Decreased autophagy: a major factor for cardiomyocyte death induced by β1-adrenoceptor autoantibodies

Cardiomyocyte death is one major factor in the development of heart dysfunction, thus, understanding its mechanism may help with the prevention and treatment of this disease. Previously, we reported that anti-β1-adrenergic receptor autoantibodies (β1-AABs) decreased myocardial autophagy, but the role of these in cardiac function and cardiomyocyte death is unclear. We report that rapamycin, an mTOR inhibitor, restored cardiac function in a passively β1-AAB-immunized rat model with decreased cardiac function and myocardial autophagic flux. Next, after upregulating or inhibiting autophagy with Beclin-1 overexpression/rapamycin or RNA interference (RNAi)-mediated expression of Beclin-1/3-methyladenine, β1-AAB-induced autophagy was an initial protective stress response before apoptosis. Then, decreased autophagy contributed to cardiomyocyte death followed by decreases in cardiac function. In conclusion, proper regulation of autophagy may be important for treating patients with β1-AAB-positive heart dysfunction.

Heart dysfunction is the terminal stage of various cardiovascular diseases, and it is characterized by a complicated etiology and high mortality. Recent studies indicate that cardiomyocyte death was a leading contributor to the development of heart dysfunction. 1 Because systolic and diastolic function is directly affected by myocardial cell loss, understanding how cardiomyocyte death occurs will inform treatment strategies to prevent or treat heart dysfunction.
Since the 1990s, studies have revealed that diverse cardiovascular diseases are correlated to anti-β 1 -adrenergic receptor autoantibodies (β 1 -AABs). 2,3 We reported that β 1 -AABs were induced by myocardial remodeling in heart dysfunction, 4 and that its long-term presence significantly decreased cardiac function in vivo. 5 β 1 -AABs also caused cell death of cultured adult rat ventricular myocytes and this was attributed to apoptosis. 6 Recently, work from our laboratory 7 and others 8 indicated that β 1 -AABs induced myocardial apoptosis. However, β 1 -AAB-induced cardiomyocyte death was not completely reversed with the caspase inhibitor Z-VADfmk, 6 indicating that other factors were involved in β 1 -AABinduced cardiomyocyte death.
Presently, we observed that β 1 -AABs decrease myocardial autophagy that maintains cellular homeostasis. 9 Deficiencies in autophagy allow the accumulation of damaged, denatured or aging proteins 10 and organelles, 11 and this will cause cell death. To date, the role of β 1 -AAB-induced changes in autophagy as related to cardiac function and cardiomyocyte death is unclear. Therefore, we characterized β 1 -AAB-induced changes in myocardial autophagy and identified a role for this in cardiac function and cardiomyocyte death. Our data will inform future studies of β 1 -AAB-positive heart dysfunction and suggest a treatment window for autophagy regulation.
Results β 1 -AABs caused the decrease of cardiac function in passively immunized rats. Rats were passively immunized by injecting β 1 -AABs (2 μg/g), once every 10 days, for 80 days. Before each immunization, serum β 1 -AABs were measured and it increased 20 days after passive immunization. Serum β 1 -AABs remained stable until the end of the experiment in the β 1 -AAB group compared with the control group (Supplementary Figure S1a).
Meanwhile, cardiac function was measured 40 and 60 days after passive immunization, and there was no significant difference in left ventricular function between the immunized and control groups. However, animals had significantly decreased left ventricular systolic pressure (LVSP), maximal positive and negative values of the instantaneous first derivative of left ventricular pressure (+dP/dt max and − dP/dt max ) and significantly increased left ventricular end diastolic pressure (LVEDP) 80 days after passive immunization in the β 1 -AAB group compared with the control group (Supplementary Figure S1b-e), indicating that a long-term presence of β 1 -AABs could cause a significant decrease in cardiac function.
Decreased autophagy had a role in decreased cardiac function induced by β 1 -AABs. In this study, Beclin-1, an important gene in autophagic regulation, and LC3, a protein marker for autophagy, were selected to measure autophagy. Significantly decreased LC3 and Beclin-1 protein were observed 20 days after passive immunization and lower levels persisted 40 and 80 days after passive immunization in the β 1 -AAB group (Figures 1a-c), indicating that β 1 -AABs may lead to decreased autophagy in myocardial tissues. To confirm the results, p62 protein, which was degraded within the autolysosomes, was used as a marker of autophagic flux. At 20 days after immunization, p62 protein in cardiac tissue increased significantly in the β 1 -AAB group compared with control groups and p62 remained relatively high at 40 and 80 days (Figures 1a and d), indicating a defect in autophagic flux in the presence of β 1 -AABs.
Rapamycin (RAPA), an mTOR inhibitor, is often used to upregulate autophagy. 12 In this experiment, it was used to improve autophagy in rat myocardial tissue (Supplementary Figure S2). Significantly improved left ventricular function was observed in rats pretreated with rapamycin compared with those only passively immunized in the β 1 -AAB group (Figures 2a-d), suggesting that upregulating autophagy may reverse decreased cardiac function induced by β 1 -AABs and that insufficient autophagy caused by the long-term presence of β 1 -AABs may be associated with decreased cardiac function. β 1 -AABs caused the death of H9c2 cells. Cell viability was measured to reflect the effects of β 1 -AABs stimulation of different durations. Significantly decreased cell viability was observed 12 h after β 1 -AABs stimulation and at a minimum of 36 h after β 1 -AABs stimulation (Figure 3a), indicating that β 1 -AABs may cause the death of myocardial cells.
LDH is released when cell membranes lyse, and it is an indicator of cell damage. Significantly increased LDH activity was observed at 6 h which lasted till the end of the experiment compared with the control group (Figure 3b), suggesting that β 1 -AABs may directly damage H9c2 cells.
Decreased autophagy was critical for cardiomyocyte death. LC3 and Beclin-1 were used to indicate autophagy, and Beclin-1 and LC3 protein and mRNA in H9c2 myocardial cells were measured 0, 12, 24, 36, and 48 h after β 1 -AABs stimulation using western blot and real-time PCR. In situ expression of LC3 protein was measured with immunostaining. Beclin-1 protein decreased 24 h after β 1 -AABs stimulation, dropped to a minimum at 36 h and recovered to normal at 48 h compared with the control group (Figures 4a and c). Beclin-1 mRNA expression significantly decreased 12 h after β 1 -AABs stimulation and it was minimal at 36 h and gradually recovered compared with the control group ( Figure 4f). Both LC3 protein (see Figures 4a and b) and LC3 mRNA (see Figure 4e) were decreased 12 h after β 1 -AABs stimulation, and were minimal at 36 h, returning to normal at 48 h compared with the control group. Immunostaining revealed that green punctate fluorescent signals representing LC3 significantly decreased 36 h after β 1 -AABs stimulation, but recovered at 48 h (Figures 4g and h). Thus, β 1 -AABs could decrease myocardial autophagy.
In addition, p62 was used to reflect autophagic flux.   (Figure 4l). To confirm these data, rapamycin or 3-methyladenine (3-MA) were used to upregulate or suppress autophagy, and cell viability of H9c2 myocardial cells was measured ( Figure 4m). Two different ways to increase or inhibit autophagy yielded similar results, suggesting that β 1 -AAB-induced decreases in autophagy have a role in cardiomyocyte death.
Increased autophagy benefitted myocardial cells with early β 1 -AABs stimulation. Autophagy is well known as a stress response, 13 so we observed changes in this stress over time, using earlier β 1 -AABs simulation. Cells were collected 0, 30 min, 1, 3, 6, and 12 h after stimulation and LC3 protein, LC3 mRNA, beclin-1 protein, beclin-1 mRNA, and p62 protein were measured. Autophagy increased early as LC3 protein (see Figures 5a and b), LC3 mRNA (see Figure 5e), beclin-1 protein (see Figures 5a and c), and beclin-1 mRNA (see Figure 5f) increased 30 min after β 1 -AABs treatment and remained high for 1 and 3 h and recovered to normal at 6 h compared with controls. At 12 h, expression decreased. Meanwhile, p62 protein was declined at 30 min, remained low at 1 and 3 h, returned to normal at 6 h, and then increased at 12 h (Figures 5a and d). These data are consistent with results mentioned above, in which    To better interpret changes in LC3, chloroquine was used to block the fusion of the autophagosome with the lysosome, and Figures 5g−i show that pretreatment with chloroquine could upregulate LC3 and p62 significantly for 3 h after β 1 -AABs stimulation, indicating that early increases in LC3 offer efficient autophagic flux.
To confirm that β 1 -AAB induced early increases in autophagy on cardiomyocyte death, autophagy was inhibited with Beclin-1 using RNA interference technology or 3-MA. In addition, because early increased autophagy recovered to almost normal 6 h after stimulation, 6 h was chosen for observation to eliminate the effect of later decreases in autophagy and diminished cell viability. Data show that cell viability did not change 6 h after β 1 -AABs stimulation of myocardial cells, but cardiomyocyte death occurred when autophagy was inhibited (Figures 5j and k).

Discussion
The objective of this study is to explore the importance of β 1 -AAB-induced reduction of myocardial autophagy on cardiac function in vivo, and discuss changes in β 1 -AABinduced autophagy over time and its significance on cardiomyocyte death. We observed that changes in β 1 -AABinduced autophagy increased early and then decreased. The early increase was cardioprotective, but the later decrease in autophagy prompted cardiomyocyte death and reduced cardiac function in vivo.
Recently, autoantibodies against the second extracellular loop of β 1 -adrenoceptor (β 1 -AR-ECII) (β 1 -AABs) were detected in the sera of patients with idiopathic dilated cardiomyopathy, 2 Chagas' heart disease, 3 and heart dysfunction caused by ischemic cardiomyopathy. 14 In our previous study, the long-term presence of autoantibodies may contribute to decreased cardiac function in a rat model immunized with the peptide corresponding to the second extracellular loop of β 1 -adrenoceptor. 5 In this study, a passively immunized rat model was established to eliminate the influence of antigen peptide itself on the body 15 and this also approximated human β 1 -AABs in the rat models. 16 During model establishment, IgG purified from actively immunized rats was administered to rats to observe the effect of β 1 -AABs on cardiac function and we found that β 1 -AABs could decrease cardiac function.
It has been reported that reduced autophagy is a major contributor to heart dysfunction, 17 and upregulated autophagy can significantly improve impaired cardiac function. 18 Autophagy is a catabolic process by which cells degrade dysfunctional cytoplasmic components and it is necessary for maintenance of cellular homeostasis. 19 Impaired autophagy causes mitochondrial dysfunction and accumulation, which is closely associated with many human diseases. 11 Generally, LC3 and Beclin-1 are used to monitor autophagy. Microtubuleassociated protein 1 light chain 3 (MAP1-LC3, LC3) is currently a more reliable biomarker to observe autophagy, 20 and the expression of soluble I-type LC3 (LC3-I) in the cytoplasm is regular. When autophagy occurs, LC3-I converts to LC3-II through ubiquitin-like modification and LC3-II then binds to and localizes in the autophagosomal membrane. LC3-II is well correlated with the number of formed autophagosomes. 21 As a component of phosphatidylinositol-3-kinase (PI3K) that is necessary in autophagic pathway, 22 Beclin-1 has an essential role in the formation of autophagosome precursors and membranes, 23 and is a common index in the observation of autophagy. In addition to LC3 and Beclin-1, p62 are markers for studying autophagic flux. 24 p62 binds directly to LC3 and is degraded within the autolysosomes, 25 so accumulation of p62 indicates inhibition of autophagy. 26 In a previous study, we confirmed that the long-term existence of β 1 -AABs reduced autophagy in myocardial tissues. 9 Similarly, in this study, we also observed that β 1 -AABs could decrease myocardial autophagy in a passively immunized rat model. Increased p62 protein indicated a defect in autophagy induced by β 1 -AABs. To observe the effect of decreased autophagy on cardiac function, the mTOR inhibitor rapamycin was used to upregulate autophagy. Data show that decreased cardiac function induced by β 1 -AABs was effectively reversed by enhanced autophagy. Thus, decreased myocardial autophagy is a major contributor to heart dysfunction.
Myocardial cells are a basic unit of cardiac systolic and diastolic function and when they are damaged or dead, contractile proteins in myocardial cells are degraded immediately, decreasing contractility. To identify a role for β 1 -AABs in myocardial cells, we purified IgG antibody in actively immunized rat serum to obtain β 1 -AABs. Next, a relatively stable H9c2 cell line was selected from embryonic rat heart tissues, and this line was used to observe the effects of β 1 -AABs on survival and autophagy of myocardial cells under the same conditions and time points. Data show that cardiac cell viability deceased at 12 h after β 1 -AABs stimulation and was minimal at 36 h, suggesting that β 1 -AABs may directly cause cardiomyocyte death. To confirm these data, LDH was measured in myocardial cells because this is documented to leak from damaged cells. 27 LDH activity in the cell culture medium significantly increased 6 h after β 1 -AABs stimulation, indicating damage. In conclusion, β 1 -AABs stimulation directly harmed the myocardial cell membrane, and caused cell death. This conclusion is consistent with previous results made by Jane-wit et al in an adult rat model. 6 Previously, β 1 -AABs were confirmed to induce apoptosis in cultured neonatal rat myocardial cells. 7 Staudt's group 28 also pointed out that β 1 -AABs could cause apoptosis in adult isolated myocardial cells. Thus, we measured β 1 -AAB effects on myocardial apoptosis over time and found increased apoptosis 6 h after β 1 -AABs stimulation and a return to normal at 24 h. Next, the caspase inhibitor Z-VAD-fmk was used to inhibit apoptosis and we observed that cardiomyocyte death recovered to a certain extent 36 h after β 1 -AABs stimulation (Supplementary Figure S5), indicating that β 1 -AAB-induced apoptosis is involved in cardiomyocyte death. Because β 1 -AAB-induced cardiomyocyte death was not completely recovered via apoptotic inhibition, other mechanisms are at play. Thus, we studied H9c2 cells with β 1 -AABs at different time points and observed decreased autophagic flux 12 h after stimulation and this was minimal at 36 h, indicating that β 1 -AABs stimulation decreased myocardial autophagic flux. Also, comparing changes of autophagy and myocardial cell viability over time after β 1 -AABs stimulation, we noted that myocardial cell viability started to decrease 12 h after stimulation, when autophagy decreased dramatically, indicating that the decline of autophagy induced by β 1 -AABs may cause cardiomyocyte death. Thus, we used recombinant plasmid pcDNA3.1-Beclin-1 and recombinant plasmidexpressing small interfering RNA targeting Beclin-1 (Beclin-1-shRNA) to upregulate and inhibit myocardial autophagy. Autophagy is reported to be upregulated by increasing or suppressing Beclin-1 expression. 29 Thus, H9c2 cells were transfected with recombinant plasmid pcDNA3.1-Beclin-1 to upregulate autophagy and we observed that cell viability was higher after β 1 -AABs stimulation in transfected cells compared with cells treated with only β 1 -AABs. In addition, Beclin-1 RNA interference plasmid was used to transfect H9c2 cells to inhibit autophagy and cell viability decreased after β 1 -AABs stimulation. However, beclin-1 could not only induce the autophagy, but also it could suppress autophagosome-lysosome fusion, 30 so data are difficult to interpret when beclin-1 manipulation is used to modulate autophagy. Therefore, we measured cell viability of H9c2 myocardial cells pretreated with rapamycin to upregulate autophagy or 3-MA to suppress autophagy and both yielded similar results. Therefore, decreased autophagy promotes cardiomyocyte death and improvements in autophagy benefit cardiac function.
Autophagy is commonly regarded as a stress response. 13 H9c2 myocardial cells were treated with β 1 -AABs at earlier time points (0, 30 min, 1, 3, and 6 h) and autophagy was found to first increase and then decrease. However, increased LC3 can be associated with either increased autophagic initiation or reduced degradation in the lysosome. To better distinguish between these two scenarios, chloroquine was applied and data show that the early increase in LC3 means efficient autophagic flux. So the increased autophagy after short-term β 1 -AABs stimulation is a stress response, and later depletion of autophagic genes and proteins decreases autophagy. Additional investigations are needed to learn whether β 1 -AABs directly induced these changes in myocardial cells.
To verify the effect of early increased autophagy after β 1 -AABs stimulation on cardiomyocyte death, H9c2 cells were transfected with Beclin-1-shRNA or 3-MA to inhibit autophagy and we found that myocardial survival was significantly reduced after early β 1 -AABs stimulation due to inhibition of autophagy. In addition, comparing changes in apoptosis and autophagy over time, autophagy increased before apoptosis. This early increase of autophagy was cardioprotective but this effect gradually disappeared as autophagy decreased and cells died.
In conclusion, autophagy is a stress response before apoptosis in β 1 -AAB-induced cardiomyocyte death, and decreased autophagy becomes a subsequent reason for cardiac dysfunction caused by β 1 -AAB-induced cardiomyocyte death. Thus, autophagic regulation is more important than apoptosis for patients with β 1 -AAB-positive heart dysfunction.
Our study is limited because we only observed that a lack of autophagy caused by β 1 -AABs decreased cardiomyocyte death. However, the role of apoptotic changes with autophagic upregulation or inhibition in cardiomyocyte death induced by β 1 -AABs requires more study. In addition, validating whether β 1 -AAB-induced cardiomyocyte death could be completely reversed by Z-VAD-fmk plus rapamycin is unknown. Still, we conclude that decreased autophagy is a major factor for cardiomyocyte death induced by β 1 -AABs. Our preliminary observations may open new insights into the pathogenesis and prevention of β 1 -AAB-positive heart dysfunction. Extraction of β 1 -AABs. First, a β 1 -AAB-positive animal model was established by actively immunizing rats with the second extracellular loop antigen peptide of β 1 -adrenergic receptor (β 1 -AR-ECII), as described in the previous studies. 9 Then, animal sera from the β 1 -AAB-positive group (actively immunized) and the control group were collected and extracted using MAbTrap Kit (GE Healthcare, 17-1128-01, Uppsala, Sweden) for affinity and purification of IgG.
Passive immunization and rapamycin treatment. Animals were randomized into passive immunization (β 1 -AAB) and control (negative IgG) groups. In the β 1 -AAB group, β 1 -AAB-positive IgG extracted as depicted above was administered to animals via the caudal vein (2 μg/g). This operation was carried out every 10 days and lasted for 80 days. Before each immunization, animal blood samples were collected via tail vein and then the sera were prepared to determine β 1 -AABs. In control group, β 1 -AAB-negative IgG was administrated by the exact same immunization and determination procedure as β 1 -AAB group. A subset of animals were treated with rapamycin (RAPA, Sigma, R0395, St. Louis, MO, USA) in β 1 -AAB group (β 1 -AAB+RAPA group). RAPA stock solution was prepared by dissolving rapamycin in DMSO (25 mg/ml) and stored until dilution with PBS for intraperitoneal injection. Because β 1 -AAB-induced decreases in myocardial autophagy occur on day 20 after passive immunization, RAPA administration started 3 days before this decrease (day 18), beginning at 0.5 mg/kg/day in the first 3 days and then adjusted to 0.25 mg/kg/day 31 until the end of this study.
ELISA. ELISA was performed using synthesized peptides corresponding to the sequence of the second extracellular loops of human β 1 -adrenoceptors as described previously. 32 First, 50 μl of the peptide (50 μg/ml) in 100 mM Na 2 CO 3 solution (pH 11.0) was coated on a microtiter plate (NUNC, Roskilde, Denmark) overnight at 4°C. Next, the wells were saturated with PMT (PBS containing 3% skimmed milk (W/V), 0.1% Tween 20 (V/V) and 0.01% thimerosal (W/V)) for 1 h at 37°C. After washing the wells three times with PBS-T, 5 μl of sera was added to 95 μl of PMT solution and incubated for 1 h at 37°C. After additional washing in PBS-T solution three more times, an affinity-purified biotinylated sheep anti-rat IgG (H+L) antibody (1 : 2000 dilution, Beijing Zhongshan Golden Bridge Biotechnology, ZB-2040, Beijing, China) was added and reacted for 1 h at 37°C. The plates were washed with PBS-T for three times. The bound biotinylated antibody was detected by incubating the plates with horseradish peroxidase streptavidin (1 : 3000 dilution, Vector, SA-5004, Burlingame, CA, USA) for 1 h. The wells were washed with PBS-T for three times. The substrate (2.5 mM H 2 O 2 , 2 mM 2, 2'-azinodi (ethylbenzthiazoline) sulfuric salt (ABTS, Bio Basic Inc., AD0002, Markham, ON, Canada)) was added. Optical density was measured after 30 min at 405 nm by a microplate reader (Molecular Devices Corp., Sunnyvale, CA, USA). Positive reactions of sera against peptides were verified as described by Liu et al. 32 In vivo measurements of cardiac function. With reference to our previous study, 4 after anesthesia, a cannula was inserted into the left ventricle via the right carotid artery. BL-410 biological signal recording and analysis system was used to record and analyze the following: the left ventricular systolic pressure (LVSP), left ventricular end diastolic pressure (LVEDP), and maximal positive and negative values of the instantaneous first derivative of left ventricular pressure (+dP/dt max and − dP/dt max ).