Havelkova, J.; Petrenko, Y.; Stehlikova, A.; Marekova, D.; Peskova, K.; Pechar, M.; Studenovsky, M.; Etrych, T.; Pola, R.; Jendelova, P.

Introduction: In this study, we developed a modular in vitro platform that integrates advanced polymer-drug conjugation chemistry with stepwise cytotoxicity screening in both 2D (monolayer) and 3D (spheroids) glioblastoma (GBM) models. Buparlisib was selected as the model therapeutic agent due to its well-characterised mechanism of action, high blood-brain barrier permeability, and relevance to PI3K-targeted therapy. Methods: Two mechanistically distinct conjugation strategies were explored using N-(2-hydroxypropyl)methacrylamide-based copolymers (pHPMA). The first strategy was based on a redox-sensitive disulphide linkage designed for intracellular glutathione-triggered release, whereas the second used an azide-bearing derivative compatible with strain-promoted azide-alkyne cycloaddition. Drug release was assessed by high-performance liquid chromatography. Biological activity was systematically evaluated in U87MG, U118MG, and T98G cells under 2D conditions using a resazurin-based metabolic activity assay. Subsequently, the more promising disulphide-based formulations were assessed in 3D spheroids by metabolic activity measurements and live-cell monitoring of spheroid growth dynamics. Results: Free buparlisib showed the strongest inhibitory effect, while its modification and polymer conjugation reduced the apparent activity. Nevertheless, the disulphide-based derivative and polymer conjugate retained concentration-dependent activity, whereas the azide-based polymer conjugate showed minimal effects. Moreover, treatment responses differed between cell lines and between 2D and 3D models. Discussion: Overall, linker chemistry, cell-line specific behaviour, and model dimensionality strongly influenced the biological performance of the polymeric buparlisib formulations. The redox-sensitive polymer conjugate therefore represents the more promising strategy for further development.