Plasmids

Guide RNAs targeting human or mouse ACAD10 and ACAD11 were inserted into the BbsI site of the vector pX458 [64]. Primer sequences are listed in Table S1.

To re-express ACAD10 or ACAD11 in knockout cell lines, we used lentiviral vectors based on the plasmid pLVX-PURO (Clontech/Takara) which expressed the puromycin resistance gene under the control of the PGK promoter. Gene expression is either driven by the CMV promotor (pUB83) or the SV40 promoter (pUB82).

To overexpress proteins in HEK293 cells we used the lentiviral vector pOH233-1 (containing a C-terminal SFB tag consisting of an S-tag, two FLAG tags and a streptavidin-binding peptide) [65] or the equivalent containing a N-terminal SFB tag (pOH147). Site-direct mutagenesis was performed by inserting two overlapping PCR products via Gibson assembly (Hifi Assembly Master Mix, New England Biolabs) [66]. Constructs for the expression of bacterial proteins were generated using a synthetic geneblock (IDT) as a template. Primers and Plasmids are presented in Tables S1 and S2, and detailed maps are available upon request.

Cell culture

HepG2 human hepatoblastoma, HEK293 and HEK293T human embryonic kidney cell lines (gifts from Eric R. Fearon, University of Michigan) were cultured in DMEM (Biowest) supplemented with 10% fetal bovine serum (FBS) (Cytiva), 1 mM Ultraglutamine (Lonza), 100 µg/mL streptomycin, and 100 U/mL penicillin (100 U/mL) at 37 °C in the presence of 5% CO2 and at 100% humidity. Mouse 3T3L1 preadipocytes were grown in the same medium but with 10% newborn calf serum instead of FBS.

Recombinant lentiviruses were produced by transient transfection of HEK293T cells with lentiviral vectors and second generation packaging plasmids psPAX2 and pMD2.G (kind gifts of Didier Trono, Addgene #12260 and #12259) as described before [67]. Culture medium was changed 6 h after transfection, and recombinant viruses were recovered in the culture supernatant 24 h later. Target cells were transduced by incubation with virus-containing culture supernatant in the presence of 4 μg/mL polybrene (Sigma). Infected cells were selected 24 h later for 4 days with puromycin (Thermofisher) at concentrations of 1.5 μg/mL for HepG2 and HEK293 cells and 1 µg/mL for 3T3L1 cells.

To generate knockout cell lines, cells were transfected in 6-well plates with a 2 µg CRISPR/Cas9—guide-RNA expression plasmids and 4 µL lipofectamine 2000 following the manufacturer’s instructions (Life Technologies). Transfected cells were selected by flow cytometrical sorting gating for GFP fluorescence on a FACSAria III flow cytometer. The specific combination of guide RNAs and vector systems used for individual clones is listed in table S1 and S2. Individual clones were expanded and analyzed by Sanger sequencing (Eurofins).

Quenching and metabolite extraction

Metabolomic analyses were performed essentially as previously reported [68]. Cell lines were plated in 6-well plates (HepG2) or 6 cm dishes (3T3L1). Lysates were obtained after quenching of metabolism with liquid nitrogen as described before (12). Briefly, culture plates were submerged in liquid nitrogen after one rapid wash with ice-cold water. 250 μL of a solution consisting of 90% methanol (Biosolve) and 10% chloroform (Fisher Scientific) was added and lysates were transferred into microcentrifuge tubes. After centrifugation for 15 min at 4 °C and 22 000 g, the supernatant was recovered, dried in a SpeedVac vacuum concentrating system (Life Technologies) and resuspended in 35 μL of 1:1 methanol:water before analysis.

To quench enzymatic reactions, we added 4 volumes of ice-cold methanol, followed by centrifugation to precipitate proteins. The supernatant was recovered for LC–MS analysis and brought to 50% methanol.

Quantification of metabolites by mass-spectrometry

Analyses by LC–MS were performed as previously described [68] based on a method by Coulier and colleagues [69]. Briefly, 5 μL of sample were analyzed with an Inertsil 3 μm particle ODS-4 column (150 × 2.1 mm; GL Biosciences) at a constant flow rate of 0.2 mL/min with an Agilent 1290 HPLC system. Mobile phase A consisted of 5 mM hexylamine (Sigma-Aldrich) adjusted to pH 6.3 with acetic acid (Biosolve) and phase B of 90% methanol (Biosolve)/10% 10 mM ammonium acetate (Biosolve) adjusted to pH 8.5 with ammonia (Merck, Darmstadt, Germany). The mobile phase profile consisted of the following steps and linear gradients: 0–2 min at 0% B; 2–6 min from 0 to 20% B; 6–17 min from 20 to 31%B; 17–36 min from 31 to 60% B; 36–41 min from 60 to 100% B; 41–51 min at 100% B; 51–53 min from 100 to 0% B; 53–60 min at 0% B.

Analytes were identified and quantified with an Agilent 6550 mass spectrometer with an electrospray ionization (ESI) source in negative mode using the following settings: ESI spray voltage 3500 V, sheath gas 350 °C at 11 L/min, nebulizer pressure 35 psig and drying gas 200 °C at 14 L/min. An m/z range from 70 to 1200 was acquired with a frequency of 1 per second by summing 8122 transients. Compound identification was based on their exact mass (< 5 ppm) and retention time compared to standards (Sigma Aldrich, and synthesized in the lab) (Supplementary Table S3). The areas under the curve (AUC) of extracted-ion chromatograms of the [M-H]− forms were integrated using MassHunter Quantitative Analysis Software (Agilent, Santa Clara, CA, USA), and normalized to the mean of the areas obtained for a series of 150 other metabolites (‘total ion current’). Further normalization steps are indicated in the figure legends.

To separate hexenoyl-CoA standards, 5 µL sample were separated with a Acquity BEH Shield RP18 column (C18, 2.1 × 100 mm, 1.7 µm; Waters) column at a constant flow of 0.2 mL/min. Mobile phase A consisted of 5 mM ammonium acetate pH 8.3 and phase B of 90% acetonitrile and 10% 5 mM ammonium acetate pH 8.3. The mobile phase profile contained the following steps and linear gradients: 0–3 min at 3% B; 3–40 min from 3 to 40% B; 40–50 min from 40 to 100% B; 50–51 min from 100 to 3% B; 51–61 min 3% B.

Affinity purification of ACAD10 with epitope tags from HepG2 cells

HepG2 cells were transduced with recombinant lentiviruses driving expression of human ACAD10 carrying an N-terminal HA-tag and a C-terminal streptavidin-binding peptide (SBP). Cell lines were selected with puromycin. Subsequently, cells were washed with phosphate-buffered saline (PBS), removed from plates in PBS using a cell scraper, and collected by centrifugation at 400 g and 4 °C. The pellets were resuspended in 300 µL lysis buffer (150 mM NaCl, 1 mM EDTA, 20 mM Tris–HCl pH 8.0, 0.2% NP40/Igepal CA-630) per subconfluent 100 mm plates. After sonication, lysates were centrifuged at 27,000 g for 15 min at 4 °C.

To purify ACAD10 via the N-terminal HA-tag, we used Thermo Scientific ™ Pierce™ anti-HA Magnetic beads. Briefly, a protein quantity of 16 mg was applied to 50 uL beads. After rotation for 1 h at 4 °C, the solution was centrifuged at 4 °C and 400 g and the supernatant was removed. Beads were washed four times in 1 ml lysis buffer and the pellet was stored at -80 °C. Purification of ACAD10 via a C-terminal SBP was performed as described below.

Proteomic analysis

16 mg of proteins were used as starting material to pulldown tagged proteins. 50 µL of Streptavidin Sepharose High-Performance beads or Thermo Scientific™ Pierce™ Anti-HA Magnetic Beads were used, accordingly. After purification (see above), beads were resuspended in 50 µL of a 5% SDS solution containing 50 mM Tris pH 8.5 and 5 mM DTT. After 15 min at 55 °C, chloroacetamide was added to a final concentration (f.c.) of 20 mM followed by another 10 min incubation at room temperature. Samples where then acidified by addition of phosphoric acid to a f.c. of 2.5%. 25 µL of each sample were transferred onto a micro S-Trap column (Protifi LLC, USA). Digestion was performed at 37 °C overnight with a 1:100 ratio of trypsin and LysC. Digested peptides were eluted in three steps using 40 µL each of 50 mM Tris pH 8.5, 0.2% formic acid and 50% acetonitrile. The eluted peptides were dried down in a vacuum concentrator (SpeedVac, Thermo Scientific) and resuspended in 2% Acetonitrile and 0.2% formic acid. Peptide concentration was determined by Pierce™ Quantitative Peptide Assay (Thermo Scientific). Peptide separation was performed using a reversed-phase analytical column (EasySpray, 0.075 × 250 mm, Thermo Scientific) with a linear gradient of 4–27.5% solvent B (0.1% FA in 80% ACN) for 37 min, 27.5–50% solvent B for 20 min, 50–95% solvent B for 10 min at a constant flow rate of 300 nL/min on a Vanquish Neo HPLC system. The peptides were analyzed by an Orbitrap Fusion Lumos tribrid mass spectrometer with an ESI source (ThermoFisher Scientific) coupled online to the nano-LC. Peptides were detected in the Orbitrap at a resolution of 120,000. Peptides were selected for MS/MS using the HCD setting at 30; ion fragments were detected in the Orbitrap at a resolution of 30,000. A data-dependent procedure alternating between one MS scan followed by MS/MS scans was applied for 3 s for ions above a threshold ion count of 50,000 in the MS survey scan with 20.0 s dynamic exclusion. The electrospray voltage applied was 2.1 kV. MS1 spectra from m/z 300 to 1800 were obtained with an AGC target of 400,000 and a maximum injection time set to custom. MS2 spectra were acquired with an AGC target of 10,000 and a maximum injection time set to custom.

The resulting MS/MS data were processed using the Sequest HT search engine within Proteome Discoverer 2.5 SP1 against a database containing only ACAD10 corresponding sequences (wild-type full length containing HA and Streptavidin tags and a shorter version containing the C-terminus). Trypsin or GluC was specified as cleaving enzyme in a setting were a specific cleavage only on one side was required, allowing up to 2 missed cleavages, 4 modifications per peptide and up to 7 charges. Mass error was set to 20 ppm for precursor ions and 0.05 Da for fragment ions. Oxidation on Met (+ 15.995 Da) and Methionine loss (− 131.040 Da) on the N-terminus of the protein and peptides were considered as variable modifications, whereas carbamidomethylation of cysteine was considered as a fixed modification. The False discovery rate (FDR) was assessed using the target decoy PSM validator node, and thresholds for the identification of proteins, peptides and modification sites were specified at 1%. Label-free quantification of peptides is based on the precursor ion intensity. Signals were normalized to the sum of all signals within each individual sample. Protein abundances were calculated as the sum of the abundances of unmodified peptides. Raw and processed mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD050316.

Western blot analysis

Western Blot analyses were performed as described before [70]. Briefly, cell lysates were prepared in in RIPA buffer [150 mM NaCl, 20 mM Tris/HCl (pH 7.5), 1% Nonidet P40, 0.5% sodium deoxycholate and 0.1% SDS] containing a Complete™ proteinase inhibitor cocktail (Merci), followed by sonication. Tisues were powdered in liquid nitrogen prior to homogenization and sonication. Protein concentration was determined using the BCA assay and equal amounts were resolved on 10% polyacrylamide gels and transferred on to PVDF membranes (Millipore) using tank-transfer system (Bio-Rad). Membranes were blocked for 1 h with 5% non-fat dried skimmed milk powder in TBST (Tris-buffered saline containing 0.1% Tween) at room temperature (22 °C). Incubations with primary antibodies were performed overnight at 4 °C in Tris-buffered saline containing 2% BSA. Antibody concentrations were 1:1000 for ACAD10 (17161-1-AP, Proteintech), 1:5000 anti-FLAG (M2, Sigma), 1:2000 anti-HA (C29F4, Cell signaling) and 1:5000 anti-β-actin (Sigma). Subsequently, membranes were washed and incubated in horseradish peroxidase (HRP)-coupled secondary antibodies in TBST containing 5% non-fat dried skimmed milk powder. Signals were revealed using chemiluminescent HRP substrates (Milipore) and detected using a chemiluminescent detection system (Cytiva). Quantifications were performed using Imagequant TL (Cytiva). Presented values represent data from at least three independent experiments.

Prediction of cleavage site for mitochondrial processing peptidase

To predict internal mitochondrial cleavage sites with algorithms designed to detect N-terminal mitochondrial targeting sequences [28, 29], we submitted a series of progressive N-terminal one amino acid deletions to the prediction servers and noted the predicted existence and identity of the cleavage site.

Purification of recombinant ACAD10 and ACAD11

HEK293 cells were infected with recombinant lentivectors driving expression of full length or fragments from ACAD10 or ACAD11. Stable cell lines were selected with Puromycin. After two washes with phosphate-buffered saline, cells were removed from plates in PBS using a cell scraper and collected by centrifugation. Pellets were resuspended in 4 volumes of lysis buffer [150 mM NaCl, 1 mM EDTA, 20 mM Tris–HCl, pH 8.0, 1% NP40/Igepal CA-630, 1 µg/mL leupeptin, 1 µg/mL aprotinin] with 5 µM FAD. After sonication, lysates were clarified by centrifugation at 27,000 g for 15 min at 4 °C. We added 100 μL of streptavidin Sepharose beads (GE healthcare) per twenty 100 mm plates. After rotation for 1 h at 4 °C, beads were washed four times in lysis buffer, followed by one wash with 1 mL buffer containing 150 mM NaCl and 20 mM Tris, pH 8). Proteins were frozen down at -80 °C bound to beads in aliquots corresponding to two 100 mm plates. Before enzymatic reactions, pellets were resuspended in 25 mM Tris–Cl pH 7.4. Protein concentrations were determined by comparing band intensities with those obtained for bovine serum albumin in polyacrylamide gels stained with Coomassie blue (Pageblue, Fisher).

Purification of bacterial enzymes

The expression construct for FadD was obtained from the ASKA collection [71], whereas the other ones are described above (see also table S2). Expression plasmids were transformed into the E. coli BL21 Rosetta (DE3) strain using electroporation.

For the expression of Pseudomonas putida LvaE and the Burkholderia pseudomallei ortholog of LvaA and LvaC, overnight cultures were diluted 1:50 in 100 to 500 mL Lysogeny broth (LB) containing the required antibiotics (30 µg/mL kanamycin or 100 µg/mL ampicillin). Cultures were incubated at 37 °C while shaking at 200 rpm until the optical density at 600 nm reached 0.5. Expression was induced with 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) followed by an incubation at 16 °C for 18 h with agitation at 200 rpm. Bacteria were collected by a 20 min centrifugation at 5000 × g and at 4 °C. Pellets were stored at − 80 °C until purification.

For the expression of His-tagged E. coli FadD, an overnight culture at 37 °C was diluted 1:200 in fresh LB containing 30 µg/mL chloramphenicol. The culture was incubated at 20 °C until the optical density at 600 nm reached 0.5. Expression was then induced with 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) followed by an incubation at 20 °C for 4 h with agitation at 200 rpm. Bacteria were collected by a 20 min centrifugation at 5,000 × g and 4 °C. Pellets were stored at − 80 °C until purification.

Purification His6-tagged Pseudomonas putida LvaE (Q88EH6) and the Burkholderia pseudomallei ortholog (Q63VL0) of LvaA (Q88G01) was adapted from Rand et al. [18]. Briefly, frozen pellets were thawed on ice and resuspended in lysis buffer containing 50 mM Na2HPO4, 300 mM NaCl, 10 mM imidazole, 2 mM DTT, pH 8.0 and 2 µL of benzonase. Cell suspensions were lysed by three cycles of sonication and 2 passages through a French press, followed by centrifugation at 25,000 × g and 4 °C for 30 min. The supernatant was filtered through a 0.45 µm filter (Sartorius). An AKTA liquid chromatography system (Cytiva) was used at a constant flow rate of 1 mL/min. A 1 mL HisTrap HP column was equilibrated with 5 column volumes (CV) of wash buffer [50 mM Na2HPO4, 300 mM NaCl, 40 mM imidazole, 2 mM DTT, pH 8.0]. After sample application, the column was washed with 15 CV wash buffer, followed by protein elution using a linear gradient reaching 100% elution buffer [50 mM Na2HPO4, 300 mM NaCl, 250 mM imidazole, 2 mM DTT, pH 7.8] over the course of 20 min, collecting 1 mL fractions. A G25 Sepharose column (PD-10 desalting, Cytiva) was used to remove imidazole after equilibration with a buffer containing 100 mM NaCl, 2 mM MgCl2 and 2 mM DTT. Absorbance at 280 nm was measured, and protein concentrations were computed using extinction coefficients calculated with the ProtParam tool on the ExPASy server. Protein was stored at − 80 °C until use.

For purification of the MBP-tagged LvaC ortholog from Burkholderia pseudomallei (Q63VL1), pellets were resuspended in 50 mM Tris–HCl pH 7.5, 50 mM NaCl, 5 mM EDTA, 2 mM β-mercaptoethanol, 2 mM MgCl2, 1 mg/mL lysozyme, 1 µg/mL aprotinin, 1 µg/mL leupeptin, and 2 mM p-toluene-sulfonyl fluoride. Cell lysis and recovery of soluble proteins was achieved as described above. The supernatant was incubated with 1 mL of prewashed amylose beads (E8021S, New England Biolabs) for 1 h on a rotating device at 4 °C. The bead suspension was spun down at 400 g for 5 min and supernatant was removed. The beads were resuspended in 5 mL lysis buffer and transferred into a 10 mL disposable column polyethylene filter (Thermo Scientific). The column was washed with 10 mL lysis buffer adjusted to 200 mM NaCl (wash buffer). Proteins were eluted by 9 successive additions of 500 µL elution buffer [wash buffer with 10 mM maltose]. To enhance elution, the column outflow was blocked for 15 min after addition of the first portion.

For purification of His-tagged E. coli FadD, frozen pellets were thawed on ice and resuspended in lysis buffer containing 50 mM Na2HPO4 pH 7.5, and 300 mM NaCl. Cell suspensions were lysed by three cycles of sonication, followed by centrifugation at 25,000 × g and 4 °C for 30 min. An AKTA liquid chromatography system was used at a constant flow rate of 1 mL/min. 1 mL HisTrap HP column was equilibrated with 10 column volumes (CV) of wash buffer [50 mM Na2HPO4, 300 mM NaCl, 10%glycerol]. After sample application, the column was washed with 30 CV wash buffer, followed by protein elution with 20 CV using a linear gradient from 2 to 100% elution buffer [50 mM Na2HPO4, 300 mM NaCl, 500 mM imidazole, 10% glycerol] over the course of 20 min, collecting 1 mL fractions. Two 1 ml fractions were dialyzed against lysis buffer to remove imidazole, using a Spectra/Por® Dialysis membrane with a 3 kDa molecular weight cut-off.

Synthesis of substrates

The 4-OH, 5-OH and 6-OH fatty acids used in this study were obtained after opening γ, δ, ε-lactones respectively (Sigma 303836, 389579, 704067, V403, W279609, W278106, W253901, D804). Saponification of lactones was performed as follows: 50 mM NaOH was added to a same volume of 50 mM lactone. The mixture was heated at 90 °C for 10 min. After cooling to room temperature, the solution was neutralized using HCl.

Production of R-4-OH-C6-CoA and S-4-OH-C6-CoA occurred in 1 ml reactions, with 10 mM R-4-OH-C6 (Sigma 75378) or S-4-OH-C6 (Sigma 77011), 8.5 ng/μL LvaE, 2.5 mM CoA, 2 U/mL inorganic pyrophophatase (Roche), and a buffer consisting of 25 mM TrisHCl pH 7.4, 10 mM ATP, 10 mM MgCl2, 200 µM TCEP, 0.1 μg/μL BSA. After 1 h incubation at 37 °C, reactions were stopped in liquid nitrogen, and stored at − 80 °C until purification.

Different isomers of hexenoyl-CoA were synthesized as described above for 4-OH-C6-CoA, but using the corresponding hexenoic acid as substrate (trans-2-hexenoic acid—Sigma W316903, cis-3-hexenoic acid Sigma-PH018432, trans-3-hexenoic acid Sigma -W317004, and a mixture of cis-4-hexenoic acid and trans-4-hexenoic acid -Santa Cruz CAS 35194-36-6) with recombinant LvaE.

R-4-OH-C6-CoA and S-4-OH-C6-CoA were purified by preparative HPLC on a Waters systems and characterized by MS. The column used was an XBridge Prep C18 19 × 150 mm reverse-phase chromatography column (Waters). A binary solvent system was used in a 30-min linear gradient of acetonitrile in water (3–18%) containing 0.1% TFA at a flow rate of 19 ml/min. The lyophilized products were diluted to a final concentration of 10 mM in Milli-Q water and stored at − 80 °C.

Enzymatic reactions

To assess ACAD10 and ACAD11 function, 4-OH-C6-CoA and 4-OH-C10-CoA were synthesized in situ using recombinant LvaE (Pseudomonas putida acyl-CoA synthetase) or FadD (E. coli acyl-CoA synthetase), respectively. Briefly, LvaE or FadD (8.5 ng/μL) was incubated in a 30 µL reaction with substrate 2.75 mM 4-OH-C6 (Sigma 303836) or 4-OH-C10 (Sigma W236004), 0.55 mM CoA (Sigma C3144) and 2.6 U/mL inorganic pyrophosphatase (Roche 10108987001) in a buffer containing 25 mM Tris–HCl pH 7.4, 2.75 mM ATP, 2.75 mM MgCl2, 200 µM TCEP, 0.05 μg/µL bovine serum albumin (BSA) and 300 µM FAD. Purified wild-type and mutant ACAD10 were used at a concentration of 50 nM (Fig. 3A, B) or 25 nM (Fig. 5C, D). The bacterial orthologs from Burkholderia pseudomallei were used at 50 nM (Fig. 3B). After 1 h incubation at 37 °C, 10 µL aliquots of each reaction were collected and quenched on ice in tubes containing 40 µL methanol. After 2 rounds of vortexing, the tubes were centrifuged at 15,000 g for 15 min at 4 °C. The supernatant was directly analyzed by LC–MS.

Activity on 4-OH-C6-CoA (Fig. 4A–C) was assessed in 30 µL reactions with 50 µM purified R-4-OH-C6-CoA, 25 nM (1.5 ng/µL) enzyme, 25 mM Tris–HCl pH 7.4, 2.75 mM ATP, 2.75 mM MgCl2, 200 µM TCEP, 0.05 μg/μL BSA. After 1 h incubation at 37 °C, 10 µL aliquots of each reaction were collected and quenched on ice in tubes containing 40 uL methanol. The tubes were centrifuged at 15,000 g for 15 min at 4 °C. The supernatant was directly analyzed by LC–MS.

To compare the production of 4-P-OH-C6-CoA by N-terminal WT and HAD mutant ACAD10 in a time course experiment (Fig. 4B), we incubated 25 µM purified R-4-OH-C6-CoA with 3.5 nM enzyme in 30 µL reactions. After 20 min incubation at 37 °C, 10 µL aliquots of each reaction were collected at the indicated timepoints, and quenched on ice in tubes containing 40 µL methanol. Tubes were centrifuged at 15,000 g for 15 min at 4 °C, and the supernatant was directly analyzed by LC–MS.

To determine the activity of the ACAD10 kinase domain at different substrate concentrations, we assessed the production of 4-P-OH-C6-CoA in 30 µL reactions containing increasing concentrations of purified R-4-OH-C6-CoA (0-10-25-50-100-200 µM) and 25 nM enzyme for 60 min at 37 °C. The experimental conditions and the quenching method were otherwise as described above.

To investigate deuterium incorporation from D2O into the intermediates of the ACAD10 reaction, we incubated 50 µM R-4-OH-C6-CoA with 25 nM enzyme, in the presence of 50% D2O. Otherwise, reaction conditions, quenching, and analysis were as described above. The abundance of the M + 1 version of 4-P-OH-hexanoyl-CoA resulting from deuterium incorporation was determined after correction for the natural isotopic distribution of carbon using the Vistaflux package within the Profinder software (Agilent).

Statistical analysis

Statistical analyses were performed in GraphPad Prism 10. Unless stated otherwise, data present the means ± SEM of three experiments containing each 2 to 3 biological replicates. Data were analysed by one-way or two-way ANOVA followed by post-hoc testing using the Dunnet or Holm-Sidak correction [72, 73].