Insights into Protein–Protein Binding by Binding Free Energy Calculation and Free Energy Decomposition for the Ras–Raf and Ras–RalGDS Complexes

https://doi.org/10.1016/S0022-2836(03)00610-7Get rights and content

Abstract

Absolute binding free energy calculations and free energy decompositions are presented for the protein–protein complexes H-Ras/C-Raf1 and H-Ras/RalGDS. Ras is a central switch in the regulation of cell proliferation and differentiation. In our study, we investigate the capability of the molecular mechanics (MM)-generalized Born surface area (GBSA) approach to estimate absolute binding free energies for the protein–protein complexes. Averaging gas-phase energies, solvation free energies, and entropic contributions over snapshots extracted from trajectories of the unbound proteins and the complexes, calculated binding free energies (Ras–Raf: −15.0(±6.3) kcal mol−1; Ras–RalGDS: −19.5(±5.9) kcal mol−1) are in fair agreement with experimentally determined values (−9.6 kcal mol−1; −8.4 kcal mol−1), if appropriate ionic strength is taken into account. Structural determinants of the binding affinity of Ras–Raf and Ras–RalGDS are identified by means of free energy decomposition. For the first time, computationally inexpensive generalized Born (GB) calculations are applied in this context to partition solvation free energies along with gas-phase energies between residues of both binding partners. For selected residues, in addition, entropic contributions are estimated by classical statistical mechanics. Comparison of the decomposition results with experimentally determined binding free energy differences for alanine mutants of interface residues yielded correlations with r2=0.55 and 0.46 for Ras–Raf and Ras–RalGDS, respectively. Extension of the decomposition reveals residues as far apart as 25 Å from the binding epitope that can contribute significantly to binding free energy. These “hotspots” are found to show large atomic fluctuations in the unbound proteins, indicating that they reside in structurally less stable regions. Furthermore, hotspot residues experience a significantly larger-than-average decrease in local fluctuations upon complex formation. Finally, by calculating a pair-wise decomposition of interactions, interaction pathways originating in the binding epitope of Raf are found that protrude through the protein structure towards the loop L1. This explains the finding of a conformational change in this region upon complex formation with Ras, and it may trigger a larger structural change in Raf, which is considered to be necessary for activation of the effector by Ras.

Introduction

The capability of proteins to form stable complexes is of fundamental importance for a wide range of biological processes, including hormone–receptor interactions, proteinase inhibition, antibody–antigen interactions, signal transduction and enzyme allostery.1 Perturbation of protein–protein interactions can result in diseases such as sickle-cell anaemia2 or the formation of tumors.3 Therefore, a promising way to interfere with biological processes is the control of protein–protein interactions by means of small molecules that modulate the formation of protein–protein complexes. Recent results demonstrate the feasibility of this approach in principle.4., 5., 6., 7., 8., 9., 10., 11. However, our limited understanding of the energetics and dynamics of the mutual binding of two proteins presents a severe limitation for the rational design of small, non-peptide inhibitors to modulate protein–protein interactions.

Investigations of crystallographically determined protein–protein complexes, thermodynamic experiments and mutational studies have contributed most to the current understanding of protein–protein interactions.1., 12., 13., 14. Of particular interest are properties of binding interfaces of hetero-complexes, which are often made and broken according to the environment or external factors, and involve proteins that must exist independently.12 Results from heat capacity measurements suggest that hydrophobic interactions are less dominant in protein–protein association15 than in protein folding,16 which points to an increased influence of (long-range) electrostatic interactions across the binding interface. Mutational studies, mostly alanine scanning mutagenesis, on protein–protein interfaces have revealed that a small number of amino acid residues in contact across the interface often yield a significant contribution to the free energy of binding.1., 17., 18., 19., 20. These occurrences of “hotspots”10., 21. thus lead to a distinction between the “structural epitope” (i.e. the interface between proteins as defined by geometrical considerations) and the “functional epitope”.22

Computational and theoretical methods provide a molecular view of the structural and energetic consequences of mutations, and can address the origin of binding in terms of contributions from electrostatic and van der Waals interactions and changes in solvation.23 Furthermore, rigorous computational approaches may be used as a means of guiding new experimental investigations.24 Recently, it has been recognized that mutations may perturb the complex (as generally assumed), as well as the unbound states.25 In the latter case, without additional structural information, interpreting energetic consequences of the mutations in terms of changes in binding interactions will thus be misleading. In contrast, virtual mutagenesis employing thermodynamic cycles26 can (in principle) provide a full description of structural and energetic influences due to mutations on the bound and unbound ensembles.

The systems we examine are the complexes between human H-Ras (Ras) and the Ras-binding domain of C-Raf1 (Raf) as well as between Ras and the Ras-interacting domain of RalGDS (RalGDS). Ras is a GTP-hydrolyzing protein (GTPase, 21 kDa) that acts as central switch in the regulation of cell proliferation and differentiation.27 The active and inactive states of Ras are coupled to the binding of GTP or GDP to the protein, respectively.3 Exchange of the guanine nucleotide induces conformational transitions mainly in two regions that have been called switch I (residues 30–37) and switch II (residues 60–76). Switch I overlaps with what has been called the effector region (residues 32–40), which is involved in interactions with the effectors.28

After activation by “upstream” signals of the signal transduction cascade, Ras itself activates a cascade of protein kinases,29 the first of which is the “downstream” effector Raf.30 The identification of other interaction partners (such as RalGDS31 or phosphatidyl-inositol-3′-kinase32) has led to the conclusion that activated Ras simultaneously induces more than one signaling pathway.3., 33. The fact that permanently activated Ras is one of the most frequent oncogenes, it is found in 30% of all human tumors,34 underlines the importance of understanding its interactions with downstream effectors.

Molecular structures for the unbound proteins35., 36., 37., 38. as well as for Ras–RalGDS39., 40. or complexes closely related to Ras–Raf41., 42. have been determined either by X-ray crystallography or by NMR spectroscopy. They show that the Ras-effector interactions involve mainly an inter-protein β-sheet, resulting in interface sizes of 1200–1300 Å.2., 3., 39. With respect to both effectors, the interaction sites on Ras mutually overlap to a large extent, but RalGDS is rotated by ≈35° when viewed from the direction of the Ras moiety compared to Raf, resulting in a shift of interaction contacts towards the switch I and switch II region of Ras.39 For the Ras–RalGDS case, comparison of the components of the complex structure with the unbound molecules indicates small (<0.5 Å rmsd of Cα atoms) or moderate (1.5 Å rmsd of Cα atoms) changes of the protein structures upon binding. It was also found in the complex with Raf, that one of two nearly equally populated states (“state 2”) of free, GTP-bound Ras with different conformations of the loop L2 region (in particular Y32) is stabilized.43., 44.

Although Raf is the best characterized Ras effector from mammalian cells in terms of biological relevance as well as structural detail,45 it has not been possible to co-crystallize full-length Raf with Ras or Ras-related proteins but only the independently folding domain from the conserved region 1 known as the Ras-binding domain (RBD), which has been investigated in this study. Although binding of the effector region of Ras to the RBD is the major site of GTP-dependent binding, there is clear evidence that contacts between Ras and Raf outside the RBD contribute to binding and kinase activation.46 Along these lines, the precise molecular mechanism for Raf activation is not fully understood. The question of whether a recruitment of Raf by Ras is sufficient for activation with Ras acting as an allosteric regulator, or whether additional events such as phosphorylation by membrane-bound kinases or a transphosphorylation by dimerized Raf itself is responsible, has not been resolved.43., 45., 47. It is consistently proposed, however, that Raf activation requires a structural change from an inactive conformation of the protein being folded between regulatory and kinase domains to an activated, open form.45., 47. Together with the structural data, thermodynamic binding data48 for wild-type or mutant Ras and effector proteins provide a solid basis for the validation of calculations.

Ras-effector interactions have been investigated repeatedly by means of theoretical methods. Among others, studies have included molecular dynamics (MD) simulations on Ras or Rap-1A bound to Raf aimed at identification of effector domains in the Raf protein49., 50. or only of Raf to investigate conformational effects of the R89K mutation.51 Zeng et al.52 performed alchemical free energy calculations on the R89K mutant of Raf, while Muegge et al.53 concentrated mainly on electrostatic contributions to the Rap-1A–Raf interaction.

Here, (absolute) free energies of protein–protein binding have been computed with the molecular mechanics (MM)-generalized Born surface area (GBSA) approach.54., 55. In this approach, gas-phase energies, solvation free energies, and entropic contributions are summed and averaged over snapshots from MD trajectories. The related MM-Poisson–Boltzmann (PB) SA approach has been applied to compare relative stabilities of different conformations of nucleic acids,55 to identify correctly folded proteins,56 and to estimate binding affinities of small molecules binding to proteins.57., 58., 59. It has been used to predict the effects of amino acid mutations on binding affinities.60., 61., 62. More recently, electrostatic contributions to solvation free energies have been calculated by GB models.55., 63., 64. Since GB is significantly faster than PB, the former is an attractive alternative. In addition, GB allows one to decompose the contributions to binding free energies on a per-residue basis which provides an interesting, structurally non-perturbing alternative to the usual “computational alanine scanning” approach.60 Although a similar idea has been presented for small-molecule solvation,65 to the best of our knowledge this is the first time that GB has been used for such an approach in the context of macromolecular association.

In this work, 5 ns MD simulations were carried out for the unbound proteins Ras, Raf, and RalGDS as well as for the complexes Ras–Raf and Ras–RalGDS. For 150 snapshots extracted from the last 3 ns of the stable trajectories, energy and entropy contributions were calculated and averaged. Energy decomposition on a per-residue basis showed convincing consistency with experimentally determined mutagenesis data. In addition, residues apart from the binding interface could be identified to contribute to the protein–protein binding (hotspots). In the Raf case, particularly, these hotspots form interaction pathways that originate in the binding interface and extend through the protein structure. Possible implications for an allosteric activation of the effector molecule are discussed. The results of this study indicate clearly the significance of the applied computational method to provide insight into binding thermodynamics of proteins on an atomic level, although limitations of the current methodology still exist.

Section snippets

Structures from MD simulations

For the systems Raf, RalGDS, Ras, Ras–Raf, and Ras–RalGDS, MD simulations with the particle mesh Ewald (PME) method were carried out in explicit water for 5 ns. For each trajectory, the time-series of the rmsd of backbone atoms from the experimental starting structure is given in Figure 1.

For all systems but Raf, these rmsd values vary between 0.7 Å and 2.0 Å. The rmsd values with respect to all protein atoms remain below 2.3 Å during the course of the simulations for these four cases. These

Conclusion

Absolute binding free energies have been calculated for the association of Ras with Raf or RalGDS, which are in fair agreement with experimental data. Following the MM-GBSA approach, gas-phase energies and entropy contributions by the solute have been complemented by solvation free energies obtained from a generalized Born model. The analysis of energetic contributions to the (absolute) binding free energy has revealed that van der Waals interactions together with non-polar contributions to the

Molecular dynamics simulations

All simulations were performed with the AMBER 7 suite of programs114 together with the Cornell et al. force-field.115 Bonded parameters for the triphosphate moiety of GTP were taken from Leach & Klein116 and atomic partial charges for GTP4− were derived using the RESP procedure.117 Non-bonded parameters for Mg2+ were taken from Aqvist118 and were adapted to the AMBER combining rules. The starting structures for the simulations of the human unbound proteins and complexes were taken from the

Acknowledgements

We thank Dr C. Herrmann (Max-Planck-Institute of Molecular Physiology, Dortmund) for providing comprehensive isothermal titration calorimetry data for Ras–Raf and Ras–RalGDS, and Dr A. Onufriev (The Scripps Research Institute, La Jolla, CA) for helpful discussions. The present study was supported by NIH grant RR12255. H.G. gratefully acknowledges a Feodor-Lynen fellowship awarded by the Alexander-von-Humboldt foundation, Germany.

References (134)

  • I.R. Vetter et al.

    Structural and biochemical analysis of Ras-effector signaling via RalGDS

    FEBS Letters

    (1999)
  • F. McCormick et al.

    Interactions between Ras proteins and their effectors

    Curr. Opin. Biotechnol.

    (1996)
  • B. Yamamori et al.

    Purification of a Ras-dependent mitogen-activated protein kinase from bovine brain cytosol and its identification as a complex of B-Raf and 14-3-3 proteins

    J. Biol. Chem.

    (1995)
  • M.G. Rudolph et al.

    Thermodynamics of ras/effector and Cdc42/effector interactions probed by isothermal titration calorimetry

    J. Biol. Chem.

    (2001)
  • W. Wang et al.

    Free energy calculations on dimer stability of the HIV protease using molecular dynamics and a continuum solvent model

    J. Mol. Biol.

    (2000)
  • C.J. Cramer et al.

    Polarization of the nucleic acid bases in aqueous solution

    Chem. Phys. Letters

    (1992)
  • T. Terada et al.

    Nuclear magnetic resonance and molecular dynamics studies on the interactions of the Ras-binding domain of Raf-1 with wild-type and mutant Ras proteins

    J. Mol. Biol.

    (1999)
  • M. Rarey et al.

    A fast flexible docking method using an incremental construction algorithm

    J. Mol. Biol.

    (1996)
  • F.B. Sheinerman et al.

    On the role of electrostatic interactions in the design of protein–protein interfaces

    J. Mol. Biol.

    (2002)
  • S.Y. Noskov et al.

    Free energy decomposition of protein–protein interactions

    Biophys. J.

    (2001)
  • D.A. Case

    Normal mode analysis of protein dynamics

    Curr. Opin. Struct. Biol.

    (1994)
  • S.D. Pickett et al.

    Empirical scale of side-chain conformational entropy in protein folding

    J. Mol. Biol.

    (1993)
  • V.K. Misra et al.

    Salt effects on ligand–DNA binding. Minor groove binding antibiotics

    J. Mol. Biol.

    (1994)
  • C.K. Vaughan et al.

    Structural response to mutation at a protein–protein interface

    J. Mol. Biol.

    (1999)
  • A.E. Mark et al.

    Decomposition of the free energy of a system in terms of specific interactions. Implications for theoretical and experimental studies

    J. Mol. Biol.

    (1994)
  • G.P. Brady et al.

    Decomposition of interaction free energies in proteins and other complex systems

    J. Mol. Biol.

    (1995)
  • S. Boresch et al.

    The meaning of component analysis-decomposition of free-energy in terms of specific interactions

    J. Mol. Biol.

    (1995)
  • G. Archontis et al.

    Binding free energies and free energy components from molecular dynamics and Poisson–Boltzmann calculations. Application to amino acid recognition by aspartyl-tRNA synthetase

    J. Mol. Biol.

    (2001)
  • B.K. Jaitner et al.

    Discrimination of amino acids mediating Ras binding from noninteracting residues affecting raf activation by double mutant analysis

    J. Biol. Chem.

    (1997)
  • J. Janin

    Wet and dry interfaces: the role of solvent in protein–protein and protein–DNA recognition

    Struct. Fold. Des.

    (1999)
  • W.E. Stites

    Protein–protein interactions: interface structure. Binding thermodynamics, and mutational analysis

    Chem. Rev.

    (1997)
  • A. Wittinghofer et al.

    Ras—a molecular switch involved in tumor formation

    Angew. Chem., Int. Ed. Engl.

    (2000)
  • S. Li et al.

    A computer screening approach to immunoglobulin superfamily structures and interactions: discovery of small non-peptidic CD4 inhibitors as novel immunotherapeutics

    Proc. Natl Acad. Sci. USA

    (1997)
  • J.W. Tilley et al.

    Identification of a small molecule inhibitor of the IL-2/IL-2R receptor interaction which binds to IL-2

    J. Am. Chem. Soc.

    (1997)
  • Z. Guo et al.

    Designing small-molecule switches for protein–protein interactions

    Science

    (2000)
  • J. Zeng et al.

    Design of inhibitors of Ras–Raf interaction using a computational combinatorial algorithm

    Protein Eng.

    (2001)
  • M. Ohnishi et al.

    Selective inhibition of Ras interaction with its particular effector by synthetic peptides corresponding to the Ras effector region

    J. Biol. Chem.

    (1998)
  • J. Zeng

    Mini-review: computational structure-based design of inhibitors that target protein surfaces

    Comb. Chem. High Throughput Screen.

    (2000)
  • S. Jones et al.

    Principles of protein–protein interactions

    Proc. Natl Acad. Sci. USA

    (1996)
  • D.R. Davies et al.

    Interactions of protein antigens with antibodies

    Proc. Natl Acad. Sci. USA

    (1996)
  • N. Brooijmans et al.

    Stability of macromolecular complexes

    Proteins: Struct. Funct. Genet.

    (2002)
  • C.J. Tsai et al.

    Studies of protein–protein interfaces: a statistical analysis of the hydrophobic effect

    Protein Sci.

    (1997)
  • K.A. Dill

    Dominant forces in protein folding

    Biochemistry

    (1990)
  • L. Jin et al.

    Dissecting the energetics of an antibody–antigen interface by alanine shaving and molecular grafting

    Protein Sci.

    (1994)
  • T. Clackson et al.

    A hot spot of binding energy in a hormone–receptor interface

    Science

    (1995)
  • Z. Hu et al.

    Conservation of polar residues as hot spots at protein interfaces

    Proteins: Struct. Funct. Genet.

    (2000)
  • A.H. Elcock et al.

    Computer simulation of protein–protein interactions

    J. Phys. Chem. ser. B

    (2001)
  • T. Selzer et al.

    Rational design of faster associating and tighter binding protein complexes

    Nature Struct. Biol.

    (2000)
  • W. Wang et al.

    Biomolecular simulations: recent developments in force fields, simulations of enzyme catalysis, protein–ligand, protein–protein, and protein–nucleic acid noncovalent interactions

    Annu. Rev. Biophys. Biomol. Struct.

    (2001)
  • H.R. Bourne et al.

    The GTPase superfamily: conserved structure and molecular mechanism

    Nature

    (1991)
  • Cited by (1046)

    • Assessment of the bound conformation of bombesin to the BB1 and BB2 receptors

      2024, International Journal of Biological Macromolecules
    View all citing articles on Scopus
    View full text