Contribution of the Sodium-Calcium Exchanger to Contractions in Immature Rabbit Ventricular Myocytes

Abstract

In immature cardiac myocytes, the sarcoplasmic reticulum is sparse. Thus, we hypothesized that sarcolemmal Ca²⁺ influx through Na⁺-Ca²⁺ exchange is the dominant mechanism for modulating intracellular Ca²⁺ during contractions in fetal and neonatal hearts. We measured Na⁺-Ca²⁺ exchange currents in neonatal and adult rabbit ventricular cells using a rapid solution switch into 0 mM external Na⁺. The current densities (mean ± SEM) were larger in 8 neonatal cells than in 10 adult cells (5.4 ± 1.38 versus 1.65 ± 0.25 pA/pF). Intracellular Ca²⁺ transients after inhibiting the sarcoplasmic reticulum with ryanodine and thapsigargin were unchanged in 15 neonatal cells, but decreased in 15 adult cells to 78.9 ± 5.6% of baseline. When the Ca²⁺ channels were also inhibited by adding nifedipine, Ca²⁺ transients from Na⁺-Ca²⁺ exchange were 30.0 ± 3.5% of baseline in neonatal cells compared with 13.4 ± 3.4% in adult cells. Simultaneous contractions were a larger percent of baseline in neonatal cells(85.7.6 ± 6.4%) than in adult cells (78.9 ± 5.6%) after inhibiting the sarcoplasmic reticulum, and were unmeasureable in many cells from both age groups after inhibiting the Ca²⁺ channels as well. The ratio of Na⁺-Ca²⁺ exchanger mRNA to sarcoplasmic reticulum Ca²⁺-ATPase mRNA levels decreased from 1.0 ± 0.13 to 0.4± 0.03 to 0.26 ± 0.02 in fetal, neonatal and adult ventricles, respectively. These measurements were consistent with a dominant role for the Na⁺-Ca²⁺ exchanger in the immature heart.

Main

The sarcoplasmic reticulum is rudimentary, and its contribution to contractions is diminished in immature hearts^(1–8). Thus, sarcolemmal Ca²⁺ influx through the Na⁺-Ca²⁺ exchanger and/or Ca²⁺ channels is likely important during contractions in fetal and neonatal hearts^{(7, 9–11)}. The Na⁺-Ca²⁺ exchanger participates in cardiac myocyte relaxation^(12–15), and probably participates in contractions as well^(16–19).

Previous studies showed sarcolemmal Ca²⁺ influx through the Na⁺-Ca²⁺ exchanger contributes to contractions in neonatal hearts. Na⁺-Ca²⁺ exchanger protein⁽²⁰⁾ and mRNA levels⁽²¹⁾ are increased in fetal and neonatal hearts, and the density of “nickel-inhibitable” currents (presumed to be I_NaCa), is increased in neonatal ventricular cells⁽¹⁰⁾. Contractions in neonatal cells are unchanged after nifedipine, suggesting Ca²⁺ influx from Na⁺-Ca²⁺ exchange but not from the Ca²⁺ channels is important for contractions⁽¹¹⁾.

We hypothesized that sarcolemmal Ca²⁺ influx through the Na⁺-Ca²⁺ exchanger contributes to contractions in fetal and neonatal hearts. Compared with adult cells, the density of I_NaCa was larger in neonatal cells. Blockade of the sarcoplasmic reticulum resulted in a smaller percent decrease in intracellular Ca²⁺ transients and contractions from baseline, and Ca²⁺ transients from Na⁺-Ca²⁺ exchange were a larger percent of baseline in neonatal cells. Expression of Na⁺-Ca²⁺ exchanger mRNA decreased with development. All of these findings were consistent with an increased contribution of the Na⁺-Ca²⁺ exchanger to contractions in the immature heart.

METHODS

Isolation of rabbit ventricular myocytes. We obtained immature ventricular myocytes from the hearts of 1-4-d old (50-100 g) New Zealand White rabbits by enzymatic digestion, using a protocol approved by the University of Utah Animal Use Committee⁽⁷⁾. All solutions used for neonatal cell isolations contained (in mM) NaCl 126, KCl 4.4, MgCl₂ 5.0, taurine 20, creatine 5.0, sodium pyruvate 5.0, sodium phosphate 1.0, HEPES 12 and dextrose 22 (pH 7.4 using NaOH). We excised the hearts after intraperitoneal injections of sodium heparin (1000 U/kg) and sodium pentobarbital (100 mg/kg), and placed them in solution with 3.5 mM Ca²⁺ to obtain vigorous contractions which cleared the hearts of blood. We then cannulated the hearts through their aortas, and perfused them at 30 torr and 37 °C for 4 min with Ca²⁺-free solution, for 6 min with solution containing 0.1 mM Ca²⁺, 0.6 mg/mL collagenase type II (Worthington, Freehold, NJ) and 0.01 mg/mL protease XIV (Sigma Chemical Co., St. Louis, MO), and for 4 min with 0.1 mM Ca²⁺ solution. Finally, we teased the digested ventricular tissue with forceps to disperse the myocytes.

We obtained adult ventricular myocytes from the hearts of 6-8-wk-old(2.0-2.5 kg) New Zealand White rabbits using a similar enzymatic digestion technique⁽²²⁾. We used MKRBB solution for adult cell isolations. This solution contained (in mM) NaCl 91.7, KCl 30, CaCl₂ 0.9, KH₂PO₄ 1.2, MgSO₄ 1.2, NaHCO₃ 19, dextrose 15, taurine 20 and adenosine 0.5 (pH 7.40 in 5% CO₂ and 95% O₂). We administered heparin and sodium pentobarbital i.v., then perfused the hearts at 60 torr for 5 min with Ca²⁺-free MKRBB, and for 20 min with MKRBB containing 0.28 mg/mL collagenase, 0.4 mg/mL hyaluronidase type I-S(Sigma Chemical Co.), and 50 μM CaCl₂. We minced the partially digested ventricular tissue in MKRBB with 50 μM CaCl₂, 1% albumin, and 0.017 mM insulin (Sigma Chemical Co.), and further digested the tissue with 10-min incubation in MKRBB with 50 μM CaCl₂, 0.28 mg/mL collagenase, and 0.0025% trypsin. We then mixed the resulting cell suspension with equal amounts of inhibitor solution containing 50 μM CaCl₂, 0.0025% trypsin inhibitor (Sigma Chemical Co.), and 12% FCS. Finally, we pelleted the cells by centrifugation at 250 rpm for 5 min.

We resuspended both neonatal and adult cells in KB medium for 1 h before studies⁽²³⁾, and gradually increased the external Ca²⁺ to 2.7 mM over 1 h. We consistently obtained large numbers of rod-shaped, Ca²⁺-tolerant cells with our cell isolations.

Measurement of Na⁺-Ca²⁺exchange current. We studied developmental changes in Na⁺-Ca²⁺ exchanger density, by measuring I_NaCa in neonatal and adult ventricular cells. We studied the cells using an Axopatch 200 A voltage clamp amplifier (Axon Instruments Inc., Foster City, CA), in the whole cell disrupted voltage-clamp mode. We pulled suction micropipettes from Corning 7052, 1.65-mm outside diameter × 1.2-mm inside diameter borosilicate glass tubing (A-M Systems, Everett, WA), coated them with Sylgard (Dow-Corning Corp., Midland MI), and fire-polished the tips. We used pipettes with a resistance of 3-4 megohms to study neonatal cells, and 1-2 megohms resistance to study adult cells. The pipette solution contained(in mM) NaCl 10, MgCl₂ 0.2, MgATP 3.0, dextrose 5.5, HEPES 10 and EGTA 14 to maintain a calculated internal Ca²⁺ of 100 nM⁽²⁴⁾. We adjusted solution pH to 7.1 with CsOH and added CsCl added to give a final Cs⁺ concentration of 130 mM. We voltage-clamped the cells with a holding potential of -40 mV and superfused them initially with normal Tyrode's solution at 30 °C. A rapid external solution change into 0 mM K⁺ and Na⁺ (Li⁺ replacement), produced an outward current from Na⁺-Ca²⁺ exchange. The rapid switch from normal solution into 0 mM external Na⁺ required 2-3 ms⁽²⁵⁾, and established a Na⁺ gradient which favored Na⁺ efflux and Ca²⁺ influx. We blocked contaminating outward K⁺ currents with 0 mM external K⁺ and Cs⁺ in the pipette.

We normalized I_NaCa for differences in surface area by dividing peak current amplitude by cell capacitance, and expressed it as current density⁽²⁶⁾. To calculate cell capacitance, we used hyperpolarizing current clamp steps and the formula C_M =τ_M/R_M (C_M = membrane capacitance,τ_M = membrane time constant, and R_M = membrane resistance) in some cells. Cell capacitance measured in this way was found to correlate with the value on the analog dial of the voltage clamp amplifier.

Measurement of contractions in field stimulated ventricular myocytes. We studied the cells in a perfusion chamber on the stage of a Nikon Diaphot TMD inverted phase-contrast microscope (Nikon Inc., Melville, NY). We attached the cells to the bottom of the cell bath with mouse laminin(Collaborative Biomedical Products, Bedford, MA), and perfused them with solutions from bottles above the cell bath. We kept the level of the bathing solution constant with continuous suction. We changed solutions with a valve, located near the inflow of the perfusion chamber. Solution changes were typically complete in ≅5-10 s. We field-stimulated the cells with platinum electrodes, using 4-6-ms pulses and voltages 1.5 × threshold for contractions.

We transmitted the image of the contracting cells, obtained with a CCD camera (Javelin Electronics, Torrance, CA) to a video recorder and video monitor, where we tracked the ends of the contracting cell with a two-edge video-based motion detector (Crescent Electronics, Salt Lake City, UT)⁽²⁷⁾. We rotated the camera to orient the long axis of the cell along a single raster line on the video monitor, and measured contractions with either a 10 × (adult cells) or 15 × (neonatal cells) projection lens between the microscope and camera to use the full range of the motion detector and maximize the signal to noise ratio during measurements. We calibrated the analog voltage output from the motion detector to indicate changes in cell length in micrometers and quantitated contractions by measuring shortening as a function of resting cell length. Temporal resolution with this system was 20 ms, and spatial resolution 1 part in 10,000, allowing the measurement of changes as small as 0.01% of the resting cell length (assuming the cell image filled the screen and the axis of contraction was aligned horizontally).

We measured contractions at room temperature (23-24 °C) rather than 37°C to obtain larger contractions and decrease cell “run-down.” The cells were superfused initially with normal Tyrode's solution containing(in mM) NaCl 138, dextrose 11.0, MgCl₂ 1.0, KCl 4.4, CaCl₂ 2.7, and HEPES 12.0 (pH adjusted to 7.4 with NaOH). To inhibit Ca²⁺ release⁽²⁸⁾ and uptake⁽²⁹⁾ from the sarcoplasmic reticulum, we perfused the cells with Tyrode's solution containing 10 μM ryanodine and 500 nM thapsigargin. To inhibit the L-type Ca²⁺ channels and sarcoplasmic reticulum, we perfused the cells with Tyrode's solution containing 20 μM nifedipine in addition to ryanodine and thapsigargin. We used 20 μM nifedipine in these studies to achieve the maximum inhibitory effect of nifedipine on contractions.

Measurement of intracellular Ca²⁺transients in field stimulated ventricular myocytes. We measured intracellular Ca²⁺ transients simultaneously in contracting neonatal and adult cells, using the membrane permeable AM form of Indo-1 (Molecular Probes Inc., Eugene, OR). A 200-W xenon light source (Oriel Corp., Stratford, CT) provided an excitation wavelength of 365 nm, and a 400-nm dichronic mirror placed beneath a 40× oil-immersion objective (Nikon, phase contrast Fluor) directed the excitation light on the cells. A beam splitter divided the light emitted from the cells, and the 405- and 495-nm fluorescence signals were measured using separate photomultiplier tubes (Hammatsu Corp., Bridgewater, NJ). We used the 405/495 ratio as an indicator of intracellular Ca²⁺ levels^(30–32).

We loaded the cells for 10 min at room temperature with 10 μM Indo-1 AM, and washed them for a minimum of 15 min (neonatal cells) or 30 min (adult cells) with normal Tyrode's solution before study. By limiting the time the cells were exposed to excitation light, bleaching and run-down of the intracellular Ca²⁺ transients were minimal for at least 20-25 min after wash-out. After nulling background fluorescence, we measured intracellular Ca²⁺ transients in field-stimulated cells at baseline, after inhibiting the sarcoplasmic reticulum (ryanodine and thapsigargin) and after inhibiting the sarcoplasmic reticulum and Ca²⁺ channels (ryanodine, thapsigargin, and nifedipine). We did not calibrate the fluorescence signals for levels of intracellular Ca²⁺.

RNase protection analysis. Plasmid construction. For pBluescript pA4E2, we inserted an approximately 380-bp Kpn I/EcoRI fragment from the 3′-end of the coding region of the dog heart Na⁺-Ca²⁺ exchanger cDNA (provided by K. Philipson, UCLA), into pBluescript (KS⁺) (Stratagene, La-Jolla, CA). To produce the antisense RNA probe, we cut the plasmid with EcoRI and transcribed it using T3 RNA polymerase.

For pBluescript Serca2a, we inserted an approximately 300-bp Bam HI/PstI fragment from the 3′-untranslated region of the rabbit sarcoplasmic reticulum Ca²⁺-ATPase cDNA (provided by D. MacLennan, University of Toronto) into pBluescript. To produce antisense RNA, we cut the plasmid with EcoRI and transcribed it using T7 RNA polymerase.

For pSP65U1, the U1 small nuclear RNA riboprobe (U1 snRNA) is derived from plasmid pSP65U1 containing the human U1 gene⁽³³⁾. We linearized the plasmid with HindIII and transcribed it with SP6 RNA polymerase.

RNA isolation and analysis. We isolated total RNA from the ventricles of 23-d fetal, 3-5-d neonatal, and 6-8-wk adult hearts, using Trizol reagent (Life Technologies, Inc., Gaithersberg, MD) as described by the manufacturer. RNA samples from five hearts for each developmental stage(fetal, neonatal, and adult) were probed for Na⁺-Ca²⁺ exchanger, sarcoplasmic reticulum Ca²⁺-ATPase, and U1. We performed RNase protection analysis with an in vitro transcription system from Promega (Madison, WI). We probed 10 μg of each sample of total RNA with the Na⁺-Ca²⁺ exchanger and the sarcoplasmic reticulum Ca²⁺-ATPase riboprobes together with the antisense U1 snRNA riboprobe as an internal control for RNA loading. After digestion with RNase A and RNase T1, we separated the protected bands for each mRNA over 2 h, on a 6% sequencing gel at 1500 V. The protected species for Na⁺-Ca²⁺ exchanger appeared as two distinct bands, and the sarcoplasmic reticulum Ca²⁺-ATPase protected species ran as a doublet. We quantified the relative amounts of Na⁺-Ca²⁺ exchanger sarcoplasmic reticulum Ca²⁺-ATPase and U1 snRNA protected species by phosphorimager analysis. We calculated the ratio of Na⁺-Ca²⁺ exchanger to sarcoplasmic reticulum Ca²⁺-ATPase to determine the relative amount of each mRNA. In addition, we normalized Na⁺-Ca²⁺ exchanger mRNA and sarcoplasmic reticulum Ca²⁺-ATPase mRNA to the U1 snRNA in each sample to determine the amount of each message relative to total RNA in the sample.

Data acquisition and analysis. We recorded the signals digitally on a Compaq Deskpro personal computer (Compaq Computer Corp., Houston, TX), on a digital/analog chart recorder (Gould Instrument Systems, Valley View, OH), and on a video/signal recorder (A. R. Vetter Co., Rebersburg, PA). We performed statistical analyses using the software package, BMDP Professional, for Windows (BMDP Statistical Software, Los Angeles CA), and expressed all results as mean ± SEM. We interpreted results as being significantly different for values of p < 0.05 using the appropriate statistical test. In studies where multiple interventions were performed on the same cells (e.g. measurements at baseline, after ryanodine and thapsigargin, and after ryanodine, thapsigargin, and nifedipine), we analyzed the data using a repeated measures analysis of variance. We compared data from neonatal cells with data from adult cells using a nonpaired t test.

RESULTS

Developmental changes in Na⁺-Ca²⁺exchange current. Peak current amplitude and cell capacitance were smaller, whereas the current density of I_NaCa was more than 3-fold larger in the neonatal cells (Fig. 1 and Table 1).

Figure 1

I_NaCa was elicited with a rapid external solution switch from normal Na⁺ into 0 mM Na⁺. The cells were voltage clamped, and held at -40 mV. Representative currents from neonatal (upper panel) and adult (lower panel) rabbit ventricular cells are shown.

Effect of sarcoplasmic reticulum and Ca²⁺ channel inhibition on intracellular Ca²⁺ transients and contractions in neonatal cells. Resting length in 25 control, field-stimulated neonatal cells was 56.6 ± 2.3 μm, and percent shortening was 4.8 ± 0.7%. Percent shortening did not decrease in cells loaded with Indo-1 AM. The intracellular Ca²⁺ transients were unchanged in neonatal cells after inhibiting the sarcoplasmic reticulum and decreased after inhibiting the Ca²⁺ channels as well (Fig. 2). Simultaneous contractions were smaller than baseline after inhibiting the sarcoplasmic reticulum and were measurable in only 2 of 15 cells after also inhibiting the Ca²⁺ channels.

Figure 2

Simultaneous intracellular Ca²⁺ transients and contractions in a neonatal cell at baseline, after treatment with ryanodine and thapsigargin, and after treatment with ryanodine, thapsigargin, and nifedipine. When the sarcoplasmic reticulum was inhibited, Ca²⁺ transients were unchanged, but contractions were decreased. The Ca²⁺ transients and contractions decreased further after sarcoplasmic reticulum and Ca²⁺ channel inhibition.

Developmental differences in the effect of sarcoplasmic reticulum and Ca²⁺ channel inhibition on intracellular Ca²⁺ transients and contractions. Neonatal cells were shorter than adult cells, and contracted less with field stimulation (adult cell length = 124.7 ± 3.8 μm and percent shortening = 9.7 ± 1.0% in 20 cells). Percent shortening in adult cells loaded with Indo-1 AM were not different from control cells. Ca²⁺ transients and contractions after inhibiting the sarcoplasmic reticulum, and residual Ca²⁺ transients from Na⁺-Ca²⁺ exchange were a larger percent of baseline in neonatal cells than in adult cells (Table 2). Simultaneous contractions from Na⁺-Ca²⁺ exchange, measurable in 2 neonatal and 9 adult cells, were not a larger percent of baseline in neonatal cells. Contractions after 5 and 10 μM nifedipine were a larger percent of baseline in neonatal cells than in adult cells, and were not different in cells from the two age groups after 20 and 30 μM nifedipine (Fig. 3).

Table 2 Contributions of the sarcoplasmic reticulum and Na⁺-Ca²⁺ exchanger to contractions in neonatal and adult rabbit ventricular myocytes

Figure 3

The full effect of nifedipine on contractions was observed after treatment with 20 μM nifedipine in neonatal cells and 10μM nifedipine in adult cells.

Developmental changes in mRNA levels for the sarcoplasmic reticulum and Na⁺-Ca²⁺exchanger. Normalized Na⁺-Ca²⁺ exchanger mRNA levels were similar in fetal and neonatal hearts, and greater than in adult hearts. Sarcoplasmic reticulum Ca²⁺-ATPase levels in fetal hearts were less than in neonatal and adult hearts (Fig. 4). The ratio of Na⁺-Ca²⁺ exchanger to sarcoplasmic reticulum Ca²⁺-ATPase mRNA levels decreased from 1.0 ± 0.13 to 0.4± 0.03 to 0.26 ± 0.02 in fetal, neonatal, and adult ventricles, respectively.

Figure 4

There was an increase in sarcoplasmic reticulum Ca²⁺-ATPase message levels, and a decrease in Na⁺-Ca²⁺ exchanger message levels with development.

DISCUSSION

Defining the mechanism of contractions in the immature heart is important to understanding the developmental process and integral to the successful treatment of maternal and fetal arrhythmias. In the absence of a well developed sarcoplasmic reticulum, sarcolemmal Ca²⁺ influx through the Na⁺-Ca²⁺ exchanger is a likely mechanism for regulating contractions in developing hearts. Our data confirm the role of the Na⁺-Ca²⁺ exchanger in neonatal cardiac myocytes.

Developmental changes in Na⁺-Ca²⁺exchange current density. We measured an increase in the density of I_NaCa in neonatal rabbit myocytes. The amplitude of I_NaCa is voltage-dependent and dependent on the transmembrane gradients of Na⁺ and Ca²⁺. Thus, a strength of our protocol was the ability to measure I_NaCa in neonatal and adult myocytes with the membrane potentials constant, and to fix internal and external Na⁺ and Ca⁺ concentrations by the contents of the pipette and external solutions respectively. Our protocol did not require the use of nonspecific inhibitors of Na⁺-Ca²⁺ exchange⁽³⁴⁾.

Developmental differences in the effect of sarcoplasmic reticulum and Ca²⁺ channel inhibition on intracellular Ca²⁺ transients and contractions. The response of neonatal cells to sarcoplasmic reticulum inhibition most likely indicates diminished sarcoplasmic reticulum function in neonatal myocytes. However, Ca²⁺ influx from the Ca²⁺ channels and Na⁺-Ca²⁺ exchanger may compensate for sarcoplasmic reticulum Ca²⁺ release. Thus, the smaller apparent response to ryanodine and thapsigargin in neonatal cells may have been related to the ability of these cells to compensate better for altered sarcoplasmic reticulum Ca²⁺ release.

The larger decrease in contractions compared with Ca²⁺ transients after ryanodine and thapsigargin is consistent with previous measurements showing relatively large changes in contractions at peak levels of intracellular Ca²⁺, and small changes in contractions at diastolic levels of intracellular Ca²⁺⁽³⁵⁾. The discordance between changes in the amplitude of the transients and contractions in the neonatal cells may be explained by decreased calcium sensitivity of the contractile elements in neonatal cells. Another explanation might be a smaller rise in intracellular calcium during neonatal contractions. Assuming equal or lower diastolic levels of intracellular calcium in neonatal cells, a larger percent of the baseline rise in intracellular calcium may be required for neonatal cells to achieve the threshold for contractions. Because contractions in cells loaded with Indo-1 were not different from contractions in control cells, it does not appear Indo-1 buffered intracellular calcium and decreased contractions.

The rise in intracellular Ca²⁺ from Na⁺-Ca²⁺ exchange produced contractions in neonatal as well as adult cells which were a small percent of baseline. Our results indicate that Ca²⁺ from sources other than Na⁺-Ca²⁺ exchange (most likely the Ca²⁺ channels), are necessary for functional contractions. Our findings are in contrast to observations that neonatal cells are not dependent on Ca²⁺ influx through the Ca²⁺ channels⁽¹¹⁾. The sarcoplasmic reticulum was not inhibited in these previous studies. Thus, the larger contractions which they observed may have been enhanced by sarcoplasmic reticulum Ca²⁺ release. In addition, the low dose of nifedipine (5μM compared with 20 μM in our studies), may have incompletely inhibited the Ca²⁺ channels.

Changes in mRNA levels for the sarcoplasmic reticulum and Na⁺-Ca²⁺exchanger in the developing heart. Our simultaneous measurements of Na⁺-Ca²⁺ exchanger and sarcoplasmic reticulum Ca²⁺-ATPase mRNA levels indicate the decrease in Na⁺-Ca²⁺ exchanger message levels coincides with an increase in sarcoplasmic reticulum Ca²⁺-ATPase message levels during cardiac development. The decline in Na⁺-Ca²⁺ exchanger mRNA occurs between the neonatal and adult periods, and the rise in sarcoplasmic reticulum Ca²⁺-ATPase mRNA occurs between the fetal and neonatal periods.

When sarcoplasmic reticulum function is diminished, as in immature hearts, sarcolemmal Ca²⁺ influx is essential for cardiac contractions. In neonatal ventricular cells Na⁺-Ca²⁺ exchange contributes to the rise in intracellular Ca²⁺ during contractions, but other sources of Ca²⁺ are required for functional contractions. If these findings in neonatal rabbit myocytes are representative of the properties of neonatal human myocytes, they may explain the adverse effects of rapidly administering Ca²⁺ channel antagonists to infants^{(36, 37)}.

Table 1 Developmental changes in Na⁺-Ca²⁺ exchange current(I_NaCa) in neonatal and adult rabbit ventricular myocytes