Mice
Table of Contents
The following mouse lines were used: Dars2fl/fl (ref. 2), Cox10fl/fl (ref. 4), Vil1-cre (ref. 7) and Villin-creERT2 (ref. 37). Sdhatm2a mice were obtained from the Knock Out Mouse Project repository (project ID: CSD48939) and bred to FLP deleter mice38 to delete the FRT-flanked region to generate Sdhafl/fl mice. IEC-specific knockout mice were generated by intercrossing mice carrying the respective loxP-flanked alleles with Vil1-cre or Villin-creERT2 transgenic mice. Both female and male mice between 1 and 12 weeks of age were used in all in vivo experiments, whereas metabolic tracing studies were performed exclusively using male mice. All mice were maintained on the C57BL/6N background. Mice were housed at the specific-pathogen-free animal facilities of the CECAD Research Center of the University of Cologne under a 12-h dark–12-h light cycle in individually ventilated cages (Greenline GM500, Tecniplast) at 22 ± 2 °C and a relative humidity of 55 ± 5%. All mice had unlimited access to water and fed a standard chow diet (Harlan diet no. 2918 or Prolab Isopro RMH3000 5P76) ad libitum. For the experiments assessing the role of dietary fat, mice were fed a FFD (E15104-3474, ssniff-Spezialdiäten) containing only traces of fat (<0.5%). All animal procedures were conducted in accordance with European, national and institutional guidelines and protocols were approved by local government authorities (Landesamt für Natur, Umwelt und Verbraucherschutz Nordrhein-Westfalen) and Animal Welfare Officers of the University Medical Center Hamburg-Eppendorf and Behörde für Gesundheit und Verbraucherschutz Hamburg. Animals requiring medical attention were provided with appropriate care and were culled humanely when reaching pre-determined termination criteria to minimize suffering. No other exclusion criteria were applied. Villin-CreERT2 recombinase activity was induced by five consecutive daily intraperitoneal administrations of 1 mg tamoxifen dissolved in corn oil and DMSO. Littermates not carrying the Vil1-cre or Villin-creERT2 transgenes were used as controls in all experiments.
Tissue preparation
The colon and SI were dissected and washed with PBS. Small pieces (about 0.5 cm) were isolated proximal (after the stomach) and distal (before the caecum) of the SI, snap-frozen on dry ice for RNA expression analysis and stored at –80 °C until further processing. The remaining SI tissue was cut longitudinally and washed in PBS to remove faeces. Intestinal tissue samples were rolled up from proximal to distal to form a Swiss roll and either fixed in 4% paraformaldehyde overnight at 4 °C or embedded in TissueTek for frozen sectioning.
H&E staining of paraffin-fixed tissues
Paraffin-embedded 3-μm-thick intestinal tissue sections were deparaffinized with xylene and rehydrated with decreasing amounts of ethanol solutions (100% ethanol, 96% ethanol and 75% ethanol). Sections were stained for 2 min in haematoxylin, differentiated in tap water for 15 min and incubated for 1 min in eosin. Stained sections were dehydrated using increasing amounts of ethanol solutions and fixed in xylene for 1 min. Slides were mounted with Entellan.
COX and SDH and ORO staining of fresh-frozen tissues
Fresh-frozen 7-μm-thick intestinal sections were sequentially stained for COX and SDH activity. Cryosections were dried and incubated for 45 min at 37 °C with COX solution. Then they were briefly washed with PBS and incubated for 40 min with SDH solution at 37 °C. Following dehydration through graded alcohol solutions, the sections were mounted with DPX and stored at room temperature. Fresh-frozen 10-μm-thick sections were fixed in 4% paraformaldehyde for 15 min at room temperature. After fixation, the sections were washed with ddH2O and stained with ORO in isopropanol/water (60:40) for 15 min. All sections were counterstained with haematoxylin for 5 min and mounted with Aquatex (EMD Millipore).
Immunohistochemistry and immunofluorescence on intestinal sections
Paraffin sections were rehydrated and heat-induced antigen retrieval was performed in 10 mM sodium citrate, 0.05% Tween-20 at pH 6.2 or with proteinase K treatment. Endogenous peroxidase was blocked in peroxidase blocking buffer for 15 min at room temperature. Sections were blocked in 1% BSA, 0.2% fish-skin gelatin, 0.2% Triton-X-100 and 0.05% Tween-20 in PBS for 1 h at room temperature. After blocking, the sections were incubated overnight at 4 °C with primary antibodies against adipophilin/PLIN2 (Progen, GP46, 1:500), Ki67 (Dako, M724901, clone 1O15, 1:1,000), OLFM4 (Cell Signaling, D6Y5A, clone D6X5A, 1:400), CC3 (Cell Signaling, 9661, 1:1,000), CC8 (Cell Signaling, 8592, 1:1,000), CD45 (BD Bioscience, 560510, clone 30-F11, 1:500) and F4/80 (AbD Serotec, MCA497, clone A3-1, 1:1,000). Sections were incubated with biotinylated anti-mouse IgG (H+L) (Vector Laboratories, BA-9200-1.5, 1:1,000), anti-rabbit IgG (H+L) (Vector Laboratories, BA-1000-1.5, 1:1,000) and anti-rat IgG (H+L) (Vector Laboratories, BA-9400-1.5, 1:1,000) secondary antibodies. Each staining was visualized using ABC Kit Vectastain Elite (Vector, PK6100) and DAB substrate (Dako and Vector Laboratories). Immunofluorescence was performed with primary antibodies against TGN38 (bio-techne, AF8059-SP, 1:200), E-cadherin (BD Biosciences, 610182, 1:1,000) and adipophilin/PLIN2 (Progen, GP46, 1:200). Nuclei were stained using DAPI (Vector Laboratories) and visualized with anti-sheep IgG NorthernLights NL557 (bio-techne, NL010, 1:300), anti-mouse Alexa 488 (Molecular Probes, A1101, 1:300) and anti-guinea pig Alexa 633 (Molecular Probes, A21105, 1:300) fluorescence-conjugated secondary antibodies. Periodic acid–Schiff (PAS) reaction was performed according to standard protocols. Endogenous alkaline phosphatase activity was visualized using a Fast Red Substrate kit according to the manufacturer’s instructions (ab64254, Abcam). For image acquisition, the intestinal sections were analysed using a light microscope equipped with a KY-F75U digital camera (JVC) (DM4000B, Leica Microsystems, Diskus 4.50 software), a TCS SP8 confocal laser scanning microscope (Inverse, DMi 8 CS, Leica Microsystems LAS X, Lightning software v.5.1.0) or a LSM Meta 710 confocal laser scanning microscope (Carl Zeiss Technology, ZEN 2009 software). Golgi quantification was performed using ImageJ software (v.2.0.0.-rc-46/1.50g) as previously described39. The number and size of TGN38-positive fluorescent objects were quantified using the ‘analyse particles’ function after applying a fixed threshold on pictures derived from maximal 2D projections of the acquired confocal stacks. Each data point corresponds to the average values from at least three randomly selected intestinal areas of a single mouse. Representative pictures from 4–5 mice per genotype per time point were analysed. More than 100 IEC profiles per mouse with visible nuclei were quantified (n = 128–527).
EM analysis
A piece of 0.5 cm proximal SI tissue was fixed overnight in 2% glutaraldehyde (Merck) and 2% paraformaldehyde (Science Services) in 0.1 M cacodylate buffer (AppliChem). Tissue samples were treated with 1% OsO4 (Science services) in 0.1 M cacodylate buffer for 2 h. After dehydration of the sample with ascending ethanol concentrations followed by propylene oxide, samples were embedded in Epon (Sigma-Aldrich). Ultrathin sections (70 nm thick) were cut, collected onto 100 mesh copper grids (Electron Microscopy Sciences) and stained with uranyl acetate (Plano) and lead citrate (Sigma Aldrich). Images were captured using a transmission electron microscope (Joel JEM2100 Plus) at an acceleration voltage of 80 kV, and pictures were acquired using a 4K-CCD camera, OneView (GATAN). Mitochondrial morphological integrity quantification was performed on randomly selected pictures of the proximal SI areas from four Dars2fl/fl and four Dars2tamIEC-KO mice. Each mitochondrial profile was classified as normal, partly affected or severely damaged based on its electron density, the appearance of the cristae and the extent of matrix loss (Fig. 2c). The relative distribution of the analysed mitochondria per mouse into the three morphological groups is presented. A total of 663 mitochondrial profiles from 69 IECs versus 707 mitochondrial profiles from 80 IECs were quantified.
Cell culture conditions and drug treatments
IEC-6 cells (ACC 111) were purchased from the Leibniz Institute DSMZ–German Collection of Microorganisms and Cell Cultures and maintained in standard conditions at 37 °C and 5% CO2. The cell culture medium was composed of 45% Dulbecco’s modified Eagle medium (ThermoFisher, 41965-039), 45% RPMI 1640 (ThermoFisher, 11875093) and 0.1 U ml–1 human insulin solution (Sigma, I9278) supplemented with 10% FCS (Bio&SELL). IEC-6 cells were routinely checked for mycoplasma contamination and tested negative. For induction of mitochondrial dysfunction, 70–80% confluent cells were treated for 48 h with 100 μM actinonin (A6671, Sigma-Aldrich) or 1 μM atpenin A5 (ab144194, Abcam). All compounds were solubilized in dimethyl sulfoxide (DMSO) (A3672, PanReac AppliChem). Control cells were treated with corresponding amounts of DMSO, which did not exceed 1% in culture medium. Treatments were renewed every 24 h. IEC-6 cells were incubated with 5 μg ml–1 brefeldin A (B6542, Abcam) for 6 h. To induce LD formation, oleic acid (O1008, Sigma-Aldrich) was complexed to fatty acid-free BSA (A6003, Sigma-Aldrich) at a ratio of 6:1 and used at a concentration of 600 μM after titration for 24 h.
Immunofluorescence of cultured cells
Immunofluorescence staining was performed on IEC-6 cells cultured on coverslips and fixed in 4% paraformaldehyde for 15 min. Reactive aldehydes were quenched with 50 mM NH4Cl for 10 min and the cells were permeabilized with 0.1% Triton-X-100 in PBS for 5 min. After 20 min in blocking solution (0.2% fish-skin gelatin diluted in PBS), IEC-6 cells were incubated with primary antibodies against TGN38 (bio-techne, AF8059-SP, 1:200) and MTCO1/COX1 (Molecular Probes, 459600, 1D6E1A8, 1:100) for 30 min at room temperature, followed by incubation with anti-sheep IgG NorthernLights NL557 (bio-techne, NL010, 1:300) or anti-mouse Alexa 488 (Molecular Probes, A1101, 1:300) fluorescence-conjugated secondary antibodies for 30 min at room temperature. When LDs were stained, 5 μM of BODIPY 493/503 (D3922, Invitrogen) diluted in PBS was applied for 30 min. Finally, IEC-6 cells were mounted in Vectashield containing DAPI. For image acquisition, a TCS SP8 confocal laser scanning microscope (Inverse, DMi 8 CS, Leica Microsystems LAS X, Lightning software v.5.1.0) was used. Quantification of Golgi morphology was performed using ImageJ software (v.2.0.0.-rc-46/1.50g) on 2D projections from Z-stack images. A total of 4–6 randomly selected viewing fields per condition, capturing at least 30 cells per image, were used. Golgi morphology was classified into five distinct categories based on TGN38-positive fluorescent objects (Extended Data Fig. 11a,b) as follows: (1) normal (juxtanuclear Golgi ribbon composed of connected stacks); (2) ring (ring-like Golgi structures surrounding the entire nucleus); (3) condensed (bulb-shaped juxtanuclear Golgi structure); (4) fragmented (Golgi ribbon replaced by more and smaller tubules and vesicles positive for TGN38); and (5) dispersed (complete loss of Golgi ribbon and dispersal of the TGN38 signal). Quantification was performed by manually classifying the TGN38 pattern in each cell in one of the five Golgi phenotypes by the same observer, who was blinded to the experimental conditions. Three independent experiments were quantified.
Measurement of serum parameters
Glucose (GLU2), total cholesterol (CHOL2), triacylglycerol (TRIGL), high-density lipoprotein (HDLC4) and low-density lipoprotein (LDLC3) levels in the blood serum from mice aged 1–12 weeks old were measured using standard assays in a Cobas C111 Biochemical Analyzer (Roche Diagnostics).
Isolation of mitochondria and analysis of mitochondrial respiratory complexes with blue native electrophoresis
Mitochondria isolation
The SI was chopped into small pieces and homogenized with a rotating Teflon potter (Potter S, Sartorius; 20 strokes, 1,000 r.p.m.) in a buffer containing 100 mM sucrose, 50 mM KCl, 1 mM EDTA, 20 mM TES and 0.2% fatty acid-free BSA, pH 7.6 followed by differential centrifugation at 850g and 8,500g for 10 min at 4 °C. Mitochondria were washed with BSA-free buffer, and protein concentrations were determined using Bradford reagent. Mitochondria were subjected to blue native polyacrylamide gel electrophoresis (BN-PAGE) followed by western blot analysis or determination of the in gel activity of respiratory complexes.
BN-PAGE
Mitochondrial protein concentrations were determined using Bradford reagent (Sigma). A total of 20 μg of mitochondria was lysed for 15 min on ice in dodecylmaltoside (5 g g–1 of protein) for individual respiratory complexes, or digitonin (6.6 g g–1 protein) for supercomplexes, and cleared from insoluble material for 20 min at 20,000g, 4 °C. Lysates were combined with Coomassie G-250 (0.25% final). Mitochondrial complexes were resolved by BN-PAGE using 4–16% NativePAGE Novex Bis-Tris mini gels (Invitrogen) in a Bis-Tris/Tricine buffering system with cathode buffer initially supplemented with 0.02% G-250 followed by the 0.002% G-250.
Complex I in-gel activity
Gels were incubated in a buffer containing 0.01 mg ml–1 NADH and 2.5 mg ml–1 nitrotetrazolium blue in 5 mM Tris-HCl pH 7.4.
Western blot analysis
Separated mitochondrial complexes were transferred onto a polyvinylidene fluoride membrane using a wet transfer methanol-free system. Membranes were immunodecorated with indicated antibodies followed by ECL-based signal detection. The following antibodies were used: anti-MTCO1 (Molecular Probes, 459600, clone 1D6E1A8, 1:5,000), anti-COX4L1 (Molecular Probes, A21348, clone 20E8C12, 1:1,000), anti-UQCRC1 (Molecular Probes, 459140, clone 16D10AD9AH5, 1:4,000), anti-NDUFS1 (Proteintech, 12444-1-AP, 1:1,000), anti-NDUFS2 (Abcam, ab96160, 1:1,000), anti-NDUFV2 (Proteintech, 15301-1-AP, 1:1,000), anti-UQCRFS1/RISP[5A5] (Abcam, ab14746, clone 5A5, 1:1,000), anti-ATP5A (Abcam, ab14748, 1:3,000), anti-SDHA (Molecular Probes, 459200, clone 2EGC12FB2AE2, 1:5,000) and anti-NDUFA9 (Molecular Probes, 459100, clone 20C11B11B11, 1:1,000).
Isolation of IECs
SI tissue was collected from mice, washed in DPBS (14190-094, Gibco) to remove faeces and cut longitudinally. IECs were isolated by sequential incubation of intestinal tissue in pre-heated 1 mM dithiothreitol and 1.5 mM EDTA solutions at 37 °C while shaking. Pellets of IECs were frozen at −80 °C for further processing.
Protein lysate preparation
IEC pellets were lysed in RIPA lysis buffer (10 mM Tris-Cl (pH 8), 140 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% Triton X-100 and 0.1% SDS). Lysis buffer was supplemented with protease and phosphatase inhibitor tablets (Roche). The protein concentration was measured using Pierce 660 nm Protein Assay reagent (22660, Thermo Scientific) and a BSA standard pre-diluted set ranging from 0 to 2,000 μg ml–1 (23208, Thermo Scientific). Cell lysates were separated on SDS–PAGE and transferred to polyvinylidene fluoride membranes (IPVH00010, Millipore). A protein size ladder (26620, Thermo Scientific) was used for size comparison. Membranes were blocked with 5% milk and 0.1% PBST and were probed overnight with primary antibodies against the following antibodies: DARS2 (Proteintech, 13807-1-AP, 1:1,200); total OXPHOS rodent WB antibody cocktail (Abcam, ab110413, 1:1,000); MTCO1 (Molecular Probes, 459600, clone 1D6E1A8, 1:5,000); COX4L1 (Molecular Probes, A21348, clone 20E8C12, 1:1,000); UQCRC1 (Molecular Probes, 459140, clone 16D10AD9AH5, 1:4,000); NDUFS1 (Proteintech, 12444-1-AP, 1:1,000); NDUFS2 (Abcam, ab96160, 1:1,000); NDUFV2 (Proteintech, 15301-1-AP, 1:1,000); UQCRFS1/RISP[5A5] (Abcam, ab14746, Clone 5A5, 1:1,000); ATP5A (Abcam, ab14748, 1:3,000); SDHA (Molecular Probes, 459200, clone 2EGC12FB2AE2, 1:5,000); NDUFA9 (Molecular Probes, 459100, clone 20C11B11B11, 1:1,000); α-tubulin (Sigma Aldrich, T6074, clone TUBA4A, 1:1,000); TOMM70 (Sigma, HPA014589, 1:500); β-actin (Santa Cruz, sc-1616, clone I-19, 1:1,000); adipophilin/PLIN2 (Progen, GP46, 1:500); FABP2 (Proteintech, 21252-1-AP, 1:500); FASN (Cell Signaling, 3189S, 1:1,000); vinculin (Cell Signaling, 13901, 1:1,000); and ApoB (Beckman Coulter, 467905, 1:500). Membranes were incubated for 1 h at room temperature with anti-rabbit IgG (GE Healthcare, NA934V, 1:5,000), anti-mouse IgG (GE Healthcare, NA931, 1:5,000), anti-goat IgG (Jackson Laboratories, 705-035-003, 1:5,000) or anti-guinea pig IgG (Progen, 90001, 1:5,000) secondary HRP-coupled antibodies and Amersham ECL Western Blotting Detection reagent (GE Healthcare) were used. The membranes were re-probed after incubation in Restore Western Blot stripping buffer (21059, ThermoFisher). The signal was measured with a Curix 60 Processor and a western blot imager (FUSION Solo X, Vilber).
RNA isolation from tissues
SI tissue samples were disrupted using a Precellys 24 tissue homogenizer (Bertin technologies). Isolation of RNA was performed using a NucleoSpin RNA isolation kit (Macherey Nagel ref. 740955.250) according to the manufacturer’s instructions.
RT–qPCR
cDNA was prepared using a Superscript III cDNA-synthesis kit (18080-044, Thermo Scientific). RT–qPCR was performed using TaqMan probes (Life Technologies) and SYBR Green (Thermo Scientific). The mRNA expression of each gene was normalized to the expression of the housekeeping genes Tbp or Hprt1. Relative expression of gene transcripts was analysed using the 2–ΔΔCt method. The RT–PCR data were collected using QuantStudio 12K Flex Software v.1.6 (Applied Biosystems). The following Taqman probes were used: Olfm4 (Mm01320260_m1, Thermo Scientific), Lgr5 (Mm00438890_m1, Thermo Scientific), Ascl2 (Mm01268891_g1, Thermo Scientific), Tbp (Mm00446973_m1, Thermo Scientific), Prominin-1 (Mm00477115_m1, Thermo Scientific) and Lrig-5 (Mm00456116_m1, Thermo Scientific). Primer sequences for SYBR Green are described in Supplementary Table 5.
C.
elegans strains, maintenance and imaging
Strains were cultured on OP50 Escherichia coli-seeded NGM plates, according to standard protocols40. Strains used in this study are Bristol N2, RT1315 unc-119(ed3); pwIs503[pvha-6::mans::gfp;cbr–unc-119], VS25 hjIs14 [vha-6p::GFP::C34B2.10(SP12) + unc-119(+)] and RT130 pwIs23 [vit-2::GFP]. RNAi knockdown was performed as previously described41. All the experiments were performed with hermaphrodite worms at days 1 and 4 of adulthood that were randomly selected and were not allocated into groups. dars-2, sar-1, sec-13 and fum-1 clones were obtained from the Ahringer RNAi library41 and confirmed by sequencing. As a control, empty L4440 vector was used. For confocal imaging, animals were immobilized on 2% agarose pads in 5 mM levamisole buffer and imaging was performed using a spinning disc confocal microscope (Inverse, Nikon TiE, UltraView VoX, Perkin Elmer, Volocity software). For fluorescence imaging, worms were immobilized on 2% agarose pads in 50 mM sodium azide buffer and imaged using the optical Zeiss Axio Imager Z1 microscope (ZEN 2009 software). Images were analysed using the open-source software Fiji (ImageJ, v.1.53c).
RNA isolation and RT–qPCR in C.
elegans
Worms were collected from a 9 cm plate and total RNA was isolated using Trizol (Invitrogen). DNAse treatment was performed using DNA-free, DNAse and removal (Ambion, Life technologies) according to the manufacturer’s protocol. RNA was quantified by spectrophotometry and 0.8 μg of total RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems). For each condition, six independent samples were prepared. qPCR was performed using a Step One Plus Real-Time PCR system (Applied Biosystems) with the following PCR conditions: 3 min at 95 °C, followed by 40 cycles of 5 s at 95 °C and 15 s at 60 °C. Amplified products were detected using SYBR Green (Brilliant III Ultra-Fast SYBR Green qPCR Master Mix, Agilent Technologies). Relative quantification was performed against Y45F10D.4.
The following primers were used: dars-2 FW1 (5′-GTTTGCTGGGGAAATTCAGA-3′); dars-2 RV1 (5′-AGTGGAGCCGTAAATGGATG-3′); Y45F10D.4 FW (5′-GTCGCTTCAAATCAGTTCAGC-3′); and Y45F10D.4 RV (5′-GTTCTTGTCAAGTGATCCGACA-3′). Data were analysed using ΔΔCt analysis.
Lipidomics
For lipid analyses, mouse tissue samples were homogenized in deionized water (10 μl per 1 mg wet weight) using a Precellys 24 homogenizer (Peqlab) at 6,500 r.p.m. for 30 s. The protein content of the homogenate was routinely determined using bicinchoninic acid.
Liquid chromatography coupled to electrospray ionization tandem mass spectrometry
Sphingolipid (ceramides and sphingomyelins) and cholesterol levels in mouse SI tissue were determined by liquid chromatography coupled to electrospray ionization tandem mass spectrometry (LC–ESI-MS/MS). For sphingolipid analyses, 50 μl of tissue homogenate was used. Lipid extraction and LC–ESI-MS/MS analysis were performed as previously described42,43. For the determination of cholesterol levels, 25 μl of tissue homogenate was extracted and processed as previously described44.
Nano-ESI-MS/MS
Levels of cholesteryl esters (CEs), diacylglycerols (DAGs), TAGs and glycerophospholipids in mouse SI tissue were determined by nano-ESI-MS/MS). Next, 10 μl (for DAGs) or 5 μl (for TAGs and CEs) of tissue homogenate was diluted to 500 μl with Milli-Q water and mixed with 1.875 ml of chloroform, methanol and 37% hydrochloric acid 5:10:0.15 (v/v/v). Next, 20 μl of 4 µM d5-TG internal standard mixture I (for TAGs), 15 μl of 256 μM CE 19:0 (for CEs) or 20 μl of 4 μM d5-DG internal standard mixtures I and II (for DAGs) (Avanti Polar Lipids) were added. Lipid extraction and nano-ESI-MS/MS analyses of DAGs and TAGs were performed as previously described45. The detection of CE species was conducted in positive-ion mode by scanning for precursors of m/z 369 Da at a collision energy of 15 eV and with a declustering potential of 100 V, an entrance potential of 10 V and a cell exit potential of 14 V. Levels of glycerophospholipids (that is, phosphatidylcholines, phosphatidylethanolamines, phosphatidylinositols, phosphatidylserines and phosphatidylglycerols) were determined by performing extraction and nano-ESI-MS/MS measurement of 10 μl of tissue homogenate as previously described46.
Metabolomics
Metabolite extraction
Metabolite extraction solution (50% methanol, 30% acetonitrile, 20% water and 5 μM valine-d8 as internal standard) was added to 10–20 mg frozen SI tissue samples at an extraction ratio of 25 μl mg–1 on dry ice. Samples were then homogenized using a Precellys 24 tissue homogenizer (Bertin Technologies). The resulting sample suspension was vortexed, mixed at 4 °C in a Thermomixer for 15 min at 1,500 r.p.m. and then centrifuged at 16,000g for 20 min at 4 °C. The supernatant was collected for LC–MS analysis.
Metabolite measurement by LC–MS
LC–MS chromatographic separation of metabolites was achieved using a Millipore Sequant ZIC-pHILIC analytical column (5 μm, 2.1 × 150 mm) equipped with a 2.1 × 20 mm guard column (both 5 mm particle size) with a binary solvent system. Solvent A was 20 mM ammonium carbonate and 0.05% ammonium hydroxide. Solvent B was acetonitrile. The column oven and autosampler tray were held at 40 °C and 4 °C, respectively. The chromatographic gradient was run at a flow rate of 0.200 ml min–1 as follows: 0–2 min: 80% solvent B; 2–17 min: linear gradient from 80% solvent B to 20% solvent B; 17–17.1 min: linear gradient from 20% solvent B to 80% solvent B; 17.1–22.5 min: hold at 80% solvent B. Samples were randomized and analysed with LC–MS in a blinded manner with an injection volume of 5 μl. Pooled samples were generated from an equal mixture of all individual samples and analysed interspersed at regular intervals within the sample sequence as a quality control. Metabolites were measured using a Thermo Scientific Q Exactive Hybrid Quadrupole-Orbitrap mass spectrometer (HRMS) coupled to a Dionex Ultimate 3000 UHPLC. The mass spectrometer was operated in full-scan, polarity-switching mode, with the spray voltage set to +4.5 kV/–3.5 kV, the heated capillary held at 320 °C and the auxiliary gas heater held at 280 °C. The sheath gas flow was set to 25 units, the auxiliary gas flow was set to 15 units and the sweep gas flow was set to 0 unit. HRMS data acquisition was performed in a range of m/z = 70–900, with the resolution set at 70,000, the automatic gain control (AGC) target at 1 × 106 and the maximum injection time at 120 ms. Metabolite identities were confirmed using two parameters: (1) precursor ion m/z was matched within 5 ppm of theoretical mass predicted by the chemical formula; (2) the retention time of metabolites was within 5% of the retention time of a purified standard run with the same chromatographic method.
Data analysis
Chromatogram review and peak area integration were performed using the Thermo Fisher software Tracefinder (v.5.0). The peak area for each detected metabolite was subjected to the ‘Filtering 80% Rule’, half minimum missing value imputation and normalized against the total ion count of that sample to correct any variations introduced from sample handling through instrument analysis. Samples were excluded after performing testing for outliers based on geometric distances of each point in the PCA score plot as part of the muma package (v.1.4)47. Afterwards, differential metabolomics analysis was performed. In detail, the R package ‘gtools’ (v.3.8.2) (cran.r-project.org/web/packages/gtools/index.html) was used to calculate the log2(fold change) using the functions ‘foldchange’ and ‘foldchange2logratio’ (parameter base = 2).The corresponding P value was calculated using the R base package ‘stats’ (v.4.0.5) (www.r-project.org) with the function ‘t.test’ (SIMPLIFY = F). The P value was adjusted using the stats base function ‘p.adjust’ (method = “bonferroni”). Volcano plots were generated using the EnhancedVolcano package48 (v.1.8.0).
QuantSeq 3′ mRNA sequencing
RNA quality was evaluated based on the RNA integrity number (RIN) and OD260/280 and OD260/230 ratios. RIN values were determined using TapeStation4200 and RNA Screen Tapes (Agilent Technologies). Gene expression was determined using a QuantSeq 3′ mRNA-Seq Library Prep kit FWD for Illumina (Lexogen). QuantSeq libraries were sequenced on an Illumina NovaSeq 6000 sequencer using Illumina RTA v.3.4.4 base-calling software. Sample exclusion criteria were OD260/280 < 1.8, OD260/230 < 1.5 and RIN < 7. Illumina adapters were clipped off the raw reads using Cutadapt with standard parameters and a minimum read length of 35 after trimming (shorter reads were discarded). QuantSeq-specific features were deliberately not removed to avoid loss of reads. Trimmed reads were mapped to a concatenation of the mouse genome (Mus_musculus.GRCm38.dna.chromosome.*.fa.gz, downloaded from ftp.ensembl.org/pub/release-100/fasta/mus_musculus/dna/) and the ERCC92 Spike In sequences (downloaded from assets.thermofisher.com/TFS-Assets/LSG/manuals/ERCC92.zip) using subread-align version v.2.0.1 with parameters -t 0 -d 50 -D 600 --multiMapping -B 5. Genomic matches were counted using featureCounts with parameters -F “GTF” -t “exon” -g “gene_id” --minOverlap 20 -M --primary -O --fraction -J -Q 30 -T 4. The genome annotation used was Mus_musculus.GRCm38.100.gtf (downloaded from ftp.ensembl.org/pub/release-100/gtf/mus_musculus/), augmented by entries for the ERCC92 Spike Ins.
All analyses were done in R-4.0.0, using the functionality of Bioconductor v.3.11. For differential gene expression analysis, the package DESeq2 was used (bioconductor.org/packages/release/bioc/html/DESeq2.html).
Pairwise comparisons were performed between genotypes TG and WT (differential_expression_DESeq2_tg_VS_wt.xlsx). Genes were excluded from a DESeq2 run if they had a zero count in more than half of the samples in either of the conditions compared. Note that DESeq2 sets the P value and the adjusted P value to NA for genes with too few counts or with extreme outlier counts. Such genes were removed after analysis from the DESeq2 output.
The output tables were augmented by gene symbols and descriptions, which were derived from the org.Mm.eg.db annotation package using the function AnnotationDbi::mapIDs (bioconductor.org/packages/release/bioc/html/AnnotationDbi.html). In addition, the raw read counts per gene and sample, as returned by featureCounts, were appended to the rows of each output table. The statistical test producing the P values is the Wald test, and P‐adjusted values were calculated using the false discovery rate and Benjamini–Hochberg approach. It was computed using the function nbinomWaldTest of the Bioconductor R package DESeq2, based on a negative binomial general linear model of the gene counts from a previously described method49.
Proteomics
In-solution digestion for MS
Samples for MS analysis were prepared by in-solution digestion. Protein (20 μg) was precipitated for at least 1 h in four volumes (v/v) of ice-cold acetone and protein pellets were extracted by centrifugation at 13,000g for 10 min and dissolved in urea buffer (6 M urea, 2 M thiourea in 10 mM HEPES, pH 8.0). Urea-containing samples were reduced by applying tris(2-carboxyethyl)phosphine at a final concentration of 10 mM, alkylated with chloroacetamide at a final concentration of 40 mM and incubated for 1 h at room temperature. Samples were then digested with 1 μl LysC for 2 h at room temperature, diluted with 50 mM ammonium bicarbonate to a urea concentration of 2 M, incubated with 1 μl 0.5 mg ml–1 trypsin overnight at room temperature, acidified to 1% formic acid and purified using Stop and Go extraction tips (StageTips)50.
MS-based proteome analysis
Proteome samples were analysed using LC–MS/MS on an Orbitrap Eclipse Tribrid mass spectrometer (Thermo Fisher) with a FAIMS Pro device using a combination of two compensation voltages of –50 V and –70 V. Chromatographic peptide separation was achieved on 50 cm reverse-phase nanoHPLC-columns (ID 75 μm, PoroShell C18 120, 2.4 μm) coupled to an EASY-nLC 1200 HPLC system and a binary buffer system A (0.1% formic acid) and B (80% acetonitrile/0.1% formic acid). Samples derived from in-solution digestion were measured over a 120 min gradient, raising the content of buffer acetonitrile from 3.2 to 22% over 102 min, from 22 to 45% over 8 min and from 45 to 76% over 2 min. The column was washed with 76% acetonitrile for 8 min. Full MS spectra (300–1,750 m/z) were recorded at a resolution of 60,000, maximum injection time of 20 ms and automatic gain control target of 6 × 105. The 20 most abundant ion peptides in each full MS scan were selected for higher-energy collisional dissociation fragmentation at nominal collisional energy of 30. MS2 spectra were recorded at a resolution of 15,000, a maximum injection time of 22 ms and an automatic gain control target of 1 × 105. This MS acquisition program was alternatively run for both FAIMS compensation voltages to cover different peptide fractions.
MS data processing and analysis
The generated MS raw data were analysed using MaxQuant analysis software and the implemented Andromeda software (v.1.6.14)51,52. Peptides and proteins were identified using the canonical mouse UniProt database (downloaded August 2019) with common contaminants. All parameters in MaxQuant were set to the default values. Trypsin was selected as the digestion enzyme, and a maximum of two missed cleavages was allowed. Methionine oxidation and amino-terminal acetylation were set as variable modifications, and carbamidomethylation of cysteines was chosen as a fixed modification. The label-free quantification algorithm was used to quantify the measured peptides and the ‘match between runs’ option was enabled to quantify peptides with a missing MS2 spectrum. Subsequent statistical analysis was performed using Perseus (1.5.8.5) software. Potential contaminants and reverse peptides were excluded, and values were log2 transformed. Raw files were assigned to two groups (TG and WT) and protein groups were filtered for four valid values in at least one group before missing values were replaced from normal distribution (width of 0.3; down shift of 1.3). Welch’s Student t-test with S0 = 0.1 and a permutation-based false discovery rate of 0.01 with 500 randomizations was performed to obtain differentially regulated proteins between the two groups. Identified proteins were annotated with the gene ontology terms biological process, molecular function, and cellular compartment, and the Reactome Pathway database. Finally, graphical visualization was achieved using Instant Clue software53 (v.0.5.3).
GSEA and data visualization
Gene set enrichment methods were applied using GSEA and over-representation analysis (ORA). In detail, GSEA was performed by using gene sets published on the MsigDB (Reactome, KEGG, Biocarta and Hallmarks)54 and from a published study55 (ATF4) using the packages fgsea56 (v.1.16.0) and GSEABase57 (v.1.52.1). Volcano plots were generated using the EnhancedVolcano package48 (v.1.8.0). The ORA was performed using the ‘enrich_GO’ function (parameters: keyType = “ENTREZID”, OrgDb = org.Mm.eg.db, ont = “ALL”, pAdjustMethod = “BH”, qvalueCutoff = 0.1) of the clusterProfiler package58(v.3.16.1). The output data were plotted using the ‘emapplot’ function of the enrichplot package (v. 1.8.1) (www.bioconductor.org/packages/release/bioc/html/enrichplot.html) (parameters: pie_scale = 1, showCategory = 40, layout = “nicely”).
Metabolic tracer studies
Postprandial glucose and fat tolerance tests
Mice were fasted for 2 h before receiving an oral gavage of 300 μl of a glucose–lipid emulsion containing triolein (3.6 g kg–1 body weight), lecithin (0.36 g kg–1 body weight) and glucose (2 g kg–1 body weight), traced with [3H]triolein (1.4 MBq kg–1 body weight) and [14C]DOG (1.7 MBq kg–1 body weight). After 2 h, mice were anaesthetized and transcardially perfused with PBS containing 10 U ml–1 heparin. Organs were collected, weighed and dissolved in 10× (v/w) Solvable (Perkin Elmer), and radioactivity (in d.p.m.) was measured by scintillation counting using a Perkin Elmer Tricarb scintillation counter. Uptake of radioactive tracers was calculated per total organ weight.
CM production
Mice were injected with tyloxapol (500 mg in 0.9% NaCl per kg body weight) to block vascular lipolysis. Mice received an oral gavage of a lipid emulsion with triolein (3.6 g kg–1 body weight) and lecithin (0.36 g kg–1 body weight) that were traced with [14C]cholesterol (1.4 MBq kg–1 body weight) and [3H]triolein (1.7 MBq kg–1 body weight). Blood was collected from the tail vein at 0, 30, 60 and 120 min after gavage. Plasma triglycerides were determined by standard colorimetric assays (Roche) and radioactivity was measured by scintillation counting.
Plasma parameters
Plasma was generated by centrifugation of EDTA-spiked blood for 10 min at 10,000 r.p.m. at 4 °C in a bench top centrifuge. Free glycerol was determined photometrically using Free Glycerol reagent (F6428, Sigma). For lipoprotein profiling, 150 μl pooled plasma was diluted with an equal amount of FPLC buffer (total 300 μl), which was separated by fast-performance liquid chromatography (FPLC) on a Superose 6 10/300 GL column (GE Healthcare) with a flow rate of 0.5 ml min–1. Forty fractions (0.5 ml each) were collected, and cholesterol and triglyceride concentrations were measured in each one.
To isolate the TRL fractions, 200 μl of plasma was mixed with 200 μl density solution 1 (0.9% NaCl, 10 mM EDTA, 10 mM Tris-Cl pH 8.6 and 0.49 g ml–1 KBr; density 1.3 g l–1). The density solution 2 (0.9% NaCl, 10 mM EDTA, 10 mM Tris-Cl pH 8.6; density 1.006 g l–1) was placed into a Beckman TL100 centrifuge tube (Beckman, 343778) and then the plasma carefully under layered. Ultracentrifugation was performed in a Beckman Optima MAX-XP ultracentrifuge for 2 h at 4 °C and 40,000 r.p.m. in a Beckman TL100 rotor. After centrifugation, 200 μl of the top containing TRL particles were collected using a syringe.
Statistical analysis
Data shown in column graphs represent the mean ± s.e.m., as indicated in the figure captions. The D’Agostino–Pearson omnibus normality test was applied to test normal (Gaussian) distribution. When data fulfilled the criteria for normality, unpaired two-sided Student’s t-tests with no assumption of equal variance were performed; otherwise, the nonparametric Mann–Whitney U-test was chosen. Multiple pairwise comparisons of groups over time by repeated measures were evaluated by two-way ANOVA with Bonferroni’s correction for multiple comparison (the corrected P values are given for comparison between genotypes at each time point). Survival curves were compared using Gehan–Breslow–Wilcoxon test. The chi-squared test was used for the comparison of the mitochondria integrity distribution between two groups and the assessment of the Golgi pattern distribution after various inhibitor treatments in IEC-6 cells. The number of mice analysed in each experiment is described in the respective figure captions. Statistical analyses were performed with GraphPad Prism 6 (v.6.01) and 9 (v.9.4.1).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.