Introduction

Persistent organic pollutants are a variegated group of chemicals that include polychlorinated biphenyls (PCBs), the latter compounds containing a high number of chlorine atoms. PCBs have been widely used in electrical and electronic industries (production of plastics, adhesives, paints, carbonless copying paper, newsprint and caulking compounds1,2), since when, in 1979, their production and use were internationally banned. However, due to their high chemical stability, PCBs are still persisting and have wide diffusion into the environment.

Growing amounts of data link their bioaccumulation through the diet with their high toxicity; more specifically, several studies have evidenced that PCBs exposure during adult life leads to an increase in the incidence and severity of cardiomyopathies3,4,5,6,7. PCBs exposure alters the expression of GATA-4, Nkx-2.5, MEF-2c, OCT-1 cardiac nuclear transcription factors and other heart-specific genes (atrial and brain natriuretic peptide, alpha- and beta-myosin heavy chain, alpha-cardiac and alpha-skeletal actin) in adult primary rat cardiomyocytes8. In guinea pig ventricular myocytes, PCB 19 exposure decreases contractile force, action potential duration and amplitude, and intracellular Ca2+ transients through inhibition of L-type voltage-gated Ca2+ channels (VGCCs)9. Also, in other adult cardiac cellular models, PCBs alter the expression and/or activity of several components of the intracellular Ca2+ signalling machinery10, such as type 2 ryanodine receptors (RyR2)11 and the Sarco-Endoplasmic-Reticulum Ca2+-ATPase (SERCA)12.

PCBs prenatal exposure through placenta determines congenital heart defects13,14 not attributable to inherited genetic mutations5. To this regard, exposed avian embryos display extensive cardiac dilation, thinner ventricle walls and reduced responsiveness to chronotropic stimuli15, whereas zebrafish embryos show heart defects, reduction of the heart beating rate and irregular and weak contractions16. Reduced heart size and functional deficits have been described in mouse foetuses, as well as cardiac hypertrophy in offsprings, the latter showing increased sensitivity to cardiovascular insults in adulthood5.

Importantly, being fat-soluble, PCBs accumulate in fat tissues and are passed to newborns through maternal milk17, increasing the myocardial wall thickness and inducing cardiac hypertrophy, thus impairing heart functionality in the adult18. Whilst the detrimental effects on the perinatal heart anatomy are evident, unknown are the PCBs effects at the cellular and molecular levels. Our study aims at dissecting the key physiological processes behind the observed heart contraction defects, by analyzing in perinatal cardiomyocytes the calcium-induced calcium-release mechanism (Ca2+ toolkit) that finely regulates contraction.

Perinatal-like cardiomyocytes (hereafter named cardiomyocytes) were obtained from the in vitro differentiation of mouse embryonic stem cells (mESCs). Functionally, these cardiomyocytes form compact beating syncytia and exhibit electrophysiological features of excitation-contraction coupling described for isolated perinatal cardiac cells19,20,21.

Following differentiation, cardiomyocytes were exposed for 24 h to Aroclor 1254 (Aroclor), a mixture of more than 80 PCBs isomers and congeners with high number of chlorine atoms (54%)2, at doses in the range of environmental contamination (1 and 2 µg/ml), then their kinematics contractile properties and the intracellular Ca2+ homeostasis were evaluated.

Results

Aroclor reduces the kinematics and dynamics properties (contractile properties) of beating syncytia

For the evaluation of the effects of Aroclor on the contractile properties of beating syncytia, we measured their kinematics and dynamics features on AVI videos, recorded from CTR samples and after 24 h exposure to 1 or 2 µg/ml Aroclor. The chronotropic (beat frequency [Hz]), inotropic (contraction force [pixel/s2] and contractility, i.e. the maximum contraction velocity, [pixel/s]) and ergotropic (consumption of ATP for kinetic energy [pixel2/s2]) features were mathematically calculated from the movement of the beating syncytia22.

After exposure to 1 µg/ml Aroclor, both beat frequency and kinetic energy showed a 1.2-fold reduction (p < 0.05), whereas inotropic parameters remain unaltered (p > 0.05) when compared to CTR (Fig. 1A,D). A more adverse effect was observed when syncytia were exposed to 2 µg/ml, since all four parameters were significantly reduced (p < 0.05). Specifically, when compared to CTR, a 1.4-, 1.2-, 1.4-, 1.8-fold decrease of beat frequency, contractility, contraction force and kinetic energy, respectively, was observed (Fig. 1 and videos supplementary materials).

Figure 1
figure 1

Kinematics contractile properties of cardiac beating syncytia. (A) Beat frequency [Hz]. (B) Contractility (maximum contraction velocity) [pixel/s]. (C) Contraction force [pixel/s2]. (D) Kinetic energy [pixel2/s2]. Horizontal bars represent the 95% confidence intervals for the differences between means according to the Least Significant Difference statistical test.

Full size image

Aroclor reduces spontaneous intracellular Ca2+ oscillations

Aroclor, as previously shown for other PCBs23,24, may induce alterations in the cardiomyocyte contraction properties through a derangement of the intracellular Ca2+ signalling machinery. Importantly, the contractile properties of beating syncytia depend on spontaneous Ca2+ oscillations, which arise even in the absence of electrical stimulation. These spontaneous Ca2+ spikes are generated by rhythmic Ca2+ release from sarcoplasmic reticulum (SR) through RyR2, which in turn evokes a Na+/Ca2+ exchanger-mediated membrane depolarization and activates VGCCs25,26.

When loading beating syncytia with the Ca2+-sensitive fluorochrome Fura-2/AM, we observed, compared to CTR unexposed cells, a significant (p < 0.05) decrease in both amplitude (Fig. 2A,B) and frequency of spontaneous Ca2+ transients (Fig. 2A,C) in cardiomyocytes exposed to 1 or 2 μg/ml Aroclor. These results show that Aroclor may reduce cardiomyocyte contraction properties by interfering with the intracellular Ca2+ handling machinery in beating syncytia.

Figure 2
figure 2

Ca2+ oscillations in cardiac beating syncytia. (A) Representative tracing of spontaneous Ca2+ spikes recorded in CTR or in the presence of either 1 or 2 µg/mL Aroclor. (B) Mean ± SE of the amplitude of the Ca2+ peaks recorded under the designated treatments. (C) Mean ± SE of the frequency of spontaneous Ca2+ spikes. *p < 0.05.

Full size image

Aroclor does not alter the molecular machinery underlying spontaneous Ca2+ activity

In CTR samples, tetracaine (100 μM), a selective RyR inhibitor27, caused a prompt inhibition of the intracellular Ca2+ spikes (Fig. 3A), thereby confirming the crucial role played by RyR2 in shaping the Ca2+ signal25,26. Similar to previous studies25, removal of external Ca2+ (0Ca2+) did not induce the abrupt interruption of the Ca2+ transients, but resulted in the progressive decline of their amplitude, determining complete arrest of Ca2+ activity after about 5 min (Fig. 3B). This finding is consistent with the notion that Ca2+ entry is triggered by spontaneous Ca2+ oscillations to refill the SR Ca2+ pool26. Conversely, the blockade of InsP3Rs with 2-APB (50 µM) failed to affect the ongoing Ca2+ transients (Fig. 3C). Overall, these findings confirm that RyR2-dependent rhythmical Ca2+ release underlies the onset of the spontaneous Ca2+ oscillations, whereas voltage-dependent Ca2+ entry maintains them over time.

Figure 3
figure 3

Intracellular Ca2+ spikes in CTR and 1 or 2 µg/ml-exposed cardiac beating syncytia. (A),(D) and (G) Tetracaine (100 µM) caused a prompt inhibition of the intracellular Ca2+ spikes; (B),(E) and (H) Extracellular Ca2+ removal (0Ca2+) did not induce the abrupt interruption of the Ca2+ transients, but resulted in the progressive decline in the amplitude of the Ca2+ transients; (C),(F) and (I) Superfusion of 2-APB (50 µM) did not affect spontaneous Ca2+ spikes.

Full size image

The same results were obtained when the beating syncytia were exposed to 1 (Fig. 3D–F) or 2 μg/ml (Fig. 3G–I) Aroclor. Therefore, RyR2 and VGCCs interact to shape the spontaneous Ca2+ oscillations also in Aroclor-exposed syncytia.

Earlier work showed that angiotensin II and endothelin 1 are able to modulate spontaneous Ca2+ oscillations in beating syncytia by inducing InsP3-induced Ca2+ release from the SR28. In agreement with these observations, the application of angiotensin II (1 μM) during ongoing oscillations caused a transient enlargement of the first Ca2+ spike, followed by a short-lasting reduction in the amplitude of the subsequent Ca2+ oscillations (Supplementary Figure 1SA). Conversely, when beating syncytia were exposed to 1 (Supplementary Figure 1SB) or 2 μg/ml (Supplementary Figure 1SC) Aroclor, angiotensin II (1 μM) caused a large Ca2+ spike whose decay to the baseline was overlapped by RyR2-dependent spontaneous Ca2+ oscillations. Of note, the Ca2+ response to angiotensin II was higher in the presence of 2 μM Aroclor. These observations suggest that the InsP3-sensitive Ca2+ pool is increased in the presence of Aroclor, which further rules out the contribution of InsP3Rs to the spontaneous Ca2+ oscillations (see Discussion).

Next, we determined how Aroclor interferes with the single components of the Ca2+ toolkit involved in the oscillatory signal, reducing spontaneous intracellular Ca2+ oscillations.

Aroclor affects the intracellular Ca2+ oscillations reducing the SR Ca2+ content

The acute addition of Aroclor in the 0Ca2+ medium did not induce any detectable variation in [Ca2+]i at both 1 (Fig. 4A) or 2 μg/ml (not shown). Instead, 24 h exposure to Aroclor caused a significant (p < 0.05) reduction in basal [Ca2+]i (Fig. 4B). Therefore, we turned our attention to the Ca2+-permeable pathways underlying the spontaneous Ca2+ oscillations, i.e., RyR2 and VGCCs. Caffeine (20 mM), a selective RyR agonist, triggered a Ca2+ response in CTR samples, while it failed in the presence of Aroclor at both concentrations (Fig. 5A,B). To further assess the effect of Aroclor on SR Ca2+ content, we used cyclopiazonic acid (CPA), known to inhibit SERCA activity thereby causing the progressive efflux of intraluminal Ca2+ through unidentified leakage channels29,30. This manoeuvre leads to the rapid depletion of the SR Ca2+ pool under 0Ca2+ conditions and, therefore, causes a transient increase in [Ca2+]i. The [Ca2+]i indeed returns to the original baseline due to the concerted interaction of the Na+/Ca2+ exchanger (NCX), plasma-membrane Ca2+-ATPase (PMCA) and mitochondria. CPA-induced intracellular Ca2+ mobilization is a well-known index of SR Ca2+ content and is widely employed to compare SR Ca2+ content between control and treated cells29,30,31. In our experiments, exposure to Aroclor caused a significant reduction in CPA-induced Ca2+ release under 0Ca2+ conditions (Fig. 5C,D), which hints at a remarkable reduction in SR Ca2+ levels.

Figure 4
figure 4

Acute effect of Aroclor on cardiac beating syncytia. (A) The acute addition of Aroclor (1 µg/mL, black arrow at 400 s) did not cause intracellular Ca2+ mobilization under 0Ca2+ conditions. (B) Mean ± SE of resting [Ca2+]i in cardiac syncytia exposed to either 1 or 2 µg/mL Aroclor for 24 h. *p < 0.05.

Full size image
Figure 5
figure 5

Ca2+ entry and RyRs-dependent Ca2+ release in 1 or 2 µg/ml-exposed cardiac beating syncytia. (A) Representative Ca2+ tracings of the Ca2+ response to caffeine (2.5 mM) in the absence or in the presence of either 1 or 2 µg/mL Aroclor. Extracellular Ca2+ was removed (0Ca2+) to prevent any contaminating effect from Ca2+ entry. (B) Mean ± SE of the percentage of Aroclor-induced inhibition of caffeine-induced Ca2+ transients. (C) Representative Ca2+ tracings of the Ca2+ response to cyclopiazonic acid (10 µM; CPA) in the absence or in the presence of either 1 or 2 µg/mL Aroclor. Extracellular Ca2+ was removed (0Ca2+) to prevent any contaminating effect from Ca2+ entry. (D) Mean ± SE of the percentage of Aroclor-induced inhibition of CPA-induced Ca2+ transients. (E) Representative tracings of the Ca2+ signals induced by sustained depolarization caused by high KCl (High-K) in the extracellular solution in the absence or in the presence of either 1  or 2 µg/mL Aroclor. (F) Mean±SE of the percentage of Aroclor-induced inhibition of High-K-induced Ca2+ transients. *p < 0.05.

Full size image

Finally, we monitored voltage-dependent Ca2+ entry by depolarizing cardiomyocytes with high-KCl solution. We found that high-KCl-induced Ca2+ entry was significantly (p < 0.05) reduced in the presence of each dose of Aroclor (Fig. 5E,F).

Overall, these data strongly suggest that Aroclor reduces the SR Ca2+ content, thereby impairing RyR2-dependent Ca2+ release, decreases the resting [Ca2+]i. and inhibits voltage-dependent Ca2+ inflow in cardiomyocytes. As a consequence, the spontaneous Ca2+ oscillations that drive cardiac contraction are inhibited, which contributes to explain Aroclor-dependent cardiac functional alterations.

Aroclor modifies the expression of genes involved in Ca2+ toolkit

To assess whether Aroclor-induced inhibition of spontaneous Ca2+ oscillations is associated with changes in the expression profile of Ca2+ toolkit genes, we analyzed Atp2a2 (coding for SERCA2A, cardiac isoform of the Ca2+-ATPase that sequesters Ca2+ back into the SR), Ryr2 (cardiac muscle SR RyR isoform), Cacna1c (α subunit of L-type VGCCs) and Itpr2 (cardiac muscle SR InsP3R isoform). The exposure to both Aroclor concentrations did not significantly alter the quantitative expression of Atp2a2 gene (p > 0.05). Conversely, the expression of Ryr2 and Itpr2 significantly increased (1.7- and 1.9-fold, respectively), whereas the expression of Cacna1c significantly decreased (0.7-fold) only in syncytia exposed to 2 µg/ml (Fig. 6).

Figure 6
figure 6

Gene expression analysis of cardiac beating syncytia. Expression profiles of Atp2a2, Ryr2, Itpr2 and Cacna1c. The expression values of CTR samples were set at 1 for the calculation of the n-fold change. Values are expressed as mean ± SD. *p < 0.05; **p < 0.001.

Full size image

Discussion

Our study demonstrates that Aroclor exerts its detrimental effects on the beating properties of perinatal-like cardiomyocytes, acting on the intracellular Ca2+ machinery. Rhythmic beating in foetal and neonatal cardiomyocytes depends on spontaneous events of RyRs-mediated Ca2+ release from the SR, which may in turn evoke small depolarizations that trigger L-type VGCCs-driven action potentials25,26. The spontaneous intracellular Ca2+ oscillations represent the main source of Ca2+ for the contractile machinery during cardiac development26,32. It has long been known that PCB toxicity is consequent to the alteration of the intracellular Ca2+ machinery10,33.

Consistent with these earlier observations, Aroclor heavily impairs spontaneous intracellular Ca2+ oscillations of cardiomyocytes, affecting both their frequency and amplitude. The signalling machinery underlying the rhythmic Ca2+ signal was not affected by the treatment, as, both in the absence and in the presence of Aroclor 1254, the intracellular Ca2+ oscillations were mainly driven by RyRs and supported by extracellular Ca2+ entry. We observed that, both in the absence or in the presence of Aroclor, the repetitive Ca2+ spikes were abruptly inhibited by tetracaine, thereby confirming that RyRs drive the spontaneous Ca2+ oscillations, and the removal of extracellular Ca2+ did not cause a prompt interruption of the Ca2+ spikes, as expected if they were mainly sustained by intracellular Ca2+ mobilization. Also, the pharmacological blockade of InsP3Rs with 2-APB, a powerful and reliable blocker of such receptors34,35, never affected the spontaneous Ca2+ oscillations, thereby ruling out InsP3Rs from the mechanisms that drive the beating activity under our conditions. Of note, the stimulation of InsP3Rs with angiotensin II caused a larger increase in [Ca2+]i in beating syncytia exposed to Aroclor, while CTR cardiomyocytes only showed a modest enlargement of the first Ca2+ spike arising after agonist addition followed by a transient reduction in the amplitude of the following Ca2+ oscillations. These data strongly suggest that: 1) InsP3Rs do not play a prominent role in inducing the spontaneous Ca2+ spikes, as the InsP3-sensitive SR Ca2+ pool is up-regulated by Aroclor exposure, while the amplitude and frequency of Ca2+ oscillations are reduced; and 2) InsP3Rs are, however, able to modulate RyRs-dependent Ca2+ release, as the transient depletion of the InsP3-sensitive SR Ca2+ pool leads to a short-lasting decrease in the amplitude of Ca2+ spikes in CTR cardiomyocytes. The increase in Itpr2 expression could explain the largest Ca2+ signal induced by angiotensin II in the presence of Aroclor. Consistently, 2 μg/ml Aroclor induces a dose-dependent elevation in both Itpr2 transcripts and in the peak Ca2+ response to angiotensin II.

These data indicate that RyRs and VGCCs work to shape the spontaneous Ca2+ oscillations also in Aroclor-exposed syncytia and suggest that the signaling machinery underlying the spontaneous Ca2+ signal was not affected by the treatment. Thus, we moved our attention to single components of the Ca2+ toolkit involved in the oscillatory signal to evaluate whether and how Aroclor interferes with them. We determined that Aroclor exposure induced a significant reduction in basal [Ca2+]i and in SR Ca2+ levels in cardiomyocytes, which was reflected in a dramatic decrease in the caffeine-releasable Ca2+ pool. Additionally, Aroclor heavily inhibited voltage-dependent Ca2+ entry, as measured in response to high-KCl extracellular solution. Therefore, Aroclor affects both the components of the Ca2+ toolkit which deliver Ca2+ to the cytosol during the spiking activity. Similar results were obtained in neuronal cells, in which it was demonstrated that Aroclor 1254 induced a depletion of SR Ca2+ store through the activation of InsP3Rs36,37.

The acute addition of Aroclor did not trigger any detectable increase in [Ca2+]i, suggesting that Aroclor is not able to cause a massive and rapid release of intraluminally stored Ca2+ which is the most intuitive mechanism to explain the depletion of the SR Ca2+ pool underpinning the rhythmic Ca2+ release (as discussed in more detail below). This result differs compared to that observed in rodent neurons38,39,40 and PC12 cells41, in which PCB-95 and Aroclor 1254 stimulate RyRs-dependent intracellular Ca2+ release. It has been demonstrated that PCB congeners sensitize RyRs by binding to one or more components of the RyR macromolecular complex, which could therefore act as receptors10,42,43. We speculate that in perinatal cardiomyocytes, characterized by an immature SR44,45, the lack of expression of components of the adult RyR macromolecular complex may explain why Aroclor fails to increase [Ca2+]i. As Aroclor did not evoke an acute Ca2+ response, we reasoned that effects observed after 24h-exposure could be explained by an interference with the Ca2+ handling machinery.

Real Time PCR analysis demonstrated that both 1 or 2 μg/ml Aroclor induced a significant increase of the expression of Ryr2, as reported also in previous studies on rat cerebellum cells46. The upregulation of Ryr2 transcript expression is predicted to increase, rather than decreasing, the release of Ca2+ from SR to the cytoplasm and thus increases the spiking activity of beating syncytia. However, it has been reported that Aroclor 1254 at 6–9 μM inhibits SR Ca2+ uptake, interfering with SERCA activity47,48, and at 10–13 μM severely compromises the SERCA-mediated uptake of Ca2+ by SR microsomes48. The Aroclor concentrations that we used, i.e. 1 and 2 μg/ml, fall within this range and, therefore, might be able to interfere with SERCA-mediated Ca2+ sequestration. This mechanism could explain the fall in SR Ca2+ levels and, consequently, in the caffeine-releasable Ca2+ pool that we observed in the presence of both 1 and 2 μg/ml Aroclor. Additionally, as SR Ca2+ continuously leaks out in the cytosol to maintain the [Ca2+]i49, this model would also explain the drop in resting Ca2+ levels caused by Aroclor-treated in the beating syncytia. Future investigations are required to confirm this hypothesis and to understand how Aroclor 1254 interacts with SERCA2A. Our data, however, rule out the down-regulation of Atp2a2 expression, that codes for SERCA2A, as the transcript levels of this gene are not altered after 24 h Aroclor exposure. The mechanism whereby Aroclor inhibits SERCA activity is yet to be elucidated47,48. However, in cardiomyocytes, SERCA activity is severely reduced by cytosolic and mitochondria-derived reactive oxygen species (ROS), which therefore lead to a dramatic reduction in SR Ca2+ concentration50,51. Interestingly, a recent work showed that ROS production through the mitochondrial pathway started to increase after 1 h of Aroclor exposure in the human lung cancer cell line A54952. This mechanism could explain why the inhibitory effect of Aroclor is not acute and requires a prolonged incubation also in the beating syncytia.

In addition to reducing SR Ca2+ levels, we found that Aroclor attenuated voltage-dependent Ca2+ entry, which is necessary to maintain the spontaneous Ca2+ oscillations over time, at both 1 and 2 μg/ml. Real Time PCR disclosed that only 2 μg/ml Aroclor 1254 decreased the expression of Cacna1c, which encodes for the α subunit of L-type VGCCs. While this mechanism is likely to play a major role in reducing voltage-dependent Ca2+ entry at the higher dose, it does not explain the inhibitory effect observed at 1 μg/ml. However, earlier reports revealed that PCBs are able to increase Ca2+ permeability across the plasma membrane in several cell types, causing a long-lasting influx of Ca2+ through unidentified Ca2+-permeable channels53 or through store-operated Ca2+ channels54. Also, Aroclor 1254 activates the Ca2+-permeable N-methyl-D-aspartate receptors55, inhibits the non-selective cation channel TRP Vanilloid 6 (TRPV6)56 and stimulates Cl–permeable γ-aminobutyric acid (GABA)A receptors55,57 in newborn rat neocortical neurons. All together, these findings clearly demonstrate that PCB congeners, including Aroclor 1254, have the potential to modulate plasma membrane channels, although it remains to be elucidated whether this interaction is direct or mediated by auxiliary partners. We hypothesize that, besides reducing Cacna1c expression, Aroclor 1254 inhibits L-type VGCCs through a non-yet identified mechanism in cardiomyocytes.

In conclusion, for the first time, we demonstrate that perinatal cardiac syncytia exposed to Aroclor 1254 undergo a dramatic functional alteration of the kinematics contractile properties, due to the disruption of the intracellular Ca2+ machinery. These findings contribute to the understanding of the molecular underpinnings of PCBs-induced cardiovascular alterations, which are emerging as an additional life-threatening hurdle associated to PCB pollution. Therefore, PCBs-dependent alteration of intracellular Ca2+ machinery is the most likely trigger of developmental cardiac functional alterations.

Methods

Cell lines

R1 mESC line (kindly provided by Dr. Nagy from Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada) and STO cell line (ATCC, CRL-2225) were cultivated as previously described58. Briefly, R1 was cultivated in Knockout DMEM added with 15% ESC Qualified FBS, 2 mM L-glutamine, 1X non-essential amino acids, 0.5% penicillin/streptomycin (all from Thermo Fisher Scientific), 0.1 mM β-mercaptoethanol (Sigma) and 500 U/ml ESGRO-LIF (Merck Millipore, Italy). The STO cell line was maintained in DMEM (Sigma) supplemented with 10% foetal bovine serum, 4 mM L-glutamine, 1X non-essential amino acids, 0.5% penicillin–streptomycin solution (all from Thermo Fisher Scientific), 0.1 mM β-mercaptoethanol (Sigma) and 0.2 mg/mL geneticin (Sigma). ESCs were routinely passaged enzymatically every 2/3 days with trypsin/EDTA 0.05%, alternating a passage on STO feeder cells with two passages on gelatin-coated p55 dish, and maintained in an incubator at 37 °C with 5% CO2 in air.

Differentiation of mESCs into perinatal-like cardiomyocytes and Aroclor treatment

mESCs were induced to differentiate through the formation of embryoid bodies (EBs) in vitro by removing the Leukemia Inhibitory Factor (LIF) from culture medium (differentiation medium), using the hanging drop method59,60. For EBs formation, about seventy 20 µl droplets of differentiation medium containing 103 mESCs were plated on the lid of p55 Petri dishes. On day 3 of culture, the developing EBs were transferred on 0.1% agarose-coated tissue dishes (Corning) and from day 5, about 5–8 EBs were plated in single 1.9 cm2 well and cultivated up to 15 days.

Aroclor 1254 (Pancreac Nova Chimica) was dissolved in 100% dimethyl sulfoxide (DMSO; Sigma) to a concentration of 20 mg/ml. This solution was added to the differentiation medium on day 15 to a final concentration of 1 or 2 µg/ml. As control (CTR) samples, cells were exposed to 0.01% DMSO. Cells were harvested 24 h after either Aroclor or DMSO exposure. This procedure was repeated for three independent experiments.

Contraction assay

On day 5 of differentiation, 25 EBs were plated onto 22 mm gelatin-coated Glass Bottom Dish (WillCo Wells) and cultured up to day 15. At day 15, cells were exposed to Aroclor and then transferred into the culture chamber of a Nikon BioStation IM at 37 °C and 5% CO2 for video recording. For each of the three independent experiments, AVI videos of the beating syncytia were recorded from 10 randomly chosen CTR or Aroclor-exposed samples, using the Snagit software and further analyzed with the Video Spot Tracker (VST) program (http://cismm.cs.unc.edu/downloads). Then, videos were processed according to the image processing algorithm based on the Matlab programming language (The MathWorks, Inc., Natick, MA)22,61.

Measurement of Ca2+ transients in mESCs-derived cardiomyocytes

Ca2+ dynamics in mESCs-derived cardiomyocytes were measured by using a conventional epifluorescence Ca2+ imaging system, as shown elsewhere25,26. At day 5 of differentiation, EBs were plated on 12 mm coverslips in 24-multiwell plates. At day 15 beating cardiac syncytia, exposed to DMSO, 1 or 2 µg/ml Aroclor for 24 h, were analyzed for Ca2+ transients. Cells were loaded with 0.5 μM Fura-2/AM (1 mM stock in DMSO; Thermo Fisher Scientific) for 20 min at 37 °C. Then, after washing with pre-warmed Ca2+-containing Tyrode’s solution (NaCl 154 mM, KCl 4 mM, MgCl2 1 mM, CaCl2 2 mM, HEPES 5 mM and Glucose 5.5 mM), the coverslips were fixed to the bottom of a Petri dish and the cells observed by an upright epifluorescence Axiolab microscope (Carl Zeiss, Oberkochen, Germany), equipped with a Zeiss 40× Achroplan objective (water-immersion, 2.0 mm working distance, 0.9 numerical aperture). Voltage-dependent Ca2+ entry was stimulated by replacing 40 mM NaCl with an equimolar amount of KCl (high-KCl solution). Intracellular Ca2+ release was monitored in the medium without Ca2+ (0 Ca2+) and supplemented with 10 mM EGTA. Cells were excited alternately at 340 and 380 nm wavelengths and the emitted light was detected at 510 nm. A first neutral density filter (1 or 0.3 optical density) reduced the overall intensity of the excitation light and a second neutral density filter (optical density = 0.3) was coupled to the 380 nm filter to reach the intensity of the 340 nm wavelength. A round diaphragm was used to increase the contrast. Excitation filters were mounted on a filter wheel (Lambda 10, Sutter Instrument, Novato, CA, USA). Custom software, working in LINUX environment, was used to drive the camera (Extended-ISIS Camera, Photonic Science, Millham, UK), the filter wheel and to measure and plot on-line the fluorescence from 30–45 rectangular “regions of interest” (ROI) enclosing 30–60 areas within each of the beating syncytia. Each ROI was identified by a number. The ratio of fluorescence emitted at 340 and 380 nm was recorded as an indicator of the changes in intracellular Ca2+ concentration ([Ca2+]i). The experiments were performed at room temperature (22 °C). Ratio measurements were performed and plotted every 1.5 s for 900 s. The resting [Ca2+]i was measured by exploiting the Grynkiewicz equation, as shown in Zuccolo et al.30.

RNA extraction, reverse transcription and quantitative Real-Time PCR

On day 16, following 24 h Aroclor exposure, RNA was extracted using the GenElute Mammalian Total RNA Kit according to the manufacturer’s instruction (Sigma) from about 200 EBs from each CTR and 1 or 2 µg/ml Aroclor-exposed samples. Three independent experiments were performed on a total of about 1800 EBs.

Reverse transcription was performed in a final volume of 20 µl reaction mixture with 1 µg of RNA, 1x PCR buffer, 5 mM MgCl2, 4 mM of each dNTP, 0.625 M oligo d(T)16, 1.875 M Random Hexamers, 20 U RNase Inhibitor, 50 U MuLV reverse transcriptase (all from Thermo Fisher Scientific). The conditions for the reverse transcription were as follows: 25 °C for 10 min, 42 °C for 15 min, 99 °C for 5 min. One twentieth of the resulting cDNA was amplified in duplicate by Real-Time PCR in 20 µl reaction mixture with 200 nM of each specific primer (designed using Primer 3 software; see Table 1S) and the MESA GREEN qPCR MasterMix Plus for SYBR assay no ROX sample (Eurogentec) at 1X as final concentration. The amplification reaction, performed in a Rotorgene 6000 (Corbett Life Science), was with the following program: 95 °C for 5 min, followed by 40 cycles at 95 °C for 10 s, 60 °C for 15 s, 72 °C for 20 s. β-2-microglobulin gene expression was used for sample normalization60. The Rotorgene 6000 Series Software 1.7 was used for the comparative concentration analysis.

Statistics

PCR expression data are presented as means ± standard deviation (SD), while the syncytium contractile parameters are expressed as means ± 95% confidence interval for the differences between means. Data were analyzed by the one-way ANOVA and by the post hoc LSD test. As to Ca2+ imaging data, the frequency of spontaneous Ca2+ transients was evaluated by dividing the number of Ca2+ transients arising over the stabilization of intracellular Ca2+ dynamics (i.e. after 1–3 min from the beginning of the recording) by 60 s. The amplitude of intracellular Ca2+ release in response to CPA, which is commonly used to estimate SR Ca2+ content, or caffeine and the amplitude of high-KCl-induced Ca2+ entry was measured as the difference between the ratio at the peak of intracellular Ca2+ mobilization and the mean ratio of 1 min baseline before the peak. Pooled data are given as means ± standard error (SE) and statistical significance (p < 0.05) was evaluated by the Student’s t-test for unpaired observations. Each Ca2+ trace is representative of 80–200 ROIs recorded from at least three embryoid bodies.