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
. 2022 Jul 8;13(1):3956.
doi: 10.1038/s41467-022-31525-8.

Lipolysis regulates major transcriptional programs in brown adipocytes

Affiliations

Lipolysis regulates major transcriptional programs in brown adipocytes

Lasse K Markussen et al. Nat Commun. .

Abstract

β-Adrenergic signaling is a core regulator of brown adipocyte function stimulating both lipolysis and transcription of thermogenic genes, thereby expanding the capacity for oxidative metabolism. We have used pharmacological inhibitors and a direct activator of lipolysis to acutely modulate the activity of lipases, thereby enabling us to uncover lipolysis-dependent signaling pathways downstream of β-adrenergic signaling in cultured brown adipocytes. Here we show that induction of lipolysis leads to acute induction of several gene programs and is required for transcriptional regulation by β-adrenergic signals. Using machine-learning algorithms to infer causal transcription factors, we show that PPARs are key mediators of lipolysis-induced activation of genes involved in lipid metabolism and thermogenesis. Importantly, however, lipolysis also activates the unfolded protein response and regulates the core circadian transcriptional machinery independently of PPARs. Our results demonstrate that lipolysis generates important metabolic signals that exert profound pleiotropic effects on transcription and function of cultured brown adipocytes.

PubMed Disclaimer

Conflict of interest statement

O.S.J and Z.G.H work or have worked, in some capacity, for Embark Biotech ApS, a company developing therapeutics for the treatment of diabetes and obesity. The remaining authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Lipolysis is sufficient and necessary for activation of a large fraction of the β-adrenergic gene program in brown adipocytes.
Mouse in vitro differentiated brown adipocytes were pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently with 100 nM isoproterenol (ISO) (blue bar) or 20 µM SR-3420 (red bar) for 3 h before harvesting. a Schematic overview of the experimental setup. Created with BioRender.com. b FA release to medium in the different conditions. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. c PCA plot of RNA-seq data. d Heatmap of RNA-seq data with K-means clustering. Only ISO-activated (ISO vs. vehicle: Log2FC ≥ 1, FDR/Benjamini-Hochberg≤0.05) genes are shown. e Models indicating the role of lipolysis in transcriptional activation of genes in C1–C4 by β-adrenergic signals. C1, C3, and C4 constitute lipolysis-activated genes, however, whereas C1 genes are strictly dependent on lipolysis for activation, C3 and C4 genes can also be activated by cAMP-independent of lipolysis. Created with BioRender.com. f mRNA expression pattern of representative genes from C1–C4 derived from RNA-seq. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. g Heatmap indicating the fold enrichment of gene pathways (FDR/Benjamini-Hochberg ≤ 0.05) in C1–C4. For all panels, error bars represent +/- SEM of 3 independent biological experiments, each carried out in technical duplicates. Statistical significance was determined by one-way ANOVA with Tukey’s multiple comparisons test for b and by DESeq2 using FDR/Benjamini-Hochberg correction for Fig. 1f (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus Vehicle, # versus ISO or SR-3420 (without Lipase inh), ¤ versus ISO (with/without Lipase inh.).
Fig. 2
Fig. 2. Lipolysis activates gene expression independent of fatty acid activation or oxidation.
Mouse in vitro differentiated brown adipocytes were pre-treated with 5 µM triacsin C (TC; ACSL inhibitor) or 50 µM etomoxir (ETO; CPT1A inhibitor) for 1 h and subsequently with 20 µM SR-3420 (red bar) for 3 h before harvest. a Schematic overview of the experimental setup. Created with BioRender.com. b Effect of triacsin C (TC) or etomoxir (ETO) on FA release. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. c Heatmap of RNA-seq data with K-means clustering. Only genes activated by SR-3420 (SR-3420 vs. vehicle Log2FC ≥ 1, pAdj < 0.05) are indicated. n = 2–3 biologically independent experiments examined, each carried out in a technical duplicate. d mRNA expression pattern of representative lipolysis-activated, FA activation-independent, and FA oxidation-dependent genes derived from RNA-seq. n = 2–3 biologically independent experiments examined, each carried out in a technical duplicate. e Intracellular ATP levels in mouse in vitro differentiated brown adipocytes pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently stimulated with 100 nM isoproterenol (ISO) or 20 µM SR-3420 for 1 h before harvest. n = 7-8 biologically independent experiments examined. f Intracellular ADP levels in mouse in vitro differentiated brown adipocytes pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently stimulated with 100 nM isoproterenol (ISO) or 20 µM SR-3420 for 1 h before harvest. n = 5 biologically independent experiments examined. g Intracellular ATP: ADP ratio in mouse in vitro differentiated brown adipocytes pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently stimulated with 100 nM isoproterenol (ISO) or 20 µM SR-3420 for 1 h before harvest. n = 4 biologically independent experiments examined. h Intracellular ATP levels in mouse in vitro differentiated brown adipocytes pre-treated with 5 µM triacsin C for 1 h and subsequently stimulated with 100 nM isoproterenol (ISO) or 20 µM SR-3420 for 1 h before harvest. n = 4–5 biologically independent experiments examined. For all panels, error bars represent ± SEM of 2–7 independent biological experiments, each carried out in technical duplicates. Statistical significance was determined by one-way ANOVA with Tukey’s multiple comparisons test for b, h and by DESeq2 using FDR/Benjamini–Hochberg correction for d (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus Vehicle/Control, # versus SR-3420/Control, ¤ versus Vehicle/Control.
Fig. 3
Fig. 3. Lipolysis-activated genes are induced by ISO by PPAR-dependent as well as -independent mechanisms.
Mouse in vitro differentiated brown adipocytes were treated with siRNA against Pparg, Ppara, or NT (negative control) for 3 days and subsequently with 100 nM isoproterenol (ISO) for 3 h before harvest. a Model indicating the potential role of PPARs in the transcriptional activation of genes in C1, C3 and C4 by β-adrenergic signals. Created with BioRender.com. b mRNA expression of adipocyte genes with or without Pparg knockdown in mature brown adipocytes quantified using qPCR. n = 2 biologically independent experiments examined, each carried out in a technical duplicate. c Micrograph of mature brown adipocytes with or without Pparg knockdown. Pictures are representative of n = 2 biologically independent experiments examined. Scale bar denotes 10 µM. d mRNA expression pattern of representative PPAR-dependent and -independent genes derived from RNA-seq after stimulation with 100 nM ISO for 3 h. n = 2 biologically independent experiments examined, each carried out in technical duplicates. e Pathways significantly enriched (FDR < 0.05) among PPAR-dependent and -independent genes in mouse-brown adipocytes stimulated with 100 nM ISO for 3 h. For all panels, error bars represent ±SEM of 2 independent biological experiments. Statistical significance was determined by DESeq2 using FDR/Benjamini-Hochberg correction for b, c (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus siNT, # versus Vehicle.
Fig. 4
Fig. 4. Machine-learning strategy infers transcription factors mediating the transcriptional effects of ISO and lipolysis.
Mouse in vitro differentiated brown adipocytes were pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently with 100 nM isoproterenol (ISO) or 20 µM SR-3420 for 3 h before harvesting for ChIP and ATAC. a Heatmap of H3K27ac levels in ATAC regions (ISO vs. vehicle and SR-3420 vs. vehicle: Log2FC ≥ 0.5 or Log2FC  ≤−0.5) of mouse-brown adipocytes stimulated ISO or SR-3420 for 3 h before harvest. The results are 2 independent biological experiments. b K-means clustering of motif activity (calculated using IMAGE) of 127 transcription factor motifs, mRNA levels, and reported function of the 23 lipolysis-dependent transcription factor motifs in mouse-brown adipocytes.
Fig. 5
Fig. 5. Activation of unfolded protein response by lipolysis.
Mouse in vitro differentiated brown adipocytes were pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently with 100 nM isoproterenol (ISO) (blue bar) or 20 µM SR-3420 (red bar) for 3 h before harvest. a Boxplot illustrating mRNA expression derived from RNA-seq, expressed as fold change activation by ISO and SR-3420 ±lipase inhibitors of 24 established UPR marker genes relative to the vehicle. Boxplots (notch, mean; box, first and third quartiles; whiskers, 1.5 times the interquartile range. n = 3 biologically independent experiments examined, each carried out in a technical duplicate b Model depicting the different UPR signaling pathways. Created with BioRender.com. c Heatmap indicating the mRNA expression of UPR marker genes derived from RNA-seq of the three UPR signaling branches in response to the different treatments. n = 3 biologically independent experiments examined, each carried out in technical duplicates. d Western blot showing levels of CHOP and HSPA5 in in vitro differentiated mouse-brown adipocytes stimulated with 20 µM SR-3420 for 1, 2, 4, and 6 h before harvest. Amido black is used as loading control. mRNA expression of spliced Xbp1 / unspliced Xbp1 derived from RT-qPCR n = 4 biologically independent experiments examined for RT-qPCR data and Western blots are representative of n = 2 biologically independent experiments examined e Heatmap indicating the enrichment of gene pathways (FDR < 0.05) among inferred ATF6, XBP1 or ATF4 _target genes in mouse-brown adipocytes. f mRNA expression of inferred ATF4 _targets genes quantified by qPCR of in vitro differentiated mouse-brown adipocytes with or without knockdown of Atf4 (siAtf4) and subsequently stimulated with 20 µM SR-3420 for 3 h. n = 2 biologically independent experiments examined, each carried out in a technical duplicate For all panels, error bars represent ±SEM of 2–4 independent biological experiments, each carried out in technical duplicates. Statistical significance was determined by Wilcoxon Signed-rank test for a and one-way ANOVA with Tukey’s multiple comparisons test for d, f (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus Vehicle/Control, # versus ISO or SR-3420 (without Lipase inh. or siNT).
Fig. 6
Fig. 6. Lipolysis overrides the intrinsic rhythmicity of brown adipocytes in a cell-autonomous manner.
a Effect of lipolysis on mRNA expression of core clock genes derived from RNA-seq in unsynchronized mouse in vitro differentiated brown adipocytes. Cells were pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently with 20 µM SR-3420 for 3 h before harvest. n = 3 biologically independent experiments examined, each carried out in a technical duplicate b Effect of lipolysis on mRNA expression of circadian clock genes quantified by qPCR in synchronized in vitro differentiated mouse-brown adipocytes. Cells were synchronized with serum and exposed to 20 µM SR-3420 (red line) or vehicle (black line) for the indicated times. n = 3 biologically independent experiments examined, each carried out in a technical duplicate c Experimental setup for d-h. Male C57BL/6 mice (12 weeks old) received 1.5 mg Atglistatin (red) or vehicle (black) by oral gavage at ZT0 and euthanized at ZT4, ZT8, ZT12, and ZT16. Created with BioRender.com. d Serum FAs levels at ZT4. n = 5 mice per condition. e Serum glycerol levels at ZT4. n = 5 mice per condition. f Correlations between serum FA levels and mRNA expression of core clock genes in BAT of vehicle-treated mice at ZT4, ZT8, ZT12, and ZT16. g Effect of Atglistatin on mRNA expression of core clock genes quantified by qPCR in brown adipose tissue at ZT4, ZT8, ZT12, and ZT16. n = 5 mice per condition per timepoint. h Effect of Atglistatin on core body temperature from ZT0-ZT16 (left), and quantification of core body temperature in the light phase from ZT2-ZT10 (right). n = 6 mice per condition. For cell culture experiments results, error bars represent + /- SEM of 3 independent biological experiments, each carried out in technical duplicates. For mice experiments, error bars represent SEM of 5-6 mice per treatment each time point. Statistical significance was determined by Pearson’s correlation coefficient test for f, one-way ANOVA with Tukey’s multiple comparisons test for a and unpaired Student’s t-test for d, e, g, h and area under the curve analysis for b (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus Vehicle/Control, # versus SR-3420 (without Lipase inh.).
Fig. 7
Fig. 7. Lipolysis activates overlapping gene programs in human brown adipocytes.
a Heatmap of RNA-seq data showing lipolysis-activated genes (Log2FC ≥ 0.7, FDR/Benjamini–Hochberg ≤ 0.05) in human in vitro differentiated brown adipocytes. Cells were stimulated with 20 µM SR-3420 or vehicle for 3 h before harvest. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. b Pathways significantly enriched (FDR ≤ 0.05) among lipolysis-activated genes in a. c Examples of lipolysis-activated genes derived from RNA-seq belonging to enriched pathways in b. d Heatmap of RNA-seq data showing dependency of lipolysis for induction of genes by ISO Log2FC ≥ 1, FDR/Benjamini-Hochberg ≤0.05) in white adipocytes. Mature mouse white adipocytes were isolated from epididymal white adipose tissue and cultured in a 3D matrix for 2 days. After 2 days, cells were pre-treated with 10 µM Atglistatin (ATGL inhibitor) and 20 µM CAY10499 (HSL inhibitor) for 1 h and subsequently with 20 µM SR-3420 for 3 h before harvest. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. e Pathways significantly enriched (FDR ≤ 0.05) among lipolysis-dependent and -independent ISO-activated genes in d. f Examples of lipolysis-dependent and -independent ISO-activated genes derived from RNA-seq in d. n = 3 biologically independent experiments examined, each carried out in a technical duplicate. g Venn diagram illustrating the overlap of lipolysis-dependent genes in mouse white (d) and brown adipocytes (Fig. 1d). For all panels, error bars represent ±SEM of 3 independent biological experiments. Statistical significance was determined by using FDR/Benjamini-Hochberg correction for Fig. 1f (p ≤ 0.05 = *, p ≤ 0.01 = **, p ≤ 0.001 = ***). * versus Vehicle, # versus ISO (without Lipase inh.).

Similar articles

Cited by

References

    1. Zechner R, et al. FAT SIGNALS–lipases and lipolysis in lipid metabolism and signaling. Cell Metab. 2012;15:279–291. doi: 10.1016/j.cmet.2011.12.018. - DOI - PMC - PubMed
    1. Cannon B, Nedergaard J. Brown adipose tissue: function and physiological significance. Physiol. Rev. 2004;84:277–359. doi: 10.1152/physrev.00015.2003. - DOI - PubMed
    1. Dawkins MJ, Hull D. Brown adipose tissue and the response of new-born rabbits to cold. J. Physiol. 1964;172:216–238. doi: 10.1113/jphysiol.1964.sp007414. - DOI - PMC - PubMed
    1. Granneman JG, Moore H-PH, Krishnamoorthy R, Rathod M. Perilipin controls lipolysis by regulating the interactions of AB-hydrolase containing 5 (Abhd5) and adipose triglyceride lipase (Atgl) J. Biol. Chem. 2009;284:34538–34544. doi: 10.1074/jbc.M109.068478. - DOI - PMC - PubMed
    1. Miyoshi H, et al. Control of adipose triglyceride lipase action by serine 517 of perilipin A globally regulates protein kinase A-stimulated lipolysis in adipocytes. J. Biol. Chem. 2007;282:996–1002. doi: 10.1074/jbc.M605770200. - DOI - PubMed

Publication types

MeSH terms

Substances

  NODES
Note 1
Project 1
twitter 2