Insulin-regulated serine and lipid metabolism drive peripheral neuropathy – Nature

Mouse experiments

All mouse experiments were approved and conducted in accordance with the Institutional Animal Care and Use Committee (IACUC) of the University of California, San Diego and the Salk Institute for Biological Studies. Mice were housed in the same room ensuring exposure to the same temperature (21 °C), humidity (ambient humidity 65%) and a 12-h light:dark cycle (06:00–18:00). In Fig. 1, 14- to 16-week-old C57BL/6J (JAX 000664) or BKS-db/db mice (JAX 000642), and 10- to 12-week-old vehicle- or STZ-treated C57BL/6J (JAX 000664) mice were fasted for 6 h prior to tissue collection. Animals were anaesthetized with isoflurane, decapitated, and tissues were rapidly collected using Wollenberger clamps pre-cooled to the temperature of liquid nitrogen and stored at −80 °C until analysis. For Fig. 2, 8-week-old C57BL/6J (JAX 000664) were fed with diets obtained from Envigo. Dietary composition is detailed in Supplementary Table 1. In dietary experiments, tissues were collected between 07:00–10:00 h—that is, in the fed state unless stated otherwise. From a separate cohort of animals, plasma samples were collected 18 weeks after dietary intervention in the fed (07:00–10:00 h) and fasted (18-h overnight fast) state. In db/db mice experiments, tissues were collected after 6-h fasting. For Fig. 3, 8-week-old C57BL/6J (JAX 000664) were fed with diets obtained from Envigo. Tissue collection took place between 07:00–10:00 h. Animals were anaesthetized with isoflurane, decapitated, and tissues were rapidly collected using Wollenberger clamps pre-cooled to the temperature of liquid nitrogen and stored at −80 °C until analysis. For Fig. 4, 6-week-old BKS-db/db mice (JAX 000642) were fed with either a control or serine-supplemented diet (provided by Envigo) for a period of 8 weeks. Tissue collection took place between 07:00–10:00 h unless stated otherwise. Animals were anaesthetized with isoflurane, decapitated, and tissues rapidly collected using Wollenberger clamps pre-cooled to the temperature of liquid nitrogen and stored at −80 °C until analysis.

Serine tolerance test

Age-matched 14- to 16-week-old wild-type and BKS-db/db, and 10- to 12-week-old vehicle- and STZ-treated C57BL/6J (JAX 000664) mice were fasted overnight with water access provided ad libitum. For a STT, animals were weighed, and serine and/or glucose were administered via oral gavage at a dose of 400 mg kg⁻¹ and 2 g kg⁻¹, respectively, with tail tip blood samples collected into EDTA-coated microvette tubes (Sarstedt) before, and 15, 30, 60, 120 and 180 min after an oral gavage. EDTA microvettes were spun at 2,000g at 4 °C for 5 min to obtain plasma, and samples stored at −80 °C until analysis. Blood glucose and serine concentrations were quantified using Contour Next glucometer (Bayer) and gas chromatography–mass spectrometry as described below, respectively. Plasma serine pharmacokinetics were determined for a 400 mg kg⁻¹ dose using PK solver⁵⁰.

To qualify downstream fate of serine, wild-type mice were fasted overnight, weighed in the morning, and [U-¹³C₃]serine administered via oral gavage at a dose of 400 mg kg⁻¹, with tissues collected, using Wollenberger clamps pre-cooled to the temperature of liquid nitrogen, before, and 15, 30, 45, 60, and 120 min after oral gavage, and samples stored at −80 °C until analysis.

Serum insulin and glucagon measurements

Commercially available kits were used to determine serum insulin (Mouse Insulin ELISA 10-1247-01, Mercodia) and glucagon (Glucagon ELISA 10-1271-01, Mercodia) following a 6-h fast in mice according to the manufacturer’s instructions.

Lipogenesis D₂O experiments

C57BL/6J mice fed diets for 18 weeks were injected intraperitoneally with D₂O (in 0.9% NaCl) at a dose of 0.027 ml per g of body weight with drinking water replaced with 6% D₂O-enriched solution for a period of ~18 h. In the morning (07:00–10:00 h) tissues were rapidly collected using Wollenberger clamps pre-cooled to the temperature of liquid nitrogen and stored at −80 °C until analysis.

Plasma D₂O enrichment was determined using deuterium–acetone exchange protocol as previously described²⁴. In brief, 5 µl of plasma were incubated with 4 µl of 5% acetone in acetonitrile solution and 4 µl of 10 M NaOH for 24 h. Next, 500 mg of Na₂SO₄ and 600 µl of chloroform were added, and samples vortex-mixed. After 2 min centrifugation at 3,000g, 80 µl was transferred in triplicate into gas chromatography–mass spectrometry (GC–MS) vials, and plasma D₂O enrichment was quantified from an external standard curve on an Agilent DB-35MS column (30 m by 0.25 mm internal diameter × 0.25 μm, Agilent J&W Scientific) installed in an Agilent 7890 A gas chromatograph (GC) interfaced with an Agilent 5975 C mass spectrometer with the following temperature program: 60 °C initial, increase by 20 °C min⁻¹ to 100 °C, increase by 50 °C min⁻¹ to 220 °C, and hold for 1 min.

To quantify tissue D₂O labelling, ~20 mg of frozen tissue was homogenized with 250 µl −20 °C methanol, 250 µl ice-cold saline and 500 µl −20 °C chloroform spiked with internal standards palmitate-d₃₁ (Cambridge Isotope Laboratories, DLM-215-PK) and coprostanol (Sigma, 7578). After a 5 min spin at 4 °C at 21,000g, the chloroform fraction was collected, dried, and resuspended with 500 µl of 2% H₂SO₄ in methanol for 2 h at 50 °C. Next, 100 µl of saturated NaCl and 500 µl of hexane were added, samples vortex-mixed, and upper hexane phase collected and transferred into a GC–MS vial. Fatty acid methyl esters were analysed using a Select FAME column (100 m × 0.25 mm internal diameter) installed in an Agilent 7890 A GC interfaced with an Agilent 5975 C MS using the following temperature program: 80 °C initial, increase by 20 °C min⁻¹ to 170 °C, increase by 1 °C min⁻¹ to 204 °C, then 20 °C min⁻¹ to 250 °C and hold for 10 min. The percent isotopologue distribution of each fatty acid and polar metabolite was determined and corrected for natural abundance using in-house algorithms adapted from a previous report⁵¹.

GTT and ITT

For GTT and ITT, C57BL/6J mice fed the diets for 18 weeks were fasted overnight with water provided ad libitum. In the morning animals were weighed and fasting blood glucose was determined from a tail bleed using a Contour Next glucometer (Bayer). For GTT, the animals were injected intraperitoneally with a bolus of glucose at a dose of 2 g kg⁻¹ of body weight, and blood glucose determined at 15, 30, 60, 120 and 180 min post-injection. For ITT, the animals were injected intraperitoneally with a bolus of insulin (100 IU ml⁻¹ Humulin Insulin, Eli Lilly) at a dose of 0.5 IU kg⁻¹, and blood glucose was quantified at 15, 30, 60 and 90 min post-injection as previously described⁵².

Determination of body composition and systemic metabolic rates

Lean and fat masses were determined using a EchoMRI 3-in-1 instrument (quantitative nuclear magnetic resonance (qNMR) imaging system). Comprehensive Laboratory Animal Monitoring System (CLAMS) (Oxymax, Columbus Instruments) was used to quantify systemic metabolic rates in individually housed mice during a period of 6 days. Water, food, and calorie intake were calculated from individually housed animals over a period of 6 days when subjected to CLAMS. Whole-body oxygen consumption (VO₂) and carbon dioxide (VCO₂) rates were normalized to corresponding total body weights, and RER was calculated as ratio of VCO₂ to VO₂.

Faecal bomb calorimetry and calorie absorption

Approximately 1 g faeces was desiccated overnight and ground using a mortar and pestle. Powdered sample was reconstituted into a pellet with 300 µl ddH₂O and weighed. Pellet was placed in bomb cylinder surrounded by 2,000 ml ddH₂O (Parr 6100 Compensated Jacket Calorimeter). Heat produced by combustion was measured by change in water temperature. The calorimeter energy equivalent, W (Cal °C⁻¹), was calculated with standardized benzoic acid. The final energy content of each pellet was calculated as follows:

Calorie absorption was calculated by subtracting gross energy (faecal calorie extraction) from calorie intake.

Microbiome analysis

DNA was extracted from 10–30 mg of stool using the MoBio PowerFecal DNA isolation kit (12830-50). Extracted DNA was quantified using a Nanodrop (ThermoFisher Scientific). The whole-genome sequencing raw data was uploaded to Qiita⁵³, where we followed their default processing workflow. In summary, the raw reads were adapter filtered using the auto-detect parameters in fastp version 20⁵⁴ and host (mouse) filtered using minimap2 version 2.17⁵⁵. The resulting sequences were aligned using Bowtie 2 version 2.4.2⁵⁶ to the Web of Life (WoL) reference database⁵⁷ via the Web of Life Toolkit App (https://github.com/qiyunzhu/woltka); this step generated tables at genus, species, per genome, and per gene tables. For all analyses we used the per genome table; then for alpha diversity we removed any samples below 1,273,062 sequences per sample and for beta-diversity analysis we rarefied at the same value. Downstream analyses were performed in QIIME 2 version 2020.11⁵⁸. To asses global microbiota alterations, alpha diversity analysis was performed through Faith’s PD⁵⁹ and beta diversity through robust principal component analysis (RPCA)⁶⁰ and resulting Aitchison distances were evaluated through permutational multivariate analysis of variance (PERMANOVA)⁶¹.

We then designed a Bayesian hierarchical model for differential abundance incorporating diet type as a fixed effect and cage as a random effect. We model the count generating process as a negative binomial distribution to account for overdispersion. Due to the sparsity of microbiome data, we also accounted for zero-inflation by assigning each microbe a probability of being unobserved separately from the count generating process:

We wrote this model using the Stan probabilistic programming language⁶² and fit the model using BIRDMAn (https://github.com/gibsramen/BIRDMAn). To account for compositionality, we fit this model using the first microbe in the table as an additive log ratio reference and converted log fold changes into centred log ratio coordinates after fitting. We used the following as prior distributions for the target parameters:

in which i is the sample, j is the feature, y is the microbial count, θ is the indicator for non-biological zero, η is the mean feature count, x is the covariate, β is the regression coefficients to be estimated (log-fold changes), π is the probability of non-biological zero, z is the cage identifier variable, u is the random effect of cage, and ϕ is the overdispersion parameter. In order to compare functional changes associated with strain level differential abundances a comparative genomics pathway completeness approach was taken. First, each genome was assessed via MetaCyc⁶³ pathway completeness, a proportion ranging from zero to one, by mapping characterized genes to reactions and finally to pathways. Each pathway was then correlated by Spearman’s rank correlation to the beta differential abundance determined from the above model. Serine biosynthesis, glycine cleavage, and fatty acid synthesis pathways were significantly correlated to betas. To validate these correlations, the log ratio of the sum of the abundance of genomes with complete pathways (completeness = 1) vs. those without (completeness < 1) were evaluated between treatment groups.

Behavioural assays

Thermal sensation

Small sensory C fibre function was quantified by behavioural responses to heat using a thermal nociception test device (UARD) as previously described⁶⁴. In brief, the apparatus surface was warmed up to 30 °C, and animals were placed in individual testing chambers for 20–30 min prior to testing. Four separate response latency measurements were performed, and the mean of the last triplicate taken to represent response latency for each animal. All measurements were made on coded animals by an observer unaware of the treatment groups.

Tactile sensation

Animals were placed on the von Frey stand and allowed to acclimate for 20–30 min. The range of manual von Frey filaments was used: 2.44, 2.83, 3.22, 3.61, 3.84, 4.08, 4.31, 4.56, 4.74 (Kom Kare). Testing began with the 3.84 filament and the pressure applied was repeated five times. If a positive response was observed, the next lower weighted filament was used in the sequence. In the case of a negative response, the next higher weighted filament was applied. All measurements were made on coded animals by an observer unaware of the treatment groups.

Nerve conduction velocity

Conduction of a motor nerve was quantified in anaesthetized mice using EZ Anesthesia Versaflex system (Braintree Scientific, Z-7150). In brief, lightly anaesthetized mice were transferred onto a water-heated pad with anaesthesia maintained via a face mask. Two recording platinum electrodes were inserted between the animal’s second, third, and fourth toes, and a grounding electrode into the skin at the neck. PowerLab stimulator delivered a 200-mV, 50-µs square-wave stimulus every 2 s. The stimulating electrode was inserted in the ankle near the Achilles tendon and subsequently into the sciatic notch at the hip, and M waves were recorded. The latency between Achilles tendon and sciatic notch was used to calculate nerve conduction velocity as described⁶⁴. All measurements were made on coded animals by an observer unaware of the treatment groups.

Corneal confocal imaging

Quantification of corneal nerves was performed in anaesthetized mice (using EZ Anesthesia Versaflex system, Braintree Scientific, Z-7150) using Retina Tomograph 3 with Rostock Cornea Module (Heidelberg Engineering) equipped with Tomocap (Heidelberg Engineering, 0220-001) as previously described⁶⁴. In brief, lightly anaesthetized mice were transferred onto a small animal platform with anaesthesia maintained via a face mask. Forty sequential images of uniform magnification and size were collected and those containing nerves of the sub-basal plexus identified. ImageJ software (ImageJ 1.53e Java 1.8.0_172) was used to quantify corneal nerve area within each image, with data presented as pixels/image. All measurements were made on coded animals and images by an observer unaware of the treatment groups.

Epidermal innervation

Quantification of epidermal innervation was performed in paw skin samples by immunostaining for the pan-neuronal protein PGP9.5, as described previously in detail⁶⁴. In brief, paw skin samples were collected into 4% buffered paraformaldehyde (Thermo scientific, J19943-K2). Staining of epidermal nerves was performed using anti-PGP9.5 antibody (ProteinTech, 14730-1-AP; 1:500 dilution). Using 40× magnification of a light microscope, the number of PGP9.5-positive profiles present in the epidermis was calculated, length of skin section calculated, and IENF profile density expressed as profiles mm⁻¹. All measurements were made on coded slides by an observer unaware of the treatment groups.

Metabolite extraction and quantification

Plasma polar metabolites were extracted from 3 µl of plasma spiked with a known amount of ¹³C- and ¹⁵N-labelled standards (Cambridge Isotope Laboratories, MSK-A2-1.2). Tissue metabolite extraction was performed as described before²¹. In brief, ~20 mg of tissue was homogenized for 2 min using Precellys beads with 500 µl −20 °C methanol, 400 µl ice-cold saline and 100 µl ice-cold water and spiked with ¹³C/¹⁵N polar metabolite standards (Cambridge Isotope Laboratories, MSK-A2-1.2), 20 pmol of sphinganine-d7 (Avanti Polar Lipids, 860658), 2 pmol of deoxysphinganine-d3 (Avanti Polar Lipids, 860474), 100 pmol of ¹³C-dihydroceramide-d7 (Avanti Polar Lipids, 330726), 200 pmol of C₁₅-ceramide-d7 (Avanti Polar Lipids, 860681), 10 pmol of d18:1-d7 glucosylsphingosine (Avanti Polar Lipids, 860695), 100 pmol of d18:1-d7/15:0 glucosylceramide (Avanti Polar Lipids, 330729), 100 pmol of d18:1-d7/15:0 lactosylceramide (Avanti Polar Lipids, 330727), 200 pmol of sphingosine-d7 (Avanti Polar Lipids, 860657), and 200 pmol of d18:1/18:1-d9 sphingomyelin (Avanti Polar Lipids, 791649). The identification of 1-deoxydihydroceramides was confirmed via retention time matching and analysis of m18:0/24:1 deoxyDHCer (Avanti Polar Lipids, 860464) and m18:0/16:0 deoxyDHCer (Avanti Polar Lipids, 860462) standards, and normalization for 1-deoxydihydroceramides was done with the 13C-dihydroceramide-d7 standard. Homogenate aliquot of 50 µl was taken to determine tissue protein content using BCA protein assay (Lambda Biotech, G1002). The remaining homogenate was transferred to a 2 ml Eppendorf tube and 1 ml of −20 °C chloroform was added. Samples were vortex-mixed for 5 min and spun down for 5 min at 4 °C at 15,000g. The organic phase was collected and 2 μl of formic acid was added to the remaining polar phase which was re-extracted with 1 ml of −20 °C chloroform. Combined organic phases were dried and the pellet was resuspended in 100 μl of buffer containing 100% methanol, 1 mM ammonium formate and 0.2% formic acid. Data represents ion counts normalized by class-specific internal standards and tissue protein content, with stacked plots to represent acyl-chain distribution.

Gas chromatography–mass spectrometry

Quantification of polar metabolites was determined after derivatization with 2% (w/v) methoxyamine hydrochloride (Thermo Scientific) in pyridine (37 °C for 60 min) and with N-tertbutyldimethylsilyl-N-methyltrifluoroacetamide (MTBSTFA) with 1% tert-butyldimethylchlorosilane (tBDMS) (Regis Technologies) (37 °C for 30 min). Polar derivatives were analysed by GC–MS using a DB-35MS column (30 m × 0.25 mm internal diameter × 0.25 μm, Agilent J&W Scientific) installed in an Agilent 7890 A gas chromatograph interfaced with an Agilent 5975 C mass spectrometer as previously described⁶⁵. Plasma glucose enrichment was determined using propionic anhydride derivatization as previously described⁶⁶. Natural isotope abundance was corrected using in-house script⁵¹.

Targeted sphingolipid quantification

Quantification of sphingolipid metabolites was determined using triple quadrupole liquid chromatography–mass spectrometry platform (Agilent 6460). Sphingolipid species were separated on a C8 column (Spectra 3 μm C8SR 150 × 3 mm inner diameter, Peeke Scientific) as previously described⁶⁷. Mobile phase A was composed of 100% HPLC-grade water containing 2 mM ammonium formate and 0.2% formic acid and mobile phase B consisted of 100% methanol containing 0.2% formic acid and 1 mM ammonium formate. The flow rate was 0.5 ml min⁻¹. The gradient elution programme consisted of the following profile: 0 min, 82% B; 3 min, 82% B; 4 min, 90% B, 18 min, 99% B; 25 min, 99%, 27 min, 82% B, 30 min, 82% B. Column re-equilibration followed each sample and lasted 10 min. The capillary voltage was set to 3.5 kV, the drying gas temperature was 350 °C, the drying gas flow rate was 10 l min⁻¹, and the nebulizer pressure was 60 psi. Sphingolipid species were analysed by selective reaction monitoring (SRM) of the transition from precursor to product ions at associated optimized collision energies and fragmentor voltages (Supplementary Table 2). Quantification of sphingolipid species was performed using spiked-in deuterated standards.

High-resolution LC–MS/MS of polar metabolites

Around 10–20 mg of frozen tissue was extracted with 800 µl of −20 °C 5:3:2 acetonitrile: methanol:water solution spike with a known concentration of norvaline as an internal standard using the Precellys Evolution Homogenizer (Bertin Technologies)⁶⁸. After extraction, a 50-µl aliquot was taken for protein quantification using BCA protein assay (Lambda Biotech, G1002), and the remaining extract was spun for 10 min at 21,000g at 4 °C. The supernatant was then transferred into a glass vial, and chromatographic separation and compound identification performed using Q Exactive Orbitrap MS with a Vanquish Flex Binary UHPLC system (ThermoFisher Scientific) on an iHILIC-(P) Classic, 150 mm by 2.1 mm, 5-mm particle, 200-Å (Hilicon) column at 45 °C. Chromatography was performed using a gradient of 20 mM ammonium carbonate, adjusted to pH 9.4 with 0.1% ammonium hydroxide (25%) solution (mobile phase A) and 100% acetonitrile (mobile phase B), both at a flow rate of 0.2 ml min⁻¹. The liquid chromatography gradient ran linearly from 80 to 20% B from 2 to 17 min and then from 20 to 80% B from 17 to 18 min and then held at 80% B from 18 to 25 min.

High-resolution LC–MS/MS of lipids

Liver samples were extracted in 400 µl of −20 °C methanol using the Precellys Evolution Homogenizer (Bertin Technologies) spiked with EquiSPLASH labelled standard (Avanti Polar Lipids, 330731) and norvaline. After extraction, 50 µl aliquot was taken to quantify protein content using BCA protein assay (Lambda Biotech, G1002), and to the remaining extract were added 400 µl of −20 °C chloroform and 400 µl of ice-cold water. After vortexing for 5 min, samples were spun for 5 min at 4 °C at 15,000g, and the organic phase was collected. Two microlitres of formic acid were added to the remaining polar phase which was re-extracted with 400 µl of −20 °C chloroform, samples were vortex-mixed, and spun as described above. Combined organic phases were dried and the pellet was resuspended in 100 μl of isopropanol.

Chromatographic separation and lipid species identification was performed using Q Exactive orbitrap mass spectrometer with a Vanquish Flex Binary UHPLC system (Thermo Scientific) equipped with an Accucore C30, 150 × 2.1 mm, 2.6 µm particle (Thermo) column at 40 °C. Five microlitres of sample was injected. Chromatography was performed using a gradient of 40:60 v/v water: acetonitrile with 10 mM ammonium formate and 0.1% formic acid (mobile phase A) and 10:90 v/v acetonitrile: propan-2-ol with 10 mM ammonium formate and 0.1% formic acid (mobile phase B), both at a flow rate of 0.2 ml min⁻¹. The liquid chromatography gradient ran from 30% to 43% B from 3–8 min, then from 43% to 50% B from 8-9 min, then 50–90% B from 9–18 min, then 90–99% B from 18–26 min, then held at 99% B from 26–30 min, before returning to 30% B in 6 min and held for a further 4 min.

Lipids were analysed in positive mode using spray voltage 3.2 kV. Sweep gas flow was 1 arbitrary units, auxiliary gas flow 2 arbitrary units and sheath gas flow 40 arbitrary units, with a capillary temperature of 325 °C. Full mass spectrometry (scan range 200–2,000 m/z) was used at 70,000 resolution with 10⁶ automatic gain control and a maximum injection time of 100 ms. Data dependent MS2 (Top 6) mode at 17,500 resolution with automatic gain control set at 10⁵ with a maximum injection time of 50 ms was used. Data were analysed using EI-Maven software, and peaks normalized to Avanti EquiSPLASH internal standard. Lipid species specific fragments used for identification and quantification are presented in the Supplementary Table 3.

Plasma sphingoid base extraction, hydrolysis and LC–MS analysis

Plasma sphingolipids were processed as previously described with minor modifications⁶⁹. In brief, 50 µl of plasma was mixed with 0.5 ml of methanol and spiked with internal standards, sphinganine-d7, sphingosine-d7 and deoxysphinganine-d3 (Avanti lipids). The samples were placed on a mixer for 1 h at 37 °C, centrifuged at 2,800g and the supernatant collected and acid hydrolysed overnight at 65 °C with 75 µl of methanolic HCl (1N HCl,10M H₂O in methanol). Next, 100 µl of 10 M KOH was added to neutralize. 625 µl of chloroform, 100 µl of 2N NH₄OH and 500 µl of alkaline water were added, samples vortex-mixed and centrifuged for 5 min at 16,000g. The lower organic phase was washed three times with alkaline water and dried under air. LC–MS analysis was performed on an Agilent 6460 QQQ LC–MS/MS. Metabolite separation was achieved with a C18 column (Luna 100 × 2.1 mm, 3 µm, Phenomenex). Mobile phase A was composed of a 60:40 ratio of methanol:water and mobile phase B consisted of 100% methanol with 0.1% formic acid with 5 mM ammonium formate added to both mobile phases. The gradient elution programme consisted of holding at 40% B for 0.5 min, linearly increasing to 100% B over 15 min, and maintaining it for 9 min, followed by re-equilibration to the initial condition for 10 min. The capillary voltage was set to 3.5 kV, the drying gas temperature was 350 °C, the drying gas flow rate was 10 l min⁻¹, and the nebulizer pressure was 60 psi. Sphingoid bases were analysed by SRM of the transition from precursor to product ions at associated optimized collision energies and fragmentor voltages¹⁶. Sphingoid bases were then quantified from spiked internal standards of known concentration.

Serine dehydratase activity assay

Frozen liver and kidney samples were extracted in an ice-cold buffer containing 50 mM KH₂PO₄, 1 mM Na₂EDTA, and 1mM DTT, pH 8.0 using glass homogenizer. Maximal enzyme activity was determined using coupled-enzyme reaction with lactate dehydrogenase (Sigma 10127230001) in the presence of 200 mM serine, 0.25 mM NADH, 0.17 mM pyridoxal phosphate, and 1 mM DTT for 3 min. Tissue homogenate protein quantification was subsequently determined using BCA protein assay (Lambda Biotech, G1002), and maximal enzyme activity expressed in international units (U) per mg of protein.

Gene expression analysis

RNA was extracted from ~20 mg of liver tissue using Direct-Zol RNA kit (Direct-Zol RNA Miniprep Plus kit, Zymo Research) according to the manufacturer’s instructions. cDNA synthesis was performed using iScript Reverse Transcription Supermix for RT–PCR (iScript Reverse Transcription Supermix, Bio-Rad) according to the manufacturer’s instructions using the following protocol: 5 min at 25 °C, 20 min 46 °C, 1 min 95 °C. PCR reactions were carried out using 96-well plates on an Applied Biosystems ViiA 7 Real-Time PCR System using the following parameters: 95 °C for 20 s, 40 cycles of 95 °C for 1 s, and 60 °C for 20 s. The final volume (10 µl) of PCR SYBR-Green reaction consisted of 5 µl fast SYBR-Green Master Mix (Applied Biosystems), 2 µl cDNA, 1 µl of 5 µM forward and reverse primers, and 1 µl of water.

Primers used are as follows. 18s forward: AGTCCCTGCCCTTTGTACACA, 18s reverse: CGATCCGAGGGCCTCACTA; Acc1 forward: AATGAACGTGCAATCCGATTTG, Acc1 reverse: ACTCCACATTTGCGTAATTGTTG; Acc2 forward: CGCTCACCAACAGTAAGGTGG, Acc2 reverse: GCTTGGCAGGGAGTTCCTC; Acly forward: AATCCTGGCTAAAACCTCGCC, Acly reverse: GCATAGATGCACACGTAGAACT; actin forward: GGCTGTATTCCCTCCATCG, actin reverse: CCAGTTGGTAACAATGCCATGT; Aldh1l1 forward: AGCCACCTATGAGGGCATTC, Aldh1l1 reverse: TGAGTGTCGAGTTGAAAAACGTC; Aldh1l2 forward: ACCAGCCGGGTTTATTTCAAA, Aldh1l2 reverse: ACTCCCACTACTCGGTGGC; Dgat1 forward: CTGATCCTGAGTAATGCAAGGTT, Dgat1 reverse: TGGATGCAATAATCACGCATGG; Dgat2 forward: GCGCTACTTCCGAGACTACTT, Dgat2 reverse: GGGCCTTATGCCAGGAAACT; Dhcr7 forward: AGGCTGGATCTCAAGGACAAT, Dhcr7 reverse: GCCAGACTAGCATGGCCTG; Dhcr24 forward: CTCTGGGTGCGAGTGAAGG, Dhcr24 reverse: TTCCCGGACCTGTTTCTGGAT; Dld forward: AGCTGGAGTCGTGTGTACC, Dld reverse: GAACCTATCACTGTCACGTCA; Fasn forward: GGAGGTGGTGATAGCCGGTAT, Fasn reverse: TGGGTAATCCATAGAGCCCAG; Fdft1 forward: GTTTGAAGACCCCATAGTTGGTG, Fdft1 reverse: CACATCTACGTTCTCTGGCTTAG; Fdps forward: GGAGGTCCTAGAGTACAATGCC, Fdps reverse: AAGCCTGGAGCAGTTCTACAC; Ggps1 forward: TTCACAGGCATTTAATCACTGGC, Ggps1 reverse: ACCACGTCGGAGCTTTGAAC; Gldc forward: CTCCTGCCCAGACACGAT, Gldc reverse: GGGACCGTCTTCTCGATGAG; Gpat1 forward: CTTGGCCGATGTAAACACACC, Gpat1 reverse: CTTCCGGCTCATAAGGCTCTC; Gpat4 forward: TCAAAGAAATTCGTCGAAGTGGT, Gpat4 reverse: CCTTTCCGGCAAAAGTAGAAGAT; Hmgcs1 forward: AACTGGTGCAGAAATCTCTAGC, Hmgcs1 reverse: GGTTGAATAGCTCAGAACTAGCC; Hmgcr forward: AGCTTGCCCGAATTGTATGTG, Hmgcr reverse: TCTGTTGTGAACCATGTGACTTC; Lss forward: TCGTGGGGGACCCTATAAAAC, Lss reverse: CGTCCTCCGCTTGATAATAAGTC; Mthfd1 forward: CTCCTGTCCCAAGTGACATTG, Mthfd1 reverse: TAGCCTTCGTTTCCCCGTACA; Mthfd2 forward: AGTGCGAAATGAAGCCGTTG, Mthfd2 reverse: GACTGGCGGGATTGTCACC; Mthfd1l forward: GCATGGCCTTACCCTTCAGAT, Mthfd1l reverse: GTACGAGCTTCCCCAGATTGA; Mthfd2l forward: AAGGACGTTGATGGATTTCACAT, Mthfd2l reverse: GATGATTTCCCAAACGGCACT; Mthfr forward: AGATGAGGCGCAGAATGGAC, Mthfr reverse: CATCCGGTCAAACCTGGAGAT; Mtr forward: TCCTCCTCGGCCTATCTTTATTT, Mtr reverse: GGTCCGAATGAGACACGCT; Mvk forward: GGTGTGGTCGGAACTTCCC, Mvk reverse: CCTTGAGCGGGTTGGAGAC; Mvd forward: ATGGCCTCAGAAAAGCCTCAG, Mvd reverse: TGGTCGTTTTTAGCTGGTCCT; Pmvk forward: CCTATGGGGCTGTGATACAGA, Pmvk reverse: TCTCCGTGGTTCTCAATGACC; Psat forward: CAGTGGAGCGCCAGAATAGAA, Psat reverse: CCTGTGCCCCTTCAAGGA; Psph forward: TGAGTACGCAGGTTTTGATGAG, Psph reverse: TGAGTACGCAGGTTTTGATGAG; Phgdh forward: ATGGCCTTCGCAAATCTGC, Phgdh reverse: AGTTCAGCTATCAGCTCCTCC; Scd1 forward: TTCTTGCGATACACTCTGGTGC, Scd2 reverse: CGGGATTGAATGTTCTTGTCGT; Scd2 forward: GATCTCTGGCGCTTACTCAGC, Scd2 reverse: CTCCCCAGTGGTGAGAACTC; Sds forward: GAAGACCCCACTTCGTGACAG, Sds reverse: TCTTGCAGAGATGCCCAATGC; Shmt1 forward: CAGGGCTCTGCTTGATGCAC, Shmt1 reverse: CGTAACGCGCTCTTGTCAC; Shmt2 forward: TGGCAAGAGATACTACGGAGG, Shmt2 reverse: AGATCCGCTTGACATCAGACA; Sqle forward: ATAAGAAATGCGGGGATGTCAC, Sqle reverse: ATATCCGAGAAGGCAGCGAAC; Srebp1a forward: TAGTCCGAAGCCGGGTGGGCGCCGG, Srebp1a reverse: GATGTCGTTCAAAACCGCTGTGTGTC; Srebp1c forward: AAGCAAATCACTGAAGGACCTGG, Srebp1c reverse: AAAGACAAGCTACTCTGGGAG; Srebp2 forward: GGATCCTCCCAAAGAAGGAG, Srebp2 reverse: TTCCTCAGAACGCCAGACTT; Tyms forward: GGAAGGGTGTTTTGGAGGAGT, Tyms reverse: GCTGTCCAGAAAATCTCGGGA.

Western blotting

To compare tissue protein levels, ~20 mg of tissue was homogenized in RIPA buffer supplemented with 5 mM EDTA solution (Thermo Scientific), protease inhibitor cocktail (cOmplete, Roche) and phosphatase inhibitor cocktail (PhosSTOP, Roche), and placed on ice for 30 min. Tissue lysate was centrifuged at 13,000g for 10 min at 4 °C, and the supernatant was stored at −80 °C. Homogenate protein content was determined using the bicinchoninic acid assay (Thermo Scientific). Protein samples were prepared in a Laemmli buffer (NuPAGE LDS Sample Buffer, Life Technologies), and were run a 4–15% precast gel (Mini-PROTEAN TGX, Bio-Rad) for 2 h at constant 100 V and transferred on a polyvinylidenedifluoride (PVDF) membrane for 2 h at constant 250 mA in an ice-chilled transfer tank. The membrane was blocked with 5% milk and incubated overnight at 4 °C with a primary antibody against ACLY (Cell Signaling 13390, 1:1,000), ACC (Cell Signaling 3662, 1:2,000), p-AKT Ser473 (Cell Signaling 9271, 1:1,000), p-AKT Ser308 (Cell Signaling 9275, 1:1,000), AKT (Cell Signaling 75692, 1:1,000), SCD1 (Cell Signaling 2794, 1:1,000), GAPDH (Cell Signaling 5174, 1:4,000), and vinculin (Cell Signaling 4650, 1:1,000). After washing with TBS-T, membranes were incubated with a horseradish peroxidase-conjugated secondary antibody (Cell Signaling 7074, 1:5,000) for 1 h at room temperature and incubated with enhanced chemiluminescence liquid (Clarity Western ECL Substrate, Bio-Rad) for 5 min. Densitometry quantification was performed using Image Lab software (Bio-Rad). For raw scans with molecular weight markers, see Supplementary Fig. 1.

Statistical analysis

Data are expressed as mean ± s.e.m. unless stated otherwise. Statistical analysis was performed with Prism software (GraphPad Prism 9.3.1) using two-sided independent t-test to compare two groups, one-way ANOVA with Fisher’s least significant difference post hoc test to compare more than two groups, two-way ANOVA with Fisher’s least significant difference post hoc test to compare two-factor study design, and PERMANOVA analysis to explore RPCA plots. For all tests, P < 0.05 was considered significant. All data points in the manuscript represent individual biological replicates. No statistical methods were used to predetermine sample size.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.