CircHIPK2 recruits SRSF1 to increase TXNIP mRNA stability and promotes autophagy-dependent ferroptosis and apoptosis in myocardial ischemia-reperfusion injury

Abstract

Background

Myocardial ischemia/reperfusion injury (MIRI) secondary to acute myocardial infarction (AMI) can lead to cardiomyocyte death and impaired cardiac function. Studies have confirmed that circular RNAs (circRNAs) play an important role in MIRI. In this study, the role and mechanism of circHIPK2 in MIRI were evaluated.

Methods

Human cardiac myocytes (HCM) were cultured under Hypoxia/Reoxygenation (H/R) condition to establish a MIRI model in vitro. Expression of circHIPK2, SRSF1 and TXNIP was assessed using RT-qPCR. Protein levels of autophagy markers (LC3II/LC3I, Beclin1, p62) and ferroptosis markers (GPX4, FTH1, ACSL4) were detected by Western blot. Cell viability and apoptosis were assessed by CCK-8 and flow cytometry. Levels of oxidative stress markers (MDA, SOD) and inflammatory factors (IL-6, IL-1β, TNF-α) were tested by ELISA assay. Iron concentration was measured with an iron detection kit. Location of circHIPK2 in cells was detected by RNA-nucleosome separation assay. RIP and ChIP assays verified the relationship between circHIPK2, SRSF1 and TXNIP. TXNIP mRNA stability was dertermined by actinomycin D. Infarct area was examined by TTC staining in myocardial ischemia/reperfusion (I/R) mouse model. HE staining evaluated myocardial injury.

Results

CircHIPK2 was increased in H/R-induced HCM cells. CircHIPK2 downregulation suppressed oxidative stress, inflammatory factors and autophagy-dependent ferroptosis in HCM cells induced by H/R. Additionally circHIPK2 recruited SRSF1 to target TXNIP and stabilized TXNIP mRNA expression. We further demonstrated that TXNIP upregulation overturned the therapeutic effects of circHIPK2 silencing on H/R model cells. In vivo, downregulation of circHIPK2 improved myocardial dysfunction caused by I/R.

Conclusion

Our results demonstrate that circHIPK2 contributes to MIRI through inducing oxidative stress and autophagy-dependent ferroptosis via SRSF1/TXNIP axis, offering new insights into MIRI treatment.

Highlights

• CircHIPK2 was increased in H/R-treated HCM cells.

• H/R activated oxidative stress and induced autophagy-dependent ferroptosis.

• CircHIPK2 silencing inhibited oxidative stress and autophagy-dependent ferroptosis.

• CircHIPK2 recruited SRSF1 to bind to TXNIP and stabilized its mRNA expression.

• Overexpressed TXNIP reversed effects of sh-circHIPK2 on H/R-treated HCM cells.

Introduction

Acute myocardial infarction (AMI) is the leading cause of mortality worldwide. Recovery of coronary blood flow as soon as possible and reperfusion therapy based on percutaneous coronary intervention are the main measures for AMI treatment [1]. However, reperfusion therapy may also lead to myocardial ischemia/reperfusion (I/R) injury (MIRI), which is characterized by myocardial metabolic disorders, cardiac local inflammatory response, myocardial cell death and cardiac remodeling and dysfunction [2]. It is reported that 40–50% of patients with AMI have myocardial necrosis due to MIRI, and there is no effective treatment to prevent MIRI. The pathophysiological processes involved in MIRI mainly include cardiomyocyte apoptosis, immune response, endothelial dysfunction and autophagy [3], but the specific molecular mechanisms of MIRI are still unclear. Therefore, it is of utmost importance to systematically understand the molecular mechanisms related to MIRI and explore methods for effective intervention against MIRI.

Circular RNA (circRNA) is a covalently closed circular non-coding RNA. Although circRNA is difficult to be translated into proteins, it can regulate the transcription and post-transcriptional modification of genes and play an important role [4]. Recent studies have found that circRNAs are involved in the ferroptosis of MIRI. For example, circFEACR participated in the regulation of cardiomyocyte ferroptosis and protected the heart function against I/R injury [5]. CircHMGA2 enhanced MIRI through promoting ferroptosis and pyroptosis [6]. CircHIPK2 (ID: hsa_circ_0001756) is a circular RNA derived from exon 2 of the HIPK2 gene. A recent report demonstrated that circPINK2 promoted apoptosis and autophagy through miR-485-5p/ATG101 pathway in hydrogen peroxide-induced oxidative damage of cardiomyocytes, indicating a new therapeutic target for myocardial injury [7]. However, the function and mechanism of circHIPK2 in MIRI are still unknown.

Ferroptosis is a cell death mode proposed in recent years, which is characterized by iron-dependent lipid peroxide accumulation leading to oxidative damage and eventually cell death [8]. Serine/arginine-rich splicing factor 1 (SRSF1), as a classic member of the alternative splicing factor SR protein family, regulates cell proliferation and apoptosis by regulating the splicing of related genes [9]. Notably, SRSF1 was reported to be involved in ferroptosis regulation in kidney I/R injury [10]. In addition, it has been reported that lncRNA HOTAIR aggravated MIRI through competitively binding SRSF1 with microRNA-126 [11]. Thioredoxin-interacting protein (TXNIP) is an ubiquitously expressed protein and plays key roles in redox homeostasis, whose overexpression leads to cellular apoptosis by oxidative stress [12]. M2 macrophage-derived exosomes play a role in protecting MIRI by carrying non-coding RNA to inhibit TXNIP expression [13]. TXNIP contributes to MIRI through exaggerating excessive autophagy during reperfusion in mice [14]. Besides, pyrimidine triazinedione can regulate TXNIP expression, and its induced glioblastoma cell death is through the interaction of autophagy and ferroptosis [15]. However, whether circHIPK2 regulates TXNIP expression by interacting with SRSF1 remains unknown, and the role of this interaction in autophagy-dependent ferroptosis in MIRI is not elucidated.

This study aimed to explore role of circHIPK2 in MIRI in vitro. We speculate that circHIPK2 may recruit SRSF1 to target TXNIP and increase its mRNA stability, thereby promoting autophagy-dependent ferroptosis, and aggravating the development of MIRI. This study indicates a novel biomarker for MIRI therapy.

Materials and methods

Cell culture and treatment

Human cardiac myocytes (HCM) were provided by Cell Bank (Shanghai, China) and cultured in DMEM (Gibco, Rockville, MD, USA) containing 10% FBS (Gibco), 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco) at 37 °C and 5% CO₂. For H/R treatment, cells were incubated in a hypoxic incubator (5% CO₂, 94% N₂, 1% O₂) for 24 h, followed by reoxygenation (5% CO₂, 74% N_2, 21% O₂) for 12 h. The control cells were cultured normally.

Cell transfection

Short hairpin RNAs (shRNAs) directed against circHIPK2 (sh-circHIPK2) with negative control (sh-NC) and TXNIP overexpressing vector (oe-TXNIP) with negative control (oe-NC) were synthesized by GenePharma (Shanghai, China). Cells were transfected with above plasmids utilizing Lipofectamine 2000 (Invitrogen, Carlsbad, CA) following manufacturer’s protocol, and collected 48 h after transfection.

Quantitative real-time PCR (RT-qPCR)

Total RNA in cells was extracted with Trizol reagent (Takara, Dalian, China), and reverse transcribed into cDNA using PrimeScript™ RT Reagent Kit (#RR600A, TaKaRa). Extracted cDNA was adopted for PCR detection utilizing DBI Bestar SybrGreen qRT-PCR Master Mix (DBI Bioscience, Shanghai, China) on 7500 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). Primers are presented as follows: circHIPK2 F: 5′-TGCTCCACCTACTTGCAGTC-3′, R: 5′-GTACCCAGTCATGTCCCAGT-3′; SRSF1 F: 5′-GCGACGGCTATGATTACGATG-3′, R: 5′-ACATACATCACCTGCTTCACGC-3′; TXNIP F: 5′-TAAGGAATGCTTGGGTGGCA-3′, R: 5′-GGTGCTGTGTCCTAAAGGGG-3′. GAPDH was used as internal reference. Data were calculated utilizing 2^−∆∆Ct method.

Western blot

Cells were lysed in RIPA buffer (Beyotime, Shanghai, China Protein was quantified using BCA assay (Beyotime). Proteins (20 μg) was transferred onto PVDF membranes after SDS–polyacrylamide gel electrophoresis (SDS-PAGE, Bio-Rad, USA). Membranes were blocked by 5% nonfat milk powder for 1 h and incubated with primary antibodies at 4 °C overnight: LC3II/LC3I (ABC929, 1:1000, Sigma, USA); Beclin1 (#3738, 1:1,000, CST, USA); p62 (ab109012, 1:2000, Abcam, USA); GPX4 (ab125066, 1:2000, Abcam); FTH1 (#3998, 1:1000, CST); ACSL4 (ab155282, 1:2000, Abcam). Subsequently, membranes were incubated with HRP labeled secondary antibody (#7074, 1:1000, CST) at 37 °C for 1 h. All bands were detected using ECL kits (Invitrogen). Protein levels were normalized to GAPDH and quantified with ImageJ software (NIH, USA).

Cell counting kit‑8 (CCK8) assay

HCM cells were seeded in 96-well plates (1 × 10³ cells/well) overnight, followed exposed to H/R. Next, 10 μl CCK-8 (Beyotime) was added to each well and incubated for 2 h at 37 °C. Optical density was determined at 450 nm wavelength on a microplate reader (Bio-Rad Laboratories, Hercules, CA, USA).

Flow cytometry

HCM cells were trypsinized, washed with PBS and resuspended in binding buffer (1 × 10⁶ cells/ml). Afterwards, cells were incubated with Annexin V-fluorescein isothiocyanate (FITC, 1.25 μl) and propidium iodide (PI, 10 μl) utilizing a cell apoptosis detection kit (BD Biosciences, San Jose, CA). After incubation in the dark for 15 min, cell apoptosis was determined by FACSCalibur™ flow cytometry (BD Biosciences). Data was quantified by FlowJo (v7.6.5, FlowJo, TreeStar, Ashland, OR, USA).

Detection of cell oxidative capacity

Levels of oxidative stress markers (MDA, SOD) in HCM cells or mouse myocardial tissues were assessed utilizing specific Enzyme-linked immunosorbent assay (ELISA) Kits (all from Nanjing JianCheng Bioengineering Institute, Nanjing, China) following manufacturer’s instructions.

DCFH-DA detection of ROS

To detect intracellular ROS, the 2′,7′‐dichlorofluorescein diacetate (DCFH-DA) probe (Sigma, St. Louis, MO, USA) was utilized. In brief, HCM cells were exposed to a 5 μM DCFH-DA solution in DMSO and incubated at 37 °C for 30 min. Subsequently, cells were washed three times with PBS and observed under a fluorescence microscope (Olympus, Tokyo, Japan).

Detection of inflammatory factors

The cells were harvested and lysed in the cell lysate on ice for 30 min, then centrifuged at 1000 × g and 4 °C for 10 min to obtain the supernatant. The concentrations of pro-inflammatory factors tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β) and interleukin-6 (IL-6) were evaluated utilizing respective ELISA kits according to manufacturer’s protocol (all from R&D Systems Inc., Minneapolis, USA). Absorbance was determined at 450 nm with a microplate reader (Bio-Rad, Hercules, CA, USA).

Detection of iron level

The cells were collected and lysed for 30 min in the cell lysate on ice, followed by centrifugation for 10 min at 1000 × g and 4 °C. The supernatant was then collected. The iron concentration in the supernatant samples was quantified utilizing an iron assay kit (ab83366, Abcam) according to the manufacturer’s instruction. The absorbance was measured with a spectrophotometer (Bio-Rad Laboratories) at 593 nm.

Subcellular fractionation

The location of circHIPK2 in cells was detected by RNA-nucleosome separation assay. Briefly, cytoplasmic and nuclear RNA isolated from HCM cells were extracted utilizing NE-PER kit (Invitrogen) according to manufacturer’s instruction. Expression of circHIPK2 was measured using RT-qPCR. U6 and GAPDH were used as controls.

RNA immunoprecipitation (RIP)

RIP experiment was conducted with a EZMagna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, Bedford, MA, USA). HEK-293T cells were lysed in RIP lysis buffer at 4 °C for 30 min. 10% of cell lysates were separated for input. Cell lysates were incubated with magnetic beads conjugated with anti-SRSF1 antibody (#42,267, 1:1000, CST) or anti-IgG (#2729, CST) at 4 °C overnight. Finally, the fold enrichment of circHIPK2 was quantified by RT-qPCR and calculated relative to the input.

Chromatin immunoprecipitation (ChIP) assay

ChIP assay was conducted with ChIP Assay Kit (Beyotime) based on manufacturer’s protocol. After H/R treatment, transfected cells were crosslinked with 1% formaldehyde. Chromatin complexes were immunoprecipitated via anti-SRSF1 (#42,267, 1:1000, CST) or IgG (#2729, CST) antibodies after sonication. Chromatin was eluted from antibody/Protein G microspheres (Roche, Mannheim, Germany). Finally, purified DNA was detected utilizing PCR with primers of TXNIP promoter.

mRNA stability assay

Transfected cells were seeded in 6-well plates and exposed to H/R condition, followed by treatment with Actinomycin D (5 μg/mL, Sigma). At 0, 2, 4, 6 h time points, TXNIP expression was assessed by RT-qPCR.

Animal experiments

A total of 20 male C57BL/6J mice (aged 8–10 weeks) were obtained from the Second Clinical College of Guangzhou University of Traditional Chinese Medicine. The mice were kept in a controlled environment, which included maintaining a constant room temperature and a 12/12-h light/dark cycle. Additionally, they were provided with a standard diet suitable for rodents. Adenovirus expressing sh-circHIPK2 and negative control (sh-NC) were designed and synthesized by GeneChem (Shanghai, China). HCM cells with adenovirus expressing either stable sh-circHIPK2 or control vector were subcutaneously injected into mice via tail veins. One week after cell injection, mice were divided into Sham group, I/R group, I/R + sh-NC group, I/R + sh-circHIPK2 group. Net, myocardial ischemia/reperfusion (MI/R) model was constructed. In brief, the mice were administered anesthesia using 2% isoflurane. To induce myocardial ischemia, a left thoracic incision was made, and the heart was temporarily exposed. A silk suture (6–0) slipknot was placed around the left anterior descending coronary artery. After 40 min of myocardial infarction, the slipknot was released, and reperfusion of the myocardium was initiated. The control group of mice underwent sham surgery, which involved performing the same surgical procedures except for tying the suture under the left anterior descending artery. The mice in the control group were kept under anesthesia using isoflurane inhalation, and after 24 h of reperfusion, they were sacrificed by severing the carotid artery. The Committee on Ethical Use of Animals approved the animal care protocols and experimental procedures.

Evans blue and tetrazolium chloride (TTC) combined staining

Twenty-four hours after reperfusion, mice were anesthetized. A in situ ligation was carried out on the left anterior descending artery. Subsequently, 1% Evans blue (Sigma) was administered via the aortic end. The cardiac tissues were rinsed with 10% potassium chloride and stored at − 20 °C for 30 min. Following this, the samples were sectioned into five 1-mm-thick slices beneath the ligation line, ensuring each slice shared the same cross-sectional area. Subsequently, the slices were immersed in a 1% TTC (Sigma) solution at 37 °C and analyzed using Image J software. The blue region represented the area not at risk (ANAR), whereas the red and white regions indicated the areas at risk (AAR). Portions unstained by TTC and Evans blue were designated as the infarcted area (n = 5).

Hematoxylin–eosin (HE) staining

The cardiac specimens were fixed with 10% paraformaldehyde, subsequently embedded in paraffin, and sliced into thin sections with a thickness of 5 μm. These thin sections were then mounted on slides for hematoxylin–eosin staining (Beyotime) and examined under an optical microscope (Olympus).

Statistical analysis

Data were presented as mean ± SD of three independent repeats. Significant differences among means were analyzed by SPSS 22.0 software (SPSS, Chicago, IL, USA). Differences between two groups were analyzed utilizing Student’s t-test, and differences among multiple groups were evaluated by one-way analysis of variance (ANOVA). P < 0.05 was considered to have statistical significance.

Results

CircHIPK2 was highly expressed in H/R-induced cells, and H/R activated oxidative stress and promoted autophagy-dependent ferroptosis

First, HCM cells were cultured under H/R condition to establish a MIRI cell model. RT-qPCR results showed that compared to control group, H/R treatment promoted circHIPK2 expression (Fig. 1A). Additionally, H/R treatment decreased cell viability but increased apoptosis in HCM cells (Fig. 1B and C). Moreover, ELISA assay showed that MDA level was significantly increased and SOD activity was inhibited in H/R-exposed cells (Fig. 1D). In the H/R model, the level of ROS in the cells and inflammatory factors (IL-6, IL-1β, TNF-α) were increased (Fig. 1E and F). Furthermore, Western blot analysis indicated that H/R treatment increased the ratio of LC3 II/I and Belin1 expression, but decreased p62 expression, suggesting that autophagy was activated (Fig. 1G). Meanwhile, protein levels of ferroptosis markers GPX4 and FTH1 were decreased, while ACSL4 protein level and iron level were increased in H/R model cells (Fig. 1H and L). Therefore, H/R triggered oxidative stress injury and induced autophagy-dependent ferroptosis in HCM cells.

Fig. 1

CircHIPK2 was highly expressed in H/R-induced cells, and H/R activated oxidative stress and promoted autophagy-dependent ferroptosis. A RT-qPCR detection of circHIPK2 in H/R-treated HCM cells. B CCK8 was used to detect cell viability. C Flow cytometry detection of cell apoptosis. D Levels of oxidative stress markers MDA and SOD were determined by ELISA. E Detection of ROS levels by immunofluorescence staining. F The changes of inflammatory factors (IL-6, IL-1β, TNF-α) were detected by ELISA. G Western blot measured protein levels of autophagy markers LC3II/LC3I, Beclin1 and p62. H Western blot assessed protein levels of ferroptosis markers GPX4, FTH1 and ACSL4. I Iron level was tested by kit. Error bars represent the mean ± SD of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

Silencing circHIPK2 inhibited oxidative stress and autophagy-dependent ferroptosis in H/R-treated cells

To explore the function of circHIPK2 in H/R-triggered cardiomyocytes, HCM cells were subjected into different groups: Control, sh-NC, sh-circHIPK2, H/R, H/R + sh-NC, H/R + sh-circHIPK2. After silencing circHIPK2 in H/R-treated cells, the expression level of circHIPK2 was significantly decreased (Fig. 2A). In H/R-exposed cells, cell viability was greatly suppressed while apoptosis was enhanced. However, circHIPK2 depletion offset these effects (Fig. 2B and C). Compared with the control groups, MDA release was enhanced in H/R-stimulated cells, and SOD activity was inhibited, while reduced circHIPK2 eliminated the effects of H/R treatment (Fig. 2D). Furthermore, the levels of inflammatory factors were decreased after silencing circHIPK2 in H/R model cells (Fig. 2E). Additionally, circHIPK2 depletion decreased LC3 II/I ratio and Belin-1 expression, but increased p62 expression in H/R-treated cells, indicating that autophagy was inhibited (Fig. 2F). After silencing circHIPK2 in H/R model cells, protein levels of ferroptosis markers GPX4 and FTH1 increased significantly, while ACSL4 and iron levels decreased greatly, suggesting that ferroptosis was also inhibited (Fig. 2G and H). Thus, circHIPK2 knockdown suppressed H/R-induced oxidative stress and autophagy-dependent ferroptosis in HCM cells.

Fig. 2

Silencing circHIPK2 inhibited oxidative stress and autophagy-dependent ferroptosis in H/R-treated cells. HCM cells were subjected into different groups: Control, sh-NC, sh-circHIPK2, H/R, H/R + sh-NC, H/R + sh-circHIPK2. A RT-qPCR detection of circHIPK2 expression. B CCK8 was used to detect cell viability. C Flow cytometry detection of cell apoptosis. D Levels of oxidative stress markers MDA and SOD were determined by ELISA. E The changes of inflammatory factors (IL-6, IL-1β, TNF-α) were detected by ELISA. F Western blot measured protein levels of autophagy markers LC3II/LC3I, Beclin1 and p62. G Western blot assessed protein levels of ferroptosis markers GPX4, FTH1 and ACSL4. H Iron level was tested by kit. Error bars represent the mean ± SD of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

CircHIPK2 recruited SRSF1 to target TXNIP and stabilized TXNIP mRNA expression

Next, bioinformatics tool StarBase predicted that there might be a binding relationship between SRSF1 and circHIPK2, and there was also a binding relationship between SRSF1 and TXNIP (Fig. 3A). RT-qPCR results showed that mRNA level of TXNIP in H/R-exposed cells was increased than that in Control group (Fig. 3B). Subcellular fractionation analysis revealed circHIPK2 was mainly located in the nucleus, not in the cytoplasm, suggesting that it may have transcriptional regulation (Fig. 3C). Next, the mRNA level of TXNIP was significantly decreased after silencing circHIPK2 in H/R model cells (Fig. 3D). Additionally, RIP assay was performed using anti-SRSF1 in HEK-293T cells. The results showed that circHIPK2 was enriched after binding to SRSF1 compared with IgG (Fig. 3E). Furthermore, the mRNA expression levels of SRSF1 and TXNIP were both decreased after silencing circHIPK2 in H/R-stimulated cells (Fig. 3F). In order to confirm whether circHIPK2 recruits SRSF1 to promote TXNIP expression, we performed ChIP experiment. The results showed that circHIPK2 silencing reduced the binding of TXNIP promoter to anti-SRSF1, indicating that circHIPK2 enhanced SRSF1 recruitment (Fig. 3G). Meanwhile, RNA stability analysis demonstrated that the half-life of TXNIP gene transcripts in H/R-treated cells with silencing circHIPK2 was shortened (Fig. 3H). These results implied that circHIPK2 recruited SRSF1 and maintain the stability of TXNIP mRNA.

Fig. 3

CircHIPK2 recruited SRSF1 to target TXNIP and stabilized TXNIP mRNA expression. A StarBase predicted the binding site of SRSF1 and circHIPK2, and the binding site of SRSF1 and TXNIP. B RT-qPCR detection of TXNIP expression. C RNA-nucleosome separation assay detected location of circHIPK2 in HCM cells. D RT-qPCR detection of TXNIP expression. E RIP assay verified the binding of circHIPK2 and SRSF1 in HEK-293T cells. F The mRNA levels of SRSF1 and TXNIP were detected by RT-qPCR. G ChIP assay verified TXNIP promoter binding to SRSF1. H Actinomycin D was used to detect TXNIP mRNA stability. Data are the means ± SD for three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

Overexpression of TXNIP reversed the therapeutic effects of sh-circHIPK2 on H/R model cells

Next, we studied the role of TXNIP in circHIPK2-mediated MIRI. We downregulated circHIPK2 and further overexpressed TXNIP in HCM cells, and then cultured under H/R condition. RT-qPCR results showed that TXNIP expression was decreased after silencing circHIPK2 in H/R-treated cells, but increased after further overexpression of TXNIP (Fig. 4A). Meanwhile, cell viability was increased and apoptosis was decreased after silencing circHIPK2 in H/R model cells, while overexpression of TXNIP partially reversed these effects (Fig. 4B and C). Moreover, sh-circHIPK2-mediated decrease in MDA expression and increase in SOD expression in H/R-exposed cells were abrogated by TXNIP upregulation (Fig. 4D). The levels of inflammatory factors were decreased after silencing circHIPK2 in H/R model cells, and the levels of inflammatory factors were increased after further overexpression of TXNIP (Fig. 4E). Additionally, circHIPK2 knockdown reduced LC3 II/I ratio and Belin1 expression, and induced p62 level in H/R-treated cells. However, these effects were overturned by TXNIP overexpression (Fig. 4F). Furthermore, overexpressed TXNIP abated the promotion of sh-circHIPK2 on GPX4 and FTH1 expression; ameliorated the inhibition of sh-circHIPK2 on ACSL4 and iron levels (Fig. 4G and H). Taken together, circHIPK2 promoted H/R-induced oxidative stress and autophagy-dependent ferroptosis via TXNIP in HCM cells.

Fig. 4

Overexpression of TXNIP reversed the therapeutic effects of sh-circHIPK2 on H/R model cells. A RT-qPCR detection of TXNIP expression. B CCK8 was used to detect cell viability. C Flow cytometry detection of cell apoptosis. D Levels of oxidative stress markers MDA and SOD were determined by ELISA. E The changes of inflammatory factors (IL-6, IL-1β, TNF-α) were detected by ELISA. F Western blot measured protein levels of autophagy markers LC3II/LC3I, Beclin1 and p62. G Western blot assessed protein levels of ferroptosis markers GPX4, FTH1 and ACSL4. H Iron level was tested by kit. Error bars represent the mean ± SD of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

Silencing circHIPK2 inhibited the damage of MIRI mouse model

To reveal the effect of circHIPK2 on MIRI in vivo, we established a MI/R mouse model. Mice were subjected to different groups: sham, I/R, I/R + sh-NC, I/R + sh-circHIPK2. RT-qPCR results showed that the expression of circHIPK2 was increased in the heart tissue of MIRI mice, while its expression level was decreased after injection of sh-circHIPK2 (Fig. 5A). Furthermore, Evans blue and TTC combined staining of mouse heart tissue indicated that knockdown of circHIPK2 reduced cardiac ischemia–reperfusion injury in I/R mice (Fig. 5B). HE staining revealed that circHIPK2 depletion reduced cardiac ischemia–reperfusion injury and inflammatory infiltration in I/R mice (Fig. 5C). After silencing circHIPK2 in MIRI mice, the MDA level was significantly decreased and the SOD activity was increased (Fig. 5D). After silencing circHIPK2 in MIRI mice, the protein ratio of LC3II/LC3I and Beclin1 protein expression were decreased, but p62 protein was increased, suggesting that autophagy was inhibited (Fig. 5E). After repressing circHIPK2 in MIRI mice, the protein levels of ferroptosis markers GPX4 and FTH1 were increased significantly, while ACSL4 protein was decreased, indicating that ferroptosis was inhibited (Fig. 5F). Finally, iron levels in heart tissue were significantly decreased after silencing circHIPK2 in MIRI mice (Fig. 5G). These results implied that circHIPK2 knockdown prevented myocardial dysfunction caused by I/R in vivo.

Fig. 5

Silencing circHIPK2 inhibited the damage of MIRI mouse model. Mice were subjected to different groups: sham, I/R, I/R + sh-NC, I/R + sh-circHIPK2. A circHIPK2 expression in mouse heart tissue was detected by RT-qPCR. B Evans blue and TTC combined staining of mouse heart tissue. C HE staining detected myocardial injury. D Levels of oxidative stress markers MDA and SOD were determined by ELISA. E Western blot measured protein levels of autophagy markers LC3II/LC3I, Beclin1 and p62. F Western blot assessed protein levels of ferroptosis markers GPX4, FTH1 and ACSL4. G Iron level was tested by kit. Error bars represent the mean ± SD. *P < 0.05, **P < 0.01. n = 5 mice/group

Discussion

MIRI is defined as a series of myocardial dysfunction and structural damage caused by the restoration of blood supply after myocardial ischemia [16]. Studies have confirmed that MIRI is involved in a variety of complex pathological processes such as inflammatory injury, oxidative stress, autophagy and ferroptosis [17]. Here, our results revealed that circHIPK2 regulated oxidative stress activation and promoted autophagy-dependent ferroptosis by targeting SRSF1/TXNIP pathway, which may provide a potential thought for MIRI treatment (Fig. 6).

Fig. 6

Molecular mechanism diagram of research

Ferroptosis is defined as iron-dependent cell death, characterized by excess reactive oxygen species generation and accumulation of lipid peroxidation [18]. It has been reported that enhanced autophagy is closely related to ferroptosis. For example, nuclear receptor coactivator 4 (NCOA4), as a selective autophagy receptor, mediates the degradation of ferritin in autophagosomes, releasing ferritin-bound iron into free iron ions, thereby promoting ferroptosis [19]. Previously, research has disclosed that ferroptosis is involved in MIRI [20]. In the present study, we confirmed the activation of oxidative stress and autophagy-dependent ferroptosis in H/R-treated HCM cells. CircRNAs are involved in the regulation of gene transcription, as well as MIRI progression [21, 22]. CircHIPK2 is a newly discovered circRNA and participates in the pathogenesis of various diseases. For instance, circHIPK2 activated lipopolysaccharide-induced astrocytic inflammation through SIGMAR1 [23]. CircHIPK2 facilitated autophagy in myocardial oxidative injury via miR-485-5p/ATG101 axis [7]. Here, we demonstrated that circHIPK2 expression was increased in H/R-exposed HCM cells. Functionally, we verified the protective role of circPINK2 knockdown in H/R-induced oxidative stress and autophagy-dependent ferroptosis in HCM cells, suggesting that circPINK2 may play an important role in MIRI. As ROS mediate lipid peroxidation and produce 4-hydroxynonenal and other related products, these products play important roles in cell death processes, including apoptosis, autophagy and ferroptosis [24]. In addition, related studies have also found that cells that trigger autophagy-dependent ferroptosis also show changes in the level of apoptosis [25]. Therefore, we also detected the level of cell apoptosis. The results showed that H/R treatment enhanced apoptosis, while circHIPK2 silencing reversed this effect. It may be that the accumulation of ROS leads to autophagy or oxidative stress, which simultaneously mediates apoptosis. Whether the autophagy-dependent ferroptosis and apoptosis mechanisms mediated by circHIPK2 have synergistic or superimposed effects is also the focus of our subsequent in-depth study.

Further bioinformatics revealed physical interaction between SRSF1and circHIPK2. SRSF1, a splicing factor and a type of RNA-binding protein, has been reported to be essential for MIRI progression. For instance, SRSF1 expression was increased in lncRNA HOTAIR-mediated MIRI regulation [11]. Additionally, lncRNA TUG1 interacted with SRSF1 to regulate ferroptosis in kidney I/R injury [10]. We further confirmed that SRSF1 was a direct target of circHIPK2 by RIP assay. Subsequently, we demonstrated that circHIPK2 could recruit SRSF1 to target TXNIP and stabilize the mRNA expression of TXNIP. Next, functional assays showed that overexpressed TXNIP reversed the sh-circHIPK2-induced inhibition of oxidative stress and autophagy-dependent ferroptosis in H/R-treated HCM cells. Recent studies have shown that TXNIP knockdown diminished NLRP3 inflammasome activation and I/R-induced injury in cardiomyocytes [26]. Formononetin ameliorated MIRI through inhibiting TXNIP-NLRP3 interaction in rats [27]. It was reported that TXNIP was associated with activation of ferroptosis in renal tubular epithelial cells [28]. Therefore, our data suggested that circHIPK2 promoted H/R-triggered oxidative stress and autophagy-dependent ferroptosis through SRSF1/TXNIP axis in HCM cells.

In conclusion, this study explored role of circHIPK2 in MIRI. We discovered that circHIPK2 was increased in H/R-exposed HCM cells. Moreover, circHIPK2 accelerated oxidative stress and autophagy-dependent ferroptosis in H/R-treated HCM cells by recruiting SRSF1 to maintain the stability of TXNIP mRNA. Therefore, circHIPK2 can be a novel target for MIRI therapy.