Lipolysis regulates major transcriptional programs in brown adipocytes

Markussen, Lasse K.; Rondini, Elizabeth A.; Johansen, Olivia Sveidahl; Madsen, Jesper G. S.; Sustarsic, Elahu G.; Marcher, Ann-Britt; Hansen, Jacob B.; Gerhart-Hines, Zachary; Granneman, James G.; Mandrup, Susanne

doi:10.1038/s41467-022-31525-8

Download PDF

Article
Open access
Published: 08 July 2022

Lipolysis regulates major transcriptional programs in brown adipocytes

Nature Communications volume 13, Article number: 3956 (2022) Cite this article

9288 Accesses
16 Citations
34 Altmetric
Metrics details

Subjects

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.

A conserved complex lipid signature marks human muscle aging and responds to short-term exercise

Article 12 April 2024

Adipose tissue macrophages secrete small extracellular vesicles that mediate rosiglitazone-induced insulin sensitization

Article 11 April 2024

Trigonelline is an NAD+ precursor that improves muscle function during ageing and is reduced in human sarcopenia

Article Open access 19 March 2024

Introduction

Adipocytes have the capacity to efficiently store large amounts of metabolic energy as triacylglycerol and mobilize this energy as fatty acids (FAs) when lipolysis is stimulated¹. During lipolysis, triacylglycerol molecules are hydrolyzed to FAs and glycerol in a process that is controlled by functional interactions of adipose triglyceride lipase (ATGL), α-β hydrolase domain containing 5 (ABHD5), and perilipins 1 and 5 (PLIN1/5). In white adipocytes, lipolysis is primarily stimulated in response to fasting, and the FAs are released to the circulation to be used as metabolic energy by other cell types, such as muscle cells. In brown adipocytes, FAs are mainly mobilized in response to cold exposure, where they provide the main fuel for thermogenesis and induce uncoupling of mitochondrial oxidative phosphorylation^2,3.

β-Adrenergic signaling represents the most powerful inducer of lipolysis in both white and brown adipocytes, acting via the cAMP/protein kinase A (PKA) signaling pathway. PKA phosphorylates hormone-sensitive lipase (HSL) and PLIN1/5, causing the release of ABHD5, thereby enabling ABHD5 to bind to and activate ATGL^4,5. Importantly, β-adrenergic signaling also induces genes supporting brown adipocyte function and differentiation (e.g., thermogenic genes, fatty acid oxidation-related genes, and mitochondrial genes) in both brown and white adipocytes. Here, PKA signaling in response to the increased level of cAMP is thought to play a major role. For example, PKA has been shown to phosphorylate the transcription factor cAMP-response element-binding protein (CREB)⁶ and to indirectly activate activating transcription factor 2 (ATF2)⁷. Thus, β-adrenergic activation of PKA activates two parallel pathways in brown adipocytes that are viewed as largely independent: gene transcription and lipolysis.

Interestingly, previous genetic loss-of-function studies have indicated that abrogation of lipolysis affects gene expression in brown adipocytes. Thus, global or adipocyte-specific deletion of ATGL or ABHD5 in mice leads to whitening of brown adipose tissue^8,9,10,11. Furthermore, pharmacological inhibition of HSL in cultured brown and white adipocytes interferes with the ability of β-adrenergic signals to activate some brown adipocyte genes^12,13. Other studies indicate that FAs released by intracellular lipolysis may activate peroxisome proliferator-activated receptors (PPARs) to regulate gene expression. Thus, lipolysis activates reporter gene expression driven by PPARα and PPARδ in brown adipocytes; and the effect of cardiac muscle-specific deletion of ATGL can be rescued by PPARα agonists^13,14.

While these studies suggest a role of lipolysis in modulating gene expression, the extent to which lipolysis signals regulate transcription and the importance of this for the acute transcriptional effects of β-adrenergic signals in adipocytes is currently unknown. A major reason for this gap in our knowledge is that it only recently has become possible to activate intracellular lipolysis without simultaneously activating cAMP-activated pathways. Thus, it has not previously been possible to determine the repertoire of lipolysis-activated genes in the absence of increased cAMP.

In this work, we use ABHD5 ligands (SR-3420 and SR-4995)^15,16,17 that activate ATGL-mediated lipolysis independently of receptor- and cAMP-mediated activation to provide global insight into the ability of lipolysis to activate transcription in brown adipocytes. We show that lipolysis is a powerful signal that exerts profound and acute effects on several major transcriptional programs in brown adipocytes and that lipolysis is required for full activation of the β-adrenergic gene program.

Results

Lipolysis is sufficient and necessary for activation of a large fraction of the β-adrenergic gene program in brown adipocytes

To investigate the importance of lipolysis for acute transcriptional regulation by β-adrenergic signals in brown adipocytes, we differentiated immortalized mouse brown preadipocytes and exposed them to the pan β-adrenergic receptor agonist isoproterenol (ISO) in the presence or absence of inhibitors of lipolysis (Fig. 1a). As previously shown for other adipocyte cell systems¹⁶, dual inhibition using lipase inhibitors of ATGL and HSL, completely blocks both basal and stimulated lipolysis, as determined by the level of FAs released into the medium. Furthermore, we included cells treated with the ABHD5 ligand, SR-3420, which is known to result in acute activation of ATGL, the rate-limiting step of lipolysis¹⁶. Mechanistically, SR-3420 specifically binds ABHD5 and stimulates lipolysis by releasing it from PLIN, thereby mimicking the effects of β-adrenergic PKA-dependent activation of ATGL^15,16,18. Importantly, consistent with previous reports^15,16, treatment of brown adipocytes with SR-3420 induces lipolysis to the same degree as ISO (Fig. 1b) without activating PKA, as determined by phosphorylation of HSL (Supplementary Fig. 1a). We have previously reported that EYFP-coupled to the LXXLL-containing domain of steroid co-activator 1 (SRC1) translocates to the nucleus of brown adipocytes in response to lipolysis, presumably as a result of the recruitment of the SRC1 fusion protein to nuclear receptors activated by increased levels of fatty acids in the nucleus¹⁷. Using this system, we can detect SRC1 accumulation in the nucleus within minutes in response to direct stimulation of lipolysis (Supplementary Fig. 1b), indicating that fatty acid levels in the nucleus are immediately increased in response to lipolysis.

To examine the acute effects of lipolysis on gene expression in brown adipocytes and determine the importance of lipolysis for transcriptional activation by β-adrenergic signals, we used the stimulation protocol depicted (Fig. 1a) and performed RNA-seq. Principal component analysis (PCA) shows that PC1 primarily separates the lipolysis-independent effect of ISO (resistant to the use of lipase inhibitor), whereas PC2 primarily separates the lipolysis-dependent effects on the transcriptome by ISO and SR-3420 (Fig. 1c). Based on the PCA analysis, SR-3420 treatment partly mimics ISO treatment with a substantial change in both PC1 and PC2, indicating that lipolysis regulates many of the same genes as ISO. Importantly, and consistent with the high specificity of SR-3420, lipase inhibitors almost completely block the effect of SR-3420. In total, 630 genes display a robust induction by ISO (Log2FC ≥ 1, pAdj ≤ 0.05), and 324 out of these also display a robust induction by SR-3420 (Log2FC ≥ 1, pAdj ≤ 0.05), indicating that lipolysis per se has profound and acute effects on transcription in brown adipocytes and that these effects mimic many of the acute transcriptional effects of β-adrenergic signals. K-means clustering of the 630 ISO-activated genes based on their dependency on lipolysis show four distinct clusters (Fig. 1d, e and Supplementary Data File 1). Cluster 1 (C1) represents genes where lipolysis is sufficient and necessary, i.e., the genes are activated by ISO in a highly lipolysis-dependent manner and to a similar extent by SR-3420. Intriguingly, this cluster is enriched for genes involved in biological processes related to cold-induced thermogenesis and UPR, and examples include Pdk4, Atf4, Plin2, and Fgf21 (Fig. 1f, g). Genes in Cluster 3 (C3) are partially dependent on lipolysis, as their induction by ISO is only partially blocked by lipase inhibitors, and since lipolysis alone (SR-3420, no lipase inhibitor) does not activate these genes to the same extent as ISO. This indicates that lipolysis-derived signals cooperate with other ISO-stimulated signals to induce maximal expression of these genes. C3 is enriched for genes involved in fat cell differentiation and cold-induced thermogenesis, and Pck1 and Elovl3 are prominent examples (Fig. 1f, g). Interestingly, we also identified genes that are strongly activated by ISO and activated to a greater extent in the absence of lipolysis (Cluster 2; C2); however, the repressive effect of lipolysis is conditional since SR-3420 treatment does not affect the expression of these genes. This indicates that lipolysis dampens the ability of ISO to activate this group of genes. C2 is enriched for genes involved in protein glycosylation, but also includes genes such as Crem and Vdr, which are known to be strongly activated by cAMP signaling^19,20. Notably, lipase inhibitors specifically increase gene programs induced by ISO/PKA, not gene programs induced by SR-3420, indicating that PKA signaling is increased by inhibition of lipolysis. Mechanisms may involve relief of FA-mediated inhibition of adenylyl cyclase^12,21,22, thereby leading to increased levels of cAMP and potentially higher expression of PKA target genes. Finally, Cluster 4 (C4) represents genes that are activated by ISO independent of whether lipolysis is inhibited or not. Importantly, however, most of these genes are also activated by SR-3420 and this activation is blocked by lipase inhibitors, indicating that lipolysis and cAMP-activated pathways can activate these genes independently. Thus, lipolytic products are sufficient, but not necessary for the induction of the ISO-activated genes in C4. This cluster is enriched for genes involved in lipid metabolism, exemplified by Acot4 and Ucp1 (Fig. 1f, g). In summary, 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 ISO independent of lipolysis (Fig. 1e).

As a complementary approach to pharmacological lipase inhibitors, we generated a brown adipocyte cell line with a stable knockdown of ABHD5, which upon doxycycline treatment re-established the expression of ABHD5 (Supplementary Fig. 1c). Consistent with ABHD5 being necessary for the activation of ATGL, the expression of ABHD5 (+Dox) significantly induces FA to release into the medium (Supplementary Fig. 1d). Clustering of genes relative to their regulation by ISO administration and ABHD5 expression revealed a group of genes strictly dependent on ABHD5 for induction by SR-3420 and ISO (Supplementary Fig. 1e, f, C3 and Supplementary Table 1). Similar to the genes activated by ISO in a highly lipolysis-dependent manner in Fig. 1d; C1+C3, this cluster of genes are enriched for genes related to lipid metabolism and thermogenesis, including Pdk4 and Plin2, (Supplementary Fig. 1g, h).

Taken together, these results show that lipolysis has acute and profound effects on the brown adipocyte transcriptome. Lipolysis is sufficient for acute activation of major gene programs and required for transcriptional regulation of a wide range of genes activated by β-adrenergic signals in brown adipocytes.

Lipolysis activates gene expression independent of fatty acid activation or oxidation

Although glycerol and mono- and di-acylglycerides can also be considered lipolytic products, the main product of lipolysis is FAs. To investigate whether activation or oxidation of FAs is required for the transcriptional effects of lipolysis, we blocked the activation of FAs using triacsin C, an inhibitor of long-chain fatty acid-CoA synthases (ACSLs), or the oxidation of FA using etomoxir, an inhibitor of carnitine palmitoyl transferase 1 (CPT1), which catalyzes the rate-limiting step for entry of activated fatty acids into the mitochondria, and thereby fatty acid oxidation (Fig. 2a). Triacsin C markedly elevates basal release of FA into the medium, whereas etomoxir has no effect (Fig. 2b), indicating that β-oxidation is not very active under unstimulated conditions. As expected, and consistent with the blockade of fatty acid metabolism, both etomoxir and triacsin C modestly increase the release of FAs in response to SR-3420 stimulation (Fig. 2b). Interestingly, the induction of most lipolysis-activated genes by SR-3420 (e.g., Pdk4 and Plin2) is not compromised by the inhibitors (Fig. 2c, d). In fact, these “lipolysis-activated, FA activation-independent” genes are modestly elevated by triacsin C alone, consistent with the ability of this inhibitor to increase FAs levels in the medium (Fig. 2b), and the notion that FAs are the main mediators of the transcriptional response to lipolysis. Only a minor fraction of the lipolysis-activated genes (e.g., Cyp26b1 and Ffar4) are dependent on acyl-CoA synthesis or FA oxidation (i.e., they are repressed by triacsin C or etomoxir treatment) (Fig. 2c, d).

Previous reports^23,24 have shown that β-adrenergic stimulation leads to a decrease in ATP levels by mechanisms that may involve inhibition of ATP synthesis by FAs²⁵ and/or consumption of ATP for FA-CoA esterification^23,26. Consistent with these reports, ISO treatment leads to a lipolysis-dependent decrease in ATP levels (Fig. 2e) and an increase in the ADP levels (Fig. 2f). Furthermore, direct lipolytic stimulation reduces ATP levels and increases ADP levels to the same degree as ISO (Fig. 2e–g). The finding that the lipolysis-induced decrease in ATP levels is not blocked by triacsin C (Fig. 2h), suggests that the decrease in ATP is primarily caused by FA-dependent inhibition of ATP synthesis. Taken together, these results indicate that lipolysis signals to the genome by mechanisms that are largely independent of FA oxidation and acyl-CoA synthesis but which might involve reduced ATP levels.

Lipolysis induces transcription by PPAR-dependent as well as -independent mechanisms

Induction of lipolysis has previously been shown to increase the expression of selected genes linked to lipid metabolism in brown adipocytes through members of the PPAR family¹³; however, the overall contribution of PPAR signaling to the lipolysis-activated transcriptional reprogramming is unknown. In addition to PPARγ, which is the master regulator of adipogenesis and the main PPAR subtype in all adipocytes, brown adipocytes express high levels of PPARα²⁷. PPARα and PPARγ have traditionally been viewed as activators of opposing metabolic fates of fatty acids, oxidation versus storage²⁸; however, it has been shown that the two PPAR subtypes bind to many of the same target enhancers in brown adipocytes²⁷. Consistent with overlapping functions of PPARγ and -α in adipocytes and a dominant role of PPARγ, recent reports suggest that PPARγ, but not PPARα, is necessary for the thermogenic function of brown adipocytes in vivo^27,29.

To investigate the role of PPARs in lipolysis-activated activation of gene expression in brown adipocytes, we performed siRNA-mediated knockdown of PPARγ and PPARα in mature brown adipocytes and applied RNA-seq to assess to what extent lipolysis-induced activation of genes in Cluster 1, 3 and 4 (Fig. 1d) is dependent on PPARs (Fig. 3a). PPARα and PPARγ knockdown resulted in ∼60% and ∼90% knockdown of Ppara and Pparg, respectively (Fig. 3b and Supplementary Fig. 2a). PPARα knockdown had minimal effect on lipolysis-activated gene expression (C1 + C3 + C4 genes) (Supplementary Fig. 2b); however, in line with previous work^30,31, PPARγ knockdown reduced the expression of a large number of adipocyte genes in unstimulated cells (Fig. 3b), an effect that could lead to an overestimation of the number of PPAR-dependent genes. Notably, PPARγ knockdown resulted in ∼60% knockdown of Ppara (Fig. 3b), consistent with previous reports indicating that expression of Ppara is dependent on PPARγ²⁹. Thus, PPARγ-knockdown in these adipocytes is functionally equivalent to double-knockdown of PPARα and PPARγ. Importantly, this transient knockdown of PPARγ/α does not significantly affect lipid content in the short time frame of the experiment (Fig. 3c).

**Fig. 3: Lipolysis-activated genes are induced by ISO by PPAR-dependent as well as -independent mechanisms.**

Notably, PPARγ/α knockdown suppressed the ability of β-adrenergic signals to activate 233 (45%) of the 525 genes from C1 + C3 + C4. This indicates that PPARs are important mediators of the transcriptional effects of ISO and lipolysis, extending far beyond the few genes previously reported to be regulated by ISO through PPAR-dependent mechanisms^13,14. Intriguingly, however, the majority of lipolysis-activated genes (292; 55%) of the 525 genes from C1 + C3 + C4) are not affected by the PPARγ/α knockdown (Fig. 3d). Pathway analysis of the 233 PPAR-dependent and the 292 PPAR-independent lipolysis-activated genes showed that PPARγ/α preferentially mediate the effects of lipolysis on metabolic genes, whereas activation of UPR and circadian clock genes by lipolysis is independent of PPARs (Fig. 3e).

Taken together, these findings show that while PPARs constitute key downstream mediators of transcriptional activation in response to ISO, more than half of the lipolysis-activated genes are activated by PPAR-independent mechanisms.

Inferring PPAR-independent mechanisms by machine-learning-based interrogation of enhancer activity and transcriptome

To make an unbiased prediction of the PPAR-independent mechanisms by which lipolysis regulates transcription in brown adipocytes, we took advantage of our recently developed bioinformatic tool Integrated analysis of Motif Activity and Gene Expression change of transcription factors (IMAGE). This tool uses a two-stage machine-learning strategy to comprehensively predict the contribution of transcription factor motifs to the observed changes in gene expression based on genome-wide profiles of enhancer activity and gene expression³². The output from IMAGE is a ‘motif activity score’ for each transcription factor motif, indicating the relative contribution of the transcription factor motif to enhancer activity and gene expression under the given condition. Since the contribution of a motif to enhancer activity reflects the activity of the transcription factor(s) binding to the particular motif under the given conditions, these analyses can be used to infer the activities of all transcription factors expressed in the cell population. We have shown that this approach accurately predicts causal transcription factors^32,33. Notably, by using a machine-learning approach we can use the sequence information in all active enhancers rather than focusing only on motifs enriched in enhancers near PPAR-independent genes. This is important because genes are regulated by many different transcription factors and signaling networks, meaning that many genes may be regulated by both PPAR-dependent as well as several PPAR-independent mechanisms.

IMAGE infers causal transcription factors based on RNA-seq and genome-wide maps of enhancer activity. As inputs into the IMAGE analysis, we performed transposase accessible chromatin sequencing (ATAC-seq) to define enhancer position and map chromatin accessibility³⁴, and H3K27 acetylation chromatin immunoprecipitation (ChIP)-seq profiles to assess enhancer activity^35,36. Mapping of chromatin accessibility by ATAC-seq in brown adipocytes stimulated or not with ISO or SR-3420 identified 13,003 robust putative enhancers (i.e., accessible regions, not overlapping (±500 bp) with annotated promoters). The ATAC-seq profiles are overall very similar in unstimulated and stimulated brown adipocytes despite substantial changes in the transcriptomes (Supplementary Fig. 3a). Nevertheless, genome-wide mapping of H3K27ac marks indicated that the activity of 28% of brown adipocyte enhancers (3725 of 13,003) is dynamically regulated in response to ISO or SR-3420 (Log2FC > 0.5 or Log2FC < -0.5) (1684 SR-3420-selective, 827 ISO-selective and 1214 common) (Fig. 4a and Supplementary Fig. 3b). Thus, consistent with the major changes in gene expression, induction of lipolysis leads to acute activation of many enhancers; however, this activation appears to involve primarily activation of pre-established enhancers.

**Fig. 4: Machine-learning strategy infers transcription factors mediating the transcriptional effects of ISO and lipolysis.**

IMAGE predicted that 617 transcription factor motifs contribute to enhancer activity in brown adipocytes under the given conditions. Of these, 127 transcription factor motifs are scored as high-to-medium confidence causal regulators of the transcriptional changes in response to ISO (Supplementary Data File 2). K-means clustering of these 127 high-to-medium confidence transcription factor motifs involved in the response to ISO based on their motif activity pattern revealed at least 4 distinct clusters (Fig. 4b; left heatmap). Cluster 2 contains many motifs that are primarily activated by ISO in a lipolysis-independent manner and that are not activated by SR-3420. These include known motifs of transcription factors known to be activated by cAMP and to be involved in thermogenesis, like CREM³⁷, CREBL2³⁸, and NOR1/NR4A3^13,39. By contrast, Cluster 1 (23 motifs) stands out as a cluster where the activity of all motifs is highly lipolysis-dependent, i.e., the contribution of these motifs to enhancer activity and gene expression is strongly increased by ISO as well as SR-3420 and abolished by lipolysis inhibitors (Fig. 4b; Cluster 1 Motif activity). Reassuringly, we find the PPAR and RXR motif in Cluster 1; however, this cluster also contains several other motifs which appear to contribute significantly to lipolysis-activated gene expression. Interestingly, and consistent with the finding that ISO stimulates gene programs related to UPR and circadian rhythm in a PPAR-independent manner (Fig. 3e), many of the transcription factors binding to motifs in Cluster 1 are known to be involved in the UPR or to regulate circadian rhythm (Fig. 4b). Thus, the motif activity as well as the expression of the transcription factors ATF4, ATF3, JunD, and C/EBPβ, known to mediate the transcriptional effects of UPR⁴⁰, are markedly increased by lipolysis and by ISO in a lipolysis-dependent manner. Similarly, the motif activity of the circadian transcription factors RORγ⁴¹, REV-ERBα⁴¹, BHLHE40⁴², and EGR⁴³ are also highly activated by lipolysis and by ISO in a lipolysis-dependent manner.

Taken together, the IMAGE analysis identified several transcription factors predicted to be involved in mediating the acute effects of lipolysis in brown adipocytes treated with ISO or SR-3420. In addition to PPARs, these include many factors involved in mediating the transcriptional response to UPR and in regulating circadian rhythm.

Lipolysis leads to UPR-mediated activation of metabolic genes

The observation that lipolysis induces genes in the UPR gene program (Fig. 1f) mainly independent of PPAR signaling (Fig. 3e) and that several UPR-activated transcriptions factors are predicted to be key mediators of the acute transcriptional activation of enhancers in response to lipolysis (Fig. 4b), prompted us to further dissect the UPR gene program. Interestingly, both SR-3420 and ISO induce the expression of 24 well-established UPR markers (Supplementary Data File 3) by approximately two-fold relative to the vehicle, and this induction is blunted by lipase inhibitors (Fig. 5a). This indicates that transcriptional effectors of UPR are activated in response to lipolysis.

**Fig. 5: Activation of unfolded protein response by lipolysis.**

In mammals, UPR activates three different stress sensors in the ER: ATF6, protein kinase R-like endoplasmic reticulum kinase (PERK) and serine/threonine-protein kinase/endoribonuclease inositol-requiring enzyme 1α (IRE1α), which each signal to distinct transcriptional mediators, ATF6, XBP1, and ATF4, respectively (Fig. 5b). Interestingly, specific target genes of all three UPR signaling branches are activated by ISO and SR-3420 in a highly lipolysis-dependent manner (Fig. 5c). Consistent with that, Xbp1 mRNA splicing is increased by ISO and SR-3420 in a highly lipolysis-dependent manner. Furthermore, protein levels of the ATF4 target gene Chop and of the ATF6 target gene Hspa5 are increased by SR-3420 (Fig. 5d). Taken together these results show that lipolysis leads to acute activation of all three UPR signaling pathways.

For each motif and associated transcription factor(s), IMAGE also accurately predicts target enhancers and target genes^32,33. We used this feature to infer target genes for ATF6, XBP1, and ATF4 in brown adipocytes and subsequently submitted these target genes to gene ontology analysis, revealing that the three arms are predicted to engage different gene programs (Fig. 5e). Thus, in response to lipolysis, ATF6 is predicted to primarily regulate genes involved in the ER signaling, XBP1 is predicted to regulate genes involved in ribosome assembly and response to increased oxygen levels, whereas ATF4 appears to be most directly linked to metabolism. Inferred ATF4 target genes are highly enriched for biological processes related to fatty acid oxidation, triglyceride biosynthesis, mitochondrial function, and thermogenesis (Fig. 5e). In addition, both XBP1 and ATF6 are predicted to activate Atf4 gene expression, suggesting that ATF4 may constitute a key transcriptional link between UPR and metabolism. Consistent with this, knockdown of ATF4 compromised the ability of lipolysis to activate the expression of inferred ATF4 target genes such as Dgat1 and Acsl1 (Fig. 5f).

Collectively, these results show that β-adrenergic activation of brown adipocytes leads to lipolysis-dependent activation of all three UPR signaling pathways. Key transcription factors in these arms, XBP1, ATF4, and ATF6 appear to target distinct downstream pathways, where ATF4 is most closely linked to metabolism.

Lipolysis overrides the intrinsic rhythmicity of brown adipocytes in a cell-autonomous manner

The finding that IMAGE identifies circadian transcription factors as putative regulators of enhancer activity and gene expression in response to lipolysis is particularly interesting, as it suggests a potential link between lipolysis and circadian regulation that is independent of PPARs. In line with these findings, SR-3420-stimulated lipolysis activates the expression of core clock genes (i.e., Reverba, Bmal1, and Cry1) (Fig. 6a).

To further explore the ability of lipolysis to regulate circadian gene expression, we serum-synchronized brown adipocytes and subsequently stimulated lipolysis with SR-3420. Interestingly, treatment with SR-3420 overrides the intrinsic rhythmicity of the core clock genes and appears to induce a new state of synchrony as evidenced by the strongly superimposed 24 h rhythmicity after SR-3420 treatment (Fig. 6b). These lipolysis-induced alterations in rhythmicity and amplitude of the core clock expression suggest that lipolytic products may play a dominant role over the normal course of oscillation.

Our in vitro results prompted us to investigate how lipolysis affects circadian gene expression in brown adipose tissue in vivo (Fig. 6c). Previous studies in mice have shown that during normal circadian regulation REV-ERBα protein levels in BAT peaks at ZT10⁴⁴ and that maximal lipolysis precedes the peak of REV-ERBα protein expression in BAT, with FAs peaking around Zeitgeber time 6-8 (ZT6-ZT8) in BAT and serum⁴⁵ (Supplementary Fig. 4a). Considering the findings in the present study, this pattern suggests that lipolysis might also induce Reverba gene expression in brown adipocytes in vivo. To investigate the effect of lipolysis on the expression of Reverba and other circadian genes, we treated male C57BL/6N mice with Atglistatin or vehicle by oral gavage at ZT0 (Fig. 6c). As expected, Atglistatin significantly decreases FAs and glycerol in serum of ad libitum-fed mice at ZT4 (Fig. 6d, e) indicating reduced systemic lipolysis. Notably, in control mice, the expression of Reverba and Cry1 in BAT significantly correlates with FAs in serum (Fig. 6f). Moreover, Atglistatin treatment reduces Reverbα expression in BAT and reciprocally increases expression of Bmal1 (Fig. 6g), whereas circadian genes in WAT and liver are not affected (Supplementary Fig. 4b). This indicates that acute manipulation of lipolysis in vivo specifically affects the circadian clock in BAT. Intriguingly and consistent with REV-ERBα being a repressor of heat production in brown adipocytes⁴⁶, we observed a transient increase in core body temperature in the light phase after Atglistatin treatment (Fig. 6h).

Collectively, these results show that lipolysis can alter circadian patterns and change the amplitude of oscillations in brown adipose tissue in vivo and brown adipocytes in vitro. These results place signals generated by lipolysis as cell-autonomous Zeitgebers that feed into the intrinsic brown adipocyte clock to influence circadian rhythmicity.

Lipolysis regulates overlapping gene programs in human brown adipocytes

Having established profound and acute transcriptional effects of lipolysis in mouse brown adipocytes, we asked whether lipolysis activates a similar gene program in human brown adipocytes. We, therefore, stimulated immortalized human brown adipocytes⁴⁷ with SR-3420 for 3 h and performed RNA-seq analysis. Similar to brown adipocytes from mice, acute activation of lipolysis by SR-3420 induces the expression of a large group of genes (149 genes, Log2FC ≥ 0.7, pAdj ≤ 0.05) in human brown adipocytes (Fig. 7a). Consistent with the data from mouse brown adipocytes, lipolysis activates the expression of genes related to brown fat cell differentiation, UPR/stress, circadian rhythm, fatty acid oxidation, and thermogenesis (Fig. 7b, c). Thus, the ability of lipolysis to acutely activate these gene programs is conserved between mouse and human brown adipocytes.

**Fig. 7: Lipolysis activates overlapping gene programs in human brown adipocytes.**

Finally, to investigate whether lipolysis regulates a similar gene program in white adipocytes from mice, we treated isolated epididymal white adipocytes with ISO for 3 h in the presence and absence of lipase inhibitors and determined the effect on the transcriptome by RNA-seq analysis. Similar to brown adipocytes, lipolysis is required for the activation of a large number of ISO-activated genes in white adipocytes (Fig. 7d). However, in contrast to brown adipocytes, the ISO-induction of genes involved in brown fat cell differentiation and fatty acid oxidation is independent of lipolysis in mouse white adipocytes (Fig. 7e, f). Moreover, and consistent with our in vivo study, lipolysis does not regulate circadian genes in mouse white adipocytes (Supplementary Fig. 4b). In line with these observations, the overlap between lipolysis-activated genes in brown (C1 + C3 + C4 genes) and white adipocytes from mice is very small (Fig. 7g). Thus, lipolysis regulates distinct gene programs in brown and white adipocytes from mice.

Discussion

In this study, we show that lipolysis has profound effects on transcriptional regulation in cultured brown adipocytes. Lipolysis is required for the acute activation of the full thermogenic gene program by β-adrenergic agonists, and lipolysis alone is sufficient to acutely induce more than 300 genes in mouse brown adipocytes, including genes involved in lipid metabolism, thermogenesis, UPR, and circadian regulation. PPARs play a major role in mediating the lipolysis-induced activation of genes involved in lipid metabolism; interestingly, however, more than half of the genes activated by ISO and lipolysis in brown adipocytes are regulated by PPAR-independent mechanisms.

Using unbiased machine-learning approaches to interrogate key transcription factors involved in mediating the activation of the PPAR-independent gene programs, we discovered that UPR-activated transcription factors, as well as circadian transcription factors, appear to play an important role. Importantly, subsequent validation experiments showed that lipolysis leads to acute activation of all three arms of the UPR response and that ATF4 appears to act as an important lipolysis-induced activator of many genes involved in lipid metabolism, some (e.g., Dgat1 and Acsl1) of which were also scored as PPAR-dependent. This indicates that several genes involved in lipid metabolism are activated in response to lipolysis by both PPAR- and UPR-mediated signals. Furthermore, we showed that induction of lipolysis leads to major changes in circadian gene programs in brown adipocytes in vitro and that lipolysis plays an important role in the regulation of circadian gene programs in brown adipose tissue, indicating that lipolysis may constitute an important link between metabolism and circadian control in brown adipocytes. Importantly, these lipolysis-activated gene programs are largely conserved between immortalized mouse and human brown adipocytes but appear to be distinct from lipolysis-regulated gene programs in white adipocytes.

Fatty acids and metabolites thereof are known to be low-affinity agonists of PPARs^48,49. PPARα is activated by a broad range of FAs including saturated FAs, whereas PPARγ is activated more efficiently by polyunsaturated FAs, prostaglandins, and hydroxylated FAs. However, both PPARs are activated by fatty acids like oleate⁵⁰, which is one of the main fatty acids released by lipolysis. Previous reports have shown that stimulation of lipolysis has the potential to activate the ligand-binding domain of both PPARα and PPARδ when ectopically expressed in cultured mouse brown adipocytes¹³. Our findings extend these findings by showing that lipolysis activates a wide range of genes, specifically gene programs related to brown fat cell differentiation and lipid metabolism in a PPARγ/α-dependent manner. Notably, knockdown of PPARα had very little effect on the ability of lipolysis to induce gene expression, whereas knockdown of PPARγ led to a marked reduction in lipolysis-activated gene programs, indicating that PPARγ is the main mediator of the lipolytic response. However, since knockdown of PPARγ in brown adipocytes leads to considerable knockdown of PPARα as well, and since the two PPAR subtypes share many functions in brown adipocytes, it is conceivable that both PPARα and PPARγ contribute to mediating the effect in a partially redundant manner. It should be noted that PPARγ is the master regulator of adipogenesis and plays a major role in gene expression in mature adipocytes³¹. Knockdown of PPARγ, therefore, leads to a reduction in the expression of many adipocyte genes, which may lead to several indirect effects, which could result in overinterpretation of the role of PPARs in lipolysis-induced activation of gene expression.

Given the major role of PPARγ in gene regulation in adipocytes, it is remarkable that over 50% of the lipolysis-induced genes are activated by ISO independent of PPARs. Our unbiased analyzes indicate that major PPAR-independent signaling pathways include UPR-activated transcription factors, and we show that lipolysis leads to acute activation of genes downstream of all three arms of the UPR response. Classically, UPR is regarded as a stress response pathway that helps cells cope with misfolded proteins and restore proteostasis in the ER. However, recently it has emerged that UPR can be triggered by lipid perturbation, independently of misfolded proteins, in a manner that is similar to its classic mechanism⁵¹. The exact mechanism by which lipolysis triggers UPR is currently not known. One possibility is that lipolysis induces the accumulation of misfolded proteins in the ER. Another possibility is that FAs liberated from lipolysis alter the fluidity of the ER membrane resulting in lipid disequilibrium of the ER membrane thereby significantly affecting membrane properties⁵², e.g. Ca²⁺ homeostasis and vesicular budding. This would be in line with other studies showing that exogenous supplementation of saturated fatty acids alters the ER phospholipid composition and activates all three UPR branches^53,54,55.

Interestingly, we found that the PERK-ATF4 branch of UPR was rapidly activated in response to lipolysis with CHOP protein levels already induced after 1 h. Inferred ATF4 target genes are highly enriched for genes involved in the metabolism of fatty acids, either by oxidation or by esterification, suggesting a key transcriptional mechanism for alleviating lipotoxicity induced by excessive FA mobilization⁵⁶. Thus, ATF4 upregulates genes that suppress fatty acid production (Hilpda) and promote re-esterification (Acsl1, Agpat9, and Dgat1) in cultured brown adipocytes. Consistent with the importance of ATF4 in the integrated response to lipolysis, cold-induced PERK activation in brown adipocytes facilitates the import of proteins that expand mitochondrial oxidative capacity⁵⁷.

The finding that circadian regulation is another major lipolysis-induced gene program activated mainly by PPAR-independent mechanisms is intriguing. Circadian rhythms are cellular oscillations that fine-tune the biology of an organism to the 24 h light/dark cycle and aid in the anticipation of environmental changes. Classically, the circadian orchestration of lipid metabolism is under the control of the central clock that resides in the suprachiasmatic nucleus (SCN) in the brain. In addition, circadian regulation of lipid metabolism is under peripheral cell type-specific control. For example, in rodent liver, cholesterol synthesis is regulated in a circadian pattern, being high during the night and low during the day^58,59. In brown adipose tissue, FA-uptake peaks at the onset of wakening⁶⁰. Here, we provide evidence that lipolysis, profoundly influences circadian rhythmicity in brown adipocytes in vitro and in vivo. This metabolic feedback regulation of the circadian clock in brown adipocytes may help link brown adipocyte function to circadian control, analogous to how contractile signals in skeletal muscle cells influences the clock in these cells⁶¹. Notably, our data indicate that lipolysis influences brown adipocyte rhythmicity by altering the expression of core clock transcriptional regulators, such as REV-ERBα and BMAL1. Metabolic regulation by REV-ERBα likely extends beyond core clock genes. For example, REV-ERBα also coordinates the rhythmicity of lipogenesis in brown adipocytes and hepatocytes by modulation the activity of sterol regulatory element-binding protein (SREBP)^44,62. Furthermore, REV-ERBα suppresses thermogenic genes in brown adipocytes⁴⁶. However, the mechanisms responsible for coordinating the oscillation of REV-ERBα activity in brown adipocytes are unknown. Here, we show that lipolytic signals activate Reverba gene expression in vitro and are required for normal oscillation of Reverba and other core clock genes in vivo. Thus, REV-ERBα may act as a bidirectional link between the regulation of the clock and lipid metabolism in brown adipocytes.

The physiological significance of lipolytic input into the clock in brown adipocytes remains to be investigated. Brown adipocytes have evolved as a natural defense system against hypothermia, and the daily oscillation in body (and BAT) temperature is modulated by REV-ERBα⁴⁶. At the light-to-dark transition phase, when lipolysis and REV-ERBα are near circadian nadir, an increase in core body temperature, energy expenditure, and lipid uptake into BAT are observed^46,60. We hypothesize that the increase in lipolysis in brown adipocytes at the onset of the light phase^45,63,64 facilitates the dark-to-light transition by activating REV-ERBα, which then downregulates brown adipocyte activity and decreases body temperature in the light phase when the animals are asleep and require minimal facultative heat production. Our hypothesis is further supported by the observation that Atglistatin treatment transiently increases core body temperature in the light phase. Similarly, mice lacking ATGL in all cells (except cardiomyocytes) or specifically in brown adipocytes exhibited a moderately higher body temperature than control mice upon cold exposure¹⁰ in accordance with REV-ERBα downregulating BAT activity.

In conclusion, we show that lipolysis acts as an important signaling nexus sufficient to regulate a broad spectrum of gene programs and is required for the acute transcriptional effects of β-adrenergic signaling in cultured brown adipocytes. While PPARs play an important role in the activation of gene programs involved in lipid metabolism by ISO, the majority of genes are regulated by PPAR-independent mechanisms involving UPR-activated transcription factors, as well as key regulators of circadian rhythm. Future investigations should delineate the exact molecular mechanisms by which lipolysis activates UPR and regulates core clock genes in brown adipocytes and should determine the role of lipolysis in the regulation of gene expression in vivo.

Methods

Chemicals and reagents

The ABHD5 ligand SR-4995 (1-phenylpropyl-3-(10-methyl-11-oxo-10,11-dihydrodibenzo [b,f][1,4]thiazepin-8-yl)urea) was synthesized at Scripps Research Institute (Jupiter, FL) but is commercially available from Sigma-Aldrich (#SML2207). SR-3420 (1-(3-(3,5-bis(trifluoromethyl)phenyl)propyl)-3-(10-methyl-11-oxo-10,11-dihydrodibenzo[b,f][1,4] thiazepin-8-yl)urea) was synthesized at Scripps Research Institute (Jupiter, FL) and is not presently commercially available but limited amounts can be provided upon request to J.G.G. SR-3420 is a structural analog of SR-4995 and was designed to increase the lipolysis efficacy of ABHD5. The murine ATGL inhibitor Atglistatin (N′-[4′-(dimethylamino)[1,1′-biphenyl]-3-yl]-N,N-dimethyl-urea), the HSL/MGL inhibitor Cay10499 (4-(5-methoxy-2-oxo-1,3,4-oxadiazol-3(2H)-yl)-2-methylphenyl]-carbamic acid, phenylmethyl ester), the carnitine palmitoyltransferase 1 inhibitor Etomoxir ((2 R)-2-[6-(4-chlorophenoxy)hexyl]-2-oxiranecarboxylic acid) and the long fatty acid acyl-CoA synthetase inhibitor Triacsin C (2E,4E,7E-undecatrienal, nitrosohydrazone) were purchased from Cayman Chemical (#15284, #10007875, #11969 and #10007448).

Antibodies to detect total CHOP (#2895, dilution 1:1000), total HSPA5 (#3177, dilution 1:1000) and phosphorylated HSL (Ser-563) (#4139, dilution 1:500) were purchased from Cell Signaling. Antibody to detect H3K27ac (#Ab4729, dilution 2 µg) was purchased from Abcam). Secondary horseradish peroxidase antibodies were from DAKO (#P0447, dilution 1:2000).

Cell culture

Mouse brown preadipocytes immortalized with SV40 large T antigen, BAT-LgT⁶⁵, were kindly provided by Patrick Seale. Cells were propagated in basal Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma-Aldrich, #D6249) with high glucose (4500 mg/L) containing 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Lonza, #09-757 F) at 37 °C in a humidified atmosphere with 5% CO₂ with the exchange of medium every second day. To induce differentiation, cells were cultured for two days post confluence (designated day 0) and then added fresh medium supplemented with 0,5 µM dexamethasone, 0,5 mM 3-isobutyl-1-methylxanthine, 1 nM T3 and 20 nM insulin. On day 2 cells were given a fresh medium containing 1 nM T3 and 20 nM insulin, and from day 4 cells were cultured in a medium without additional inducers. Cells were considered fully differentiated by day 6–7. Immortalized human brown adipocytes (TERT-hBA) were cultured as described in^47,66. Briefly, TERT-hBA cells were cultured in advanced DMEM/F12 basal medium (Gibco, #11320033) supplemented with 10% FBS, 2mM L-glutamine (Gibco, #25030081), 1% penicillin/streptomycin and 2.5 ng/ml FGF2 (Gibco, #PHG0367). Two days post confluence (day 0), TERT-hBA cells were induced to differentiate in advanced DMEM/F12 basal medium supplemented with 2% FBS, 2mM L-glutamine, 1% penicillin/streptomycin, 5 µg/ml insulin, 1 µM dexamethasone, 0,5 mM 3-isobutyl-1-methylxanthine, 1 nM T3 and 1 µM rosiglitazone. On day 3, the medium was refreshed with exactly same medium used on day 0. At day 6 and 9 only T3 was added. On day 12, the adipocytes were considered mature. Brown adipocytes with stable knockdown of ABHD5 have previously been described^15,17. Briefly, lentiviral shRNA to Abhd5 was created from a plasmid purchased from Sigma-Aldrich (TRCN0000032737 NM_026179.1-1063s1c1) and used to stably transduce mouse-brown adipocytes. To create an inducible rescue cell line, we mutated the DNA sequence targeted by the shRNA while maintaining the wild-type amino acid sequence of ABHD5, and we tagged the construct with mCherry. Doxycycline-inducible expression was achieved by transferring the shRNA-resistant construct into pINDUCER20 and the resulting lentivirus was used to infect the brown adipocytes in which endogenous ABHD5 was stably knocked down. Propagation and differentiation of brown adipocytes with stable knockdown of ABHD5 were done similar to BAT-LgT but supplemented with 0.125 mM indomethacin from day 0-2. Re-established expression of ABHD5 was achieved by treating cells with 1 µg/ml doxycycline for 24-48 hrs.

Culture of primary mature white adipocytes in 3D culture

eWAT was isolated from lean male C57/BL6J mice (Charles River), all non-adipose tissue was removed, and the tissue was minced with a scalpel. The tissue was digested for 15 min at 37 °C and 130 rpm in degradation buffer (collagenase type 1 (Worthington, #LS004196) in Krebs–Ringer buffer (adjusted to pH = 7,4 containing 1% BSA). The cell suspension was then filtered through a double 250 µm filter to a sterile tube. The filter and suspension were washed with Krebs-Ringer buffer containing 1% BSA to stop the digestion of the tissue. A long syringe was inserted, and the tube was left for 2–5 min to allow the mature adipocytes to float to the top before removing the infranatant with the syringe. Cells were distributed in a 24-well plate (50,000 cells per well) in a gel mix consisting of 1:1 gelin-S and glycosil-S (ESI-BIO HyStem®-C Hydrogel Kit, #GS1005) and 30 μL Extralink (ESI-BIO HyStem®-C Hydrogel Kit, #GS1005). Gelation occurred after 60 min, and 700 μL DMEM (Sigma-Aldrich, #D6249) with high glucose (4500 mg/L) –with 10% FBS and 1% penicillin/streptomycin (Lonza, #09-757 F) was added to the wells. The cells were incubated at 37 °C, 5% CO₂ for 2 days prior to investigations.

Cell synchronization

Synchronization of in vitro differentiated BAT-LgT was performed by incubating adipocytes on day 7 with DMEM containing 50% FBS for 2 hrs. Following two washes in warm PBS, cells were placed in DMEM containing 0.5% FBS for the indicated periods of time.

Animals

All animal studies were performed under Approval #2018-15-0201-01459 from The Danish Animal Experient Inspectorate and complied with the ARRIVE guidelines. Male C57BL/6 N mice (12 weeks old) were housed on a 12:12-h light-dark cycle (lights on at 6 AM, lights off at 6 PM) at room temperature. All mice used in the studies were single housed with ad libitum access to a chow diet (Altromin, #30404) and water. For Atglistatin administration experiments, mice received 1.5 mg Atglistatin (dissolved in oil) in a volume of 100 µL per mouse by oral gavage at ZT0. For implantation of telemetric temperature and activity monitoring devices, mice were anesthetized with isofluorane and kept on heated pads during the procedure. G2 E-Mitter Telemetry System devices (Starr Life Sciences) were surgically implanted in the visceral cavity. ER4000 Receivers were placed under the cages within the TSE cabinets. Temperature data was integrated into the Phenomaster software.

Lipolysis assay

FAs were quantified in the medium or serum using the WAKO NEFA assay (Wako Diagnostics, #999-34691, #995-34791, #991-34891, and #993-35191), following the instructions of the manufacturer. Glycerol was quantified in serum using the free glycerol reagent (Sigma-Aldrich, #F6428).

ATP and ADP assay

Intracellular ATP and ADP were quantified using the ATP Assay Kit and ADP Assay Kit (Abcam, #ab83355 and #ab83359), following the instruction of the manufacturer. Cells were washed twice in cold PBS before harvesting.

Microscopy

Images were acquired using an Olympus IX-81 microscope equipped with a spinning-disc confocal unit. Microscope control and data acquisition were performed using CellSens Dimensions (Olympus) software. Brown adipocytes grown and differentiated on 35 mm glass coverslips were imaged in HEPES-Krebs-Ringer buffer supplemented with 1% BSA. EYFP-reported images were acquired every minute, with basal fluorescence recorded for the first 3–5 min, followed by stimulation of lipolysis. The region of interest (ROI; lipid droplets, nucleus, or cytoplasm) for each frame was quantified using ImageJ and the lipid droplet or nucleus ROI was normalized to the cytoplasm to correct for any nonspecific changes in fluorescent intensity due to changes in excitation or focal plane.

Western blotting and ECL detection

Whole-cell extracts were prepared in SDS-containing lysis buffer and subject to Western blotting as previously described⁶⁷. Briefly, cells were washed once with PBS, then scraped and collected in ice-cold lysis buffer: Tris-HCl (50 mM), glycerol (10%), SDS (2.5%), dithioerythritol (10 mM), β-glycerophosphate (10 mM), NaF (10 mM), sodium orthovanadate (0.1 mM), phenylmethylsulfonylfluoride (0.1 mM) and Complete protease inhibitor mixture (1/50 tablet per ml). After lysis, cells were instantly heated at 95 °C for 5 min and afterward treated with benzonase (Merck, #1.01654.0001), and lysates were clarified by centrifugation at 13,000 g for 5 min at 4 °C. Immunoblotting was performed using NuPage 4–12% Bis-Tris gradient gels (Invitrogen, #NP0323BOX) for 2hrs at 90 V. After gel electrophoresis, proteins were blotted onto PVDF membranes. Blots were then probed with antibodies to detect proteins of interest diluted in 5% non-fat milk. Following overnight incubation membranes were washed 3 times for 10 min each in TBS-T and immunoreactive bands were detected using peroxidase-conjugated secondary antibodies and enhanced chemiluminescence. Amido black staining of the membrane was used as loading control.

RNA extraction, cDNA synthesis and quantitative real-time PCR

RNA extraction, cDNA synthesis, and quantitative real-time PCR (qPCR) were performed as previously described⁶⁷. Briefly, total RNA was harvested using TRIzol reagent and isolated using the chloroform-isopropanol method. cDNA was transcribed using the QuantiTect Reverse Transcription Kit (Qiagen, #205311). qPCR was conducted using FastStart Essential DNA Green Master Mix (Roche, #06924204001) and the reaction was performed using LightCycler 480 Instrument II (Roche). Data were normalized to transcript levels of transcription factor II B (Tfiib). All primer sequences can be provided upon request.

RNA-seq library construction, sequencing, and processing

RNA-seq was performed according to the instructions by the manufacturer (TruSeq2, Illumina) using 500ng-1µg of total RNA for the preparation of cDNA libraries. Sequencing reads were mapped to the mouse genome mm9 or hg19 using STAR⁶⁸ (v2.4.2a), tag counts were summarized at the gene level using HOMER⁶⁹, allowing only one read per position per length. Quantification of the number of mapped reads of each gene was performed using iRNA-seq⁷⁰ (v1.1) specifying the -count gene option. Data analysis was performed in R studio (v2022.2.0) and K-means clustering using 4 clusters was used to cluster differentially expressed genes as determined by DESeq2⁷¹ (v1.24.0). Gene ontology analysis was performed using GO-seq^72,73 (v1.36.0). RNA-seq data were generated from three independent experiments, except PPAR knockdown RNA-seq which is from two independent experiments.

ChIP- and ATAC sequencing and analysis

ChIP experiments were performed in brown adipocytes at D7 essentially as previously described⁷⁴. Approximately 10–20 ng of immunoprecipitated DNA was prepared for sequencing according to the instructions from the manufacturer (Illumina). Sequence tags were aligned to the mouse reference genome mm9 using STAR⁶⁸, and ChIP peaks were called using idr ENCODE. STAR derived bam files were used as input for MACS2 at a low threshold. MACS2 peak calling was done on a pooled library of both replicates for each condition. Input DNA was used as background for peak detection. Peaks with an IDR ≤ 0.05 were kept and merged using HOMER (v4.10) mergePeaks function with distance -d given. ChIP-seq data were generated from two independent experiments. ATAC-sequencing (ATAC-seq) was performed on 50,000 nuclei as previously described³⁴. Library amplification was performed according to the instructions from the manufacturer (Illumina). Sequencing reads were mapped to the mouse genome using STAR⁷¹, and sequencing tags in peaks were quantified using HOMER⁶⁹. Only reads with a fragment length of ≤ 100 bp were kept for further analyzes corresponding to the reads located in nucleosome-free regions. For visualization purposes, the individual tag directories were converted into bedgraphs using HOMER makeUCSC specifying -fragLength 70 and -fsize 20e ATAC-seq data were generated from two independent experiments. For identification of ATAC peaks, peaks were called in each condition with HOMER findPeaks using the following settings: ‘peaks’, -fragLength 70, -style factor, minDist 140, -size 70. Tags were then counted in a 200 bp window around the peak centers for each condition. From this, mitochondrial peaks were removed, and high confident peaks were identified as having at least five tag per 200 bp window for both replicates in at least one of the conditions. Codes and scripts can be found in Supplementary Note 1 and on GitHub: https://github.com/LasseKMarkussen/Lipolysis-regulates-major-transcriptional-programs-in-brown-adipocytes.

Modeling of motif and prediction of binding sites and target genes (IMAGE)

IMAGE analysis was performed according to instructions³². Briefly, sequence information in enhancers (defined by ATAC), in combination with enhancer activity (defined by H3K27ac) and transcriptome data were used in a two-step machine-learning algorithm to calculate ‘motif activity’, i.e. the contribution of transcription factor motifs to overall enhancer activity and gene expression under given cellular conditions. IMAGE does not group enhancers but uses machine learning to regress the contribution of a single motif to experimentally determined enhancer activity on the basis of an additive model. Furthermore, IMAGE uses an additive and distance-dependent algorithm to model the contribution of distinct enhancer motifs to experimentally determined gene expression. We have shown that the accuracy of IMAGE exceeds that of other prediction strategies or tools³². The activity of a particular motif for a given sample represents the average contribution of that motif to the activity of all enhancers (step 1) or the expression of all genes (step 2). Moreover, IMAGE predicts target enhancers and target genes for all motifs on the basis of an estimated error per enhancer and gene when comparing the modeled enhancer activity with and without a particular motif (leave-one-out analysis).

siRNA transfection

Mature brown adipocytes (day 6) were reverse transfected with siRNA as previously described⁶⁷ by replating cells in 96-culture plates (80,000 cells/well). Cells were harvested 4 days after transfection for analyses. Pre-designed siRNAs were obtained from Sigma-Aldrich. siRNA Universal Negative Control #1 (Sigma-Aldrich, #SIC001) was used as a control in all experiments.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

The datasets generated in this study have been deposited at NCBI GEO under accession code GSE202833. The processed data generated in this study are in Source data and its supplementary materials. Some figures were created using BioRender.com. Source data are provided with this paper.

Code availability

Codes and scripts used to process and analyze data have been deposited to GitHub: https://github.com/LasseKMarkussen/Lipolysis-regulates-major-transcriptional-programs-in-brown-adipocytes.git

References

Zechner, R. et al. FAT SIGNALS–lipases and lipolysis in lipid metabolism and signaling. Cell Metab. 15, 279–291 (2012).
Article CAS PubMed PubMed Central Google Scholar
Cannon, B. & Nedergaard, J. Brown adipose tissue: function and physiological significance. Physiol. Rev. 84, 277–359 (2004).
Article CAS PubMed Google Scholar
Dawkins, M. J. & Hull, D. Brown adipose tissue and the response of new-born rabbits to cold. J. Physiol. 172, 216–238 (1964).
Article CAS PubMed PubMed Central Google Scholar
Granneman, J. G., Moore, H.-P. H., 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. 284, 34538–34544 (2009).
Article CAS PubMed PubMed Central Google Scholar
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. 282, 996–1002 (2007).
Article CAS PubMed Google Scholar
Reusch, J. E., Colton, L. A. & Klemm, D. J. CREB activation induces adipogenesis in 3T3-L1 cells. Mol. Cell Biol. 20, 1008–1020 (2000).
Article CAS PubMed PubMed Central Google Scholar
Cao, W., Medvedev, A. V., Daniel, K. W. & Collins, S. beta-Adrenergic activation of p38 MAP kinase in adipocytes: cAMP induction of the uncoupling protein 1 (UCP1) gene requires p38 MAP kinase. J. Biol. Chem. 276, 27077–27082 (2001).
Article CAS PubMed Google Scholar
Ahmadian, M. et al. Desnutrin/ATGL is regulated by AMPK and is required for a brown adipose phenotype. Cell Metab. 13, 739–748 (2011).
Article CAS PubMed PubMed Central Google Scholar
Haemmerle, G. et al. Defective lipolysis and altered energy metabolism in mice lacking adipose triglyceride lipase. Science 312, 734–737 (2006).
Article ADS CAS PubMed Google Scholar
Schreiber, R. et al. Cold-induced thermogenesis depends on ATGL-mediated lipolysis in cardiac muscle, but not brown adipose tissue. Cell Metab. 26, 753–763.e757 (2017).
Article CAS PubMed PubMed Central Google Scholar
Shin, H. et al. Lipolysis in brown adipocytes is not essential for cold-induced thermogenesis in mice. Cell Metab. 26, 764–777.e765 (2017).
Article CAS PubMed PubMed Central Google Scholar
Mottillo, E. P. & Granneman, J. G. Intracellular fatty acids suppress β-adrenergic induction of PKA-targeted gene expression in white adipocytes. Am. J. Physiol. Endocrinol. Metab. 301, E122–E131 (2011).
Article CAS PubMed PubMed Central Google Scholar
Mottillo, E. P., Bloch, A. E., Leff, T. & Granneman, J. G. Lipolytic products activate peroxisome proliferator-activated receptor (PPAR) alpha and delta in brown adipocytes to match fatty acid oxidation with supply. J. Biol. Chem. 287, 25038–25048 (2012).
Article CAS PubMed PubMed Central Google Scholar
Haemmerle, G. et al. ATGL-mediated fat catabolism regulates cardiac mitochondrial function via PPAR-α and PGC-1. Nat. Med. 17, 1076–1085 (2011).
Article CAS PubMed PubMed Central Google Scholar
Sanders, M. A. et al. Endogenous and synthetic ABHD5 ligands regulate ABHD5-perilipin interactions and lipolysis in fat and muscle. Cell Metab. 22, 851–860 (2015).
Article CAS PubMed PubMed Central Google Scholar
Rondini, E. A. et al. Novel pharmacological probes reveal ABHD5 as a locus of lipolysis control in white and brown adipocytes. J. Pharm. Exp. Ther. 363, 367–376 (2017).
Article CAS Google Scholar
Mottillo, E. P., Zhang, H., Yang, A., Zhou, L. & Granneman, J. G. Genetically-encoded sensors to detect fatty acid production and trafficking. Mol. Metab. 29, 55–64 (2019).
Article CAS PubMed PubMed Central Google Scholar
Lum, K. M. et al. Mapping protein targets of bioactive small molecules using lipid-based chemical proteomics. ACS Chem. Biol. 12, 2671–2681 (2017).
Article CAS PubMed PubMed Central Google Scholar
Haus-Seuffert, P. & Meisterernst, M. Mechanisms of transcriptional activation of cAMP-responsive element-binding protein CREB. Mol. Cell Biochem 212, 5–9 (2000).
Article CAS PubMed Google Scholar
Krishnan, A. V. & Feldman, D. Cyclic adenosine 3’,5’-monophosphate up-regulates 1,25-dihydroxyvitamin D3 receptor gene expression and enhances hormone action. Mol. Endocrinol. 6, 198–206 (1992).
CAS PubMed Google Scholar
Fain, J. N. & Shepherd, R. E. Free fatty acids as feedback regulators of adenylate cyclase and cyclic 3’:5’-AMP accumulation in rat fat cells. J. Biol. Chem. 250, 6586–6592 (1975).
Article CAS PubMed Google Scholar
Ho, R. J., Russell, T. R., Asakawa, T. & Sutherland, E. W. Cellular levels of feedback regulator of adenylate cyclase and the effect of epinephrine and insulin. Proc. Natl Acad. Sci. USA 72, 4739–4743 (1975).
Article ADS CAS PubMed PubMed Central Google Scholar
Gauthier, M. S. et al. AMP-activated protein kinase is activated as a consequence of lipolysis in the adipocyte: potential mechanism and physiological relevance. J. Biol. Chem. 283, 16514–16524 (2008).
Article CAS PubMed PubMed Central Google Scholar
Winther, S. et al. Restricting glycolysis impairs brown adipocyte glucose and oxygen consumption. Am. J. Physiol. Endocrinol. Metab. 314, E214–e223 (2018).
Article PubMed CAS Google Scholar
Angel, A., Desai, K. S. & Halperin, M. L. Reduction in adipocyte ATP by lipolytic agents: relation to intracellular free fatty acid accumulation. J. Lipid Res. 12, 203–213 (1971).
Article CAS PubMed Google Scholar
Rognstad, R. & Katz, J. The balance of pyridine nucleotides and ATP in adipose tissue. Proc. Natl Acad. Sci. USA 55, 1148–1156 (1966).
Article ADS CAS PubMed PubMed Central Google Scholar
Shen, Y. et al. Shared PPARα/γ target genes regulate brown adipocyte thermogenic function. Cell Rep. 30, 3079–3091.e3075 (2020).
Article CAS PubMed Google Scholar
Poulsen, L., Siersbæk, M. & Mandrup, S. PPARs: fatty acid sensors controlling metabolism. Semin. Cell Developmental Biol. 23, 631–639 (2012).
Article CAS Google Scholar
Lasar, D. et al. Peroxisome proliferator activated receptor gamma controls mature brown adipocyte inducibility through glycerol kinase. Cell Rep. 22, 760–773 (2018).
Article CAS PubMed Google Scholar
Nielsen, R. et al. Genome-wide profiling of PPARgamma: RXR and RNA polymerase II occupancy reveals temporal activation of distinct metabolic pathways and changes in RXR dimer composition during adipogenesis. Genes Dev. 22, 2953–2967 (2008).
Article CAS PubMed PubMed Central Google Scholar
Lefterova, M. I., Haakonsson, A. K., Lazar, M. A. & Mandrup, S. PPARγ and the global map of adipogenesis and beyond. Trends Endocrinol. Metab. 25, 293–302 (2014).
Article CAS PubMed PubMed Central Google Scholar
Madsen, J. G. S. et al. Integrated analysis of motif activity and gene expression changes of transcription factors. Genome Res. 28, 243–255 (2018).
Article CAS PubMed PubMed Central Google Scholar
Rauch, A. et al. Osteogenesis depends on commissioning of a network of stem cell transcription factors that act as repressors of adipogenesis. Nat. Genet 51, 716–727 (2019).
Article CAS PubMed Google Scholar
Buenrostro, J. D., Wu, B., Chang, H. Y. & Greenleaf, W. J. ATAC-seq: a method for assaying chromatin accessibility genome-wide. Curr. Protoc. Mol. Biol. 109, 21.29.21–21.29.29 (2015).
Article Google Scholar
Creyghton, M. P. et al. Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proc. Natl Acad. Sci. USA 107, 21931–21936 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Rada-Iglesias, A. et al. A unique chromatin signature uncovers early developmental enhancers in humans. Nature 470, 279–283 (2011).
Article ADS CAS PubMed Google Scholar
Thonberg, H., Lindgren, E. M., Nedergaard, J. & Cannon, B. As the proliferation promoter noradrenaline induces expression of ICER (induced cAMP early repressor) in proliferative brown adipocytes, ICER may not be a universal tumor suppressor. Biochem J. 354, 169–177 (2001).
Article CAS PubMed PubMed Central Google Scholar
Ma, X. et al. CREBL2, interacting with CREB, induces adipogenesis in 3T3-L1 adipocytes. Biochem J. 439, 27–38 (2011).
Article CAS PubMed Google Scholar
Yang, A., Mottillo, E. P., Mladenovic-Lucas, L., Zhou, L. & Granneman, J. G. Dynamic interactions of ABHD5 with PNPLA3 regulate triacylglycerol metabolism in brown adipocytes. Nat. Metab. 1, 560–569 (2019).
Article CAS PubMed PubMed Central Google Scholar
Hetz, C., Zhang, K. & Kaufman, R. J. Mechanisms, regulation and functions of the unfolded protein response. Nat. Rev. Mol. Cell Biol. 21, 421–438 (2020).
Article CAS PubMed PubMed Central Google Scholar
Solt, L. A., Kojetin, D. J. & Burris, T. P. The REV-ERBs and RORs: molecular links between circadian rhythms and lipid homeostasis. Future Med Chem. 3, 623–638 (2011).
Article CAS PubMed Google Scholar
Honma, S. et al. Dec1 and Dec2 are regulators of the mammalian molecular clock. Nature 419, 841–844 (2002).
Article ADS CAS PubMed Google Scholar
Riedel, C. S., Georg, B., Jørgensen, H. L., Hannibal, J. & Fahrenkrug, J. Mice lacking egr1 have impaired clock gene (BMAL1) oscillation, locomotor activity, and body temperature. J. Mol. Neurosci. 64, 9–19 (2018).
Article CAS PubMed Google Scholar
Adlanmerini, M. et al. Circadian lipid synthesis in brown fat maintains murine body temperature during chronic cold. Proc. Natl Acad. Sci. USA 116, 18691–18699 (2019).
Article CAS PubMed PubMed Central Google Scholar
Dyar, K. A. et al. Atlas of circadian metabolism reveals system-wide coordination and communication between clocks. Cell 174, 1571–1585.e1511 (2018).
Article CAS PubMed PubMed Central Google Scholar
Gerhart-Hines, Z. et al. The nuclear receptor Rev-erbα controls circadian thermogenic plasticity. Nature 503, 410–413 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Markussen, L. K. et al. Characterization of immortalized human brown and white pre-adipocyte cell models from a single donor. PLoS One 12, e0185624 (2017).
Article PubMed PubMed Central CAS Google Scholar
Schupp, M. & Lazar, M. A. Endogenous ligands for nuclear receptors: digging deeper. J. Biol. Chem. 285, 40409–40415 (2010).
Article CAS PubMed PubMed Central Google Scholar
Poulsen, Ll. C., Siersbæk, M. & Mandrup, S. PPARs: fatty acid sensors controlling metabolism. Semin Cell Dev. Biol. 23, 631–639 (2012).
Article PubMed CAS Google Scholar
Xu, H. E. et al. Molecular recognition of fatty acids by peroxisome proliferator-activated receptors. Mol. Cell 3, 397–403 (1999).
Article CAS PubMed Google Scholar
Ho, N., Xu, C. & Thibault, G. From the unfolded protein response to metabolic diseases - lipids under the spotlight. J. Cell Sci. 131, jcs199307 (2018).
Article PubMed CAS Google Scholar
Fun, X. H. & Thibault, G. Lipid bilayer stress and proteotoxic stress-induced unfolded protein response deploy divergent transcriptional and non-transcriptional programmes. Biochim Biophys. Acta Mol. Cell Biol. Lipids 1865, 158449 (2020).
Article CAS PubMed Google Scholar
Volmer, R., van der Ploeg, K. & Ron, D. Membrane lipid saturation activates endoplasmic reticulum unfolded protein response transducers through their transmembrane domains. Proc. Natl Acad. Sci. USA 110, 4628–4633 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Cunha, D. A. et al. Initiation and execution of lipotoxic ER stress in pancreatic beta-cells. J. Cell Sci. 121, 2308–2318 (2008).
Article CAS PubMed Google Scholar
Robblee, M. M. et al. Saturated fatty acids engage an ire1α-dependent pathway to activate the nlrp3 inflammasome in myeloid cells. Cell Rep. 14, 2611–2623 (2016).
Article CAS PubMed PubMed Central Google Scholar
Chitraju, C. et al. Triglyceride synthesis by DGAT1 protects adipocytes from lipid-Induced ER stress during lipolysis. Cell Metab. 26, 407–418.e403 (2017).
Article CAS PubMed PubMed Central Google Scholar
Latorre-Muro, P. et al. A cold-stress-inducible PERK/OGT axis controls TOM70-assisted mitochondrial protein import and cristae formation. Cell Metab. 33, 598–614.e597 (2021).
Article CAS PubMed PubMed Central Google Scholar
Mayer, D. The circadian rhythm of synthesis and catabolism of cholesterol. Arch. Toxicol. 36, 267–276 (1976).
Article ADS CAS PubMed Google Scholar
Ho, K. J. Circadian rhythm of cholesterol biosynthesis: dietary regulation in the liver and small intestine of hamsters. Int J. Chronobiol. 6, 39–50 (1979).
CAS PubMed Google Scholar
van den Berg, R. et al. A diurnal rhythm in brown adipose tissue causes rapid clearance and combustion of plasma lipids at wakening. Cell Rep. 22, 3521–3533 (2018).
Article PubMed CAS Google Scholar
Gabriel, B. M. & Zierath, J. R. Circadian rhythms and exercise - re-setting the clock in metabolic disease. Nat. Rev. Endocrinol. 15, 197–206 (2019).
Article PubMed Google Scholar
Le Martelot, G. et al. REV-ERBalpha participates in circadian SREBP signaling and bile acid homeostasis. PLoS Biol. 7, e1000181–e1000181 (2009).
Article PubMed PubMed Central CAS Google Scholar
Shostak, A., Meyer-Kovac, J. & Oster, H. Circadian regulation of lipid mobilization in white adipose tissues. Diabetes 62, 2195–2203 (2013).
Article CAS PubMed PubMed Central Google Scholar
Stubblefield, J. J. et al. Temporal Control of Metabolic Amplitude by Nocturnin. Cell Rep. 22, 1225–1235 (2018).
Article CAS PubMed PubMed Central Google Scholar
Harms, M. J. et al. Prdm16 is required for the maintenance of brown adipocyte identity and function in adult mice. Cell Metab. 19, 593–604 (2014).
Article CAS PubMed PubMed Central Google Scholar
Breining, P. et al. Metformin targets brown adipose tissue in vivo and reduces oxygen consumption in vitro. Diabetes, Obes. Metab. 20, 2264–2273 (2018).
Article CAS Google Scholar
Markussen, L. K., Winther, S., Wicksteed, B. & Hansen, J. B. GSK3 is a negative regulator of the thermogenic program in brown adipocytes. Sci. Rep. 8, 3469–3469 (2018).
Article ADS PubMed PubMed Central CAS Google Scholar
Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013).
Article CAS PubMed Google Scholar
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).
Article CAS PubMed PubMed Central Google Scholar
Madsen, J. G. et al. iRNA-seq: computational method for genome-wide assessment of acute transcriptional regulation from total RNA-seq data. Nucleic Acids Res. 43, e40 (2015).
Article PubMed PubMed Central CAS Google Scholar
Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550–550 (2014).
PubMed PubMed Central Google Scholar
Ashburner, M. et al. Gene ontology: tool for the unification of biology. The gene ontology consortium. Nat. Genet. 25, 25–29 (2000).
Article CAS PubMed PubMed Central Google Scholar
The Gene Ontology Consortium. The Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res. 47, D330–D338 (2019).
Article CAS Google Scholar
Larsen, B. D., Madsen, M. R., Nielsen, R. & Mandrup, S. Chromatin immunoprecipitation for identification of protein-dna interactions in human cells. Methods Mol. Biol. 1794, 335–352 (2018).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We thank the members of the Mandrup lab, Granneman lab, and Gerhart-Hines lab for fruitful discussions. Work in the Mandrup laboratory was supported by grants from The Independent Research Fund Denmark (project grant: 7016-00279), The Novo Nordisk Foundation to Center for Adipocyte Signaling (ADIPOSIGN) (project grant: NNF18OC0033444), and The Danish National Research Foundation to the Center for Functional Genomics and Tissue Plasticity (ATLAS) (Project grant: 141). Work in the Granneman lab was supported by National Institutes of Health grants R01DK076629, R01DK062292, and R01DK105963. The Mandrup/Granneman collaboration was also supported by the Danish Diabetes Academy supported by the Novo Nordisk Foundation. Work in the Gerhart-Hines laboratory was supported by funding from the European Research Council under the European Union’s Horizon 2020 Research and Innovation Programme (grant agreement no. 639382).

Author information

These authors contributed equally: James G. Granneman, Susanne Mandrup.

Authors and Affiliations

Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense, Denmark
Lasse K. Markussen, Jesper G. S. Madsen, Ann-Britt Marcher & Susanne Mandrup
Center for Adipocyte Signaling (AdipoSign), Odense, Denmark
Lasse K. Markussen, Olivia Sveidahl Johansen, Ann-Britt Marcher, Zachary Gerhart-Hines & Susanne Mandrup
Center for Functional Genomics and Tissue Plasticity (ATLAS), Odense, Denmark
Lasse K. Markussen, Jesper G. S. Madsen, Ann-Britt Marcher & Susanne Mandrup
Center for Molecular Medicine and Genetics, Wayne State University, Detroit, MI, USA
Elizabeth A. Rondini & James G. Granneman
Novo Nordisk Foundation Center for Basic Metabolic Research, University of Copenhagen, Copenhagen, Denmark
Olivia Sveidahl Johansen, Elahu G. Sustarsic & Zachary Gerhart-Hines
Embark Biotech ApS, Copenhagen, Denmark
Olivia Sveidahl Johansen & Zachary Gerhart-Hines
Department of Biology, University of Copenhagen, Copenhagen, Denmark
Jacob B. Hansen

Authors

Lasse K. Markussen
View author publications
You can also search for this author in PubMed Google Scholar
Elizabeth A. Rondini
View author publications
You can also search for this author in PubMed Google Scholar
Olivia Sveidahl Johansen
View author publications
You can also search for this author in PubMed Google Scholar
Jesper G. S. Madsen
View author publications
You can also search for this author in PubMed Google Scholar
Elahu G. Sustarsic
View author publications
You can also search for this author in PubMed Google Scholar
Ann-Britt Marcher
View author publications
You can also search for this author in PubMed Google Scholar
Jacob B. Hansen
View author publications
You can also search for this author in PubMed Google Scholar
Zachary Gerhart-Hines
View author publications
You can also search for this author in PubMed Google Scholar
James G. Granneman
View author publications
You can also search for this author in PubMed Google Scholar
Susanne Mandrup
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

L.K.M., E.A.R., J.B.H, Z.G.H, J.G.G. and S.M. conceived and designed the study. L.K.M., E.A.R., A.B.M, E.G.S. and O.S.J. performed the experiments. L.K.M. and J.G.S.M. performed computational analyses. L.K.M. and S.M. wrote the manuscript with input from the other authors.

Corresponding authors

Correspondence to James G. Granneman or Susanne Mandrup.

Ethics declarations

Competing interests

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.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of Additional Supplementary Files

Supplementary Data 1

Supplementary Data 2

Supplementary Data 3

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as 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. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Markussen, L.K., Rondini, E.A., Johansen, O.S. et al. Lipolysis regulates major transcriptional programs in brown adipocytes. Nat Commun 13, 3956 (2022). https://doi.org/10.1038/s41467-022-31525-8

Download citation

Received: 22 April 2021
Accepted: 17 June 2022
Published: 08 July 2022
DOI: https://doi.org/10.1038/s41467-022-31525-8

This article is cited by

Direct effects of adipocyte lipolysis on AMPK through intracellular long-chain acyl-CoA signaling
- Abir A. Rahman
- Andrew J. Butcko
- Emilio P. Mottillo
Scientific Reports (2024)
p21-activated kinase 4 counteracts PKA-dependent lipolysis by phosphorylating FABP4 and HSL
- Hwang Chan Yu
- Yong Geun Jeon
- Byung-Hyun Park
Nature Metabolism (2024)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.