Protein:Ligand binding free energies: A stringent test for computational protein design

Abstract

A computational protein design method is extended to allow Monte Carlo simulations where two ligands are titrated into a protein binding pocket, yielding binding free energy differences. These provide a stringent test of the physical model, including the energy surface and sidechain rotamer definition. As a test, we consider tyrosyl-tRNA synthetase (TyrRS), which has been extensively redesigned experimentally. We consider its specificity for its substrate l-tyrosine (l-Tyr), compared to the analogs d-Tyr, p-acetyl-, and p-azido-phenylalanine (ac-Phe, az-Phe). We simulate l- and d-Tyr binding to TyrRS and six mutants, and compare the structures and binding free energies to a more rigorous "MD/GBSA" procedure: molecular dynamics with explicit solvent for structures and a Generalized Born + Surface Area model for binding free energies. Next, we consider l-Tyr, ac- and az-Phe binding to six other TyrRS variants. The titration results are sensitive to the precise rotamer definition, which involves a short energy minimization for each sidechain pair to help relax bad contacts induced by the discrete rotamer set. However, when designed mutant structures are rescored with a standard GBSA energy model, results agree well with the more rigorous MD/GBSA. As a third test, we redesign three amino acid positions in the substrate coordination sphere, with either l-Tyr or d-Tyr as the ligand. For two, we obtain good agreement with experiment, recovering the wildtype residue when l-Tyr is the ligand and a d-Tyr specific mutant when d-Tyr is the ligand. For the third, we recover His with either ligand, instead of wildtype Gln.

Redesigning the stereospecificity of tyrosyl-tRNA synthetase

“…However, in an earlier attempt to increase AsnRS activity for aspartate, selection for Asp/Asn specificity or Asp affinity led to similar designs. Detailed analysis of the present design simulations and errors is left to another article . The main structural differences between the CPD models and the MD simulations have to do with the motion of d ‐Tyr and Asp41 in the MD, described above (when comparing the X‐ray and MD structures).…”

“…For each pair and combination, we compute an interaction energy, which is stored in an “energy matrix” The energy includes a contribution from the protein and an l ‐ or d ‐Tyr ligand, described by molecular mechanics, from the solvent, described implicitly, and from the unfolded state. The unfolded state is described with a simple “tripeptide” model, plus an empirical correction that leads to a reasonable overall amino acid composition . To alleviate the rotamer approximation, each interaction energy is computed after a short energy minimization, of conjugate gradient steps, where only the pair of interest can move and the energy is limited to the pair interaction plus the interaction with the backbone .…”

“…This CPD round led to our set of candidate mutations, which were studied with experiments and MD simulations. A second round using an improved energy function was done recently and extended here. It employed the Amber protein force field and a Generalized Born solvent model .…”

“…A second round using an improved energy function was done recently and extended here. It employed the Amber protein force field and a Generalized Born solvent model . For this round, we used either the TyRS X‐ray structure or a structure sampled during MD for the TyrRS: d ‐Tyr complex, which differs in the d ‐Tyr position (see below).…”

Computational protein design for given backbone: recent progresses in general method-related aspects

“…Such molecular simulation-derived energies, retrained using quantitatively measured experimental data if possible, may be used as an effective metric for assessing sequences proposed through CPD. Also based on molecule mechanics energies with implicit solvation, Druart et al proposed a constant-activity Monte Carlo approach by which one or several residues at a binding pocket can be redesigned to achieve desired relative binding free energies for different ligands [42].…”