Elsevier

Journal of Biotechnology

Volume 96, Issue 1, 13 June 2002, Pages 103-118
Journal of Biotechnology

Phage display selects for amylases with improved low pH starch-binding

https://doi.org/10.1016/S0168-1656(02)00041-XGet rights and content

Abstract

Directed evolution of secreted industrial enzymes is hampered by the lack of powerful selection techniques. We have explored surface display to select for enzyme variants with improved binding performance on complex polymeric substrates. By a combination of saturation mutagenesis and phage display we selected α-amylase variants, which have the ability to bind starch substrate at industrially preferred low pH conditions. First we displayed active α-amylase on the surface of phage fd. Secondly we developed a selection system that is based on the ability of α-amylase displaying phages to bind to cross-linked starch. This system was used to probe the involvement of specific β-strands in substrate interaction. Finally, a saturated library of α-amylase mutants with one or more amino acid residues changed in their Cβ4 starch-binding domain was subjected to phage display selection. Mutant molecules with good starch-binding and hydrolytic capacity could be isolated from the phage library by repeated binding and elution of phage particles at lowered pH value. Apart from the wild type α-amylase a specific subset of variants, with only changes in three out of the seven possible positions, was selected. All selected variants could hydrolyse starch and heptamaltose at low pH. Interestingly, variants were found with a starch hydrolysis ratio at pH 4.5/7.5 that is improved relative to the wild type α-amylase. These data demonstrate that useful α-amylase mutants can be selected via surface display on the basis of their binding properties to starch at lowered pH values.

Introduction

In recent years the possibilities for creating genetic diversity have been expanded enormously with the advent of the PCR technique (error prone PCR, PCR with spiked oligo's, staggered extension process) and DNA shuffling (Crameri et al., 1997, Zhao et al., 1998, Matsumura et al., 1999). The combination of gene pool diversification and selection for function is nowadays considered as one of the strongest strategies for protein engineering that is collectively termed ‘directed evolution’. In evolution two processes play a key role: diversification of the gene pool and selection of the best individual of that pool. Most of the newly described PCR techniques have been exemplified on intracellular enzymes with in vivo selectable functions, such as β-lactamase that can be selected for by increasing the antibiotic concentrations (Stemmer, 1994). For non-essential or for extracellular proteins, which represent the majority of industrial enzymes, an in vivo functional selection is less obvious. The industrially important enzyme α-amylase is secreted from the cell and no cell bound enzymatic activity can be detected in the host Bacillus licheniformis. Living cells do not take up complex insoluble substrates, such as starch, cellulose, xylanase, and lipids and therefore enzymatic conversion without exception occurs in the extracellular medium. This leads to a loss of the physical link between the genotype—the DNA—and the phenotype—the enzymatic activity—that impairs the possibility for a selection.

In the field of antibody engineering the phage display technique has been well developed as a selection technique that maintains the coupling between externally displayed peptides and the genotype (for review see Hoogenboom et al., 1998). Phage display has also been used successfully to select for complete proteins, such as better ligands of extracellular receptors (Lowman and Wells, 1993), however in the field of enzyme engineering the method of phage display has been barely applied (Jestin et al., 1999, Demartis et al., 1999). A reason for this limitation may be that phage display typically selects for a binding interaction between the protein and its target, while for the improvement of an enzyme apart from binding also catalysis is needed. It follows that phage display selection may be successfully used for those enzymes where the rate limitation is in substrate binding (Km) rather than in maximal velocity (Vmax). In many industrial processes involving the conversion of complex polymeric substrates indeed the binding of an enzyme to the substrate is a particularly rate limiting step. Often these enzymes have evolved specific substrate binding domains at a location remote from the active site of the enzyme.

In order to explore this opportunity, we decided to investigate the possibility to use phage display for a distinct problem in the grain wet milling industry. α-Amylase (1,4 α-d-glucanhydrolase, EC 3.2.11) catalyses the hydrolysis of α-1,4 glucan linkages in starch to produce larger oligosaccharides and maltose. The enzyme is produced by a variety of microorganisms including Bacillus and Aspergillus. The enzyme from B. licheniformis is very heat stable and therefore it is particularly suited for the large-scale liquefaction of cornstarch in industry. In the liquefaction process, the α-amylase is mixed with starch slurry and heated to a temperature of 110 °C in order to obtain liquefaction. Before this process the pH of the slurry has to be adjusted from pH 4.5, the natural value, to a value between pH 5.8 and 6.5 (Crabb and Mitchinson, 1997). The process cannot be operated properly at a value below a pH of 5.9, since the α-amylase does not form oligosaccharides below that value. After the liquefaction the pH of the starch has to be adjusted again—now to pH 4.2—in order to provide the conditions for the saccharification process by the fungal exo 1-4 glycanohydrolase, that splits the oligosaccharides into glucose units. Therefore the availability of an α-amylase, which has a good performance at low pH, would eliminate process steps and reduce the costs for chemicals and ionic exchange resins related to these pH adjustments.

The structure of B. licheniformis α-amylase has been resolved and the co-ordinates have been released recently (Machius et al., 1995; Hwang et al., 1997). The αβ-barrel central domain including the active site residues has been extensively discussed. The C-terminal region is characterized by a number of β-sheet regions designated Cβ1 to Cβ8; however, to this domain of the enzyme no function has been ascribed (Machius et al., 1995). The α-amylase shows homology with other α-amylases and with the well-studied cyclodextrin transferases. Notably in the latter enzymes the C-terminal domain is significantly extended and in this domain β-sheet residues have been identified that are responsible for the interaction with starch (Penninga et al., 1996). Despite the fact that no specific starch-binding domain has been assigned to the B. licheniformis α-amylase structure, functional studies show that the enzyme can scroll along the starch surface to break down the complete polymer into oligosaccharides (Helbert et al., 1996). This suggests a specific interaction between the enzyme and its complex starch substrate distinct from the hydrolytic reaction. In most polysaccharide degrading enzymes the catalytic and substrate binding domains are separated, implying that the starch-binding property of the protein can be independent from the catalytic action of the enzyme. Therefore we decided to study if the C-terminal domain of the B. licheniformis α-amylase is indeed involved in starch-binding and to explore the possibility to develop a phage display selection method for α-amylase mutants which have altered starch-binding. The starch-binding property was, without relating this property to a specific part of the protein molecule, already used to develop a large-scale purification procedure, which involves affinity chromatography to raw starch or cross-linked starch (Weber et al., 1976, Rozie et al., 1991, Satoh et al., 1993, Somers et al., 1995). Interestingly at pH 4.5 the enzyme shows a reasonable hydrolytic activity on the small substrate heptamaltose, whereas the activity on polymeric starch is zero. Based on this observation we hypothesized that a decreased starch-binding capacity hampers the application of α-amylase at pH values lower than six. In the past the complete gene was subjected to low intensity random mutagenesis and the variants with altered halo formation capacity were selected. The mutations selected in those experiments reside in the αβ-barrel central domain (Quax et al., 1994, Mitchinson et al., 1998). We decided to address specifically the C-terminal domain by a saturation mutagenesis method that probes every residue over a limited region. In order to explore a large sequence space a powerful selection is essential. The phage display technique offers such a possibility.

The technique was originally developed by Smith (1985), who showed that small peptides can be expressed on the surface of the filamentous phage fd as a fusion to the N-terminus of g3p coat protein resulting in the ‘display’ of that peptide. It was shown that it is possible to select for a phage encoding a specific peptide by affinity binding. It has been reported that a single phage particle can be selected out of a population of 109 by multiple rounds of selection and enrichment (For a review see Cortese et al., 1994). The selection of a desired peptide, e.g. by binding it to an antibody, automatically leads to the co-selection and isolation of the DNA that codes for the peptide. Initially the application of this technique was restricted to small peptides. Later it was found that protein domains (Roberts et al., 1992) and even complete proteins could be functionally expressed as fusion proteins to g3p. Also β-lactamase was displayed on the surface of phage fd as a fusion with g3p coat protein. A suicide inhibitor was used to selectively enrich a mutant enzyme (Soumillion et al., 1994). Recently we have demonstrated that also penicillin G acylase, an enzyme requiring post-translational processing in the periplasm, can be expressed actively on the surface of a phage particle (Verhaert et al., 1999).

So by displaying α-amylase on the surface of a phage and by setting up a selection system that discriminates between α-amylases (linked to phages) on their starch-binding ability, we here demonstrate that it is possible to isolate amylases with improved substrate binding. Obviously this method is not restricted to α-amylase. We foresee that in a similar fashion improved variants of polymeric substrate converting enzymes such as cellulases, xylanases, chitinases, ligninases, lipases and proteases can be selected.

Section snippets

Materials

The following bacterial strains were used: E. coli GM-1 (Miller et al., 1977), E. coli TG1: Sup E, K 12 Δ(lac-pro), thi, hsd D5/F', traD 36, pro AB, lacIq, lac Z ΔM15 and E. coli HB 2151: K12 Δ(lac-pro), ara, nalr,thi/F', pro AB, lacIq, lac Z ΔM15. Bacteriophage M13KO7 and phagemid pCANTAB 5E were bought from Pharmacia, phagemid pMc/NdeI was used as described (Stanssens et al., 1989). Purified α-amylase from B. licheniformis was kindly provided by Gist-brocades N.V. (Delft, The Netherlands).

Display of α-amylase of B. licheniformis on phage fd

The α-amylase gene of B. licheniformis was isolated from the plasmid pLAT3 (27) using the primers Af1 and Ar1 and incorporated in the pCANTAB 5E plasmid yielding a g3p fusion (pRα1, Fig. 1). To ensure proper transport of the g3p-amylase fusion the modified g3p signal sequence with an optimised maturation site that has been used successfully for the display of penicillin acylase was used (Verhaert et al., 1999). E. coli non–amber suppressor cells were transformed with the vector and after growth

Discussion

Phage display has been used extensively as a tool to select for protein with improved affinities. Although it was suggested that display techniques could also be used to select improved enzymes, the demonstration of a practical example has still to be realized. As a major reason for this failure, it has been put forward that display techniques can only be used to select for binding, not for catalysis (Forrer et al., 1999).

As binding can be considered as an essential part of catalysis we have

Acknowledgements

We thank Dr. Wolfgang Aehle (Genencor International B.V.) for introducing us to the problem of poor starch hydrolysis of amylase at low pH. This project was partly funded by the Foundation for Technological Sciences, STW, which is a division of the Netherlands Science Foundation, NWO.

References (31)

Cited by (31)

  • Protein engineering: Methods and applications

    2022, Advances in Protein Molecular and Structural Biology Methods
  • Heterogeneous catalysis on the phage surface: Display of active human enteropeptidase

    2013, Biochimie
    Citation Excerpt :

    In addition to various applications such as analysis of protein–protein or protein–DNA interactions, generation of target-specific antibodies, and screening for receptor agonists and antagonists, the phage display technology [39] has been successfully used as an enzyme engineering tool to determine the substrate specificity of enzymes, develop modulators of their active and allosteric sites, search for enzymes with novel specificities, etc. [27,40,41]. The examples of enzymes that have been phage-displayed to study their action mechanisms include trypsin [42], β-lactamases [43], metallo-β-lactamase β LII from Bacillus cereus [44], amylases [45], subtilisin [46,47], and carboxylesterase of Bacillus subtilis [48]. The mutant variant of 4′-phosphopantetheinyl transferase showing a more than 300-fold increase in catalytic efficiency toward 3′-dephospho-CoA was selected by co-displaying peptide substrates and enzyme mutants on the M13 phage surface as fusions to the phage capsid protein pIII [49].

  • Microbial acid-stable α-amylases: Characteristics, genetic engineering and applications

    2013, Process Biochemistry
    Citation Excerpt :

    A string of substitutions have been made in BLA to stabilize the protein and improve its performance at acidic pHs [88–90,93,94,103–105]. Site-directed mutagenesis and saturated mutagenesis have been employed to tailoring the optimal pH of a number of enzymes, including the α-amylase from B. licheniformis [106] and soybean β-amylase [107], but the catalytic rate of these enzymes was affected. Liu et al. [104] described the relationships between acid resistance and the structural features of different mutants of BLA with changes at two crucial residues, L134 and S320 (Fig. 3).

  • Oxidation of cornstarch using oxygen as oxidant without catalyst

    2011, LWT
    Citation Excerpt :

    Major properties affected by such modifications are solubility, viscosity, performance and resistance to ageing of native starch and of solutions or pastes prepared from them. Modified starch is widely used in various industries (Ishida & Mitsuo, 2004; Krishnan, Bhosale, & Singhal, 2005; Krishnan, Kshirsagar, & Singhal, 2005; Mostafa, & Morsy, 2004; Razem, Katusin-Razem, Starcevic, & Galekovic, 1990; Verhaert, Beekwilder, Olsthoorn, van Duin, & Quax, 2002). A commonly used method of starch modification is its oxidation with chemical agents differing in their oxidation power (Achremowicz, Gumul, Bala-Piasek, Tomasik, & Haberko, 2000; Li & Vasanthan, 2003; Veelaert, de Wit, & Tournois, 1994; Zhu, Sjoholm, Nurmi, & Bertoft, 1998).

  • Phage display: Concept, innovations, applications and future

    2010, Biotechnology Advances
    Citation Excerpt :

    Phage display has been used in enzymology to determine the substrate specificity and to develop modulators of both the active and allosteric sites of the enzyme (Diamond, 2007; Kehoe and Kay, 2005; Kay et al., 2001; Benhar, 2001). The method can be used to display mutants of enzymes to study their mechanisms of action (Vanwetswinkel et al., 2000; Ponsard et al., 2001; Verhaert et al., 2002). Since filamentous phage is resistant to broad range of proteases, it has been used in identification of substrates of various proteases (Matthews and Wells, 1993; Diamond, 2007).

View all citing articles on Scopus
1

Present address: Cargill R&D Europe Lelyweg 31, 4612, PS Bergen op Zoom, The Netherlands.

2

Present address: Department of Molecular Biology, Plant Research International, P.O. Box 16, 6700, AA Wageningen, The Netherlands.

View full text