1. Introduction
Angiogenesis, a critical physiological process, plays a key role in embryonic development and the maintenance of tissue homeostasis. Studies have shown that promoting angiogenesis immediately after myocardial infarction (MI) enhances blood supply to ischemic myocardial areas, preserves cardiomyocytes, and mitigates adverse cardiac remodeling [
1,
2]. As such, stimulating therapeutic angiogenesis represents a promising strategy to prevent heart failure and accelerate post-MI tissue repair.
Following MI, metabolic imbalance and cell death in the ischemic zone activate inflammatory responses, resulting in immune cell infiltration and the release of pro-inflammatory cytokines and chemokines, which in turn recruit and activate myeloid cells
[3],
[4],
[5]. Monocytes are mobilized and recruited to the site of injury, where they differentiate into macrophages. These macrophages adopt two major polarization phenotypes: pro-inflammatory M1-type macrophages that release interleukin-1B (IL1B), interleukin-6 (IL6), and tumor necrosis factor-α (TNF-α); and reparative M2-type macrophages that secrete interleukin-10 (IL-10), vascular endothelial growth factor (VEGF), and insulin-like growth factor-1 (IGF-1), thereby contributing to tissue repair after MI. Previous studies have shown that macrophages can interact with endothelial tip cells, guiding vessel sprouting as cellular companions [
6,
7]. However, the pathways and mechanisms by which macrophages promote endothelial angiogenesis remain incompletely understood.
Exosomes are small membrane vesicles (30–100 nm in diameter) that contain a variety of biomolecules, including DNA, RNA, lipids, metabolites, and surface proteins. Due to their low immunogenicity, stability, and organ-targeting capabilities, exosomes have gained attention as therapeutic delivery vehicles. Macrophage-derived exosomes, especially those from M1-type macrophages, have been shown to aggravate cardiac inflammation and injury, whereas M2-type macrophage exosomes exhibit cardioprotective properties [
8,
9]. Wang et al. [
10] identified that M1 macrophage-derived exosomes are enriched in miR-155, which can be taken up by cardiac fibroblasts during MI, exacerbating inflammation and contributing to myocardial rupture. Furthermore, M1 exosomes impair angiogenesis by inhibiting the sirtuin 1 (SIRT1)/adenosine 5′-monophosphate-activated protein kinase α2 (AMPKα2)–endothelial nitric oxide synthase (eNOS) and Rac family small GTPase 1 (RAC1)/P21-activated kinase 2 (PAK2) signaling pathways [
9]. In contrast, exosomes derived from M2-type macrophages have been reported to contain miR-148a, which alleviates myocardial ischemia-reperfusion injury through modulation of the thioredoxin-interacting protein (TXNIP) and Toll-like receptor 4 (TLR4)/nuclear factor-κB (NF-κB)/NLR family, pyrin domain containing 3 (NLRP3) signaling pathways [
11]. These findings underscore the therapeutic potential of M2 macrophage-derived exosomes in MI.
Nicotinamide adenine dinucleotide (NAD
+) is an essential intracellular coenzyme involved in energy metabolism and cellular homeostasis. It acts as an electron carrier, cycling between oxidized (NAD
+) and reduced (NADH) forms, thereby enabling the redox reactions required for adenosine triphosphate (ATP) production [
12,
13]. Beyond its metabolic role, NAD
+ is also a critical signaling molecule involved in DNA repair, epigenetic regulation, and post-translational modifications [
14,
15]. Given the heart’s high energy demands, disruptions in NAD
+ metabolism have been increasingly implicated in cardiovascular pathology [
13,
16,
17]. Our previous research showed that NAD
+ promotes angiogenesis in murine MI hearts by inducing M2 macrophage polarization [
18]. Nicotinamide mononucleotide (NMN), a precursor of NAD
+, is readily taken up by cells [
19,
20]. However, whether exosomes from NMN-stimulated macrophages exert therapeutic effects on MI remains unknown. Building on our prior findings, this study investigates the potential of NMN-induced macrophage exosomes in promoting angiogenesis and treating MI.
The regulation of angiogenesis occurs at multiple levels, including pretranscriptional, transcriptional, and posttranscriptional mechanisms [
5]. Alternative splicing (AS), a key pre-transcriptional mechanism, generates multiple messenger RNA (mRNA) isoforms from a single pre-mRNA transcript, contributing to tissue-specific transcriptomic and proteomic diversity
[21],
[22],
[23]. AS is regulated by diverse splicing factors, including serine and arginine-rich splicing factor 1 (SRSF1), serine and arginine-rich splicing factor 6 (SRSF6), and U2 small nuclear RNA auxiliary factor 1 (U2AF1)
[24],
[25],
[26]. Our previous work demonstrated that NAD
+ promotes the proangiogenic VEGF165 isoform while suppressing the antiangiogenic VEGF165b isoform in macrophages by modulating the expression of splicing factors SRSF1 and SRSF6 [
27]. Yang et al. [
28] reported that circ-0001052 modulates miR-148a-3p/miR-124-3p and recruits proteins to elevate homeodomain-interacting protein kinase 3 (Hipk3) levels, ultimately worsening myocardial hypertrophy. Xu et al. [
29] found that alternative splicing factor/splicing factor 2 (ASF/SF2) directs AS patterns essential for the maintenance of cardiomyocyte electrophysiology during development, with mis-splicing of Ca
2+/calmodulin-dependent protein kinase II δ (CaMKIIδ) leading to a hypercontractile phenotype. U2AF1, a cofactor of U2 small nuclear RNA, is essential for spliceosome assembly by recognizing and binding to 3′ splice sites (AG dinucleotides) [
30,
31]. Somatic mutations in
U2af1 have been linked to various myeloid malignancies [
30,
32,
33]. Regulatory function of U2AF1 in cardiovascular pathology and assesses whether NMN-induced macrophage exosomes influence angiogenesis via AS mechanisms.
Yes1-associated transcriptional regulator (YAP1), a key co-activator in the Hippo signaling pathway, plays a central role in various physiological and pathological processes. In tumors, YAP1 cooperates with hypoxia-inducible factor 1α (HIF1α) to activate the transcription of vascular endothelial growth factor A (VEGFA), thereby promoting tumor angiogenesis [
34]. He et al. [
35] found that YAP1 enhances angiogenesis in both postnatal retinas and tumors. Similarly, Fan et al. [
36] demonstrated that endothelial-specific deletion of
Yap1 impairs angiogenesis and worsens cardiac function in MI model mice. The
Yap1 gene comprises nine exons and gives rise to at least eight alternatively spliced isoforms. Among these, the
Yap1-1 and
Yap1-2 isoforms are distinguished by the inclusion or exclusion of exon 4, which encodes a second WW domain, serving as a crucial motif for protein–protein interactions [
37]. Isoforms containing two WW domains exhibit higher transcriptional activity. The α, β, γ, and δ isoforms arise from exon 6 deletion and the use of an alternative splice site in intron 5 of
Yap1-1 and
Yap1-2, leading to functional diversity. Previous studies suggest that YAP1-2γ modulates YAP1’s oncogenic activity via interaction with protein tyrosine phosphatase non-receptor type 14 (PTPN14) [
38], while YAP1-2α acts as a more potent co-transcriptional activator of the carboxyl-terminal fragment of ErbB-4 than YAP1-1β [
39]. However, the role of
Yap1 isoforms in neovascularization following MI remains unclear. This study aims to investigate the effects of different
Yap1 isoforms on post-MI neovascularization, potentially offering novel therapeutic strategies.
Our current findings in the current study demonstrate that exosomes derived from NMN-treated macrophages confer cardioprotective effects in MI by enhancing angiogenesis. These exosomes specifically encapsulate U2AF1, which is delivered to vascular endothelial cells to promote angiogenesis by modulating the AS of Yap1, thereby facilitating cardiac repair. Furthermore, plasma U2AF1 levels in MI patients are significantly correlated with the quality of coronary collateral vessels (CCVs), suggesting U2AF1 may serve as a promising prognostic biomarker for MI.
2. Materials and methods
2.1. Study population and data collection
The prospective registration was conducted between June, 2022 and August, 2023, involving 40 patients with unstable angina (UA) pectoris and 70 with acute myocardial infarction (AMI). The patients’ ages ranged from 18 to 85 years. UA pectoris AMI is defined as continuous chest pain lasting at least 30 min and arrival at the hospital within 12 h after symptoms onset. The diagnostic criteria included 12-lead electrocardiogram (ECG) changes with or without segment T (ST) segment changes and elevated cardiac biomarker levels (creatine kinase MB or troponin T/I). To avoid the potential impact of other factors on U2AF1 plasma protein levels, we excluded patients with malignant tumors and severe liver or renal dysfunction. The research adhered to the principles of the Declaration of Helsinki and received approval from the Ethics Committee of Harbin Medical University (No. IRB5016723).
2.2. Coronary angiography
Angiography of the coronary arteries was accomplished by accessing either the femoral or radial artery. The off-line quantitative coronary angiography program (CAAS 5.10; Pie Medical Imaging B.V., the Netherlands) was used to analyze coronary angiography images before intervention, after thrombectomy, and at the end of operation. Coronary angiography (CAG) images were digitally archived in a database by two experienced researchers who were insulated from patient information. When there was disagreement between observers, consensus would be reached. The presence/absence and extent of collateral circulation were determined using the Rentrop score for 70 patients with AMI, with 0 representing a linear opaque collateral, 1 a faint distal vessel, 2 moderate filling of indirect channels, and 3 large and brightly filled collateral channels, immediately visible throughout the distal vessel > 10 mm. Subsequently, patients were divided into two groups based on their Rentrop score: poor collateral circulation (grade 0 or 1) and good collateral circulation (grade 2 or 3).
2.3. Enzyme-linked immunosorbent assay (ELISA)
ELISA kit (Cat# MM-62596H1; Meimian, China) was used following manufacturer’s instructions. U2AF1 levels in human serum samples were measured. Briefly, sera from patients categorized by Rentrop scores of 0, 1, 2, and 3 were dispensed into microplates and incubated at room temperature for 2 h. Following washing steps, the detection antibody was added to the microplates and incubated at room temperature for an additional 2 h interval. Streptavidin-horseradish peroxidase (HRP), appropriately diluted, was introduced into each well and incubated at room temperature for 20 min. Subsequently, the colored substrate was added to each well and incubated for 20 min in the dark at room temperature. After that, the stop solution was added to each well, and within 30 min, absorbance was measured using a microplate reader (ABP 00380; SpectraMax, USA). The content of U2AF1 was determined by calculating against the standard curve. U2AF1 levels were determined in mouse serum samples followed above-described procedures.
2.4. Cell cultures
RAW264.7 cells and human umbilical vein endothelial cells (HUVECs) were obtained from Shanghai Zhongqiao Xinzhou Biotechnology Co., Ltd. (China). Cell culture was carried out using Dulbecco’s modification of Eagle’s medium (DMEM; Gibco, Cat# C11965500BT; Thermo Fisher Scientific, USA) as the basal medium. This medium was supplemented with a combination of 10% fetal bovine serum (Thermo Fisher Scientific), 1% penicillin (HyClone, USA), and 1% streptomycin (HyClone). The cells were maintained in an incubator set at 37 °C with a 5% carbon dioxide atmosphere. For macrophage basal medium in which exosomes were collected, exosome free serum is required to configure the medium. The fetal bovine serum was centrifuged at 100 000g for 2 h, and the supernatant was collected to prepare exosome free serum. Subsequently, exosome free serum was used and configured according to the above formula of basal medium.
Bone marrow primary macrophages were obtained from tibia and femur of C57BL/6 mice. Bone marrow was flushed out with DMEM, centrifuged at 1000 r∙min–1 for 5 min, and red blood cell lysate (Cat# R1010; Solarbio, China) was resuspended for 2 min and centrifuged to remove red blood cells. Macrophage colony stimulating factor (MCSF; 20 ng∙mL–1; Cat# 315-02-10UG; PeproTech, USA) stimulated bone marrow cells to differentiate into macrophages.
2.5. Exosome purification, characterization, and analysis
Each batch of macrophage exosomes was purified by continuous ultra-high speed centrifugation. Briefly, RAW264.7 cells were seeded into six-well plates. The cells had reached 70%–80% confluency culture for 48 h, after which the supernatant was collected. For NMN dosing group, NMN was added at concentrations of 0.5 mmol∙L–1 for 48 h, after which the supernatant was collected. The supernatant was collected after centrifugation at 300g for 10 min and at 12 000g for 30 min. Following another run of centrifugation at 4 °C using a low-temperature ultracentrifuge at 160 000g for 2 h, macrophage exosomes were subsequently obtained. Macrophage exosomes were characterized using nanoparticle tracking analysis, examination of exosomal marker expression, and transmission electron microscopy.
2.6. In vivo fluorescence tracing of exosomes
RAW264.7-derived exosomes were isolated following a previously reported protocol. Exosomes at a concentration of 1 µg∙µL–1 were labeled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindotricarbocyanine iodide (DiR). Specifically, they were incubated with the 1 mmol∙L–1 DiR dye at a volume ratio of 500:1 for 30 min. To track exosomes in vivo, 200 µg of DiR-labeled exosomes were injected into the tail vein of mice. After 2 h, the localization of exosomes in various organs was detected using the IVIS LuminaXRMS in-vivo imaging system (PerkinElmer, USA). For ex-vivo fluorescence tracing of exosomes, after in-vivo luminescence imaging, mice were sacrificed. The liver, heart, spleen, lung, and kidney were dissected and separately imaged again using the IVIS LuminaXRMS in-vivo imaging system.
2.7. Mice
The animal-use protocol in this study was sanctioned by the Ethics Committee of Harbin Medical University and adhered to the Guidelines for the Care and Use of Laboratory Animals issued by the National Institutes of Health (NIH Publication No. 85-23, 1996 revision). C57BL/6 male mice (Liaoning Changsheng Biotechnology Co., Ltd., China) aged 6–8 weeks were used for the construction of left anterior descending (LAD) coronary artery ligation model and hindlimb ischemia model. Systemic endothelial-specific U2af1 heterozygous knockout mice (U2af1flox/+/cadherin 5 (Cdh5)-Cre heterozygous transgenic (Cre+/−), hereafter referred to as U2af1-CKO) were generated by crossing the U2af1flox/flox line with the Cdh5-Cre+/− line. Mice homozygous for a conditional allele, which becomes activated upon polyinosinic-polycytidylic acid (pIpC) induction, manifested impaired survival and functionality of hematopoietic stem cells. Embryonic lethality limited us to construct U2af1-CKO heterozygotes. Genotyping polymerase chain reaction (PCR) was performed using the following primers: F1: 5′-CACTAAAAGTATCTATGGGACGGC-3′, R1: 5′-CCACAAAGCTCTGCTCAGATACTC-3′ (homozygotes: one band with 224 bp, heterozygotes: two bands with 224 and 155 bp, wildtype allele: one band with 155 bp); Cdh5-Cre-F: 5′-CCCAGGCTGACCAAGCTGAG-3′ Cdh5-Cre-R: 5′-GTGAAACAGCATTGCTGTCACTT-3′ (Cre amplicon: 583 bp). Anesthesia was performed by intraperitoneal injection of 2% aftin (10 mL∙kg–1; Cat# T48402; Sigma-Aldrich, USA). Animals were euthanized using a two-step procedure. First, they were placed in a carbon dioxide chamber, after which cervical dislocation was carried out.
2.8. Construction of LAD coronary artery ligation model in mice
Anesthesia followed by endotracheal intubation connecting to a small animal respirator. Thoracotomy was performed in the second to third intercostal spaces on the left side, the pericardium was opened, and the heart and LAD were exposed. The LAD was ligated using 7-0 Prolene suture passing approximately 1–2 mm below the tip of the left auricle. Successful establishment of MI model was confirmed by color change of ischemia area and ST segment elevation of ECG. Subsequently, the chest cavity was closed, and the skin wound was sutured. Exosome treatment groups was given control exosomes from untreated macrophages (Nor-Exo; 10 μg∙g–1∙d–1) or NMN-induced macrophage exosomes (NMN-Exo; 10 μg∙g–1∙d–1) via tail vein injection for seven consecutive days.
2.9. Echocardiographic assessment of cardiac function
Mice of the MI group and the sham group were anesthetized on the 7th or 28th day post-MI. They were then placed in supine position on an ultrasound table. Cardiac imaging in M-mode echocardiography was performed using a small-animal Doppler ultrasound Vevo2100 imaging system (Visualistics, Canada). Fractional shortening (FS) and ejection fraction (EF) were calculated as the primary indices for assessing cardiac function.
2.10. Hindlimb ischemia model in mice by ligation of femoral artery
Male C57BL/6 mice aged 6–8 weeks were injected with U2af1-overexpressing virus (Tie promoter) three weeks before surgery, and 2% aftin (10 mL∙kg–1; Cat# T48402; Sigma-Aldrich) was used for anesthesia at surgery. The left femoral artery and vein were ligated. Blood flow measurements were performed on days 0, 3, and 7 post-surgery to assess the effects of U2AF1 on arteriogenesis, angiogenesis, and prognosis.
2.11. Laser Doppler imaging
Hindlimb blood flow was evaluated serially before surgery, immediately after surgery, and on days 3 and 7 following femoral artery ligation. A real-time microcirculation imager (PeriCam PSI HR system; Perimed, Sweden) based on laser speckle contrast analysis technology (LASCA) was utilized for this assessment. Blood perfusion was quantified in arbitrary perfusion unit (PU). At each time point, three images were acquired, and their averaged values were calculated. The results were presented as the ratio of blood perfusion in the ischemic hindlimb to that in the non-ischemic hindlimb.
2.12. 2,3,5-Triphenyltetrazolium chloride (TTC) staining
To assess infarct size, mice after MI model establishment were euthanized, and the hearts were isolated, washed three times with saline, placed on glass slide, and immediately frozen at –80 °C for 5 min, following by slicing into cross-sections (1 mm thick) and incubation in TTC (2%; Cat# G3005; Solarbio) solution at 37 °C for 20 min. Subsequently, the infarcted areas of the heart turned white, while the non-infarcted areas remained red. Images were post-processed using ImageJ software for photo analysis. MI area ratio = white area/total myocardial area × 100%.
2.13. Masson staining
In accordance with the experimental design, following the establishment of the MI model and treatment of the small animals, the mice were sacrificed. Their hearts were then carefully isolated in an intact state and fixed in paraformaldehyde for 48 h. Subsequently, the heart tissue underwent dehydration and was embedded in paraffin. Paraffin-embedded blocks were sectioned into 5 μm-thick slices for further analysis. Masson’s trichrome stain kit (Cat# G1340; Solarbio) was used to stain paraffin sections according to the manufacturer’s instructions. Then dehydrated and sealed, the slices were taken out of xylene and dried slightly and sealed with neutral resin. Panoramic scanning was performed using the Aperio VERSA scanning system and image analysis was performed using ImageScope software.
2.14. Microfil perfusion and three-dimensional (3D) micro-computed tomography (CT)
As per the experimental protocol, the mice were sacrificed. Their hearts were then isolated intact and placed in large dishes containing physiological saline. The aorta of each heart was carefully dissected free, after which the aorta was cannulated and secured in place using a ligature. The other end of the aortic cannula was connected with a syringe filled with 4% paraformaldehyde. The heart was slowly perfused, excess blood in the blood vessels was drained. Subsequently, attach a syringe containing angiography agent (Microfil MV 122; Microfil extended release solution = 1:1.25, 5% Microfil sclerosing agent) to fill the cardiac vessels with contrast medium and ligate the aortic vessel at the cannula interface with ligature. The distribution of blood vessels around the left anterior descending branch of the heart was recorded by stereomicroscopy. Micro-CT (Quantum GX II; Perkinelmer) imaging system for small animals visualizes cardiac vessels.
2.15. Immunofluorescence
After frozen sections of heart tissue, fixation with 4% paraformaldehyde for 1 h, 0.1% Triton X-100 for 15 min at room temperature, blocking with 10% goat serum + 1% bovine serum albumin (BSA) for 1 h at 37 °C, followed by platelet and endothelial cell adhesion molecule 1 (CD31) antibody (1:500; Cat# 14-0311-82; Invitrogen, USA), α-smooth muscle actin (α-SMA) antibody (1:500; Cat# AF1032; Affinity Biosciences, USA), incubation overnight at 4 °C in the dark. Subsequently, after washing three times with phosphate-buffered saline (PBS), unbound primary antibody was removed, fluorescent secondary antibody was added, incubated at 37 °C for 1 h, and slides were mounted after 4′,6-diamidino-2-phenylindole (DAPI) staining. Confocal laser microscopy (Cat# 0972818; Olympus, Japan) was used to obtain images of the infarct margin.
2.16. Cell counting kit-8 (CCK8) analysis
HUVECs were seeded into 96-well plates and the cell culture medium was discarded after treatment according to the experimental design. Add 100 μL culture solution to each well and add 10 μL CCK8 reagent to each well. After incubation in an incubator at 37 °C for 2–4 h, the absorbance value at 450 nm was detected with a microplate reader.
2.17. Tube formation assay
In vitro angiogenic capacity of HUVECs was measured by Matrigel. Briefly, microscope dishes were coated with Matrigel (150 μL∙well–1; Cat# 356234; Corning, USA) and U2af1 small interfering RNA (siRNA) or negative control (NC) pretreated HUVECs were seeded at a cell density of 5 × 104 cells∙well–1 after Matrigel had solidified and incubated with 5% CO2 for 4–6 h at 37 °C for tube formation. Images of the tubes in each well were taken using an inverted microscope. The length of the tube was calculated by Image J, and the branching nodes of the tubule were analyzed by Image J software.
2.18. Scratch assay
Confluent HUVECs to 100%, draw a straight line along the central axis of the cell culture plate with a sterile white pipette head, add serum-free ECM, and immediately take scratches with an inverted microscope, that is, record the scratch area at 0 h. The cells were then transferred to a cell incubator at 37 °C for further culture for 12 h, and the scratches were photographed under an inverted microscope, that is, the scratch area for 12 h was recorded. Image J software was used to evaluate wound healing rate quantitatively.
2.19. Particle size analysis
Macrophage-derived exosome samples were diluted in double-distilled water. The particle size and polydispersity index (PDI) were measured at 25 °C using a Zetasizer Nano ZS90 (Malvern Instruments, UK).
2.20. Flow cytometry
HUVECs were digested with trypsin to obtain a single-cell suspension, washed with PBS, and fixed overnight in 70% ethanol at 4 °C. Cells were then resuspended in buffer and stained with 25 μL propidium iodide and 10 μL ribonuclease (RNase) A. After a 30 min incubation at 37 °C in the dark, samples were analyzed by flow cytometry (Beckman Coulter, USA) using a 488 nm excitation wavelength. ModFit software was used to analyze DNA ploidy and cell cycle distribution.
2.21. RNA binding protein immunoprecipitation (RIP)
RIP was performed using the Magna RIP RNA binding protein immunoprecipitation kit (Cat# RIP-12RXN; Sigma-Aldrich) following the manufacturer’s protocol. Briefly, HUVECs were lysed in buffer containing protease and RNase inhibitors. Cell lysates were incubated overnight at 4 °C with magnetic beads conjugated to anti-U2AF1 antibody (Cat# ab172614; Abcam, UK). RNA was extracted, reverse-transcribed into complementary DNA (cDNA), and analyzed by quantitative real-time PCR (qRT-PCR). Enrichment was calculated as a percentage relative to input.
2.22. Agarose gel electrophoresis
First, weigh 1 g of agarose powder and pour it into a 250 mL conical flask, then add 1× tris-acetate-EDTA buffer (TAE) 100 mL, heat in a microwave until completely dissolved, cool at room temperature to 60 °C, add 10 μL loading buffer with ethidium bromide (LEB), pour it into the electrophoresis tank gel plate, and prepare 1% agarose gel after cooling. Then put agarose gel in the electrophoresis tank and pour 1× DNA markers and DNA samples for electrophoresis after TAE. After electrophoresis, observe the bands in the gel under the ultraviolet light, and record and take photos.
2.23. Data analysis
Statistical analysis was performed using SPSS v26.0 and GraphPad Prism V9.0.1 (GraphPad Software, USA). Categorical variables were expressed as counts and percentages. Normality of continuous variables was assessed with the Kolmogorov–Smirnov test. Data were presented as mean ± standard deviation (SD) for normally distributed variables or median (interquartile range) for non-normally distributed variables. Appropriate statistical tests were used for group comparisons, including Student’s t-test, Mann–Whitney U test, one-way analysis of variance (ANOVA) with Tukey’s test, two-way ANOVA with Tukey’s test, Log-rank (Mantel-Cox) test, Kruskal–Wallis test, χ2 test, or Fisher’s exact test. Binary logistic regression was conducted to assess the relationship between serum U2AF1 levels and the presence of good collateral circulation, adjusting for clinically relevant covariates. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Statistical significance was defined as P < 0.05 (two-tailed).
3. Results
3.1. NMN-Exo treatment enhances cardiac function in MI mice by promoting angiogenesis
Our previous study demonstrated that NAD
+ promotes macrophage polarization toward the M2 phenotype [
18], and exosomes derived from M2 macrophages exert protective effects in the context of MI [
11]. Given the large molecular weight of NAD
+, NMN was used as a precursor supplement. As shown in Fig. S1(a) in Appendix A, NMN treatment increased NAD
+ levels in macrophages in a dose-dependent manner. To investigate the effects of NMN on macrophage polarization, we employed Luminex liquid suspension chip technology to assess the expression levels of secreted factors from NMN-treated macrophages (Fig. S1(b) in Appendix A). Compared to the control group, NMN (1mmol∙L
–1)-treated macrophages exhibited decreased levels of pro-inflammatory cytokines (CXCL1, IL1B, and TNF) and elevated levels of anti-inflammatory factors (VEGF, IL10, and CXCL12) (Figs. S1(c) and (d) in Appendix A). Moreover, NMN promoted M2-type macrophage polarization by downregulating M1-associated markers (
Nos2,
Tnf, and
Ccl3) and upregulating M2-associated markers (
Il10,
Vegf) (Figs. S1(e)–(i) in Appendix A). In lipopolysaccharide (LPS)-induced macrophages, NMN also reduced the expression of pro-inflammatory cytokines (
Il1b,
Il6,
Tnf, and
Ccl3) while enhancing the expression of the anti-inflammatory cytokine (
Il10) (Figs. S1(j)–(n) in Appendix A). These findings suggest that NMN facilitates macrophage polarization toward the M2 phenotype.
To further explore whether macrophage-derived exosomes stimulated by NMN are functional, we first isolated them via differential centrifugation and conducted morphological analyses. For clarity, we refer to these exosomes as NMN-Exo. Nanoparticle tracking analysis revealed a uniform size distribution centered around 100 nm (Fig. S2(a) in Appendix A), and transmission electron microscopy showed that the exosomes possessed a transparent bilayer membrane structure with diameters ranging from 30 to 150 nm (Fig. S2(b) in Appendix A). Compared to cell lysates, NMN-Exo exhibited markedly high expression of the exosomal markers CD63 and CD81 (Fig. S2(c) in Appendix A). Immunofluorescence analysis further demonstrated increased levels of the exosome-specific marker CD63 in NMN-treated macrophages (Fig. S2(d) in Appendix A). These results indicate that NMN enhances both the production and cargo loading efficiency of macrophage-derived exosomes.
To investigate the impact of NMN-Exo on cardiac function, we intravenously administered NMN-Exo or Nor-Exo into MI model mice (
Fig. 1(a)). Fluorescence imaging of DiR-labeled exosomes revealed predominant accumulation in the liver and spleen, with a smaller portion reaching the heart (
Fig. 1(b)). Compared to Nor-Exo, NMN-Exo significantly improved cardiac function, as evidenced by increased EF and FS, reduced infarct size, and decreased collagen deposition 7 days post MI (
Figs. 1(c)–(e)). These improvements in cardiac function persisted at 28 days post MI (Fig. S3 in Appendix A). We further compared the therapeutic effects of direct NMN administration (at 50 and 500 mg∙kg
–1) versus NMN-Exo treatment. All treatment groups showed enhanced cardiac function and reduced infarct areas, but NMN-Exo exhibited the most pronounced therapeutic effect (Fig. S4 in Appendix A). Additionally, NMN-Exo increased neovascularization in the infarct border zone, as observed through Microfil perfusion (
Fig. 1(f)), and upregulated CD31 expression in heart tissues (
Fig. 1(g)). These effects were not observed with Nor-Exo, suggesting a specific proangiogenic effect of NMN-Exo in MI mice. To examine the
in vivo behavior of macrophage-derived exosomes, we quantified their retention within coronary vasculature. Dil-labeled macrophage exosomes were injected via the tail vein, and co-localization with CD31-positive vascular endothelial cells was observed (
Fig. 1(h)), indicating that these exosomes can be internalized by endothelial cells and contribute to post-MI angiogenesis.
To elucidate the underlying mechanisms through which NMN-Exo treatment promotes angiogenesis in MI mice, we explored the uptake of NMN-Exo by endothelial cells (Fig. S5(a) in Appendix A). PKH26-labeled NMN-Exo were internalized by HUVECs and localized near the nucleus (Fig. S5(b) in Appendix A). NMN-Exo significantly promoted HUVEC proliferation at concentrations of 25 and 125 μg∙mL
–1, whereas Nor-Exo had no noticeable effect (Fig. S5(c) in Appendix A). NMN-Exo also facilitated the transition of HUVECs into the S phase, further supporting enhanced proliferation (Fig. S5(d) in Appendix A). Scratch assays and tube formation assays confirmed that NMN-Exo, but not Nor-Exo, significantly promoted endothelial cell migration and tubulogenesis (
Figs. 1(i) and
(j)). Collectively, these findings demonstrate that NMN-Exo improves cardiac function in MI mice by promoting angiogenesis.
3.2. Characteristic protein expression profile in NMN-Exo
Quantitative proteomic analysis revealed differentially expressed proteins between NMN-Exo and Nor-Exo (
Fig. 2(a)). Gene Ontology (GO) analysis identified 16 upregulated proteins involved in processes such as vesicle synthesis, vesicle transport, and AS (e.g., SEC22B, U2AF1, and SLC7A5), and eight downregulated proteins (e.g., AOC3, VNN1, ITIH3, KNG1, TALOD1, MFAP4, TUBGCP3, and SERPING1) (
Figs. 2(b)–(d)). qRT-PCR analysis confirmed that most of the 16 upregulated genes were also elevated in NMN-treated macrophages, with
U2af1 showing the most prominent upregulation (Fig. S6 in Appendix A). Among the upregulated proteins, SEC22B and SLC7A5 are known to participate in vesicle transport and amino acid transport, both of which have been implicated in angiogenesis. Therefore, we further evaluated the levels of these proteins, with Western blotting ultimately confirming the presence of elevated levels of SEC22B and U2AF1 in NMN-Exo samples, consistent with the quantitative proteomic results (
Fig. 2(e)). Similarly elevated levels of these proteins were noted in NMN-treated macrophages (
Fig. 2(f)). SEC22B is recognized for its role in vesicle trafficking, while U2AF1 contributes to mRNA splicing. These findings suggest that SEC22B and U2AF1 may be mechanistically involved in NMN-Exo-mediated angiogenesis.
3.3. U2AF1 mediates cardiac repair in MI by promoting angiogenesis
To identify the primary contributor to heart disease when comparing U2AF1 and SEC22B, we reanalyzed serum RNA-seq data (GSE208194) from healthy volunteers (HVOLs) and heart failure (HF) patients. Serum U2AF1 levels were significantly lower in HF patients than in the HVOL group (
Fig. 3(a)), whereas no significant changes were observed in SEC22B. The protein sequence of U2AF1 showed high conservation between mice and humans (Fig. S7 in Appendix A). Further validation using a mouse MI model revealed that U2AF1 expression was downregulated in both serum and cardiac tissue at 3, 7, and 28 days post MI (
Figs. 3(b) and
(c)), suggesting a potential role for U2AF1 in MI pathology. To explore U2AF1’s functional role, we employed an adeno-associated virus serotype 9 (AAV9)-based endothelial-specific overexpression system (U2AF1-v) along with a control virus (Vector-v) in MI mice (Fig. S8 in Appendix A). U2AF1 overexpression significantly improved cardiac function and upregulated CD31 expression, a marker of angiogenesis (
Figs. 3(d) and
(e)), indicating that U2AF1 enhances angiogenesis. To further validate this, we established a posterior limb ischemia model in U2AF1-v and Vector-v mice. Laser Doppler imaging revealed that U2AF1 overexpression accelerated blood flow recovery at 3 and 7 days post ischemia compared to the control group. Additionally, U2AF1-overexpressing mice showed increased capillary density (CD31) and arteriole numbers (α-SMA) in the ischemic hind limbs (
Figs. 3(f) and
(g)).
In vitro, HUVECs exposed to hypoxia and H
2O
2—conditions mimicking the ischemic and oxidative stress environment of MI—exhibited decreased U2AF1 expression (
Figs. 4(a) and
(b)). Transfection with a U2AF1 expression vector significantly elevated
U2af1 mRNA and protein levels in HUVECs (
Figs. 4(c) and
(d)). Moreover, U2AF1 overexpression increased the proportion of hypoxia- or H
2O
2-treated HUVECs in the S and G2/M phases (
Figs. 4(e) and
(f)) and enhanced the migration of H
2O
2-treated HUVECs (
Fig. 4(g)). We also investigated the effect of U2AF1 on cardiomyocytes. Western blot analysis revealed that U2AF1 protein levels were significantly reduced in primary neonatal mouse cardiomyocytes after 12 and 24 h of hypoxia-induced injury (Fig. S9(a) in Appendix A). A U2AF1-overexpressing plasmid was successfully constructed and verified for transfection efficiency (Fig. S9(b) in Appendix A). Flow cytometry demonstrated that U2AF1 overexpression markedly reduced hypoxia-induced apoptosis in cardiomyocytes (Figs. S9(c) and (d) in Appendix A). These findings indicate that U2AF1 promotes angiogenesis by enhancing the proliferation and migration of HUVECs under ischemic and oxidative stress conditions and also exhibits a protective effect against cardiomyocyte apoptosis.
3.4. Knockout of U2af1 inhibits cardiac angiogenesis and impairs cardiac injury repair in MI model mice
Endothelial-specific
U2af1 knockdown mice (
U2af1-CKO) exhibited efficient knockdown and reduced U2af1 expression in aortic tissue (
Figs. 5(a) and
(b), Fig. S10 in Appendix A). At eight weeks of age,
U2af1-CKO mice displayed significantly reduced cardiac function compared to
U2af1f/+ mice (
Fig. 5(c)).
U2af1-CKO + MI mice (
U2af1-CKO+MI) showed reduced survival rates compared to the
U2af1f/++MI group (
Fig. 5(d)). Notably, with prolonged myocardial ischemia (MI for 7, 14, and 28 days), cardiac function in
U2af1-CKO mice declined more severely than that in
U2af1f/+ mice (
Figs. 5(e) and
(f)). Moreover, the infarcted area in
U2af1-CKO mice was larger than that in
U2af1f/+ mice at 28 days post MI (
Fig. 5(g)). Collagen deposition was more severe in
U2af1-CKO mice than in
U2af1f/+ mice at 7 days post MI (
Fig. 5(h)). 3D micro-CT and CD31 immunofluorescence confirmed a significant reduction in cardiac vascular density and CD31-labeled endothelial cells in
U2af1-CKO+MI group (
Figs. 5(i) and
(j)).
In vitro, small interfering RNA (siRNA)-mediated knockdown of
U2af1 (si
U2af1) was validated by qRT-PCR (
Fig. 5(k)). As shown in
Figs. 5(l) and
(m), knocking down
U2af1 significantly reduced the migration and tube formation abilities of HUVECs compared to siRNA negative control (siNC)-transfected cells. These results underscore the critical role of U2AF1 in angiogenesis and post-MI repair when specifically knocked down in endothelial cells.
3.5. U2AF1 regulates the variable splicing of Yap1 to promote angiogenesis
As a splicing cofactor, U2AF1 primarily regulates AS events. To uncover the molecular mechanisms underlying U2AF1’s role in angiogenesis, whole-genome RNA sequencing was performed on U2AF1-overexpressing HUVECs and vector-transfected controls. Five major AS types were analyzed, including skipped exon (SE), alternative 5′ splice site (A5SS), alternative 3′ splice site (A3SS), mutually exclusive exon (MXE), and retained intron (RI) events, revealing 2696 significant AS events in U2AF1-overexpressing cells compared to controls. Among these, 1294 AS events were upregulated and 1402 were downregulated, with SE events accounting for the largest proportion (53.9%) (
Fig. 6(a)). Filtering for |△percent spliced in (PSI)| > 0.1 and
p < 0.05, we identified U2AF1-associated SE events. GO analysis showed that these events were enriched in processes such as cell proliferation, migration, wound healing, cell cycle regulation, autophagy, and mRNA splicing (
Fig. 6(b)). Immunofluorescence confirmed that U2AF1 was primarily localized in the nucleus (
Fig. 6(c)), consistent with its role in pre-mRNA splicing. Given U2AF1’s involvement in splicing and angiogenesis, we conducted gene regulatory network analysis focusing on YAP1 (
Fig. 6(d)).
Yap1, composed of nine exons and existing as two primary splice variants (
Yap1-1 and
Yap1-2), was identified as a central target. These isoforms further divide into α, β, γ, and δ subtypes (
Fig. 6(e)). Sequencing data revealed that U2AF1 overexpression induced exon skipping in
Yap1 at chromosome (ch) 11:102056749–102056862, specifically affecting exon 4 (
Fig. 6(f)).
To confirm the direct regulatory effect of U2AF1 on
Yap1 variable splicing, we designed primers at the junction of intron 3 and exon 4 for RNA immunoprecipitation experiments, confirming that U2AF1 binds to the exon 4 splicing site of
Yap1 (
Fig. 6(g)). Analyses of the exon 4 SE event in
Yap1 using primers upstream of exon 3 and downstream of exon 5 further revealed that U2AF1 overexpression induced relative
Yap1-2 upregulation and
Yap1-1 downregulation compared to the NC group (
Figs. 6(h) and
(i)).
3.6. U2AF1 promotes the formation of Yap1-2γ spliceosome and enhances angiogenesis
Yap1-1 and
Yap1-2 give rise to four splicing isoforms: α, β, γ, and δ. Analysis of transcript expression in U2AF1-overexpressing cells revealed upregulation of the
Yap1-2β and
Yap1-2γ isoforms and downregulation of
Yap1-1δ and
Yap1-2δ (
Fig. 6(j)). Using upstream primers at exon 3 and downstream primers at exon 7, eight splice variants associated with
Yap1-1 and
Yap1-2 were amplified. Agarose gel electrophoresis showed distinct DNA bands, with
Yap1-2γ,
Yap1-2α, and
Yap1-1δ being the most abundant (
Fig. 6(k)). A Venn diagram comparing high-abundance
Yap1 splice variants and significantly altered transcripts from whole-genome sequencing suggested that U2AF1 regulates
Yap1 function by inhibiting exon 4 skipping, thereby promoting the expression of the
Yap1-2γ isoform while suppressing the
Yap1-1δ isoform (
Fig. 6(l)). To assess the impact of alternative
Yap1 splicing on angiogenesis, overexpression plasmids for
Yap1-2γ and
Yap1-1δ were constructed. Agarose gel electrophoresis confirmed significantly increased expression of both splice variants in transfected cells (
Fig. 6(m)). Cell scratch assays demonstrated an enhanced wound closure rate following
Yap1-2γ overexpression, whereas
Yap1-1δ showed no comparable effect (
Fig. 6(n)). In tubule formation assays,
Yap1-2γ overexpression significantly increased both relative tube length and the number of tubule nodes. In contrast, although
Yap1-1δ overexpression also extended tube length, it had no significant effect on tubule node formation (
Fig. 6(o)). To determine whether U2AF1 regulates vascular endothelial cell migration via modulation of
Yap1 splicing, we used the AS inhibitor Madrasin. Scratch assays showed that U2AF1 overexpression enhanced HUVEC migration compared to the control group; however, this effect was abolished in cells pretreated with Madrasin (Fig. S11 in Appendix A). These findings indicate that U2AF1 regulates vascular endothelial cell migration by modulating the AS of
Yap1.
3.7. U2AF1 as a circulating biomarker for MI prognosis
These findings highlight the role of U2AF1 in MI repair through the promotion of angiogenesis. Previous studies have shown that timely development of collateral circulation is a critical step in the repair of myocardial injury. Based on this, we investigated the association between circulating U2AF1 levels and MI. In a study involving 110 patients (40 with UA and 70 with AMI), U2AF1 protein levels were significantly lower in AMI patients than in controls. U2AF1 levels showed a positive correlation with collateral circulation, an essential component in the repair of myocardial ischemic injury (
Fig. 7). Baseline characteristics differed between the two groups (
Table 1). Specifically, the AMI group had a higher proportion of male patients (75.7% vs 37.5%,
p < 0.001), as well as significantly elevated levels of creatine kinase-MB (CK-MB) (110.65 (47.85, 225.05) vs 0.55 (0.40, 0.80) μ∙L
–1,
P < 0.001), NT-pro B-type natriuretic peptide (BNP) (1404.50 (47.85, 225.05) vs 47.61 (0.40, 0.80) pg∙mL
–1,
P < 0.001), and left ventricular ejection fraction (LVEF) (53.0% (42.0%, 59.0%) vs 62.5% (60.0%, 69.0%),
p < 0.001), along with other indicators of MI and heart failure. Notably, U2AF1 protein levels were significantly reduced in AMI patients compared to controls ((2.51 ± 1.87) vs (4.35 ± 1.48) ng∙mL
–1,
p < 0.001) (
Table 1 and
Fig. 7(a)). No significant differences were found between UA and AMI in other measured parameters. Collateral circulation in AMI patients was assessed by Rentrop scoring of contrast images. Rentrop scores of 0, 1, 2, and 3 were recorded for 7, 19, 30, and 14 patients, respectively. A total of 26 patients (37%) had poor collateral circulation (Rentrop scores of 0 or 1), whereas 44 patients (63%) had well-developed collateral circulation (scores of 2 or 3) (
Fig. 7(b)). Further analysis confirmed a positive correlation between U2AF1 protein levels and collateral circulation (
Figs. 7(c) and
(d)). These results support the potential of U2AF1 as a circulating biomarker for MI prognosis.
4. Discussion
Disturbance in coronary collateral microcirculation increases the risk of MI and contributes to adverse cardiac remodeling. However, the specific targets and mechanisms for restoring coronary microcirculation during MI remain unclear. The results of this study identified NMN as a promoter of cardiac angiogenesis that regulates macrophage-endothelial cell interactions during MI progression. NMN acts on macrophages to stimulate the release of exosomes with distinct protein profiles that enhance angiogenesis and improve cardiac function. Importantly, U2AF1 was identified as a critical regulator of post-MI angiogenesis via its role in alternative Yap1 splicing. Manipulating U2AF1 expression specifically in vascular endothelial cells can either promote or inhibit angiogenesis. Clinical samples from patients with UA and AMI exhibited significantly reduced serum U2AF1 levels in the AMI group, with a clear correlation to the extent of collateral circulation.
Conventional MI treatments such as revascularization via bypass surgery, stenting, or thrombolytic therapy primarily address large-vessel blockages. While these methods rapidly restore primary blood flow, they often fail to fully resolve ischemia and hypoxia at the cardiomyocyte level. Therefore, the timely formation of coronary collateral circulation has emerged as a promising therapeutic strategy
[40],
[41]. Research indicates that increasing NAD
+ levels protects against heart failure, MI, dilated cardiomyopathy, and ischemia-reperfusion injury by preventing cardiomyocyte apoptosis and improving metabolic, mitochondrial, and lysosomal functions
[20],
[42],
[43],
[44],
[45],
[46],
[47],
[48]. However, most prior studies utilized systemic administration, and the specific mechanisms by which NAD
+ improves collateral circulation post MI remain unclear. Our previous work showed that NAD
+ enhances angiogenesis in infarcted mouse hearts by promoting M2 macrophage polarization [
18]. NMN, a NAD
+ precursor with better cellular uptake [
19,
20], was chosen to investigate the molecular mechanisms underlying collateral vessel formation following MI. This study also aimed to identify both circulating biomarkers and therapeutic targets related to coronary collateral circulation.
Angiogenesis is a complex process involving multiple cell types, with macrophages playing a central role [
49]. The results of this study confirm that NMN inhibits polarization toward the pro-inflammatory M1 phenotype and promotes M2-type macrophages. Due to their endogenous origin and heterogeneous nature, exosomes possess distinct advantages in diagnostics and therapeutics
[50],
[51],
[52]. NMN-derived exosomes significantly improved cardiac function, reduced infarct size, and inhibited excessive collagen deposition in MI mouse models. NMN-Exo-treated HUVECs showed enhanced proliferation, migration, and tube formation. Proteomic analysis of NMN-Exo revealed the upregulation of 16 proteins related to vesicle transport and AS. U2AF1 was further identified as a critical factor in myocardial repair, based on blood sample analysis from HF patients. U2AF1, a splicing cofactor involved in the development of various myeloid malignancies, was found to exhibit reduced expression in MI mouse heart tissue. Overexpression of U2AF1 improved cardiac function and promoted angiogenesis
in vivo. Whole-genome sequencing further confirmed its role in regulating AS events, especially exon skipping events. The transcriptional coactivator YAP1, a key regulator of organ size and vascular development, was subsequently identified as a downstream target of U2AF1.
Yap1 has multiple isoforms, and U2AF1 overexpression led to the enhancement of
Yap1-2γ expression while suppressing
Yap1-1δ. These splicing changes modulated endothelial cell migration and tubule formation.
As a non-invasive diagnostic tool, serum markers have demonstrated significant potential in assessing cardiac collateral circulation and predicting disease progression. However, studies have shown that certain serum markers, such as VEGF and fibroblast growth factor (FGF), promote pathological angiogenesis and increase the risk of vascular leakage
[53],
[54],
[55]. Therefore, the identification of novel physiological angiogenic factors is necessary. In this study, by analyzing changes in serum U2AF1 levels in 110 clinical patients (40 with UA and 70 with AMI), we found that U2AF1 protein levels were significantly reduced in AMI patients compared to the control group. Furthermore, we evaluated the collateral circulation in AMI patients using the Rentrop score and found a strong correlation between U2AF1 levels and collateral circulation, suggesting that U2AF1 may serve as a potential peripheral blood marker for MI prognosis. In line with our findings, Bao et al. investigated the effects of angiopoietin-2 on angiogenesis in myocardial-ischemic diabetic mice and assessed the relationship between serum vasostatin-2 levels and CCVs in diabetic patients with chronic total occlusions (CTOs) [
56]. However, the application of circulating U2AF1 levels as a peripheral blood marker for MI prognosis remains limited. Serum U2AF1 levels may fluctuate with disease progression and therapeutic interventions, making it difficult to capture its dynamic changes accurately. To obtain more reliable results, these factors must be carefully considered, and appropriate measures should be taken to control or eliminate potential confounders. Additionally, future studies should investigate the association between serum protein levels and disease using larger sample sizes, extended follow-up periods, broader proteomic profiling, and more rigorous study designs. While this study provides important insights into the role of NMN and U2AF1 in promoting angiogenesis during MI, it also acknowledges several limitations. Targeted therapeutic drug screening for U2AF1 remains to be conducted, and clinical validation of these findings is necessary before they can be translated into therapeutic interventions. Overall, this study highlights the importance of therapeutic strategies targeting microcirculatory dysfunction during MI and underscores the pivotal role of U2AF1 in regulating angiogenesis, offering a promising clinical marker for MI related to microcirculatory disorders.
CRediT authorship contribution statement
Manyu Gong: Writing – original draft, Formal analysis, Data curation, Conceptualization. Haodong Li: Formal analysis, Data curation. Lei Jiao: Formal analysis, Data curation, Conceptualization. Tong Liu: Formal analysis, Data curation. Yanwei Zhang: Formal analysis. Jie Liu: Data curation. Siyu Wang: Data curation. Hao Wang: Data curation. Dongping Liu: Formal analysis. Zhaoyue Li: Formal analysis. Zhiyuan Du: Data curation. Lihua Sun: Methodology. Lina Xuan: Conceptualization. Shihua Lv: Data curation. Xuewen Yang: Data curation. Yanying Wang: Data curation. Yingfeng Tu: Data curation. Mengmeng Li: Formal analysis. Haodi Wu: Data curation. Xin Li: Formal analysis. Xue Feng: Data curation. Juan Xu: Formal analysis, Conceptualization. Wenzhi Li: Methodology. Yong Zhang: Writing – review & editing, Data curation, Conceptualization. Ying Zhang: Writing – review & editing, Data curation, Conceptualization. Baofeng Yang: Writing – review & editing, Funding acquisition, Formal analysis, Data curation, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study is supported by the National Natural Science Foundation of China (82370417, 82330011, and U21A20339), the Science Fund for Distinguished Young Scholars of Heilongjiang Province (JQ2024H001), and the Heilongjiang Provincial Postdoctoral Science Foundation (LBH-Z23212).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2025.06.006.
PDF
(7536KB)
