PKCɛ mediates serine phosphorylation of connexin43 induced by lysophosphatidylcholine in neonatal rat cardiomyocytes

Toxicology 314 (2013) 11–21

Contents lists available at ScienceDirect

Toxicology journal homepage: www.elsevier.com/locate/toxicol

PKC␧ mediates serine phosphorylation of connexin43 induced by lysophosphatidylcholine in neonatal rat cardiomyocytes Chih-Kai Liao a , Hsiang-Hsi Cheng a , Sheng-De Wang a , Dong-Feng Yeih b,∗∗ , Seu-Mei Wang a,∗ a b

Department of Anatomy and Cell Biology, College of Medicine, National Taiwan University, 1-1 Jen-Ai Road, Taipei 10051, Taiwan Department of Internal Medicine, College of Medicine, National Taiwan University, 1-1 Jen-Ai Road, Taipei 10051, Taiwan

a r t i c l e

i n f o

Article history: Received 27 May 2013 Received in revised form 17 July 2013 Accepted 2 August 2013 Available online 20 August 2013 Keywords: Cardiomyocytes Lysophosphatidylcholine Gap junction Connexin43 Serine 368 phosphorylation PKC␧

a b s t r a c t Lysophosphatidylcholine (LPC) is a potent pro-arrhythmic derivative of the membrane phosphotidylcholine, which is accumulated in heart tissues during cardiac ischemia. However, the cellular mechanism underlying LPC-induced cardiomyocyte damage remains to be elucidated. This study focuses on the effects of LPC on cardiomyocyte gap junction. At 30 ␮M, LPC decreased the spontaneous contraction rates of cardiomyocytes, and caused arrhythmic contraction without affecting cell viability. Connexin43 (Cx43) was seen as large plaques at cell junctions in control cells, whereas upon LPC treatment, the intensity of Cx43 staining was decreased in a concentration-sensitive manner and Cx43 staining appeared as tiny dots at cell junctions with a corresponding increase in cytoplasmic punctate staining. This distributional change of Cx43 was accompanied by an impairment of the gap junction intercellular communication (GJIC). Further, LPC treatment induced protein kinase C (PKC) activation, and PKC-dependent Cx43 phosphorylation at serine (Ser) 368. Pre-treatment with a specific PKC␧ inhibitor, eV1-2, prevented the LPC-induced Cx43 phosphorylation at Ser368 and the loss of Cx43 from gap junctions, both of which may disturb GJIC functions. Furthermore, siRNA knockdown of PKC␧ in H9c2 cells prevented LPC-induced serine phosphorylation of Cx43, confirming the role of PKC␧ in Cx43 serine phosphorylation. Double labeling immunofluorescence showed that LPC increased the colocalization of Cx43 with ubiquitin, and pretreatment with MG132 effectively prevented LPC-induced gap junction disassembly. LPC increased the ubiquitination of Cx43, which was blocked by eV1-2 pretreatment, suggesting that LPC accelerated the intracellular degradation of Cx43 via the ubiquitin-proteasomal pathway. It can be concluded that LPC destroyed the structure and function of gap junctions via PKC␧-mediated serine phosphorylation of Cx43. PKC␧ inhibitors might therefore be effective in prevention of LPC-related diseases. © 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Phophatidylcholine, an essential phospholipid in eukaryotes, is the major component of cell membranes and circulating plasma lipoproteins such as high density lipoprotein (Li and Vance, 2008). Lysophosphatidylcholine (LPC) is an oxidized product of phophatidylcholine by low density lipoprotein (LDL)-associated phospholipase A2 (Parthasarathy and Barnett, 1990). The physiological concentrations of LPC in the human peripheral venous blood are 55.3 ± 8.6 ␮M (Sedis et al., 1990). However, under

Abbreviations: LPC, lysophosphatidylcholine; GJIC, gap junction intercellular communication; PKC, protein kinase C; Cx43, connexin43; ZO-1, zonula occludens1. ∗ Corresponding author. Tel.: +886 2 23123456x88179; fax: +886 2 23915192. ∗∗ Corresponding author. Tel: +886 2 23123456x65259; fax: +886 2 23215443. E-mail addresses: [email protected] (D.-F. Yeih), [email protected] (S.-M. Wang). 0300-483X/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tox.2013.08.001

pacing-induced ischemia, LPC concentrations are increased to 178 ± 18.0 ␮M at peak pacing in the human coronary sinus (Sedis et al., 1990). Thus, overproduction of LPC is considered to be an inflammatory factor of atherosclerosis or a toxic metabolite of ischemia (Arnsdorf and Sawicki, 1981; Matsumoto et al., 2007). Arrhythmogenesis and contractile dysfunction in heart failure is thought to be related to the high levels of LPC (Clarkson and Ten Eick, 1983; Corr et al., 1982; Hoque et al., 1995; Pogwizd et al., 1986; Ver Donck et al., 1992). Man et al. (1983) have shown that high extracellular LPC concentrations cause arrhythmia (Man et al., 1983). LPC modulates a variety of channels in the plasma membrane. In isolated cardiomyocytes, LPC increases intracellular calcium ions and changes cellular morphology (Ma et al., 1999). It also activates protein kinase C (PKC) to increase L-type Ca2+ current, which might be related to the formation of atherosclerosis (Jung et al., 2008). LPC stimulates T-type calcium currents in a calcium-dependent manner via PKC␣ activation, which may play a role in triggering arrhythmias in pathophysiological conditions of the heart (Zheng et al., 2010).

12

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

Gap junction channels, which are made up of two hemichannels (connexons) in the opposing plasma membrane, are responsible for the metabolic and electrical coupling between adjacent cells, a process also known as gap junction intercellular communication (GJIC) (Solan and Lampe, 2009). Each hemichannel is formed by six connexin molecules including connexin43 (Cx43), which is the predominant gap junction protein in the ventricular myocardium (Severs et al., 2008). Cx43 is known as a phosphoprotein with three electrophoretic isoforms including one non-phosphorylated form (P0) and two phosphorylated forms (P1 and P2) (Solan and Lampe, 2005). It is well-known that posttranslational modifications of Cx43 such as phosphorylation and ubiquitination can regulate its trafficking, assembly/disassembly, internalization, and GJIC (Solan and Lampe, 2009; Kjenseth et al., 2010). GJIC between cardiomyocytes contributes to the rapid propagation of electrical conductance, impulse generation and rhythmic contraction in synchrony through the heart (Jansen et al., 2010). In comparison to the wild type Cx43 (Cx43+/+), the heterozygous Cx43 (Cx43+/−) caused a reduction in ventricular conduction velocity, suggesting that Cx43 is essential for the electrical coupling in ventricular myocardium (Guerrero et al., 1997). Several previous studies have demonstrated that Cx43 remodeling is related to arrhythmogenesis (Kieken et al., 2009; Li et al., 2012). Alternating status of Cx43 phosphorylation has been found to be responsible for the slow contraction of cardiomyocytes as induced by uremic toxin p-cresol (Peng et al., 2012). Rat cardiomyocytes express PKC␣, ␤II, ␦, ␧, ␩, and ␨ (Mackay and Mochly-Rosen, 2001). PKC␣ and PKC␧ are thought to mediate fibroblast growth factor-induced Cx43 phosphorylation in cardiomyocytes (Doble et al., 2000). In human cardiomyocytes, PKC␣ and PKC␧ can form complexes with Cx43 and promote Cx43 phosphorylation (Bowling et al., 2001). Phosphorylation of Cx43 controls the structure, function, and degradation of gap junctions (Lampe and Lau, 2000; Solan and Lampe, 2009). The PKC activator 12O-tetradecanoylphorbol 13-acetate increases the colocolization of Cx43 with early endosome antigen 1 (EEA1), indicating that PKC activation promotes the endocytosis of Cx43 gap junctions (Fykerud et al., 2012). Phosphorylation of the C-terminus of Cx43 controls GJIC function (Solan and Lampe, 2005). PKC is believed to directly phosphorylate Cx43 at serine (Ser) 368, which decreases GJIC (Lampe and Lau, 2000). Application of a specific PKC␧ inhibitor eV1-2 was observed to completely abrogate Cx43 phosphorylation at Ser368 induced by oleic acid in neonatal cardiomyocytes, providing further evidence of the essential role of PKC␧ in Cx43 Ser368 phosphorylation (Huang et al., 2004). During cardiac ischemia, the accumulation of LPC in ischemic tissue causes a reduction of gap junction coupling (Daleau, 1999). Moreover, a recent study reported that LPC downregulates endothelial Cx43 expression, resulting in inhibition of GJIC (Jia et al., 2009). Using dual whole-cell voltage-clamp technique, LPC has been shown to reduce gap junctional coupling in cardiac cells (Daleau, 1999). Recent electrophysiological study also reported that LPC decreases the beating rate, reduces the peak-peak field potential, and delays the signal propagation (Gizurarson et al., 2012). However, to our knowledge, the involvement of specific kinase activation in LPC-induced Cx43 remodeling and gap junction uncoupling remains unknown. The purpose of this study is to investigate the effect of LPC on gap junction distribution and function, and to provide evidence for LPC-mediated Cx43 phosphorylation. 2. Materials and methods 2.1. Cell culture Neonatal cardiomyocyte cultures were prepared from 1- to 2-day-old Sprague–Dawley rats of both sexes, as described previously (Huang et al., 2004). The ventricles were minced in Ca2+ /Mg2+ -free Hank’s balanced saline solution (HBSS; Invitrogen, Carlsbad, CA, USA), and incubated for 15 min at 37 ◦ C with 0.5 mg/ml of

collagenase type II (Sigma, St. Louis, MO, USA) and 0.6 mg/ml of pancreatin (Sigma) in HBSS. After repetitive dissociation, the combined cell suspension was centrifuged at 1000 × g for 10 min at room temperature, re-suspended and pre-plated in complete culture medium containing minimal essential medium (MEM, Gibco, Grand Island, NY, USA), 10% fetal bovine serum, 100 IU/ml of penicillin, and 100 ␮g/ml of streptomycin for 1 h at 37 ◦ C in a 10 cm culture dish in a 5% CO2 incubator to remove fibroblasts. On the next day, the medium was changed to growth medium (MEM plus 10% calf serum). All experiments were performed on day-2 cultures. The embryonic rat ventricular myocardial H9c2 cells were grown and maintained at 37 ◦ C in a 5% CO2 incubator in Dulbecco’s Modified Eagle’s Medium (DMEM) containing 10% fetal bovine serum, 55 mg/ml sodium pyruvate, 0.2 mM glutamine, 100 IU/ml of penicillin, and 100 ␮g/ml of streptomycin. The medium was renewed every 3 days. Cells reaching a subconfluent state were used for the experiments. 2.2. Reagents and antibodies LPC, 4 -6-diamidino-2-phenylindole (DAPI), propidium iodide (PI), and 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), were obtained from Sigma–Aldrich (St. Louis, MO, USA). Fluoro-3/AM was purchased from Invitrogen life science (Carlsbad, USA). The PKC␧ translocation inhibitor peptide (eV1-2) was bought from Calbiochem (La Jolla, CA, USA).The primary mouse antibodies included anti-total PKC␧, anti-total Cx43 (BD Transduction Laboratories, Lexington, KY, USA) and anti-EEA1 (Cell signaling, Danvers, MA). The primary rabbit antibodies included anti-phospho-Ser729-PKC␧, anti-phospho-Ser368-Cx43 (Cell signaling, Danvers, Ma) and anti-total Cx43 (Abcam Inc., Cambridge, MA; Santa Cruz Biotechnology, Santa Cruz, CA). For Western blotting, all primary antibodies were used at a 1:500 dilution and the secondary antibodies were horseradish peroxide (HRP)-conjugated goat anti-rabbit or anti-mouse IgG antibodies (1:7500 dilution, Millipore). For immunofluorescence studies, the secondary antibodies were DyLight 488-conjugated goat anti-rabbit IG and Texas red-conjugated goat anti-mouse IgG (Sigma–Aldrich, St. Louis, MO, USA). 2.3. Cell survival assay MTT assay was used to examine the cell viability. Briefly, cardiomyocytes (2 × 104 cells/well) were suspended with 200 ␮l media and plated on 24-well cell culture plates. The cells were treated with 0.1% methanol or 10–30 ␮M LPC for 12 h to observe the dose–response relationship. The cells were incubated with MTT substrate solution (0.5 mg/ml of MTT in PBS) for 4 h, and then dissolved in dimethyl sulfoxide. Then the optical density was measured at 590 nm using an ELISA reader (Molecular Devices, San Francisco, CA, USA). 2.4. DAPI and PI staining After treatment with 30 ␮M LPC for 12 h, cells were washed with PBS once, and then vitally stained with PI solution (50 ␮g/ml PI in PBS) for 30 min. The cells were washed with PBS once again and fixed with 2% paraformaldehyde and treated with PBS containing 0.5% Triton X-100 for 10 min. Finally, cells were counterstained with DAPI solution (1 ␮g/ml DAPI in 0.9% NaCl) for 15 min. 2.5. Measurement of the cardiomyocyte contraction rate Confluent day-2 cardiomyocytes were treated with 0.1% methanol and 30 ␮M LPC for 30 min, and examined under an inverted phase microscope. Four to ten synchronously beating cardiomyocytes were monitored in one microscopic field, and the number of contractions in a 30 sec period was counted. Five fields were selected for one coverslip, and three coverslips were used for one experimental group. For estimation of contraction frequency, cardiomyocytes were recorded for 30 s by a camera, and the contraction frequency was counted by a stopwatch online clock. 2.6. Immunofluorescence After various treatments, cardiomyocytes were fixed for 10 min at room temperature in 10% formalin in phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 1.5 mM KH2 PO4 , 8 mM Na2 HPO4 , pH 7.4), then the non-specific binding was blocked by incubation with 5% non-fat milk and 0.1% Triton X-100 in PBS for 30 min. The cells were then incubated at 4 ◦ C overnight with rabbit anti-total Cx43 (Abcam Inc., Cambridge, MA), washed with PBS (3 × 5 min), and incubated for 1 h at 37 ◦ C with FITC-conjugated goat anti-rabbit IgG (Jackson Immunoresearch, West Grove, PA, USA). For double labeling, mixtures of two different primary antibodies [rabbit anti-total Cx43 (Abcam Inc., Cambridge, MA) and mouse anti-ubiquitin (Epitomics, Burlingame, CA, USA)] or rabbit anti-EEA1 (Epitomics, Burlingame, CA, USA) and mouse anti-total Cx43 (BD Transduction Laboratories, Interchim, Montluc¸on, France) were incubated with cells at 4 ◦ C overnight. Cells were washed with PBS, and reacted with secondary antibodies (Dylight-488 conjugated goat anti-rabbit IG and Texas red-conjugated goat anti-mouse IgG) at room temperature for 1 h. Images were taken on a Zeiss Axiophot epifluorescence microscope (Carl Zeiss, Oberkochen, Germany) equipped with a Nikon D1X digital camera (Nikon, Tokyo, Japan).

C.-K. Liao et al. / Toxicology 314 (2013) 11–21 2.7. Scrape loading/dye transfer analysis We evaluated the GJIC by the scrape loading dye transfer technique, which is based on studying the transfer of the fluorescent dye through a cut in the cell monolayer (Opsahl and Rivedal, 2000). Confluent cells in 35 mm-dishes were washed with PBS, and incubated with 1 ml of 0.05% 6-carboxyfluorescein. Linear wounds were made with a knife blade. After 3 min, the cells were washed with PBS several times and the images were recorded under an inverted fluorescence microscope. 2.8. Western blot analysis To prepare whole cell lysates, cardiomyocytes were scraped off the culture dish in a lysis buffer (0.15% Triton X-100, 10 mM EGTA, 2 mM MgCl2 , 60 mM PIPES, 25 mM HEPES, pH 6.9) and sonicated for twenty 10-s pulses, and then the protein concentration of the lysates was determined using a protein assay kit (Bio-Rad, Hercules, CA, USA). Fifty micrograms of protein was applied to each lane of a 10% SDS-polyacrylamide gel, subjected to electrophoresis, and transferred to nitrocellulose paper (Schleicher and Schuell BioSciences, Boston, MA, USA). Membrane strips were blocked for 1 h at room temperature with 5% non-fat milk in Tris-buffered saline (TBS; 150 mM NaCl, 50 mM Tris base, pH 8.2) containing 0.1% Tween 20, then incubated overnight at 4 ◦ C with diluted primary antibodies. They were then incubated for 1 h at room temperature with alkaline phosphatase-conjugated secondary antibodies. The immunoreactive bands were developed using nitro blue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP) as chromogen (3.3 mg/ml of NBT and 1.65 mg/ml of BCIP in 100 mM NaCl, 5 mM MgCl2 , 100 mM Tris, pH 9.5). The blots were then stripped with 25 mM glycine–HCl, pH 2.0, 1% (w/v) SDS, reprobed with mouse antibodies against total PKC␧ or rabbit antibodies against total Cx43, and followed by incubation with peroxidase-conjugated secondary antibodies for chemiluminescence detection. After a buffer wash, blots were reacted in ECL substrate developing solution (Millipore). Densitometry was performed using Gel Pro 3.1 (Media Cybernetics, Silver Spring, MD, USA). All experiments were performed at least three times and the values were expressed as the mean ± SD.

13

centrifuged at 16,000 × g for 10 min at 4 ◦ C. The supernatants in 1.5 ml eppendorf were precleared with 10 ␮l Protein G Mag Sepharose (GE healthcare) for 30 min at 4 ◦ C on a shaker. The precleared cell lysates were incubated with primary antibodies (1 ␮g of affinity-purified antibodies for 400 ␮g of protein) against total Cx43 (Santa Cruz Biotechnology, Santa Cruz CA) at 4 ◦ C for 16 h on a shaker, then Protein G Mag Sepharose (GE healthcare) was added and the mixture incubated for 1 h at 4 ◦ C on a shaker. After centrifugation at 16,000 × g for 30 s, the immunoprecipitates were washed, re-suspended in RIPA buffer, dissolved in SDS gel sample buffer, applied to a lane of an SDS gel, and analyzed by Western blotting using antibodies against mouse anti-ubiquitin (Cell signaling, Danvers, MA). 2.10. Triton X-100 soluble and insoluble extraction Primary cultured cardiomyocytes were harvested in PBS and centrifuged at 3000 × g for 10 min at 4 ◦ C. The pellet was resuspended in an extraction buffer (50 mM Tris–base, 150 mM NaCl, 1% Nonidet-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM PMSF, 1 mM NaF, 1 mM Na3 VO4 , 1 ␮g/ml of aprotinin, leupeptin, and pepstatin, pH 7.4) containing 1% Triton X-100, sonicated, incubated for 30 min on ice, and centrifuged at 14,000 × g for 30 min at 4 ◦ C. The supernatant was defined as the Triton X-100 soluble fraction. The Triton X-100 insoluble pellet was lysed in an extraction buffer containing 0.5% SDS, which was subsequently sonicated and incubated for 30 min on ice. Equal volumes of Triton X-100-soluble and -insoluble fractions were analyzed by western blotting. 2.11. Small Interfering RNA (siRNA) PKC␧ siRNAs targeting the mRNA coding sequence and negative control siRNA were designed by Ambion (Silencer® Select Pre-designed siRNA, Ambion, Austin, TX, USA). Electroporation was used to transfect siRNAs according to the manufacturer’s instructions (Amaxa, Germany). Briefly, 4 × 105 H9c2 cells were trypsinized and resuspended in 100 ␮l of Nucleofector solution (Amaxa, Germany), and 100 nM of siRNA duplexes were electroporated. Protein expression was measured at 24 h after electroporation.

2.9. Immunoprecipitation 2.12. Statistics Cultured cardiomyocytes were extracted on ice for 30 min with RIPA (Radioimmunoprecipitation assay) buffer (50 mM Tris–HCl, 1% Nonidet-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM PMSF, 1 mM NaF, 1 mM Na3 VO4 , 1 ␮g/ml of aprotinin, leupeptin, and pepstatin, pH 7.4) containing 0.1% SDS, ultra-sonicated, and

Results are presented as the mean ± SD for at least three replicate experiments. Statistical differences between means were evaluated using Student’s t-test. Differences were considered significant when p < 0.05.

Fig. 1. Effect of different LPC concentrations on cardiomyocyte survival and Cx43 distribution. (A) Cardiomyocytes were treated with 0.1% methanol (vehicle control), 10 ␮M, 20 ␮M, or 30 ␮M LPC for 12 h and then assayed by MTT test (N = 3). (B) The cells were treated with 30 ␮M LPC for 12 h, and vitally stained with DAPI (a and c) and PI (b and d). Bar = 40 ␮m. (C) Cardiomyocytes were treated with 0.1% methanol (a), 20 ␮M (b), 30 ␮M (c) or 100 ␮M (d) LPC for 12 h, immunostained for total Cx43 and counterstained with DAPI for nuclei. N, nucleus. Arrows, gap junction plaques. Bar = 20 ␮m. (D) Quantitative data on optical density analysis of Cx43 staining. ** p < 0.01 compared with the methanol-treated groups. N = 10.

14

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

3. Results 3.1. Effects of LPC on the distribution of Cx43 MTT test showed that treatment with 10, 20 or 30 ␮M LPC for 12 h did not affect cell survival (Fig. 1A). Nuclear morphology revealed by DAPI and PI vital staining showed that there was no significant cell death occurring at 30 ␮M as compared to the control (Fig. 1B, a–d). The dose-dependent effect on GJIC was examined by a scrapeloading assay with the 6-CF fluorescence dye. A considerable dye transfer was observed in 0.1% methanol-treated cardiomyocytes (Supplementary Fig. 1A). The transfer distance was significantly decreased after incubation with 30 ␮M LPC for 30 min, whereas 10 and 20 ␮M LPC did not affect the GJIC (Supplementary Fig. 1B-E). Thus, 30 ␮M LPC was the minimal effective concentration to affect cardiomyocyte function. The concentration- and time-dependent immunofluorescence studies were also performed to examine the effects of LPC on junctional Cx43 distribution in cardiomyocytes. In 0.1% methanoltreated cardiomyocytes, Cx43 staining appeared as large plaques

at intercellular junctions (Supplementary Fig. 2A; Fig. 1C-a). In response to 30 ␮M LPC treatment, junctional Cx43 staining was slightly decreased in the 1 h- and 6 h-treated groups, and became tiny dotted signals in 12 h-treated group (Supplementary Fig. 2B, 1C). When cardiomyocytes were incubated with different concentrations of LPC for 12 h, we detected significant changes of staining intensity and subcellular distribution of Cx43. The intensity of junctional Cx43 staining was decreased at 20 ␮M LPC (Fig. 1C-b). At 30 ␮M, Cx43 staining appeared in the cytoplasm as tiny dots in addition to the dotted staining pattern at cell junctions (Fig. 1C-c). At 100 ␮M, the Cx43 dots became much smaller both in the cytoplasm and at cell junctions (Fig. 1C-d). The LPC-induced decrease in junctional Cx43 staining was supported by the quantitative data (Fig. 1D). 3.2. Effect of LPC on cardiomyocyte beating This defect in GJIC was consistent with the slow and arrhythmic contraction of cardiomyocytes. In the control groups, the contraction rate of cardiomyocytes was 42 ± 5/30 s. LPC treatment not only reduced the contraction rate of cardiomyocytes (18.5 ± 6.3/30 sec)

Fig. 2. Effects of LPC treatment on cardiomyocyte contraction frequency and on dye coupling. Cells were treated with 0.1% methanol, 30 ␮M LPC or 30 ␮M LPC plus 1 ␮M eV1-2 for 30 min, and analyzed for contraction frequency. (A) A set of typical examples was shown. (B) Contraction rate. N = 8. (C) Cardiomyocytes treated with 0.1% methanol (a), 30 ␮M LPC (b) or 30 ␮M LPC plus 1 ␮M eV1-2 for 30 min (c) were processed for scrape loading/dye transfer assay. Bar = 100 ␮m. (d) Quantitative data on the dye-spreading distance. ** p <0.01 compared with the methanol-treated groups. ## p < 0.01 compared with the LPC group. N = 3.

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

(Fig. 2A and B), but also caused an irregular contraction (Fig. 2A and B). Pretreatment with eV1-2, a specific PKC␧ translocation inhibitor (Johnson et al., 1996), completely prevented LPC-induced decrease in contraction rates (47.3 ± 4.8/30 s) (Fig. 2A and B). 3.3. Effect of LPC on gap junction intercellular communication (GJIC) We used dye coupling assay to measure the diffusion distance of fluorescein dyes from the cutting edge to adjacent cells. Cardiomyocytes were treated with 0.1% methanol or 30 ␮M LPC for 30 min, and assayed for GJIC. LPC decreased the function of GJIC by 50% as compared to methanol-treated control (Fig. 2C-a, C-b and C-d). Pretreatment with a specific PKC␧ inhibitor eV1-2 (1 ␮M) effectively blocked LPC-induced dye uncoupling (Fig. 2C-c and C-d), suggesting the involvement of PKC␧ in this event. 3.4. Effect of PKC␧ inhibition on Cx43 distribution Previous studies have shown that activation of different PKC isoforms is responsible for Cx43 phosphorylation and function of GJIC (Lampe and Lau, 2000; Saez et al., 1997; Shah et al., 2002). PKC␧ has been demonstrated to induce phosphorylation of Cx43 at Ser368 and to cause disassembly of gap junction (Huang et al., 2004; Lampe and Lau, 2000). Several PKC isoform inhibitors were used to examine their effects on preventing LPC-induced loss of junctional Cx43 staining. Among them, a PKC␧ inhibitor eV1-2, showed a significant blockade effect. In control cardiomyocytes, Cx43 localized to gap junction plaques at appositional plasma membrane (Fig. 3A-a). In the LPC-treated groups, the size of Cx43-containing gap junction plaques were reduced; meanwhile, the cytoplasmic dots were apparently increased (Fig. 3A-b). Treatment with eV1-2 alone yielded a normal distribution of Cx43 at intercellular junctions (Fig. 3A-c). Compared to the weak Cx43 staining in LPC-treated groups (Fig. 3A-b), pretreatment with eV1-2 effectively prevented LPC-induced decrease in Cx43

15

junctional staining (Fig. 3A-d). Quantitative analyses supported the immunofluorescence observations (Fig. 3B). Prior research has shown that Triton X-100-insoluble Cx43, mostly the Cx43-P2 form, is localized in the gap junctional plaques, whereas Triton X-100-soluble Cx43-P1 and -P0 are localized in the subcellular compartment (Musil and Goodenough, 1991). Compared to the methanol-treated control, LPC treatment for 1 h increased Triton X-100-soluble Cx43 levels and decreased Triton X-100-insoluble Cx43 levels (Fig. 3C). Quantitative analyses also indicated that LPC treatment induced an increase in the ratio of Triton X-100-insoluble Cx43 to Triton X-100-soluble Cx43. This phenomenon was effectively blocked by pretreatment of eV1-2 (Fig. 3D). 3.5. LPC activated PKC␧ and induced Cx43 phosphorylation To examine the activation of PKC␧ by LPC, a time course study was carried out. Phospho-PKC␧ levels started to increase to 2–2.5 fold at 15 min after LPC treatment (Fig. 4A). Accordingly, the levels of phospho-Ser368-Cx43 were increased to 1.5-fold at 30 min after LPC treatment (Fig. 4B). The LPC-induced increase in phospho-Ser368-Cx43 levels was blocked by pretreatment with eV1-2 (Fig. 4C). 3.6. Effects of LPC on PKC␧ activation and Cx43 phosphorylation in H9c2 cells We next set out the siRNA approach to further confirm the roles of PKC␧ in LPC-induced gap junction disassembly in a cardiomyocyte cell line H9c2. Immunoblotting analysis using phospho-specific antibodies against phospho-Ser368-Cx43 and phospho-PKC␧ showed that LPC-induced PKC␧ activation peaked at 30 min (Fig. 5A). The same trend was also observed for the phosphorylation of Ser368-Cx43 (Fig. 5A). Pre-treatment with eV1-2 blocked the LPC-induced phosphorylation of PKC␧ and Ser368-Cx43 (Fig. 5B). These results confirm that LPC also induced

Fig. 3. Inhibition of PKC␧ prevents the distributional change of Cx43 induced by LPC. (A) Cardiomyocytes were treated for 12 h with 0.1% methanol (a), 30 ␮M LPC (b), 1 ␮M eV1-2 alone (c) or 30 ␮M LPC plus 1 ␮M eV1-2 (d), and immunostained for total Cx43. N, nucleus. Arrows, gap junction plaques. Bar = 20 ␮m. N = 3. (B) Quantitative data on optical density analysis of Cx43 staining. ** p < 0.01 compared with the methanol groups. ## p < 0.01 compared with the LPC-treated groups. N = 10. (C) Cardiomyocytes were treated for 1 h with 0.1% methanol, 30 ␮M LPC, or 30 ␮M LPC plus 1 ␮M eV1-2. The cell lyates were collected and separated into Triton X-100 soluble fractions (Tx-S) and insoluble fractions (Tx-I), as described in Materials and Methods. The equal volumes of two fractions were analyzed by immunoblotting with antibodies against total Cx43 and GAPDH. (D) Quantitative data on Tx-S Cx43/Tx-I Cx43. * p <0.05 compared with the methanol-treated controls. # p < 0.05 compared with the LPC groups. N = 3.

16

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

phosphorylation of Cx43 at Ser368 through PKC␧ activation in H9c2 cells. LPC treatment induced Cx43 phosphorylation at Ser368 in the siRNA negative groups (Fig. 5D). Although the siRNA knockdown efficiency of PKC␧ was around 55% (Fig. 5C), PKC␧ knockdown effectively blocked LPC-induced Cx43 phoshorylation at Ser368 (Fig. 5D). 3.7. LPC induced Cx43 gap junction disassembly via the proteasomal pathway To investigate the role of LPC in gap junction disassembly, double immunofluorescence for Cx43 and EEA1 was performed to examine LPC-induced gap junction endocytosis. As shown in Fig. 6A, EEA1 staining was rarely colocalized with Cx43 in control cardiomyocytes (Fig. 6A, control), whereas LPC treatment for 12 h induced the disassembly of gap junctions and augmented the colocalization of Cx43 with EEA1 (Fig. 6A, LPC). A previous study has demonstrated that Cx43 was degraded intracellularly by both proteasomes and lysosomes (Laing et al., 1997). In order to examine how cytoplasmic Cx43 was degraded, we used specific inhibitors for proteasomes and lysosomes. Bafilomycin A1 is a specific inhibitor of vacuolar proton-ATPase, and has been used to inhibit the endosome-lysosome system (Bayer et al., 1998). Pretreatment with bafilomycin A1 (10 ␮M) showed partial prevention effect on LPCinduced gap junction disassembly (Fig. 6B, LPC + BFA1). MG132 can reversibly block the activity of the 26S proteasome (TomicCarruthers and Gorden, 1998). It was noted that MG132 (5 ␮M) effectively blocked the LPC-induced junctional Cx43 disassembly and prevented the increase of cytoplasmic dots (Fig. 6B, LPC and LPC + MG132). Treatment with MG132 or bafilomycin A1 alone did not affect the distribution of Cx43 at gap junctions (Fig. 6B, MG132 and BFA1). Quantitative data on junctional Cx43 staining supported the immunofluorescence observations (Fig. 6C). Moreover, double immunofluorescence staining for ubiquitin and Cx43 revealed that LPC treatment increased the colocalization of these two proteins (Fig. 7A, LPC). The expression of ubiquitinated Cx43 was further examined by immunoprecipitation using antibodies against Cx43 and immunoblotting using antibodies against ubiquitin. LPC caused a significant increase in ubiquitinated Cx43 levels (Fig. 7B, C, LPC 1 h). Pretreatment of eV1-2 effectively prevented the LPC-induced ubiquitination of Cx43 (Fig. 7B, C, LPC + eV12). These results indicate that the disassembly of Cx43 gap junctions induced by LPC was mediated mainly by the ubiquitinproteasomal pathway, and partially by the endosome-lysosome pathway. 4. Discussion This study elucidates the underlying mechanism responsible for the LPC-induced reduction in GJIC. LPC induced activation of PKC␧, which caused phosphorylation of Cx43 at Ser368, and resulted in the spatial redistribution of gap junctional Cx43. LPC is released from cell membranes during ischemia, and the accumulation of LPC in ischemic myocardium has been associated

Fig. 4. LPC induces activation of PKC␧ and increases the levels of phospho-Ser368Cx43. (A) Whole cell lysates from cardiomyocytes treated with 30 ␮M LPC for 0, 15, 30, 60 or 90 min were analyzed for phospho-Ser729-PKC␧ (pPKC␧) or total PKC␧ (total PKC␧). Lower panel: Densitometric analyses. The optical density of phosphoSer729-PKC␧ was normalized to that of total PKC␧ loading control. * p < 0.05 compared to the time zero. (B) Cardiomyocytes were treated with 30 ␮M LPC for

0, 15, 30 or 60 min. Whole cell lysates were analyzed for phospho-Ser368-Cx43 (pSer368-Cx43) or total Cx43 (total Cx43). Lower panel: Densitometric analyses. The optical density of pSer368-Cx43 was normalized to that of total Cx43 (P0 plus P1 plus P2) loading control. * p < 0.05 compared with the time zero. (C) Cardiomyocytes were treated for 30 min with 0.1% methanol, 30 ␮M LPC, or 30 ␮M LPC plus 1 ␮M eV1-2. Whole cell lysates were analyzed for phospho-Ser368-Cx43 (pSer368Cx43) or total Cx43 (total Cx43). Lower panel: Densitometric analyses. The optical density of pSer368-Cx43 was normalized to that of total Cx43 (P0 plus P1 plus P2) loading control. ** p < 0.01 compared with the methanol-treated controls. ## p < 0.01 compared with the LPC-treated groups.

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

17

Fig. 5. LPC-mediated PKC␧ activation induces phosphorylation of Ser368-Cx43 in H9c2 cells. (A) Whole cell lysates from H9c2 cells treated with 30 ␮M LPC for 0, 15, 30, or 60 min were analyzed for phospho-Ser729-PKC␧ (pPKC␧), total PKC␧ (total PKC␧), phospho-Ser368-Cx43 (pSer368-Cx43) or total Cx43 (total Cx43). Lower panel: Densitometric analyses. The optical density of the phosphoprotein was normalized to that of total protein. * p < 0.05, ** p < 0.01 compared with the time-zero controls. N = 3. (B) H9c2 cells were treated for 30 min with 0.1% methanol, 30 ␮M LPC, 30 ␮M LPC plus 1 ␮M eV1-2, or 1 ␮M eV1-2 alone. Whole cell lysates were analyzed for phospho-Ser729PKC␧ (pPKC␧), total PKC␧ (total PKC␧), phospho-Ser368-Cx43 (pSer368-Cx43), or total Cx43 (total Cx43). Lower panel: Densitometric analyses. The optical density of the phosphoprotein was normalized to that of total protein. * p < 0.05 compared with the methanol-treated controls. # p < 0.05, ## p < 0.01 compared with the LPC-treated groups. N = 3. (C and D) H9c2 cells were transfected with non-targeting siRNA or PKC␧ siRNA for 24 h followed by treatment for 0.1% methanol or 30 ␮M LPC for 60 min. Whole cell lysates were analyzed for phospho-Ser729-PKC␧ (pPKC␧), total PKC␧ (total PKC␧) (C) or phospho-Ser368-Cx43 (pSer368-Cx43), total Cx43 (total Cx43) (D). Lower panel: Densitometric analyses. The optical density of the phosphoprotein was normalized to that of total protein. ** p < 0.01 compared with the methanol-treated non-targeting siRNA controls. ## p < 0.01 compared with the LPC-treated non-targeting siRNA groups. N = 4.

with arrhythmia (Sobel et al., 1978). It has been suggested that calcium overload and electrolyte abnormalities induced by LPC could contribute to irregular contraction in ischemic hearts (Hashizume and Abiko, 1999; Woodley et al., 1991; El-Sherif and Turitto, 2011). Moreover, LPC could upregulate T-type Ca2+ channels and lead to an elevation of intracellular calcium levels and PKC␣ activity, thereby triggering arrhythmia in neonatal rat cardiomyocytes (Zheng et al., 2010). Here, we demonstrated that LPC-induced PKC␧-dependent serine phosphorylation of Cx43 plays a pathological role in the generation of arrhythmia. Several studies have addressed the role of LPC in gap junction uncoupling and disassembly. For example, LPC treatment noticeably downregulated Cx43 expression and disturbed GJIC function in cultured human umbilical vein endothelial cells (Jia et al.,

2009). Electrophysiological studies indicated that LPC accumulation in ischemic hearts impaired the electrical coupling of gap junctions (Daleau, 1999). Cx43 phosphorylation may regulate GJIC, via assembly/disassembly, trafficking, degradation, and channel opening or closure (Solan and Lampe, 2009). Classical PKC isoforms (PKC␣, ␤, ␥) and novel PKC isoforms (PKC␦, ␧, ␩, ␪) are known to be activated by diacylglycerol (Carter, 2000). It is possible that LPC directly activates PKC␧ or indirectly induces PKC␧ activation by diacylglycerol (Prokazova et al., 1998). Previous studies have shown that PKC␧ activation can elicit a loss of functional gap junctions in cardiomyocytes (Huang et al., 2004; Liang et al., 2008), suggesting that the PKC␧ pathway may participate in LPC-mediated inhibition of GJIC. In this study, we provided novel evidence that PKC␧ activation is responsible for gap junction uncoupling and

18

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

Fig. 6. LPC induces Cx43 internalization and disassembly in rat cardiomyocytes. (A) Cardiomyocytes treated with 0.1% methanol or 30 ␮M LPC for 12 h were double-stained with antibodies against total Cx43 (red, left panels) and EEA1 (green, central panels), and were counterstained with DAPI (blue) for nuclei. N, nucleus. Merged images show colocalization. Arrows indicate Cx43 gap junction plaques. Inset: enlarged Cx43-positive early endosomes (arrowheads) (bars = 10 ␮m). (B) Cardiomyocytes treated for 12 h with 0.1% methanol (control), 5 ␮M MG132 (MG132), 10 ␮M BafilomycinA1 (BFA1), 30 ␮M LPC (LPC), 30 ␮M LPC plus 5 ␮M MG132 (LPC + MG132) or 30 ␮M LPC plus 10 ␮M BafilomycinA1 (LPC + BFA1) were immunostained for total Cx43 and counterstained with DAPI for nuclei. N, nucleus. Arrows indicate Cx43 gap junction plaques. Bar = 10 ␮m. (C) Quantitative data on Cx43 staining. ** p < 0.01 compared with the methanol groups. # p < 0.05, ## p < 0.01 compared with the LPC-treated groups. N = 10.

disassembly in LPC-treated cardiomyocytes. Our finding that LPC treatment slowed down the contraction of cardiomyocytes implies that aberrant gap junction function may contribute to the pathological process of arrhythmia. Thus, the development of specific PKC␧ inhibitors will be a great benefit to antiarrhythmic therapy. Increased Cx43 phosphorylation by PKC is associated with decreased gap junction coupling in some cell types (Lampe and Lau, 2004). PKC␧ colocalizes with Cx43 in neonatal rat cardiomyocytes (Doble et al., 2000), and can directly phosphorylate Cx43 at Ser368 (Lampe and Lau, 2000; Saez et al., 1997; Shah et al., 2002) and Ser372 (Saez et al., 1997). Ser368 phosphorylation has been associated with decreased GJIC (Lampe and Lau, 2000), as well as gap junction disassembly (Huang et al., 2004). Mutagenesis analysis also revealed that PKC activation mediates the downregulation of gap junction channels via Ser368 phosphorylation (Lampe

et al., 2000), and leads to GJIC suppression in ischemic hearts (EkVitorin et al., 2006; Hund et al., 2007). Using siRNA and a specific inhibitor peptide eV1-2 to inhibit PKC␧ activity, we demonstrated that LPC induces Cx43 Ser368 phosphorylation via PKC␧. Activation of PKC␣, a conventional PKC isoform, by uremic toxin p-cresol leads to the decrease of GJIC and arrhythmic contraction (Peng et al., 2012). Thus, we cannot rule out the possibility that other PKC isoforms may also be involved in the inhibitory effect of LPC on Cx43-mediated GJIC in cardiomycytes. PKC activation causes a strong increase in Cx43 ubiquitination, which then regulates the endocytosis and post-endocytic delivery of Cx43 (Fykerud et al., 2012). Ubiquitin is a small protein formed by 76 conserved amino acids, whose molecular weight is about 8 kD (Kjenseth et al., 2010). Three phosphorylated forms (P0, P1, P2) of Cx43 migrate at the 42–46 kD range (Musil and Goodenough,

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

19

Fig. 7. LPC induces Cx43 ubiquitination in rat cardiomyocytes. (A) Methanol- or LPC-treated cells were double-stained for total Cx43 (green, left panels) and ubiquitin (red, central panels). Nuclei were counterstained with DAPI (blue). N, nucleus. Arrows indicate ubiquitinated Cx43. Bar = 10 ␮m. (B) Whole cell lysates from cardiomyocytes treated for 1 h with 0.1% methanol, 30 ␮M LPC or 30 ␮M LPC plus 1 ␮M eV1-2 were immunoprecipitated with antibodies against total Cx43. The immunoprecipitates were analyzed for ubiquitin or rabbit IgG. (C) Densitometric analyses. The optical density of ubiquitinated Cx43 was normalized to that of rabbit IgG loading control. * p < 0.05 compared with the methanol-treated controls. # p < 0.05 compared with the LPC-treated group. N = 3.

1991). In our study, the ubiquitinized Cx43 was located at 50–70 kD range, indicating that at least one to three ubiquitins are covalently bound to Cx43. Fusion of Cx43 to ubiquitin has been shown to significantly decrease the half-life of endogenous Cx43 and to promote the internalization of Cx43 (Catarino et al., 2011). Therefore, it is difficult to visualize ubiquitinated Cx43 by immunofluorescence. However, a specific PKC␧ inhibitor eV1-2 prevented the LPC-induced increase of ubiquitinated Cx43, suggesting that ubiquitination of Cx43 occurred later to the serine phosphorylation of Cx43. Thus, the exact mechanisms for PKC␧ activation and PKC␧induced phosphorylation of Cx43 and ubiquitination of Cx43 by LPC require further investigation. Our immunofluorescence results indicate that LPC decreases the size of gap junction plaques, demonstrating that LPC induced Cx43 gap junction disassembly. LPC also increased the levels of Triton X-100 soluble Cx43-P1 and -P0, which agrees with the increase in the cytoplasmic Cx43 dots after LPC treatment. Under normal physiological condition, older intercellular channels are removed from the center region of exsiting plaques into one of the neighboring cells to form double-membrane vesicles called annular gap junctions via the clathrin-dependent endocytic pathway (Gumpert et al., 2008). In addition to direct fusion with lysosomes, the internalization of Cx43 gap junctions has multiple endocytic pathways en route to lysosomes mediated by autophagosomallysosomal fusion and endosomal-lysosomal system (Leithe et al., 2012). In our immunofluorescence study using antibodies against

total Cx43 and early endosome marker EEA1, the internalized Cx43 were partially colocalizes with early endosomes, suggesting that other post-endocytic pathways may also involve in LPC-induced gap junction disassembly. It has been thoroughly researched that both lysosomes and proteasomes are involved in the degradation of Cx43 gap junction (Laing et al., 1997). The degradation process of lysosomes and proteasomes is associated with the ubiquitination of Cx43, which acts as a sorting signal in the regulation of Cx43 trafficking during endocytosis (Fykerud et al., 2012). In the present study, pretreatment with the acidification inhibitor bafilomycin A1, which blocks the process of early endosomes to late endosomes, partially prevented the effect of LPC on gap junction disassembly. This suggests that internalized Cx43 might undergo recycling to gap junctions. On the other hand, a previous study showed that the inhibition of proteasome activity by MG132 was able to cause Cx43 accumulation at the gap junction due to the increase in the interaction between Cx43 and tight junction protein zonula occludens-1 (ZO-1), which indicates that proteasomes play a regulatory role in stability of gap junctional channels (Girao and Pereira, 2007). Our observation that the proteasome inhibitor MG132 effectively prevented LPC-induced gap junction disassembly in cardiomyocytes suggests that the LPC-mediated activation of the ubiquitin-proteasome system was responsible for the formation of small gap junction plaques, presumably by enhancing the interaction between Cx43 and its partner proteins such as ZO-1.

20

C.-K. Liao et al. / Toxicology 314 (2013) 11–21

In conclusion, we have demonstrated that LPC induces gap junction disassembly and uncoupling, leading to an arrhythmic contraction via a pathway involving PKC␧ activation in rat primary cardiomyocytes. These findings also provide novel evidence indicating that LPC-induced PKC␧ activation causes Cx43 phosphorylation at Ser368. All LPC-mediated effects on Cx43 gap junction are prevented by the specific PKC␧ inhibitor eV1-2. Further study on detailed signaling mechanisms will facilitate the pharmacological exploitation of gap junctions for arrhythmia. Conflict of interest The authors declare that there are no conflicts of interest. Acknowledgment This study was supported by the grant from the National Science Council (Grant number NSC-98-2314-B-002-14). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.tox.2013.08.001. References Arnsdorf, M.F., Sawicki, G.J., 1981. The effects of lysophosphatidylcholine, a toxic metabolite of ischemia, on the components of cardiac excitability in sheep Purkinje fibers. Circ. Res. 49, 16–30. Bayer, N., Schober, D., Prchla, E., Murphy, R.F., Blaas, D., Fuchs, R., 1998. Effect of bafilomycin A1 and nocodazole on endocytic transport in HeLa cells: implications for viral uncoating and infection. J. Virol. 72, 9645–9655. Bowling, N., Huang, X., Sandusky, G.E., Fouts, R.L., Mintze, K., Esterman, M., Allen, P.D., Maddi, R., McCall, E., Vlahos, C.J., 2001. Protein kinase C-alpha and -epsilon modulate connexin-43 phosphorylation in human heart. J. Mol. Cell. Cardiol. 33, 789–798. Catarino, S., Ramalho, J.S., Marques, C., Pereira, P., Girao, H., 2011. Ubiquitinmediated internalization of connexin43 is independent of the canonical endocytic tyrosine-sorting signal. Biochem. J. 437, 255–267. Carter, C.A., 2000. Protein kinase C as a drug target: implications for drug or dietprevention and treatment of cancer. Curr. Drug Targets 1, 163–183. Clarkson, C.W., Ten Eick, R.E., 1983. On the mechanism of lysophosphatidylcholineinduced depolarization of cat ventricular myocardium. Circ. Res. 52, 543–556. Corr, P.B., Snyder, D.W., Lee, B.I., Gross, R.W., Keim, C.R., Sobel, B.E., 1982. Pathophysiological concentrations of lysophosphatides and the slow response. Am. J. Physiol. 243, H187–H195. Daleau, P., 1999. Lysophosphatidylcholine, a metabolite which accumulates early in myocardium during ischemia, reduces gap junctional coupling in cardiac cells. J. Mol. Cell. Cardiol. 31, 1391–1401. Doble, B.W., Ping, P., Kardami, E., 2000. The epsilon subtype of protein kinase C is required for cardiomyocyte connexin-43 phosphorylation. Circ. Res. 86, 293–301. Ek-Vitorin, J.F., King, T.J., Heyman, N.S., Lampe, P.D., Burt, J.M., 2006. Selectivity of connexin 43 channels is regulated through protein kinase C-dependent phosphorylation. Circ. Res. 98, 1498–1505. El-Sherif, N., Turitto, G., 2011. Electrolyte disorders and arrhythmogenesis. Cardiol. J. 18, 233–245. Fykerud, T.A., Kjenseth, A., Schink, K.O., Sirnes, S., Bruun, J., Omori, Y., Brech, A., Rivedal, E., Leithe, E., 2012. Smad ubiquitination regulatory factor-2 controls gap junction intercellular communication by modulating endocytosis and degradation of connexin43. J. Cell. Sci. 125, 3966–3976. Girao, H., Pereira, P., 2007. The proteasome regulates the interaction between Cx43 and ZO-1. J. Cell. Biochem. 102, 719–728. Gizurarson, S., Shao, Y., Miljanovic, A., Ramunddal, T., Boren, J., Bergfeldt, L., Omerovic, E., 2012. Electrophysiological effects of lysophosphatidylcholine on HL-1 cardiomyocytes assessed with a microelectrode array system. Cell. Physiol. Biochem. 30, 477–488. Guerrero, P.A., Schuessler, R.B., Davis, L.M., Beyer, E.C., Johnson, C.M., Yamada, K.A., Saffitz, J.E., 1997. Slow ventricular conduction in mice heterozygous for a connexin43 null mutation. J. Clin. Invest. 99, 1991–1998. Gumpert, A.M., Varco, J.S., Baker, S.M., Piehl, M., Falk, M.M., 2008. Double-membrane gap junction internalization requires the clathrin-mediated endocytic machinery. FEBS. Lett. 582, 2887–2892. Hashizume, H., Abiko, Y., 1999. Cardiac cell injury induced by lysophosphatidylcholine. Nihon Yakurigaku Zasshi 114, 287–293. Hoque, A.N., Hoque, N., Hashizume, H., Abiko, Y., 1995. A study on dilazep: II. Dilazep attenuates lysophosphatidylcholine-induced mechanical and metabolic derangements in the isolated, working rat heart. Jpn. J. Pharmacol. 67, 233–241.

Huang, Y.S., Tseng, Y.Z., Wu, J.C., Wang, S.M., 2004. Mechanism of oleic acid-induced gap junctional disassembly in rat cardiomyocytes. J. Mol. Cell. Cardiol. 37, 755–766. Hund, T.J., Lerner, D.L., Yamada, K.A., Schuessler, R.B., Saffitz, J.E., 2007. Protein kinase Cepsilon mediates salutary effects on electrical coupling induced by ischemic preconditioning. Heart Rhythm. 4, 1183–1193. Jansen, J.A., van Veen, T.A., de Bakker, J.M., van Rijen, H.V., 2010. Cardiac connexins and impulse propagation. J. Mol. Cell. Cardiol. 48, 76–82. Jia, S.J., Zhou, Z., Zhang, B.K., Hu, Z.W., Deng, H.W., Li, Y.J., 2009. Asymmetric dimethylarginine damages connexin43-mediated endothelial gap junction intercellular communication. Biochem. Cell. Biol. 87, 867–874. Johnson, J.A., Gray, M.O., Chen, C.H., Mochly-Rosen, D., 1996. A protein kinase C translocation inhibitor as an isozyme-selective antagonist of cardiac function. J. Biol. Chem. 271, 24962–24966. Jung, S., Lee, Y., Han, S., Kim, Y., Nam, T., Ahn, D., 2008. Lysophosphatidylcholine increases Ca current via activation of protein Kinase C in rabbit portal vein smooth muscle cells. Kor. J. Physiol. Pharmacol. 12, 31–35. Kieken, F., Mutsaers, N., Dolmatova, E., Virgil, K., Wit, A.L., Kellezi, A., Hirst-Jensen, B.J., Duffy, H.S., Sorgen, P.L., 2009. Structural and molecular mechanisms of gap junction remodeling in epicardial border zone myocytes following myocardial infarction. Circ. Res. 104, 1103–1112. Kjenseth, A., Fykerud, T., Rivedal, E., Leithe, E., 2010. Regulation of gap junction intercellular communication by the ubiquitin system. Cell. Signal. 22, 1267– 1273. Laing, J.G., Tadros, P.N., Westphale, E.M., Beyer, E.C., 1997. Degradation of connexin43 gap junctions involves both the proteasome and the lysosome. Exp. Cell Res. 236, 482–492. Lampe, P.D., Lau, A.F., 2000. Regulation of gap junctions by phosphorylation of connexins. Arch. Biochem. Biophys. 384, 205–215. Lampe, P.D., Lau, A.F., 2004. The effects of connexin phosphorylation on gap junctional communication. Int. J. Biochem. Cell. Biol. 36, 1171–1186. Lampe, P.D., TenBroek, E.M., Burt, J.M., Kurata, W.E., Johnson, R.G., Lau, A.F., 2000. Phosphorylation of connexin43 on serine368 by protein kinase C regulates gap junctional communication. J. Cell. Biol. 149, 1503–1512. Leithe, E., Sirnes, S., Fykerud, T., Kjenseth, A., Rivedal, E., 2012. Endocytosis and post-endocytic sorting of connexins. Biochim. Biophys. Acta 1818, 1870– 1879. Li, C., Meng, Q., Yu, X., Jing, X., Xu, P., Luo, D., 2012. Regulatory effect of connexin 43 on basal Ca2+ signaling in rat ventricular myocytes. PloS One 7, e36165. Li, Z., Vance, D.E., 2008. Phosphatidylcholine and choline homeostasis. J. Lipid Res. 49, 1187–1194. Liang, J.Y., Wang, S.M., Chung, T.H., Yang, S.H., Wu, J.C., 2008. Effects of 18glycyrrhetinic acid on serine 368 phosphorylation of connexin43 in rat neonatal cardiomyocytes. Cell Biol. Int. 32, 1371–1379. Ma, H., Hashizume, H., Hara, A., Yazawa, K., Abiko, Y., 1999. Protective effect of quinaprilat, an active metabolite of quinapril, on Ca2+-overload induced by lysophosphatidylcholine in isolated rat cardiomyocytes. Jpn. J. Pharmacol. 79, 17–24. Mackay, K., Mochly-Rosen, D., 2001. Arachidonic acid protects neonatal rat cardiac myocytes from ischaemic injury through epsilon protein kinase C. Cardiovasc. Res. 50, 65–74. Man, R.Y., Wong, T., Choy, P.C., 1983. Effects of lysophosphoglycerides on cardiac arrhythmias. Life Sci. 32, 1325–1330. Matsumoto, T., Kobayashi, T., Kamata, K., 2007. Role of lysophosphatidylcholine (LPC) in atherosclerosis. Curr. Med. Chem. 14, 3209–3220. Musil, L.S., Goodenough, D.A., 1991. Biochemical analysis of connexin43 intracellular transport, phosphorylation, and assembly into gap junctional plaques. J. Cell. Biol. 115, 1357–1374. Opsahl, H., Rivedal, E., 2000. Quantitative determination of gap junction intercellular communication by scrape loading and image analysis. Cell Adhes. Commun. 7, 367–375. Parthasarathy, S., Barnett, J., 1990. Phospholipase A2 activity of low density lipoprotein: evidence for an intrinsic phospholipase A2 activity of apoprotein B-100. Proc. Natl. Acad. Sci. U. S. A. 87, 9741–9745. Peng, Y.S., Ding, H.C., Lin, Y.T., Syu, J.P., Chen, Y., Wang, S.M., 2012. Uremic toxin pcresol induces disassembly of gap junctions of cardiomyocytes. Toxicology 302, 11–17. Pogwizd, S.M., Onufer, J.R., Kramer, J.B., Sobel, B.E., Corr, P.B., 1986. Induction of delayed afterdepolarizations and triggered activity in canine Purkinje fibers by lysophosphoglycerides. Circ. Res. 59, 416–426. Prokazova, N.V., Zvezdina, N.D., Korotaeva, A.A., 1998. Effect of lysophosphatidylcholine on transmembrane signal transduction. Biochemistry (Mosc.) 63, 31– 37. Saez, J.C., Nairn, A.C., Czernik, A.J., Fishman, G.I., Spray, D.C., Hertzberg, E.L., 1997. Phosphorylation of connexin43 and the regulation of neonatal rat cardiac myocyte gap junctions. J. Mol. Cell. Cardiol. 29, 2131– 2145. Sedis, S.P., Sequeira, J.M., Altszuler, H.M., 1990. Coronary sinus lysophosphatidylcholine accumulation during rapid atrial pacing. Am. J. Cardiol. 66, 695– 698. Severs, N.J., Bruce, A.F., Dupont, E., Rothery, S., 2008. Remodelling of gap junctions and connexin expression in diseased myocardium. Cardiovasc. Res. 80, 9– 19. Shah, M.M., Martinez, A.M., Fletcher, W.H., 2002. The connexin43 gap junction protein is phosphorylated by protein kinase A and protein kinase C: in vivo and in vitro studies. Mol. Cell. Biochem. 238, 57–68.

C.-K. Liao et al. / Toxicology 314 (2013) 11–21 Sobel, B.E., Corr, P.B., Robison, A.K., Goldstein, R.A., Witkowski, F.X., Klein, M.S., 1978. Accumulation of lysophosphoglycerides with arrhythmogenic properties in ischemic myocardium. J. Clin. Invest. 62, 546–553. Solan, J.L., Lampe, P.D., 2005. Connexin phosphorylation as a regulatory event linked to gap junction channel assembly. Biochim. Biophys. Acta 1711, 154–163. Solan, J.L., Lampe, P.D., 2009. Connexin43 phosphorylation: structural changes and biological effects. Biochem. J. 419, 261–272. Tomic-Carruthers, N., Gorden, P., 1998. Effects of proteasomal inhibitors on the maturation of the insulin proreceptor: an anatomical paradox. Biochem. Biophys. Res. Commun. 244, 728–731.

21

Ver Donck, L., Verellen, G., Geerts, H., Borgers, M., 1992. Lysophosphatidylcholineinduced Ca(2+)-overload in isolated cardiomyocytes and effect of cytoprotective drugs. J. Mol. Cell. Cardiol. 24, 977–988. Woodley, S.L., Ikenouchi, H., Barry, W.H., 1991. Lysophosphatidylcholine increases cytosolic calcium in ventricular myocytes by direct action on the sarcolemma. J. Mol. Cell. Cardiol. 23, 671–680. Zheng, M., Wang, Y., Kang, L., Shimaoka, T., Marni, F., Ono, K., 2010. Intracellular Ca(2+)- and PKC-dependent upregulation of T-type Ca(2+) channels in LPC-stimulated cardiomyocytes. J. Mol. Cell. Cardiol. 48, 131– 139.