Glyoxal induces arterial injury through the inactivation of receptors and signaling pathways Authors: Ming-Zhang Xie, Xue Yang, Zhi-Wei Wu, Lin-Ping Zhou, Hai-Xia Chu, Qun-Mei Zhang, Yu-Chao Zhang, Xiao-Chun Zhou, Meng-Meng Zhao, Qing-Jiang Mo, Xiao-Fang Wang, Wei-Dong Xv, Wen-Yan Bian, Ya-Nan Li Published: 2026 ABSTRACT Glyoxal exposure is associated with a spectrum of adverse health outcomes, including arterial injury. While existing evidence confirms that glyoxal targets oxidative stress and the mitogen-activated protein kinase pathway, the precise underlying mechanism of glyoxal-induced arterial damage remains elusive. In this study, we aimed to elucidate the molecular mechanism driving glyoxal-induced arterial injury through a combination of in vitro and in vivo experiments. In vitro, we treated human aortic endothelial cells with glyoxal, then performed cell viability assays, tandem mass tag (TMT)-based quantitative proteomics, parallel reaction monitoring (PRM), glutathione quantification for oxidative stress assessment, western blotting, and quantitative polymerase chain reaction (qPCR) analysis. In vivo, we administered glyoxal to C57BL/6 mice, after which we conducted biochemical, histological, and immunohistochemical assays to evaluate arterial pathological changes. In vitro, the proteomic profiling results were further validated via PRM and western blotting, which confirmed that glyoxal exposure induces energy metabolism dysfunction, aberrant vascular endothelial growth factor receptor (VEGF-R) expression, and signaling pathway inactivation. Consistent with the proteomic results, we detected reduced protein levels of VEGF-R (36.4% reduction), insulin-like growth factor receptor (IGF-R, 21.6% reduction ), and insulin receptor (INS-R, 16.4% reduction) relative to control after glyoxal treatment. Compared to the control group, the glyoxal-exposed group exhibited a 48.7% reduction in the phosphorylation levels of ErbB2 and ErbB4, while total protein levels of ErbB2 and ErbB4 (106.0% of the control level) were comparable between the two groups. In addition, the levels of Wnt signaling-related proteins—including β-catenin (52.5% of the control level) and Wnt ligands—were significantly decreased following glyoxal treatment, relative to the control. Consistent with our in vitro results, we observed that the protein levels of VEGF-R (47.7% reduction), β-catenin (25.7% reduction), and INS-R (66.0% reduction) were decreased relative to the control group after in vivo glyoxal treatment. Collectively, our findings outline a comprehensive molecular portrait of glyoxal-induced arterial injury, which provides a foundational framework that may advance the development of precision medicine strategies for glyoxal-related vascular disease.Graphical abstract Our previous study demonstrated that oxidative stress and MAPK pathways are important targets of glyoxal (GX); however, the specific mechanism is not completely understood at present. Therefore, this study investigates the AI mechanism induced by GX based on the proteomics. These novel mechanisms include INS-R, IGF-R, and VEGF-R damage and systematic changes in the Wnt and ErbB signaling pathways. These injuries correspond to abnormal energy metabolism and function of HAECs, causing aberrant vascular function, inflammation, and abnormal vascular structure.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s11010-026-05719-3. Introduction Hyperglycemia, a metabolic endocrine disorder characterized by high blood sugar levels, has emerged as a major global public health concern. Hyperglycemia serves as the primary clinical manifestation of diabetes globally [1, 2]. The leading cause of disability and death in patients with diabetes is cardiovascular complications, including major vascular complications, such as cerebrovascular, cardiovascular, and lower limb vascular lesions, and microvascular complications, such as kidney and fundus lesions. The dialdehyde glyoxal (GX) is a colorless or pale yellow crystalline solid. In the body, methylglyoxal and GX are produced as a result of metabolic disorders, such as high blood sugar levels in patients with diabetes [2]. Exogenous sources of GX include processed foods, such as barbecued, fermented, and fried foods [1]. Endogenous GX is primarily produced under hyperglycemia or after sugar intake, and high levels of endogenous GX can cause critical vascular damage [3] including signal transduction pathway activation, DNA injury, oxidative damage, immune damage, and abnormal lipid metabolism, in patients with diabetes [3, 4]. Our previous study demonstrated that oxidative stress and the mitogen-activated protein kinase pathway are crucial targets of GX [3]; however, the underlying mechanism has only been partially elucidated. Therefore, in this study, we aimed to investigate the mechanism of GX-induced arterial injury (AI) using proteomic techniques and determine its potential as a therapeutic target to advance precision medicine, particularly in patients with diabetes. Methods Materials Human aortic endothelial cells (HAECs; BNCC337727) were procured from BeNa Culture Collection, Beijing, China. Microtubule-associated protein light chain 3-II (LC3-II; ab192890), glyceraldehyde-3-phosphate dehydrogenase (GAPDH; ab181602), Wnt3 (ab172612), Wnt5 (ab235966), insulin-like growth factor receptor (IGF-R; ab131476), phospho-ErbB2 and phospho-ErbB4 (ab201013), and ErbB2 and ErbB4 (ab219208) antibodies were purchased from Abcam (Cambridge, UK). Vascular endothelial growth factor receptor (VEGF-R; 340401) was procured from Zen-Bioscience Technology Co., Ltd. (Chengdu, China). Insulin-like growth factor receptor (INS-R; 17509-1-AP), Wnt1 (27935-1-AP), Wnt2 (66656-1-Ig), Wnt6 (24201-1-AP), Wnt7 (10605-1-AP), CD31 (11265-1-AP), and β-catenin (51067-2-AP) antibodies were purchased from Proteintech Group (Rosemont, IL, USA). C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). HAEC culture and survival assays We cultured HAECs with or without GX (0, 50, 100, 150, and 200 µM) in Dulbecco’s modified Eagle’s medium containing 10% inactivated fetal bovine serum, as described previously [3], and counted them (50/colony). Doses that induced the death of 50% of the cells (LD50) were determined using survival curves (Online Resource Fig. 1) [3]. Tandem mass tag-based proteomic analysis Briefly, HAECs were treated with GX (median lethal dose [LD50] = 51.3 µM); control cells were not exposed to GX. Proteomic analysis was conducted following a previously described method [5]. Protein preparation and TMT labeling All experimental samples were fully homogenized in pre-prepared lysis buffer (the composition of the buffer was 4% SDS, 1 mM DTT, 150 mM Tris–HCl with pH adjusted to 8.0, supplemented with protease inhibitor cocktail to prevent protein degradation). After homogenization, the mixture was heat-treated in boiling water for 5 min, followed by ice-bath sonication to fully disrupt cells and release proteins. Subsequently, the crude extract was subjected to secondary heat treatment in boiling water, then centrifuged at 16,000×g under 25 °C for 10 min to remove precipitated impurities. After measuring the protein concentration of the obtained supernatant, all samples were stored in a −80 °C ultra-low temperature refrigerator for subsequent labeling experiments. For TMT labeling pretreatment, 200 µg of protein from each independent sample was mixed with 30 µL of SDT buffer (formula: 4% SDS, 100 mM DTT, 150 mM Tris-HCl, pH 8.0). To remove detergent, DTT and other low-molecular-weight impurities, repeated ultrafiltration was performed with UA buffer (8 M Urea, 150 mM Tris–HCl, pH 8.0) using a 10 kD Pall ultrafiltration unit. After purification, 100 µL of 0.05 M iodoacetamide prepared with UA buffer was added to the system to block the reduced cysteine residues, and the reaction was carried out at room temperature in the dark for 20 min to avoid iodoacetamide inactivation. Next, the ultrafiltration membrane was washed three times with 100 µL UA buffer, then washed twice with 100 µL DS buffer (50 mM triethylammonium bicarbonate, pH 8.5) to replace the buffer system. Finally, the treated protein was re-suspended in 40 µL DS buffer, digested with 2 µg of sequencing-grade trypsin (purchased from Promega) at 37 °C overnight. The resulting peptide fragments were collected as filtrate after digestion, and then lyophilized by vacuum centrifugation for subsequent TMT labeling. Peptide fractionation via strong cation exchange chromatography Separation of the digested peptides was carried out using strong cation exchange (SCX) chromatography with the AKTA Purifier modular high-performance liquid chromatography system (manufactured by GE Healthcare, Illinois, USA). The mobile phase system used for gradient elution was composed of Buffer A (10 mM KH2PO4 and 25% acetonitrile, pH 2.7, no additional potassium chloride) and Buffer B (500 mM KCl, 10 mM KH2PO4, 25% acetonitrile, pH 2.7). Gradient elution was implemented at a constant flow rate of 1 mL/min following the gradient procedure: 0–10% Buffer B for 2 min, 10–20% Buffer B for 25 min, 20–45% Buffer B for 5 min, and 45–100% Buffer B for 5 min. Eluted peptide fractions were collected automatically following the standard operational protocol provided by the instrument manufacturer. All collected peptide fractions were then merged and desalted using C18 solid-phase extraction cartridges with standard bed density (Empore SPE, specification 7 mm/3 mL, purchased from Sigma-Aldrich, Missouri, USA). After desalting, the desalted peptides were concentrated via vacuum centrifugation, and finally re-dissolved in a dilute trifluoroacetic acid solution for subsequent mass spectrometry analysis. Liquid chromatography-electrospray ionization tandem mass spectrometry analysis Nano-liquid chromatography tandem mass spectrometry analysis was performed on a Q Exactive HF-X orbital trap mass spectrometer, which was coupled online with an Easy nLC nano-flow liquid chromatography system (both purchased from Thermo Fisher Scientific, Massachusetts, USA). For each separated SCX fraction, 5 µL of the reconstituted peptide solution was loaded for nanoLC-MS/MS data acquisition. The peptide mixture was first loaded onto an in-house packed C18 reversed-phase analytical column (specifications: 15 cm in length, 75 μm inner diameter, filled with 5 μm RP-C18 resin) equilibrated with 0.1% formic acid aqueous solution. Peptide separation was carried out with a linear gradient of mobile phase B (80% acetonitrile supplemented with 0.1% formic acid), and the entire gradient separation procedure lasted for 90 min, with a constant flow rate of 250 nL/min maintained by IntelliFlow automatic flow control technology. MS/MS data were collected with the data-dependent acquisition mode, in which the top 20 most abundant precursor ions were automatically selected from each full scan survey (mass range set to 300–1800 m/z) for higher-energy collisional dissociation (HCD) fragmentation. The target ion abundance for fragmentation was determined by predictive automatic gain control (pAGC), and the dynamic exclusion duration was set to 60 s to avoid repeated sequencing of the same precursor ion. Sequence database searching and data processing Raw mass spectrometry data was searched against protein sequence databases for peptide identification using Mascot version 2.2 (developed by Matrix Science, London, UK) integrated in Proteome Discoverer version 1.4 (from ThermoFisher Scientific). The target database was the UniProt Mus musculus (house mouse) reference protein database, which was downloaded on March 26th 2020 and contained 87,874 sequences in total. A reversed decoy sequence database was constructed to evaluate the false discovery rate of identification. The key parameters for database searching were set as follows: precursor mass tolerance was set to 20 ppm, tandem MS fragment ion mass tolerance was set to 0.1 Da; the cleavage enzyme was specified as trypsin, allowing up to 2 missed cleavage sites; fixed modifications were set as carbamidomethylation on cysteine residues, and TMT16 labeling on lysine residues and peptide N-termini; methionine oxidation was set as a variable modification. Only peptide and protein identifications with a false discovery rate (FDR) ≤ 0.01 were retained for subsequent analysis. Bioinformatic analysis Functional annotation and enrichment analysis were performed with the following workflow: Gene Ontology (GO) annotation results were obtained and GO annotation histograms were plotted using the Blast2GO online bioinformatics platform (available at https://www.blast2go.com/). For pathway-level functional analysis, protein sequences were mapped to reference canonical pathways using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, and pathway annotation was completed via the KEGG Automatic Annotation Server (KAAS, accessible at http://www.genome.jp/kaas-bin/kaas_main). Oxidative stress detection based on glutathione level The biochemical principle underlying this protocol is as follows using a GSH detection kit (A006-1-1; Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China) as previously described [5]: dithionitrobenzoic acid reacts with sulfhydryl compounds to produce a yellow-colored product, and this reaction is catalyzed by reduced glutathione (GSH). The specific experimental procedure is outlined below: After treatment with GX at the LD₅₀ dose, HAECs (1 × 106 cells) were harvested, resuspended, and subjected to ultrasonic lysis. The cell lysate samples were mixed with Reagent 1, and 1 mM GSH was also mixed with Reagent 1 at the same time, followed by incubation at 37 °C for 5 min. Subsequently, a mixed working solution containing Reagents 2 to 4 was added to the reaction system, and the mixture was incubated at 37 °C for an additional 15 min. Finally, GSH concentration was detected using an absorbance microplate reader (Molecular Devices, San Jose, CA, USA) at a wavelength of 420 nm. Western blotting After incubating HAECs without GX and chloroquine (CQ, control), with GX (LD50) alone, or with both GX (LD50) and CQ (20 µmol/L; GX-CQ), the protein expression levels of IGF-R, LC3-II, VEGF-R, INS-R, Wnt1, Wnt2, Wnt3, Wnt5, Wnt6, Wnt7, β-catenin, and phospho-ErbB (30 µg), and total protein levels (ErbB2 and ErbB4) were determined following a previously described method [5]. HAECs were washed with phosphate-buffered saline (PBS) and harvested following treatment with the protein inhibitor (purchased from Hefei, China). 30 mg of total protein was separated on a 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) resolving gel. The separated proteins were then transferred to polyvinylidene difluoride (PVDF) membranes using a Bio-Rad blotting system. After membrane transfer, the membranes were blocked with 5% bovine serum albumin in TBST buffer (20 mM Tris–HCl, pH 7.6, 140 mM NaCl, 0.1% Tween-20) for 2 h at 37 °C, followed by overnight incubation with the corresponding primary antibody at 4 °C. After three washes with TBST, the membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG secondary antibody (dilution 1:5000) for 2 h at 37 °C. Following an additional three washes with TBST, the membranes were incubated with enhanced chemiluminescence (ECL) western blotting substrate. Chemiluminescence signals were captured and quantified using the ChemiDoc XRS+ system (Bio-Rad, Hercules, CA, USA), and data analysis was performed with ImageJ software (version 1.47). Quantitative polymerase chain reaction Total RNA was extracted from HAECs divided into two groups: an untreated control group and a GX-treated group (LD50). Extraction was conducted strictly following the manufacturer’s instructions for the TRIzol reagent kit (Thermo Fisher Scientific, Waltham, MA, USA). Briefly, cells were washed twice with PBS, then lysed with TRIzol, followed by chloroform extraction, isopropanol precipitation, and 75% ethanol washing before RNA concentration quantification. First-strand complementary DNA (cDNA) was synthesized via reverse transcription according to the manufacturer’s protocol of the reverse transcription kit from TIANGEN BIOTECH (Beijing, China) Co., Ltd. The reverse transcription reaction was performed under the following conditions: 15 min at 42 °C, followed by 3 min at 95 °C to inactivate the reverse transcriptase. The resulting cDNA was stored at −20 °C until further use. Quantitative real-time polymerase chain reaction (qRT-PCR) primers for target genes were designed with Primer 5.0 software and synthesized by GenePharma Co., Ltd. (Suzhou, China). Detailed primer sequences are listed in Online Resource Table 1. The qRT-PCR was carried out with a reaction system, which contained SYBR Premix Ex Taq, 2.5 µL each of 10 µmol/L forward and reverse primers, 2 µL of cDNA template, and 18 µL of nuclease-free sterile water. The thermal cycling program was set as follows: initial pre-denaturation for 10 s at 95 °C, followed by 40 cycles of 5 s denaturation at 95 °C, 20 s annealing at 61 °C, and 10 s extension at 72 °C. The relative mRNA level of target genes was calculated using the comparative threshold cycle (Ct) method (2-ΔΔCt) and normalized to the corresponding GAPDH control. Animal experiments The Xinxiang Medical University Institutional Animal Care and Use Committee approved the study (Approval ID: XYYFY-2022-021), and all experiments were performed following the ARRIVE guidelines. An mouse model was developed using 7–11-week-old male C57BL/6 mice. Female mice were not used because estrogen can protect against atherosclerosis and other similar vascular damage. Three mice each were housed in 29 × 17.8 × 16 cm cages (JBS-NY-PL01, JOYBOS Inc., Shanxi, China) with an adequate supply of food and water, and 18 mice each were randomly assigned to the GX and control groups. GX doses were determined according to the concentration used in experiments on minimizing the likelihood of liver and kidney damage. Animals in the treatment group received intraperitoneal injection of GX every 2 days for 16 weeks, with the dosage adjusted based on individual body weight; the control group animals were injected with only normal saline for 16 weeks. Thereafter, blood, common artery, liver, lung, and kidney samples were obtained with the mice under anesthesia. We conducted a biochemical comparison between the GX-treated experimental and untreated control groups to evaluate the effects of GX. Briefly, we quantified K, Na, and Cl using potentiometry with an ion-selective electrode. The levels of alanine transaminase, aspartate transaminase, creatine kinase, glutamine, blood urea nitrogen, creatine, and protein- and lipid-related indicators were determined via colorimetry. Immunohistochemistry The mice were euthanized via cervical detachment. The dissected abdominal aortae were fixed, permeabilized with 1,000× diluted Triton X-100, rinsed with PBS at each step, and sectioned. The sections were mounted on slides and incubated with H2O2, and 5% bovine serum was used to reduce background noise. The sections were incubated with 1:50 diluted primary antibodies in PBS for at least 12 h (4 °C), followed by incubation with 1:300 diluted secondary antibodies (in PBS) for 2 h. Subsequently, the sections were reacted with 3,3′-diaminobenzidine substrate and washed. Protein levels were determined using ImageJ version 1.47 (National Institutes of Health, Bethesda, MD, USA; https://imagej.net/ij/, 1997–2018). Parallel reaction monitoring Diagnostic and therapeutic targets were confirmed using parallel reaction monitoring (PRM) analysis. The target peptides was prepared following the methodology described in “Tandem mass tag-based proteomic analysis ” and the label-free protocol was used to prepare 72 peptides. The stable isotope AQUA peptide was used to prepare the standard internal reference. Tryptic peptides were loaded onto the stage tips of C18 columns for desalting prior to reversed-phase chromatography on an nLC-1200 Easy system (Thermo Scientific). Subsequently, liquid chromatography was performed with 5%–35% acetonitrile gradients for 45 min, and Q Exactive Plus MS was used for conducting PRM analysis. An optimized protocol was used to process data on the energy of collision, state of charge, and retention time of the target peptides, including unique peptides with high intensities. The Skyline application (MacCoss Lab Software, Seattle, WA, USA) was used to analyze raw data, in which the intensity of the signal produced by a certain peptide sequence can be quantified according to each sample and referenced to standards via normalization of each protein. Statistical analysis All data were analyzed using IBM Statistics for Windows (version 22.0; IBM Corp., Armonk, NY, USA). The results were obtained from three to five repetitions (mean ± standard deviation) (SD). Comparisons between two groups were performed using the two-tailed Student’s t-test. For comparisons among three or more groups, one-way analysis of variance (one-way ANOVA) was applied, followed by Bonferroni correction for multiple comparisons. Categorical variables were analyzed using the chi-square test or Fisher’s exact test when appropriate. GO and KEGG enrichment analyses were evaluated using Fisher’s exact test. Analyses were performed on GraphPad Prism 9.0.0 for Windows (GraphPad Software). P < 0.05 was considered statistically significant. Results Determination of LD50 based on HAEC survival HAEC colonies (50/colony) were counted, and the dose that induced the death of 50% of cells (LD50 = 51.3 µM) was determined from survival curves (Online Resource Fig. 1). Proteomics-based analysis of HAECs Protein expression in HAECs was assessed after treatment with (LD50) or without (control) GX using liquid chromatography–tandem mass spectrometry. The results revealed that the expression of 104 and 142 (total 246) differentially expressed proteins (DEPs) was upregulated and downregulated, respectively, with fold-changes (FCs) of > 1.5 or < 0.67 compared with those in the controls, FCs of key DEPs were verified using PRM analysis (Online Resource Table 2). KEGG pathway and GO term analyses To investigate the biological functions and potential mechanisms underlying GX-induced cellular alterations, DEPs were analyzed via KEGG pathway enrichment and GO enrichment analyses. Several pathways were enriched, with protein export, cysteine and methionine metabolism, complement and coagulation cascades, and estrogen signaling pathways representing affected pathways (Fig. 1a). Consistent with the pathway enrichment analysis, statistical evaluation of the enriched KEGG pathways revealed that protein export and cysteine and methionine metabolism showed significance, with additional enrichment observed in complement and coagulation cascades, estrogen signaling, and other biological pathways (Fig. 1b). Fig. 1Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of key proteins. a The ordinate is the specific pathway classification and name; the left side shows the link of the selected signaling pathway in the database, its p-value, and the number of associated genes; the right side shows their corresponding KEGG classification and the horizontal coordinate is the percentage of genes. b KEGG pathways. The red and blue lines indicate p-values of < 0.01 and < 0.05 (n = 3), respectively Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of key proteins. a The ordinate is the specific pathway classification and name; the left side shows the link of the selected signaling pathway in the database, its p-value, and the number of associated genes; the right side shows their corresponding KEGG classification and the horizontal coordinate is the percentage of genes. b KEGG pathways. The red and blue lines indicate p-values of < 0.01 and < 0.05 (n = 3), respectively To further characterize the functional significance of the DEPs identified after GX treatment, GO enrichment analysis was performed. In the biological process category, the most significantly enriched terms included multicellular organismal process, cellular component organization or biogenesis, developmental process, and negative regulation of biological process (Fig. 2a). These results indicate that GX-induced proteomic alterations are associated with broad changes in cellular organization, development-related processes, and biological regulation. In the cellular component category, the enriched GO terms were primarily associated with intracellular structures and protein localization, including organelle part, intracellular part, organelle lumen, membrane-enclosed lumen, cytoplasm, and intracellular organelle part (Fig. 2b). This suggests that GX treatment affects proteins involved in diverse cellular compartments and structural components. For the molecular function category, the most significantly enriched terms were binding, protein binding, and macromolecular complex binding, followed by enzyme binding, receptor binding, and RNA-related binding functions (Fig. 2c). These findings indicate that the DEPs are mainly involved in protein interaction networks and regulatory molecular functions. Together, GO enrichment analysis demonstrates that GX induces extensive proteomic changes affecting cellular processes, protein localization, and molecular interactions, providing a basis for further investigation of specific signaling pathways involved in GX-mediated endothelial dysfunction. Fig. 2Gene Ontology term enrichment. The relationship between the top 10 processes and the percentage of protein enrichment is shown. The y-axis is arranged according to the maximum level from low to high, where each level is arranged according to the p-value, and the x-axis shows the percentage of protein enrichment. a Biological processes, b cellular components, and c molecular functions (n = 3) Gene Ontology term enrichment. The relationship between the top 10 processes and the percentage of protein enrichment is shown. The y-axis is arranged according to the maximum level from low to high, where each level is arranged according to the p-value, and the x-axis shows the percentage of protein enrichment. a Biological processes, b cellular components, and c molecular functions (n = 3) Abnormal energy metabolism and autophagy are associated with changes in IGF-R and INS-R expression The key DEPs indicated that oxidative stress occurred following GX treatment (Online Resource Tables 2 and 3). The GSH level in the treatment group decreased by 58% (Fig. 3a) compared with that in the group not treated with GX. Fig. 3Energy metabolism abnormalities and changes in VEGF-R expression induced by GX and autophagy in HAECs. The results were based on a reduced GSH level and b and c changes in VEGF-R, INS-R, and IGF-R protein expression induced by GX; autophagy was inhibited by CQ. d The expression of VEGF-R, IGF-R, and INS-R was altered by GX. e LC3-II levels confirmed GX-induced autophagy in HAECs. n = 3 and *p < 0.05. CA, chloroquine; GX, glyoxal; GX-CQ, glyoxal with chloroquine; GSH, glutathione; HAEC, human aortic endothelial cell; IGF-R, insulin-like growth factor 1 receptor; INS-R, insulin-like growth factor receptor; LD, lethal dose; VEGF, vascular endothelial growth factor; WT, wild-type Energy metabolism abnormalities and changes in VEGF-R expression induced by GX and autophagy in HAECs. The results were based on a reduced GSH level and b and c changes in VEGF-R, INS-R, and IGF-R protein expression induced by GX; autophagy was inhibited by CQ. d The expression of VEGF-R, IGF-R, and INS-R was altered by GX. e LC3-II levels confirmed GX-induced autophagy in HAECs. n = 3 and *p < 0.05. CA, chloroquine; GX, glyoxal; GX-CQ, glyoxal with chloroquine; GSH, glutathione; HAEC, human aortic endothelial cell; IGF-R, insulin-like growth factor 1 receptor; INS-R, insulin-like growth factor receptor; LD, lethal dose; VEGF, vascular endothelial growth factor; WT, wild-type IGF-R and INS-R are important regulators of glucose metabolism and energy homeostasis. Therefore, we examined whether GX treatment affected the expression of these receptors in HAECs. Western blot analysis showed that GX treatment significantly decreased the protein levels of IGF-R and INS-R compared with untreated control cells, with reductions of 21.65% and 16.45%, respectively (Fig. 3b, c). To determine whether these changes resulted from altered gene expression, the mRNA levels of IGF-R and INS-R were measured by qPCR. The transcript levels of IGF-R and INS-R remained unchanged following GX treatment, reaching 99.51% and 98.69% of control levels, respectively (Fig. 3d), indicating that GX-mediated reduction of IGF-R and INS-R protein expression was not associated with changes in transcription. To investigate whether autophagy contributed to the GX-induced changes in receptor protein levels, HAECs were treated with GX alone or GX in combination with chloroquine (CQ). LC3-II protein levels, a commonly used marker of autophagosome formation, were significantly increased following GX treatment (4.15-fold compared with control) and further assessed in the presence of CQ (3.40-fold compared with control) (Fig. 3e). In contrast, the mRNA levels of IGF-R and INS-R were not reduced in cells treated with GX or GX-CQ compared with control cells; instead, their expression levels were maintained or increased following CQ treatment (IGF-R: 116.45%; INS-R: 121.65%) (Fig. 3d). These results demonstrate that GX treatment induces LC3-II accumulation and alters IGF-R and INS-R protein expression in HAECs, while the regulation of these receptors is associated with autophagy-related processes. Role of VEGF-R in autophagy VEGF-R expression after GX treatment decreased by 36.45% (Fig. 3b and c) compared with that in the control (Online Resource Table 3). VEGF and VEGF-R are crucial factors in angiogenesis and vascular repair capacity [6]. We investigated the mechanism(s) underlying the decrease in VEGF-R protein levels after GX treatment. First, To determine whether the reduction in VEGF-R protein resulted from transcriptional regulation, VEGF-R mRNA expression was measured by qPCR. No significant difference in VEGF-R transcript levels was observed between the control and GX-treated groups (approximately 99% of the control level; Fig. 3d), indicating that GX-induced VEGF-R downregulation occurred primarily at the post-transcriptional level. Thereafter, western blot analysis confirmed that GX markedly decreased the protein expression of VEGF-R (Fig. 3b, c). In contrast, inhibition of autophagy with chloroquine (CQ) largely restored the expression of these receptors, suggesting that their downregulation was associated with GX-induced autophagy. Wnt and ErbB signaling pathways were inactivated by GX To investigate whether GX affects signaling pathways involved in endothelial cell function, we examined the expression and activation status of proteins associated with the ErbB and Wnt signaling pathways. Western blot analysis showed that GX treatment significantly reduced the phosphorylation levels of ErbB compared with control cells, while total ErbB protein levels remained unchanged (Fig. 4a, b). These results indicate that GX inhibits the phosphorylation levels of ErbB activation without affecting its overall protein abundance. Fig. 4Wnt and ErbB signaling pathways were inactivated by glyoxal (GX) in human aortic endothelial cells (HAECs. a and b Phospho-ErbB2 and phospho-ErbB4 levels, the total protein levels (ErbB2 and ErbB4). c Wnt1, Wnt2, Wnt3, Wnt5, Wnt6, Wnt7, and β-catenin expression after GX treatment (n = 3, *p < 0.05) Wnt and ErbB signaling pathways were inactivated by glyoxal (GX) in human aortic endothelial cells (HAECs. a and b Phospho-ErbB2 and phospho-ErbB4 levels, the total protein levels (ErbB2 and ErbB4). c Wnt1, Wnt2, Wnt3, Wnt5, Wnt6, Wnt7, and β-catenin expression after GX treatment (n = 3, *p < 0.05) We next evaluated the expression of Wnt pathway-related proteins following GX treatment. Compared with control cells, GX exposure markedly decreased the protein levels of multiple Wnt ligands, including Wnt1 (37.63%), Wnt2 (31.65%), Wnt3 (75.01%), Wnt5 (64.82%), Wnt6 (38.15%), and Wnt7 (19.07%) (Fig. 4c). In addition, β-catenin, a key downstream effector of canonical Wnt signaling, was significantly reduced following GX treatment (52.51% of control levels; Fig. 4c). Collectively, these findings demonstrate that GX suppresses both ErbB signaling activation and Wnt pathway-related protein expression in HAECs. Wnt signaling pathway and INS-R and VEGF-R expression were disrupted in mice treated with GX To determine whether GX-induced alterations in signaling pathways observed in HAECs were also present in vivo, we examined the expression of VEGF-R, INS-R, and β-catenin in mouse vascular tissues using immunohistochemical staining. Compared with control mice, GX-treated mice exhibited markedly reduced staining intensity for VEGF-R, INS-R, and β-catenin (Fig. 5a). Quantitative analysis of immunohistochemical signals confirmed that the expression levels of VEGF-R and INS-R decreased to 49.51% and 68.75% of control levels, respectively, while β-catenin expression was reduced to 24.55% of control levels (Fig. 5b). CD31 staining was used to identify endothelial cells within blood vessels (Fig. 5c). These findings demonstrate that GX treatment decreases the expression of VEGF-R, INS-R, and β-catenin in vascular tissues, consistent with the inhibitory effects of GX on endothelial signaling pathways observed in vitro. In addition, biochemical analysis of serum markers showed no significant evidence of liver or kidney injury following GX treatment (Online Resource Table 4), suggesting that the observed vascular effects were not accompanied by detectable systemic toxicity under the experimental conditions. Fig. 5Immunohistochemical analysis showed a GX-induced decrease in the protein expression of β-catenin, vascular endothelial growth factor receptor (VEGF-R), and insulin receptor (INS-R) in mice. a and b Immunohistochemical staining of β-catenin, VEGF-R, and INS-R proteins in mice. c Location of endothelial cells in blood vessels. n = 3 and *p < 0.05. CD31 is highly expressed on the surface of endothelial cells as a transmembrane glycoprotein Immunohistochemical analysis showed a GX-induced decrease in the protein expression of β-catenin, vascular endothelial growth factor receptor (VEGF-R), and insulin receptor (INS-R) in mice. a and b Immunohistochemical staining of β-catenin, VEGF-R, and INS-R proteins in mice. c Location of endothelial cells in blood vessels. n = 3 and *p < 0.05. CD31 is highly expressed on the surface of endothelial cells as a transmembrane glycoprotein Discussion Proteomic and pathway analyses revealed that GX disrupts key signaling pathways, including the Wnt and ErbB pathways, and that it is associated with abnormal energy metabolism, oxidative stress, and autophagy, particularly involving IGF-R, INS-R, and VEGF-R. These molecular disruptions were confirmed both in vitro and in vivo, suggesting a potential mechanistic link to impaired cellular function relevant to angiogenesis and metabolic regulation. GX is primarily derived from food via a series of chemical reactions involving high concentrations of carbohydrates and lipids [1]. The main environmental sources of exogenous GX include cigarette and fire smoke or automobile exhaust fumes. Exogenous GX is rapidly absorbed and metabolized in the human body. It is produced via metabolic auto-oxidation, such as vitamin, amino acid, and lipid peroxidation, in the body [7]. Exposure to high concentrations of GX may promote the development of metabolic diseases, such as diabetes [8] and AI [2, 3]. In this study, we confirmed the novel mechanisms underlying GX-induced AI in vitro and in vivo using proteomics. These novel mechanisms include changes in the expression of INS-R, IGF-R, and VEGF-R and systematic changes in the Wnt and ErbB signaling pathways. Although our previous research revealed that GX induces AI [3], the specific underlying mechanisms remained poorly understood. In this study, we analyzed the effects of GX treatment on HAECs compared with untreated controls using tandem mass tag (TMT)-based proteomic analysis. Key protein targets were screened and analyzed using KEGG pathway, GO term. We identified 19 key DEPs that are closely related to proteins such as receptors, abnormal energy metabolisms (such as glucose metabolism), and the Wnt and ErbB signaling pathways, which were found to be inactivated after treatment with GX. Glucose is important in energy metabolism; therefore, we examined the metabolic effects of GX on HAECs. Methylglyoxal negatively affects pancreatic and beta cells [9], insulin receptor substrate 1 expression [10] and insulin signaling in adipose tissue [11], but not arterial glucose metabolism [12] and insulin receptor signaling. Moreover, microvascular alterations post-acrolein treatment [13] and insulin synthesis and degradation impairment caused by alcohol have been reported [14, 15]; However, there are no reports on GX causing such dysfunctions. Specifically, to the best of our knowledge, no previous studies have reported the effects of GX and blood glucose metabolism (IGF-R and INS-R) on vascular damage. We confirmed that the GSH level decreased by 58% following GX treatment. GSH has an active sulfhydryl group (-SH), which is a crucial functional group that participates in various biochemical reactions and protects the protein sulfhydryl group from oxidation and inactivation, thereby preserving energy metabolism and cell utilization [3]. Our qPCR analysis revealed that the mRNA expression of IGF-R and INS-R in the GX-treated group was similar to that in the controls. The protein levels of IGF-R and INS-R were also similar between HAECs treated with both GX (LD50) and CQ (20 µmol/L; GX-CQ) and GX or CQ controls; however, their levels significantly decreased in HAECs treated with GX (LD50) only compared with those in GX or CQ controls. LC3-II is an autophagy marker. ‌To determine the role of autophagy in IGF-R and INS-R expression changes, we investigated the protein expression of LC3-II after treatment with GX-CQ and found that it was higher than that in the controls. In this study, analysis of the key DEP indicators revealed significant changes in the VEGF-R level after GX treatment. VEGF-R promotes vascular development mechanisms such as vascular permeability, cell migration, and cell proliferation. Our results showed a decrease in the expression of VEGF-R (36.4%) after GX treatment, and we confirmed that autophagy causes VEGF-R expression changes through changes in the LC3-II protein level. Changes in VEGF expression cause diabetic retinopathy. Studies have shown that acrolein [6]; 4-hydroxynonenal [16], glyceraldehydes, and glycolaldehydes [17, 18] alter VEGF-R expression. However, the effects of these compounds and those of GX on HAECs have not been demonstrated. In this study, analysis of the key DEP indicators revealed that GX inactivated the ErbB signaling pathway. The phospho-ErbB2 and phospho-ErbB4 levels decreased following GX treatment; however, the total protein levels were not significantly different from those in the controls. ErbB2 is involved in vascular proliferation [19], increasing VEGF expression and promoting angiogenesis [20]; however, no study has determined the role of GX in the ErbB signaling pathway. Analysis of the key DEP indicators further revealed that the effects of GX are closely associated with the Wnt signaling pathway. We systematically investigated the levels of Wnt signaling pathway proteins, including Wnt1, Wnt2, Wnt3, Wnt5, Wnt6, Wnt7, and β-catenin, and found that the levels of these proteins decreased compared with those in the controls. To the best of our knowledge, this is the first study on the relationship between the inactivation of Wnt signaling pathway-related proteins after treatment with GX and the development of AI, myocardial hypertrophy, vascular calcification, arteriosclerosis, angiogenesis, and inflammation, as well as disruption of the blood–brain barrier. Antagonism between Wnt1 and Wnt3a can regulate the angiogenesis process [21]; Wnt1 promotes vascular stability, whereas Wnt5a does the opposite [22]. The decrease in the Wnt1 level observed post-GX treatment in this study implies that Wnt1 cannot effectively antagonize Wnt3a and Wnt5a. Inadequate Wnt2 levels significantly impair the ability of embryonic stem cells to form cardiogenic progenitor cells [23], inhibiting vascular endothelial cell differentiation and tubule formation [23, 24]. Decreased levels of Wnt3a reduce its ability to stimulate cell development processes, including differentiation and survival [21, 25], as well as its effective antagonization of Wnt1 [21] and Wnt5a [26]. Some studies have reported a relationship between Wnt5 expression and angiogenesis, inflammation, and atherosclerotic and vascular calcification. Wnt5 has different functions. For example, human endothelial angiogenesis is suppressed by Wnt5 signaling through decreased endothelial cell proliferation and survival [22]. However, a previous study has shown that Wnt5 has an enhancing effect [27]. A decrease in the level of Wnt5 results in the disruption of cytoskeletal organization and cell polarity, leading to decreased cell polarization and increased permeability, which in turn impairs the regulation of Wnt-mediated blood–brain barrier function [24]. Enhanced permeability drives the endothelium into an early inflammatory phase [28], implying that Wnt5 signaling under GX treatment exerts pro-inflammatory effects. Wnt5a expression is high in atherosclerotic lesions [29], suggesting that a decrease in the level of Wnt5 is involved in atherosclerosis. Inactivation of the Wnt5 pathway is important for vascular calcification [30, 31]. Wnt6 plays a central role in the Wnt signaling pathway [32], and suppression of Wnt6 expression under treatment with GX suppresses corneal neovascularization [33]. Siglec6 can induce downregulation of Wnt6 expression, leading to vascular endothelial cell damage [34]. Wnt7 is essential for the blood–brain barrier function [35]. Wnt7–LRP6 signaling is crucial to vascular calcification [36]; highlighting the strong correlation between Wnt7 and inflammatory effects [37]. In this study, we confirmed that the Wnt signaling pathway involving Wnt1, Wnt2, Wnt3, Wnt5, Wnt6, and Wnt7 and its downstream protein β-catenin is impaired after treatment with GX. Similarly, we identified changes in the Wnt signaling pathway in HAECs after GX treatment, highlighting its key role in mediating blood–brain barrier development [24], vascular calcification, arteriosclerosis [38], angiogenesis [25], and inflammation [26]. Therefore, we investigated the mechanism underlying the relationship between the Wnt signaling pathway in HAECs after GX treatment and its effects on the proliferation, migration, differentiation, and survival of HAECs. A limitation of this study is that we used laboratory-cultured HAECs rather than clinical specimens, and this could have led to differences between the results of this study and those of clinical studies. Another limitation is that antibodies used to detect the proteins of the Wnt signaling pathway, such as Wnt8 and Wnt9, could not be acquired at the time of this study. These limitations will be addressed in future studies. Conclusions Overall, we confirmed the mechanisms underlying GX-induced AI in vitro and in vivo using proteomics. These novel mechanisms include changes in the expression of INS-R, IGF-R, and VEGF-R and systematic changes in the Wnt and ErbB signaling pathways. These alterations correspond to abnormal energy metabolism and function of HAECs, causing aberrant vascular function, inflammation, and vascular structure. Our findings could be used to develop a new molecular map of GX-induced AI, thereby advancing precision medicine in patients with diabetes. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4