Autophagy regulates lipid metabolism

Abstract

The intracellular storage and utilization of lipids are critical to maintain cellular energy homeostasis. During nutrient deprivation, cellular lipids stored as triglycerides in lipid droplets are hydrolysed into fatty acids for energy. A second cellular response to starvation is the induction of autophagy, which delivers intracellular proteins and organelles sequestered in double-membrane vesicles (autophagosomes) to lysosomes for degradation and use as an energy source. Lipolysis and autophagy share similarities in regulation and function but are not known to be interrelated. Here we show a previously unknown function for autophagy in regulating intracellular lipid stores (macrolipophagy). Lipid droplets and autophagic components associated during nutrient deprivation, and inhibition of autophagy in cultured hepatocytes and mouse liver increased triglyceride storage in lipid droplets. This study identifies a critical function for autophagy in lipid metabolism that could have important implications for human diseases with lipid over-accumulation such as those that comprise the metabolic syndrome.

Figure 1. Inhibition of autophagy leads to increased TG accumulation

a, TG levels in hepatocytes untreated (None) or treated with 3-methyladenine (3MA) and cultured in regular medium (RM) or oleate (OL) (*P < 0.02, **P < 0.002, n = 3). b, TG levels in vector-infected (VEC) and siAtg5 cells in RM, OL or in MCDM (*P < 0.001, n = 5). OL values are in mM. c, TG levels in wild-type (WT) or Atg5 knockout mice embryonic fibroblasts (Atg5^−/−)(*P < 0.01 or **P < 0.0001, n = 5). d–f, Cells from a and b stained with BODIPY 493/503 (d), oil red O (e) or visualized by electron microscopy (f). Right: quantifications of LD number and size (*P < 0.01, **P < 0.001 with untreated wild-type cells; #P < 0.01, ##P < 0.001 with cells in RM; n = 4–6). Nuclei are highlighted with 4,6-diamidino-2-phenylindole (DAPI; d). Error bars, s.e.m.

Figure 2. Inhibition of autophagy decreases TG β-oxidation and decay

a, b, VEC and siAtg5 cells cultured with oleate (OL) or in MCDM were examined for their rates of TG synthesis (a) and β-oxidation (b) as compared to cells in regular medium alone (*P < 0.03, **P < 0.004 with VEC cells in the same medium, n = 3–4). c, d, Rates of TG decay in OL (c) and MCDM (d) (*P < 0.05, **P < 0.01, ***P < 0.001 as compared to VEC cells, n = 3–7). e, TG levels in wild-type cells treated with dimethyl sulphoxide vehicle (DMSO), 3-methyladenine (3MA) or diethylumbelliferyl phosphate (DEUP) (*P < 0.00001 with DMSO-treated cells, #P < 0.003 with 3MA-treated cells, n = 6). Error bars, s.e.m. f, Co-localization of BODIPY 493/503 (green) with LAMP1 (red) or LC3 (red) in hepatocytes in MCDM. g, High-magnification regions of hepatocytes in MCDM alone (none) or treated with vinblastine and stained as labelled. Arrows indicate co-localization.

Figure 3. Lipid droplet content is delivered to lysosomes in autophagosomes

a, Electron micrographs of cultured hepatocytes. dm, double-membrane cytosolic vesicles. Arrows indicate membranes in LDs. Insets show the double membrane in nucleus (N), lipid-containing vesicles (L) and mitochondria (mit), but not in LDs. b, Mouse liver LC3 immunogold. Insets show higher magnification. Arrowheads indicate gold particles (black) and LC3-labelled bilayer membranes (white). c, Percentages of autophagic vacuoles (AVs) containing only lipid (Lipids, *), other cargo (Other) or mixed cargo (Mixed) in cells treated as in b (*P < 0.01, **P < 0.001 with cells in RM, n = 4–6). Left: representative examples. Error bars, s.e.m.

Figure 4. Effects of starvation, HFD feeding and a hepatocyte-specific blockage of autophagy on hepatic lipid accumulation

a, Immunoblots of cellular homogenates (Hom) and LDs from fed (F) or 24-h starved (S) mice. IκB, inhibitor of the nuclear factor of kappa light polypeptide gene enhancer; GPDH, glyceraldehyde-3-phosphate dehydrogenase. b, Percentages of autophagic vacuoles (AVs) containing only lipid, other cargo or mixed cargo calculated from samples processed as in Supplementary Fig. 16b (*P < 0.001, **P < 0.0001, n = 4–6). c, Immunoblots of Hom, AV and lysosomes (Lys). L-1, LAMP1. d, Immunoblots of liver Hom and LDs from HFD-fed mice. e, Liver homogenate immunoblots. PDI, protein disulphide isomerase. f, Total hepatic TG and cholesterol content (*P < 0.01, **P < 0.00001, n = 8–17). g, Hepatic TG concentration (*P < 0.05, n = 3). h, Percentage of cellular cholesterol in lysosomes (*P < 0.02, n = 4). i, Immunoblots of homogenates and isolated LDs. Error bars, s.e.m.