Computational identification and mechanistic characterization of natural product binders targeting the PDE6D prenyl binding tunnel

RAS oncogenesis remains a significant clinical challenge due to the difficulty of directly targeting RAS proteins. PDE6D, a prenyl-binding chaperone involved in the RAS membrane trafficking pathway, represents an indirect yet tractable target whose modulation has been proposed to influence RAS localization and signaling. This study employed a multiscale structure-based in silico workflow to identify natural compounds with putative binding potential toward the prenyl-binding tunnel of PDE6D. A curated natural product library was screened using molecular docking, followed by density functional theory–based geometry optimization. Additional 500 ns molecular dynamics simulations were then used to analyze the stability of binding and the overall conformational behavior of the most promising complexes. Further PCA and free energy landscape mapping highlighted distinct low-energy conformational states adopting unique structural transitions, where the MolPort-039-052-621 complex presented the most compact and well-defined low-energy basin. In that respect, superimposing free energy minima with the respective initial docked poses demonstrated minimal structural deviation from the binding orientation during dynamic evolution. Finally, QM/MM calculations qualitatively described the electronic stabilization of the ligands within the PDE6D environment. The present study identifies natural compounds with computationally favourable tunnel-binding characteristics and provides mechanistic insights that may guide future experimental validation.