The Ties That Bind: Peptide Display Technology

Date: March 15, 1999 Phage Display Systems and Vectors Structure of the T7 phage particle. The negative-stained pattern from polyheads showing capsid hexamer and pentamer units has been fitted onto the surface of the icosahedral particle. A single monomer of the capsid protein is shaded in red. Figure provided by Novagen. Back in the early '50s, at a time when Elvis Presley was beginning his undisputed reign as the king of rock 'n' roll, bacteriophage were rearing their ugly heads (so to spe

Debra Swanson
Mar 14, 1999

Date: March 15, 1999 Phage Display Systems and Vectors
Structure of the T7 phage particle. The negative-stained pattern from polyheads showing capsid hexamer and pentamer units has been fitted onto the surface of the icosahedral particle. A single monomer of the capsid protein is shaded in red. Figure provided by Novagen.
Back in the early '50s, at a time when Elvis Presley was beginning his undisputed reign as the king of rock 'n' roll, bacteriophage were rearing their ugly heads (so to speak) to take center stage as the rising stars of molecular biology. Isolation of these parasitic bacterial viruses prompted much interest in the study of their astonishing growth cycles, but early experiments provided little insight as to whether the nucleic acid or a protein component of the viral coat carried the genetic specificity in progeny production. In 1952, Alfred Hershey and Martha Chase, while working with 32P-labeled bacteriophage T2 in Cold Spring Harbor, N.Y., showed that only the nucleic acid of phage entered the host bacteria and that the protein coat material was left behind. It was the birth of the blues--for that bacteria!

While to some the term "cloning" may conjure up images of hordes of identical creatures, peptide display cloning, which combines cloning and combinatorial chemistry, in contrast generates diversity--and not just a little. Peptide display technology, as it has evolved over the past decade, can create vast libraries with upwards of 1011 different peptides. Such libraries have been used to screen for novel agonists or antagonists for a variety of receptors, and to define the exact polypeptide target of protein-binding compounds, protein and not.

Phage display technology had its beginnings in 1985 when George Smith of the University of Missouri at Columbia cloned a restriction enzyme digest of plasmid DNA into the gene III insertion site of filamentous phage f1 and created a fusion protein with the foreign sequence in the middle, which was displayed in immunologically accessible form on the virion surface (hence the term "phage display").1 Smith showed that after transfection in E. coli, the "fusion phage" clones produced contained the insert, which encoded Eco RI endonuclease, and could be affinity purified from a library of random inserts using antibody to Eco RI endonuclease.

Cwirla et al.2 further expanded phage display technology by creating a large and diverse oligonucleotide library using inserts with randomly synthesized residues representing many variations in the tip of the coat protein. In this study, the phage isolated from the infected E. coli were "panned," or affinity purified, using a monoclonal antibody against L-enkephalin, collected, and used to infect E. coli, followed by two more rounds of panning. The pooled phage population collected during each panning procedure was sequenced, as were DNA samples of 51 individual clones from the third round of panning. The results indicated that the panning method was highly specific as shown by the relatedness of the sequences obtained: nearly all the 51 phage sequenced displayed Tyr-Gly on the N terminus of the variable peptide, in agreement with the known specificity of the monoclonal. And even more important, this screening approach did not require previous knowledge of the structure of the peptide (ligand) or its antibody specificity. Researchers have since capitalized on these findings by using both degenerate and rearranged oligonucleotides to synthesize phage display libraries for creating entities such as high-affinity hormones or antibodies.

One of the advantages of peptide display illustrated by the Cwirla study is the linkage between the peptide and the genetic information for it provided by the phage particle itself. While attempting to determine the sequence of hundreds of peptides is an onerous, if not impossible task, it is entirely feasible to sequence multiple (nay, hundreds) of phage inserts in order to home in on a consensus target structure

Phage display offers a powerful and general method to change and refine the properties of a protein or peptide that can be displayed on the phage surface. The capability of this technique to correlate protein structure and function in an interactive fashion makes possible new methods of finding novel drugs with greater therapeutic potential. Without knowing anything about a particular binding site, it is possible with peptide display technology to determine the combination of amino acids that gives the best fit to the binding site of the receptor, and following that, to select, or even design a small peptide that binds as well as a natural ligand. This approach has been used to create molecular mimics for the two clinically significant polypeptide hormones, EPO and TPO, for example. These mimetics, as they are called, have been used to determine the three-dimensional conformation of the binding site, and further have been shown to bind and even activate the appropriate receptor.3,4

Phage display can also be an attractive alternative to the traditional methods of antibody studies, such as hybridoma technology and animal immunization, enabling the building of antibodies in bacteria by mimicking characteristics of immune selection. The screening process can be performed on any number of solid phases--plastic ELISA plates or culture dishes, glass beads, agarose beads, and magnetic particles. While initial pannings tend to produce regions with low binding affinities, by going through several rounds of selection it is possible to select for peptides with increasingly higher affinity.

Libraries of peptides on phage can also be used to map antibody epitopes, as a much larger number of potential epitopes can be screened than is practical with methods based on chemical generation of candidate peptides.

The displayPHAGE System is a versatile, high-throughput, phage-based method. It provides a head start in the screening process by using three premade E. coli libraries, a 6-mer, 7-mer, and 8-mer, eliminating the initial amplification required by library construction. The pSKAN library vector can generate 3 x 107 phage, each displaying a different pIII phage coat fusion protein on its surface, so the likelihood of quickly finding affinity attractants is increased from the start. The phage are cultured, pooled, and exposed to a microtiter plate coated with the "bait"--the target of interest, which can be a protein, nucleic acid, lipid, or even intact cells. Unbound phage are removed with repeated washing, after which phage bound to the bait molecule are eluted with an excess of a known ligand for the target, or with a lowering of wash-buffer pH. The eluted pool of high-binding phage is amplified in culture, and the biopanning process is repeated for a total of three or four rounds. The recovered enriched pool of clones is used to infect the E. coli host strain, and individual phagemids expressing ligands for the target are isolated using standard techniques. Individual libraries are then cultured and isolated. Specific positive clones are identified with ELISAs using one of two available murine monoclonal antibodies to components of the pSKAN vector--anti-PSTI IgG and anti-PIII IgG. The insert of the clones of choice can then be sequenced with displayPHAGE sequencing primers for final identification.

The versatility offered by this system is evident in the turnaround time from start to finish--results in less than three weeks. Since it is a well-known but unproven "fact" that one week in lab time correlates to one day in real-life time, the quick processing is indeed impressive.

Expression and display of a FliTrx peptide fusion. Figure provided by Invitrogen.
The FliTrx Random Display Library employs the same basic route of expression--fusion protein display-- as does phage display, but without the phage. The vehicle of choice is the pFliTrx vector, which positions a diverse library (1.77 x 108 primary clones) of random dodecapeptides in a flagella (Fli) thioredoxin (Trx) fusion protein (FliTrx). The recombinant protein is exported and assembled into partially functional E. coli flagella and displayed on the cell surface in a conformationally constrained manner due to insertion of the peptide into the active-site loop of the thioredoxin protein. Peptides inserted within this loop have both their N and C termini tethered by the rigid and stable tertiary fold of the thioredoxin molecule. With this system, phage infection and isolation steps are eliminated. For those researchers with a peptide sequence in hand who wish to construct their own library to study protein-protein interactions, the pFliTrx Peptide Display Vector is available as a solo kit. Also included are the FliTrx Forward and Rsr Reverse Sequencing primers for that final identification of positive clones.

The screening process is the standard panning technique described in phage-based methodology involving immobilization of the target on a solid surface and repeated probing with large numbers of individual peptide clones. But it should be noted that the elution step of the FliTrx procedure is not dependent upon the binding of the peptides but instead involves the physical shearing of the flagella. Therefore, all washing steps need to be performed very gently. The FliTrx Panning Kit is designed specifically to simplify the screening process using the FliTrx Random Display Library or the pFliTrx Peptide Display Vector and provides all the buffers and medium necessary for optimizing the system.

Epitope mapping of an anti-beta-endorphin monoclonal antibody with the Ph.D.-12 library. The Ph.D.-12 library was panned against anti-beta-endorphin antibody. Selected 12-mer sequences from each round are shown aligned with the first 12 residues of beta-endorphin; consensus elements are boxed. The results show that the epitope for this antibody spans the first seven residues of beta-endorphin, and that the bulk of the antibody-antigen binding energy is contributed by the first four residues (YGGF), with some flexibility allowed in the third position. Figure provided by New England BioLabs.
The three Ph.D. Phage Display Peptide Library Kits engineered by New England BioLabs are based on random peptide libraries displayed on bacteriophage M13, a nonlytic entity that is closely related to filamentous phage fd and f1. The Ph.D.-7 and Ph.D.-12 kits employ linear, unconstrained peptide libraries of seven and 12 residues, respectively, while the Ph.D.-C7C kits utilize a seven-residue peptide library in which a pair of cysteine molecules flank the randomized insert sequence. All three libraries express the randomized sequence as a fusion of the pIII coat protein via a Gly-Gly-Gly-Ser flexible linker, resulting in display of five copies of the peptide on the exterior surface of the phage virion with the genetic information located inside the virion.

The specific applications of each library are determined by their individual characteristics. The Ph.D.-7 Peptide Library Kit consists of randomized linear 7-mer peptides in which the first residue of the mature displayed fusion product is the first randomized position. The library contains 2.8 x 109 independent clones, enough to encode most, if not all, of the 207 (1.28 x 109) possible 7-mer sequences. It is the best characterized and most recommended library of the three, useful to identify targets with a binding motif localized in a short stretch of amino acids.

The Ph.D.-C7C Disulfide Constrained Peptide Library Kit contains 3.7 x 109 independent clones and can encode the same number of possible 7-mer sequences as the Ph.D.-7 library, but the displayed peptides are cyclic rather than linear. Under nonreducing conditions, the flanking cysteine residues will spontaneously form a disulfide bond, resulting in each peptide in the library being constrained in a disulfide loop. This constrained structure can be useful for identification of antibody structural epitopes whose native ligands are in the context of a surface loop.

The Ph.D.-12 Peptide Library Kit contains 1.9 x 109 independent clones, only a small fraction of the 2012 (4.1 x 1015) possible 12-mer sequences. This library permits interaction with targets requiring seven or fewer defined peptides that are spaced farther apart than the 7-mer "window" of the Ph.D. library. For example, the motif ASDXXXTXPY has six defined positions but cannot be present in the Ph.D.-7 library. Additional bonuses of the longer randomized segment are (1) the allowance of structural folding of the peptide that may be a necessary component for target binding and (2) selection for sequences with many weak binding contacts instead of few strong contacts. The M13KE vector, used to construct all three Ph.D. libraries, is available as part of the Ph.D. Library Cloning System, which also includes an extension primer for second-strand synthesis of the randomized library insert.

As previously mentioned, M13 is a nonlytic phage; therefore, contamination from E. coli (ER2537) host proteins is not a problem as in many phage purification schemes, so less time is spent isolating the phage between rounds of panning. Selection of the appropriate biopanning strategy will be dependent upon the intended application. The kits contain detailed instructions for a variety of protocols, including surface panning, solution panning, and ELISA and/or DNA characterization of selected clones (sequencing primers are included). NEB has successfully used all three libraries to identify consensus peptide ligands for streptavidin, RNase A, chitin, protein kinases, cell-surface receptors, and monoclonal antibodies.

Novagen's T7Select Phage Display System is uniquely based on bacteriophage T7, a double-stranded nonfilamentous DNA phage that expresses peptides or proteins as fusions to the C terminus of the 10B capsid protein. This feature of the T7 system is especially useful when working with cDNA as source material, since the presence of stop codons within the insert won't interfere with the display as it would if located at the N-terminus of the M13 minor coat protein used in some other peptide display systems. T7 phage assembly happens in the E. coli cytoplasm and mature phage are released by cell lysis. Therefore, unlike filamentous phage assembly that involves extrusion, displayed molecules do not need to be capable of secretion through the cell membrane, providing another measure of flexibility that can be useful with cDNA inserts.

The T7 vector in addition has many attractive features that make it an excellent general cloning vector: it grows easily and replicates more efficiently than either filamentous phage, plaques form within three hours at 37°C, and cultures lyse one to two hours after infection. Actual cloning experiments have routinely produced >108 recombinant plaques per microgram of arms, cloning efficiency 10- to 50-fold greater than that of bacteriophage lambda cloning systems and rivaling that of plasmid systems. The T7 phage is a hardy infective creature, able to withstand conditions that would inactivate lesser phage, thereby expanding the variety of biopanning agents and conditions that can be used in the screening process to include reagents such as 1% SDS, 5 M NaCl, 4 M urea, 2 M guanidine-HCl, 10 mM EDTA, reducing conditions, and a pH range of 4 to 10.

Three T7Select phage display vectors are now available in T7Select Cloning Kits. The T7Select415 vector can display high copy number (415 per phage) peptides of up to 50 amino acids in size on the surface of the capsid, and the phage grows on a female E. coli host, BL21. The strong signal obtained from high copy display is useful for attracting peptides that bind with low affinity to their target or for epitope mapping. The T7Select1 vectors display low copy number (0.1 to 1 per phage) peptides or proteins of up to 1,200 amino acids in size, and the phage grow on a complementing host (BLT5403 or BLT5615). Low copy number display is achieved by reduced expression of the capsid protein by prior removal of the capsid gene promoter and the translation initiation site. The capsid mRNA is still produced from phage promoters farther upstream of the gene, but 10B capsid protein production is minimized and replaced with 10A capsid protein supplied by the host. Low copy display can be useful for the selection of proteins that bind strongly to their targets. For intermediate levels of display (five to 15 copies per phage) Novagen has introduced the T7Select10-3b vector. Its midcopy display level is achieved by mutagenesis of the ribosome binding site upstream of T7 gene 10. T7Select1-1 vectors can display proteins up to 1,200 amino acids while the T7Select1-2 series vectors have a somewhat smaller capacity (900 amino acids) but have multiple cloning sites in all three reading frames.

The T7Select Biopanning Kit is available to screen for T7 phage displaying target ligands using 96-well microtiter plates or similar solid phases. The phage, in lysate, are allowed to bind to a protein coated plate, the phage displaying the protein are eluted and used to infect the host, and the selection cycle is repeated for up to four rounds. Target-labeled phage in the lysate are detected following each round by chemiluminescent visualization of plaque lifts. Control experiments with phage displaying an S-tag sequence have shown nearly a 106-fold enrichment after two rounds of selection, and almost all of the phage contained the desired sequence after four rounds of selection.

Novagen offers complete systems including vector, packaging extracts, and components for five libraries as well as vectors and stand-alone reagents. A kit including the T7 Select vectors and all the reagents required for cDNA synthesis by either oligo-dT priming or directional random-primer cDNA synthesis is also available.

Immunologists should take note of Amersham Pharmacia's Recombinant Phage Antibody System (RPAS)--an integrated modular system designed for cloning recombinant antibody fragments from mice or cell culture and expressing them in bacteria via phage display. The components are available as individual modules and support products that provide a systematic approach for cloning, expressing, detecting, and purifying single-chain fragment variable (ScFv) antibodies. The core system consists of the following three modules: Mouse ScFv, Expression, and Detection Modules.

The RPAS Mouse ScFv Module is designed to PCR-amplify the heavy (VH) and light (VL) genes of the antibody cDNA using mRNA from mouse hybridoma or spleen cells and primers specific for the variable region of each chain. The reaction produces a small amount of the 750 base pair ScFv gene in which the VH chain is joined to the VL chain with a special linker. The fragment is then amplified using a set of restriction site primers that add Sfi I and Not I sites to its 5' and 3' ends, respectively, in preparation for enzyme digestion and ligation into the pCANTAB 5 expression vector in the Expression Module.

The RPAS Expression Module provides for the cloning of the antibody ScFv gene into Sfi I/Not I-digested pCANTAB 5 E phagemid for transformation of competent E. coli TG1 cells. The recombinant phage that are produced with the aid of M13KO7 helper phage contain a single-strand copy of the phagemid DNA encoding a specific antibody ScFv gene, displayed on the phage tips as fully functional antibody ScFv fusion proteins. Selection and enrichment of the desired phage is accomplished by panning against the target antigen either on a solid phase (Expression Module instructions) or in solution using biotin-labeled antigen (optional Recombinant Phage Selection Module instructions). The resultant antigen-positive clone can be used to infect E. coli HB2151 cells to produce soluble antibodies for use as immunological reagents, and even further purified from other E. coli proteins by affinity chromatography using the optional RPAS Purification Module.

The RPAS Detection Module uses HRP/antiM13 Conjugate in an ELISA to detect phage-displayed antibodies bound to antigen in 96-well plates.

Optional modules previously mentioned may be used in conjunction with the core system. The Recombinant Phage Selection Module provides an alternative method for selecting recombinant phage antibodies using biotinylated antigen in a solution-phase system. Protein antigens are labeled with NHS-LC-biotin and incubated with the phage-displayed antibodies in solution to form antigen/antibody complexes that are captured using streptavidin coupled to a crosslinked beaded agarose. This method allows for not only easy manipulation of antigen concentration but, based on the quantity of antigen used, selection of antibodies that exhibit high affinities or kinetics of dissociation. The RPAS Purification Module can be used to affinity purify "E-tagged" soluble ScFv recombinant antibodies, produced specifically by the pCANTAB cloning procedure, that are expressed in E. coli periplasm or secreted into the culture medium. Included in the kit are pre-packed HiTrap Anti-E Tag Sepharose columns and preformulated buffers.

For those who have opted for a phage display system but have limited personnel, time, or equipment, Syncomm Corporation provides the technology as a service, from beginning to end, offered in component form. Phage display is to Syncomm as antibody is to antigen; after all, Syncomm's specialty is the engineering or reconstruction of human recombinant antibodies using phage display systems. This includes humanization of murine monoclonal antibodies with specific or unique binding features, cloning of human antibody heavy and light variable regions into custom-defined patterns, phage display library construction of human recombinant antibodies, or randomly synthesized peptides with or without cyclic loops. The resulting recombinant antibodies and peptides can be used for new drug discovery or as diagnostic reagents.

Following a free consultation, the libraries (either peptide or recombinant antibody libraries) to be panned against targets of interest can be provided either by the client or constructed by Syncomm. The targets can be peptides, proteins, antibodies, DNA fragments, or sugars. Clones are enriched through three rounds of standard biopannings (a fourth round is an available option) and three amplifications. Clones from the final output are either followed by an ELISA assay or DNA sequencing, and clients will receive clones chosen with either known relative affinity or a panel of DNA sequences or both.

The phage display system is a powerful technology that has been, and continues to be used extensively in biotechnology research and development. The basis of its power is the ease it provides in screening enormous numbers of clones--greater than 1011 different displayed sequences. The more, the merrier!

  • G.P. Smith, "Filamentous fusion phage: novel expression vectors that display cloned antigens on the virion surface," Science, 228:1315-7, 1985.

  • S. Cwirla et al., "Peptides on phage: a vast library of peptides for identifying ligands," Proceedings of the National Academy of Sciences, 87:6378-82, 1990.

  • N.C. Wrighton et al., "Small peptides as potent mimetics of the protein hormone erythropoietin," Science, 273:458-63, 1996.

  • O. Livnah et al., "Functional mimicry of a protein hormone by a peptide agonist: the EPO receptor complex at 2.8 Å," Science, 273:464-71, 1996.

    The author can be contacted at

    Phage Display Systems and Vectors