Elsevier

Methods in Enzymology

Volume 503, 2012, Pages 101-134
Methods in Enzymology

Chapter five - Designed Ankyrin Repeat Proteins (DARPins): From Research to Therapy

https://doi.org/10.1016/B978-0-12-396962-0.00005-7Get rights and content

Abstract

Designed ankyrin repeat proteins (DARPins) have been developed into a robust and versatile scaffold for binding proteins. High-affinity binders are routinely selected by ribosome display and phage display. DARPins have entered clinical trials and have found numerous uses in research, due to their high stability and robust folding, allowing many new molecular formats. We summarize the DARPin properties and highlight some biomedical applications. Protocols are given for labeling with dyes and polyethylene glycol, for quantitatively measuring binding to cell surface receptors by kinetics and thermodynamics, and for exploiting new engineering opportunities from using “click chemistry” with nonnatural amino acids.

Introduction

To embark on developing a new protein scaffold class as a general binding module, there has to be a strong motivation, especially when carried out in a laboratory with a long-standing focus on antibody engineering and technology (Glockshuber et al., 1990, Skerra and Plückthun, 1988). As the designed ankyrin repeat protein (DARPin) technology was developed in an academic setting (Binz et al., 2003, Forrer et al., 2003), the driving force was the desire to create a scaffold with superior technological properties.

Through the design of a fully synthetic antibody library (Knappik et al., 2000) and the development of ribosome display (Hanes and Plückthun, 1997, Hanes et al., 1998), which incorporates affinity maturation and directed evolution directly in the workflow, the basic operation of both natural antibody generation and somatic mutation had been replicated in the laboratory. Ironically, the antibody molecule was then no longer needed, and the technology had become independent of its roots. From extensive experiments with the engineering of antibodies and their fragments (Wörn and Plückthun, 2001), it had become evident that the molecules themselves contain some technical limitations.

Today, recombinant antibody scFv or Fab fragments are usually converted back to an IgG for therapeutic applications (Plückthun and Moroney, 2005), and the biophysical properties (or “developability”) of a particular antibody are an important consideration. When using the antibody fragments in more ambitious formats (as fusions to other aggregation-prone proteins, in multimers, or in the absence of disulfide bonds), the limitations in their biophysical properties become even more apparent.

Repeat proteins appeared very attractive as a choice for a general binding protein. Repeat proteins (Kobe and Kajava, 2000) contain modules, whose number can be chosen freely, which stack up on each other to create a rigid protein domain. They have different architectures, but within one family, they use modules of almost identical structure but with individual surfaces to specifically bind their target. After engineering work on the repeat proteins had started (Binz et al., 2003, Forrer et al., 2003, Stumpp et al., 2003), the discovery that jawless vertebrates use an adaptive immune system made of leucine-rich repeat proteins (Pancer et al., 2004) came as a surprising validation of the concept.

Repeat proteins seem to follow rules which can be derived from biophysical considerations: a large interaction surface is usually a prerequisite for tight binding. Such a binding interface should be rigid, in order to not lose entropy upon binding, which would otherwise decrease the overall achievable binding free energy (equivalent to a loss in affinity). Also, the surface should be modular and thus “patches” should be individually exchangeable or modifiable by affinity maturation. Furthermore, the protein should also not require disulfides, to have the option of expressing it in the bacterial cytoplasm, and to later introduce unique cysteines for site-specific coupling with drugs, fluorescent labels, or polyethylene glycol (PEG), just to name a few.

Ankyrin repeat proteins (Bork, 1993, Li et al., 2006) are built from tightly joined repeats of usually 33 amino acid residues. Each repeat forms a structural unit consisting of a β-turn followed by two antiparallel α-helices (Fig. 5.1), and up to 29 consecutive repeats can be found in a single protein (Walker et al., 2000). Yet, ankyrin repeat domains usually consist of four to six repeats, which stack onto each other, leading to a right-handed solenoid structure with a continuous hydrophobic core and a large solvent accessible surface (Kobe and Kajava, 2000, Sedgwick and Smerdon, 1999).

We exploited the huge available sequence information of the repeats within the protein family (different protein variants each with several repeats) by using a “consensus” strategy (Forrer et al., 2004). The underlying assumption is that residues important for maintaining the fold will be more conserved and thus show up prominently in an alignment, while residues involved in interactions of individual members of the protein family with their specific target will not be conserved. The ankyrin repeats seem to belong to one predominant single family, such that consensus design is comparatively straightforward.

The repeat modules are held together by a hydrophobic interface, and thus the first (N-capping repeat or N-cap) and last repeat (C-capping repeat or C-cap) have to be special and need to present a hydrophilic outside surface exposed to the solvent. In the original design, both caps were taken from a natural protein (Binz et al., 2003). More recently, this C-cap has been redesigned, based on molecular dynamics calculations, to make it more similar to the consensus, and it has been experimentally verified that the new C-cap is indeed much more resistant against thermal and denaturant-induced unfolding (Interlandi et al., 2008). Crystallography (Kramer et al., 2010) and NMR (Wetzel et al., 2010) show that this is due to better packing of the C-cap against the internal repeats.

The “full-consensus” DARPins, where all the residues are chosen from consensus considerations, show remarkable properties. They express very well in Escherichia coli as soluble monomers, their stability increases with length, and those with more than three internal repeats are resistant to denaturation by boiling or guanidine hydrochloride. Full denaturation requires high temperature in 5 M guanidine hydrochloride (Wetzel et al., 2008). Hydrogen/deuterium exchange experiments of DARPins with two and three internal repeats indicate that this high stability of the full-consensus ankyrin repeat proteins is due to the strong coupling between repeats. Some amide protons require more than a year to exchange at 37 °C (Wetzel et al., 2010), highlighting the extraordinary stability of the proteins. The location of these very slowly exchanging protons indicates a very stable core structure in the DARPins that combines hydrophobic shielding with favorable electrostatic interactions.

We can consider these full-consensus DARPins as the hypothetical diversification point of a library (even though, historically, the library had been constructed first; Binz et al., 2003). Thus, when diverging from a point of extremely high stability, many changes in the protein, necessary for function but potentially detrimental to stability, can be tolerated, and the outcome is still a very good protein. It appears that the experimental results confirm this hypothesis (see e.g., Binz et al., 2003, Binz et al., 2006, Kohl et al., 2003, Wetzel et al., 2008).

A second reason, besides stability, for basing the DARPin library on a consensus ankyrin, as opposed to a particular naturally occurring ankyrin, was to make the library modules self-compatible (Binz et al., 2003) (Fig. 5.1A and B) such that they can be assembled in any order. Such a designed repeat library module comprises fixed and variable positions. The fixed positions mainly reflect conserved framework positions, while the six variable positions mainly reflect nonconserved surface-exposed residues that can be potentially engaged in interactions with the target, as they are located on the target-binding (concave) face. The theoretical diversities of the DARPin libraries are 5.2 × 1015 or 3.8 × 1023 for two- or tree-module binders, respectively, and the actual sizes of the libraries are equal to the number of different molecules present. We can estimate them as 1012 in ribosome display (Plückthun, 2011) and 1010 in phage display (Steiner et al., 2008).

DARPins have been selected from the synthetic libraries by ribosome display and phage display. Ribosome display is a potent in vitro method to select and evolve proteins or peptides from a naïve library with very high diversity to bind to any chosen target of interest (Hanes and Plückthun, 1997, Hanes et al., 1998, Hanes et al., 2000a, Mattheakis et al., 1994). A key feature of ribosome display, in contradistinction to most other selection technologies, is that it incorporates PCR into the procedure and thus allows a convenient incorporation of a diversification (“randomization”) step. Thereby, ribosome display allows refinement and affinity maturation not only of preexisting binders (Dreier et al., 2011, Hanes et al., 2000b, Luginbühl et al., 2006, Zahnd et al., 2007b) but also of the whole pool during selection from a complex library, if desired. It appears that, as the DARPins fold very well, also in cell free translation, binders are enriched somewhat faster than binders from a comparable scFv ribosome display library (Dreier and Plückthun, 2010, Hanes et al., 2000a).

Using ribosome display, DARPins have been evolved to bind various targets with affinities all the way down to the picomolar range (Amstutz et al., 2005, Binz et al., 2004, Dreier et al., 2011, Huber et al., 2007, Schweizer et al., 2007, Veesler et al., 2009, Zahnd et al., 2006, Zahnd et al., 2007b). The theoretical considerations for designing efficient off-rate selection experiments were recently summarized (Zahnd et al., 2010a).

DARPins have also been selected by phage display. In filamentous phage display using fusions to the minor coat protein g3p, the protein of interest is first produced as a membrane-bound intermediate, with the domains of interest secreted to the E. coli periplasm but remaining still attached to the inner membrane. The g3p fusion is then taken up by the extruding phage. As DARPins fold very fast in the cytoplasm (Wetzel et al., 2008), they fold before they can be transported across the membrane via the posttranslational Sec system, the normal way of secreting E. coli proteins. Thus, very low display rates would be observed using Sec-dependent signal sequences. However, E. coli has another secretion system, the signal recognition particle (SRP) dependent one, which is essentially cotranslational (Bibi, 2011, Fekkes and Driessen, 1999). Using phagemids with SRP-dependent signal sequences, phage display selection of DARPins leads to enrichment just as fast as phage display of slow folding proteins, for example, scFv fragments, using Sec-dependent phage display (Steiner et al., 2006). Binders with subnanomolar KD could be obtained from the phage display library without affinity maturation for a variety of targets (Steiner et al., 2008). For completeness, we also mention that previous attempts to achieve functional display of g3p fusions via the Tat route have proven unsuccessful (Dröge et al., 2006, Nangola et al., 2010, Paschke and Höhne, 2005), as the full-length p3 protein may itself be incompatible with the Tat system. However, a truncated version of p3 can support Tat-mediated phage display (Speck et al., 2011).

Most DARPins express well in E. coli in soluble form constituting up to 30% of total cellular protein (up to 200 mg per 1 l of shake flask culture), and in the fermenter, multigram quantities per liter culture can be achieved (www.molecularpartners.com). Purification is thus straightforward, and for laboratory use, IMAC purification is the standard method used (Binz et al., 2004). Usually, in the initial research, only few milligrams are needed, and thus the extreme overloading of an IMAC column of small capacity leads to very pure protein in a single step, as most E. coli contaminants are thereby competed out.

Additional purification steps are of course required when the protein is derivatized (e.g., with PEG, or fluorescent dyes) (see Section 3). For animal experiments, still higher purity is needed and absence of endotoxins needs to be secured, requiring additional washing steps and chromatography for endotoxin removal (Section 3.13).

Not only the full-consensus DARPin molecules but also most library members showed high thermodynamic stability during unfolding induced by heat or denaturants (Binz et al., 2004, Kohl et al., 2003) and can be brought to very high protein concentrations.

Section snippets

DARPins in diagnostics

DARPins have been tested for their suitability in quantitative immunohistochemistry (Theurillat et al., 2010), which requires high specificity in complex tissue. A DARPin specific for epidermal growth factor receptor 2 (HER2) with picomolar affinity was compared to an FDA-approved rabbit monoclonal antibody in paraffin-embedded tissue sections in tissue microarrays. HER2 gene amplification status is an important criterion to determine the optimal therapy in breast cancer. As an external

Protocols for DARPins in Biomedical Applications

The selection of DARPins from libraries using ribosome display (Dreier and Plückthun, 2011, Zahnd et al., 2007a) or phage display (Steiner et al., 2008) as well as their straightforward biochemical characterization (Binz et al., 2003, Binz et al., 2004) have been described elsewhere. Thus, we will concentrate on methods for coupling DARPins with fluorescent labels and PEG, using both conventional coupling reactions and the introduction of nonnatural amino acids for the use of “click chemistry.”

Acknowledgment

Work in the author's laboratory on establishing the methods was supported by the Swiss National Science Foundation and Swiss Anti-Cancer League (Krebsliga Schweiz; KFS 02448-08-2009).

References (98)

  • E.B. Getz et al.

    A comparison between the sulfhydryl reductants tris(2-carboxyethyl)phosphine and dithiothreitol for use in protein biochemistry

    Anal. Biochem.

    (1999)
  • J. Hanes et al.

    Selecting and evolving functional proteins in vitro by ribosome display

    Methods Enzymol.

    (2000)
  • R.E. Hansen et al.

    Quantification of protein thiols and dithiols in the picomolar range using sodium borohydride and 4,4′-dithiodipyridine

    Anal. Biochem.

    (2007)
  • T. Huber et al.

    In vitro selection and characterization of DARPins and Fab fragments for the co-crystallization of membrane proteins: The Na(+)-citrate symporter CitS as an example

    J. Struct. Biol.

    (2007)
  • G. Interlandi et al.

    Characterization and further stabilization of designed ankyrin repeat proteins by combining molecular dynamics simulations and experiments

    J. Mol. Biol.

    (2008)
  • A. Knappik et al.

    Fully synthetic human combinatorial antibody libraries (HuCAL) based on modular consensus frameworks and CDRs randomized with trinucleotides

    J. Mol. Biol.

    (2000)
  • B. Kobe et al.

    When protein folding is simplified to protein coiling: The continuum of solenoid protein structures

    Trends Biochem. Sci.

    (2000)
  • M.A. Kramer et al.

    Structural determinants for improved stability of designed ankyrin repeat proteins with a redesigned C-capping module

    J. Mol. Biol.

    (2010)
  • B. Luginbühl et al.

    Directed evolution of an anti-prion protein scFv fragment to an affinity of 1 pM and its structural interpretation

    J. Mol. Biol.

    (2006)
  • R.C. Münch et al.

    DARPins: An efficient targeting domain for lentiviral vectors

    Mol. Ther.

    (2011)
  • S. Nangola et al.

    Appraisal of translocation pathways for displaying ankyrin repeat protein on phage particles

    Protein Expr. Purif.

    (2010)
  • S.A. Nicklin et al.

    The influence of adenovirus fiber structure and function on vector development for gene therapy

    Mol. Ther.

    (2005)
  • M. Paschke et al.

    A twin-arginine translocation (Tat)-mediated phage display system

    Gene

    (2005)
  • P. Reichelt et al.

    Single step protocol to purify recombinant proteins with low endotoxin contents

    Protein Expr. Purif.

    (2006)
  • A. Schweizer et al.

    Inhibition of caspase-2 by a designed ankyrin repeat protein: Specificity, structure, and inhibition mechanism

    Structure

    (2007)
  • S.G. Sedgwick et al.

    The ankyrin repeat: A diversity of interactions on a common structural framework

    Trends Biochem. Sci.

    (1999)
  • J.E. Seely et al.

    Use of ion-exchange chromatography and hydrophobic interaction chromatography in the preparation and recovery of polyethylene glycol-linked proteins

    J. Chromatogr. A

    (2001)
  • D.E. Shafer et al.

    Reaction of tris(2-carboxyethyl)phosphine (TCEP) with maleimide and alpha-haloacyl groups: Anomalous elution of TCEP by gel filtration

    Anal. Biochem.

    (2000)
  • C.B. Siegall et al.

    Functional analysis of domains II, Ib, and III of Pseudomonas exotoxin

    J. Biol. Chem.

    (1989)
  • N. Stefan et al.

    DARPins recognizing the tumor-associated antigen EpCAM selected by phage and ribosome display and engineered for multivalency

    J. Mol. Biol.

    (2011)
  • D. Steiner et al.

    Efficient selection of DARPins with sub-nanomolar affinities using SRP phage display

    J. Mol. Biol.

    (2008)
  • M.T. Stumpp et al.

    Designing repeat proteins: Modular leucine-rich repeat protein libraries based on the mammalian ribonuclease inhibitor family

    J. Mol. Biol.

    (2003)
  • J.P. Theurillat et al.

    Designed ankyrin repeat proteins: A novel tool for testing epidermal growth factor receptor 2 expression in breast cancer

    Mod. Pathol.

    (2010)
  • G.M. Thurber et al.

    Antibody tumor penetration: Transport opposed by systemic and antigen-mediated clearance

    Adv. Drug Deliv. Rev.

    (2008)
  • M. Trzpis et al.

    Epithelial cell adhesion molecule: More than a carcinoma marker and adhesion molecule

    Am. J. Pathol.

    (2007)
  • D. Veesler et al.

    Crystal structure of a DARPin neutralizing inhibitor of lactococcal phage TP901-1: Comparison of DARPin and camelid VHH binding mode

    J. Biol. Chem.

    (2009)
  • D. Vercauteren et al.

    The use of inhibitors to study endocytic pathways of gene carriers: Optimization and pitfalls

    Mol. Ther.

    (2010)
  • S.K. Wetzel et al.

    Folding and unfolding mechanism of highly stable full-consensus ankyrin repeat proteins

    J. Mol. Biol.

    (2008)
  • S.K. Wetzel et al.

    Residue-resolved stability of full-consensus ankyrin repeat proteins probed by NMR

    J. Mol. Biol.

    (2010)
  • J. Willuda et al.

    Tumor targeting of mono-, di- and tetravalent anti-p185HER-2 miniantibodies multimerized by self-associating peptides

    J. Biol. Chem.

    (2001)
  • A. Wörn et al.

    Stability engineering of antibody single-chain Fv fragments

    J. Mol. Biol.

    (2001)
  • C. Zahnd et al.

    Selection and characterization of Her2 binding-designed ankyrin repeat proteins

    J. Biol. Chem.

    (2006)
  • C. Zahnd et al.

    A designed ankyrin repeat protein evolved to picomolar affinity to Her2

    J. Mol. Biol.

    (2007)
  • T. Zimmerman et al.

    Simultaneous metal chelate affinity purification and endotoxin clearance of recombinant antibody fragments

    J. Immunol. Methods

    (2006)
  • G.P. Adams et al.

    Highly specific in vivo tumor targeting by monovalent and divalent forms of 741F8 anti-c-erbB-2 single-chain Fv

    Cancer Res.

    (1993)
  • G.P. Adams et al.

    High affinity restricts the localization and tumor penetration of single-chain Fv antibody molecules

    Cancer Res.

    (2001)
  • A. Amalfitano et al.

    Separating fact from fiction: Assessing the potential of modified adenovirus vectors for use in human gene therapy

    Curr. Gene Ther.

    (2002)
  • P. Amstutz et al.

    Rapid selection of specific MAP kinase-binders from designed ankyrin repeat protein libraries

    Protein Eng. Des. Sel.

    (2006)
  • T.M. Bandeiras et al.

    Structure of wild-type Plk-1 kinase domain in complex with a selective DARPin

    Acta Crystallogr. D Biol. Crystallogr.

    (2008)
  • Cited by (108)

    View all citing articles on Scopus
    View full text