microRNA‐454‐mediated NEDD4‐2/TrkA/cAMP axis in heart failure: Mechanisms and cardioprotective implications

Abstract The current study aimed to investigate the mechanism by which miR‐454 influences the progression of heart failure (HF) in relation to the neural precursor cell expressed, developmentally downregulated 4‐2 (NEDD4‐2)/tropomyosin receptor kinase A (TrkA)/cyclic adenosine 3',5'‐monophosphate (cAMP) axis. Sprague‐Dawley rats were used to establish a HF animal model via ligation of the left anterior descending branch of the coronary artery. The cardiomyocyte H9c2 cells were treated with H2O2 to stimulate oxidative stress injury in vitro. RT‐qPCR and Western blot assay were subsequently performed to determine the expression patterns of miR‐454, NEDD4‐2, TrkA, apoptosis‐related proteins and cAMP pathway markers. Dual‐luciferase reporter gene assay coupled with co‐immunoprecipitation was performed to elucidate the relationship between miR‐454, NEDD4‐2 and TrkA. Gain‐ or loss‐of‐function experiments as well as rescue experiments were conducted via transient transfection (in vitro) and adenovirus infection (in vivo) to examine their respective functions on H9c2 cell apoptosis and myocardial damage. Our results suggested that miR‐454 was aberrantly downregulated in the context of HF, while evidence was obtained suggesting that it targeted NEDD4‐2 to downregulate NEDD4‐2 in cardiomyocytes. miR‐454 exerted anti‐apoptotic and protective effects on cardiomyocytes through inhibition of NEDD4‐2, while NEDD4‐2 stimulated ubiquitination and degradation of TrkA protein. Furthermore, miR‐454 activated the cAMP pathway via the NEDD4‐2/TrkA axis, which ultimately suppressed cardiomyocyte apoptosis and attenuated myocardial damage. Taken together, the key findings of the current study highlight the cardioprotective role of miR‐454, which is achieved through activation of the cAMP pathway by impairing NEDD4‐2‐induced TrkA ubiquitination.


| INTRODUC TI ON
As a significant consequence of cardiovascular disease, heart failure (HF) has been reported to affect approximately 26 million people worldwide, many of whom have a dismal prognosis and quality of life. 1 The elder population is generally more susceptible to HF, which explains the expected increase in incidence in this population owing to the growing aging population as well as increased life expectancy. 2 HF is characterized by the loss of autonomic balance with the degeneration of vagal activity accompanied by an elevation in sympathetic activity. 3 HF has been suggested to predominately arise from progressive myocardial disorders coupled with myocardial remodelling. 4 Cardiomyocyte apoptosis and oxidative stress represent two crucial factors widely documented to participate in the pathogenesis of HF. 5,6 Thus, further investigation into the molecular mechanisms associated with cardiomyocyte apoptosis and oxidative stress is highly necessary for the development of therapeutic targets for HF. microRNAs (miRs) represent endogenous non-coding RNAs (~22 nucleotide) capable of mediating gene expression in a posttranscriptional manner, some of which have been shown to possess the ability to regulate myocardial remodelling post-HF by targeting genes. 7 Of note, miR-454 has been previously reported to exhibit an aberrant decrease in patients with diastolic dysfunction, an important symptom of cardiovascular disorder. 8 With miR-454 as a research focus, we performed prediction of its target genes, based on which neural precursor cell expressed, developmentally downregulated 4-2 (NEDD4-2) was predicted to be the target gene of miR-454. NEDD4-2 is an ubiquitin E3 ligase capable of disposing target proteins for degradation. 9 NEDD4-2 has been previously reported to exert regulatory functions on high-glucose-treated rat cardiomyocytes to affect diabetic cardiomyopathy. 10 Importantly, NEDD4-2 downregulates voltage-gated sodium channel Nav1.5 through ubiquitination in the progression of HF. 11 The activation of NEDD4-2 is correlated with oxidative stress in hypertension. 12 Existing literature has highlighted the ubiquitination of tropomyosin receptor kinase A (TrkA) by NEDD4-2 as a crucial degradation pathway required in the maintenance of neuronal networks. 13 Depletion of NEDD4-2 is also reported to activate TrkA signalling. 14 Likewise, a previous investigation concludes that impaired binding to NEDD4-2 results in a diminished ubiquitination level of the TrkA neurotrophin receptor in a mouse model, which ultimately influences neuronal survival and their sensitivity to pain. 15 Intriguingly, TrkA is a receptor of nerve growth factor (NGF), which has been linked to sympathetic neuronal function and neuroanatomy in congestive HF. 16,17 To our knowledge, TrkA is capable of activating the 3',5'-cyclic adenosine monophosphate (cAMP) pathway. 18 cAMP represents a pivotal second messenger that possesses the capacity to modulate heart function by means of contributing to subcellular microdomains in the setting of chronic HF. 19 A recent study has demonstrated the protective role of this pathway in the cardiac function following HF. 20 Based on the aforementioned findings and reports, we proposed that miR-454 may play a regulatory role in HF through mediating the NEDD4-2/TrkA/cAMP axis. An in vivo HF model in addition to an in vitro oxidative stress model were established in order to identify the effects associated with these mediators on myocardial damage as well as their interaction during the progression of HF.

| Ethics statement
This study was conducted with the approval of the medical ethics committee of the Daping Hospital, Army Medical University. Signed   Table 1. A total of 72 healthy individuals received coronary angiography at Daping Hospital, Army Medical University, during the same period were randomly selected as the controls, including 37 males and 35 females, with a median age of 58.5 ± 5.4 years. The inclusion criteria of the experimental group as well as the control group were as follows: age <70 years old, no history of chronic diseases such as hypertension, type 2 diabetes, malignant tumour, lung disease, cirrhosis, renal failure, etc A total of 3 mL of peripheral venous blood was collected and placed into ethylenediaminetetracetic acid (EDTA) tubes from all selected individuals who had undergone a period of fasting early in the morning, with the plasma component subsequently separated within 2 hours via centrifugation at 4°C and 3000 g for 10 minutes and stored at −80°C. The rats were placed in a supine position with their limbs and heads fixed on the operating table; the hair on their neck and chest was shaved off, followed by disinfection with iodine and 70% ethanol.

| Establishment of HF rat model
The skin of the middle part of the neck was cut open, after which the subcutaneous muscle tissues were surgically removed in a layer-bylayer fashion until the trachea was exposed. Endotracheal intubation was performed using intravenous indwelling needles; the inner core was removed and fixed, followed by connection with a small animal ventilator (DW-2000) with the ventilation volume maintaining at 1.5 mL-2.0 mL/time, with a respiratory rate of 80 breaths/min. The skin, intercostal muscle and pleura were cut open in turn between the fourth and fifth intercostal spaces of the left thoracic cavity in order to open the thoracic cavity. The heart was fully exposed with the pericardium separated. For HF modelling, the left anterior descending (LAD) coronary artery was ligated approximately 2 mm from the lower edge of the left auricle using 8-0 nylon suture. LAD ligation was deemed to be successful if the anterior wall of the left ventricle near the apex of the heart was grey white or blue purple, accompanied by ventricular wall abnormality and reduced movement contractile force. An identical operation with no ligation was performed in the sham-operated rats. The rib, muscle and skin were sutured in turn to close the thoracic cavity; iodine and 70% ethanol were subsequently employed for disinfection in turn. The tracheal intubation was removed following recovery of spontaneous respiration; the skin of the neck was sutured and placed back into the cage for 7 days of standard feeding. Twelve rats were sham-operated, while the remaining 48 rats underwent LAD ligation for modelling purposes. Among the 48 rats after modelling, 12 rats were administered with an injection of 50 μL of sterile phosphate-buffered saline (PBS) after 30 minutes of LAD ligation, 12 rats received injection with adeno-associated virus-NC (AAV-NC), respectively, in situ after 30 minutes of LAD ligation, and 12 rats received injection with AAV-miR-454 (the AAV-DJ vector was provided by GenePharma) in situ after LAD ligation. The remaining 12 rats were administrated with an injection of AAV-miR-454 in situ following LAD ligation, after which they were intraperitoneally injected with cAMP pathway inhibitor H-89 at a concentration of 10 μmol/L, 30 mg/kg. 20 The dosage of the AAVs was 1.5 × 10 14 vector genomes/kg.

| Colour Doppler ultrasound
The rats were mildly and continuously anaesthetized with 2.0% iso- and fractional shortening (FS) were measured and calculated on the short-axis view of the left ventricular papillary muscle. Heart rate was evaluated by simultaneous electrocardiographic monitoring.

| Haematoxylin-eosin (HE) staining
The rats were euthanized following a 7-day period of modelling, with their heart tissues collected, paraffin-embedded and sectioned.
HE staining was performed to examine the pathological changes. After the heart tissues were fixed, dehydrated, waxed, sliced (5 μm in thickness) and placed onto slides, the slides were subsequently baked in a 60°C incubator for 30 minutes, dewaxed, immersed in xylene I and II, each for 20 minutes and hydrated. Following immersion in gradient alcohol (100%, 95%, 80% and 70% each for a 5-minutes period), the slices were stained with haematoxylin for 2-3 minutes and then immersed in 1% ammonia water for 1-2 minutes. The slices were immersed in eosin dye for 1 minutes, followed by dehydration in 70%, 80%, 95% and 100% alcohol for 5 minutes, respectively, and then cleared in xylene for 10 minutes. The slices were finally sealed with neutral resin and observed by microscopy.

| Masson's Trichrome staining
The paraffin-embedded heart sections were baked in a 60°C in- Glyceraldehyde phosphate dehydrogenase (GAPDH) and U6 were employed as internal controls, with the relative quantitative method (2 − ΔΔC t method) applied for the calculation of the relative expression of the target genes.

| Western blot assay
The tissues and cells were lysed, respectively, using radioimmu- and p-cAMP response element-binding protein (CREB; PA1-850; 2 µg/mL), which were purchased from Thermo Fisher Scientific Inc The proteins were incubated with rabbit antibody against immunoglobulin G (IgG; 1:1000, Santa Cruz) labelled with horseradish peroxidase at room temperature for 1 hour and colour was developed in the enhanced chemiluminescence reagent (Thermo Fisher Scientific Inc), followed by development and fixation (Bio-Rad ChemiDoc Imaging system, Bio-Rad Laboratories). GAPDH (mouse anti-GAPDH, 1:1000, Santa Cruz Biotechnology) was regarded as an internal reference, while the protein blot images were subsequently analysed using imageJ2x software.

| Immunochemistry
The sample was fixed with 10% neutral formalin, embedded in paraffin and sectioned with an ultramicrotome. The sections were deparaffinized with xylene, rehydrated with graded alcohol and incubated with 3% hydrogen peroxide to block endogenous peroxidase activity. The sections were boiled in 10 mmol/L sodium citrate (pH 6.0) for 30 minutes, then blocked in 10% normal goat serum for 15 minutes and incubated with antibodies against TrkA (PA5-109216, Invitrogen) and NEDD4-2 (ab46521, Abcam) overnight in a wet room at 4°C. Next day, the sections were washed with PBS and incubated with the secondary antibody for 1 hour at room temperature. The immunoreactivity was detected with DAB kit (Invitrogen).

| Cell culture and transfection
The cardiomyocyte H9c2 cell line was obtained from American Type

| Oxidative stress cell model
Upon reaching 80% confluence following plasmid transfection, the H9c2 cells were treated with 200 μm H 2 O 2 for 12 hours to induce oxidative stress. At the same time, the cells without treatment served as the control.

| Flow cytometry
Apoptosis was assessed by Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) via the double-staining method. Briefly, the H9c2 cells were treated with H 2 O 2 for 12 hours, rinsed twice using 4°C precooled PBS, trypsinized and placed into a 15-mL centrifuge tube for centrifugation at 800 g. Following supernatant removal, the precipitates were rinsed twice with PBS, after which the cells were re-suspended in 500-μL binding buffer in accordance with the Annexin V-FITC apoptosis detection kit instructions (BD Biosciences). Afterwards, 5 μL FITC and 5 μL PI were added into the buffer, followed by sufficient mixing. After incubation for 15 minutes, apoptosis was detected using a flow cytometer (BD).

| Reactive oxygen species (ROS) determination
The H9c2 cells or myocardial tissues that had been treated with

| Enzyme linked immunosorbent assay (ELISA)
The H9c2 cells were treated with H 2 O 2 , rinsed twice with pre-cooled PBS and trypsinized. The cell precipitates were collected and homogenized at 4°C with precooled PBS, and subsequently centrifuged at

| Dual-luciferase reporter gene assay
The as an internal reference, the activation degree of the target reporter gene was expressed as the ratio between firefly luciferase and renilla luciferase.

| RNA immunoprecipitation (RIP) assay
RIP assay was performed using the Magna RIP Kit (Millipore) as per the manufacturer's instructions. Briefly, the H9c2 cells were lysed in RIP lysis buffer, then 100 μL of whole cell extract was incubated with RIP buffer containing magnetic beads conjugated with human anti-Ago2 antibody, negative control normal mouse IgG (Millipore). Next, to digest the protein, the samples were incubated with proteinase K with shaking, after which the immunoprecipitated RNA was isolated with the use of phenol-chloroform-isoamyl alcohol (Sigma Aldrich) and finally the RNA pellet was visualized using sodium acetate/ethanolprecipitated together with RNA-grade glycogen (Thermo Scientific) and resuspended in RNase-free water. Finally, the levels of miR-454 and NEDD4-2 mRNA in the precipitates were detected by RT-qPCR.

| Co-immunoprecipitation (Co-IP)
The cells were harvested 48 hours post-transfection and lysed with cell lysis buffer containing protease inhibitor on ice for 30 minutes. One part of the lysate (30 μg of total protein) was used as the input or incubated with 2 μg of NEDD4-2 antibody (anti-rabbit, Sigma), while 2 μg of IgG (anti-rabbit, Sigma) was added to the other part as the NC. Next, 10 μL protein A agarose beads were washed with lysis buffer three times and centrifuged at 3000 rpm for 3 minutes. The beads were subsequently incubated with the cell lysate that had been incubated overnight with the antibody for 3 hours at 4°C, to couple the antibody. Following immunoprecipitation reaction, the agarose beads were centrifuged to the bottom of the tube via centrifugation at 4°C and 3000 rpm for 3 minutes. Following protein concentration determination, 15 μL of 2 × sodium dodecyl sulphate buffer was added to the protein, followed by boiling for 5 minutes, followed by Western blot assay.

| Analysis of protein ubiquitination
Next, 20 μmol/L MG-132, a protease inhibitor, was added to the H9c2 cells following 48 hours of plasmid transfection. After 6 hours of treatment with MG-132, the cells were lysed using a lysis buffer that had been supplemented with protease inhibitor for 30 minutes on ice. The protein immunoprecipitation complex was obtained using 2μg TrkA antibody (anti-rabbit, Sigma). The expression of TrkA (anti-rabbit, 1:100, Sigma) and ubiquitin (anti-rabbit, 1:10 000, Abcam) was analysed by Western blot assay.

| Statistical analysis
SPSS version 21.0 (IBM) was utilized to evaluate data in this study. All data are summarized by the mean ± standard deviation, with P < .05 as a level of statistically significance. Unpaired t-test was conducted for data comparison between two groups, while one-way analysis of variance (ANOVA) was applied for data comparison among multiples in combination with a Tukey's posthoc test. Two-way ANOVA was conducted for data comparison at different time points, while a Pearson's correlation analysis was performed to analyse the correlation between indicators.

| miR-454 is downregulated in HF and negatively correlated with the grade of HF
RT-qPCR was performed to determine the expression of miR-454 in peripheral blood of 72 patients with AMI complicated with HF and 72 healthy individuals (control). In comparison with the peripheral blood of healthy individuals, the expression of miR-454 was notably downregulated in peripheral blood of patients with acute HF ( Figure 1A). The 72 patients with acute HF were classified into grade II, III and IV based on the Killip classification.
Our results also revealed that the expression of miR-454 was negatively correlated with the different grades of HF (P < .05; Figure 1B).

| miR-454 targets NEDD4-2 in H9c2 cells
In an attempt to further identify the mechanism by which miR-454 influences HF, TargetScan was employed to predict the downstream targets of miR-454. Bioinformatics analysis provided data identifying the miR-454 binding sites in NEDD4-2 ( Figure 3A). Hence, we set out to evaluate whether miR-454 could suppress NEDD4-2.
Subsequently, we assessed the expression of miR-454 and NEDD4-2 in patients with HF using Immunohistochemistry, which identified F I G U R E 1 miR-454 is downregulated in HF and negatively correlated with the grade of HF. A, The expression of miR-454 in peripheral blood of HF patients with AMI complicated with HF and healthy individuals determined by RT-qPCR; n = 72; *P < .05. B, miR-454 expression in peripheral blood of healthy individuals and patients with different grades of HF determined by RT-qPCR; *P < .05. Measurement data are displayed as mean ± standard deviation. Independent sample t-tests were used for comparing data between the two groups, and one-way ANOVA for comparison of data among multiple groups

| miR-454 targets NEDD4-2 to repress H9c2 cell apoptosis and injury
The rescue experiments were subsequently performed to ascertain whether miR-454 affects H9c2 cell apoptosis and injury by targeting

| NEDD4-2 induces ubiquitination and degradation of TrkA protein in H9c2 cells
Existing literature has suggested that NEDD4-2 ubiquitinated TrkA and TrkA are aberrantly downregulated in HF tissues. 15,17,21 We subsequently set out to investigate whether NEDD4-2 plays a role in H9c2 cells by regulating TrkA expression. As reflected by immunohistochemistry, a negative correlation was observed between the expression of NEDD4-2 and miR-454 ( Figure 5A) As expected,

| miR-454 overexpression protects H9c2 cells from apoptosis by activating the cAMP pathway
Next, to ascertain whether miR-454 could suppress the apoptosis and injury of H9c2 cells through the NEDD4-2/TrkA/cAMP axis, miR-454 was overexpressed in H 2 O 2 -exposed H9c2 cells,

| miR-454 activates cAMP pathway through NEDD4-2/TrkA axis to delay HF in vivo
The emphasis was subsequently shifted to the involvement of

| D ISCUSS I ON
HF represents a complex disorder that is responsible for a huge societal burden in regard to both cost and overall fatality. 22 Accumulating evidence continues to implicate miRs in the progression of various    with diastolic dysfunction. 8 The overexpression of miR-454 has been speculated to be a promising potential therapeutic target for the alleviation of acute lung injury. 28 We have uncovered that in the Previous research has also revealed that NEDD4-2 can bind to TrkA via a PPXY motif leading to the ubiquitination and downregulation of TrkA in NGF-dependent sensory neurons. 35 The regulatory role of TrkA in cardiovascular diseases has been highlighted in previous research. TrkA has also been shown to prevent cardiomyocyte apoptosis 36 and protect them against oxidative stress. 37 In addition, the activation of TrkA/Akt pathway exerts protective functions against hypoxia/reoxygenation-induced cardiomyocyte apoptosis. 38 During the current study, NEDD4-2-induced TrkA ubiquitination in H9c2 cells to downregulate TrkA, a gene which exerts protection against H9c2 cell apoptosis and injury.
Furthermore, our data revealed that miR-454 regulated the NEDD4-2/TrkA/cAMP axis resulting in the suppression of H9c2 cells apoptosis and oxidative stress injury both in vitro and in vivo.
Accumulating evidence continues to implicate cAMP pathway activation with the attenuation of HF. 39 Diminished cAMP generation is regarded as a contributor to chronic HF with its relation to contractile dysfunction. 40 A previous study emphasized reduced CREB as a responsible entity behind oxidative stress, ultimately facilitating the progression of HF by influencing cardiac growth and apoptosis. 41 Increased formation of cAMP has been reported to be beneficial for chronic HF treatment by maintaining the TH1/ TH2 phenotype. 42 Additionally, the activation of the cAMP/PKA pathway and interaction with the GLP-1 receptor have been reported to be cardioprotective in H9c2 cells exposed to hypoxia/ reoxygenation. 43 Our data suggested that the miR-454/NEDD4-2/ TrkA axis activated cAMP pathway as evidenced by increased expression of cAMP pathway marker genes PKA and CREB, to protect against the apoptosis of H9c2 cells and subsequent myocardial damage. Evidence has previously been presented suggesting that hesperetin-mediated TrkA could activate PKA and CREB in PC12 cells. 18 Additionally, Karkoulias et al 44  In conclusion, the current study provided evidence that miR-454 elevation could target NEDD4-2 to attenuate NEDD4-2-induced ubiquitination of TrkA and activate cAMP pathway for protection against H9c2 cell apoptosis and oxidative stress injury (Figure 9).
These findings highlight the miR-454-mediated NEDD4-2/TrkA/ cAMP axis as a potential cardioprotective mechanism in progression of HF. Nevertheless, further validation through large-scale samples and the use of primary neonatal cardiomyocytes is still required. Additionally, the cell type as the main source of circulating miR-454 requires further investigation. In future investigations, we aim to explore the effects of miR-454 inhibition, TrkA inhibition and NEDD4-2 on the deterioration of cardiac function in non-stress conditions and HF in vivo.

ACK N OWLED G EM ENTS
This study was supported by National Natural Science Foundation of China (No. 81860643).

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.