JNK1 mediates serine phosphorylation of STAT3 in response to fatty acids released by lipolysis

TL;DR

The paper shows that JNK1 is a key sensor linking fatty acids to energy expenditure in fat cells during lipolysis.

Contribution

Identifies JNK1 as the kinase mediating STAT3 phosphorylation and regulating oxidative metabolism in lipolytic adipocytes.

Findings

Fatty acids from lipolysis activate JNK1, which phosphorylates STAT3 and suppresses fatty acid re-esterification.

JNK1 promotes mitochondrial uncoupling and energy expenditure without affecting lipolysis rates.

Genetic and pharmacological inhibition of JNK1 reduces lipolysis-driven respiration in adipocytes.

Abstract

Adipocytes play a central role in energy balance and metabolic health by storing excess nutrients as triglycerides in white adipose tissue. During physiological stress, sympathetic activation triggers lipolysis, releasing fatty acids and glycerol to meet systemic energy demands. Lipolytic activation in white adipocytes also increases their rate of oxygen consumption. Phosphorylation of signal transducer and activator of transcription 3 (STAT3) at Ser727 is a key regulatory event in lipolysis-driven respiration. Here, we identify c-Jun N-terminal kinase 1 (JNK1) as the kinase responsible for this essential phosphorylation event and a key regulator of oxidative metabolism in lipolytic adipocytes. We show that fatty acids produced by lipolysis activate JNK, which phosphorylates lipid droplet-associated STAT3, leading to inhibition of glycerol-3-phosphate acyltransferase 3 and suppression of fatty acid re-esterification. This shift in lipid handling promotes mitochondrial uncoupling and increases energy expenditure. Pharmacological inhibition of JNK1 markedly reduced lipolysis-driven respiration without altering the rate of lipolysis. The critical role of JNK1 in promoting respiration in lipolytic adipocytes was verified using genetic knockdown studies. Notably, canonical upstream mitogen-activated protein kinase kinases were not required for JNK1 activation, suggesting a noncanonical pathway that senses acute increases in intracellular fatty acid levels. Together, these findings identify JNK1 as a metabolic sensor linking intracellular fatty acid levels to STAT3-mediated oxidative metabolism in adipocytes, with potential implications for energy balance and metabolic disease.

Figures

Increased respiration in lipolytic adipocytes is dependent on STAT3 suppression of esterification. A: Change in oxygen consumption rate (OCR) from baseline after stimulation with vehicle or CL at 100 nM in differentiated primary adipocytes isolated from adipocyte-specific knockout (SAKO) mice and floxed littermate control (SAWT) mice. B: Relative TMRM staining intensity normalized to baseline, 5 μM triacsin C (ACSL inhibitor), and 10 μM PF-04620110 (DGAT inhibitor) primary adipocytes after 40 min of stimulation with 100 nM CL or vehicle. C: Change in OCR from baseline after stimulation with ve

JNK is activated by lipolysis in adipocytes. A: Western blot analysis of inguinal white adipose tissue (iWAT) and epididymal white adipose tissue (eWAT) from mice stimulated with a 1 mg/kg dose of CL or vehicle via intraperitoneal (IP) injection for 20 min. Vehicle and CL blots are from the same blot and exposure, but bands in the middle have been removed. B: Western blot analysis of eWAT from mice stimulated with a 1 mg/kg dose of CL via IP injection for 30 and 90 min or vehicle injection for 60 min. C: Western blot analysis was performed on treated PPDIVs following stimulation with 1 μM and

JNK activity is required for STAT3 Ser^727^ phosphorylation and GPAT inhibition. A: Western blot analysis was performed on fractionated PPDIVs following stimulation with 1 μM CL or vehicle for 15 min. B: Treatment of PPDIVs from SAWT or SAKO mice with and without 2% BSA and CL at 30 and 60 min or vehicle. C: Western blot analysis was performed on fractionated lipid droplet (LD) PPDIVs following stimulation with 1 μM CL or vehicle for 15 min. D: Rate of FFA release from 3T3-L1 and primary adipocytes pretreated with 20 μM JNK inhibitor or vehicle, then stimulated with 100 nM CL or vehicle with (

JNK1 activity is required for lipolysis-driven oxidative metabolism. A: Oxygen consumption rate (OCR) in primary adipocytes pretreated with 5 μM and 50 μM SP600125 (JNK inhibitor [JNKi]) at the 15-min time point before baseline measurements. Port A: 0.5 μM CL or vehicle at 20 min. B: Change in OCR in primary adipocytes from SAKO and littermate control SAWT mice following stimulation with 5 μM JNKi or vehicle. ∗p < 0.05 WT control versus KO control, # p < 0.05 WT control versus WT JNKi, ∼p < 0.05 WT JNKi versus KO JNKi. KO control versus KO JNKi not significant. C: OCR in primary adipocytes pre

ERK is not required for lipolysis-driven respiration. A: Change in oxygen consumption rate (OCR) in primary adipocytes following a treatment with ravoxertinib (ERK inhibitor) at the indicated doses, measured from baseline to 20 min after stimulation with 100 nM CL. B: OCR in primary adipocytes treated with 100 nM ERK inhibitor, 20 μM JNK inhibitor, or vehicle, then stimulated with 100 nM CL or vehicle in port A at 20 min. C: Change in OCR in primary adipocytes following treatment with 5 μM and 10 μM MEK inhibitor and 20 μM JNK inhibitor, measured from baseline to 20 min after stimulation with

MAPK signaling upstream of JNK activation in lipolytic adipocytes. A: Western blot analysis was performed on treated PPDIVs following stimulation with 100 nM CL or vehicle, 5 μM ASK1 inhibitor or vehicle, and 50 μM ATGLi or vehicle. B: Oxygen consumption rate (OCR) normalized to baseline in primary adipocytes pretreated with 10 μM ASK1 inhibitor or vehicle, then stimulated with 100 nM CL or vehicle in port A. Treatment in port B: 2 μM oligomycin or vehicle; treatment in port C: 1 μM FCCP; treatment in port D: 1 μM rotenone/antimycin A (R/A). C: OCR in MLK3 knockdown (KD) and control primary ad

Full text

Obesity is a medical condition characterized by the excessive accumulation of body fat and is associated with serious health complications, including cardiovascular disease, type 2 diabetes, chronic inflammation, and several cancers (1, 2, 3). In individuals with obesity, particularly those with an excess of visceral fat, immune cells in adipose tissue are activated and secrete proinflammatory cytokines, contributing to a state of chronic, low-grade inflammation. This persistent inflammatory state plays a key role in the development of obesity-related metabolic diseases. In contrast, acute inflammation increases energy expenditure and may represent an early adaptive response to obesity (4). Acute inflammatory activation induces a metabolic shift marked by heightened energy expenditure driven by immune system activation, fever generation, and altered nutrient handling (5).

Excess nutrients are stored as triglycerides within white adipose tissue (WAT). Fatty acids are esterified onto a glycerol backbone through the glycerolipid synthesis pathway; the first and rate-limiting enzyme in this pathway in WAT is glycerol-3-phosphate acyltransferase 3 (GPAT3). During physiological stress, such as fasting or cold exposure, sympathetic stimulation mobilizes adipocyte lipid stores. Catecholamines released within adipose tissue activate lipolysis, liberating triglyceride stores as fatty acids and glycerol. Notably, not all fatty acids released during lipolysis enter the circulation; at least one-third are retained within WAT. Under basal conditions, white adipocytes strongly favor fatty acid esterification and storage. However, during lipolysis, the activity of the esterification pathway is suppressed, resulting in increased oxidative metabolism (6, 7, 8, 9). One mechanism underlying this shift involves serine phosphorylation of STAT3. Elevated intracellular fatty acids during lipolysis stimulate phosphorylation of lipid droplet-associated STAT3 at Ser^727^, which enhances interaction with and inhibition of GPAT3. This phosphorylation event is essential for the induction of lipolysis-driven oxidative metabolism (6).

The protein kinases responsible for STAT3 Ser^727^ phosphorylation have been studied extensively (10, 11, 12, 13, 14, 15, 16, 17, 18). Proline residues surrounding the Ser^727^ site match the mitogen-activated protein kinase (MAPK) consensus sequence [P-X-S/T-P] (19, 20, 21). Several MAPKs, including p38, extracellular signal-regulated kinase (ERK), and c-Jun N-terminal kinase (JNK), have been implicated in STAT3 serine phosphorylation (10, 11, 17, 22, 23, 24). Here, we investigate the kinase responsible for STAT3 Ser^727^ phosphorylation during lipolysis. Although multiple MAPKs are activated during lipolysis, our results specifically implicate JNK1 in STAT3 serine phosphorylation in this context. JNK1 has previously been shown to directly phosphorylate STAT3 in vitro (16). Catecholamine-stimulated JNK activation is mediated by fatty acids, as is STAT3 Ser^727^ phosphorylation (25, 26, 27, 28). Surprisingly, the signaling pathway linking fatty acids to JNK activation does not appear to require the MAPK kinases 4/7 (MKK4/7) or upstream MAPK kinase kinases (MAP3Ks). This report specifically investigates whether JNK1 phosphorylates STAT3 at Ser^727^ in response to fatty acids released during lipolysis.

Materials and Methods

Reagents

The following reagents were used in this study: Amphotericin B (Sigma, A2411), atglistatin (adipose triglyceride lipase inhibitor, Sigma, SML1075), BSA (Sigma, A028), BSJ-04-122 (MKK4/7 inhibitor, MedChemExpress, HY-152185), carbonyl cyanide 4 (trifluoromethoxy)phenylhydrazone (FCCP, Sigma C2920), CL-316,243 (CL, Sigma C5979), collagenase (Sigma, C6885), DB07268 (JNK1 inhibitor, MedChemExpress, HY-15737), dexamethasone (Sigma, D4902), DMEM/F-12 50/50 (Corning, 15-090), extracellular matrix gel from Engelbreth-Holm-Swarm murine sarcoma (Sigma, E1270), FBS (Corning, 35-010), fibronectin (Sigma, F1141), GS-444217 (apoptosis signal-regulating kinase [ASK1] inhibitor, MedChemExpress, HY-100844), insulin (Sigma, I6634), 3-isobutyl-1-methylxanthine (Sigma, I5879), Lenti-X™ 293T (Takara, 632180), Lenti-X Packaging Single Shots (Takara, 631282) MitoTracker Green (Fisher, M7514), oligomycin A (Sigma, 75351), penicillin/streptomycin-glutamine (Gibco, 10378-016), PF-04620110 (diacylglycerol acyltransferase [DGAT] inhibitor, Sigma, PZ0207), pLKO.1 puro (Addgene, Plasmid #8453), ravoxertinib (MedChemExpress, GDC-0994), rosiglitazone (Sigma, 557366), rotenone (Sigma, R8875), SP600125 (JNK inhibitor, MedChemExpress, HY-12041), TCSJNK5a (JNK IX inhibitor, Selleckchem, S7508), tetramethylrhodamine, methyl ester (TMRM, Fisher, I34361), triacsin C (Sigma, T4540), and U0126 (MedChemExpress, HY-12031A).

Animals

Animals homozygous for the Stat3 floxed allele (stock no.: 016923) were bred to Adipoq-promoter-driven Cre mice (stock no.: 028020) to generate mice homozygous for the Stat3 floxed allele both with and without the Adipoq-Cre. Animals with Adipoq-Cre expression lose Stat3 in mature adipocytes and are referred to in the article as SAKO (adipocyte-specific knockout) animals, whereas floxed littermate controls without Adipoq-Cre are referred to as SAWT (floxed littermate control of SAKO). All strains of mice were on the C57BL/6J background (stock no.: 000664). Animals for experiments were bred in-house. Animals in each cohort were produced from multiple breeding pairs to minimize the birth date range. Extra attention was paid to housing arrangements to ensure that each cage accommodated multiple treatment groups to minimize potential confounding by the cage effect. During animal studies, ear tag numbers were used to identify animals. Within an experiment, the genotype and/or treatment groups were both littermates and cagemates. Researchers performing tests and collecting data were blinded during experiments. Sample sizes were determined using a power analysis with the expected effect size but were sometimes limited by availability. Sex as a biological variable was considered; female and male cohorts were analyzed separately. Mice were housed in a specific pathogen-free facility with a 12-h light-dark cycle and were given free access to food and water. Mice were fed a normal diet (5053, Labdiet Picolab). All animal use was approved by the IACUC at Weill Cornell Medicine.

Cell culture

Primary adipocytes

Primary preadipocytes were isolated from inguinal fat pads as follows: Following fine mincing, the tissue was digested with 1 mg/ml collagenase and 2% BSA in a 37°C water bath with shaking for 20–35 min. The digestion was stopped by adding FBS, then the slurry was passed through a 100 μm filter and spun at 500 g for 5 min. The pellet was washed and resuspended in culture medium (DMEM/F-12 with 15% FBS and penicillin/streptomycin-glutamine), plated with 2.5 mg/L amphotericin B, and placed in a 10% CO_2_ incubator. Nonadherent cells were washed away 3 days later. When the cells reached ∼80% confluence, they were passaged from their original 10 cm culture plate to a 15 cm plate and incubated for an additional 2–5 days. Cells from the second passage were plated to confluence for experiments on extracellular matrix and fibronectin-coated plates. Differentiation was initiated with 500 μM 3-isobutyl-1-methylxanthine, 5 μM dexamethasone, 1 μg/ml insulin, and 1 μM rosiglitazone for 4 days, followed by insulin alone for at least 3 days. Cells were used for experiment 8–12 days after the initiation of differentiation. Twenty-four hours prior to the start of an assay, insulin was removed from the culture media. Only cultures in which >90% of cells displayed adipocyte morphology were used.

Lentiviral knockdowns

Gene-specific guide RNA sequences were cloned into the pLKO.1 plasmid, JNK1 (Mapk8): 5′-CGGGACTTAAAGCCTAGTAAT-3′ and MLK3 (Map3k11): 5′-GCTGTAAACAAGTTAACGTTA-3′. Lentivirus was produced in Lenti-X 293T (Takara) cells, which were seeded in 10 cm culture plates to reach 70–80% confluence by the time of transfection the next morning. Media were removed and replaced with 5 ml high-glucose DMEM with 10% Tet system-approved FBS and no antibiotics. pLKO.1 plasmid DNA (7 μg) with shRNA sequence or nonhairpin control was mixed with Lenti-X Packaging Single Shots and added dropwise to Lenti-X 293T plates. After at least 4 h, another 6 ml of media were added to each plate. About 48–72 h post transfection, all media were removed from 293T plates and centrifuged (500 g for 10 min). The supernatant was filtered through a 0.45 μm polyether sulfone syringe filter. Before use, 10 μg/ml hexadimethrine bromide was added to the viral supernatant media. Preadipocyte isolations used for JNK1 knockdown experiments were passaged once into 10 cm plates (2:5 split) and grown for 1 day before lentiviral transduction with JNK1 shRNA or control medium overnight. After 16–18 h incubation, the plates were washed and replaced with normal culture medium. Starting the next day, transduced cells were selected with 2 μg/ml puromycin for at least 3 days. Once grown back to confluence, the preadipocytes were plated for differentiation as normal.

Fractionation

Adipocytes were fractionated using differential centrifugation in fractionation buffer: 20 mM Hepes (pH 7.4), 10 mM KCl, 2 mM MgCl_2_, 1 mM EDTA, 1 mM EGTA, 200 μM DTT, 10 mM NaF, 1 mM NaVO_4_, 2.5 mM β-glycerophosphate, and a phosphatase inhibitor tablet (Roche). All steps were performed at 4°C. Cells and tissues were lysed with a glass Dounce homogenizer, then spun at 1,000 g for 10 min to separate lipid droplets (floating fat cake) and nuclear pellets. Mitochondria were then pelleted from the supernatant at 10,000 g, and then the membrane was pelleted at 21,000 g after incubation with 8 mM CaCl_2_. The remaining proteins in the solution were cytosol. Each fraction (except cytosol) was resuspended in a fractionation buffer and spun a second time to wash before the addition of a lysis buffer. Each supernatant was spun a second time at the same speed before being transferred to a new tube to pellet the next fraction. Protein content in each fraction was normalized, and 20 μg of protein was loaded per well for Western blot analysis.

GPAT activity assay

GPAT-specific activity was assayed for 15 min at room temperature in a 200 μl reaction mixture containing 100 μg total lysate protein in GPAT assay buffer (75 mM Tris-HCl, pH 7.5, 4 mM MgCl_2_, 1 mg/ml BSA [FA-free], 1 mM DTT, 8 mM NaF, 77 μM [2 μCi per reaction] glycerol 3-phosphate, and 50 μM lauryl-CoA). Phosphatase and protease inhibitor cocktail (IV) cell lysates were prepared by 10 passages through a 28G needle in GPAT assay lysis buffer (250 mM sucrose, 10 mM Tris [pH 7.5], 1 mM EDTA, and 8 mM NaF). Inguinal WAT tissue lysates were prepared by mechanical homogenization (Roto Star) with 1 μl GPAT assay lysis buffer per mg tissue. Crude lysates were spun at 1,000 g for 10 min, followed by 17,000 g for 15 min. Supernatant was transferred to a new tube, and protein concentration was measured before dilution to 1 mg/ml and addition of 2x GPAT assay buffer. N-ethyl maleimide treatments were performed for 15 min prior to the addition of 2x GPAT assay buffer. Reactions were terminated by the addition of 2:1 chloroform:methanol to extract lipids and precipitate protein. Lipid extraction was repeated an additional two times, and lipids were dried and separated by TLC with 6:35:8 chloroform:methanol:water to isolate phosphatidic acid. The spot corresponding to phosphatidic acid was identified based on a standard 16:0 PA standard (Avanti Polar Lipids, 830855) and visualized with I_2_ to be scraped off the plate, and radioactivity was quantified in a liquid scintillation counter (PerkinElmer, MicroBeta TriLux 1450).

Seahorse respiration assays

Extracellular oxygen consumption rates were measured with a Seahorse XFe96 analyzer. Primary preadipocytes were differentiated in 96-well Seahorse Xfe96 culture plates. After differentiation, cells were switched to Seahorse XF base DMEM (Agilent, 102353) supplemented with 2 mM glutamine, 1 mM pyruvate, and 8 mM glucose. Pretreatments were added to the base medium. During the assays, drug treatments were injected sequentially using the ports. Unless otherwise indicated, port A contained 100 nM CL or vehicle control, port B: 2 mM oligomycin, port C: 1 mM FCCP, and port D: 1 mM rotenone and 1 mM antimycin A.

Mitochondrial membrane potential

TMRM perchlorate staining was performed and imaged with an ImageXpress MICRO Confocal Automated High-Content Analysis System to visualize live mitochondrial membrane potential. Cells were stained with 200 nM TMRM for 30 minutes, then washed three times with PBS. Imaging was performed in a live cell imaging solution (Invitrogen, A59688DJ). After baseline images were obtained, cells were treated with either vehicle control or CL. Finally, cells were treated with 1 mM FCCP as a negative control for membrane potential.

Lipolysis assay

FA concentration was measured using 10 μl conditioned media with the NEFA kit (WAKO), using 75 μl reagent A and 150 μl reagent B. Absorbance was measured at 550 nm (reference 660 nm) using the manufacturer's protocol. Free glycerol concentration was measured by reacting 25 μl of sample with 175 μl of Free Glycerol Reagent (Sigma), and absorbance was measured at 540 nm according to the manufacturer's protocol.

Western blot analysis

We homogenized tissues and cells in lysis buffer (50 mM Tris [pH 7.5], 150 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 1 mM DTT, 1 mM Na_3_VO_4_, 5 mM NaF, 1 mM phenylmethanesulfonyl fluoride, 25 mM glycerol 2-phosphate, and a freshly added protease inhibitor tablet) and then incubated them for 1 h at 4°C. We centrifuged crude lysates at 17,000 g for 15 min twice and determined the protein concentration using Bio-Rad Protein Assay Dye Reagent. Samples were diluted in the SDS sample buffer, resolved by SDS-PAGE, and transferred to nitrocellulose membranes (Bio-Rad). Individual proteins were detected with specific antibodies and visualized on film using horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) and Western Lightning Enhanced Chemiluminescence (PerkinElmer Life Sciences). Primary antibodies against the following were used at a 1:1,000 dilution unless otherwise specified and were purchased from Cell Signaling: JNK (9252, lot no.: 18, Research Resource Identifier [RRID]: AB_2250373), JNK1 (3708, lot no.: 4, RRID: AB_1904132), pJNK (T183/Y185; 9251, lot no.: 17, RRID: AB_2307321), HSL (4107, lot no.: 1, RRID: AB_2798800), pHSL (S563, 4139, RRID: AB_2135495), STAT3 1:2,000 (9139, Lot no.: 16, RRID: AB_331757), pSTAT3 1:500 (S727; 9134, lot no.: 21, RRID: AB_331589), c-JUN (9165, lot no.: 13, RRID: AB_2130165), pcJUN (S73; 9164, lot no.: 5, RRID: AB_2129575), P38 (8690, lot no.: 9, RRID: AB_10999090), pP38 (T180/Y182; 4511, lot no.: 13, RRID: AB_2139682), MKK4 (9152, lot no.: 3, RRID: AB_330905), and pMKK4 (S257/T261, 9156, lot no.: 2, RRID: AB_2297420). Goat anti-mouse IgG (31430, lot no.: YF375332, RRID: AB_228307) and Goat anti-rabbit IgG (31460, lot no.: 33, RRID: AB_228341) secondary antibodies were purchased from ThermoFisher and were used at a concentration of 1:10,000. Antibodies were validated by the manufacturer and confirmed in the laboratory by Western blot using knockout, knockdown, or expected signaling response to stimulation. All blots are from separate membranes, unless otherwise noted, but were run, transferred, and blotted in parallel using the same power source and antibody dilutions.

Statistical analysis

Two-way or one-way ANOVA was performed to evaluate statistical significance, followed by the Holm-Sidak post hoc analysis to determine specific between-group and time-dependent differences. In each case, significance was set at α = 0.05, and p values were adjusted for multiple comparisons. Statistical analyses were performed in Prism 10 (GraphPad).

Results

Cellular assays were performed using adipocyte progenitor cells differentiated in vitro, which display robust lipolytic activity and strongly induce oxidative metabolism upon stimulation. As expected, activation of lipolysis with the β3-adrenergic receptor agonist, CL, caused a rapid increase in oxygen consumption rate, which was markedly blunted in STAT3-knockout adipocytes (Fig. 1A). STAT3 promotes oxidative metabolism by suppressing fatty acid re-esterification via GPAT3 (9). Although re-esterification increases ATP demand, it does not drive the respiration observed during lipolysis (6). Instead, intracellular fatty acids stimulate respiration by promoting ADP/ATP carrier-mediated uncoupling (9). The first step of re-esterification is fatty acid activation by long-chain acyl-CoA synthetase, whereas DGAT catalyzes the final step of triglyceride synthesis. Inhibition of either long-chain acyl-CoA synthetase or DGAT increases mitochondrial uncoupling and reduces membrane potential (Fig. 1B). Consistent with STAT3-mediated suppression of re-esterification promoting uncoupled respiration, STAT3-knockout adipocytes exhibited attenuated mitochondrial depolarization, which was restored upon inhibition of re-esterification (Fig. 1B). Moreover, STAT3 deficiency did not impact the increase in oxidative metabolism caused by DGAT inhibition, further supporting that STAT3 acts upstream of esterification (Fig. 1C). Sequestration of free fatty acids by 0.2% albumin suppressed respiration in wild-type adipocytes to levels observed in STAT3-knockout adipocytes, whereas 2% albumin completely blocked lipolysis-driven respiration in both genotypes (Fig. 1D). Together, these findings show that STAT3-dependent suppression of re-esterification is required for fatty acid-driven respiration during lipolysis.Fig. 1. Increased respiration in lipolytic adipocytes is dependent on STAT3 suppression of esterification. A: Change in oxygen consumption rate (OCR) from baseline after stimulation with vehicle or CL at 100 nM in differentiated primary adipocytes isolated from adipocyte-specific knockout (SAKO) mice and floxed littermate control (SAWT) mice. B: Relative TMRM staining intensity normalized to baseline, 5 μM triacsin C (ACSL inhibitor), and 10 μM PF-04620110 (DGAT inhibitor) primary adipocytes after 40 min of stimulation with 100 nM CL or vehicle. C: Change in OCR from baseline after stimulation with vehicle or 10 μM PF-04620110 (DGAT inhibitor) in differentiated primary adipocytes isolated from SAKO mice and floxed littermate control (SAWT) mice. D: OCR in SAKO and SAWT adipocytes in the presence of no, 0.2, or 2% BSA in the media. Treatments: 1 nM CL, 2 μM oligomycin (oligo), 1 μM FCCP, and 1 μM rotenone/antimycin A (R/A). ∗p < 0.05 SAWT versus SAKO, ^#^p < 0.05 V versus CL, ^∼^p < 0.05 versus control. n = 3. ACSL, long-chain acyl-CoA synthetase.

In response to elevated intracellular fatty acids during lipolysis, we have observed increased Ser^727^ phosphorylation of STAT3. This phosphorylation site matches the MAPK consensus sequence, which, in combination with sensitivity to fatty acids, leads us to investigate the role of JNK. First, we investigated whether lipolysis activates JNK. In vivo, activation of lipolysis with CL increased JNK phosphorylation in adipose tissue (Fig. 2A, B). In cultured adipocytes, JNK activation was induced by lipolysis, peaking at 1 h—distinct from the rapid and transient phosphorylation of HSL (Fig. 2C-E). These results align with earlier observations in 3T3-L1 adipocytes and other mouse strains in which JNK was activated during lipolysis (28).Fig. 2JNK is activated by lipolysis in adipocytes. A: Western blot analysis of inguinal white adipose tissue (iWAT) and epididymal white adipose tissue (eWAT) from mice stimulated with a 1 mg/kg dose of CL or vehicle via intraperitoneal (IP) injection for 20 min. Vehicle and CL blots are from the same blot and exposure, but bands in the middle have been removed. B: Western blot analysis of eWAT from mice stimulated with a 1 mg/kg dose of CL via IP injection for 30 and 90 min or vehicle injection for 60 min. C: Western blot analysis was performed on treated PPDIVs following stimulation with 1 μM and 10 μM of CL or vehicle. D and E: Quantification of Western blots in 2C. ^#^p < 0.05 V versus 1 μM and V versus 10 μM CL. PPDIV, phosphatase and protease inhibitor cocktail (IV).

Although Ser^727^ phosphorylation of lipid droplet-associated STAT3 increases with lipolysis, no change in STAT3 localization to the lipid droplet is observed, suggesting that the kinase that phosphorylates STAT3 does so in the vicinity of the lipid droplet. Subcellular fractionation revealed JNK in all compartments, with the highest levels in the cytosol and lower levels in mitochondria (Fig. 3A). Importantly, all fractions, including the lipid droplets, showed robust increases in JNK phosphorylation after CL treatment (Fig. 3A). JNK activation occurred normally in STAT3-knockout cells but was suppressed by extracellular sequestration of fatty acids with BSA (Fig. 3B), suggesting that JNK activation lies upstream of STAT3. Pharmacological inhibition of JNK reduced STAT3 Ser^727^ phosphorylation in the lipid droplet fraction (Fig. 3C). Inhibition of JNK did not impact the rate of either fatty acid or glycerol release from adipocytes (Fig. 3D, E).Fig. 3JNK activity is required for STAT3 Ser^727^ phosphorylation and GPAT inhibition. A: Western blot analysis was performed on fractionated PPDIVs following stimulation with 1 μM CL or vehicle for 15 min. B: Treatment of PPDIVs from SAWT or SAKO mice with and without 2% BSA and CL at 30 and 60 min or vehicle. C: Western blot analysis was performed on fractionated lipid droplet (LD) PPDIVs following stimulation with 1 μM CL or vehicle for 15 min. D: Rate of FFA release from 3T3-L1 and primary adipocytes pretreated with 20 μM JNK inhibitor or vehicle, then stimulated with 100 nM CL or vehicle with (5% BSA) or without BSA media. ∗p < 0.05 V versus CL. E: Rate of glycerol release. ∗p < 0.05 V versus CL. F: GPAT activity assay on in vitro differentiated adipocytes with 1 μM CL or vehicle and 20 μM JNK inhibitor II or vehicle. ∗p < 0.05 V versus CL. PPDIV, phosphatase and protease inhibitor cocktail (IV); SAKO, adipocyte-specific knockout; SAWT, floxed littermate control of SAKO.

Because STAT3 Ser^727^ phosphorylation is required for STAT3-mediated inhibition of GPAT3, we next tested whether JNK inhibition blocked GPAT3 suppression. Indeed, JNK inhibition with SP-600125 prevented the decrease in GPAT activity during lipolysis (Fig. 3F). Furthermore, JNK inhibition dose-dependently suppressed CL-stimulated respiration (Fig. 4A). This effect was absent in STAT3-knockout adipocytes (Fig. 4B), indicating that STAT3 is the functional JNK substrate in this context.Fig. 4JNK1 activity is required for lipolysis-driven oxidative metabolism. A: Oxygen consumption rate (OCR) in primary adipocytes pretreated with 5 μM and 50 μM SP600125 (JNK inhibitor [JNKi]) at the 15-min time point before baseline measurements. Port A: 0.5 μM CL or vehicle at 20 min. B: Change in OCR in primary adipocytes from SAKO and littermate control SAWT mice following stimulation with 5 μM JNKi or vehicle. ∗p < 0.05 WT control versus KO control, # p < 0.05 WT control versus WT JNKi, ∼p < 0.05 WT JNKi versus KO JNKi. KO control versus KO JNKi not significant. C: OCR in primary adipocytes pretreated with 5 μM, 10 μM, 20 μM, and 50 μM JNKi IX (JNK 2-3i) at the 15-min time point before baseline measurements. Port A: 100 nM CL or vehicle control. ∗p < 0.05 V versus CL, # p < 0.05 versus control. D: Change in OCR in primary adipocytes following a 15-min pretreatment with DB07268 (JNK1i) at the indicated doses, measured from baseline to 20 min after stimulation with CL. ∗p < 0.05 V versus CL, ^#^p < 0.05 versus control. n = 6–8. E: OCR in 0.5% BSA media, with pretreatment using 20 μM JNK1i or vehicle, followed by stimulation with 10 μM CL or vehicle in port A at 20 min. F: Change in OCR from baseline to 30 min after stimulation with 100 nM CL or vehicle control in primary adipocytes in 0.2% BSA or vehicle media. ∗p < 0.05 V versus CL, ^#^p < 0.05 V versus JNKi, ^∼^p < 0.05 V versus BSA. G: Western blot analysis with and without BSA, after stimulation of 1 μM CL or vehicle and 20 μM JNKi or vehicle. H: Western blot analysis in PLKO control or JNK1 knockdown (KD) primary adipocytes, after stimulation with 10 nM CL or vehicle. I: OCR in JNK1 KD and control female primary adipocytes stimulated with 10 nM CL or vehicle. J: Change in OCR in JNK1 KD and control female primary adipocytes pretreated with 5 μM JNKi or vehicle, measured from baseline to 20 min after stimulation with 10 nM CL. ∗p < 0.05 V versus CL, # p < 0.05 KD CL versus control CL. ∼p < 0.05 inhibitor CL versus control CL. SAKO, adipocyte-specific knockout; SAWT, floxed littermate control of SAKO.

Adipocytes express both JNK1 and JNK2. JNK inhibitor IX, which selectively inhibits JNK2 and JNK3 (29), had no effect on lipolysis-induced respiration (Fig. 4C). In contrast, the JNK1-selective inhibitor DB-07268 dose-dependently suppressed respiration (Fig. 4D–F). Consistent with fatty acid-dependent activation, BSA suppressed JNK activation during lipolysis (Fig. 4G). Residual activation in the presence of JNK1 inhibition suggests that JNK2 is also activated but not required. Lentiviral knockdown of JNK1 significantly attenuated lipolysis-induced respiration (Fig. 4H, I). And the effect of the JNK1 inhibitor was lost in knockdown cells (Fig. 4J). Together, these data indicate that fatty acids activate JNK1, which then phosphorylates STAT3 at Ser^727^ to suppress esterification and promote oxidative metabolism.

Several MAPKs can phosphorylate STAT3 at Ser^727^; however, only JNK1 was required in this setting. Although p38 was activated during lipolysis, it was unaffected by extracellular fatty acid sequestration (Fig. 4G), and STAT3 Ser^727^ phosphorylation during lipolysis is insensitive to p38 inhibition (6). ERK inhibition also had no effect on lipolysis-driven respiration (Fig. 5A, B). The highest dose tested exhibited slight attenuation; however, at this concentration, the inhibitor exhibits off-target effects (30). Inhibition of mitogen-activated protein kinase kinase, the kinase upstream of ERK, also had no impact on respiration in lipolytic adipocytes (Fig. 5C). Thus, JNK1 is uniquely required for STAT3 Ser^727^ phosphorylation during adipocyte lipolysis.Fig. 5ERK is not required for lipolysis-driven respiration. A: Change in oxygen consumption rate (OCR) in primary adipocytes following a treatment with ravoxertinib (ERK inhibitor) at the indicated doses, measured from baseline to 20 min after stimulation with 100 nM CL. B: OCR in primary adipocytes treated with 100 nM ERK inhibitor, 20 μM JNK inhibitor, or vehicle, then stimulated with 100 nM CL or vehicle in port A at 20 min. C: Change in OCR in primary adipocytes following treatment with 5 μM and 10 μM MEK inhibitor and 20 μM JNK inhibitor, measured from baseline to 20 min after stimulation with 100 nM CL. ∗p < 0.05 V versus CL, #p < 0.05 inhibitor versus control. MEK, mitogen-activated protein kinase kinase.

The mechanism by which fatty acids activate JNK remains unclear. Inhibition of ATGL prevented increases in intracellular fatty acids and blocked JNK activation (Fig. 6A), supporting regulation by fatty acids rather than upstream adrenergic inputs. We next examined the MAPK cascade upstream of JNK. The MAP3K, ASK1, has been implicated in JNK activation in response to ER stress, which can result from elevated intracellular fatty acid levels (31, 32, 33). However, we did not observe an impact of ASK1 inhibition on lipolysis-induced oxidative metabolism in adipocytes (Fig. 6B). This result is consistent with previous reports that ASK1 does not mediate fatty acid stimulation of JNK (34). Although the MAP3K, MLK3, has been implicated in fatty acid stimulation of JNK in human embryonic kidney 293 cells (35), we did not observe any impact of MLK3 knockdown (Fig. 6C). MKK4 and MKK7 are the canonical kinases that phosphorylate JNK (36). Phosphorylation of MKK4 in response to lipolytic activation was observed after only 15 min and blocked by the MKK4/7 inhibitor BSJ-04-122. While MKK4/7 inhibition suppressed baseline JNK phosphorylation, it did not block JNK activation by CL (Fig. 6D, E). Furthermore, the MKK4/7 inhibitor had no impact on lipolysis-driven oxidative metabolism (Fig. 6F). These results suggest that JNK activation during lipolysis may involve noncanonical upstream signaling.Fig. 6MAPK signaling upstream of JNK activation in lipolytic adipocytes. A: Western blot analysis was performed on treated PPDIVs following stimulation with 100 nM CL or vehicle, 5 μM ASK1 inhibitor or vehicle, and 50 μM ATGLi or vehicle. B: Oxygen consumption rate (OCR) normalized to baseline in primary adipocytes pretreated with 10 μM ASK1 inhibitor or vehicle, then stimulated with 100 nM CL or vehicle in port A. Treatment in port B: 2 μM oligomycin or vehicle; treatment in port C: 1 μM FCCP; treatment in port D: 1 μM rotenone/antimycin A (R/A). C: OCR in MLK3 knockdown (KD) and control primary adipocytes stimulated with 100 nM CL or vehicle. D: Western blot analysis of treated PPDIVs from SAKO and littermate control SAWT mice, following stimulation with 1 μM CL or vehicle and 5 μM and 10 μM MKK4/7i or vehicle. E: Western blot analysis was performed on PPDIV-differentiated adipocytes stimulated with 1 μM CL or vehicle, in addition to 5 μM MKKi or vehicle. F: OCR in differentiated primary adipocytes, measured from baseline to 30 min after stimulation with either 5 μM MKK4/7i or vehicle in port A. Treatment in port B: 10 nM CL or vehicle; treatment in port C: 1 μM FCCP; treatment in port D: 1 μM rotenone/antimycin A (R/A). PPDIV, phosphatase and protease inhibitor cocktail (IV); SAKO, adipocyte-specific knockout; SAWT, floxed littermate control of SAKO.

Discussion

Inflammation plays a multifaceted role in metabolic health and energy balance. Here, we identify a critical role for JNK in promoting energy expenditure in white adipocytes during lipolysis. In this acute context, JNK activation appears metabolically beneficial. This contrasts with chronic JNK activation in obesity, where elevated fatty acids induce JNK-mediated IRS1 phosphorylation and contribute to insulin resistance (25, 37). The acute impact of JNK activation during lipolysis on insulin signaling was not investigated in the current studies, all of which were performed in the absence of insulin. Because insulin and catecholamine signaling are mutually antagonistic, acute JNK activation during lipolysis may enhance catecholamine-driven pathways by relieving the suppressive effect of insulin signaling.

Using selective inhibitors and genetic knockdown, we demonstrate that JNK1, rather than JNK2, is required for activation of the respiratory program initiated during lipolysis. This aligns with prior work highlighting a prominent metabolic role for JNK1 in adipose tissue (38, 39). While JNK2 may contribute to total JNK activity, its role in this pathway appears to be dispensable. The broader role of JNK1 in obesity is complex. Whole-body JNK1 knockout protects against diet-induced obesity, whereas adipocyte-specific knockout does not, even though these mice display increased UCP1-dependent thermogenesis (40). The divergent outcomes likely reflect differences in tissue-specific versus systemic signaling as well as qualitative differences between chronic and acute JNK activation.

The mechanism by which fatty acids activate JNK in adipocytes remains unresolved. Canonical upstream MAPKKs appear dispensable, suggesting a noncanonical mode of JNK activation. In addition, the spatial specificity of STAT3 phosphorylation at the lipid droplet raises further mechanistic questions. Lipolysis dramatically alters the lipid-droplet proteome, and it is possible that remodeling of this compartment enhances accessibility of STAT3 to JNK or promotes assembly of a local signaling complex.

Finally, these studies were performed using adipocytes derived from both male and female C57BL/6 mice; future work will be needed to determine whether these mechanisms are conserved across strains and species.

Data Availability

All data are provided within the article.

Supplemental Data

This article contains supplemental data.

Conflict of Interest

The authors declare that they have no conflicts of interest with the contents of this article.