Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 May 7;29(5):1119-1134.e12.
doi: 10.1016/j.cmet.2019.01.005. Epub 2019 Feb 7.

Glycogen Synthase Kinase-3α Promotes Fatty Acid Uptake and Lipotoxic Cardiomyopathy

Michinari Nakamura  1 , Tong Liu  2 , Seema Husain  3 , Peiyong Zhai  1 , Junco S Warren  4 , Chiao-Po Hsu  5 , Takahisa Matsuda  1 , Christopher J Phiel  6 , James E Cox  7 , Bin Tian  3 , Hong Li  2 , Junichi Sadoshima  8
Affiliations

Glycogen Synthase Kinase-3α Promotes Fatty Acid Uptake and Lipotoxic Cardiomyopathy

Michinari Nakamura et al. Cell Metab. .

Abstract

Obesity induces lipotoxic cardiomyopathy, a condition in which lipid accumulation in cardiomyocytes causes cardiac dysfunction. Here, we show that glycogen synthase kinase-3α (GSK-3α) mediates lipid accumulation in the heart. Fatty acids (FAs) upregulate GSK-3α, which phosphorylates PPARα at Ser280 in the ligand-binding domain (LBD). This modification ligand independently enhances transcription of a subset of PPARα targets, selectively stimulating FA uptake and storage, but not oxidation, thereby promoting lipid accumulation. Constitutively active GSK-3α, but not GSK-3β, was sufficient to drive PPARα signaling, while cardiac-specific knockdown of GSK-3α, but not GSK-3β, or replacement of PPARα Ser280 with Ala conferred resistance to lipotoxicity in the heart. Fibrates, PPARα ligands, inhibited phosphorylation of PPARα at Ser280 by inhibiting the interaction of GSK-3α with the LBD of PPARα, thereby reversing lipotoxic cardiomyopathy. These results suggest that GSK-3α promotes lipid anabolism through PPARα-Ser280 phosphorylation, which underlies the development of lipotoxic cardiomyopathy in the context of obesity.

Keywords: GSK-3α; PPARα; diabetic cardiomyopathy; fatty acid metabolism; fibrates; lipid accumulation; lipotoxic cardiomyopathy; lipotoxicity; metabolic syndrome; obesity.

PubMed Disclaimer

Conflict of interest statement

Declaration of Interests

The authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Cardiac-specific GSK-3α haploinsufficiency ameliorates high-fat diet (HFD)-induced lipotoxic cardiomyopathy.
(A) Area-proportional Venn diagram representing the 26 genes significantly (p < 0.05) upregulated both in ob/ob (GSE16790) and db/db (GSE36875) mouse hearts, but not in streptozotocin-induced diabetic hearts (Type I DM) (GSE5606). (B) Pie chart illustrating the percent composition of Gene Ontology biological processes of the 26 common genes found in (A). Metabolic process (GO: 0008152) contains the largest gene set (13 genes), among which only GSK-3α is a kinase. (C) Immunoblots to evaluate nuclear GSK-3α activity in the hearts of wild-type (WT) mice fed a HFD or normal chow (NC) for the indicated periods. GSK-3α was immunoprecipitated from the nuclear fraction of heart lysates, followed by in vitro kinase assays with recombinant β-catenin. Recombinant GSK-3α protein was used as a positive control and immunoprecipitation with IgG was used as a negative control. (D) Quantification of the nuclear GSK-3α activity in (C) (n = 3). (E to L) GSK-3α cardiac-specific heterozygous knockout (GSK-3α cHKO) mice and heterozygous floxed (control) mice were fed a HFD or NC for 14 weeks. (E) Photograph of the hearts of control and GSK-3α cHKO mice fed a HFD or NC. (F) Left ventricular (LV) weight normalized by tibia length, a marker of cardiac hypertrophy (n = 8 (NC) and 22–24 (HFD)). (G and H) Diastolic function, as indicated by deceleration time (n = 8–15) (G), and the slope of the end-diastolic pressure-volume (PV) relation (EDPVR) (n = 5 (NC) and 9 (HFD)) (H). (I) Lipid accumulation in the hearts (Oil Red O staining, left). Scale bar, 100 μm. Inset scale bar, 20 μm. Quantification of myocardial lipid accumulation (right) (n = 6). (J) Palmitate oxidation in the hearts (n = 8–9 (NC) and 17 (HFD)). (K) Picric acid sirius red (PASR) staining, indicating cardiac fibrosis (left). Scale bar, 100 μm. Percentage of PASR positive areas (right) (n = 4). (L) mRNA expression related to cardiac metabolism, inflammation, and transcription factors in the hearts (n = 6). (M) Gene set enrichment analysis plot of Kyoto encyclopedia of genes and genomes (KEGG). PPAR signaling signatures in GSK-3α S21A knock-in (KI) and WT mice fed NC. NES denotes normalized enrichment score. FDR denotes false discovery rate. Error bars indicate s.e.m. * p<0.05, ** p<0.001. See also Figures S1 and S2 and Table S1.
Figure 2.
Figure 2.. GSK-3α physically interacts with and phosphorylates PPARα at Ser280 in cardiomyocytes (CMs) and in the heart.
(A and B) Immunoprecipitation assays to test the interaction between endogenous GSK-3α and exogenously expressed PPARα. YFP-tagged PPARα or FLAG-tagged PPARα was overexpressed in CMs using adenovirus (A) or in transgenic mouse hearts under the control of the αMHC-promoter (B), respectively. (C) Co-immunoprecipitation assays testing the interaction between endogenous GSK-3α and endogenous PPARα in CMs. (D) In vitro binding assays testing the direct interaction between recombinant (r) GSK-3α and rPPARα. (E to G) Immunoprecipitation assays to identify the amino acids in PPARα responsible for the interaction with endogenous GSK-3α. (E) Schematic representation of rGST-fused PPARα fragments. (F) Schema of the immunoprecipitation assays. rGST-fused-PPARα-full length (FL) or truncated PPARα (T1 to 5) was incubated with lysates extracted from cultured CMs, followed by pull-down with glutathione-sepharose and immunoblotting with anti-GSK-3α antibody. (G) Coomassie Brilliant Blue staining of rGST-PPARα-FL or truncated rGST-PPARα (T1 to T5) (left). Immunoblots testing the binding of endogenous GSK-3α to rGST-PPARα-FL or T1 to T5 (right). (H) Mass spectrometry analysis of the rGST-PPARα protein phosphorylated by GSK-3α in a kinase reaction. The MS/MS spectrum of the PPARα residue corresponding to Ser280 was increased at 80 Da, indicating phosphorylation. (I) Immunoblots showing Ser280 phosphorylation of endogenous PPARα in the hearts of WT mice fed a high-fat diet (HFD) or normal chow (NC) for 3 weeks. α-sarcomeric actinin was used as a loading control. (J) Immunoblots showing pPPARα (S280) in the hearts of control or GSK-3α cHKO mice fed a HFD for the indicated period. See also Figure S3.
Figure 3.
Figure 3.. Ser280 phosphorylation stimulates a subset of PPARα targets and promotes fatty acid uptake in cardiomyocytes.
(A) Immunoblots showing subcellular localization and expression of p-PPARα (Ser280) in cultured cardiomyocytes (CMs) treated with BSA-palmitic acid (PA) (0–500 μM) for 9 hours. (B) Immunoblots examining the involvement of GSK-3α in the BSA-PA-induced increase in PPARα phosphorylation at Ser280 in CMs. CMs transduced with adenovirus harboring shRNA-GSK-3α or scramble were treated with the indicated concentrations of PA, followed by nuclear extraction and immunoblotting. (C) Immunoblots showing the expression of p-PPARα (Ser280) in the nucleus of CMs treated with BSA-fatty acid cocktail for 9 hours. (D) Clustergram heat map of RNA-sequencing data. H9C2 cells were transduced with PPARα-wild type (WT), -S280A (SA) or -S280D (SD) mutant, or YFP alone as a control. Gene sets having 1) a fold difference of 1.5 or more between SA and SD and 2) a WT expression level located between SD and SA are shown in the heat map. (E) Gene set enrichment analysis plots of PPARα-SD (vs WT) showing upregulated signatures related to energy metabolism. Gene expression was determined by RNA-seq data. (F) qRT-PCR validation of expression of genes related to lipid metabolism and oxidative phosphorylation in CMs transduced with adenovirus harboring a YFP-PPARα-S280A or -S280D mutant or YFP alone as a control (n = 8–10). (G) Fatty acid uptake into CMs transduced with the indicated adenovirus. 3H-palmitate (left) and combined 3H-palmitate and 3H-oleate (right) incorporation into CMs was measured by scintillation counting (n = 4–5). (H) Relative oxygen consumption rate (OCR) of CMs transduced with the indicated adenovirus was measured in a 24-well Seahorse experiment in the presence of 500 μM of BSA-PA. Mitochondrial fatty acid oxidation (FAO) was evaluated by etomoxir-inhibitable OCR. Histograms show FAO rate (the ratio of FAO versus total OCR) (n = 5). (I) FAO rate in CMs transduced with the indicated adenovirus, measured in a 96-well Seahorse experiment in the presence of 500 μM of BSA-fatty acid cocktail (n = 8). (J and K) Oil Red O staining of CMs transduced with the indicated adenovirus in the presence or absence of BSA-PA (500 μM) (n = 5) (J) or in the presence of BSA-fatty acid cocktail (500 μM) (n = 6) (K). BSA alone was used as a control (PA 0 μM). Scale bars, 100 μm (upper panel) and 20 μm (lower panel). Error bars indicate s.e.m. * p < 0.05, ** p < 0.001. See also Figure S4 and Table S2.
Figure 4.
Figure 4.. PPARα-Ser280 phosphorylation enhances both interaction with RXRα and PPRE binding.
(A) Representative immunoblots showing the interaction between YFP-PPARα-wild type (WT), -S280D (SD) or -S280A (SA) mutant and RXRα in cardiomyocytes (CMs) in vitro. YFP alone was used as a control. (B) PPRE-luciferase reporter assay using a series of alanine mutations to evaluate the effect of the indicated basic residues on the activity of PPARα-S280D (n = 12). ## p<0.001 compared to PPARα-S280D. (C) Chromatin immunoprecipitation (ChIP) assays using CMs transduced with adenovirus (Ad)-YFP-PPARα-WT, -SD, -SA, or YFP alone as a control. DNA was amplified by PCR with specific primers flanking the promoter of the indicated genes containing the PPARα-binding motif. PCR using input DNA as template served as an internal control. The data shown are representative of three independent experiments. (D) ChIP assays using specific primers flanking the promoter of the indicated genes containing the PPARα-binding motif. The data shown are representative of three independent experiments. (E) Double-stranded oligo pull-down assays, using biotinylated oligos containing the specific PPRE sequences in the indicated gene promoters. Recombinant GST-PPARα-WT or -SA was subjected to in vitro kinase assays using recombinant GSK-3α prior to the oligo pull-down assays. An oligo containing a PPRE with mutations in four base pairs was used as a negative control. (F) Venn diagram showing the number of genes in the rat genome containing the specific PPRE/DR1 motif (shown in Figure S5F) in their promoters. The numbers of overlapping genes in the Venn diagram (red circle) were significantly different (p=0.0002, Fisher’s exact test). See also Figure S5.
Figure 5.
Figure 5.. Phosphorylation of PPARα at Ser280 is critical for the development of lipotoxic cardiomyopathy.
(A) Heatmap of metabolites associated with fatty acids/glycerol, glycolysis and BCAA metabolism in the hearts of wild-type (WT) and heterozygous PPARα-S280A knock-in (KI) mice at baseline, measured by metabolomics. (B) Histograms showing myocardial free fatty acids in the hearts of WT and heterozygous PPARα-S280A KI mice (n = 5). (C to E) WT and heterozygous PPARα-S280A KI mice were fed a high-fat diet (HFD) or normal chow (NC) for 8 weeks. (C) Left ventricular (LV) weight normalized by tibia length (n = 5). (D) The slope of the end-diastolic pressure-volume (PV) relation (EDPVR), a marker of diastolic function (n = 5–8). (E) Oil Red O staining of the heart sections (left). Scale bar, 50 μm. Quantification of myocardial lipid accumulation (right) (n = 6). (F to M) Either PPARα-WT or PPARα-S280D (SD) mutant was expressed in the hearts of WT mice (C57BL/6J background) fed NC for 8 weeks via adeno-associated virus (AAV)-mediated gene delivery. AAV-empty injection (Emp or Empty) was performed as a control. (F) Representative immunoblots showing the expressions of p-PPARα (Ser280), PPARα-S280D and total PPARα in the heart and skeletal muscle. Histogram indicates the expression levels relative to Histone H3 (n = 8). (G) LV weight normalized by tibia length (n = 5–7). (H) LV posterior wall thickness on diastole (LVPWd), a marker of hypertrophy (n = 5–6). (I) Deceleration time, a marker of diastolic function (n = 5–6). (J) The slopes of EDPVR (left) and Tau (right), both markers of diastolic function (n = 4–6). (K) Fatty acid oxidation (FAO) rate in CMs isolated from the hearts of mice transduced with the indicated AAV (n = 5). (L) Oil Red O staining of the heart sections (left). Scale bars, 50 μm. Quantification of myocardial lipid accumulation (right) (n = 4–6). (M) Picric acid sirius red (PASR) staining indicating cardiac fibrosis in the heart sections (left) and the percentage of PASR positive areas (right). Scale bars, 200 μm (n = 4–7). Error bars indicate s.e.m. * p<0.05, ** p<0.001. See also Figure S6 and Table S3.
Figure 6.
Figure 6.. Fenofibrate, a PPARα ligand, inhibits GSK-3α-mediated PPARα-Ser280 phosphorylation, thereby ameliorating lipotoxic cardiomyopathy.
(A) Immunoblots examining the effect of fenofibrate, a PPARα agonist, on PPARα-Ser280 phosphorylation in cardiomyocytes (CMs) in the presence or absence of 500 μM of BSA-palmitic acid (PA). (B to E) Wild-type (WT) mice were fed a high-fat diet (HFD) in the presence or absence of fenofibrate (Feno) for the indicated periods, as shown in Figure S7C. (B) Immunoblots examining the effect of fenofibrate on PPARα-Ser280 phosphorylation in the hearts. (C) Left ventricular (LV) weight normalized by tibia length (n = 8–12). (D) The slope of the end-diastolic pressure-volume (PV) relation (EDPVR), a marker of diastolic function (n = 5–11). (E) Oil Red O staining of heart sections after 14 weeks of HFD in the presence or absence of fenofibrate (left). Scale bar, 100 μm. Quantification of myocardial lipid accumulation (right) (n = 5). (F) The differential effect of fenofibrate on PPRE-luciferase reporter activity in H9C2 cells transduced with PPARα-WT or PPARα-S280D mutant in the presence of a high concentration of fatty acid (500 μM of PA) or BSA control. YFP alone was used for background extraction (n = 6). (G) Fatty acid oxidation (FAO) rate in CMs in the presence of 50 μM or 500 μM of PA. Total oxygen consumption rate (OCR) in CMs transduced with the indicated adenoviruses was measured by 96-well Seahorse experiment and mitochondrial FAO was evaluated by etomoxir-inhibitable OCR (n = 18–23 (WT) and 6–7 (S280D)). (H) Oil Red O staining of rat neonatal CMs transduced with the indicated adenovirus in the presence or absence of fenofibrate with 50 μM or 500 μM of BSA-PA (left). Scale bar, 50 μm. Quantification of myocardial lipid accumulation (right) (n = 6). (I and J) Immunoprecipitation assays showing the interaction between endogenous GSK-3α and YFP-PPARα or YFP alone as a control in CMs treated with 100 μM or 500 μM of BSA-PA in the presence or absence of 10 μM of fenofibrate (I) and quantification of the data (n = 4) (J). (K) Proximity Ligation Assay (PLA) showing the in situ interaction between endogenous GSK-3α and YFP-PPARα in CMs treated with 100 μM or 500 μM of PA in the presence or absence of 10 μM of fenofibrate. PLA was performed using anti-GSK-3α and anti-GFP-YFP antibodies. Red color indicates localization in close proximity. Scale bar, 10 μm. (L) Immunoblots examining GSK-3α activity in failing human hearts in the presence (n = 7) or absence (n = 14) of diabetes. See also Figure S7 and Table S4.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ackers-Johnson M, Li PY, Holmes AP, O’Brien SM, Pavlovic D, and Foo RS (2016). A Simplified, Langendorff-Free Method for Concomitant Isolation of Viable Cardiac Myocytes and Nonmyocytes From the Adult Mouse Heart. Circulation research 119, 909–920. - PMC - PubMed
    1. Azoulay-Alfaguter I, Yaffe Y, Licht-Murava A, Urbanska M, Jaworski J, Pietrokovski S, Hirschberg K, and Eldar-Finkelman H (2011). Distinct molecular regulation of glycogen synthase kinase-3alpha isozyme controlled by its N-terminal region: functional role in calcium/calpain signaling. The Journal of biological chemistry 286, 13470–13480. - PMC - PubMed
    1. Banks AS, McAllister FE, Camporez JP, Zushin PJ, Jurczak MJ, Laznik-Bogoslavski D, Shulman GI, Gygi SP, and Spiegelman BM (2015). An ERK/Cdk5 axis controls the diabetogenic actions of PPARgamma. Nature 517, 391–395. - PMC - PubMed
    1. Bugger H, and Abel ED (2009). Rodent models of diabetic cardiomyopathy. Disease models & mechanisms 2, 454–466. - PubMed
    1. Choi JH, Banks AS, Estall JL, Kajimura S, Bostrom P, Laznik D, Ruas JL, Chalmers MJ, Kamenecka TM, Bluher M, et al. (2010). Anti-diabetic drugs inhibit obesity-linked phosphorylation of PPARgamma by Cdk5. Nature 466, 451–456. - PMC - PubMed

Publication types

MeSH terms