INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ACUTE MYOCARDIAL ISCHEMIA is now known to produce within the first few minutes abnormalities of cytosolic calcium that may play a role in the genesis of arrhythmias (2, 4, 6, 20, 24, 27). The best method of studying this phenomenon is to directly record the intracellular Ca²⁺ concentration ([Ca²⁺]_i) transient using an intracellular calcium indicator that can be loaded into the myocytes of an intact heart or papillary muscle. This has been accomplished using aequorin (1), as well as the fluorescent calcium indicators indo 1 (4, 24, 27) and fura 2 (2).

Previous studies of the [Ca²⁺]_i transient during ischemia have involved small, saline-perfused hearts that were denervated and studied in vitro. This is due partly to the high cost of in vivo experiments in large animals and partly to the difficulty of recording [Ca²⁺]_i transients in the presence of blood. A number of studies suggest that the full effects of ischemia on cellular electrophysiology are best observed in vivo (3) and may involve interaction of the myocardium with factors released from blood (10, 14). Blood cannot be present when indo 1 is used, because both the excitation and emission wavelengths are strongly absorbed by hemoglobin. Additionally, changes in the oxygenation of hemoglobin or myoglobin can produce changes in indo 1 fluorescence that are independent of [Ca²⁺]_i (13).

Recently, some longer-wavelength fluorescent calcium indicators have been developed that can be loaded into myocytes as cell-permeant acetoxymethyl esters (AM). In preliminary studies of several such compounds, Fura Red (22) gave particularly good results. Fura Red has an emission peak at 660 nm, which is beyond the absorption range for hemoglobin and myoglobin. The excitation wavelengths for Fura Red are also favorable for use with blood.

We now report the properties of calcium transients recorded from rabbit hearts using Fura Red. We found that high-quality [Ca²⁺]_i transients can be obtained in the presence of blood. We sought to determine whether the known effects of ischemia on the [Ca²⁺]_i transient, such as the development of [Ca²⁺]_i transient alternans, are more severe in the presence of blood. We have also developed a method for loading and calibrating Fura Red that would be suitable for use in the open-chest in vivo heart.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Isolated heart preparation. New Zealand rabbits (2.0-3.0 kg) of either sex were killed by an overdose of pentobarbital. Heparin (1,000 U/kg) was injected into the ear marginal vein before removal of the heart. The heart was rapidly removed and perfused with an oxygenated saline solution (Tyrode solution) via the aorta at a constant pressure of 100 cmH₂0. The perfusate contained (in mM) 115 NaCl, 4.7 KCl, CaCl₂ 2.0, 0.7 MgCl₂, 28 NaHCO₃, 0.5 NaH₂PO₄, 20 glucose, and 0.3 probenecid as well as insulin (10 U/l) and 1 ml/l fetal calf serum. Probenecid has been shown to prevent loss of tetracarboxylate fluorescent indicators from cells by blocking an organic anion transport system (9). The pH was adjusted to 7.4 by equilibration with 95% O₂-5% CO₂ and heated to maintain the heart at 30 ± 1°C. Left ventricular pressure was recorded using an intracavitary latex balloon. Global ischemia was produced by complete cessation of coronary flow. The temperature of the heart was maintained at 30 ± 1°C by a space heater during ischemia.

Fluorescence recordings. Calcium transients were recorded from the epicardial surface of the left ventricles. Illumination from a 100- or 500-W mercury vapor lamp was filtered at 546 ± 10 nm and directed via a liquid light guide (Oriel, Long Beach, CA) that was attached to the heart by a plastic sleeve and a rubber girdle to minimize relative motion. Filtration at 546 nm was chosen because this wavelength is an isosbestic point for light absorption by hemoglobin and myoglobin during the oxygenation/deoxygenation reactions. Excitation illumination was confined to a circular region of the left ventricular epicardial surface 1 cm in diameter. Fluorescence emissions were collected by a ring of eight coaxial fiber optics and were directed through a beam splitter into two photomultipliers that recorded a fluorescence signal and a reference signal. The fluorescence signal (F_>630 or F_>645) was obtained using a 630- or 645-nm long-pass filter, whereas the reference signal (F₅₄₆) was reflected excitation light obtained using a band-pass filter. The output of the photomultipliers was passed into an electronic ratio circuit to obtain a motion-corrected fluorescence ratio, usually F_>645/F₅₄₆. The fluorescence and ratio signals were filtered at settings of 30-70 Hz and recorded on a Gould Brush strip-chart recorder (Cleveland, OH).

Calibration procedure. A new method was developed for calibration of Fura Red fluorescence in situ, which could potentially be used in vivo. This method involves infusion of high- and low-calcium solutions through the same needle that was used to infuse Fura Red-AM. The first solution to be infused contained 100 mM CaCl₂ along with 1.5 µM ionomycin, 5 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES; pH 7.4), and 6% fetal calf serum to prevent binding of ionomycin to the tubing. This solution was infused at 1 ml/min until a minimum fluorescence level was achieved, which generally occurred within 1 min. The infusate was then switched to a zero-calcium solution containing 30 mM EGTA until fluorescence reached a maximum value, which typically occurred after 2 min. The zero-calcium solution contained 70 mM HEPES (substituted for NaCl) to control pH during chelation of calcium. To calculate [Ca²⁺]_i according to the standard equation, the fluorescence achieved after high-calcium solution was designated F_max and the fluorescence obtained after zero-calcium solution was designated F_min. Because fluorescence intensity decreases when calcium is bound, F_max is the lowest fluorescence level recorded and F_min is the highest level. Then, for a given fluorescence F All calculations were made using the dissociation constant (K_D) value of 190 nM, which is derived from work of Karon et al. (18). Karon et al. reported the association constant (K_A) of Fura Red as 5.27 × 10⁶ M¹. As seen from the equations below, K_A is the reciprocal of K_D The reciprocal of 5.27 × 10⁶ M¹ is 190 nM.

Mechanism of extravascular loading and calibration. The extravascular loading and calibration procedures involved injection of 4-10 ml of fluid into the tissue through a small needle. This fluid is presumably taken up by capillaries or lymphatics and leaves the heart as part of the venous effluent. The region of tissue that is stained by Fura Red can be approximated as a 6-mm-thick disk with a 5-mm radius. Given that one-third of heart tissue volume is extracellular space, then the total extracellular volume in the stained region is 0.16 ml. This fluid volume exchanges several times per minute during calibration.

Blood perfusion experiments. Just before removal of the heart from the chest, we aspirated blood from the right ventricle using a 21-gauge needle. Blood was diluted 30% with saline, and heparin was added (50 U/ml). The flask containing the blood was placed in a 30°C water bath and aerated with 95% O₂-5% CO₂. Hearts were perfused with blood only after they had been loaded with Fura Red-AM and preliminary recordings had been obtained.

Determination of signal depth for Fura Red and indo 1. To determine the tissue depth from which the Fura Red signals arise, the optical attenuation per millimeter of tissue was determined using appropriate wavelengths. Optical attenuation was measured by a double transillumination technique in which a glass tube containing Fura Red or indo 1 was inserted into the right ventricle of a saline-perfused heart. The fluorescence change produced by insertion of the dye-filled tube was compared with the signal obtained by placing the tube an equivalent distance from the fiber-optic apparatus in the absence of interposed heart tissue. Wavelengths used were 360 ± 10 (excitation) and 400 ± 10 (emission) nm for indo 1 and 546 ± 10 (excitation) and >645 (emission) nm for Fura Red. With a tube containing 0.2 mM Fura Red, transillumination of the right ventricle produced a 28-fold reduction in signal strength. If right ventricular thickness is taken as 3 mm, the effective attenuation of Fura Red fluorescence is 3.1-fold/mm of heart tissue (cube root of 28). This measurement was unsuccessful with 0.2 mM indo 1 because of insufficient signal strength, but similar measurements with brightly fluorescent latex spheres (Fluo spheres, Molecular Probes, Eugene, OR) gave an attenuation of 2,300-fold or 13.2-fold/mm of tissue. These values were used to calculate the percentage of the signal arising within a given distance from the surface.

Chemicals. Fura Red-AM was obtained from Molecular Probes. Bradykinin was obtained from Calbiochem-Novabiochem (La Jolla, CA). Ionomycin was obtained from Behring Diagnostics (La Jolla, CA). All other chemicals were obtained from Sigma Chemical (St. Louis, MO).

Statistics. Data are expressed as means ± SE. Statistical significance was determined using Student's t-test.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Properties of calcium transients. Calcium transients recorded from a Fura Red-loaded heart are shown in Fig. 1A. Transients show a brisk onset with continuous decay throughout diastole, similar to transients obtained with indo 1-AM. The transients in Fig. 1A (middle trace) are 29% of end-diastolic fluorescence. The motion artifact is typically ~5% of total signal (Fig. 1B, bottom trace). To reduce motion artifact further the F_>645-to-F₅₄₆ ratio was computed using an analog circuit (Fig. 1, A and B; top traces). The ratio can be shown to adequately cancel artifacts produced by vibration.

View larger version (19K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Fig. 1.   A: Fura Red fluorescence transients from a saline-perfused heart. Heart is illuminated at 546 nm. Middle trace, fluorescence at >645 nm (F>645). End-diastolic fluorescence level is defined as 100% (vertical scale) with 0 being signal obtained with illumination shutter closed. Calcium transients are 29% of total signal and ~40% of autofluorescence-corrected signal (see text). Top trace, F>645 divided by reflected excitation light (546 nm; F546). Bottom trace, left ventricular developed pressure in mmHg. B: calcium transients recorded during perfusion of same heart with blood. Blood is oxygenated and diluted 30% with saline. Reduction of F>645 from preblood value (preblood end-diastolic value = 100%) is caused by absorbance of excitation light. F>645 (middle trace) and F546 (bottom trace) are reduced by a comparable amount so that F>645-to-F546 ratio (top trace) is same for blood and saline perfusion. This result indicates that there is no significant change in intracellular Ca2+ concentration ([Ca2+]i) on switching from saline to blood. Pacing cycle length is increased slightly in B.

5

Estimation of subcellular compartmentation. A potential drawback of cell-permeant indicators is that some of the indicator may be trapped in organelles. An additional problem is the production of partly deesterified indicator, which is fluorescent but does not respond to changes in [Ca²⁺]_i. The extent of these problems can be studied using Mn²⁺, which abolishes the fluorescence of Fura Red and can be transported across cell membranes by ionomycin. Miyata et al. (26) found that exposure of indo 1-loaded cells to 0.l mM Mn²⁺ for 30 min produces selective quenching of indo 1 in the cytoplasm, with no effect on contraction. The effects of Mn²⁺ on a Fura Red-loaded heart are shown in Fig. 2A. Perfusion with 0.1 mM Mn²⁺ for 45 min (middle panel) abolishes the transients and reduces fluorescence to 45% of the initial end-diastolic level. Changes in left ventricular systolic pressure are minimal. Because the transients are completely abolished under these conditions, it is concluded that 55% of the end-diastolic fluorescence is due to Fura Red trapped in the cytosol, whereas the remaining 45% is due to Fura Red trapped in organelles plus Mn²⁺-insensitive fluorescence. In the right panel of Fig. 2A, the heart has been perfused with 30 mM Mn²⁺ plus 1.5 µM ionomycin and 6% fetal calf serum for an additional 2 min. High Mn²⁺ causes further reduction in fluorescence that is ascribed to quenching of Fura Red in organelles. The remaining fluorescence is ascribed to autofluorescence plus fluorescence of partly deesterified Fura Red.

View larger version (7K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Fig. 2.   A: effect of Mn2+ on Fura Red fluorescence transients. Perfusion of heart with 0.1 mM Mn2+ for 45 min abolishes transients and reduces fluorescence (>630 nm, F>630) by 55%. Contraction pressure is minimally reduced. Fluorescence is further reduced by infusion of 30 mM Mn2+ + 1.5 µM ionomycin for 2 min. Change in fluorescence from middle to right panel is due to quenching by Mn2+ of Fura Red in nuclei and mitochondria. Remaining signal is due to autofluorescence plus partially deesterified Fura Red-acetoxymethyl ester (AM). B: calibration of [Ca2+]i transients. Fura Red-loaded cells are exposed to calibration solutions in situ by infusion through same needle that was used to infuse Fura Red-AM. High-calcium solution (100 mM) containing ionomycin (1.5 µM) is infused at a rate of 1 ml/min until fluorescence reaches steady state (top traces). Infusate is then switched to calcium-free solution containing 30 mM EGTA + ionomycin. End-diastolic [Ca2+]i (shown as 100% on vertical scale) is 260 nM, and peak systolic [Ca2+]i is 500 nM (calculated using dissociation constant of 190 nM). Reflected excitation light (F546, bottom traces) barely changes during experiment.

Measurement of Fura Red washout. One factor that reduces the amount of fluorescence arising from the cytoplasm is extrusion of the dye into the extracellular space by organic anion transport. If the rate of transport (termed "washout") is rapid, the proportion of dye trapped in organelles will be relatively high. In Table 1, the rate of washout is expressed as the percentage of end-diastolic fluorescence remaining after 30, 45, 60, and 75 min (means ± SE; n = 5 hearts). Results show that there is no detectable reduction in fluorescence after 30 min of washout and a reduction of only 8% after 60 min. Measurements with indo 1-AM in both rat (29) and rabbit (R. Mohabir and W. T. Clusin, unpublished observations) hearts show that about one-third of the fluorescence is washed out in 30 min. It follows that the proportion of the dye retained in the cytoplasm may be greater for Fura Red than for indo 1. 

In situ calibration of Fura Red fluorescence transients. [Ca²⁺]_i transients can be calibrated by exposure of cells to high extracellular calcium followed by zero calcium in the presence of a calcium ionophore. To test the feasibility of calibrating calcium transients in a blood-perfused heart, we infused calibration solutions into the extracellular space through the same needle that was used to infuse Fura Red-AM. The infusion rate for calibration (1.0 ml/min) was faster than the rate for Fura Red-AM (0.35 ml/min) so that the calibration solutions would diffuse over a wider area. As shown in Fig. 2B, infusion of high-calcium solution (100 mM Ca²⁺ + 1.5 µM ionomycin) causes disappearance of the transients, together with an increase in [Ca²⁺]_i to a steady state. Subsequent infusion of a zero-calcium EGTA solution (30 mM) causes a decline in [Ca²⁺]_i to a level far below the original end-diastolic level. There is no appreciable change in reflected excitation light (Fig. 2B, bottom traces) or in developed left ventricular pressure (not shown).

Calcium transients in blood-perfused hearts. Calcium transients recorded from a heart perfused by blood are shown in Fig. 1B. Perfusion with blood causes a 32% reduction in the end-diastolic fluorescence (middle trace) and an equivalent reduction in reflected excitation light (F₅₄₆, bottom trace). There is no change in the F_>645-to-F₅₄₆ ratio signal. Because blood is transparent at wavelengths >645 nm, the stability of the ratio signal with addition of blood means that the reduction in the fluorescence signal is entirely due to increased absorption of excitation light and that there is no change in systolic or diastolic [Ca²⁺]_i. The shape of the transients is also unaffected by blood and remains constant when the perfusate is switched back to saline. Several switches can be performed with no effect on the transients.

Effect of metabolic inhibitors on [Ca^2+]_i transients. Metabolic inhibitors increase both the systolic and the diastolic level of [Ca²⁺]_i in single cardiac myocytes loaded with fluorescent calcium indicators (17). To test for this effect in intact hearts, 2,4-dinitrophenol (DNP; 0.1 mM) plus iodoacetate (IAA; 0.1 mM) was abruptly introduced into the perfusate. As shown in Fig. 3, infusion of DNP + IAA for 1 min caused a downward shift in the peak systolic and end-diastolic levels of the [Ca²⁺]_i transients, indicating a rise in systolic and end-diastolic [Ca²⁺]_i. There is a reduction in peak pressure (bottom traces) together with a rise in end-diastolic pressure. The rise in end-diastolic pressure presumably reflects the increase in [Ca²⁺]_i, whereas the reduced peak pressure is caused by reduced sensitivity of the myofilaments to calcium. The experiment in Fig. 3 was repeated in a total of 10 hearts. An increase in end-diastolic [Ca²⁺]_i was found in eight hearts (80%), whereas a net increase in peak systolic [Ca²⁺]_i was detected in five hearts (50%).

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Effect of metabolic inhibitors on calcium transients. A: control. B: heart is perfused with 0.1 mM 2,4-dinitrophenol (DNP) + 0.1 mM iodoacetic acid (IAA) saline for 1 min. Ratio of F>645 to F546 (top traces) falls, indicating an increase in systolic and end-diastolic [Ca2+]i. Left ventricular developed pressure (bottom traces) falls, and end-diastolic pressure increases. Heart is paced at 150 beats/min. No alternans is observed.

Effects of ischemia in saline-perfused hearts. The effects of 2.2 min of ischemia on a saline-perfused heart are shown in Fig. 4. The heart is paced at 150 beats/min. Ischemia produces a marked fall in peak systolic pressure (Fig. 4B, bottom trace). The falling phase of the calcium transients is slowed so that the duration of the transients at half-amplitude is longer, but there is no change in the peak systolic or end-diastolic level. Broadening of the transients is better illustrated in Fig. 5, which shows results obtained at faster speed. The upstroke of the transients is rapid and is not perceptibly changed by ischemia. The half-relaxation time is therefore approximately equal to the width of the transient at half-amplitude. Ischemia increases this value from 177 ± 4 to 227 ± 4 ms (P < 0.0001) in Fig. 5. Similar results have been obtained in 14 hearts. Ischemia sometimes produces alternans in the peak amplitude of the [Ca²⁺]_i transient (see Effects of ischemia in blood-perfused hearts). Broadening of the peak of the [Ca²⁺]_i transients by ischemia and occurrence of alternans have been observed with indo 1 (4, 24, 27). However, the absence of a change in peak systolic or end-diastolic fluorescence levels contrasts with indo 1 results. Possible reasons for this are discussed in Effects of ischemia and metabolic inhibition on Fura Red signals.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of ischemia on a saline-perfused heart. A: control. B: F>645/F546 and left ventricular developed pressure after 2.2 min of ischemia. There is no change in end-diastolic level of F>645/F546 (top traces) compared with control, but left ventricular developed pressure (bottom traces) drops dramatically. Total duration of ischemia is 2.5 min. Same experiment was performed in 14 hearts. Ischemia produced no significant change in autofluorescence in 2 hearts not loaded with Fura Red.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effect of ischemia on duration of calcium transients in a saline-perfused heart. Dotted lines show half-relaxation time of transients before (A) and after (B) 2 min of ischemia. Ischemia increases half-relaxation time from 177 ± 4 to 227 ± 4 ms (SE; P < 0.0001).

Signal depth as determinant of fluorescence response to ischemia. An important difference between ischemia and metabolic inhibition is that ischemia reduces cytosolic pH. Recent studies have shown that the K_D of tetracarboxylate calcium indicators is increased by acidification (see Ref. 23). As a result, the change in fluorescence that would result from an increase in [Ca²⁺]_i during ischemia could be cancelled by the increase in K_D. It is also known that the degree of acidification during ischemia is about twofold smaller in the subepicardium (cells <300 µm from the surface) than in midmyocardium (28). This difference is caused, at least in part, by the high diffusibility of carbon dioxide in tissue. As a result, recordings of fluorescence that were obtained solely from the subepicardium during 1-2 min of ischemia could be converted to [Ca²⁺]_i with no correction for change in intracellular pH (pH_i = 0.1; see data in Refs. 23 and 28), but this would not be true for longer periods of ischemia in cells at greater depth.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   Relationship between percentage of signal collected and heart tissue depth. Calculations are given in text. For indo 1, most of signal arises from superficial tissue, 90% of signal from within 850 µm of surface. Tissue giving rise to Fura Red signal is considerably deeper. For Fura Red, 90% of signal arises within 2,100 µm of surface.

Effects of ischemia in blood-perfused hearts. To determine whether blood perfusion modifies the effects of ischemia on [Ca²⁺]_i transients, 11 hearts were made ischemic after 2 min of perfusion with blood. Hearts were paced at 150 beats/min before and during ischemia. Effects of ischemia were similar to those in saline-perfused hearts. However, the incidence and magnitude of calcium transient alternans was much greater in the blood-perfused hearts (Fig. 7). To quantify the magnitude of [Ca²⁺]_i transient alternans in saline- versus blood-perfused hearts, an alternans ratio was defined as 1  B/A, where A is the net amplitude (end-diastolic value to peak systolic value) of the taller transient and B is the net amplitude of the smaller transient (Fig. 7). To determine the alternans ratio, the eight consecutive beats with the largest degree of alternans (i.e., 4 consecutive pairs of beats) were identified and the resulting alternans ratios were averaged. Trials having no alternans had an alternans ratio of 0. Alternans occurred (ratio > 0) in 9 of 11 blood-perfused ischemic hearts (82%) and in 6 of 14 saline-perfused hearts (43%). The mean alternans ratio was 22.8 ± 4.5% in the blood-perfused hearts versus 7.3 ± 2.6% in the saline-perfused hearts (P < 0.005). If only hearts showing alternans are included, the mean alternans ratio for the blood-perfused hearts is 27.9 ± 3.7% (n = 9) compared with 17.0 ± 2.7% (n = 6) for the saline-perfused hearts (P = 0.05). It is concluded that both the incidence and magnitude of [Ca²⁺]_i transient alternans are greater in blood-perfused ischemic hearts than in saline-perfused hearts.

View larger version (10K): [in this window] [in a new window]   Fig. 7.   Calcium transient alternans induced by ischemia in same blood-perfused heart as in Fig. 1. Heart was perfused with blood for 2 min and then rendered ischemic for 2.5 min. Recording shown is at 2.3 min of ischemia. For each pair of beats, an alternans ratio can be calculated as 1  (B/A), where B is net amplitude of smaller transient and A is net amplitude of larger transient. Alternans ratio for marked pair of beats, expressed as a percentage, is 48%. Mean alternans ratio (see text) was significantly larger in blood-perfused ischemic hearts.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Favorable loading of cardiac myocyte cytoplasm with Fura Red-AM. The initial objective of this study was to find a suitable indicator for recording calcium transients in blood-perfused hearts, in which short-wavelength indicators cannot be used. We have found that Fura Red gives excellent [Ca²⁺]_i transients in blood-perfused hearts and that it can be used at wavelengths at which tissue absorbance is not affected by oxygen desaturation. Fura Red may therefore permit acquisition of [Ca²⁺]_i transients in open-chest models of acute myocardial infarction.

Effects of ischemia and metabolic inhibition on Fura Red signals. The effects of ischemia and metabolic inhibition on Fura Red fluorescence are similar to those observed with indo 1, except for the absence of a shift in the absolute levels of the transients toward higher [Ca²⁺]_i during ischemia. A rapid increase in end-diastolic [Ca²⁺]_i during metabolic inhibition was observed previously in chick embryonic myocardial cells loaded with indo 1 (17) and is associated with less complete relaxation. The early rise in diastolic [Ca²⁺]_i during metabolic inhibition is considered to be a feature of rapidly contracting cardiac cells.

Importance of [Ca^2+]_i transient alternans in blood-perfused ischemic hearts. An important potential application of Fura Red is to study ischemic cardiac arrhythmias in vivo. The most reliable marker for impending ventricular fibrillation in the in vivo heart is the development of beat-to-beat alternans of the S-T segment. S-T segment alternans occurs just before the onset of ischemic ventricular fibrillation (15) and is known to be caused by alternations in action potential duration (8). Alternations in action potential duration are associated with alternations in contraction strength and in the amplitude of the [Ca²⁺]_i transient (24). Alternans of the S-T segment or action potential duration can be localized to small regions of myocardium and can be out of phase (discordant) in adjacent regions (21). Localized or discordant alternans would produce dispersion of refractoriness that is believed to be critical in the genesis of ventricular fibrillation.

Relation of [Ca^2+]_i transient alternans to calcium-activated ion currents. One reason that [Ca²⁺]_i transient alternans could be considered a marker for fluctuation of action potential duration is that some of the ion currents that govern action potential duration are activated by [Ca²⁺]_i. A rise in [Ca²⁺]_i can produce inward current through nonselective cation channels or through electrogenic sodium/calcium exchange. Based on these mechanisms, a larger [Ca²⁺]_i transient would produce a longer plateau. More recently, cardiac chloride (30) and potassium (33) currents activated by [Ca²⁺]_i have been described. Increased activation of these currents would abbreviate the plateau. Because the distribution of ion currents varies in different types of cardiac tissue, the relation between [Ca²⁺]_i transient amplitude and action potential duration may not always be the same. Lee et al. (24) used a monophasic action potential (MAP) electrode to record action potentials in indo 1-loaded rabbit hearts. When the MAP electrode was placed immediately adjacent to the fiber-optic probe, the taller [Ca²⁺]_i transients during alternans were associated with a longer action potential plateau. Fluctuations in action potential duration were therefore ascribed to variations in calcium-activated inward current.