iPLA2β Contributes to ER Stress-Induced Apoptosis during Myocardial Ischemia/Reperfusion Injury

Both calcium-independent phospholipase A2 beta (iPLA2β) and endoplasmic reticulum (ER) stress regulate important pathophysiological processes including inflammation, calcium homeostasis and apoptosis. However, their roles in ischemic heart disease are poorly understood. Here, we show that the expression of iPLA2β is increased during myocardial ischemia/reperfusion (I/R) injury, concomitant with the induction of ER stress and the upregulation of cell death. We further show that the levels of iPLA2β in serum collected from acute myocardial infarction (AMI) patients and in samples collected from both in vivo and in vitro I/R injury models are significantly elevated. Further, iPLA2β knockout mice and siRNA mediated iPLA2β knockdown are employed to evaluate the ER stress and cell apoptosis during I/R injury. Additionally, cell surface protein biotinylation and immunofluorescence assays are used to trace and locate iPLA2β. Our data demonstrate the increase of iPLA2β augments ER stress and enhances cardiomyocyte apoptosis during I/R injury in vitro and in vivo. Inhibition of iPLA2β ameliorates ER stress and decreases cell death. Mechanistically, iPLA2β promotes ER stress and apoptosis by translocating to ER upon myocardial I/R injury. Together, our study suggests iPLA2β contributes to ER stress-induced apoptosis during myocardial I/R injury, which may serve as a potential therapeutic target against ischemic heart disease.


Introduction
Cardiovascular disease remains the leading cause of death worldwide, imposing major burdens on our healthcare and socioeconomic systems [1]. Myocardial infarction is the irreversible injury of myocardium due to prolonged ischemia caused by coronary arteries narrowing or spasm. Myocardial ischemia/reperfusion (I/R) injury refers to circumstances wherein the ischemic myocardium resumes normal perfusion, but tissue damage is worse. Function preservation of infarcted myocardium remains a challenge for physicians [2]. Although great effort has been devoted to basic and clinical research about I/R injury, pathological mechanisms remain poorly defined [3].
I/R injury is a result of multiple factors, such as oxidative stress, intracellular Ca 2+ overload, inflammation and metabolic derangements [4], most of which perturb the normal function of the endoplasmic reticulum (ER) and cause ER stress. ER is a subcellular organelle involved in the synthesis, maturation and transport of various proteins. Accumulation of misfolded protein in the ER leads to the unfolded protein response (UPR) to rescue

Primary Neonatal Rat Ventricular Myocyte (NRVM) Isolation, Culture and Transfection
Isolation and culture of NRVMs were conducted as before [18]. In brief, hearts from 1or 2-day-old Sprague Dawley rats were harvested (provided by the experimental animal center of Sir Run Run Shaw Hospital, affiliated with Zhejiang University). Cardiac ventricles were digested with 0.114% collagenase II (Biosharp, Shanghai, China) and 0.036% trypsin (Solarbio, Beijing, China) at 37 • C. We first pre-plated cells for 90 min to remove noncardiomyocytes. We then cultured cardiomyocytes in Dulbecco's modified Eagle's medium (DMEM) (Cienry, Huzhou, China) supplemented with 10% fetal bovine serum (FBS) (Gemi-niBio, Woodland, CA, USA) and 0.1 mmol/L bromodeoxyuridine (BrdU) (Sigma, St. Louis, MO, USA). BrdU was used to reduce the proliferation of non-cardiomyocytes. After 48 h, NRVMs were cultured in DMEM without serum for another 24 h, and then they were subjected to various treatments. The Animal Welfare Ethics Committee of Sir Run Run Shaw Hospital, affiliated with Zhejiang University, approved all animal experiments. For iPLA 2 β knockdown, NRVMs were transfected with control scrambled siRNA or siiPLA 2 β (Sigma, St. Louis, MO, USA) mixed with Lipofectamine RNAiMax (Invitrogen, Waltham, MA, USA), based on the manufacturer's recommendations.

Simulated Ischemia/Reperfusion (sI/R) and Cell Death Assay
To simulate ischemia/reperfusion in vitro, NRVMs were washed with 1X Phosphate Buffer Saline (PBS) two times and then cultured in freshly prepared ischemic Esumi buffer [19]. Next, cells were placed in a hypoxic chamber (StemCell Technologies, Vancouver, Canada) filled with 94% N 2 /5% CO 2 /1% O 2 for 15 min at a rate of 20 ml/min. The chamber was sealed and kept at 37 • C for an additional 3 h. Next, we replaced the ischemic Esumi buffer with DMEM. The control group was cultured in normal DMEM. Cell death was assessed by lactate dehydrogenase (LDH) assay using Cytotoxicity LDH Assay kits according to the manufacturer's recommendations (Dojindo, Kumamoto, Japan).

Generation of iPLA 2 β Knockout (KO) Mice
iPLA 2 β knockout (Pla2g6 −/− ) mice were generated by Cyagen Biosciences (Suzhou, China) with global deletion of exons 3 and 4 in the Pla2g6 gene. iPLA 2 β-KO mice were bred to the C57BL/6 background. Homozygous male iPLA 2 β-KO mice of 12 to 14 weeks old were selected as the experimental group and littermate male wild-type (WT) mice were used as the control group. Genotyping primers for the Pla2g6 null allele were: forward: 5 -GTCCACACTGGCTTGTTCCTTTAG-3 and reverse: 5 -TGTGCCGATGTCTCTGCTAGGTAG-3 . All animal experiments were approved by the Animal Welfare Ethics Committee of Sir Run Run Shaw Hospital, affiliated with Zhejiang University.

Cardiac Ischemia/Reperfusion
Male mice (12-14-week-old) mice were subjected to cardiac ischemia/reperfusion (I/R) surgery as previously reported [20]. In brief, mice were anesthetized with isoflurane and placed on a heating pad to keep their body temperature between 34 • C and 37 • C. After intubation with an 18-G tracheal tube, mice were ventilated using a MiniVent mouse ventilator (tidal volume: 260 µl/stroke; ventilation rate: 100 strokes per min). An oblique incision was made from the left sternal border to observe the fourth intercostal space. Under the exposure of the left anterior descending artery, a 6-0 silk suture was tied at a position 2-mm lower than the tip of the left auricle in a loose double knot, and the thread was kept outside the chest. After 30 min of coronary occlusion, the ligation was untied in order to reestablish flow. Sham groups were subjected to only thoracotomy and closure. After 6, 24, or 72 h, animals were sacrificed for tissue collection or 2,3,5-triphenyltetrazolium chloride (TTC) staining. Infarcted area and area at risk (AAR) were quantified, and relative ratios were calculated. For tissue harvesting, left ventricles were divided into ischemic region, boundary region, and distal region. All tissues were snap-frozen in liquid nitrogen and stored at −80 • C until use.

Echocardiography
A Vevo 2100 system (MS400C probe, VisualSonics, Toronto, ON, Canada) was used to examine cardiac contractile function with conscious, mildly restrained mice. M-mode recordings were collected and analyzed to calculate left ventricular end-diastolic dimension (LVIDd) and left ventricular end-systolic dimension (LVIDs). Ventricular fractional shortening (FS%) was calculated as (LVIDd-LVIDs)/LVIDd. All measurements were performed at the papillary muscle level.

TUNEL Staining
Frozen heart sections and cultured cardiomyocytes were stained with terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) using an In Situ Cell Death detection kit (Roche, Basel, Switzerland to detect apoptosis. Cryosections (5-µm thickness) were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. Cell nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI). Three fields were selected randomly for each sample to calculate the percentage of TUNEL-positive cells.

Collection of Human Blood Samples
Blood samples were collected from patients with acute myocardial infarction (AMI) within 24 h after percutaneous coronary intervention (PCI) at Sir Run Run Shaw Hospital. A control group included healthy people without AMI who visited the hospital for physical examination. Clinical diagnosis of AMI has followed WHO diagnostic criteria: 1. Symptoms of ischemic chest pain; 2. Dynamic evolution of ST-segment and T-wave (ST-T) characteristic in the electrocardiogram, or abnormal Q wave; 3. Serum myocardial enzyme spectrum increased and decreased. If two of these are satisfied, a diagnosis is made. The myocardial enzyme spectrum of patients with chest pain was measured immediately after admission, and serum cardiac troponin T (cTnT) level was much higher than 0.5ug/L, Creatine Kinase MB Isoenzyme (CK-MB) was much higher than 25IU/L. The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Medical Ethics Committee of Sir Run Run Shaw Hospital, affiliated with Zhejiang University (20200210-113).

ELISA Assay
Blood samples clotted for 2 h at ambient temperature and were centrifuged for 15 min at 12,000 rpm 4 • C. Serum was collected and frozen at −20 • C until use. The culture medium of NRVMs was centrifuged to remove cell debris and then stored at −20 • C until use. ELISA kits for lysophosphatidylcholine (LPC) were purchased from CLOUD-CLONE CORP. (CCC, USA). LPCs in the serum of AMI patients and the culture medium of the NRVMs were then determined. ELISA kits for iPLA 2 β were purchased from GIVE (Shanghai, China) and serum iPLA 2 β concentration was evaluated according to the manufacturer's recommendations. Concentrations were calculated through standard curves using a specific curve fitting tool CurveExert 1.4. The study of human blood samples was performed according to the principles outlined in the Declaration of Helsinki. The research protocol was approved by the Medical Ethics Committee of Sir Run Run Shaw Hospital affiliated with Zhejiang University.

Flow Cytometry
For quantification of cell death, NRVMs were digested with 0.05% trypsin (without EDTA) after washing with cold PBS. Cells were concentrated and suspended in 1mL PBS after neutralizing with fetal bovine serum. Annexin V-FITC and propidine iodide (PI) staining solution (Yeasen, Shanghai, China) were added, and suspensions were evaluated using flow cytometry.

Cell Surface Protein Biotinylation and Detection
NRVMs were first subjected to sI/R. Then, cells were rinsed twice with cold PBS. EZ-link ® Sulfo-NHS-LC-Biotin (Thermo Fisher Scientific, Waltham, MA, USA) at a concentration of 0.5 mg/mL was added. As a control, the same volume of PBS was used. Plates were gently shaken (30 min at 4 • C). Tris-Cl, pH 7.5 at 100 nM was used to stop biotinylation reaction. After rinsing twice with cold PBS, the cells were lysed with radioimmune precipitation buffer (RIPA). To purify surface proteins, neutravidin agarose beads were mixed with the lysate (overnight at 4 • C). The beads were then washed six times with PBS. Cell surface proteins were released by mixing with SDS-PAGE loading buffer and heating (100 • C for 5 min). Immunoblotting was then conducted to detect iPLA 2 β biotinylation. ATP1A (ATPase Na + /K + transporting subunit alpha 1) was used as a cell membrane protein marker and loading control.

Assay for [Ca 2+ ]i Levels
For the analysis of [Ca 2+ ]i levels, cells were loaded with 5 µmol/L Fluo-4AM (Yeasen, China) at 37 • C for 30 min. The cells were digested and collected, [Ca 2+ ]i levels were detected by flow cytometry.

Statistics
Data were expressed as mean ± SEM and Student's t-test was used to calculate statistical difference between two groups. Differences among multiple groups were analyzed by one-way ANOVA, followed by Tukey's test. p < 0.05 was considered statistically significant.

Myocardial Ischemia/Reperfusion (I/R) Injury Induces iPLA 2 β In Vivo
Previous studies have shown that myocardial I/R injury increases the production of lysophosphatidylcholine (LPC) [22][23][24]. To determine whether LPC is stimulated by myocardial I/R injury in humans, we collected serum samples from 12 patients for ELISA assay. The basic demographic and clinical characteristics of the AMI patients and healthy donors are shown in Table 1. ELISA assay for LPC showed that the serum LPC level was higher in the AMI patients than the non-AMI controls (Figure 1a). Next, blood samples from 12-14-week-old mice subjected to myocardial ischemia for 30 min, followed by reperfusion for 6, 24, 72 h were collected for ELISA assay. Our results showed that LPC level was increased in mice after 24 h of reperfusion compared to the sham group (Figure 1b). LPC is generated by partial hydrolysis of phosphatidylcholines, which removes one of the fatty acid groups. This hydrolysis is generally the result of the enzymatic action of phospholipase A2 (PLA 2 ). Previous studies have shown that the subtypes of PLA 2 , including cPLA 2 α, sPLA 2 -V, Lp-PLA 2 , iPLA 2 β and iPLA 2 γ, are associated with myocardial I/R injury [25][26][27][28][29][30]. Here, we first examined the correlation between PLA 2 enzymes and I/R. We found that the mRNA level of iPLA 2 β was elevated significantly after 24 h of reperfusion, which was also consistent with the increase of LPC concentrations (Figure 1c). In contrast, the increase of other PLA 2 subtypes peaked at 6 h instead of 24 h after I/R. Next, we asked whether iPLA 2 β was also increased at the protein level upon I/R. Indeed, our iPLA 2 β ELISA data showed that serum levels of iPLA 2 β in the AMI patients were significantly higher than those in the non-AMI patients ( Figure 1d). Consistently, the protein level of iPLA 2 β in mouse hearts underwent I/R was also dramatically increased ( Figure 1e). Taken together, these data indicate that myocardial I/R injury induces iPLA 2 β expression in the heart. patients. Twelve serum samples were collected, six of which were in the AMI group. LPC ELISA was performed to detect the LPC concentration in serum. (b) Upregulation of LPC in mouse sera after I/R injury. Mice at 12-14 weeks of age were subjected to myocardial ischemia for 30 min, followed by reperfusion for 6, 24, 72 h. Serum samples were collected, and LPC ELISA was performed to detect the LPC concentration. n = 6 per group. (c) The mRNA level of phospholipase A2 in mouse hearts upon myocardial I/R was detected. Quantitative RT PCR (qRT-PCR) was conducted to measure mRNA level changes, with 18S as an internal control. n = 5 per group. (d) iPLA 2 β was upregulated in the sera of the AMI patients. iPLA 2 β ELISA was performed to detect the iPLA 2 β concentration. (e) The expression of iPLA 2 β in mouse hearts was increased after I/R. Mice were subjected to myocardial ischemia for 30 min, followed by 24 h of reperfusion. Immunoblot analysis was performed for iPLA 2 β. β-actin was used as a loading control. n = 5 per group. Data are represented as mean ± SEM. All experiments were independently replicated in triplicate. * p < 0.05.

sI/R Induced iPLA 2 β In Vitro
We treated NRVMs using the Esumi ischemic buffer, and incubated the cells in a 94% N 2 /5% CO 2 /1% O 2 hypoxic chamber for 3 h. Then, the buffer was replaced with a normal medium. We next determined cardiomyocyte death by assessing LDH release. We found that sI/R induced a time-dependent increase of cardiomyocyte death (Figure 2a). We found that, consistent with the elevated LDH release, NRVMs subjected to simulated 3-h ischemia/3-h reperfusion (sI/R) showed a significant increase of iPLA 2 β expression at both mRNA and protein levels (Figure 2b,c). In contrast, the mRNA levels of other types of PLA 2 , including cPLA 2 α, sPLA 2 -V, Lp-PLA 2 and iPLA 2 γ, had no significant changes ( Figure 2b). Besides, LPC concentrations were also upregulated in both supernatants and intracellular extracts (Figure 2d).Following sI/R, we observed that GRP78, an ER stress marker, was increased at both mRNA and protein levels (Supplementary Figure S1a,b). Moreover, NRVMs subjected to sI/R showed an increase of the spliced form of XBP1 mRNA (Supplementary Figure S1c). In summary, these findings suggest that sI/R leads to upregulation of iPLA 2 β and induction of the UPR. Simulated ischemia/reperfusion (sI/R) induces iPLA 2 β upregulation in neonatal rat ventricular myocytes (NRVMs). (a) sI/R caused significant cell death. NRVMs were subjected to 3 h of ischemia followed by 3, 6, 16 h of reperfusion. Lactate dehydrogenase (LDH) was determined in culture medium and in cell lysates, respectively. Relative cell death was calculated based on the ratio of released LDH in the medium. n = 6 per group. (b) sI/R led to strong induction of phospholipase A2 at the mRNA level. NRVMs were subjected to ischemia for 3 h, followed by different times of reperfusion. qRT-PCR was conducted to measure mRNA changes with 18S as an internal control. n = 6 per group. (c) sI/R induced the expression of iPLA 2 β at the protein level. Immunoblotting showed that iPLA 2 β was significantly increased at the protein level (left). β-actin was used as a loading control. Quantification is shown on the right. n = 5 per group. (d) Lysophosphatidylcholine (LPC) was increased during sI/R. NRVMs were subjected to 3 h of ischemia and 3 h of reperfusion. LPC ELISAs were performed to detect LPC concentrations in culture medium and in cell lysates. Relative LPC concentration in cell lysates was normalized based on total cellular protein concentrations. n = 5 per group. Data are represented as mean ± SEM. All experiments were independently replicated in triplicate. * p < 0.05.

Knockout of iPLA 2 β Alleviated Heart I/R Injury
To explore the role of iPLA 2 β in vivo, we generated a Pla2g6 (the gene encoding iPLA 2 β) knockout (KO) mouse model (C57BL/6N) by CRISPR/Cas9-mediated genome engineering (Figure 3a). Positive mice were identified by PCR screening, and the size of the PCR products of Pla2g6 −/− mice was 650 bp, whereas PCR products for the wild-type (WT, Pla2g6 +/+ ) mice were 573 bp. Mice designated #1-8 were homozygous Pla2g6 −/− mice (Figure 3b), and mice designated #9-13 were homozygous Pla2g6 +/+ mice. To confirm the knockout of Pla2g6, we performed qRT-PCR and Western blotting. iPLA 2 β-KO mice showed significantly lower iPLA 2 β expression at both mRNA level and protein levels (Figure 3c,d). Furthermore, WT and Pla2g6 −/− mice had similar body weights (Supplementary Figure S2a), heart rates (Supplementary Figure S2b), ventricular sizes and contractile functions (Supplementary Figure S2c) at baseline. Next, 30-min of myocardial ischemia and 24-h of reperfusion were performed with 12-14-week-old WT control mice and KO mice. We used TTC staining to identify and quantify damaged tissue areas (Figure 3e). The relative ratios of infarct area to area at risk (AAR) of the WT and KO hearts were calculated and compared. iPLA 2 β deletion led to a significantly smaller infarct size (Figure 3f). In addition, there was no significant difference in the relative ratio of AAR to the left ventricle between control group and KO group, suggesting that I/R surgery was similarly conducted between groups. Moreover, echocardiographic analysis showed iPLA 2 β deficiency resulted in improved systolic and diastolic function 24 h after I/R surgery (Table 2), which indicates that knockout of iPLA 2 β preserves cardiac performance in respond to I/R injury. Furthermore, heart sections stained with TUNEL showed less myocardial apoptosis in iPLA 2 β KO mice than control littermates (Figure 3g). Taken together, our data indicate that knockout of iPLA 2 β protects the heart against I/R injury.

Inhibition of iPLA 2 β Ameliorates NRVMs Death Triggered by sI/R In Vitro
To investigate the role of iPLA 2 β in I/R in vitro, we adopted an inhibitor-based approach. First, after 3 h of ischemia, we inhibited iPLA 2 β in NRVMs with 15 µM bromoenol lactone (BEL) for 15 min, which is a potent and irreversible inhibitor of iPLA 2 β [31]. DMSO was used as a negative control. Our results showed that inhibition of iPLA 2 β significantly reduced cell death caused by sI/R (Figure 4a). LPC ELISAs verified that BEL pretreatment reduced sI/R induced supernatant and intracellular LPC release, which was mainly due to the reduced activity of iPLA 2 β (Figure 4b). Additionally, the decreased number of TUNELpositive cells in the BEL-treated group confirmed that the inhibition of iPLA 2 β protected NRVMs from the sI/R-induced cell death (Figure 4c,d). Recent studies have shown that BEL inhibits other key enzymes in phospholipid metabolism as well [31]. To further test the function of iPLA 2 β in sI/R injury directly, we transfected NRVMs with siiPLA 2 β to knock down iPLA 2 β. SiiPLA 2 β efficiently decreased target gene expression at both protein and mRNA levels (Figure 4e,f). Knockdown of iPLA 2 β modestly but significantly decreased cell death, as shown by an LDH release assay and the LPC concentration measurement (Figure 4g,h). Furthermore, knockdown of iPLA 2 β mitigated NRVM apoptosis caused by sI/R as shown by TUNEL assays (Figure 4i). Collectively, these results demonstrate that inhibition of iPLA 2 β protects cardiomyocytes from sI/R injury in vitro.  Total RNA was isolated from ventricular tissues of WT and iPLA 2 β KO mice, respectively. Real-time PCR was conducted to assess iPLA 2 β at the mRNA level. n = 4-5 for each group. (d) Confirmation of iPLA 2 β deletion in iPLA 2 β-KO mice. Total proteins were extracted from the ventricular tissues of wild-type (WT) and iPLA 2 β-KO mice. Western blotting was conducted to assess iPLA 2 β at protein level. n = 5 for each group. (e,f) Knockout of iPLA 2 β protected the heart from I/R injury. WT and iPLA 2 β-KO mice at 12-14 weeks of age were subjected to myocardial ischemia for 30 min, followed by 24 h of reperfusion. Then, 2,3,5-triphenyltetrazolium chloride (TTC) staining was conducted to detect and quantify the injured cardiac tissue. The white region depicts the ischemic infarct zone; the pink/red area is the border zone; and the blue region represents the remote zone. The relative ratio of the infarct region (IF) to the area at risk (AAR, ischemic and border) was compared between WT and iPLA 2 β-KO mice. The relative ratio of AAR to the left ventricle did not differ between WT and KO groups, suggesting that I/R surgery was performed similarly across genotypes. n = 4 per group. (g) Representative photographs of heart sections stained with terminal-deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) from WT and KO mice after myocardial I/R (left). Scale bar = 25 µm. Quantification is shown on the right. n = 10-11 per group. Data are represented as mean ± SEM. All experiments were independently replicated in triplicate. * p < 0.05.

The Reduction of I/R Injury Is Mainly Achieved by Reducing Apoptosis
Previous work suggests that apoptosis is involved in myocardial I/R injury [32]. Since prolonged ER stress leads to apoptosis, we hypothesized that the inhibition of iPLA 2 β might alleviate I/R injury caused-myocardial damage by reducing apoptosis. Indeed, flow cytometric analysis showed that the inhibition of iPLA 2 β protected NRVMs from sI/R injury, mainly by decreasing Annexin V+/PI+-positive apoptotic cells (Figure 5a,b). We observed that the expression of CHOP was markedly reduced in the BEL-pretreated and siiPLA 2 β-transfected NRVMs under the sI/R condition, respectively (Figure 5c,d). Moreover, the expression of CHOP in iPLA 2 β-KO mouse hearts after I/R injury was downregulated compared to that in WT mouse hearts (Figure 5e). Additionally, there was no significant difference of CHOP expression at baseline between WT and iPLA 2 β-KO mice (Supplementary Figure S3a).

iPLA 2 β Translocation to the ER Aggravates ER Stress-Induced Apoptosis during I/R
The function of iPLA 2 β is influenced by its organelle localization, which is regulated by the intracellular environment [16]. To determine the localization of iPLA 2 β in cardiomyocyte upon I/R injury, a biotinylation experiment for cell surface proteins was conducted. sI/R did not increase the cell membrane localization of iPLA 2 β, which actually showed a trend of decrease (Figure 6a). Immunofluorescence was next used to investigate whether iPLA 2 β was present on the ER membrane. We found that colocalization of iPLA 2 β and an ER marker calreticulin was significantly increased during sI/R. In contrast, iPLA 2 β inhibition significantly disrupted iPLA 2 β and ER colocalization (Figure 6b,c). The upregulation of ER stress markers, apoptotic markers and Ca 2+ regulated related proteins induced by sI/R was decreased by the inhibition of iPLA 2 β in vitro and in vivo (Figure 6d,e). Quantification of these blots were shown (Supplementary Figure S4a,b). However, significant differences were not observed in the other two UPR marker proteins, ATF6 and IRE1α (Supplementary Figure S5a,b). When NRVMs were subjected to 0.5 µM thapsigargin (TG), an ER stress activator, for 6 h, we also found that colocalization of iPLA 2 β with ER was significantly increased under ER stress condition (Supplementary Figure S5d). Inhibition of iPLA 2 β reduced intracellular Ca 2+ during sI/R (Figure 6f,g). Moreover, inhibition of ER stress by 5 mM 4-phenylbutyric acid (4-PBA) for 30 min ameliorated cardiomyocyte death caused by sI/R (Supplementary Figure S5c). Collectively, these findings suggest that I/R induces iPLA 2 β expression and promotes its translocation to the ER, which leads to unbalanced lipid metabolism homeostasis, aggravates calcium overload, triggers ER stress and eventually causes cardiomyocyte apoptosis.  . I/R promotes iPLA 2 β translocate to the ER, leading to cardiomyocyte apoptosis induced by ER stress. (a) sI/R caused no significant increase in iPLA 2 β on the cell surface membrane. NRVMs were subjected to sI/R and then placed on ice. EZ-link Sulfo-NHS-LC-Biotin was used to modify cell surface-localized proteins. After isolation of biotinylated surface proteins with neutravidin, immunoblotting was conducted to detect iPLA 2 β. ATP1A was used as loading control of cell-surface-localized proteins. Quantification (below) revealed no increase in cardiomyocyte surface-localized iPLA 2 β after sI/R. n = 3-4 for each group. (b) iPLA 2 β was localized in the ER. iPLA 2 β was detected by confocal immunofluorescence staining (green). ER was revealed by confocal immunofluorescence staining for calreticulin (red). The areas of interest are labeled with white arrows. The signaling intensity for both channels was scanned and recorded. Results are shown on the right. Scale bar = 20 µm. (c) Colocalization of iPLA 2 β and ER was quantified by Mander's colocalization coefficients (MCCs). The comparison suggests that the inhibition of iPLA 2 β significantly disrupts iPLA 2 β and ER colocalization. n = 8-12 per group. (d) Knockdown of iPLA 2 β relieved ER stress caused by I/R. Immunoblotting analysis was performed to detect the ER stress markers and Ca 2+ -regulated related proteins. β-actin was used as a loading control. The protein levels of calnexin, ATF4, p-PERK, p-eIF2a, CHOP, cleaved-caspase3, Bax, p-CaMKII and SERCA2 were downregulated in the siiPLA 2 β -transfected NRVMs. (e) Knockout of iPLA 2 β relieved the ER stress caused by I/R. Immunoblotting analysis of ER stress markers and Ca 2+ -regulated related proteins were performed. β-actin was used as a loading control. The protein levels of calnexin, ATF4, p-PERK, p-eIF2a, CHOP, p-CaMKII and SERCA2 were downregulated in the Pla2g6 KO mouse heart tissues. (f) Inhibition of iPLA 2 β by BEL reduced calcium overload upon sI/R. Intracellular Ca 2+ was measured using 5 µM Fluo-4 AM. Mean fluorescence intensity indicates the concentration of intracellular Ca 2+ . Quantification is shown on the right. n = 3 per group. (g) Knockdown of iPLA 2 β reduced calcium overload during sI/R. Intracellular Ca 2+ was measured using 5µM Fluo-4 AM. Mean fluorescence intensity indicates the concentration of intracellular Ca 2+ . Quantification is shown on the right. n = 4-7 per group. Data are represented as the mean ± SEM. All experiments were independently replicated in triplicate. * p < 0.05.

Discussion
Coronary heart disease is caused by atherosclerotic lesions in the coronary arteries that narrow or block the lumen of the vessels, resulting in myocardial ischemia, hypoxia, or necrosis. Reports from the World Health Organization's website in 2017 show that cardiovascular disease (CVD) is the primary cause of death globally: more people die annually from CVD than from any other causes. Of these deaths, 85% are due to heart attack and stroke, with ischemic heart disease as the predominating one. The current development of PCI technology makes it possible to reduce the myocardial infarction area effectively and improve clinical outcomes. However, the mortality rate of patients with myocardial infarction still remains high. Studies have shown that even the restoration of blood flow to the ischemic myocardium cannot significantly reduce the infarct area, instead, further damage has been observed. This phenomenon is referred to as cardiac ischemia/reperfusion (I/R) injury. The pathological mechanism of cardiac I/R injury includes the accumulation of oxygen free radicals [33], calcium iron overload [34], energy metabolism failure [35], cell apoptosis [36] and other pathological factors. Cardiac I/R injury is a complex and multi-factorial pathological process that needs further investigations.
ER stress has been proven to be critical for the progression of cardiac diseases, including I/R [37][38][39]. In our study, we revealed the important effects of iPLA 2 β on ER stress during myocardial I/R injury. The levels of iPLA 2 β and ER stress markers were induced by sI/R injury, with a concomitant increase of apoptotic markers. Besides, inhibition of iPLA 2 β in cardiomyocytes decreased cardiomyocyte death in response to sI/R injury via the ATF4/CHOP pathway. These results indicate that iPLA 2 β increased upon I/R injury caused ER stress and further induced apoptosis. Moreover, iPLA 2 β-deficient mice showed alleviated myocardial I/R injury. We also found that iPLA 2 β participated in ER stress-induced apoptosis by translocating to the ER (Figure 7). Based on these, we propose that iPLA 2 β is critical for the control of ER stress and cardiac function by its dynamic cellular localization. iPLA 2 β, which is also referred to as PLA 2 group VI and encoded by PLA2G6, contributes to the regulation of the concentration of phosphatidylcholine, a metabolite abundant in the cell membrane. iPLA 2 β is also involved in multiple processes of phospholipid remodeling, arachidonic acid release [40], Fas receptor-mediated apoptosis [41,42], and transmembrane ion flux [43]. Mutations in PLA2G6 have been proven to lead to mitochondrial dysfunction [44] and related neurological diseases [45][46][47][48], including Parkinson's disease 14 [49] and hereditary spastic paraplegia [50]. Intracellular localization of iPLA 2 β is dynamic and tissue-specific [51,52]. Previous studies have shown that proinflammatory cytokines induce iPLA 2 β upregulation-caused cell apoptosis by stimulation of ER stress in β-cells, an effect blocked by iPLA 2 β inhibitor bromoenol lactone (BEL) [53]. iPLA 2 β and cardiovascular function are closely related. iPLA 2 β activation in the heart may result in both positive and negative effects [14,54,55]. The inhibition of iPLA 2 signaling pathway blocks the negative inotropic effect of HIV gp120 on cardiac myocytes [56], and activation of iPLA 2 β has been reported to result in reduced conduction of arrhythmias caused by diabetic heart ischemia [57]. iPLA 2 β also regulates calcium flow in various cells [58][59][60][61]. For example, iPLA 2 and its metabolite LPC play a key role in the control of endothelial store-operated Ca 2+ entry and arterial tone [62].
It has been reported that the UPR is closely associated with lipid metabolism imbalance in the ER [63]. The UPR has been shown to be activated by changes in different lipids, which is independent of protein folding homeostasis in the ER lumen [64]. In this study, we established the heart I/R injury model in vivo and in vitro and observed that the activation of ER stress was accompanied by the upregulation of iPLA 2 β. In view of the metabolic function of membrane phospholipids, we hypothesized that changes of iPLA 2 β in the ER would lead to the disruption of ER function and thus cause ER stress. We confirmed that the increase of iPLA 2 β in the ER lumen induced ER stress, caused calcium overload, and ultimately led to apoptosis. However, we did not determine the mechanism of iPLA 2 β's translocation into ER under I/R, nor did we identify the ER protein that acts directly with it. Further studies are required to answer these key questions. iPLA 2 β is intensively studied in the field of the nervous system [65], and it has been reported that iPLA 2 β's pathogenic causes are related to the structural and functional changes of the ER [66] and mitochondria [67]. Scoot et al. showed that iPLA 2 β localizes in the mitochondria and affects the phospholipid metabolism of the mitochondrial membrane during myocardial I/R injury, which leads to the impairment of mitochondrial structure and function and consequently causes cardiomyocyte death [68]. Although many studies have confirmed the relationship between iPLA 2 β and ER stress, our study contributes to the understanding of the connection between iPLA 2 β and myocardial I/R injury. In addition, studies have reported that iPLA 2 β is essential for maintaining basal ER homeostasis, suggesting that this enzyme performs different functions under pathological and physiological conditions. Illustration of the role of iPLA 2 β on ER stress in cardiomyocyte upon I//R. PLA 2 expression is up-regulated in cardiomyocyte after I/R, which consequently leads to an increase of phosphatidylcholine (PC) metabolism. The generation of LPC is elevated and cardiomyocyte damage is augmented. The protein level of intracellular iPLA 2 β, one of the PLA 2 enzymes, is upregulated. The increased iPLA 2 β translocates to the ER and induces ER stress, which further leads to apoptosis through the PERK/ATF4/CHOP pathway. Additionally, iPLA 2 β may also increase intracellular calcium flow in cardiomyocyte.
In summary, our study identified a key role of iPLA 2 β in ER stress-induced apoptosis and in intracellular calcium homeostasis upon cardiac I/R. Inhibition of iPLA 2 β reduces myocardial I/R injury and preserves cardiac function. Therefore, inactivation of iPLA 2 β may serve as a potential therapeutic strategy for myocardial ischemia/reperfusion injury.

Limitation
We detected LPC concentration in the sera of patients with acute myocardial infarction by ELISA. Although previous reports have shown no significant difference in the detection results between ELISA and HPLC methods [69,70], it is undeniable that HPLC is more accurate, stable and specific than ELISA [71]. LPCs are a class of metabolites derived from phosphatidylcholines, which are composed of a variety of species [72]. Due to the unavailability of HPLC, we could not detect and analyze the various LPC components.
The neonatal cardiomyocytes were chosen for in vitro experiments in our study. There are some limitations in using neonatal cardiomyocytes as a model system. First, the phenotypic immaturity of neonatal cardiomyocytes limits the full translation of the findings to the in vivo conditions, for example, metabolic characteristics of neonatal cells are different from those in adult cardiomyocytes [73]. Secondly, the lack of extracellular matrix in the culture system may alter the behavior of these cells, resulting in a certain discrepancy between in vitro and in vivo findings [74].