All animal studies performed in this work were conducted in accordance with the guidelines and approval of the Animal Care and Use Committees at Jinan University (No. 20241015-0025, Guangzhou, China) and Zhongshan Hospital, Fudan University (No. 20210220, Shanghai, China). Adult male C57BL/6J mice (8 weeks old, 22-24 g) were housed in a pathogen-free environment. The animals were kept on a 12-h light/dark cycle and provided with ad libitum access to water and food. To establish type 2 diabetes mellitus, mice were fed a 60% high-fat diet (percent fat calorie) for 28 days, followed by 5 consecutive days of intraperitoneal streptozotocin (STZ, 40 mg/kg/day, in citrate buffer, 0.1 M, pH 4.5) injections. Mice were maintained on a high-fat diet throughout STZ administration and for the remainder of the study. The following 4 animal groups were designed: (1) Control: standard chow diet and equivalent volume of citrate buffer for injection; (2) Diabetic: high-fat diet for 28 days, then STZ injection for 5 days and a continued high-fat diet for an additional 6 weeks; (3) Control-FPS-ZM1: standard chow diet plus the RAGE inhibitor FPS-ZM1 (10 mg/kg i.p., twice per week)^[26] starting on day 33; and (4) Diabetic-FPS-ZM1: high-fat diet-STZ regimen as described, with concurrent FPS-ZM1 treatment (10 mg/kg i.p., twice per week) beginning on day 33 (end of STZ injection). Mice were weight-matched and randomized into 4 experimental groups (10 mice per group). Mice with overnight fasting blood glucose levels greater than 11.1 mM on two readings were considered diabetic^[27]. Circulating levels of fasting glucose, triglycerides, cholesterol, and AGEs were measured at the completion of the entire high-fat diet intake regimen using commercial colorimetric assay kits (MyBioSource, San Diego, CA, USA). FPS-ZM1 is chosen primarily for its unparalleled specificity for the RAGE receptor compared to broader first-generation inhibitors. By targeting the V-domain of RAGE (Kd ≈ 25 nM), FPS-ZM1 effectively decouples the pathological signaling initiated by AGEs. Furthermore, its superior pharmacokinetic profile ensures adequate tissue distribution in vivo. In addition, intraperitoneal injections are the standard route for ensuring consistent systemic bioavailability^[28,29].

The GSE123975 dataset was retrieved from the Gene expression omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) using STZ-induced diabetic mice. Data were generated using the Agilent-028005 SurePrint G3 Mouse GE 8x60K Microarray platform on left ventricular tissues from 6 hyperglycemic mice and 6 normoglycemic mice. Raw data were processed in R using the affy package. Background correction and normalization were performed using the Oligo and robust multi-array average (RMA) algorithms, while missing values were imputed using the k-Nearest Neighbor method^[30]. For genes represented by multiple probes, expression levels were collapsed to the median value. Differentially expressed genes (DEGs) were detected by the “limma” algorithm. Statistical significance was determined based on the following predefined criteria: |log2 fold change| greater than 0.7 (1.6-fold difference) and an adjusted P-value < 0.05^[7].

The clusterProfiler software was utilized to perform Gene Ontology (GO) functional annotation on the identified DEGs. Genes were categorized into three principal functional categories: Molecular Functions (MF), Cellular Components (CC), and Biological Processes (BP) for enrichment analysis. This hierarchical classification enabled the identification of significantly enriched functional clusters, providing a comprehensive overview of biological landscapes altered in experimental models^[31]. Simultaneously, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed using clusterProfiler to map DEGs to biochemical and signaling networks. This analysis was instrumental in elucidating the complex interactions of DEGs within established biological pathways and their coordinated roles in driving pathological phenotypes^[32]. To complement these threshold-based methods, the fgsea package was used to perform Gene Set Enrichment Analysis (GSEA). Unlike traditional enrichment, GSEA assesses entire gene sets across an ordered list of all detected genes, allowing for detection of subtle yet biologically significant coordinated shifts. Enrichment scores were calculated to reveal functional pathways that were significantly correlated with the observed gene expression patterns between groups^[33]. GO and KEGG enrichment analyses were performed using over-representation analysis based on Fisher’s exact test. To control the false discovery rate, P-values were adjusted using the Benjamini-Hochberg method, with a significance threshold set at adjusted P < 0.05. GSEA was conducted with 1,000 phenotype permutations, and significance was determined by normalized enrichment score (NES) with an false discovery rate (FDR) q value < 0.25. For differential expression analysis, between-group comparisons were evaluated using a two-sided Student’s t-test with Benjamini-Hochberg correction across all genes; an adjusted P < 0.05 was considered significant.

Cardiac structure and contractile function were evaluated in sedated mice (ketamine: 80 mg/kg; xylazine: 12 mg/kg, i.p.) using a Vevo 2100 high-resolution imaging system (FUJIFILM VisualSonics, Toronto, Canada) equipped with a 22-55 MHz linear array transducer. Briefly, the heart was visualized in the parasternal long-axis view to obtain M-mode tracings perpendicular to the interventricular septum and the left ventricular (LV) posterior wall. Key parameters, including LV wall and septal thicknesses, as well as LV end-diastolic (LVEDD) and end-systolic (LVESD) diameters, were recorded. Functional indices, including fractional shortening and ejection fraction, were averaged over five consecutive cardiac cycles. To ensure objectivity, all echocardiographic assessments and data analyses were performed by an investigator blinded to the experimental groups^[20,34].

Following the induction of deep sedation, hearts were rapidly removed and fixed in 10% neutral-buffered formalin for 24 h. Subsequently, tissues underwent standard processing and were embedded in paraffin wax. Serial tissue sections (5 μm) were stained with hematoxylin and eosin (H&E) to evaluate cardiomyocyte morphology. Cross-sectional areas were quantified using ImageJ software (version 1.34S). In parallel, Masson’s trichrome staining was applied to evaluate the degree of myocardial interstitial fibrosis, expressed as the portion of the collagen-positive (blue colored) section within the entire microscopic field^[7,35].

MG-derived AGEs (MG-AGE) were generated by incubating 100 mM methylglyoxal with 7.2 mg/mL bovine serum albumin (BSA) in a 100 mM phosphate buffer (pH 7.4) at 37 °C for 48 h under sterile conditions. A control BSA solution was processed in parallel in the absence of methylglyoxal. Following incubation, residual unreacted carbonyl species were removed via extensive dialysis against ammonium bicarbonate buffer (30 mM, pH 7.9). Following preparation, MG-AGE and control BSA were sterilized using a 0.22 μm filter and stored at -20 °C until further use. Concentrations were determined based on the equivalent BSA levels supplied^[18]. Adult murine cardiomyocytes were isolated via retrograde Langendorff perfusion, followed by enzymatic digestion using Liberase Blendzyme 4. Only cells exhibiting distinct striations and an absence of spontaneous twitches were selected for functional assessment. For assessment of RAGE, cuproptosis, and the S100 calcium-binding protein A13 (S100A13) Cu²⁺ binding protein in AGE- or diabetes-induced cardiomyopathy, adult C57BL/6 mouse cardiomyocytes were incubated with either MG-AGE (5 μM)^[18] or high glucose (25.5 mM, to mimic hyperglycemia)^[36] for 12 h with or without the RAGE blocker FPS-ZM1 (25 μM)^[26], the cuproptosis blocker tetrathiomolybdate (TTM, 20 μM)^[20], the S100A13 blocker amlexanox (AMX, 20 μM)^[37,38], or the Cu²⁺ ionophore elesclomol (40 nM)^[20]. Control BSA (5 μM) and normal glucose (5.5 mM) were employed as controls for MG-AGE and high-glucose environment, respectively^[18,36].

Cardiomyocyte relengthening and shortening were examined using a SoftEdge MyoCam detection system (IonOptix, Milton, MA, USA) in a 1 mM CaCl₂ environment. Indices evaluated encompassed maximal rate of cell relengthening (-dL/dt), shortening (+dL/dt), peak shortening (PS), time-to-90% relengthening (TR₉₀), and time-to-peak shortening (TPS), in cardiomyocytes with 0.5 Hz field-stimulation^[36,39].

Isolated cardiomyocytes were incubated with intracellular fluorescence probe 5-(6)-chloromethyl-2’,7’-dichlorodihydrofluorescein diacetate (DCF; 5 μM; Molecular Probes, Eugene, OR, USA) at 37 °C for 30 min to detect intracellular ROS. After dye loading, fluorescence images were acquired using an Olympus fluorescence microscope under identical exposure settings for all groups^[7,20].

For ultrastructural characterization, ventricular tissue was sectioned into small cubes (≤ 1 mm³) and fixed in 2.5% glutaraldehyde (0.1 M sodium phosphate buffer, pH 7.4) at 4 °C. After primary fixation, specimens were dehydrated through a graded series of ethanol and embedded in Epon-Araldite resin. Ultrathin sections (approximately 50 nm) were obtained using an Ultracut E ultramicrotome (Leica) and double-stained with lead citrate and uranyl acetate for contrast. Samples were visualized using a Hitachi H-7000 transmission electron microscope (Pleasanton, CA, USA) equipped with a high-resolution digital imaging system^[20].

MMP was assessed using the ratiometric fluorescence dye JC-1 (5 μM). Isolated cells were exposed to JC-1 at 37 °C for 20 min in the dark. Following incubation, fluorescence intensities were quantified using a spectrofluorimeter (SpectraMax Gemini XS). The JC-1 monomer fluorescence (green, indicating mitochondrial depolarization) was recorded at an emission wavelength of 530 nm, while the JC-1 aggregate fluorescence (red, indicating polarized, healthy mitochondria) was recorded at 590 nm. The aggregate-to-monomer ratio was calculated and utilized as a sensitive index of MMP^[20].

Myocardial AGE levels were evaluated via immunohistochemistry on 8 µm thick paraffin-embedded myocardial sections. Following deparaffinization and rehydration, sections underwent antigen retrieval with 0.05% proteinase K in phosphate-buffered saline (PBS) (pH 7.4) for 30 min. Background peroxidase activity was blocked with H₂O₂, followed by systematic PBS washes. To visualize AGE deposition, slides were incubated with the primary anti-AGE antibody 1H7G5 (1:500), followed by an anti-mouse fluorescein isothiocyanate (FITC)-conjugated immunoglobulin G (IgG) secondary antibody. Fluorescence signals (Excitation: 490 nm/Emission: 517 nm) were captured on an Olympus BX51 fluorescence microscope, coupled with a high-resolution cooled CCD camera. The intensity and distribution of FITC-labeled AGEs were quantified using Image-Pro Plus software^[40].

Total RNA was isolated from myocardial tissues utilizing TRIzol® reagent (Invitrogen, Carlsbad, CA, USA). cDNA was produced using the SuperScript^TM III system (Invitrogen) from 2 μg of RNA. Quantitative real-time PCR was performed using SYBR Green-based detection on an iCycler thermal cycler (Bio-Rad). Specific primers were used for RAGE (F: 5’-GGT CAT GCC GGA GTG TCT ATT-3’; R: 5’-GTG TAG AAC CCG TCG GTG AGG-3’) and the internal control glyceraldehyde-3-phosphate dehydrogenase (GAPDH; F: 5’-CTG GAA AGC TGT GGC GTG ATG-3’; R: 5’-GCC AGT GAG CTT CCC GTT CAG-3’). Relative mRNA levels were determined using the ΔCT protocol^[39].

Total protein was extracted from myocardial tissues, and concentrations were determined by a bicinchoninic acid (BCA) assay to ensure consistent loading. Protein extracts were resolved by sodium dodecyl sulfate (SDS) gel electrophoresis using the Mini-PROTEAN II system (Bio-Rad) and subsequently electrotransferred onto nitrocellulose membranes. After blocking, membranes were incubated with primary antibodies against uncoupling protein 2 (UCP2), peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α), B-cell lymphoma 2 (Bcl2), Bcl-2-associated X protein (Bax), solute carrier family 31 member 1 (SLC31A1, CTR1), ATPase copper transporting beta (ATP7B), ferredoxin 1 (FDX1), dihydrolipoamide S-acetyltransferase (DLAT), aconitase 2 (ACO2), NADH dehydrogenase [ubiquinone] iron-sulfur protein 8 (NDUFS8), and S100A13, each at a dilution of 1:1,000. Housekeeping proteins, including GAPDH and α-tubulin (1:5,000), were used as internal loading controls to ensure accurate normalization and quantitative reliability. Immunoreactive bands were detected using enhanced chemiluminescence (ECL) and quantified via a Bio-Rad densitometry system. All experimental antibodies were sourced from Abcam or Cell Signaling Technology^[41].

Intracellular Cu²⁺ was quantified using a commercial detection kit (ab272528, Abcam, Boston, MA, USA). Briefly, cells were homogenized and diluted with ultrapure water to ensure consistent sample viscosity. A 100 μL aliquot of the homogenized sample was dispensed into a 96-well UV-transparent plate. The reaction was initiated by sequential addition of Reagent A, Reagent B, and Reagent C. Following gentle mixing, fluorescence signals were measured using a BioTek Synergy 2 microplate reader. To capture the specific Cu²⁺-binding signal, the emission was detected at a 359 nm wavelength^[20].

Mouse HL-1 atrial cardiomyocytes were cultured in medium supplemented with 1% penicillin-streptomycin and 10% fetal bovine serum in a humidified 5% CO₂ atmosphere at 37 °C. For gene silencing, a specific small interfering RNA (siRNA) targeting murine S100A13 (sense: 5’-CGUGGGCUCUCUAGAUGAAtt-3’; antisense: 3’-UUCAUCUAGAGAGCCCACGtt-5’) was synthesized by Guangzhou IGE Biotechnology Co., Ltd. (Guangzhou, China). HL-1 cells were transfected with scrambled control or S100A13 siRNA using Lipofectamine 3000. Knockdown efficiency was rigorously validated by immunoblotting 48 h post-transfection. Following validation, HL-1 cells (scrambled control or S100A13-silenced) were incubated with 5 µM MG-AGE for 12 h with or without the S100A13 blocker AMX (20 μM) or FPS-ZM1 (25 μM) prior to assessment of intracellular Cu⁺ levels. HL-1 cells were incubated with the Cu⁺-selective fluorescent probe CS1 (5 μM, FocusBio, China) at 37 °C for 30 min. Fluorescence signals were captured using a laser scanning confocal microscope (Excitation: 543 nm; Emission: 550-650 nm). Mean fluorescence intensity, representing intracellular Cu⁺ level, was determined via ImageJ. All experimental groups were imaged under identical acquisition parameters to ensure the comparability of labile Cu⁺ levels^[42].

The putative complex structure between RAGE and S100A13 was predicted using AlphaFold3 (https://alphafoldserver.com/). Structure-based interaction interfaces were subsequently analyzed with PDBe-PISA (https://www.ebi.ac.uk/msd-srv/prot_int/pistart.html) and visualized using PyMOL (https://pymol.org/)^[41].

Physical protein-protein interactions were evaluated using the Thermo Fisher Pierce^TM Co-Immunoprecipitation Kit. Briefly, 50 μg of purified anti-RAGE, anti-S100A13, or anti-Flag antibodies were immobilized onto the provided amine-reactive resin. Myocardial protein lysates (1 mg total protein) were then incubated with the antibody-conjugated resin for 2 h to facilitate the formation of stable immune complexes. Following extensive washing, the captured protein complexes were eluted using an acidic elution buffer. The resulting eluates were neutralized and subjected to immunoblot analysis using complementary primary antibodies^[39].

Results are expressed as the mean ± standard error of the mean (SEM). Statistical analyses were performed using GraphPad Prism software (GraphPad Software, San Diego, CA, USA). Differences between two groups were evaluated using a two-tailed Student’s t-test, while comparisons among multiple groups were conducted using one-way or two-way analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple comparisons. Statistical significance was established at P < 0.05.

GO analysis of the DEGs in diabetic hearts revealed a distinct functional landscape. MF analysis showed pronounced enrichment of metal ion transmembrane transporter activity, phospholipid binding, and GTPase regulator activity. Regarding CC, significant enrichment was observed in the monoatomic ion channel complex, mitochondrial protein-containing complex, and mitochondrial matrix. Furthermore, BP analysis identified key pathways related to the regulation of metal ion transport, response to metal ions, and response to oxidative stress [Figure 1A]. KEGG pathway analysis further revealed significant enrichment in metabolic and structural networks, including V-type ATPase, the citrate cycle (TCA/Krebs cycle), and inositol phosphate metabolism [Figure 1B]. To achieve a higher resolution of coordinated gene shifts, GSEA was performed, revealing a close correlation with the GOBP: cellular response to Cu²⁺ ion pathway in diabetic myocardial injury (enrichment Score = -0.677, NES = -1.833, P = 1.29e-03, adjusted P = 0.039) [Figure 1C]. To further delineate specific genes driving this Cu²⁺ dysregulation, we utilized a Venn diagram to identify overlapping DEGs and generated scatter plots to visualize the expression distribution of key regulators within the Cu²⁺ ion-binding pathway [Figure 1D and E].

To determine if the AGE-RAGE axis, a primary driver of diabetic complications, acts as the upstream trigger for this observed cuproptosis signature in diabetic hearts, previously reported^[43], we next evaluated the cardiac expression of AGE and RAGE in diabetic mouse hearts. Figure 1F and G revealed that diabetes increased AGE immunofluorescence by 1.86-fold and RAGE mRNA levels by 2.16-fold (P < 0.0001) compared with controls. In this context, we treated diabetic mice with the RAGE inhibitor FPS-ZM1 (10 mg/kg, i.p., twice per week) beginning after completion of STZ injection (day 33 following study initiation) and continuing for an additional 6-week high-fat diet. Neither diabetes nor FPS-ZM1, alone or in combination, affected body weight or liver and kidney weight. Diabetes increased circulating levels of fasting glucose by 2.84-fold (P < 0.001), triglycerides by 1.45-fold (P < 0.001), cholesterol by 0.80-fold (P < 0.001) and AGEs by 1.31-fold (P < 0.001), the effect of which was unaffected by FPS-ZM1. Treatment of FPS-ZM1 alone did not exert any effect on biometric or plasma profiles [Figure 1H-N].

Echocardiographic evaluation revealed that diabetes enhanced septal thickness by 0.15-fold (P < 0.001), LVESD by 0.41-fold (P < 0.0001), LVEDD by 0.15-fold (P < 0.0001), absolute LV mass by 0.38-fold (P < 0.001), and normalized LV mass by 0.34-fold (P < 0.0001), and dampened ejection fraction by 0.12-fold (P < 0.0001) and fractional shortening by 0.23-fold (P < 0.0001) without affecting LV wall thickness or heart rate. Gross heart weight and size (normalized to body weight) were also elevated by 0.16-fold (P < 0.001) and 0.22-fold (P < 0.0001), respectively, in experimental diabetes. Although FPS-ZM1 alone failed to alter echocardiographic or heart parameters tested, it effectively mitigated diabetes-evoked changes in myocardial geometry and function [Figure 2A-L].

Sustained diabetes decreased cardiomyocyte shortening capacity, as manifested by a reduction in PS (0.52-fold, P < 0.001), +dL/dt (0.49-fold, P < 0.01), and -dL/dt (0.43-fold, P < 0.01), and prolongation of TR₉₀ (0.41-fold, P < 0.0001) without altering TPS and resting cell length. FPS-ZM1 rescued diabetes-induced cardiomyocyte anomalies with little response itself [Figure 3A-F].

H&E, Masson trichrome, and DCF staining were conducted to assess the possible effects of RAGE inhibition on diabetes-induced morphological pathology and oxidative damage. As shown in Figure 4A-F, diabetes provoked evident rises in cardiomyocyte cross-sectional area (by 0.19-fold, P < 0.0001), interstitial fibrosis (by 6.75-fold, P < 0.0001), and oxidative injury (by 2.53-fold, P < 0.0001, assessed using DCF staining), the effects of which were ablated by FPS-ZM1 with subtle effects itself.

TEM results showed that diabetes induced localized ultrastructural damage, characterized by increased mitochondrial area (0.19-fold, P < 0.0001) and circularity (0.23-fold, P < 0.0001), along with localized myofibrillar disarray. Consistent with myocardial functional changes, FPS-ZM1 canceled out diabetes-induced ultrastructural pathology in the absence of a notable effect itself [Figure 5A-C]. In addition, diabetes led to the collapse of mitochondrial membrane potential (MMP, 0.51-fold, P < 0.0001) and downregulation of mitochondrial proteins PGC1α (0.36-fold, P < 0.001) and UCP2 (0.32-fold, P < 0.05); the response was abolished by FPS-ZM1 without any effect of its own [Figure 5D-H].

Given the bioinformatics finding of prominent cell death features, including cuproptosis, in diabetic hearts, protein markers of apoptosis, cuproptosis and intercellular Cu²⁺ were evaluated. Our results revealed evident apoptosis [increased Bax (0.37-fold, P < 0.05) and decreased Bcl2 (0.54-fold, P < 0.01)], cuproptosis [increased Cu²⁺ importer SLC31A1 (CTR1, 0.55-fold, P < 0.0001), FDX1 (0.99-fold, P < 0.01) and DLAT (1.33-fold, P < 0.0001), with unchanged Cu²⁺ exporter ATP7B], Cu²⁺ binding protein S100A13, loss of Fe-S cluster proteins (NDUFS8 and ACO2) along heightened intracellular Cu²⁺ levels (by 0.65-fold, P < 0.0001) following diabetic insult. FPS-ZM1 reversed diabetes-induced alterations in Bax, Bcl2, FDX1, DLAT, ACO2, NDUFS8, S100A13, and intracellular Cu²⁺ levels, without affecting SLC31A1 (CTR1) and ATP7B. FPS-ZM1 did not affect any of the apoptotic, Cu²⁺-related proteins or intracellular Cu²⁺ levels [Figure 6A-K].

Diabetes results in elevated levels of methylglyoxal (MG-AGE), the precursor of AGE, prompting pathogenesis of diabetic complications^[43]. To decipher the possible involvement of RAGE, cuproptosis and Cu²⁺ binding protein S100A13 in diabetic cardiomyopathy, C57BL/6 mouse cardiomyocytes were exposed to MG-AGE (5 μM) for 12 h in vitro with or without the RAGE blocker FPS-ZM1 (25 μM), the cuproptosis blocker TTM (20 μM)^[43], and the S100A13 inhibitor AMX (20 μM)^[37]. Our findings displayed in Figure 7A-F revealed that MG-AGE overtly compromised cardiomyocyte mechanical properties, manifested as decreased PS (by 0.38-fold, P < 0.05) and ±dL/dt (by 0.35- and 0.38-fold, respectively, P < 0.05 for both) along with prolonged TR₉₀ (by 0.55-fold, P < 0.0001) without influencing resting cell length and TPS. Intriguingly, these MG-AGE-evoked cardiomyocyte mechanical anomalies were effectively mitigated by FPS-ZM1, TTM, and amlexanox, with no discernable response from the inhibitors themselves. These results suggest a role for RAGE, cuproptosis, and S100A13 in MG-AGE-elicited cardiomyocyte dysfunction.

To determine whether S100A13 is the functional target of amlexanox, we conducted a gene-silencing study in HL-1 cells. Cells transfected with either scramble or S100A13 siRNA were incubated with 5 μM MG-AGE for 12 h, with or without amlexanox (20 μM) or the RAGE inhibitor FPS-ZM1 (25 μM). As shown in Figure 7G and H, MG-AGE challenge significantly increased intracellular labile Cu⁺ accumulation (1.93-fold, P < 0.0001), a response that was effectively mitigated by amlexanox. Notably, siRNA knockdown of S100A13, but not scramble RNA, essentially abrogated the protective effect of amlexanox against MG-AGE-induced labile Cu⁺ buildup. S100A13 silencing [efficiency validated in Figure 7I] failed to induce a discernible effect itself without MG-AGE exposure. Furthermore, pharmacological inhibition of RAGE with FPS-ZM1 obliterated the rise in intracellular labile Cu⁺, phenocopying the effect of amlexanox. These findings demonstrate that S100A13 is likely a requisite mediator for amlexanox-induced benefit against MG-AGE-driven Cu⁺ dysregulation.

To understand the processes underlying AGE-RAGE-instigated governance of cuproptosis, interactions between RAGE and cuproptosis regulatory components were evaluated. Co-IP assay demonstrated pronounced physical interaction between RAGE and S100A13, a known Cu²⁺ binding protein [Figure 7J]. The interaction interface was modeled by AlphaFold3 software^[44] and a detectable protein interaction interface between RAGE and S100A13 was identified in the model, as evaluated by PDBe-PISA software (interaction interface of 650.3 square angstrom, Figure 7K).

To discern the possible roles of RAGE, cuproptosis and S100A13 in high glucose (mimicking hyperglycemia in diabetes)-induced myocardial mechanical anomalies, a high glucose-supplemented culture medium was employed as an in vitro model for diabetes^[36]. Freshly isolated murine cardiomyocytes were exposed to normal glucose (NG, 5.5 mM) or high glucose (HG, 25.5 mM) for 12 h with or without the RAGE blocker FPS-ZM1 (25 μM), the cuproptosis blocker TTM (20 μM)^[43], the S100A13 blocker amlexanox (20 μM)^[37,38] or the Cu²⁺ ionophore elesclomol (40 nM)^[43]. Our results showed that high glucose disrupted cardiomyocyte mechanical properties, shown as compromised PS (by 0.44-fold, P < 0.001) and ±dL/dt (by 0.44- and 0.51-fold, respectively, P < 0.01 for both) in conjunction with prolonged TR₉₀ (by 0.46-fold, P < 0.0001) and unchanged TPS and resting cell length. Intriguingly, the high glucose-instigated cardiomyocyte anomalies were rescued by FPS-ZM1, TTM, and amlexanox. Moreover, Cu²⁺ ionophore elesclomol canceled out the FPS-ZM1-elicited beneficial effect against high glucose concentrations. None of these pharmacological agents had any notable effect on cardiomyocyte function under normal glucose culture conditions [Figure 8A-F]. These observations suggest a vital role for RAGE, cuproptosis and the Cu²⁺-binding protein S100A13 in high glucose culture-elicited myocardial pathologies.