RUNX1 facilitates heart failure progression through regulating TGF-β-induced cardiac remodeling

Animal experiment

This study was approved by the Animal Ethics Committee of Shandong University Qilu Hospital (IACUC approval number, KYLL-2022(ZM)-741). C57BL6 mouse was chosen as heart failure model animal. All the C57BL6 mice were purchased from animal experiment center of Shandong University. The C57 mice were fed with normal chow in SPF condition. The C57 mice involved in this study were raised in quiet environment at density of six mice per cage in normal circadian rhythm. The heart failure model was carried out using transverse aortic arch constriction method (TAC).

Mice anesthesia was performed using tracheal intubation method. Tracheal intubation anesthesia was carried out by using inhalation 5% Isoflurane at speed of 0.8–1.0 L/min and maintain dose using 3% Isoflurane at speed of 0.8–1.0 L/min. The mice neck skin were cut and trachea were exposed by blunt dissection method. Tracheal intubation need to be inserted from midline of mandible until tracheal carina position. A ventilator was used continuously during tracheal intubation anesthesia and tidal volume was set at 4 ml/min with respiratory rate at 75–90 times/min.

The C57 mice in TAC model group were fixed and performed thoracotomy and aortic arch ligation under anesthesia with tracheal intubation. The C57 mice in control group were only performed thoracotomy without aortic arch ligation under anesthesia with tracheal intubation. All the surviving animals at the conclusion of this experiment were fed at SPF condition until naturally dead. Data collection was performed eight weeks after TAC experiment.

Cell lines

The HL-1 Cardiac Muscle Cell Line was purchased from Sigma-Aldrich (No. SCC065) and was used for luciferase reporter assay. Cell line provenance was provided as a Supplemental File.

Groups and administration

At four weeks after heart failure model was established, the mice were randomly divided into three groups, TAC+lenti-NC group, TAC+lenti-shRUNX1 group and control group. The mice in control group were treated with thoracotomy only. The mice in TAC+lenti-NC group accepted thoracotomy and aortic arch ligation and were treated with nonsense lentivirus vector by tail vein injection. The mice in TAC+lenti-shRUNX1 group were treated with TAC and lentivirus vector which carried RUNX1-shRNA by tail vein injection. In addition, RUNX1-shRNA sequences were as follows, CCTCGAAGACATCGGCAGAAA; and lenti-NC sequences were, CCTAAGGTTA AGTCGCCCTCG.

Cardiac function and structure noninvasive assessment

We used cardiac ultrasound to detect cardiac function and left ventricle structure changes in mice under anesthesia by 3% pentobarbital sodium intraperitoneal injection. The evaluating parameters including left ventricular internal diameter diastole (LVIDd), left ventricular internal diameter systole (LVIDs), ejection fraction (EF) value and fractional shortening of ventricle (FS) value. Both EF and FS value are important pathophysiological parameters for myocardial systolic function assessment. Besides, LVIDd combing with LVIDs and EF value can be used as parameters for systolic heart failure assessment.

Construction of lentivirus vector

HEK-293T cells in the logarithmic phase were used for lentivirus package. We seeded 293T cells into 15 cm² culture dish at density of 5 × 10⁵ cells/mL and cultured cells using bovine serum-free medium 2 h before cell transfection. Cell transfection was carried out using Lipofectamine 2,000 when cell growth density reached to 70–80%. Four vectors system was chosen in this lentivirus transfection assay which includes PACK vectors (pPACKH1-GAG, pPACKH1-REV, pPACKH1-VSV), small hairpin RNA targeting RUNX1 (pLVX-sh-RUNX1) or nonsense small hairpin RNA control vectors (pLVX-sh-NC). All the PACK vectors, shRNA recombinant vectors were mixed with Lipofectamine, Polybrene and bovine serum-free medium at room temperature for 20 min. The HEK-293T cells were washed by PBS (phosphate buffer solution) for three times before transfection. We added mixture of vectors and Lipofectamine into 293T cells drop by drop. After that, the 293T cells were cultured using medium containing 10% fetal bovine serum and puromycin for 48 h at 37 °C and 5% CO₂ condition. Subsequently, the supernatant of HEK-239T cells was collected and slightly centrifuged at 3,000×g at 4 °C for 10 min. We filtered the gained cell supernatant using 0.45 μm filter membrane and froze the virus liquid at −80°C refrigerator.

Detection of blood BNP level

BNP (natriuretic peptide type B) is generally accepted as assessment marker for heart failure. In this study, we collected blood from ventricle of mice that were killed by cervical dislocation. A total of 0.5 ml serum was gained from mice blood and equivalent buffer was added. The specific procedures were carried out following the protocol of BNP ELISA kit (RAB0386, minimum detectable concentration is 1.66 pg/mL, detection range, is 0.1–1,000 pg/mL; Sigma-Aldrich, St. Louis, MO, USA) and absorbance value at 450 nm were detected by microplate reader.

Pathomorphological assessment of myocardial tissue

The pathomorphological tissue changes of myocardium were evaluated by H&E staining. All the myocardial tissues were dehydrated by gradient ethanol step by step and embedded by paraffin. The embedded tissue were cut into serial histologic sections and dewaxed using xylene step by step. The tissue slides were stained by eosin dye and Hematoxylin according to H&E staining procedures. The changes of myocardial fibrosis were assessed by Masson staining. For Masson staining, tissue slides were stained by Weigert hematoxylin, Ponceau and Aniline blue step-by-step. In addition, the pale blue in Masson staining image is for collagen fibers, pink for muscle fibers and blue brown represents nuclei.

Detection of apoptotic cells

TUNEL staining was applied to apoptotic rate detection in myocardial tissues. For TUNEL staining, we stained myocardial tissue slides with TUNEL Assay kit (BrdU-Red, ab66110; Abcam, Cambridge, UK) following the procedure instructions and the fluorescence images were captured by using Fluorescence microscope. At the same time, we stained cardiomyocyte in tissue slides with Desmin antibody (Alexa Fluor 555 Anti-Desmin antibody, ab203422; Abcam, Cambridge, UK shown in green) at 1/100 dilution by immunofluorescence staining assay. Moreover, the nuclei in tissue slides were stained by using DAPI (ab285390; Abcam, Cambridge, UK). DAPI was used for nuclei stain at concentration of 10 μg/ml under room temperature for 15 min avoid of light.

Dual-luciferase reporter assay

HL-1 cells were plated in 96-well plates and were transiently transfected with 0.15 μg of pGL4.48[luc2P/SBE/Hygro] vector (Promega, Madison, WI, USA) and 0.015 μg of Renilla luciferase control reporter vector (Promega, Madison, WI, USA). The transfection assay above was carried out with FuGENE six transfection reagents (Promega, Madison, WI, USA) following the manufacturer’s instructions. The SBE in pGL4.48 vector was the Smad binding element and luciferase activity was normalized to Renilla activity. After 24 h of transfection, HL-1 cells were treated with TGFβ1 (20 ng/ml, Peprotech, Rocky Hill, CT, USA) for 24 h. In the RUNX1 treated group, cells were treated with TGFβ1 (20 ng/ml) for 12 h and then co-treated with recombinant RUNX1 protein (Origene Technologies, Rockville, MD, USA) for 12 h. In the end, cells were collected and put into luciferase reporter assay.

Protein extraction and western blot

The whole heart was gained from mice and the total protein was extracted using T-PER Tissue Protein Extraction Reagent (No.78510; Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s instructions. Besides, protein quantitative detection was carried out with BCA protein detection Kit following the procedures provided by the manufacturer (Pierce^™ BCA protein detection kit, No. 23227; Thermo Fisher Scientific, Waltham, MA, USA). The expressions of RUNX1, Smad2, Smad3, Smad4 and GAPDH were detected by western blot assay. The concentration of separation gel (15%) and SDS-PAGE electrophoresis voltage were adjusted according to molecular weight of target proteins. Furthermore, goat anti-Mouse IgG H&L Adsorbed Secondary Antibody (Alexa Fluor^™ Plus 680, A32729; Invitrogen, Waltham, MA, USA), and goat anti-Rabbit IgG H&L Highly Cross-Adsorbed Secondary Antibody (Alexa Fluor^™ Plus 680, A32734; Invitrogen, Waltham, MA, USA) were used for fluorescence imaging. All the fluorescence images were captured by using Odyssey LI-COR infrared imaging system. The primary RUNX1 antibody was purchased from Abcam (Rabbit to RUNX1, ab229482, Reactivity: Mouse, Rat, Human) and Smads antibodies from CST (Smad2 Rabbit mAb (dilutation ratio, 1:2,000; #5339S), Smad3 Rabbit mAb (dilutation ratio, 1:2,000; #9523), Smad4 Rabbit mAb (dilutation ratio, 1:2,000;#46535), Phospho-Smad2 (dilutation ratio, 1:2,000; Ser465/Ser467, Rabbit mAb#18338), Phospho-Smad3 (dilutation ratio, 1:2,000; Ser423/Ser425, Rabbit mAb#9520), GAPDH (Rabbit mAb#5174, dilutation ratio, 1:2,000; Cell Signaling Technology; Danvers, MA, USA)). All the fluorescence images were converted into monochrome images and grey value semi-quantitative analysis was carried out using ImageJ software.

RNA extraction and qPCR

RNA extraction from myocardial tissues was carried out using Trizol Reagent (Invitrogen, Waltham, MA, USA) following manufacturer’s instructions. All the PCR primers were designed and synthesized by Shanghai Sangon Co. (Shanghai, China). Primers for RUNX1 gene were, 5′-3′ CAGGCAGGACGAATCACACT; 3′-5′ CTCGTG CTGGCATCTCTCAT. Primers for GAPDH were designed as 5′-3′ GGTGAAG GTCGGTGTGAACG; 3′-5′ CTCGCTCCTGGAAGATGGTG. cDNA synthesis was performed using RT Revert Aid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA). PCR reaction was carried out by using Platinum SYBR Green PCR Super Mix- UDG w/ROX Kit (No. C11744100; Invitrogen, Waltham, MA, USA) according to the manufacturer’s instructions. Fluorescence signals were captured by ABI7500 real time PCR instrument and cycles to threshold (CT value) were calculated and recorded. 2^−ΔΔCT value was used to reflect relative gene expression levels.

Statistical analysis

Data analysis was carried out using SPSS 20.0 software (Chicago, IL, USA). Student’s t-test was used for comparison between two groups of numerical variable data. Besides, One-way ANOVA method was used for comparison among three groups and more numerical variable data. P < 0.05 was considered as the difference with statistical significance. P < 0.01 was considered as the difference with obvious statistical significance.

Inhibition of RUNX1 lead to morphological changes of heart in TAC mouse

We carried out heart failure animal model by TAC induced method and found that heart size in TAC-induced group (TAC+lenti-NC) was much bigger than that in control group (Fig. 1A). Compared with lenti-NC group, mouse heart in lenti-shRUNX1 group became much smaller in size (Fig. 1A). These results revealed that the inhibition of RUNX1 caused alleviation of cardiac enlargement induced by TAC.

Inhibition of RUNX1 lead to improvement of heart function in TAC mouse

We performed noninvasive ultrasound examination to assess cardiac function changes in different groups (Fig. 1B) and found that LVIDs and LVIDd values in lenti-shRUNX1 group were much lower than those in TAC+lenti-NC group (Fig. 1C, i, ii). Besides, EF and FS values which can reflect cardiac systolic function in lenti-shRUNX1 group mice were much higher than those in TAC+lenti-NC group (Fig. 1C, iii, iv). These results showed that the inhibition of RUNX1 can lead to improvement of cardiac function in TAC mouse. This result also means that heart function can be improved by RUNX1 inhibition in TAC induced model.

Inhibition of RUNX1 can ameliorate myocardial fibrosis after TAC

We compared histological changes of mouse heart in different groups through H&E staining. The results showed that myocardial cells in control group were arranged with clear and regular transverse striations with normal cardiomyocyte morphology (Figs. 2A and 2B). Nevertheless, myocardial cells in TAC+lenti-NC group had irregular shape and significant disordered arrangement compared with TAC+lenti-shRUNX1 group and control group. In addition, the transverse striations of heart tissues were broken in TAC+lenti-NC group. The tissue structure and cell morphological changes in TAC+lenti-shRUNX1 group were much slighter than that in TAC+lenti-NC group.

We assessed myocardial fibrosis of mouse heart tissues in different groups by Masson’s trichrome staining and found that myocardial tissue in control group showed no obvious collagen accumulation. However, myocardial tissue in TAC+lenti-NC group appeared significant collagen accumulation and a decrease of myocardial cells number compared with TAC+lenti-shRUNX1 group (Figs. 3A and 3B). Therefore, these results revealed that the inhibition of RUNX1 could lead to significant alleviation of myocardial fibrosis and cardiac remodeling in TAC mouse.

Inhibition of RUNX1 reduces cell apoptosis after TAC induced heart failure

Decrease of myocardial cells caused by apoptosis is one of the important etiologies in heart failure pathophysiological process. We detected cell apoptosis in myocardial tissues using TUNEL staining. The results showed that the myocardial cell apoptosis in TAC+lenti-NC group was significantly higher than that in TAC+ lenti-shRUNX1 group and control group (Fig. 4). This result suggested that the inhibition of RUNX1 could lead to reduction of cell apoptosis in TAC induced model. Besides, this result partly explained why RUNX1 could regulate heart failure process in animal experiments.

Inhibition of RUNX1 lead to changes of heart weight/body weight ratio and BNP levels in TAC mouse

Furthermore, we recorded the body weight and heart weight of mice in different groups and compared the heart weight/body weight ratio among them. The results revealed that heart weight/body weight ratio in TAC+lenti-NC group was significantly higher than that in TAC+lenti-shRUNX1 group (Figs. 5A–5C). These macroscopical morphological changes in the heart of TAC mouse indicated that RUNX1 may play an important role in regulating cardiac remodeling that induced by TAC. Moreover, we detected plasma BNP levels in different groups and the results showed that BNP levels in TAC+lenti-NC group were much higher than those in TAC+ lenti- shRUNX1 group (Fig. 5D).

RUNX1 promotes activation of TGF-β/Smads signaling in mouse heart

In order to illustrate the underlying molecular mechanisms of how RUNX1 regulates myocardial fibrosis, we examined whether the canonical Smad3-dependent TGF-β signaling could be activated by RUNX1 through luciferase reporter assay. The results showed that RUNX1 had significantly stimulated Smad transcriptional activity in HL-1 cells (Fig. 6A). Western blotting results further revealed that the inhibition of RUNX1 caused reduction of phosphorylated Smad2/3 levels in mouse myocardial tissues (Fig. 6B). These results suggested that RUNX1 exerted its effects in cardiac remodeling through regulating the canonical TGF-β signaling pathway.