Date Published: June 30, 2017
Publisher: Public Library of Science
Author(s): Vincenzo Prestigiacomo, Anna Weston, Simon Messner, Franziska Lampart, Laura Suter-Dick, Partha Mukhopadhyay.
Currently most liver fibrosis research is performed in vivo, since suitable alternative in vitro systems which are able to recapitulate the cellular events leading to liver fibrosis are lacking. Here we aimed at generating a system containing cells representing the three key players of liver fibrosis (hepatocyte, Kupffer cells and stellate cells) and assess their response to pro-fibrotic compounds such as TGF-β1, methotrexate (MTX) and thioacetamide (TAA).
Human cell lines representing hepatocytes (HepaRG), Kupffer cell (THP-1 macrophages) and stellate cells (hTERT-HSC) were co-cultured using the InSphero hanging drop technology to generate scaffold-free 3D microtissues, that were treated with pro-fibrotic compounds (TGF-β1, MTX, TAA) for up to 14 days. The response of the microtissues was evaluated by determining the expression of cytokines (TNF-α, TGF-β1 and IL6), the deposition and secretion of ECM proteins and induction of gene expression of fibrosis biomarkers (e.g. αSMA). Induction of Nrf2 and Keap1, as key player of defence mechanism, was also evaluated.
We could demonstrate that the multicellular 3D microtissue cultures could be maintained in a non-activated status, based on the low expression levels of activation markers. Macrophages were activated by stimulation with LPS and hTERT-HSC showed activation by TGF-β1. In addition, MTX and TAA elicited a fibrotic phenotype, as assessed by gene-expression and protein-deposition of ECM proteins such as collagens and fibronectin. An involvement of the antioxidant pathway upon stimulation with pro-fibrotic compounds was also observed.
Here, for the first time, we demonstrate the in vitro recapitulation of key molecular and cellular events leading to liver fibrosis: hepatocellular injury, antioxidant defence response, activation of Kupffer cells and activation of HSC leading to deposition of ECM.
Liver fibrosis and cirrhosis are canonical endpoint of many chronic liver diseases, including virus infections (HBV, HCV), non-alcoholic steatohepatitis or damage due to alcohol consumption . In addition, liver fibrosis is also a relevant toxicological outcome and has been identified as an Adverse Outcome Pathway (AOP), a novel tool in human risk assessment designed to provide mechanistic representation of critical toxicological effects [2,3]. Liver fibrosis is characterized by an accumulation of fibrillar extracellular matrix (ECM), leading to liver failure, portal hypertension, and increased risk of cancer. The pathophysiology of fibrosis requires chronic liver damage (including chronic alcohol consumption, chemically-induced hepatocyte damage, and viral infections) and involves the interplay of several hepatic cell types; it requires hepatocyte injury and cell death, activation of Kupffer cells (KC), activation of hepatic stellate cells (HSC), and chronic inflammation [4,5].
In this work, we provide experimental evidence demonstrating that multicellular MTs generated with well characterized human cell lines can recapitulate the key cellular and molecular events leading to hepatic fibrosis. We reproducibly generated human liver MT using HepaRG, hTERT-HSC and THP-1, three human cell lines representing hepatocytes, macrophages and stellate cells. The use of stellate cells in culture has been attempted previously but often not considered a good option for the study of fibrosis in vitro due to the fact that stellate cells often undergo spontaneous activation when grown on plastic dishes . This is mainly due to the physical properties of cell culture dishes, which have a tissue tension greater than that of the fibrotic/cirrhotic liver (20 KPa) and much larger than that of normal liver tissue (5 KPa) . hTERT-HSC have been reported to revert to a more quiescent status when cultured on extracellular matrix components . In our hands, however, hTERT-HSC showed low levels of expression of the stellate cell activation marker αSMA before induction with TGF-β1. In both 2D culture and in the scaffold-free 3D MT, hTERT-HSC responded to pro-fibrotic stimuli such as LPS, TNF-α, TGF-β1 and pro-fibrotic compounds (MTX and TAA) by attaining an activated status. In the scaffold-free 3D liver MTs that we generated, the hTERT-HSC are kept in a more physiological environment, surrounded by other relevant cell types (HepaRG and THP1) as shown by IHC staining for αSMA and vimentin. In addition, the co-culture of HSC with hepatocytes and Kupffer cells is a model that incorporates all three cell types involved in the AOP leading to fibrosis, and is therefore very well suited to recapitulate cell-cell interactions leading to fibrosis in vitro. The results we obtained after stimulation with TGF-β1 clearly demonstrate that the hTERT-HSC cell line is able to respond to well established pro-fibrotic stimuli by increasing the expression of the key factors: the activation marker αSMA and ECM proteins (collagen I and IV, fibronectin, and CD44). Similarly, in our hands differentiated THP-1 macrophages served as an excellent surrogate for KC, as they were able to produce cytokines (in particular TNF-α) upon stimulation with LPS in both 2D and 3D cultures. In our system, we used HepaRG cells as equivalent of hepatocytes. These cells are known to display many characteristics of human hepatocytes  including retention of metabolic activity . Similarly to reported data with primary murine hepatocytes , HepaRG MTs responded to LPS and TNF-α by increasing transcription of IL6, showing for the first time that also HepaRG cells are able to produce IL6 following injuring stimuli. HepaRG were also able to produce albumin in the 3D culture, indicating a functional phenotype up to three weeks in culture. However, albumin was significantly decreased after treatments with MTX and TGF-β1 suggesting hepatocytes damage and onset of fibrosis in the microtissues. The higher biological relevance of this model system allows studying responses in an integrated biological system with intricate crosstalks between the main contributing cell types to fibrosis. For example, biological response to LPS stimulation is only possible in presence of inflammation responsive cells (THP-1) and resulted in high activation of HSCs. This would not be possible in conventional monocultures. IHC staining for vimentin (as marker for NPCs) shows a physiological liver cell ratio of Hep/NPCs with approximately 80% hepatocytes, 20% NPCs including inflammatory cells and stellate cells prior treatments. Thus, we were able to generate a complex cellular model system with cell types and ratios which mimic in vivo liver organization. Moreover, with such model systems, it is for the first time possible to study in detail the contribution of the different cell types on induction and progression of fibrosis in an in vitro model system.
Summarizing, in this study, we have demonstrated that a 3D-liver MT co-culture containing HepaRG, THP-1 and hTERT-HSC is able to recapitulate the known cellular events leading to the fibrotic phenotype .