Design of an Optimized Scaffold for Affibody Molecules

https://doi.org/10.1016/j.jmb.2010.03.002Get rights and content

Abstract

Affibody molecules are non-immunoglobulin-derived affinity proteins based on a three-helical bundle protein domain. Here, we describe the design process of an optimized Affibody molecule scaffold with improved properties and a surface distinctly different from that of the parental scaffold. The improvement was achieved by applying an iterative process of amino acid substitutions in the context of the human epidermal growth factor receptor 2 (HER2)-specific Affibody molecule ZHER2:342. Replacements in the N-terminal region, loop 1, helix 2 and helix 3 were guided by extensive structural modeling using the available structures of the parent Z domain and Affibody molecules. The effect of several single substitutions was analyzed followed by combination of up to 11 different substitutions. The two amino acid substitutions N23T and S33K accounted for the most dramatic improvements, including increased thermal stability with elevated melting temperatures of up to + 12 °C. The optimized scaffold contains 11 amino acid substitutions in the nonbinding surface and is characterized by improved thermal and chemical stability, as well as increased hydrophilicity, and enables generation of identical Affibody molecules both by chemical peptide synthesis and by recombinant bacterial expression. A HER2-specific Affibody tracer, [MMA-DOTA-Cys61]-ZHER2:2891-Cys (ABY-025), was produced by conjugating MMA-DOTA (maleimide-monoamide-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid) to the peptide produced either chemically or in Escherichia coli. ABY-025 showed high affinity and specificity for HER2 (equilibrium dissociation constant, KD, of 76 pM) and detected HER2 in tissue sections of SKOV-3 xenograft and human breast tumors. The HER2-binding capacity was fully retained after three cycles of heating to 90 °C followed by cooling to room temperature. Furthermore, the binding surfaces of five Affibody molecules targeting other proteins (tumor necrosis factor α, insulin, Taq polymerase, epidermal growth factor receptor or platelet-derived growth factor receptor β) were grafted onto the optimized scaffold, resulting in molecules with improved thermal stability and a more hydrophilic nonbinding surface.

Introduction

Affibody molecules are small and robust affinity ligands based on the three-helical-bundle Z domain, which is a stabilized variant of the B domain of staphylococcal protein A (SPA).1 Thus, Affibody molecules are not related to and do not share sequence or structural homology with antibodies. The Z domain was chosen as the starting point for the construction of novel binding proteins because this small protein (58 amino acids, approximately 6.5 kDa) was known to have excellent biophysical properties—including high melting temperature, reversible and rapid folding, a binding surface as large as that of an antibody, high solubility in aqueous solutions—and because it could be produced at high levels in Escherichia coli.2, 3, 4 The exceptional nature of the Z domain scaffold is further highlighted by the shortest folding time yet reported for a protein, that is, 3 μs.5 Combinatorial libraries containing different Affibody molecules have been generated by randomizing 13 surface-exposed amino acids located in helices 1 and 2 of the Z domain.4, 6 Affibody molecules with nanomolar and picomolar affinities have been selected from these libraries to a large range of targets, including human epidermal growth factor receptor 2 (HER2),7, 8 epidermal growth factor receptor (EGFR),9, 10 tumor necrosis factor α (TNFα)11, 12 and amyloid-β (Aβ) peptide.6, 13

Multimeric Affibody molecules (i.e., head-to-tail gene fusions of two or more Affibody molecules),12, 14, 15 bispecific Affibody molecules (i.e., fusion of two Affibody molecules having separate target specificities16 and fusions of Affibody molecules with other proteins and toxins17, 18, 19, 20 have been shown to be functionally active. Since the Affibody scaffold lacks cysteines, homogenous site-specific modifications are possible by the introduction of a unique cysteine. This has been done to achieve site-specific labeling with, for example, different radionuclides and fluorescent dyes.21, 22, 23, 24, 25 Thus, Affibody molecules have been shown to be amenable for a wide range of additional modifications, including fusions at the N- or the C-terminus.

Each domain of protein A contains a large surface interacting with the Fc portion of immunoglobulin (Ig) G and a partially overlapping surface responsible for interaction with the variable domain of antibodies containing a heavy chain from the VH3 family. The latter surface has been defined in the crystal structure of the D domain of SPA in complex with a Fab fragment of a human IgM.26 The amino acids most important for the VH3 interactions are G29, F30, S33, D36, D37 and E47. These amino acids are conserved in all five domains of protein A, although the B domain has been reported to have little or no Fab-binding activity.27 The Z domain was based on the B domain and stabilized by the G29A substitution, removing a hydroxylamine-sensitive site in position N28–G29.1 This replacement in the Z domain scaffold has been shown to essentially remove the Fab binding.28 Amino acid residues in the Fc-binding surface are randomized in Affibody libraries, which removes the Fc binding. The only amino acid (Q32) implicated to be directly involved in both the Fc and the VH3 interactions is among the 13 randomized residues. Although Affibody molecules are derived from the Z domain, weak interaction with immunoglobulins cannot be excluded, for example, through interaction with the VH3 domain.

Three-dimensional structures of the Z domain determined by NMR spectroscopy have shown that the protein is composed of three nearly perfectly antiparallel α-helices.29, 30 This fold is indeed conserved in the NMR-derived structures of the free ZTaq and anti-ZTaq Affibody molecules.31 Furthermore, structural analysis of the anti-idiotypic Affibody molecule ZSPA-1 in complex with the Z domain32, 33 and the ZTaq–anti-ZTaq complex31 has confirmed the ability of these Affibody molecules to retain the three-helical-bundle structure upon target binding. Thus, the Z domain has been shown to accommodate an exchange of 22% (13 of 58) of all amino acids.

The goal of the present investigation was to remodel the nonbinding surface of the Affibody molecules to further reduce similarity to the Z domain and to reduce any residual interactions with immunoglobulins while increasing the overall hydrophilicity, improving thermal and storage stability, enhancing amenability for peptide synthesis and retaining target binding capability and folding properties. Here, we report the design of an optimized scaffold for Affibody molecules. In the new scaffold, 11 amino acids have been substituted in the nonbinding surface. Together with the 13 randomized positions, 41% (24 of 58) of all amino acids in the parent Z domain have been substituted.

Section snippets

Rationale, molecular modeling and procedure

The development of an optimized scaffold was an iterative process based on substitutions of one or several amino acids of the HER2-binding Affibody molecule ZHER2:342.8 The results of several single-site substitutions were used as a starting point, followed by constructs with various combinations of 2, 3, 4, 5, 6, 8, 9 and 11 amino acid exchanges. Several points were considered in the selection of alternative amino acids for a given position. A structural model for ZHER2:342 was generated on

Discussion

Here we describe the design of an optimized Affibody molecule scaffold with improved properties and a surface distinctly different from that of the parent Z domain. The improvements include increased hydrophilicity, higher thermal stability and further reduced interactions with immunoglobulins, improved photometric detection and full production flexibility.

The effects of different scaffold amino acid replacements were studied in the context of the HER2-binding Affibody molecule ZHER2:342.

Cloning

Affibody molecules were cloned into expression vectors derived from pAY00442 containing a T7 promoter,50 a sequence coding for a hexahistidine (His6) tag, a multiple cloning site and a kanamycin resistance gene. Affibody molecules without His6 tag were cloned into pAY00457, which is similar to pAY00442 but does not contain the sequence coding for a His6 tag. All oligonucleotides used in this study were from Thermo Electron Corp. (Ulm, Germany).

ZHER2:342 variants were created by three different

Acknowledgements

We thank Fredrik Frejd and Karin Nord for critically reading the manuscript.

References (53)

  • LendelC. et al.

    Structural basis for molecular recognition in an affibody:affibody complex

    J. Mol. Biol.

    (2006)
  • TavernaD.M. et al.

    Why are proteins so robust to site mutations?

    J. Mol. Biol.

    (2002)
  • DecristoforoC. et al.

    99m-Technetium-labelled peptide-HYNIC conjugates: effects of lipophilicity and stability on biodistribution

    Nucl. Med. Biol.

    (1999)
  • LacroixE. et al.

    Elucidating the folding problem of alpha-helices: local motifs, long-range electrostatics, ionic-strength dependence and prediction of NMR parameters

    J. Mol. Biol.

    (1998)
  • HaackT. et al.

    Serine derived oxazolidines as secondary structure disrupting, solubilizing building blocks in peptide synthesis

    Tetrahedron Lett.

    (1992)
  • NilssonB. et al.

    A synthetic IgG-binding domain based on staphylococcal protein A

    Protein Eng.

    (1987)
  • DeisenhoferJ.

    Crystallographic refinement and atomic models of a human Fc fragment and its complex with fragment B of protein A from Staphylococcus aureus at 2.9- and 2.8-Å resolution

    Biochemistry

    (1981)
  • GoudaH. et al.

    Three-dimensional solution structure of the B domain of staphylococcal protein A: comparisons of the solution and crystal structures

    Biochemistry

    (1992)
  • NordK. et al.

    A combinatorial library of an alpha-helical bacterial receptor domain

    Protein Eng.

    (1995)
  • AroraP. et al.

    Fast and faster: a designed variant of the B-domain of protein A folds in 3 microsec

    Protein Sci.

    (2004)
  • WikmanM. et al.

    Selection and characterization of HER2/neu-binding affibody ligands

    Protein Eng. Des. Sel.

    (2004)
  • OrlovaA. et al.

    Tumor imaging using a picomolar affinity HER2 binding affibody molecule

    Cancer Res.

    (2006)
  • FriedmanM. et al.

    Phage display selection of Affibody molecules with specific binding to the extracellular domain of the epidermal growth factor receptor

    Protein Eng. Des. Sel.

    (2007)
  • KronqvistN. et al.

    A novel affinity protein selection system based on staphylococcal cell surface display and flow cytometry

    Protein Eng. Des. Sel.

    (2008)
  • JonssonA. et al.

    Generation of tumour-necrosis-factor-alpha-specific affibody molecules capable of blocking receptor binding in vitro

    Biotechnol. Appl. Biochem.

    (2009)
  • SteffenA.C. et al.

    In vitro characterization of a bivalent anti-HER-2 affibody with potential for radionuclide-based diagnostics

    Cancer Biother. Radiopharm.

    (2005)
  • Cited by (133)

    • Engineering high affinity antigen-binders: Beyond conventional antibodies

      2024, Advances in Protein Chemistry and Structural Biology
    View all citing articles on Scopus

    The phrase “Affibody molecule” is used in this publication instead of “Affibody® molecule.”

    1

    Present address: C. Lendel, Department of Molecular Biology, SLU, Box 590, SE-751 24 Uppsala, Sweden.

    View full text