講演情報
[17a-PB3-9]A Machine Learning Framework for Deciphering Ferroelectric Surface-Cell Membrane Interactions
〇ALEXIS BOROWIAK1,2, YOHEI KONO1,2, TAKESHI SHIMI1,2, TAKESHI FUKUMA1,2 (1.Kanazawa Univ., 2.NanoLSI)
キーワード：
Machine Learning、Biomechanics、Ferroelectric surface
This study presents a biocompatible, circuit-free approach employing ferroelectric materials to generate localized electric fields, thereby mitigating biocompatibility concerns associated with conventional metallic electrodes in contact with biological tissues. We investigate cellular electromechanics by culturing HeLa cells engineered with a Fluorescent Ubiquitination-based Cell Cycle Indicator (FUCCI) on lithium niobate crystals with distinct polarization domains.Utilizing Atomic Force Microscopy (AFM) with Quantitative Imaging (QI), Piezoresponse Force Microscopy (PFM), and fluorescence microscopy, combined with a machine learning approach, we examined cells in the G1 phase. We acquired over 100,000 force-distance measurements using commercial AFM tips to quantify cellular elasticity.Our results demonstrate that the spontaneous polarization of the ferroelectric substrate profoundly alters the local ionic environment, subsequently affecting cellular mechanical properties. To analyze this high-dimensional dataset, machine learning algorithms—including supervised and unsupervised methods—were employed to uncover subtle relationships between substrate polarization states and cellular elasticity metrics. The data reveal that ferroelectric surface polarization modulates cell membrane electromechanical behavior, likely through the redistribution of ionic species near the surface. These findings contribute to the development of ML-driven, biocompatible platforms for the precise control and characterization of cellular mechanics in bioelectric research.