The beneficial effect of hepatic ER stress-associated protein PDI on obesity-associated glucose dysregulation

Abstract

Background

The liver plays critical roles in glucose metabolism regulation. Accumulating evidence supported that endoplasmic reticulum (ER) stress in liver tissue may involve in the development of type 2 diabetes. However, the role of ER stress-associated proteins in diabetes still needs to be clarified.

Methods

Genome-wide DNA methylome and proteome in the liver biopsies from patients with or without type 2 diabetes were performed and further validated by pyro-sequencing, real-time PCR and western blots. Circulating protein disulfide isomerase (PDI) levels at baseline and postoperative follow-up were measured by ELISA. The glucose tolerance, metabolic gene expression, glycogen deposition, glycogenesis and ER stress-associated proteins were detected in adeno-associated virus (AAV)-treated high fat-diet (HFD) mice.

Results

Based on methylome and proteome analysis, we identified the hypermethylation of PDI gene in liver biopsies, concomitant with decreased mRNA expression and protein levels in diabetic group compared with non-diabetic group. Circulating PDI levels were lower in patients with diabetes and elevated after metabolic surgery. The decreased PDI expression was correlated with increased gluconeogenesis and reduced glycogen synthesis in human liver tissue. Furthermore, hepatic PDI downregulation aggravated hyperglycemia, whereas PDI overexpression ameliorated glucose intolerance, decreased glycogen deposition and increased glycogenesis in HFD mice.

Conclusions

We identified the beneficial effect of ER stress-associated protein PDI on the regulation of hepatic glucose metabolism, which is expected to be a potential therapeutic target against type 2 diabetes, and provided an important clue for better understanding ER stress in the pathogenesis of diabetes.

Trial registration

The study was registered on ClinicalTrials.gov (NCT03296605).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12986-025-01041-9.

Background

The prevalence of diabetes worldwide is growing rapidly, which is closely associated with the epidemic of obesity and insulin resistance [1]. Metabolic-dysfunction associated steatotic liver disease (MASLD), previously known as nonalcoholic fatty liver disease (NAFLD) affects over 30% of adults worldwide and is associated with insulin resistance and type 2 diabetes [2]. The global prevalence of type 2 diabetes in individuals with MASLD is 28.3%, with an incidence density of 24.6 per 1000 person-years [3]. Thus, MASLD has emerged as a crucial target for preventing and managing type 2 diabetes.

The liver is the major contributor to endogenous glucose production via gluconeogenesis and glycogenolysis, and is the greatest reserve of glucose as glycogen. Dysregulation of hepatic regulation of glucose metabolism is a hallmark feature of MASLD [4]. Hepatic insulin resistance leads to the abnormality in gluconeogenesis and glycogen synthesis, and contributes to the development of type 2 diabetes [5], but very little is known about the underlying pathological mechanisms. Endoplasmic reticulum (ER) is an organelle in charge of protein synthesis, folding, modification, and transport [6]. Previous studies reported that ER stress has a versatile role on hepatic glucose metabolism [7, 8]. Sustained ER stress promotes hyperglycemia by increasing glucagon action and hepatic gluconeogenesis through the deubiquitinating enzyme ubiquitin-specific peptidase 14. Notably, the effect of hepatic ER stress in glucose metabolism was only studied in mice models [9], few studies addressing the issue in humans.

The pathogenesis of type 2 diabetes is multifactorial and influenced by both genetic and environmental factors. DNA methylation is an epigenetic mark reflecting the effects of environment [10]. Recent studies have suggested that epigenetic modification in the liver is associated with diabetes [11, 12], and methylome analysis was used as the tool to identify novel molecules linking to the disease [13]. In this study, we identified Protein Disulfide Isomerase (PDI), an ER-resident chaperone protein as a potential key molecule linked to hepatic glucose metabolism through genome-wide DNA methylome and proteome analysis. PDI plays a vital role in maintaining cellular homeostasis by facilitating proper protein folding [14, 15]. Recent research has demonstrated that PDI is involved in several cellular processes, including chronic liver inflammation and lipid metabolism [16, 17]. However, its specific role in glucose metabolism remains underexplored. The results will help us better characterize the role of hepatic ER stress contributing to type 2 diabetes and identify potential therapeutic targets.

Methods

Clinical study

The study was conducted in compliance with the guidelines of the Declaration of Helsinki and approved by the ethics committee of Nanjing Drum Tower Hospital. All individuals provided written informed consent and the study was registered on ClinicalTrials.gov (NCT03296605). In total, liver biopsies were collected from 66 obese patients during metabolic surgery and followed up for 6 months, from March 2016 to October 2018, in Drum Tower Hospital Affiliated to Nanjing University Medical School. Liver histology was assessed independently by two pathologists, according to the Nonalcoholic Steatohepatitis Clinical Research Network scoring system [18].

All enrolled participants underwent a 75 g oral glucose tolerance test (OGTT) and then divided to type 2 diabetes group and non-diabetic control group [19]. The exclusion criteria for the study were as follows: history of alcohol consumption (≥ 140 g/week for males or 70 g/week for females); history or current use of antidiabetic medication; other liver diseases, including chronic hepatitis B or C infection, biliary obstructive diseases and autoimmune hepatitis; pregnancy; type 1 diabetes mellitus; and malignant tumors.

Methylome array and proteomics

The liver biopsies from age, gender, BMI and stage of liver steatosis matched subjects with or without diabetes were used as the discovery cohort (10 in each), for methylome and proteomics analysis. The methylome was run on the Illumina Infinium Human Methylation 850 K BeadChip arrays (Illumina, San Diego, CA, USA) and analyzed by the limma package of R/Bioconductor (http://www.bioconductor.org) [20, 21]. The DNA methylation levels were then qualified by pyrosequencing in validation cohort (n = 46). iTRAQ labeling was performed using the iTRAQ Reagent-8plex Multiplex Kit (Applied Biosystems, Foster City, USA). We set a ± 1.2-fold change of the abundance ratio as the threshold with a p value < 0.05 to identify significantly different expression [22].

Immunohistochemical assays

Liver sections were probed with a primary antibody against protein disulfide isomerase (PDI) (ab3672, Abcam, Cambridge, USA). The images were captured using a Panoramic 250 Slide Scanner (3DHISTECH Ltd., Hungary), and brown staining of PDI was considered positive.

Clinical measurements

The plasma glucose concentration was measured using a hexokinase method (TBA-200FR, Tokyo, Japan). The level of hemoglobin A1c (HbA1c) was measured by HPLC (HLC-73G8, Tosoh, Japan). The serum triglyceride (TG), total cholesterol (TC), low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and gamma-glutamyltransferase (γ-GT) levels were detected using an autoanalyzer (Abbott Laboratories, Parsippany, USA). Fasting insulin (FINS) was detected by an electrochemiluminescent immunoassay (Roche). Insulin resistance (IR) was evaluated by a homeostasis model assessment of the IR index as fasting serum insulin (iU/mL) × fasting plasma glucose (mmol/L)/22.5. Plasma PDI level was measured with a commercial ELISA kit (Enzo Life, New York, USA) following the manufacturer’s instructions.

Animal experiment

Male C57BL/6 mice aged 7 weeks were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). All mice were housed at 21 °C ± 1 °C with a humidity of 55% ± 10% and a 12-hour light/dark cycle. The high fat-diet (HFD) contained 60% kcal from fat, 20% kcal from carbohydrate, and 20% kcal from protein. The normal diet contained 10% kcal from fat, 70% kcal from carbohydrate, and 20% kcal from protein. All animal studies were performed in accordance with the guidelines of the National Institutes of Health and were approved by Nanjing University Medical School Institutional Animal Care and Use Committee.

Model 1: 8-week-old male C57BL/6 mice were injected with AAV for PDI downregulation (sgPDI group, n = 6) or for control (sgCtl group, n = 6). The mice were kept on an HFD until the endpoint (8 weeks after vector administration). The sequence of sgRNA is TGTGGCATCCACCTTTGCTAG. Model 2: 8-week-old male C57BL/6 mice were injected with AAV for PDI overexpression (AAV-PDI group, n = 6) or for control (AAV-GFP group, n = 6). The mice were kept on an HFD until the endpoint (12 weeks after vector administration). All AAVs (AAV9 vector system) were purchased from GeneChem (Shanghai, China), and administrated by tail vein injection at a final dose of 2 × 10¹¹ genome copies per mouse. Glucose tolerance tests (GTTs) were performed by intraperitoneal injection of D-glucose (Sigma, St. Louis, MO, USA) at a dose of 2.0 mg/g body weight after a 8-hr fast (IPGTT). For insulin tolerance tests (ITTs), mice were injected with regular human insulin (Eli Lily, Indianapolis, Indiana, USA) at a dose of 0.75 U/kg body weight after an 8-hr fast. Liver tissues from overnight fasted mice collected 5 min after intravenous insulin (2 IU/mouse) injection. The amount of intracellular glycogen in the liver tissue of mice was then determined using a glycogen assay kit (Bio Vision, Milpitas, CA, USA).

Quantitative real-time PCR (qPCR)

Total RNA was extracted from liver tissues of human and mice using TRIzol Reagent (#15596018; Invitrogen, Carlsbad, CA, USA). RNA was reverse transcribed into cDNA using a PrimeScript RT Reagent Kit (TaKaRa Bio, Kusatsu, Shiga, Japan). QPCR was performed on Light Cycler 480 (Roche, Basel, Switzerland). The relative mRNA level for each sample was calculated according to the 2^−ΔΔCt method by normalization to β-actin [23]. The primers used are listed in Supplementary Table 3.

Western blot experiments

Western blot analyses were performed according to standard protocols. The following primary antibodies were used: anti-PDI (#3501), anti-p-protein kinase B (AKT, #4060), anti-t-AKT (#4685), anti-p- Glycogen synthase kinase-3β (GSK3β, Ser9, #5558), anti-t-GSK3β (#12456), and anti- Glyceraldehyde-3-phosphate dehydrogenase (GAPDH, #5174) and were purchased from Cell Signaling Technology (CST, Danvers, MA, USA; rabbit polyclonal, all diluted 1:1000).

Statistical analysis

Data with non-normal distributions were presented as medians with an interquartile range (IQR). The Mann-Whitney U test was used to test nonnormally distributed data between two groups. Kruskal-Wallis analysis was applied to compare nonnormally distributed data among three groups. Categorical data were analyzed using Chi-square tests or Fisher’s exact test as indicated. The association between the circulating PDI levels and the risk of diabetes was analyzed using the multivariate logistic regression method. The Spearman correlation coefficient was used to evaluate correlations between PDI expression and the expressions of genes involved in hepatic lipid metabolism. For animal experiments, data were analyzed by an unpaired t test or one-way analysis of variance (ANOVA). All analyses were performed using SPSS software (Version 20.0; SPSS Inc., USA). The significance level was set at p < 0.05, and p values were provided for two-sided tests.

Results

PDI was identified based on methylome and proteome data in the liver tissues from patients with diabetes

Anthropometric and clinical parameters of the discovery and replication cohort were presented in Supplementary Tables 1, and the study flow chart was shown in Figure S1. In combination analysis of methylome and proteome data on ER function pathway, Protein disulfide isomerase (PDI) was identified with an increased DNA methylation at the site cg09221994 (promoter region) and decreased protein expression in type 2 diabetes (Fig.S2A-B, Table S4). Then we performed pyrosequencing and qPCR for further validation, and found that the DNA methylation levels (at the cg09221994 site) were higher and mRNA levels of PDI were lower in the diabetic group than that in the non-diabetic group (p = 0.025 and 0.004, respectively) (Fig. 1A-B, S2C). Moreover, the DNA methylation levels at cg09221994 in the PDI gene were associated with the PDI mRNA levels (p = 0.007, r = 0.337, Fig. 1C), supporting hepatic PDI as a target for epigenetic modification. Consistently, the results of immunostaining and western blots showed that the protein expression of PDI was decreased in patients with diabetes compared with that in patients without diabetes (p<0.001, Fig. 1D -E).

Fig. 1.

Reduced hepatic PDI expression is associated with increased DNA methylation in MASLD subjects with T2DM. (A) The methylation levels of the cg09221994 site at the PDI gene in MASLD patients with (n = 18) or without (n = 30) T2DM. (B) Hepatic mRNA expression of PDI in patients with MASLD and T2DM (n = 28) or age- and BMI-matched nondiabetic patients (n = 38). (C) Spearman correlation analysis between the mRNA level of PDI with the methylation levels of the cg09221994 site. (D) Immunostaining for hepatic PDI in diabetic patients (n = 8) and age- and BMI-matched nondiabetic patients (n = 9). Top panel: immunohistochemical analysis. Bottom panel: immunofluorescence analysis. Scale bar = 100 μm. (E) Hepatic protein expression of PDI in diabetic patients (n = 3) and nondiabetic patients (n = 3). Statistical analyses were based on a t test. Data are presented as the mean ± standard error of the mean

Hepatic and circulating PDI were associated with glucose metabolism

Hepatic PDI mRNA expression was negatively associated with the fasting blood glucose (FBG, r=−0.440, p < 0.001), postprandial blood glucose (PBG, r=−0.381, p = 0.002), and HbA1c levels (r=−0.261, p = 0.042) (Fig. 2A-C). Moreover, hepatic PDI mRNA levels were negatively associated with the expression of gluconeogenesis genes, including glucose − 6- phosphatase catalytic subunit (G6PC) and phosphoenolpyruvate carboxykinase (PEPCK) (r=−0.539, p < 0.001; r=−0.262, p = 0.033, respectively), and positively associated with the expression of glycogen synthase (GS) (r = 0.308, p = 0.012) (Fig. 2D-F).

Fig. 2.

Decreased liver PDI expression is associated with hyperglycemia and impaired glucose metabolism in MASLD patients. (A - C) Correlation analysis between the mRNA level of PDI with (C) the methylation levels of the cg09221994 site or with glycemic parameters, including (D) FBG, (E) PBG, and (F) HbA1c. (D - F) Correlation analysis of the mRNA levels between PDI and (G) G6PC, (H) PEPCK and (I) GS. Spearman correlation analysis was used based on the PCR data of 28 diabetic and 38 nondiabetic patients with MASLD

Furthermore, spearman analysis indicated a positive association between hepatic PDI mRNA expression and circulating PDI levels (r = 0.337, p = 0.007) (Fig. 3A). Patients with diabetes had lower circulating PDI levels compared with patients without diabetes (7.65 ± 2.69 vs. 10.60 ± 3.25 ng/ml, p < 0.001) (Fig. 3B). In addition, the circulating PDI levels presented a negative association with FBG and HbA1c in enrolled subjects (Fig. 3C-D). Based on the tertiles of circulating PDI concentrations, enrolled patients were divided into three groups (T1, < 7.35 ng/ml; T2, 7.35–9.86 ng/ml; and T3, >9.86 ng/ml) (Table S5). A gradual decrease was noted in the prevalence of diabetes with the increased tertile of PDI (T1, 71.4%; T2, 38.1%; T3, 23.8%, p = 0.006). The odds ratios (ORs) (95% CIs) were significantly lower in the T2 (OR 0.218, 95% CI 0.055–0.854) and T3 (OR 0.126, 95% CI 0.030–0.526) groups than in the T1 group after adjusting for age, gender and BMI (Fig. 3E). Additionally, the ROC curve showed that the AUC is 0.763 for the PDI and the optimal cut point was about 7.0 ng/ml (sensitivity, 0.500; specificity 0.914, Fig. 3F), indicating plasma PDI had prediction performance in classifying patients with diabetes. We also determined the circulating PDI levels in patients with diabetes before and 6 months after metabolic surgery, the clinical characteristics are presented in Table S6. Circulating PDI levels were elevated after surgery (7.58 ± 2.90 ng/ml vs. 10.24 ± 3.00 ng/ml, p < 0.001). The postoperative change in circulating PDI level was correlated with the change in HbA1c (r=−0.485, p = 0.022).

Fig. 3.

Decreased circulating PDI levels are associated with an increased risk of T2DM in MASLD patients. (A) Spearman correlation analysis between the plasma levels of PDI with the hepatic mRNA expression of PDI. (B) Plasma PDI levels in diabetic (n = 28) and nondiabetic (n = 35) subjects. Spearman correlation analysis between the plasma levels of PDI with (C) FBG and (D) HbA1c. (E) Adjusted ORs of diabetes according to tertiles of plasma PDI levels in patients. The ORs with corresponding 95% confidence intervals (CIs) were adjusted for age, sex and BMI. The tertile ranges of T1, T2 and T3 of the plasma PDI level were < 7.35, 7.35–9.86, and > 9.86 ng/ml, respectively. T1 is the reference group. A binary logistic regression study was applied. (F) ROC curves of classification of participants with and without T2D

Hepatic PDI regulated glucose homeostasis in HFD-fed mice

In HFD-fed mice with PDI knockdown, the body weight and FBG were increased compared with that in the control mice (Fig. 4A-C, S3A-B). The areas under the IPGTT and ITT curves were significantly increased in mice with diminished PDI levels (Fig. 4D-E). Moreover, the mRNA levels of gluconeogenesis - targeted genes, e.g., Pepck and G6PC were increased, and the glycogen levels were decreased in the livers of HFD-fed mice with PDI knockdown (Fig. 4F-G). In addition, the p-AKT and p-GSK3β protein levels were reduced in the liver tissues of mice with diminished PDI levels (Fig. 4H).

Fig. 4.

Decreased hepatic PDI expression disrupts glucose homeostasis in HFD-fed mice. (A) Body weights, (B) circulating PDI levels and (C) FBG were measured, and (D) IPGTT and (E) ITT analyses were performed in HFD-fed mice with or without AAV against PDI. (F) Glycogen deposition and (G) the mRNA levels of G6PC and PEPCK were determined in the livers of AAV-administered mice and their controls. (H) Protein expression of PDI, p-AKT and p-GSK3β in HFD-fed mice with PDI silencing. Data are presented as the mean ± standard error of the mean (n = 6), ^*p < 0.05

We next investigated the effect of PDI overexpression on glucose metabolism (Fig. S4A-C). In PDI-overexpressing HFD-fed mice exhibited decreased body weight (Fig. 5A), improved insulin resistance (Fig. 5B-C), and lower fasting glucose (Fig. 5D). Moreover, PDI overexpression led to a significant reduction in the Pepck and G6PC mRNA levels, and an increase in the cellular glycogen levels (Fig. 5E-F). HFD decreased expression of p-AKT and p-GSK3β, and these effects were diminished by PDI overexpression (Fig. 5G). Collectively, these results indicated that hepatic PDI serve as a regulator in the homeostasis between glycogen synthesis and gluconeogenesis, improving glucose intolerance under HFD challenge.

Fig. 5.

PDI overexpression prevents the development of insulin resistance in HFD-challenged mice. (A) Body weights (B) IPGTT, (C) ITT and (D) FBG were measured in HFD-fed mice after PDI overexpression. (E) Glycogen deposition and (F) the mRNA levels of G6PC and PEPCK and (G) protein expression of PDI, p-AKT and p-GSK3β were determined in the livers of mice overexpressing PDI and in the controls

Discussion

ER stress is a prominent feature of MASLD, which promotes lipogenesis and IR. Evidence from animal studies showed that ER stress is involved in the development of type 2 diabetes [24, 25], however, the effects of the proteins involved in the ER stress signaling on glucose homeostasis remains complex and paradoxical. For example, the spliced form of X-box-binding protein-1 (XBP-1s) plays a critical role in alleviating ER stress by regulating the expression of genes involved in protein folding. Evidence showed that XBP1 decreases gluconeogenesis via suppressing FOXO1 activity [26]. However, other studies have reported aberrantly high levels of XBP-1 in obese livers. Overexpression of XBP-1 in the liver is associated with hyperglycemia, increased hepatic glucose production, and insulin resistance [27]. Thus, a better understanding of the mechanisms of ER stress governing hepatic glucose metabolism can identify novel therapeutic targets for type 2 diabetes.

In our study, using liver tissues from patients with or without diabetes, we performed a genome-wide methylome analysis, which is a tool to identify novel molecules linking to the disease. There were several studies performed methylome analysis to explore the linking of hepatic epigenetic-modification with diabetes [28–31], but few information regarding ER stress signaling was reported. Our study for the first time identified 124 differentially methylated regions mapping to 22 ER-associated genes in human livers in diabetes. Considering DNA methylation modify gene expression in response to environmental challenges [32], then we conduct combination analysis with proteome data in the same human cohort. PDI, the only gene expressed differently in the methylation and protein levels, was identified as a target.

PDI is an important protein responsible for the oxidative folding and chaperone-mediated quality control of protein processing, and its dysfunction serves as a vital contributor to the development of ER stress [14, 15]. Previous studies showed that PDI is a multifunctional protein associated with several diseases, including metabolic disorders, cardiovascular diseases, cancer and autoimmune diseases [33–36]. A very recent study demonstrated that hepatic PDIA1 critically regulates lipid homeostasis by differentially mediating the secretion of saturated and unsaturated fatty acid esters [16]. However, the role of PDI in diabetes was not fully understood. An animal study suggested that PDI may be essential in maintaining β cell function and viability [37]. Prior studies have highlighted the critical role of PDIA1 in β cell stress responses and its involvement in β cell function and viability during diabetes progression [38, 39]. In our study, the methylation levels of cg09221994 site the promoter region of PDI gene, as an epigenetic marker, were showed negatively associated with gene expression in the liver. Importantly, the expression level of hepatic PDI was negatively correlated with FBG and HbA1c. In addition, a low circulating level of PDI was associated with an increased risk of diabetes. After surgery with a 6-month follow up, the plasma PDI levels were elevated in diabetic group. These results indicated that PDI was tightly associated with glucose homeostasis.

Previous studies have demonstrated that the dysregulation of gluconeogenesis and glycogen synthesis in the liver plays a key role in the development of type 2 diabetes [5, 37]. Sustained ER stress induces the upregulation of the deubiquitinating enzyme ubiquitin-specific peptidase 14 (USP14), which promotes gluconeogenesis by deubiquitinating CBP/p300 and upregulating key gluconeogenic genes [7]. Ufmylation of UFBP1 ameliorated hepatic insulin resistance by enhancing glycogen synthesis, thereby alleviates hepatic ER stress [40]. Then, we investigate the mechanism underlying the effect of PDI on glucose metabolism. The decreased PDI expression was correlated with increased gluconeogenesis and reduced glycogen synthesis in the liver. In our AAV-mediated HFD-fed mice models, PDI downregulation aggravated the development of hyperglycemia and increased hepatic gluconeogenesis, while PDI overexpression ameliorated glucose intolerance and suppressed gluconeogenesis. These findings indicate that hepatic PDI, an ER stress-associated protein, regulates the balance between glycogen synthesis and gluconeogenesis, contributing to the improvement of glucose intolerance under HFD conditions.

There are several strengths in the present study. Firstly, we performed methylome and proteome with human liver tissues for identifying the target, and validated in another cohort. Secondly, the multidimensional investigation of PDI, including DNA methylation, mRNA, and protein levels, along with circulating levels, provided a comprehensive perspective on its role in diabetes. There are also some limitations. Firstly, we did not explore the precise mechanisms underlying the regulation of ER stress by PDI. Future studies should investigate these pathways to provide a more comprehensive understanding of PDI’s role in glucose metabolism. Secondly, the study was conducted in obese patients, limiting its applicability to lean individuals. Further studies including lean individuals would be necessary to validate the broader relevance of our findings. In addition, we did not directly assess hepatic de novo lipogenesis in this study; future studies incorporating direct measurements will be necessary to further clarify the relationship between glucose dysregulation and lipid metabolism.

In conclusion, we identified a novel role of ER stress-associated protein PDI in regulating hepatic glucose metabolism, expanding the knowledge of ER stress in the pathogenesis of diabetes and providing a new therapeutic target against type 2 diabetes.

Abbreviations

AKT

Protein kinase B

ER

Endoplasmic reticulum

FOXO1

Fork head box protein O1

FINS

Fasting insulin levels

HbA1c

Hemoglobin A1c

HFD

High fat-diet

HDL

High-density lipoprotein

γ-GT 

Gamma-glutamyltransferase

GTTs

Glucose tolerance tests

GSK3β

Glycogen synthase kinase-3β

G6PC

Glucose -6- phosphatase catalytic subunit

GAPDH

Glyceraldehyde-3-phosphate dehydrogenase

GS

Glycogen synthase

ITTs

Insulin tolerance tests

IPGTT

Intraperitoneal glucose tolerance test

IR

Insulin resistance

LDL

Low-density lipoprotein

MASLD

Metabolic-dysfunction associated steatotic liver disease

NAFLD

Nonalcoholic fatty liver disease

OGTT

Oral glucose tolerance test

PBG

Postprandial blood glucose

FBG

Fasting blood glucose

PEPCK

Phosphoenolpyruvate carboxykinase

PDI

Protein disulfide isomerase

TG

Triglyceride

TC

Total cholesterol

XBP1

X-box binding protein 1

Author contributions

DF contributed to the study design, data acquisition and interpretation, drafting of the article, and revision of the article. XY and PZZ contributed to data acquisition, data analysis, and the drafting and revision of the article. WHF contributed to the acquisition and analysis of the data and the drafting of the article. TH contributed to the study design, data interpretation, and the drafting and revision of the article. TWG contributed to the study design, data interpretation, and the revision of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China Grant Awards (82370841, 82401007), Nanjing Health Science and Technology Development Project (ZKX22027), the China International Medical Foundation (No.2023 -N -03 -01), Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024YFA1307001), the Natural Science Foundation of Jiangsu Province of China (BK20240232).

Data availability

Data regarding the manuscript are available upon reasonable request made to the corresponding author.

Declarations

Ethics approval and consent to participate

The study was conducted in compliance with the guidelines of the Declaration of Helsinki and approved by the ethics committee of Nanjing Drum Tower Hospital. All individuals provided written informed consent and the study was registered on ClinicalTrials.gov (NCT03296605).

Competing interests

The authors declare no competing interests.