Organobodies: a robust and size-controllable system for generating scalable hiPSC-derived liver organoids for drug toxicity screening

Abstract

Human hepatic organoids derived from pluripotent or adult stem cells offer powerful platforms for disease modeling and drug discovery. However, developing robust and scalable organoids capable of sustaining long-term functionality remains challenging. Here, we developed a novel, semi-defined approach using a self-assembling peptide and collagen I to create highly uniform human induced pluripotent stem cell-derived hepatic organoids in droplet format, which we term hepatic organobodies (OBs). This method enabled rapid, reproducible production of threedimensional (3D) liver tissues, which remained structurally, metabolically, and functionally stable for several weeks. OBs adopted hallmark hepatic morphology and expressed key hepatocyte genes, several at levels approaching freshly isolated primary human hepatocytes (PHHs). OBs secreted substantially higher albumin and A1AT compared with parallel two dimensional cultures, and transcriptomic profiling revealed marked enhancement of hepatic maturation, including elevated expression ofCYP3A4, CYP2C9, andCYP1A2, and enrichment of PPAR signaling and fatty acidβ-oxidation pathways. Additionally, OBs exhibited drug metabolizing activity comparable to classical Matrigel-based organoids and demonstrated CYP3A4 and CYP2C9 activities comparable to the 'gold standard' 3D PHH microtissues. Critically, OBs accurately predicted hepatotoxicity of more than 10 reference compounds, outperforming HepG2 cells and matching PHH-based benchmarks. Overall, we present OBs, a novel, and scalable 3D liver model that delivers advanced maturation and robust metabolic function. This platform offers a powerful and reproducible alternative to existing organoid systems as it avoids animal-derived, undefined matrices such as Matrigel, requires no specialized equipment, and relies on rapid self-curation of the hydrogel triggered by physiological salt concentrations, making the process fast, reproducible, broadly accessible, and scalable.