Journal of Molecular Biology
Volume 343, Issue 3, 22 October 2004, Pages 685-701
Journal home page for Journal of Molecular Biology

Substantial Energetic Improvement with Minimal Structural Perturbation in a High Affinity Mutant Antibody

https://doi.org/10.1016/j.jmb.2004.08.019Get rights and content

Here, we compare an antibody with the highest known engineered affinity (Kd=270 fM) to its high affinity wild-type (Kd=700 pM) through thermodynamic, kinetic, structural, and theoretical analyses. The 4M5.3 anti-fluorescein single chain antibody fragment (scFv) contains 14 mutations from the wild-type 4-4-20 scFv and has a 1800-fold increase in fluorescein-binding affinity. The dissociation rate is ∼16,000 times slower in the mutant; however, this substantial improvement is offset somewhat by the association rate, which is ninefold slower in the mutant. Enthalpic contributions to binding were found by calorimetry to predominate in the differential binding free energy. The crystal structure of the 4M5.3 mutant complexed with antigen was solved to 1.5 Å resolution and compared with a previously solved structure of an antigen-bound 4-4-20 Fab fragment. Strikingly, the structural comparison shows little difference between the two scFv molecules (backbone RMSD of 0.6 Å), despite the large difference in affinity. Shape complementarity exhibits a small improvement between the variable light chain and variable heavy chain domains within the antibody, but no significant improvement in shape complementarity of the antibody with the antigen is observed in the mutant over the wild-type. Theoretical modeling calculations show electrostatic contributions to binding account for −1.2 kcal/mol to −3.5 kcal/mol of the binding free energy change, of which −1.1 kcal/mol is directly associated with the mutated residue side-chains. The electrostatic analysis reveals several mechanistic explanations for a portion of the improvement. Collectively, these data provide an example where very high binding affinity is achieved through the cumulative effect of many small structural alterations.

Introduction

The mechanisms differentiating very high affinity protein interactions (sub-picomolar dissociation constants) from high affinity (nanomolar dissociation constants) have seldom been possible to study,1, 2 because few protein interaction families that span this range of affinity have been identified from either nature or engineering efforts. Antibodies provide a natural affinity series of related proteins but, in vivo, rarely surpass a 0.1 nanomolar dissociation constant (Kd) due to physiological limits on the selection.3 Further affinity improvements are often desired for pharmaceutical and biotechnology uses, and can be attained through directed evolution protein engineering techniques.4, 5, 6, 7 Computational methods have also provided some limited predictions of specific changes needed to improve high affinity interactions,8 although they have yet to realize substantial improvements experimentally. To date, the transitions from micromolar to nanomolar9, 10, 11, 12, 13, 14 and from femtomolar back to picomolar2, 15 affinity have been studied, providing little guidance into the mechanistic means to engineer an enhancement from nanomolar to femtomolar affinity.

Previous studies of antibody maturation from weak binding (micromolar) to high affinity binding (nanomolar) have shown that a number of factors are involved in improving affinity across this range. Several studies have noted that decreased loss of entropy on binding, with creation of a lock and key fit mechanism, contribute to affinity improvements.11, 14, 16, 17 In a structural, thermodynamic, and computational energetic analysis of an esterolytic antibody bound to a p-nitrophenyl phosphonate hapten, affinity maturation occurred through a reorganization of the binding site geometry so as to optimize gaining favorable electrostatic interactions with the hapten and losing those with solvent during the binding process.9, 18 In a study of an anti-testosterone antibody affinity maturation, improvement was derived from small structural changes providing more comprehensive packing around the antigen.12 A study of an anti-(4-hydroxy-3-nitrophenyl) acetate affinity matured antibody suggests that “relief of cramped contacts” in the wild-type provided the increased affinity in the mutant.19 In the first structural mechanistic study of affinity matured antibodies to a protein antigen, the higher affinity was attributed to an increased burial of hydrophobic surface area on binding and improved shape complementarity between antigen and antibody.13 Across these affinity maturation studies, mutations were found in antigen contact and non-contact regions; however, it was not always clear whether all the cited mutations were contributing to the improvements. It is perhaps to be expected that insights derived from studies of micromolar to nanomolar affinity improvements may not translate directly to mechanisms involved in fine-tuning an interaction from nanomolar to femtomolar affinity levels.

The monoclonal antibody 4-4-20 directed against the hapten fluorescein was originally isolated to facilitate studies of immunological recognition.20 The 4-4-20 binding has been extensively studied as the whole IgG, the Fab, the Fv, and the scFv by thermodynamic, kinetic, structural, computational, spectroscopic, and mutational methods.21, 22, 23, 24, 25, 26, 27, 28 A recent comparative structural study compared the 4-4-20 Fab structure with an idiotypically related lower affinity Fab, 9-40.29 This study revealed a reorganization of the binding site and a more closed structure with fewer water molecules at the binding pocket for 4-4-20, compared with 9-40, to account for the change in binding affinity.

Ten femtomolar affinity mutants of 4-4-20 were previously identified through directed evolution with random mutagenesis over the whole Fv followed by screening with yeast surface display.7 Clone 4M5.3 from that study was chosen for further study here. This clone contains 14 mutations, including all ten of the consensus mutations observed in the final directed evolution round clones.

Here, we compare 4M5.3 to 4-4-20 by kinetic, thermodynamic, structural, and theoretical analyses. No individual structural change was unambiguously associated with the −4.5 kcal/mol improvement in binding free energy. However, these results implicate a sum of many small structural changes in combination with specific mutations that improve electrostatics, which together result in a large binding free energy change.

Section snippets

Kinetics and thermodynamics

The fluorescein-biotin equilibrium binding constant for 4-4-20 in both the soluble form and displayed on the surface of yeast has been determined previously, with a Kd of 0.7 nM at 25 °C for both.7 The 4M5.3 binding constant with fluorescein-biotin was previously determined to be 270 fM through the ratio of dissociation and association kinetic rates.7 To test for the presence of kinetic intermediates that would cause the ratio of the measured association and dissociation rates to differ from the

Discussion

The entropic and enthalpic components of protein–ligand binding free energy include: (1) covalent, van der Waals, hydrogen-bonding, and electrostatic interactions within and between protein and ligand; (2) solvation effects, including desolvation costs and the hydrophobic effect; and (3) overall translational and rotational, as well as internal (often represented as vibrational) degrees of freedom within the protein and ligand. It remains unclear how best to create substantial enhancements in

Over-expression and purification of scFv proteins

The 4-4-20 and 4M5.3 scFvs7 were subcloned into the pRS316 backbone with the Gal1-10 promoter.40 A Flag tag (amino acid sequence Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys) was inserted between the EagI and NheI sites N-terminal to the scFv sequence and the first expressed amino acid was changed to alanine by site-directed mutagenesis (Stratagene, La Jolla, CA). The sequence expressed was Ala-Ala-Arg-Pro-(Flag-tag)-(scFv). The scFvs were expressed solubly in Saccharomyces cerevisiae strain YVH10, which

Acknowledgements

K.M. thanks Balaji Rao for extremely helpful discussions about the competition assays for Kd determination on the surface of yeast. Funding was provided from NCI CA96504 (to K.D.W.). S.M.L. was funded by an NSF Graduate Fellowship. Funding was provided from the Alfred P. Sloan Foundation (to C.L.D.). H.H.H. was funded by an NIH biotechnology training grant (T32-GM08334). Portions of this research were carried out at the Stanford Synchrotron Radiation Laboratory, a national user facility

References (56)

  • D.M. Kranz et al.

    Partial elucidation of an anti-hapten repertoire in BALB/c mice: comparative characterization of several monoclonal anti-fluorescyl antibodies

    Mol. Immunol.

    (1981)
  • D.M. Kranz et al.

    Mechanisms of ligand binding by monoclonal anti-fluorescyl antibodies

    J. Biol. Chem.

    (1982)
  • J.N. Herron et al.

    High resolution structures of the 4-4-20 Fab-fluorescein complex in two solvent systems: effects of solvent on structure and antigen-binding affinity

    Biophys. J.

    (1994)
  • L.K. Denzin et al.

    Mutational analysis of active site contact residues in anti-fluorescein monoclonal antibody 4-4-20

    Mol. Immunol.

    (1993)
  • S. Terzyan et al.

    Three-dimensional structures of idiotypically related Fabs with intermediate and high affinity for fluorescein

    J. Mol. Biol.

    (2004)
  • M.C. Lawrence et al.

    Shape complementarity at protein/protein interfaces

    J. Mol. Biol.

    (1993)
  • F. Dullweber et al.

    Factorising ligand affinity: a combined thermodynamic and crystallographic study of trypsin and thrombin inhibition

    J. Mol. Biol.

    (2001)
  • S.K. Burley et al.

    Amino–aromatic interactions in proteins

    FEBS Letters

    (1986)
  • B.M. Baker et al.

    Dissecting the energetics of a protein–protein interaction: the binding of ovomucoid third domain to elastase

    J. Mol. Biol.

    (1997)
  • P. Mendes

    Biochemistry by numbers: simulation of biochemical pathways with Gepasi 3

    Trends Biochem. Sci.

    (1997)
  • Z. Otwinowski et al.

    Processing of X-ray difraction data collected in oscillation mode

    Methods Enzymol.

    (1997)
  • J. Foote et al.

    Kinetic and affinity limits on antibodies produced during immune responses

    Proc. Natl Acad. Sci. USA

    (1995)
  • E.T. Boder et al.

    Directed evolution of antibody fragments with monovalent femtomolar antigen-binding affinity

    Proc. Natl Acad. Sci. USA

    (2000)
  • R.W. Dixon et al.

    Theoretical and experimental studies of biotin analogues that bind almost as tightly to streptavidin as biotin

    J. Org. Chem.

    (2002)
  • G.J. Wedemayer et al.

    Structural insights into the evolution of an antibody combining site

    Science

    (1997)
  • H. Wu et al.

    Stepwise in vitro affinity maturation of Vitaxin, an alphav beta3-specific humanized mAb

    Proc. Natl Acad. Sci. USA

    (1998)
  • J. Yin et al.

    A comparative analysis of the immunological evolution of antibody 28B4

    Biochemistry

    (2001)
  • Y. Li et al.

    X-ray snapshots of the maturation of an antibody response to a protein antigen

    Nature Struct. Biol.

    (2003)
  • Cited by (108)

    • Photoswitchable CAR-T Cell Function In Vitro and In Vivo via a Cleavable Mediator

      2021, Cell Chemical Biology
      Citation Excerpt :

      MALDI-TOF-MS: m/z C59H57N15NaO16S [M+Na]+1286.37, Found 1286.85; C59H57KN15O16S [M+K]+ 1302.35, Found 1302.83. A gene cassette containing the 4M5.3 anti-FITC scFv (Ma, et al., 2016; Midelfort, et al., 2004; Vaughan, et al., 1996), the CD8α hinge and transmembrane region, and the cytoplasmic domains of 4-1BB and CD3ζ was synthesized by Icartab Co., Ltd. (SuZhou, China) and cloned into a lentivirus vector. Lentivirus production and transduction of human T cells were performed as previously described (Kim, et al., 2015; Ma, et al., 2016).

    View all citing articles on Scopus
    View full text