Cardioprotective mechanism of SGLT2 inhibitor against myocardial infarction is through reduction of autosis

Sodium-glucose cotransporter 2 (SGLT2) inhibitors reduce cardiovascular mortality in patients with diabetes mellitus but the protective mechanism remains elusive. Here we demonstrated that the SGLT2 inhibitor, Empagliflozin (EMPA), suppresses cardiomyocytes autosis (autophagic cell death) to confer cardioprotective effects. Using myocardial infarction (MI) mouse models with and without diabetes mellitus, EMPA treatment significantly reduced infarct size, and myocardial fibrosis, thereby leading to improved cardiac function and survival. In the context of ischemia and nutritional glucose deprivation where autosis is already highly stimulated, EMPA directly inhibits the activity of the Na+/H+ exchanger 1 (NHE1) in the cardiomyocytes to regulate excessive autophagy. Knockdown of NHE1 significantly rescued glucose deprivation-induced autosis. In contrast, overexpression of NHE1 aggravated the cardiomyocytes death in response to starvation, which was effectively rescued by EMPA treatment. Furthermore, in vitro and in vivo analysis of NHE1 and Beclin 1 knockout mice validated that EMPA’s cardioprotective effects are at least in part through downregulation of autophagic flux. These findings provide new insights for drug development, specifically targeting NHE1 and autosis for ventricular remodeling and heart failure after MI in both diabetic and non-diabetic patients. Electronic supplementary material The online version of this article (10.1007/s13238-020-00809-4) contains supplementary material, which is available to authorized users


Generation of Nhe1 KO mice.
NHE1 KO mice (Nhe1 -/-) were generated using CRISPR/Cas9 system and were performed by Tongji University and BIOCYTOGEN. The brief process is as follows: sgRNA was transcribed in vitro. Cas9 mRNA and sgRNAs were microinjected into fertilized embryos of C57BL/6N mice. Fertilized eggs were transplanted to obtain positive F0 mice which were confirmed by PCR and sequencing. A stable F1 generation mouse model was obtained by mating positive F0 generation mice with C57BL/6N mice. Chimeric mice were generated and bred with C57BL6/N females and germline transmission was identified by PCR analysis. PCR genotyping analysis of Nhe1 KO mice was carried out using genomic DNA from tail biopsies and the following primers ( Figure  6J

In vivo myocardial I/R and TTC staining
WT were anesthetized by spontaneous inhalation and maintained under general anesthesia with 1-2% isoflurane and subjected to left coronary artery occlusion for 30 min, followed by 3 h /24 h of reperfusion. Hearts were then excised and stained with Evans blue and triphenyl tetrazolium chloride to measure the ischemic area at risk and the area of necrosis, respectively.

In vivo myocardial infarction and Masson staining
Mice were anesthetized by spontaneous inhalation and maintained under general anesthesia with 1-2% isoflurane. Animals were mechanically ventilated using a rodent ventilator (Harvard Apparatus) connected to an endotracheal tube. The heart was exposed by a left side limited thoracotomy and the LAD was ligated with a 6-0 polyester suture 1 mm from the apex of the normally positioned left auricle and samples were analyzed 7 days after injury. Tissue from the left ventricle fixed in 4% paraformaldehyde were embedded in opti-mum cutting temperature compound (OCT) and cut into 10μm thick sections. The sections were stained separately with Masson's trichrome. The morphology of the cardiomyocytes and the deposition of collagen were observed by microscope.

Echocardiography
Echocardiography was conducted on patients using an iE33 Doppler ultrasonography system (Philips) in this study. The images were obtained with the subjects at rest and lying in the lateral decubitus position. Standard echocardiographic studies, including left ventricular maximal wall thickness, left atrial and left ventricular dimensions at end-systole and end-diastole, septal thickness and posterior wall thickness in end-diastole, were performed according to previously published methods. The heart was imaged in the two-dimensional parasternal short-axis view, and an M-mode echocardiogram of the midventricle was recorded at the level of the papillary muscles.
Heart rate, intraventricular septum and posterior wall thickness, and end-diastolic and end-systolic internal dimensions of the left ventricle were obtained from the M-mode image.

Cell culture and Isolation of neonatal rat cardiomyocytes
H9c2 rat cardiomyoblast cells were obtained from ATCC and cultured in high glucose (4,500 mg/l) Dulbecco's modified Eagle's medium (DMEM) supplemented with 10%FBS. Cardiomyocytes were isolated from neonatal Wistar rats (1 to 2 days). In brief, after dissection, hearts were washed and minced in phosphate buffer saline.
Tissues were then dispersed in a series of incubations at 37 ℃ in 1.2 mg/ml pancreatin and 0.14 mg/ml collagenase. Subsequent supernatants were collected and centrifuged at 200 g for 5 min. After centrifugation, cells were resuspended in Dulbecco's modified Eagle medium/F-12 (11330032, Gibco) containing 5% heating activated horse serum, 0.1 mmol/l ascorbate, insulin-transferring-sodium selenite media supplement, 100U/ml penicillin, 100 mg/ml streptomycin, and 0.1 mmol/l bromodeoxyuridine. The dissociated cells were pre-plated at 37 ℃ for 1 h. The cells were then diluted to 1x 10 6 cells/ml and plated in 10 mg/ml laminin-coated different culture dishes according to the specific experimental requirements.

Membrane Transporter/Ion Channel Compound library screen
A small-molecule compound library with 387 candidates (HY-1011 ， HY-LD-000001445，MCE), targeting membrane transporters and ion channels, was used for pharmacological screening with cardiomyocytes protection of anti-glucose deprivation (GD). Cultured cardiomyocytes were incubated with each of candidate compound at concentration of 10μM or 100μM prior to GD for 24 hours, viability of cardiomyocytes were determined at 450nm using the Cell Counting Kit-8 (MCE).
Duplicate treatment were averaged and normalized to control from the same plate.
Normalized values were then processed by calculating the relative fold change (compound/control) for each compound at the dosage. We then obtained the compounds that were more than 1.5 fold different for further analysis according to their target specificity.

Contraction Force Measurements
Neonatal rat cardiomyocytes were seeded on confocal dishes 2-3 days before and measured in their normal culture medium with or without EMPA at 2, 4, 12, 24h in an environment at 37°C. Video-based motion edge detection system was used to assess cellular contractility of contracting CMs. Briefly, CMs were visualized using a Zeiss CFM-500 inversion fluorescence microscope coupled to FelixGX video microscopy system software (PTI). Spontaneous contraction traces were recorded and the resting cell length and the peak cell shortening amplitude of each cell were determined using commercially available data analysis software (FelixGX, PTI). Only isolated cardiomyocytes with appropriate morphology and function were used, and non-shortening cells were excluded from analysis.

Western blotting
The cardiac cells (H9c2 and neonatal rat cardiomyocytes) and mouse heart tissue were

Immunofluorescence staining
Cells and cardiac tissue sections were fixed with 4% paraformaldehyde, permeabilized with 0.05% Triton X-100, blocked with 1% BSA, and incubated with appropriate primary antibodies for 24h at 4°C. The samples were then incubated with Alexa Fluor 594 or 488 conjugated secondary antibodies (Invitrogen) at 37°C for 1-2h and subsequently counterstained with DAPI. Neonatal rat cardiomyocytes were validated by immunostaining of α-actinin (Abcam). For characterization of mouse tissue sections, immunofluorescent staining were performed using antibodies for sarcomeric α-actinin (Abcam), ki67 (Abcam) and LC3 (Abcam)). Labeled cells or cardiac tissue sections were examined and imaged with a fluorescence microscope (Leica). Cell surface area size was quantified using the ImageJ software package.

Quantitative RT-PCR
Total RNA was isolated from the left ventricle or cultured cardiomyocytes for analysis using the miRNeasy Mini Kit (1038703, QIAGEN). The mRNA levels were determined by quantitative RT-PCR. For reverse transcription and amplification, we used iScript Reverse Transcription Supermix for RT-qPCR (1708841, BIO-RAD) and the SsoFast EvaGreen Supermix (1725201, BIO-RAD), respectively. The PCR primers were obtained from Sangon (Table S3). We constructed quantitative PCR standard curves using the corresponding complementary DNA, and all data were normalized to Gapdh mRNA content.

HPLC analysis
HPLC analysis was performed with different types of sample. HPLC was performed on an Agilent 1260 (Agilent) equipped with a Waters C18 (5 µm, 4.6 * 250 mm) (Waters). Elution of EMPA was performed with 35% acetonitrile in water for 20 min at a flow rate of 1 ml/min. Samples were analyzed at specific absorption wavelength of 230 nm. Elution of tamoxifen was performed with 1% triethylamine (PH 8.5) in water for 20 min at a flow rate of 1 ml/min. Samples were analyzed at specific absorption wavelength of 250 nm.
Sample Preparation was obtained from H9c2-NHE1-KO cell or the neonatal rat cardiomyocytes that had been treated with 100μM Empagliflozin or 10μM Tamoxifen for 24h or 48h. After a period of time, the cell culture medium is treated as extracellular fluid. Then cells were washed with phosphate-buffered saline and trypsinized, neutralized with complete medium. The suspension was repeatedly frozen and thawed three times. The intracellular fluid was obtained by centrifugation at 16000g for 5 minutes. The sediment is added to the cell membrane lysate, incubated on ice for 5 min, and centrifuged at 8000g for 5 min to form the cell membrane fraction. The three components of cells were subjected to HPLC analysis.

RNA-seq analysis.
The quality of the reads was evaluated with Fast QC. The samples information of clean data was shown in Tab

Molecular Docking
A homology model of the protein structure of NHE-1 was prepared using a human NHE1 protein structure (from UniProt (https://www.uniprot.org/) with ID: P19634) as a template. A molecular docking study was carried out using the following method to explore the possible interaction between NHE and SGLT2i. (a) Since homology below 30% sequence identity limits single-sequence based searches, the molecular Based on the interaction studies, Empaglifozin, Dapagliflozin and Canagliflozin were selected as the hit molecules for further analysis in the wetlab.

Statistical analysis.
Results are shown as the mean ± S.D. Paired data were evaluated using a Student's t-test. A one-way analysis of variance with the Bonferroni post hoc test was used for multiple comparisons. The Kaplan-Meier method with a log-rank test was used for survival analysis. Data were analyzed using GraphPad Prism 7.