Mice and drug treatments

Wild-type SPF-grade male C57BL/6J male mice (WT) were purchased from Vital River Experimental Animal Technology Co., LTD. SPF-grade Male B6.Cg-dock7m Leprdb/++/J (db/db) mice (strain#: 000699) were purchased from The Jackson Laboratory (JAX). The mice were professionally and securely delivered to Laboratory Animal Research Center, Tsinghua University (THU-LARC). The mice were kept stable for a week before undergoing the studies to ensure that they acclimated to their new habitat. The animals were housed in a sterile barrier environment at 23 °C with a 12-h cycle of light and darkness. After arrival, mice on a normal diet (ND) were fed with irradiation-sterilized JAX-standard breeder chow (SHOOBREE®, Xietong Pharmaceutical Biotechnology Co., Ltd., 1010058) and sterilized water during the entire study. An adequate number of 8-week-old WT mice were chosen for diet-induced obesity (DIO) and fed a high-fat diet (HFD; ResearchDiet, D12492) for 8 weeks. Mice with a bodyweight of more than 30 g were chosen to participate in the experiments.

The young, undamaged, and ultrapure rMSA and treatments used in this study were the same as previously described [22] unless stated otherwise herein. Briefly, mice in the rMSA-treated group were i.v. injected once every 3 weeks with rMSA in a dose of 1.5 mg per gram of body weight, while the control group mice were injected with isometric saline. All animal studies were approved by the Institutional Animal Care and Use Committee of Tsinghua University (Beijing, China).

Cell culture, treatments, and transfections

Mouse hepatocytes AML12 were purchased from ATCC (#CRL-2254) and cultured with DMEM/F12 medium (Gibco) supplemented with 10% FBS, 10 μg/ml insulin, 5.5 mg/ml transferrin, 5 ng/ml selenium, 40 ng/ml dexamethasone, and 15 mM HEPES. Mouse insulinoma cell line MIN6 (Beijing Crisprbio, CE18728) was maintained under a 5% CO2 atmosphere at 37 °C in RPMI 1640 medium supplemented with 10% FBS, 10 mM HEPES, 2 mM l-glutamine, 1 mM sodium pyruvate, and 0.05 mM 2­mercaptoethanol. Mouse primary hepatocyte isolation was performed as previously described [23].

For mass labelling of AML12 hepatocytes, AML12 hepatocytes were washed twice with warm PBS and starved in serum-free DMEM overnight. Labelled compounds [U-13C]-glucose (Sigma, 389374) were added to serum-free DMEM (without glucose). Indicated concentrations of rMSA with or without insulin were added at the same time.

For signaling pathway analysis, AML12 hepatocytes or mouse primary hepatocytes were washed twice with warm PBS and starved in serum-free DMEM overnight. For the AKT signaling, after the pretreatment with vehicle, Wortmannin (Selleck, S2758), PI-103 (Selleck, S1038), or ZSTK474 (Selleck, S1072), respectively for 30 min, indicated concentrations of rMSA with or without insulin were added at the same time in the absence or presence of PI3K inhibitors. For the IGF-1R signaling, after the pretreatment with vehicle or OSI-906 (Selleck, S1091) for 30 min, indicated concentrations of rMSA with or without insulin were added at the same time in the absence or presence of IGF-1R inhibitor. For the SRC signaling, after the pretreatment with vehicle, PP1 (Selleck, S7060), or Dasatinib (Selleck, S1021) for 30 min, indicated concentrations of rMSA were added at the same time in the absence or presence of SRC inhibitors. For the EGFR signaling, after the pretreatment with vehicle, BDTX-189 (Selleck, S9786), or Varlitinib (Selleck, S2755) for 30 min, indicated concentrations of rMSA were added in the absence or presence of EGFR inhibitors.

For induction of MIN6 cell apoptosis, sodium palmitate (PA) was dissolved to 250 mM in 50% ethanol at 60 °C, conjugated to 10% fatty acid-free bovine serum albumin (BSA; Beyotime, ST025), and added to serum-free RPMI 1640 medium to a final concentration of 0.2 mM PA and < 0.05% ethanol. PA, rMSA, sulfo-N-succinimidyl oleate sodium (SSO; MCE, HY-112847A), Fumonisin B1 (MCE, HY-N6719), CCT020312 (MCE, HY-119240), and GSK2656157 (Selleck, S7033) of specific concentrations were dissolved as indicated or manufacturer’s instructions and used to treat MIN6 cells for 16 h.

For immobilized rMSA and PA treatment, NHS-activated agarose (Sangon Biotech, C600024) was used to bind rMSA following the manufacturer’s instruction. Briefly, the agarose was washed and balanced with 1 mM HCl and sodium bicarbonate buffer (pH = 8.0) respectively. Then rMSA and the agarose, or the agarose alone, were co-incubated overnight in sodium bicarbonate buffer (pH = 8.0) at 4 °C, following blocking with 0.1 M Tris–HCl Buffer (pH = 8.5). FITC labeled rMSA was also immobilized with the same method to examine whether the rMSA could be dissociated from the agarose and uptake by cells. MIN6 cells were cultured in 6-well plates for 80% confluence. PA or vehicle, and immobilized rMSA or empty agarose were added to Transwell chambers with 4 μm pore membranes hanging above the cells. Cells were treated for 16 h and subjected to western blot for apoptosis assessment.

Transfection was performed with FuGene HD (Roche Applied Science). Cav1 (5′-CUGCGAUCCACUCUUUGAATT-3′ and 5′-CGCUUGUUGUCUACGAUCUTT-3′), and control small interfering RNAs (siRNAs) were from Genechem.

Assessment of glucose homeostasis in animals

Mice fasted for 6 h before fasting blood glucose level (FBGL) values were measured using a glucometer (Roche, ACCU-CHEK) on blood from tail snip. For intraperitoneal glucose tolerance tests (IPGTT) and oral glucose tolerance tests (OGTT), blood samples from mice fasted for 15 h were collected at 0, 15, 30, 60, and 120 min after an i.p. injection of glucose (2 mg/g body weight) or intragastric administration (1 mg/g body weight). For in vivo glucose-stimulated insulin secretion (GSIS), the insulin was measured from serum collected at 0 and 15 min after an i.p. injection of glucose (2 mg/g body weight). Insulin concentrations were determined using a radio immune assay (RIA) kit by the manufacturer (Beijing Victor Biotechnology Co., Ltd., RK-146). For insulin tolerance tests, mice fasted for 4 h and were i.p. injected with insulin (Beyotime, P3375), following blood glucose measurements at 0, 15, 30, and 60 min.

In omics and histology studies, the serum, pancreas, and liver were collected in fed status for sample preparation or 4% paraformaldehyde (PFA) fixation and paraffin embedding. Major blood biochemical parameters of serum samples were determined with an automatic biochemistry analyzer (Sysmex, BX3010).

Assessments of energy budget in animals

For energy budget analysis, mice were placed into each Metabolic PhenoCage (TSE Systems) loaded with feed and drinking water in advance. Consumptions of feed, drinking water, and oxygen, CO2 production, and other parameters were tracked (about every 20 min). The monitoring data from the 4th day was gathered for statistics.

Apoptosis and cell viability assays

For the determination of apoptosis in the paraffin-embedded pancreatic section, a TUNEL assay (Beyotime, C1088) was performed according to the manufacturer’s instructions. Images were captured and TUNEL+ nuclei surrounded by insulin+ staining were determined with the Automated Quantitative Pathology Imaging System and inForm Software (Vectra Polaris).

For determination of apoptosis in PA/rMSA-treated MIN6 cells, an Annexin V-FITC/PI apoptosis detection kit (Mei5bio, MF124) and a GreenNuc™ Caspase-3 Assay Kit for Live Cells (Beyotime, C1168) were used following manufacturers’ instructions. The apoptosis level of cells was analyzed with a flow cytometer (BD, Aria II).

For determination of cell viability, 5,000 MIN6 cells per well were seeded in a 96­well plate and subjected to specific treatments. Cell viability was determined with a Cell Counting Kit-8 (CCK-8; Mei5bio, MF128) according to the manufacturer’s instructions.

H&E staining and immunofluorescence analyses

Paraffin-embedded tissues were sectioned at 5 μm thickness. Paraffin sections with the largest tissue surface area were used. Sections were dewaxed and rehydrated.

For H&E staining, sections were stained with hematoxylin and eosin successively. Whole images from pancreas slices were captured with the Automated Quantitative Pathology Imaging System (Vectra Polaris). Areas of islets, exocrine tissues, other tissues, and image background were quantified with the inForm Software (Vectra Polaris).

For immunofluorescence (IF) staining, BrdU (Beyotime, ST1056) was i.p. injected at 1 mg per mouse for five continuous days before mouse sacrifice. After antigen retrieval and blocking, appropriate primary antibodies were incubated overnight at 4 °C, followed by incubation with FITC/TRITC conjugated secondary antibodies (ZSbio). Dilution ratio of primary antibodies used: anti-insulin (1:200; Abcam, ab181547), anti-cleaved caspase-3 (CC3, 1:50; CST, 9661S), and anti-BrdU (1:200; Abcam, ab6326). Secondary antibodies were applied for 1 h at room temperature, and nuclei were counterstained with DAPI. Imaging was performed with Nikon A1 laser scanning confocal microscope and NIS-Elements Software (Nikon) and analyzed with inForm Software (Vectra Polaris) and Image-ProPlus Software.

Transmission electron microscopy

Cell pellets were fixed with glutaraldehyde followed by en bloc fixation with 2% (vol/vol) uranyl acetate, fixed in osmium tetroxide, dehydrated by alcohol gradient, and then infiltrated with Spurr’s resin. Transmission electron micrographs of these samples were acquired with an HT7800 instrument (HITACHI Japan). For quantitation of caveolae, only distinctly flask-shaped, noncoated vesicles (50–100 nm in diameter) found on the plasma membranes were scored as caveolae. Total caveolae counts were normalized to the unit length of the plasma membrane (10 μm) and measured using Image J software.

RNA sequencing and data analysis

A total of 6 liver samples from mice treated with saline or rMSA for 9 weeks were randomly selected to explore the chronic effects of rMSA on the hepatic gene expression profile. AML12 hepatocytes were washed twice with warm PBS and starved in serum-free DMEM overnight. The AML12 hepatocytes were respectively treated with vehicle, insulin (10 nM), rMSA (600 μM), or insulin + rMSA for 24 h. The mirVana RNA Isolation Kit was used to isolate total hepatic RNAs (Life Technologies). For the construction of the cDNA library, RNA samples having a RNA integrity number greater than 8.5 were employed, and rRNA was depleted. A BGISEQ 500 platform was used to do the mRNA sequencing (BGI, Hong Kong). Differentially expressed genes (DEGs) were determined with the R package “DESeq2” (v 1.18.0). With an absolute log2 fold change (shrunken) ≥ 1 and a Benjamini–Hochberg adjusted p < 0.05, a gene was considered as differentially expressed. The Dr. Tom web tool (https://biosys.bgi.com/) was used to visualize the DEGs and enriched pathways. Gene set enrichment analysis (GSEA) was performed to identify significantly enriched pathways in the KEGG databases using the Dr. Tom web tool.

Mass spectrometry analysis

Proteomic analysis was performed as described in the previous work [24]. Briefly, an equal weight of the pancreas or livers from each mouse was homogenized in an 8 M urea containing protease and phosphatase inhibitor cocktail (Beyotime, P1048). The mixture was centrifuged, and the concentration of protein was determined in the supernatant. Two hundred micrograms of proteins were desalinized and digested with trypsin. Each sample was labeled with a unique tandem mass tag (TMT) (ThermoFisher, 90110) and subjected to LC–MS/MS analyses respectively. Differentially expressed proteins were determined with the R package “limma” (v 3.42.2). With an absolute log2 fold change ≥ 1 and an adjusted p < 0.05, a protein was considered as differentially expressed. For gene set enrichment analysis (GSEA), GSEA Software v4.3.2 was used according to the mouse hallmark (MH) gene set that was obtained from MSigDB mouse collections [25, 26].

For lipidomics analysis, MIN6 cells in 6 cm dishes of 80% confluence were treated with 0/0.2 mM PA and 0/300 μM rMSA for 16 h. Cells were harvested, counted, and lysed with 300 μL methanol. Lipids were extracted by adding 600 μL dichloromethane to the mixture and thorough vortex. The organic phase was placed under a nitrogen blower at room temperature until the solvent volatilized, and then redissolved by an equal number of cells. LC–MS/MS was performed as previously described [27, 28].

For the Flux method, AML12 hepatocytes in 6 cm dishes were treated with 0/10 nM insulin and 0/600 μM rMSA for 6 h. After the treatments, AML12 hepatocytes were lysed with 3 ml 80% methanol. The lysate was centrifuged at 12,000×g at 4 °C for 20 min. The supernatant was spin-dried and dissolved for LC–MS/MS analyses.

Glucose uptake assay

AML12 hepatocytes were washed twice with warm PBS and starved in serum-free DMEM (without glucose) overnight. After the pretreatment with vehicle or inhibitors for 30 min, indicated concentrations of rMSA with or without insulin were added. After the treatment, all culture medium was removed from each well and the glucose uptake was detected using a Cell Meter™ 2-NBDG Glucose Uptake Assay Kit (AAT Bioquest, 23500) following the manufacturer’s instruction.

Fatty acid uptake analysis

To examine the effects of rMSA on fatty acid (FA) uptake using flow cytometry, MIN6 cells cultured in 6-well plates of 80% confluence were treated for 30 min at 37 °C with serum-free RPMI-1640 medium containing rMSA of specific concentrations and 15 μM BODIPY™ FL-C12 to simulate the uptake and metabolism of PA [29].

For FA uptake rate assessments, MIN6 cells were incubated with serum-free RPMI-1640 medium containing 0.2 mM PA and rMSA of specific concentrations. Media at 0, 0.5, 1, 2, 4, 8, 12, and 24 h time points were collected to determine the concentration of FA using an FFA Quantitative Detection Kit (Boxbio, AKFA008M) following the manufacturer’s instruction.

Western blot analysis

Cells subjected to specific treatments were harvested and lysed in lysis buffer for western blot (Beyotime, P0013) containing protease inhibitor cocktail and phosphatase inhibitor cocktail (Beyotime, P1048). Protein extracts were resolved on an SDS-PAGE gel and transferred to a PVDF membrane. Membranes were incubated overnight at 4 °C with appropriate primary antibodies. Dilution ratio of primary antibodies used: anti-p-AKT (1:1000; CST, 4058S), anti-AKT (1:1000; CST, 9272S), anti-p-GSK3β (1:1000; Proteintech, 67558-1-Ig), anti-GSK3β (1:1000; Proteintech, 22104-1-AP), anti-p-FOXO1/FOXO3a (1:1000; Beyotime, AF605), anti-FOXO1 (1:1000; Beyotime, AF603), anti-β-actin (1:1000; ZSbio, TA-09), anti-p-IGF-1Rβ/IRβ (1:1000; CST, 3024T), anti-IGF-1Rβ (1:1000; CST, 9750T), anti-p-SRC (1:1000; CST, 6943S), anti-SRC (1:1000; CST, 2109S), anti-p-EGFR (1:1000; CST, 2234S), anti-EGFR (1:1000; CST, 4267S), anti-CAV1 (1:1000; CST, 3267S), anti-CC3 (1:1000; CST, 9661S), anti-CHOP (1:1000; Proteintech, 15204-1-AP), anti-p-PERK (1:1000; CST, 3179S), anti-PERK (1:1000; CST, 3192S), anti-β-Tubulin (1:2000; ZSbio, TA-10). Detection of proteins was carried out with a CCD scanner (Tanon, 4600SF) by incubations with horseradish peroxidase (HRP) conjugated secondary antibodies (CST) followed by enhanced chemiluminescence detection reagents (ThermoFisher, 34095).

Phospho-antibody array analysis

For comparison of the activation and expression of phosphoproteins by rMSA treatment, AML12 hepatocytes were treated with or without rMSA (600 μM) for 24 h. Then, the samples were treated with cell lysis buffer (Full Moon BioSystems, USA). Next, phosphatase inhibitor and protease inhibitor were added to each sample at a volume ratio of 1:50, and each sample was treated with a tube of Full Moon cell lysis beads. After the above steps, the proteins of the samples were extracted, and then, Phospho Explorer Antibody Array (Full Moon BioSystems, USA) was used to react with the protein according to the standard procedures provided by Full Moon. The array contained 1318 antibodies, and each of them had 2 replicates. Wayen Biotechnologies (Shanghai) performed the experiment and analyzed the data. The extent of protein phosphorylation was calculated by the following equation:

Co-immunoprecipitation

Co-immunoprecipitation was performed using the Immunoprecipitation Kit with Protein A+G Magnetic Beads (Beyotime, P2179S) following the manufacturer’s instruction. The pre-cleared cell lysate was incubated with CAV1 antibody (1:200; CST, 3267S) immobilized on magnetic beads overnight at 4 °C, after which the beads were washed extensively with lysis buffer, eluted, and solubilized in SDS sample loading buffer.

Glycolysis stress tests

Five thousand AML-12 hepatocytes were seeded on a Seahorse 96-well tissue culture plate (Agilent Technologies). The plate was stood for 30 min for the cells to settle before placing it in the incubator overnight. The adhered cells were treated with or without 600 μM for 6 h before Seahorse analysis. The cells were switched to serum-free Seahorse media before the assay according to the manufacturer’s instructions. For cells subjected to Seahorse XF Glycolysis Stress Test Kit (103020-100, Agilent Technologies), glucose (20 mM), oligomycin (2 mM), and 2-DG (50 mM) were injected sequentially. At the beginning of the assay, the medium was changed to unbuffered, glucose-free DMEM (Sigma-Aldrich Cat D5030, pH 7.35 at 37 °C) supplemented with 2 mM glutamine.

Statistical analysis

Unless otherwise stated, data are presented as mean ± s.e.m. Data are derived from experiments repeated at least three times unless stated otherwise. Animal studies were performed with the blinding of the experiment operator. If not mentioned otherwise in the figure legend, statistical significance is indicated by *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Statistical analysis was carried out using unpaired, two­tailed t­test or two­way ANOVA unless otherwise noted. GraphPad Prism 8 and R 4.2.2 were used for statistical analysis.