Animals

Institute of Cancer Research (ICR) mice aged 8 weeks were purchased from Shanghai SLAC Laboratory Animal Co. (Shanghai, China). The method for establishing the GDM model was modified from previous work [12]. Virgin female mice were mated with male mice. Once the vaginal plug was observed, the day was considered gestational day 0.5 (GD0.5) for the mother and embryonic day 0.5 (ED0.5) for the foetus. Time-pregnant mice were assigned to the control (CTR) or GDM group. Each GDM female was treated with 100 mg/kg streptozotocin (Sigma‒Aldrich, St. Louis, MO, USA; S0130) dissolved in citrate buffer (Beijing Solarbio Science & Technology Co., Ltd., China; C1013) at GD6.5 and GD12.5 after overnight fasting. CTR mice received an equal volume of buffer. Blood glucose was detected daily from GD13.5 via the tail vein. GDM was defined only when the random glucose level was higher than 16.8 mmol/L. The offspring of GDM were breastfed by normal female mice for 3 weeks.

Physiological measurements and metabolic testing

After weaning, the weights of the offspring were recorded weekly. Intraperitoneal glucose (2 g/kg body weight) tolerance tests (GTTs) and insulin (0.8 U/kg body weight) resistance tests (ITTs) were performed after 16-h and 6-h fasts, respectively. In vivo oxygen consumption and carbon dioxide emission were measured over 24 h in a TSE LabMaster System (TSE Systems, Bad Homburg, Germany) when the animals had access to food and water. The body composition of the conscious mice was detected through a nuclear magnetic resonance instrument (Niumag Corporation, China, QMR06-090H). Gripping force was evaluated by a grip strength metre (Xin Run Corporation, China, XR501): the mouse was held by the tail and placed on the net. Then, we pulled the mice gently back by the tail so that their forepaws grabbed the net, and the grip in Newtons (N) was recorded until the net was released. There was no acclimation period for the test. The process was repeated three times, the highest force was recorded and we calculated the average. The exercise endurance capacity (EEC) test was performed as follows: Mice were acclimated to the device for 4 consecutive days for 10 min at a speed of 15 meter per minute (m/min). The mice were exercised at the start of the dark cycle, such as at 5:00 pm. The speeds were adjusted using following protocol:15 m/min for 45 min, 18 m/min for 15 min, 20 m/min for 15 min, 22 m/min for 15 min and 25 m/min for up to 60 min. Mice were constantly monitored and were motivated to exercise via an electric stimulus of 0.5 mA. Exhaustion was considered after 6 s of permanence on the electric grid, and the mouse did not run anymore even if it was tapped. Maximum exercise capacity was estimated using parameters: the duration of the run (min) and the distance (m).

Immunohistochemical and immunofluorescence staining

The frozen skeletal muscle was cut into 8 millimetre (mm) thick cross sections without fixation. The Tetrazole Salt Method was used according to the kit protocol (Solarbio Science & Technology Co., Ltd., China, G2000) to stain for succinate dehydrogenase (SDH) inside the muscle. 8 mm-thick-sections were fixed with 4% paraformaldehyde, and antigen retrieval was performed accordingly (Beyotime Biotech. Inc., China, P0090). After they were incubated in 0.05% Triton X-100 (Thermo Fisher Scientific, USA, BP151-500) and washed with PBS, the sections were incubated with primary antibodies overnight at 4 °C followed by secondary antibodies (Supplementary Table 1) for 1 h at room temperature. Images were observed and captured with an Olympus IX83-FV3000-OSR. More than six sections were randomly selected to estimate the myofiber cross-sectional area (CSA) and distribution using image analysis software (Image-Pro Plus, Media Cybernetics, Silver Spring, MD, USA).

Transmission electron microscopy (TEM)

Skeletal muscle was minced into small pieces and fixed in 2.5% glutaraldehyde at 4 °C overnight. After being fixed in 1% osmium tetroxide for 1 h and stained with uranyl acetate for half an hour, the muscle minces were gradient dehydrated. Then, the muscle was embedded and stained with uranyl acetate and lead citrate. Images were acquired by a 120 kV cryo-transmission electron microscope (Thermo FEI, Czech, Tecnai G2 spirit).

RNA sequencing and proteomics

Foetal limb muscle was extracted after the skin and bones were removed. Total RNA was isolated and purified using TRIzol reagent (TaKaRa, Japan, 9108) following the manufacturer’s procedure. The RNA concentration and purity of each sample were quantified using a NanoDrop ND-1000 (NanoDrop, Wilmington, DE, USA). An RNA-Seq library preparation kit was used to generate cDNA. Finally, we performed 2 × 150 bp paired-end sequencing (PE150) on an Illumina NovaSeq™ 6000 (LC-Bio Technology Co., Ltd., Hangzhou, China). After removing low-quality and undetermined bases, HISAT2 software was used to map the reads to the genome (Mus musculus, Ensembl v101). The total differentially expressed genes are listed in Data Set 1. Compared to those in the CTR group, the significant differentially expressed genes (DEGs) with a fold change ≥2 or ≤0.5 and a P ≤ 0.05 were regarded as significant. The DEGs were further subjected to prediction of potential upstream regulators via Ingenuity Pathway Analysis (IPA; QIAGEN, Valencia, CA, USA). Muscle protein was prepared by SDT lysis and homogenised. After labelling using TMT reagent (Thermo Fisher Scientific), liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis was performed on a Q Exactive plus mass spectrometer coupled to an Easy nLC, and the raw files were processed using the MASCOT engine (Matrix Science, London, UK; version 2.6) embedded into Proteome Discoverer 2.2. The total differentially expressed proteins are listed in Data Set 2. Significant differentially expressed proteins (DEPs) between groups were defined on the basis of a fold change ≥1.2 or ≤0.833 with a P ≤ 0.05. Histograms and bubble diagrams were generated with OmicStudio at http://www.omicstudio.cn/tool.

Isolation of primary myoblasts from skeletal muscle

Myoblasts were isolated as previously reported [13]. Briefly, skeletal muscle was removed, washed, and minced into small pieces. Then, collagenase II (Worthington Biochemical, USA, LS004176, final concentration 400 U/ml) was added to the media for digestion in a 15 ml tube. The tube was shaken at 37 °C at most speeds for 1 h with another 5 s vortex. The tubes were spun at 1400 × g for 5 min, after which the supernatant was discarded. The pellet was resuspended in media and pipetted several times using a sterile 10 ml pipette. The resuspended mixture was collected and passed through prewet 70 and 30 μm strainers. The tubes were spun at 1400 × g for 5 min. The pellet was resuspended in DMEM (American Type Culture Collection, USA, 30-2002) and seeded onto a 6-cm dish. After 24 h, the supernatant was collected and spun at 930 × g for 5 min, after which the pellet was resuspended in DMEM supplemented with bFGF (Thermo Fisher Scientific, USA, 100-18B), after which the mixture was transferred to 10% Matrigel-coated dishes (Corning, Inc., USA, 354234).

In vivo and in vitro myoblast experiments

The oxygen consumption rate (OCR) was measured with a Seahorse XFe96 Analyser (Agilent Technologies, USA). Briefly, cells were seeded on culture microplates (Agilent Technologies, 101085-004) the day before. The following mitochondrial respiratory inhibitors were injected: oligomycin (Merck SA, Germany; 75351, final concentration in the well: 1 μM), fluorocarbon cyanide phenylhydrazone (Merck SA; C2920, final concentration in the well: 1 μM), rotenone (Merck SA; R8875, final concentration in the well: 1 μM) and antimycin A (Merck SA; A8674, final concentration in the well: 5 μM). The basal respiration, adenosine triphosphate (ATP) production, maximal respiration, proton leakage and spare capacity were calculated.

Specific small interfering RNAs (siRNAs) targeting mouse Ppargc1α (RiboBio Co., Ltd., Guangzhou, China; siB12322140811) were purchased. C2C12 myoblast at 30–40% confluency was transfected with 50 nM siRNA using Lipofectamine™ RNAiMAX (Thermo Fisher Scientific, Inc., 13778150) according to the manufacturer’s protocol.

When C2C12 cells (ATCC® CRL-1772™, RRID: CVCL_0188) were grown in DMEM (Biological Industries, Israel, 06-1055-57-1ACS) supplemented with 10% (v/v) foetal bovine serum (Biological Industries, Israel, 04-001-1ACS) to 80–90% confluence, pECMV-Ppargc1α-m-FLAG (MiaoLingBio, China, P8854) and pECMV-MCS-FLAG (MiaoLingBio, China, P0787) were obtained, and 2.5 μg plasmids were transfected with Lipofectamine 3000 (Thermo Fisher Scientific, Inc., L3000015) according to the manufacturer’s protocol. The transfected cells were harvested after 48 h.

H89 (MedChemExpress, HY-15979) and Forskolin (MedChemExpress, HY-15371) were dissolved in dimethyl sulfoxide (DMSO), diluted to 50 μM in medium and cultured for another 48 h.

C2C12 cells and myoblasts were cultured in DMEM supplemented with low glucose (HyClone, USA; SH30021.01) or high glucose (HyClone, USA; SH30243.01) when necessary. High glucose media contained D-glucose at a concentration of 4500 mg/L, and the D-glucose concentration was 1000 mg/L in low glucose media.

Real-time quantitative RT‒PCR

Total RNA was isolated from skeletal muscle of different ages and from myoblasts by using TRIzol (Takara, Otsu, Shiga, Japan, 9108). Total RNA (at most 1 mg) was reverse transcribed to cDNA with a PrimeScript RT Reagent Kit (Takara, RR036A). TB Green Premix Ex Taq (Takara, RR420A) was used for real-time quantitative PCR. The relative quantification of each mRNA was calculated using Actin or 18S as an internal reference. The primers used are listed in Supplementary Table 2.

Western blot

Referring to the phosphorylation status of the insulin pathway, every male mouse aged 8 weeks was intraperitoneally injected with 0.08 U/kg insulin after fasting for 6 h. Fifteen minutes after the key proteins in the pathway were fully activated by insulin, the mice were sacrificed, and limb muscles, specifically the quadriceps femoris (QUA), tibialis anterior (TA), gastrocnemius (GAS) and soleus (SOL) muscles, were collected. Whole limb muscle was collected when it referred to the foetal experiment. A total of 15–30 mg of skeletal muscle protein was separated using 10% polyacrylamide gels and then transferred to a polyvinylidene fluoride membrane. The membranes were incubated with the antibodies listed in Supplementary Table 1. The bands were visualised by a chemiluminescence system and was quantified by ImageJ software.

DNA methylation

Genomic DNA was collected from foetal limb muscle using the TIANamp Genomic DNA Kit (TIANGEN Biotech Co., Ltd., Beijing, China, DP304). Bisulfite was converted using the EpiTect Fast DNA Bisulfite Kit (Qiagen, 59802). The primers were designed with Qiagen PyroMark Assay Design 2.0 software and covered specific CpG islands in the target genes (Supplementary Table 2). Purified converted DNA was amplified with a PyroMark PCR Kit. Pyrosequencing was conducted on a pyrosequencer (Qiagen, PyroMark Q24).

Chromatin immunoprecipitation (ChIP)

Each immunoprecipitation assay was performed using fresh foetal limb muscle according to the protocol of the SimpleChIP® Plus Enzymatic Chromatin IP Kit (Magnetic Beads) (Cell Signaling Technology (CST), Danvers, Massachusetts, USA, 9005). The samples were first fixed with 1.5% formaldehyde for crosslinking for 20 min at room temperature, after which the reaction was terminated by the addition of glycine. The compact tissues were isolated into cell suspensions using Dounce homogenizers. Micrococcal nuclease (CST, 10011) was added together with ultrasonication (Diagenode, Belgium, Bioruptor Pico) to digest DNA to approximately 150–900 bp in length, followed by EDTA (CST, 7011) to stop digestion. Ten microlitres of a total 500 μl of chromatin sample were used for a 2% input control. Primary antibody or anti-IgG (CST, 2729) 1.5 μg was added to 200 μl of chromatin sample, which was subsequently incubated with rotation overnight at 4 °C. 25 μl of Protein G Magnetic Beads (CST,9006) was added to each IP and input reaction, and the mixture was rotated at 4 °C for another 2 h. DNA was ultimately purified after elution from the chromatin-antibody-bead mixture. The primers used are listed in Supplementary Table 2. ChIP sequencing was conducted by Novogene (Beijing, China). The data were processed through the Integrative Genomics Viewer (IGV).

Statistical analysis

The data are shown as the mean ± SEM. In the F1 generation, the differences in body composition, muscle weight ratio, area under the curve (AUC) of GTT and ITT, grip force, gene expression, and western blot data between the CTR and GDM groups were evaluated by a two-tailed Student’s t test. GTTs, ITTs, and weight monitoring in F1 generation were generally analysed via two-way ANOVA. The results were subsequently compared according to the time points in detail by Sidak’s multiple comparisons test. In the F2 generation, body composition, muscle weight ratio, the area under the curve (AUC) of the GTT and ITT, and gene expression were compared among the four groups by one-way ANOVA. Weight monitoring, GTT, and ITT were performed through two-way ANOVA, generally followed by Tukey’s multiple comparisons test, specifically at the time points. A P < 0.05 was considered to indicate statistical significance. GraphPad Prism software (GraphPad Software, San Diego, CA, USA, version 8.3.1) was utilised for statistical analysis and graphics.