Tightly regulated Ca2+ homeostasis is a prerequisite for proper cardiac function. To dissect the regulatory network of cardiac Ca2+ handling, we performed a chemical suppressor screen on zebrafish tremblor embryos, which suffer from Ca2+ extrusion defects. Efsevin was identified based on its potent activity to restore coordinated contractions in tremblor. We show that efsevin binds to VDAC2, potentiates mitochondrial Ca2+ uptake and accelerates the transfer of Ca2+ from intracellular stores into mitochondria. In cardiomyocytes, efsevin restricts the temporal and spatial boundaries of Ca2+ sparks and thereby inhibits Ca2+ overload-induced erratic Ca2+ waves and irregular contractions. We further show that overexpression of VDAC2 recapitulates the suppressive effect of efsevin on tremblor embryos whereas VDAC2 deficiency attenuates efsevin's rescue effect and that VDAC2 functions synergistically with MCU to suppress cardiac fibrillation in tremblor. Together, these findings demonstrate a critical modulatory role for VDAC2-dependent mitochondrial Ca2+ uptake in the regulation of cardiac rhythmicity.https://doi.org/10.7554/eLife.04801.001
The heart is a large muscle that pumps blood around the body by maintaining a regular rhythm of contraction and relaxation. If the heart loses this regular rhythm it works less efficiently, which can lead to life-threatening conditions.
Regular heart rhythms are maintained by changes in the concentration of calcium ions in the cytoplasm of the heart muscle cells. These changes are synchronised so that the heart cells contract in a controlled manner. In each cell, a contraction begins when calcium ions from outside the cell enter the cytoplasm by passing through a channel protein in the membrane that surrounds the cell. This triggers the release of even more calcium ions into the cytoplasm from stores within the cell. For the cells to relax, the calcium ions must then be pumped out of the cytoplasm to lower the calcium ion concentration back to the original level.
Shimizu et al. studied a zebrafish mutant—called tremblor—that has irregular heart rhythms because its heart muscle cells are unable to efficiently remove calcium ions from the cytoplasm. Embryos of the tremblor mutant were treated with a wide variety of chemical compounds with the aim of finding some that could correct the heart defect.
A compound called efsevin restores regular heart rhythms in tremblor mutants. Efsevin binds to a pump protein called VDAC2, which is found in compartments called mitochondria within the cell. Although mitochondria are best known for their role in supplying energy for the cell, they also act as internal stores for calcium. By binding to VDAC2, efsevin increases the rate at which calcium ions are pumped from the cytoplasm into the mitochondria. This restores rhythmic calcium ion cycling in the cytoplasm and enables the heart muscle cells to develop regular rhythms of contraction and relaxation. Increasing the levels of VDAC2 or another similar calcium ion pump protein in the heart cells can also restore a regular heart rhythm.
Efsevin can also correct irregular heart rhythms in human and mouse heart muscle cells, therefore the new role for mitochondria in controlling heart rhythms found by Shimizu et al. appears to be shared in other animals. The experiments have also identified the VDAC family of proteins as potential new targets for drug therapies to treat people with irregular heart rhythms.https://doi.org/10.7554/eLife.04801.002
During development, well-orchestrated cellular processes guide cells from diverse lineages to integrate into the primitive heart tube and establish rhythmic and coordinated contractions. While many genes and pathways important for cardiac morphogenesis have been identified, molecular mechanisms governing embryonic cardiac rhythmicity are poorly understood. The findings that Ca2+ waves traveling across the heart soon after the formation of the primitive heart tube (Chi et al., 2008) and that loss of function of key Ca2+ regulatory proteins, such as the L-type Ca2+ channel, Na/K−ATPase and sodium-calcium exchanger 1 (NCX1), severely impairs normal cardiac function (Rottbauer et al., 2001; Shu et al., 2003; Ebert et al., 2005; Langenbacher et al., 2005), indicate an essential role for Ca2+ handling in the regulation of embryonic cardiac function.
Ca2+ homoeostasis in cardiac muscle cells is tightly regulated at the temporal and spatial level by a subcellular network involving multiple proteins, pathways, and organelles. The release and reuptake of Ca2+ by the sarcoplasmic reticulum (SR), the largest Ca2+ store in cardiomyocytes, constitutes the primary mechanism governing the contraction and relaxation of the heart. Ca2+ influx after activation of the L-type Ca2+ channel in the plasma membrane induces the release of Ca2+ from the SR via ryanodine receptor (RyR) channels, which leads to an increase of the intracellular Ca2+ concentration and cardiac contraction. During diastolic relaxation, Ca2+ is transferred back into the SR by the SR Ca2+ pump or extruded from the cell through NCX1. Defects in cardiac Ca2+ handling and Ca2+ overload, for example during cardiac ischemia/reperfusion or in long QT syndrome, are well known causes of contractile dysfunction and many types of arrhythmias including early and delayed afterdepolarizations and Torsade des pointes (Bers, 2002; Choi et al., 2002; Yano et al., 2008; Greiser et al., 2011).
Ca2+ crosstalk between mitochondria and ER/SR has been noted in many cell types and the voltage-dependent anion channel (VDAC) and the mitochondrial Ca2+ uniporter (MCU) serve as primary routes for Ca2+ entry through the outer and inner mitochondrial membranes, respectively (Rapizzi et al., 2002; Bathori et al., 2006; Shoshan-Barmatz et al., 2010; Baughman et al., 2011; De Stefani et al., 2011). In the heart, mitochondria are tethered to the SR and are located in close proximity to Ca2+ release sites (García-Pérez et al., 2008; Boncompagni et al., 2009; Hayashi et al., 2009). This subcellular architecture exposes the mitochondria near the Ca2+ release sites to a high local Ca2+ concentration that is sufficient to overcome the low Ca2+ affinity of MCU and facilitates Ca2+ crosstalk between SR and mitochondria (García-Pérez et al., 2008; Dorn and Scorrano, 2010; Kohlhaas and Maack, 2013). Increase of the mitochondrial Ca2+ concentration enhances energy production during higher workload and dysregulation of SR-mitochondrial Ca2+ signaling results in energetic deficits and oxidative stress in the heart and may trigger programmed cell death (Brandes and Bers, 1997; Maack et al., 2006; Kohlhaas and Maack, 2013). However, whether SR-mitochondrial Ca2+ crosstalk also contributes significantly to cardiac Ca2+ signaling during excitation-contraction coupling requires further investigation.
In zebrafish, the tremblor (tre) locus encodes a cardiac-specific isoform of the Na+/Ca2+ exchanger 1, NCX1h (also known as slc8a1a) (Ebert et al., 2005; Langenbacher et al., 2005). The tre mutant hearts lack rhythmic Ca2+ transients and display chaotic Ca2+ signals in the myocardium leading to unsynchronized contractions resembling cardiac fibrillation (Langenbacher et al., 2005). In this study, we used tre as an animal model for aberrant Ca2+ handling-induced cardiac dysfunction and took a chemical genetic approach to dissect the Ca2+ regulatory network important for maintaining cardiac rhythmicity. A synthetic compound named efsevin was identified from a suppressor screen due to its potent ability to restore coordinated contractions in tre. Using biochemical and genetic approaches we show that efsevin interacts with VDAC2 and potentiates its mitochondrial Ca2+ transporting activity and spatially and temporally modulates cytosolic Ca2+ signals in cardiomyocytes. The important role of mitochondrial Ca2+ uptake in regulating cardiac rhythmicity is further supported by the suppressive effect of VDAC2 and MCU overexpression on cardiac fibrillation in tre.
Homozygous tre mutant embryos suffer from Ca2+ extrusion defects and manifest chaotic cardiac contractions resembling fibrillation (Ebert et al., 2005; Langenbacher et al., 2005). To dissect the regulatory network of Ca2+ handling in cardiomyocytes and to identify mechanisms controlling embryonic cardiac rhythmicity, we screened the BioMol library and a collection of synthetic compounds for chemicals that are capable of restoring heartbeat either completely or partially in tre embryos. A dihydropyrrole carboxylic ester compound named efsevin was identified based on its ability to restore persistent and rhythmic cardiac contractions in tre mutant embryos in a dose-dependent manner (Figure 1A,E, and Videos 1–4). To validate the effect of efsevin, we assessed cardiac performance of wild type, tre and efsevin-treated tre embryos (Nguyen et al., 2009). Fractional shortening of efsevin treated tre mutant hearts was comparable to that of their wild type siblings and heart rate was restored to approximately 40% of that observed in controls (Figure 1B–D). Periodic local field potentials accompanying each heartbeat were detected in wild type and efsevin-treated tre embryos using a microelectrode array (Figure 1F–H). Furthermore, while only sporadic Ca2+ signals were detected in tre hearts, in vivo Ca2+ imaging revealed steady Ca2+ waves propagating through efsevin-treated tre hearts (Figure 1I, Videos 5–7), demonstrating that cardiomyocytes are functionally coupled and that efsevin treatment restores regular Ca2+ transients in tre hearts.
We next examined whether efsevin could suppress aberrant Ca2+ homeostasis-induced arrhythmic responses in mammalian cardiomyocytes. Mouse embryonic stem cell-derived cardiomyocytes (mESC-CMs) establish a regular contraction pattern with rhythmic Ca2+ transients (Figure 2A,B,E,F). Mimicking Ca2+ overload by increasing extracellular Ca2+ levels was sufficient to disrupt normal Ca2+ cycling and induce irregular contractions in mESC-CMs (Figure 2C,E,F). Remarkably, efsevin treatment restored rhythmic Ca2+ transients and cardiac contractions in these cells (Figure 2D–F). Similar effect was observed in human embryonic stem cell-derived cardiomyocytes (hESC-CMs) (Figure 2G). Together, these findings suggest that efsevin targets a conserved Ca2+ regulatory mechanism critical for maintaining rhythmic cardiac contraction in fish, mice and humans.
To identify the protein target of efsevin, we generated a N-Boc-protected 2-aminoethoxyethoxyethylamine linker-attached efsevin (efsevinL) (Figure 3A,C). This modified compound retained the activity of efsevin to restore cardiac contractions in ncx1h deficient embryos (Figure 3B,D) and was used to create efsevin-conjugated agarose beads (efsevinLB). A 32kD protein species was detected from zebrafish lysate due to its binding ability to efsevinLB and OK-C125LB, an active efsevin derivative conjugated to beads, but not to beads capped with ethanolamine alone or beads conjugated with an inactive efsevin analog (OK-C19LB) (Figure 3A–E). Furthermore, preincubation of zebrafish lysate with excess efsevin prevented the 32kD protein from binding to efsevinLB or OK-C125LB (Figure 3E). Mass spectrometry analysis revealed that this 32kD band represents a zebrafish homologue of the mitochondrial voltage-dependent anion channel 2 (VDAC2) (Figure 3F and Figure 3—figure supplement 1).
VDAC2 is expressed in the developing zebrafish heart (Figure 4A), making it a good candidate for mediating efsevin’s effect on cardiac Ca2+ handling. To examine this possibility, we injected in vitro synthesized VDAC2 RNA into tre embryos and found that the majority of these embryos had coordinated cardiac contractions similar to those subjected to efsevin treatment (Figure 4B, Videos 8–11). In addition, we generated myl7:VDAC2 transgenic fish in which VDAC2 expression can be induced in the heart by tebufenozide (TBF) (Figure 4C). Knocking down NCX1h in myl7:VDAC2 embryos results in chaotic cardiac movement similar to tre. Like efsevin treatment, induction of VDAC2 expression by TBF treatment restored coordinated and rhythmic contractions in myl7:VDAC2;NCX1h MO hearts (Figure 4D, Videos 12,13). Conversely, knocking down VDAC2 in tre hearts attenuated the suppressive effect of efsevin (Figure 4E, Videos 14–16). Furthermore, we generated VDAC2 null embryos by the Zinc Finger Nuclease gene targeting approach (Figure 4G). Similar to that observed in morpholino knockdown embryos, homozygous VDAC2LA2256 embryos do not exhibit noticeable morphological defects, but the suppressive effect of efsevin was attenuated in homozygous VDAC2LA2256; NCX1MO embryos (Figure 4F). These findings demonstrate that VDAC2 is a major mediator for efsevin’s effect on ncx1h deficient hearts.
VDAC is an abundant channel located on the outer mitochondrial membrane serving as a primary passageway for metabolites and ions (Figure 5A) (Rapizzi et al., 2002; Bathori et al., 2006; Shoshan-Barmatz et al., 2010). At its close state, VDAC favours Ca2+ flux (Tan and Colombini, 2007). To examine whether efsevin would modulate mitochondrial Ca2+ uptake via VDAC2, we transfected HeLa cells with VDAC2. We noted increased mitochondrial Ca2+ uptake in permeabilized VDAC2 transfected and efsevin-treated cells after the addition of Ca2+ and the combined treatment further enhanced mitochondrial Ca2+ levels (Figure 5B).
Mitochondria are located in close proximity to Ca2+ release sites of the ER/SR and an extensive crosstalk between the two organelles exists (García-Pérez et al., 2008; Hayashi et al., 2009; Brown and O'Rourke, 2010; Dorn and Scorrano, 2010; Kohlhaas and Maack, 2013). We examined whether Ca2+ released from intracellular stores could be locally transported into mitochondria through VDAC2 in VDAC1/VDAC3 double knockout (V1/V3DKO) MEFs where VDAC2 is the only VDAC isoform being expressed (Roy et al., 2009a). While treatments with ATP, an IP3-linked agonist, and thapsigargin, a SERCA inhibitor, stimulated similar global cytoplasmic [Ca2+] elevation in intact cells, only ATP induced a rapid mitochondrial matrix [Ca2+] rise (Figure 5—figure supplement 1). This finding is consistent with observations obtained in other cell types (Rizzuto et al., 1994; Hajnóczky et al., 1995) and suggests that Ca2+ was locally transferred from IP3 receptors to mitochondria through VDAC2 at the close ER-mitochondrial associations. We next investigated whether this process could be modulated by efsevin. In permeabilized V1/V3DKO MEFs, treatment with efsevin increased the amount of Ca2+ transferred into mitochondria during IP3-induced Ca2+ release (Figure 5C). Also, in intact V1/V3 DKO MEFs, efsevin accelerated the transfer of Ca2+ released from intracellular stores into mitochondria during stimulation with ATP (Figure 5D,E).
We next examined the effect of efsevin on cytosolic Ca2+ signals in isolated adult murine cardiomyocytes. We found that efsevin treatment induced faster inactivation kinetics without affecting the amplitude or time to peak of paced Ca2+ transients (Figure 6A). Similarly, efsevin treatment did not significantly alter the frequency, amplitude or Ca2+ release flux of spontaneous Ca2+ sparks, local Ca2+ release events, but accelerated the decay phase resulting in sparks with a shorter duration and a narrower width (Figure 6B). These results indicate that by activating mitochondrial Ca2+ uptake, efsevin accelerates Ca2+ removal from the cytosol in cardiomyocytes and thereby restricts local cytosolic Ca2+ sparks to a narrower domain for a shorter period of time without affecting SR Ca2+ load or RyR Ca2+ release. Under conditions of Ca2+ overload, single Ca2+ sparks can trigger opening of neighbouring Ca2+ release units and thus induce the formation of erratic Ca2+ waves (Figure 6C). Efsevin treatment significantly reduced the number of propagating Ca2+ waves in a dosage-dependent manner (Figure 6C,D), demonstrating a potent suppressive effect of efsevin on the propagation of Ca2+ overload-induced Ca2+ waves and suggesting that efsevin could serve as a pharmacological tool to manipulate local Ca2+ signals.
We hypothesize that efsevin treatment/VDAC2 overexpression suppresses aberrant Ca2+ handling-associated arrhythmic cardiac contractions by buffering excess Ca2+ into mitochondria. This hypothesis predicts that activating other mitochondrial Ca2+ uptake molecules would likewise restore coordinated contractions in tre. To test this model, we cloned zebrafish MCU and MICU1, an inner mitochondrial membrane Ca2+ transporter and its regulator (Perocchi et al., 2010; Baughman et al., 2011; De Stefani et al., 2011; Mallilankaraman et al., 2012; Csordas et al., 2013). In situ hybridization showed that MCU and MICU1 were expressed in the developing zebrafish heart (Figure 7A) and their expression levels were comparable between the wild type and tre hearts (Figure 7—figure supplement 1). Overexpression of MCU restored coordinated contractions in tre, akin to what was observed with VDAC2 (Figure 7B). In addition, tre embryos injected with suboptimal concentrations of MCU or VDAC2 had a fibrillating heart, but embryos receiving both VDAC2 and MCU at the suboptimal concentration manifested coordinated contractions (Figure 7C), demonstrating a synergistic effect of these proteins. Furthermore, overexpression of MCU failed to suppress the tre phenotype in the absence of VDAC2 activity and VDAC2 could not restore coordinated contractions in tre without functional MCU (Figure 7B,D). Similar results were observed by manipulating MICU1 activity (Figure 7E,F). Together, these findings indicate that mitochondrial Ca2+ uptake mechanisms on outer and inner mitochondrial membranes act cooperatively to regulate cardiac rhythmicity.
In summary, we conducted a chemical suppressor screen in zebrafish to dissect the regulatory network critical for maintaining rhythmic cardiac contractions and to identify mechanisms underlying aberrant Ca2+ handling-induced cardiac dysfunction. We show that activation of VDAC2 through overexpression or efsevin treatment potently restores rhythmic contractions in NCX1h deficient zebrafish hearts and effectively suppresses Ca2+ overload-induced arrhythmogenic Ca2+ events and irregular contractions in mouse and human cardiomyocytes. We provide evidence that potentiating VDAC2 activity enhances mitochondrial Ca2+ uptake, accelerates Ca2+ transfer from intracellular stores into mitochondria and spatially and temporally restricts single Ca2+ sparks in cardiomyocytes. The crucial role of mitochondria in the regulation of cardiac rhythmicity is further supported by the findings that VDAC2 functions in concert with MCU; these genes have a strong synergistic effect on suppressing cardiac fibrillation and loss of function of either gene abrogates the rescue effect of the other in tre.
The regulatory roles of mitochondrial Ca2+ in cardiac metabolism, cell survival and fate have been studied extensively (Brown and O'Rourke, 2010; Dorn and Scorrano, 2010; Doenst et al., 2013; Kasahara et al., 2013; Kohlhaas and Maack, 2013; Luo and Anderson, 2013). Our study provides genetic and physiologic evidence supporting an additional role for mitochondria in regulating cardiac rhythmicity and reveals VDAC2 as a modulator of Ca2+ handling in cardiomyocytes. Our findings, together with recent reports of the physical interaction between VDAC2 and RyR2 (Min et al., 2012) and the close proximity of outer and inner mitochondrial membranes at the contact sites between the mitochondria and the SR (García-Pérez et al., 2011), suggest an intriguing model. We propose that mitochondria facilitate an efficient clearance mechanism in the Ca2+ microdomain, which modulates Ca2+ handling without affecting global Ca2+ signals in cardiomyocytes. In this model, VDAC facilitates mitochondrial Ca2+ uptake via MCU complex and thereby controls the duration and the diffusion of cytosolic Ca2+ near the Ca2+ release sites to ensure rhythmic cardiac contractions. This model is consistent with our observation that efsevin treatment induces faster inactivation kinetics of cytosolic Ca2+ transients without affecting the amplitude or the time to peak in cardiomyocytes and the reports that blocking mitochondrial Ca2+ uptake has little impact on cytosolic Ca2+ transients (Maack et al., 2006; Kohlhaas et al., 2010). Further support for this model comes from the observation of the Ca2+ peaks on the OMM (Drago et al., 2012) and the finding that downregulating VDAC2 extends Ca2+ sparks (Subedi et al., 2011; Min et al., 2012) and that blocking mitochondrial Ca2+ uptake by Ru360 leads to an increased number of spontaneous propagating Ca2+ waves (Seguchi et al., 2005). Future studies on the kinetics of VDAC2-dependent mitochondrial Ca2+ uptake and exploring potential regulatory molecules for VDAC2 activity will provide insights into how the crosstalk between SR and mitochondria contributes to Ca2+ handling and cardiac rhythmicity.
Aberrant Ca2+ handling is associated with many cardiac dysfunctions including arrhythmia. Establishing animal models to study molecular mechanisms and develop new therapeutic strategies are therefore major preclinical needs. Our chemical suppressor screen identified a potent effect of efsevin and its biological target VDAC2 on manipulating cardiac Ca2+ handling and restoring regular cardiac contractions in fish and mouse and human cardiomyocytes. This success indicates that fundamental mechanisms regulating cardiac function are conserved among vertebrates despite the existence of species-specific features and suggests a new paradigm of using zebrafish cardiac disease models for the dissection of critical genetic pathways and the discovery of new therapeutic approaches. Future studies examining the effects of efsevin on other arrhythmia models would further elucidate the potential for efsevin as a pharmacological tool to treat cardiac arrhythmia associated with aberrant Ca2+ handling.
Zebrafish of the mutant line tremblor (tretc318) were maintained and bred as described previously (Langenbacher et al., 2005). Transgenic lines, myl7:gCaMP4.1LA2124 and myl7:VDAC2LA2309 were created using the Tol2kit (Esengil et al., 2007; Kwan et al., 2007; Shindo et al., 2010). The VDAC2LA2256 was created using the zinc finger array OZ523 and OZ524 generated by the zebrafish Zinc Finger Consortium (Foley et al., 2009a, 2009b).
Full length VDAC2 cDNA was purchased from Open Biosystems (Huntsville, AL) and cloned into pCS2+ or pCS2+3XFLAG. Full length cDNA fragments of zebrafish MCU (Accession number: JX424822) and MICU1 (JX42823) were amplified from 2 dpf embryos and cloned into pCS2+. For mRNA synthesis, plasmids were linearized and mRNA was synthesized using the SP6 mMESSAGE mMachine kit according to the manufacturers manual (Ambion, Austin, TX.).
VDAC2 mRNA and morpholino antisense oligos (5′-GGGAACGGCCATTTTATCTGTTAAA-3′) (Genetools, Philomath, OR) were injected into one-cell stage embryos collected from crosses of tretc318 heterozygotes. Cardiac performance was analyzed by visual inspection on 1 dpf. The tre mutant embryos were identified either by observing the fibrillation phenotype at 2–3 dpf or by genotyping as previously described (Langenbacher et al., 2005).
Chemicals from a synthetic library (Castellano et al., 2007; Choi et al., 2011; Cruz et al., 2011) and from Biomol International LP (Farmingdale, NY) were screened for their ability to partially or completely restore persistent heartbeat in tre embryos. 12 embryos collected from crosses of tretc318 heterozygotes were raised in the presence of individual compounds at a concentration of 10 µM from 4 hpf (Choi et al., 2011). Cardiac function was analyzed by visual inspection at 1 and 2 dpf. The hearts of tretc318 embryos manifest a chaotic movement resembling cardiac fibrillation with intermittent contractions in rare occasion (Ebert et al., 2005; Langenbacher et al., 2005). Compounds that elicit persistent coordinated cardiac contractions were validated on large number of tre mutant embryos and NCX1h morphants (>500 embryos).
Videos of GFP-labelled myl7:GFP hearts were taken at 30 frames per second. Line-scan analysis was performed along a line through the atria or the ventricles of these hearts (Nguyen et al., 2009). Fraction of shortening was deduced from the ratio of diastolic and systolic width and heart rate was determined by beats per minute. Cardiac parameters were analyzed in tremblortc318 and VDAC2LA2256 at 2 dpf.
36 hpf myl7:gCaMP4.1 embryos were imaged at a frame rate of 30 ms/frame. Electromechanical isolation was achieved by tnnt2MO (Milan et al., 2006). The fluorescence intensity of each pixel in a 2D map was normalize to generate heat maps and isochronal lines at 33 ms intervals were obtained by identifying the maximal spatial gradient for a given time point (Chi et al., 2008).
The mouse E14Tg2a ESC and human H9 ESC line were cultured and differentiated as previously described (Blin et al., 2010; Arshi et al., 2013). At day 10 of differentiation, beating mouse EBs were exposed to external solution containing 10 mM CaCl2 for 10 min before DMSO or efsevin (10 μM) treatment. Human EBs were differentiated for 15 days and treated with 5 mM CaCl2 for 10 min before DMSO or efsevin (5 μM) treatment. Images of beating EBs were acquired at a rate of 30 frames/s and analyzed by motion-detection software. For calcium recording, the EBs were loaded with 10 μM fluo-4 AM in culture media for 30 min at 37°C. Line-scan analysis was performed and fluorescent signals were acquired by a Zeiss LSM510 confocal microscope.
2-day-old wild type, tre, and efsevin-treated tre embryos were placed on uncoated, microelectrode arrays (MEAs) containing 120 integrated TiN electrodes (30 μm diameter, 200 μm interelectrode spacing). Local field potentials (LFPs) at each electrode were collected for three trials per embryo type over a period of three minutes at a sampling rate of 1 kHz using the MEA2100-HS120 system (Multichannel Systems, Reutiligen, Germany). Raw data was low-pass filtered at a cutoff frequency of 10 Hz using a third-order Butterworth filter. Data analysis was carried out using the MC_DataTool (Multichannel Systems) and Matlab (MathWorks).
Murine ventricular cardiomyocytes were isolated as previously described (Reuter et al., 2004). Cells were loaded with 5 µM fluo-4 AM in external solution containing: 138.2 mM NaCl, 4.6 mM KCl, 1.2 mM MgCl, 15 mM glucose, 20 mM HEPES for 1 hr and imaged in external solution supplemented with 2, 5 or 10 mM CaCl2. For the recording of Ca2+ sparks and transients, the external solution contained 2 mM CaCl2. For Ca2+ transients, cells were field stimulated at 0.5 Hz with a 5 ms pulse at a voltage of 20% above contraction threshold. For all measurements, efsevin was added 2 hr prior to the actual experiment. Images were recorded on a Zeiss LSM 5 Pascal confocal microscope. Data analysis was carried out using the Zeiss LSM Image Browser and ImageJ with the SparkMaster plugin (Picht et al., 2007). Cells were visually inspected prior to and after each recording. Only those recordings from healthy looking cells with distinct borders, uniform striations and no membrane blebs or granularity were included in the analysis.
For pull down assays mono-N-Boc protected 2,2'-(ethylenedioxy)bis(ethylamine) was attached to the carboxylic ester of efsevin and its derivatives through the amide bond. After removal of the Boc group using TFA, the primary amine was coupled to the carboxylic acid of Affi-Gel 10 Gel (Biorad, Hercules, CA). 2-day-old zebrafish embryos were deyolked by centrifugation before being lysed with Rubinfeld's lysis buffer (Rubinfeld et al., 1993). The lysate was precleaned by incubation with Affi-Gel 10 Gel to eliminate non-specific binding. Precleaned lysate was incubated with affinity beads overnight. Proteins were eluted from the affinity beads and separated on SDS-PAGE. Protein bands of interest were excised. Gel plugs were dehydrated in acetonitrile (ACN) and dried completely in a Speedvac. Samples were reduced and alkylated with 10 mM dithiotreitol and 10 mM TCEP solution in 50 mM NH4HCO3 (30 min at 56°C) and 100 mM iodoacetamide (45 min in dark), respectively. Gel plugs were washed with 50 mM NH4HCO3, dehydrated with ACN, and dried down in a Speedvac. Gel pieces were then swollen in digestion buffer containing 50 mM NH4HCO3, and 20.0 ng/μl of chymotrypsin (25°C, overnight). Peptides were extracted with 0.1% TFA in 50% ACN solution, dried down and resuspended in LC buffer A (0.1% formic acid, 2% ACN).
Extracted peptides were analyzed by nano-flow LC/MS/MS on a Thermo Orbitrap with dedicated Eksigent nanopump using a reversed phase column (New Objective, Woburn, MA). The flow rate was 200 nl/min for separation: mobile phase A contained 0.1% formic acid, 2% ACN in water, and mobile phase B contained 0.1% formic acid, 20% water in ACN. The gradient used for analyses was linear from 5% B to 50% B over 60 min, then to 95% B over 15 min, and finally keeping constant 95% B for 10 min. Spectra were acquired in data-dependent mode with dynamic exclusion where the instrument selects the top six most abundant ions in the parent spectra for fragmentation. Data were searched against the Danio rerio IPI database v3.45 using the SEQUEST algorithm in the BioWorks software program version 3.3.1 SP1. All spectra used for identification had deltaCN>0.1 and met the following Xcorr criteria: >2 (+1), >3 (+2), >4 (+3), and >5 (+4). Searches required full cleavage with the enzyme, <4 missed cleavages and were performed with the differential modifications of carbamidomethylation on cysteine and methionine oxidation.
In situ hybridization was performed as previously described (Chen and Fishman, 1996). DIG-labeled RNA probe was synthesized using the DIG RNA labeling kit (Roche, Indianapolis, IN).
HeLa cells were transfected with a C-terminally flag-tagged zebrafish VDAC1 or VDAC2 in plasmid pCS2+ using Lipofectamine 2000 (Invitrogen). After staining with MitoTracker Orange (Invitrogen) cells were fixed in 3.7% formaldehyde and permeabilized with acetone. Immunostaining was performed using primary antibody ANTI-FLAG M2 (Sigma Aldrich, St. Luis, MO) at 1:100 and secondary antibody Anti-Mouse IgG1-FITC (Southern Biotechnology Associates, Birmingham, AL) at 1:200. Cells were mounted and counterstained using Vectashield Hard Set with DAPI (Vector Laboratories, UK).
HeLa cells were transfected with zebrafish VDAC2 using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). 36 hrs after transfection, cells were loaded with 5 µM Rhod2-AM (Invitrogen), a Ca2+ indicator preferentially localized in mitochondria, for 1 hr at 15°C followed by a 30 min de-esterification period at 37°C. Subsequently, cells were permeabilized with 100 µM digitonin for 1 min at room temperature. Fluorescence changes in Rhod2 (ex: 544 nm, em: 590 nm) immediately after the addition of Ca2+ (final free Ca2+ concentration is calculated to be approximately 10 µM using WEBMAXC at http://web.stanford.edu/∼cpatton/webmaxcS.htm) were monitored in internal buffer (5 mM K-EGTA, 20 mM HEPES, 100 mM K-aspartate, 40 mM KCl, 1 mM MgCl2, 2 mM maleic acid, 2 mM glutamic acid, 5 mM pyruvic acid, 0.5 mM KH2PO4, 5 mM MgATP, pH adjusted to 7.2 with Trizma base) using a FLUOSTAR plate reader (BMG Labtech, Germany).
V1/V3 DKO MEFs were cultured as previously described (Roy et al., 2009a). Efsevin-treated (15 μM for 30 min) or mock-treated MEFs were used for measurements of [Ca2+]c in suspensions of permeabilized cells or imaging of [Ca2+]m simultaneously with [Ca2+]c in intact single cells. Permeabilization of the plasma membrane was performed by digitonin (40 μM/ml). Changes in [Ca2+] in the cytoplasmic buffer upon IP3 (7.5 μM) addition in the presence or absence of ruthenium red (3 μM) was measured by fura2 in a fluorometer (Csordás et al., 2006; Roy et al., 2009b). To avoid endoplasmic reticulum Ca2+ uptake 2 μM thapsigargin was added before IP3. For imaging of [Ca2+]m and [Ca2+]c, MEFs were co-transfected with plasmids encoding polycistronic zebrafish VDAC2 with mCherry and mitochondria-targeted inverse pericam for 40 hr. Cells were sorted to enrich the transfected cells and attached to glass coverslips. In the final 10 min, of the efsevin or mock-treatment, the cells were also loaded with fura2AM (2.5 μM) and subsequently transferred to the microscope stage. Stimulation with 1 μM ATP was carried out in a norminally Ca2+ free buffer. Changes in [Ca2+]c and [Ca2+]m were imaged using fura2 (ratio of ex:340 nm–380 nm) and mitochondria-targeted inverse pericam (ex: 495 nm), respectively (Csordas et al., 2010).
All values are expressed as mean ± SEM, unless otherwise specified. Significance values are calculated by unpaired student's t-test unless noted otherwise.
Rigid microenvironments promote cardiac differentiation of mouse and human embryonic stem cellsScience and Technology of Advanced Materials, 14, pii: 025003, 10.1088/1468-6996/14/2/025003.
Ca2+-dependent control of the permeability properties of the mitochondrial outer membrane and voltage-dependent anion-selective channel (VDAC)The Journal of Biological Chemistry 281:17347–17358.https://doi.org/10.1074/jbc.M600906200
A purified population of multipotent cardiovascular progenitors derived from primate pluripotent stem cells engrafts in postmyocardial infarcted nonhuman primatesThe Journal of Clinical Investigation 120:1125–1139.https://doi.org/10.1172/JCI40120
Mitochondria are linked to calcium stores in striated muscle by developmentally regulated tethering structuresMolecular Biology of the Cell 20:1058–1067.https://doi.org/10.1091/mbc.E08-07-0783
Small-molecule inhibitors of protein geranylgeranyltransferase type IJournal of the American Chemical Society 129:5843–5845.https://doi.org/10.1021/ja070274n
Zebrafish tinman homolog demarcates the heart field and initiates myocardial differentiationDevelopment 122:3809–3816.
Diversity through phosphine catalysis identifies octahydro-1,6-naphthyridin-4-ones as activators of endothelium-driven immunityProceedings of the National Academy of Sciences of USA 108:6769–6774.https://doi.org/10.1073/pnas.1015254108
Structural and functional features and significance of the physical linkage between ER and mitochondriaThe Journal of Cell Biology 174:915–921.https://doi.org/10.1083/jcb.200604016
Cardiac metabolism in heart failure: implications beyond ATP productionCirculation Research 113:709–724.https://doi.org/10.1161/CIRCRESAHA.113.300376
Mitochondrial Ca2+ uptake contributes to buffering cytoplasmic Ca2+ peaks in cardiomyocytesProceedings of the National Academy of Sciences of USA 109:12986–12991.https://doi.org/10.1073/pnas.1210718109
Calcium extrusion is critical for cardiac morphogenesis and rhythm in embryonic zebrafish heartsProceedings of the National Academy of Sciences of USA 102:17705–17710.https://doi.org/10.1073/pnas.0502683102
Small-molecule regulation of zebrafish gene expressionNature Chemical Biology 3:154–155.https://doi.org/10.1038/nchembio858
Targeted mutagenesis in zebrafish using customized zinc-finger nucleasesNature Protocols 4:1855–1867.https://doi.org/10.1038/nprot.2009.209
Physical coupling supports the local Ca2+ transfer between sarcoplasmic reticulum subdomains and the mitochondria in heart muscleThe Journal of Biological Chemistry 283:32771–32780.https://doi.org/10.1074/jbc.M803385200
Alignment of sarcoplasmic reticulum-mitochondrial junctions with mitochondrial contact pointsAmerican Journal of Physiology Heart and Circulatory Physiology 301:H1907–H1915.https://doi.org/10.1152/ajpheart.00397.2011
Alterations of atrial Ca(2+) handling as cause and consequence of atrial fibrillationCardiovascular Research 89:722–733.https://doi.org/10.1093/cvr/cvq389
Three-dimensional electron microscopy reveals new details of membrane systems for Ca2+ signaling in the heartJournal of Cell Science 122:1005–1013.https://doi.org/10.1242/jcs.028175
The Tol2kit: a multisite gateway-based construction kit for Tol2 transposon transgenesis constructsDevelopmental Dynamics 236:3088–3099.https://doi.org/10.1002/dvdy.21343
Mutation in sodium-calcium exchanger 1 (NCX1) causes cardiac fibrillation in zebrafishProceedings of the National Academy of Sciences of USA 102:17699–17704.https://doi.org/10.1073/pnas.0502679102
Mechanisms of altered Ca2+ handling in heart failureCirculation Research 113:690–708.https://doi.org/10.1161/CIRCRESAHA.113.301651
Zebrafish as a model for cardiovascular development and diseaseDrug Discovery Today Disease Models 5:135–140.https://doi.org/10.1016/j.ddmod.2009.02.003
SparkMaster: automated calcium spark analysis with ImageJAmerican Journal of Physiology Cell Physiology 293:C1073–C1081.https://doi.org/10.1152/ajpcell.00586.2006
Recombinant expression of the voltage-dependent anion channel enhances the transfer of Ca2+ microdomains to mitochondriaThe Journal of Cell Biology 159:613–624.https://doi.org/10.1083/jcb.200205091
Mice overexpressing the cardiac sodium-calcium exchanger: defects in excitation-contraction couplingThe Journal of Physiology 554:779–789.https://doi.org/10.1113/jphysiol.2003.055046
VDAC, a multi-functional mitochondrial protein regulating cell life and deathMolecular Aspects of Medicine 31:227–285.https://doi.org/10.1016/j.mam.2010.03.002
Defective Ca2+ cycling as a key pathogenic mechanism of heart failureCirculation Journal 72:A22–A30.https://doi.org/10.1253/circj.CJ-08-0070
Jodi NunnariReviewing Editor; University of California, Davis, United States
eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.
Thank you for sending your work entitled “Activation of VDAC2 regulates mitochondrial Ca2+ uptake and cardiac rhythmicity” for consideration at eLife. Your article has been favorably evaluated by Vivek Malhotra (Senior editor), a Reviewing editor, and 2 reviewers.
The following individuals responsible for the peer review of your submission have agreed to reveal their identity: Jodi Nunnari (Reviewing editor); Rosario Rizzuto (peer reviewer 1). Reviewer 2 remains anonymous.
The Reviewing editor and the reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission. All agree that the work is of high quality and represents a significant advance in our understanding of the role of VDAC in calcium handling and cardiac physiology. A majority of the comments pasted below can be addressed by a careful revision of the manuscript. However, as pointed out by reviewer 2, the experiments do not adequately test the authors' conclusion that Ca2+ is directly transferred between the ER and mitochondria. Additional experimentation, for example examining the effect of addition of a Ca2+ chelator.
The authors point to mitochondrial Ca2+ buffering as a protective mechanism against erratic electrical activity, and indicates a prominent role of VDAC2 as a “facilitator” of Ca2+ transfer through MCU channels. In this respect, the authors could consider investigating whether additional, possibly isoform-specific mechanisms add to the role of VDAC2 in Ca2+ permeation across the outer mitochondrial membrane.
1) Does VDAC 2 interact with MCU and/or with some of its regulators (MICU1/2/3, EMRE or MCUR1)?
2) Is the expression of MCU and its regulators (“MCU complex”) altered in the tre mutants?
3) Does pharmacological or molecular stimulation of VDAC2 affect the expression of the MCU complex?
The manuscript reports exciting results but needs to be improved in the following ways:
1) The results in Figure 3F show a sample of the identification of one VDAC peptide. The mass numbers are not visible and the color code of red and blue is not clear at all especially the blue. The reader is asked to trust the authors and that is not acceptable. If space is a problem then a clear figure should be added to the supplemental file. Further, the authors report the peptides identified that are a match to VDAC 2 but they do not report what peptides were detected that did not match. What fraction of the peptides matched to VDAC2? There is an unfortunate history of attributing action to VDAC that was not due to VDAC and of identifying VDAC as a binding site when another protein was responsible. This manuscript provides other evidence that supports the conclusion that VDAC2 is the target but that does not excuse the failure to honestly report on other proteins present.
2) The authors state “... efsevin restores rhythmic cardiac contraction in tre by potentiating VDAC2 activity.” The authors do not measure VDAC2 activity and thus this conclusion is unnecessary speculation. VDAC2 is not an enzyme and thus its action is not so easily defined in this context. The first part of the last sentence on that page is the appropriate summarizing statement.
3) The authors state: “VDAC is an abundant channel located on the outer mitochondrial membrane serving as a primary passageway for metabolites and ions including Ca2+ (Fig.5A) (Rapizzi et al., 2002; 162 Bathori et al., 2006; Shoshan-Barmatz et al., 2010).“ The authors fail to point out that it is the closed state of VDAC that favors Ca++ flux. That seminal publication was not cited. (Biochim Biophys Acta. 2007 Oct;1768(10):2510-5.) The state of VDAC2 as well as its presence is critical.
4) Figure 5 presents some very interesting findings but experimental details are sometimes missing and unclear. In the methods section it talks about perfusion whereas in the figure legend it is called superfusion. It is my understanding that the Fluorostar plate reader does not have perfusion capabilities but rather has the ability to make fluid additions while recording. Is that what was done? Also in the methods section it does not specify the total amount of Ca++ present in the solution, just the presumably calculated free Ca++ concentration. The actual amount of calcium present should be stated and whether the free Ca++ concentration reported is a calculated value of a measured value. Either way it is unclear why the reported value is an exact value. Is this a guess, an approximation?
In addition, the authors state “We examined whether Ca2+ released from intracellular stores could be directly transported into mitochondria through VDAC2 and whether this process could be modulated by efsevin...” The experiments reported do not distinguish between Ca++ release into the medium and quickly taken up my mitochondria or Ca++ directly traveling between the ER and the mitochondria. That could have been tested in 3C by having a strong chelator in the medium. Thus, with the current data the authors cannot claim evidence for direct transfer. They can claim a more effective uptake of Ca++ into mitochondria after efsevin treatment.
Based on the results reported in 5B, the authors conclude that Ca++ has been transported into mitochondria in the permeabilized cells, ignoring the ER. The choice of Rhod2 was presumably based on its affinity for Ca++ and its likely saturation with Ca++ in the ER, so that observed changes are likely due to the mitochondria. The authors make no effort to point this out to the reader. It should be addressed.
5) In the conclusions the authors state: “In this model, VDAC-dependent Ca2+ uptake controls the duration and the diffusion of cytosolic Ca2+ near the Ca2+ release sites...” How, physically, can VDAC control the “duration of cytosolic Ca++” and the “diffusion of cytosolic Ca++”? VDAC cannot control diffusion nor can it control the duration of diffusion. What I hope the authors want to say is that VDAC2 facilitates Ca++ uptake via MCU thus reducing the local Ca++ concentration more rapidly than otherwise.https://doi.org/10.7554/eLife.04801.029
We appreciate the reviewers’ insightful and constructive suggestions to help us improve our manuscript. We have taken the reviewers’ advice and revised the manuscript accordingly. In addition to clarifying the text, we revised Figure 3, present new data in Figure 7, and added three supplementary figures. Below, we address each reviewer’s critiques specifically.
Reviewer #1 questioned whether the expression of MCU complex would be affected by NCX1 mutation or by the molecular stimulation of VDAC2.
We address this question by in situ hybridization. We find that the expression levels of MCU and MICU1 are comparable between wild type and tremblor hearts with and without efsevin treatment. These findings are presented in Figure 7– figure supplement 1.
Reviewer#1 also questioned whether VDAC2 interacts with MCU and its regulators.
Our data that restoration of rhythmic contractions in tremblor requires the activity of both VDAC2 and MCU provides genetic evidence supporting a functional interaction between VDAC2 and MCU (Figure 7B-D). As the reviewer pointed out, mitochondrial Ca2+ transport involves MCU and its regulators. Our model would predict that manipulating the activity of other components of the MCU complex would affect the rescue effect of VDAC2 overexpression. In this revised manuscript, we provide new data showing that: 1) MICU1 is expressed in the developing heart (Figure 7A), 2) overexpression of MICU1 restores rhythmic contractions in tremblor and this effect requires functional VDAC2 (Figure7E), and 3) there is a synergistic effect of VDAC2 and MICU1 overexpression (Figure 7F). These findings provide evidence to support the genetic interaction between VDAC2 and the MCU complex and strengthen our hypothesis that mitochondrial Ca2+ buffering is beneficial to maintain cardiac rhythmicity. Future studies on the biochemical interaction between these proteins would provide further mechanistic insights.
Reviewer #2 made a few suggestions to Figure 3:
The reviewer thought that the sample Mass Spec image presented in Figure 3F was too small for inspection, and questioned the identities of peptides detected in our Mass Spec analysis.
We thank the reviewer for pointing out this problem. In the original Figure 3F, we presented the sample Mass Spec image in the top panel and indicated eight peptides that match VDAC2 in the lower panel. In this revised manuscript, we rearranged Figure 3 as suggested; we present peptide identity in new Figure 3F and provide an enlarged Mass Spec image in Figure 3–figure supplement 1.
The reviewer also questioned whether we identified any other proteins interacting with efsevin.
In the course of our study, we inspected six sets of data. VDAC2 was the only protein that was not present in the controls but was consistently identified from the efsevin affinity column. Vitellogenin 1 (vtg1) and NADPH quinine 1 (nqo1) were frequently found in both the control and efsevin affinity columns and thus were considered as contaminants. More importantly, our genetic analysis showed that overexpression of VDAC2 recapitulates the rescue effect of efsevin on tremblor and knocking down VDAC2 abolishes the responsiveness of tremblor to efsevin. Together, these findings support VDAC2 as a target of efsevin.
Reviewer #2 raised a few questions regarding the mitochondrial Ca2+ uptake analyses.
The reviewer questioned the experimental detail about the mitochondrial Ca2+ uptake experiment in HeLa cells.
We apologize for the confusing description. In this experiment, we added Ca2+ into the samples to make the final Ca2+ concentration as approximately 10 μM. We have now clarified this information in the Methods and Figure Legend.
The reviewer thought it would be helpful to point out how Rhod2 functions as a mitochondrial Ca2+ indicator.
We appreciate the reviewer’s concern and include this information in the Methods section.
The reviewer thought that “examine whether Ca2+ released from intracellular stores could be ‘locally’ transported into mitochondria through VDAC2” would be a better description of our analysis.
We thank the reviewer for the suggestion. We have revised the manuscript accordingly. In addition, we provide new data to support a local IP3R-mitochondrial Ca2+ transfer through VDAC2 (Figure 5–figure supplement 1). Previous reports have provided multiple lines of evidence in various cell types to support local Ca2+ transfer between IP3 receptors and mitochondria. A commonly used approach is that while both IP3-mediated and SERCA inhibition-induced discharge of the ER Ca2+ store cause similar global cytoplasmic [Ca2+] signal, only the IP3-induced Ca2+ release is rapidly propagated to the mitochondria (Rizzuto et al. 1994, Hajnoczky et al. 1995). We now present a similar experiment with V1/V3DKO MEFs to specifically support local Ca2+ delivery from IP3 receptors to VDAC2/mitochondria (Figure 5–figure supplement 1).
Finally, Reviewer #2 made a few editorial suggestions (points 2, 3, 5).
We thank the reviewer for the suggestions. We have revised the manuscript accordingly.https://doi.org/10.7554/eLife.04801.030
- Jau-Nian Chen
- Ohyun Kwon
- Hirohito Shimizu
- Fei Lu
- Fei Lu
- Jau-Nian Chen
- Johann Schredelseker
- Atsushi Nakano
- Thomas M Vondriska
- Sarah Franklin
- Joshua I Goldhaber
- György Hajnóczky
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
The authors thank Kenneth D Philipson, James N Weiss and Adam D Langenbacher for comments on the manuscript, Janice Ahn for assisting the initial chemical screen and Lingling Peng for the synthesis and Yi Chiao Fan for the characterization of efsevin and its derivatives. We also thank Jing Huang, James N Weiss and the UCLA cardiovascular research laboratory for reagents and infrastructure, and Jinghua Tang of UCLA-BSCRC for technical assistance on human ES cell works. We thank William Craigen for providing V1/V3 DKO MEFs.
Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols of the University of California, Los Angeles and the Cedars-Sinai Hospital. The protocols were approved by the Cedars-Sinai Institutional Animal Care and Use Committee (#003574 for the use of mouse cardiomyocytes), the Office of Animal Research Oversight that oversees the Ethics of Animal Experiments (ARC# 2000-051-43B for the use of zebrafish) and Embryonic Stem Cell Research Oversight (#2009-006-06 for the use of ES cells) of the University of California, Los Angeles. Every effort was made to minimize suffering.
- Jodi Nunnari, University of California, Davis, United States
- Received: September 17, 2014
- Accepted: December 23, 2014
- Version of Record published: January 15, 2015 (version 1)
© 2015, Shimizu et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.