Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Aug 1.
Published in final edited form as: J Cell Physiol. 2019 Apr 1;234(10):18371–18381. doi: 10.1002/jcp.28471

Evidence for the Impact of BAG3 on Electrophysiological Activity of Primary culture of Neonatal Cardiomyocytes

Farzaneh G Tahrir 1, Jennifer Gordon 1, Arthur M Feldman 2, Joseph Cheung 2,3, Kamel Khalili 1,†, Taha Mohseni Ahooyi 1,†
PMCID: PMC6830737  NIHMSID: NIHMS1052564  PMID: 30932190

Abstract

Homeostasis of proteins involved in contractility of individual cardiomyocytes and those coupling adjacent cells is of critical importance as any abnormalities in cardiac electrical conduction may result in cardiac irregular activity and heart failure. Bcl2-associated athanogene 3 (BAG3) is a stress induced protein whose role in stabilizing myofibril proteins as well as protein quality control (PQC) pathways, especially in the cardiac tissue, has captured much attention. Mutations of BAG3 have been implicated in the pathogenesis of cardiac complications such as dilated cardiomyopathy. In this study, we have used an in vitro model of neonatal rat ventricular cardiomyocytes (NRVCs) to investigate potential impacts of BAG3 on electrophysiological activity by employing microelectrode array (MEA) technology. Our MEA data showed that BAG3 plays an important role in the cardiac signal generation as reduced levels of BAG3 led to lower signal frequency and amplitude. Our analysis also revealed that BAG3 is essential to the signal propagation throughout the myocardium, as the MEA data-based conduction velocity, connectivity degree, activation time and synchrony were adversely affected by BAG3 knock-down. Moreover BAG3 deficiency was demonstrated to be connected with the emergence of independently beating clusters of cardiomyocytes. On the other hand, BAG3 overexpression improved the activity of cardiomyocytes in terms of electrical signal amplitude and connectivity degree. Overall, by providing more in-depth analyses and characterization of electrophysiological parameters, this study reveals that BAG3 is of critical importance for electrical activity of neonatal cardiomyocytes.

Keywords: BAG3, Cardiomyocytes, Conduction Velocity, Microelectrode Array

Introduction

Electrical potential change as a result of sequential activation and inactivation of specific ion channels resulting in ion movement across plasma membrane forms the molecular basis of action potential (AP) [Grant et al., 2009]. APs are propagated throughout the myocardium via the cable properties of specific conduction fibers (His bundle and Purkinje fibers) and between cells via gap junctions (GJs) [Epifantseva et al., 2018]. At the organ level, electrical activity of the heart is measured by electrocardiogram (ECG), while changes in electrical activity at the level of individual myocytes is best measured with patch-clamp and Microelectrode array (MEA) which measures extracellular ion currents known as field potentials (FPs) [Navarrete et al., 2013]. Alterations in electrophysiological parameters form the basis for increased arrhythmogenesis and are commonly associated with heart failure [Sameni et al., 2010]. Among various techniques used to measure cardiac electrical activity, MEA screening mapping technique is a high resolution and non-invasive method that has been widely used for drug screening in pharmacological studies to identify cardiotoxic drugs [Johnstone et al., 2010, Jans et al., 2017]. Precise measurement and analysis of MEA-derived FP recordings leads to a better understanding of electrophysiological activity of cardiac cells; as FP parameters correlate with those measured by ECG or patch-clamp. For example, FP duration (FPD) correlates with AP duration (APD) measured by patch-clamp and QT intervals measured by ECG [Navarrete et al., 2013].

Each adult cardiomyocyte is coupled to ~ 11 adjacent cells to ensure anisotropic electrical conduction in the myocardium [Rohr et al., 2004]. During electrical activation of the myocardium, excitatory electrical currents spread through membrane channels and depolarize cardiomyocytes in atria and ventricles that results in contraction of the heart muscle in a coordinated manner [Grant et al., 2009]. The functionality of this complex machinery is guaranteed by maintaining the homeostasis of the constituent subunits underlying cardiac electrical conduction; as their dysregulation results in cell-to-cell uncoupling and has been implicated in the pathogenesis of arrhythmia and infarction [Vaidya et al., 2001, Schulz et al., 2015]. Furthermore, previous studies demonstrated that cardiac conduction defects in mutant embryos were associated with aberrant cardiac remodeling and uncoordinated ventricular contraction [Chi et al., 2010]. Therefore, understanding the precise mechanisms and key molecular players underlying electrical impulse propagation in the myocardium is of great importance toward the development of novel clinical interventions for cardiac complications such as arrhythmia.

BAG3 is a member of the Bcl2-associated athanogene (BAG) family proteins that is composed of 575 amino acids. BAG3 is highly expressed in cardiac and skeletal muscles and plays an important role in myocardial homeostasis. Mutations of BAG3 have been reported in patients with cardiomyopathy [Myers et al., 2018]. BAG3 deficiency in mice led to muscle degeneration and lethal cardiomyopathy followed by death at 4 weeks of age [Homma et al., 2006]. Further studies demonstrated that BAG3 through interaction with the actin capping protein, CapZ, maintains structural stability of F-actin. Deficiency of BAG3 under mechanical tension resulted in degradation of CapZpi leading to disruption of the myofibril structure and diminished contractile activity of heart muscle [Hishiya et al., 2010]. Abnormal accumulation of proteins has been implicated in the pathogenesis of myofibrillar myopathies [Selcen et al., 2011]. BAG3 has been extensively investigated as a key regulator of protein homeostasis in the heart. For this purpose, BAG3, through interaction with heat shock proteins, targets dysfunctional proteins and accumulated aggregates for further degradation and clearance via PQC machinery [Myers et al., 2018]. Considering the role of BAG3 in regulating cardiac function and the impact of PQC failure in cardiac abnormalities, we hypothesized that BAG3 deficiency may impact electrical conductivity and subsequent contractility of the myocardium.

In this work, we have studied the impact of BAG3 dysregulation on electrical signal generation and propagation using an in vitro culture of spontaneously beating neonatal cardiomyocytes. Acquired data were analyzed using custom-designed MATLAB algorithms, and important electrophysiological parameters such as spike frequency, spike amplitude, velocity of impulse propagation and network connectivity were extracted and compared. The detailed experimental procedures are outlined in the Materials and Methods section. The Results section provides our observations on the effects of BAG3 alteration on different aspects of cardiac electrophysiological activity. Finally, the concluding remarks are provided in Discussion.

Experimental Procedure

Isolation and culture of neonatal cardiomyocytes

All animal protocols used in this study were reviewed and approved by the Institutional Animal Care and Use Committee at Temple University. NRVCs were isolated from 2–3 day old Harlan Sprague-Dawley rats (Charles River) using the protocol previously described [Gupta et al., 2016]. Isolated cardiomyocytes were plated on EmbyoMax® 0.1% Gelatin solution (Millipore)-coated plates in minimal essential medium α (MEMα, Life Technologies) with 10% fetal bovine serum (FBS, Denville Scientific Inc., Hollistion, MA) and then 24 hours post seeding, the medium was replaced with Dulbecco’s modified Eagle medium (DMEM, Life Technologies, Carlsbad, CA) with 2% FBS and 25 μg/mL gentamicin (Life Technologies).

Adenovirus transduction

Twenty-four hours after isolation, NRVCs were transduced with either Ad-control (Vector Biolabs) or Ad-siBAG3 (Vector Biolabs) or Ad-BAG3 (prepared in-house) in reduced volumes of FBS-free DMEM for 2 hours. The transduction medium was then replaced with DMEM supplemented with 2% FBS and 25 μg/mL gentamicin.

MEA recording

NRVCs were plated at a density of 1000 cells/mm2 on MEA glass arrays (Multichannel Systems, Germany) pre-coated with EmbyoMax® 0.1% Gelatin solution (Millipore) and were allowed to stabilize for 48 hr. Each MEA plate consists of 60 titanium nitrate (TiN) electrodes with a diameter of 30 μm embedded in the bottom of the dish and positioned in a rectangular grid. Two days post seeding, extracellular recordings of untreated cardiomyocytes were performed as initial conditions (Day 0) using the MEA-1060 system (Multichannel systems, Germany). Then afterwards, cardiomyocytes were transduced with either Ad-siBAG3, Ad-BAG3 or Ad-control and electrophysiological activities were recorded over time via MEA (Day 1, Day 2 and Day 3) and reported as Local Field Potentials (LFPs). Extracellular recordings were performed at a sampling frequency of 2 kHz using MC_Rack software, which generates data in the format of .mcd. Then, by using the MC_Data Tool, data were converted to ASCII format for further analysis by MATLAB®. Each recording was performed for 10 min and cardiomyocytes were maintained in an incubator at 37 °C supplied with 5% CO2 throughout the experiments. As a first step to analyze data, signals were filtered by a low-pass filter to enhance signal-to-noise ratio by attenuating the electrical signals with frequency higher than cutoff frequency (fc). In order to determine signal fc, Fast Fourier Transform (FFT) was applied to decompose the time-domain signal into its constituent frequencies. The resulting transformation represents a signal in the frequency domain with the intensity associated with each frequency (Figure S1A). Typically the background noise amplitude determined from the reference electrode was about 12 μV (Figure S1B).

Conduction velocity estimation

Conduction velocity (CV) is the speed at which electrical activity is transmitted among the electrodes throughout the culture. Local signal CV between electrode i and j, Vi,j is calculated by dividing the distance between two electrodes (Δli,j) by their activation time distance (Δti,j) and mean CV for electrode i (Vi) is calculated using the equation below:

Vi,j=Δli,j|Δti,j|,Vi=∑j=1,j≠i60Vi,jn−1

n denotes the number of electrodes.

The overall signal CV is calculated as an average signal velocities calculated for 60 electrodes:

SignalCV=∑i=160Vi60

Network connectivity analysis

Connectivity maps were calculated based on the partial correlation of pairs of electrodes controlling the rest of the array according to the method described previously [Ahooyi et al., 2017]. Partial correlation at any given period of time, Δt, indicates whether two signals are correlated regardless of the rest of the network. The adjacency matrix derived by this method was used to construct a connectivity graph which provides information on which parts of the network were likely to affect each other directly. The connectivity order is calculated as follows:

Connectivityorder=∑numberofconnectingedgesmaximumpossibleconnectivity(=1800for60electrodes)

Synchrony analysis

MATLAB peak detection algorithm was applied to the filtered MEA data to identify peaks of each individual channel over the period of recording. After all channels peaks were plotted next to each other, the resulting raster plots were used to investigate whether the peaks of a selection of electrodes at the onset of a spiking activity showed synchrony.

Cluster analysis

To study the formation of independently beating clusters of cardiomyocytes, the pairwise Spearmen’s rho rank correlation of channels data were calculated. Hierarchical clustering was then applied to the resulting 60-by-60 positive semidefinite matrix to identify clusters of channels that show high level of correlation similarities over the course of recording. This method is particularly useful as there are distinctly beating clusters with independent beating times. For cases when the low conduction velocity leads to delayed activation of electrodes, rank correlation may not be able to detect distantly correlated channels. For such cases cross-rank correlation is a more accurate measure of segregated clusters.

Statistical analysis

Student’s t-test was used to calculate the statistical differences between each two independent groups. p values less than 0.05 were considered as statistically significant.

Results

BAG3 suppression reduces signal frequency and amplitude

NRVCs were transduced with either Ad-control or Ad-siBAG3 and extracellular recordings were performed using MEA technology. Results from western blot verified a significant decrease in the level of BAG3 in cells upon transduction with Ad-siBAG3 (Figure 1A–B). Each spike in FP recordings is characterized by a biphasic peak with positive and negative potentials. Spike frequency was calculated as the number of spikes divided by the recording time. Our results showed that adenovirus transduction enhanced spike frequency 1 day post transduction, which might occur due to the activation of inflammatory responses as a result of adenovirus transduction as reported previously [Byrnes et al., 1995]. Then over time, BAG3 knock down led to lower spike frequency as compared to the control cardiomyocytes. 3 days post transduction, spike frequency was 1.7-fold higher in control cells compared to their initial frequency; while it was reduced to 0.37 compared to initial frequency in BAG3-suppressed cardiomyocytes. We also calculated spike amplitude as the voltage difference between positive and negative values of each spike. Data indicated that signal amplitudes in both control and BAG3-suppressed cardiomyocytes were significantly reduced 3 days post transduction. However, a decrease in spike amplitude in BAG3-suppressed cells is likely underestimated since signals from silent electrodes were not included in spike amplitude calculation (Figures 1C–E).

Figure 1. BAG3 suppression attenuates electrophysiological activity of cardiomyocytes.

Figure 1.

Open in a new tab

(A) NRVCs were transduced with either Ad-control or Ad-siBAG3 and the levels of BAG3 protein were evaluated 24, 48 and 72 hr post transduction using western blot. (B) BAG3 protein levels were quantified based on the data presented in (A). *p<0.05, ***p<0.001 (n=4). (C) Electrical activities of NRVCs were recorded as initial conditions (Day 0). Cells were then transduced with either Ad-control or Ad-siBAG3 and extracellular FPs were recorded via MEA over time (Day 1, Day 2 and Day 3). (D) The impact of BAG3 suppression on spike frequencies was quantified based on the data presented in (C). (E) The impact of BAG3 suppression on spike amplitudes was quantified based on the data presented in (C). ***p<0.001 (n=3).

BAG3 knockdown dysregulates synchronous activity of cardiomyocytes

In order to further analyze the spiking activity of cardiomyocytes recorded simultaneously by 60 electrodes, spike raster plots of control and BAG3-suppressed cardiomyocytes were plotted over time. For this purpose, a MATLAB algorithm was designed to detect spikes of each channel from the filtered signal and plot them over the course of recordings. Figure 2A shows two dimensional spiking events of 60 electrodes over a 10 seconds time span. The Y-axis corresponds to the electrode number and the X-axis corresponds to time. The occurrence time of a spike for each channel is illustrated by a dot at a certain time. According to this analysis, spike activation at initial time point (Day 0) indicates similar patterns for both conditions; while BAG3 suppression led to dysregulation of spike activity over time as the number of spikes reduced and irregular spontaneous activity increased as a result of BAG3 reduction. We then magnified spike occurrence within 1000 ms of recording and found that spontaneous firing pattern and spike timing were dysregulated as a result of BAG3 suppression over time (Figure 2B).

Figure 2. Spike rasters depicting simultaneous activity of 60 electrodes.

Figure 2.

Open in a new tab

NRVCs were transduced with either Ad-control or Ad-siBAG3 and spike rasters were plotted within (A) 10,000 ms and (B) 1000 ms of recording. Data indicate that the number of firing and regular activity of 60 electrodes embedded throughout the culture reduced as a result of BAG3 suppression.

BAG3 suppression dysregulates signal propagation across the culture

By knowing the activation time of each electrode, we then plotted color-coded activation maps to exhibit the effect of BAG3 suppression on electrical signal initiation and propagation within the culture of spontaneously beating cardiomyocytes. To this end, an 8-by-8 grid represents activation time of 60 electrodes arranged throughout the monolayer culture and each segments of the grid indicates local activation times. Activation maps indicate where the action potential originates (lighter regions) and how it propagates throughout the rest of the array (red regions). 3 days post-transduction, activation maps indicated faster signal propagation in control cells as compared to Day 0; while areas with delayed activation (darker regions) increased under BAG3 knock down. In the control culture, total activation time of 60 electrodes reduced from 58.27±12.36 ms to 43.78±11.23 ms by 3 days post-transduction; while activation time increased from 47.52±13.84 ms to 160.97±63.58 ms under BAG3 suppression (Figure 3A–B).

Figure 3. BAG3 suppression reduces electrode activation and impulse progression.

Figure 3.

Open in a new tab

(A) Electrodes are represented within an 8-by-8 grid in which each segment represents local activation time of each electrode in control and BAG3-suppressed cardiomyocytes on day 1 and day 3 post transduction. Color scale bar depicts regions with earliest (yellow) and latest (red) electrical activation (0 to 60 ms). (B) Quantification analysis indicated that signal activation time reduced from 58.27±12.36 ms to 43.78±11.23 ms 3 days post transduction in control cardiomyocytes; while signal activation time increased from 47.52±13.84 ms to 160.97±63.58 ms 3 days post transduction in BAG3-suppressed cardiomyocytes. ***p<0.001 (n=3).

BAG3 suppression promotes asynchrony in the cardiomyocytes culture (no physical discontinuity)

Next, we hypothesized that BAG3 suppression may dysregulate synchronous beating across the cardiomyocyte culture. For this purpose, we assessed Spearman’s rank correlation coefficient between each pair of electrodes (Ei and Ej, i,j=1:60) in the MATLAB environment. Generally, correlation coefficient varies from −1 to +1. Correlation coefficient of 0 indicates no correlation between two electrodes. Correlation coefficient of +1 indicates perfect relationship, positive correlation, and correlation coefficient of −1 indicates perfect negative correlation between two electrodes (in this case negative correlations are nonexistent). Once all pairwise correlations were calculated, hierarchical clustering identifies clusters of electrodes with similar pairwise correlation patterns with other electrodes. The results were then plotted as a symmetrical 60-by-60 matrix in which each element represents the extent of correlation between each pair of electrodes. In this representative matrix, green indicates no correlation and red indicates high correlation between two electrodes. Data analysis indicated that 3 days post transduction, two distinct clusters of beating cardiomyocyte populations which were highly correlated within each cluster were found in BAG3-suppressed cardiomyocytes, while there was only one larger cluster of highly correlated electrodes 3 days post transduction in control cells. This observation suggests the important role of BAG3 in maintaining the cardiac tissue as a single contracting unit, whose deficiency exerts detrimental effects on the cardiac tissue integrity and function (Figure 4).

Figure 4. Correlation matrices representing the impact of BAG3 suppression on coordinated activity of primary cardiomyocytes.

Figure 4.

Open in a new tab

Hierarchical clustering of MEA recording-based correlation matrices for primary neonatal cardiomyocytes reveals that control cardiomyocytes (top row) exhibit one single cluster; while BAG3 suppression (bottom row) results in the formation of two distinct clusters of cardiomyocyte 3 days post transduction.

BAG3 suppression alters signal conduction velocity within the culture

Figure 5A demonstrates average wave conduction velocity (CV) calculated based on the data recorded from 60 electrodes. CV was estimated for the control and BAG3-knocked down cardiomyocytes from initial condition (Day 0, no transduction) up to 3 days post-transduction (Day 3). We then calculated the mean of maximum CV (CVmax) for each culture and found that in control cells CVmax reduced by 8% from 36.1±5.6 cm/s on Day 0 to 33.28±2.6 cm/s on Day 3; while in BAG3-suppressed cardiomyocytes CVmax significantly reduced by 42% from 34.0±1.52 cm/s on Day 0 to 20.01±1.89 cm/s on Day 3 (Figure 5B).

Figure 5. Suppression of BAG3 results in signal CV reduction.

Figure 5.

Open in a new tab

(A) Electrophysiological activities of cardiomyocytes were recorded at initial condition (Day 0, no transduction) until 3 days post transduction for control and BAG3-suppressed cardiomyocytes. CVs were then calculated and represented within 60 seconds of recording. (B) CVmax was calculated based on the data presented in (A) and results showed that CVmax was reduced by 8% in control cells while BAG3 suppression resulted in a 42% reduction of CVmax 3 days post transduction. ***p<0.001 (n=3).

BAG3 knock down adversely affects the network connectivity

We estimated the network connectivity map through detecting the highly correlated electrodes using the MEA data. The connectivity map density increases from the initiation step (sparsely connected) to the propagation step and culminates at the contraction peak (Figure 6A). The network connectivity degree defined by the summation of all edges at each period of time, At, provides a measure of how correlated regions of the culture are. As shown in Figure 6B, the control cardiomyocytes exhibit high level of correlation with every spiking activity on Day 0 (pre-treatment). This high degree of connectivity continues with a slight decrease 3 days post-transduction. In contrast, BAG3-knocked down cardiomyocytes undergo a drastic deformation and reduction in the average connectivity over the course of treatment as compared to pretreatment (Figure 6C).

Figure 6. Connectivity map showing the degree of correlation among different electrodes.

Figure 6.

Open in a new tab

(A) During the generation of spikes typically three different levels of connectivity are observed; initiation with sparsely connected network, propagation with increasing number of connections, and contraction with maximum number of connecting edges. From this point on, the connectivity degree returns to resting state. (B) Normalized connectivity degree of cardiomyocytes transduced with Ad-control showing a slight increase in the maximum connectivity degree during spiking activity followed by a slight decrease. (C) Normalized connectivity degree of cardiomyocytes undergoing BAG3 KD shows a gradual disruption of average connectivity degree until 3 days post transduction. The heat-map indicates the difference between outward and inward connections for each electrode. The brighter channels have a higher number of outward connections.

BAG3 overexpression improved cardiac connectivity

NRVCs were transduced with either A-control or Ad-BAG3 and BAG3 levels were evaluated using western blot analysis (Figure 7A–B). To investigate the effects BAG3 overexpression on cardiomyocytes, some key markers of cardiomyocyte electrophysiological activity at a network scale were analyzed using MEA recordings. As shown in Figure 7C, BAG3 overexpression improved the cardiac spiking amplitude as compared to the control group of Figure 1C. Also the frequency of spikes increased more than 42% 3 days post Ad-BAG3 transduction. BAG3 overexpression has also had no adverse effect on the maximal connectivity measure compared to Day 0. Considering that this measure relatively reduced in the control group, BAG3 overexpression improved the connectivity over the course of recordings. Also, average connectivity increased 3 days after treatment with Ad-BAG3 (Figure 7D). Overexpressing BAG3 did not change the signal conduction velocity throughout the network of NRVCs (Figure 7E) as compared to Day 0, which is similar to the control group.

Figure 7. NRVC spiking activity under BAG3 overexpression.

Figure 7.

Open in a new tab

(A) NRVCs were transduced with either Ad-control or Ad-BAG3 and the levels of BAG3 protein were evaluated 3 days post transduction using western blot. (B) BAG3 protein levels were quantified based on the data presented in (A). **p<0.01 (n=4). (C) Spiking activity of 60 electrodes shown together exhibits an increase (~42%) in the number of spikes and improvement in spiking amplitude (compared to the control group) 3 days post Ad-BAG3 transduction. (D) Connectivity map shows a more regular pattern as BAG3 is overexpressed in NRVCs compared to the pretreatment condition. (E) Unlike BAG3 suppression, BAG3 overexpression does not have any negative effect on the conduction velocity.

Discussion

BAG3 is highly expressed in cardiac and skeletal muscles and has been reported as a key regulator of PQC in cardiomyocytes [Tahrir et al., 2017, Su et al., 2016]. BAG3 deficiency in mice was associated with the development of dilated cardiomyopathy [Fang et al., 2017]. Recently, BAG3 was reported to play an important role in quality control of connexin 43, a key component in gap junctions in the heart which modulates intercellular electrical connectivity, in neonatal cardiomyocytes [Tahrir et al., 2019]. Furthermore, BAG3 was demonstrated to co-localize with and regulate L-type Ca2+ channel activity in adult mouse left ventricular myocytes [Feldman et al., 2016]. However, the effects of BAG3 on initiation and propagation of cardiac electrical signals have not been described in detail. As an initial approach to this complex problem, we have used multisite, simultaneous and non-invasive MEA technology to monitor and analyze the impact of BAG3 downregulation on the electrophysiological activity of primary cardiomyocytes. NRVCs were transduced with adenovirus and extracellular electrical activities were recorded over a period of 3 days post-transduction and analyzed to extract important electrophysiological components.

The first major finding using raster plots (which indicates population events underlying each spike occurrence) was a significant reduction in the number of spikes over time as a result of BAG3 suppression. Furthermore, spike distribution and spike timing among the electrodes were altered, suggesting that the number of random firings increased within the culture as a result of BAG3 reduction. Since plasma membrane ion channels mediate action potential initiation, amplitude and morphology, our observations suggest that BAG3 may regulate the levels of ion channels or their function, as has been demonstrated for L-type Ca2+ channels [Feldman et al 2016]. The physiological relevance of our observation is that extracellular action potential frequency correlates with heart beat obtained from ECG, suggesting that BAG3 reduction may lead to Bradyarrhythmias (decreased spike frequency) and increased premature beats (increased random spike firings). The second major finding is that BAG3 suppression delayed activation time and reduced conduction velocity, which together resulted in slowed signal propagation across the NRVC monolayer. Alterations in connexin 43 homeostasis associated with BAG3 downregulation may partly account for this phenomenon [Tahrir et al., 2019]. Finally, synchronous activation pattern throughout the monolayer culture reduced as a result of BAG3 knock down, which may lead to increased arrhythmic foci in the intact heart.

BAG3 downregulation not only reduced the intensity of spiking activity, but also promoted random firings and disrupted the transmission of electrical signals in NRVC monolayers. We have recently reported that BAG3 suppression did not cause significant cell death in neonatal cardiomyocyte [Tahrir et al., 2017]; therefore, this suggests that BAG3 plays a key role not only in the Z-disc integrity and mechanical activity within cardiomyocytes, but also in electrical signal initiation and propagation in a population of cardiomyocytes. Since BAG3 levels are reduced in many models of cardiomyopathy including human heart failure [Feldman et al., 2014], and increased arrhythmogenesis is commonly observed with heart failure, our novel observations support the hypothesis that BAG3 is a rational therapeutic target for heart failure, not only in terms of enhancing contractility, but also in reducing risks of arrhythmias. On the other hand, BAG3 overexpression improved spiking amplitude and connectivity degree among the cardiomyocytes, as compared to the control which signifies the role of BAG3 in maintaining the electrophysiological activity of the myocardium.

Supplementary Material

Supp

Acknowledgments

The authors wish to thank past and present members of the Department of Neuroscience/Center for Neurovirology for their support, and sharing of ideas and reagents.

Funding Statement

This work was made possible by grants R01HL123093 awarded to KK, JYC and AMF, and P30MH092177 awarded to KK.

Footnotes

Conflicts of Interests

The authors declare that there are no competing financial interests.

REFERENCES

  1. Ahooyi TM, Shekarabi M, Decoppet EA, Langford TD, Khalili K. (2018). Network analysis of hippocampal neurons by microelectrode array in the presence of HIV-1 Tat and cocaine. The Journal of Cellular Physiology, 233, 9299–9311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Asai Y, Tada M, G Otsuji T, Nakatsuji N. (2010). Combination of functional cardiomyocytes derived from human stem cells and a highly-efficient microelectrode array system: an ideal hybrid model assay for drug development. Current Stem Cell Research and Therapy, 5, 227–232. [DOI] [PubMed] [Google Scholar]
  3. Byrnes AP, Rusby JE, Wood MJ, Charlton HM (1995). Adenovirus gene transfer causes inflammation in the brain. Neuroscience, 66, 1015–24. [DOI] [PubMed] [Google Scholar]
  4. Epifantseva I, Shaw RM (2018). Intracellular trafficking pathways of Cx43 gap junction channels. Biochimica et Biophysica Acta (BBA)-Biomembranes, 1860, 40–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Fang X, Bogomolovas J, Wu T, Zhang W, Liu C, Veevers J, Stroud MJ, Zhang Z, Ma X, Mu Y, Lao DH (2017). Loss-of-function mutations in co-chaperone BAG3 destabilize small HSPs and cause cardiomyopathy. The Journal of Clinical Investigation, 127, 3189–3200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Feldman AM, Gordon J, Wang J, Song J, Zhang XQ, Myers VD, Tilley DG, Gao E, Hoffman NE, Tomar D, Madesh M. (2016). BAG3 regulates contractility and Ca2+ homeostasis in adult mouse ventricular myocytes. The Journal of Molecular and Cellular Cardiology, 92, 10–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Feldman AM, Begay RL, Knezevic T, Myers VD, Slavov DB, Zhu W, Gowan K, Graw SL, Jones KL, Tilley DG, Coleman RC (2014). Decreased levels of BAG3 in a family with a rare variant and in idiopathic dilated cardiomyopathy. The Journal of Cellular Physiology, 229, 1697–1702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chi NC, Bussen M, Brand-Arzamendi K, Ding C, Olgin JE, Shaw RM, Martin GR, Stainier DY (2010). Cardiac conduction is required to preserve cardiac chamber morphology. Proceedings of the National Academy of Sciences, 107, 14662–14667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Grant AO (2009). Cardiac ion channels. Circulation: Arrhythmia and Electrophysiology, 2, 185–94. [DOI] [PubMed] [Google Scholar]
  10. Gupta MK, Tahrir FG, Knezevic T, White MK, Gordon J, Cheung JY, Khalili K, Feldman AM. (2016). GRP78 interacting partner BAG5 responds to ER stress and protects cardiomyocytes from ER stress-induced apoptosis. The Journal of Cellular Biochemistry, 117, 1813–1821. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hishiya A, Kitazawa T, Takayama S. (2010). BAG3 and Hsc70 interact with actin capping protein CapZ to maintain myofibrillar integrity under mechanical stress. Circulation Research, 107, 1220–1231. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Homma S, Iwasaki M, Shelton GD, Engvall E, Reed JC, Takayama S. (2006). BAG3 deficiency results in fulminant myopathy and early lethality. The American Journal of Pathology, 169, 761–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jans D, Callewaert G, Krylychkina O, Hoffman L, Gullo F, Prodanov D, Braeken D. (2017). Action potential-based MEA platform for in vitro screening of drug-induced cardiotoxicity using human iPSCs and rat neonatal myocytes. The Journal of Pharmacological and Toxicological Methods, 87, 48–52. [DOI] [PubMed] [Google Scholar]
  14. Johnstone AF, Gross GW, Weiss DG, Schroeder OH, Gramowski A, Shafer TJ (2010). Microelectrode arrays: a physiologically based neurotoxicity testing platform for the 21st century. Neurotoxicology, 31, 331–350. [DOI] [PubMed] [Google Scholar]
  15. Knezevic T, Myers VD, Su F, Wang J, Song J, Zhang XQ, Gao E, Gao G, Madesh M, Gupta MK, Gordon J. (2016). Adeno-associated virus serotype 9-driven expression of BAG3 improves left ventricular function in murine hearts with left ventricular dysfunction Ssecondary to a myocardial infarction. JACC: Basic to Translational Science, 1, 647–656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Myers VD, McClung JM, Wang J, Tahrir FG, Gupta MK, Gordon J, Kontos CH, Khalili K, Cheung JY, Feldman AM (2018). The multifunctional protein BAG3: A novel therapeutic target in cardiovascular disease. JACC: Basic to Translational Science, 3, 122–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Natarajan A, Stancescu M, Dhir V, Armstrong C, Sommerhage F, Hickman JJ, & Molnar P. (2011). Patterned cardiomyocytes on microelectrode arrays as a functional, high information content drug screening platform. Biomaterials, 32, 4267–4274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Navarrete EG, Liang P, Lan F, Sanchez-Freire V, Simmons C, Gong T, Sharma A, Burridge PW, Patlolla B, Lee AS, Wu H. (2013). Screening drug-induced arrhythmia using human induced pluripotent stem cell-derived cardiomyocytes and low-impedance microelectrode arrays. Circulation, 128(11 suppl 1), S3–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Nygren A, Kondo C, Clark RB, Giles WR (2003). Voltage-sensitive dye mapping in Langendorff-perfused rat hearts. The American Journal of Physiology-Heart and Circulatory Physiology, 284, H892–902. [DOI] [PubMed] [Google Scholar]
  20. Papadaki M, Bursac N, Langer R, Merok J, Vunjak-Novakovic G, Freed LE (2001). Tissue engineering of functional cardiac muscle: molecular, structural, and electrophysiological studies. The American Journal of Physiology-Heart and Circulatory Physiology, 280, H168–178. [DOI] [PubMed] [Google Scholar]
  21. Peters NS, Wit AL (1998). Myocardial architecture and ventricular arrhythmogenesis. Circulation, 97, 746–754. [DOI] [PubMed] [Google Scholar]
  22. Rohr S. (2004). Role of gap junctions in the propagation of the cardiac action potential. Cardiovascular Research, 62, 309–322. [DOI] [PubMed] [Google Scholar]
  23. Sameni R, Clifford GD (2010. A review of fetal ECG signal processing; issues and promising directions. The Open Pacing, Electrophysiology and Therapy Journal, 3, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Selcen D, Engel AG (2011). Myofibrillar myopathies In Handbook of Clinical Neurology Vol. 101, pp. 143–154, Elsevier. [DOI] [PubMed] [Google Scholar]
  25. Schulz R, Gorge PM, Gorbe A, Ferdinandy P, Lampe PD, Leybaert L. (2015). Connexin 43 is an emerging therapeutic target in ischemia/reperfusion injury, cardioprotection and neuroprotection. Pharmacology and Therapeutics, 153, 90–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Su F, Myers VD, Knezevic T, Wang J, Gao E, Madesh M, Tahrir FG, Gupta MK, Gordon J, Rabinowitz J, Ramsey FV (2016). Bcl-2-associated athanogene 3 protects the heart from ischemia/reperfusion injury. JCI Insight, 1, e90931 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Tahrir FG, Knezevic T, Gupta MK, Gordon J, Cheung JY, Feldman AM, Khalili K. (2017). Evidence for the role of BAG3 in mitochondrial quality control in cardiomyocytes. The Journal of Cellular Physiology, 232, 797–805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tahrir FG, Gupta MK, Myers V, Gordon J, Cheung JY, Feldman AM, Khalili K. (2019). Role of BAG3 in quality of gap junction protein, connexin 43, in cardiomyocytes. Scientific Reports, in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Tse G, Yeo JM (2015). Conduction abnormalities and ventricular arrhythmogenesis: the roles of sodium channels and gap junctions. IJC Heart and Vasculature, 9, 75–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Vaidya D, Tamaddon HS, Lo CW, Taffet SM, Delmar M, Morley GE, Jalife J. (2001). Null mutation of connexin43 causes slow propagation of ventricular activation in the late stages of mouse embryonic development. Circulation Research, 88, 1196–1202. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp
NIHMS1052564-supplement-Supp.pdf (337.2KB, pdf)

RESOURCES