Salvianolic Acid B Upregulates miR‐744‐5p Expression in Peripheral Blood Mononuclear Cell–Derived Extracellular Vesicles to Alleviate Uremic Cardiomyopathy

ABSTRACT

Uremic cardiomyopathy (UCM) is a severe complication of uraemia that lacks effective treatments. The role of immune dysfunction in haemodialysis patients with UCM remains unclear. Peripheral blood mononuclear cells (PBMCs) are major components of the immune system; however, they do not directly contact cardiomyocytes. In general, extracellular vesicles (EVs) function as intercellular communication mediators. Therefore, in this study, we first investigated the role of PBMC‐derived EVs (PBMC‐EVs) in UCM and identified EV‐miR‐744‐5p as a key molecule involved in PBMC–cardiomyocyte communication. Mechanistically, indoxyl sulphate (IS) downregulated miR‐744‐5p expression in PBMC‐EVs, leading to IGF2R upregulation in cardiomyocytes, thereby exacerbating myocardial injury via induction of hypertrophy, apoptosis and inflammatory pathway activation in the cardiomyocytes. We also explored the potential of natural products to treat UCM by modifying PBMC‐EVs and found that the traditional Chinese medicine monomer salvianolic acid B (Sal B) could bind to YY1, enhancing miR‐744‐5p expression in PBMC‐EVs and thus mitigating myocardial injury in UCM. Taken together, these findings indicate the critical role of PBMC‐EVs in UCM and suggest the cardioprotective effects of Sal B via PBMC‐EV modification. They also provide novel insights into immune mechanisms underlying UCM, particularly indicating that targeting PBMC‐EVs may be a promising UCM treatment strategy.

Uremic cardiomyopathy (UCM) is a life‐threatening complication of uremia. We find that extracellular vesicle (EV)‐miR‐744‐5p mediates peripheral blood mononuclear cell (PBMC)‐cardiomyocyte communication in UCM. Salvianolic acid B treats UCM by modulating EV‐miR‐744‐5p expression. These findings suggest potential immune mechanisms in UCM and suggest PBMC‐EVs modifying as a potential therapy.

1. Introduction

Patients with uraemia demonstrate 10‐ to 20‐fold higher cardiovascular mortality than the general population, and uremic cardiomyopathy (UCM) contributes significantly to adverse cardiovascular outcomes in these patients (Winterberg et al. 2019; Sárközy et al. 2023; Nguyen and Schulze 2023; Dinh et al. 2024). Indoxyl sulphate (IS) is a prototypical, protein‐bound uremic toxin that mainly drives UCM progression (Curaj et al. 2024; Li et al. 2024). Many studies have focused on the direct damage of IS to cardiomyocytes through mechanisms involving oxidative stress amplification, sustained inflammatory activation and fibrotic tissue remodelling (Barisione et al. 2024; Yamaguchi et al. 2022; Yang et al. 2022; Lekawanvijit et al. 2010). However, >90% of IS in circulation is protein‐bound (Leong and Sirich 2016), which cannot directly cross blood vessel walls and the extracellular matrix to attach to cardiomyocytes. Thus, it is reasonable to propose that IS may induce damage in circulating cardiomyocytes via intercellular communication.

Immune dysfunction is prevalent in patients with uraemia (Acree et al. 2024). Peripheral blood mononuclear cells (PBMCs), the main immune system components, are pivotal in UCM‐associated myocardial injury (Dounousi 2021; Xu et al. 2022; Han et al. 2023). IS increases tissue factor secretion in PBMCs, promoting atherosclerotic thrombosis (Grover and Mackman 2020; Gondouin et al. 2013). This suggests a pathogenic role of PBMCs in the development of cardiovascular complications. Because circulating PBMCs lack direct contact with cardiomyocytes, paracrine mechanisms may be the primary pathways governing intercellular crosstalk between PBMCs and cardiomyocytes. Extracellular vesicles (EVs) are critical agents mediating intercellular communication (Wang et al. 2024; Pinheiro et al. 2024). The role of EVs in uraemic myocardial injury promotion via PBMC–cardiomyocyte crosstalk remains unknown.

Modifying PBMC‐derived EVs (PBMC‐EVs) for therapeutic applications, including suboptimal isolation purity, undefined safety thresholds and the absence of standardized protocols, is clinically challenging (Ma et al. 2025; Liu et al. 2022; Zhang et al. 2021). Thus, natural compounds may be viable alternatives. Salvianolic acid B (Sal B)—the predominant, safe and bioactive polyphenol in the Salvia miltiorrhiza (Danshen)—has multifaceted pharmacological activities including anti‐inflammatory properties, antioxidant capabilities and cardioprotective effects (Fu et al. 2024; He et al. 2023). Emerging evidence indicates that Sal B can modulate immune responses, particularly macrophage‐derived exosome regulation in acute ischemic stroke (Liu et al. 2025).

Here, we, for the first time, assessed the critical role of EV‐mediated interactions between PBMCs and cardiomyocytes in UCM and elucidated pharmacological mechanisms underlying the effects of Sal B on PBMC‐EVs during UCM therapy. Our results confirmed that lowered miR‐744‐5p expression in PBMC‐EVs substantially contributes to UCM development and that Sal B exerts its anti‐UCM effects by binding to the transcription factor YY1 and subsequently upregulating miR‐744‐5p expression in PBMC‐EVs. Therefore, transcriptional modifications of PBMC‐EVs are a promising strategy for UCM treatment.

2. Materials and Methods

2.1. Human Subjects

The PBMC‐EVs from 60 participants—including healthy controls, uraemic patients and patients with uraemic cardiomyopathy—were collected for the detection of miR‐744‐5p expression levels. Uremic patients and uremic cardiomyopathy patients were recruited from the haemodialysis centre of our institution, and all uraemic patients were diagnosed according to the KIDGO 2024 guidelines (KDIGO 2024). Since left ventricular hypertrophy (LVH) is a typical manifestation of UCM, UCM patients were screened from the uraemic cohort based on the left ventricular mass index (LVMI; ≥115 g/m² for males, ≥95 g/m² for females) (Noels et al. 2025; Lang et al. 2015; Edwards et al. 2026). Blood sample collection was performed before dialysis sessions. The following exclusion criteria were applied: (1) left cardiac structural and functional changes secondary to diseases such as hypertension or diabetes; (2) cardiac structural and functional abnormalities caused by primary cardiomyopathies; (3) active infections, malignant tumours or acute severe cardiac diseases. Healthy control subjects were recruited in an age‐ and sex‐matched manner. All participants provided written informed consent. Sample size was determined by resource constraints rather than power calculations. Human data collection was approved by the Ethics Committee of the First Affiliated Hospital of Dalian Medical University. For further details, see the Ethical Statement.

2.2. Animals and Animal Models

Animal use in this study was approved by the Animal Ethics Committee (for details, see Ethics Statement). Specific pathogen–free ICR mice (wildtype, aged 6–8 weeks, weighing 20–30 g) and C57BL/6 mice (wildtype, aged 6–8 weeks, weighing 15–25 g) were sourced from Liaoning Changsheng Co., Ltd (Benxi, China). The animals were housed under constant room temperature (22°C–26°C), 12‐h light–dark cycle, abundant and clean food and drinking water, and ample cage space.

Under isoflurane anaesthesia, a 5/6 nephrectomy (Nx) model was induced by a two‐step surgical procedure in male C57BL/6 mice. We removed two‐thirds of each mouse's left kidney for the first surgical stage. After 1 week of recovery, the mouse was subjected to the second surgical stage, its right kidney was removed, and it was fed with 1% salt water.

In the Sal B treatment group, Sal B (Solarbio, Beijing, China) was administered to the mice at 25 mg/kg daily intraperitoneally for 8 weeks (Fu et al. 2024; Li et al. 2024). For PBMC‐EV‐miR‐744‐5p mimic or inhibitor interventions, PBMCs cultured in 10‐cm² dishes (cell count = 8 × 10⁶) were initially transfected with an miR‐744‐5p mimic or inhibitor, and EVs derived from transfected PBMCs were isolated and delivered weekly to the mice via their tail vein for 8 weeks.

2.3. Cell Culture and Transfection

PBMCs were isolated from 6 to 8‐week‐old ICR mice. Clinical subjects' PBMCs were isolated from autologously obtained blood samples. In brief, the mice were anaesthetized and sacrificed. Next, their eyeballs were excised for blood drawing. We diluted the obtained blood by mixing it with an equal volume of blood Hank's solution in vacuum tubes. Then, the diluted blood was combined with an equal volume of Ficoll separation medium (TBD, Tianjin, China) in the vacuum tubes. The mixture underwent centrifugation at a speed of 1000 revolutions per minute (rpm) for 30 min. PBMCs were then aspirated and transferred into a centrifuge tube filled with phosphate‐buffered saline (PBS) and then centrifuged at 800 rpm for 5 min at room temperature, repeated twice. The collected PBMCs were plated in a 10‐cm² cell culture dish at 37°C under 5% CO². PBMCs were maintained in RPMI‐1640 medium (Cytiva, Marlborough, MA, USA) supplemented with foetal bovine serum (FBS; Gibco, Invitrogen, New York, USA) at a final concentration of 10%, along with penicillin (100 U/mL) and streptomycin (100 µg/mL). The in vitro dose of IS used to stimulate PBMCs was set at its 25% inhibitory concentration (IC₂₅), which induces roughly 25% PBMC death (Figure S1A, B). This concentration was chosen to ensure that PBMCs retain sufficient vitality and bioactivity after stimulation, allowing them to continue secreting EVs under stress. This dosing strategy is widely used to establish appropriate stimulation conditions in cell experiments (Bekeschus et al. 2022; Romero‐Trejo et al. 2025). Cell counting kit‐8 (CCK8) assays identified 50 µM SalB as the optimal effective concentration for counteracting IS‑induced PBMC damage, representing the minimal dose that elicits the maximal therapeutic effect (Figure S1C). In the IS stimulation group, PBMCs were exposed to 100 µM IS (Sigma–Aldrich, St. Louis, MO, USA), whereas those in the Sal B treatment group were administered 50 µM Sal B (Solarbio, Beijing, China) (Han et al. 2011).

Neonatal mouse cardiomyocytes (NMCMs) were harvested from newborn ICR mouse hearts. In brief, we removed the hearts and placed them into cell culture medium containing PBS and 2% penicillin–streptomycin. Next, heart tissues were shredded with ophthalmic scissors in 5‐mL Eppendorf tubes, followed by addition of 1 mL of 0.4% collagenase IV and 0.5 mL of 0.05% pancreatin and then by incubation in a 37°C water bath for 10 min. After centrifuging the supernatant at 3000 rpm for 5 min, NMCMs were collected. These cells were then mixed with complete Dulbecco's modified Eagle medium (DMEM; Cytiva), and the remaining heart tissue was further mixed with collagenase and pancreatin. The aforementioned procedure was repeated four to five times until complete tissue digestion. The purified NMCMs were then cultured in DMEM supplemented with 10% FBS, 1% penicillin–streptomycin and 5‐bromo‐2′‐deoxyuridine (Sigma–Aldrich) to inhibit fibroblast proliferation. The cells were then maintained at 37°C under 5% CO₂. Spontaneous contractions of NMCMs were observed within 24–48 h. The NMCMs were finally cultured in DMEM supplemented with 10% FBS, penicillin (100 U/mL) and streptomycin (100 µg/mL).

When cells reached 50%–70% confluency, oligo complexes were prepared by combining oligos with Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's protocol. The mixture was subsequently applied to six‐well plates containing 2 × 10⁵ cells per well. miR‐744‐5p mimics, inhibitors and corresponding NC oligos were purchased from GenePharma (Suzhou, China). For IGF2R silencing in NMCMs, specific siRNA oligos (GenePharma) were transiently transfected. Moreover, YY1–siRNA (small interfering RNA) oligos and OE plasmids were procured from PPL (Nanjing, China) and YouBio (Changsha, China), respectively.

2.4. Western Blotting

Cells and tissues were collected and lysed with a buffer consisting of RIPA and PMSF (100:1). Next, separating and stacking gels, as well as a sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS‐PAGE) buffer, were prepared. Processed protein samples were loaded at equal amounts (30 mg) per lane into a polyacrylamide gel and subjected to SDS‐PAGE. Then, proteins were electrophoresis‐separated and subsequently transferred onto polyvinylidene fluoride membranes and blocked with gelatin blocking buffer. Subsequently, incubation of the membranes with primary antibodies specific for cleaved caspase 3 (1:500; 9664; Cell Signalling Technology), cleaved caspase 9 (1:500; 20750S; Cell Signalling Technology), p65 (1:500; 8242; Cell Signalling Technology), p‐p65 (1:500; 3033; Cell Signalling Technology), p‐IκKα/β (1:500; 2697; Cell Signalling Technology), IκKα (1:500; 11930; Cell Signalling Technology), IκBα (1:500; 4814; Cell Signalling Technology), p‐IκBα (1:500; 2859; Cell Signalling Technology), p50 (1:500; 13586; Cell Signalling Technology), Erk (1:500; ab17942; Abcam), p‐Erk (1:500; 9101; Cell Signalling Technology), CD9 (1:500; 98327; Cell Signalling Technology), CD81 (1:500; 10037; Cell Signalling Technology), Alix (1:200; sc53540; Santa Cruz), Calnexin (1: 1000, 10427‐2‐AP; Proteintech), p38 (1:1000; 9212; Cell Signalling Technology), p‐p38 (1:1000, 4511; Cell Signalling Technology), Bax (1:500; sc7480; Santa Cruz), Bcl2 (1:500; sc7382; Santa Cruz), caspase 3 (1:500; 19677‐1‐AP; Proteintech), caspase 9 (1:500; ab202068; Abcam), IGF2R (1:500; 14364; Cell Signalling Technology), YY1 (1:500; 66281‐1‐Ig; Proteintech), GAPDH (1:500; 2118; Cell Signalling Technology) was performed following blocking for 24 h. Next, secondary antibodies (1:3,000; 7074, 7076; Cell Signalling Technology) were applied to the membranes and incubated for 4 h. Each blot was detected using Super RX‐N film (Fujifilm, Tokyo, Japan).

2.5. Immunofluorescence Staining

Harvested heart slices from the mice were fixed, embedded in paraffin and sectioned. The cells or tissue samples to be detected were grouped for immunofluorescence staining. In brief, cell samples were prepared using 4% paraformaldehyde and 0.2% Triton X‐100, whereas tissue sample preparation was performed using xylene, anhydrous ethanol, and 0.01 mol/L sodium citrate buffer (pH 6.0). Next, the samples were blocked with goat serum for 30 min and then coincubated with the primary antibodies against Cyt c (1:100; 12963; Cell Signalling Technology), p50 (1:100; 13586; Cell Signalling Technology), p65 (1:100; 8242; Cell Signalling Technology), p‐p38 (1:100, 4511; Cell Signalling Technology), α‐actinin (1:50, 11313‐2‐AP, Proteintech) and cardiac troponin T (1:100; ab209813; Abcam) for 12 h. Thereafter, the samples were incubated with secondary antibodies for approximately 60 min. This was followed by DAPI counterstaining for nucleus visualization. Finally, observations of each sample section were conducted via a confocal laser scanning microscope (Leica SP8; Leica, Wetzlar, Germany).

2.6. Electron Microscopy and Nanoparticle Tracking Analyses

EVs were fixed with 2% paraformaldehyde, incubated with uranyl acetate solution, washed in 100 µL distilled water for 2 min, and observed on a transmission electron microscope (HT‐7700; Hitachi, Tokyo, Japan). For nanoparticle tracking analysis (NTA), we diluted 10 µL of EVs (dissolved in PBS) to 30 µL. Then, the EVs were loaded on an NTA instrument (N30E; NanoFCM, Xiamen, China) to measure their particle size and concentration according to the appropriate standards.

2.7. Transwell Coculture

Transwell inserts equipped with 0.4‐µm pore polycarbonate membranes compatible with six‐well plates were procured from Corning (New York, USA) and employed according to the manufacturer's standardized operational protocols. NMCMs were then plated onto a six‐well transwell plate. Primary PBMCs were resuspended using Diluent C and stained with the PKH67 stain in the PKH67 kit (Sigma–Aldrich). The mixed cell suspension was evenly plated in the upper chamber of the six‐well transwell plate, followed by incubation for 24 h. Finally, the NMCMs were collected for further analysis.

2.8. PBMC‐EV Extraction

EVs were isolated using a standardized differential centrifugation protocol (Singh et al. 2025). In brief, the culture medium containing PBMCs was gathered and underwent a series of sequential ultracentrifugation procedures to isolate EVs. First, samples were ultracentrifuged at 120,000 × g at 4°C for 2 h in an Optima XE‐100/90 Ultracentrifuge (Beckman Coulter, USA) equipped with an SW 32 Ti swinging‐bucket rotor and 38.5‐mL polyallomer tubes. Next, the resulting pellet was gently re‐dispersed in cold PBS, followed by another round of ultracentrifugation under the same conditions (120,000 × g, 4°C, 2 h) to increase EV purity. The supernatant was aspirated entirely, and the final PBMC‐EV fraction was collected.

2.9. Flow Cytometry Analysis for Apoptosis and AO/EB Fluorescent Staining

For flow cytometry analysis, NMCMs were stimulated and harvested, as required. Next, the AV/PI Detection Kit (Keygen Biotech, Nanjing, China) was utilized to quantitatively analyse cell apoptosis. The cell sample was reconstituted with 500 µL of Binding Buffer and subsequently incubated with 5 µL of AV reagent for 10 min. Then, 5 µL of PI dye was added to the solution, which was then left to incubate at room temperature for an additional 5 min. After filtering the cells, we detected the fluorescence signals in the mixture.

For AO/EB fluorescent staining, stimulated NMCMs were collected, as required. A fresh AO/EB working solution was prepared by mixing the components of an AO/EB Double Staining Kit (Maokang Biotechnology, Shanghai, China) in predetermined ratios, according to the manufacturer's instructions. This solution served the purpose of staining the cells. Finally, the cells that had been stained were examined through a fluorescence microscope.

2.10. MiRNA Microarray Assay

Differently expressed miRNAs in EVs from IS‐treated PBMCs were identified using miRNA microarray assays with the assistance of Novegene Bioinformatics Technology (Beijing, China). In brief, the EVs were collected by ultracentrifugation, and total EV‐derived RNA was purified using an exoRNeasy Maxi Kit (Qiagen). Next, assessment of RNA quality and quantity was performed with a NanoPhotometer spectrophotometer (Implen, CA, USA) and Qubit 2.0 Flurometer (Life Technologies, CA, USA), respectively. Small RNA sequencing (SmallRNA‐seq) libraries were prepared using a NEBNext Multiplex Small RNA Library Prep Set for Illumina (NEB, USA), with an RNA input of 10 ng per sample and index codes incorporated for sample multiplexing, according to the manufacturer's protocol. The integrity and quality of the final library were confirmed using the Agilent Bioanalyzer 2100 equipped with DNA High Sensitivity Chips.

2.11. MiRNA Extraction and qRT‐PCR

Total RNA was isolated from cells using the TRIzol reagent (Invitrogen), following the manufacturer's recommended procedure. MiRNAs from PBMC‐EVs and mouse serum were extracted using the Exosome miRNA Easy Kit and Plasma/Serum miRNA Extraction Kit from BIOG (Changzhou, China), respectively, according to the manufacturer's instructions. The NanoDrop 2000 system (Agilent Technologies) was employed to measure the concentration and purity of RNA samples.

The extracted miRNAs were reverse‐transcribed into cDNAs by using a FastKing gDNA Dispelling RT SuperMix kit (Tiangen, Beijing, China), followed by cDNA amplification with the Talent qPCR PreMix (SYBR Green) kit (Tiangen, Beijing, China) on the 7500 FAST Real‐Time PCR system (Thermo Fisher Scientific, Waltham, MA, USA). Relative quantification of miRNA expression was performed using the 2^−ΔΔCt method, with data normalized to a stably expressed internal control gene (Livak and Schmittgen 2001). The results are presented in terms of fold changes compared to the pertinent control group.

We used the following qRT‐PCR primers, purchased from Sangon Biotech (Shanghai, China): u6 (mouse), 5′‐CTCGCTTCGGCAGCACA‐3′ (forward), 5′‐AACGCTTCACGAATTTGCGT‐3′ (reverse); mmu‐miR‐744‐5p, 5′‐CTATCATATGCGGGGCTAGGGC‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTGCTGT‐3 (RT); mmu‐miR‐193b‐5p, 5′‐AACAGACGGGGTTTTGAGGGC‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCATCT‐3′ (RT); mmu‐miR‐181d‐5p, 5′‐AGGCGTGCAACATTCATTGTTGT‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACACCCAC‐3′ (RT); mmu‐miR‐466b‐3p, 5′‐ACGGCACATACATACACGCACA‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCTTAT‐3′ (RT); mmu‐miR‐466a‐3p, 5′‐ACGGCACTATACATACACGCACA‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTCTTAT‐3′ (RT). u6 (human): 5′‐ CTCGCTTCGGCAGCACAT‐3′ (forward), 5′‐TTTGCGTGTCATCCTTGCG‐3′ (reverse); hsa‐miR‐744‐5p, 5′‐CTATCATATGCGGGGCTAGGGC‐3′ (forward), 5′‐ATCCAGTGCAGGGTCCGAGG‐3′ (reverse), 5′‐GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACTGCTGT‐3′ (RT).

2.12. Luciferase Reporter Assay

2.12.1. Assessment of Whether IS Downregulates miR‐744‐5p in PBMC‐EVs and Targets Upregulation IGF2R

NMCMs were transfected with a 3′‐ untranslated region (UTR) luciferase reporter construct (pmirGlo‐IGF2R‐3′ UTR‐WT (wild‐type) or pmirGlo‐IGF2R‐3′ UTR‐Mut (mutant); GenePharma) and stimulated using PBMC‐EVs, transfected with a miRNA mimic (miR‐744‐5p NC or miR‐744‐5p mimic; GenePharma). The luciferase activities were measured using the Dual‐Luciferase Reporter Assay System (Promega, Madison, WI, USA).

2.12.2. Assessment of Whether Sal B Directly Binds to YY1 to Upregulate miR‐744‐5p Expression in PBMC‐EVs

PBMCs were cotransfected with the mutant promoter of mouse miR‐744‐5p vector (pGL3‐Basic Vector; YouBio), OE‐YY1 plasmid (PPL) or si‐YY1 plasmid (PPL). Renilla luciferase is used as the normalization control. The luciferase activity of the cells was then quantified with the Dual Luciferase Assay Kit (Promega).

2.13. Histology and Immunohistochemistry (IHC)

Formalin‐fixed, paraffin‐embedded (FFPE) myocardium sections (3 µm thick) were subjected to haematoxylin and eosin (HE) staining to assess cardiac hypertrophy.

For IHC, the FFPE myocardium sections were dewaxed in xylene, hydrated in ethanol and washed in PBS. Then, we subjected the sections to antigen retrieval by boiling them in 0.01 mol/L sodium citrate buffer (pH 6.0) for 20 min. The sections were subsequently incubated in Endogenous Peroxidase Blocker (ZSGB‐BIO, Beijing, China) for 30 min and then in goat serum for 30 min for blocking. The sections were then coincubated with primary antibodies against IGF2R (1:100; 20253‐1‐AP; Proteintech), ANP (1:100; 27426‐1‐AP; Proteintech), p‐p65 (1:100; sc136548; Santa Cruz), p50 (1:100; sc8414; Santa Cruz) and caspase 3 (1:100; 19677‐1‐AP; Proteintech) at 4°C overnight and then incubated with IHC secondary antibodies (ZSGB‐BIO, Beijing, China) for 40 min. Finally, colour was developed with DAB chromogen, followed by counterstaining with haematoxylin.

2.14. Transthoracic Echocardiography

Transthoracic echocardiography was conducted on anaesthetized mice with the VEVO 2100 echocardiography system (VisualSonics Inc, Toronto, Canada) by using standardized protocols. In brief, after anaesthesia induction, the thoracic region of the mice was completely shaved and coated with acoustic coupling gel to optimise imaging quality. Mice were then placed in the supine decubitus position on a heating pad thermostatically controlled at 37°C to maintain physiological normothermia. Two‐dimensional and M‐mode echocardiographic examinations were systematically performed. Left ventricular posterior wall thickness (LVPWT) was measured across three consecutive cardiac cycles to ensure measurement reproducibility.

2.15. WGA and TUNEL Staining

To examine alterations in cardiomyocyte morphology with WGA staining, the FFPE myocardium sections were processed according to standard immunofluorescence protocols through the antigen retrieval and blocking stages. Then, the sections were stained with 5 µg/mL FITC‐WGA (GeneTex, Irvine, CA, USA) at room temperature for 15 min, followed by three 5‐min PBS washes to remove unbound probes. Fluorescence signals were captured under a confocal laser scanning microscope (Leica SP8, Leica, Wetzlar, Germany) equipped with DAPI counterstaining and appropriate filter sets for multichannel visualization.

For TUNEL staining, the FFPE myocardium sections were treated with xylene and ethanol, followed by 20‐min treatment with protease K (Invitrogen). The TUNEL reaction mixture (Roche, Basel, Switzerland) was prepared according to the manufacturer's protocol. The sections were then incubated with this mixture at 37°C for 60 min in a humidified dark chamber. Next, DAPI reagent was used to stain nuclei. After PBS washing, the sections were sealed with glycerin gelatin.

2.16. Ex Vivo Animal Imaging

The 5/6 Nx model was surgically induced in mice, and EVs were fluorescently labelled using DiR (MCE, NJ, USA). Subsequently, DiR‐labelled EVs were intravenously administered to the 5/6 Nx mice via the tail vein at 100 µg per mouse. In vivo EV biodistribution was monitored using the FCZAA8‐E1JHAA imaging system (Bruker, Bremen, Germany), and whole‐body fluorescence imaging was performed 24 h after injection.

2.17. Surface Plasmon Resonance (SPR) Assay

Target protein was immobilized on a sensor chip by using amine coupling methodology under optimized buffer conditions (coupling buffer: 1× PBS‐P, pH 7.4; interaction buffer: 1× PBS‐P containing 5% v/v dimethyl sulfoxide, pH 7.4). After solvent background subtraction, serial dilutions of test compounds were prepared in 96‐well plates and injected across the chip surface at a flow rate of 30 µL/min and contact time per concentration of 150 s. Concentration gradients were sequentially introduced from low to high concentrations to monitor binding progression. Between analyte injections, surface regeneration was achieved through 5‐min exposure to 10 mM glycine‐HCl (pH 2.0), which ensured complete dissociation of bound ligands. Real‐time binding data were acquired through BIAcore T200 Control (version 2.0; GE Healthcare, Chicago, IL, USA) and subsequently processed using BIAcore T200 Evaluation (version 2.0; GE Healthcare) for kinetic parameter calculation. The sensorgram data underwent global fitting using a 1:1 Langmuir binding model, which enabled the determination of association and dissociation rate constants. These constants were then used to calculate the equilibrium dissociation constant (KD).

2.18. Cellular Thermal Shift Assay (CETSA)

Cell lysates were treated with 200 µM Sal B or PBS (NC) at 4°C for 120 min. Subsequently, temperature‐dependent treatments were performed by incubating the lysates at varying temperatures ranging from 37°C to 73°C at 4°C increments (i.e., 37°C, 41°C, 45°C, 49°C, 53°C, 57°C, 61°C, 65°C, 69°C, and 73°C) for 5 min each, followed by centrifugation. The resulting supernatants were subjected to Western blotting using YY1 antibodies.

2.19. Drug Affinity Responsive Target Stability (DARTS) Assay

Cell lysates were pretreated with 200 µM Sal B or PBS (NC) for 30 min, aliquoted into five portions, and digested with Pronase E (MCE) at varying concentrations at room temperature for 15 min. Next, digestion was terminated, followed by centrifugation at 12,000 × g at 4°C for 10 min to pellet undigested cellular debris. The resulting supernatant was analyzed through Western blotting with YY1 antibodies.

2.20. CCK8 Assay

PBMCs were seeded in 96‐well plates at a density of 5000–10,000 cells/well and treated with increasing concentrations of IS and Sal B for 24 h. Cell viability was assessed using the CCK‐8 assay (Solarbio, Beijing, China), and absorbance was measured at 450 nm. The 25% inhibitory concentration (IC25) was calculated using non‐linear regression analysis with GraphPad Prism (version 9.0; GraphPad Software, CA, USA).

2.21. Chromatin Immunoprecipitation (ChIP) Assays

The ChIP assay was performed in accordance with the manufacturer's instructions (Thermo Fisher Scientific, Waltham, MA, USA). Briefly, a total of 4 × 10⁶ PBMCs were fixed with 1% formaldehyde, neutralized by glycine, and then resuspended in lysis buffer supplemented with protease inhibitors. The cells were lysed using an ultrasonic cell disruptor (SCIENTZ, Ningbo, China) at 30% power for 30 s on and 30 s off, repeated a total of 40 cycles. Subsequently, chromatin was immunoprecipitated with 10 µg of anti‑YY1 antibody (22156‐1‐AP, Proteintech), with 1 µg of IgG serving as the negative control. DNA extracts were amplified via qPCR using primers purchased from Sangon Biotech (Shanghai, China) with the following sequences for the miR‑744‑5p promoter: 5′‐GGCAGTTAAAGTAGGTGATGCC‐3′ (forward);5′‐GAACAGAGCAGTGGTGTTGC‐3′ (reverse).

2.22. Statistical Analysis

All experimental data, presented as means ± SDs, were subjected to statistical analysis using Student's t test for direct comparisons between pairs of groups or using one‐way analysis of variance (ANOVA) plus Tukey's post hoc test to evaluate differences across multiple groups in GraphPad Prism (version 9.0; GraphPad Software, CA, USA). For the clinical patients, the clinical characteristics were compared among the three groups. The p values for continuous variables were calculated using ANOVA or the Kruskal–Wallis test, while the p values for categorical variables were computed using either the χ² test or Fisher's exact test. The Pearson correlation analysis was employed to assess the relationship between miRNA expression and LVMI. All data were analyzed and plotted using GraphPad Prism. A p value less than 0.05 is generally interpreted as indicating statistical significance, ns means no difference.

3. Results

3.1. IS‐Stimulated PBMC‐EVs Trigger Cardiomyocyte Damage

We first extracted mouse PBMCs and NMCMs (Figure S1D‐F). Treatment with IS‐PBMC culture medium triggered cardiomyocyte hypertrophy, apoptosis and activation of proinflammatory signalling cascades (Figure 1A‐D). Next, EVs were extracted from mouse PBMC supernatant (Figure S2A‐C). EVs derived from IS‐stimulated PBMCs (PBMC^IS‐EVs) were internalized by NMCMs, which induced cardiomyocyte hypertrophy and a reduction in spontaneous beating rate (Figure 1E, F; Figure S2D‐F) and increased cardiomyocyte apoptosis considerably (Figure 1G, H). It also upregulated the expression of the proapoptosis proteins Bax, cleaved caspase 3 and cleaved caspase 9 and downregulated that of the antiapoptosis protein Bcl2 in NMCMs (Figure 1I), along with increasing Cyt c release from mitochondria (Figure 1J). Furthermore, IκKα/β, p‐IκKα/β, IκBα, p‐IκBα, p50, p65, p‐p65 and p‐p38 expression (Figure 1K; Figure S2G, H), as well as p50 and p65 translocation into the nucleus (Figure 1L), increased. However, treatment with GW4869 (an exosome biogenesis inhibitor) restored cardiomyocyte damage induced by PBMC^IS‐EVs (Figure 1E‐L; Figure S2E‐H). Thus, PBMC‐EVs may play the primary role of IS‐induced cardiomyocyte damage in uraemia.

FIGURE 1.

IS‐stimulated PBMC‐EVs are critical mediators of cardiomyocyte damage. (A) Immunofluorescence staining using phalloidin (red), with nuclei counterstained using DAPI (blue), to assess cardiomyocyte hypertrophy in NMCMs treated with conditioned medium from IS‐stimulated PBMCs (bar = 50 µm, n = 3). (B) Flow cytometry analysis of cardiomyocyte apoptosis using Annexin V‐FITC/propidium iodide (AV/PI) staining in NMCMs treated with conditioned medium from IS‐stimulated PBMCs (n = 3). (C and D) Western blotting of apoptosis‐related proteins (cleaved caspase 3, Bax and Bcl2; n = 3; C) and inflammation‐related proteins (p65, p‐p65, Erk, p‐Erk, p38 and p‐p38; n = 3; D) in NMCMs treated with conditioned medium from IS‐stimulated PBMCs. (E) Immunofluorescence staining using phalloidin (red), with nuclei counterstained using DAPI (blue), to assess cardiomyocyte hypertrophy in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (bar = 50 µm, n = 3). (F) Quantitative analysis of beating frequency and amplitude in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both, observed through optical microscopy (n = 3). (G) Cardiomyocyte apoptosis analysis through flow cytometry using AV/PI staining in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (n = 3). (H) Representative images of fluorescence staining using acridine orange/ethidium bromide in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (bar = 50 µm). (I) Western blotting of apoptosis‐related proteins (pro caspase 9, cleaved caspase 9, pro caspase 3, cleaved caspase 3, Bax and Bcl2) in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (n = 3). (J) Representative immunofluorescence images of Cyt c (green) expression in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (bar = 50 µm, n = 3). (K) Western blotting of NF‐κB pathway proteins (IκKα/β, p‐IκKα/β, IκBα, p‐IκBα, p50, p65 and p‐p65) in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (n = 3). (L) Representative immunofluorescence staining demonstrating nuclear translocation of p65 (green) and p50 (red), with nuclei counterstained using DAPI (blue), in NMCMs treated with EVs derived from IS‐treated PBMCs, GW4869‐treated PBMCs or both (bar = 50 µm, n = 3). For each biological replicate in A, E, J and L, two independent fields were analyzed as technical replicates. Student's t test for pairwise comparisons in A–D and one‐way ANOVA for multiple comparisons in E–G and I–L. The data are presented with average values and standard deviations (means ± SDs).

3.2. PBMC^IS‐EVs Downregulate miR‐744‐5p Expression

Through smallRNA‐seq, we screened out miR‐744‐5p and determined it to be the most significantly downregulated miRNA in the PBMC^IS‐EVs through species‐specific qRT‐PCR identification (Figure 2A, B; Figure S3A). Moreover, the miR‐744‐5p sequence is highly conserved across species (Figure 2C).

FIGURE 2.

MiR‐744‐5p downregulation in IS‐stimulated PBMC‐EVs is pivotal in cardiomyocyte damage. (A and B) The differential expression of miRNAs was visualized using a volcano plot (A) and a Heatmap (B), comparing IS‐stimulated and control PBMC‐EVs via smallRNA‐seq (n = 3). (C) Sequence conservation analysis of miR‐744‐5p between humans and mice using the miRDB database. (D) qRT‐PCR quantification of miR‐744‐5p in PBMC‐EVs after protease K and RNase treatment (n = 3). (E) Representative immunofluorescence images showing uptake of GFP‐miR‐744‐5p‐transfected PBMC‐EVs (PKH26‐labelled) in NMCMs in a transwell coculture system (NMCMs and PBMCs in the lower and upper chambers, respectively; bar = 30 µm, n = 3). Left: Schematic of the transwell system; Right: Fluorescence images (GFP: green; PKH26: red; DAPI: blue). (F) Immunofluorescence staining using phalloidin (red), with nuclei counterstained using DAPI (blue), to assess cardiomyocyte hypertrophy in NMCMs stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p mimics or inhibitors (bar = 50 µm, n = 3). (G) Flow cytometry analysis of cardiomyocyte apoptosis using AV/PI staining in NMCMs stimulated with IS‐PBMC‐EVs received miR‐744‐5p mimics or inhibitors for transfection (n = 3). (H) Western blotting of apoptosis‐related proteins (pro caspase 9, cleaved caspase 9, pro caspase 3, cleaved caspase 3, Bax and Bcl2) in NMCMs stimulated with IS‐PBMC‐EVs received miR‐744‐5p mimics or inhibitors for transfection (n = 3). (I) Representative immunofluorescence images of Cyt c (green) expression in NMCMs stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p mimics or inhibitors (bar = 50 µm, n = 3). (J) Western blotting of NF‐κB pathway proteins (IκKα/β, p‐IκKα/β, IκBα, p‐IκBα, p50, p65 and p‐p65) in NMCMs stimulated with IS‐PBMC‐EVs received miR‐744‐5p mimics or inhibitors for transfection (n = 3). (K) Representative immunofluorescence staining demonstrating nuclear translocation of p65 (green) and p50 (red), with nuclei counterstained using DAPI (blue), in NMCMs stimulated with IS‐PBMC‐EVs received miR‐744‐5p mimics or inhibitors for transfection (bar = 50 µm, n = 3). For each biological replicate in F, I and K, two independent fields were analyzed as technical replicates. Student's t test for pairwise comparisons in D and one‐way ANOVA for multiple comparisons in F–K. The data are presented with average values and standard deviations (means ± SDs).

Our EV RNA degradation assay revealed that miR‐744‐5p in PBMCs was predominantly present in PBMC‐EVs (hereinafter, PBMC‐EV‐miR‐744‐5p; Figure 2D). By using a transwell coculture system (Figure 2E), we confirmed that miR‐744‐5p was encapsulated and transported by PBMC‐EVs to cardiomyocytes. Moreover, the role of PBMC‐EV‐miR‐744‐5p in UCM was identified using miR‐744‐5p mimics or inhibitors (Figure S3B). The results demonstrated that NMCMs induced by PBMC‐EVs transfected with miR‐744‐5p mimics (hereinafter, miR‐744‐5p overexpression PBMC‐EVs) demonstrated decreases in hypertrophy, apoptosis, NF‐κB and p38 MAPK pathway activity (Figure 2F‐K; Figure S3C‐E). In contrast, treatment with PBMC‐EVs transfected with miR‐744‐5p inhibitors (miR‐744‐5p knockdown PBMC‐EVs) considerably aggravated cardiomyocyte damage (Figure 2F‐K; Figure S3C‐E). Thus, miR‐744‐5p, the PBMC‐EV cargo, is critical in UCM‐related cardiomyocyte damage.

Subsequently, we collected peripheral blood samples from UCM patients for the detection of PBMC‐EV‐miR‐744‐5p. The demographic characteristics of the UCM patients are detailed in Table S1. qPCR results indicated a significant reduction in circulating PBMC‐EV‐miR‐744‐5p levels in UCM patients compared to healthy controls (p < 0.0001; Figure S3F). A decreasing trend in PBMC‐EV‐miR‐744‐5p expression was also observed in UCM patients compared with non‐cardiac hypertrophy uraemic patients, but this difference did not reach statistical significance (Figure S3F). Notably, PBMC‐EV‐miR‐744‐5p expression exhibited a moderate negative correlation with LVMI (r = –0.4123, p = 0.0011; Figure S3G). Collectively, these findings suggest that the downregulation of PBMC‐EV‐miR‐744‐5p is associated with cardiac hypertrophy in UCM patients.

3.3. IGF2R Is an miR‐744‐5p Target

IGF2R was predicted as a candidate target of miR‐744‐5p through integrated screening across three online miRNA target prediction platforms: mirDIP, RNA22 and miRWalk (Figure 3A). Moreover, miR‐744‐5p was predicted to directly interact with the 3′‐UTR of IGF2R mRNA (Figure 3B). The luciferase activity of the IGF2R WT reporter was notably decreased in NMCMs by PBMC‐EV‐miR‐744‐5p overexpression, whereas no such effect was observed on the IGF2R Mut reporter (Figure 3C). Furthermore, miR‐744‐5p overexpression PBMC‐EVs significantly reduced IGF2R expression in NMCMs, whereas miR‐744‐5p knockdown PBMC‐EVs upregulated it (Figure 3D). Therefore, IGF2R is a direct target of miR‐744‐5p.

FIGURE 3.

IGF2R is the direct target gene of miR‐744‐5p, and it exacerbates cardiomyocyte damage. (A) A Venn diagram showing the intersection of predicted targets from three online miRNA databases (miRDIP: yellow, RNA22: green, miRWalk: red) with NF‐κB pathway‐related genes (blue), identifying IGF2R as an miR‐744‐5p target. (B) Bioinformatic analysis using miRanda (an online miRNA/target mRNA binding prediction tool) demonstrating strong binding between miR‐744‐5p and IGF2R mRNA. (C) The dual‐luciferase reporter assay was conducted with plasmids containing either the WT or MUT IGF2R 3′‐UTR. NMCMs transfected with WT or MUT plasmids were stimulated with PBMC‐EVs transfected with miR‐744‐5p mimics. The Dual‐Luciferase Reporter Assay system was used for measuring luciferase activity (n = 3). (D) Western blotting of IGF2R expression in NMCMs treated with PBMC‐EVs transfected with miR‐744‐5p mimics or inhibitors (n = 3). (E) Immunofluorescence staining using phalloidin (red) to assess cell size of NMCMs (transfected with si‐IGF2R or negative control [NC]) stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors or NC (bar = 50 µm, n = 3). (F) Flow cytometry analysis using AV/PI staining for cardiomyocyte apoptosis in NMCMs (transfected with si‐IGF2R or NC) stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors or NC (n = 3). (G) Western blotting of IGF2R, p50 and p‐p65 NMCMs (transfected with si‐IGF2R or NC) stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors or NC (n = 3). (H) Representative immunofluorescence staining demonstrating nuclear translocation of p65 (green) and p50 (red), with nuclei counterstained using DAPI (blue), in NMCMs (transfected with si‐IGF2R or NC) stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors or NC (bar = 50 µm, n = 3). For each biological replicate in E and H, two independent fields were analyzed as technical replicates. One‐way ANOVA for multiple comparisons. The data are presented with average values and standard deviations (means ± SDs).

To further investigate the role of IGF2R in PBMC^IS‐EV‐induced cardiomyocyte damage, IGF2R siRNA was transfected into NMCMs (Figure S4A). Rescue experiments demonstrated that IGF2R silencing partially attenuated cardiomyocyte hypertrophy, apoptosis increase, IGF2R expression upregulation, p38 MAPK and NF‐κB pathway activation mediated by miR‐744‐5p knockdown PBMC‐EVs (Figure 3E‐H; Figure S4B‐D). Taken together, these findings indicated that IGF2R upregulation is the downstream mechanism underlying decreased miR‐744‐5p expression in PBMC‐EVs, which drives UCM.

DiR‐labelled PBMC‐EVs were then administered to 5/6 Nx mice via the tail veins, and the results data revealed that PBMC^IS‐EVs exhibited preferential cardiac tissue accumulation compared with the control groups (Figure 4A). Moreover, PKH26‐labelled PBMC‐EVs transported GFP‐miR‐744‐5p to the hearts of 5/6 Nx mice (Figure 4B). In 5/6 Nx mice administered PBMC‐EV‐miR‐744‐5p mimics via the tail vein, isolated serum‐derived EVs demonstrated strong miR‐744‐5p expression (Figure 4C). Thus, PBMC‐EV‐miR‐744‐5p may have been transported to the UCM mouse hearts via circulation.

FIGURE 4.

Administration of miR‐744‐5p overexpression PBMC‐EVs downregulates cardiac IGF2R expression and exerts cardioprotective effects in mice. (A) Images of organs (heart, liver, spleen, lung, kidney and brain) from a 5/6 Nx mouse 24 h after tail vein injection of PBS (solvent for PBMC‐EVs; negative control), PBMC^Ctrl‐EVs and PBMC^IS‐EVs, as visualized on an ex vivo imaging system. (B) Representative immunofluorescence images showing cardiomyocytes uptake of PKH26‐labelled PBMC‐EVs (red) transfected with GFP‐miR‐744‐5p (green) in 5/6 Nx mouse heart (bar = 100 µm). (C) qRT‐PCR qualification of miR‐744‐5p in serum EVs from sham and 5/6 Nx mice injected with miR‐744‐5p‐transfected PBMC‐EVs (n = 6). (D) Schematic of the in vivo experimental procedure. Tissues were collected at the endpoint for sample preparation. (E) Western blotting of IGF2R in mouse cardiac tissues (n = 6). (F) Morphological observation and HE staining of heart tissue. (G and H) Immunofluorescence staining with α‐actinin (G, green) and WGA (H, green) to assess cardiomyocyte hypertrophy in mouse heart tissue (scale bars: 10 µm for α‐actinin, 100 µm for WGA; n = 6). (I) Echocardiographic evaluation of cardiac function, including LVPWT, during systole (n = 6). (J) IHC of IGF2R, ANP, p‐p65, p50 and caspase 3 in mouse heart tissue (bar = 50 µm). (K) Representative immunofluorescence staining demonstrating nuclear translocation of p65 (green) and p50 (red), with nuclei counterstained using DAPI (blue; bar = 40 µm). (L) Representative immunofluorescence staining showing TUNEL‐positive cells (red; apoptotic cells), with quantification of apoptotic cells (bar = 75 µm, n = 6). For each biological replicate in H and L, two independent fields were analyzed as technical replicates. One‐way ANOVA for multiple comparisons. The data are presented with average values and standard deviations (means ± SDs).

Administration of miR‐744‐5p overexpression PBMC‐EVs further downregulated IGF2R expression in 5/6 Nx mouse hearts (Figure 4D,E); it also alleviated cardiac hypertrophy and sarcomere pathological disarray (Figure 4F‐H; Figure S4E), reduced hypertrophy marker ANP (Figure 4J), improved cardiac function (Figure 4I), decreased NF‐κB pathway and p38 MAPK pathway activation (Figure 4J, K; Figure S4F, G), and lowered cardiomyocyte apoptosis (Figure 4J, L).

3.4. Sal B Exerts Therapeutic Effects in UCM by Enhancing miR‐744‐5p Expression in PBMC‐EVs

The aforementioned results suggested that injecting miR‐744‐5p overexpressing PBMC‐EVs is a promising strategy for UCM treatment. To improve the safety and efficacy of clinical treatment strategies, we investigated whether Sal B, a traditional Chinese medicine (TCM) monomer with proven safety, exerts cardioprotective effects by modifying PBMC‐EVs. First, we noted that miR‐744‐5p expression in Sal B‐treated PBMC‐EVs (PBMC^{IS+Sal B}‐EVs) was upregulated (Figure S5A), indicating that Sal B may play a protective role by modifying PBMC‐EVs through transcriptional upregulation of miR‐744‐5p (Figure S5B‐M). We further found that Sal B significantly reversed the decreased miR‐744‐5p expression in miR‐744‐5p knockdown PBMC‐EVs (Figure 5A). Moreover, Sal B partially abolished the IGF2R expression increase and cardiomyocyte damage induced by miR‐744‐5p knockdown PBMC‐EVs (Figure 5B‐G; Figure S6A‐C).

FIGURE 5.

PBMC‐EV‐miR‐744‐5p is a key molecule mediating cardioprotective effects of Sal B. NMCMs were stimulated with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors, treated with Sal B, or both. (A) Expression of miR‐744‐5p in PBMC‐EVs was measured through qRT‐PCR (n = 3). (B) NMCM size observed using phalloidin (red) staining via immunofluorescence staining (bar = 50 µm, n = 3). (C) Flow cytometry analysis of cardiomyocyte apoptosis using AV/PI staining (n = 3). (D) Western blotting of apoptosis‐related proteins (pro caspase 9, cleaved caspase 9, pro caspase 3, cleaved caspase 3, Bax and Bcl2) in NMCMs (n = 3). (E) Representative immunofluorescence staining of Cyt c (green) in NMCMs (bar = 50 µm, n = 3). (F) Western blotting analysis of IGF2R and NF‐κB pathway‐related proteins (IκKα/β, p‐IκKα/β, IκBα, p‐IκBα, p50, p65 and p‐p65) in NMCMs (n = 3). (G) Representative immunofluorescence staining images demonstrating nuclear translocation of p65 (green) and p50 (red), with nuclei counterstained using DAPI (blue, bar = 50 µm, n = 3). For each biological replicate in B, E and G, two independent fields were analyzed as technical replicates. One‐way ANOVA for multiple comparisons. The data are presented with average values and standard deviations (means ± SDs).

We next administered miR‐744‐5p knockdown PBMC‐EVs and Sal B to 5/6 Nx mice (Figure 6A) and noted that Sal B significantly upregulated miR‐744‐5p expression in their serum EVs (Figure 6B) but downregulated IGF2R expression in their hearts (Figure 6C). These changes were accompanied by cardiac hypertrophy attenuation (Figure 6D‐F; Figure S6D), significant cardiac function improvement (Figure 6G), NF‐κB and p38 MAPK pathway activation decrease (Figure 6H, I; Figure S6E, F), and cardiomyocyte apoptosis reduction (Figure 6H, J). The cardioprotective effects of Sal B in mouse hearts were significantly reduced after treatment with miR‐744‐5p knockdown PBMC‐EVs (Figure 6B‐J; Figure S6D‐F). Thus, Sal B may exert cardioprotective effects mainly by upregulating miR‐744‐5p expression in PBMC‐EVs.

FIGURE 6.

Sal B exerts cardioprotective effects against cardiac damage in UCM mice by upregulating PBMC‐EV‐miR‐744‐5p expression. 5/6 Nx mice were injected with Sal B, with IS‐PBMC‐EVs transfected with miR‐744‐5p inhibitors, or with both. Heart tissues and serum were collected at the modelling endpoint for subsequent analyses. (A) Schematic of the experimental procedure. (B) Expression of miR‐744‐5p in serum EVs detected by qRT‐PCR (n = 6). (C) Western blotting of IGF2R in heart tissue (n = 6). (D) Morphological observation and HE staining of heart tissue sections. (E and F) α‐actinin (E, green) and WGA (F, green) immunofluorescence staining for cardiomyocyte size and sarcomere structure (scale bars: 10 µm for α‐actinin, 100 µm for WGA; n = 6). (G) Echocardiographic evaluation of cardiac function, including LVPWT, during systole (n = 6). (H) IHC of IGF2R, ANP, p‐p65, p50 and caspase 3 in heart tissue sections (bar = 50 µm). (I) Representative immunofluorescence staining for nuclear translocation of p65 (green) and p50 (red; bar = 40 µm). (J) TUNEL staining (red) showing apoptotic cells in heart tissue sections (bar = 75 µm, n = 6). For each biological replicate in F and J, two independent fields were analyzed as technical replicates. One‐way ANOVA for multiple comparisons. The data are presented with average values and standard deviations (means ± SDs).

3.5. Sal B Modulates miR‐744‐5p/IGF2R Axis by Binding to YY1 in PBMCs

We next investigated how Sal B upregulates miR‐744‐5p expression in PBMC‐EVs by using bioinformatic databases and software. The results identified that the transcription factor YY1 binds to the potential promoter region of the miR‐744‐5p gene (Figure 7A). Subsequently, the dual‐luciferase reporter assay demonstrated YY1 overexpression significantly suppressed miR‐744‐5p expression in PBMCs, whereas YY1 knockdown increased it (Figure 7B). Furthermore, we performed ChIP assays to assess YY1 binding to the miR‐744‐5p promoter. YY1 exhibited significant enrichment at this promoter; Sal B treatment markedly reduced this binding (Figure 7C). Therefore, YY1 can negatively regulate transcriptional miR‐744‐5p gene activation in PBMCs, while Sal B can inhibit this process.

FIGURE 7.

Sal B directly binds to YY1 and upregulates miR‐744‐5p expression in PBMC‐EVs to alleviate cardiomyocyte damage. (A) Schematic of transcription factor YY1 binding to the promoter region of the miR‐744‐5p gene. Utilize JASPAR (https://jaspar.elixir.no/) and Promoter ‐ 2.0 (https://services.healthtech.dtu.dk/services/Promoter‐2.0/) to predict potential promoter regions. (B) Dual luciferase reporter assay showing the fluorescence intensity of miR‐744‐5p promoter activity in PBMCs transfected with NC, YY1 overexpression plasmids (OE) or YY1 siRNA (n = 3). (C) The effect of PBS or Sal B on YY1‐mediated regulation of miR‐744‐5p expression in PBMCs was examined using ChIP‐qPCR. (D) AutoDock simulation illustrating direct binding between Sal B and YY1. (E) SPR of Sal B–YY1 interaction (n = 3). (F and G) CETSA and DARTS assay confirming Sal B–YY1 interaction (n = 3). (H) qRT‐PCR of miR‐744‐5p expression in PBMC‐EVs (n = 3). (I) Immunofluorescence staining using phalloidin (red), with nuclei counterstained using DAPI (blue), to assess cardiomyocyte hypertrophy in NMCMs stimulated with IS‐PBMC‐EVs treated with Sal B, transfected with OE YY1/si‐YY1 plasmids, or both (bar = 50 µm, n = 3). (J) Flow cytometry analysis of cardiomyocyte apoptosis using AV/PI staining in NMCMs stimulated with IS‐PBMC‐EVs treated with Sal B, transfected with OE YY1/si‐YY1 plasmids, or both (n = 3). (K) Western blotting of apoptosis‐related proteins (pro caspase 9, cleaved caspase 9, pro caspase 3, cleaved caspase 3, Bax and Bcl2) in NMCMs stimulated with IS‐PBMC‐EVs treated with Sal B, transfected with OE YY1/si‐YY1 plasmids, or both (n = 3). (L) Western blotting of NF‐κB pathway proteins (IκKα/β, p‐IκKα/β, IκBα, p‐IκBα, p50, p65 and p‐p65) in NMCMs stimulated with IS‐PBMC‐EVs treated with Sal B, transfected with OE YY1/si‐YY1 plasmids, or both (n = 3). For each biological replicate in I, two independent fields were analyzed as technical replicates. One‐way ANOVA for multiple comparisons. The data are presented with average values and standard deviations (means ± SDs).

We then investigated the binding affinity between Sal B and YY1 by using molecular docking analysis and SPR assays. Sal B and YY1 were noted to exhibit a direct interplay, with the equilibrium dissociation constant (KD) of approximately 3.95 × 10⁻⁶ M (Figure 7D, E). We then performed CETSA to assess the thermal stabilization of YY1, as well as DARTS assay to evaluate alterations in protease susceptibility of YY1. The results demonstrated that after binding to Sal B, the thermal stability of YY1 decreased, but its antiprotease activity increased (Figure 7F, G).

PBMCs were transfected with YY1 overexpression plasmid and siRNA and then treated with IS or Sal B treatment (Figure S7A, B). We noted that YY1 overexpression significantly reduced miR‐744‐5p expression in PBMC‐EVs, and Sal B reversed this reduction (Figure 7H). Subsequent analysis revealed that EVs extracted from YY1 overexpression PBMCs exacerbated cardiomyocyte hypertrophy, increased apoptosis, upregulated p38 MAPK and NF‐κB‐related protein expression; these changes were attenuated in YY1‐overexpressing PBMCs treated with Sal B (Figure 7I‐L; Figure S7C‐E). However, combined treatment with Sal B and YY1 knockdown in PBMC did not have synergistic cardioprotective effects compared with YY1 knockdown alone in protecting NMCMs from damage (Figure 7H ‐ L; Figure S7C‐E). Therefore, Sal B–YY1 interactions in PBMCs and subsequent upregulation of miR‐744‐5p expression may be a pivotal mechanism through which Sal B ameliorates cardiomyocyte damage.

4. Discussion

Immunopathological mechanisms underlying UCM pathogenesis remain unclear. Our findings revealed that PBMCs are critical to cardiomyocyte damage in UCM (Figure 1A‐D). EVs are pivotal molecules in the intercellular crosstalk between IS‐stimulated PBMCs and cardiomyocytes, which ultimately leads to cardiomyocyte damage (Figure 1E‐L). To our knowledge, this is the first study to explore PBMC–cardiomyocyte communication mechanisms in UCM pathogenesis. Given the nanoscale features of EVs (van Niel et al. 2022), we propose that the pathological mechanism involving cardiomyocyte damage triggered by IS‐stimulated PBMC‐EVs is more physiologically relevant than direct cardiac damage induced by circulating macromolecular IS–protein complexes.

In the current study, we conducted in vivo experiments using a 5/6 Nx mouse model. 5/6 Nx is a validated surgical induction method for UCM, which induces renal insufficiency and secondary myocardial damage (McMahon et al. 2006; Yang et al. 2022; Chen et al. 2021). Moreover, the 5/6 Nx mouse model exhibits significant pathological concordance with human uraemia‐associated cardiac pathology (Navarro‐García and Ruiz‐Hurtado 2022). Our results confirmed efficient construction of this model, and subsequent PBMC‐EV administration resulted in cardiac targeting, leading to cardioprotective effects (Figure 4).

We next elucidated the role of miR‐744‐5p in UCM through a series of experiments. First, IS stimulation led to downregulation of miR‐744‐5p expression in PBMC‐EVs (Figure 2A, B; Figure S3A), and miR‐744‐5p overexpression or knockdown PBMC‐EVs correspondingly alleviated or aggravated UCM (Figures 2F‐K and 4). Second, miR‐744‐5p was specifically transported by PBMC‐EVs to cardiomyocytes or mouse myocardial tissues (Figures 2E and 4B). Studies have found that diminished miR‐744‐5p expression is negatively correlated with pulmonary fibrosis, colorectal carcinoma and osteosarcoma severity (Li et al. 2021; Zhang et al. 2021; Xie et al. 2024), whereas the highly expressed miR‐744‐5p exerts a therapeutic effect in obstructive sleep apnoea–induced cardiomyocyte damage, acute lung injury and glioblastoma (Wang et al. 2024; Qian et al. 2024; Fan et al. 2021). MiR‐744‐5p has critical regulatory roles in various diseases. As such, manipulating miR‐744‐5p expression may not only provide new targets for UCM treatment but also have wide, multidisease therapeutic potential without negative effects in other organs. miRNAs hold great promise as biomarkers across a broad spectrum of diseases (Lee et al. 2023; Rohm et al. 2025). Therefore, we investigated the expression of PBMC‐EV‐miR‐744‐5p in UCM patients. Our findings revealed a significant reduction in PBMC‐EV‐miR‐744‐5p levels in UCM patients compared to healthy controls, which aligns with the results observed in both in vitro IS‐stimulated PBMCs and in vivo 5/6 nephrectomy mouse model (Figure S3F). Furthermore, PBMC‐EV‐miR‐744‐5p exhibited a moderate negative correlation with LVMI, an indicator of myocardial hypertrophy (Figure S3G). These findings not only confirm that PBMC‐EV‐miR‐744‐5p is a pivotal molecule in the pathogenesis of UCM but also highlight its significant potential as a novel clinical diagnostic marker for UCM. However, we observed considerable individual variability in PBMC‐EV‐miR‐744‐5p expression among uraemia and UCM patients. While there was a decreasing trend in PBMC‐EV‐miR‐744‐5p expression in UCM patients compared to uraemic patients without myocardial hypertrophy, this difference did not reach statistical significance. Consequently, larger‐scale clinical studies are required in the future to verify the stability and reliability of PBMC‐EV‐miR‐744‐5p as a novel biomarker for UCM diagnosis.

We identified IGF2R as a pivotal molecule involved in UCM pathogenesis by comparing the candidate target genes of miR‐744‐5p in three databases with NF‐κB‐related genes, followed by dual luciferase reporter assays and rescue experiments to further validate the findings (Figure 3). Studies have focused on the role of IGF2R, as a myocardial injury‐related molecule, in myocardial injury induced by high glucose (Feng et al. 2018), high salt (Chang et al. 2019) and hypoxic stimulation (Lin et al. 2015); however, its involvement in UCM has not been explored thus far.

EVs, as nanoscale vesicles, participate in a wide range of physiological and pathological processes within the human body. EVs also play a pivotal role in UCM. Recent studies have shown that kidney‐derived EVs circulating in chronic kidney disease (CKD) patients carry distinct renal miRNAs, which contribute to the development of cardiac failure in CKD (Li et al. 2026). Targeted injection of engineered exosomes into the tibialis anterior muscle of CKD mice has been demonstrated to prevent CKD‐induced muscle wasting and mitigate cardiomyopathy (Wang et al. 2019). Although the mechanisms underlying EVs' role in UCM have been elucidated to a certain extent, significant challenges remain for their clinical translation. Technically, low loading efficiency compromises the standardization of therapeutic outcomes; heterogeneity is difficult to control; and insufficient production limits the feasibility of large‐scale manufacturing, engineered EV strategies in current clinical settings (Li et al. 2023; Xu et al. 2025). One study demonstrated that EVs derived from human liver stem cells (HLSC‐EVs) could mitigate myocardial fibrosis and improve cardiac function in CKD mouse models (Ceccotti et al. 2024). Similarly, strategies that employ physiologically derived human EVs or natural products for EV modification are poised to become critical breakthroughs for the future clinical treatment of UCM.

To expedite the translational application of PBMC‐EV modification, we selected Sal B as the therapeutic agent for experimental intervention and investigated its pharmacological effects. Sal B has demonstrated notable therapeutic effects on various cardiac diseases (Ho and Hong 2011), including diabetic cardiomyopathy (Luo et al. 2023), myocardial infarction (Li et al. 2022) and subacute myocardial infarction (Wang et al. 2011; Li et al. 2019). However, research on the use of Sal B in UCM treatment is limited, and the underlying mechanisms remain unclear. Our results indicated that Sal B exerts therapeutic effects on UCM‐associated hypertrophy, apoptosis and myocardial inflammation by modulating PBMC‐EVs via upregulation of miR‐744‐5p expression (Figures 5 and 6). Thus, Sal B may be a therapeutic candidate for targeting EV modification in UCM pathogenesis. Recent studies have revealed that various TCM monomers interact with transcription factors (Yang et al. 2024; Hou et al. 2025). Through an in silico method (molecular docking) and various in vitro techniques (SPR, CETSA and DARTS assay), we confirmed that Sal B directly binds to the transcriptional repressor YY1 in PBMCs (Figure 7D‐G). At the transcriptional level, Sal B binds to YY1 in PBMC nuclei, enhancing miR‐744‐5p expression in PBMC‐EVs and mediating the cardioprotective role of the miRNA in UCM (Figure 7). Targeting YY1 for drug screening may thus be a promising strategy for UCM treatment in the future.

In conclusion, we elucidated the novel immunopathological mechanisms underlying UCM development, establishing a novel therapeutic paradigm for UCM therapy. Our results revealed that IS stimulation reduces miR‐744‐5p expression in PBMC‐EVs, increasing IGF2R levels and promoting hypertrophy, apoptosis and NF‐κB pathway activation in cardiomyocytes. Moreover, Sal B increases PBMC‐EV‐miR‐744‐5p levels by target binding to YY1 in PBMCs, ultimately ameliorating UCM. Therefore, we propose that inhibition of aberrant EV production by using Sal B may be an effective strategy for UCM treatment.

Author Contributions

Ziran Wang: software, validation, investigation, data curation, writing – original draft, writing – review and editing, visualization. Jiaojiao Hao: conceptualization, methodology, investigation, data curation, writing – review and editing. Yaxi Shang: software, validation, formal analysis, investigation, data curation, visualization. Wenli Ma: software, validation, investigation, visualization. Nan Wang: investigation, resources, writing – review and editing. Weidong Wang: investigation, resources. Qingzhu Tang: investigation, resources. Xiangning Du: investigation, resources. Fan Yang: investigation, resources. Biaojie Qin: investigation, resources. Shuni Chen: investigation, resources. Kun Xiao: investigation, resources. Longkai Li: investigation, resources. Xianan Guo: investigation, resources. Xin Qi: investigation, resources. Shuang Meng: investigation, resources. Huiyi Song: investigation, resources. Hongli Lin: conceptualization, methodology, data curation, writing – review and editing, supervision, project administration, funding acquisition.

Funding

This work was granted by the National Natural Science Foundation of China (NSFC) [grant numbers 82270735, 82470721 to H.L.], the Liaoning Province Xing Liao Ying Cai Plan Project [grant number XLYC1908029] and the Dalian Leading Talents Funding Project [grant number 2021RD01].

Ethics Statement

Human subjects: The authorization for this clinical study was provided by the Ethics Committee of the First Affiliated Hospital of Dalian Medical University (Approval Number: PJ‐KS‐KY‐2026‐233). Animals: The research involving animals was conducted in adherence to the ethical standards and regulations established by Dalian Medical University and was subsequently endorsed by the university's Ethics Committee (protocol number: AEE24009).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information: jev270331‐sup‐0001‐FigureS1‐S7.docx

Supporting Information: jev270331‐sup‐0002‐TableS1.docx

Supporting Information: jev270331‐sup‐0005‐SuppMat.pdf

Supporting Information: jev270331‐vid‐0003‐Video1.zip

Supporting Information: jev270331‐vid‐0004‐Video2.zip

Wang, Z. , Hao J., Shang Y., et al. 2026. “Salvianolic Acid B Upregulates miR‐744‐5p Expression in Peripheral Blood Mononuclear Cell–Derived Extracellular Vesicles to Alleviate Uremic Cardiomyopathy.” Journal of Extracellular Vesicles 15, no. 7: e70331. 10.1002/jev2.70331

Data Availability Statement

All the data are available in the article and Supplementary Files. The RNA‐seq data for IS‐stimulated PBMC‐EVs are provided in the Gene Expression Omnibus under accession code GSE298327 (https://www.ncbi.nlm.nih.gov/geo/).