Introduction and Objective: The liver is a central regulator of metabolic homeostasis, which is disrupted in diabetes and MASLD. Commonly used in vitro cultures of primary hepatocytes rapidly lose key aspects of biochemical signaling and metabolic competence and fail to maintain multicellular interactions. Therefore, experimental models that preserve native liver architecture while enabling high experimental control and throughput are needed. Our objective was to develop a reproducible precision cut liver slices (PCLS) platform suitable for obesity research by supporting high levels of culture oxygenation. Methods: Cylindrical cores of mouse liver tissue were generated using manual 5-mm biopsy punches and stored on ice in oxygenated, pH-stabilized Krebs-Henseleit buffer prior to sectioning with a Krumdieck tissue slicer. Optimization included systematic evaluation of coring methods, cutting parameters, slice thickness and weight. Alternative preparation techniques involving fatty liver perfusion using low-gelling agarose were tested separately. To enhance slice culture oxygenation, a custom biaxial shaker capable of supporting a full-size commercial incubator was engineered and validated. Results: The optimized procedure yielded uniform slices with consistent thickness (~250 µm) and weight (4-6 mg). Standardization of workflow significantly reduced preparation related variability. A fully programmable biaxial shaker prototype was manufactured and validated to enable oscillations of up to 100 rpm, support loads of up to 275 lb, and allow continuous operation over multiple days. High oxygen tension in the cultures was achieved using a tri-gas incubator in combination with constant shaking at 90 rpm. Conclusion: Our PCLS protocol supports simultaneous targeted metabolomics and ex vivo metabolic assays, offering an efficient tool for studying hepatic metabolism in diabetes and MASLD. This approach may help identify new therapeutic targets for obesity-related liver disease. Disclosure M.M. Fydryszewski: None. J. Rajendran: None. S. Deja: None.
FYDRYSZEWSKI et al. (Fri,) studied this question.