Cardiotoxicity of environmental contaminant tributyltin involves myocyte oxidative stress and abnormal Ca²⁺ handling

Abstract

Tributyltin (TBT) is an organotin environmental pollutant widely used as an agricultural and wood biocide and in antifouling paints. Countries began restricting TBT use in the 2000s, but their use continues in some agroindustrial processes. We studied the acute effect of TBT on cardiac function by analyzing myocardial contractility and Ca²⁺ handling. Cardiac contractility was evaluated in isolated papillary muscle and whole heart upon TBT exposure. Isolated ventricular myocytes were used to measure calcium (Ca²⁺) transients, sarcoplasmic reticulum (SR) Ca²⁺ content and SR Ca²⁺ leak (as Ca²⁺ sparks). Reactive oxygen species (ROS), as superoxide anion (O2^•−) was detected at intracellular and mitochondrial myocardium. TBT depressed cardiac contractility and relaxation in papillary muscle and intact whole heart. TBT increased cytosolic, mitochondrial ROS production and decreased mitochondrial membrane potential. In isolated cardiomyocytes TBT decreased both Ca²⁺ transients and SR Ca²⁺ content and increased diastolic SR Ca²⁺ leak. Decay of twitch and caffeine-induced Ca²⁺ transients were slowed by the presence of TBT. Dantrolene prevented and Tiron limited the reduction in SR Ca²⁺ content and transients. The environmental contaminant TBT causes cardiotoxicity within minutes, and may be considered hazardous to the mammalian heart. TBT acutely induced a negative inotropic effect in isolated papillary muscle and whole heart, increased arrhythmogenic SR Ca²⁺ leak leading to reduced SR Ca²⁺ content and reduced Ca²⁺ transients. TBT-induced myocardial ROS production, may destabilize the SR Ca²⁺ release channel RyR2 and reduce SR Ca²⁺ pump activity as key factors in the TBT-induced negative inotropic and lusitropic effects.

1. Introduction

Organotin compounds (OTs), such as tributyltin (TBT), are persistent organometallic environmental contaminants that have been used as components of antifouling paints. TBT has low water solubility and is highly lipophilic. Its high octanol–water partition coefficient (K_OW) defined ([TBT] in octanol)/([TBT] in water) is ~10⁴ (Bangkedphol et al., 2009). This means that TBT has high bioavailability and is readily absorbed across biological lipid membranes (and may also accumulate in lipophilic compartments such as membranes, lipid droplets and fat). This property gives TBT ready access to intracellular systems, including Ca²⁺handling processes that we study here. This also means that even low aqueous concentrations (<1 μM) can result in high local membrane concentrations and cytotoxicity.

Although banned by the International Maritime Organization in 2008, TBT is still used in agriculture and other industrial process, especially in some world regions that are not included in the international convention on the control of harmful anti-fouling systems on ships (AFS Convention) or that have limited environmental monitoring (Kotrikla, 2009; Graceli et al., 2013). The main sources of exposure to OTs for humans are through diet, such as ingestion of seafood and contaminated water exposed to antifouling agents or fungicides (Takahashi et al., 1999). Toxicity levels are related to the concentration, time of exposure, bioavailability and sensitivity of the biota, as well as the persistence of OTs in the environment (de Carvalho and Santelli, 2010).

TBT belongs to the obesogens, an endocrine-disrupting chemical (EDC) subclass that can disrupt sensitive metabolic processes if exposure occurs during early development. Obesogens act as metabolic programmers to increase obesity risk (Grun et al., 2006; Janesick and Blumberg, 2011). Several studies have reported important additional toxic TBT effects, including abnormal sex hormone metabolism in rodents (Kishta et al., 2007; Omura et al., 2001), reproductive dysfunctions, irregular estrous cycling, impairment of rat ovary morphophysiology (Delgado Filho et al., 2011; Graceli et al., 2013; Lang et al., 2012), inhibited mitochondrial function (Aldridge et al., 1977), depressed thymus dependent immunity (Miller et al., 1980) and abnormal heme metabolism (Rosenberg et al., 1982).

OTs also affect the cardiovascular system, leading to vascular reactivity abnormalities (dos Santos et al., 2012; Rodrigues et al., 2014; Ronconi et al., 2018) and endothelial dysfunction (Ronconi et al., 2018). We previously demonstrated that even low TBT exposure (100 and 500 ng/kg/bw/day for 15 days) increased reactive oxygen species (ROS) and induced endothelial dysfunction in aorta and mesenteric vessels of female rats (Ribeiro Junior et al., 2016; Rodrigues et al., 2014; Ximenes et al., 2017). OTs appear to inhibit oligomycin-sensitive mitochondrial ATPase/synthase (Mehrotra et al., 1985) the sarcoplasmic reticulum Ca-ATPase (SERCA) (Kodavanti et al., 1991) and plasma membrane Na⁺/K⁺ ATPase (Cameron et al., 1991). These studies indicate that triorganotin compounds including TBT may affect the Ca²⁺ handling by altering the function of key cardiac ion transporters, excitation-contraction coupling and mitochondrial function.

Myocardial SERCA is responsible for the re-uptake of cytosolic Ca²⁺ that is required for cardiac relaxation between beats (diastole). During the cardiac action potential, Ca²⁺ enters the cell via the L-type channel and induces further Ca²⁺ release from the sarcoplasmic reticulum (SR) via ryanodine receptors (RyR2). That Ca²⁺-induced Ca²⁺-release amplifies the rise in intracellular [Ca²⁺]_i that activates systolic cardiac contraction (Bers, 2002). Synchronized SR Ca²⁺ release and re-uptake are fundamental for normal systolic and diastolic functioning of cardiomyocytes.

TBT induces reactive oxygen species (ROS) generation and oxidative damage in various systems such as vascular, neural, reproductive, immune, renal, in both in vivo and in vitro (Ishihara et al., 2012; Katika et al., 2011; Kato et al., 2013; Isomura et al., 2013; Mitra et al., 2015; Nishimura et al., 2015; Rodrigues et al., 2014). ROS can also alter the function of key signaling proteins, e.g. Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), protein kinase A (PKA), protein kinase C (PKC) and RyR2 (Oda et al., 2015; Wagner et al., 2011). ROS may be harmful when produced in excess due in part to altered ion transport balance (Na⁺ and Ca²⁺) which alters Ca²⁺ handling and promotes both contractile dysfunction and arrhythmias (Kohler et al., 2014).

The present study aims to analyze acute TBT toxicity to myocardial contractility and intracellular Ca²⁺ handling. The hypothesis of this study is that TBT induces cardiotoxicity involving excitation-contraction coupling mechanisms and ROS generation. In order to test this hypothesis we examined the inotropic and lusitropic effects of TBT and its consequence on SR Ca²⁺ leak (assessed as Ca²⁺ sparks), SR Ca²⁺ content in rat ventricular myocytes and the ROS from cytosol and mitochondrial. The myocyte studies support the multicellular data, demonstrating that acute exposure to the environmental contaminant TBT can cause cardiotoxicity. The TBT effects on Ca²⁺ regulatory proteins increase diastolic Ca²⁺ by increasing Ca²⁺ leak, leading to reduced SR Ca²⁺ content and reduced systolic Ca²⁺ transients and contraction.

2. Methods

2.1. Evaluation of the acute effect of TBT on isometric force of papillary muscle

Female Wistar rats (200–250 g) were supplied by the Central Animal Facility of the Federal University of Espírito Santo (Universidade Federal do Espirito Santo, UFES), Espirito Santo, Brazil and were kept in cages with free access to water and food, under temperature control and a light-dark cycle of 12 h. Females were chosen because TBT causes abnormal female sexual development (Grote et al. 2004). All procedures were approved by the UFES Committee for Animal Experiments (CEUA number 27/2016) or NIH Guide for the Care and Use of Laboratory Animals and were approved by Institutional Animal Care and Use Committee at the University of California Davis.

Animals were anesthetized with intraperitoneal injection of ketamine (40 mg/kg, IP) and xylazine (8 mg/kg, IP) and then submitted to thoracotomy. The heart was removed and retrograde perfused through the aorta with nutrient solution to allow adequate dissection of the anterior and posterior papillary muscles from left ventricle (LV). Papillary muscles were attached to rings and connected to an isometric force transducer (TSD125 - Biopac Systems, Inc.) in a water-jacketed bath, maintained at 30 ± 0.5 °C and containing 20 mL of Krebs-Henseleit solution (KHS): 130 mM NaCl, 1.2 mM Cacl₂, 5.4 mM KCl, 2.5 mM MgSO₄, 2 mM NaH₂ PO₄, 18 mM NaHCO₃, 1.2 mM Na₂SO₄ and 11 mM glucose, gassed with 5% O₂ and 95% CO₂, with pH verified as 7.38 ± 0.02. Field stimulation was provided by isolated rectangular pulses (20 V, 10 ms duration, 0.5 Hz). After gradual fiber stretch to achieve L_max (the length at which maximum developed force was attained) and stabilized, isometric force was measured and 1 nM to 10 mM TBT was added to the superfusate, in cumulative fashion, waiting 3 min between increasing concentration and correcting for chamber volume change. TBT was diluted > 100-fold from a stock in 0.4% ethanol.

2.2. Evaluation of cardiac parameters in Langendorff perfused heart system

We used the Langendorff isolated heart perfusion technique to monitor intraventricular pressure during cardiac contraction. The aorta is cannulated and perfused in a retrograde manner to drive coronary artery perfusion throughout the heart. This allows measurement of both coronary perfusion pressure and left ventricular (LV) pressure during constant coronary flow controlled by a peristaltic pump (10 mL/min). Changes in coronary resistance are detect as changes in coronary perfusion pressure (CPP, mmHg) measured by a pressure transducer (TSD 104A-Biopac connected to a Funbec MP-100 preamplifier) connected to the perfusion system. For LV pressure, the left atrium was opened and a water inflatable latex balloon connected to a pressure transducer (Biopac TSD 104A-Biopac connected to a preamplifier) was introduced into the LV. This also allowed control of preload by changing end diastolic pressure (DP, mmHg). Left ventricular isovolumetric systolic pressure (LVISP, mmHg) was measured via a data acquisition system (MP 100, Biopac Systems: Inc; CA, including Biopac Student Lab software), using a sampling rate of 2000 samples/s. Hearts were randomly grouped as Control (perfused with normal KHS) and TBT (exposed 5 min to 50 μM TBT). TBT was diluted in ethanol 0.4% plus 99.6% saline.

2.3. Ventricular function in isolated perfused heart

To analyze heterometric (Frank-Starling) effects, ventricular function curves were obtained by measuring LVISP while the DP was increased from 0 to 30 mmHg in 5 mmHg increments. Frank-Starling relationships were obtained at different perfusate [Ca²⁺] and upon β-adrenergic stimulation. To evaluate inotropic (homeometric) responses to [Ca²⁺]_o, DP was maintained constant at 10 mmHg and LVISP was recorded during heart perfusion with solutions with [Ca²⁺] at 0.62; 1.25; 1.87; 2.5 and 3.12 mM. The inotropic response to β-adrenergic activation was analyzed at DP 10 mmHg for a single bolus of 100 μL of 100 μM isoproterenol hydrochloride. Heterometric and homeometric mechanisms were analyzed in the presence and absence of antioxidants: 500 μM Tiron (a superoxide anion (O₂ ^•−) scavenger; 100 μM Tempol (4-hydroxy-2,2,6,6-tetramethyl-piperidine-N-oxyl, a superoxide dismutase (SOD) mimetic; 30 μM apocynin (NADPH oxidase inhibitor) and also 10 μM losartan (angiotensin II AT1 receptor blocker).

2.4. Myocardial oxidative stress assay (ROS) by dihydroethidium fluorescence (DHE)

The in vitro effects of TBT on myocardial ROS production were measured in the presence of 500 μM Tiron, 100 μM Tempol, 30 μM apocynin and 10 μM losartan. Oxidative fluorescence of dihydroethidium (DHE) was used to evaluate the in situ production of O₂ ^•−as previously described (Fiorim et al., 2012, Ribeiro et al., 2013). Following the isolated heart procedures, left ventricular (LV) free wall samples approximately 5 mm in length were immersed in Krebs-HEPES buffer solution with 30% sucrose for 1 h, then embedded in cryostat inclusion medium and maintained at −80 °C until cryostat sectioning (10 μm thick), with sections placed on gelatinized slides and frozen. On the analysis day slides were warmed 1 h at 37°C to remove the inclusion medium and then incubated with Krebs-HEPES buffer, for 30 min at 37 °C, dried and incubated with DHE-containing Krebs-HEPES buffer (2 μM) in a chamber protected from light for 30 min. Sections were assembled with mounting medium (Erv-Mount, Easy Path) and coverslip. Heart sections were imaged with an inverted fluorescence microscope (Leica DM 2500) using a 568 nm emission filter at 100 × magnification and photographed with the microscope coupled camera (Leica DFC 310 FX). The mean fluorescence density was calculated using ImageJ.

2.5. Cytosolic, mitochondrial ROS production and mitochondrial membrane potential

Cells were incubated with 5 μM MitoSOX Red (Molecular Probes, Eugene, OR, USA) for 20 min at 37 °C, with 1 μM CM-H₂DCFDA (Molecular Probes, Eugene, OR, USA) for 5 min or 100 nM TMRM (Molecular Probes, Eugene, OR, USA) for 30 min and were subsequently washed with an extracellular solution (Tyrode) that contained 1.2 mM Ca²⁺ to remove the excess of dye. The images were recorded with confocal line-scan imaging using a Zeiss LSM Pascal microscope, with a 40X water-immersive objective. Two-dimensional imaging mode was used, with excitation at 488 and emission at > 517 nm, for CM-H₂DCFDA, alternating with excitation at 532 nm, and emission at > 560 nm for TMRM. Time-lapse x,y images were acquired at 1024 bit resolution and at the sampling rate of 507 ms per frame.

2.6. Cardiomyocyte isolation

Ventricular myocytes were isolated as previously described (Uchinoumi et al., 2016). Briefly, excised hearts were retrograde perfused with collagenase-containing buffer. After myocyte dissociation [Ca²⁺] was gradually increased from 20 μM to 1 μM over 1 h, in otherwise normal Tyrode's solution (below). Isolated rat ventricular myocytes were transferred to laminin-coated disposable glass coverslips.

2.7. Ca²⁺ transients, SR Ca²⁺ content and Ca²⁺ spark measurements

Intact cardiomyocytes were loaded with 5–10 μM Fluo-4 AM (Molecular Probes, Eugene, OR) for 20–24 min, then perfused with a standard Tyrode's solution containing (in mM) 140 NaCl, 4 KCl, 0–1.2 CaCl₂, 1 MgCl₂, 5.5 glucose and 10 HEPES, pH 7.4. The indicator was excited at 488 nm (argon laser) with fluorescence measured at >500 nm, using a Biorad Radiance 2100 confocal system. Longitudinal line scans (166 pixels per line) were acquired at 6 ms intervals. Myocytes were electrically stimulated at 0.5 Hz, with just suprathreshold field intensity and pulse duration 2 ms using a pair of platinum electrodes. To determine SR Ca²⁺ content, electrical stimulation was ceased and the cell was rapidly super-fused with Tyrode's solution with 10 mM caffeine included (after 5 min exposure to TBT at either 10 or 100 nM). Ca²⁺ transients are reported as fluorescence intensity ratio (ΔF/F₀), where F₀ is the resting fluorescence of the cell. The amplitude of caffeine-induced Ca transients was taken as an index of SR Ca²⁺ content. In some experiments 10 μM dantrolene, an inhibitor of Ca²⁺ efflux through RyR, or 10 mM Tiron, a cell permeable superoxide scavenger, was included.

Ca²⁺ spark frequency (CaSpF) was measured in transverse line-scan mode (2 ms/line) before and after exposure to TBT 10 and 100 nM (in sparks.μm⁻¹.s⁻¹). Spark amplitude, width and duration (as full width or duration at half-maximum) were also measured using the SparkMaster plug-in with ImageJ (Picht et al., 2017) with visual confirmation.

2.8. Permeabilized cardiomyocytes

Cardiomyocytes were suspended in internal solution containing (in mM) 120 K⁺-aspartate, 0.5 EGTA, 1.0 MgCl₂, 5.0 Na₂-ATP, 10.0 glutathione reductase, 10 creatine phosphokinase, CaCl₂ to achieve free Ca²⁺ 10 nM, 5 U/mL creatine phosphokinase, 8% dextran (40,000 Da), 5.0 phosphocreatine di-Na⁺ and 10 HEPES (pH 7.2), and placed in the disposable glass chamber for 15 min. The cell membrane was then permeabilized by the addition of saponin 50 μg/mL (quillaja bark purified,-Sigma), with free [Ca²⁺] buffered at 10 nM. After 30 s, the bath solution was exchanged for an internal solution without saponin and loaded with Fluo-4, 5 K⁺ 10 μM, with free [Ca²⁺] increased to 50 nM Ca²⁺sparks were recorded with the above mentioned Biorad confocal microscope, using a 40 × oil immersion objective. Fluo-4 was excited at 488 nm and fluorescence was measured at >515 nm. Line scans were made, along the longitudinal axis of the cell, in a central focal plane (scanning speed 166 lines/sec), before and 5 min after TBT exposition. All experiments were performed at room temperature of 25 °C.

2.9. Data analysis

Data are reported as the mean ± SEM values. T-test, One or Two-Way ANOVA test were used to analyze differences between the different groups and followed by Tukey or Bonferroni post-hoc test. p < 0.05 was considered significant.

3. Results

3.1. Acute effect of TBT on rat isolated papillary muscle

The effect of increasing TBT concentrations (1 nM–10 mM) on the developed force in papillary muscle is shown in Fig. 1A. There was a gradual negative inotropic response after sequential 3 min exposure to TBT increases up to 100 μM, and a precipitous decline at [TBT] = 1 mM and 10 mM (to 26% and 1% of the initial force, respectively; *p < 0.05, n = 6).

Fig. 1.

Isolated papillary muscle: A) Inotropic response to TBT (1 nM–10 mM; 3 min exposure) on developed force (g/mg) (n = 6). Langendorff perfused heart: B) Change in the left ventricle isovolumic systolic pressure (LVISP, mmHg), C) First positive derivative of pressure (dP/dt+, mmHg/s), D) First negative derivative of pressure (dP/dt−, mmHg/s) during increments of extracellular Ca²⁺ from 0.62 to 3.12mM in the Langendorff perfused hearts, Control group (Ctl, n = 7 rats),TBT (50 μM, n = 8). E) Representative record of LVISP during increment of extracellular Ca²⁺ from 0.62 to 3.12 mM for Ctl and TBT groups during increment of extracellular Ca²⁺ from 0.62 to 3.12 mM. F) Diastolic pressure (DP, mmHg) during increment of Ca²⁺ in the Ctl and TBT group (TBT 50 μM, n = 8 rats). Results are reported as mean ± SEM. *p < 0.05 TBT vs Ctl, two-way ANOVA plus Tukey post hoc.

3.2. Acute effect of 50 μM TBT on intact isolated hearts

We also measured LV function in Langendorff-perfused hearts upon 5 min exposure to 50 μM TBT at different [Ca²⁺]_o (0.62–3.12 mM) to test whether TBT inhibited Ca²⁺-dependent (homeometric) cardiac function. Fig. 1B and E shows that, in the presence of TBT,LVISP was strongly depressed at normal [Ca²⁺]_o (1.25 mM: Ctl = 115.1 ± 9.2 vs TBT = 66.6 ± 4* mmHg, n = 7–8, *p < 0.05) and at all higher [Ca²⁺]_o (1.87, 2.5 and 3.12 mM). Fig. 1C and D shows similar reductions in both maximal LVISP rise (dP/dt+; Ca²⁺ 1.25 mM: Ctl = 3072 ± 184 vs TBT = 1990 ± 227* mmHg/s, n = 7–8, *p < 0.05) and relaxation (dP/dt−; Ca²⁺ 1.87 μM: Ctl = 2530 ± 235 vs TBT = 1792 ± 253* mmHg/s, n = 7–8, *p < 0.05). Fig. 1F demonstrates that DP was also elevated by 50 μM TBT (e.g. at 3.12 mM [Ca²⁺]_o: Ctl = 8.7 ± 0.26 vs TBT = 14.3 ± 3.03 mmHg, n = 6–7,*p < 0.05).

3.3. Myocardial oxidative stress

We examined whether ROS was involved in the negative inotropic effect of TBT at different [Ca²⁺]_o in hearts perfused with antioxidants: Tiron, Tempol and apocynin and also with the angiotensin II AT1 receptor blocker losartan, a pathway known to induce ROS production. Fig. 2A-D shows that only Tiron, an antioxidant with metal-chelating properties, partially improved the contractility ([Ca²⁺]_o = 1.25 mM; Ctl = 115 ± 9.3 vs TBT = 66.6 ± 4.1 mmHg *, TBT + Tiron = 86.4 ± 5.1 mmHg^#; *p < 0.05 vs Ctl or ^#TBT). Neither Tempol, apocynin nor losartan had any remedial effect against TBT. Cardiac O₂ ^•− in situ production is shown in Fig. 2E. There was a significant increase in O₂ ^•− production in the hearts perfused with TBT, and that was prevented by all of the antioxidants and losartan. These results indicate that the TBT-induced negative inotropy is largely independent of the increased ROS production.

Fig. 2.

Effects of antioxidants during increment of extracellular Ca²⁺ from 0.62 to 3.12 mM in the Langendorff perfused hearts. A) Groups Ctl, TBT (50 μM) and TBT + Tiron (500 μM); B) Groups Ctl, TBT and TBT + Tempol (100 μM); C) Groups Ctl, TBT and TBT + Apocynin (30 μM) and D) Ctl, TBT and TBT + Losartan (10 μM). E) In vitro detection of superoxide anion (O2•⁻) after TBT (50 μM) incubation in the presence and absence of antioxidants. Heart fluorescence micrographs were stained with DHE (red fluorescence) sensitive dye for the Ctl (n = 4 rats), TBT (n = 4 rats), TBT + Tiron (500 μM) (n = 4 rats) groups, TBT + Tempol (100 μM) (n = 4 rats) and TBT + Losartan (10 μM) (n = 4 rats). Results are reported as mean ± SEM. *p < 0.05 vs Ctl, #p < 0.05 vs TBT, two way ANOVA plus Tukey post hoc.

Fig. 3 tests whether TBT (or its induced ROS) influences the Frank-Starling curves as LV preload was increased from 0 to 30 mmHg at low (0.62 mM) and normal [Ca²⁺]_o (1.25 mM). TBT perfusion (50 μM) significantly reduced LVISP (Fig. 3A and B), but when the curves were normalized (to LVISP at 30 mmHg DP as 100%), the preload-dependence of LVISP was unaltered by TBT (CD). This indicates that the acute TBT effect is not due to a loss of length-dependent activation, such that the heterometric Frank-Starling is preserved.

Fig. 3.

Left ventricular isovolumetric systolic pressure (LVISP, mmHg) and diastolic pressure (DP, mmHg) in 0,62 mM Ca²⁺ (A, C, E) and 1,25 mM Ca²⁺ (B, D, F). A and B represent absolute increment on LVISP (mmHg). C and D represent % increment in the LVISP considering its value at 30 mmHg DP as 100%. E and F show the increment on LVISP in the groups control (Ctl), TBT (50 μM) and TBT + Tiron (500 μM); B: Groups Ctl, TBT and TBT + Tempol (100 μM); C: Groups Ctl, TBT and TBT + Apocynin (30 μM) and D: Ctl, TBT and TBT + Losartan groups (10 μM). Results are reported as mean ± SEM. *p < 0.05 TBT vs Ctl, #p < 0.05 TBT + Tempol vs TBT, two way ANOVA; n = 7–8 per group.

Perfusion with either antioxidants or AT1 receptor block had no influence on TBT's suppression of contraction at normal [Ca²⁺]_o (1.25 mM) or high preload (Fig. 3F). However, at lower preloads there was some apparent mitigation of the negative effect of TBT on LVISP (especially for Tiron). At reduced [Ca²⁺]_o (0.62 mM; Fig. 3E) the negative inotropic effect of TBT was more consistently mitigated by all three antioxidants and losartan. Thus, at low [Ca²⁺]_o, the protective effect of antioxidants and losartan was most evident at low cardiac preload. However, at normal [Ca²⁺]_o and higher preload levels, ROS suppression may not suffice to protect myocardial contractile strength.

3.4. β-adrenergic response

The β-adrenergic receptor agonist isoproterenol (ISO) was used to test whether acute TBT exposure limited the physiological inotropic and lusitropic responses to ISO, at [Ca²⁺]_o of 0.62 mM, where baseline contractility was least altered by TBT (Fig. 4). We added a 100 μL bolus of 100 μM ISO to the perfusate and monitored peak functional response during the ensuing 1–5 min. While control hearts showed robust ISO-induced increases in LVISP, it was strongly suppressed by TBT pretreatment (Fig. 4A, D and 4E). The change in LVISP for Control and TBT were 133 ± 9 and 27.1 ± 4.6 mmHg, respectively (Fig. 4A). In addition, the activation time (AT) and the relaxation time 90% (RT 90%) after ISO were longer in hearts exposed to TBT (Fig. 4B, C, 4E). Neither antioxidants nor losartan prevented the TBT-induced suppression of β-adrenergic-induced inotropic effect (Fig. 4D), although some benefit accrued, again, especially for Tiron.

Fig. 4.

Effect of 100 μM Isoproterenol (ISO) in the Langendorff isolated perfused heart: A) Change in the left ventricular isovolumetric systolic pressure (LVISP, mmHg) before and after ISO; B) AT: Activation Time (ms); C) RT: Relaxation Time 90% (ms); D) LVISP in the control group (Ctl) and TBT perfusion, before and after ISO, in the presence and absence of antioxidants: Tiron, Tempol, Apocynin and angiotensin II blocker, Losartan; E) Representative record of LVISP in the Ctl and TBT 50 μM before and after 100 μM ISO, showing also the way activation and relaxation time (AT and RT) were measured (*p < 0.05 vs Ctl; #p < 0.05 TBT + ISO vs TBT + ISO + Tiron); n = 7–8 per group.

3.5. Myocyte Ca²⁺ transients, SR Ca²⁺ content and Ca²⁺ sparks

To test whether myocyte Ca²⁺ handling is involved in the negative inotropic effects of TBT, we measured Ca²⁺ transients in isolated ventricular myocytes that were electrically stimulated at 0.5 Hz (Fig. 5). TBT (100 nM) depressed Ca²⁺ transient peaks (Fig. 5A and B), and the rates of twitch [Ca²⁺]_i decline, indicated by longer time constants (τ) (Ctl = 0.406 ± 0.0445 s, n = 20 vs TBT 100 nM = 0.601 ± 0.057 s n = 17, p < 0.01). Higher TBT concentrations tended to cause myocyte death. SR Ca²⁺ content was evaluated by rapid application of 10 mM caffeine to intact ventricular myocytes (Fig. 5C). Cardiomyocytes exposed to 100 nM TBT had significantly reduced SR Ca²⁺ content compared to the control group. The rate of caffeine-induced [Ca²⁺]_i decline, indicative of Na/Ca exchange dysfunction were also slowed by TBT (τ for Ctl = 3.038 ± 0.385 s, n = 13 vs TBT 100 nM = 4.18 ± 0.34 s, n = 14, p < 0.03).

Fig. 5.

Ca²⁺ transient (CaT) amplitude during 0.5 Hz pacing and SR Ca²⁺ content (10mM Caffeine) in rat intact cardiomyocytes of Control (Ctl), TBT groups in the presence and absence of dantrolene 10 μM (DAN) and 10 mM Tiron. A) Representative confocal images of caffeine-induced CaT and SR Ca²⁺ content in the absence (Ctl, Ctl + DAN and Ctl + Tiron) and presence of 100 nM (TBT, TBT + DAN and TBT + Tiron). B) Peak of Ca²⁺ Transient (ΔF/F₀) and C) Caffeine-iInduced SR Ca²⁺ release (ΔF/F₀). Dantrolene and Tiron suppressed the TBT effect on Ca²⁺ transient and caffeine-induced SR Ca²⁺ release. The fluorescence data were normalized as ΔF/F₀. Results are reported as mean ± SEM. *p < 0.05 TBT vs Ctl; &TBT vs TBT + DAN, n = number of cells.

To further test whether there might be enhanced pathological RyR leak (which could cause the observed lower SR Ca²⁺ content), we co-treated isolated cardiomyocytes with 100 nM TBT and 10 μM dantrolene (diastolic RyR channel stabilizer) and the antioxidant Tiron. Fig. 5 shows that both dantrolene and Tiron restored normal SR Ca²⁺ content and twitch Ca²⁺ transient amplitude. Thus, while 100 nM TBT may partially inhibit SERCA2a function (based on twitch [Ca²⁺]_i decline), the SR Ca leak via RyR is sufficient to explain the reduced SR Ca²⁺ content and twitch Ca²⁺ transients.

We also measured SR Ca²⁺ leak more directly as Ca²⁺ sparks. Ca²⁺ sparks measured in saponin-permeabilized cardiomyocytes are shown in Fig. 6. CaSpF was significantly higher in permeabilized myocytes exposure to 100 nM TBT (Time 5 min: Ctl = 9.7 ± 1.6 vs TBT 16.02 ± 1.3 μm/s). In addition, CaSpF amplitude decreased progressively after 5 min of 100 nM TBT exposure (Ctl = 0.568 ± 0.068 vs TBT 0.342 ± 0.059). This would be consistent with a TBT-induced increase in SR Ca²⁺ leak that reduces SR Ca²⁺ content available for release, explaining smaller twitch Ca²⁺ transients and Ca²⁺ spark amplitude. Dantrolene prevented the TBT-induced Ca²⁺ spark effects on frequency (Ctl = 14.5 ± 0.54 vs TBT 13.6 ± 1.65/sec) and amplitude (Ctl = 0.463 ± 0.017 vs TBT 0.432 ± 0.05). The fractional contribution of SR and NCX/Slow flux were similar in the twitch and caffeine decay in the presence of TBT (data not shown).

Fig. 6.

Ca²⁺ Sparks analysis in rat permeabilized cardiomyocytes at 0, 3, 5 and 10 min in control condition (Ctl) and in the presence of 10 nM and 100nM TBT: A) Calcium Sparks Frequency (CaSF, Sparks/100 μm/s) and B) Calcium Sparks Amplitude (CaS Amplitude, ΔF/F₀), n = 5–6 cells per group; C) Representative records of calcium sparks 10 min after, control condition (Ctl), 10 nM and 100 nM TBT; D) Calcium Sparks Amplitude and E) Calcium Sparks Frequency in the presence of 10 μM Dantrolene after 5 min TBT exposition. Results are reported as mean ± SEM. *p < 0.05 vs Ctl, two way ANOVA, n = number of cells.

3.6. TBT increases cytosolic, mitochondrial ROS production and decreases mitochondrial membrane potential

Since redox homeostasis is determined by the balance between the pro-oxidant and antioxidant, attention must also be given to the mechanisms responsible for intracellular ROS production. We examined cytosolic ROS production by incubating the cells with CM-H₂DCFDA. As shown in Fig. 7A and B, 100 nM TBT increased cytosolic ROS production after 5 min. TBT also decreased mitochondrial membrane potential as shown in Fig. 7C and D and increased mitochondrial O2^•−production after 5 min (Fig. 7E and F). Note that the smaller number of cells in the TBT groups reflects in part poorer overall survival of these cells, such that the TBT-induced mitochondrial dysfunction may be underestimated here.

Fig. 7.

Cytosolic (DCF), mitochondrial (MitoSOX) ROS production and mitochondrial membrane potential (TMRM) in the control Ctl and TBT groups treated or not with 10 mM Tiron. Cells were incubated for 5 min with: A and B) 1 μM DCF; C and D) 100 nM TMRM and E and F) 5 μM MitoSOX Red, The images were recorded with confocal line-scan imaging using a Zeiss LSM Pascal microscope, with a 40X water-immersive objective. *p < 0.05 vs Ctl.; #p < 0.05 vsTBT, n = number of cells.

4. Discussion

In this study, we investigated the acute effects of TBT on the heterometric and homeometric mechanisms of cardiac contraction using isolated hearts, papillary muscle and isolated ventricular myocytes from rats. The main result indicates that the acute exposure to TBT is responsible for an intense negative inotropic and lusitropic effect involving impairment SR Ca²⁺ handling.

The heterometric Frank-Starling mechanism is the increase in contractility in response to increasing preload, and it is due both to enhanced myofilament overlap and enhanced myofilament Ca²⁺sensitivity (ter Keurs, 2012). We analyzed the TBT effects on intact heart length-dependent activation during normalized length-tension curves, by increasing DP and measuring LVISP (Stefanon et al., 1990; Stefanon et al., 1994). TBT induced an acute negative inotropic effect for all DP (Fig. 1B). However, the relative increment on LVISP at both [Ca²⁺]_o was preserved. We conclude that TBT did not affect the intrinsic Frank-Starling mechanism. While this does not rule out myofilament effects of TBT, we also assessed whether homeometric modulation of cardiac contraction and relaxation were altered (using β-adrenergic activation and altered [Ca²⁺]_o).

The inotropic responses to both increasing [Ca²⁺]_o and isoproterenol were reduced by TBT compared to the control group in isolated hearts (Figs. 3 and 4). These results suggest that the reduction of contractility might involve alterations of intracellular Ca²⁺. OTs are known to affect Ca²⁺ transport in diverse cell models (Gennari et al., 2000; Isomura et al., 2013) and also inhibit adenylate cyclase in rat brain, thereby reducing cAMP levels (Leow et al., 1979). Therefore, changes in β-adrenergic responses found in the present study may also be due to decreased cAMP levels or effects on Ca²⁺ handling in the SR. TBT also delayed the kinetics of cardiac muscle contraction (Fig. 4D and E). Three independent variables determine myocardial contractility: the number of active cross-bridges, the rate of the cross-bridge cycle, and the time courses of the onset and offset of cross-bridge interactions. In addition, excitation contraction mechanisms, involving calcium handling, determine the amplitude and kinetics of contraction and relaxation. Thus, changes in the time course of Ca²⁺ delivery and removal can modify the ability of cardiac muscle to do work, and so, control myocardial contractility. It reflects the myocardial fiber shortening and relaxation velocity involving the events of excitation-contraction coupling and contractility mechanisms. Activation time is the time from the onset of contraction to peak tension. It depends on the time to peak of [Ca²⁺]_i, mainly during the phase 2 of action potential and activation kinetics of cross-bridge cycling. The relaxation time is the time from peak tension to the point when tension returns to baseline. During relaxation time, Ca²⁺ is being pumped back into the SR, and the muscle is stretching back to its original length. In the present study, TBT slowed both the activation and relaxation time isoproterenol-activated, demonstrating its cardiotoxicity on the contractility.

Impaired Ca²⁺ reuptake can elevate diastolic [Ca²⁺]_i and affecting diastolic function and pressure. Indeed, in our study, left ventricle DP increased in the presence of TBT, consistent with this idea.

Kodavanti et al. (1991) found inhibition of cardiac SERCA function by TBT (IC₅₀ = 2 μM), and other OTs (triethyltin and trime-thyltin) which would limit Ca²⁺ pumping into the SR, slow relaxation and reduce SR Ca²⁺ content, and could explain the negative inotropic and lusitropic effects of TBT. Consistent with that biochemical data, we found that in intact isolated cardiomyocytes, TBT decreased the amplitude of twitch Ca²⁺ transients, the content of Ca²⁺ in the SR and the apparent rate of SR Ca²⁺ uptake (slower rate of [Ca²⁺]_i decline). However, TBT also increased CaSpF in permeabilized cardiomyocytes, indicating increased SR Ca leak, which could also explain the TBT-induced reduction of SR Ca²⁺ content (Bers, 2014). We used dantrolene to inhibit Ca²⁺ efflux via RyR, to test whether the TBT-induced increase in SR Ca²⁺ leak was sufficient to explain the reduction in systolic Ca²⁺ transients and the negative inotropic effect of acute TBT exposure in intact hearts. In fact, dantrolene inhibited the TBT-induced SR Ca²⁺ leak and restored normal twitch Ca²⁺ transients (Figs. 5B and 6D). Thus, while TBT inhibited SERCA function by ~50% (based on prolongation in t of [Ca²⁺]_i decline), the increased SR Ca²⁺ leak may be the major cause of TBT's effect to suppress Ca²⁺ transients and contraction. Such diastolic SR Ca²⁺ leak can also be arrhythmogenic (Bers, 2014) and limit Ca²⁺ transient amplitude (Vassalle and Lin, 2004). We and others have also shown that dantrolene does not inhibit RyR2 in normal physiological conditions, but that dantrolene can restore stable diastolic RyR2 closure during diastole (Oda et al. 2015; Uchinoumi et al. 2016; Maxwell et al., 2012). Indeed, our hypothesis is that TBT is inducing a pathological RyR2 conformation that is partially restored by dantrolene.

The rise in left ventricular DP and diastolic dysfunction during TBT treatment may also be explained by the combination of leaky RyR and reduced SERCA function. In addition, TBT slowed caffeine-induced Ca²⁺ transient decline by the 33%, indicative of a weaker Na/Ca exchange to extrude Ca²⁺ from the myocyte (Bers, 2002). This could be due to either depression of the Na/Ca exchanger function, or elevated [Na⁺]_i (secondary to Na/K-ATPase inhibition (Cameron et al., 1991)). In either case, this could also contribute to TBT-induced diastolic dysfunction. A disadvantage of high diastolic Ca²⁺ is the high energy cost to the myocyte and TBT is thought to inhibit mitochondrial ATP production (Matsuno-Yagi and Hatefi, 1993; von Ballmoss et al., 2004).

The mechanism by which TBT acts on RyR2 needs further investigation. RyR2 oxidation can increase SR Ca²⁺ leak and reduce SR Ca²⁺ content, and Ca²⁺ transients (Kohler et al., 2014). Uchinoumi et al. (2016) also showed that H₂O₂-induced RyR2 oxidation reduces the RyR2 calmodulin affinity and leads to a leaky RyR2 conformation. Since TBT promoted ROS production this seems plausible, but ROS scavengers did not prevent the TBT-induced contractile dysfunction or reduced Ca²⁺ transients (Figs. 2 and 5). Thus, TBT might have some other direct or indirect effect to cause SR Ca²⁺ leak.

Other studies have also shown that exposure to TBT increases ROS generation with subsequent oxidative damage in various systems such as vascular, neural, reproductive, immune, T-lymphocyte cells and neuroblastoma in in vivo and in vitro models (Katika et al., 2011; Ishihara et al., 2012; Kato et al., 2013; Isomura et al., 2013; Mitra et al., 2015; Nishimura et al., 2015; Rodrigues et al., 2014). Our data are consistent with these findings showing an increase in O₂^•− production in hearts perfused with 50 μM TBT as compared to the control group. The increment on O₂^•− production was prevented by the antioxidants Tiron, losartan and Tempol. But only Tiron had partial effects to limit contractile dysfunction at [Ca²⁺]_o above 0.62 mM (Fig. 2). On the other hand, the negative inotropic effect of TBT at low pre-load and low [Ca²⁺]_o was ameliorated by all antioxidants (Fig. 3). In fact, hearts perfused with either antioxidants or AT1 receptor block, losartan, were not protected against TBT's suppression of contraction at normal [Ca²⁺]_o (1.25 mM) or high preload. Thus, at low [Ca²⁺]_o, the protective effect of antioxidants and losartan was most evident at low cardiac preload. The present results demonstrated that, at normal [Ca²⁺]_o and higher preload, ROS suppression may not be enough to protect myocardial contractile against TBT cardiotoxicity. It suggests that TBT might have many different targets in the cardiomyocyte involving its deleterious effect.

These results suggest that the TBT-induced negative inotropy may be at least in part ROS-dependent, especially at low workloads and Ca transients. However, at higher workloads and [Ca²⁺]_o levels these antioxidants cannot prevent dysfunction. It seems likely that TBT may also have direct effects on the ion transporters discussed, and there is evidence that OTs may interact with SERCA thiol groups resulting in inhibition (Sahib and Desaiah, 1987).

TBT has been identified as a potent inhibitor of mitochondrial ATP synthase (Matsuno-Yagi and Hatefi, 1993, von Ballmoss et al., 200), and that may also participate in the observed contractile dysfunction (both Ca²⁺ handling and contraction rely on ATP). However, the beneficial effects of dantrolene, lead us to conclude that the acute alterations in Ca²⁺ handling and contractile function described here may occur at lower TBT concentration or exposure times than required for profound ATP limitation. But such ATP limitation (suggested by the observed mitochondrial depolarization in Fig. 7D) would certainly further depress cardiac function.

In conclusion, acute exposure to the environmental contaminant TBT causes cardiotoxicity within minutes. We demonstrated that TBT causes negative inotropic and lusitropic effects on the heart, increases ROS and SR Ca²⁺ leak leading to reduced SR Ca²⁺ content. The lower SR Ca²⁺ and slowed reuptake may mediate the negative inotropic and lusitropic effects of TBT.