Mouse models and animal experiments
All animal protocols were approved by the Washington University Institutional Animal Care and Use Committee. C57BL/6J WT mice were obtained from the Jackson Laboratory (strain no. 000664). FASLox/Lox mice were previously described in ref. 34. To generate mice with brown-adipose-specific knockout of FAS (FAS-BKO), FASLox/Lox animals were crossed with Ucp1-Cre mice obtained from the Jackson Laboratory (stock no. 024670). FASLox/Lox mice without Cre were used as a control for FAS-BKO mice. ACOX2−/− mice have been previously described in ref. 36 and were generated using two TALEN pairs targeting beginning and end of exon 4 that led to deletion of 176 base pairs covering the entire exon 4, resulting in a frameshift. Founder mice were generated by cytoplasmic injection of TALENs into C57BL/6N mouse zygotes and then backcrossed into the WT C57BL/6J background. ACOX2Lox/Lox mice on the C57BL/6J background were generated using the CRISPR–Cas9 system. CRISPR-mediated mutagenesis was done by the Genome Engineering and IPSC Center at Washington University. To generate mice with adipose-specific knockout of ACOX2 (ACOX2-AKO), ACOX2Lox/Lox animals were crossed with adiponectin-Cre mice obtained from the Jackson Laboratory (stock no. 028020). ACOX2Lox/Lox mice without Cre were used as a control for ACOX2-AKO. To generate mice with adipose-specific overexpression of ACOX2 (ACOX2Adipo-OE), murine ACOX2 was cloned downstream of a previously characterized 5.4-kilobase adiponectin promoter51. The adiponectin promoter plasmid was a generous gift from P. Scherer (University of Texas Southwestern Medical Center). The transgenic construct was microinjected into the pronucleus of a newly fertilized egg from a C57BL/6J × CBA hybrid mouse and implanted into a pseudopregnant female by the Mouse Genetics Core at Washington University. The founders were backcrossed more than seven generations into the WT C57BL/6J genetic background.
UCP1−/− mice were obtained from the Jackson Laboratory (stock no. 003124). To generate mice with adipose-specific overexpression of ACOX2 on the UCP1−/− background, ACOX2Adipo-OE mice were crossed with UCP1−/− mice. An inbred strain of these mice on the C57BL/6J genetic background was developed and used for all experiments. Mice were fed either normal chow diet (Purina 5053) or a HFD (D12492, Research Diets). All animals were randomly allocated into different groups. For metabolic phenotyping studies, both male and female animals were studied, and all data were disaggregated by sex. Mice were maintained under constant temperature (23–25 °C), circulating air and humidity (45–65%) with a 12-h light/dark cycle and provided ad libitum access to food and water. Body composition (fat and lean mass) was measured using an EchoMRI system.
To assess glucose homeostasis, intraperitoneal glucose tolerance tests and intraperitoneal insulin tolerance tests were performed based on body weight or lean body mass, as indicated. Mice were injected intraperitoneally with glucose (2.5 g kg−1 body weight or 2 g kg−1 lean mass) or insulin (0.75 U kg−1 body weight or 1 U kg−1 lean mass) after being fasted for 6 h or 4 h, respectively. Blood glucose levels were measured at 0 min, 15 min, 30 min, 60 min and 120 min after injection by a glucometer.
VO2, VCO2 and respiratory exchange ratio were measured by indirect calorimetry using a PhenoMaster (TSE Systems) metabolic cage system and analysed using CalR web-based software (v.1.3). For measurement of CL316,243-induced energy expenditure, mice were anaesthetized using pentobarbital (90 mg kg−1 intraperitoneally) and acclimated to the environment for 60 min, allowing oxygen consumption to stabilize. The mice were then injected with CL316,243 at a dose of 1 mg kg−1, and data were collected for 2 h in mice housed at 23 °C or 30 °C as previously described52 To assess cold tolerance, body temperature was measured at time 0 and hourly for 6 h during cold exposure using implantable IPTT-300 temperature-sensitive transponders and a DAS-8007 programmable reader from Bio Medic Data Systems as previously described52.
Cell culture and treatments
Immortalized mouse brown preadipocytes were established and differentiated as previously described13. Immortalized human brown preadipocytes, kindly provided by Y.-H. Tseng (Joslin Diabetes Center), were cultured and differentiated as previously described53,54. For isolation of mature pig adipocytes from subcutaneous WAT, fresh iWAT was digested in a collagenase buffer containing HBSS (GIBCO 14065-45), 12.6 mM CaCl2, 4.9 mM MgCl2, 2% BSA and 800 U g−1 (3 mg g−1 tissue) of type 2 collagenase for roughly 40 min in a 37 °C water bath. The digested material was passed through a 250-μm strainer and washed 3 times with a total of 1 l of KRHB (1× KBH, 25 mM HEPES, 2 mM glucose, 2% BSA) to separate the floating mature adipocytes. The cells were centrifuged at 50g for 3 min. The mature adipocytes (floating cells) were collected and cultured with DMEM/F12 medium.
Mouse brown adipocytes and pig iWAT mature adipocytes were cultured in Dulbecco’s modified eagle medium F12 (DMEM/F12) supplemented with 10% (v/v) fetal bovine serum (FBS), 1% (v/v) penicillin/streptomycin, 1% (v/v) l-glutamine and 1% (v/v) sodium pyruvate. Human embryonic kidney 293T (HEK293T) cells and immortalized human brown adipocytes were cultured in DMEM. All cells were cultured in a humidified incubator at 37 °C with 5% CO2 in air.
Palmitic acid, C15:0 OCFA, C17:0 OCFA and BCFA iso-C17:0 were dissolved in ethanol, then diluted with serum-free medium containing 0.1% fatty acid-free BSA, then added into culture medium at final concentrations of 0.2 μM, 1 μM or 2 μM based on experimental need. Norepinephrine was first dissolved in 1× PBS and diluted into the culture medium at final concentrations of 0.25 μM, 1 μM or 2 μM for different experiments. H2O2 was dissolved in 1× PBS and diluted into the culture medium at a final concentration of 500 μM. Isotope-labelled glucose (U-13C6 glucose) was dissolved in 1× PBS and diluted into the culture medium at a final concentration of 17.5 mM.
Plasmid constructs
Genome-wide guide RNA (gRNA) databases were used to design gRNA oligonucleotides against ACOX2, CRAT, Bckdha, CAT and CKB. The oligonucleotides were ordered from Integrated DNA Technologies and subcloned into lentiCRISPRv2 plasmid. ACOX2 and Hyper7 complementary DNA (cDNA) clones were purchased from transOMIC (BC021339) and Addgene (136466), respectively, and subcloned into pLJM1 lentiviral overexpression plasmid. Sequences of primers used for cloning are listed in Supplementary Table 1. Pexo-TEMP plasmid was generated by inserting PTS2-Sirius-T2A fragment by PCR using primers below. Forward: CGG CGA CCG GTG CCA CCA TGC ACC GGC TGC AGG TGG TGC TGG GCC ACC TGG CCG GCC GGC CCG AGT CCT CCT CCG CCC TGC AGG CCG CCC CCT GCA GCT CGG ATC CCA CCA TG, and reverse: TCA CCA TGA GCT CGG GGC C and inserting mT-Sapphire fragment by PCR using primers below. Forward: CGG CCG AGC TCA TGC ACC GGC TGC AGG TGG TGC TGG GCC ACC TGG CCG GCC GGC CCG AGT CCT CCT CCG CCC TGC AGG CCG CCC CCT GCG TGA GCA AGG GCG AGG AGC T, reverse: TGT GAT GGA TAT CTG CAG AAT TC based on the gTEMP_pcDNA3 plasmid (Addgene no. 89583). To generate recombinant lentiviruses, the overexpression plasmid (pLJM1), gRNA plasmid (lentiCRISPRv2) or short-hairpin RNA plasmid (pLKO.1-puro) together with packaging plasmids (pMD2.G and psPAX2) were cotransfected into HEK293T cells. Lentiviral particles were collected 48 h after transfection and stored at −80 °C until they were used to transduce immortalized BAT SVF cells.
Extraction and mass spectrometric analysis of fatty acids
BAT or brown adipocytes homogenate containing 50 μg of total protein content was hydrolysed in an acid hydrolysis buffer (CH3CN:37% HCl, 4:1) in 90 °C water bath for 2 h. Then, hexane was used to extract the fatty acids. The samples were dried under a stream of nitrogen and redissolved in chloroform, methanol, H2O and 25% NH4OH (50:45:5:0.01). After extraction, the electrospray ionization-mass spectrometry images of the fatty acids in the adipose tissues and adipocytes were obtained by a Thermo Fisher LTQ Orbitrap Velos in the negative-ion mode scanning from 200 m/z to 600 m/z with a resolution of 100,000 (at m/z 400 Da). Data were processed by built-in Xcalibur software as previously described14 and the exogenous docosanoic-22, 22, 22-D3 acid added to samples before extraction was used as an internal standard for quantitation. To verify the iso-form of the fatty acid structures, dried fatty acids were derivatized to the N-(4-aminomethylphenyl) pyridinium derivatives, which were subjected to higher-energy collisional dissociation tandem mass spectrometry for structural identification as described previously55. Data were analysed using R (v.4.2.1).
Carbon flux tracing and lipidomic analysis
To assess incorporation of 13C-label into mmBCFA, BAT SVF cells were differentiated into adipocytes in normal DMEM/F12 medium and then cultured in DMEM/F12 in which glucose was replaced with [U13C6]-glucose, as previously described in ref. 9. Briefly, after differentiation, sgACOX2 or control adipocytes were cultured in the presence of [U13C6]-glucose in DMEM/F12 medium lacking glutamine and pyruvate containing 10% FBS for an extra 3 days, followed by 6 h of treatment with 10 nM CL316,243 or vehicle in a normal DMEM/F12 media and the lipids with labelled mmBCFA were extracted and subjected to high resolution electrospray ionization-mass spectrometry analysis as described above.
Immunofluorescence analysis
Frozen sections or cell samples were fixed with ethanol or 4% paraformaldehyde, followed by primary antibody and the corresponding secondary antibody incubation. Nuclei were counterstained with 4′6-diamidino-2-phenylindole. Samples were subjected to immunofluorescence analysis using rabbit polyclonal anti-ACOX2 antibody (1:100), rabbit polyclonal anti-PMP70 antibody (1:100), rabbit polyclonal anti-FASN antibody (1:100) and rabbit polyclonal anti-CRAT antibody (1:100). Slides were imaged using a Nikon A1Rsi Confocal Microscope. Images were analysed using NIS (v.5.21). Fluorescence intensity and colocalization were calculated using ImageJ (v.1.53) and Colocalization Finder (v.1).
Oil Red O staining
Adipocytes were fixed with 10% formalin overnight and then washed twice with 60% isopropanol. Oil Red O working solution was added, and the cells were incubated for 10 min at room temperature. The Oil Red O solution was removed, and the wells were washed four times with diH2O. The stained cells were photographed under a ×2 objective lens using light microscopy.
Quantitative real-time PCR analysis
Total RNA was isolated using PureLink RNA Mini Kit (Invitrogen, 12183018A) and 2 μg of total RNA was reverse transcribed into cDNA using the iScriptcDNA Synthesis Kit (Bio-Rad) as previously reported23. Quantitative real-time PCR was conducted using PowerUp SYBR Green Master Mix. Relative mRNA expression level was determined using the 2(−ΔΔCT) method and L32 was used as an internal reference. Primers used in PCR analyses are listed in Supplementary Table 1.
Western blot and immunoprecipitation analyses
Cells or mouse tissue samples were homogenized in RIPA buffer (Cell Signaling Technology, 9806S) or homogenization buffer (0.25 M sucrose, 20 mM HEPES in distilled H2O) containing a protease and phosphatase inhibitor cocktail (Sigma, P8465 and P0001). Total protein was extracted and quantitated using a bicinchoninic acid (BCA) kit. Denatured protein was separated by SDS–PAGE electrophoresis and transferred onto polyvinyl difluroride or Nylon membrane, blocked with 5% non-fat milk, incubated overnight with primary antibodies and finally incubated with horse radish peroxidase-conjugated secondary antibodies as reported. Tissue or cell lysates were subjected to immunoblot analysis using rabbit polyclonal anti-ACOX2 antibody (1:1,000), rabbit polyclonal anti-FASN antibody (1:10,000), mouse monoclonal anti-catalase antibody (1:1,000), rabbit polyclonal anti-UCP1 antibody (1:1,000), rabbit polyclonal anti-CRAT antibody (1:500), rabbit polyclonal anti-Ap2 antibody (1:1,000), rabbit polyclonal anti-AKT (1:1,000), rabbit polyclonal anti-PMP70 antibody (1:1,000), mouse monoclonal anti-COX 4 antibody (1:1,000), mouse polyclonal anti-oxidative phosphorylation cocktail (1:1,000), rabbit polyclonal anti-pAMPK (T172) antibody (1:1,000), rabbit polyclonal anti-AMPK antibody (1:1,000), rabbit polyclonal anti-pACC (S79) antibody (1:1,000), rabbit monoclonal anti-ACC antibody (1:1,000), rabbit polyclonal anti-BCKDHA antibody (1:1,000), rabbit polyclonal anti-pAK(T473) antibody (1:1,000), rabbit polyclonal anti-AKT antibody (1:1,000) and mouse monoclonal anti-CKB antibody (1:10,000). Rabbit polyclonal anti-β actin (1:1,000), rabbit polyclonal anti-β Tubulin (1:1,000) and rabbit polyclonal anti-Vinculin (1,000) were used as loading controls. For western blot analysis of ACOX2, initial experiments used either a rabbit polyclonal antibody (Invitrogen; catalogue no. PA5-114814) or a goat polyclonal antibody (Novus; catalogue no. 06011). Subsequently, a rabbit polyclonal anti-ACOX2 antibody from Millipore Sigma (catalogue no. HPA038280) was used. All antibodies used in western blot analyses are listed in the Reporting summary. Proteins were detected with the Odyssey Infrared Imaging System (LI-COR Biosciences). Uncropped raw blots are presented in Supplementary Fig. 1.
Interaction between FASN and PEX7 was assessed by co-immunoprecipitation analysis as described previously56. Briefly, HEK293 cells transiently transfected with Flag-FASN were homogenized using RIPA Lysis Buffer System (Santa Cruz) supplemented with protease and phosphatase inhibitors. Cell lysates were centrifuged at 13,000 rpm in a microcentrifuge for 10 min to remove unlysed cells. Supernatants were collected and subjected to protein quantification using the BCA assay. Cell lysates were then incubated with Anti-Flag M2 Affinity gel (Sigma-Aldrich, A220) overnight. Immunoprecipitates were washed three times with Tris-buffered saline before elution with SDS–PAGE sample buffer and subjected to SDS–PAGE. Cell lysates were subjected to immunoblot analysis using mouse monoclonal anti-HA antibody (1:1,000) and rabbit polyclonal anti-Flag antibody (1:1,000).
Subcellular fractionation of adipocytes
Adipocytes were lysed in Peroxisome Extraction Buffer (Sigma) using a Dounce homogenizer. Lysed cells were centrifuged at 1,000g for 10 min to pellet the nuclear fraction. The supernatant was transferred to a new tube and centrifuged at 12,000g for 15 min to collect the mitochondrial pellet. The supernatant was transferred to a new tube and centrifuged at 21,000g for 40 min to collect the peroxisome pellet. The remaining supernatant was saved as the cytosol fraction.
OCR
OCR in cultured adipocytes was measured using a Seahorse XFe Extracellular Flux Analyzer (Agilent) in a 24-well plate. For mitochondrial stress test, cells were treated with oligomycin (3 μM), FCCP (1.5 μM) and antimycin (2 μM) plus rotenone (1 μM). For measurement of fatty acid-induced respiration, cells were treated with BSA-conjugated 1 mM C15:0, C17:0, iso-C17:0 or BSA alone.
Measurement of mitochondrial respiration in BAT
Mitochondrial respiration in mouse BAT was measured using a previously described method57. Briefly, BAT depots from ACOX2Adipo-OE and WT mice were thawed in ice-cold PBS, minced in an Eppendorf tube and mechanically homogenized with 10–20 strokes using Teflon-glass in MAS buffer (70 mM sucrose, 220 mM mannitol, 10 mM KH2PO4, 5 mM MgCl2, 2 mM HEPES, 1.0 mM EGTA and 0.2% (w/v) fatty acid-free BSA). The homogenates were centrifuged at 1,000g for 10 min at 4 °C. The supernatant was collected and loaded into a Seahorse XF96 microplate in 20 µl of MAS (5 µg BAT). The loaded plate was centrifuged at 2,000g for 5 min at 4 °C. An extra 130 µl of MAS buffer was added to each well. Substrate injections were as follows: 5 mM succinate + 2 μM rotenone for port A, 2 μM rotenone + 4 μM antimycin for port B, 0.5 mM TMPD + 1 mM ascorbic acid for port C and 50 mM azide for port D.
Infrared thermal imaging
The surface temperature in the region surrounding the BAT was measured using a high-speed mid-wave infrared camera (Telops FAST M3k) equipped with a 50 mm lens (Telops) and a spacer ring that facilitates a long working distance, resulting in a resolution of 0.25 mm per pixel. The temperature distribution was captured from a top-down perspective at 30 frames per second and a resolution of 320 × 256 pixels. The recorded thermal images were processed using the Reveal-IR software suite v.1.13.0 (Telops), assuming a fur emissivity of 0.75 (refs. 58,59).
MTT assay
To determine whether β-oxidation of mmBCFA results in oxidative stress leading to cytotoxicity, cell viability was assessed using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide) assay (Cayman Chemicals). Control and catalase knockout brown adipocytes were treated with iso-C17:0, C16:0, H2O2 or vehicle for 2 h. Cell viability was assessed according to the manufacturer’s instructions.
Lipid peroxidation assay
To determine the effect of ACOX2 overexpression on oxidative stress, BAT samples were collected from ACOX2AdipoOE and WT mice. Lipid peroxidation (4-HNE) protein adducts were quantified in the samples using a 4-HNE assay kit (Abcam) according to the manufacturer’s instructions.
Fecal lipid extraction and measurement
Feces were collected for 24 h from each cage and dried in an incubator for 16 h at 42 °C. One hundred milligrams of feces were homogenized with 1 M NaCl, followed with a chloroform:methanol solution (2:1). Chloroform layers were collected after centrifuge and evaporated under N2 flow until dry. The 2% Triton X-100 in chloroform was added and evaporated again under N2 flow. The samples were dissolved with ddH2O and measured by a commercial kit (FFA, Wako; HR Series NEFA-HA(2)).
Ex vivo lipolysis assay
BAT was collected from WT and ACOX2Adipo OE mice and placed into a well of a 24-well plate containing 1 ml of prewarmed KRBH-BSA with 2% fatty acid-free BSA. The tissues were treated with vehicle or with 10 μM isoproterenol for 2 h at 37 °C. The glycerol levels in the media were measured using a commercial kit (Glycerol Assay Kit; Sigma, MAK117).
ATP assay
The ATP levels in differentiated adipocytes were measured using a commercial kit (ATP Detection Assay Kit; Cayman, 700410). Briefly, adipocytes were homogenized in the ATP detection sample buffer supplied in the kit and cell lysates were transferred to prechilled polypropylene tube. The samples were treated with Reaction Mix at room temperature for 15 min and the luminescence was read in a plate reader.
Bile acid measurement
Bile acids were measured by the Metabolomics Innovation Centre using an LC-multiple-reaction monitoring-MS method, as previously described60, with necessary modifications. Tissue samples were precisely weighed into safe-lock Eppendorf tubes, and water (2 μl per mg of tissue) was added. Samples were homogenized using two metal beads at a shaking frequency of 30 Hz for 2 min on a MM 400 mill mixer. Acetonitrile (8 μl per mg of tissue) was then added, followed by a second homogenization under the same conditions for 3 min. Samples were further subjected to ultrasonication for 2 min in an ice-water bath and centrifuged at 21,000g and 5 °C for 10 min. The clear supernatant (100 μl) was collected and mixed with 75 μl of an internal standard solution containing 14 isotope-labelled bile acids. The mixtures were dried under nitrogen gas, and the residues were reconstituted in 75 μl of 50% methanol. A standard solution containing all targeted bile acids was prepared in the same internal standard solution and serially diluted to generate 10 calibration standards. Aliquots (10 μl) of the calibration and sample solutions were injected into an Agilent 1290 UHPLC system coupled to a 6495B Agilent QQQ mass spectrometer. The mass spectrometer was operated in multiple-reaction monitoring mode with negative-ion detection. Chromatographic separation was performed on a C18 column (150 × 2.1 mm, 1.7 μm) using a binary-solvent gradient with mobile phase A (0.01% formic acid in water) and mobile phase B (0.01% formic acid in acetonitrile). Calibration curves for individual bile acids were generated using linear regression, and bile acid concentrations in the samples were determined by interpolating the calibration curves with the analyte-to-internal standard peak ratios obtained from sample injections, ensuring quantification within the appropriate concentration range.
Statistics and reproducibility
Data are reported as mean ± standard error of the mean (s.e.m.) unless stated otherwise. Statistical comparisons between two groups were performed by using unpaired t-test. Analysis of variance (ANOVA) was used for more than two groups. A P value of less than 0.05 was considered statistically significant. Statistical analysis and graphs were generated using GraphPad Prism software (v.9). All in vitro and in vivo experiments were repeated independently at least twice, and key experiments were performed many times using separate cohorts. All replication attempts yielded consistent results. Representative immunoblot images reflect two independent experiments, and microscopy images are representative of n = 3 per group, unless otherwise specified in the figure legends.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
