Fig. 3: Unbiased coarse-grained (CG) molecular dynamics simulations of inhibitors binding to ECF transporter.
From: Identification of inhibitors targeting the energy-coupling factor (ECF) transporters
a Crystal structure of ECF FolT2 (PDB ID 5JSZ), in complex with EcfA and EcfA’ shown in blue and red, EcfT in white, and S-component in gray. Pocket P2, used in the structure-based virtual screening campaign is highlighted in yellow. Compound 1 densities obtained from the simulations are shown in green, which indicates the existence of two other pockets, P9 and the entrance of P11 (pocket names according to Supplementary Fig. 1). The green isosurfaces correspond to regions with high occupancy (100 times higher than compound 1 density in the membrane). As compound 1 binds only with low affinity to P2, no high occupancy density is observed. The figure also shows bar plots with calculated binding free energies (ΔGbind) of compounds 13 and 14 for pockets P2, P9, and P11 in relation to water and membrane. In the case of compound 13 in pocket P2, the binding events are so reduced that it is only possible to estimate the upper limit (indicated by “≥” in the plots). For compound 14, as the affinity for P2 is very low, the number of frames in which the ligand stays bound is low compared to the other binding sites. As a consequence, the error bars are larger in this case. Estimates of the partitioning free energies of the inhibitors between water and lipids (ΔGlipids/water) are also shown in the figure. The error bars represent the mean absolute error obtained by the block-averaging approach. b, c Representative poses of compounds 1 and 14 inside pocket P9, respectively, backmapped from the CG to the atomistic resolution. Some residues from EcfT (salmon) and S–component (gray) are highlighted, including the hydrogen-bond network (black dashed lines). Hydrogen atoms have been omitted for clarity.
