A free energy calculation study of the effect of H→F substitution on binding affinity in ligand–antibody interactions

Abstract

Changes in binding affinity to catalytic antibody 6D9 of chloramphenicol phosphonate derivatives (CPDs) containing H or F were investigated by performing free energy calculations based on molecular dynamics simulations. We calculated the binding free energy, enthalpy, and entropy changes (DeltaDeltaG, DeltaDeltaH, and -TDeltaDeltaS) attributable to H-->F substitution by comparing results for CPDs containing a trifluoroacetylamino group (CPD-F) or an acetylamino group (CPD-H). The calculated DeltaDeltaG, DeltaDeltaH, and -TDeltaDeltaS values were -2.9, -6.3, and 3.5 kcal mol(-1) and close to experimental values observed for a series of similar ligands, chloramphenicol phosphonates with F and H (-1.4, -3.5, and 2.1 kcal mol(-1)). Therefore, CPD-F binds more strongly to 6D9 than does CPD-H. To clarify the origin of the large difference in DeltaDeltaG, we apportioned the calculated values of DeltaDeltaG and DeltaG for the associated and dissociated states into contributions from various atomic interactions. We found that the H-->F substitution increased the binding affinity mainly by decreasing the hydration free energy and not by increasing favorable interactions with the antibody. The decreased hydration free energy of the ligand was mainly due to unfavorable coulombic interactions between the trifluoroacetylamino group and solvent waters, which increased the free energy of the dissociated state (by about 3.7 kcal mol(-1)). Also, the trifluoroacetylamino group slightly increased the free energy level of the associated state (about 0.8 kcal mol(-1)) because favorable van der Waals interactions compensated for unfavorable coulombic interactions with antibody atoms. In addition, the enthalpy and entropy changes, DeltaDeltaH and -TDeltaDeltaS (computationally -6.3 and 3.5 kcal mol(-1)), originated mainly from a decrease in hydration free energy in the dissociated state. The CPD-F and CPD-H ligands had substantially different structures in the dissociated and complexed states.

Contribution of the trifluoroacetyl group in the thermodynamics of antigen–antibody binding

“…The free energy calculations based on MD simulations were performed according to the same procedures as our previous simulation study of 6D9 by replacing 6D9 with 7C8 and the derivative of CP with CP (Saito et al, 2005). The initial structure of the associated state was prepared as follows.…”

“…One of the authors (MS) and a coworker showed that these simplifications caused serious calculation errors in the free energy values, although only a few atoms in the whole system consisting of tens of thousands of atoms were perturbed. Furthermore, they demonstrated that the free energy perturbation calculations based on MD simulations with long-range coulombic interactions and with all degrees of freedom remarkably decreased calculation errors and gave reliable free energy difference values (Saito and Tanimura, 1995;Horii et al, 2001;Saito, 2002;Saito and Sarai, 2003;Saito et al, 2005). On the other hand, there is no straightforward approach to precisely calculate the absolute binding free energy or the relative binding enthalpy and entropy of large systems such as a protein in water (Rodinger and Pomè s, 2005).…”

“…However, this concept does not always apply to ligand-protein interactions in general (Bö hm et al, 2004). We have recently shown that the binding affinity of an antibody, namely, 6D9, to its antigen chloramphenicol phosphonate (CP) is higher when it contains the trifluoroacetyl group (CP-F) than when it contains the acetyl group (CP-H) (Figure 1) (Saito et al, 2005). Interestingly, free energy calculations based on molecular dynamics (MD) simulations indicated that the increase in the binding affinity of the antigen upon substitution of H by F is mainly due to the decrease in the hydration free energy of the dissociated state and not due to the more favorable hydrophobic interactions between the antibody and the antigen, despite the binding of 6D9 to the trifluoroacetyl group in the hydrophobic pocket.…”

“…The ordinary free energy perturbation method cannot suppress large statistical errors in these thermodynamics quantities for the large systems in contrast to their relative free energies. However, in the previous studies, one of the authors (MS) and coworkers valuably discussed the binding thermodynamics by complementarily utilizing the relative binding enthalpy and entropy values calculated and the experimental values observed by calorimetry for the absolute binding free energy, enthalpy, and entropy (Saito and Sarai, 2003;Saito et al, 2005). Therefore, in this study, we compared DDH MD and TDDS MD with DDH ITC TDDS ITC on the binding thermodynamics of the 7C8 antibody to antigens, and discussed the effects of the F-to-H-substitution in the antigen, which can be applied in the rational design of ligands in biomolecular interactions.…”

Application of a polarizable force field to calculations of relative protein–ligand binding affinities

“…It is surprisingly difficult to assess the current state of the art in protein-ligand free-energy calculations. Unquestionably, a number of impressively accurate results have been published for both relative and more recently absolute binding free-energy calculations (6)(7)(8)(9)(10)(11)(12)(13)(14), with a general trend toward increased ambition and accuracy as computer power and theoretical understanding have increased. However, it is equally true that binding affinity calculations by simulation methods are not widely used in the pharmaceutical industry (3) and have not yet had the same impact that simpler docking and scoring methods and general molecular modeling have had.…”