RhoA rescues cardiac senescence by regulating Parkin-mediated mitophagy

Heart failure is one of the leading causes of death worldwide. RhoA, a small GTPase, governs actin dynamics in various tissue and cell types, including cardiomyocytes; however, its involvement in cardiac function has not been fully elucidated. Here, we generated cardiomyocyte-specific RhoA conditional knockout (cKO) mice, which demonstrated a significantly shorter lifespan with left ventricular dilation and severely impaired ejection fraction. We found that the cardiac tissues of the cKO mice exhibited structural disorganization with fibrosis and also exhibited enhanced senescence compared with control mice. In addition, we show that cardiomyocyte mitochondria were structurally abnormal in the aged cKO hearts. Clearance of damaged mitochondria by mitophagy was remarkably inhibited in both cKO cardiomyocytes and RhoA-knockdown HL-1 cultured cardiomyocytes. In RhoA-depleted cardiomyocytes, we reveal that the expression of Parkin, an E3 ubiquitin ligase that plays a crucial role in mitophagy, was reduced, and expression of N-Myc, a negative regulator of Parkin, was increased. We further reveal that the RhoA–Rho kinase axis induced N-Myc phosphorylation, which led to N-Myc degradation and Parkin upregulation. Re-expression of Parkin in RhoA-depleted cardiomyocytes restored mitophagy, reduced mitochondrial damage, attenuated cardiomyocyte senescence, and rescued cardiac function both in vitro and in vivo. Finally, we found that patients with idiopathic dilated cardiomyopathy without causal mutations for dilated cardiomyopathy showed reduced cardiac expression of RhoA and Parkin. These results suggest that RhoA promotes Parkin-mediated mitophagy as an indispensable mechanism contributing to cardioprotection in the aging heart.

Heart failure is one of the leading causes of death worldwide. RhoA, a small GTPase, governs actin dynamics in various tissue and cell types, including cardiomyocytes; however, its involvement in cardiac function has not been fully elucidated. Here, we generated cardiomyocyte-specific RhoA conditional knockout (cKO) mice, which demonstrated a significantly shorter lifespan with left ventricular dilation and severely impaired ejection fraction. We found that the cardiac tissues of the cKO mice exhibited structural disorganization with fibrosis and also exhibited enhanced senescence compared with control mice. In addition, we show that cardiomyocyte mitochondria were structurally abnormal in the aged cKO hearts. Clearance of damaged mitochondria by mitophagy was remarkably inhibited in both cKO cardiomyocytes and RhoA-knockdown HL-1 cultured cardiomyocytes. In RhoA-depleted cardiomyocytes, we reveal that the expression of Parkin, an E3 ubiquitin ligase that plays a crucial role in mitophagy, was reduced, and expression of N-Myc, a negative regulator of Parkin, was increased. We further reveal that the RhoA-Rho kinase axis induced N-Myc phosphorylation, which led to N-Myc degradation and Parkin upregulation. Re-expression of Parkin in RhoA-depleted cardiomyocytes restored mitophagy, reduced mitochondrial damage, attenuated cardiomyocyte senescence, and rescued cardiac function both in vitro and in vivo. Finally, we found that patients with idiopathic dilated cardiomyopathy without causal mutations for dilated cardiomyopathy showed reduced cardiac expression of RhoA and Parkin. These results suggest that RhoA promotes Parkinmediated mitophagy as an indispensable mechanism contributing to cardioprotection in the aging heart.
Heart failure is a multifaceted disease with a complex etiology. It remains a major public health problem and is the leading cause of death worldwide, with high morbidity and mortality rates (1)(2)(3)(4). Heart failure is a chronic pathophysiological state in which the heart muscle is unable to pump an adequate supply of blood to the whole body due to progressive loss of myocardial contractile function over time (5). Despite medical advances, the prognosis of patients with heart failure remains poor (6), and current therapeutic approaches seem palliative as the underlying mechanisms contributing to heart failure are still not fully addressed.
The heart is a highly metabolic organ in which mitochondrial dynamics are precisely regulated to ensure optimal mitochondrial function (7,8). Given the high energetic demand of the heart, age-related defects in mitochondrial bioenergetics can have detrimental effects on normal cardiac pumping. Accumulation of dysfunctional mitochondria is associated with suppression of mitophagy (9,10), leading to a defect in mitochondrial quality control. Mitophagy is an evolutionarily conserved mechanism that plays a crucial role in the mitochondrial quality control (11). It enables the degradation of damaged and superfluous mitochondria in response to cardiac stress, including senescence.
RhoA is a small GTPase that regulates diverse cellular events, including actin cytoskeleton organization, cell adhesion, migration, invasion, apoptosis, extracellular matrix remodeling, and smooth muscle contractility (12,13). RhoA is ubiquitously expressed in almost all tissue and cell types, including cardiomyocytes. RhoA signaling plays a pivotal role in processes leading to cardiovascular diseases, such as pulmonary hypertension, vasospastic angina, and heart failure (14,15). Thus, RhoA function in the heart remains an interesting focus among molecular cardiologists as well as biologists. However, the understanding of the molecular signaling of RhoA in the heart is still incomplete.
In this study, we found that cardiomyocyte-specific RhoA conditional knockout (cKO) mice had a significantly shorter lifespan with features of early senescence, severely impaired cardiac function, and build-up of many structurally disorganized and enlarged mitochondria compared with control mice. These phenotypes suggest a causative link between cardiac aging and mitochondrial dysfunction with regard to RhoA signaling. We further revealed the molecular mechanisms of cardiac RhoA in regulating mitochondrial dynamics, which may protect the heart from senescence-mediated dysfunction.

Deterioration of cardiac function and early death in RhoA cKO mice
RhoA cKO mice were healthy at birth with normal growth. The mice were similar in weight, and no obvious phenotypic abnormalities were observed at around 10 weeks after birth, compared with the littermate control mice (Fig. 1A). However, after 10 weeks of age, the body weight of RhoA cKO mice did not increase further and was significantly lower than that of control mice. RhoA expression was confirmed to be absent in cardiomyocytes from RhoA cKO mice (Fig. 1B). Strikingly, RhoA cKO mice experienced early death from around 30 weeks of age compared with control mice (Fig. 1C). To investigate the cause of early death in RhoA cKO mice, we assessed the cardiac function of these mice by echocardiography. The left ventricular ejection fraction of RhoA cKO mice was initially normal after birth, but it decreased significantly with age ( Fig. 1, D and E). In addition to the lower left ventricular ejection fraction (LVEF), LV dilatation and increased LV mass without LV wall thickening during the experimental period were observed in the RhoA cKO hearts compared with the control hearts ( Fig. 1, F-H), suggesting age-dependent cardiomyopathy caused by loss of RhoA in the heart. We also measured heart rate (HR) and blood pressure (BP). HR was similar between RhoA cKO mice and control mice, while RhoA cKO mice exhibited an age-dependent lower systolic BP than control mice (Fig. 1, I and J). Collectively, these results indicate the severe low cardiac output condition and an accelerated transition to heart failure in RhoA cKO mice, resulting in a shorter lifespan.

Accelerated aging and fibrosis in the RhoA cKO heart
To explore how cardiac function rapidly declined with aging in RhoA cKO mice, we examined the progression of cardiac senescence using several markers because cardiac senescence impairs cardiac function (16,17). Cellular senescence markers, including p16, p21, and senescence-associated β-galactosidase, were more highly detected in the RhoA cKO hearts than the control hearts (Fig. 2, A-F). Consistent with these results, the histological analysis by hematoxylin and eosin (H-E) staining revealed severe myocardial pathology, including increased myofiber disarray and interstitial space in LV of aged RhoA cKO mice (Fig. 2G). The RhoA cKO hearts also exhibited significantly augmented LV fibrosis (Fig. 2, H and I). In agreement with echocardiography, these results suggest that cardiac RhoA depletion accelerates cardiac aging and induces cardiac structural changes with abnormally increased fibrosis.

Mitochondrial dysfunction and mitophagy dysregulation in the RhoA cKO heart
Mitochondrial dynamics in the heart are closely related to aging (7), and abnormal mitochondrial dynamics result in an insufficient energy supply in the heart, suppressing cardiac function (8,10). Thus, we examined the morphology of the mitochondria in the heart by transmission electron microscopy (TEM). The mitochondria in the RhoA cKO hearts were severely damaged by aging, which occurred in parallel with the accumulation of many swollen and fragmented mitochondria with cristae disruption (Fig. 3A). We also found that the expression of ATP5A, a subunit of the mitochondrial ATP synthase, decreased in the heart of RhoA cKO mice compared with control mice (Fig. 3, B-E), validating the functionally defective mitochondria that resulted from RhoA knockout. Next, to examine the effect of RhoA on mitochondrial function in in vitro experiments, RhoA expression was knocked down in HL-1 cardiomyocytes. When two siRNAs for RhoA were transfected into HL-1 cells to check their efficiency for RhoA inhibition, siRhoA #2 significantly reduced RhoA expression, while siRhoA #1 did not. Thus, siRhoA #2 was used for further experiments (Fig. 3, F-H). Similar to the RhoA cKO hearts, ATP5A expression was significantly reduced in RhoAknockdown HL-1 cardiomyocytes (Fig. 3, I-L).
Mitochondrial function and homeostasis are mainly regulated by (1) mitophagy and (2) fission and fusion (18,19). Mitophagy is an important regulatory mechanism for clearing damaged mitochondria by proteasomal degradation. We first detected impaired mitophagy in siRhoA-treated HL-1 cells compared with scramble RNA-treated cells after exposure to carbonyl cyanide m-chlorophenyl hydrazone (Fig. 3, M and N). These results suggest defective mitophagy regulation in the absence of RhoA, leading to impaired removal and abnormal accumulation of damaged mitochondria in the heart in response to cardiac stress, such as aging. In contrast, the expression and phosphorylation of the mitochondrial fission marker Drp1 were not different between the RhoA cKO and control hearts or between RhoA-knockdown and control HL-1 cells (Fig. S1), suggesting that mitochondrial biogenesis in cardiomyocytes is normal, regardless of the absence of RhoA.

Reduced Parkin expression and ubiquitinated mitochondrial proteins by loss of RhoA in cardiomyocytes
To delineate the mitochondrial abnormality and mitophagy dysregulation in cardiomyocytes with loss of RhoA, we focused on Parkin, an E3 ubiquitin (Ub) ligase, which mediates the ligation of Ub to the damaged mitochondria for proteasomal degradation (20). The loss of RhoA resulted in the reduction of Parkin expression with a decrease in ubiquitinated mitochondrial proteins in the hearts of younger (18-week-old) and older (55-week-old) mice (Fig. 4, A-F). Similarly, Parkin expression and ubiquitinated mitochondrial proteins in HL-1 cells were suppressed by RhoA knockdown (Fig. 4, G-L). These data suggest that RhoA plays a role in the expression of Parkin in cardiomyocytes, which regulates ubiquitination of mitochondrial proteins.
Parkin is phosphorylated, and its function is regulated by PTEN-induced putative kinase 1 (PINK1) (21). We then examined PINK1 expression in the presence and absence of RhoA in the mouse heart. PINK1 protein expression was almost identical between control and RhoA cKO hearts and was not different between young (18-week-old) and old (53week-old) mice (Fig. S2, A and B). Similar to this, the gene expression of Park6, encoding PINK1, as well as the protein expression of PINK1 was not changed by RhoA knockdown in HL-1 cardiomyocytes, as shown by quantitative PCR and Western blotting (Fig. S2, C-E). Immunostaining of HL-1 cells also showed that the PINK1-positive area after siRhoA transfection was equal to that after scramble RNA transfection (Fig. S2, F and G). Thus, RhoA does not seem to affect PINK1 expression in the heart.       RhoA-mediated N-Myc-Parkin pathway regulation N-Myc is a negative transcription factor for the Parkin gene expression (22), and the expression of N-Myc is reduced by phosphorylation-dependent degradation (23). We determined the endogenous expression of N-Myc in both the mouse heart and HL-1 cells. By depletion of RhoA, the expression of N-Myc was increased (Fig. 5, A-F), together with a remarkable reduction of its phosphorylation (Fig. 5, A-D), indicating an inverse correlation between N-Myc and Parkin expressions. To further examine how RhoA regulates N-Myc phosphorylation, we focused on Rho kinase (ROCK), which is an effector of RhoA (14). ROCK in HL-1 cells was confirmed to be inhibited by treatment with a ROCK inhibitor Y-27632 (Fig. 5G). In the presence of Y-27632, N-Myc phosphorylation was decreased, and N-Myc expression was increased, resulting in the reduction of Parkin and ATP5A expressions (

Restoration of mitophagy and cardiac function by supplementation of Parkin expression in the RhoA-depleted cardiomyocytes
To demonstrate the essential effects of Parkin on the rescue of mitophagy and cardiac function in the RhoA cKO hearts and RhoA-knockdown cardiomyocytes, we used the adenoassociated virus (AAV) serotype 6 gene transfer system to introduce the Parkin gene in cardiomyocytes. First, we infected AAV-Parkin-T2A-green fluorescent protein (GFP) and control AAV-GFP into HL-1 cells and examined how the infection increased Parkin expression. AAV-Parkin-T2A-GFP infection recovered the siRhoA-mediated decrease in Parkin expression back to the basal level (Fig. 6, A and B). Similarly, AAV-Parkin-T2A-GFP restored mitochondrial protein ubiquitination and mitophagy ( Fig. 6, C-F). In the fluorescence microscopy, we found that all of the HL-1 cells were infected with AAV-Parkin-T2A-GFP or AAV-GFP as monitored by GFP fluorescence, although the level of GFP fluorescence was variable in each cell (Fig. 6E).
Next, we examined the in vivo function of AAV-Parkin-T2A-GFP in RhoA cKO mice by intravenous injection of AAV through the tail vein. When AAV-Parkin-T2A-GFP was administered in control mice to assess the in vivo efficiency of the AAV-mediated Parkin gene transfer, Parkin expression was increased in the heart compared with the administration of AAV-GFP (Fig. S3, A and B). In contrast, the increase was not observed in the brain (Fig. S3A), confirming adequate gene transfer by the AAV serotype 6 system. Four weeks after AAV infection in mice, cardiac function was unchanged, as evaluated by LVEF and hemodynamics, such as HR and systolic BP (Fig. S3C). This suggests that Parkin gene was safely transferred by AAV in vivo. Because Parkin expression in the heart was restored for approximately 25 weeks after AAV-Parkin-T2A-GFP administration in RhoA cKO mice (Fig. S3D), we injected AAV in RhoA cKO mice twice (10 and 32 weeks after birth) for a total 1-year (53 weeks) observation period. After the injection of AAV-Parkin-T2A-GFP, the deterioration of LVEF was attenuated, and the lifespan was prolonged compared with mice injected with AAV-GFP (Fig. 7, A and B). Mice were sacrificed at around 55 weeks after birth. The heart was enlarged in RhoA cKO mice injected with control AAV-GFP due to heart failure, which was clearly recovered by AAV-Parkin-T2A-GFP injection (Fig. 7, C and D). Similarly, the lung weight, which was also increased by heart failureinduced pulmonary edema in RhoA cKO mice, was reduced by AAV-Parkin-T2A-GFP injection (Fig. 7D). The treatment maintained the expression of Parkin in the RhoA cKO hearts, and the results were comparable to those of the control hearts ( Fig. 7, E-H). H-E staining showed an improvement of the severe myocardial damage in the RhoA cKO hearts after AAV-Parkin-T2A-GFP injection (Fig. 7I). The increased cardiac fibrosis in the RhoA cKO hearts was also attenuated by the injection (Fig. 7, J and K).
Further ultrastructural analysis using TEM revealed a remarkable reduction of damaged mitochondria in the RhoA cKO hearts after treatment with AAV-Parkin-T2A-GFP (Fig. 8A). This was justified by the restored expression of ATP5A and the increase in mitochondrial protein ubiquitination in the AAV-Parkin-T2A-GFP-treated RhoA cKO hearts (Fig. 8, B-G). Cellular senescence in RhoA cKO cardiomyocytes was also suppressed by AAV-Parkin-T2A-GFP treatment (Fig. 8, H and I). These findings suggest that Parkin, as the downstream molecule of RhoA, could compensate for RhoA deficiency by maintaining mitochondrial homeostasis, resulting in the prevention of age-related acceleration of cardiac dysfunction and heart failure in the absence of RhoA.
Considering the clinical implication of Parkin expression supplementation in patients with heart failure caused by reduction or loss of RhoA, we additionally examined the effect M, fluorescence images of mitophagy in viable HL-1 cells after CCCP induction. Nuclei were counterstained with Hoechst. Scale bar: 20 μm. N, summary graph of the percentage of mitophagy area. The data in each graph are shown as the mean ± SD. Comparisons of the data between groups were performed using one-way ANOVA (C and E) or t test (F, H, J, L, and N). ***p < 0.001 versus control or scramble; † † † p < 0.001 versus week 18. cKO, conditional knockout; TEM, transmission electron microscopy. of the Parkin gene transfer on RhoA cKO mice when cardiac function was mildly impaired. After AAV-Parkin-T2A-GFP was administered once in 30-week-old RhoA cKO mice, the impairment of LVEF tended to be prevented for 15 weeks after the administration (Fig. S4A), and the lifespan was significantly prolonged compared with after AAV-GFP administration (Fig. S4B). Parkin expression in 50-week-old mice after a single AAV-Parkin-T2A-GFP administration was higher than after AAV-GFP administration, whereas the expression was lower than after AAV-Parkin-T2A-GFP administration twice (Fig. S4, C-F). The histological analysis showed that dysregulation of myocardial tissue and the degree of fibrosis were similar between mice treated with AAV-Parkin-T2A-GFP and mice treated with AAV-GFP (Fig. S4, G-I). However, cardiac senescence determined by senescence-associated β-galactosidase and mitochondrial function evaluated by ATP5A expression were significantly improved after a single administration of AAV-Parkin-T2A-GFP compared with AAV-GFP  (Fig. S4, J-O), suggesting the benefit of the Parkin gene transfer in RhoA cKO mice even after mild heart failure begins.

Reduced RhoA expression in aged patients with idiopathic dilated cardiomyopathy
There is no definitive knowledge about the RhoA expression in aged patients who suffer from heart failure without known hereditary gene mutations. In this context, we examined the cardiac RhoA expression in the heart samples obtained from adult patients with severe heart failure caused by idiopathic dilated cardiomyopathy (DCM). The heart samples were obtained at the time of heart transplantation. The clinical characteristics of the patients are shown in Table 1. Because all patients underwent LV assist device (LVAD) implantation prior to heart transplantation, the data in Table 1 were obtained just before LVAD implantation. The average period between LVAD implantation and heart transplantation was 4.4 ± 0.9 years. Cardiac RhoA expression was significantly decreased in patients with idiopathic DCM compared with control subjects (average age: 38.5 ± 10.8 years; male/female (n): 12/3) who died accidentally without cardiovascular diseases (Fig. 9, A-C). Concomitant with these results, the significant reduction of Parkin expression was observed in the hearts of patients with DCM (Fig. 9, A-C). These findings may validate our hypothesis that reduced RhoA expression in the heart attenuates Parkin expression, not only in mice but also in humans. In contrast to RhoA and Parkin expressions, PINK1 expression in the heart was almost equal between DCM patients and control subjects (Fig. 9, B and C). We confirmed the severe myocardial damage and fibrosis in patients with DCM by histological analysis (Fig. 9, D-F). TEM also detected an abundance of disrupted mitochondria in the hearts of patients with DCM compared with control subjects (Fig. 9G). Finally, we found the significant decrease in ATP5A expression in the hearts of patients with DCM (Fig. 9, H and I). These results support our conclusion that RhoA plays a role in cardiac mitochondrial function via Parkin and that the defect of RhoA expression results in mitophagy dysregulation, leading to accelerated cardiac senescence and heart failure.

Discussion
This study provides an important insight into the function of RhoA in the aging heart, as well as the molecular mechanism by which RhoA regulates cardiac function through Parkin-mediated mitochondrial homeostasis. RhoA cKO mice showed earlier death from around 30 weeks of age and a dramatic reduction of LVEF with accelerated senescence and age-dependent cardiac fibrosis. Concomitant with the severe deterioration of cardiac function, we found that loss of RhoA in the heart induced excess accumulation of severely damaged mitochondria in cardiomyocytes. In patients with idiopathic DCM who had no hereditary gene mutations, both RhoA and Parkin expressions in the heart were markedly reduced, and the morphology of the cardiac mitochondria was disturbed. This suggests that RhoA has cardioprotective effects and is crucial for the maintenance of healthy mitochondria, resulting in the prevention of heart failure with aging.
In support of our findings, another research group has recently reported that in myocardial infarction, cardiac RhoA signaling plays a role in mitochondrial quality control by regulating the function and expression of Parkin and PINK1, a protein kinase that phosphorylates and activates Parkin (24). In our study, we further revealed the mechanism of RhoA in the regulation of Parkin expression through N-Myc in cardiomyocytes. N-Myc is a member of the Myc family. It is a transcription factor that is critically involved in diverse physiological and pathological events, including neuronal development and tumor progression (25,26). This protein binds to the E-box motif at the Parkin transcription initiation site and transcriptionally inhibits Parkin expression in neuroblastoma cell lines (22). In this study, we observed downregulation and upregulation of N-Myc expression in RhoA-intact and RhoAdepleted cardiomyocytes, respectively. N-Myc expression has also been shown to be suppressed by its phosphorylation and subsequent ubiquitination (23,27). GSK-3β was identified to be a kinase that phosphorylates N-Myc, but other kinases that contribute to N-Myc phosphorylation have not been well documented. Using a ROCK inhibitor Y-27632, we discovered that ROCK, which is an effector of RhoA, functions as a kinase that phosphorylates N-Myc to reduce its expression. Thus, we propose that the RhoA-ROCK axis negatively regulates N-Myc to maintain sufficient Parkin expression in cardiomyocytes.
As for ROCK, there are two isoforms ROCK1 and ROCK2, and the disruption of both ROCK isoforms has been reported to be cardioprotective by promoting autophagy and reducing cardiac fibrosis during aging (28). ROCKs are well-known effectors of RhoA, while other proteins, such as mDia, also function downstream of RhoA. Inhibition of mDia in the heart markedly suppressed the cardiac function and induced heart failure (29). In addition, a single deletion of ROCK2 in cardiomyocytes was profibrotic and reduced autophagy (28), suggesting that only ROCK1 deletion is favorable for cardiomyocytes and overwhelms the ROCK2 deletion-mediated deteriorative cardiac phenomena. Collectively, because RhoA regulates a variety of molecules including ROCKs, it might be reasonable that the phenomena observed in RhoA cKO mice are different from those in mice in which double cardiac ROCKs are ablated.
Mutations in the Parkin gene are intimately related to familial Parkinson's disease (PD) (30). PD is the common neurodegenerative disorder that involves loss of dopaminergic examined in (J). The data in each graph are shown as the mean ± SD. In (A), two-way ANOVA and one-way ANOVA were applied to compare the data between groups and weeks, respectively, and in (C), the data were analyzed using the Kaplan-Meier method. One-way ANOVA (D, F, and H) or t test (K) was used to compare the data between groups. *p < 0.05, **p < 0.01, and ***p < 0.001 versus AAV-GFP; † p < 0.05, † † p < 0.01, and † † † p < 0.001 versus Week 9; § § § neurons in the substantia nigra (31,32). In addition to neuronal system dysfunction in PD, PD is associated with the risk of cardiovascular disease, including congestive heart failure (33). Although the role of Parkin in the brain has been extensively studied (34,35), the understanding of Parkin regulation in the aging heart downstream of RhoA remains elusive. Mitophagy is the system that clears the damaged mitochondria in various cell types, including cardiomyocytes, and is fundamental for constitutive mitochondrial housekeeping to maintain cardiac homeostasis and prevent heart failure (36). Several reports have demonstrated the pathophysiological importance of mitophagy in the heart, in which Parkin exerts cardioprotection in response to ischemic stress (37,38). In addition, PINK1 contributes to the maintenance of cardiac function because PINK1 knockout mice develop LV dysfunction and pathological cardiac hypertrophy with impaired mitochondrial function (39). Although the present study showed that loss of RhoA in cardiomyocytes attenuated Parkin expression, PINK1 expression was not changed. Moreover, the mitochondrial fission marker Drp1 and its phosphorylated form were not disturbed in the RhoA cKO hearts and RhoA-knockdown HL-1 cells. These findings suggest that RhoA specifically regulates Parkin in cardiomyocytes, independent of PINK1, and that it does not affect mitochondrial biogenesis. Parkin is a cytosolic E3 Ub ligase that selectively ubiquitinates proteins located on dysfunctional mitochondria for mitophagy (40,41). To prevent unnecessary cell death, dysfunctional mitochondria, which are harmful to cells, should be cleared, and mitophagy is one of the systems responsible for this clearance. In our study, the hearts from 18-week-old RhoA cKO mice had normal mitochondria, while the hearts from 55week-old RhoA cKO mice had swollen and disorganized mitochondria with broken cristae, which was quite different from the hearts from 55-week-old control mice that had morphologically normal mitochondria. Our data suggest that RhoA deficiency in the heart causes a defect in the clearance of dysregulated mitochondria due to reduced Parkin expression. Similar to RhoA cKO mice in the present study, young 12-week-old Parkin −/− mice had normal cardiac function under baseline conditions in a previous study. However, Parkin −/− mice were quite sensitive to the cardiac stress induced by myocardial infarction (42). According to another previous study, mitochondrial DNA mutations in mice accelerated cardiac aging, and overexpression or deletion of Parkin in the mice did not rescue or worsen the cardiac phenotype (43). These results differ from ours; however, different mouse models may demonstrate different degrees of mitochondrial damage and different regulatory mechanisms to compensate for the defect in mitophagy in the aged heart, which may affect the rate of transition to cardiomyopathy.
One advantage of our study is that we demonstrated the significant reduction of both RhoA and Parkin expressions in patients with DCM compared with normal subjects. Although it might be difficult to strictly identify which of RhoA or Parkin reduction is the primary and specific cause of DCM, it is possible to interpret that RhoA is involved in cardiac homeostasis cooperatively with Parkin. Mitochondrial morphology and function as well as mitophagy were disturbed in patients with DCM in the present study. Furthermore, several novel functions of RhoA, which are mediated by Parkin, were observed not only in the mouse heart but also in the human heart. To date, several gene mutations associated with heart failure have been listed (44,45). Loss or mutation of the RHOA and PARK2 genes can be added to the list in line with the findings from our and other research groups.
In conclusion, we showed the functional role of RhoA in regulating Parkin expression through ROCK and N-Myc and Parkin-dependent mitophagy for the clearance of damaged mitochondria in the heart, resulting in the maintenance of mitochondrial homeostasis and prevention of cardiac senescence (Fig. S5). Thus, we conclude that loss of RhoA in the heart induces heart failure due to early cardiac senescence and cardiomyopathy. Further understanding of RhoA signaling in the aged heart will help to develop future therapies for the prevention and treatment of heart failure.

Generation of RhoA cKO mice
RhoA-floxed mice (RhoA fl/fl : C57BL/6 background), in which exon 3 of the Rhoa gene was flanked by loxP sites, were generated and used in our previous study (46). The mice were then mated with C57BL/6 mice expressing Cre recombinase under the control of the α-myosin heavy chain promoter (Myh6-Cre; Jackson Laboratory) to generate cardiomyocytespecific RhoA cKO mice. In the Myh6-Cre mice, Cre exerts its recombination activity specifically in cardiomyocytes but not in other tissues such as the liver, lung, skeletal muscle, and spleen (47), and the recombinase functions from embryonic day 9.5 (48). Mice harboring RhoA fl/fl alleles alone were used as controls. The mice were housed in specific pathogen-free conditions at the Research Centre for Animal Life Science of Shiga University of Medical Science. All animal protocols were in accordance with institutional guidelines, including Animal Research Reporting of In Vivo Experiments (ARRIVE) Human heart sample collection All protocols using human heart samples were approved by the Research Ethics Committee of Osaka University Graduate School of Medicine and Shiga University of Medical Science and conformed to the principles of the Declaration of Helsinki. Heart tissues were obtained from (1) subjects who died accidentally without cardiovascular diseases and were sent to Division of Legal Medicine, Shiga University of Medical Science, for forensic autopsy and (2) patients with idiopathic DCM at the time of heart transplantation during the period of November 2017 through June 2021. All of the patients provided written informed consent for the use of heart tissues in this study.

BP measurement
Arterial BP and HR of conscious mice were assessed using the noninvasive plethysmographic tail-cuff method (model BP-98-AL; Softron). Mice were weighed and warmed at 37 C in a cylindrical thermostat supplemented on the BP-98-AL machine before and during the assessment. Measurements were taken at 2-min intervals, and an average of five BP and HR measurements was taken as the true BP and HR of each mouse, respectively.

Histological analysis of the heart
Fresh mouse and human hearts were fixed with 4% paraformaldehyde and 10% formaldehyde, respectively, followed by embedding in paraffin blocks overnight. Otherwise, the hearts were frozen in water-soluble medium (Surgipath FCS22; Leica Biosystems) in liquid nitrogen. Formalin-fixed paraffinembedded heart tissues were sectioned at 4-μm thickness using a microtome (Leica Biosystems). The frozen heart tissues were sectioned at a thickness of 10 μm using a cryostat (Leica CM1800; Leica Biosystems) at −20 C. The sections were layered on poly-L-lysine-coated slides. The formalin-fixed paraffin-embedded heart sections were deparaffinized before being subjected to H-E or Picro-Sirius red staining (51).

Confocal microscopy
Cells on poly-L-lysine-coated cover slides were fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100 in phosphate-buffered saline (PBS) for 30 min at 37 C. Primary antibodies were applied in 2% bovine serum albumin (BSA) plus 1% or 3% skimmed milk in PBS overnight, followed by a 1-h incubation with the fluorescent dye-labeled secondary antibody. The cells were imaged using the Leica SP8 X confocal microscope (Leica Microsystems). Similar staining techniques were performed on cross-sections of the frozen hearts. The percentage of positive area in the images was quantified using ImageJ software (National Institute of Health). After converting the composite fluorescent image (with three colors) into individual RGB images, the individual threshold level for each fluorescent marker was determined to generate the percentage of fluorescence positive area for each marker.

TEM
Mouse and human hearts were fixed with 2.5% glutaraldehyde in 0.1 mol/l cacodylate buffer; postfixed in 1% osmium tetroxide; treated with 0.5% tannic acid, and 1% sodium sulfate; cleared in 2-hydroxypropyl methacrylate; and embedded in LX112 (Ladd Research). Sections were mounted on copper slot grids coated with parlodion and stained with uranyl acetate and lead citrate for examination on the H-7500 electron microscope (Hitachi High-Tech Corporation).

Mitophagy assay
Viable cells were stained with 100 nmol/l Mitophagy Dye (Dojindo Laboratories) for 30 min and washed in Hank's Hepes buffer solution. The attached cells were stimulated with 10 μmol/l carbonyl cyanide m-chlorophenyl hydrazone (Nacalai Tesque) for 24 h before observation, as described in the manufacturer's protocol. Fluorescent images were obtained using the Leica SP8 X confocal microscope.

Cellular senescence detection
Frozen sections were fixed with 4% paraformaldehyde for 3 min and incubated with SPiDER-βGal solution (Dojindo Laboratories) at 37 C for 30 min. After removing the solution, the sections were washed with PBS and mounted with mounting medium including DAPI (Vector Laboratories).

Isolation of the mitochondrial fraction
Cells were washed in ice-cold PBS and resuspended in subcellular fractionation buffer containing 20 mmol/l Hepes (pH 7.4), 10 mmol/l KCl, 2 mmol/l MgCl 2 , 200 mmol/l sucrose, 1 mmol/l ethylenediaminetetraacetic acid, 1 mmol/l ethyleneglycol tetraacetic acid, 2 mmol/l phenylmethylsulfonyl fluoride (PMSF), and 1 mg/l leupeptin (52). The hearts extracted from mice were washed with PBS, transferred into the subcellular fractionation buffer, and homogenized in the buffer with 15 strokes using the Potter-Elvehjem tissue homogenizer (DWK Life Sciences). Cell and heart samples were then passed through a 26-gauge needle attached to a 1-ml syringe ten times for lysis, followed by centrifugation at 800g at 4 C for 5 min. The supernatant including the mitochondria was transferred into a new tube and centrifuged at 10,000g at 4 C for 5 min. The pellets were resuspended in radioimmunoprecipitation assay buffer containing 50 mmol/l Tris-HCl (pH 7.5), 150 mmol/l NaCl, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 1% Nonidet P-40, 1 mmol/ l PMSF, and 1 μg/ml leupeptin to obtain the mitochondrial fraction.

Western blotting
Mouse and human hearts were homogenized mechanically in radioimmunoprecipitation assay (RIPA) buffer containing 50 mmol/l Tris-HCl (pH7.5), 150 mmol/l NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 1% Nonidet P-40, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 mmol/l PMSF, 5 mmol/l NaF, and 1 mmol/l Na 3 VO 4 . HL-1 cells were also lysed in RIPA buffer. The homogenates and lysates were centrifuged at 14,000 rpm for 15 min, and the supernatant was used for further analysis. After the protein concentration in the supernatant was measured by Quick Start Bradford (Bio-Rad Laboratories) using BSA standards, 10 μg of protein samples were separated by 10% or 12% SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane (Bio-Rad Laboratories). The membrane was then blocked for 1 h at room temperature in 5% BSA or 5% skimmed milk in Trisbuffered saline with Tween 20. The membrane was incubated with primary antibody overnight in 5% skimmed milk at 4 C, followed by incubation with horseradish peroxidase (HRP)-labeled secondary antibody (GE Healthcare) for 1 h in 5% skimmed milk. The membrane was incubated with HRP substrate (Luminata Forte) for 5 min and observed on a luminescent image analyzer (Fusion Solo 6S Edge; Vilber Bio imaging). The band densities were analyzed using ImageJ software.

Primary antibodies
The detailed information of the primary antibodies used in this study is summarized in Table S1.

ROCK kinase assay
HL-1 cells (6 × 10 5 cells) were treated with or without 10 μmol/l Y-27632 for 1 h and were lysed in RIPA buffer. After centrifugation at 16,000g, the clear supernatant was applied to Cyclex Rho-kinase Assay kit (Medical & Biological Laboratories) for measurements of the kinase activity. The procedures were carried out according to the manufacturer's instructions, and the optical absorbance was measured at 450 nm with MultiSkan JX (Thermo Fisher Scientific). The background-subtracted values were used for data presentation.

AAV serotype 6-mediated Parkin expression
Viral particles containing the AAV serotype 6 vector harboring the Parkin and EGFP genes linked with the T2A sequence (AAV-Parkin-T2A-GFP) driven by the cytomegalovirus promoter were generated using Vector Builder. The AAV serotype 6 vector carrying only the EGFP gene (AAV-GFP) was similarly generated and used as the control. For the recombinant AAVs manufacturing, the plasmid carrying the cDNA of Parkin-T2A-GFP or GFP was transfected into HEK293T packaging cells, together with Rep-cap plasmid and helper plasmid (Vector Builder) encoding adenovirus genes (E4, E2A, and VA) that mediate AAV replication. After a short incubation period, viral particles were harvested from the cell lysate and concentrated by polyethylene glycol precipitation. The viral particles were further purified and concentrated by cesium chloride gradient ultracentrifugation. For measurements of the AAVs titer, digital PCR-based approach was applied. Parkin and GFP expressions in HL-1 cells and mouse hearts were performed by infecting the above viral particles. For the administration of AAV into mice, 1 × 10 11 viral particles of AAV-Parkin-T2A-GFP or AAV-GFP were intravenously injected through the mouse tail vein after mice were anesthetized with 2% isoflurane. Following AAV serotype 6 injection, HL-1 cells, as well as mouse hearts and lungs, were harvested and isolated, respectively, at the appropriate time points for further analysis.

Statistical analysis
All numerical values are shown as the mean ± standard deviation. All experiments were performed at least three times independently. Statistical differences between experimental groups were evaluated by two-tailed unpaired Student's t test or one-way or two-way analysis of variance with Bonferroni's multiple comparison test. The Kaplan-Meier analysis was conducted to evaluate the lifespan of mice in the two groups. The survival curves were compared using the log-rank test. For all analyses, p < 0.05 was considered statistically significant.

Data availability
All of the data are contained within the article and are available from the corresponding author on reasonable request.
Supporting information-This article contains supporting information.