Cell culture, reagents and antibodies
HAP1 cells were maintained in IMDM medium supplemented with 10% FBS, L-glutamine, penicillin and streptomycin, and cultured at 37 °C in 5% CO2. 293T, A549, HeLa, HT29, RPE1, U251 and U2OS cells (purchased from ATCC) were maintained in DMEM under the same conditions. Cell lines used in this study were routinely monitored for mycoplasma contamination. BODIPY 493/503 and 665/676 neutral lipid dyes, TMRM and MitoTracker Red CM-H2XROS were from ThermoFisher. Brefeldin A, CCCP, cycloheximide, etomoxir, lipoprotein-deficient fetal calf serum, oleic acid, PF-06424439 (a DGAT2 inhibitor) and phosphatase-inhibitor cocktail were from Sigma-Aldrich. A-922500 (a DGAT1 inhibitor) was from Selleck. Protease inhibitor cocktail was from Roche. TAG (triolein) and DAG (diolein) were from Avanti Polar Lipids. 14C-DAG (1,2,-dioleoyl-rac-glycerol) was from American Radiolabeled Chemicals. Antibodies used in this study were anti-ACTB (Abcam, ab6276, WB 1:10,000), anti-α-tubulin (Santa Cruz Biotechnology, sc-32293, WB 1:1,000), anti-AMPK (Cell Signaling Technology, 2532, WB 1:1,000), anti-AMPK pT172 (Cell Signaling Technology, 2535, WB 1:1,000), anti-CANX (Abcam, ab22595, immunofluorescence (IF) 1:100, WB 1:1,000), anti-CLTC (Thermo Fisher, PA5-17347, WB 1:1,000), anti-EIF4G (Cell Signaling Technology, 2498, WB 1:1,000), anti-FASN (Santa Cruz Biotechnology, sc-55580, WB 1:1,000), anti-HA (Biolegend, 901503, IF 1:200, WB 1:1,000), anti-HSPA5 (Cell Signaling Technology, 3177, WB 1:1,000), anti-LAMP1 (Santa Cruz Biotechnology, sc-19992, WB 1:1,000), anti-LC3B (Cell Signaling Technology, 2775, WB 1:1,000), anti-LDHA (Cell Signaling Technology, 3582, WB 1:5,000), anti-PDI (Abcam, ab2792, IF 1:500, WB 1:20,000), anti-S6 (Cell Signaling Technology, 2137, WB 1:1,000), anti-S6 pS235/pS236 (Cell Signaling Technology, 4856, WB 1:1,000), anti-TMX1 (Atlas Antibodies, HPA003085, IF 1:100, WB 1:1,000 and Origene, TA507042, WB 1:1,000), anti-TOMM20 (Abcam, ab186735, WB 1:10,000) and anti-V5 (ThermoFisher, 14-6796-82, IF 1:500, WB 1:1,000).
Plasmids and cloning
We purchased sgRNAs as short single-stranded DNA with sticky ends. These oligonucleotides were then annealed and cloned into pX330 and pLentiCRISPRv2 (with puromycin, blasticidin or mCherry selection markers) that had been cut with BbsI or BsmBI (New England BioLabs), respectively. As a non-targeting control sgRNA (sgCTRL), a sequence targeting the zebrafish tia gene was used28. A complete list of sgRNAs used in this study, generated using the Broad Institute’s Genetic Perturbation Platform (portals.broadinstitute.org/gpp/public), is provided in Supplementary Table 1. 3×HA-tagged DIESL (Q96MH6-1, WT and N0) and V5-tagged TMX1 (Q9H3N1), TMX2 (Q9Y320-1), TMX3 (Q96JJ7-1) and TMX4 (Q9H1E5) were purchased as linear DNA flanked by NheI and AgeI restriction sites from Integrated DNA Technologies. These were then digested and cloned into pLEX305 backbones. DIESL(N5Q) and DIESL(H130A) mutants were generated by site-directed mutagenesis using the QuikChange II kit (Agilent). For recombinant expression of DIESL, human DIESL was codon-optimized for E. coli expression and pET24 constructs encoding DIESL, pelB–DIESL and pelB–3×HA–DIESL were purchased as plasmids from Twist Bioscience.
Generation of clonal knockout cell lines
HAP1 and 293T cells were transfected with pX330 encoding the sgRNA of choice, along with a plasmid either carrying an integration cassette28 or encoding a blasticidin or puromycin resistance gene, using Xfect (Takara), according to the manufacturer’s instructions. Transfected cells were selected with 25 µg ml−1 blasticidin (HAP1) or 1 µg ml−1 puromycin (HAP1 and 293T). After selection, the medium was replaced with complete medium and clones were allowed to form colonies, which were eventually picked by micropipette and transferred to 24-well plates. Genetic modification of individual clones was detected by PCR, in some cases using a primer annealing to the sequence of the blasticidin (5′-CCGACATGGTGCTTGTTGTCCTC-3′) or puromycin (5′-GCAACCTCCCCTTCTACGAGC-3′) resistance gene. See Supplementary Table 2 for a list of primers used to amplify genomic loci. Disruption of the locus was confirmed by Sanger sequencing, either directly or, in the case of compound heterozygotes, using TIDE analysis29 for 293T cell lines. The targeted genetic modifications in cell lines used in this study are summarized in Supplementary Table 3.
Lentiviral transduction
Lentivirus was produced in 293T cells transfected with p∆VPR, pVSVg and either pLEX305 or pLentiCRISPRv2 lentiviral plasmids, as well as pAdVAntage. Then, 2 days after transfection, virus was collected from conditioned medium passed through a 40 µm filter. Viral supernatants, supplemented with 8 µg ml−1 protamine sulfate, were applied directly to cells for 24 h. In some cases, a second collection and transduction was done to improve efficiency. Transduced cells were selected with puromycin (in HAP1, 1 µg ml−1; A549, 0.5 µg ml−1; HeLa, 2 µg ml−1; HT29, 3 µg ml−1; U251, 2 µg ml−1; U2OS, 2 µg ml−1) or transduction efficiency was assessed by expression of a fluorescent marker (mCherry). RPE1 cells were transduced with a large viral titre and were not selected. Transduced cells were analysed between 7 and 28 days after transduction.
Mutagenesis screening
All genetic screens reported here used gene-trap mutagenesis in haploid HAP1 cells, as described previously17. Typically, 2–3 × 109 gene-trapped HAP1 cells of the indicated genotype were collected by trypsinization and fixed in Fix Buffer I (BD Biosciences) for 10 min at 37 °C. For the oleic acid-loaded screen, cells were first cultured for 24 h in complete medium supplemented with 200 µM oleic acid and then chased in medium lacking oleic acid for another 24 h before collection. Cells were treated with 1 mg ml−1 RNase A (Qiagen) diluted in FACS buffer (10% FBS in PBS) at 37 °C for 30 min before staining with 1 µg ml−1 BODIPY 493/503 and 10 µg ml−1 propidium iodide (Sigma-Aldrich), diluted in FACS buffer, for 1 h at room temperature. Cells were washed twice in FACS buffer before being passed through a 40 µm cell strainer. Cell sorting was done using an S3 Sorter (Bio-Rad), collecting 107 cells (per population) representing the lowest and highest 5% of BODIPY signal from haploid cells in G1 phase. The isolation of genomic DNA, preparation of sequencing libraries and the analysis were done as described17. Reads were aligned to the reference genome (GRCh37 genome assembly), tolerating one mismatch, and sense insertions in the ‘low’ and ‘high’ populations were compared using the Fisher’s exact test to determine significant differences (P < 0.05). The mutational index represents the ratio of the occurrence of unique, disruptive (that is, insertions (ins.) of the gene-trap in the sense orientation) mutations in the body of a given gene (5′ untranslated region, exon and intron) in the high compared with the low population, normalized by the total of other unique, disruptive mutations in each population17: (sense ins. of gene X in high/(total sense ins. in high − sense ins. of gene X in high))/(sense ins. of gene X in low/(total sense ins. in low − sense ins. of gene X in low)). The number of recovered mutations for each population in each screen is reported in Supplementary Table 4.
SDS–PAGE and immunoblot analysis
Unless otherwise specified, cells were lysed in RIPA buffer (25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS and protease-inhibitor cocktail) on ice. Lysates were sonicated twice (40% amplitude for 4 s), cleared by centrifugation and protein concentrations were determined by BCA assay (Thermo Scientific). Equivalent amounts of protein were separated by SDS–PAGE over Bolt 4–12% Bis–Tris gels (Invitrogen) then transferred to nitrocellulose membranes. Membranes were blocked with 5% milk–TBST and incubated overnight at 4 °C in primary antibody diluted in 3% BSA–TBST. The next day, membranes were washed three times in TBST and then incubated in HRP-conjugated secondary antibodies (Invitrogen 65-6120 and Bio-Rad 1706516), diluted 1:10,000 in blocking buffer for 1 h at room temperature. After three more washes in TBST, signal was developed using Clarity Western ECL substrate (Bio-Rad) and imaged on a Universal Hood II GelDoc system (Bio-Rad).
Recombinant expression in E.
coli
BL21 E. coli were transformed with pET24 or pET28a plasmids expressing the indicated construct. Single clones were cultured up to a 5 ml volume in LB broth at an optical density of 0.5 at 37 °C. Cultures were then transferred to 25 °C for 1 h and expression was induced with 0.2 mM IPTG for 16 h at 25 °C. For SDS–PAGE and immunoblot analysis, 100 µl of culture was pelleted for 5 min at 3,500g before lysis in SDS–PAGE sample buffer.
TLC and DAG acylation assays
Equal numbers of human cells grown to confluence in six-well plates were washed with PBS. Cells were then incubated twice in 400 µl extraction buffer (hexane:isopropanol at 3:2) for 10 min30. E. coli cell pellets from 2 ml of the overnight culture were incubated in 500 µl extraction buffer (only once) for 30 min while being rotated. Extractions were pooled and dried under a stream of nitrogen to a volume of around 10 µl and then spotted on silica gel 60 TLC plates (Sigma-Aldrich). Plates were placed in a TLC tank containing a mobile phase to separate neutral lipids (hexane:diethyl ether:acetic acid at 80:20:1) or polar lipids (chloroform:methanol:water:acetic acid at 60:50:4:1). After separation, plates were air dried and stained with 0.2% Amido Black 10B (Sigma-Aldrich) dissolved in 1 M NaCl for 15–30 min (ref. 31). After staining, plates were rinsed with water and then washed several times with 1 M NaCl before drying overnight. Plates were imaged the next day on a Universal Hood II or EZ Imager GelDoc system (Bio-Rad).
NBD–DAG acylation in HAP1 cells
Unless otherwise specified, 25 µM NBD–DAG (Avanti Polar Lipids) was first conjugated to 0.125% fatty acid-free BSA in serum-free IMDM for 1 h at 37 °C. Cells were then incubated in this medium for 1 h before lipid extraction. TLC was done as described above. After drying, NBD fluorescence was monitored on a Universal Hood GelDoc system (Bio-Rad) using the fluorescein channel.
Cell-free 14C-DAG acylation
4KO HAP1 cells (naive or expressing 3×HA-DIESL) were washed in ice-cold PBS, collected in ice-cold buffer (20 mM HEPES pH 7.4, 250 mM sucrose, 2 mM MgCl2) and lysed by passing through a glass homogenizer 60 times on ice. Lysates were cleared by centrifugation at 600g for 10 min at 4 °C. Protein content was assayed and lysates were diluted to 2 mg ml−1 and stored at −80 °C. Reconstitution assays were based on assays described previously32 and were performed at the Radionuclides Centre of the Netherlands Cancer Institute according to its guidelines. First, 1.4 µl of 1.8 mM 14C-DAG, 3.6 µl buffer and 5.0 µl of 6.6 mg ml−1 fatty acid-free BSA were incubated on a heat block for 1 h at 37 °C with shaking at 850 rpm. We then added 100 µl of the indicated lysate and assays were incubated at 37 °C with shaking at 850 rpm for the indicated time (the final 14C-DAG concentration was 50 µM and the total radioactivity per assay was 0.14 µCi). For some controls, the reaction was incubated instead on wet ice or the lysate was heat-inactivated for 20 min at 90 °C before being added to the reaction. Reactions were quenched by performing a modified Bligh and Dyer extraction33; 110 µl chloroform and 110 µl methanol were added to each reaction, mixed and centrifuged at 20,000g for 2 min at room temperature. The bottom organic layer was transferred to a new tube, dried and reconstituted in 8 µl ethanol and separated by neutral-lipid TLC as described above. Plates were dried and imaged using a Typhoon FLA 9500 phosphorimager (General Electric, software v.1.1.0.187) after exposure to a BAS-TR2949 imaging plate (Fuji) for 3 days.
Immunoprecipitation and crosslinking
Formaldehyde crosslinking
∆DIESL and ∆TMX1∆DIESL HAP1 cells reconstituted with 3×HA-DIESL or WT HeLa cells expressing 3×HA-DIESL were washed with PBS and incubated in 1% paraformaldehyde for 20 min at room temperature. Crosslinking reactions were quenched by incubation in 0.2 M glycine for 10 min. Cells were then lysed in RIPA buffer on ice as described above, sonicated, and protein concentrations were determined. Samples were separated by SDS–PAGE and immunoblotted as described above.
Co-immunoprecipitation
∆DIESL HAP1 cells reconstituted with 3×HA-DIESL and grown to confluence in a six-well plate were lysed in 400 µl IP buffer (25 mM HEPES pH 7.0, 150 mM NaCl, 1% detergent (Tween-20 unless otherwise indicated) and protease- and phosphatase-inhibitor cocktails) on ice, sonicated and cleared by centrifugation. Then 250 µl of the lysate was incubated with 10 µl PBS-washed anti-HA magnetic beads (Pierce) overnight while being rotated at 4 °C. The next day, beads were washed three times in IP buffer before elution in SDS–PAGE sample buffer at 90 °C for 20 min. One-quarter of this eluate (along with the corresponding amount of input sample) were separated by SDS–PAGE and subjected to immunoblot analysis as described above.
Deglycosylation assay
∆DIESL HAP1 cells reconstituted with 3×HA-DIESL were lysed in buffer (25 mM Tris-HCl pH 7.5, 150 mM NaCl, 1% NP-40, protease-inhibitor cocktail) on ice, sonicated and protein levels were quantified as described above. N-glycans were removed using PNGase F (New England BioLabs) according to the manufacturer’s instructions. Denatured lysates (1 µg µl−1) were incubated for 15 min at 37 °C while being shaken at 850 rpm, in the presence or absence of 25 units µl−1 PNGase F. Reactions were separated by SDS–PAGE and analysed by immunoblotting as described above.
Subcellular fractionation and MAM purification
The MAM of the ER was purified from six confluent 15 cm plates of HAP1 cells, as described previously34, with all steps performed on ice or at 4 °C. Cells were washed in ice-cold PBS, collected in 750 µl per plate ice-cold homogenization buffer (10 mM HEPES pH 7.4, 250 mM sucrose) and passed through a glass homogenizer 100 times. The homogenate was centrifuged for 5 min at 600g. The pellet was resuspended in 2 ml homogenization buffer, homogenized and centrifuged again. Both supernatants were pooled and centrifuged at 10,300g for 20 min. The supernatant was put aside and the pellet (crude mitochondria) was resuspended in 1 ml isolation buffer 1 (5 mM HEPES pH 7.4, 250 mM mannitol, 0.5 mM EGTA). Next, 500 µl of resuspended mitochondrial pellet was layered on top of 1.5 ml Percoll solution (25 mM HEPES pH 7.4, 225 mM mannitol, 1 mM EGTA, 30% (v/v) Percoll) in two ultracentrifuge tubes. Mitochondria were fractionated by centrifugation at 95,000g for 30 min using a TLS-55 swinging-bucket rotor (Beckman). Pure mitochondria and crude MAM layers were collected using a syringe and resuspended in four volumes of isolation buffer 1 and isolation buffer 2 (25 mM HEPES pH 7.4, 225 mM mannitol, 1 mM EGTA), respectively. Purified mitochondria were centrifuged at 10,500g for 10 min. The crude MAM was cleared by centrifugation at 6,300g for 10 min and then both the MAM and the postmitochondrial supernatant from earlier were centrifuged for 1 h at 100,000g. The pellets (microsome and pure MAM) were collected and the cytosolic supernatant was concentrated by centrifugal filtration (10 kDa filter, Amicon). Protein concentration was determined by BCA assay (Thermo Scientific) and fractions were stored at −80 °C before immunoblot analysis.
Electron microscopy
Cells were prepared for transmission electron microscopy as previously described35. Grids were imaged on a Tecnai 12 G2 (ThermoFisher).
Immunofluorescence and confocal microscopy
Cells were grown on glass coverslips and fixed in 4% formaldehyde in PBS for 15 min at room temperature and then washed three times in PBS. Cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min and then washed three times in PBS and blocked in 1% BSA in PBS for 20 min. Cells were stained with primary antibodies, diluted in blocking buffer, for 1 h at room temperature. After three washes in blocking buffer, cells were stained with Alexa Fluor 488-, 568- and 647-conjugated secondary antibodies (Invitrogen A10042, A11011, A11004, A11008, A21245 and A31571) and/or neutral lipid dyes, diluted 1:500 in blocking buffer, for 1 h at room temperature. Cells were washed three times in PBS, counterstained with Hoechst 33342 (Invitrogen) diluted in PBS, and then washed three more times in PBS, before being mounted on glass slides using Aqua Poly/Mount (Polysciences). Cells were imaged by confocal laser-scanning microscopy using a Leica SP5 microscope with a 60× (1.4 NA) objective lens. Images were analysed using ImageJ (NIH).
Flow cytometry
BODIPY 493/503 measurements in fixed HAP1 cells
HAP1 cells grown on 10 cm plates were collected by trypsinization and fixed in Fix Buffer I (BD Biosciences) for 10 min at 37 °C. Cells were pelleted, washed with FACS buffer (10% FBS in PBS), resuspended in FACS buffer and counted. Next, about 10 million cells were stained with 1 µg ml−1 BODIPY 493/503 and 5 µg ml−1 DAPI (Invitrogen), diluted in FACS buffer, for 1 h at room temperature. Cells were washed once in FACS buffer then passed through a 35 µm nylon mesh cell strainer into a FACS tube. Fluorescence was analysed on an LSR Fortessa (BD Biosciences) analytical flow cytometer, using lasers with wavelengths of 405 nm and 488 nm to detect DAPI and BODIPY 493/503, respectively. Data were analysed using FlowJo (BD Life Sciences). Fluorescence plots represent the fluorescent signal measured in single haploid G1 cells, as determined by DAPI fluorescence intensity.
Mitochondrial measurements in live RPE1 cells
RPE1 cells were cultured in the presence or absence of lipoproteins, supplemented with 50 µM oleic acid where indicated, and pulsed for 30 min with either 600 nM TMRM or 250 nM MitoTracker Red CM-H2XROS in medium depleted of lipoproteins. In TMRM experiments, cells were then incubated in 150 mM TMRM for an additional 30 min, using 20 µM carbonyl cyanide m-chlorophenylhydrazone (CCCP) as a positive control for depolarization. Cells were collected by trypsinization and stored on ice in FACS buffer. Fluorescence (using a 561 nm laser) and data were analysed as described above. Fluorescence plots represent the fluorescent signal measured in single cells. Membrane potential was calculated by subtracting the fluorescence mean of depolarized (CCCP-treated) cells.
ATP measurements
RPE1 cells were plated in 96-well plates and ATP levels were measured using CellTitre-Glo (Promega) according to the manufacturer’s instructions. ATP levels were normalized by measuring protein content in a parallel plate by BCA assay (Thermo Scientific).
Mice
Generation and genotyping
Animal experiments were performed with the approval of the National Ethics Committee for Animal Experiments of the Netherlands, in accordance with the relevant guidelines and regulations (including laboratory and biosafety regulations). Frozen embryos (morula stage, from C57/BL6J mice) carrying a single disruptive allele in Diesl (also known as Tmem68) tm1a(EUCOMM)Wtsi (herein referred to as gt(lacZ-neo)), were purchased from the European Conditional Mouse Mutagenesis Program (EUCOMM). After thawing, embryos were developed into blastocysts overnight in KSOM medium in an incubator at 37 °C, and were then implanted into C57/BL6N foster female mice and carried to term. Mice were genotyped by PCR, using primer combinations to detect the WT (5′-GCTCCCTTCCATTTACTCTG-3′ and 5′-CCGGTGAGATAGCTAACAAG-3′) and mutant (5′-CTTATCATGTCTGGATCCGG-3′ and 5′-CCGGTGAGATAGCTAACAAG-3′) alleles. Because C57/BL6J mice have a deletion in the Nnt gene that influences DIESL protein levels36, only mice with 6N alleles at the Nnt locus were used in this study, and this was monitored by PCR, using primer combinations to detect the 6N (5′-GGGCATAGGAAGCAAATACCAAGTTG-3′ and 5′-GTAGGGCCAACTGTTTCTGCATGA-3′) and 6J (5′-GTGGAATTCCGCTGAGAGAACTCTT-3′ and 5′-GTAGGGCCAACTGTTTCTGCATGA-3′) alleles. To generate mice for analysis, mice heterozygous for the mutant Diesl allele were intercrossed to generate knockout and control mice. Mice were born at the expected Mendelian ratio (P = 0.4279, chi-squared test, n = 341 mice) and maintained on a diet of RM3-SAFE (4.2% fat, Special Diet Services). Mice were maintained in a certified animal facility at 21 °C and 55% humidity in 12 h:12 h light:dark cycles. No sample-size calculations were performed. No randomization was performed because animals were assayed based on genotype. In experiments where subjectivity could be introduced, the experimenter was blinded to the details.
PCR with reverse transcription
RNA from mouse liver tissue was isolated using the Qiagen RNeasy Mini Kit, and cDNA libraries were synthesized with SuperScript III reverse transcriptase (Invitrogen) using random hexamers. The DIESL transcript was detected by PCR using a primer pair spanning exons 4 and 5 (5′-GAGCCATCCCCATAGACTTTTACTACTTC-3′ and 5′-CCCGGTGAGATAGCTAACAAGTGAC-3′). In the knockout, the disruptive cassette was integrated between these two exons. The ACTB transcript was used as a control and was amplified using the following primer pair: 5′-ATCCTGACCCTGAAGTACCCCA-3′ and 5′-CCTCTCAGCTGTGGTGGTGAAGCTGTAGCCACGCT-3′. Although DIESL has previously been reported to be a brain-specific protein19, we readily detected DIESL expression in the mouse liver, in agreement with proteomic data36. Furthermore, expression data from the Human Protein Atlas indicates that DIESL expression could be detected in all tissues that were tested37.
Histology
Tissues and organs were collected and fixed in EAF fixative (ethanol:acetic acid:formaldehyde:saline at 40:5:10:45 v:v) and embedded in paraffin. Sections were prepared at 2 µm thickness from the paraffin blocks and stained with haematoxylin and eosin according to standard procedures. The sections were reviewed with a Zeiss Axioskop2 Plus microscope (Carl Zeiss Microscopy) and images were captured with a Zeiss AxioCam HRc digital camera and processed with AxioVision 4 software (both from Carl Zeiss Vision).
Indirect calorimetry
Indirect calorimetry was done using fully automated metabolic cages (LabMaster, TSE systems). Mice were housed individually in a 12 h:12 h light:dark cycle (light, 07:00–19:00 hours) with ad libitum access to water and a standard chow diet (10% fat, 23% protein and 67% carbohydrate; V1554-703, Ssniff Spezialdiäten). After at least 24 h of acclimatization, O2 consumption (({V}_{{{rm{O}}}_{2}})), CO2 production (({V}_{{{rm{CO}}}_{2}})) and caloric intake were measured for at least three consecutive days followed by 12 h fasting and refeeding. The RER was calculated as ({V}_{{{rm{CO}}}_{2}}/{V}_{{{rm{O}}}_{2}}). Glucose oxidation was calculated using the formula (((4.585times {V}_{{{rm{CO}}}_{2}})-(3.226times {V}_{{{rm{O}}}_{2}}))times 4) and fat oxidation was calculated as (((1.695times {V}_{{{rm{O}}}_{2}})-(1.701times {V}_{{{rm{C}}{rm{O}}}_{2}}))times 9). Values were corrected for the lean mass of each individual mouse.
Body composition
Fat mass and lean tissue mass were assessed at several time points in non-anaesthetized mice using NMR (MiniSpec LF90 BCA-analyser).
Shotgun lipidomics
Sample preparation
4KO (∆DGAT1∆DGAT2∆DIESL∆TMX1) HAP1 cells, reconstituted with WT or H130A DIESL, were cultured in complete medium. Prior to collection, cells were cultured in serum-free IMDM for 30 min. Cells were then collected by trypsinization, washed twice with PBS and 3 million cells (in triplicate) were pelleted and stored at −80 °C. For E. coli samples, 2 ml of induced culture (as described above, in triplicate) was pelleted and incubated overnight at −80 °C. The next day, bacterial pellets were thawed and incubated in 25 µg ml−1 lysozyme in PBS for 30 min. Cells were pelleted, washed in PBS and pelleted again before resuspension in distilled water. Cells were sonicated at 60% amplitude for 5 min, in continuous cycles of 5 s on and 25 s off. The insoluble fraction was pelleted by centrifugation and the lysate (supernatant) was quantified for protein and stored at −80 °C. For mouse tissue samples, tissues were collected from mice fed ad libitum and stored at −80 °C. Tissues were then thawed on ice, weighed and homogenized in PBS (50 mg ml−1 tissue) using a glass homogenizer with 20 and 40 strokes for brain and liver tissue, respectively. Homogenates were diluted 10-fold further in PBS and homogenized again with 20 strokes. Homogenates (5 mg ml−1) were stored at −80 °C. Lipid abundancies are expressed either in weight (where indicated, corrected for the amount of protein in the sample) or as a molar percentage of the total lipid species in the sample (mol%).
Lipid extraction for MS lipidomics
MS-based lipid analysis was performed by Lipotype as described38. Lipids were extracted using a two-step chloroform–methanol procedure39. Samples were spiked with internal lipid standard mixture containing: cardiolipin 14:0/14:0/14:0/14:0 (CL), ceramide 18:1;2/17:0 (Cer), DAG 17:0/17:0, hexosylceramide 18:1;2/12:0 (HexCer), lyso-phosphatidate 17:0 (LPA), lyso-phosphatidylcholine 12:0 (LPC), lyso-phosphatidylethanolamine 17:1 (LPE), lyso-phosphatidylglycerol 17:1 (LPG), lyso-phosphatidylinositol 17:1 (LPI), lyso-phosphatidylserine 17:1 (LPS), phosphatidate 17:0/17:0 (PA), phosphatidylcholine 17:0/17:0 (PC), phosphatidylethanolamine 17:0/17:0 (PE), phosphatidylglycerol 17:0/17:0 (PG), phosphatidylinositol 16:0/16:0 (PI), phosphatidylserine 17:0/17:0 (PS), cholesterol ester 20:0 (CE), sphingomyelin 18:1;2/12:0;0 (SM) and TAG 17:0/17:0/17:0. Chain composition is given as length:saturation;oxidation. After extraction, the organic phase was transferred to an infusion plate and dried in a speed vacuum concentrator. The first-step dry extract was resuspended in 7.5 mM ammonium acetate in chloroform:methanol:propanol at 1:2:4 (v:v:v) and the second-step dry extract in 33% ethanol solution of methylamine:chloroform:methanol at 0.003:5:1 (v:v:v). All liquid-handling steps were performed using the Hamilton Robotics STARlet robotic platform with the anti-droplet control feature for organic solvents pipetting.
MS data acquisition
Samples were analysed by direct infusion on a QExactive mass spectrometer (Thermo Scientific) equipped with a TriVersa NanoMate ion source (Advion Biosciences). Samples were analysed in both positive and negative ion modes with a mass-to-charge ratio resolution (Rm/z) of Rm/z = 200 = 280,000 for MS and Rm/z = 200 = 17,500 for tandem MS (MS/MS) experiments, in a single acquisition. MS/MS was triggered by an inclusion list encompassing corresponding MS mass ranges scanned in 1-Da increments40. Both MS and MS/MS data were combined to monitor CE, DAG and TAG ions as ammonium adducts; PC, PC O- as acetate adducts; and CL, PA, PE, PE O-, PG, PI and PS as deprotonated anions. MS alone was used to monitor LPA, LPE, LPE O-, LPI and LPS as deprotonated anions; Cer, HexCer, SM, LPC and LPC O- as acetate adducts.
Data analysis and postprocessing
Data were analysed by Lipotype with in-house lipid-identification software based on LipidXplorer41,42. Data postprocessing and normalization were performed using an in-house data management system. Only lipid identifications with a signal-to-noise ratio greater than 5 and a signal intensity 5-fold higher than in corresponding blank samples were considered for further data analysis.
Homology modelling and data analysis
The DIESL AlphaFold22 model was accessed via the EMBL–EBI portal (https://alphafold.ebi.ac.uk/) and visualized using PyMOL. Data wrangling, statistical analyses and plot generation were done using Prism (GraphPad Software) and RStudio (https://www.rstudio.com/).
Statistics and reproducibility
The number of replicates is indicated in the figure legends. All attempts at replication were successful.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.