Abstract 1320: Nanotopographical cues enhance glioblastoma migration: Harnessing biomimetic matrices and machine learning to uncover cancer invasion

Abstract Glioblastoma (GBM), the most aggressive brain cancer, invariably recurs after surgery and rapidly develops resistance to radiation therapy and chemotherapy. The invasiveness and genetic variation (heterogeneity) of GBM cells are major challenges for developing effective therapies. Furthermore, the study of glioblastoma invasion is particularly challenging due to the lack of robust experimental models that recapitulate the tumor microenvironment. To simulate in vivo brain tissue extracellular matrix (ECM), we have established nanotopographically defined extracellular matrix (ECM)-mimetic platforms. Moreover, to preserve the GBM cancer stem cell (GSC) phenotype and mimic tumor microenvironment, GBM cells were cultured on these nanopatterned surfaces in neuro stem cell media under hypoxia. The migration, motility, and morphological changes of the cells were analyzed using live cell microscopy and machine learning (ML). The migratory abilities of GBM cells were enhanced on these nanopatterned surfaces (cells elongate and move along the ridges) in comparison to unpatterned surfaces. Approximately 9 times more number of GBM cells migrated distances exceeding 400 micrometer along the direction of the ridges in the nanopatterned surface than the unpatterned surfaces. The MARS-Net ML analysis demonstrates that stromal nanotopography induces directional migration of GBM cells, and a 64% increase in skewness from rounded to more elongated cell shapes over time when compared to unpatterned surfaces. The genetic landscape of the GBM cells was characterized using whole exome sequencing (WES). About 55% of the single base substitutions (SBS) in GBM cells are missense mutations causing amino acid changes. The most prevalent SBS types are C>T/G>A, followed by T>C/A>G and C>A/G>T. We have developed radiation resistant GBM cells (rGBM) by expanding surviving cells after a series of cytotoxic x-ray treatments. rGBM cells exhibit higher migratory properties and colony formation abilities than normal GBM cells, implying they can be more invasive and oncogenic. Our results underscore the important role of nanotopographical cues in guiding GBM migration and invasiveness. To our knowledge, this study is the first to apply MARS-Net ML analysis to investigate changes in cell motility and morphology within an ECM-mimetic and hypoxic environment. The study has implications in the preclinical screening of anti-migratory chemotherapeutic agents and the molecular classification of glioblastoma. Citation Format: Christopher P. Miller, Satvik R. Kethireddy, Seung-Hyuk T. Lee, Patrick A. Ting, Paul Lazarchuck, Deok-Ho Kim, Raymond J. Monnat, Eun Hyun Ahn. Nanotopographical cues enhance glioblastoma migration: Harnessing biomimetic matrices and machine learning to uncover cancer invasion [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 1320.