Functional characterization of PPAR targets HILPDA and SLC25A47

Abnormal lipid metabolism has been associated with a wide range of chronic and infectious diseases including non-alcoholic fatty liver disease, viral hepatitis C infection, atherosclerosis, diabetes, and cancer. Peroxisomal proliferator activator receptor alpha (PPARα) is a ligand-activated transcription factor that plays a major role in the regulation of lipid metabolism. Gene expression studies performed on PPARα null mice have shed light into a variety of genes regulated by PPARα. A comprehensive understanding of the physiological and molecular function of PPARα target genes is needed for the accurate development of therapeutical strategies. The research contained in this thesis first sought to improve our current knowledge on the transcriptional regulation by peroxisome proliferator-activated receptors (PPAR)-α activation on human liver in vivo using a novel humanized-liver mouse model. Then, we pursued to expand our understanding on the physiological and molecular function of two PPARα target genes, Hilpda and Slc25a47. We studied the function of HILDPA in adipose tissue macrophages (ATM) in the context of obesity-induced inflammation and in hepatocytes during non-alcoholic steatohepatitis (NASH). Ultimately, we sought to identify the physiological function of the liver-specific mitochondrial carrier Slc25a47 in liver lipid metabolism and energy expenditure. The role of PPARα in gene regulation in mouse liver is well characterized. However, less is known about the effect of PPARα activation in human liver in vivo. Model systems to study PPARα in human liver vary from hepatoma cell lines, human primary hepatocytes, human precision cut liver slices, and mice expressing human PPARα. A novel model to study in vivo PPARα activation in human liver is a chimeric mouse carrying human liver cells. These mice are generated by transplanting human hepatocytes into albumin enhancer–driven urokinase-type plasminogen activator transgenic/severe combined immunodeficiency (uPA/SCID) mice, leading to replacement of the host hepatocytes. In this research, we performed transcriptomics analysis on the effect of fenofibrate in mice with hepatocyte-humanized livers and compared the results with other relevant transcriptomics datasets. We observed that human hepatocytes exhibited excessive lipid accumulation. Fenofibrate increased the size of the mouse but not human hepatocytes and tended to reduce steatosis in the human hepatocytes. Quantitative PCR indicated that induction of PPARα targets by fenofibrate was less pronounced in the human hepatocytes than in the residual mouse hepatocytes. Comparison with other transcriptomics datasets indicated that hepatocyte humanized livers recapitulate the principal effects of PPARα activation on lipid metabolism as revealed by other model systems of human liver. In contrast, pathways connected to DNA synthesis were downregulated by fenofibrate in chimeric mice with hepatocyte humanized livers yet upregulated by fenofibrate in normal mouse livers. Our results support the major role of PPARα in regulating hepatic lipid metabolism, highlight the more modest effect of PPARα activation on gene regulation in human liver compared to mouse liver, and
indicate that PPARα may have a suppressive effect on DNA synthesis in human liver. In the second part of this thesis, we characterized the physiological and molecular function of HILDPA in ATM in the context of obesity-induced inflammation and in hepatocytes during NASH. To this end we used a HILPDA tissue-specific knockout mice model in macrophages and hepatocytes generated by using LysM-Cre and Alb-Cre transgenic mice, respectively. In diet-induced obese mice, HILPDA deficiency in macrophages markedly reduced lipid accumulation in macrophages yet it did not alter any measured inflammatory or metabolic
parameters. Mechanistically, HILPDA acts an inhibitor of ATGL-mediated lipolysis in macrophages. Treatment with the ATGL inhibitor Atglistatin rescued lipid accumulation inside lipid droplets in HILPDA-deficient macrophages. This research questions the contribution of lipid droplet accumulation in adipose tissue macrophages in obesityinduced inflammation and metabolic disfunction. Similarly, in diet-induced NASH mice, HILPDA hepatocyte deficiency modestly yet significantly reduced liver triglyceride accumulation and plasma ALT levels. However, expression of macrophage/inflammatory markers and fibrosis were not different between HILPDA knockout and floxed mice. In hepatoma cell lines, fatty acids increase Hilpda expression and protein levels. Hilpda overexpression in turn induces triglyceride accumulation inside lipid droplets. Mechanistically, HILPDA interacts and increases DGAT1 protein level and activity as indicated by FRET-FLIM analysis, western blot and Dgat1 activity assay. These findings propose a novel regulatory mechanism by which fatty acids promote triglyceride synthesis and storage. In the last part of this thesis, we sought to characterize the liver-specific mitochondrial carrier SLC25A47. We identified Slc25a47 from a transcriptome data analysis of three independent studies in which human primary hepatocytes were treated with the PPARα agonist Wy14643 or GW7647 for 24 hours. Slc25s47 was consistently induced by PPARα agonists in the three datasets. To study the functional role of SLC25A47 in mouse liver, we overexpressed Slc25a47 in liver using adeno-associated virus. We first studied the effect of Slc25a47 overexpression in fed and fasted mice. No significant effects were observed on liver triglyceride levels, nor on plasma glucose, triglyceride, cholesterol, glycerol, nonesterified fatty acids and β-hydroxybutyrate levels. Similarly, after feeding mice a high fat diet for 8 weeks, we did not observe any effect of Slc25a47 overexpression in any of the beforementioned parameters. However, transcriptome analysis at the pathway level showed that several gene sets related to cholesterol synthesis were significantly enriched among the upregulated genes, suggesting that SLC25A47 might stimulate cholesterol synthesis. Next, we reasoned a knockout model would provide a clearer phenotype than the overexpression given that Slc25a47 is already expressed at high levels in liver. We next performed indirect calorimetry in wildtype and Slc25a47 knockout mice in the fed and fasted state yet there was no difference in energy expenditure, activity level nor respiratory exchange ratio between genotypes. We subsequently placed the WT and Slc25a47-/- mice on a high fat diet for 20 weeks to induce obesity and insulin resistance and another cohort of WT and Slc25a47-/- mice on a low fat diet for the same duration. No significant differences were observed in body weight gain, food intake, liver or white adipose tissue weight, either on the low fat diet or the high fat diet. By contrast, glucose tolerance was significantly improved in the Slc25a47-/- mice compared to the WT mice. Similarly, no significant differences were observed between WT and Slc25a47-/- mice in hepatic triglyceride nor in plasma glucose, cholesterol, triglycerides, glycerol and non-esterified fatty acids in either condition. At the molecular level, we explored if SLC25A47 might led to mitochondria uncoupling as previously reported. However, we did not observe and change in mitochondria respiration in mice liver overexpressing Slc25a47, nor in transiently transfected Hepa 1-6 cells as measured by high-resolution respirometry with Oroboros. However, image intensity analysis showed a significant decrease on the intensity of Mitrotrack Red, a membrane-potential dependent dye, on Slc25a47 overexpressing mitochondria. This research does not support an uncoupling role of SLC25A47 yet it identifies SLC25A47 as novel PPARα-regulated gene in human and mouse hepatocytes. Further studies are needed to identify SLC25A47 function in liver metabolism.