Phage Display and Biotechnological Applications

Embed Size (px)

Citation preview

  • 8/6/2019 Phage Display and Biotechnological Applications

    1/33

    Research review paper

    Biotechnological applications of phage and cell display

    Itai Benhar*

    Department of Molecular Microbiology and Biotechnology, The George S. Wise Faculty of Life Sciences,Green Building, Room 202, Tel-Aviv University, Ramat Aviv 69978, Israel

    Abstract

    In recent years, the use of surface-display vectors for displaying polypeptides on the surface of

    bacteriophage and bacteria, combined with in vitro selection technologies, has transformed the way in

    which we generate and manipulate ligands, such as enzymes, antibodies and peptides. Phage display is

    based on expressing recombinant proteins or peptides fused to a phage coat protein. Bacterial display

    is based on expressing recombinant proteins fused to sorting signals that direct their incorporation on

    the cell surface. In both systems, the genetic information encoding for the displayed molecule is

    physically linked to its product via the displaying particle. Using these two complementary

    technologies, we are now able to design repertoires of ligands from scratch and use the power of

    affinity selection to select those ligands having the desired (biological) properties from a large excess

    of irrelevant ones. With phage display, tailor-made proteins (fused peptides, antibodies, enzymes,

    DNA-binding proteins) may be synthesized and selected to acquire the desired catalytic properties or

    affinity of binding and specificity for in vitro and in vivo diagnosis, for immunotherapy of human

    disease or for biocatalysis. Bacterial surface display has found a range of applications in the expression

    of various antigenic determinants, heterologous enzymes, single-chain antibodies, and combinatorial

    peptide libraries. This review explains the basis of phage and bacterial surface display and discusses

    the contributions made by these two leading technologies to biotechnological applications. This review

    focuses mainly on three areas where phage and cell display have had the greatest impact, namely,

    antibody engineering, enzyme technology and vaccine development. D 2001 Elsevier Science Inc. All

    rights reserved.

    Keywords: Antibody engineering; Bacterial surface display; Enzyme evolution; Immobilized enzymes; Phage

    display; Vaccine development

    * Tel.: +972-3-6407511; fax: +972-3-6409407.

    E-mail address: [email protected] (I. Benhar).

    Biotechnology Advances 19 (2001) 133

    0734-9750/01/$ see front matterD 2001 Elsevier Science Inc. All rights reserved.

    PII: S 0 7 3 4 - 9 7 5 0 ( 0 0 ) 0 0 0 5 4 - 9

  • 8/6/2019 Phage Display and Biotechnological Applications

    2/33

    1. Introduction

    As a result of ongoing genome projects, a large-scale effort to measure and analyze

    protein protein interactions using experimental methods is under way. These include

    biochemical techniques such as co-immunoprecipitation or crosslinking (Silver and Hunt,

    1993), molecular biological techniques such as the two-hybrid system (Fashena et al., 2000)

    or the phage- and cell-display systems (reviewed herein), and genetic techniques such as

    mutant isolation and analysis (Gargus, 1987). Therefore, determining functions of novel gene

    products identified in genome sequencing efforts has become an important scientific

    challenge. Identifying proteins that interact with novel gene products can help in deducing

    their activity, regulation, and potential role in disease.

    Traditional biochemical analyses of enzyme structure and function and elucidation ofreceptor or antibody recognition processes are based typically on the utilization of

    efficient model systems. These are characterized by practical considerations, such as:

    enrichment of the molecules being studied, a source of tissue or cells that can be obtained

    in abundance with reasonable ease and the stability of the components to biochemical

    manipulations. The reductionist's approach is to systematically weed out the molecules of

    interest until purity.

    The advent of molecular genetics has enhanced the study of proteinligand interactions

    both by providing means to produce rare molecules in large quantities via gene-expression

    systems and by focusing on specific regions or residues of the proteins using site-directedmutagenesis. However, the most commonly used technologies for identifying proteinligand

    interactions, such as the yeast two-hybrid system (Fashena et al., 2000) and expression library

    screening (Maser and Calvet, 1995; McCarrey and Williams, 1994) are labor-intensive and

    often not amenable to high-throughput evaluation of multiple gene products.

    Two leading methodologies have been added to those mentioned above that provide new

    horizons for the study of proteinprotein interactions. These technologies are based on the

    production and presentation of enormous collections of peptides or proteins to be screened by

    corresponding `binders.' This enables the identification of binding pairs or the selection of

    molecules by virtue of their function. The approaches are based on the concept of generating

    a complete set of all possible combinations of the peptides or proteins of interest, collectively

    referred to as `combinatorial display.' The respective approaches involve the display of

    (poly)peptides on the surface of bacteria and the combinatorial display of synthetic peptides,

    organic structures and recombinant proteins on bacteriophage.

    2. Phage display

    The generation of new drugs has long involved the search amongst hundreds of thousands

    of components using well-defined in vitro screening tests, the output of which was chosen tomimic as closely as possible the desired in vivo activity of the new drug. Today, new library

    methodologies offer many alternative and at least as powerful routes, by combining the

    generation of billions of components with a fast screening or selection procedure to identify

    the most interesting lead candidates.

    I. Benhar / Biotechnology Advances 19 (2001) 1332

  • 8/6/2019 Phage Display and Biotechnological Applications

    3/33

    Phage display has proven to be a very powerful technique for obtaining libraries contain-

    ing millions or even billions of different peptides or proteins. Phage display (Smith, 1985) has been used for affinity screening of combinatorial peptide libraries to identify ligands for

    peptide receptors, define epitopes for monoclonal antibodies, select enzyme substrates (Kay

    et al., 1996), and screen cloned antibody repertoires (Griffiths and Duncan, 1998). Regulated

    associations between select proteins are the basis of cellular signal transduction pathways,

    and defining a signal transduction pathway can, in many cases, be reduced to tracing a chain

    of proteinprotein interactions.

    One of the most successful applications of phage display has been the isolation of

    monoclonal antibodies, and fragments thereof, using large phage antibody libraries (Winter et

    al., 1994). Indeed, in the last few years, very efficient techniques have been developed to

    design and build large libraries of antibody fragments, and ingenious selection procedureshave been established to derive antibodies with the desired characteristics. Much progress has

    been made in this rapidly developing field and many possible applications of phage

    technology have been developed (Hoogenboom et al., 1998). These include the creation

    and screening of libraries to discover novel therapeutic targets and methods for selection of

    biologically active ligands. The most widely used library methodology is based on the use of

    filamentous phage (Smith, 1985), a bacteriophage that infects male Escherichia coli. Display

    systems based on other phage were also developed, including display on phage l heads

    (Sternberg and Hoess, 1995; Mikawa et al., 1996; Santini et al., 1998) and tail (Maruyama et

    al., 1994; Kuwabara et al., 1997), P4 capsid (Lindqvist and Naderi, 1995), T7 capsid(Houshmand et al., 1999; Yamamoto et al., 1999), T4 capsid (Jiang et al., 1997; Ren and

    Black, 1998; Mullaney and Black, 1996, 1998), and MS2 coat (Mastico et al., 1993; Heal et

    al., 1999). The phage-display systems and the molecules that were displayed using them are

    listed in Table 1.

    Filamentous phage display is based on cloning DNA fragments encoding millions of

    variants of certain ligands (e.g. peptides, proteins or fragments thereof) into the phage

    genome, fused to the gene encoding one of the phage coat proteins (usually pIII, but also pIV,

    PVI or pVIII, Table 1). Upon expression, the coat protein fusion is incorporated into new

    phage particles that are assembled in the periplasmic space of the bacterium. Expression ofthe gene fusion product and its subsequent incorporation into the mature phage coat results in

    the ligand being presented on the phage surface, while its genetic material resides within the

    phage particle. This connection between genotype and phenotype allows the enrichment of

    specific phage, e.g. using selection on an immobilized target. Phage that display a relevant

    ligand are retained by virtue of their binding to the target, while nonadherent phages are

    washed away. Bound phage can be recovered from the surface, used to reinfect bacteria and

    reproduced for further enrichment, and eventually for analysis of binding. The success of

    ligand phage display hinges on the combination of this display and enrichment method, with

    the synthesis of large combinatorial repertoires on phage.

    Phage selection is not limited to the isolation of antibodies or short peptides as describedabove. As listed in Table 1, this approach has also been instrumental in studies and

    manipulation of a variety of other biologically active molecules, e.g. cytokines (Gram et

    al., 1993; Saggio et al., 1995; Buchli et al., 1997), receptors (Lowman et al., 1991; Scarselli et

    al., 1993), enzymes (McCafferty et al., 1991; Soumillion et al., 1994; Pedersen et al., 1998;

    I. Benhar / Biotechnology Advances 19 (2001) 133 3

  • 8/6/2019 Phage Display and Biotechnological Applications

    4/33

    Table 1

    Phage display systems

    Phage Coat

    protein

    Displayed

    molecules

    References

    M13 pIII (minor) Peptides Numerous reports. Recent reviews: Daniels and Lane, 1996;

    Lowman, 1997; Rodi and Makowski, 1999; Cabilly, 1999

    Antibodies Numerous reports. Recent reviews: Winter et al., 1994;

    Burton and Barbas, 1994; Rader and Barbas, 1997;

    Hoogenboom, 1997; Hoogenboom et al., 1998;

    Griffiths and Duncan, 1998

    Cytokines Gram et al., 1993; Saggio et al., 1995; Buchli et al., 1997

    Receptors Lowman et al., 1991; Scarselli et al., 1993

    Enzymes McCafferty et al., 1991; Soumillion et al., 1994;Pedersen et al., 1998; Demartis et al., 1999;

    Forrer et al., 1999

    Enzyme inhibitors Roberts et al., 1992; Pannekoek et al., 1993;

    van Meijer et al., 1996; Huang et al., 1998

    Catalytic antibodies Janda et al., 1994; Sastry et al., 1995; Baca et al., 1997;

    Fujii et al., 1998

    DNA-binding

    proteins

    Rebar and Pabo, 1994; Jamieson et al., 1994;

    Choo and Klug, 1994; Choo et al., 1997

    Cellulose-binding

    proteins

    Smith et al., 1998; Berdichevsky et al., 1999

    M13 pVI Enzyme inhibitors Jespers et al., 1995Enzymes Hufton et al., 1999

    cDNA libraries Hufton et al., 1999

    M13 pVIII (major) Peptides Numerous reports. Recent reviews: Felici et al., 1995;

    Ladner, 1995; Cortese et al., 1996; Lowman, 1997;

    Wilson and Finlay, 1998

    Antibodies Kang et al., 1991; Wang et al., 1997

    Enzymes Corey et al., 1993

    Enzyme inhibitors Markland et al., 1991, 1996

    M13 pVII/pIX Antibodies Gao et al., 1999

    l D (Head protein) Peptides Sternberg and Hoess, 1995

    IgG-binding protein Sternberg and Hoess, 1995Enzymes Mikawa et al., 1996

    Protein A Mikawa et al., 1996

    cDNA libraries Santini et al., 1998

    pV (Tail protein) Peptides Maruyama et al., 1994; Kuwabara et al., 1997

    Enzymes Maruyama et al., 1994

    P4 Psu capsid protein Peptides Lindqvist and Naderi, 1995

    T7 10B capsid protein P eptides Houshmand et al., 1999

    cDNA libraries Yamamoto et al., 1999

    T4 Hoc capsid protein P eptides Jiang et al., 1997

    Antibodies Ren and Black, 1998

    Soc capsid protein Peptides Jiang et al., 1997Antibodies Ren and Black, 1998

    Internal protein III Enzymes Mullaney and Black, 1996, 1998

    Green fluorescent

    protein

    Mullaney and Black, 1996

    MS2 Coat protein Peptides Mastico et al., 1993; Heal et al., 1999

    I. Benhar / Biotechnology Advances 19 (2001) 1334

  • 8/6/2019 Phage Display and Biotechnological Applications

    5/33

    Demartis et al., 1999; Forrer et al., 1999), substrates (Matthews and Wells, 1993; Ohkubo et

    al., 1999) enzyme inhibitors (Roberts et al., 1992; Pannekoek et al., 1993; van Meijer et al.,1996; Huang et al., 1998), catalytic antibodies (Sastry et al., 1995; Janda et al., 1994; Baca et

    al., 1997; Fujii et al., 1998) DNA-binding proteins (Rebar and Pabo, 1994; Jamieson et al.,

    1994; Choo and Klug, 1994; Choo et al., 1997), receptor domains for constructing binding

    molecules with novel properties (reviewed in Nygren and Uhlen, 1997), and cellulose-

    binding domains (Smith et al., 1998; Berdichevsky et al., 1999). Scaffolds different from

    antibodies have been reported to form suitable binding ligands for many types of molecules.

    There are ample examples of host scaffolds that contain sufficient permissive regions to

    accommodate a reasonable numbers of substitutions, which may be used to generate a library

    of localized variability. Alternative scaffolds reported to date include b-sheet proteins

    (Tramontano et al., 1994), a-helical bundle proteins (Ku and Schultz, 1995), combinationsof these two (Tramontano et al., 1994), a separate group of highly constrained protease

    inhibitors (Markland et al., 1996; Tramontano et al., 1994; Rottgen and Collins, 1995), and,

    most recently, Green fluorescent protein (GFP) (Abedi et al., 1998) and cellulose-binding

    domains (Smith et al., 1998).

    Since secreted as well as cytoplasmic and nuclear proteins have been displayed on phage,

    the use of phage display is often the first strategy to define permissive sites for randomization

    and to generate ligand-binding variants (Choo et al., 1997). Alternatively, the use of lambda

    phage (Mikawa et al., 1996), bacterial cell display (Georgiou et al., 1997), and eukaryotic

    cell-display methods have been reported. With regard to functional selection methods,choosing other types of molecules besides antibodies is validated by the fact that antibody

    expression and folding may be impaired in the subcellular location where the desired

    functional activity is required. It would be advantageous to engineer the antibody for

    intracellular expression, for example by building stable disulfide-free antibodies (Proba et

    al., 1997), or to use libraries of scaffolds that are naturally produced in the targeted cell

    organelle, provided that effective and sufficient structural diversity may be obtained.

    3. Bacterial surface display

    The targeting and anchoring of heterologous proteins to the outer surface of yeast and

    mammalian cells (Schreuder et al., 1991) have been utilized for various applications.

    However, this review is focused on bacterial surface-display systems that are becoming an

    increasingly important research area (Georgiou et al., 1997; Hofnung, 1991; Little et al.,

    1993; Francisco and Georgiou, 1994; Fischetti et al., 1996). Surface display of heterologous

    proteins in bacteria has been employed as a tool for basic and applied research in

    microbiology, molecular biology, vaccinology and biotechnology.

    The first examples of heterologous surface display were reported a decade ago, when short

    gene fragments were inserted into the genes for the E. coli outer membrane proteins LamB,OmpA and PhoE, and the gene fusion products were found to be accessible on the outer

    surface of the recombinant bacteria (Charbit et al., 1988; Agterberg et al., 1990a,b;

    Tommassen et al., 1994; Janssen and Tommassen, 1994). Those display systems, however,

    were not suitable in most cases for the display of large proteins (Tommassen et al., 1994).

    I. Benhar / Biotechnology Advances 19 (2001) 133 5

  • 8/6/2019 Phage Display and Biotechnological Applications

    6/33

    Table 2

    Bacterial display systems

    Bacterium Membrane protein Displayed molecules References

    A. Gram-negative bacteria

    E. coli LamB Peptides Charbit et al., 1988; Leclerc et al., 1991;

    Bradbury et al., 1993; Sousa et al., 1996;

    Persic et al., 1999

    OmpA Peptides Pistor and Hobom, 1988; Agterberg et al.,

    1990a,b

    Metal binding

    proteins

    Kotrba et al., 1999

    Pal Antibodies Fuchs et al., 1991, 1996; Dhillon et al., 1999

    Pap Protein A Steidler et al., 1993PhoE Peptides Tommassen et al., 1994; Janssen and

    Tommassen, 1994

    TraT Peptides Chang and Lo, 2000; Chang et al., 1999

    Type 1 fimbriae Peptides Hedegaard and Klemm, 1989

    FliC Peptides Lu et al., 1995

    FimH Peptides Pallesen et al., 1995

    Lpp-OmpA Enzymes Francisco et al., 1992

    Antibodies Francisco and Georgiou, 1994;

    Francisco et al., 1993b;

    Chen et al., 1996

    Receptors Benhar et al., 2000Toxin domain Benhar et al., 2000

    Cellulose binding

    proteins

    Francisco et al., 1993a

    AIDA-I Peptides Maurer et al., 1997

    Toxin domain Maurer et al., 1997

    F pilin Peptides Malmborg et al., 1997

    Pseudomonas Inp Enzymes Jung et al., 1998a,b

    HIV GP120 Kwak et al., 1999

    ExtB Peptides Sewani et al., 1998

    OmpC Peptides Xu and Lee, 1999

    Salmonella Flagellin Peptides Wu et al., 1989; Newton et al., 1989Neisseria IgAb Toxin domain Klauser et al., 1990

    E. coli OmpA Peptides Schorr et al., 1991

    Caulobacter RsaA Peptides Bingle et al., 1997

    B. Gram-positive bacteria

    Streptococcus M1 protein Peptides Pozzi et al., 1992

    M6 protein Peptides Oggioni et al., 1999

    Staphylococcus Synthetic anchor Protein A Schneewind et al., 1992, 1993

    Enzymes Schneewind et al., 1992, 1993

    Protein A Peptides Hansson et al., 1992

    Receptors Hansson et al., 1992Antibodies Gunneriusson et al., 1996, 1999

    FnBPB Enzymes Strauss and Gotz, 1996

    B. anthracis S-layer protein EA1 Peptides Mesnage et al., 1999

    I. Benhar / Biotechnology Advances 19 (2001) 1336

  • 8/6/2019 Phage Display and Biotechnological Applications

    7/33

    Since then, not only outer membrane proteins but also lipoproteins, fimbria proteins and

    flagellar proteins, as well as dedicated systems with coupled translocation and surfaceanchoring, have been employed to achieve heterologous surface display on gram-negative

    bacteria (Hofnung, 1991; Little et al., 1993; Francisco and Georgiou, 1994), for example E.

    coli and Salmonella spp. In parallel, systems have been described for heterologous surface

    display on gram-positive bacteria such as staphylococci (Hansson et al., 1992), streptococci

    (Pozzi et al., 1992; Hanski et al., 1992) and mycobacteria (Stover et al., 1993). The bacterial

    display systems and the molecules that were displayed using them are listed in Table 2.

    The most common application of bacterial surface display has been in the development of

    live bacterial vaccine-delivery systems. In this context, the cell-surface display of hetero-

    logous antigenic determinants has been considered advantageous for the induction of antigen-

    specific antibody responses when using live recombinant cells for immunization (Stover etal., 1993; Georgiou et al., 1997; Nguyen et al., 1995). Bacterial surface display has also been

    applied in generating whole-cell bioadsorbents for environmental purposes (Sousa et al.,

    1996), novel microbial biocatalysts [surface display of active enzymes on E. coli (Francisco

    et al., 1992, 1993a,b) and staphylococci (Strauss and Gotz, 1996)], diagnostic tools [bacteria

    with surface-displayed antibody fragments (Fuchs et al., 1991; Francisco et al., 1992)] and for

    the display of entire peptide libraries (Lu et al., 1995) as an alternative to the rapidly

    developing phage-display technology. It should, however, be specified that combinatorial

    display is commonly applied using phage systems, and rarely using bacterial display systems.

    The latter are usually dedicated to the display of functionally defined molecules.

    4. Surface display in gram-negative bacteria

    Display systems have also been developed for the expression of a number of heterologous

    proteins on the surface of gram-negative bacteria. Most often, foreign gene products have

    been fused to outer membrane proteins, such as the maltoporin LamB (Charbit et al., 1988),

    the phosphate-inducible porin PhoE (Tommassen et al., 1994), the outer membrane protein

    OmpA (Agterberg et al., 1990a,b), and to lipoproteins such as the major lipoprotein Lpp, theTraT lipoprotein and the peptidoglycan-associated lipoprotein PAL (Fuchs et al., 1991, 1996;

    Dhillon et al., 1999; Chang and Lo, 2000; Chang et al., 1999). Proteins from the filamentous

    structures present on gram-negative bacteria have also been employed for surface-expression

    purposes. These include fimbria proteins such as the FimA protein (Hedegaard and Klemm,

    1989) and the FimH adhesin of type I fimbriae (Pallesen et al., 1995), the flagellar protein

    flagellin (Wu et al., 1989; Newton et al., 1989) and pili proteins such as the PapA pilus

    subunit (Steidler et al., 1993) and F Pilin (Malmborg et al., 1997). Certain display systems

    have been based on proteins that have specific mechanisms for translocation and surface

    anchoring. The lipoprotein pullulanase from Klebsiella pneumoniae has been employed for

    surface-display purposes in E. coli (Kornacker and Pugsley, 1990). Using this system, thetarget protein becomes transiently anchored to the outer membrane by its N-terminal fatty

    acid moiety and is subsequently released into the culture medium. The b-domain of the

    Neisseria gonorrhoeae IgA protease precursor (Igab) and the E. coli AIDA-I have also been

    employed for surface exposure of various heterologous proteins. When used as a C-terminal-

    I. Benhar / Biotechnology Advances 19 (2001) 133 7

  • 8/6/2019 Phage Display and Biotechnological Applications

    8/33

    fusion partner, the Igab and AIDA-l mediate attachment of hybrid proteins on the outer

    surface ofSalmonella typhimurium and E. coli strains without the OmpT protease (Maurer etal., 1997; Klauser et al., 1990).

    One particularly interesting system is based on the combined features of an Lpp-OmpAH

    chimera (Francisco and Georgiou, 1994; Francisco et al., 1992, 1993a,b; Stathopoulos et al.,

    1996; Georgiou et al., 1997; Daugherty et al., 1998). The Lpp-OmpAH chimera is comprised

    of the signal sequence and first nine amino acids of Lpp, fused to a region encompassing

    either three or five transmembrane helices of OmpA. This Lpp-OmpA H display system has

    been shown to give efficient translocation and surface anchoring of the fused gene products,

    resulting in a high number of chimeric surface proteins present in an accessible form on E.

    coli cells.

    5. Surface display in gram-positive bacteria

    Gram-positive bacteria have also been considered suitable for bacterial surface-display

    purposes (Hansson et al., 1992; Pozzi et al., 1992; Fischetti et al., 1996; Georgiou et al.,

    1997). A range of proteins, including antigenic determinants, heterologous enzymes and

    single-chain Fv (scFv) antibodies, have been targeted and anchored to the cell surface of

    gram-positive bacteria. Several gram-positive bacterial hosts have been investigated in this

    context, but until recently the research has focused on nonpathogenic staphylococci used infood-fermentation processes (Hansson et al., 1992; Pozzi et al., 1992; Hanski et al., 1992;

    Samuelson et al., 1995). Schneewind et al. (1992, 1993, 1995) have elucidated the

    mechanisms of cell-surface targeting and subsequent anchoring of surface proteins on

    staphylococcal cells several years after heterologous surface display had been achieved on

    staphylococci and streptococci. They found that Staphylococcus aureus protein A was sorted

    to the cell surface by its C-terminal surface-anchoring region. That region consists of a

    charged repetitive region, postulated to interact with the peptidoglycan cell wall. The C-

    termini of numerous gram-positive bacterial surface receptors are highly homologous,

    suggesting that this or a similar mechanism is utilized for their targeting to the cell surface(Pozzi et al., 1992; Schneewind et al., 1993, 1995).

    The surface-display systems developed for Staphylococcus xylosus (Hansson et al.,

    1992) and Staphylococcus carnosus (Samuelson et al., 1995) both take advantage of the

    cell-surface-anchoring regions of protein A. The Staphylococcus gordinii surface-display

    system (Pozzi et al., 1992) uses the similar C-terminal region of the M6 protein of

    Streptococcus pyogenes to achieve surface exposure of various chimeric surface proteins.

    While the staphylococcal systems utilize gene expression from plasmid vectors, the

    streptococcal system is based on incorporation of the target genes into the streptococcal

    chromosome by homologous recombination (Hansson et al., 1992). Recently, heterologous

    surface expression was achieved in Bacillus anthracis by integrating into the chromosomea translational fusion harboring the DNA fragments encoding the cell wall-targeting

    domain of the S-layer protein EA1 tetanus toxin fragment C. This construct was

    expressed under the control of the promoter of the S-layer component gene (Mesnage

    et al., 1999).

    I. Benhar / Biotechnology Advances 19 (2001) 1338

  • 8/6/2019 Phage Display and Biotechnological Applications

    9/33

    The major applications of bacterial surface display (see below) are antibody and enzyme

    display, and vaccine development. In addition, other less common applications includepeptide display for mapping epitopes of monoclonal antibodies. In one example, a random

    peptide library in a conformationally constrained thioredoxin region was introduced into

    the flagellin gene of E. coli and thus surface-exposed on the E. coli flagellum (Lu et al.,

    1995). The epitopes for three different immobilized monoclonal antibodies were mapped

    by the selection of bacterial clones and the identified consensus epitopes had several amino

    acids in common with motifs found in the original antigens used to generate the

    monoclonal antibodies. A very different application of bacterial surface display was

    recently suggested by Sousa et al. (1996), who displayed polyhistidyl peptides on the

    surface of E. coli using the LamB system. E. coli cells with surface-exposed histidine tags

    adsorbed approximately 11 times more Cd2 + ions than control cells. Such bacteria could perhaps be used for bioadsorption of heavy metal ions, potentially valuable for environ-

    mental (bioremediation) applications.

    Powerful techniques to monitor and characterize the exposed target proteins being

    displayed are crucial for the full exploitation of any display technology. Traditional

    techniques for studying (recombinant) surface-displayed proteins include immunofluores-

    cence and immunogold methods. Although these methods reveal whether a heterologous

    protein is produced in an accessible form or not, they fail to give reliable quantitative

    information about the number of exposed chimeric surface proteins per bacterial cell.

    Immunofluorescent staining can give misleading results by positive staining of partiallylysed cells and should thus be accompanied by additional assays. Flow cytometry, in the form

    of fluorescence-activated cell sorting (FACS), has been used as the method of choice to

    characterize surface display on bacteria (Fuchs et al., 1991; Stover et al., 1993; Francisco et

    al., 1993b; Nguyen et al., 1995; Samuelson et al., 1995).

    6. Surface display of antibodies

    Phage display is by far the major tool for the isolation and engineering of recombinantantibodies, while bacterial display plays a minor role in that field. Antibodies in the form of

    recombinant antibody fragments were the first proteins to be successfully displayed on the

    surface of phage (McCafferty et al., 1990). This was achieved by fusing the coding sequence

    of the antibody variable (V) regions encoding for a single-chain Fv (scFv) fragment to the

    amino terminus of the phage gene III, coding for the phage minor coat protein pIII. Initial

    attempts to display FabH fragments fused to pVIII, the phage major coat protein, were also

    successful (Gram et al., 1992). However, the pVIII site, although very popular for peptide

    phage display, is not suitable for the efficient display of large polypeptides such as antibodies.

    For this reason, most antibody phage-display systems utilize the pIII site.

    Antibodies were first displayed using a phage vector, based on the genome of fd-tet(Zacher et al., 1980) and its gene III as fusion partner. In this vector, the genes coding for

    antibody scFv fragments were cloned in-frame with gene III and downstream of the gene III

    signal sequence, which normally directs the export of the phage-coat protein to the periplasm.

    Here, the antibody VH and VL domains may fold correctly, both stabilized by an

    I. Benhar / Biotechnology Advances 19 (2001) 133 9

  • 8/6/2019 Phage Display and Biotechnological Applications

    10/33

    intramolecular disulfide bridge, and pair to form a functional scFv (Skerra and Pluckthun,

    1988; Better et al., 1988).Initially, phage vectors that carried all the genetic information required for the phage life

    cycle were used (McCafferty et al., 1990; Clackson et al., 1991). Later, phagemids became

    the popular type of vector for display. Phagemids are small plasmid vectors that carry gene III

    with appropriate cloning sites (for the insertion of cloned antibody genes) and the phage

    intergenic region (for rolling-circle replication and packaging) (Bass et al., 1990; Hoogen-

    boom et al., 1991; Garrard et al., 1991). In phagemids, the scFv may be fused at the N-

    terminus of an intact gene III (McCafferty et al., 1990; Hoogenboom et al., 1991) or at the N-

    terminus of a truncated gene III, lacking the first two N-terminal domains (Bass et al., 1990;

    Barbas et al., 1991). Phagemids have high transformation efficiencies and are therefore

    ideally suited for generating very large repertoires. They may also be formatted for directsecretion of the unfused antibody fragment, without subcloning (Hoogenboom et al., 1991).

    Many phagemids utilize the lac promoter to control the expression of the antibodypIII

    fusion product (Hoogenboom et al., 1991; Barbas et al., 1991). For display of the antibody

    pIII product, glucose, acting in catabolite repression of the lac promoter, is removed or

    depleted, leading to the expression of sufficient fusion product to generate monovalent phage

    particles. Whenever expression-mediated toxicity is an issue [which is the case for some,

    mostly hybridoma-derived, antibody fragments (Thompson et al., 1996)], tighter regulation of

    expression levels may be required. The use of a lac promoter with an additional transcrip-

    tional terminator (Krebber et al., 1996), the tightly regulated tet promoter (Zahn et al., 1999)or of the phage shock promoter (psp) (Rakonjac et al., 1997) may allow display of relatively

    toxic products and reduce expression-mediated library biases.

    The phagemid DNA encoding the antibodypIII fusion are preferentially packaged into

    phage particles using a helper phage such as M13KO7 or VCS-M13, which supplies all

    structural proteins. Since the helper phage genome encodes wild-type pIII, typically over 90%

    of rescued phage display no antibody at all and the vast majority of the rescued phage

    particles that do display the fusion product will only contain a single copy. More efficient,

    multivalent display may be preferable when selecting very large antibody libraries, to

    guarantee selection when a limited number of phage particles per clone are available.Monovalent display, on the other hand, may be essential when selecting antibodies for

    higher affinity. Therefore, the use of inducible promoters (Lutz and Bujard, 1997) or the use

    of a helper phage with gene III deleted (Griffiths et al., 1993), may allow the modulation of

    the valency of displayed antibodies.

    Phage antibody selection involves the sequential enrichment of specific binding phage

    from a large excess of non-binding ones. This, as shown in Fig. 1, is achieved by multiple

    rounds of phage binding to the target, washing to remove nonspecific phage and elution to

    retrieve specific binding phage. Any method that separates phage that bind from those that

    do not, can be used for phage selection, and indeed, many different selection methods have

    been used.The most popular selection methods include affinity selection (also called biopanning) on

    immobilized antigen coated onto solid supports, columns or BIAcore sensor chips (Clackson

    et al., 1991; Marks et al., 1991; Griffiths et al., 1994; Malmborg et al., 1996), selection using

    biotinylated antigen (Hawkins et al., 1992), panning on fixed prokaryotic cells (Bradbury et

    I. Benhar / Biotechnology Advances 19 (2001) 13310

  • 8/6/2019 Phage Display and Biotechnological Applications

    11/33

    al., 1993) and on mammalian cells (Cai and Garen, 1996), subtractive selection using sorting

    procedures (de Kruif et al., 1995), enrichment on tissue sections or pieces of tissue (van Ewijk

    et al., 1997), and, finally, selections using living animals, as reported for peptide phage

    libraries (Pasqualini and Ruoslahti, 1996). Phage antibodies bound to antigen may be eluted

    in different ways: with (one step or gradients of) acidic solutions such as HCl or glycine

    buffers, with basic solutions like triethylamine, with chaotropic agents, with DTT when biotin

    is linked to antigen by a disulfide bridge (Marks et al., 1991), by enzymatic cleavage of aprotease site engineered between the antibody and gene III, or by competition with excess

    antigen (Marks et al., 1991; Bradbury et al., 1993; Griffiths et al., 1994; de Kruif et al., 1995).

    Antibody phage display has been the leading tool in antibody engineering during the last

    decade. Initially applied to the isolation of antibodies against various target molecules, the

    Fig. 1. The phage affinity-selection (biopanning cycle). Recombinant DNA techniques are used to generate a

    library consisting of millions of different antibodies, or of variants (mutants) of an existing antibody. The resulting

    phage library is subjected to several cycles of affinity selection including capture of phage with antigen, washing

    to remove unbound phage, elution to release antigen-bound phage, and amplification in E. coli. When adequate

    enrichment has been obtained, individual antigen-specific clones are isolated and characterized.

    I. Benhar / Biotechnology Advances 19 (2001) 133 11

  • 8/6/2019 Phage Display and Biotechnological Applications

    12/33

    technology has evolved into an efficient tool for the in vitro manipulation of antibody affinity,

    specificity and stability. The most successful application of antibody phage display has beenthe isolation of antibodies from large libraries. For that purpose, three types of libraries were

    used. First, immune libraries that were generated by cloning antibody genes from B cells

    (usually from spleens) of immunized donors. This approach takes advantage of the fact that

    antibodies directed against the immunogen are enriched in such donors, and that they have

    been subjected to in vivo affinity maturation by the host immune system (Clackson et al.,

    1991; Burton et al., 1991). A further advantage of immune libraries is the ability to isolate

    many antibodies that are immunogen-specific from a single library. Immune libraries have

    been produced from a variety of species, including mouse (Clackson et al., 1991; Chester et

    al., 1994; Kettleborough et al., 1994; Berdichevsky et al., 1999), human (Barbas et al., 1993;

    Cai and Garen, 1995), rabbit (Lang et al., 1996; Li et al., 2000a), camelids (Arbabi Ghahroudiet al., 1997; Muyldermans and Lauwereys, 1999), sheep (Li et al., 2000b), and chicken

    (Davies et al., 1995; Yamanaka et al., 1996).

    Immunization, however, is not always feasible, due to ethical considerations, possible

    toxicity or lack of immunogenicity of the antigen. Consequently, non-immune libraries were

    developed. Non-immune libraries are further divided into naive and synthetic antibody

    libraries. Naive libraries were constructed using the same methodology that was applied in the

    construction of immune libraries. The difference was that here B cells from non-immunized

    donors were used as sources for antibody genes. Several naive libraries have been

    constructed, and because they are not biased towards any specific antigen, were used toisolate antibodies against multiple antigens (Marks et al., 1991; Gram et al., 1992; Vaughan et

    al., 1996). With naive libraries, the affinities of isolated antibodies correlated with the library

    size. Hence it was necessary to construct very large libraries to obtain high-affinity antibodies

    (Vaughan et al., 1996; de Haard et al., 1999).

    Synthetic or semisynthetic libraries were the next generation of non-immune libraries.

    These were constructed artificially by assembling in vitro antibody V genes with synthetic D

    and J segments. The DJ segments contained information for coding CDR3 loops of random

    sequence and varying length. The resulting libraries were sometimes referred to as `single-

    pot' libraries, because each such library was a source of antibodies directed against multipleantigens (Hoogenboom and Winter, 1992; Garrard and Henner, 1993; Nissim et al., 1994; de

    Kruif et al., 1995).

    Antibodies selected from phage libraries are not always suitable for direct application as

    therapeutics or biotechnological reagents. In some cases, manipulation of the antibody

    affinity, valency, specificity or stability was required. For such cases, phage display was

    applied, in a manner similar to the process of producing synthetic libraries and selecting the

    best binders from them. Such secondary libraries contained variants of the initially isolated

    antibodies, where mutations were introduced either randomly, or following rational design.

    Mutations were introduced into the antibody genes using one of several methods, including

    site-directed mutagenesis (Schier et al., 1996; Chowdhury and Pastan, 1999) error-prone PCR(Hawkins et al., 1992), Chain shuffling (Clackson et al., 1991; Marks et al., 1992), DNA

    shuffling (Crameri et al., 1996), and mutator E. coli strains (Low et al., 1996; Irving et al.,

    1996). Using such approaches it was possible to obtain antibodies having extremely high

    affinities, well below 100 pM (Schier et al., 1996; Yang et al., 1995). Advanced phage-

    I. Benhar / Biotechnology Advances 19 (2001) 13312

  • 8/6/2019 Phage Display and Biotechnological Applications

    13/33

    selection schemes in combination with secondary antibody libraries were useful in the

    isolation of antibodies having improved stability and folding properties (Spada et al., 1998;Forrer et al., 1999; Jung et al., 1999). Finally, phage display enabled the isolation of catalytic

    antibodies, where catalysis or binding to transition-state analogs were used as the selective

    criteria (Sastry et al., 1995; Janda et al., 1994; Baca et al., 1997; Fujii et al., 1998).

    The expression of functional antibodies on the surface of E. coli (Fuchs et al., 1991; Little

    et al., 1993; Francisco et al., 1993b; Maurer et al., 1997) and staphylococci (Gunneriusson et

    al., 1996, 1999) has led to discussion whether this strategy would be a way to create

    inexpensive diagnostic tools or alternatives to the rapidly developing phage technology for

    the selection of peptides or recombinant antibody fragments from large libraries (Little et al.,

    1993; Georgiou et al., 1997). Most studies of bacterial display of antibodies involved the

    Lpp-OmpAH system developed by Georgiou and coworkers (Francisco and Georgiou, 1994;Francisco et al., 1993b; Georgiou et al., 1997; Chen et al., 1996) (schematically drawn in Fig.

    2). The Lpp-OmpAH system was applied for the development of a quantitative immunoassay

    that utilizes E. coli bacteria expressing scFv antibody fragments attached to the cell surface

    (Chen et al., 1996). Scatchard analysis demonstrated that the antibodies on the surface of the

    cells retained full binding activity and that about 60 000 scFv molecules were displayed per

    cell. The bacteria were used as the antibody reagent in that assay, and, following incubation

    with analyte simple centrifugation could separate the antibody-bound from unbound analyte.

    The Lpp-OmpAH system was also applied for antibody affinity maturation by bacterial

    surface display of scFv libraries utilizing FACS (Daugherty et al., 1998). This system wasemployed to isolate clones with high affinity to a fluorescently labeled hapten from libraries

    Fig. 2. Schematic representation of the Lpp-OmpAH display system. Surface display on E. coli is facilitated by the

    Lpp signal peptide (residues 19) followed by residues 46159 of OmpA. The OmpA fragment is composed of

    five membrane-spanning b-strands, with the C-terminus located outside the cell. A protein linked to the C-

    terminus of the construct is thus located at the cell surface.

    I. Benhar / Biotechnology Advances 19 (2001) 133 13

  • 8/6/2019 Phage Display and Biotechnological Applications

    14/33

    constructed by randomizing heavy and light-chain residues in the anti-digoxin 26-10 derived

    scFv-antibody. The use of flow cytometry enabled the detection of rare library membersdirectly in heterogeneous populations and the optimization of selection conditions prior to

    sorting. A similar approach was used for in vitro scanning saturation mutagenesis of the anti-

    digoxin 26-10 antibody's binding site that allows the high throughput production and

    characterization of antibody mutants with improved binding affinity (Chen et al., 1999),

    and to investigate the effects of mutation frequency on the affinity maturation of the displayed

    antibody (Daugherty et al., 2000).

    A possible advantage of the bacterial-display systems over the phage-display techniques

    lies in the fact that bacterial selection can be accomplished through FACS technology, using

    fluorescently labelled antigens. This avoids crucial steps in phage-display selection proce-

    dures such as immobilization of the antigen, elution of bound phage and reinfection ofbacteria with eluted phage. Filamentous phages are too small to be compatible with current

    FACS technology (Francisco and Georgiou, 1994). It has, however, been shown that E. coli

    cells expressing functional cell-surface-anchored antibody fragments can he separated by

    FACS (Francisco and Georgiou, 1994; Francisco et al., 1993b; Fuchs et al., 1996; Chen et al.,

    1999; Daugherty et al., 2000). Moreover, by manipulating the concentration of the fluorescent

    ligand it is possible to obtain higher and lower affinity antibodies in a single experiment based

    on signal intensity. Using this approach it was possible to analyze quantitatively the effect of

    mutation frequency (inserted by error-prone PCR) on the frequency of clones that remain

    functional and on the likelihood of isolating gain-of-function mutants of a high-affinitysingle-chain antibodies (Daugherty et al., 2000). The results demonstrated that large protein

    libraries can be surface displayed on bacteria, that scFv fragments can be functionally

    expressed on bacterial cells, and that such cells can be efficiently enriched by FACS

    technology. This approach should promote future use of bacterial display as an alternative

    or at least a complement to the phage systems.

    7. Surface display of enzymes, substrates and inhibitors

    Both phage and cell display have contributed to the field of enzyme technology. A number

    of enzymes have been displayed on phage (McCafferly et al., 1991; Corey et al., 1993;

    Soumillion et al., 1994; Maruyama et al., 1994; Mikawa et al., 1996; Mullaney and Black,

    1996; Pedersen et al., 1998; Demartis et al., 1999; Forrer et al., 1999; Hufton et al., 1999;

    Legendre et al., 2000) and bacteria (Francisco et al., 1992; Schneewind et al., 1992, 1993;

    Strauss and Gotz, 1996; Jung et al., 1998a, b). Initially attempts were made to apply phage

    display for the isolation of `improved' enzymes, or for enzymes with altered substrate

    specificity by selecting for improved binding to transition-state analogs. This approach

    resulted in modest improvements for catalytic antibodies displayed on phage (Fujii et al.,

    1998; Baca et al., 1997; Hansson et al., 1997; Arkin and Wells, 1998). However, binding oftransition-state analogs does not correlate well with improved catalysis (Fujii et al., 1998;

    Arkin and Wells, 1998). Specialized selection schemes using reactive substrates (Janda et al.,

    1994), inhibitors (Light and Lerner, 1995), active-site ligands (Widersten and Mannervik,

    1995), and reactive products (Janda et al., 1997) have also been used in attempts to select for

    I. Benhar / Biotechnology Advances 19 (2001) 13314

  • 8/6/2019 Phage Display and Biotechnological Applications

    15/33

    improved catalysts. The selection from libraries of phage-displayed enzymes based on

    catalytic activity of the mutants is more complex than selection based on affinity for aligand, which is how phage display is routinely applied.

    Several strategies have been proposed to alleviate this problem (Soumillion et al., 1994;

    Hansson et al., 1997; Atwell and Wells, 1999; Demartis et al., 1999; Jestin et al., 1999). These

    novel approaches are based on linking the enzymatic reaction product to the surface of the

    phage that displays the active enzyme. It is thus possible to identify the reaction products (and

    the phage they are linked to) with an appropriate antibody. Recently, a technique based on

    phage display for the directed evolution of enzymes of industrial interest was proposed. This

    technique is based on the use of relevant suicide inhibitors specifically designed to

    irreversibly bind the active pocket of the enzyme following their cleavage (Janda et al.,

    1997). By applying phage display for enzyme evolution, it should also be possible to focusselection on particular properties besides catalysis. The robustness of the phage themselves

    (their tolerance to a wide pH range and buffer conditions) makes it is possible to improve the

    stability and the catalytic properties of the displayed enzymes by altering library construction

    and selection conditions.

    In addition to the display of enzymes, phage were used to display enzyme substrates and

    enzyme inhibitors. The term `substrate phage' was coined by Matthews and Wells (1993) to

    describe a prototype approach for identifying novel substrates for enzymes (mostly proteases).

    A library of fusion proteins was constructed containing an amino-terminal domain used to bind

    to an affinity support, followed by a randomized protease substrate sequence and the carboxyl-terminal domain of M13 gene III. Each fusion protein was displayed as a single copy on

    filamentous phagemid particles (substrate phage). Phage were then bound to an affinity support

    and treated with the protease of interest. Phage that displayed good protease substrates were

    released, whereas phage with substrates that resisted proteolysis remained bound. This

    approach was valuable in identifying novel substrates or optimizing existing ones (Matthews

    and Wells, 1993; O'Boyle et al., 1997; Ke et al., 1997a,b; Ohkubo et al., 1999).

    A direct approach similar to that used for the isolation of antibodies from phage libraries

    was undertaken for the isolation of enzyme inhibitors. In these studies, engineered variants,

    based on small proteins derived from natural inhibitors (Roberts et al., 1992; Pannekoek etal., 1993; van Meijer et al., 1996; Huang et al., 1998) or peptide libraries (Ploug et al., 1998),

    were displayed on phage. Inhibitors were isolated by selecting phage on immobilized

    enzymes. Potent inhibitors of trypsin (Markland et al., 1991; Roberts et al., 1992), tPA

    (Pannekoek et al., 1993; van Meijer et al., 1996; Ploug et al., 1998), beta-lactamase (Huang et

    al., 1998), serine proteases (Jespers et al., 1995), human plasma kallikrein and human

    thrombin (Markland et al., 1996) were readily isolated using this approach. Enzyme inhibitors

    that are based on novel protein scaffolds were isolated by Sollazzo and coworkers (Martin et

    al., 1997; Dimasi et al., 1997). They isolated inhibitors of hepatitis C virus NS3 protease from

    an antibody `camelized VH' library (Martin et al., 1997) and from a `Minibodi' library

    (Dimasi et al., 1997).While phage display was used primarily for the molecular evolution of enzymes,

    substrates and inhibitors, with cell display the emphasis was different. Certain enzymes

    have been expressed with retained activity on the surface of gram-negative and gram-

    positive bacteria (Table 2), and the potential use of such recombinant bacteria as novel

    I. Benhar / Biotechnology Advances 19 (2001) 133 15

  • 8/6/2019 Phage Display and Biotechnological Applications

    16/33

    microbial biocatalysts has been investigated. E. coli b-lactamase, which is normally a

    periplasmic enzyme, has been successfully exposed on the outer surface of E. coli using thepullulanase system (Kornacker and Pugsley, 1990) or the Lpp-OmpAH system (Francisco et

    al., 1992) as has been the Cex exoglucanase of Cellulomonas fimi (Francisco et al., 1993a).

    Bacteria displaying heterologous enzymes on their surface hold great potential as whole-

    cell biocatalysts. For industrial applications, such surface-engineered cells need to be killed

    and chemically fixed to prevent disintegration and leakage of the displayed proteins under

    process conditions. It is also highly desirable to couple the chemically stabilized cells onto

    a solid support matrix for additional mechanical stability, flexibility in reactor choice, and

    easy separation from processed medium. Using such an approach, a method designed to

    prepare an immobilized whole cell biocatalyst from beta-lactamase-displaying E. coli cells

    was recently described (Freeman et al., 1999). The method is based on crosslinking carriedout in the presence of beta-lactamase inhibitors, that protects the active site from chemical

    modification resulting in up to threefold higher specific activities without affecting the cell-

    stabilizing effect of the glutaraldehyde treatment.

    However, not all enzymes can be efficiently exposed on the bacterial surface. Alkaline

    phosphatase, which is normally periplasmic, was found to be retained in the periplasm of E.

    coli when expressed via the Lpp-OmpAH system (Stathopoulos et al., 1996). The efficacy of

    using E. coli with surface-displayed heterologous enzymes as novel microbial biocatalysts

    remains to be seen. Surface expression is undoubtedly an inexpensive way to produce

    immobilized enzymes, but gram-negative bacteria might suffer from practical drawbacks suchas inhibition of growth or cell lysis. The E. coli display systems seem to be limited in the size

    of proteins they are capable of exporting to the surface (Tommassen et al., 1985; Stathopoulos

    et al., 1996). Recently, the ice-nucleation protein (Inp), a glycosyl phosphatidylinositol-

    anchored outer membrane protein found in some gram-negative bacteria was used for enzyme

    display in E. coli. By using Pseudomonas syringae Inp as an anchoring motif, the functional

    display ofZymomonas mobilis levansucrase (LevU) on the cell surface was obtained (Jung et

    al., 1998a,b). The cells expressing Inp-LevU were found to retain both the ice-nucleation and

    whole-cell levansucrase enzyme activities. Viability of the cells was also maintained over 48

    h in the stationary phase. When the levansucrase-displayed cells were used as the enzymesource, levan was efficiently synthesized from sucrose with 34% conversion yield, generating

    glucose as a by-product.

    Regarding enzyme display on gram-positive bacteria, a lipase from Staphylococcus hyicus

    and E. coli b-lactamase were expressed on the outer cell surface ofS. carnosus with retained

    activity (Strauss and Gotz, 1996). Approximately 10 000 enzyme molecules were found to be

    present on each cell and it was suggested that the rigid structure of gram-positive bacteria

    would make them particularly appropriate as microbial catalysts.

    More recently, bacterial surface display was applied for enzyme evolution. The E. coli

    surface-display system based on Inp (Jung et al., 1998a,b) was used to selectively screen for

    improved variants of carboxymethyl cellulase (CMCase) (Kim et al., 2000). A library ofmutated CMCase genes generated by DNA shuffling was fused to the ice nucleation protein

    gene so that the resulting fusion proteins were displayed on the bacterial cell surface. Some

    cells displaying mutant proteins grew more rapidly on carboxymethyl cellulose plates than

    controls, forming heterogeneous colonies. In contrast, cells displaying the parent CMCase

    I. Benhar / Biotechnology Advances 19 (2001) 13316

  • 8/6/2019 Phage Display and Biotechnological Applications

    17/33

    formed uniform tiny colonies. These variations in growth rate were assumed to result from

    altered availability of glucose caused by differences in the activity of variant CMCases at thecell surface. Staining assays indicate that large, rapidly growing colonies exhibited increased

    CMCase activity. Increased CMCase activity was confirmed by assaying the specific

    activities of cell extracts after the expression of unfused forms of the variant genes in the

    cytoplasm. The best-evolved CMCases showed about a 5- and 2.2-fold increase in activity in

    the fused and free forms, respectively. Sequencing of evolved CMCase variant genes showed

    that most amino acid substitutions occurred within the catalytic domain of the enzyme. These

    results demonstrate that the bacterial surface display of enzyme libraries may provide a direct

    way to correlate evolved enzyme activity with cell growth rates. This technique may provide

    a useful technology platform for directed evolution and high-throughput screening of

    industrial enzymes, including hydrolases.

    8. Vaccine development

    Both phage- and cell-display systems have been used for vaccine development, but with

    differing approaches. Phage display has been used mostly for the identification of immuno-

    genic epitopes or mimotopes on displayed peptides. These, in turn could become lead

    molecules for the design of peptide-based vaccines. In only a handful of cases have peptide-

    displaying phage themselves been used for immunization.A number of groups used monoclonal of polyclonal antibodies to select peptide libraries.

    At the dawn of the phage-display era, George Smith, who pioneered the phage-display

    technology (Smith, 1985), proposed that foreign sequences could be inserted within the minor

    coat protein, pIII, of filamentous phage. This will result in the creation of a fusion protein that

    is incorporated into the virion; the virions display the foreign amino acids encoded in the

    insert on their surface.

    Phage, bearing specific antigenic determinants from a target gene, can be purified in

    infectious form from a vast excess of phage that carry other determinants, by virtue of their

    affinity to the antibody directed against the gene product. Fusion phage can be used as asource of antigen, as a carrierhapten conjugate for obtaining immunological reagents in

    rabbits, and for B epitope mapping. By generating a fusion-phage library expressing virtually

    all possible short, amino acid sequences, it may be possible to study epitopes on immuno-

    logically important proteins without the use of synthetic peptides and without ever having

    cloned the genes (Parmley and Smith, 1989). Indeed, de la Cruz et al. (1988) showed that

    cloning of repeat regions of the circumsporozoite protein gene of Plasmodium falciparum

    into the pIII gene of a filamentous phage results in the display of the recombinant proteins on

    the phage surface. The displaying phage were both antigenic and immunogenic when injected

    to rabbits. Shortly thereafter phage display in the form of combinatorial peptide libraries was

    used to map linear epitopes of monoclonal antibodies (Scott and Smith, 1990; Cwirla et al.,1990; Devlin et al., 1990; Felici et al., 1991, reviewed in Lane and Stephen, 1993; Daniels

    and Lane, 1996).

    The potential of phage display for identification of neutralizing viral epitopes was

    demonstrated by Perham (di Marzo Veronese et al., 1994) who inserted peptide sequences

    I. Benhar / Biotechnology Advances 19 (2001) 133 17

  • 8/6/2019 Phage Display and Biotechnological Applications

    18/33

    from the V3 loop of gp120 from HIV-1 into the N-terminal region of the major coat protein

    (pVIII) of filamentous bacteriophage fd. This approach led to their display in multiple copieson the surface of the virion. Peptides displayed in this way were shown to be remarkably

    effective structural mimics of the natural epitope. They were recognized by human HIV

    antisera and evoked high titres of antibodies in mice, which cross-reacted with other strains of

    HIV and were capable of neutralizing the virus.

    Soon it was realized, however, that peptide sequences selected from phage library do not

    always correspond to the exact (linear) sequence of the original epitope. It was postulated that

    such peptides mimic the structural features of the epitope in being recognized by epitope-

    specific antibodies, and by eliciting upon injection into animals, antibodies capable of

    recognizing the original epitope. Such peptide mimics were thus named `mimotypes' (Cortese

    et al., 1994; Folgori et al., 1994; Mennuni et al., 1996; Demangel et al., 1996). Mimotypeswere later identified for various nonprotein molecules such as the O-antigen part of the

    human pathogen Shigella flexneri serotype 5a lypopolysaccharide (Phalipon et al., 1997), the

    Neisseria meningitidis group B polysaccharide (Moe et al., 1999) and meningococcal group

    C polysaccharide (Grothaus et al., 2000).

    While most of the studies were carried out using filamentous phage peptide libraries, other

    phage systems were later developed for similar purposes. A 36-amino-acid PorA peptide from

    N. meningitidis was cloned into the display vectors to generate fusions at the N terminus of

    Hoc or Soc capsid genes of phage T4 (Jiang et al., 1997). The PorA-Hoc and PorA-Soc

    fusion proteins retained the ability to bind to the capsid surface, and the bound peptide wasdisplayed in an accessible form as shown by its reactivity with specific monoclonal

    antibodies. Both the PorA-Hoc and PorA-Soc recombinant phages were highly immunogenic

    in mice and elicited strong antipeptide antibody titers.

    Once it was established that phage display can be useful for epitope identification, the use

    of displaying phage as immunogens was considered. A number of groups compared the

    magnitude of antibody response that was elicited by various immunization protocols. Tested

    criteria included display valency (monovalent vs. polyvalent display), the presence and nature

    of added adjuvants, and the route of administration.

    Willis et al. (1993) displayed the antigenic determinants of the circumsporozoite proteinof the malaria parasite, P. falciparum in multiple copies on pVIII of Fd phage. The peptide

    epitopes in the hybrid phage were found to be strongly immunogenic in four different

    strains of mice without the use of external adjuvants, and the antibodies were highly

    specific to the individual epitopes. When tested in nude mice, the immune response was

    found to be T-cell dependent and to undergo class-switching from IgM to IgG. Proliferation

    assays of T cells, taken from lymph nodes of BALB/c mice injected with bacteriophage

    particles in the presence or absence of Freund's complete adjuvant, indicated no difference

    in the immune response.

    Meola et al. (1995) tested if mimotopes could be useful in developing a vaccine against the

    human hepatitis B virus. They compared the humoral immune response of animals immunizedeither with a recombinant HbsAg vaccine or with mimotopes. Immunogens were prepared by

    fusing the mimotypes on different carrier molecules (phage coat protein pIII and pVIII,

    recombinant human H ferritin, HBV core peptide) and by synthesizing multiple antigenic

    peptides carrying the mimotypes' amino acid sequences. These immunogens were injected

    I. Benhar / Biotechnology Advances 19 (2001) 13318

  • 8/6/2019 Phage Display and Biotechnological Applications

    19/33

    into mice and rabbits and sera were collected and tested for the presence of HBsAg-specific

    antibodies. The results confirmed that mimotopes could induce a humoral immune responseresembling that induced by the original antigen, and HBsAg mimotopes displayed on phage

    prove to be the best immunogens, inducing the most repro- ducible and potent immunization.

    Delmastro et al. (1997) showed that intranasal or intragastric administration of phage into

    mice induces an immunological response both to the wild type phage proteins and to

    mimotopes displayed on them. Using mimotopes of human hepatitis B virus surface antigen

    and of human hepatitis C virus peptides, they showed that the response induced by oral

    administration is specifically cross-reactive with the original antigen.

    Bastien et al. (1997) investigated whether a recombinant bacteriophage displaying a

    disease-specific protective epitope could be experimentally used as a vaccine to confer

    protection of immunized animals against infection. They genetically engineered a recombi-nant phage, fd, displaying at its surface a chimeric pIII coat protein (monovalent display)

    fused to the previously identified protective epitope 173187 from the glycoprotein G of the

    human respiratory syncytial virus. A selected recombinant fd phage elicited a strong immune

    response in mice, inducing a high level of circulating RSV-specific antibodies. Mice

    immunized with the recombinant phage acquired a complete resistance to RSV infection as

    evidenced by the lack of detectable virus particles in their lungs following intranasal

    challenge with live RSV. This was the first report of the ability of a phage presenting an

    immunogenic peptide to prevent infection of immunized animals by a pathogen.

    Recently, Zuercher et al. (2000) tested the effect of exposure of displaying phage to gastricfluid. This approach was examined, since an essential requirement for oral vaccines is the

    ability to survive the harsh environment of the stomach in an antigenically intact form. The

    feasibility of this approach was tested in a simulated gastric fluid using two different

    mimotopes as well as an anti-idiotypic Fab of the non-anaphylactogenic monoclonal anti-IgE

    antibody BSW17. All phage clones remained infective after this treatment. However, only

    epitopes displayed on the pVIII protein (polyvalent display) were still recognized by BSW17

    whereas pIII-expressed (monovalent display) epitopes were rapidly inactivated. Surprisingly,

    when used for oral immunization of mice all phage clones induced anti-IgE antibodies. In

    contrast, oral immunization with the purified, pVIII protein that displayed the mimotope alsoinduced anti-phage antibodies, but no anti-IgE antibodies were obtained. After feeding a

    single dose of mimotope-displaying bacteriophage, phage DNA could be detected in mouse

    feces for 10 days. These results have shown that epitope-displaying bacteriophage can be

    used to induce an epitope-specific antibody response via the oral route.

    In conclusion, phage that display epitopes or mimotopes may emerge as useful tools for the

    development of effective vaccines or may serve as vaccine-delivery vehicles themselves. For

    the application of phage as vaccines it seems that polyvalent rather than monovalent display

    are more likely to elicit an efficient immune response, that oral administration is the more

    effective route for immunization, and that adjuvant is usually not required. Such immuniza-

    tion protocols can be less expensive and more effective than the conventional method ofsynthesis and coupling peptide ligands to particulate synthetic carriers for immunization.

    The most common application of bacterial surface display has been in the development of

    live bacterial vaccine delivery systems. In contrast to phage display, which was a tool for

    epitope discovery and for vaccine delivery, displaying immunogenic peptides or proteins on

    I. Benhar / Biotechnology Advances 19 (2001) 133 19

  • 8/6/2019 Phage Display and Biotechnological Applications

    20/33

    bacteria for the purpose of vaccine development followed a different scheme. Here, an

    already identified vaccine candidate (an immunogenic peptide or protein) was displayed on bacteria to be used as carriers for vaccine delivery. Gram-positive cells were earlier

    considered for vaccine delivery. Nguyen et al. (1993) developed a system for the surface

    expression of heterologous receptors on the cell surface ofS. xylosus. That expression system

    was constructed by subcloning gene fragments encoding for peptides to be displayed on the

    cell surface into an E. coli staphylococci shuttle vector. The vector was designed for

    targeting of recombinant fusion proteins to the outer cell surface of the gram-positive host

    cell. A specific humoral immune response could be elicited in mice by oral immunization

    with the recombinant S. xylosus cells, suggesting that this type of gram-positive bacteria

    might offer potential vehicles for oral vaccination. Liljeqvist et al. (1999) assayed the surface

    expression in S. carnosus of three different fibronectin binding domains (FNBDs), derivedfrom fibronectin binding proteins of Streptococcus dysgalactiae and S. aureus. All three

    surface-displayed FNBDs were demonstrated to bind fibronectin in whole-cell enzyme-linked

    binding assays. Furthermore, for one of the constructs, intranasal immunizations with the

    recombinant bacteria resulted in improved antibody responses to a model immunogen present

    within the chimeric surface proteins. To avoid the use of engineered pathogens for vaccine

    delivery, Fischetti et al. (1996) developed systems that allow the expression of heterologous

    antigens in commensal gram-positive bacteria. In some cases, both a serum IgG and secretory

    IgA response were induced to the recombinant protein after vaccination, verifying the validity

    of the approach. These live recombinant bacteria may be used in the future to introduce aprotective immune response to pathogenic microorganisms after mucosal colonization.

    Many of the described gram-negative display systems have initially been evaluated in E.

    coli for surface display of various antigenic determinants, and subsequently applied to

    Salmonella spp. These species are of interest for the live, oral delivery of heterologous

    antigens for immunization. An early study demonstrated that bacteria that express recombi-

    nant proteins (although not at the surface in this particular example) could be used as live oral

    vaccines. In this study, Poirier et al. (1988) investigated the use of attenuated strains of

    Salmonella to deliver cloned antiphagocytic virulence determinants of unrelated bacteria.

    Thus, the aroA strain of S. typhimurium SL3261 was transformed with a low-copy plasmidthat contained the cloned gene spm5, encoding streptococcal M protein (the major virulence

    factor of these organisms). The transformed Salmonella expressed type-5 M protein in the

    cytoplasmic fraction, and when fed orally to BALB/c mice, evoked both serum and salivary

    IgA, IgG, and IgM antibodies directed against type-5 M protein. The orally immunized mice

    were completely protected against both intranasal and intraperitoneal challenge infections

    with virulent S. typhimurium SL1344 or M5 streptococci. In a second study, mice were

    immunized with live Salmonella dublin, expressing a flagellum with an inserted cholera

    toxin, hepatitis B virus or influenza virus haemaglutinin epitopes. The immunized mice

    produced antigen-specific antibodies and in some cases partial protection from virus

    challenge (Wu et al., 1989; Newton et al., 1989). It was suggested that epitopes exposedat the bacterial cell surface induce antibody responses in a T-cell-independent manner and that

    the lipopolysaccharides may serve as adjuvants. These studies provided early evidence that an

    attenuated strain ofSalmonella can be used effectively as a general vaccine vehicle to deliver

    antiphagocytic virulence determinants of unrelated bacteria.

    I. Benhar / Biotechnology Advances 19 (2001) 13320

  • 8/6/2019 Phage Display and Biotechnological Applications

    21/33

    Schorr et al. (1991) used E. coli OmpA as a carrier for the expression of foreign antigens on

    the surface of gram-negative bacteria. Immunogenic moieties of P. falciparum blood stageantigens were expressed in an attenuated S. typhimurium strain. Live bacteria expressing the

    fusion proteins reacted with malarial antigen sera. Mice that were immunized orally with S.

    typhimurium cells expressing malarial antigens on their surface showed a humoral immune

    response. Leclerc et al. (1991) targeted two foreign B cell antigenic viral epitopes to different

    locations in the E. coli cell to examine what effect this had on antibody responses elicited by

    the recombinant bacteria. The two epitopes were peptides derived from the PreS2 region of

    hepatitis B virus and the C3 neutralization epitope of poliovirus type 1. They were each

    expressed in two forms either on the surface, as part of the outer-membrane protein LamB, or

    soluble in the periplasm, as part of the periplasmic protein MalE. When live bacteria

    expressing the foreign epitopes at the cell surface were used for immunization of mice, theyinduced T-cell-independent antibody responses characterized by a rapid induction of IgM and

    IgG antibodies. In contrast, when the same foreign epitopes were inserted into the MalE

    protein, the antibody response was only detectable after 3 weeks, belonged only to the IgG

    class and was strictly T-cell dependent. This study has therefore identified two major pathways

    by which epitopes expressed by bacterial cells can stimulate specific antibody responses. The

    first pathway is mediated by direct activation of B cells by bacterial cell-surface antigen and

    does not require T-cell help. The second pathway is T-cell dependent and concerns antigen that

    can be released from the bacteria in a soluble form. Stentebjerg-Olesen et al. (1997) assessed

    the potential of the major structural protein of type 1 fimbriae as a display system forheterologous sequences. As a reporter-epitope, a heterologous sequence mimicking a

    neutralizing epitope of the cholera toxin B chain was inserted into the fimA gene. The chimeric

    proteins were exposed on the bacterial surface and the cholera toxin epitope was authentically

    displayed, i.e. it was recognized on bacteria by specific antiserum. Immunization of rabbits

    with purified chimeric fimbriae resulted in serum, which specifically recognized cholera toxin

    B chain, confirming the utility of the employed strategy. It can be concluded that several of the

    developed bacterial surface-display systems, of both gram-positive and gram-negative bacteria

    merit additional future consideration for application as live or attenuated vaccines.

    9. Phage and cell display come together

    As described in the previous sections, phage display has been used mostly for affinity

    screening of combinatorial peptide and protein libraries, to define epitopes for monoclonal

    antibodies, to select enzyme substrates and to screen cloned antibody repertoires. Phage

    libraries are enriched for specific binding clones by subjecting the phage to repetitive rounds

    of affinity selection (biopanning) that include binding, washing and elution steps, reinfection

    of bacteria and growth to re-express the binding molecule on the phage surface (Fig. 1).

    Several selection schemes may be applied to isolate antigen-specific phage from aheterogeneous population composed mostly of nonbinders. They include capture with

    immobilized antigen, capture with soluble ligand followed by isolation of phage ligand

    complexes, selection on antigen columns, selection on living or fixed cells and selection in

    whole animals (Bradbury et al., 1993; Hoogenboom, 1997; Hoogenboom et al., 1998; Persic

    I. Benhar / Biotechnology Advances 19 (2001) 133 21

  • 8/6/2019 Phage Display and Biotechnological Applications

    22/33

    et al., 1999; Pasqualini and Ruoslahti, 1996; Poul and Marks, 1999; Becerril et al., 1999).

    With most selection approaches, successive panning results in the diminution of diversitywith a few dominant phage clones (usually those having the highest affinity or display

    efficiency) being recovered. It is therefore desirable to isolate binders following fewer,

    preferably a single panning cycle. This, in principle, could be achieved by linking antigen

    recognition with phage infectivity, as has been demonstrated in SIP (selectively infective

    phage) and similar selection strategies (Malmborg et al., 1997; Krebber et al., 1997).

    However, even with those powerful selection strategies, enrichment factors rarely exceed

    104 per cycle.

    Recently we developed a novel selection scheme that combines bacterial surface display and

    phage display of polypeptides for the highly efficient isolation of binders, where antigen-

    displaying E. coli cells are used to capture antibody-displaying phage. Bradbury et al. (1993)used a similar approach to select phage on living or fixed antigen-displaying F bacteria. In

    that case, an epitope derived from the oncoprotein p21ras was expressed within theE. coli outer

    membrane protein, LamB. A library of single-chain Fvs was selected on the displaying bacteria.

    Bound phage were eluted at a high pH and used to infect F + bacteria for phage expansion. In the

    reported example, an enrichment factor of 13 000 was obtained in a single cycle.

    Our recently described system (Benhar et al., 2000) was named DIP (delayed infectivity

    panning). In this approach, the enrichment factor exceeded 106 per cycle. DIP takes

    advantage of the versatile Lpp-OmpAH bacterial surface-display system (Georgiou et al.,

    1997) where protein-coding sequences are fused to an Lpp-OmpAH

    hybrid sequence. Inaddition to the successful display of enzymes, antibodies, and cellulose-binding proteins that

    were described in the earlier sections, we could display receptor domains and a bacterial toxin

    fragment by this system.

    DIP combines bacterial display with conventional phage-display libraries (Fig. 3) where

    the selecting bacterial cells are also being infected by the captured phage. This is made

    possible by the fact that F + E. coli cells grown at low temperatures do not express the F pilus,

    which is required for phage infection (Novotny and Fives-Taylor, 1974; Novotny and Lavin,

    1971). After phage capture and washing procedures, the bacteria are transferred to 37C so

    pilus expression is induced and infection ensues. Some antigens (in particular cell-surfacereceptors) are notoriously difficult to produce as recombinant proteins. In DIP, however, a

    purified antigen is not required for phage selection and identification. Using two model

    antibodyantigen pairs, we demonstrated enrichment of over 106-fold in a single DIP cycle

    for phage, which display a specific antibody, over those displaying an irrelevant antibody. We

    further showed the successful isolation of anti-toxin and anti-receptor antibodies from an

    immune phage library, a shuffled library and a large synthetic human library.

    DIP selection is far more efficient than panning with immobilized antigen or compared

    to other systems where antigen-displaying cells are used to capture antibody-displaying

    phage. Several factors may contribute to this observation. First, the selection `matrix' is

    bacteria that are essentially identical to those used to propagate the phage (except that thelatter do not display the antigen). This may be regarded as a `counter-selection' step, where

    most of the antibody-displaying phage that recognize E. coli cell-surface epitopes adhere to

    the cells that were used to produce them and are thus eliminated. Second, we may have

    been fortunate in selecting the Lpp-OmpAH bacterial display format as the platform for DIP,

    I. Benhar / Biotechnology Advances 19 (2001) 13322

  • 8/6/2019 Phage Display and Biotechnological Applications

    23/33

    in that the antigen density and accessibility to antibodies may be better than antigens

    adsorbed to solid surfaces. Third, the antigen concentration used in DIP is relatively low:

    assuming 104105 displayed molecules per bacterium (Francisco and Georgiou, 1994;

    Francisco et al., 1993b; Georgiou et al., 1997; Daugherty et al., 1998), with 1010 bacteria

    per milliliter we are in the nanomolar concentration range. Thus, our selection is more

    stringent than using micromolar concentrations of antigen, which is the general practice

    with immobilized antigens. The effectiveness of DIP makes it suitable for the efficient

    isolation of rare clones present in large libraries. Since a multitude of phage- and cell-basedsystems for surface display of polypeptides exists, DIP should be suitable for the isolation

    and characterisation of various proteinprotein interacting pairs.

    References

    Abedi MR, Caponigro G, Kamb A. Green fluorescent protein as a scaffold for intracellular presentation of

    peptides. Nucleic Acids Res 1998;26:62330.

    Agterberg M, Adriaanse H, van Bruggen A, Karperien M, Tommassen J. Outer-membrane PhoE protein

    of Escherichia coli K-12 as an exposure vector: possibilities and limitations. Gene 1990;88:37 45.Agterberg M, Adriaanse H, Lankhof H, Meloenn R, Tommassen J. Outer membrane PhoE protein of

    Escherichia coli as a carrier for foreign antigenic determinants: immunogenicity of epitopes of foot-and-

    mouth disease virus. Vaccine 1990;8:85 91.

    Arbabi Ghahroudi M, Desmyter A, Wyns L, Hamers R, Muyldermans S. Selection and identification of single

    domain antibody fragments from camel heavy-chain antibodies. FEBS Lett 1997;414:5216.

    Fig. 3. Scheme of the DIP selection method. (A) Map of plasmid pIB-Tx used for bacterial display of antigens. (B)

    Scheme of DIP selection: (1) Antigen displaying E. coli grown at 16C, when the F pilus is not expressed, are

    mixed with antibody displaying phage. Antigen-specific phages bind to the cells. (2) Unbound phages are washed

    out. (3) Cells are shifted to 37C, so F pilus expression is induced, bound phage dissociate from antigen and infect

    the cell through the F pilus. (4) Infected cells where bacterial display plasmid and phage genome co-reside.

    I. Benhar / Biotechnology Advances 19 (2001) 133 23

  • 8/6/2019 Phage Display and Biotechnological Applications

    24/33

    Arkin MR, Wells JA. Probing the importance of second sphere residues in an esterolytic antibody by phage

    display. J Mol Biol 1998;284:108394.

    Atwell S, Wells JA. Selection for improved subtiligases by phage display. Proc Natl Acad Sci USA 1999;

    96:9497502.

    Baca M, Scanlan TS, Stephenson RC, Wells JA. Phage display of a catalytic antibody to optimize affinity for

    transition-state analog binding. Proc Natl Acad Sci USA 1997;94:100638.

    Barbas CF, Kang AS, Lerner RA, Benkovic SJ. Assembly of combinatorial antibody libraries on phage surfaces:

    the gene III site. Proc Natl Acad Sci USA 1991;88:797882.

    Barbas CF 3rd, Collet TA, Amberg W, Roben P, Binley JM, Hoekstra D, Cababa D, Jones TM, Williamson RA,

    Pilkington GR, Haigwood NL, Cabezas E, Satterthwait AC, Sanz I, Burton DR. Molecular profile of an

    antibody response to HIV-1 as probed by combinatorial libraries. J Mol Biol 1993;230:81223.

    Bass S, Greene R, Wells JA. Hormone phage: an enrichment method for variant proteins with altered binding

    properties. Proteins 1990;8:309 14.

    Bastien N, Trudel M, Simard C. Protective immune responses induced by the immunization of mice with arecombinant bacteriophage displaying an epitope of the human respiratory syncytial virus. Virology 1997;

    234:11822.

    Becerril B, Poul MA, Marks JD. Toward selection of internalizing antibodies from phage libraries. Biochem

    Biophys Res Commun 1999;255:38693.

    Benhar I, Nahary L, Shaky S, Azriel R, Tamarkin A, Wels W. DIP selection of phage-antibodies mediated by

    antigen expression on live bacteria. J Mol Biol 2000;301:893904.

    Berdichevsky Y, Ben-Zeev E, Lamed R, Benhar I. Phage display of a cellulose binding domain from Clos-

    tridium thermocellum and its application as a tool for antibody engineering. J Immunol Methods 1999;

    228:15162.

    Better M, Chang CP, Robinson RR, Horwitz AH. Escherichia coli secretion of an active chimeric antibody

    fragment. Science 1988;240:10413.Bingle WH, Nomellini JF, Smit J. Cell-surface display of a Pseudomonas aeruginosa strain K pilin peptide within

    the paracrystalline S-layer of Caulobacter crescentus. Mol Microbiol 1997;26:27788.

    Bradbury A, Persic L, Werge T, Cattaneo A. Use of living columns to select specific phage antibodies. Biotech-

    nology 1993;11:1565 9.

    Buchli PJ, Wu Z, Ciardelli TL. The functional display of interleukin-2 on filamentous phage. Arch Biochem

    Biophys 1997;339:79 84.

    Burton DR, Barbas CF. Human antibodies from combinatorial libraries. Adv Immunol 1994;57:191280.

    Burton D, Barbas CF, Persson M, Koenig S, Chanock R, Lerner RA. A large array of human monoclonal

    antibodies to type 1 human immunodeficiency virus from combinatorial libraries of asymptomatic seropositive

    individuals. Proc Natl Acad Sci USA 1991;88:101347.

    Cabilly S. The basic structure of filamentous phage and its use in the display of combinatorial peptide libraries.Mol Biotechnol 1999;12:1438.

    Cai X, Garen A. Anti-melanoma antibodies from melanoma patients immunized with genetically modified auto-

    logous tumor cells: selection of specific antibodies from single-chain Fv fusion phage libraries. Proc Natl Acad

    Sci USA 1995;92:653741.

    Cai X, Garen A. A melanoma-specific VH antibody cloned from a fusion phage library of a vaccinated melanoma

    patient. Proc Natl Acad Sci USA 1996;93:62805.

    Chang H, Lo SJ. Modification with a phosphorylation tag of PKA in the TraT-based display vector of Escherichia

    coli. J Biotechnol 2000;78:11522.

    Chang HJ, Sheu SY, Lo SJ. Expression of foreign antigens on the surface ofEscherichia coli by fusion to the outer

    membrane protein traT. J Biomed Sci 1999;6:6470.

    Charbit A, Molla A, Saurin W, Hofnung M. Versatility of a vector for expressing foreign polypeptides at thesurface of gram-negative bacteria. Gene 1988;70:1819.

    Chen G, Cloud J, Georgiou G, Iverson BL. A quantitative immunoassay utilizing Escherichia coli cells possessing

    surface-expressed single chain Fv molecules. Biotechnol Prog 1996;12:5724.

    Chen G, Dubrawsky I, Mendez P, Georgiou G, Iverson BL. In vitro scanning saturation mutagenesis of all the

    specificity determining residues in an antibody binding site. Protein Eng 1999;12:34956.

    I. Benhar / Biotechnology Advances 19 (2001) 13324

  • 8/6/2019 Phage Display and Biotechnological Applications

    25/33

    Chester KA, Begent RH, Robson L, Keep P, Pedley RB, Boden JA, Boxer G, Green A, Winter G, Cochet O,

    Hawkins RE. Phage libraries for generation of clinically useful antibodies. Lancet 1994;343:4556.

    Choo Y, Klug A. Toward a code for the interactions of zinc fingers with DNA: selection of randomized fingers

    displayed on phage. Proc Natl Acad Sci USA 1994;91:111637.

    Choo Y, Castellanos A, Garcia-Hernandez B, Sanchez-Garcia I, Klug A. Promoter-specific activation of gene

    expression directed by bacteriophage-selected zinc fingers. J Mol Biol 1997;273:52532.

    Chowdhury PS, Pastan I. Improving antibody affinity by mimicking somatic hypermutation in vitro. Nat Bio-

    technol 1999;17:56872.

    Clackson T, Hoogenboom HR, Griffiths AD, Winter G. Making antibody fragments using phage display libraries.

    Nature 1991;352:6248.

    Corey DR, Shiau AK, Yang Q, Janowski BA, Craik CS. Trypsin display on the surface of bacteriophage. Gene

    1993;128:12934.

    Cortese R, Felici F, Galfre G, Luzzago A, Monaci P, Nicosia A. Epitope discovery using peptide libraries

    displayed on phage. Trends Biotechnol 1994;12:2627.Cortese R, Monaci P, Luzzago A, Santini C, Bartoli F, Cortese I, Fortugno P, Galfre G, Nicosia A, Felici F.

    Selection of biologically active peptides by phage display of random peptide libraries. Curr Opin Biotechnol

    1996;7:61621.

    Crameri A, Cwirla S, Stemmer WP. Construction and evolution of antib