Fig. 4: V991W and T998W stabilize HexaPro-SS-Δstalk in the closed conformation.

a Thermodynamic stabilization of the closed prefusion conformation as imparted by each tryptophan variant relative to the base system (HexaPro-SS-Δstalk). For each variant, the value of ∆∆G_{mutation-folding} relative to the base construct is plotted with a colored bar indicating the free energy difference estimate (in kcal/mol) ± standard error as calculated from 1000 (forward and backward) non-equilibrium alchemical simulations. The standard errors, denoted by thin black lines, were calculated via bootstrapping with 100 drawn samples. b Distribution of the number of interprotomer contacts established per frame by residues at positions 991 + 998 from the ensemble of closed conformations extracted from respective WE simulations. Distributions are shown as kernel densities. c Relative frequency of occurrence (%) of the top six residue–residue interprotomer contacts occurring in the closed conformations observed in HexaPro-SS-Δstalk and HexaPro-SS-2W WE simulations. Relative frequencies are calculated with respect to the total number of closed conformations extracted from the respective WE simulation. d, e Molecular representation of the most important interprotomer contacts in HexaPro-SS-2W in (d) the closed conformation and (e) the partially open conformation. Chain-A is depicted in light purple, chain-B in purple, chain-C in dark purple, and interacting residues are highlighted with sticks. Solid lines connect residues that form hydrophobic (π–π) interactions, whereas dashed lines indicate electrostatic (cation–π and salt bridge) interactions. A representative closed conformation of HexaPro-SS-2W was selected for this purpose upon clustering of closed conformations retrieved from the respective WE simulation (Supplementary Fig. 18).