Cellular Signalling 26 (2014) 2591–2600

Contents lists available at ScienceDirect

Cellular Signalling journal homepage: www.elsevier.com/locate/cellsig

Protein kinase RNA-like endoplasmic reticulum kinase (PERK)/calcineurin signaling is a novel pathway regulating intracellular calcium accumulation which might be involved in ventricular arrhythmias in diabetic cardiomyopathy Zhongwei Liu a,1, Hui Cai b, Haitao Zhu c, Haroldo Toque d, Na Zhao a, Chuan Qiu e, Gongchang Guan a, Yonghui Dang f,⁎, Junkui Wang a,⁎ a

Department of Cardiology, Shaanxi Provincial People's Hospital, China Department of Anesthesiology, First Affiliated Hospital of Xi'an Jiaotong University, China School of Medicine, Xi'an Jiaotong University, China d Department of Pharmacology and Toxicology, GA Regents University, USA e School of Public Health & Tropical Medicine, Tulane University, USA f Department of Forensic Science, Xi'an Jiaotong University School of Medicine, China b c

a r t i c l e

i n f o

Article history: Received 31 July 2014 Accepted 17 August 2014 Available online 22 August 2014 Keywords: Endoplasmic reticulum stress Arrhythmias PERK Diabetic cardiomyopathy

a b s t r a c t We previously found that endoplasmic reticulum (ER) stress was involved in ventricular arrhythmias in diabetic cardiomyopathy. The present study was aimed to investigate the possible mechanism. In the in vivo study, diabetes cardiomyopathy (DCM) was induced by streptozotocin (STZ) injection. Hemodynamic and plasma brain natriuretic peptide (BNP) detections were used to evaluate cardiac functions; ECG was used to assess the vulnerability to arrhythmias by recording ventricular arrhythmia events (VAEs). In the in vitro study, highglucose incubation was employed to mimic the diabetic environment of myocytes. Immunofluorescent staining was used to investigate the nuclear factor of activated T cells (NFAT) nuclear translocation and (FK506-binding protein 12.6) FKBP12.6 disassociation. [3H]-ryanodine binding assay was implemented to assess the channel activity of ryanodine receptor. In both in vivo and in vitro studies, activity of calcineurin was determined by colorimetric method, and western blotting was used to detect protein expression levels. In the in vivo study, we found that inhibition of both of ER stress and PERK activation decreased the VAEs in DCM rats, accompanied by reduced activity of calcineurin in myocardial tissue. In the in vitro study, in high-glucose incubated myocytes, the depletion of PERK reduced activity of calcineurin, decreased NFAT translocation and FKBP12.6 disassociation from ryanodine receptor 2 (RyR2). Furthermore, PERK deletion also reduced RyR2 channel activity and consequently impaired intracellular calcium accumulation. We concluded that PERK/calcineurin-pathway was involved in intracellular calcium regulation in myocytes in diabetic heart, which might be the mechanism inducing arrhythmias in DCM. © 2014 Elsevier Inc. All rights reserved.

1. Introduction As the process of industrialization and economy development worldwide, more and more people are accommodating static lifestyles and western diets with excess energy. As a result, the morbidity and mortality of diabetes mellitus are increasing rapidly in recent decades [1]. It was believed that cardiovascular complications of diabetes were the most important causes leading to death [2]. Diabetic cardiomyopathy (DCM) which is considered as one of the key manifestations of diabetic cardiovascular complications is characterized by impaired cardiac ⁎ Corresponding authors. E-mail addresses: [email protected] (Y. Dang), [email protected] (J. Wang). 1 Dr. Zhongwei Liu and Dr. Hui Cai contributed equally to this study.

http://dx.doi.org/10.1016/j.cellsig.2014.08.015 0898-6568/© 2014 Elsevier Inc. All rights reserved.

systolic and diastolic functions. DCM is a specific cardiomyopathy because it is irrelevant with coronary heart disease, congenital heart disease, valvular heart disease and hypertensive heart disease even under circumstances of diabetes [3]. Studies showed increased heterogeneous cardiac electrophysiological features in diabetic patients who were highly susceptible to ventricular arrhythmias [4]. Some malignant ventricular arrhythmias often directly lead to sudden cardiac death (SCD) [5]. In our previous study, increased incidence of ventricular arrhythmias in DCM rats was also identified [6]. However, till now, there are rare studies exploring mechanism(s) regarding ventricular arrhythmias in DCM. Accumulating evidences indicated calcium mishandling in diabetic heart, participating in occurrence and development of DCM [7]. Accompanied by cardiac function loss, myocytes are in condition of calcium

2592

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

overload [8]. Several previous studies confirmed that the resting intracellular calcium ([Ca2+]i) increased dramatically in both myocytes isolated from diabetic rats and myocytes incubated with high-glucose medium [9]. Some arrhythmias were found associated with elevated myocardial [Ca2+]i resulted from gene mutations of certain calcium ion channels such as sarcoplasmic reticulum Ca2+-ATPase (SERCA), L-type calcium channel (LTCC) and ryanodine receptor 2 (RyR2) [10–12]. In this regard, a hypothesis that [Ca2+]i elevation in myocytes might also be the mechanism causing arrhythmias in diabetic hearts was formed. Furthermore, the molecular events connecting calcium homeostasis and arrhythmias in DCM aroused our interests. Endoplasmic reticulum (ER) stress is the intracellular response to stressful challenges which occurred in ER which is a specialized and important cellular organelle responsible for maintaining normal functions of protein folding, protein post-translational modifications, lipid synthesis and calcium homeostasis [13]. It was believed that ER stress was intensified in DCM which could be initiated by multiple stimuli such as excessive reactive oxygen species (ROS) generated by mitochondrial glucose oxidation [14]. Previously, we found that the occurrence of ventricular arrhythmic events (VAEs) in DCM rats increased along with the intensified ER stress in cardiac tissue, indicating ERS may play an important role in inducing arrhythmias [6]. Protein kinase RNA-like ER kinase (PERK) is one of the transducers sensing and conducting ER stress signals. Results from a recent study suggested that calcineurin is a down-stream molecule of PERK signaling by direct interaction with PERK [15]. Calcineurin is previously known to regulate ER calcium uptake and release by dissociating FK506-binding protein 12.6 (FKBP12.6) from RyR2–FKBP12 complex to regulate opening of this channel [16]. Thus, PERK/calcineurin signaling regulated RyR2 opening is the possible mechanism of [Ca2+]i elevation in myocytes which further induces arrhythmias in diabetic hearts. The present study undertaken was aimed to testify this presupposition. In the in vivo study, we investigated the association of ER stress and PERK with occurrences of arrhythmias by treating animals with specific inhibitors of ER stress and PERK. In the in vitro part, PERK/calcineurin regulated RyR2 opening and [Ca2+]i were also monitored. We believe that results in this study would not only elucidate the link between DCM and arrhythmia and increase understanding about the mechanisms, but also provide clue and theoretical basis for developing novel drugs for DCM. 2. Materials and methods 2.1. Animals, grouping and treatments 24 male and 16 female Sprague–Dawley (SD) rats were provided by Animal Experimental Center of Xi'an Jiaotong University (SPF class, 9 weeks old, body weight 254 ± 10 g). Every 10 rats were raised in an independent polypropylene cage under controlled conditions (12hour artificial light–dark cycle, temperature at 25 ± 1 °C, humidity at 56% ± 4%) for a week before the main experiments. Rats were fed

with standard formulated food and fresh tap water. Animal experimental procedures were in accordance with instructions and protocols approved by the Medical Animal Research Ethics Committee at Xi'an Jiaotong University. 40 rats were evenly and randomly divided into 4 groups namely control group (Ctrl), diabetic cardiomyopathy group (DCM), DCM rats treated with 4-phenyl butyric acid (4-PBA) (DCM + 4-PBA) and DCM rats treated with PERK inhibitor GlaxoSmithKline 2606414 (GSK 2606414, Toronto Research Chemicals) (DCM + GSK). Treatments in each group are shown in Table 1. 2.2. Plasma brain natriuretic peptide (BNP) assay Plasma was acquired from whole blood sample by centrifugation at 1500 rpm for 20 min at room temperature. Plasma BNP concentration was determined by Triage BNP assay (Biosite) according to the manufacturer's instructions. 2.3. Hemodynamic and electrocardiographic evaluations Rats were anesthetized by intraperitoneal injection of chloral hydrate (10%, 0.03 ml/kg bodyweight). According to descriptions in a previous study [17], the invasive hemodynamic determination was implemented. After intubation into the left ventricle through carotid artery, the cardiac function parameters including left ventricular end-diastolic pressure (LVEDP) and left ventricular systolic pressure (LVSP) were evaluated by a Mikro Tip catheter transducer (Millar Instruments) connected to Powerlab 4/25 Biological Analysis System (AD Instruments). After anesthetized rats from each group were fixed in a supine position, standard limb lead II probes were fixed to upper and lower limbs then connected to Powerlab 4/25 Biological Analysis System (AD Instruments) which was used to monitor and record ECG. Occurring number of ventricular arrhythmic events (VAEs) was introduced to evaluate the arrhythmic vulnerability of rats which was presented mainly as ventricular tachycardia and premature beats. 2.4. Primary myocytes isolation and culturing Primary myocytes were isolated from 2-day-old neonatal SD rats provided by Experimental Animal Center of Xi'an Jiaotong University. Hearts were harvested after rats were anesthetized by intraperitoneal injection of chloral hydrate (10%, 0.03 ml/kg) and sacrificed by cervical dislocation. After ventricular tissue was preserved and minced into small pieces which were then processed by enzymic digestion of Liberase (4.5 mg/ml, Roche). Then the cells were purified and cultured according to the protocol described previously [18]. Cells were incubated with medium containing minimum essential medium (MEM) supplemented with Hank's buffered solution: MEM (Gibco), 5% bovine calf serum (Gibco), 1.8 mmol/L CaCl2, 2 mmol/L L-glutamine (Invitrogen), 10 mmol/L 2,3butanedione monoxine (Invitrogen) and 100 U/ml penicillin–streptomycin mix (Invitrogen). After incubation on laminin-coated dishes at 37 °C

Table 1. Treatments of rats in different groups. Groups

Treatment 1

Treatment 2

Reagent

Description

Reagent

Description

Ctrl

Physiological saline

Single intraperitoneal injection

Physiological saline

DCM

Streptozotocin (STZ)

Physiological saline

DCM + 4-PBA

Streptozotocin (STZ)

65 mg/kg bodyweight; Single intraperitoneal injection 65 mg/kg bodyweight; Single intraperitoneal injection

DCM + GSK

Streptozotocin (STZ)

65 mg/kg bodyweight; Single intraperitoneal injection

GSK2606414

Oral administration; following continuous 21 Oral administration; following continuous 21 500 mg/kg bodyweight; Oral administration; following continuous 21 150 mg/kg bodyweight; Oral administration; following continuous 21

4-PBA

days days

days

days

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

in humidified incubator containing 5% CO2, the medium was replaced by fresh culture media containing MEM, 0.1 mg/ml myocyte bovine serum albumin (Sigma Aldrich), 2 mmol/L L-glutamin (Invitrogen) and 100 U/ml penicillin–streptomycin mix (Invitrogen). 2.5. Target gene knockdown by small interfering RNA (siRNA) and cell treatment Specific siRNA targeting PERK (5′-CACAACTGTATAACCGTTA-3′) synthesized by TaKaRa was used to knockdown the expression of PERK in primary myocytes. By using HiperFectsiRNA transfection reagent (Qiagen), siRNA (final concentration at 12.5 nmol/L) was transfected into cultured myocytes (at 60% confluence). Transfected cells were incubated for 24 h before cell treatments. Equal amount of normal myocytes and PERK knockdown cells at 60% confluence were then incubated with either normal glucose medium (NG, glucose concentration at 5.5 mmol/L) or high glucose medium (HG, glucose concentration at 33 mmol/L). 2.6. Colocalization and translocation analysis by immunofluorescence staining Myocytes from each group were placed on fibronectin (40 μg/ml) pre-coated cover slips and incubated in PBS for 30 min at room temperature. Then cells were fixed by 4% paraformaldehyde for 15 min and then permeabilized by incubation with 0.2% Triton. Non-specific binding was eliminated by incubation with blocking buffer (PBS containing 2.5% bovine serum albumin) for 30 min. 2.6.1. Colocalization analysis of RyR2 and FKBP12.6 Then the cells were subjected to incubate with anti-RyR2 antibody (Abcam, 1:50) and anti-FKBP12.6 antibody (Abcam, 1:100) at 4 °C for 12 h. After being washed by PBS, the cells were incubated with either Alexa Fluor 488-conjugated second antibody (Invitrogen) or Alexa Fluor 594-conjugated second antibody (Invitrogen). SlowFade Light Antifade Kit (Molecular Probes) was then added to alleviate fluorescence quenching. After exiting at 488 nm and 594 nm, cells were observed under Axio Imager 2 fluorescence inverse microscope (Zeiss) at 505 nm and 585 nm respectively. The colocalizations of RyR2 and FKBP12.6 were analyzed by Zeiss Physiology software (version 3.2). 2.6.2. Nuclear translocation analysis of NFATc1 Fixed cells were subjected to incubate with anti-NFATc1 antibody (Cell Signaling Tech, 1:250) at 4 °C for 12 h. The cells were then incubated with Alexa 488-conjugated secondary antibody (Invitrogen) after PBS washing. Nuclei were stained by 4,6-Diamidino-2-phenylindole (DAPI, Sigma Aldrich) for 1 min. SlowFade Light Antifade Kit (Molecular Probes) was then added to alleviate fluorescence quenching. After exiting at 519 and 340 nm, cells were observed under Axio Imager 2 fluorescence inverse microscope (Zeiss) at 495 nm and 442 nm respectively. The unclear translocations of NFATc1 were analyzed by Zeiss Physiology software (version 3.2). 2.7. Acquisition of cytosol fraction and ER membrane fraction by subcellular fractionation The method of subcellular fraction isolation was in accordance with previous studies [16]. Cells were homogenized in isolation buffer containing 0.29 mmol/L sucrose, 3 mmol/L imidazole, 1 mmol/L benzamidine, 0.5 mmol/L phenylmethanesulfonyl fluoride, 2 μg/ml pepstain A, 2 μg/ml leupeptin and 2 μg/ml aprotinin. Then cells were transferred into lysis buffer containing 50 mmol/L HEPES, 25 mmol/L Tris, 2.5 mmol/L dithiothreitol, 0.5% soya-bean phosphatidylcholine, 1% 3-[3-(cholamidopropyl) dimethylammonio]-1-propanesulfonic acid, 1 mmol/L benzamidine, 0.5 mmol/L phenylmethanesulfonyl fluoride, 2 μg/ml pepstain A, 2 μg/ml leupeptin and 2 μg/ml aprotinin. After

2593

centrifugation at 16,000 g twice at 4 °C for 30 min, the total cell lysate was collected. The collected cell lysate was then centrifuged at 1000 g at 4 °C for 20 min, and then the resulted pellet was homogenized and further centrifuged at 27,000 g at 4 °C for 15 min. Then the resulted supernatant was further centrifuged at 100,000 g at 4 °C for 15 min. The resulted supernatant was cytosol fraction while the resulted pellet could be collected as ER membrane fraction. 2.8. Western blotting Western blotting assay was implemented according to procedures described previously [19]. Proteins were extracted from cardiac tissue, total cell lysate, ER membrane fraction and cytosol fraction respectively. Then the concentrations of sample proteins were detected by BCA protein assay kit (Santa Cruz). Then 20 μg protein was loaded and subjected to vertical sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then the separated proteins were transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore). Antibodies including GRP78 (Abcam), PERK (Cell Signaling Tech), phosphorylated PERK (p-PERK, Cell Signaling Tech), FKBP12.6 (Santa Cruz), Sigma receptor 1 (Santa Cruz) and beta-Actin (Santa Cruz) were used to detect the targeted proteins by incubation with PVDF membranes at 4 °C for 8 h. After being washed by 0.02% Tris-buffered saline-Tween 20 (TBST), corresponding second antibodies conjugated to HRP (Santa Cruz) were used to incubate the membranes at room temperature for 2 h. The membranes were eventually visualized by using Super Signal West Pico chemiluminescence reagent (Thermo Scientific). The densities of the immunobands were quantified and analyzed by ImageJ2x software (Rawak Software). 2.9. Calcineurin activity assay The enzymatic activity of calcineurin was calculated by a colorimetric approach with extracted total protein from myocytes in a 96-well plate by Calcineurin Activity Assay Kit (Merck). All procedures in this assay followed the manufacturer's instructions. 2.10. Intracellular [Ca2+]i evaluation Same amount cells from each group were incubated with 10 μmol/L fura-2/AM (Beyotime) to load this fluorescent calcium indicator at room temperature for 30 min. After being washed by physiological saline for 3 times, cells were exited at 340 nm, and then observed at 510 nm under Axio Imager 2 fluorescence inverse microscope (Zeiss). Quantification of [Ca2+]i was represented as mean fluorescent intensity (MFI). 2.11. [3H]-ryanodine binding assay Protocol of this assay was in accordance with previous studies [20]. 20 nmol/L [3H]-ryanodine solution (PerkinElmer) was used to incubate cell lysate at 37 °C for 3 h in binding buffer containing 25 mmol/L Tris, 50 mmol/L HEPES, 100 μmol/L CaCl 2 , 1 mmol/L benzamidine, 0.5 mmol/L phenylmethanesulfonyl fluoride, 2 μg/ml pepstain A, 2 μg/ml leupeptin and 2 μg/ml aprotinin. Then this solution was further washed with buffer containing 25 mmol/L Tris and 250 mmol/L KCl and filtered through washing buffer soaked membrane filter (Millipore). The radioactivity of the filter represented the binding of ryanodine which was determined by liquid scintillation counter (Bioscan). 2.12. Statistical analysis Data in this study are presented in a (mean ± SD) manner and analyzed by SPSS software (SPSS, version 16.0). Significances of differences were decided by student's t test or one-way analysis of variance (ANOVA) with LSD test. P b 0.05 was set as statistically significant.

2594

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

3. Results

3.3. Effects of 4-PBA and GSK2606414 on GRP78 expression, PERK phosphorylation and calcineurin activity in diabetic rats

3.1. Fasting blood glucose and cardiac function evaluations As generally accepted, rats with fasting plasma glucose levels exceeding 300 mg/dL were considered diabetic while ranging from 90 to 130 mg/dL were considered normal [21]. As shown in Fig. 1, fasting blood glucose concentrations of rats in DCM, DCM + 4-PBA and DCM + GSK exceeded 300 mg/dL and were significantly higher than Ctrl which ranged from 90 to 130 mg/dL. Cardiac function loss was evaluated by both hemodynamic parameters and plasma BNP levels in this study. As shown in Fig. 1b, LVEDP was found increased while LVSP was found decreased significantly in DCM compared with Ctrl. However, administration of 4-PBA or GSK2606414 alleviated LVEDP and LVSP in DCM + 4-PBA and DCM + GSK respectively. Plasma BNP levels were also employed to indicate cardiac functions in each group. As demonstrated in Fig. 1c, plasma BNP levels in DCM increased significantly when compared with Ctrl, which were then reduced significantly in DCM + 4-PBA and DCM + GSK.

3.2. 4-PBA and GSK2606414 alleviated VAEs in diabetic rats The VAEs in each group were recorded by a standard limb lead II ECG as shown in Fig. 2. Compared with Ctrl, the number of VAEs was found to increase significantly in DCM. In both DCM + 4-PBA and DCM + GSK, numbers of VAEs were decreased significantly compared with DCM.

The activation of ER stress was indicated by expression level of GRP78. As one of the down-stream molecule of PERK, calcineurin activity is believed to increase after interaction with activated PERK (phosphorylated PERK). In the present study, as shown in Fig. 3, expression levels of GRP78 were elevated in DCM compared with Ctrl, which was reduced by 4-PBA treatment in DCM + 4-PBA. Significantly increased phosphorylation of PERK (p-PERK/PERK) was found in DCM when comparing with Ctrl. However, both 4-PBA and GSK2606414 inhibited PERK phosphorylation in diabetic hearts. The calcineurin activity was found dramatically increased in DCM compared with Ctrl, while was correspondingly decreased in DCM + 4-PBA and DCM + GSK compared with DCM. 3.4. PERK knockdown and its effects on GRP78 expression, calcineurin activity and NFATc1 unclear translocation in high-glucose incubated primary myocytes High-glucose incubation was used to mimic the environment of myocytes in diabetic heart. By utilizing RNAi technique, as shown in Fig. 4a, both PERK expression and phosphorylation were found absent in primary myocytes. After high-glucose incubation, GRP78 expression was elevated significantly in HG and PERK−/− (Fig. 4b). As shown in Fig. 4c, calcineurin activity was found to increase significantly in HG compared with Ctrl but decreased dramatically in PERK−/− compared with HG. NFATc1 nuclear translocation was evaluated by immunofluorescence

Fig. 1. Fasting blood glucose concentration and cardiac function determination. Columns indicate values of Ctrl, DCM, DCM + 4-PBA and DCM + GSK in (mean ± SD) manner respectively. (A) Fasting glucose concentrations were examined after modeling and treatments. (B) Plasma BNP concentrations (pg/ml) were determined by Triage BNP assay in collected blood sample. (C) and (D) LVEDP and LVSP in different groups were evaluated by invasive hemodynamic determinations. a values are significantly different from Ctrl; b values are significantly different from DCM [(#) P b 0.05; (*) P b 0.01].

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

2595

Fig. 2. Electrocardiogram analysis. (A) Graphics captured during electrocardiogram recording including (a) normal ECG graphics, (b) ventricular premature beat and (c) ventricular tachycardia. (B) The severity of arrhythmia was presented by occurring number of ventricular arrhythmia events (VAEs) in 15-minute ECG recording. Occurring numbers were presented as columns in a (mean ± SD) manner in Ctrl, DCM, DCM + 4-PBA and DCM + GSK respectively. a values are significantly different from Ctrl; b values are significantly different from DCM [(*) P b 0.01].

which is demonstrated in Fig. 4d. We found that NFATc1 unclear accumulation was enhanced in high-glucose incubated myocytes which was reduced dramatically by PERK depletion in PERK−/−. 3.5. PERK knockdown impaired FKBP12.6 translocation from ER membrane to cytosol and disassociation from RyR2 Western blot results in Fig. 5a indicate that the translocation of FKBP12.6 from ER membrane to cytosol was augmented after

high-glucose incubation in HG compared with Ctrl. However, the PERK depletion significantly mitigated this translocation significantly in PERK−/− compared with HG. Furthermore, the colocalization analysis was based on calculating the Pearson's correlation coefficients which could indicate the degree of contact between two detected molecules. In this study, the association of FKBP12.6 and RyR2 was evaluated by double immunofluorescent colocalization which is demonstrated in Fig. 5b. Evidenced by correlation coefficient analysis, the colocalization of FKBP12.6 and RyR2 was dramatically impaired in HG

Fig. 3. Assessments of GRP78 expression, PERK phosphorylation and calcineurin activity in in vivo study. (A) Immunoblots of antibodies against GRP78, phosphorylated PERK, PERK and β-Actin in cardiac tissue from Ctrl, DCM, DCM + 4-PBA and DCM + GSK by Western Blotting respectively. (B) Myocardial calcineurin activity was detected by using a colorimetric method in different groups. (C) PERK phosphorylation was quantified by normalizing expression of phosphorylated PERK to PERK. (D) Relative GRP78 expressions were normalized to β-Actin which was introduced as the internal reference in different groups. Values are presented as columns in a (mean ± SD) manner in Ctrl, DCM, DCM + 4-PBA and DCM + GSK respectively. a values are significantly different from Ctrl; b values are significantly different from DCM [(*) P b 0.01].

2596

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

Fig. 4. Assessment of GRP78 expression, PERK phosphorylation, calcineurin activity and NFATc1 translocation in in vitro study. (A) Immunoblots of phosphorylated PERK and PERK after PERK was knocked down by siRNA in normal glucose (NG) and high-glucose (HG) incubated primary myocytes. (B) The upper part demonstrates the immunoblots of GRP78, phosphorylated PERK, PERK and β-Actin in NG, HG and PERK−/− myocytes respectively. The lower part shows the quantification of GRP78 expression and PERK phosphorylation. The GRP78 expression was normalized to β-Actin while the PERK phosphorylation was quantified by normalizing expression of phosphorylated PERK to PERK. Values were presented as (mean ± SD) as columns. (C) Calcineurin activity in primary myocytes was detected by using a colorimetric method. Values were presented as (mean ± SD) as columns. (D) The left part demonstrated immunofluorescent staining of NFATc1 (green), DAPI (blue) and merged images captured in NG, HG and PERK−/− respectively. The NFAT nuclear translocation was indicated by the ratio of nuclear translocated cell count against total cell count which was demonstrated on the right part. Values were presented as (mean ± SD) as columns in NG, HG and PERK−/− respectively. a values are significantly different from NG; b values are significantly different from HG [(*) P b 0.01].

compared with Ctrl but significantly alleviated in PERK−/− compared with HG.

in [Ca2+]i in PERK+/+myocytes. However, the high-glucose incubation induced [Ca2 +]i increase was inhibited by PERK depletion in PERK−/− cells.

3.6. PERK knockdown inhibited RyR2 activity in high-glucose incubated myocytes 4. Discussion The disassociation of FKBP12.6 from RyR2 may lead to increase RyR2 activity. Results from [3H]-ryanodine binding assay provided relatively direct evidence for PERK −/− on RyR2 activity. As shown in Fig. 6a and 6b, after high-glucose incubation, the maximal binding (Bmax) was increased significantly compared with Ctrl. In PERK−/−, however, Bmax decreased significantly compared with HG. 3.7. PERK knockdown decreased [Ca2 +]i in high-glucose incubated myocytes The effects of PERK depletion on [Ca2+]i were further investigated by fluorescence detection of calcium indicator fura-2/AM which is demonstrated in Fig. 7. High-glucose incubation induced dramatic increase

According to the data from the World Health Organization (WHO), it was estimated that there would be 300 million diabetic patients by the year 2025 because of the spread of sedentary lifestyle and obesity in the modern society [22]. Among the acute and chronic complications of diabetes, the cardiovascular complications take major responsibility for diabetes related death. DCM is an important clinical manifestation of diabetes-induced cardiac injury which is independent of coronary artery disease, valvular heart disease, hypertension, etc. [23]. Pathological changes including cell apoptosis [24], fibrosis [25] and cardiac hypertrophy [26] are involved in DCM. More importantly, during the development of DCM, in addition to cardiac pump dysfunction due to loss of contractile units, malignant ventricular arrhythmias such as sudden

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

2597

Fig. 5. Evaluation of FKBP12.6 disassociation in in vitro study. (A) Immunofluorescent staining of FKBP12.6 (red), RyR2 (green), DAPI (blue) and merged imaged captured in NG, HG and PERK−/− respectively. Light yellow arrows are pointing disassociation image pattern of FKBP12.6 from RyR2. (B) Quantification of the degree of colocalization of FKBP12.6 and RyR2 by coefficients analysis in each group. Coefficients were presented in a (mean ± SD) manner as columns. (C) The upper part demonstrated immunoblots of FKBP12.6, β-Actin and Sigma R1 in protein extracted from cytosol and ER membranes in myocytes. The lower part demonstrated the relative expression of FKBP12.6 in cytosol (normalized to β-Actin) and in ER membrane (normalized to Sigma R1) in NG, HG and PERK−/− respectively. Values are presented in a (mean ± SD) manner as columns. a values are significantly different from NG; b values are significantly different from HG [(#) P b 0.05; (*) P b 0.01].

cardiac death (SCD) were considered contributing more to the mortality in diabetic patients [27,28]. In our previous study, although very preliminary, we revealed that there was a correlation between ER stress and occurrence of ventricular arrhythmia in DCM [6]. ER is one of the vital organelles in eukaryotic cells executing fundamental biological functions including directing protein folding, protein post-translational modification, protein sorting and trafficking, lipid synthesis and calcium homeostasis [29,30]. Unfold protein response (UPR) which performs reactions to redirect protein folding and eliminate misfolded proteins, is a protective compensatory response to many stimuli [31]. However, when ER encounters severe stimuli leading accumulation of misfolded or unfolded proteins in ER lumen, decompensated UPR would result in a condition called ER stress. In this study, intraperitoneal injection of STZ was used to induce DCM which was confirmed by hemodynamic examination of cardiac dysfunction. After rats were treated with 4-PBA, a specific ER stress inhibitor, evidenced by decreased expression of GRP78, inhibited ER stress was found in cardiac tissue from DCM rats. Importantly, according to ECG, occurring number of VAEs which mainly presented as ventricular tachycardia and premature beats was found dramatically decreased in 4-PBA treated DCM rats. This result re-confirmed the notion that ER stress was involved in inducing ventricular arrhythmias in DCM. In addition, we also recorded ECG in DCM rats treated by GSK2606414 which is a specific PERK inhibitor by inhibiting PERK self phosphorylation. Interestingly, without affecting GRP78 expression, decreased occurring number of VAEs was also found in GSK2606414 treated DCM rats. Thus, in ER stress-induced arrhythmia in DCM, PERK was identified as the possible mediator. Calcium is believed as an important ion in the heart coupling cardiac pump and electrophysiological functions [32,33]. The diastolic/systolic dysfunctions, intracellular/intercellular electrophysiological changes and even cell apoptosis may be underscored by calcium homeostasis

disturbance. Elevated cytosolic calcium concentration was found in hearts which suffered from lethal ventricular arrhythmias [34]. Under certain conditions, calcium would be excessively released from ER to increase intracellular calcium concentration which was reportedly involved in the formation of early after depolarization (EAD) and delayed after depolarization (DAD) [35]. After depolarization (AD), also with the name of triggered activity (TA), was described as the featured electrophysiological changes causing arrhythmias [36]. As the supposed mediator of ER stress and arrhythmia mentioned above, we also supposed PERK to be a regulator of calcium homeostasis. According to the results from a recent study, the coimmunoprecipitation assay suggested an upstream and downstream association between PERK and calcineurin with direct contact [15]. Indeed, in DCM rats, we found that both 4-PBA and GSK2606414 treatment impaired the enzymatic activity of calcineurin. This result confirmed the regulatory relationship between PERK and calcineurin. Furthermore, it was considered that calcineurin could initiate and facilitate the disassociation of FKBP12.6 from RyR2 to increase the channel activity of RyR2 [16]. Eventually, calcium stored in ER would be relapsed to cytosol. In this regard, we proposed that during ER stress in DCM, PERK could induce [Ca2+]i elevation in myocytes by increasing activity of calcineurin which further prompt FKBP12.6 to separate from RyR2. In order to testify the above calcium regulatory mechanism in DCM hearts undergoing ER stress, the in vitro study was implemented. In accordance with the in vivo study, after PERK was removed from cardiac myocytes by RNAi technique, enzymatic activity of calcineurin decreased significantly. This notion was reinforced by the observation of changes of translocation of NFATc1. As a downstream effecter of calcineurin, NFAT is activated by calcineurin and forms calcineurin/NFAT signaling pathway [37,38]. It was suggested that after calcineurin/NFAT signaling was activated by multiple stimuli, NFAT would be translocated into the cell nucleus to regulate related gene expressions in response to

2598

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

Fig. 6. RyR2 channel activity assessment by [3H]-ryanodine binding assay. (A) [3H]ryanodine binding assay was conducted in cell lysate from NG, HG and PERK−/− respectively. The binding ability was assessed by radioactivity detection which was represented in the line chart. (B) Scatchard analysis of [3H]-ryanodine binding assay. Equation of each group was indicated. Bmax and dissociation constant (KD) were calculated according to equation of each group.

these stimuli [39]. Thus NFAT nuclear translocation is the hallmark event of activation of calcineurin signaling [40]. In this study, NFATc1 nuclear translocation increased in response to high-glucose incubation but attenuated by PERK depletion, indicating that PERK activation was the prior molecular event requited for calcineurin signaling. The fundamental function of FKBP12.6 is to regulate intracellular release by adjusting RyR activity [41]. Under physiological conditions, FKBP12.6 binds to RyR to form FKBP–RyR complex with low RyR channel activity. However, under pathological conditions, calcineurin binds to FKBP12.6 and facilitates the disintegration of FKBP–RyR complex [42]. In this study, in high-glucose incubated myocytes, FKBP12.6 was found disassociated from ER membrane and dis-colocalized with RyR2. We also found that the disassociation and dis-colocalization were impaired because of decreased calcineurin activity in PERK −/− high-glucose incubated myocytes. By [3H]-ryanodine binding assay, we found that high-glucose incubation resulted in increased activity of RyR2 dramatically in PERK+/+myocytes. However, in comparison, the RyR2 activity decreased significantly in PERK −/− myocytes after high-glucose incubation. There are three subtypes of RyR, namely RyR1, RyR2 and RyR3. RyR2 is highly expressed in cardiac tissue while RyR1 in skeletal muscle and RyR3 in the nervous system. Located on ER membrane, RyR2 is the major calcium release channel in myocytes [43]. Over activation of RyR2 was considered associated with ventricular tachycardia [44], atrial fibrillation [45], arrhythmic right ventricular cardiomyopathy/dysplasia (ARVC/D) [46] and so on. Excessive calcium release from RyR2 was believed to be the initiator of several myocardial electrophysiological changes including EAD and DAD [47]. In the present study, we found that high-glucose incubation treatment increased [Ca2+]i in myocytes on the premise of increased RyR2 activity. In high-glucose incubated PERK −/− myocytes however, [Ca2 +]i was significantly reduced because of decreased RyR2 activity. These results are consistent with the theory of RyR2's regulatory effects on intracellular calcium accumulation.

Fig. 7. Detection of intracellular calcium in primary myocytes. (A) Images captured after myocytes in NG, HG and PERK−/− were stained by fura-2/AM which is used as a fluorescent calcium indicator (green stained). (B) The columns represented the mean fluorescent intensity (MFI) of fura-2/AM which indicated the [Ca2+]i in myocytes from NG, HG and PERK−/− respectively.

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

2599

Fig. 8. Schematic diagram demonstrating speculated calcium regulating mechanism related to arrhythmias. Located on ER membrane and as an ER signaling transducer, PERK governed signaling is activated by metabolic disturbance induced UPR in DCM. After GRP78 disassociates from PERK, PERK is activated by self phosphorylation. As the activated form, phosphorylated PERK increases the enzymatic activity of calcineurin by direct interaction. Under normal physiological conditions, FKBP12.6 binds to RyR2 to inhibit its ion channel activity, blocking calcium outflow from ER. However, calcineurin would facilitate the disassociation of FKBP12.6 from RyR2 by its increased enzymatic activity. Since Elevated [Ca2+]i was considered the cause of electrophysiological changes such as DAD and EAD which induce arrhythmias, the PERK/calcineurin/FKBP12.6 signaling regulated intracellular calcium elevation could be considered one of the possible mechanisms inducing arrhythmias in DCM.

Taken together, evidences we collected in this study may provide a novel RyR2 related calcium regulating mechanism leading to arrhythmias. The schematic diagram demonstrating this mechanism was demonstrated in Fig. 8. During development of DCM, stimuli from metabolic disturbance induce excessive ER stress in myocytes. Then calcineurin activity is increased by PERK which is activated by its self phosphorylation during conducting signals in ER stress. Calcineurin facilitates degradation of FKBP–RyR2 complex by promoting disassociation of FKBP12.6 from RyR2. Thus, after the channel activity is enhanced, calcium stored in ER is released to cytosol. Electrophysiological changes such as EAD and DAD might be induced in the circumstance of intracellular calcium accumulation, leading to arrhythmias which could be detected in ECG eventually. The present study shows that PERK/calcineurin/RyR2 signaling is the calcium regulatory mechanism which is also related to arrhythmias in DCM. However, there are still several limitations which should be elicited in the future. Utility of optical calcium mapping (OCM) would be helpful in establishing direct evidence concerning calcium levels in DCM hearts undergoing arrhythmias. In addition, further results from electrophysiological study with whole-cell patch clamp would be a replenishment, which has become the topic for our team in the near future. Acknowledgments This study is supported by grants from the National Natural Science Foundation of China (NSFC) No. 81171262 and Fundamental Research Funds for the Central Universities No. 2011JDHZ59. References [1] D.W. Lam, D. LeRoith, Curr. Opin. Endocrinol. Diabetes Obes. 19 (2012) 93–96. [2] G.S. Hillis, M. Woodward, A. Rodgers, C.K. Chow, Q. Li, S. Zoungas, A. Patel, R. Webster, G.D. Batty, T. Ninomiya, G. Mancia, N.R. Poulter, J. Chalmers, Diabetologia 55 (2012) 1283–1290. [3] S. Boudina, E.D. Abel, Circulation 115 (2007) 3213–3223. [4] O. Casis, E. Echevarria, Curr. Vasc. Pharmacol. 2 (2004) 237–248.

[5] S. Peters, H. Peters, L. Thierfelder, Int. J. Cardiol. 71 (1999) 243–250. [6] Z.W. Liu, H.T. Zhu, K.L. Chen, X. Dong, J. Wei, C. Qiu, J.H. Xue, Cardiovasc. Diabetol. 12 (2013) 158. [7] L. Ligeti, O. Szenczi, C.M. Prestia, C. Szabo, K. Horvath, Z.L. Marcsek, R.G. van Stiphout, N.A. van Riel, J. Op den Buijs, Van der Vusse GJ, T. Ivanics, Int. J. Mol. Med. 17 (2006) 1035–1043. [8] A. Ziegelhoffer, T. Ravingerova, J. Styk, N. Tribulova, K. Volkovova, J. Sebokova, A. Breier, Diabetes Res. Clin. Pract. 31 (1996) S93–S103 (Suppl.). [9] H. Qi, Y. Cao, W. Huang, Y. Liu, Y. Wang, L. Li, L. Liu, Z. Ji, H. Sun, PLoS One 8 (2013) e65147. [10] R. Alcalai, H. Wakimoto, M. Arad, D. Planer, T. Konno, L. Wang, J.G. Seidman, C.E. Seidman, C.I. Berul, J. Cardiovasc. Electrophysiol. 22 (2011) 316–324. [11] I. Splawski, K.W. Timothy, N. Decher, P. Kumar, F.B. Sachse, A.H. Beggs, M.C. Sanguinetti, M.T. Keating, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 8089–8096 (discussion 8086–8088). [12] M. Scoote, A.J. Williams, Cardiovasc. Res. 56 (2002) 359–372. [13] T. Minamino, I. Komuro, M. Kitakaze, Circ. Res. 107 (2010) 1071–1082. [14] Y. Tagawa, N. Hiramatsu, A. Kasai, K. Hayakawa, M. Okamura, J. Yao, M. Kitamura, Free Radic. Biol. Med. 45 (2008) 50–59. [15] R. Wang, B.C. McGrath, R.F. Kopp, M.W. Roe, X. Tang, G. Chen, D.R. Cavener, J. Biol. Chem. 288 (2013) 33824–33836. [16] B. Liao, Y.M. Zheng, V.R. Yadav, A.S. Korde, Y.X. Wang, Antioxid. Redox Signal. 14 (2011) 37–47. [17] L. Bocchi, M. Savi, G. Graiani, S. Rossi, A. Agnetti, F. Stillitano, C. Lagrasta, S. Baruffi, R. Berni, C. Frati, M. Vassalle, U. Squarcia, E. Cerbai, E. Macchi, D. Stilli, F. Quaini, E. Musso, PLoS One 6 (2011) e17750. [18] L. Zuo, D.J. Youtz, L.E. Wold, PLoS One 6 (2011) e23116. [19] Z.W. Liu, H.T. Zhu, K.L. Chen, C. Qiu, K.F. Tang, X.L. Niu, Biol. Trace Elem. Res. 156 (2013) 262–270. [20] T. Hamada, M.L. Bannister, N. Ikemoto, Biochemistry 46 (2007) 4272–4279. [21] Y. Li, Y. Zhang, D.B. Liu, H.Y. Liu, W.G. Hou, Y.S. Dong, Int. J. Med. Sci. 10 (2013) 377–381. [22] H. King, R.E. Aubert, W.H. Herman, Diabetes Care 21 (1998) 1414–1431. [23] A.G. Bertoni, D.C. Goff Jr., R.B. D'Agostino Jr., K. Liu, W.G. Hundley, J.A. Lima, J.F. Polak, M.F. Saad, M. Szklo, R.P. Tracy, D.S. Siscovick, Diabetes Care 29 (2006) 588–594. [24] F. Kuethe, H.H. Sigusch, S.R. Bornstein, K. Hilbig, V. Kamvissi, H.R. Figulla, Horm. Metab. Res. 39 (2007) 672–676. [25] C.S. Samuel, T.D. Hewitson, Y. Zhang, D.J. Kelly, Endocrinology 149 (2008) 3286–3293. [26] C. Aguilar, F. Ventura, L. Rodriguez-Delfin, Rev. Peru. Med. Exp. Salud Publica 28 (2011) 264–272. [27] G.A. Suarez, V.M. Clark, J.E. Norell, T.E. Kottke, M.J. Callahan, P.C. O'Brien, P.A. Low, P.J. Dyck, J. Neurol. Neurosurg. Psychiatry 76 (2005) 240–245. [28] J. Matejikova, J. Kucharska, D. Pancza, T. Ravingerova, Physiol. Res. 57 (Suppl. 2) (2008) S55–S60.

2600

Z. Liu et al. / Cellular Signalling 26 (2014) 2591–2600

[29] D.W. Tang, Y. Fang, Z.X. Liu, Y. Wu, X.L. Wang, S. Zhao, G.C. Han, S.M. Zeng, Biol. Reprod. 89 (2013) 124. [30] I. Braakman, D.N. Hebert, Cold Spring Harb. Perspect. Biol. 5 (2013) a013201. [31] R.Y. Hampton, Curr. Biol. 10 (2000) R518–R521. [32] L. Wang, S.J. Zhang, H.P. Wang, C. Zou, Z.H. Liu, Sheng Li Xue Bao 61 (2009) 551–558. [33] A. Tinker, J.L. Sutko, L. Ruest, P. Deslongchamps, W. Welch, J.A. Airey, K. Gerzon, K.R. Bidasee, H.R. Besch Jr., A.J. Williams, Biophys. J. 70 (1996) 2110–2119. [34] M.G. Chelu, X.H. Wehrens, Biochem. Soc. Trans. 35 (2007) 952–956. [35] W.T. Clusin, Crit. Rev. Clin. Lab. Sci. 40 (2003) 337–375. [36] F.I. Marcus, J. Electrocardiol. 38 (2005) 60–63. [37] H. Diedrichs, M. Chi, B. Boelck, U. Mehlhorn, R.H. Schwinger, Eur. J. Heart Fail. 6 (2004) 3–9. [38] W. Ding, M. Dong, J. Deng, D. Yan, Y. Liu, T. Xu, J. Liu, Am. J. Physiol. Heart Circ. Physiol. (2014). [39] C. Sommerer, S. Meuer, M. Zeier, T. Giese, Clin. Chim. Acta 413 (2012) 1379–1386.

[40] V. Rajapurohitam, F. Izaddoustdar, E. Martinez-Abundis, M. Karmazyn, Cell. Signal. 24 (2012) 2283–2290. [41] T. Guo, R.L. Cornea, S. Huke, E. Camors, Y. Yang, E. Picht, B.R. Fruen, D.M. Bers, Circ. Res. 106 (2010) 1743–1752. [42] A.R. Marks, Front. Biosci. 7 (2002) d970–d977. [43] A.R. Marks, J. Mol. Cell. Cardiol. 33 (2001) 615–624. [44] J. Shan, W. Xie, M. Betzenhauser, S. Reiken, B.X. Chen, A. Wronska, A.R. Marks, Circ. Res. 111 (2012) 708–717. [45] V.V. Fedorov, O.F. Sharifov, O.P. Piatakova, G.G. Beloshapko, A.V. Iushmanova, L.V. Rozenshtraukh, Kardiologiia 42 (2002) 59–71. [46] X.H. Wehrens, S.E. Lehnart, A.R. Marks, Ann. N. Y. Acad. Sci. 1047 (2005) 366–375. [47] S. Kobayashi, M. Yano, T. Suetomi, M. Ono, H. Tateishi, M. Mochizuki, X. Xu, H. Uchinoumi, S. Okuda, T. Yamamoto, N. Koseki, H. Kyushiki, N. Ikemoto, M. Matsuzaki, J. Am. Coll. Cardiol. 53 (2009) 1993–2005.