Lysosomal acid lipase regulates VLDL synthesis and insulin sensitivity in mice
Lysosomal acid lipase (LAL) hydrolyses cholesteryl esters and triacylglycerols (TG) within lysosomes to mobilise NEFA and cholesterol. Since LAL-deficient (Lal-/-) mice suffer from progressive loss of adipose tissue and severe accumulation of lipids in hepatic lysosomes, we hypothesised that LAL deficiency triggers alternative energy pathway(s).
We studied metabolic adaptations in Lal-/- mice.
Despite loss of adipose tissue, Lal-/- mice show enhanced glucose clearance during insulin and glucose tolerance tests and have increased uptake of [3H]2-deoxy-D-glucose into skeletal muscle compared with wild-type mice. In agreement, fasted Lal-/- mice exhibit reduced glucose and glycogen levels in skeletal muscle. We observed 84% decreased plasma leptin levels and significantly reduced hepatic ATP, glucose, glycogen and glutamine concentrations in fed Lal-/- mice. Markedly reduced hepatic acyl-CoA concentrations decrease the expression of peroxisome proliferator-activated receptor α (PPARα) target genes. However, treatment of Lal-/- mice with the PPARα agonist fenofibrate further decreased plasma TG (and hepatic glucose and glycogen) concentrations in Lal-/- mice. Depletion of hepatic nuclear factor 4α and forkhead box protein a2 in fasted Lal-/- mice might be responsible for reduced expression of microsomal TG transfer protein, defective VLDL synthesis and drastically reduced plasma TG levels.
Our findings indicate that neither activation nor inactivation of PPARα per se but rather the availability of hepatic acyl-CoA concentrations regulates VLDL synthesis and subsequent metabolic adaptations in Lal-/- mice. We conclude that decreased plasma VLDL production enhances glucose uptake into skeletal muscle to compensate for the lack of energy supply.
KeywordsGlucose tolerance Lipolysis Lysosomes VLDL
Cholesteryl ester storage disease
Fast protein liquid chromatography
Haematoxylin and eosin
Insulin tolerance test
Lysosomal acid lipase
Peroxisome proliferator-activated receptor α
subcutaneous white adipose tissue
White adipose tissue
Lysosomal acid lipase (LAL) hydrolyses cholesteryl esters (CE) and triacylglycerols (TG), delivered to the lysosome mainly via LDL particle uptake , to release mono- and diacylglycerols and mobilise cholesterol and NEFA for membrane assembly, steroidogenesis and energy production. Genetic mutations of human LAL (also known as LIPA) cause an autosomal recessive lysosomal storage disorder with accumulation of CE predominantly in hepatocytes, adrenal glands, intestine and cells of the monocyte-macrophage system throughout the body . Mutations in LAL cause Wolman disease (WD) or cholesteryl ester storage disease (CESD) [3, 4, 5]. WD is a rare, neonatal-onset disease with less than 1% of LAL activity, characterised by massive hepatosplenomegaly, adrenal calcifications, malabsorption, growth retardation and cachexia. Affected patients die within the first three to 12 months of life [2, 6, 7]. CESD patients have up to 5% residual LAL activity , which keeps the syndrome mostly unrecognised until adulthood. These patients suffer from progressive lysosomal CE and TG accumulations, which lead to the characteristic liver pathology and elevated concentrations of serum transaminases, serum LDL and TG. Premature death of individuals with CESD is due to liver failure and/or accelerated atherosclerosis rather than to chronic hyperlipidaemia [2, 9, 10].
In contrast to humans, complete loss of LAL activity in mice phenotypically resembles CESD rather than WD. LAL-deficient (Lal-/-) mice appear normal at birth and reach the median life span of approximately 1 year . Severe accumulations of CE and TG are found predominantly in the liver, spleen, small intestine and adrenals [11, 12, 13]. Lal-/- mice exhibit reduced size and body weight (BW) compared with wild-type (WT) littermates as well as progressive loss of white adipose tissue (WAT) and brown adipose tissue [12, 13]. In this study, we explored the metabolic changes caused by impaired lysosomal TG and CE hydrolysis in Lal-/- mice.
Age- and sex-matched Lal-/- and WT littermates  on a C57BL/6J background were maintained with unlimited access to chow and water in regular 12 h light/12 h dark cycles. For fenofibrate treatment, 5-week-old Lal-/- mice were administered 0.2% fenofibrate/chow for 4 weeks. Experimenters were blind to group assignment and outcome assessment. Animal experiments were approved by the Federal Ministry of Science, Research and Economy, Vienna, Austria. See electronic supplementary material (ESM) Methods for further details.
Lipid and hormone concentrations and fast protein liquid chromatography
Cholesterol, TG, glycerol and markers of liver injury were quantified enzymatically. Lipoprotein profiles were analysed by fast protein liquid chromatography (FPLC) . Hormones were determined by ELISA. See ESM Methods for further details.
Polar metabolites from mouse liver were extracted as described previously . HPLC was performed on a 1100 Agilent capillary LC (Agilent Technologies, Santa Clara, CA, USA) equipped with a polyhydroxyethyl column (PolyLC Inc, Columbia, MD, USA). Solvent A (10 mmol/l ammonium acetate/water) and solvent B (10 mmol/l ammonium acetate/90% acetonitrile) were used in varying gradients. Selected ions/fragments were detected by TSQ Quantum Ultra Mass Spectrometry (Thermo Fisher Scientific, Waltham, MA, USA) in negative mode. See ESM Methods for further details.
De novo lipid synthesis
Animals were i.p. injected with [14C]acetate (5 μCi in 200 μl PBS), killed 1 h post-injection, and livers were isolated and lyophilised for 48 h. Lipid extracts were separated by thin-layer chromatography (TLC) (n-hexane:diethylether:acetic acid; 80:20:2, vol.:vol.:vol.). Radioactivity in bands corresponding to NEFA, TG, non-esterified cholesterol and CE was determined by liquid scintillation counting.
Acyl-CoAs were extracted from liver lysate homogenates using 0.5 ml of buffer (50% 0.1 mol/l KH2PO4, 50% 2-propanol; 4°C), 30 μl saturated (NH4)2SO4 and 0.5 ml acetonitrile, and centrifuged (2,500 g, 10 min, 4°C). Acyl-CoAs were determined by liquid chromatography-mass spectrometry as described previously . See ESM Methods for further details.
Neutral TG hydrolase activity
Mitochondria isolation and respirometry
Liver mitochondria were isolated as described previously  and resuspended in medium. Mitochondria were diluted in medium plus 0.2% NEFA-free BSA, 5 mmol/l glutamate, 1 mmol/l malate, or 5 mmol/l pyruvate, 5 mmol/l succinate, 1 μmol/l rotenone. Oxygen consumption rates were measured at 37°C (Oxygraph-2k; Oroboros Instruments, Innsbruck, Austria). State 3-respiration (mitochondrial respiration due to ADP supply) and state 4o-respiration (after ATP synthase inhibition by oligomycin) were induced by 2 mmol/l ADP and 1 μg/ml oligomycin. Maximal uncoupled respiration was determined after titration of carbonyl cyanide p-trifluoromethoxyphenylhydrazone. See ESM Methods for further details.
RNA isolation and quantitative real-time PCR analysis
Mice (8 h-fasted) were i.p. injected with 500 mg/kg BW tyloxapol in PBS. Plasma TG were determined every hour post-injection.
Mice were i.p. injected with glucose (2 g/kg BW), insulin (0.25 U/kg BW), glucagon (140 μg/kg BW), glycerol (2 g/kg BW), pyruvate (2 g/kg BW) and glutamine (2 g/kg BW). Blood glucose levels were determined using Accu-Chek Active glucometer (Roche Diagnostics, Mannheim, Germany). See ESM Methods for further details.
Mice (6 h-fasted) were i.p. injected with 30 mg glucose and 0.5 μCi glucose per 30 g BW. Radioactivity in 40 μl plasma (15, 30 and 60 min post-injection) and tissue lysates was determined by β-counting. See ESM Methods for further details.
Quantification of metabolites by nuclear magnetic resonance spectroscopy
Liver and skeletal muscle lysates (200 μl) were mixed with methanol (400 μl), incubated at −20°C (30 min) and centrifuged. Supernatants were dried and re-dissolved in 500 μl D2O. 1H-one-dimensional-nuclear magnetic resonance experiments were performed at 310 K. Reference chemical shifts were taken from the Madison-Qingdao Metabolomics Consortium Database (http://mmcd.nmrfam.wisc.edu/) . Bruker Topspin3.1 (Rheinstetten, Germany) and MestReNova10.0 software (http://mestrelab.com) were used for data acquisition, processing and analyses. See ESM Methods for further details.
In vivo MRI for body fat
MR images of anesthetised mice were acquired by 3T-MRI (Siemens Tim-Trio, Erlangen, Germany) with an eight-channel multipurpose coil (Noras MRI products, Hoechenberg, Germany).
Protein samples were separated by SDS-PAGE and transferred to polyvinylidene-difluoride membranes. Blots were incubated with primary antibodies followed by HRP-conjugated secondary antibodies. See ESM Methods for further details.
Subcutaneous WAT sections
Tissue samples were fixed in 4% paraformaldehyde and embedded in paraffin. Tissue sections (4 μm) were stained with haematoxylin and eosin (H&E) and evaluated by light microscopy.
Liver sections were fixed in phosphate buffer/2.5% glutaraldehyde, washed, post-fixed in phosphate buffer/OsO4 for 1 h and 4 × 10 min in phosphate buffer. After dehydration, tissues were infiltrated (acetone and agar 100 epoxy resin, pure agar 100 epoxy resin) for 4 h, placed in agar 100 epoxy resin (8 h), transferred into embedding moulds, and allowed to polymerise (48 h, 60°C). Sections stained with lead citrate and uranyl acetate were imaged. See ESM Methods for further details.
Since Lal-/- mice were obviously much smaller than WT mice, in vivo experiments were not blind. In vitro experiments were blind to group assignment and outcome assessment. Statistical analyses were performed using GraphPad Prism 5.04 (GraphPad Software, San Diego, CA, USA). Significant outliers were detected by Grubb’s test (http://graphpad.com/quickcalcs/Grubbs1.cfm). Significances were determined by Student’s unpaired t test and Welch correction (in case of unequal variances) for two group comparison and ANOVA followed by Bonferroni correction for multiple group comparison. Data are presented as means ± SD. Differences were considered statistically significant at p < 0.05.
Decreased acyl-CoA concentrations and peroxisome proliferator-activated receptor α signalling in Lal-/- liver
Intact mitochondrial function despite decreased ATP concentrations in Lal-/- livers
Diminished VLDL secretion in Lal-/- mice
We next investigated VLDL secretion after inhibition of peripheral lipolysis by tyloxapol and found strongly reduced VLDL release in fasted Lal-/- mice (Fig. 3e). As expected, Lal-/- mice showed markedly increased concentrations of aminotransferases (ESM Fig. 2), an indication of liver damage. Hepatic mRNA expression of microsomal TG transfer protein (Mttp), a key player in VLDL assembly, was decreased (Fig. 3f). Drastically reduced mRNA and protein expression of the hepatic transcription factors hepatocyte nuclear factor 4a (Hnf4a) and forkhead box protein a2 (Foxa2) (Fig. 3f, g), which are involved in VLDL synthesis [21, 22], reflect a causal link between low expression levels of these transcription factors and decreased VLDL secretion in Lal-/- livers.
Increased insulin sensitivity in Lal-/- mice
Decreased plasma leptin concentrations and altered expression of hepatic adiponectin and leptin receptors in Lal-/- mice
Reduced liver glucose, glycogen and glutamine concentrations in Lal-/- mice
PPARα activation reduces liver glucose and glycogen as well as plasma TG concentrations in Lal-/- mice
The appearance of hepatic ‘fatty lysosomes’ in insulin-sensitive Lal-/- mice contrasts with hepatosteatosis, in which accumulation of cytoplasmic lipid droplets is accompanied by insulin resistance [25, 26, 27]. Liver is the main organ for VLDL assembly and secretion and thus influences whole body TG and cholesterol homeostasis . Skop et al have already suggested the involvement of LAL in VLDL synthesis . The authors linked the process to autophagy, since inhibition of lysosomal activity by chloroquine decreased VLDL secretion in vitro. Our in vivo data, however, indicate that defective VLDL synthesis in Lal-/- mice is a consequence of decreased hepatic availability of acyl-CoAs, which then leads to downregulation of PPARα signalling, nuclear exclusion of HNFα and FOXA2, decreased Mttp expression, reduced TG synthesis, and eventually futile lipidation of ApoB [21, 22, 30, 31]. In general, PPARα activation mediates lipid oxidation and reduces ectopic lipid storage, thereby counteracting insulin resistance . Interestingly, Lal-/- mice show reduced expression of PPARα target genes, yet produce less VLDL, which contrasts with the increased secretion of VLDL observed in Ppara-/- mice . However, fenofibrate treatment did not normalise the phenotype but further decreased plasma TG (and hepatic glucose and glycogen) concentrations in Lal-/- mice. These findings indicate that neither activation nor inactivation of PPARα per se but rather the availability of hepatic acyl-CoAs regulates VLDL synthesis and subsequent metabolic adaptations.
HNF4α, a key regulator of various metabolic pathways, is classified as an orphan receptor despite NEFA being suggested as its endogenous ligand . In fact, Yuan et al demonstrated that HNF4α is selectively associated with linoleic acid in mammalian cells and in the liver of fed mice , indicating that linoleic acid is at least one possible endogenous ligand for HNF4α. Nuclear exclusion of HNF4α together with a 60% reduction in 18:2-CoA in livers of Lal-/- mice suggest that lysosomal mobilisation of linoleic acid is involved in VLDL synthesis via the HNF4α pathway. Accordingly, protein expression of FOXA2, which is controlled by insulin signalling  and promotes VLDL synthesis , was significantly reduced in Lal-/- liver. Besides reduced VLDL secretion, Lal-/- mice had increased insulin sensitivity compared with their WT littermates. This finding is in accordance with studies in animal models [37, 38] and type 2 diabetes patients , indicating a connection between insulin resistance and VLDL overproduction. We thus propose that reduced plasma VLDL is one of the features to induce a shift from NEFA to glucose utilisation, resulting in improved insulin sensitivity as observed by enhanced uptake of glucose in skeletal muscles of these mice. Although non-significant, a 35% increase of skeletal muscle glycogen in fed Lal-/- mice indicates enhanced storage of glucose for energy supply during fasting.
WAT is a primary contributor to metabolic regulation during feeding and fasting. Although lipodystrophy is generally associated with insulin resistance , Lal-/- mice show reduced plasma glucose and enhanced glucose usage. Despite substantial loss of WAT mass in Lal-/- mice, plasma concentrations of adiponectin were comparable, whereas leptin levels were profoundly decreased after feeding. These findings indicate that Lal-/- mice have a constant energy demand due to the unavailability of NEFA from WAT cytosolic lipid droplets. Diminished VLDL synthesis induces depletion of liver energy storage pools as reflected by reduced liver glycogen, glucose and glutamine concentrations. Decreased liver glucose levels may be causative for lower hepatic mRNA expression of genes involved in glycolysis, particularly of Gck and Pk1 (also known as Pklr). The liver provides glucose through gluconeogenesis from non-carbohydrate precursors during prolonged fasting . Lal-/- mice, however, are unable to produce glucose from external glutamine, suggesting that the degradation of glutamine to pyruvate is defective. However, gluconeogenesis per se is functional, as shown by unaltered pyruvate tolerance tests. Glycerol tolerance test revealed earlier maximal glucose levels and glucose was cleared faster from the circulation, confirming enhanced glucose utilisation in Lal-/- mice.
Open access funding provided by Medical University of Graz. The authors thank S. Rainer and A. Ibovnik (Medical University of Graz, Austria) for excellent technical assistance, R. Schreiber (University of Graz, Austria) for providing the PLIN2 antibody, E. Bernhart for providing the Lamin A/C antibody (Medical University of Graz, Austria) and I. Hindler (Medical University of Graz, Austria) for the care of the mice.
This work was supported by the Austrian Science Fund FWF (DK-MCD W1226, P27070, P22832, SFB-LIPOTOX F3004) (DKr), the PhD programme ‘Molecular Medicine’ of the Medical University of Graz and BioTechMed-Graz. TM was supported by the Bavarian Ministry of Sciences, Research and the Arts (BioSysNet), the German Research Foundation (MA 5703/1-1), the Centre for Integrated Protein Science, Munich, the President’s International Fellowship Initiative of CAS (No: 2015VBB045) and the National Natural Science Foundation of China (No. 31450110423).
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
BR and DK conceived and designed the study and wrote the paper. BR, NV, CL, SS, MG, JVP, MK, DKo, JR, MW, TT, MS, LG, CM, CD and TM contributed to data collection and analyses. RBG, SF, ES, HD, WFG and TM contributed to study design and data interpretation. All authors critically revised the manuscript and have read and approved the final version. DK is responsible for the integrity of the work as a whole.
- 2.Assmann G, Seedorf U (1995) Acid lipase deficiency: Wolman disease and cholesteryl ester storage disease. In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds) The metabolic and molecular basis of inherited disease, 7th edn. McGraw and Hill, New York, pp 2563–2587Google Scholar
- 5.Sloan HR, Fredrickson DS (1972) Enzyme deficiency in cholesteryl ester storage disease. J Clin Invest 51:1923–1926Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.