The novel endocannabinoid noladin ether holds putative anti-fibrotic properties by selectively inducing cell death in hepatic stellate cells

Abstract

Background: We could previously demonstrate that the endogenous cannabinoid 2-arachidonoyl glycerol (2-AG) is an antifibrotic lipid mediator that blocks proliferation and induces apoptotic cell death in hepatic stellate cells (HSCs), but not in hepatocytes. Recently a novel endocannabinoid termed noladin ether (NE, 2-arachidonyl glyceryl ether) has been identified, which represents a more stable entity against enzymatic hydrolysis than 2-AG. Aims: To test the effects of NE on hepatic cell populations. Methods: Primary hepatocytes and HSCs were isolated from healthy rat or mouse liver by collagenase perfusion. Cell death was analyzed by LDH assay and western blot for caspase 3- and PARP-cleavage. Glutathione (GSH) was depleted by DL-buthionine-(S,R)-sulfoximine (BSO) treatment. Antioxidant resistance was enhanced by Trolox application. ROS formation was detected by DCFDA fluorescence. Membrane cholesterol was depleted by methyl-beta-cyclodextrin (MCD). Results: NE dose-dependently induced cell death in activated primary HSCs after 2–4h, starting from 2µM. In contrast to 2-AG, which induced late apoptosis in HSCs, NE induced hallmarks of necrosis, lack of caspase-3 or PARP cleavage. ROS formation in HSCs was significantly and even more increased by NE than by 2-AG. The antioxidant Trolox significantly decreased NE-induced death, thus indicating an ROS-dependent cell death mechanism. Although the cannabinoid receptors CB1, CB2 and TRPV1 were expressed in HSCs, pharmacological blockade failed to block NE-mediated cell death, indicating a cannabinoid-receptor independent mechanism. Interestingly, membrane cholesterol depletion with MCD completely inhibited NE-induced cell death. Similarly to 2-AG, NE did not induce cell death in hepatocytes. The resistance toward NE in hepatocytes was due to high levels of antioxidants such as GSH, since GSH depletion with BSO significantly increased NE-induced cell death in hepatocytes. Intriguingly, adenoviral overexpression of the putative endocannabinoid degradation enzyme fatty acid amide hydrolase (FAAH) in HSCs did not rescue from NE-induced cell death, underlining the stability of NE against enzymatic hydrolysis. Conclusion: NE effectively caused cell death in activated HSCs, but not in hepatocytes. NE-induced cell death was mediated through membrane cholesterol-rich lipid rafts and pronounced ROS formation, but independently from cannabinoid receptors. The selective induction of cell death in HSCs and its high stability against enzymatic hydrolysis proposes NE as a potential endogenous antifibrogenic mediator.