VD₃ was procured from Sigma Aldrich (St. Louis, MO, USA). Dulbecco’s modified eagle medium (DMEM) and fetal bovine serum (FBS) were obtained from GIBCO (Grand Island, NY, USA) and LONSERA (Auckland, New Zealand), respectively. 1,25(OH)₂D₃ and Ferrostatin-1(Fer-1) were purchased from GLPBIO (Montclair, CA, USA), and Dacomitinib (Dac) was acquired from MedChemExpress (Monmouth Junction, NJ, USA). The sources of antibodies were as follows: NRG1 (10527-1-AP), SLC7A11 (26864-1-AP), TFR1 (84766-4-RR), NCOA4 (83394-4-RR), β-actin (20536-1-AP) were purchased from Proteintech (Wuhan, China), ErbB4 (RT1276), YAP (RT1664), Phospho-YAP (ET1611-69), ACSL4 (ET7111-43), Ferritin (R1601-9), GPX4 (ET1706-45) were purchased from HUABIO (Hangzhou, China). Phospho-ErbB4 (Tyr1284) (AF3445) were purchased from Affinity Biosciences (Cincinnati, USA).

Four-week-old specific pathogen free (SPF) male C57BL/6J mice (n=12) and KKAy mice (n=48) were purchased from Beijing Huafukang Bio-technology Co. The animals were housed in a controlled barrier facility with a 12h light/dark cycle at 22 ± 1°C and 55 ± 5% humidity, with ad libitum access to food and water. Following a 2-week acclimatization period, C57BL/6J mice were designated as the normal control group (NC) and fed a standard chow diet (fat: 4.0 gm%), while KKAy mice received a HFD (17.9 gm%) for 6 weeks to establish a prediabetic model. The modeling criteria were defined as fasting blood glucose (FBG) levels maintained within 6.7 to 16.7 mmol/L, combined with 2h post-oral glucose tolerance test (2h-OGTT) blood glucose levels ranging from 11.1 to 30 mmol/L. Successfully modeled KKAy mice were randomly divided into four groups according to body weight and FBG levels: model group (MC), low-dose VD₃ intervention group (LVD, 0.42 IU/g/w), medium-dose VD₃ intervention group (MVD, 1.68 IU/g/w), and high-dose VD₃ intervention group (HVD, 4.20 IU/g/w). VD₃ dissolved in 0.1 mL corn oil was administered via intraperitoneal injection, and the NC and MC groups received an equivalent volume of corn oil. After 16 weeks of VD₃ intervention, mice were anesthetized by intraperitoneal injection of sodium pentobarbital at a dose of 50 mg/kg. Heart tissues and blood samples were then immediately collected and preserved for further analysis. All animal experiments were approved by the Life Sciences Ethics Review Committee of Zhengzhou University (Ethics No. ZZUIRB2021-GZR0141).

After 16 weeks of VD₃ intervention, echocardiography was performed by using Vevo2100 imaging system. Parasternal short-axis M-mode echocardiographic images were obtained at papillary muscles level. The following parameters were recorded: left ventricular (LV) mass, left ventricular volume at diastole and systole (LV Vol;d and LV Vol;s), left ventricular internal dimension at diastole and systole (LVID;d and LVID;s), left ventricular ejection fraction (EF), fractional shortening (FS), and interventricular septal thickness at systole and diastole (IVS;s and IVS;d).

FBG, total cholesterol (TC), low-density and high-density lipoprotein cholesterol (LDL-C and HDL-C), creatine kinase (CK) and lactate dehydrogenase (LDH) levels were measured using the Nanjing JianCheng Assay Kits. Serum fasting insulin (INS) and 25(OH)D were measured using Mlbio Enzyme-linked immunosorbent assay Kits.

Heart tissue (50 mg) was precisely weighed and placed into a grinding tube containing grinding beads. Subsequently, 450 mL of physiological saline was added, and the tissue was thoroughly ground using a grinder. The resulting mixture was then centrifuged in a pre-cooled centrifuge for 10 min, after which the supernatant was collected for measurement. The protein concentration was determined, and Fe, GSH, and MDA levels were measured as per the provided instructions. Subsequently, values were calculated and analyzed using the designated formulas.

Following fixation in 4% paraformaldehyde, cardiac tissues were sequentially subjected to graded ethanol dehydration, xylene clearing, paraffin infiltration, and sectioning into 3-5 μm slices using a rotary microtome. For histological analysis, sections were dewaxed in xylene (3 × 5 min), rehydrated through descending ethanol gradients (100%, 95%, 75%; 2 min each), and stained with H&E per standard protocols. For WGA staining, the sections underwent dewaxing, antigen retrieval, and incubation with primary and secondary antibodies. Simultaneously, the WGA working solution was drop-stained, counterstained with DAPI for nuclear localization, and imaged under a laser scanning confocal microscope. The cross-sectional area of cardiomyocytes was measured using Image J software, and the resulting data were analyzed. Furthermore, sections were deparaffinized, stained with Prussian blue and DAB, and images were observed and captured using a light microscope.

The tissues preserved at -80°C were embedded using OCT compound to prepare frozen sections. Subsequently, these sections were stained with oil red O following the provided guidelines, and the resulting images were observed and captured using a light microscope. For further analysis, the frozen sections were allowed to thaw to room temperature, ensuring moisture control and subsequent drying. Tissue autofluorescence was minimized, and a staining solution for ROS was meticulously added to the sections. After restaining the nuclei with DAPI, the sections were sealed and observed using fluorescence microscopy, with images captured for subsequent analysis.

Cardiac paraffin sections were deparaffinized in xylene (3 × 10 min) and rehydrated through a graded ethanol series into distilled water. Antigen retrieval was performed in citrate buffer at 95°C for 20 min. After cooling to room temperature, sections were washed in PBS (3 × 5 min), permeabilized with 0.1% Triton X-100 for 10 min, and treated in 3% hydrogen peroxide in methanol for 20 min in the dark. To suppress lipofuscin autofluorescence, sections were incubated with 10% acetic acid in methanol for 10 min and rinsed in PBS (3 × 5 min). Following equilibration in commercial equilibration buffer, sections were incubated with TUNEL reaction mixture at 37°C for 90 min in a humidified dark chamber. After washing in stop/wash buffer (3 × 10 min), Converter-POD was applied for 30 min at 37°C. DAB chromogen was used for signal development under microscopic monitoring. Sections were counterstained with Mayer’s hematoxylin for 45 sec, dehydrated through graded ethanol, cleared in xylene, and mounted with DPX mounting medium.

H9c2 cells were obtained from the First Affiliated Hospital of Zhengzhou University and cultured in DMEM medium supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C in humidified 5% CO₂ conditions. The cells were used in the experiments once they reached 70-80% confluence. High glucose and palmitic acid medium (HGPA, 33.3 mM d-glucose, 0.25 mM PA) was used to construct injury models by culturing cells for 24h. Addition of 10 nM 1,25(OH)₂D₃ and/or 73.7 nM Dacomitinib or 1 μM Ferrostatin-1 pretreated the cells for 2h while co-culturing the cells with HGPA for 24h ( Figure 1 ).

Cells were seeded in 96-well plates, allowing them to adhere before the addition of medium containing various concentrations of intervening substances. Subsequently, 10 μL of Cell Counting Kit-8 (ApexBio, Houston, USA) solution was introduced into each well, and the plates were incubated at 37°C in an incubator for 1-4h. Absorbance at 450 nm was measured, and cell viability was calculated based on the recorded measurements.

Add a minimum of 1 mL of 10 μM DCFH-DA to the cultured cells, enabling its free passage through the cell membrane. Subsequently, intracellular lipase hydrolyzes DCFH-DA into DCFH, which is unable to traverse the cell membrane. Active oxygen within the cell oxidizes DCFH to form DCF, which exhibits fluorescence. The fluorescence intensity of DCF serves as an indicator to measure intracellular active oxygen levels. Incubate the cells for 20 minutes at 37°C in the cell incubator to facilitate the process. Remove any residual DCFH-DA that hasn’t entered the cells, then observe and capture images using a fluorescence microscope.

Proteins extracted from murine hearts and cells were initially separated using SDS-PAGE gel and subsequently transferred onto polyvinylidene fluoride membranes. Following this, the membranes were blocked using either 5% skim milk or bovine serum albumin. Later, they underwent incubation with primary and secondary antibodies and were detected using an ECL detection system.

The obtained results underwent statistical analysis via SPSS 25.0 software. For quantitative data, normality and tests for normality and variance homogeneity were performed. Quantitative data were presented as mean ± standard deviation (Mean ± SD). Between-group comparisons were performed using ANOVA for one-way multilevel quantitative data and ANOVA for repeated measures across multiple time points. Additionally, the Tukey’s HSD was employed for two-way comparisons between groups, Р<0.05 was considered statistically significant.

Basic index analysis showed that compared with the C57BL/6J control group (NC group), KKAy mice showed remarkable metabolic abnormalities (P<0.01), including: accelerated body weight gain ( Figure 2A ), decreased serum 25(OH)D concentrations ( Figure 2B ), elevated FBG levels ( Figure 2C ), impaired glucose tolerance ( Figures 2D–F ), hyperinsulinemia evidenced by increased fasting insulin and HOMA-IR ( Figures 2G, H ), and dyslipidemia manifested by altered TC, LDL-C, and HDL-C levels ( Figures 2J–L ). Notably, 16-week VD₃ supplementation significantly ameliorated all these metabolic abnormalities except for body weight and HDL-C changes (P<0.01).

Compared with the NC group, MC group mice exhibited significant cardiac remodeling (P<0.01), characterized by increased heart weight, cardiomyocyte hypertrophy, and marked disorganization of myocardial fibers with prominent vacuolization ( Figures 3A, B ). Echocardiographic assessment confirmed severe cardiac dysfunction in MC mice (P<0.01), demonstrating elevated LV mass ( Figure 3C ), increased end-diastolic and end-systolic LV volumes (LV Vol;d and LV Vol;s) ( Figures 3E, F ), and enlarged LV internal diameters during both diastole and systole (LVID;d and LVID;s) ( Figures 3G, H ). Concurrently, systolic function was significantly impaired, as evidenced by reduced ejection fraction (EF%) and fractional shortening (FS%) ( Figures 3I, J ) (P<0.01). VD₃ intervention substantially ameliorated these pathological changes, yielding improved myocardial histoarchitecture with reduced hypertrophy, fewer structural discontinuities, and better organized fiber arrangement (P<0.01) ( Figures 3A, B ). Corresponding improvements were observed in echocardiographic parameters of ventricular structure and function. Additionally, VD₃ treatment significantly attenuated the elevated serum levels of cardiac injury markers CK and LDH observed in MC mice (P<0.01) ( Figures 3M, N ). Collectively, these findings demonstrate that VD₃ intervention effectively mitigates both structural and functional cardiac abnormalities in prediabetic mice.

Prediabetic mice exhibited characteristic hyperglycemia and hyperlipidemia, metabolic disturbances known to promote excessive ROS generation and lipid peroxidation, ultimately leading to ferroptosis. Histochemical analyses revealed striking pathological alterations in cardiac tissues of MC group mice, including abundant ferric hemoflavin-containing particles, excessive lipid accumulation, and marked ROS deposition ( Figures 4A–C ). These pathological features were significantly attenuated by VD₃ intervention (P<0.05). The TUNEL assay confirmed increased cardiomyocyte death in MC mice, which was effectively mitigated by VD₃ treatment (P<0.05). ( Figures 4D, E ). Quantitative assessments demonstrated that MC mice displayed elevated cardiac iron content and MDA levels, coupled with depleted GSH reserves compared to NC group (P<0.01) ( Figures 4F–H ). Western blot analysis further elucidated the molecular mechanisms underlying these observations: cardiac tissues from MC mice showed upregulated expression of ferroptosis-promoting proteins (TFR1, ACSL4, NCOA4, and Ferritin), while exhibiting downregulation of the ferroptosis inhibitors SLC7A11 and GPX4 (P<0.01) ( Figures 4I–O ). VD₃ intervention consistently reversed these expression patterns (P<0.05). Collectively, these data demonstrate that ferroptosis contributes to cardiac pathology in prediabetic mice and that VD₃ administration exerts protective effects against this form of programmed cell death.

To further delineate the molecular mechanisms responsible for the cardioprotective effects of VD₃, the expression of key signaling proteins in myocardial tissues were evaluated. Quantitative analysis revealed a significant upregulation of both NRG1 and phosphorylated ErbB4 (p-ErbB4) in MC mice compared to NC group (P<0.01) ( Figures 5A, B ), an effect that was substantially attenuated following VD₃ intervention (P<0.05). Notably, VD₃ treatment also induced a marked increase in phosphorylated YAP (p-YAP) levels, indicating effective activation of YAP (P<0.05) ( Figure 5C ). Collectively, these findings demonstrate that VD₃ ameliorates prediabetic myocardial injury by downregulating NRG1 expression and subsequently affecting the activation of both ErbB4 and YAP signaling cascades.

To further elucidate the mechanisms underlying cardiomyocyte injury, an in vitro model was established using H9c2 cardiomyocytes exposed to HGPA. As depicted in supplementary Figure 1 , exposure to a combination of 33.3 mM HG and 0.25 mM PA for 24h (designated as HGPA) significantly reduced cell viability (P<0.01). Notably, pre-treatment with 10 nM 1,25(OH)₂D₃ for 2h remarkably enhanced cell viability followed by co-culture with HGPA for 24h (P<0.01).

Subsequently, we explored whether inhibiting ferroptosis could alleviate cardiomyocyte injury induced by HGPA. In vitro, cells were initially exposed to 1 μM Fer-1 for 2h, followed by co-incubation with HGPA for 24 h. A significant increase in the cross-sectional area was observed in HGPA+Fer-1 cells compared to HGPA group ( Figures 6A, B ). Additionally, the number of dead cells was notably decreased ( Figures 6C, D ), indicating a degree of recovery from damage. Furthermore, elevated levels of Fe²⁺, MDA, and ROS, along with a substantial decrease in GSH levels, were detected in HGPA cells compared to normal cells ( Figures 6E–I ). Notably, these alterations were reversed upon application of the Fer-1. Western blot analysis was subsequently performed to assess key proteins implicated in ferroptosis. The results revealed a substantial upregulation in the relative expression levels of TFR1, ACSL4, and Ferritin, along with a notable reduction in the relative expression of SLC7A11 and GPX4 following HGPA exposure ( Figures 6J–O ). Intriguingly, these changes were reversed upon treatment with the Fer-1. This study thus delineates the induction of ferroptosis in H9c2 cells by HGPA and highlights the potential of ferroptosis inhibition in mitigating HGPA-triggered cardiomyocyte injury.

During the progression of prediabetic myocardial injury, our investigations revealed concurrent ErbB4 activation, YAP activity suppression, and ferroptosis occurrence. To elucidate the precise interrelationship among these three factors, we conducted in vitro experiments involving pretreatment with 73.7 nM Dac for 2h followed by 24-h co-culture with HGPA to specifically inhibit ErbB4 phosphorylation and monitor subsequent changes. Comparative analysis demonstrated that Dac-treated cells exhibited significantly reduced cell mortality and markedly diminished cellular cross-sectional area ( Figures 7A–D ) relative to HGPA-treated cells. Western blot analysis confirmed effective ErbB4 phosphorylation inhibition ( Figures 7E, F ), which was accompanied by a substantial elevation in phosphorylated YAP levels ( Figures 7G, H ) and concurrent attenuation of ferroptosis markers ( Figures 7I–S ). Notably, the phosphorylation profiles of both ErbB4 and YAP showed no statistically significant differences between Fer-1-treated and Dac-treated groups. These collective findings establish a hierarchical relationship wherein ErbB4 operates upstream of YAP, and its reduced phosphorylation state appears to mitigate YAP-mediated ferroptotic processes in the context of prediabetic myocardial injury.

Our previous in vivo study demonstrated a decrease in p-ErbB4 expression following VD₃ intervention. Subsequently, we explored whether VD mitigates myocardial injury by modulating ErbB4/YAP-mediated ferroptosis. Therefore, in vitro, we added 10 nM 1,25(OH)₂D₃ and 73.7 nM Dacomitinib for 2h pretreatment followed by 24h co-culture with HGPA [1,25(OH)₂D₃+Dac]. Compared with HGPA group, HGPA+1,25(OH)₂D₃ group showed a significant decrease in phosphorylated ErbB4 levels ( Figure 8E ), a significant increase in phosphorylated YAP levels ( Figure 8F ), and an inhibition of ferroptosis ( Figures 8G-Q ). Compared with HGPA+1,25(OH)₂D₃ group, the above indicators of HGPA+1,25(OH)₂D₃+Dac group were reversed more significantly, the cell cross-sectional area was significantly reduced ( Figures 8A, B ), and the number of dead cells was reduced ( Figures 8C, D ). The above analyses suggest that 1,25(OH)₂D₃ inhibits ErbB4/YAP-mediated ferroptosis and rescues myocardial injury, and that p-ErbB4 inhibitor enhance the therapeutic effect of 1,25(OH)₂D₃ on myocardial injury.