All animal experiments were carried out in accordance with the Guidelines of Animals Experiments from Ethical Committee for Animal Research of China Pharmaceutical University. Adult male Sprague–Dawley (SD) rats (weight 220–250 g) were purchased from the Model Animal Research Center of Nanjing University. The rats were raised under pathogen-free conditions, under the 12 light-dark cycle, 25 ± 2°C condition and free to water and food. Isoprotereno (ISO) (Sigma-Aldrich, United States) was used for cardiac fibrosis modeling. The rats were randomized into control group, ISO group, ISO + valsartan group and ISO + sacubitril/valsartan group. Valsartan was purchased from Changzhou Siyao Pharm, China; and sacubitril/valsartan was purchased from Novatis Pharma, Schweiz AG. Rats in control group were subcutaneously injected with saline, while the other three groups were subcutaneously injected with ISO (7.5 mg/kg/day) for 14 days. From the 6th day to the 14th day, the rats in control group and ISO group were given saline by gavage (0.01 mL/g body weight), the rats in ISO + valsartan group were given valsartan by gavage (0.01 mL/g body weight, dissolved in 0.5% CMC-Na solution), and the rats in ISO + sacubitril/valsartan group were given sacubitril valsartan by gavage (0.01 mL/g body weight, dissolved in 0.5% CMC-Na solution). On the 15th day, rats were anesthetized with 2% isoflurane and kept on a 37°C heated platform. The chest hairs were removed using depilatory cream and acoustic coupling gel was applied to the thorax, then echocardiogram tests were performed using small animal ultrasound imaging system vevo3000 (VisualSonics, Canada). The cardiac function related indicators, including cardiac output, ejection fraction, fractional shortening, and stroke volume, were analyzed through the echocardiogram results. Then all rats were sacrificed and heart tissues were taken for follow-up analysis. The left ventricular apexes were resected and divided into two parts: one part was fixed in 10% formalin solution and then subjected to Masson trichrome staining to observe the histological changes, while the other part was frozen at −80°C for subsequent detection of mRNA and protein.

Mouse embryonic fibroblast (MEF) and rat embryonic cardiomyocyte cell line H₉C₂ were cultured in DMEM (Gibco, United States) containing 10% FBS (Gibco, United States), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco, United States). Cells were digested with 0.25% trypsin-EDTA (Gibco, United States) and subcultured for subsequent experiments when grew to 90% confluence.

Si-TRPM7 and si-NC were synthesized by GenePharm (Shanghai, China). MEF were seeded in 12-well cell culture plate (NEST, China) and transfected with small interfering RNAs (siRNAs) using Lipofectamine^TM 3000 Transfection Reagent (ThermoFisher, United States) in serum-free Opti-MEM medium (Gibco, United States). The cells were transfected for 8 h and then cultured in DMEM containing 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin. The sequences of si-TRPM7 are as following:

Total RNA of cells or tissues was extracted with Trizol (Invitrogen, United States) according to the manufacturer’s instructions, then reverse transcription was performed using a reverse transcription kit (Vazyme, China). The quantification of gene transcripts was performed by real-time PCR using SYBR Green PCR mix (Vazyme, China). All values were normalized to the level of 18S mRNA. The primers used were listed below:

Cell protein was extracted by sonication and tissue protein was extracted by grinding. After centrifugation at 12,000 rpm for 15 min, the supernatant was collected for subsequent western blot analysis. Protein samples were boiled for 5 min, electrophoresed in SDS polyacrylamide gel, and then transferred onto PVDF membranes (Bio-Rad, United States). The blots were blocked with 5% skimmed milk in Tris-buffered saline solution-Tween 0.1% (TBST) (Sigma-Aldrich, United States) for 2 h at room temperature and probed with primary antibodies overnight at 4°C. The blots were washed four times with TBST for 8 min each time and incubated for 1 h at room temperature with the HRP-conjugated secondary antibodies (dilution 1:10,000; Sigma-Aldrich, United States), then developed with chemiluminescence (Vazyme, China). MMP9 antibody (10375-2-AP) was purchased from Proteintech, China, dilution 1:1,000; α-SMA (ab32575) and MLKL (ab243142) antibodies were purchased from Abcam, United States, dilution 1:1,000; TRPM7 antibody (BM5443) was purchased from Boster, China, dilution 1:200; LC3B (3868S), caspase3 (14220S) and cleaved caspase3 (9664S) were purchased from Cell Signaling Technology, United States, dilution 1:1,000; RIPK1 (AF7877), RIPK3 (AF7942) and phospho-MLKL (Ser358) (AF7420) antibodies were purchased from Affinity, China, dilution 1:1,000. The densitometry of protein bands was quantified using ImageJ software. All values were normalized to the level of β-actin.

The crystal structure of TRPM7 (PDB ID: 5ZX5) (Duan et al., 2018) was retrieved from the RCSB Protein Data Bank. The protein structure was prepared for Glide docking calculations using the Protein Preparation Wizard utility. Water molecules were removed and the structure was treated and prepared by employing a protein preparation panel in the Schrödinger enterprise to magnify H-bond interactions. Structures of investigated compound (NEPi and Diprotic acid ARB) were built recruiting MAESTRO (Friesner et al., 2004) build panel and subsequently energy minimization by LigPrep module employing the OPLS2005 forcefield. Ionizable compounds were converted to their most probable charged forms at pH 7.0 ± 2.0 and explicit hydrogen atoms were added. The grid file for TRPM7 was generated by the Glide Grid Generation wizard in Schrödinger determining the co-crystallized ligands as the center point. Van der Waals scaling factor of the non-polar atoms was adjusted to 1.0 to enhance flexibility. The other parameters were fixed as defaults. The top conformations of investigated compounds were ranked by the corresponding values of docking score.

Hearts from the neonatal rats (1–3 days) were rapidly removed and washed in D-hanks solution and minced into 1 mm³ pieces, then shake in 0.1% trypsin and 0.1% collagenase type II for several times to digest. The fully digested cell suspension was centrifuged and resuspended in Dulbecco’s modified eagle’s medium (DMEM) (Life Technologies, United States) containing 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin, and incubated in a humidified atmosphere of 5% CO₂ at 37°C for 2 h. After removing non-adhered cells, the attached cells were cultured and inherited, the second or third generations were used in our experiments.

The third passage of neonatal rat cardiac fibroblasts was used in our experiments. TRPM7-like current was recorded by whole cell patch clamp at room temperature. The pipette solution contained (in mM) CsCl 145, NaCl 8, HEPES 10, EGTA 10, and CsOH was used to adjust to pH7.2. The bath solution contained (in mM) NaCl 145, KCl 5, HEPES 10, glucose10 and CaCl₂ 2, and NaOH was used to adjust to pH7.4. With a holding potential at 0 mV, TRPM7 current was recorded under the voltage stimulation from −100 mV to + 120 mV in 400 ms.

H₉C₂ were divided into normoxic group, hypoxia group, hypoxia + BAPTA-AM group and hypoxia + LBQ657 group. BAPTA-AM and LBQ657 were purchased from MedChemExpress, China and Sigma-Aldrich, United States, respectively. Cells were seeded in 6-well cell culture plate (NEST, China) at a density of 5 × 10⁵ per well. 8 h later, cells were treated with BAPTA-AM or LBQ657 for 8 h and then moved to a hypoxia environment by incubating into three-gas incubator (93% N₂, 5% CO₂ and 2% O₂) for 24 h to achieve, while the control groups were kept in normoxic incubator. All cell pellets were collected and total protein were extracted for western blot test.

H₉C₂ were loaded with the Ca²⁺ fluorescent dye, Fluo-4-AM (Invitrogen, United States) at 37°C for 1 h incubation, then washed by PBS solution and incubated at 37°C for 30 min. Cells were fixed by 4% paraformaldehyde and screened in 488 nm excitation light using Laser scanning confocal microscope LSM800 (Zeiss, Germany). The fluorescence ratio was analyzed using Image J software.

Statistical analysis was performed using GraphPad Prism 6.0. The results were expressed as mean ± SD for experiments conducted at least in triplicates. Unpaired t test was used for comparison between two groups, and One-way ANOVA was used for three or more groups, and P < 0.05 was considered to be significantly different.

The rats were subcutaneous injected ISO to induce cardiac fibrosis and treated with normal saline, valsartan or sacubitril/valsartan, respectively (Figure 1A). Valsartan, a drug used for clinical treatment of heart failure, was used as positive control (Suematsu et al., 2016). The valsartan group and the sacubitril/valsartan group had lower heart weight to body weight ratio (Figure 1B). Echocardiogram showed that the cardiac output, ejection fraction, fractional shortening and stroke volume of ISO-induced rats were significantly increased after valsartan and sacubitril/valsartan treatment (Figures 1C,D). Especially, sacubitril/valsartan ameliorated stroke volume more efficiently than valsartan. Additionally, the valsartan group and the sacubitril/valsartan group had smaller area of fibrosis (Figure 1E) compared to non-treatment control group. These results indicate that sacubitril/valsartan effectively protected cardiac dysfunction in ISO-stimulated rats (P < 0.01 compared with control group).

Matrix metalloproteinase 9 (MMP9) is a biomarker associated with collagen degradation to cause heart failure (Zile et al., 2019). Alpha-smooth muscle actin (α-SMA) is an activation marker of fibrotic pathways in fibroblast differentiation (Jing et al., 2017). Accordingly, quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was used to analyze the expression of these cardiac fibrosis related genes in the left ventricular apexes. We found that ISO induced a notable increase in Mmp9 and α-Sma mRNA expression levels, while sacubitril/valsartan or valsartan treatment diminished the effect (Figure 1F). The protein expression levels of MMP9 and α-SMA detected by western blotting analysis showed a similar trend (Figure 1G). Moreover, sacubitril/valsartan significantly decreased α-Sma mRNA expression compared with valsartan. Taken together, the above results suggest that sacubitril/valsartan effectively ameliorated ISO induced cardiac fibrosis and improved cardiac function in vivo, superior to valsartan therapy alone. This indicates a potential anti-fibrotic function of sacubitril.

Transforming growth factor-β 1 (TGF-β1) is considered as a typical stimulus to mediate the differentiation of cardiac fibroblasts into myofibroblasts to acquire a fibrotic phenotype (Porter and Turner, 2009). To further examine the effect of sacubitril on cardiac fibrosis, we used TGF-β1 (10 ng/mL) to stimulate MEF for 24 h to induce a profibrotic phenotype. LBQ657, the metabolite of sacubitril, was used to treat MEF cells at a concentration of 50 μM prior to TGF-β1 stimulation. We found both the mRNA and protein expression of MMP9 and α-SMA was dramatically up-regulated by TGF-β1 stimulation, but this was prevented by LBQ657 pre-treatment (P < 0.01 comparison with the TGF-β1 group), indicating LBQ657 potently inhibited fibroblast activation in vitro (Figures 2A,B).

Because the previous reports suggested a close relationship between TRPM7 and cardiac fibrosis (Du et al., 2010; Wei et al., 2020), we further investigated the changes of TRPM7 expression in TGF-β1 induced cardiac fibrosis. We designed and selected a pair of TRPM7 siRNA with the best interference effect for follow-up experiments. The TRPM7 mRNA levels were reduced by 74% (Figure 2C). MEF were subjected to si-TRPM7 transfection and LBQ657 (50 μM) treatment, respectively, before TGF-β1 stimulation. The qRT-PCR and western blotting analysis indicated that si-TRPM7 suppressed the expression of TRPM7, but LBQ657 pre-treatment did not show discernible effects on TRPM7 protein expression in TGF-β1 treated MEF (Figures 2E,F). Meanwhile, both si-TRPM7 and LBQ657 reduced the expression of MMP9 and α-SMA induced by TGF-β1 (Figures 2D,E). These results suggest that LBQ657 attenuated TGF-β1-induced cardiac fibroblast activation without reducing the expression levels of TRPM7.

The inhibition of cardiac fibroblast activation by LBQ657 was not associated with reduced TRPM7 mRNA and protein levels, ruling out the possibility that LBQ657 functions by regulating TRPM7 channel current. Molecular docking was used to determine the interaction between TRPM7 (PDB ID: 5ZX5) and sacubitril/valsartan. Compound NEPi and Diprotic acid ARB of sacubitril/valsartan docked to the active site of TRPM7 with a free energy of −6.20 kcal/mol and −10.12 kcal/mol, respectively. This indicated that both sacubitril and valsartan docked to TRPM7 protein. The docking structure suggested that LBQ657, as the NEPi, occupied a similar position in TRPM7 compared with natural cholesterol hemisuccinateon ligand (Figure 3A).

Aimed to validate the blocking of TRPM7 channel by LBQ657, TRPM7-like current was recorded by whole cell patch-clamp in neonatal rat cardiac fibroblasts. We observed that the current of TRPM7 channel was decreased by LBQ657 treatment in a clear dose-dependent manner (Figure 3B), strongly implying its direct influence on the channel function. And then we used 2-aminoethoxydiphenyl borate (2-APB), an inhibitor of TRPM7 channel, as a positive control to explore the anti-fibrotic effect of LBQ657 (Chokshi et al., 2012). As shown in Figures 3C,D, 2-APB (100 μM) indeed caused dramatic reduction of the MMP9 and α-SMA expression by ∼50%, so did LBQ657, in TGF-β1 stimulated MEF cells. We concluded from this set of experiments that the LBQ657 extenuated cardiac fibroblast activation through inhibiting the function of TRPM7 channel, rather than suppressing the expression of TRPM7.

The causal importance of hypoxia in cardiac fibrosis has been strongly supported by literature. Hypoxia occurs in myocardial infarction and causes large-scale loss of cardiomyocytes, which is the most common trigger of cardiac fibrosis. Cardiomyocyte death induces inflammatory responses, promoting the activation of fibroblasts into myofibroblasts, leading to cardiac fibrosis (Kong et al., 2014; Kumar and Choi, 2015). TRPM7 channel is mainly concentrated in myocardium during embryonic development and is indispensable for cardiac automaticity of cardiomyocytes (Sah et al., 2013). Therefore, we detected that TRPM7 was abundantly expressed in cardiomyocytes, approximately 1.5-fold greater than that in fibroblasts (Figure 4A). And then we measured the intracellular calcium concentration in H₉C₂ under hypoxia stimulation. The Figure 4B showed that Ca²⁺ influx was dramatically increased by hypoxia, but significantly deceased after LBQ657 treatment. These results demonstrated that LBQ657 inhibited Ca²⁺ entry though TRPM7 channel in cardiomyocytes.

While some studies had previously shown that activation of TRPM7 channels and produced Ca²⁺ overload significantly contributes to cell death (Wei et al., 2007; Cai et al., 2014), it was still to be determined whether LBQ657 attenuates cardiomyocyte PCD by TRPM7 mediated Ca²⁺ influx. We first studied the effect of LBQ657 on hypoxia-induced cardiomyocyte PCD. As shown in Figure 4C, there were no significant changes on the protein expression level of caspase3, cleaved caspase3, LC3B-I and LC3B-II in hypoxia-stimulated H₉C₂ pre-treated with LBQ657, suggesting that LBQ657 was incapable of regulating cardiomyocyte apoptosis or autophagy in hypoxic environment. The investigation of cell necrosis related proteins RIPK1/RIPK3/MLKL/p-MLKL expression in hypoxia-induced H₉C₂ under LBQ657 pre-treatment indicated the expression level of detected proteins declined to 70% or less compared with that in the hypoxia group (Figure 4D). We further found LBQ657 did not cause obvious changes in the expression level of RIPK1/RIPK3/MLKL/p-MLKL in normoxia cardiomyocytes (Figure 4E). Next, BAPTA-AM (5 μM) was used to chelate the intracellular calcium and caused down-regulated expression of RIPK1/RIPK3 in hypoxia-treated cardiomyocytes, but MLKL and p-MLKL protein expression were not notably altered (Figure 4F). These results suggested that the reduction of cardiomyocyte necrosis was closely associated with calcium influx. The cardiomyocyte necrosis was significantly reduced accompanied with decrease of TRPM7-mediated Ca²⁺ influx. Based on the above results, we conclude that LBQ657 could ameliorated cardiomyocyte necrosis induced by hypoxia via inhibiting TRPM7-mediated Ca²⁺ influx.