Background/Aim: Obesity and metabolic injury often result in the development of liver fibrosis. Despite extensive studies on liver fibrosis, new therapies are urgently needed. The progress in anti-fibrotic target discovery and drug development relies on translational studies, which are limited by the availability of all human liver cell types, and the methodologies to study high-quality 3D spheroids that mimic metabolically injured livers. In this study, we introduce our system using 3D human spheroids. It has broad applications and has been successfully used to study activity, interaction, signaling, and disease-associated crosstalk between various human liver cells.

 

Methods: Hepatic cells were isolated from donor livers (acquired from Organ Procurement Organizations). Livers were perfused, digested and cells isolated. Hepatocytes, non-parenchymal cells (NPCs), and Hepatic Stellate Cells (HSCs) were phenotyped and cryopreserved. For the 3D human liver spheroids, different human hepatic cells (5,300 cells per well/round-bottom ultra-low attachment plate) are combined in physiological ratios; 60% hepatocytes and 40% NPCs containing myeloid, T and B cells, natural killer cells, epithelial cells, and enriched HSCs. To replicate the complex hepatic microenvironment in the human liver, our lab has optimized 3D human liver spheroids by using a controlled growth factor-enriched defined media. Each of the liver cell types maintained a well-differentiated phenotype. To induce metabolic dysfunction-associated steatohepatitis (MASH) condition, we incubated the spheroids for an additional 7 days in the presence of a MASH conditioned media that contains a 1:1 mixture of oleate and palmitate (320 µM), 5.5 mM glucose, 10 mM fructose, 1 ng/ml TGF-β1, and 2 µg/ml LPS.

 

Results: Using complimentary single nucleus (sn)RNA- and snATAC-sequencing analysis of human NORMAL, MASL, and MASH HSCs, we identified a group of genes in which transcriptional and epigenetic activity was closely linked to regulation of extracellular matrix organization network of human HSCs in MASH livers. The role of selected genes in fibrogenic activation of human HSCs was investigated using 3D human liver spheroids composed of all liver cells. Human HSCs were transfected with dsi-negative (control) or dsiRNA for the target genes and used to generate human liver spheroids. 7 days after spheroid formation, the spheroids were exposed to MASH-conditioned media. The efficiency of the transfection was approximately 90%, and we could confirm the knockdown in spheroids after 14 days. It is critical for target identification, as it validates the responses of “MASH human livers in a dish”. We observed the increase of fibrogenic genes (COL1A1 and ACTA2) and the accumulation of lipid droplets in MASH spheroids. Knockdown of SERPINE1, SPON1, and LARP6 in human HSCs effectively suppressed the expression of COL1A1 and ACTA2 in MASH spheroids. We confirmed the decrease of fibrogenic markers both in mRNA and protein levels.

 

Conclusion: Our system of human liver spheroids mimics human livers in a dish. We successfully induced human MASH in the liver spheroids. In addition, we knocked down specific genes in human HSCs and used the spheroids to investigate their role in the pathogenesis of MASH. The spheroid system is a useful tool to translate mouse research into humans.