GLP-1 analogs reduce hepatocyte steatosis and improve survival by enhancing the unfolded protein response and promoting macroautophagy

Abstract

Background: Nonalcoholic fatty liver disease (NAFLD) is a known outcome of hepatosteatosis. Free fatty acids (FFA) induce the unfolded protein response (UPR) or endoplasmic reticulum (ER) stress that may induce apoptosis. Recent data indicate ER stress to be a major player in the progression of fatty liver to more aggressive lesions. Autophagy on the other hand has been demonstrated to be protective against ER stress-induced cell death. We hypothesized that exendin-4 (GLP-1 analog) treatment of fat loaded hepatocytes can reduce steatosis by autophagy which leads to reduced ER stress-related hepatocyte apoptosis.

Methodology/principal findings: Primary human hepatocytes were loaded with saturated, cis- and trans-unsaturated fatty acids (palmitic, oleic and elaidic acid respectively). Steatosis, induced with all three fatty acids, was significantly resolved after exendin-4 treatment. Exendin-4 sustained levels of GRP78 expression in fat-loaded cells when compared to untreated fat-loaded cells alone. In contrast, CHOP (C/EBP homologous protein); the penultimate protein that leads to ER stress-related cell death was significantly decreased by exendin-4 in hepatocytes loaded with fatty acids. Finally, exendin-4 in fat loaded hepatocytes clearly promoted gene products associated with macroautophagy as measured by enhanced production of both Beclin-1 and LC3B-II, markers for autophagy; and visualized by transmission electron microscopy (TEM). Similar observations were made in mouse liver lysates after mice were fed with high fat high fructose diet and treated with a long acting GLP-1 receptor agonist, liraglutide.

Conclusions/significance: GLP-1 proteins appear to protect hepatocytes from fatty acid-related death by prohibition of a dysfunctional ER stress response; and reduce fatty acid accumulation, by activation of both macro-and chaperone-mediated autophagy. These findings provide a novel role for GLP-1 proteins in halting the progression of more aggressive lesions from underlying steatosis in humans afflicted with NAFLD.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Exendin-4 treatment reduces fatty acid load in hepatocytes.

Primary human hepatocytes were treated with fatty acids for 24 h and then supplemented with exendin-4 in fresh media. Hepatocytes treated with fatty acids were cultured for additional 24 h. (A) Primary human hepatocytes treated with exendin-4 (+) demonstrate reduction of fat load as quantified by Oil Red O staining of intracellular lipid droplets (B). (C: control; Ex: exendin-4; OA: oleic acid; OAE: oleic acid with exendin-4; PA: palmitic acid; PAE: palmitic acid with exendin-4; EA: elaidic acid; EAE: elaidic acid with exendin-4). Three independent experiments with three replicates each were conducted *p<0.001, Tukey-Kramer.

Figure 2. Exendin-4 suppresses apoptosis in primary human hepatocytes.

(A) Hepatocyte survival increased after exendin-4 treatment as measured by XTT assay. (B) DNA condensation as observed by DAPI staining of cells treated with different fatty acids and exendin-4. Damaged DNA is marked by white arrows (C). (D) Exendin-4 treatment led to suppression of apoptosis as demonstrated by reduction in cleaved caspase 3. These experiments were performed three times in triplicate, #p<0.001, *p<0.05, Tukey-Kramer.

Figure 3. GLP-1 analogs reduce ER stress in hepatocytes.

(A). Immunoblot of ER stress marker proteins GRP78 and CHOP. (B) GRP78 protein levels were increased in following primary hepatocyte cultures in presence of exendin-4 for 24 h. GRP78 levels were significantly increased by oleic acid more than by either palmitic or elaidic acid. (C) CHOP expression was significantly reduced by exendin-4, with maximum reduction from lysates obtained from elaidic acid-loaded hepatocytes. (D–E) Percent change over control in the expression levels of GRP78 and CHOP in human hepatocytes as calculated by RT-qPCR analysis. Representative blots from different independent experiments are presented here. #p<0.001, *p<0.05, Tukey-Kramer.

Figure 4. GLP-1 analogs induce macroautophagy and lysosomal turnover.

(A) Immunoblot of human hepatocytes for autophagy related proteins – beclin-1, LC3B and LAMP2A. Note the increased conversion of LC3B-I to LC3B-II in samples treated with exendin-4. (B) Densitometric analysis of immunoblots. For LC3B 16 kDa band was quantified. (C) Beclin, LC3B and LAMP2A gene expression levels were elevated in exendin-4 treated cells, as quantified by RT-qPCR. In (B) and (C) significance was determined by comparing exendin-4 treated samples vs. its respective fat loaded sample for each gene. These experiments were repeated three times in triplicate, #p<0.001, *p<0.05, Students t test.

Figure 5. Exendin-4 stimulates autophagic vacuoles formation in human hepatocytes.

(A) (i) Immunoblot for LC3 in palmitic acid and/or exendin-4 treated samples in presence or absence of bafilomycin A1. (B: bafilomycin A1; BPA: bafilomycin with palmitic acid; BEx: bafilomycin with exendin-4; PABEx: palmitic acid with bafilomycin and exendin-4). (ii) Autophagic flux depicted as values obtained after taking the ratio of densitometry values of LC3-II values from bafilomycin treated samples to their respective bafilomycin-free counterparts. Note the increase in autophagic flux in response to exendin-4 in the presence of palmitic acid against palmitic acid alone. (B) Electron microscopy of primary human hepatocytes (i) showing fat loading, (N: nucleus; LD: lipid droplet), (ii) structure of lipid droplets, (iii) autophagosomes containing lipid droplets (red arrow), (iv) autophagolysosome with lipid autophagosome and lipid droplet (iv) and, (v) shriveled lipid droplet (red arrow). Magnification: (i) – (ii) x3.5E102, (iii) – (v) x25E103. (C) Quantification of autophagic vacuoles (autophagosomes containing lipid droplets – dark bands, and autophagolysosomes containing lipid droplets – light bands) by integer scoring. #p<0.001, *p<0.05, Student's t-test, three independent experiments were performed.

Figure 6. Liraglutide treatment reduces steatosis and ER stress in high fat diet fed mice.

Mice were fed American Lifestyle-Induced Obesity Syndrome (ALIOS) diet and high fructose corn syrup for 8 weeks. One set of animals was subsequently treated with 200 µg/kg body weight liraglutide for next 4 weeks while maintaining other on high fat high sucrose diet. (A). Oil Red O staining of liver sections from mice kept on normal chow and treated with saline [C], liraglutide [L] and those given ALIOS diet alone [A] and subsequently liraglutide [AL]. (B) Immunoblotting of mouse liver lysates for GRP78 and CHOP. (C) Densitometric analysis of immunoblot (C: control, L: liraglutide treated, A: ALIOS fed, A+L: ALIOS fed and then liraglutide injected). (D) Real time mRNA quantification of GRP78 and CHOP. (E) Immunohistochemical staining of mouse liver sections for GRP78 and CHOP. All studies were conducted at least three times in triplicate with fresh lysates, *p<0.05 Student's t-test.

Figure 7. ALIOS-fed mice treated with liraglutide confirm GLP-1 enhanced lipoautophagy in liver.

(A). Immunoblot comparing beclin, LC3B conversion and LAMP2 levels in liver lysates from mice treated with liraglutide or ALIOS diet or combination thereof. (B) Quantification of immunoblot by densitometry. Differences between protein levels were significant for respective genes between ALIOS and ALIOS+Liraglutide samples (C) Real time PCR analysis of Beclin, LC3B and LAMP2 gene expression. Histogram shows fold change in ALIOS-fed, liraglutide treated animals versus ALIOS fed animals. These experiments were performed with lysates from six mice, *p<0.05 Student's t-test.