Integration of metabolomics and transcriptomics to reveal metabolic characteristics and the role of mTORC1 in β-cell proliferation induced by a short-term high-fat diet

Abstract

Background

Pancreatic β-cell proliferation is essential for maintaining the balance of β-cell mass, and an elevated metabolic load can stimulate their proliferation. Numerous studies have shown that a short-term high-fat diet increases metabolic load without affecting insulin sensitivity, thereby promoting the proliferation of pancreatic β-cells. However, the underlying mechanisms of this effect remain to be fully elucidated.

Results

A model has been constructed in our study to emulate pancreatic β-cell proliferation induced by a short-term high-fat diet, aiming to scrutinize the underlying mechanisms. Integrated transcriptomic and metabolomic analyses suggest that the mTORC1 signaling pathway may be crucial in this induced proliferation. Further analysis revealed that rapamycin, a specific inhibitor of the mTORC1 pathway, can inhibit proliferation induced by the short-term high-fat diet.

Conclusion

Our study confirms the significant role of the mTORC1 signaling pathway in pancreatic β-cell proliferation induced by a short-term high-fat diet.

Introduction

Diabetes mellitus is a prevalent chronic disease, affecting approximately 536.6 million people worldwide as of 2021, with projections estimating a rise to 783.2 million by 2045 (Nair, Tzanakakis & Hebrok, 2020; Sun et al., 2021). The ongoing loss of functional β-cell mass is a critical factor in both type 1 and type 2 diabetes (Abel et al., 2024; Eizirik, Pasquali & Cnop, 2020). To maintain internal homeostasis, β-cell mass undergoes dynamic changes, relying on a delicate equilibrium of cell neogenesis, proliferation, hypertrophy, and apoptosis (Bartolome et al., 2022; McNelis et al., 2015). Numerous investigations have evidenced that the amplification of β-cell mass predominantly originates from the replication of pre-existing β-cells (Granger & Kushner, 2009; Meier et al., 2008; Shcheglova, Blaszczyk & Borowiak, 2021). Thus, understanding the mechanisms underlying β-cell replication and discovering effective and relatively safe therapeutic agents for diabetes is crucial, despite the inherently low replicative capacity of β-cells (Shcheglova, Blaszczyk & Borowiak, 2021; Wang et al., 2015).

Under conditions of increased metabolic stress, such as obesity, pregnancy, and pancreatic injury, β-cell mass undergoes a series of compensatory proliferative changes, a phenomenon observed in both rodents and humans (Kondegowda et al., 2015; Sachdeva & Stoffers, 2009; Tschen et al., 2009). Mice fed a high-fat diet (HFD) represent an excellent in vivo animal model for exploring the mechanisms of β-cell compensatory proliferation due to its wide availability, low cost, and easy genetic manipulation (Roat et al., 2014; Woodland et al., 2016). Previous research has indicated that β-cell proliferation begins within the first 7 days following the initiation of high-fat feeding (Stamateris et al., 2013). Moreover, Mosser et al. (2015) observed that the increased β-cell proliferation occurs even earlier when mice are exposed to HFD for only 3 days. Additionally, other studies have confirmed that short-term treatment with HFD is sufficient to induce β-cell proliferation without pronounced effects on systemic insulin resistance (Kitao et al., 2018; Woodland et al., 2016). However, the mechanisms underlying these associations remain unclear. This study employed an integrated transcriptomic-metabolomic approach to prioritize candidate genes regulating high-fat diet-induced β-cell proliferation, leveraging gene-metabolite network analysis. Notably, the mechanistic target of rapamycin (mTORC1), a nutrient sensor with significant roles in modulating β-cell mass (Chau et al., 2017; Esch et al., 2020), was found to be involved in the proliferation of pancreatic β-cells induced by the short-term high-fat diet.

Materials and Methods

Seven-day HFD feeding protocol

Animal care and experiments were approved by the Animal Ethics Committee, Tongren Hospital, Shanghai Jiao Tong University School of Medicine (2021-076-01). In this study, 24 healthy 8-week-old C57BL/6N male mice were used, the number of which was determined based on previous studies and statistical analysis to ensure the statistical significance of the experimental results while adhering to the “replacement, reduction, and optimization” (3R principle). The mice were purchased from Nanjing Gempharmatech Co. Ltd and housed in the animal room of Shanghai Jiaotong University School of Medicine, with a temperature of 22 ± 2 °C, humidity of 50–60%, and a 12-hr light cycle, and free access to feed and drinking water. The mouse housing environment utilized standard mouse cages with external dimensions of 1,500 * 480 * 1,970 mm. Each cage houses five mice to ensure that each mouse has ample space for movement. Animals were checked daily, and veterinary staff were ready to evaluate any animal showing signs of illness or distress, which included weight loss (≥10%), lethargy, ruffled hair, or difficulty breathing, although no such cases occurred in this study. After weighing and measuring random blood glucose, the mice were randomly divided into two groups: normal chow diet (13% kcal from fat; Jiangsu Synergy Biological Engineering Co. Ltd, Nanjing City, Jiangsu Province, China) or high-fat diet (60% kcal from fat; Research Diets, Inc, NJ, USA). These mice were maintained on normal chow or high-fat feeding for 7 days. To minimize the impact of potential confounding factors, all experimental treatments and measurements were performed in a completely randomized design. On day 8, the mice were anesthetized (sodium pentobarbital, 50 mg/kg, intraperitoneal injection), and cervical dislocation was performed only after the mice were completely unconscious. Subsequently, the pancreas was removed for tissue sectioning or islet isolation. As for the mTORC1 inhibition experiment, two additional groups of 24 male mice were purchased and intraperitoneally injected with either the vehicle or rapamycin continuously for 7 days. Rapamycin (Immunoway, CA, USA) was dissolved into a final concentration of 0.2 μg/μL, and a single daily intraperitoneal injection dose was 0.5 mg/kg. The vehicle contained the same solvent (5% PEG400, 5% Tween 80, and 4% ethanol) as in the rapamycin solution. This procedure was conducted in strict compliance with relevant ethical guidelines and operational standards to minimize animal discomfort and suffering. Euthanasia at the end of the experiment was performed based on clear criteria, including but not limited to: reaching the predetermined research endpoint, the presence of irreversible disease or injury affecting welfare, or humane necessity. Post-experiment, surviving mice meeting predefined humane endpoints were euthanized as specified; those in good health and not meeting these criteria might be used in other research projects. All subsequent actions complied with applicable laws, regulations, and ethical committee requirements to ensure animal welfare.

Glucose and insulin tolerance and insulin secretion assays

Intraperitoneal insulin tolerance test (IPITT) and intraperitoneal glucose tolerance test (IPGTT) were conducted on day 7 to assess insulin sensitivity and glucose homeostasis, respectively. For the IPITT, mice were fasted for 6 h and then administered insulin (0.75 U/kg; Wanbang Bio-pharm, China) via intraperitoneal injection. Blood glucose concentrations were measured at various time points (0, 15, 30, 60 min) by glucometers (Ascensia Diabetes Care Inc, Leverkusen, Germany). The IPGTT was performed after a 16-hour fasting period. Mice received an intraperitoneal injection of glucose (2 g/kg; Otsuka Pharmaceutical Co., Ltd., China), and blood glucose concentrations were monitored at 0, 15, 30, 60, and 120 min using the same tail vein sampling method as in the IPITT. In the in vivo glucose-stimulated insulin secretion (GSIS) experiment, serum samples were re-collected at 0, 15, and 30 min following glucose administration. Fasting and glucose-induced serum insulin secretion were collected and measured by the Insulin Mouse Ultra Sensitive ELISA kit (Crystal Chem, Itasca, IL, USA). In the determination of pancreatic insulin content, the pancreas was isolated and cut into pieces before being homogenized in an acidified ethanol solution. This solution consisted of 75% absolute ethanol, 1.5% concentrated hydrochloric acid, and 23.5% water. The resulting mixture was then shaken overnight at 4 °C and the supernatant was collected through centrifugation. The fasting blood was collected and placed on ice before being centrifuged for 20 min at 5,000 rpm at 4 °C. The serum layer was obtained and frozen at −80 °C, waiting for non-targeted metabolic analysis.

Pancreatic islet isolation

In order to isolate pancreatic islets, the mice were anesthetized with sodium pentobarbital using intraperitoneal injection, and the anesthetized mice were fixed to the dissection using tape. The abdominal cavity of the mice was cut open along the midline of the abdomen, the duodenal papilla was found first, the common bile duct was found along the papilla, the common bile duct was clamped closed with a small hemostatic forceps, and then collagenase (0.5 mg/mL; Sigma) was gently pushed into the syringe to fill the whole pancreas along the common bile duct. Eventually, the pancreas was separated starting along the cecum, and the intact pancreas was placed in a centrifuge tube containing collagenase and digested at 37 °C. After digestion was terminated, the pancreas was centrifuged and sieved (Li et al., 2016; Yin et al., 2020). Finally, islets were hand-picked under a stereomicroscope. These islets were stored at −80 °C for subsequent RNA and protein analysis.

Transcriptomics analysis

The fresh islets were extracted with Trizol (Ambion, USA) and sent to Genedenovo Biotechnology Co. Ltd (Guangzhou, China) for further transcriptomics analysis. The RNA-seq analysis features a sequencing depth of 50X, includes three biological replicates, and achieves a statistical power of 0.999971, as calculated by RNASeqPower. The detailed experimental protocols were the same as before (Youlten et al., 2021). The extracted RNA quality was checked by Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) and agarose electrophoresis. According to Ribo-ZeroTM Magnetic Kit (Epicentre, USA) instructions, the general procedure was as follows: After total RNA was extracted, eukaryotic mRNA was enriched by Oligo (dT) beads. The poly-A-containing mRNA molecules were purified using poly-T oligomer-attached magnetic beads. After purification, the mRNA was fragmented by treatment with divalent cations at 94 °C for 8 min, and the cleaved RNA fragments were reverse transcribed into first-strand cDNA using reverse transcriptase and random primers. Second-strand cDNA was then synthesized using DNA polymerase I and RNase H. These cDNA fragments were then subjected to an end-repair process where a single A base was added and ligated to the adapter. The products were then purified and amplified by PCR to create the final cDNA library. The purified library was quantified using a Qubit® 2.0 fluorometer (Life Technologies, Waltham, MA, USA) and validated with an Agilent 2100 Bioanalyzer (Agilent) to determine the size of the insert and calculate the molar concentration. The resulting cDNA library was sequenced using Illumina Novaseq6000. Pre-process raw reads for sequencing by filtering out rRNA reads, sequencing adapter reads, short fragment reads and other low-quality reads. The processed reads were aligned to the reference genome using HISAT2.2.4. After gene annotation, FPKM values of known genes were analyzed by running StringTie v1.3.1 for integral calculations. Finally, the differentially expressed genes (DEGs) were identified. The criteria used for the screening of DEGs was fold change ≥ 1.5 or ≤0.67 and P value <0.05. All data analysis was conducted by Genedenovo Biotechnology Co., Ltd (Guangzhou, China).

Metabolomics analysis

The serum samples of mice (n = 6 per group) were further selected and sent to BioNovoGene Co., Ltd (Suzhou, China) for untargeted metabolic profiling analysis. Firstly, serum metabolite extraction was performed as described below. A total of 400 µL of methanol was added to 100 µL serum samples and shaken for 60 s. After being centrifuged for 10 min (12,000 rpm, 4 °C), the supernatant was transferred to a new tube, concentrated, and dried under vacuum. The prepared samples for LC-MS were obtained by dissolving the samples in a 150 µL solution of 2-chlorobenzalanine (4 ppm) and 80% methanol. The supernatant was then filtered through a 0.22 µm membrane. Taking 20 µL from each sample to the quality control samples and the remaining samples were used for LC-MS. The mass spectrometer used an electrospray ionization source (ESI) in positive and negative ionization mode, with a full scan at a resolution of 60,000 and a scan range of 81 to 1,000, while using dynamic exclusion to remove unnecessary MS/MS information. The following were the chromatographic conditions: column temperature, 40 °C; flow rate, 0.25 mL/min. Gradient elution of analytes was carried out with 0.1% formic acid in water (C) and 0.1% formic acid in acetonitrile (D) or 5 mM ammonium formate in water (A) and acetonitrile (B). After equilibration, each sample was injected in a volume of 2 µL. The optimal chromatographic gradient was as follows: 0~1 min, 2% B/D; 1~9 min, 2~50% B/D; 9~12 min, 50~98% B/D;12~13.5 min, 98% B/D; 13.5~14 min, 98~2% B/D; 14~20 min, 2% D-positive model. Data were gathered with the aid of the software Thermo Excalibur 2.2 (Thermo Fisher Science, MA, USA) in centroid mode. The screening criteria for differential metabolites was fold change ≥ 2 or ≤0.5 and P value <0.05.

Functional network analysis

To screen for functional networks among the differential metabolites and genes, the ingenuity pathway analysis (IPA), an online integrated analysis platform available at the following website: http://www.ingenuity.com, was employed to integrate metabolomic and transcriptomic data.

Western blot analysis

Total proteins were extracted from islets with RIPA lysis buffer (Shanghai Beyotime Biotechnology Co. Ltd, Shanghai, China) supplemented with PMSF (Shanghai Beyotime Biotechnology Co. Ltd). The protein concentration was detected by the Pierce BCA protein assay kit (Thermo Fisher Scientific, WA, MA, USA). A total of 30 μg proteins were loaded and separated by 10% SDS-PAGE and then transferred to the membranes of polyvinylidene fluoride (Millipore, Burlington, MA, USA). Rabbit anti-pS6 (Ser-235/236) antibody (1:1000; CST, Danvers, MA, USA), β-Tubulin antibody (1:20000; Affinity, USA), β-actin antibody (CST, 1:1000), Anti-rabbit HRP Antibody (1:2000; CST, USA), Anti-mouse HRP Antibody (1:2000; CST) were used as indicated. The blotted membrane was developed with ECL Advance (Biovision, Milpitas, CA, USA) and imaged with a LAS-4000 Super CCD Remote Control Science Imaging System (Fuji film, Tokyo, Japan). Intensities of bands were quantified by Image J, expressed in arbitrary units of optical density, and normalized for tubulin intensity in the same blot.

Tissue preparation and histology

After overnight fasting, mice were anesthetized and euthanized by cervical dislocation. The entire pancreas was promptly excised, rinsed in PBS, gently blotted dry, and immersed in 4% paraformaldehyde for fixation at 4 °C overnight. The pancreatic tissues were dehydrated and defatted in 50%, 70%, 80%, 95%, and 100% alcohol in sequence. Pancreatic tissues were embedded in paraffin and serially sectioned at 5 μm intervals. Dewaxed and rehydrated pancreatic sections were subjected to antigen retrieval, permeabilization, blocking, and subsequent immunofluorescence staining. The sections were immunolabeled with rabbit anti-ki67 (1:200; Abcam) or rabbit anti-pS6 (1:200; Ser235/236, CST, USA) and guinea pig anti-insulin (1:2; Dako). Fluorescent secondary antibodies were purchased from Jackson ImmunoResearch and used at 1:200 dilution. β-Cell proliferation was determined by counting the number of insulin-ki67 dual-positive cells. Co-immunofluorescence staining of pS6 and insulin was performed to assess mTORC1 activation in pancreatic β-cells. For the assessment of β-cell mass (n = 6 mice/group), pancreatic tissues from each mouse were processed as follows: representative sections were selected at intervals of 10 consecutive slices (with a thickness of 5 μm per section) to minimize the risk of repeated sampling from the same anatomical region. All selected sections underwent full-slide scanning, and images were captured using a NanoZoomer S360 instrument with a ×20 objective lens. Quantitative analysis was performed using ImageJ software. The insulin-positive area and the total pancreatic area were measured in each sampled section. β-cell mass was then calculated by scaling the relative β-cell area to the total pancreatic weight. The size of pancreatic β-cells was determined by immunofluorescence staining of pancreatic tissue sections. Rabbit anti-β-catenin (1:100; CST) and guinea pig anti-insulin (1:200; Dako) were used as primary antibodies.

RNA preparation and real-time PCR

Islets were isolated from mice in both the normal diet and high-fat diet groups, with three mice per group subjected to qRT-PCR analysis. Total RNA from the islets was extracted using Trizol reagent (Ambion), and the concentration and purity of the RNA were assessed using a NanoDrop One spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Complementary DNA (cDNA) was synthesized using a reverse transcription kit (TOYOBO, Japan). The RNA template amount was 500 ng, employing the RT Master Mix II. The reaction volume was 10 µL, with the thermal cycling conditions consisting of 37 °C~15 min, 50 °C~5 min, 98 °C~5 S. The primers were designed online using Primer-Blast software (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). The design required a product amplification length of 100–250 bp, Optimal Tm 57–62 °C, and Primer GC% 40–60%. The primer sequences were listed in Table 1. qRT-PCR was conducted with a SYBR Green Realtime PCR Master Mix kit (TOYOBO, Japan) on a QuantStudio 5. The protocol included an initial 94 °C denaturation for 30 s, followed by 40 cycles of 94 °C for 5 s and 60 °C for 30 s. A melt curve analysis was conducted to verify the specificity of the PCR amplification. Gene expression was quantified using the 2^−ΔΔCt method with Gapdh as the reference gene. Substrate specificity was confirmed through melt curve analysis, while PCR efficiency was ascertained by constructing standard curves from serial dilutions of cDNA. Efficiencies for each primer pair were determined to be within the range of 90% to 110%. All experiments were conducted in triplicate to ensure reproducibility.

Table 1. Primer sequences.

Gene	Forward	Reverse
Mtor	GCATAACAGATCCTGACCCTGAT	TTCAGAGCCACAAACAGAGC
Rps6kb1	GAGCTGGAGGAGGGGG	ACAATGTTCCATGCCAAGTTCA
Rps6	GCCCTTAAACAAAGAAGGTAAGAAG	CGCTTTTCTTTGGCTTCCTTCA
Eif4ebp1	ACTCACCTGTGGCCAAAACA	TTGTGACTCTTCACCGCCTG
Eif4b	ACTGTGGAACCGTGGGAGAT	AGTGGTAGGGGTGGTTTCCT
Mki67	ACCATCATTGACCGCTCCTT	TTGACCTTCCCCATCAGGGT
Cdk2	CGGCTCGACACTGAGACTG	TTCTTGAGGTCCTGGTGCAG
Gapdh	CATCACTGCCACCCAGAAGACTG	ATGCCATGAGCTTCCCGTTCAG

Statistical analysis

Data were analyzed between the two groups using Student’s t-test. All graphical representations were created using GraphPad Prism version 10.0 software. Statistical data are commonly presented as mean ± standard error of the mean (SEM). P < 0.05 was considered statistically significant. Our experimental design was well-structured, procedures were precisely conducted, statistical methods were appropriately applied, and data were complete and ethically obtained; therefore, there was no need to exclude any data. During the experimental group allocation, conduct, outcome assessment, and data analysis phases, only essential personnel were privy to the group assignments. A double-blind design was implemented to minimize bias, ensuring the objectivity and reliability of the results.

Results

Compensatory $\text{β}$-cell proliferation occurred after 7 days of high-fat feeding

Male C57BL/6N mice, aged 8 weeks, were subjected to either a high-fat diet (HFD) or a normal chow diet (CD) over a 7-day period, during which alterations in body weight were meticulously observed. Consistent with expectations, high-fat feeding induced a 10.18% augmentation in body weight, notably surpassing that of the CD group (Fig. 1A). This elevation in weight was concurrently associated with an accrual of epididymal fat. After dissecting the mice to weigh their epididymal fat mass, the epididymal fat mass of the HFD group was about twice that of the CD group (Fig. 1B). Despite presenting normal fasting and random blood glucose levels (Figs. 1C and 1D), the HFD mice displayed a deviation in glucose homeostasis as evidenced by the IPGTT (Fig. 1E). Our analysis showed that blood glucose levels were significantly higher in HFD-fed mice than in CD-fed mice at 15 and 30 min (P < 0.001 and P < 0.05, respectively), indicating mildly impaired glucose tolerance after short-term HFD exposure. Subsequent IPITT tests were conducted to examine potential alterations in insulin sensitivity due to high-fat feeding. The result suggested that the HFD mice do not exhibit insulin resistance (Fig. 1F). Since high-fat feeding did not result in insulin sensitivity, plasma insulin concentrations were assessed to determine whether the impaired glucose intolerance was attributed to insufficient insulin secretion. The glucose-stimulated insulin secretion (GSIS) experiment indicated a modest decrease in insulin secretion in mice fed a high-fat diet (Fig. 1G). In addition, pancreatic insulin content was also not remarkably affected by 7 days of HFD (Fig. 1H). Taken together, the 7-day high-fat diet led to impaired glucose tolerance and insulin secretion.

Figure 1. Compensatory β cell proliferation occurred after 7 days of high-fat feeding.

Short-term high-fat feeding induced significant β cell replication and impaired glucose tolerance. (A and B) 7 days following the high-fat diet (HFD) or chow diet (CD) feeding, the body weight (A) and epididymal fat mass (B) of mice were measured. (C-D) Fasting blood glucose and random blood glucose were monitored in two groups; n = 6. (E and F) Intraperitoneal glucose tolerance test (IPGTT) and Intraperitoneal insulin tolerance test (IPITT) were performed to evaluate the homeostasis of glucose. (G) Plasma insulin during IPGTT and insulin content (H) were measured; n = 6. (I) Representative microscopic images of the ki67/insulin double-positive cells revealed the percentage of proliferating β cells (J), n = 3; scale bars, 20 µm. (K and L) H&E and insulin staining were carried out to assess β cell mass, n = 7; scale bars, 50 µm . (M-N) The size of pancreatic β-cells was determined by immunofluorescence staining. At least three sections from three mice per group were used; scale bars, 20 µm. Data were presented as mean ± SEM. Student’s t-test: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

To understand the response of β-cells to short-term high-fat feeding, proliferation marker ki67 and insulin immunolabeling were performed on pancreas sections, and the percentage of ki67/insulin double-positive cells was calculated to compare the proliferation rate of β-cells in the CD and HFD groups. The analysis revealed that the proliferation of β-cells in HFD mice was significantly increased, which was 1.76 times that of CD mice (Figs. 1I and 1J). To further investigate whether the 7-day high-fat diet was capable of inducing β-cell mass expansion, the pancreas mass and insulin-positive area divided by section area were quantified. As shown in Figs. 1K, 1L, high-fat diet fed mice exhibited a significant increase in β-cell mass compared with normal chow diet fed mice (P < 0.01), indicating that the compensatory response of β-cells is already initiated within 7 days of HFD exposure. To determine whether this expansion arose from cellular hypertrophy or hyperplasia, β-cell size was assessed. No significant difference in individual β-cell size was observed between HFD and CD mice (Figs. 1M and 1N). This finding indicates that the increase in β-cell mass occurs independently of cellular enlargement. Instead, the early adaptive expansion of β-cells under short-term HFD is driven predominantly by increased cell proliferation, highlighting hyperplasia as the primary mechanism.

HFD-induced changes in Gene-expression profiles

To elucidate the molecular mechanisms of the effects of short-term high-fat feeding on pancreatic β-cell proliferation, RNA sequencing analysis of islets from HFD and CD mice was performed. Compared to the control group, a total of 169 genes (fold change ≥ 1.5 or ≤0.67 and P value <0.05) were significantly differentially expressed in the HFD group, which included 39 up-regulated and 130 down-regulated genes (Figs. 2A, 2B). A heat map was used to visualize the top 20 upregulated and downregulated genes in Figs. 2D. It was noted that the differentially expressed genes (DEGs) were related to the regulation of cell proliferation and insulin secretion, such as Bmp6, Arhgdib, Rab31, and so on. As shown in Fig. 2D, the significant downregulation of Bmp6 and Arhgdib in the high-fat diet group suggests that a short-term high-fat diet may impair pancreatic β-cell function. The Rab31 gene was upregulated in the high-fat diet group and may promote cell proliferation through the mTOR/p70S6K/Cyclin D1 signaling pathway in vitro (Yang et al., 2020). Gene ontology (GO) analyses revealed cellular process, single-organism process, biological regulation, regulation of the biological process, response to stimulus, metabolic process, and growth as the main GO biological processes. In the molecular function subontology, terms including binding, catalytic activity, molecular transducer activity, signal transducer activity, and molecular function regulator were significantly enriched. The cellular components of these key targets mainly were cell part, cell, organelle, membrane and membrane part (Fig. 2C). Next, KEGG pathway analysis of the freshly isolated mouse islets transcriptomics showed the top 20 significantly enriched pathways (Fig. 2E), and some of them were involved in the mTOR signaling pathway, breast cancer, hepatocellular carcinoma, autophagy, insulin signaling pathway, PI3K-Akt signaling pathway, wnt signaling pathway, and FoxO signaling pathway.

Figure 2. HFD-induced changes in Gene-expression profiles.

Gene-expression profile of the HFD and CD groups. (A) Volcano plot showing differentially expressed genes (up- and down-regulated) from islets in the chow diet group compared with the short-term high-fat diet group. The red dots represent significantly upregulated genes, and the blue dots represent downregulated genes, n = 3 per group. (B) There were 169 discrepant genes detected, including 39 upregulated and 130 downregulated genes. (C) GO term enrichment analysis. (D) Heatmap of top 20 differentially expressed genes in a comparison of HFD and CD groups. (E) Top 20 signaling pathways in the HFD and CD groups, as determined by KEGG.

HFD-induced changes in serum untargeted metabolomic profiles

Subsequently, the metabolomics of serum following HFD exposure was further analyzed using untargeted LC-MS/MS-based metabolomics, based on the same RNA sequencing mice. Orthogonal partial least squares discriminant analysis (OPLS-DA) was performed to obtain the OPLS-DA score chart reflecting the variety of metabolites between and within the groups. As shown in Fig. 3A, the OPLS-DA score showed a clear clustering of the six biological replicates and distinct differences between the HFD and CD groups. To better visualize the differential metabolites of the two subgroups, metabolites (upregulated and downregulated) that passed the filter criteria (fold change ≥ 2 or ≤0.5 and P value <0.05) were presented by a volcano plot (Fig. 3B). There were 55 differential metabolites between the HFD and CD groups and the top 10 most significantly up-or down-regulated metabolites were clustered and displayed as a heatmap (Fig. 3C). Notably, serum sphingosine-1-phosphate (S1P) levels were markedly elevated in the HFD group. The reason for this might be that soy, a natural plant source rich in sphingolipids from high-fat diets, may be responsible for the accumulation of S1P (Degagne et al., 2014; Fyrst et al., 2009; Wang et al., 2021). Sphingosine-1-phosphate (S1P) is a bioactive sphingolipid metabolite that has been shown to improve pancreatic β-cell function, enhance insulin secretion, and reduce apoptosis through activation of the ERK, p38-MAPK, and ROCK/Rho signaling pathways (He et al., 2021; Liu et al., 2021). Serum levels of 2-keto-glutaramic acid (αKG) and Chenodeoxycholic Acid (CDCA) were also significantly altered. αKG is a well-known multifunctional molecule that is involved in regulating cell growth, proliferation, senescence, and metabolic processes (Lushnikova et al., 2023). It has been reported that αKG showed a strong correlation with nutrient-regulated insulin secretion in pancreatic β-cells (Hoang & Joseph, 2020). Numerous studies indicate that αKG can also modulate mTOR activity and autophagy, although results are mixed (Yoon et al., 2017). Among the identified differential metabolites, CDCA was also associated with mTOR, which can activate the AKT-mTOR pathway, resulting in the disruption of autophagic flux and the accumulation of YAP in cancer cells (Zhao et al., 2022). Pathway enrichment analysis of the differential metabolites indicated predominant involvement in sphingolipid signaling, rheumatoid arthritis, glycolysis, vitamin digestion and absorption, neuroactive ligand−receptor interaction, calcium signaling, and phospholipase D signaling pathways (Fig. 3D).

Figure 3. HFD-induced changes in serum untargeted metabolomic profiles.

Serum untargeted metabolomic profile of the HFD and CD groups. (A) OPLS-DA score plot of serum metabolomic was utilized to display the differences between the samples, n = 6 per group. (B) Volcano plot of significant serum metabolites (P < 0.05 and FC ≥ 2 or ≤0.5) after HFD treatment. Metabolites up-regulated or down-regulated were shown in red and blue, respectively. (C) Heatmap of differentially expressed metabolites (the top 10 in upregulated metabolites and the top 10 in downregulated metabolites) between the HFD and CD groups. (D) Top 20 signaling pathways in the HFD and CD groups, as determined by KEGG.

Short-term exposure to a high-fat diet upregulated mTOR activity in C57BL/6N mice

To explore the potential relationship between differentially expressed genes and metabolites, a gene-metabolite network model integrating transcriptomics and metabolomics was constructed with ingenuity pathway analysis (IPA). The results were visualized as a network (Fig. 4A) and the central nodes in this network included the MAPK9, Caspase, CSN2A1, HARS, CARD11, and PLBD1, directly or indirectly interacting with genes and metabolites whose GO biological process functions were primarily correlated with proliferation, growth, apoptosis, and phospholipid catabolic process. Notably, mTOR, which was an important canonical pathway in this network, regulated cell growth, lipid, and protein synthesis via the phosphorylation of several ribosomal proteins (Saxton & Sabatini, 2017; Um, D’Alessio & Thomas, 2006). The mTOR pathway can be activated by various stimuli, including growth factors, hormones, chemokines, and nutrients (Chan et al., 2007). In particular, extracellular nutritional conditions, such as glucose, amino acids, and free fatty acids, regulate pancreatic β-cell mass and insulin secretion via the stimulation of mTOR signaling pathways (Asahara et al., 2022). A long-term high-fat feeding may result in an increase in mTOR protein activity that has already been reported in other studies (Chen et al., 2022; da Cruz et al., 2022). Thus, the mRNA expression of Mtor and its downstream genes in pancreatic β-cells was first determined by qRT-PCR. As shown in Fig. 4B, high-fat feeding mice presented higher Mtor mRNA levels as compared to those of normal chow diet feeding mice, while Rps6kb1, Rps6, Eif4ebp1, and Eif4b mRNA levels did not change. As a measure of mTOR activity, the phosphorylated level of ribosomal protein S6 (S6) was assessed by immunofluorescent staining and Western blot. Our results revealed that short-term HFD treatment promoted the phosphorylation of S6 (Figs. 4C, 4D), indicating increased activity of mTORC1. Moreover, the mRNA expression levels of Mki67 and cell cycle regulatory molecule (Cdk2) were also elevated in HFD mice (Fig. 4B). Hence, we speculated boldly that the promoting effect of high-fat diet on β-cell proliferation may be related to mTORC1 activity.

Figure 4. Short-term exposure to a high-fat diet upregulated mTOR activity in C57BL/6N mice.

Short-term exposure to a high-fat diet upregulated mTOR signaling in C57BL/6N mice. (A) Gene‐metabolite network analysis was constructed with ingenuity pathway analysis (IPA). (B) The mRNA expression of mTOR and mTOR downstream genes in the pancreatic β cells were detected by qRT-PCR, n = 3. (C) mTORC1 activity was measured by immunofluorescence labeling for phosphorylated S6 (pS6); scale bars, 20 µm. (D) mTORC1 activity was measured by western blot, n = 3. Data were presented as mean ± SEM. Student’s t-test: *P < 0.05.

mTOR inhibition resulted in a decreased proliferative response to HFD in $\text{β}$-cells

To further confirm whether the increase in β-cell proliferation induced by HFD was dependent upon the activation of the mTOR signaling pathway, C57BL/6N mice were intraperitoneally injected with rapamycin or vehicle daily with concurrent HFD feeding for 7 days. As we reported earlier (Li et al., 2016), daily treatment with 0.5 mg/kg rapamycin could effectively suppress mTORC1 activity in β-cells without altering systemic insulin sensitivity, therefore a dose of 0.5 mg/kg/day rapamycin was chosen for the subsequent experiments. After 7 days of rapamycin treatment, immunofluorescent staining and western blot analysis were carried out using the Anti-pS6 (Ser235/236) antibody to evaluate the inhibitory efficiency. Immunostaining for pS6 in the pancreas from rapamycin-treated HFD mice was partially abolished compared with vehicle-treated HFD mice (Fig. 5A), implying that mTORC1 activity was effectively suppressed. Corresponding to the immunofluorescent staining findings, western blot of islet lysates from rapamycin-treated HFD mice also showed decreased mTORC1 activity (Fig. 5B). Administration of rapamycin did neither change body weight or epididymal fat weight nor alter fasting blood glucose or random blood glucose in HFD mice, as shown in Figs. 5C–5F. To further investigate the impact of mTORC1 inhibition on glucose metabolism in vivo, the IPGTT and IPITT tests were performed. As predicted, the HFD+rapamycin mice and HFD+vehicle mice exhibited similar glucose tolerance and insulin sensitivity (Figs. 5G and 5H). With regard to insulin release and insulin content, no significant differences were observed between the two groups (Figs. 5I and 5J). However, the increased β-cell proliferation observed in HFD-fed mice was markedly attenuated by mTORC1 inhibition (Figs. 5K and 5L). Furthermore, rapamycin treatment reduced β-cell mass without altering individual β-cell size (Figs. 5M–5P). These findings indicate that mTORC1 signaling specifically regulates β-cell replication, highlighting its essential role in mediating the proliferative expansion of pancreatic β-cells in response to a short-term high-fat diet.

Figure 5. mTOR inhibition resulted in a decreased proliferative response to HFD in β cells.

Rapamycin decreased compensatory β cell proliferation induced by short-term high-fat feeding. (A and B) Immunofluorescent staining and western blot were performed to evaluate the mTORC1 inhibition efficiency; scale bars, 20 µm. (C and D) Body weight (C) and epididymal fat mass (D) were monitored in HFD + vehicle or HFD + rapamycin mice; n ≥ 3. (E and F) The changes in the fasting and random blood glucose levels were plotted; n = 6. (G–J) IPGTT, IPITT, plasma insulin during IPGTT, and insulin content were measured on day 6 following the vehicle or rapamycin injection; n = 6. (K) Immunofluorescence for insulin, ki67 and dapi in pancreatic sections; scale bars, 20 µm. (L) The rate of β cell proliferation was assessed quantitatively based on ki67/insulin double-positive cells; n = 3. (M and N) The β cell mass was analyzed in sections of pancreas by immunohistochemistry staining; scale bars, 50 µm, n = 6. (O and P) The size of pancreatic β-cells was determined by immunofluorescence staining; scale bars, 20 µm , n = 3. Data were presented as mean ± SEM. Student’s t-test: *P < 0.05, ***P < 0.001.

Discussion

The ability of pancreatic β-cells to expand in response to metabolic demands plays a crucial role in the development of type 2 diabetes. Studies conducted on rodents have demonstrated that there is a correlation between the increase of metabolic load, such as pregnancy and obesity, and the compensatory expansion of β-cell mass (Moullé, Ghislain & Poitout, 2017). This compensatory expansion is essential for maintaining glucose homeostasis and preventing, or even reversing, the onset of early-stage diabetes. However, this compensatory expansion is not permanent and eventually declines, leading to the development of diabetes. In rodents, β-cell proliferation is considered the primary mechanism for increasing pancreatic β-mass (Weir, Gaglia & Bonner-Weir, 2020). Thus, understanding the mechanisms underlying β-cell proliferation to metabolic load is critical to devising new strategies to prevent the progression to overt type 2 diabetes.

Compared to the long-term high blood glucose and high blood lipid levels causing damage to the pancreatic β-cells through glucolipotoxicity, short-term high glucose and lipid levels can stimulate the proliferation of pancreatic β-cells. We have reported that short-time high glucose infusion can stimulate pancreatic β-cell proliferation (Guan et al., 2016). Previous literature reported that 7 days high-fat diets can induce proliferation of pancreatic β-cells (Mosser et al., 2015; Stamateris et al., 2013). Nevertheless, limited studies have been done so far to explore the mechanism of 7 days high-fat diets in regulating the β-cell proliferation. Stamateris et al. (2013) demonstrated that early compensatory β-cell proliferation is associated with increased cyclin D2 protein expression. A glucokinase- and Irs2-independent pathway have also been shown to play an essential role in short-term HFD-induced β-cell proliferation (Kitao et al., 2018). In addition, Seferovic et al. (2018) found that HFD-mediated β-cell expansion is accompanied by significant changes in complex and global metabolite. However, how pancreatic β-cell perceive metabolic stress or changes in metabolic signals induced by 7 days of high-fat diet, and consequently promote proliferation, has not yet been elucidated.

Our experiments revealed that after 7 days of high-fat diet feeding, mice exhibited significantly increased body weight and epididymal fat mass compared to those on a normal diet. These findings indicate that short-term HFD exposure elevates metabolic load, which may serve as a stimulus for pancreatic β-cell proliferation. Indeed, β-cell mass increased from 0.8 to 1.4 mg over this period, reflecting a rapid and substantial adaptive expansion. To assess the impact of short-term HFD on insulin sensitivity, an insulin tolerance test was performed. The results showed no significant difference in insulin sensitivity between HFD-fed and control mice. Given that glucose levels are a key regulator of β-cell proliferation, we further evaluated fasting and random blood glucose levels. No significant differences were observed between the two groups in either parameter. However, as shown by the intraperitoneal glucose tolerance test, glucose tolerance was significantly impaired in HFD-fed mice, indicating early β-cell dysfunction despite compensatory mass expansion.

We propose several potential explanations for this phenomenon. First, the newly generated β-cells may be immature or in a state of hypofunction. Metabolomic changes indicate an imbalance in energy metabolism, such as enhanced fatty acid metabolism, which may interfere with glucose-sensing ability (Yaney & Corkey, 2003). RNA-seq data further shows that although no downregulation was observed in genes directly related to insulin synthesis and secretion (e.g., Mafa, Glut2), genes such as Bmp6 and Arhgdib were significantly downregulated after short-term high-fat diet. Bmp6 has been demonstrated to promote glucose homeostasis and insulin secretion in the T2D mouse model (Pauk et al., 2018). Arhgdib, as a member of the Rho (or ARH) protein family, is involved in multiple biological processes, including proliferation, cytoskeletal remodeling, and insulin release (Thamilselvan & Kowluru, 2019; Wang & Thurmond, 2009). Both genes have been reported to be closely associated with pancreatic β-cell function, supporting the possibility of functional impairment. Therefore, despite the significant increase in β-cell mass, their overall function did not improve synchronously and even exhibited decompensation, which may be the reason for the suboptimal results of IPGTT. Further in vivo glucose-stimulated insulin secretion experiments also confirmed this finding, demonstrating that short-term high-fat diet leads to mild impairment of pancreatic β-cell function.

Untargeted metabolomics revealed significant serum metabolic changes after short-term HFD, highlighting key bioactive metabolites in early β-cell adaptation. S1P, markedly elevated in HFD mice, promotes β-cell proliferation, survival, and function via ERK/p38-MAPK signaling (He et al., 2021; Liu et al., 2021). Serum αKG and CDCA were significantly altered by short-term HFD. αKG, a key metabolite linked to nutrient sensing and insulin secretion, modulates mTOR and autophagy (Lushnikova et al., 2023; Yoon et al., 2017). CDCA activates the AKT-mTOR pathway, impairing autophagic flux (Zhao et al., 2022). Both suggest potential roles in mTOR-driven β-cell adaptation.

Integrated transcriptomic and metabolomic analyses further highlight a critical role for the mTORC1 signaling pathway in short-term high-fat diet-induced pancreatic β-cell proliferation. mTORC1 complex is a central regulator of growth metabolism and homeostasis in response to external stimuli in cells. mTORC1 is activated on the surface of lysosomes in cells and phosphorylates downstream substrates such as S6K kinase, S6, 4EBP1, and transcription factor TFEB to promote cell synthesis metabolism. This process plays a crucial role in maintaining cellular homeostasis and responding to changes in the environment. In transgenic mouse models, the lack of raptor in β-cells results in reduced β-cell mass (Blandino-Rosano et al., 2017; Ni et al., 2017). Selective hyperactivation of mTORC1 in β-cells, achieved through β-cell-specific overexpression of Rheb or deletion of TSC1 or TSC2, leads to increased β-cell size and mass, accompanied by enhanced insulin secretion and improved glucose tolerance (Hamada et al., 2009; Rachdi et al., 2008; Xie & Herbert, 2012). Several reports have demonstrated that mTORC1 signaling serves as a master regulator of β-cell expansion and cell cycle progression by modulating cyclin D2/D3 levels and Cdk4 activity (Balcazar et al., 2009; Jaafar et al., 2019). Additionally, multiple studies have demonstrated that rapamycin, an mTOR inhibitor, exerts significant detrimental effects on β-cell function, survival, and peripheral insulin resistance (Barlow, Nicholson & Herbert, 2013; Fraenkel et al., 2008). These findings confirm that mTORC1 may be a key regulator of β-cell mass and function. Long-term HFD feeding increases metabolic stress and nutrient surplus signals in pancreatic β-cells, which may impair functional β-cell mass via mTORC1-dependent mechanisms (Aggarwal et al., 2022; Yuan et al., 2016). However, whether 7 days of HFD feeding can activate mTOR and thereby influence β-cell mass remains unclear. This was validated by our next experiments. qRT-PCR analysis revealed upregulated expression of genes associated with the mTORC1 signaling pathway. However, since mTORC1 signaling is primarily regulated at the post-translational level, we assessed mTORC1 activation by examining pS6 expression levels using immunofluorescence and western blot. Our results show that short-term high-fat diet activates the mTORC1 signaling pathway. To further confirm this finding, intervention experiments were conducted, and it was found that short-term high-fat diet-induced pancreatic β-cell proliferation could be inhibited by the mTORC1 signaling pathway inhibitor, rapamycin.

Studies have suggested that mTORC1 activation may be regulated by a number of factors, including nutrient availability, energy status, and cellular stress (Moullé, Ghislain & Poitout, 2017). Specifically, in β-cells, mTORC1 is inhibited under fasting and activated by glucose (Israeli et al., 2022). The effect of lipids, however, exerts a more complex and context-dependent influence on mTORC1 signaling. While saturated fatty acids like palmitate exhibit toxicity to β-cells, monounsaturated fatty acids, namely oleate, have been observed to promote β-cell proliferation. Researchers found that fatty acids can promote the proliferation of pancreatic β-cells in vitro under both low and high glucose conditions. Furthermore, an increase in metabolic stress load can promote the activation of the mTORC1 signaling pathway, and our experiments have confirmed this. Further exploration is needed to determine whether the activation of the mTORC1 signaling pathway in pancreatic β-cells, induced by a 7-day high-fat diet as observed in our experiments, is mediated through alterations in lipid metabolism or through metabolic stress. Understanding the role of metabolic stress and lipid metabolism in mTORC1 signaling could provide insights into the response of the β-cell to nutrient availability and energy balance. Additionally, this information could be valuable for developing new therapeutic strategies for diabetes.

Conclusion

In this study, a short-term high-fat diet was employed to induce pancreatic β-cell proliferation, and the pivotal role of the mTORC1 signaling pathway was identified. These findings shed new light on the mechanisms underlying the effects of short-term high-fat diets on pancreatic β-cells and may have important implications for the development of new treatments for diabetes.

Supplemental Information

Supplemental Information 3. Original dataset for Fig. 1.

A comprehensive range of measurements, including mouse body weight, blood glucose levels, epididymal fat pad weight, insulin content, insulin secretion profiles, results from intraperitoneal glucose tolerance tests (IPGTT), and intraperitoneal insulin tolerance tests (IPITT).

Supplemental Information 12. Original dataset for Fig. 4.

The original dataset for Fig. 4 consists of the raw data from qRT-PCR experiments.

Funding Statement

This study was supported by the National Key Research and Development Program of China (2021YFC2701900, 2021YFC2701903), National Natural Science Foundation of China (81770769), Master and Doctor innovation talent base for endocrine and metabolic diseases (RCJD2021S03), Pudong New Area Health Commission clinical characteristic discipline construction project (PWYts2021-13), and Shanghai Tongren Hospital Firefly program (TRYXJH06). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Additional Information and Declarations

Competing Interests

The authors declare that they have no competing interests.

Author Contributions

Jiajia Wang conceived and designed the experiments, performed the experiments, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Jing Li performed the experiments, prepared figures and/or tables, and approved the final draft.

Yunshan Li analyzed the data, prepared figures and/or tables, and approved the final draft.

Mengran Liu performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft.

Shan Huang conceived and designed the experiments, authored or reviewed drafts of the article, and approved the final draft.

Wenyi Li conceived and designed the experiments, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft.

Animal Ethics

The following information was supplied relating to ethical approvals (i.e., approving body and any reference numbers):

Animal care and experiments were approved by the Animal Ethics Committee, Tongren Hospital, Shanghai Jiao Tong University School of Medicine (2021-076-01).

DNA Deposition

The following information was supplied regarding the deposition of DNA sequences:

The raw sequence data are available at NCBI: PRJNA1199729.

Data Availability

The following information was supplied regarding data availability:

The raw data are available in the Supplemental Files.

The raw sequence data are available at NCBI: PRJNA1199729.