P02-09

Computational assessment of the binding mode of Verteporfin, an inhibitor targeting the YAP-TEAD protein-protein interaction

Yurika IKEGAMI *1Genki KUDO2Takumi HIRAO1Ryunosuke YOSHINO3, 4Takatsugu HIROKAWA3, 4

1Degree programs in Comprehensive Human Sciences, Graduate School of Comprehensive Human sciences Doctoral Program in Medical Sciences, University of Tsukuba
2b Doctoral Program in Physics Degree Programs in Pure and Applied Sciences, Graduate School of Science and Technology, University of Tsukuba
3Transborder Medical Research Center, University of Tsukuba
4Division of Biomedical Science, Faculty of Medicine, University of Tsukuba
( * E-mail: s2230361@s.tsukuba.ac.jp )

Purpose:
Inhibiting overexpressed yes-associated protein (YAP) through YAP inhibitors is expected to improve the prognosis of patients with malignant tumor. Visudyne (Verteporfin: VP) has been reported to inhibit the formation of the YAP-TEA domain family member (TEAD) complex. Previous docking studies have suggested that VP bind to the WW domain of YAP or the TEAD interaction surface of YAP1,2). However, the precise binding poses have not been elucidated. Additionally, it remains unclear which of the four VP isomers is the most effective against YAP. In this study, we re-evaluate the interaction between VP and YAP using in silico techniques such as docking simulation and AlphaFold2(AF2), a highly accurate protein structure prediction method. Furthermore, we aim to identify the isomer with the strongest inhibitory activity against YAP to enhance its efficacy as an adjuvant-like drug.
Methods:
We performed the in silico techniques in the following steps; (1) The YAP-TEAD complex was modeled using ColabFold, (2) Docking simulations were conducted between the four VP isomers, named Ia-1, Ia-2, Ib-1, and Ib-2 and YAP using Glide, and (3) The binding pose clusters were analyzed. The docking sites were determined at the YAP-TEAD interaction interface and the druggable pocket, generating two grids at the docking sites. The docking results were rescored using binding free energy calculations using MM/GBSA method. Clustering of the binding modes was performed using Protein-Ligand Interaction Fingerprint (PLIF).
Results:
As the results of docking calculation and clustering analysis, the binding poses were classified into nine clusters. The Ia-2 isomer showed the lowest binding free energy. The pose generated from the grid centered on the druggable pocket formed hydrogen bonds with Gln200, Thr197, and Lys254; however, since these residues are not involved in protein-protein interactions (PPI) with TEAD, PPI inhibition did not occur. On the other hand, another pose generated from the grid centered on the YAP-TEAD interaction interface formed hydrogen bonds with Met86 and Arg87 of YAP. This result suggests that VP competitively binds to Met86 in place of TEAD residues. Therefore, this binding pose of Ia-2 is considered the mechanism by which the YAP-TEAD interaction is inhibited.
Conclusions:
Ia-2 tends to form more stable hydrogen bonds with Arg and Lys in YAP compared with the other isomers. The combination of carboxyl methyl chirality in Ia-2 creates a stable binding pose with Met86 and Arg87 of YAP. The binding mode suggested by in silico analysis indicates that if VP was to be used as an anti-cancer drug in the future, administering the isolated Ia-2 might be more effective than administering the Visudyne which is a mixture of the four VP isomers.

References:
(1) Kandoussi I. et al. Bioinformation. 2017 Jul 31;13(7):237-240.
(2) Wei C & Li X. Mol Med Rep. 2020 Nov;22(5):3955-3961.