Theses Doctoral

Physiology and Pathophysiology of Retinoid and Lipid Storage in Mouse Hepatic Stellate Cell Lipid Droplets

D'Ambrosio, Diana N.

Retinoids are important mediators of many physiological processes in the body, including vision, reproduction, embryonic development, immunity and bone growth. Thus, the storage and metabolism of retinoids in the body has immediate implications for the overall health and metabolic homeostasis of the animal. This thesis research focused on two retinoid metabolites: retinyl ester, the form in which retinoids are stored, and retinoic acid, the transcriptionally active retinoid metabolite. Approximately 70% of retinoid in the body is stored in the liver, and, of this fraction, 80-90% is stored in the hepatic stellate cell (HSC) lipid droplets as retinyl ester. These lipid droplets are a distinguishing feature of the HSC, and they have recently been proposed to be specialized organelles for the storage of retinoid based on their unique retinoid content and responsiveness to dietary retinoid status. It is also known that the ability to synthesize and store retinyl ester in HSCs is necessary for the presence of HSC lipid droplets. Interestingly, it is well established that, with the progression of liver disease in human patients, there is a progressive loss of total hepatic retinoid content. As hepatic disease progresses, the HSCs transition from a quiescent to an activated phenotype, accompanied by the loss of their lipid droplet and retinoid content. The ultimate goal of this dissertation was to further elucidate the factors that regulate HSC retinoid storage as retinyl esters in lipid droplets and to define the factors that regulate HSC lipid droplet genesis and dissolution. The first aim of this research was to investigate the heterogeneity of HSCs and their lipid droplets in healthy, uninjured liver. Our observations suggest that the HSC population in a healthy, uninjured liver is heterogeneous. One subset of the total HSC population, which expresses early markers of HSC activation, may be primed and ready for rapid response to acute liver injury. We show that these "pre-activated" HSCs have: (i) increased expression of typical markers of HSC activation; (ii) decreased retinyl ester levels, accompanied by reduced expression of the enzyme needed for hepatic retinyl ester synthesis (LRAT); (iii) decreased triglyceride levels; (iv) increased expression of genes associated with lipid catabolism; and (v) an increase in expression of the retinoid-catabolizing cytochrome, CYP2S1. The second aim of this research was to investigate HSC lipid droplet formation and maintenance in healthy, but genetically-modified liver: specifically, we studied HSC lipid droplets in the LRAT KO mouse model, a system where HSC lipid droplets do not form. Our findings indicate that there are not global differences in retinoid-related gene expression, suggesting that the formation and maintenance of HSC lipid droplets is likely regulated entirely by the synthesis and storage of retinyl ester and not by more profound changes in retinoid metabolism. Our data also shows that the LRAT KO HSCs have significant differences in expression of genes related to lipid metabolism; overall, lipid biosynthesis is down-regulated and lipid catabolism is up-regulated in LRAT KO HSCs, which likely contributes to the complete absence of lipid droplets in the HSCs of these animals. Importantly, we show for the first time, to our knowledge, that the lipid droplet-associated proteins may be post-transcriptionally regulated. A final aim of this research was to investigate HSC lipid droplet dissolution in HSC activation and hepatic fibrosis, systems where HSC lipid droplets form, but are subsequently lost. We employed two standard models of HSC activation, the in vivo model of carbon tetrachloride (CCl4) treatment and the in vitro model, the culture of purified HSCs on plastic cell culture dishes. Additionally, we studied the effects of hypervitaminosis A since there is evidence in the literature that dietary vitamin A toxicity can cause hepatic fibrosis. Our studies suggest that, despite being unable to synthesize and store retinyl ester in lipid droplets, LRAT KO mice are not more susceptible than WT to the development of diet- or chemically-induced hepatic fibrosis. We found that, while the culture of HSCs on plastic results in the typical hallmark events of HSC activation, including the upregulation of Col1a1, the decrease in retinyl ester and the loss of lipid droplets, it does not regulate gene expression as HSC activation does in vivo. Thus, all future studies on HSC activation and its effects on retinoid storage should be conducted in vivo. We also present preliminary data on the alterations in the lipidome of activated HSCs, specifically with regard to the potent lipid signaling molecules, endocannabinoids, sphingolipids and ceramides. Our findings allow us to hypothesize that endocannabinoids and sphingolipids may function in activated HSCs as mediators of apoptosis. Importantly, this study demonstrates the ability to detect these lipids in very small aliquots of in vivo-activated HSCs and provides a strong foundation upon which all future studies may be built.



  • thumnail for DAmbrosio_columbia_0054D_10379.pdf DAmbrosio_columbia_0054D_10379.pdf application/pdf 20.3 MB Download File

More About This Work

Academic Units
Nutritional and Metabolic Biology
Thesis Advisors
Blaner, William S.
Ph.D., Columbia University
Published Here
September 28, 2011