A chemically optimized, GalNAc-conjugated siRNA targeting HSD17B13 demonstrates efficacy in a human 3D organoid model of MASH, showcasing an alternative to animal testing.

Abstract

Animal models have limited ability to predict human metabolically dysfunction-associated steatohepatitis (MASH), creating a major obstacle in drug development. To address this challenge, we introduce an integrated drug discovery approach that combines rational small-interfering RNA (siRNA) design and chemical modifications to improve stability, validated in a complex human 3D NAC-organoid MASH model. Using this system, we created new GalNAc-conjugated siRNAs targeting the 17β-hydroxysteroid dehydrogenase 13 (HSD17B13). The lead candidates were evaluated for stability in human serum and liver microsomes, off-target risks, and efficacy in the 3D organoid model that mimics key disease features such as steatosis and fibrosis. Our main candidate, si-R5-42, demonstrated greater stability in human serum and liver microsomes, as well as fewer off-target risks. Most importantly, in the 3D organoid model, si-R5-42 successfully reduced disease markers. Its ability to lower hepatic steatosis and fibrogenesis was comparable to that of the clinical-phase candidate ARO-HSD. This research provides a promising therapeutic candidate and a solid, human-relevant preclinical siRNA testing framework, reducing dependence on animal models.