Free Radical Research, September 2014; 48(9): 1085–1094 © 2014 Informa UK, Ltd. ISSN 1071-5762 print/ISSN 1029-2470 online DOI: 10.3109/10715762.2014.932114

ORIGINAL ARTICLE

Subsarcolemmal mitochondrial flashes induced by hypochlorite stimulation in cardiac myocytes W. Zhang1, K. Li1, X. Zhu1, D. Wu1, W. Shang1, X. Yuan1, Z. Huang1, M. Zheng2, X. Wang1, D. Yang3, J. Liu4 & H. Cheng1 1State

Free Radic Res Downloaded from informahealthcare.com by University of Guelph on 02/03/15 For personal use only.

Key Laboratory of Biomembrane and Membrane Biotechnology, Beijing Key Laboratory of Cardiometabolic Molecular Medicine, Institute of Molecular Medicine, Peking-Tsinghua Center for Life Sciences, Peking University, Beijing, China, 2Department of Physiology and Pathophysiology, Health Science Center, Peking University, Beijing, China, 3Department of Chemistry and Department of Surgery, The University of Hong Kong, Hong Kong, China, and 4Department of Pathophysiology, School of Medicine, Shenzhen University, Shenzhen, China Abstract Mitochondrial superoxide flash (mitoflash) reflects quantal and bursting superoxide production and concurrent membrane depolarization triggered by transient mitochondrial permeability transition in many types of cells, at the level of single mitochondria. Here we investigate reactive oxygen species (ROS)-mediated modulation of mitoflash activity in cardiac myocytes and report a surprising finding that hypochlorite ions potently and preferentially triggered mitoflashes in the subsarcolemmal mitochondria (SSM), whereas hydrogen peroxide (H2O2) elicited mitoflash activity uniformly among SSM and interfibrillar mitochondria (IFM). The striking SSM mitoflash response to hypochlorite stimulation remained intact in cardiac myocytes from NOX2-deficient mice, excluding local NOX2-mediated ROS as the major player. Furthermore, it occurred concomitantly with SSM Ca2⫹ accumulation and local Ca2⫹ and CaMKII signaling played an important modulatory role by altering frequency and unitary properties of SSM mitoflashes. These findings underscore the functional heterogeneity of SSM and IFM and the oxidant-specific responsiveness of mitochondria to ROS, and may bear important ramifications in devising therapeutic strategies for the treatment of oxidative stress-related heart diseases. Keywords: mitochondrion, superoxide flash, ROS signaling, calcium signaling, heart

Introduction More than a powerhouse generating energy by oxidative phosphorylation, the mitochondrion in eukaryotic cells plays pivotal roles in calcium signaling, redox homeostasis, and cell fate regulation. In a cardiac myocyte, approximately 6,000 mitochondria occupy 30–40% of cellular volume and generate approximately 90% of the ATP supply. In addition, mitochondria are the major source of reactive oxygen species (ROS)—up to 1–2% of electrons in the electron transfer chain (ETC) are leaked to molecular oxygen to form superoxide anion (.O2⫺), which is quickly converted into hydrogen peroxide (H2O2) through spontaneous or enzymatic dismutation. The production of ROS by mitochondria can lead to oxidative damage that underlies many pathologies including malignant diseases, diabetes mellitus, atherosclerosis, ischemia–reperfusion injury, chronic inflammatory processes, and neurodegenerative diseases [1]. A paradigm-shifting concept in recent years, however, is that mitochondrial ROS also play signaling roles in a variety of pathways in differentiation and organogenesis [2], cell fate regulation [3], and stress response [4]. Recently, we have shown that respiratory mitochondria undergo stochastic, intermittent bursts of superoxide

production, triggered by reversible opening of membrane permeability transition pore (mPTP) and associated with transient depolarization of mitochondrial membrane potential [5]. These discrete events were named “superoxide flashes”, for the free radical production, and “mitochondrial flashes (mitoflashes)”, for the entirety of the multifaceted concurrent signals [6]. The bursting superoxide production in a mitoflash differs from the flashless basal ROS production in terms of ignition mechanism as well as the level, duration, and total amount of ROS signals involved. It depends on intact ETC activity, but can also be regulated independently of respiration [5]. Thus, the mitoflash represents a novel mode of mitochondrial ROS signaling that is intimately interwoven with other core functions of the mitochondria. Massive basal mitochondrial ROS production has been shown under pathological cardiovascular conditions such as ischemia–reperfusion injury. At the same time, the mitochondria are a common target of a variety of oxidative stressors. Acting in a frequency-encoded manner, mitoflashes are a robust and sensitive responder to oxidative stresses including reoxygenation after hypoxia and anoxia [5,7], menadione and selenite stimulation [8,9], and photolytically produced ROS from Killer Red [6]. Under

Correspondence: Jie Liu, M.D., Ph.D, Department of Pathophysiology, School of Medicine, Shenzhen University, Shenzhen 518060, China. Tel: ⫹ 86 755 8667 1912. Fax: ⫹ 86 755 8667 1906. E-mail: [email protected] (Received date: 28 April 2014; Accepted date: 30 May 2014; Published online: 30 June 2014)

Free Radic Res Downloaded from informahealthcare.com by University of Guelph on 02/03/15 For personal use only.

1086 W. Zhang et al. extremely stressful oxidative conditions, mitochondria amplify the damaging oxidative stress by generating bursts of ROS production through the ROS-induced ROS release upon irreversible mPTP opening [10]. Spatially regenerative ROS waves involving the inner membrane anion channel have also been reported in metabolically stressed cardiac myocytes [11,12]. Functional heterogeneity among populations of mitochondria in a single cell has been observed in several cell types including neurons [13] and cardiac myocytes [14–16]. Mitochondria in the outmost cortical layer of a cardiac myocyte, known as subsarcolemmal mitochondria (SSM), differ from those enwrapped by myofibrils, called interfibrillar mitochondria (IFM), in many important ways. Compared with IFM, SSM are not only lower in oxidative rates but also are relatively fragile and display less capacity for Ca2⫹ retention and increased susceptibility to Ca2⫹triggered damage [14,17]. Furthermore, oxidative defects are more pronounced in SSM than IFM isolated from hearts subjected to short periods of ischemia [18]. While these studies were all performed on isolated mitochondria, recent advance of mitoflash visualization would provide an unprecedented means to investigate how subgroups of mitochondria respond to different stresses and insults in intact cardiac myocytes. In the present study, we systematically investigated the effects of hydrochloride ions (ClO⫺) and H2O2, two major endogenous ROS in biological system, on the frequency and spatial distribution as well as unitary properties of mitoflashes in rat and mouse cardiac myocytes, at high spatial and temporal resolution. Our results revealed distinctive mitoflash responses to different ROS and, more importantly, differential modulation of SSM and IFM by ClO⫺. Uncovering heterogeneous SSM and IFM mitoflash responses to different ROS would be helpful in devising oxidant-selective, subpopulational mitochondrion-targeted therapeutic strategies for the treatment of oxidative stressrelated heart diseases.

Methods Culture and adenoviral infection of adult rat ventricular myocytes All animal procedures conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996) and were approved by the Institutional Animal Care and Use Committee of Peking University. Adult male Sprague–Dawley rats weighing 200–250 g were anesthetized by intraperitoneal injection of trichloroacetaldehyde monohydrate (0.5 g/kg). Single ventricular myocytes were isolated by a standard enzymatic technique as described previously [19]. Isolated adult rat ventricular myocytes (ARVMs) were cultured on laminin-coated coverslips in M199 medium (Sigma) and infected with adenovirus carrying the mt-cpYFP-coding sequence at an m.o.i. of 50–100 for 48–72 hrs.

Generation of mt-cpYFP transgenic and mt-cpYFP Cybb⫺/Y mouse models Pan-tissue mt-cpYFP transgenic mice in a C57BL/6 background were generated as previously described [20]. The mt-cpYFP transgenic mice were crossed with NOX2 knockout (Cybb⫺/Y) mice from the Jackson Laboratory to generate mt-cpYFP Cybb⫺/Y mice, with mt-cpYFP Cybb⫹/Y littermates used as controls. Isolation and culture of adult mouse ventricular myocytes Male mice of 8–12 weeks of age were anesthetized by intraperitoneal injection of trichloroacetaldehyde monohydrate (0.5 g/kg) and were given heparin by intraperitoneal injection. Mouse ventricular myocytes were isolated by an enzymatic method reported previously [21]. Isolated cells were plated on laminin-coated coverslips in M199 medium supplemented with 10 mM 2,3-butanedione monoxime and confocal imaging experiments were performed within 6 h of culture. Confocal imaging Fluorescent imaging was carried out on a Zeiss LSM 510 or 710 confocal microscope equipped with a 40⫻, 1.3NA oil immersion objective. To detect mitoflashes by mtcpYFP, images were acquired by excitation alternately at 488 and 405 nm, and collection of emission at greater than 505 nm in mt-cpYFP expressing cells. Visualization of intracellular ClO⫺ was achieved by preloading ARVMs with 20 μM of HKOCl-1 for 30 min followed by imaging with 514 nm excitation and greater than 530 nm emission. To examine mitoflash and mitochondrial calcium signals simultaneously, mt-cpYFP expressing ARVMs were preloaded with 5 μM Rhod-2 AM (Molecular Probes, Eugene, OR, USA) at 37°C for 30 min. Laser excitation at 488, 405, and 543 nm was applied in tandem, and corresponding fluorescence collection was at 505–530, 505–530, and greater than 560 nm, respectively. In a given cell, usually 100 frames of 900 ⫻ 256 (xy) pixels were collected at 1.0 frame/s in a bidirectional scanning mode. The axial resolution was set to 1.0 μm and the size of the imaging field was 105 μm ⫻ 30 μm. All experiments were performed at room temperature (22–26°C). Electrical pacing and ROS stimulation Cells were plated in a custom-designed pacing chamber (Figure 1A) and bathed in Tyrode solution consisting of (in mmol/L) 137 NaCl, 5.4 KCl, 1.2 MgCl2, 1.2 NaH2PO4, 0, 1, or 5 CaCl2, 10 glucose, and 20 HEPES or Cl⫺-free solution consisting of 137 Na aspartate, 5.4 K aspartate, 1.2 MgSO4⭈7H2O, 1 CaSO4, 10 glucose, 20 HEPES (pH 7.35, adjusted with NaOH). ARVMs were electrically paced at 5 Hz (10 ms square wave of 7 V/cm) while perfused with Tyrode solution or Cl⫺-free solution at a rate of 2 ml/min. For mitoflash analysis, cells displaying no

Free Radic Res Downloaded from informahealthcare.com by University of Guelph on 02/03/15 For personal use only.

ROS modulation of mitoflashes in heart

1087

Figure 1. Electrical pacing and ETS triggered mitoflashes predominantly in SSM. A) Imaging mitoflashes during field electrical pacing (5 Hz, left panel) or perfusion with ETS (right panel). As shown in the right panel, the electrolyzed Tyrode solution was produced by pacing Tyrode solution in another plate. ARVMs: adult rat ventricular myocytes; Red stars indicate region where ARVMs were imaged. ⫹ and -: anode and cathode of the electrodes. Arrows show the inlets and outlets of perfused Tyrode solution. B) Subsarcolemmal localization of pacing-elicited mitoflashes. Surface plots of mitoflashes in a 100-s period are overlaid on confocal micrograph of the corresponding ARVM. Note that all mitoflashes arose from SSM except one from an interfibrillar mitochondrion (IFM; arrowhead). C) Averaged traces of SSM and IFM mitoflashes in paced ARVMs (n ⫽ 613 SSM mitoflashes and 1393 IFM mitoflashes). Individual events are aligned by the time of onset. D–F). Frequency (D), amplitude (E) and duration indexed by full FDHM (F) of SSM and IFM mitoflashes. n ⫽ 194, 82, 15, 27 cells or 61, 472, 100, 15 SSM mitoflashes and 388, 1081, 252, 152 IFM mitoflashes for control (CTRL), pacing, ETS, and pacing with Cl⫺-free solution groups, respectively. G) Normalized distance to cell edge averaged over all SSM and IFM mitoflashes. Dashed lines in this and following figures denote the expected value if mitoflashes are uniformly distributed inside the cells. Note that near identical results are obtained with pacing and ETS stimulation and that Cl⫺ is required for pacing-induced SSM mitoflashes. **p ⬍ 0.01 vs. control (CTRL); #p ⬍ 0.05; ##p ⬍ 0.01 SSM vs. IFM.

discernible or minimal shortening were selected to avoid motion artifact. For electrolyzed Tyrode solution (ETS) stimulation, rat or mouse ventricular myocytes were applied with 1 ml of Tyrode solution subjected to a 20 mA D.C. current for 20 s. H2O2 was diluted in Tyrode solution, and ClO⫺ was diluted in a non-reducing physiological buffer HBSS consisting of (in mmol/L) 137 NaCl, 2.7 KCl, 0.5 MgCl2, 8.1 Na2HPO4, 1.47 KH2PO4, 1 CaCl2, and 10 glucose (pH 7.35, adjusted with HCl). Concentrations of H2O2 and ClO⫺, both purchased from Sinopharm Chemical Reagent Beijing Co., Ltd, were calibrated using a UV spectrophotometer right before use.

each mitoflash was calculated as the ratio between the distance of the mitoflash’s centroid to the nearest surface and the half transversal width of the cell at the corresponding location, with value varying between 0 (at the surface) and 1 (at the center of the local width). Mitochondria within 1 μm of the sarcolemma were considered SSM, and all the rest IFM. Data are reported as the mean ⫾ SEM. Student’s t-test was applied to determine the significance of differences. A p value less than 0.05 was considered statistically significant.

Results Data analysis and statistics Mitoflash data were analyzed with custom-developed software, FlashSniper [22], which is implemented in IDL (Interactive Data Language) (ITT, New York, NY). Parameters of individual mitoflashes were calculated as previously described [22]. Mitoflash frequency was reported as number of events in a 100-s period normalized by the imaged cellular area. The normalized distance to edge of

SSM mitoflashes induced by hypochlorite in cardiac myocytes In search for the relationship between cardiac mitoflash and contractile activity, we unexpectedly found that electrical pacing elicited hyperactive mitoflashes in the subset of ARVMs located near the anode (Figure 1A). Strikingly, pacing-induced mitoflashes were predominantly localized

Free Radic Res Downloaded from informahealthcare.com by University of Guelph on 02/03/15 For personal use only.

1088 W. Zhang et al. immediately beneath the sarcolemma, that is, among SSM (Figure 1B). The rate of mitoflash occurrence was increased by 13-fold in SSM, reaching approximately 30/(100 s*1000 μm2) (Figure 1D). Mitoflash activity in IFM was also significantly elevated after pacing, but it rose to only 37% of that in SSM (Figure 1D). In order to quantify the heterogeneous SSM and IFM mitoflash responses, we measured the normalized distance to edge (see Methods) and the mean value was markedly decreased from 0.44 in quiescence to 0.23 after pacing (Figure 1G). It has been shown that electrolysis of physiological saline causes massive production of a variety of ROS, including ClO⫺, ⭈O2⫺, H2O2, ⭈OH, and 1O2 at the anode, through many putative electrochemical reactions [23]. We thus examined whether such electrochemically produced ROS were responsible for pacing-induced SSM mitoflashes. We found that effluents of ETS (See Methods) induced robust heterogeneous SSM and IFM mitoflash activities in quiescent cells (Figures 1A and D). On average, ETS stimulation caused 17- and 7-fold increases of mitoflash activity in SSM and IFM, respectively (Figure 1D), and a 47% decrease of the averaged distance to edge (Figure 1G), mimicking the pacing-induced SSM and IFM mitoflash responses. Because oxidation of the chloride ion (Cl⫺) at the anode is thought to be the first step of a series of chemical reactions that are initiated via electrolyzing a physiological saline [23], we hypothesized that ClO⫺ was the major ROS responsible for pacing and ETS stimulation of SSM and IFM mitoflashes. Indeed, in the Cl⫺-free solutions, pacing failed to augment the mitoflash frequency in either SSM or IFM, and the averaged distance to edge was comparable to that in the absence of electrical pacing (Figures 1D and G). Collectively, these data indicate that electrochemically generated ROS in ETS, especially ClO⫺, potently trigger mitoflashes preferentially in SSM than IFM in ARVMs. Analysis of the unitary properties of evoked SSM and IFM mitoflashes revealed that both electrical pacing and ETS stimulation significantly increased the amplitude and prolonged the duration of both the SSM and IFM mitoflashes, with the effects slightly greater for SSM than IFM mitoflashes (Figures 1C, E and F). However, the magnitude of the changes was moderate as compared to that of mitoflash frequency, consistent with the notion that mitoflash modulation occurs predominantly in the frequencymodulated manner [24]. The above data suggest that ClO⫺ is most likely the key ROS species responsible for pacing- and ETS-stimulated mitoflash responses. Given that ClO⫺ can be produced via endogenous mechanisms in pathological conditions, especially in inflammation, we therefore sought to characterize the possible effects of ClO⫺ on mitoflash production in cardiac myocytes. Figure 2 shows that 50 μM ClO⫺ alone was sufficient to induce mitoflashes of increased amplitude and duration, and the evoked mitoflashes were predominantly confined to SSM. At high (100 μM) ClO⫺, the mitoflash frequency was increased by two-orders of magnitude (Figure 2C), the amplitude of individual events was increased by 3.2-fold (Figure 2D), the full width at half

maximum (FDHM) was markedly prolonged from 8.6 to 21.0 s, and the normalized distance to edge was decreased to 0.11 (Figure 2E), demonstrating the high potency of ClO⫺ stimulation of mitoflashes. Kinetic analysis showed that the SSM mitoflash response displayed a rapid onset with peak response attained within 5 min (Figure 2F), but lasted for only approximately 15 min, due perhaps to mitochondrial damaging effects of extremely high mitoflash activity. Regardless of the magnitude of the frequency response, they were all localized to the cell surface with averaged normalized distance to edge approximately 0.12. Likewise, changes in unitary properties, including increased amplitude (Figure 2D) and FDHM appeared to be dose independent at 50 μM and higher ClO⫺ stimulation. Thus, ClO⫺ stimulation recapitulated essential features of mitoflash response to pacing and ETS depicted in Figure 1. Together with the requirement of Cl⫺ in the mitoflash response to pacing (Figure 1D–G), this result strongly supports the interpretation that ClO⫺ produced at the anode is mainly responsible for mitoflash response to electrical pacing and ETS. H2O2 elicited uniform SSM and IFM mitoflash response H2O2 is one of the most important ROS signaling molecules in many vital biological processes, and excessive H2O2 has been implicated in multiple diseases including heart injury [25]. It is thus important to determine whether and how SSM and IFM respond to H2O2-mediated oxidative stress. We found that, similar to ClO⫺, H2O2 dose dependently elicited stochastic mitoflashes in both SSM and IFM (Figures 3A and B). However, H2O2-mediated mitoflash regulation appeared to be distinct from that by ClO⫺. First, mitoflash responses to H2O2 were similar in both SSM and IFM subpopulations (Figure 3B), and the mean distance to edge was essentially unaltered over the entire range of H2O2 concentration examined (Figure 3D). Second, H2O2 exerted little effect on unitary properties of mitoflashes, evidenced by unchanged SSM and IFM mitoflash amplitudes at the concentrations up to 200 μM. Compared with the potent effect of ClO⫺, H2O2 was only a moderate mitoflash activator, and the maximal effect at 100 μM H2O2 was merely a 4-fold elevation from basal activity (Figure 3B). Moreover, H2O2 effect displayed a very slow onset (latency ∼20 min, at 100–200 μM H2O2) but long endurance (up to 120 min after treatment; Figure 3E). These results indicate that SSM and IFM mitoflashes differentially respond to oxidative stresses in oxidant-sensitive manner. NOX2 was not required for the heterogeneous hypochlorite-stimulated SSM and IFM mitoflash responses Next, we investigated possible mechanisms underlying the distinctive SSM response to ClO⫺. Previous in vitro experiments have shown that cardiac SSM and IFM differ in the activities of respiratory enzymes, coupling status, and Ca2⫹ accumulation. An intrinsic SSM-IFM

Free Radic Res Downloaded from informahealthcare.com by University of Guelph on 02/03/15 For personal use only.

ROS modulation of mitoflashes in heart

1089

Figure 2. SSM mitoflashes triggered by ClO⫺ perfusion. A) Intracellular distribution of mitoflashes in a representative ARVM stimulated by 100 μM ClO⫺. Contours mark sites of individual mitoflashes. B) Averaged traces of SSM (n ⫽ 42) and IFM mitoflashes (n ⫽ 11) obtained with 50–75 μM ClO⫺ perfusion. C–E) Dose-dependence of ClO⫺ modulation of SSM and IFM mitoflash frequency (C), amplitude (D) and normalized distance to edge (E). Data are collected within 50 min after ClO⫺ perfusion. n ⫽ 27, 11, 8, and 6 cells with 4, 24, 18, 109 SSM mitoflashes and 38, 5, 6, and 84 IFM mitoflashes for each dose, respectively. F) Time-dependence of SSM and IFM mitoflash response to 100 μM ClO⫺ stimulation. n ⫽ 3–8 cells for each data point. *p ⬍ 0.05; **p ⬍ 0.01 vs. control; #p ⬍

Subsarcolemmal mitochondrial flashes induced by hypochlorite stimulation in cardiac myocytes.

Mitochondrial superoxide flash (mitoflash) reflects quantal and bursting superoxide production and concurrent membrane depolarization triggered by tra...
2MB Sizes 3 Downloads 3 Views