Computational Characterization of Multistep Opening Mechanism and Heterogeneous Closing Pathways of Periplasmic Binding protein Ps_PtxB

Abstract Periplasmic binding proteins (PBPs) undergo large-scale domain opening and closing during substrate recognition, capture, and delivery. Although ligand-binding mechanisms in PBP have been extensively discussed in terms of conformational selection and induced fit, less is known about how a ligand-depleted closed conformation opens and whether opening and closing transitions follow corresponding conformational routes. Here, we used the probable phosphite transport system-binding protein PtxB from Pseudomonas stutzeri WM88 (Ps_PtxB) to investigate the transition between the ligand-removed closed structure (closed-apo) and the open apo structure. By combining conventional molecular dynamics (CMD), targeted molecular dynamics(TMD), and umbrella sampling, we reconstructed a TMD-derived opening pathway and characterized metastable conformational regions along the transition. The results suggest that Ps_PtxB opening proceeds through a multistep process involving early Lobe1-Lobe2 interface decoupling, hinge-associated stabilization, twisting rearrangement, and subsequent coordinated domain motion. Residue-interaction network and dynamic cross-correlation matrices (DCCM) analyses further revealed stage-dependent changes in local contacts, network topology, and interdomain coupling. Based on these dynamic-coupling features, DCCM-selected interface hotspots were used as targeted MSA-masking sites for AlphaFold3 conformational sampling, which expanded the sampled conformational range compared with random masking under controlled MSA conditions. In addition, elevated-temperature CMD simulations initiated from the open-apo state revealed at least two representative closing-like pathways with different conformational overlap and dynamic-coupling patterns. Together, these results provide a pathway-level view of Ps_PtxB opening and closing dynamics and suggest that molecular simulation-derived dynamic information can guide interpretable conformational exploration.