Human participants

Forty five patients with type 2 diabetes (male: 28, female: 17) and 20 healthy control individuals (male: 12, female: 8) were recruited from Beijing Anzhen Hospital. The clinical profiles of the diabetic patients are detailed in Supplementary Table 1. The study protocol received approval from the Ethics Committee of Beijing Anzhen Hospital (No. 2017005) and adhered to the principles of the Declaration of Helsinki. Written and verbal informed consents were obtained from all participants. Venous blood samples were collected upon admission without the use of anticoagulants. After centrifugation at 4 °C, plasma was promptly separated and stored at − 80 °C for subsequent analysis. To ensure data integrity, we excluded patients with a history of stroke, type 1 diabetes mellitus (T1DM), valvular heart disease, severe cardiovascular disease (CVD) classified as level III or IV by New York Heart Association criteria, significant hepatic or renal insufficiency, recent infectious diseases within the past two months, active liver disease, hemodialysis, malignancies, pregnancy, or hyperthyroidism.

Animal models

All experiments in this study were conducted in compliance with the NIH Guidelines for the Use of Laboratory Animals and were approved by the Capital Medical University Committee on Animal Care. APPL1 knockout mice (male APPL1−/−, 8 weeks old) and C57BL/6J mice were used in the study. The mice were randomly assigned to either a high-fat diet (HFD) (60% kcal fat, D12492i; Research Diets Inc.) or a standard diet (ND, D12450Bi) for 12 weeks to induce type 2 diabetes, as determined by fasting blood glucose levels (Figure S1). After 12 weeks, the mice were administered either vehicle or globular adiponectin treatment (0.25 μg/g/day intraperitoneally, PeproTech, Cranbury, NJ, United States) via a miniosmotic pump (ALZET, DURECT Corp, Cupertino, CA, United States) for an additional two weeks.

Packaging of ECAAV9-shOcln/ECAAV9-shFoxo1/ECAAV9-hOcln-Y467A

Packaging of ECAAV9-shOcln/ECAAV9-shFoxo1/ECAAV9-hOcln-Y467A: pAAV-Ocln/Foxo1 shRNAs were designed as follows: Ocln siRNA: 5ʹ-CCAUGGCAUACUCUUCCAATT-3ʹ; Foxo1 siRNA: 5ʹ-GAGCGUGCCCUACUUCAAGTT-3ʹ. The pAAV-hOcln-Y467A construct was based on NM_002538.3. The pECAAV9 helper plasmid (for packaging with SIGYPLP peptide) was developed by Beijing Likely Biotechnology Co., Ltd. To prepare retargeted AAV vectors, 24 plates of 293 cells at 80% confluence were co-transfected with 50 µg of pAAV-Ocln/Foxo1 shRNAs, adenovirus helper plasmid (pHelper), and pECAAV9 at a mass ratio of 2:2:1. After 60 h, cells were harvested, pelleted by centrifugation, and resuspended in 150 mM NaCl, 50 mM Tris–HCl (pH 8.5). The cells underwent multiple freeze–thaw cycles. Cell debris was cleared by centrifugation and filtration through a 0.45 µm PVDF filter. The viral supernatant was purified using ViraBind™ AAV Purification Kits (Cell BioLabs, VPK-141) and finally eluted with 1 ml of buffer. Eight-week-old C57BL/6J mice were administered a single dose of ECAAV9-scramble, ECAAV9-shOcln, or ECAAV9-shFoxo1 (5 × 1011 GC/mouse) via the retro-orbital route.

Cell apoptosis assay

Cell apoptosis was assessed using flow cytometry with a FITC-Annexin V Apoptosis Detection Kit (Yeasen Biotech). Cells were harvested using mild trypsin digestion. Double staining was performed with FITC-Annexin V and propidium iodide following the manufacturer’s protocol, and analysis was conducted using Gallios flow cytometry (Beckman Coulter). Data acquisition was performed using Kaluza for Gallios software (Beckman Coulter), collecting a minimum of 10,000 events per analysis. We categorized cells into living, dead, early apoptotic, and late apoptotic stages. The late apoptotic cells served as the target for our comparative analysis.

In vitro phosphorylation assays and mass spectrometry analysis

Phosphorylation reactions were conducted in a 50-μL reaction volume containing 1 × kinase buffer, 0.2 mmol/L ATP, 250 ng APN, 900 ng Occludin (Proteintech, Ag26173), and ddH2O. Reactions were started for 40 min at 30 °C and were terminated by the addition of 6 × sample loading dye (10 μL). Samples were heated for 5 min at 95 °C. Then, each reaction mixture was loaded onto 10% Tris–glycine gel. After electrophoresis followed by Coomassie bright blue stain, protein bands of interest were excised from the gel slab and processed for in-gel digestion by trypsin. Where specified, recovered tryptic peptides were analyzed. Recovered peptides from each protein were analyzed using nanoelectrospray tandem mass spectrometry on a Q-TOF hybrid mass spectrometer (QSTAR-Pulsar, Applied Biosystems/Sciex and Bruker-Daltonics AutoFlex TOF-TOF LIFT). The mass profiles of tryptic peptides were subsequently assessed by querying the protein sequence databases (NCBI Nonredundant Protein Database).

Small interfering RNA transfection

siRNA duplex oligonucleotides were designed to target specific gene sequences, effectively silencing the expression of APPL1 and Foxo1 genes. Human umbilical vein endothelial cells (HUVEC) were transfected using the Lipofectamine® 3000 Transfection Kit (Thermo Fisher Scientific, Inc) following the manufacturer's instructions, employing siRNA duplexes against APPL1 (5ʹ-UCUCACCUGACUUCGAAACUdTdT-3ʹ) and Foxo1 (5ʹ-GAGCGUGCCCUACUUCAAGTT-3ʹ), as well as universal control oligonucleotides (AllStars, Westerville, OH, USA). Cells were initially seeded in six-well plates and allowed to reach 80% confluence before siRNA transfection, with a final concentration of 50 nM applied to each well.

Chromatin immunoprecipitation-qPCR

Chromatin immunoprecipitation (ChIP) assays were conducted using an EZ-ChIP kit (Abcam, Cambridge, MA, USA) following the manufacturer's protocol. Briefly, three biological replicates of HUVECs were cross-linked with 1% paraformaldehyde and subsequently lysed in cold phosphate-buffered saline (PBS) containing a protease inhibitor cocktail. Chromatin was fragmented by sonication using a Covaris 220 instrument, with 10-s pulses interspersed with 30-s pauses, repeated 10 times. Immunoprecipitation was carried out overnight at 4 °C using 1 μg of anti-Foxo1 antibody (#2880, CST) and IgG control antibody (#2729, CST). Each immunoprecipitated sample was incubated with protein G agarose beads for 3 h at 4 °C. Following pull-down, beads were washed five times to remove non-specific binding. The immunocomplexes were then eluted and subjected to reverse cross-linking, proteinase K digestion, and DNA precipitation. The recovered DNA was analyzed by quantitative PCR (qPCR). Primers targeting the promoter regions of the Ocln genes were utilized to amplify both input and immunoprecipitated DNA (5ʹ-GAAGTGGGTGGGATTGGATAG-3ʹ). All experiments were conducted in triplicate from at least two independent experiments, and the data were normalized to percent input.

ChIP-qPCR data analysis

Normalize each ChIP DNA fraction's Ct value to the Input DNA fraction's Ct value for the same qPCR assay (ΔCt) to account for variations in chromatin sample preparation. Calculate the %Input for each ChIP fraction using the formula: %Input = 2^(CtInput—CtChIP) × Fd × 100%. Here, Fd represents the Input dilution factor. In our study, a 100 μL sonicated sample was used for ChIP, while a 20 μL sonicated sample served as Input, resulting in Fd = 1/5. To determine Fold Enrichment, use the formula: Fold Enrichment = [% (ChIP/Input)]/[% (IgG/Input)]. The fold enrichment value for the Vehicle group is standardized to “1”, with other groups being compared relative to the Vehicle group.

Transcription factor (TF) profiling array

The Transcription Factor Activation Profiling Plate Array II (Signosis, Sunnyvale, CA, USA) was utilized to assess the activity of transcription factors in HUVEC cells following the manufacturer's guidelines. Nuclear proteins from HUVEC cells were extracted using a Nuclear Extraction Kit (Signosis, Sunnyvale, CA, USA). Biotin-labeled probes, designed based on consensus sequences of transcription factor DNA-binding sites, were combined with 15 μg of nuclear protein extract to form transcription factor/probe complexes. The bound probes were then separated from the complex and hybridized to a plate pre-coated with sequences complementary to the probes. The captured DNA probe was detected using streptavidin-HRP, and the signal intensity was quantified using a microplate luminometer. Relative gene expression levels were determined using the ΔΔCT method, normalizing to the average expression level of five housekeeping genes.

RT2 Profiler™ PCR array for human transcription factors and quantitative real time-PCR (qRT-PCR) analysis

The Human Transcription Factors Array (Qiagen, USA) was utilized to assess the mRNA levels of human transcription factors in HUVEC cells in accordance with the manufacturer's instructions. Total RNA was extracted using the Trizol reagent method (Invitrogen) and employed for first-strand cDNA synthesis. qRT-PCR was conducted using RT2 SYBR Green Mastermix (PARN-026Z, QIAGEN) on a 7500 Real-Time PCR system (Thermo Fisher Scientific, Inc). The PCR protocol consisted of an initial denaturation step at 95 °C for 10 min, followed by 40 cycles of denaturation at 95 °C for 15 s and annealing/extension at 60 °C for 1 min. Gene expression levels were determined based on the CT values, and fold-changes in expression were calculated using the 2−ΔΔCT method. All samples were analyzed in triplicate.

Immunoprecipitation and Western blot analysis

Cells were rinsed once with PBS and subsequently lysed using cold 1 × lysis buffer supplemented with a protease inhibitor cocktail (Thermo Fisher Scientific, 78438). For immunoprecipitation, 1 μg of the appropriate control IgG was added to the cell lysate, followed by 20 μL of resuspended Protein A/G Plus-Agarose beads. The mixture was incubated at 4 °C for 30 min. The cleared lysate was then combined with normal immunoglobulin G (IgG) and anti-Occludin primary antibodies (Invitrogen, #71-1500), along with 15 μL of pre-washed Protein A beads, and incubated overnight at 4 °C. The immunoprecipitated proteins were subsequently released from the beads using an elution buffer. Samples were boiled, separated by SDS-PAGE, and analyzed via Western blotting.

Mouse aortic vascular tissues and cells were harvested and lysed to extract total protein. Protein concentrations were determined using the BCA Protein Assay Kit (Thermo Fisher Scientific, Inc. 23227). The total proteins were separated by gel electrophoresis and transferred onto a poly-vinylidene fluoride (PVDF) membrane. The membranes were then blocked with 5% nonfat milk for 1 h, followed by overnight incubation at 4 °C with primary antibodies. After washing, the membranes were incubated with HRP-conjugated secondary antibodies at room temperature for 1 h. Protein bands were visualized using the BioRad Imaging System. Antibodies against GAPDH (#5174), Phospho (#2981), and HA (#5017) were obtained from Cell Signaling Technology (Beverly, MA), while the antibody against Flag (Cat. No: F9291) was purchased from Sigma.

Confocal microscopic analysis

Endothelial cells were fixed in paraformaldehyde/PBS for 15 min and then washed with PBS. The cells were incubated with primary antibodies (rabbit anti-Occludin or mouse anti-LAMP2 antibody) at a 1:200 dilution. Subsequently, they were treated with either tetramethyl rhodamine (TRITC)-conjugated anti-mouse IgG or FITC-conjugated anti-rabbit IgG at a 1:100 dilution. After another wash with PBS, coverslips were mounted using an anti-fade solution (KPL, Gaithersburg, MD). Negative controls, which lacked primary antibodies, were also processed similarly. Fluorescent images were captured using Fluoview software (Olympus) on an FV3000 confocal microscope (Olympus, Tokyo, Japan).

Enzyme-linked immunosorbent assay (ELISA)

Blood samples were collected at the time of inclusion to measure Adiponectin and Occludin levels. Plasma samples obtained at enrolment were promptly stored at − 80 °C in a dedicated biologic resource center. The levels of plasma Adiponectin and Occludin were quantified using commercial ELISA kits (Cat #SEA605Mu, Cat#SEC228Hu, Cloud-Clone Corp) following the manufacturer's protocols. All sample preparations, as well as the utilization of reagents and buffers, strictly conformed to the manufacturer’s guidelines.

Mouse hindlimb ischemia model

Hindlimb ischemia (HLI) was induced by ligating the left femoral artery, starting from the distal point of its bifurcation down to the saphenous artery. Blood flow in the hindlimbs was assessed both before and immediately after ligation using laser Doppler flowmetry (LDF; PeriCam PSI). Blood flow was assessed and analyzed in a blinded manner. Mice exhibiting a reduction in blood flow of no less than 50% in the left hindlimb post-ligation were included in this study. For the mice with a successful ligation, another four measurements of blood flow were performed at day 0, 4, and 7.

Statistical analyses

Quantitative results are presented as mean ± SEM or median with interquartile range. Comparisons between two groups were assessed using an unpaired t-test. For comparisons involving more than two groups, one-way ANOVA or two-way ANOVA was employed, followed by post hoc analysis using the Tukey test. Prior to conducting parametric tests such as t-tests and ANOVA, assumptions of equal variance and normality were checked. Equal variance was examined using Bartlett’s test, while normality was assessed using the Kolmogorov–Smirnov test. In cases where data exhibited a normal distribution but had unequal variances, the Welch t-test or Brown-Forsythe and Welch ANOVA were utilized. For data that did not meet the normality assumption, the Kruskal–Wallis nonparametric test was applied. Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software, Inc, San Diego, CA, USA) and SPSS 24.0 (SPSS, Inc, Chicago, IL, USA). A p-value of less than 0.05 was considered statistically significant.