Non-alcoholic fatty liver disease (NAFLD) is the most common liver disorder in western countries. Despite the high prevalence of NAFLD, the underlying biology of the disease progression is not clear, and there are no approved drugs to treat non-alcoholic steatohepatitis (NASH), the most advanced form of the disease. Thus, there is an urgent need for developing advanced in vitro human cellular systems to study disease mechanisms and drug responses. We attempted to create an organoid system genetically predisposed to NAFLD and to induce steatosis and fibrosis in it by adding free fatty acids. We used multilineage 3D spheroids composed by hepatocytes (HepG2) and hepatic stellate cells (LX-2) with a physiological ratio (24:1). HepG2 and LX-2 cells are homozygotes for the PNPLA3 I148M sequence variant, the strongest genetic determinant of NAFLD. We demonstrate that hepatic stellate cells facilitate the compactness of 3D spheroids. Then, we show that the spheroids develop accumulations of fat and collagen upon exposure to free fatty acids. Finally, this accumulation was rescued by incubating spheroids with liraglutide or elafibranor, drugs that are in clinical trials for the treatment of NASH. In conclusion, we have established a simple, easy to handle, in vitro model of genetically induced NAFLD consisting of multilineage 3D spheroids. This tool may be used to understand molecular mechanisms involved in the early stages of fibrogenesis induced by lipid accumulation. Moreover, it may be used to identify new compounds to treat NASH using high-throughput drug screening.

There is a high unmet need for developing treatments for nonalcoholic fatty liver disease (NAFLD), for which there are no approved drugs today. Here, we used a human in vitro disease model to understand mechanisms linked to genetic risk variants associated with NAFLD. The model is based on 3D spheroids from primary human hepatocytes from five different donors. Across these donors, we observed highly reproducible differences in the extent of steatosis induction, demonstrating that inter-donor variability is reflected in the in vitro model. Importantly, our data indicates that the genetic variant TM6SF2 E167K, previously associated with increased risk for NAFLD, induces increased hepatocyte fat content by reducing APOB particle secretion. Finally, differences in gene expression pathways involved in cholesterol, fatty acid and glucose metabolism between wild type and TM6SF2 E167K mutation carriers (N = 125) were confirmed in the in vitro model. Our data suggest that the 3D in vitro spheroids can be used to investigate the mechanisms underlying the association of human genetic variants associated with NAFLD. This model may also be suitable to discover new treatments against NAFLD.