The PIWI-interacting RNA CRAPIR alleviates myocardial ischemia‒reperfusion injury by reducing p53-mediated apoptosis via binding to SRSF1

Abstract

Ischemia-reperfusion (I/R) injury refers to the secondary damage that occurs when blood flow is restored to heart tissues and organs following a period of prolonged ischemia. This damage is exacerbated primarily through mechanisms such as oxidative stress, inflammatory responses and apoptosis, all of which can severely impact patient prognosis. PIWI-interacting RNAs (piRNAs) represent a novel class of small noncoding RNAs that play pivotal roles in regulating gene expression and cellular functions. However, the precise role and underlying mechanisms of piRNAs in I/R injury remain poorly understood. In this study, we investigated the role and molecular mechanisms of a cardiac regeneration-associated PIWI-interacting RNA (CRAPIR), previously identified by our team, in I/R injury. An I/R injury model was established in adult male mice. The protein levels of cleaved caspase-3, Bax, Bcl2 and p53 were assessed using Western blotting, and cardiomyocyte apoptosis was detected via TUNEL staining. Our study revealed that, in I/R-damaged heart tissues and hypoxia‒reoxygenation (H/R)-induced cardiomyocyte models, CRAPIR was upregulated 24 h after I/R and H/R but was markedly downregulated at 72 h after I/R injury and 48 h after H/R injury. In the I/R mouse model, agomir-mediated overexpression of CRAPIR alleviated heart dysfunction and reduced cardiomyocyte apoptosis caused by I/R injury. Conversely, CRAPIR knockdown via an antagomir exacerbated I/R-induced cardiac dysfunction and increased the number of apoptotic cardiomyocytes. Mechanistically, CRAPIR interacts with serine/arginine-rich splicing factor 1 (SRSF1), triggering the upregulation of murine double minute 2 (MDM2) expression. The increased MDM2 promoted p53 ubiquitination, leading to reduced p53 levels. Furthermore, silencing SRSF1 or MDM2 attenuated the protective effect of CRAPIR against cardiomyocyte apoptosis following H/R injury. These findings suggest that CRAPIR serves as a critical regulator of I/R injury via the SRSF1/MDM2/p53 signaling pathway.

Introduction

Cardiovascular diseases (CVDs) are a major threat to human health, with a mortality rate that exceeds that of other causes, and myocardial infarction (MI) is the leading cause of death among CVD patients [1, 2]. Currently employed clinical treatments, such as percutaneous coronary intervention (PCI), are associated with the potential to induce myocardial ischemia‒reperfusion (I/R) injury, which strongly affects clinical outcomes and prognosis [3]. Apoptosis plays a crucial role in various physiological processes and pathological conditions, including myocardial I/R injury [4–6]. Currently, no effective method exists for preventing or treating myocardial I/R injury in clinical practice. Therefore, further investigations of the molecular mechanisms underlying myocardial I/R injury, the development of effective preventive strategies for reperfusion-induced cardiac injury, and the identification of novel cardioprotective approaches are urgently needed.

PIWI-interacting RNAs (piRNAs) are small RNAs ranging from 23 to 32 nucleotides in length that play crucial roles in maintaining the integrity and stability of the animal genome [7, 8] Like other small noncoding RNAs, piRNAs are involved in a range of pathophysiological processes. PiRNAs are abundantly expressed in the heart and are involved in cardiac differentiation, repair, circulatory system regeneration, and cardiac hypertrophy [9–13]. For example, the piRNA CHAPIR regulates myocardial hypertrophy by modulating the METTL3-dependent N⁶-methyladenosine methylation of Parp10 mRNA [14]. CFAPIR inhibits myocardial fibrosis through the regulation of the muscleblind-like protein MBNL2 [15]. PiRNA-6426 alleviates heart failure by increasing DNMT3B-mediated SOAT1 methylation, whereas HNEAP regulates cardiomyocyte necrosis by inhibiting the m⁵C methylation of the ATF7 mRNA [16, 17]. Furthermore, piRNAs show promise as potential biomarkers and therapeutic targets for cardiovascular disorders [18]. In contrast to their roles in the reproductive system, the current understanding of the biological functions of piRNAs in cardiac physiology, particularly their regulatory mechanisms in myocardial I/R injury, remains limited. Thus, the potential role of piRNAs in I/R injury remains largely unexplored. In our previous study, we identified the highly conserved piRNA-CRAPIR as a key regulator of cardiomyocyte proliferation and revealed that CRAPIR promotes cardiomyocyte proliferation by competing with NF110 for binding to the RNA-binding protein PA2G4 [19]. The role of CRAPIR in myocardial I/R injury remains unknown.

Serine/arginine-rich (SR) proteins are a family of proteins with conserved domains consisting of 12 canonical members [20]. Serine/arginine-rich splicing factor 1 (SRSF1, also known as ASF/SF2) was the first identified family member, is considered the prototypical member, and remains one of the best-studied proteins. In addition to its roles in mRNA transcription, export, translation, and nonsense-mediated decay (NMD), SRSF1 also plays a crucial role in maintaining genomic stability and nucleolar stress [21–25]. Further studies have shown that SRSF1 is involved in vascular smooth muscle cell proliferation, immune system regulation, and the maintenance of skeletal muscle development [26–28]. LINC00174 directly interacts with SRSF1 to inhibit p53 expression and attenuate I/R injury [29]. In summary, SRSF1 plays a critical role in cardiovascular diseases. However, the mechanism by which CRAPIR regulates SRSF1 in the context of I/R injury remains to be elucidated.

These findings demonstrate that CRAPIR plays a crucial role in regulating cardiomyocyte apoptosis during I/R injury. CRAPIR binds to SRSF1, thereby increasing MDM2 stability and promoting its expression, which subsequently leads to reduced p53 expression.

Materials and methods

Animals

C57BL/6 adult mice (8–10 weeks old) were purchased from Liaoning Changsheng Biotechnology Co. Ltd. (Liaoning, China). The mice were housed under a 12-h light/12-h dark cycle at 20 ± 4 °C with 30%–60% humidity. Food and water were provided ad libitum throughout the experiment. All the animal experiments were approved by the guiding principles for the care and use of laboratory animals of the Second Affiliated Hospital of Harbin Medical University. All the animal experiments were approved by the Ethics Committee of the Second Affiliated Hospital of Harbin Medical University (No. IRB3002724).

Mouse model of myocardial ischemia‒reperfusion injury

A mouse model of I/R injury was established by inducing 45 min of ischemia followed by 24 h of reperfusion. Briefly, male mice were anesthetized with 2% 2,2,2-trifluoroethanol (0.1 mL/10 g) via intraperitoneal (i.p.) injection. The ventilation system (Ugo Basile Mouse Ventilator Cat. No. 28025) was established with a tidal volume of 250 μL and a frequency of 108 breaths/min. After the heart was exposed, the left anterior descending (LAD) coronary artery was ligated with a 7–0 silk thread. After 45 min of ischemia, the ligature was opened and reperfused for 24 h. For the sham-operated mice, the same procedure was performed without LAD ligation. The agomir/antagomir mixture was mixed with the transfection reagent Entranster (Engreen Biosystem Co., Ltd., Beijing, China; Cat. No. 18668-11-1). In accordance with the manufacturer’s instructions, agomir/antagomir (μg) and transfection reagent (μL) were mixed at a 2:1 ratio to prepare a 30 μL mixture, which was then randomly injected into the left ventricle at six points. The negative control (NC) group was injected with the same volume of agomir/antagomir negative control mixture mixed with the transfection reagent.

Isolation and culture of neonatal mouse ventricular cardiomyocytes (NMVCs)

Cardiomyocytes were isolated from neonatal mice (1–3 d) and cultured. Briefly, after the mice were disinfected with 75% alcohol, the hearts were collected and transferred to a Petri dish placed on ice with Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, USA). The heart tissues were digested with 0.25% trypsin (Meilunbio, Dalian, China) and then purified on the basis of cell adhesion. The obtained NMVCs were cultured in DMEM supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin‒streptomycin. NMVCs were maintained at 37 °C in a 5% CO₂ atmosphere with 95% humidified air. For hypoxia/reoxygenation (H/R) treatment, the medium was initially replaced with sugar-free DMEM/F12 (Meilunbio, Dalian, China) and the cells were then cultured in a low-oxygen environment at 37 °C in 5% CO₂ and 95% N₂ for 12 h. After hypoxia treatment, the medium was replaced with DMEM supplemented with 10% fetal bovine serum and 1% penicillin‒streptomycin, followed by incubation in a 5% CO₂ atmosphere with 95% humidified air at 37 °C for 24 h.

Culture of the AC16 cardiomyocyte cell line

Human AC16 cells were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin‒streptomycin. AC16 cells were maintained at 37 °C in a 5% CO₂ atmosphere with 95% humidified air. The H/R damage model was established as NMVCs.

Isolation of adult mouse cardiomyocytes and cardiac fibroblasts

The mice were anesthetized via intraperitoneal injection of 0.1 mL/10 g of 2% 2,2,2-trifluoroethanol. The hearts were rapidly excised and mounted on a Langendorff apparatus and perfused with calcium-free Tyrode’s solution for 5 min, followed by perfusion with calcium Tyrode’s solution containing collagenase type II (Gibco, USA) and bovine serum albumin (BSA) (BioFroxx, Germany). After the heart tissue softened, the left ventricle was excised, and cardiomyocytes were gently isolated in stop buffer. The composition of Tyrode’s solution was as follows: 126 mM NaCl, 5.4 mM KCl, 10 mM HEPES, 0.33 mM NaH₂PO₄, 1.0 mM MgCl₂·6H₂O and 10 mM glucose (pH 7.4, adjusted with NaOH).

For the isolation of adult cardiac fibroblasts, the mice were anesthetized and disinfected with 75% alcohol, and the heart tissue from the ischemia/reperfusion area was excised. The heart tissue was minced into small, uniform pieces and incubated in 3–4 mL of collagenase type II solution at 37 °C for 15–20 min. The supernatant was collected, and this process was repeated until the heart tissue was completely digested. Adult cardiac fibroblasts were isolated for subsequent experimental analysis.

Transfection of the CRAPIR agomir and antagomir

The CRAPIR agomir/antagomir and their negative controls were designed and synthesized by Gene Pharma Co., Ltd. (Suzhou, China), and were transfected into cardiomyocytes using Lipofectamine 2000 (Invitrogen, NY, USA). For example, agomir (5 μL)/antagomir (10 μL) or the same volume of Lipofectamine 2000 with 100 μL of DMEM was added to 6 wells, incubated for 5 min, and then mixed. The combined mixture was incubated for 15‒20 min, mixed with 1,800 μL of DMEM supplemented with 10% fetal bovine serum, and added to a well of a 6-well plate. The sequences are as follows:

CRAPIR agomir:

sense: 5’-GCAUUGGUGGUUCAGUGGUAGAAUUCUCGCC-3’

antisense: 5’-CGAGAAUUCUACCACUGAACCACCAAUGCUU-3’

Agomir negative control (ago-NC):

sense: 5’-UGGCAACCUGCCGGUGAAUUGUUCUGUGUGA-3’

antisense: 5’-ACACAGAACAAUUCACCGGCAGGUUGCCAUU-3’

CRAPIR antagomir:

sense: 5’-GGCGAGAAUUCUACCACUGAACCACCAAUGC-3’

Antagomir negative control (anta-NC):

sense: 5’-AGAGCACGAGAGCAUUCCCAUCCUUAGACAC-3’

Transfection of small interfering RNAs and plasmids

Small interfering RNA (siRNA) oligonucleotides specific for SRSF1, MDM2 and SRSF1 plasmids were designed and synthesized by Gene Pharma Co., Ltd., and were transfected into cardiomyocytes using Lipofectamine 2000. The sequences of the siRNAs used were as follows:

SRSF1 siRNA:

sense: 5’-GCAGUUCGAAAACUGGAUATT-3’

antisense: 5’-UAUCCAGUUUUCGAACUGCTT-3’

Negative control (si-NC):

sense: 5’-UUCUCCGAACGUGUCACGUTT-3’

antisense: 5’-ACGUGACACGUUCGGAGAATT-3’

MDM2 siRNA:

sense: 5’-GAAGAUAAAGGGAAAGAUA-3’

antisense: 5’-UAUCUUUCCCUUUAUCUUC-3’

Echocardiography

Mouse cardiac function was evaluated using transthoracic echocardiography on a Vevo 1100 Ultrasound machine (VisualSonics, Toronto, ON, Canada). Three days post-surgery, the mice were anesthetized and positioned on a heated platform. M-mode imaging was employed to assess the ejection fraction (EF) and fractional shortening (FS).

Evans blue/triphenyl tetrazolium chloride (TTC) staining

After the mice were anesthetized, the LAD coronary artery was via thoracotomy. Subsequently, 2% Evans blue (Biosharp, China, BS177-1g) was injected into the abdominal aorta. The hearts were harvested immediately, serially sectioned into 2-mm-thick slices, and stained with 2% TTC (Solarbio, China, G3005) at 37 °C for 20 min. After fixation with 4% paraformaldehyde for 15 min, images were captured. The infarct area and area at risk were quantified via ImageJ.

TUNEL staining

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was conducted using the TUNEL BrightRed Apoptosis Detection Kit (Vazyme, China, A113-01) in accordance with the manufacturer’s instructions. After TUNEL staining, the cell nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI) (Solarbio, China). TUNEL-positive cells were observed under a fluorescence microscope. The percentage of TUNEL-positive cells was calculated.

Lactate dehydrogenase (LDH) and creatine kinase isoenzyme MB (CKMB) activity detection assay

Lactate dehydrogenase (LDH) activity in the culture supernatant of cardiomyocytes and mouse serum was measured using an LDH detection kit (Nanjing JianCheng, Jiangsu, China, A020-1) following the manufacturer’s instructions. The absorbance at 450 nm was quantified via a microplate reader (BioTek, Richmond, VA, USA). Serum creatine kinase isoenzyme MB (CK-MB) activity was evaluated using an automatic biochemical analyzer (Beckman Coulter, USA).

RNA pulldown assay

Biotinylated CRAPIR and CRAPIR-NC were designed and synthesized by Gene Pharma Co., Ltd. The interaction between CRAPIR and SRSF1 was detected using the Pierce Magnetic RNA‒Protein Pull-Down Kit (Thermo Scientific, USA, 20164) according to the manufacturer’s protocols. Ten percent of the total cell lysate volume was used as input, and the remaining mixture was incubated with streptavidin magnetic beads with protein A/G magnetic beads (Thermo Scientific, USA, 88803). The magnetic beads were eluted and analyzed by Western blotting.

The sequences are as follows:

Biotinylated-CRAPIR: 5’-GCATTGGTGGTTCAGTGGTAGAATTCTCGCC-3’

Biotinylated-negative control (NC): 5’-UUGUACUACACAAA-3’

RNA-binding protein immunoprecipitation (RIP) assay

RNA-binding protein immunoprecipitation (RIP) experiments were conducted according to the following protocol. Briefly, cells were harvested and resuspended in a 1.5-fold cell volume of RIP lysate. A total of 50 μL of magnetic beads was washed three times with RIP wash buffer and then incubated with SRSF1 antibody (Proteintech, China, 12929-2-AP) at 4 °C overnight. The magnetic bead-antibody complex was then incubated with the lysed cells at 4 °C overnight. After incubation, the magnetic beads were washed six times with RIP wash buffer, and the supernatant was discarded. The beads were then eluted with 300 μL of elution buffer at 50 °C for 40 min. RNA from the three groups was extracted using the TRIzol method and analyzed via qRT‒PCR. Ten percent of the cell lysate was used for the input group, and the remaining 90% was evenly divided between the IgG (negative control) and the SRSF1-treated groups.

Fluorescence in situ hybridization (FISH)

Fluorescence-labeled oligonucleotide probes complementary to CRAPIR junction sequences were designed and synthesized by Gene Pharma Co., Ltd. An RNA FISH assay was performed via an RNA FISH kit (Gene Pharma, Shanghai, China) according to the manufacturer’s protocols. The nuclei were counterstained with DAPI. Fluorescence staining was then examined using a confocal microscopy (Nikon C2, Japan).

JC-1 staining

The mitochondrial membrane potential (Δψm) was assessed using the Enhanced Mitochondrial Membrane Potential assay kit with JC-1 according to the manufacturer’s protocols (Beyotime, Shanghai, China, C2003S). The results were observed using a fluorescence microscope.

Western blot

Proteins were extracted from cells or tissues using RIPA lysis buffer (Beyotime, Shanghai, China, P0013B) containing protease inhibitors. Protein concentrations were measured via an Enhanced BCA Protein Assay Kit (Beyotime, Shanghai, China, P0009). The samples were separated by SDS‒PAGE, and images were captured via an Odyssey Infrared Imaging System (Odyssey, USA) [30]. The target antigens were detected with the following antibodies: Bax (Proteintech, 50599-2-Ig, 1:1000), cleaved caspase-3 (Wanleibio, WL02117, 1:1000), Bcl2 (Proteintech, 68128-1-Ig, 1:1000), p53 (Proteintech, 21891-1-AP, 1:1000), SRSF1 (Proteintech, 12929-2-AP, 1:500), MDM2 (Abways, AY2958, 1:500), phospho-p53 (ABclonal, AP0083, 1:1000), ubiquitin (Santa Cruz, sc-8017, 1:500) and β-actin (Abways, AB0035, 1:1000).

Real-time quantitative PCR (qRT‒PCR)

Total RNA was extracted from cells or tissues with TRIzol reagent (Invitrogen, USA, Carlsbad, CA) according to the manufacturer’s protocol. The RNA concentration was determined with a NanoDrop 2000c (Thermo Scientific, USA). Total RNA was reverse transcribed into cDNA via a ReverTra Ace qPCR RT Kit (TOYOBO, Japan, FSQ-101) following the manufacturer’s guidelines. The expression levels of RNA were measured with an ABI 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, USA) using SYBR Green PCR Master Mix (Roche, Switzerland). Relative RNA levels were analyzed by using the 2^−ΔΔct method, with U6 and 18S rRNA as internal controls.

The primers used were as follows:

CRAPIR: Forward primer 5'-AACAAGGCATTGGTGGTTCAG-3'

Reverse primer 5'-GTCGTATCCAGTGCAGGGT-3'

U6: Forward primer 5'-GCTTCGGCAGCACATATACTAAAAT-3'

Reverse primer 5'-CGCTTCACGAATTTGCGTGTCAT-3'

SRSF1: Forward primer 5'-GTTTACCGAGATGGCACTGGTG-3'

Reverse primer 5'-ACCCGGATGTAGGCAGTTTCTC-3'

MDM2: Forward primer 5'-CCGAGTTTCTCTGTGAAGGAGC-3'

Reverse primer, 5'-GTCTGCTCTCACTCAGCGATGT-3'

18 s: Forward primer: 5'-GGCGGCTTGGTGACTCTAGATAAC-3'

Reverse primer: 5'-CCTGCTGCCTTCCTTGGATGTG-3'

Statistical analysis

All the data are presented as the means ± SEMs. GraphPad Prism 9.4 was used for data analysis. Comparisons between two groups were conducted using an unpaired t test, whereas comparisons involving multiple groups were conducted using one-way ANOVA followed by Tukey’s multiple comparisons test. P < 0.05 was considered statistically significant.

Results

CRAPIR is prominently expressed in cardiomyocytes under I/R injury conditions and is markedly upregulated 24 h after I/R and H/R injury

Our research team previously identified a piRNA associated with cardiac regeneration, which was named CRAPIR [19]. Therefore, we wanted to further explore the role of CRAPIR in I/R injury. To explore the dynamics of CRAPIR expression during I/R injury, we first assessed its levels in heart tissue at 24 and 72 h post-I/R. Our results revealed that, compared with that in the sham group, CRAPIR was upregulated at 24 h following I/R but downregulated at 72 h compared with that in the reperfusion group at 24 h (Fig. 1a). Furthermore, CRAPIR expression in cardiomyocytes isolated from the I/R-treated tissues was greater than that in the sham group (Fig. 1b). Notably, CRAPIR expression was more prominent in cardiomyocytes than in fibroblasts from the I/R model (Fig. 1c). We subsequently validated the changes in the expression of CRAPIR in vitro. The data revealed an upregulation of CRAPIR at 24 h, followed by a downregulation at 48 h, compared with that of the reoxygenation group at 24 h (Fig. 1d). Fluorescence in situ hybridization (FISH) further confirmed an increase in CRAPIR expression at 24 h after H/R, with CRAPIR being localized in both the nucleus and cytoplasm of cardiomyocytes (Fig. 1e). Overall, these data suggest that CRAPIR is enriched in cardiomyocytes following I/R injury and is involved in regulating I/R injury.

Fig. 1. CRAPIR is elevated in cardiac tissue subjected to I/R and H/R-treated NMVCs.

a The level of CRAPIR in I/R-treated heart tissues compared with that in sham-operated heart tissues. n = 20 or n = 13 or n = 10. b The expression of CRAPIR in cardiomyocytes (CMs) isolated from I/R- and sham-operated hearts. n = 10 or n = 8. c The level of CRAPIR in CMs and cardiac fibroblasts (CFs) isolated from I/R hearts. n = 10 or n = 8. d The expression of CRAPIR in H/R-treated and control (CTL) NMVCs. n = 15  or n = 6 or n = 7. e Fluorescence in situ hybridization (FISH) was used to detect changes in CRAPIR in NMVCs after H/R injury. Scale bar = 20 μm. n = 3. All the data are presented as the mean ± SEM. ^***P < 0.001 versus the sham or sham-CM or I/R-CF or CTL group. ^###P < 0.001 versus the I/R–24 h or H/R-12/24 h group.

Overexpression of CRAPIR alleviates I/R injury

To further explore the role of CRAPIR in the pathophysiological process of I/R injury, we established an I/R model in adult mice. The model was constructed by performing 45 min of ligation of the left anterior descending (LAD) artery, followed by concurrent injection of the CRAPIR agomir into the myocardium and subsequent reperfusion (Fig. 2a). We subsequently evaluated CRAPIR expression at 1, 2, and 3 d post-injection via qRT‒PCR. Our results revealed that CRAPIR expression in heart tissue was elevated on Days 1, 2, and 3 post-injection, with the highest expression observed on Day 1 compared with that in the ago-NC group (Fig. 2b). FISH staining revealed that the level of CRAPIR overexpression also markedly increased, and CRAPIR was distributed in the nucleus and cytoplasm (Fig. S1). Three days after injection, echocardiographic analysis revealed that CRAPIR alleviated the reductions in the ejection fraction (EF) and fractional shortening (FS) induced by I/R (Fig. 2c, d). Moreover, Evans blue and TTC staining demonstrated that CRAPIR overexpression reduced the infarct size (Fig. 2e, f). Moreover, CRAPIR overexpression suppressed the I/R-induced increase in the serum levels of creatine kinase MB (CK-MB) and lactate dehydrogenase (LDH) (Fig. 2g, h).

Fig. 2. Overexpression of CRAPIR reduces cardiomyocyte apoptosis and normalizes cardiac function following I/R.

a Schematic diagram of the experimental procedure. b After 1 d, 2 d and 3 d of treatment with the intramyocardial agomir-CRAPIR or agomir-NC, qRT‒PCR was used to determine the expression of CRAPIR. n = 3. c, d Representative echocardiographic images and statistical analysis of EF and FS after 72 h of intramyocardial injection of the CRAPIR agomir in the I/R injury model. n = 6. e, f Infarct area after I/R, as determined by Evans blue/TTC staining. Scale bar = 5 mm. n = 4. g Serum CK-MB levels after intracardiac injection of the CRAPIR agomir after I/R injury. n = 6. h Serum LDH levels after intracardiac injection of the CRAPIR agomir after I/R injury. n = 6. i The percentage of cardiomyocytes was assessed using TUNEL staining. Scale bar = 50 μm. n = 4. j–m The protein levels of cleaved caspase-3, caspase-3, Bax, Bcl2, phospho-p53 and p53, as shown by Western blotting. n = 3 or n = 4 or n = 5. EF, ejection fraction. FS, fractional shortening. The data are presented as the means ± SEMs. ^**P < 0.01, ^***P < 0.001 versus the ago-NC or sham group; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the ago-NC or I/R + ago-NC group; ^$P ＜ 0.05 versus the ago-NC group.

We further explored whether CRAPIR overexpression affects cardiomyocyte apoptosis induced by I/R injury. TdT-mediated dUTP nick-end labeling (TUNEL) staining revealed that CRAPIR overexpression reduced I/R-induced cardiomyocyte apoptosis (Fig. 2i). Finally, Western blot analysis further verified that, compared with those of the I/R + NC group, CRAPIR overexpression did not affect total caspase-3 expression but significantly reduced the levels of cleaved caspase-3 protein (Fig. 2j) and BCL2-associated X protein (Bax) (Fig. 2k) and upregulated the expression of the antiapoptotic protein B-cell leukemia/lymphoma 2 (Bcl2) (Fig. 2l). Moreover, CRAPIR overexpression inhibited the expression of phospho-tumor suppressor protein 53 (p-p53) and p53 compared with that in the I/R + NC group (Fig. 2m). These findings suggest that CRAPIR overexpression mitigates the reduction in cardiac function and cardiomyocyte apoptosis induced by I/R in vivo.

CRAPIR overexpression reduces NMVCs injury caused by H/R

Next, we investigated whether the effects of CRAPIR overexpression in vitro were consistent with those observed in vivo. We isolated ventricular myocytes from neonatal mice and established a hypoxia/reoxygenation (H/R) model [31]. The expression of CRAPIR in NMVCs increased after transfection (Fig. 3a). CRAPIR overexpression reduced LDH release and the TUNEL-positive apoptosis rate in NMVCs following H/R injury (Fig. 3b, c). Myocardial I/R injury induces oxidative stress in cardiomyocytes. Oxidative stress causes mitochondrial dysfunction, which in turn affects the mitochondrial membrane potential, leading to apoptosis. Therefore, we measured changes in the mitochondrial membrane potential using the JC-1 method [32]. CRAPIR overexpression reversed the H/R-induced decrease in the mitochondrial membrane potential and reduced cell apoptosis during the early stage (Fig. 3d). Consistent with these findings, CRAPIR overexpression downregulated cleaved caspase-3 and Bax expression (Fig. 3e, f) and upregulated Bcl2 expression (Fig. 3g) in H/R-stimulated cells compared with those in the H/R + NC group. In vitro, CRAPIR overexpression also downregulated the protein expression of p53 after H/R injury (Fig. 3h). These findings demonstrated that CRAPIR overexpression reduced H/R-induced apoptosis in NMVCs and decreased LDH release into the medium. In summary, CRAPIR serves as a crucial regulator of myocardial I/R injury, and increased expression of CRAPIR effectively mitigated I/R injury and NMVC injury triggered by H/R.

Fig. 3. Overexpression of CRAPIR reduced H/R-induced injury to NMVCs.

a Transfection of NMVCs with agomir-CRAPIR or agomir-NC was verified. n = 3. b Effects of CRAPIR on LDH release in NMVCs compared with H/R injury. n = 4. c Representative TUNEL staining and statistical analysis of NMVCs transfected with the agomir CRAPIR before H/R injury. Scale bar = 50 μm. n = 4. d Representative JC-1 staining image. Scale bar = 50 μm. n = 3. e–h Cleaved caspase-3, Bax, Bcl2 and p53 levels were evaluated via Western blot analysis. n = 4 or n = 5. The data are expressed as the means ± SEMs. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 versus the ago-NC or CTL group; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the H/R + ago-NC group.

Inhibition of CRAPIR exacerbates I/R-induced injury

To further confirm the role of CRAPIR knockdown in I/R injury, we concurrently injected the CRAPIR antagomir into the myocardium, followed by reperfusion (Fig. 4a). We subsequently assessed CRAPIR expression at 1, 2, and 3 d post-injection via qRT‒PCR. The results revealed that CRAPIR expression in heart tissue was suppressed on Days 1, 2, and 3 post-injection, with the lowest expression observed on Day 1 compared with that in the anta-NC group (Fig. 4b). Compared with those of the I/R + NC group, CRAPIR inhibition reduced EF and FS, thereby exacerbating cardiac function injury caused by I/R in mice (Fig. 4c, d). Evans blue/TTC staining further confirmed that CRAPIR inhibition increased infarct sizes (Fig. 4e, f) and increased CK-MB and LDH expression levels in mouse serum (Fig. 4g, h). TUNEL staining revealed that CRAPIR knockdown increased the positive rate of cardiomyocyte apoptosis following I/R injury (Fig. 4i). Western blot analysis revealed that CRAPIR inhibition increased cleaved caspase-3 protein and Bax expression (Fig. 4j, k), reduced Bcl2 protein expression (Fig. 4l), and increased phospho-p53 and p53 protein expression in heart tissue after I/R (Fig. 4m). Taken together, these results suggest that the inhibition of CRAPIR expression in mouse hearts exacerbates the injury induced by I/R.

Fig. 4. Inhibition of CRAPIR exacerbates I/R injury to the heart in mice.

a Schematic diagram of the experimental procedure. b The levels of CRAPIR in the hearts of mice after intracardiac injection of the antagomir-CRAPIR or antagomir-NC. n = 4. c, d Representative echocardiography and statistical analysis of EF and FS after intracardiac injection of the CRAPIR antagomir after I/R injury. n = 6. e, f Representative heart sections from mice with I/R-injured hearts stained with Evans blue/TTC after in situ CRAPIR antagomir injection were statistically analyzed for mean infarct size. Scale bar = 5 mm. n = 4. g Serum CK-MB levels after intracardiac injection of the CRAPIR antagomir after I/R injury. n = 5. h Serum LDH levels after intracardiac injection of the CRAPIR antagomir after I/R injury. n = 6. i Representative TUNEL staining and statistical analysis after intracardiac injection of the CRAPIR antagomir after I/R injury. Scale bar = 50 μm. n = 4. j–m Cleaved caspase-3, Bax, Bcl2, phospho-p53 and p53 levels were evaluated via Western blot analysis. n = 3 or n = 4 or n = 5. The data are expressed as the means ± SEMs. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 versus the anta-NC or sham group; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the anta-NC or I/R + anta-NC group; ^$$$P < 0.001 versus the anta-NC group.

Suppressing CRAPIR worsens the damage induced by H/R

As demonstrated above in vivo, the inhibition of CRAPIR expression exacerbated myocardial I/R injury in mice. Next, we investigated whether the inhibition of CRAPIR exacerbates myocardial H/R-induced injury in vitro. CRAPIR expression was successfully inhibited in NMVCs transfected with the CRAPIR antagomir (Fig. 5a). Inhibiting the expression of CRAPIR not only increased the release of LDH (Fig. 5b) but also significantly increased the percentage of TUNEL-positive cells compared with that in the H/R + anta-NC group (Fig. 5c). In addition, inhibition of CRAPIR exacerbated the decrease in the mitochondrial membrane potential induced by H/R (Fig. 5d). Similar to the results of the in vivo experiments, compared with the H/R + anta-NC treatment, the CRAPIR antagomir significantly increased the expression levels of cleaved caspase-3, Bax and p53 while reducing the protein expression of Bcl2 (Fig. 5e–h). These results further suggest that the inhibition of CRAPIR exacerbates H/R-induced apoptosis in NMVCs in vitro.

Fig. 5. Inhibition of CRAPIR increased H/R-induced NMVCs damage.

a NMVCs were transfected with antagomir-CRAPIR or antagomir-NC, and the results were verified. n = 3. b Effects of inhibiting CRAPIR on LDH release in NMVCs compared with H/R injury. n = 4. c Representative TUNEL image and statistical analysis of NMVCs transfected with the antagomir CRAPIR before H/R injury. Scale bar = 50 μm. n = 4. d Representative JC-1 staining image. Scale bar = 50 μm. n = 3. e–h The cleaved caspase-3, Bax, Bcl2 and p53 levels were evaluated via Western blot analysis. n = 4 or n = 5. The data are expressed as the means ± SEMs. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 versus the anta-NC or CTL group; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the H/R + anta-NC group.

In summary, the above data suggest that the overexpression of CRAPIR protects against cardiac dysfunction induced by I/R injury, whereas the inhibition of CRAPIR accelerates cardiomyocyte apoptosis mediated by I/R injury, thereby exacerbating cardiac damage.

CRAPIR binding to SRSF1 mediates the effects of CRAPIR on H/R injury

One mechanism by which piRNAs exert their biological function is through binding to specific proteins [33]. To further elucidate the molecular mechanism by which CRAPIR attenuates I/R injury, we identified CRAPIR-interacting proteins using RNA pulldown combined with mass spectrometry (MS). In the MS results, the intersection of the top 20 proteins ranked by log fold change≥2, coverage percentage, and unique peptides≥2 was identified. Venn diagram analysis revealed that SRSF1, CNBP, and PPIB were selected (Fig. 6a). Previous studies have shown that vascular endothelial cell-derived exosome LINC00174 directly interacts with SRSF1 to inhibit p53 expression and attenuate I/R-induced myocardial injury [29]. According to these studies, SRSF1 plays a crucial role in cardiovascular diseases and contributes to cardiomyocyte apoptosis. Therefore, we further investigated the relationship between CRAPIR and SRSF1. CRAPIR-interacting proteins were validated using an RNA pulldown assay with a biotinylated CRAPIR probe. The results demonstrated that CRAPIR binds to SRSF1 (Fig. 6b). Additionally, an RNA immunoprecipitation (RIP) assay confirmed that SRSF1 could interact with CRAPIR (Fig. 6c). We subsequently treated NMVCs transfected with the CRAPIR agomir with actinomycin D (ActD). The results suggested that overexpression of CRAPIR had no effect on the mRNA stability of SRSF1 (Fig. S2). These experiments further confirmed that CRAPIR binds to SRSF1 in NMVCs in vitro. We then delved deeper into the mechanistic link between CRAPIR and SRSF1 to determine whether the former plays a cardioprotective role through the latter. Furthermore, we found that under H/R injury conditions, the overexpression of CRAPIR increased the protein expression of SRSF1, whereas the inhibition of CRAPIR decreased the protein expression of SRSF1 (Fig. 6d, e). Next, SRSF1 siRNA was transfected into cardiomyocytes, and the protein expression of SRSF1 was significantly reduced, as confirmed by Western blotting (Fig. S3). We performed rescue experiments to determine whether CRAPIR mediates cardiomyocyte apoptosis through SRSF1. NMVCs were cotransfected with an agomir of CRAPIR and SRSF1 siRNA, followed by H/R-induced cardiomyocyte injury. Overexpression of CRAPIR protected cardiomyocytes from H/R injury; however, cotransfection of CRAPIR and siSRSF1 increased LDH release into the culture medium and the percentage of TUNEL-positive cardiomyocytes (Fig. 6f, g). In contrast, the transfection of siSRSF1 counteracted the effect of CRAPIR on the mitochondrial membrane potential (Fig. S4). Moreover, cleaved caspase-3, Bax and p53 were upregulated, and Bcl2 was downregulated (Fig. 6h–k). These data suggest that SRSF1 silencing reversed the cardioprotective effect of CRAPIR overexpression in cardiomyocytes and that SRSF1 is a downstream molecule of CRAPIR-mediated cardiomyocyte injury.

Fig. 6. CRAPIR binding to SRSF1 mediates the effects of CRAPIR on H/R injury.

a Venn diagram. b Western blot analysis of SRSF1 pulled down by CRAPIR. n = 3 c RNA immunoprecipitation (RIP) of CRAPIR with an anti-SRSF1 antibody in NMVCs vs. IgG. n = 5. d, e SRSF1 levels were evaluated via Western blot analysis of NMVCs transfected with agomir-CRAPIR or antagomir-CRAPIR before H/R injury. n = 4. f Effects of NMVCs transfected with agomir-CRAPIR and siSRSF1 on LDH release in NMVCs before H/R injury. n = 6. g Representative TUNEL staining and statistical analysis of NMVCs transfected with agomir-CRAPIR or siSRSF1 before H/R injury. Scale bar = 50 μm. n = 4. h–k Cleaved caspase-3, Bax, Bcl2 and p53 levels were evaluated via Western blot analysis. n = 4 or n = 6. The data are expressed as the means ± SEMs. ^**P < 0.01, ^***P < 0.001 versus the lgG or CTL or H/R+ ago-NC group; ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the H/R + ago-NC or H/R + anta-NC or H/R + ago-CRAPIR + si-NC group.

CRAPIR mediates the ubiquitination of p53 by MDM2 through SRSF1

In previous in vivo and in vitro experiments, CRAPIR was shown to regulate the protein expression of p53. Previous studies have shown that SRSF1 is involved in the mRNA and protein expression of MDM2 in tumor cells and human fibroblasts but has not been reported in cardiomyocytes [34–37]. Many studies have shown that p53 protein activation is an important factor in cell apoptosis and that the E3 ubiquitin ligase MDM2 is a key negative regulator of p53 that is capable of ubiquitinating p53 and targeting it for proteasome-mediated degradation. Silencing MDM2 reduces the ubiquitination-mediated degradation of the p53 protein when the proteasome-mediated degradation of proteins is blocked by the proteasomal inhibitor MG132 (Fig. S5). We hypothesize that CRAPIR regulates MDM2-mediated effects on p53 by binding to SRSF1, thereby reducing cardiomyocyte apoptosis. The SRSF1 overexpression plasmid was subsequently constructed and transfected into cardiomyocytes for 48 h. Compared with that in the NC group, the expression of SRSF1 was significantly increased (Fig. S6). Consistent with our hypothesis, the overexpression of SRSF1 increased the stability of MDM2 mRNA expression when ActD was added (Fig. 7a). We thus tested the effect of SRSF1 on the stability of MDM2 by adding cycloheximide (CHX), which demonstrated that SRSF1 overexpression reduced the degradation of the MDM2 protein (Fig. 7b). Moreover, we found that CRAPIR overexpression was able to increase the protein level of MDM2, but inhibiting CRAPIR decreased the protein expression of MDM2 following H/R injury (Fig. 7c, d). We also observed that SRSF1 knockdown reversed the upregulation of the MDM2 protein under H/R injury caused by CRAPIR overexpression (Fig. 7e). After MDM2 was successfully knocked down, the protective effect of CRAPIR on cardiomyocytes after H/R injury was confirmed by cotransfection with MDM2 siRNA (Fig. S7, Fig. 7f–j). Taken together, these data suggest that CRAPIR regulates the ubiquitination of p53 by MDM2 by binding SRSF1 and participates in the pathophysiological process of I/R and H/R injury, which further explains the involvement of CRAPIR in cardiomyocyte apoptosis.

Fig. 7. CRAPIR mediates the ubiquitination of p53 by MDM2 through SRSF1.

a The effect of ActD on MDM2 stability was assessed at 0, 2, 4, and 6 h post-treatment via qRT‒PCR. n = 3. b The effect of CHX on MDM2 stability was assessed at 0 and 60 min post-treatment using Western blotting. n = 4. c, d MDM2 levels in NMVCs before H/R injury were evaluated via Western blot analysis following transfection with agomir-CRAPIR or  antagomir-CRAPIR. n = 4 or n = 5. e MDM2 levels were evaluated via Western blot analysis of NMVCs transfected with agomir-CRAPIR or siSRSF1 before H/R injury. n = 5. f LDH release in NMVCs before H/R injury after transfection with agomir-CRAPIR and siMDM2. n = 4. g Representative TUNEL staining and statistical analysis of NMVCs transfected with agomir-CRAPIR or siMDM2 before H/R injury. Scale bar = 50 μm. n = 3. h–j Cleaved caspase-3, Bax and Bcl2 were evaluated via Western blot analysis. n = 3. The data are expressed as the means ± SEMs. ^**P < 0.01, ^***P < 0.001 versus the NC or CTL or H/R+ ago-NC group, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001 versus the NC+CHX or H/R+ ago-NC or H/R+ anta-NC or H/R + ago-CRAPIR + si-NC group.

Effects of CRAPIR on AC16 cells exposed to H/R injury

We evaluated the effects of CRAPIR on human AC16 cells exposed to H/R injury. CRAPIR was also upregulated in AC16 cells after H/R (Fig. S8a). The overexpression of CRAPIR further reduced LDH release in AC16 cells (Fig. S8b), reduced cell apoptosis (Fig. S8c), downregulated the expression of cleaved caspase-3 and Bax (S8d, e) and upregulated the expression of Bcl2 following H/R injury (Fig. S8f). These results suggest that CRAPIR has the same protective effect on human AC16 cells.

Discussion

In this study, we identified the PIWI-interacting RNA CRAPIR as a key regulator of myocardial I/R injury. Both in vivo and in vitro experiments demonstrated that CRAPIR overexpression alleviated I/R-induced cardiomyocyte apoptosis, normalized cardiac function and reduced infarct size. Conversely, CRAPIR inhibition exacerbated cardiomyocyte apoptosis and myocardial injury. Mechanistically, CRAPIR binds to SRSF1, thereby upregulating the expression of MDM2. Elevated MDM2 subsequently facilitates the ubiquitination of p53. This change ultimately leads to a reduction in p53 expression, conferring protection against myocardial injury in the hypoxia-reoxygenation (H/R) model (Fig. 8). These findings not only elucidate the molecular mechanism of myocardial I/R injury but also suggest that CRAPIR may serve as a novel therapeutic target for myocardial I/R injury.

Fig. 8. Schematic diagram of the regulatory effect of CRAPIR on myocardial I/R injury.

Schematic diagram showing that upon ischemia/reperfusion stress, CRAPIR overexpression stabilizes the expression of the SRSF1 protein, and SRSF1 upregulates MDM2 mRNA and protein expression. MDM2 targets p53 for ubiquitination-mediated degradation, which reduces cardiomyocyte apoptosis induced by ischemia‒reperfusion injury and alleviates myocardial injury. Created at https://BioRender.com.

PiRNAs have previously been reported to be involved in the pathophysiological processes of cardiac hypertrophy, cardiac fibrosis, and cardiomyocyte necrosis [14, 17, 38]. To date, the functions of most piRNAs have not been determined, nor have their pathophysiological processes in cardiovascular disease. Although a related piRNA, HAAPIR, has been reported to regulate cardiomyocyte death following myocardial infarction via NAT10-mediated ac4C acetylation of Tfec mRNA, the involvement of piRNAs in cardiac I/R injury remains largely unexplored. In our previous research, we identified piRNAs related to cardiomyocyte proliferation and reported that the piRNA-CRAPIR is essential for cardiomyocyte proliferation [19]. In this study, we found that CRAPIR responds to oxidative stress, with its expression was upregulated within 24 h post-I/R injury, likely as a protective mechanism, but downregulated after 72 h compared with 24 h, indicating attenuation of this protective effect. Cardiomyocytes are the most important cells in the heart and are nonregenerative, and apoptosis is the main feature of myocardial injury [39, 40]. Cardiomyocyte apoptosis occurs in a model of myocardial I/R injury related to oxidative stress in cells, and this phenomenon occurs during reperfusion rather than during the nonischemic myocardial response [41]. After I/R injury, the expression of CRAPIR in cardiomyocytes was greater than that in fibroblasts, suggesting that CRAPIR may play a role in the apoptosis of cardiomyocytes. We established a CRAPIR agomir overexpression model via intracardiac myocardial injection. We found that CRAPIR overexpression inhibited cardiomyocyte apoptosis and normalized cardiac function after I/R injury, whereas the opposite effect was observed with suppression of CRAPIR in mice in vivo. Furthermore, our experiments confirmed that overexpressing CRAPIR can effectively suppress H/R-mediated cardiomyocyte apoptosis. In contrast, CRAPIR knockdown exacerbated cardiomyocyte apoptosis in vitro. Moreover, we discovered that both the long and short sequences of the agomir-NC and antagomir-NC had no discernible influence on H/R-mediated cardiomyocyte apoptosis (Fig. S9).

To elucidate the molecular mechanisms of CRAPIR in I/R injury, we examined its protein interactions. Since the biogenesis of piRNAs is species specific, the process in humans has not been well studied. The present study revealed that this molecule can regulate the selective splicing of transposons by binding to proteins, silence transposons by binding to target RNAs, regulate DNA methylation and histone modification, and interact with transcription factors to participate in posttranscriptional modifications and other biological functions [42–47]. Since proteins are the ultimate executors of gene function, we further investigated the binding proteins of CRAPIR and the molecular mechanism by which CRAPIR is involved in cardiomyocyte apoptosis. We sequenced the proteins bound to CRAPIR by mass spectrometry. By analyzing CRAPIR-binding proteins, we found that SRSF1 directly interacts with CRAPIR. In H/R-induced injury, we observed that overexpression of CRAPIR upregulated SRSF1; conversely, inhibition of CRAPIR had the opposite effect. SRSF1 is implicated in the p53 signaling pathway, where its overexpression decreases p53 levels, and its silencing increases p53 expression [29]. p53 is a sequence-specific transcription factor that is widely involved in the damaging process of apoptosis [48, 49]. The transcription of Bax induced by p53 as a transcription factor also directly activates Bax to permeabilize mitochondria and participate in apoptosis [50]. MDM2, an E3 ubiquitin ligase, mediates p53 ubiquitination and proteasomal degradation. SRSF1 stabilizes MDM2 mRNA and protein, thus influencing p53 levels; however, its role in cardiomyocytes has not been previously reported [34, 35]. The results of our research revealed that SRSF1 stabilizes the mRNA and protein of MDM2 and affects the ubiquitination-mediated degradation of p53. This finding further explains why CRAPIR protects cardiomyocytes from apoptosis and normalizes cardiac function. Under normal conditions, p53 levels in the heart are low because p53 is ubiquitinated by the E3 ubiquitin ligase MDM2, which targets p53 for proteasomal degradation [51, 52]. When the heart undergoes a stress response, p53 accumulates rapidly and promotes apoptosis [53]. Loss of MDM2 can stabilize p53 and further influence the apoptotic role of the p53 signaling pathway. Overexpression of CRAPIR restored MDM2 expression. Conversely, inhibition of CRAPIR reduces the expression of MDM2 and further affects the stability of p53. In the rescue assay, we further confirmed that silencing SRSF1 and MDM2 reversed the protective effect of CRAPIR. In summary, these results suggest that CRAPIR affects the ubiquitination of p53 by MDM2 through binding with SRSF1.

In addition, we further explored the possible upstream signaling pathways of CRAPIR. Previous studies have shown that the PI3K/AKT signaling pathway plays an important role in reducing myocardial infarction size and restoring cardiac function after I/R injury [54–57]. Activation of the PI3K/AKT pathway provides myocardial protection through synergistic promotion of antioxidant, anti-inflammatory and autophagic activities and inhibition of mitochondrial dysfunction and cardiomyocyte apoptosis. The results showed that inhibition of PI3K/AKT attenuated H/R-mediated upregulation of CRAPIR (Fig. S10). Therefore, PI3K/AKT signaling might mediate the upregulation of CRAPIR during I/R injury.

However, the use of intracardiac injection of the CRAPIR agomir in this study is acknowledged as a potential constraint. The use of intracardiac injection post-I/R injury presents limitations in its applicability to clinical interventions. The protective efficacy of CRAPIR overexpression against I/R has been substantiated in vivo and in vitro. Research on stem cell therapy and biomaterials has led to new strategies for the treatment of ischemic cardiomyopathy, such as the transplantation of platelet gel and cardiosphere-derived cells [58–61]. In the future, we will explore new dosage forms of nanotherapeutic drugs or exosomes encapsulating ago-CRAPIR to achieve clinical therapeutic strategies for intravenous administration and heart-specific targeting [62, 63].

CRAPIR is a newly discovered cardioprotective factor that inhibits apoptosis in I/R injury, highlighting its potential as a therapeutic target for I/R treatment through the SRSF1/MDM2/p53 signaling axis.

Conclusions

CRAPIR protects against myocardial I/R-induced injury by binding to SRSF1 and increasing the expression of MDM2, thereby promoting the modulation of p53 ubiquitination through MDM2 and inhibiting the activation of the p53 signaling pathway. The findings of the present study highlight the novel regulatory role of CRAPIR in regulating apoptosis and the molecular mechanism underlying the pathogenesis of myocardial I/R injury.

Author contributions

HYC, YNL, BZC, XXW and YL conceived and designed the project. HY and HL drafted the manuscript and analyzed the data; YH, DHY, ZZZ, QYZ, GYZ, and ZYR helped with the study data collection; HY and YH revised this article critically for important intellectual content; WYM and HYC made a final decision on the manuscript. All the authors have read and approved the final paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-025-01534-6.