation. At the present, only histidine and arginine tags have been found to be suitable for this process as they maintain their matrix binding ability under denaturing conditions (Stempfer et al., 1996; Glansbeek et al., 1998) Recently, a CBD derived from Clostridium thermocellum was used as a tag for matrix-assisted refolding of a single-chain antibody expressed in Escherichia coli. This CBD binds to cellulose in the presence of 6 M urea. The method was shown to provide a threefold increase in protein yields when compared to standard refolding procedures (Berdichevsky et al., 1999a).

Phage display technology is a proven tool for isolating biologically active molecules

(Cortese et al., 1996; Forrer et al., 1999; Johnsson and Ge, 1999; Rodi and Makowski, 1999; Gaskin et al., 2001). One of the limitations preventing extensive implementation of this technology is the relatively high proportion of clones that lack insertions within the library. In a recent study, CBD from Clostridium thermocellum was fused to a single-chain antibody (scFv) and expressed as scFv–CBD phage display library. The CBD tag allowed for rapid recovery of phages that displayed functional inserts, thus increasing the efficiency of the screening process for recombinant antibodies (Berdichevsky et al., 1999b).

Direct passive coating of proteins to plastic can result in partial or total denaturation of the adsorbed molecule. This can be attributed to hydrophobic interaction between the protein and the solid phase. (Suter and Butler, 1986; Schwab and Bosshard, 1992). Recently, Levy and Shoseyov (2002) used phage display of a random peptide library to screen for peptides that enable indirect immobilization of proteins to a solid surface via CBD. It was demonstrated that the affinity between the four amino acids tag and a CBD could be employed to immobilize horseradish peroxidase (HRP) on CBD-precoated cellulosic surfaces. This technique enables researches to tailor-make fusion tags that will mediate indirect noncovalent immobilization of proteins to solid matrixes.

The binding of CBD to cellulose can be classified as reversible (Family I) or irreversible

(Families I and II). When Family I CBD is used as an immobilizing tag, a low-rate column leakage is often observed (Linder et al., 1996). In order to overcome this problem, Linder et al. (1998) constructed a chimerical protein that was composed of CBDHII and CBHI from T. reesei, and a single-chain antibody. A significant decrease in protein leakage was observed with this unique structure, indicating that CBD–antibody fused proteins are suitable tools for affinity chromatography.

(Parte 1 de 6)