home > ebr > autumn 2010 > the evolutions of molecular display
European Biopharmaceutical Review

The Evolutions of Molecular Display

Chris Ullman of Isogenica Ltd gives an insight into molecular display methods and how they are rejuvenating protein and peptide discovery, enabling the selection of novel molecules with greater therapeutic potential

Twenty-five years have passed since the publication of George Smith’s seminal paper in Science describing the display of a foreign peptide on the coat of filamentous phage. Expression and specific recovery of the engineered phage using a ‘panning’ process was termed ‘phage display’ and gave rise to a more general methodology of molecular display. Since then, phage display has become an essential component of the protein engineer’s toolkit where diverse pools (or ‘libraries’) of peptides and proteins containing hundreds of millions of mutations can be rapidly created and the best candidates selected; a process akin to Darwinian evolution or, more precisely, ‘molecular evolution’. The versatility and flexibility of phage display has been a major factor in its adoption in many laboratories. Thousands of papers have been published describing its application, and its contribution to the study of ligandreceptor and protein-DNA interactions cannot be underestimated. However, phage display has its limitations, particularly in terms of speed, library size, expressed protein size and the intractability of some targets. Realisation of these limits has spawned a number of new methods that use the same principles but exploit the cellular machinery in a cell-free environment. These technologies interrogate greater sequence space leading to the rapid discovery of even better peptides and proteins. This article briefly reviews the impact of phage display and assesses the performance of the next generation of display technologies.

Phage display was invented approximately one year before the first high-throughput combinatorial chemistry approach was described. Both techniques synthesise chemical compounds as ensembles, or libraries, which are screened for desirable properties. However, phage display harnesses the power of a natural system to synthesise peptides and proteins. Cycles of selection and enrichment, prior to screening, allow the interrogation of massive libraries (approximately 1010 members) which dwarf the capacity of high-throughput screening (HTS) techniques (see Figure 1); even at a rate of 100,000 compounds per day, it would take HTS hundreds of years to screen similarly sized libraries (1,2). The key to the effective functioning of phage display, and all molecular display technologies, is the formation of a robust link between the encoding nucleic acid (genotype) and the expressed protein (phenotype), such that when the expressed protein is selected, its sequence can simply be translated from its accompanying nucleic acid code.


Phages are viruses that infect bacterial cells. The rod-shaped filamentous phages are composed of a proteinaceous tube containing a single stranded DNA genome. The coat proteins each have been adapted to display foreign peptides, thus demonstrating the versatility and the adaptability of the virus, but the most common fusions are with the abundant P8 protein or P3 (a minor coat protein expressed at the tip of the phage and essential for infection). The peptides that are displayed as fusions with phage coat proteins are panned for improved or altered characteristics. The panning procedure applies specific pressure that drives the selection of the fittest candidates in the gene pool towards a desired outcome.

In practice, the selection pressure is usually a target molecule which is immobilised or captured onto a surface. The protein library is exposed to the target and a subpopulation of the proteins binds. Diversity is important as most of the phages are lost from the system under stringent washing conditions, leaving bound peptides and their associated phage particles to be enriched by infection and re-growth in bacteria (see Figure 2). Those that are best adapted for survival will eventually dominate the selection environment. Panning conditions are important for the selection of the best clones. Conditions ideally need to be balanced so that stringency is sufficiently high enough to apply pressure for the selection of optimal peptides without diminishing the yield. Not all foreign proteins are amenable to display, and the expression of larger proteins has been less successful due to folding and secretion problems (1,3-5).


Short peptides of random sequence are the simplest combinatorial phage libraries (1). These have successfully produced hits for diverse applications. Peptides that have been selected by phage display to agonise or antagonise protein interactions are now reaching the market or the latter stages of development (see Table 1). Technological advances include the selection of tissuespecific peptides by in vivo panning experiments (biopanning) in a human patient and similar biopanning experiments have also shown success for peptide translocation across skin in rats (6,7).

Phage display has revolutionised the ability to engineer larger protein biologics, in particular antibodies. However, as whole antibodies could not be functionally expressed in bacteria, segments had to be amplified from variable region genes and then engineered into single-chain variable fragment antibody (scFv) libraries. It was not until 1990 that these fragment libraries were first displayed on the surface of phage fused to P3. Further work led to the cloning of antigen specific ‘immune’ libraries from immunised mice and humans, ‘naïve’ libraries from healthy individuals and ‘synthetic’ antibody libraries constructed from synthetic oligonucleotides. These advances enabled phage display to compete with hybridoma technology and offered credible alternatives to immunisation. The speed at which hits with novel specificity could be generated, the lack of bias introduced by immune tolerance, and the ability to mimic the natural processes of somatic hypermutation and V-gene rearrangement in human framework sequences in vitro, all provided key advantages over hybridomas. To date, the most successful antibody libraries have been those based upon the natural variable domains of the antibody in which rational optimisation of the framework regions and diversification of the complementarity-determining regions (CDRs) has been achieved.

Humira (adalimumab), a TNFα inhibitor for the treatment of rheumatoid arthritis, was the first antibody derived from phage display to be approved and there is a growing pipeline (see Table 1) (3,8). Closely related techniques are yeast and bacterial display which enable the display of scFvs or scFvs and Fabs, respectively, upon the surface of cells (4).

Phage display has contributed greatly to the understanding and application of protein-DNA interactions, and in particular zinc fingers. The zinc finger is the most abundant DNA-binding domain, small in size and possessing a compact structure that coordinates a zinc ion. Their small structure, tolerance to engineering and modular nature has enabled multiple domains to be expressed on phage and a code for DNA recognition to be determined. These results have opened the door for possible site-specific control and correction of deleterious genes in humans (9,10).

However, despite success, phage display is limited by a laborious and inefficient process known as transformation, which is necessary for phage production in E. coli. Often, it takes months to construct libraries of 1010 members.


Acellular in vitro molecular display technologies, which avoid the need for transformation, enable more of the sequence landscape to be displayed, ultimately leading to greater diversity and an increased probability of higher affinity hits. These systems use the transcription and translation machinery extracted from prokaryotic or eukaryotic cells, thereby enabling theoretical library sizes up to 1014 to be investigated. They also do not rely on a secretory pathway for display, can select for a wider variety of antimicrobial peptides and have less non-specific interactions. The distinguishing features of these technologies are shown in Figure 3.

The first cell-free, in vitro molecular display technique – ribosome display – was originally developed in an E. coli lysate for the display of peptides. The ribosomes were stalled on the mRNA template and the nascent peptide remained in a complex, which could then be disrupted by EDTA. The released RNA was subsequently amplified by an RT-PCR step. Later, scFvs were successfully displayed on ribosomes using both bacterial and eukaryotic systems (8,11,12). However, the system is very sensitive to RNase degradation and mechanisms have been employed to reduce this effect (13). Ribosome display methods have also been used in the selection of peptides containing non-natural amino acids, and non-antibody structures such as ankyrin domains (14-16).

A related technique, mRNA (or in vitro virus) display differentiates itself from ribosome display by the formation of a covalent link between the template and the expressed protein via puromycin. Puromycin is carried on a DNA primer appended to the mRNA template and mimics amino-acyl tRNA, binding covalently to the nascent peptide as a result of the peptidyl transferase activity of the ribosome (17). The DNA primer is then used in a reverse transcription step to stabilise the RNA template in a RNA/ DNA hybrid. Coupling efficiency between protein and mRNA has been reported to be 10 to 40 per cent, but seven panning rounds were typically needed in the selections (18). The mRNA display technique has been used for a number of different applications including the incorporation of non-natural amino acids (19).

DNA based systems have advantages of speed and stability over RNA templates as the DNA template is less sensitive to degradation, therefore libraries can be generated quickly by standard PCR procedures. One system, CIS display, harnesses the ability of a DNA-binding protein, RepA, that exclusively binds back to its encoding DNA (termed cis-activity). Coupling of the protein and DNA is non-covalent and is approximately 40 per cent efficient, supporting effective library sizes over 1013, with enrichment factors between 103 and 105 fold per round. CIS display was the first published example of recovery of a specific binder from a 1 in 1010 dilution, therefore demonstrating potential for unprecedented library sizes (20). This technique has been used to select in the presence of proteases (21), and to select from scaffold and scFv libraries (unpublished data). Another system using a cis-acting DNA binding protein that links covalently to its template has been used for scFvs to select for tetanus toxin binders from an immune human library with enrichment between 14- and 300-fold (22).

In vitro compartmentalisation (IVC) provides an alternative way of linking phenotype and genotype and mimics the natural compartments of living organisms, entrapping DNA and ITT components in water-in-oil emulsions by directly mixing mineral containing surfactants in water; each droplet having approximately one gene. The method is particularly advantageous for the direct selection of enzymes; however, efficiencies may be reduced by the incomplete separation of DNA molecules into individual droplets or by fusion of the compartments. IVC has been further adapted for use with microfluidics (8,23-25).


With over 25 years of development behind phage display methods, it’s utility in the discovery of novel pharmaceutical biomolecules and in the understanding of protein function cannot be underestimated. It has been the pathfinder of molecular display, leading to a new generation of selection technologies. These new technologies are becoming more commonplace and the improvements that they offer are gradually being realised. Amongst these improvements, rapid library design and selection and interrogation of larger library diversity have rejuvenated peptide and protein discovery. The ease in which new protein designs can be validated and optimised allows greater exploration of the structural space which, in turn, will facilitate a greater understanding of the molecular interactions of the proteome and enable the selection of novel molecules with greater therapeutic potential.


  1. Smith GP and Petrenko VA, Phage Display, Chem Rev 97: pp391-410, 1997
  2. Houghten RA, Dooley CT and Appel JR, In vitro and direct in vivo testing of mixture-based combinatorial libraries for the identification of highly active and specific opiate ligands, AAPS J 8: ppE371-382, 2006
  3. Hoogenboom HR, Selecting and screening recombinant antibody libraries, Nat Biotechnol 23: pp1,105-1,116, 2005
  4. Levin AM and Weiss GA, Optimizing the affinity and specificity of proteins with molecular display, Mol Biosyst 2: pp49-57, 2006
  5. Fernandez-Gacio A, Uguen M and Fastrez J, Phage display as a tool for the directed evolution of enzymes, Trends Biotechnol 21: pp408-414, 2003
  6. Arap W, Kolonin MG, Trepel M, Lahdenranta J, Cardo-Vila M et al, Steps toward mapping the human vasculature by phage display, Nat Med 8: pp121-127, 2002
  7. Chen Y, Shen Y, Guo X, Zhang C, Yang W et al, Transdermal protein delivery by a coadministered peptide identified via phage display, Nat Biotechnol 24: pp455-460, 2006
  8. Rothe A, Hosse RJ and Power BE, In vitro display technologies reveal novel biopharmaceutics, FASEB J 20: pp1,599-1,610, 2006
  9. Isalan M and Choo Y, Engineering nucleic acid-binding proteins by phage display, Methods Mol Biol 148: pp417-429, 2001
  10. Urnov FD, Miller JC, Lee YL, Beausejour CM, Rock JM et al, Highly efficient endogenous human gene correction using designed zinc-finger nucleases, Nature 435: pp646-651, 2005
  11. He M and Taussig MJ, Ribosome display: cell-free protein display technology, Brief Funct Genomic Proteomic 1: pp204-212, 2002
  12. Zahnd C, Amstutz P and Pluckthun A, Ribosome display: selecting and evolving proteins in vitro that specifically bind to a target, Nat Meth 4: pp269-279, 2007
  13. Villemagne D, Jackson R and Douthwaite JA, Highly efficient ribosome display selection by use of purified components for in vitro translation, J Immunol Methods 313: pp140-148, 2006
  14. Nakajima E, Goto Y, Sako Y, Murakami H and Suga H, Ribosomal synthesis of peptides with C-terminal lactams, thiolactones, and alkylamides, Chembiochem 10: pp1,186-1,192, 2009
  15. Bonetto S, Spadola L, Buchanan AG, Jermutus L and Lund J, Identification of cyclic peptides able to mimic the functional epitope of IgG1-Fc for human Fc gammaRI, FASEB J 23: pp575-585, 2009
  16. Amstutz P, Koch H, Binz HK, Deuber SA and Pluckthun A, Rapid selection of specific MAP kinasebinders from designed ankyrin repeat protein libraries, Protein Eng Des Sel 19: pp219-229, 2006
  17. Hosse RJ, Rothe A and Power BE, A new generation of protein display scaffolds for molecular recognition, Protein Sci 15: pp14-27, 2006
  18. Takahashi TT, Austin RJ and Roberts RW, mRNA display: ligand discovery, interaction analysis and beyond, Trends in Biochemical Sciences 28: pp159-165, 2003
  19. Millward SW, Fiacco S, Austin RJ and Roberts RW, Design of cyclic peptides that bind protein surfaces with antibody-like affinity, ACS Chem Biol 2: pp625-634, 2007
  20. Odegrip R, Coomber D, Eldridge B, Hederer R, Kuhlman PA et al, CIS display: In vitro selection of peptides from libraries of protein-DNA complexes, Proc Natl Acad Sci USA 101: pp2,806-2,810, 2004
  21. Eldridge B, Cooley RN, Odegrip R, McGregor DP, Fitzgerald KJ et al, An in vitro selection strategy for conferring protease resistance to ligand binding peptides, Protein Eng Des Sel 22: pp691-698, 2009
  22. Reiersen H, Lobersli I, Loset GA, Hvattum E, Simonsen B et al, Covalent antibody display: an in vitro antibody-DNA library selection system, Nucleic Acids Res 33: pE10, 2005
  23. Tawfik DS and Griffiths AD, Manmade cell-like compartments for molecular evolution, Nat Biotechnol 16: pp652-656, 1998
  24. Sepp A, Tawfik DS and Griffiths AD, Microbead display by in vitro compartmentalisation: selection for binding using flow cytometry, FEBS Lett 532: pp455-458, 2002
  25. Griffiths AD and Tawfik DS, Directed evolution of an extremely fast phosphotriesterase by in vitro compartmentalization, EMBO J 22: pp24-35, 2003


Read full article from PDF >>

Rate this article You must be a member of the site to make a vote.  
Average rating:

There are no comments in regards to this article.

Chris Ullman is Chief Scientific Officer at Isogenica Ltd, a small biotech company that is using its CIS display technology to engineer peptides and proteins. Chris has been at Isogenica since 2002, where he has been involved in the development of CIS display. Prior to joining Isogenica, Chris was Team Leader for Gendaq Ltd (later acquired by Sangamo Biosciences Inc), developing customised transcription factors. The company was spun-out from Sir Aaron Klug’s Laboratory at the LMB, Cambridge.
Chris Ullman
Print this page
Send to a friend
Privacy statement
News and Press Releases

ACG appoints Shankar Gupta as Chief Sales Officer

ACG, the world’s largest integrated pharmaceutical and nutraceutical manufacturing solutions provider, is delighted to announce that Shankar Gupta will be re-joining the business in the newly created role of chief sales officer.
More info >>

White Papers

Quality By Design

Biopharma Group

Quality by Design (QbD) is a systematic approach to development that begins with the defined objectives and emphasizes science and quality risk management in order to achieve those goals. QbD ensures that the quality of the product is built into production processes from the outset, rather than being tested after development has already commenced.
More info >>




©2000-2011 Samedan Ltd.
Add to favourites

Print this page

Send to a friend
Privacy statement