Inhibition of the sodium–calcium exchanger via SEA0400 altered manganese-induced T₁ changes in isolated perfused rat hearts

Abstract

Manganese (Mn²⁺)-enhanced MRI (MEMRI) provides the potential for the in vivo evaluation of calcium (Ca²⁺) uptake in the heart. Recent studies have also suggested the role of the sodium–calcium (Na⁺–Ca²⁺) exchanger (NCX) in Mn²⁺ retention, which may have an impact on MEMRI signals. In this study, we investigated whether MEMRI with fast T₁ mapping allowed the sensitive detection of changes in NCX activity. We quantified the dynamics of the Mn²⁺-induced T₁ changes in isolated perfused rat hearts in response to SEA0400, an NCX inhibitor. The experimental protocol comprised 30 min of Mn²⁺ perfusion (wash-in), followed by a 30-min wash-out period. There were three experimental groups: 1, NCX inhibition by 1 μM SEA0400 during Mn²⁺ wash-in only (SEAin, n = 6); 2, NCX inhibition by 1 μM SEA0400 during Mn²⁺ wash-out only (SEAout, n = 6); 3, no NCX inhibition during both wash-in and wash-out to serve as the control group (CNTL, n = 5). Rapid T₁ mapping at a temporal resolution of 3 min was performed throughout the perfusion protocol using a triggered saturation–recovery Look–Locker sequence. Our results showed that NCX inhibition during Mn²⁺ wash-in caused a significant increase in relaxation rate (R₁) at the end of Mn²⁺ perfusion. During the wash-out period, NCX inhibition led to less reduction in R₁. Further analysis of Mn²⁺ content in myocardium with flame atomic absorption spectroscopy was consistent with the MRI findings. These results suggest that Mn²⁺ accumulation and retention in rat hearts are, in part, dependent on NCX activity. Hence, MEMRI may provide an imaging method that is also sensitive to changes in NCX activity.

INTRODUCTION

Cardiac excitation–contraction (EC) coupling, the process that converts an electrical stimulus to muscle contraction, is fundamental to ventricular function. Calcium (Ca²⁺), the ubiquitous second messenger that directly activates myofilaments, is of key importance in EC coupling (1). During a cardiac action potential, Ca²⁺ influx, mostly via the voltage-sensitive L-type Ca²⁺ channels, triggers Ca²⁺ release from the sarcoplasmic reticulum (SR), leading to a transient increase in cytosolic Ca²⁺ concentration during systole. During diastole, Ca²⁺ is removed from the cytosol, mostly through the Ca²⁺-ATPase located in the SR and the sarcolemmal sodium–calcium (Na⁺–Ca²⁺) exchanger (NCX) (2,3). Myocardial contractility is dependent on the total Ca²⁺ concentration which must be supplied to and removed from the cytosol during each heart beat. Abnormal Ca²⁺ cycling has been implicated in contractile dysfunction (4).

The current investigation of Ca²⁺ cycling largely relies on the characterization of isolated cells or ex vivo hearts using fluorescent dyes or electrophysiological methods (5–7). Recently, manganese (Mn²⁺)-enhanced MRI (MEMRI) has been proposed for the in vivo evaluation of Ca²⁺ cycling in the heart (8). Mn²⁺ is a potent T₁-shortening agent (9). Unlike a gadolinium (Gd)-based contrast agent that is confined to the extracellular space, Mn²⁺ enters the cell through L-type Ca²⁺ channels (10,11) and perhaps NCX (12). Thus, it offers the unique opportunity for the in vivo delineation of an important cellular process that initiates EC coupling. Several studies have shown that the MEMRI signal reflects changes in Ca²⁺ uptake in the myocardium in vivo (13,14). More recently, Waghorn et al. (15,16) have also observed T₁ changes associated with Mn²⁺ efflux via the NCX 1 h after the withdrawal of Mn²⁺. Their observation suggests that MEMRI may also be used for the evaluation of NCX activity, another important determinant of Ca²⁺ cycling. However, significantly higher temporal resolution and sensitivity are required to evaluate Mn²⁺ transport via NCX without prolonged imaging times.

In the current study, we evaluated the potential of MEMRI for the sensitive delineation of NCX activity via fast T₁ mapping. We hypothesized that Mn²⁺ accumulation and retention in cardiomyocytes were dependent on Mn²⁺ efflux via the NCX, which can be calculated from Mn²⁺-induced T₁ changes. Using a fast T₁ mapping method, we quantified the dynamics of Mn²⁺-induced T₁ changes in isolated perfused rat hearts in response to SEA0400, an NCX inhibitor (17), during Mn²⁺ perfusion (wash-in) and wash-out. Myocardial Mn²⁺ content was also measured by flame atomic absorption spectrophotometry to validate the findings by MRI. In addition, the cardiotoxicity of Mn²⁺ was evaluated in isolated cardiac myocytes exposed to 50–500 μM Mn²⁺. These results will contribute to our understanding of the cellular processes that may have an impact on MEMRI measurements.

MATERIALS AND METHODS

Heart perfusion protocol

Male Sprague–Dawley rats (8–10 weeks) were heparinized (1000 units/kg, intraperitoneally) and anesthetized by sodium pentobarbital (85 mg/kg, intraperitoneally). The heart was excised, cannulated and perfused with Krebs–Henseleit (KH) buffer containing (in mM): NaCl, 118.5; KCl, 4.7; MgSO₄, 1.2; KH₂PO₄, 1.2; CaCl₂, 1.5; glucose, 11.1; NaHCO₃, 25. The perfusate was maintained at 37 °C and equilibrated with 95% O₂–5% CO₂. The perfusion pressure was maintained at a constant level (100 cmH₂O). A water-filled latex balloon was inserted into the left ventricle and connected to a pressure transducer to record the left ventricular end-diastolic pressure (LVEDP), left ventricular systolic pressure (LVSP) and heart rate (HR). The left ventricular developed pressure (LVDP) was calculated as the difference between LVSP and LVEDP. The rate–pressure product (RPP), i.e. the product of LVDP and HR, was calculated as an index of the workload.

Hearts were paced at 360 beats/min using a Grass stimulator (Grass Technologies, West Warwick, RI, USA). Once the heart rate and pressure were stabilized, the perfusate was switched to modified KH buffer containing 30 μM MnCl₂ for 30 min (the wash-in period), followed by a 30-min wash-out period with Mn²⁺-free buffer. During Mn²⁺ perfusion, phosphate and sulfate were replaced with chloride in the modified KH buffer to prevent Mn²⁺ precipitation. There were three experimental groups: 1, NCX inhibition by 1 μM SEA0400 during Mn²⁺ wash-in only (SEAin, n = 6); 2, NCX inhibition by 1 μM SEA0400 during Mn²⁺ wash-out only (SEAout, n = 6); 3, no NCX inhibition during both wash-in and wash-out to serve as the control group (CNTL, n = 5).

Image acquisition

The perfusion column was placed inside a 9.4-T vertical-bore spectrometer (Bruker Biospin Co., Billerica, MA, USA). Image acquisition used a 20-mm volume coil. A 1-mm-thick short-axis slice at the midventricular level was prescribed for imaging. A triggered saturation–recovery Look–Locker sequence was used for rapid T₁ mapping (18). Pacing signals were used to trigger the image acquisition. Prior to Mn²⁺ perfusion, two baseline T₁ maps were acquired. To delineate the kinetics of Mn²⁺-induced contrast enhancement, sequential T₁ maps were acquired at a temporal resolution of 3 min during the Mn²⁺ wash-in and wash-out periods. Imaging parameters were as follows: TE, 2 ms; TR, trigger interval, 166 ms; flip angle, 10°; field of view, 2.5 × 2.5 cm²; matrix size, 128 × 64. At the end of imaging acquisition, hearts were frozen in liquid nitrogen for the quantification of Mn²⁺ content by flame atomic absorption spectrophotometry. To quantify the Mn²⁺ content at the end of Mn²⁺ wash-in, an additional set of hearts was frozen at the end of the wash-in period.

Image analysis

Image analysis used in-house-developed MATLAB-based software described in detail previously (18). T₁ maps of the whole heart were generated by performing pixel-wise curve fitting. The myocardial free wall was selected as the region of interest to quantify the changes in the T₁ relaxation time during the imaging protocol.

Mn²⁺ quantification by flame atomic absorption spectrophotometry

Frozen ventricular tissues were burned in a furnace at 600 °C for 2 h. The ashes were dissolved in 20% nitric acid. The Mn²⁺ content was measured by a flame atomic absorption spectrophotometer (Buck Scientific, Norwalk, CT, USA).

Mn²⁺ toxicity on isolated myocytes

To evaluate the cardiotoxicity of Mn²⁺, myocyte shortening and Ca²⁺ transients were measured in isolated mouse myocytes. Briefly, mice were anesthetized with pentobarbital (85 mg/kg) and heparin (1000 units/kg). The heart was excised, cannulated and perfused with Ca²⁺-free Tyrode solution containing 0.8 mg/mL collagenase type II (Worthington Biochemical Co., Lakewood, NJ, USA) for 5 min. The original Tyrode solution contained (in mM): NaCl, 136; KCl, 5.4; MgCl₂, 1.0; HEPES, 10; NaH₂PO₄, 1.2; glucose, 5.6; L-glutamine, 2; taurine, 5. The heart was then removed from the perfusion column. The ventricles were minced, gently agitated and rinsed. Isolated myocytes were collected and incubated in Media 199 (GIBCO, Grand Island, NY, USA) containing 1.8 mM Ca²⁺ and 0, 50, 100 and 500 μM of MnCl₂, respectively, for 1 h. The temperature and pH were maintained at 37°C and pH 7.2–7.4, respectively.

Myocytes were placed in a glass-bottomed Petri dish on the stage of an Olympus IX71 inverted fluorescence microscope (Olympus America, Center Valley, PA, USA). Myocyte contractility was evaluated by calculating the maximum fractional shortening from changes in cell length during 1-Hz stimulation using a Grass stimulator (Grass Technologies). To measure Ca²⁺ transients, myocytes were incubated at 37 °C with 1 μM fura-2-acetoxymethyl ester for 15 min, and the extra dye was washed out. The dye was excited at 340 and 380 nm using a xenon arc lamp through a computer-controlled high-speed random access monochromator (Photon Technology International, Birmingham, NJ, USA). The fluorescent signals were detected at 510 nm by an analog/photon counting photomultiplier detector (Photon Technology International). Ca²⁺ transients were calculated as the ratio of the detected fluorescence in response to 340 and 380-nm excitation wavelengths (F₃₄₀/F₃₈₀), respectively, using in-house-developed MATLAB software.

Statistical analysis

All the data are expressed as the mean ± standard deviation. Mean values in the CNTL, SEAin and SEAout groups were compared by one-way analysis of variance. If there were statistical differences, multiple pairwise comparisons were performed using Tukey’s test. P < 0.05 was considered to be statistically significant.

RESULTS

Impact of Mn²⁺ on myocyte shortening and Ca²⁺ transients

At a concentration of 50 μM, Mn²⁺ had no obvious impact on myocyte contractility and Ca²⁺ handling (Fig. 1). Both myocyte shortening and Ca²⁺ transients were similar to those of the control myocytes [p = not significant]. At 100 μM Mn²⁺ concentration, peak Ca²⁺ transients remained normal. However, myocyte shortening was decreased significantly from 8.65 ± 1.15% without Mn²⁺ to 6.21 ± 1.43% (p < 0.005). At 500 μM Mn²⁺ concentration, both myocyte shortening and peak Ca²⁺ transients were significantly reduced (p < 0.001). In addition, the baseline fluorescence ratio increased significantly (p < 0.001).

Figure 1.

Manganese (Mn²⁺) toxicity on isolated myocytes. (a) Representative recordings of changes in myocyte length during electrical stimulation and the calculated fractional shortening. (b) Peak calcium (Ca²⁺) transients. (c) Resting intracellular Ca²⁺ concentration. #p < 0.005 compared with the control; *p <0.0001 compared with the other groups.

Animal characteristics and contractile function

The age, body weight and heart weight of the animals are listed in Table 1. There were no significant differences among the three groups. Figure 2a shows the time course of RPP changes during the imaging protocol. Ventricular pressure and RPP before Mn²⁺ perfusion (baseline), as well as during Mn²⁺ wash-in and wash-out, are listed in Table 2. Mn²⁺ perfusion showed no impact on ventricular function. However, NCX inhibition with SEA0400 induced a significant increase in both LVSP and LVEDP, with LVSP increased to a greater extent (p < 0.001). As a result, both LVDP and RPP increased significantly in the SEAin and SEAout groups when compared with the controls (p < 0.001).

Table 1.

Animal characteristics

	Age (weeks)	Body weight (g)	Heart weight (g)
Control (n = 5)	8.89 ± 0.91	317.20 ± 22.13	1.54 ± 0.12
SEA0400 wash-in (n = 6)	9.40 ± 0.76	334.01 ± 16.87	1.58 ± 0.27
SEA0400 wash-out (n = 6)	9.86 ± 0.62	332.33 ± 26.55	1.66 ± 0.27

Figure 2.

Time courses of rate–pressure product (RPP) (a) and relaxation rate (R₁) (b) during imaging protocol. Shaded areas indicate the wash-out period. Sodium–calcium (Na⁺–Ca²⁺) exchanger (NCX) inhibition was induced by 1 μM SEA0400 during either manganese (Mn²⁺) perfusion (SEAin) or wash-out (SEAout). *p < 0.05 compared with control (CNTL).

Table 2.

Ventricular function during the imaging protocol

		Baseline	Wash-in	Wash-out
Control (n = 5)	LVSP (mmHg)	90.58 ± 5.85	84.16 ± 10.81	74.95 ± 8.30b
	LVEDP (mmHg)	1.25 ± 1.68	3.04 ± 1.83	3.45 ± 1.30
	LVDP (mmHg)	88.40 ± 6.97	81.96 ± 10.61	71.40 ± 8.70b
	RPP (×104 mmHg/min)	3.32 ± 0.20	3.06 ± 0.39	2.60 ± 0.32b
SEA0400 wash-in (n = 6)	LVSP (mmHg)	90.94 ± 8.25	113.20 ± 19.10a	102.09 ± 26.45a
	LVEDP (mmHg)	2.92 ± 0.79	1.73 ± 1.44a	1.50 ± 1.61a
	LVDP (mmHg)	87.97 ± 9.02	111.47 ± 19.72a	100.59 ± 26.84a
	RPP (×104 mmHg/min)	3.21 ± 0.33	4.07 ± 0.72a	3.67 ± 0.98a
SEA0400 wash-out (n = 6)	LVSP (mmHg)	88.68 ± 5.25	84.94 ± 10.83	101.51 ± 22.07a
	LVEDP (mmHg)	1.19 ± 2.07	2.78 ± 1.71	1.18 ± 2.38a
	LVDP (mmHg)	86.58 ± 5.75	81.12 ± 11.62	99.31 ± 23.39a
	RPP (×104 mmHg/min)	3.16 ± 0.21	2.96 ± 0.42	3.61 ± 0.86a

LVDP, left ventricular developed pressure; LVEDP, left ventricular end-diastolic pressure; LVSP, left ventricular systolic pressure; RPP, rate–pressure product.

^ap < 0.001 compared with the control group at the same time point.

^bp < 0.05 compared with baseline in the same group.

T₁ changes during wash-in and wash-out

Figure 3a shows representative T₁ maps at baseline, at the end of Mn²⁺ wash-in and at the end of the wash-out period. The time courses of R₁ changes during the imaging protocol are shown in Fig. 2b. All three groups showed a progressive increase in R₁ during the wash-in period. However, the SEAin group exhibited an accelerated increase in R₁. After 3 min of Mn²⁺ perfusion, R₁ in the SEAin group was significantly higher than that in the other two groups (p < 0.05). At the end of Mn²⁺ wash-in, the R₁ values were 1.61 ± 0.18, 1.88 ± 0.28 and 1.62 ± 0.21 s⁻¹ for the CNTL, SEAin and SEAout groups, respectively (Fig. 3b).

Figure 3.

Longitudinal relaxation time (T₁), rate (R₁) and manganese (Mn²⁺) content. (a) Representative T₁ maps before Mn²⁺ perfusion (baseline), at the end of Mn²⁺ wash-in and at the end of wash-out. Relaxation rate (R₁) (b) and Mn²⁺ content (c) at the corresponding time points. Sodium–calcium (Na⁺–Ca²⁺) exchanger (NCX) inhibition was induced by 1 μM SEA0400 during either Mn²⁺ perfusion (SEAin) or wash-out (SEAout). *p < 0.05 compared with the other two groups at the same time points. CNTL, control.

All three groups showed a slight decrease in R₁ during the wash-out period (Fig. 2b). The R₁ decrease was smallest in the SEAout group. At the end of wash-out, R₁ in both the SEAout and SEAin groups was significantly higher than that in the CNTL group (Fig. 3b, p <0.05). The R₁ values at the end of wash-out were 1.36 ± 0.17, 1.70 ± 0.32 and 1.65 ± 0.10 s⁻¹ for CNTL, SEAin and SEAout groups, respectively.

The wash-out curves were fitted to two exponential functions, i.e. y = A e^−t/B and y = A e^−t/B + C, respectively. The fitted parameters are listed in Table 3. Using a simple exponential curve fitting (y = A e^−t/B), the mean half-times for R₁ reduction were 3.18, 4.25 and 6.66 h for the CNTL, SEAin and SEAout groups, respectively. Using an expoenntial function with a constant term (y = A e^−t/B + C), the fitted time constant was reduced significantly for all three groups. The C/A ratio ranged from 5.87 to 10.71.

Table 3.

Time constants of R₁ decrease during the wash-out period

y = A e−t/B				y = A e−t/B + C
	Control	SEA0400 wash-in	SEA0400 wash-out		Control	SEA0400 wash-in	SEA0400 wash-out
A	1.63	1.80	1.72	A	0.24	0.16	0.23
B (h)	4.57	6.09	9.62	B (h)	0.42	0.22	1.07
				C	1.40	1.67	1.49
				C/A	5.87	10.71	6.44
R2	0.94	0.82	0.74	R2	0.96	0.88	0.74

Myocardial Mn²⁺ content

Consistent with MRI findings, NCX inhibition increased significantly the Mn²⁺ content in SEAin hearts (12.9 ± 1.6 μg/g wet weight) at the end of wash-in when compared with controls (8.2 ± 0.2 μg/g wet weight) (Fig. 3c, p <0.001). Perfusion with Mn²⁺-free buffer caused a significant decrease in myocardial Mn²⁺ content at the end of wash-out in CNTL hearts (6.3 ± 0.9 μg/g wet weight, p < 0.005). NCX inhibition led to increased Mn²⁺ retention in the SEAout group (9.5 ± 3.0 μg/g wet weight) when compared with the CNTL group (Fig. 3c, p < 0.05). As a result of increased Mn²⁺ accumulation during the wash-in period, SEAin hearts also showed elevated Mn²⁺ content at the end of wash-out (9.1 ± 2.5 μg/g wet weight) when compared with the CNTL group (Fig. 3c, p < 0.05).

DISCUSSION

The major findings of the present study were that, in perfused rat hearts, the Mn²⁺ content and T₁ mapping with MEMRI were dependent on SEA0400 and thus, presumably, on NCX activity. Previously, Waghorn et al. (15,16) have investigated Mn²⁺ retention in an in vivo mouse study. They observed that treatment with SEA0400 reduced significantly the rate of decrease in ΔR₁ hours after Mn²⁺ infusion was withdrawn. In the current study, we used a fast T₁ mapping method to follow the dynamic changes in R₁ during Mn²⁺ perfusion and wash-out. Our results suggest that altered Mn²⁺ efflux via NCX is reflected in R₁ changes during both wash-in and wash-out.

Physiologically, Ca²⁺ entry into myocytes via L-type Ca²⁺ channels is balanced by its efflux through NCX (1). Although Mn²⁺ also enters the myocytes through L-type Ca²⁺ channels, its efflux via NCX has been considered to be negligible because of the long Mn²⁺ retention time. However, recent studies by Waghorn et al. (15,16) have shown that differences in R₁ changes induced by NCX inhibition can be observed 4 h after Mn²⁺ withdrawal. With NCX inhibition, these authors observed an increase in the half-time of ΔR₁ reduction from 3.4 h (without SEA0400) to 5.6 h (with SEA0400) in mouse hearts. In the current study, the estimated half-time for R₁ reduction increased from 3.18 h with-out SEA0400 to 6.66 h in the presence of SEA0400, which was similar to that reported by Waghorn et al. The average half-time for R₁ reduction in the SEAin group (4.25 h) was also higher than that in the controls because of the incomplete elimination of the inhibitor during the wash-out period (19). Consistent with the MRI findings, the Mn²⁺ content at the end of the wash-out period was also significantly higher in hearts perfused with SEA0400 than in the control group.

It is interesting to note that, when the wash-out curves were fitted to an exponential function with a constant term, the fitted time constant was significantly reduced by an order of magnitude (Table 3). Moreover, the ratio of the constant term to the exponential term (C/A) was more than five-fold in all three groups, which may suggest the existence of a large Mn²⁺ pool that is washed away very slowly.

With NCX inhibition by SEA0400, the R₁ values at the end of the wash-in period increased in the SEAin group when compared with hearts without the inhibitor (SEAout and CNTL groups), suggesting increased Mn²⁺ accumulation in the presence of the NCX inhibitor during the wash-in period. Atomic absorption spectroscopy analysis of hearts at the end of the wash-in period also showed that the Mn²⁺ content in the SEAin group increased by 57% compared with the other two groups. As SEA0400 does not enhance Ca²⁺ channel activity under normal physiological conditions (20), this increased Mn²⁺ accumulation presumably reflects reduced Mn²⁺ efflux via NCX. These data suggest that MEMRI with fast T₁ mapping may provide a tool for the evaluation of NCX activity.

The contribution of NCX to Ca²⁺ extrusion from myocytes is species dependent (21,22). Sham et al. (21) evaluated the rates of Ca²⁺ removal in isolated myocytes from rat, guinea-pig, hamster ventricles and human atria. Their study suggests that NCX activity in the rat ventricle is lower than that in hamsters, guinea-pigs and humans. As such, larger animals and humans may manifest a greater Mn²⁺ efflux rate than that observed in rats in the current study. However, careful further investigations are necessary to evaluate whether this difference in NCX contribution to Ca²⁺ extrusion will render MEMRI more sensitive to alterations in NCX activity in large animals and humans.

The inhibition of NCX also induced a small increase in LVDP. This result is consistent with previous studies on perfused rat hearts (23,24). Previously, Acsai et al. (20) have reported an increase in myocyte shortening in isolated rat myocytes incubated with SEA0400. In addition, they observed an increase in Ca²⁺ transients and a trend of increase in diastolic Ca²⁺ level. However, this increase in cytosolic Ca²⁺ was not accompanied by an increase in the L-type Ca²⁺ current. Therefore, the observed increase in Ca²⁺ transients is probably caused by the blockage of Ca²⁺ efflux via NCX, leading to increased ventricular contractility.

Because both Mn²⁺ uptake and efflux can have an impact on the dynamics of R₁ during Mn²⁺ infusion, care must be taken in interpreting the data. In general, the determination of both influx and efflux rates from a single wash-in curve is underdetermined unless other constraints are imposed in parameter estimation. In the current experimental settings, the measured R₁ dynamics during wash-out can provide additional data to constrain curve fitting. For in vivo studies, the measurement of the Mn²⁺ content in blood, i.e. the arterial input function, can also provide additional constraints for curve fitting. However, such an approach requires fast and accurate T₁ mapping in both myocardium and blood.

In summary, we have investigated the sensitivity of MEMRI to NCX inhibition via SEA0400. Our results show that, in rat hearts, MEMRI with fast T₁ mapping is sensitive to SEA0400-dependent changes in Mn²⁺ accumulation and retention. Although these results suggest that MEMRI has the potential to detect altered NCX activity, further investigation is needed to assess the sensitivity of MEMRI to more subtle changes in NCX activity under disease conditions.

Abbreviations used

EC

excitation–contraction

HR

heart rate

KH

Krebs–Henseleit

LVDP

left ventricular developed pressure

LVEDP

left ventricular end-diastolic pressure

LVSP

left ventricular systolic pressure

MEMRI

manganese-enhanced MRI

NCX

Na⁺–Ca²⁺ exchanger

RPP

rate–pressure product

SR

sarcoplasmic reticulum