Background: A shortage of donor livers for transplant has led to an increased interest in cell therapeutics as an alternative to whole organ transplant to treat end-stage liver disease. Primary human hepatocytes have been used in cell-based therapies. However, hepatocytes do not proliferate in vitro so it is challenging to grow enough cells for a successful transplant. Many have suggested using hepatocyte-like Adipose-derived mesenchymal Stromal/ stem Cells (ASCs) differentiated into hepatocyte-like cells as a substitute. Here we evaluate how closely these cells resemble primary hepatocyte cell morphology and function. Methods: Human ASCs were mechanically isolated from lipoaspirates. The stem cell nature of ASCs was characterized using flow cytometry and tri-lineage differentiation into osteocytes, adipocytes, and chondrocytes. ASCs were differentiated into hepatocyte-like cells in culture using various protocols that included combinations of growth factors and small molecules. Primary ASCs quickly attached and proliferated in vitro, forming a homogeneous spindle-like cell monolayer. Mesenchymal stem cells showed high expression of the markers CD73, CD90, CD271, CD44, CD166, CD105, and successfully differentiated into osteocytes, chondrocytes, and adipocytes. ASCs were cultured on type I collagen coated plates and differentiated into hepatocyte-like cells using 5 different protocols. Results: The ASCs differentiated into hepatocyte-like cells, using protocol C (induction with FGF4 and maturation with HGF, ITSPre, Dex, OncM and 2% serum), displayed a cuboidal morphology. Bioactivity assays demonstrated their ability to synthesize urea, uptake LDL, and metabolize glucose; all cardinal characteristics of hepatocytes, not present in undifferentiated ASCs. Gene expression analysis also showed the expression of several genes known to play an important role in liver function including, TDO2, ALB, HNF1B1, HNF6b, HNF4a, and AFP. However, even the best hepatocyte-like induced ASC obtained in this study had much inferior hepatocyte-related gene expression levels compared to primary human hepatocytes. Conclusion: We successfully differentiated ASCs into hepatocyte-like cells; Protocol C produced the best hepatocyte-like cells based on morphology and function typically seen in primary hepatocytes. Although the results displayed some hepatocyte-related function, comparison of bioactivity and gene expression of hepatocyte-like cells were drastically lower than those of primary human hepatocytes, suggesting that caution should be taken when considering using differentiated hepatocyte-like ASCs to replace hepatocytes. Further studies are needed to better understand the functional capacity of hepatocyte-like ASCs, and which specific metabolic function could potentially offer therapeutic applications.
