INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

INCREASED EXTRACELLULAR MG²⁺ concentration ([Mg²⁺]_o) has protective effects during reperfusion of the ischemic human heart (9, 43) and in different experimental protocols of ischemia-reperfusion (5, 12). However, the underlying mechanisms of action of elevated [Mg²⁺]_o are still obscure. Ca²⁺ overload (in addition to oxidative stress) is probably a major factor causing tissue injury during ischemia/hypoxia and reperfusion/reoxygenation of the heart (6, 39), and strategies to reduce Ca²⁺ overload have been shown to protect the myocardium in experimental models of reperfusion and reoxygenation (21, 42). Intracellular organelles like the sarcoplasmatic reticulum (SR) and mitochondria might accumulate excess Ca²⁺ during reperfusion/reoxygenation. Data from Lochner et al. (23) showed that Ca²⁺ content was increased two- to threefold in mitochondria isolated from the reperfused heart. The role of the SR is more uncertain because Ca²⁺-ATPase activity of the SR (the mechanism responsible for Ca²⁺ uptake to the SR) is actually reduced during reoxygenation and reperfusion (44). Both Ca²⁺ overload and a fall in mitochondrial membrane potential (_m) may induce mitochondrial permeability transition (mPT), which is an initiating step for apoptosis and necrosis (20). Mg²⁺, being known as nature's own Ca²⁺ antagonist, has shown inhibitory action on both Ca²⁺ channels (1, 11) and Na⁺/Ca²⁺ exchange (17, 40). Beneficial effects of high Mg²⁺ concentrations in cardioplegic solutions have been associated with reduced free cytosolic Ca²⁺ concentration ([Ca²⁺]_i) in a simulated ischemia model with neonatal rat myocytes (14) and in an ischemia-reperfusion model with both mature and aged rabbit hearts (41). Mg²⁺ has been shown to reduce binding of Ca²⁺ to the mitochondrial membrane and to inhibit mPT (19). Recently, in a report (34) from our laboratory, we showed that elevation of [Mg²⁺]_o to 5 mM reduced cell Ca²⁺ accumulation during reoxygenation of hypoxic rat cardiomyocytes. This effect could be due to decreased influx and/or increased efflux of Ca²⁺ by Mg²⁺. The first goal of the present study was consequently to test the effect of increased [Mg²⁺]_o on the influx and efflux kinetics of Ca²⁺ during reoxygenation of hypoxic cardiomyocytes. The second goal was to test whether the inhibitory effect of increased [Mg²⁺]_o on Ca²⁺ accumulation occurs with a reduction in [Ca²⁺]_i and is linked to an increase in free cytosolic Mg²⁺ concentration ([Mg²⁺]_i). The third goal was to investigate the effect of increased [Mg²⁺]_o on Ca²⁺ accumulation in cells treated with an inhibitor of SR or mitochondrial Ca²⁺ transport in an attempt to identify the locus (loci) of action. The latter was also the reason for investigating the effect of increased [Mg²⁺]_o on _m during reoxygenation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals and materials. The following chemicals were obtained from Sigma (St. Louis, MO): BSA (essentially fatty acid-free), DL-carnitine, dibutyl phthalate, ouabain, trypsin, cyclopiazonic acid, calcimycin, and SDS. Joklik's minimum essential medium (MEM) was obtained from Life Technologies (Paisley, UK). Collagenase and deoxyribonuclease were obtained from Worthington Biochemical (Lakewood, NJ). ⁴⁵Ca²⁺ was obtained from DuPont de Nemours, NEN Division (Dreiech, Germany). Micro bicinchoninic acid (BCA) Protein Assay Reagent was obtained from Pierce (Rockford, IL). 3,4-Dichlorobenzamil (DCB), fluo 3-AM, magnesium green (MgG), calcium calibration buffer kit No. 2, and 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) were obtained from Molecular Probes (Eugene, OR). Di-isononyl phthalate was obtained from Fluka Chemie (Buchs, Switzerland). Opti-Fluor was obtained from Packard (Groningen, The Netherlands). Solutions for calibration of the MgG signal were made due to a recipe obtained from Molecular Probes. DMSO was obtained from Merck (Darmstadt, Germany). Clonazepam and polyethylene glycol (PEG) 6000 were obtained from Norwegian Medicinal Depot (Oslo, Norway).

Buffers and solutions. Normal physiological buffer (NPB) was composed of (in mM) 120.0 NaCl, 3.3 KCl, 1.2 KH₂PO₄, and 24.0 NaHCO₃; pH 7.4. NPB-1 contained NPB supplemented with 0.5 mM CaCl₂, 0.8 mM MgSO₄, and 1% (wt/vol) BSA. NPB-2 contained NPB supplemented with 1.0 mM CaCl₂, 0.8 mM MgSO₄, and 0.1% (wt/vol) BSA. Aliquots of MgSO₄ (1 M, aqueous solution) was added to the incubation buffer to increase [Mg²⁺]_o at the times indicated. The stock solution of DCB (5 mM) was made in PEG (20% wt/vol, aqueous solution) and further diluted to 1:500 in the incubation buffer at the time of addition (10 µM DCB and 0.04% PEG). Clonazepam was dissolved in DMSO (100 mM) and diluted further to 1:1,000 in the incubation buffer at the start of experiments (100 µM clonazepam and 0.1% DMSO). In the experiments in which clonazepam was tested, 0.1% DMSO was added to all test conditions to exclude any effects of DMSO alone. The stock solution of JC-1 was made (7.7 mM in DMSO) and kept at 20°C. Before measurements were made, the JC-1 stock solution was diluted to 1:100 with NPB-2 and added to the cell suspension to give the final concentration of 12 µM JC-1 (0.13% DMSO).

Isolation of cardiomyocytes. Adult male Wistar rats (200-400 g) were obtained from Møllegaard (Skensved, Denmark) and housed in accordance with the conditions set by the Norwegian Council for Animal Research. The investigation conformed with the guidelines of the Norwegian National Institute for Public Health. Cardiomyocytes were isolated by trypsin-collagenase perfusion as previously described (33). Cell suspensions used for experiments contained at least 70% viable cells (trypan blue exclusion), of which at least 95% had an elongated shape. Cell protein was measured according to Smith et al. (36) by Micro BCA Protein Assay Reagent using BSA as the standard.

Incubation. Cells were incubated in 25-ml Erlenmeyer flasks in NPB-2 supplemented with trace amounts of ⁴⁵Ca²⁺ (5 × 10⁶ disintegrations · min¹ · ml¹) in a 37°C water bath. In normoxia (control cells), the cell suspension was gassed with 95% air-5% CO₂ (PO₂  12-14 kPa) and supplied with 5.5 mM glucose. In hypoxia, cells were added to NPB-2 (no glucose added) equilibrated with 95% N₂-5% CO₂ (PO₂  1 kPa). At reoxygenation, gas was changed to 95% air-5% CO₂, and buffer was supplied with 5.5 mM glucose. Oxygen tension in the test flasks was checked using an oxygen electrode (WTW Microprocessor Oximeter Oxi 96, Wissenschaftliche Technische Werkstätten; Weilheim, Germany) near the bottom of the Erlenmeyer flask whereto cells were added. Each Erlenmeyer flask was supplied with an inlet and outlet for gas and continuously gassed for the indicated time.

Determination of cell Ca²+ and influx kinetics of Ca²⁺. Cell Ca²⁺ was determined as rapidly exchangeable Ca²⁺ by uptake of ⁴⁵Ca²⁺ and calculated from the ⁴⁵Ca²⁺ content of the cells and the known specific activity of ⁴⁵Ca²⁺ in the uptake buffer. Cell membrane Ca²⁺ transport was stopped by pipetting samples into centrifuge tubes with ice-cold buffer layered over an oil mixture, and cells were separated from buffer by subsequent centrifugation as previously described (34). The data were normalized to the protein content in each cell pellet. The Ca²⁺ influx rate was determined from the initial, apparently linear uptake of ⁴⁵Ca²⁺ after the addition of isotope and exchangeable cell Ca²⁺ from the apparent equilibrium of ⁴⁵Ca²⁺ uptake (plateau level).

Determination of efflux kinetics of Ca²+. Decline of ⁴⁵Ca²⁺ content in cells exposed to hypoxia and reoxygenation was used to evaluate Ca²⁺ efflux. Briefly, cells were loaded with trace amounts of ⁴⁵Ca²⁺ during 1 h of hypoxia and 0.5 h of reoxygenation. The cell suspension was then centrifuged (250 g for 30 s), the supernatant was aspirated, and the cells were resuspended in oxygenated buffer (37°C) to dilute extracellular ⁴⁵Ca²⁺ and start efflux of the isotope. After 5 min, the centrifugation and resuspension steps were repeated. For each dilution step, the extracellular isotope was diluted by a factor of ~10. To characterize the kinetics of the decline, the radioactivity remaining in the cells was normalized to protein content and subtracted the calculated equilibrium value (the prewash value divided by the product of both dilution factors).

Flow cytometric analysis of [Ca²+]_i and [Mg²⁺]_i. Levels of [Ca²⁺]_i and [Mg²⁺]_i were measured by flow cytometry as the fluorescence of fluo 3 and MgG, respectively. Fluo 3-AM and MgG are membrane permeable. After entering the cells, these esters are cleaved by intracellular esterases to yield the relatively cell-impermeant fluorescent indicators fluo 3 and MgG, respectively. Fluo 3-AM and MgG were dissolved in DMSO right before each experiment, added to the incubation buffer (NPB-2) to give the final concentration of 4.5 µM fluo 3 or 5 µM MgG (0.5% DMSO), and were present during the whole experiment. Samples were directly applied to the flow cytometer without any washing step to remove the extracellular indicators. Preliminary testing showed that the level of fluorescence in all groups (normoxic control, hypoxia/reoxygenation ± Mg²⁺) was slightly reduced, and to the same degree by a washing step, probably due to some loss of probe from the cells. A FACSort (Becton-Dickinson; Rutherford, NJ) flow cytometer equipped with a 488-nm argon ion laser and supplied with the Cell Quest software was used to measure fluorescence signals of fluo 3 and MgG in the cells. Fluo 3 and MgG, unlike ultraviolet light-excited indicators (e.g., fura 2), do not exert any spectral shift. Signals were obtained using a 530-nm bandpass filter (FL-1 channel) for both indicators. Each determination was based on a mean fluorescence intensity of 5,000 cells in arbitrary units. The fluorescence intensity of unloaded cells (background signal) of both indicators was ~2% of signals of indicator-loaded cells.

Calibration of fluo 3 and MgG fluorescence. Cells were loaded with the indicator for 1 h in NPB-2 (normoxic, 37°C and 5.5 mM glucose). The cell suspension was then centrifuged (250 g for 30 s). The cell pellet was dispersed in 25-ml NPB-2 (nominally Ca²⁺ and Mg²⁺ free) and centrifuged (250 g for 30 s). The cell pellet was then washed twice with 5 ml of zero-Ca²⁺ buffer (calcium calibration buffer kit No. 2, Molecular Probes) in the case of fluo 3-loaded cells and Mg²⁺ calibration buffer component A (see MATERIALS AND METHODS) for MgG-loaded cells. Fluo 3-loaded cardiomyocytes were then exposed to various Ca²⁺ concentrations (0, 58, 135, 315, and 1,215 nM). [Ca²⁺]_i was equilibrated with extracellular Ca²⁺ by addition of the ionophore calcimycin (50 µM). MgG-loaded cells were exposed to various Mg²⁺ concentrations (0, 0.18, 0.36, 0.72, 1.44, 2.88, and 5.76 mM). [Mg²⁺]_i was equilibrated with extracellular Mg²⁺ by the addition of calcimycin (50 µM). Different Mg²⁺ concentrations were made by mixing Mg²⁺ calibration buffer components A and B [component A was composed of (in mM) 115 KCl, 20 NaCl, and 10 Tris (pH 7); and component B was composed of (in mM) 35 MgCl₂, 115 KCl, 20 NaCl, and 10 Tris (pH 7.05)]. These were made using the recipe of the Mg²⁺ calibration standard kit (M-3120), which is no longer produced by Molecular Probes. Each determination was based on a mean fluorescence intensity of 5,000 cells in arbitrary units. The particular indicator (4.5 µM fluo 3 or 5 µM MgG) was present in all solutions used in the signal calibration protocol in likeness with the experiments. Ca²⁺ binding [dissociation constant (K_d) of 325 nM] results an increase in fluorescence intensity of fluo 3 by ~100-fold. Fluo 3 has an eight times higher affinity for Ca²⁺ than Mg²⁺ (data given by Molecular Probes). However, MgG, like other tricarboxylate aminophenol triacetic acid chelators, binds Ca²⁺ with high affinity. The interference with Mg²⁺ measurements due to Ca²⁺ binding becomes significant when Ca²⁺ concentrations exceed ~1 µM (13, 18). Leyssens et al. (22) found that the K_d of MgG for Ca²⁺ was 4.7 µM in rat cardiomyocytes, whereas the K_d of MgG for Mg²⁺ is ~1 mM. Consequently, we did exactly the same calibration experiments mentioned above but incubated instead the MgG-loaded cells with Ca²⁺ calibration buffers and fluo 3-loaded cells with Mg²⁺ calibration buffers.

_m measurements with JC-1. The dye JC-1 has recently been evaluated as an optimal dye for measuring _m in cardiomyocytes (24). JC-1 is a lipophilic and cationic dye that exhibits potential-dependent accumulation in negatively charged mitochondria. At low concentrations (low _m), JC-1 exists mainly in a monomeric form, which emits green fluorescence. JC-1 at high concentrations (high _m) forms aggregates called "J" complexes, which emit red fluorescence at 590 nm. Thus a reduction in the ratio of red to green fluorescence indicates a fall in _m. Cells were loaded with 12 µM JC-1 for 10 min at 37°C and then immediately applied to the flow cytometer for signal recording using the 530-nm (FL-1 channel) and 585-nm (FL-2 channel) bandpass filters simultaneously. Each determination was based on the ratio of red to green mean fluorescence intensities measured in arbitrary units from 5,000 cells.

Statistics. All experiments were performed as paired comparisons. Statistical analysis was performed using Statgraphics Plus software (version 4.0, Manugistics; Rockville, MD). Statistical analysis of multigroup comparisons was performed by one-way ANOVA, and the method to discriminate among the means was Fisher's least significant difference (LSD) method. Statistical analysis of two-sample comparisons was performed by Student's t-test (two-sided). P < 0.05 was considered to be significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell Ca²+ during normoxia and reoxygenation of hypoxic cardiomyocytes. Cell Ca²⁺ was in the range of 2.5-3 nmol/mg protein at apparent equilibrium in the normoxic cells. ⁴⁵Ca²⁺ uptake reached a plateau after 1-2 min irrespective of adding ⁴⁵Ca²⁺ to the buffer at 10 or 120 min and was virtually unchanged during the time course of the experiments (Fig. 1). In cells exposed to hypoxia and reoxygenation, cell Ca²⁺ significantly increased after reoxygenation from 3 to 12 nmol/mg protein (Fig. 1). The equilibrium values for exchangeable cell Ca²⁺ in reoxygented cells at 180 min were similar irrespective of adding ⁴⁵Ca²⁺ at the start of incubation or at 120 min (Fig. 1). ⁴⁵Ca²⁺ apparently equilibrated more slowly with the pool of exchangeable Ca²⁺ in reoxygenated cells than in normoxic cells at 120 min (Fig. 1, inset).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Cell Ca2+ during normoxia and reoxygenation of hypoxic cardiomyocytes. 45Ca2+ was added 10 min after start of experiments to the incubation buffer of normoxic control cells () or added from start to cells exposed to 1 h of hypoxia and 2 h of reoxygenation (). Alternatively, 45Ca2+ was added after 2 h to control cells () and cells exposed to 1 h of hypoxia and 1 h of reoxygenation (X). Each point represents the mean ± SE of 4-8 determinations from 2 different cell isolations. Inset: plot on a log vertical scale of the uptake kinetics of Ca2+ in control cells () and hypoxia-reoxygenation (H/R)-treated cells (X) when 45Ca2+ was added at 2 h, expressed as the percent remaining to be taken up as a function of time.

Effect of increased [Mg²+]_o on cell Ca²⁺ during hypoxia and reoxygenation. To investigate whether the effect of Mg²⁺ was exerted during hypoxia or reoxygenation, [Mg²⁺]_o was increased to 5 mM either from the start of the experiments or at the onset of reoxygenation. Both treatments reduced reoxygenation-mediated Ca²⁺ accumulation to the same degree (50%) compared with hypoxia- and reoxygenation-treated cells in the presence of normal [Mg²⁺]_o (0.8 mM). Lowering [Mg²⁺]_o to 0.3 mM during hypoxia-reoxygenation did not alter cell Ca²⁺ accumulation compared with normal [Mg²⁺]_o (0.8 mM) (Fig. 2).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Effect of increased extracellular Mg2+ concentration ([Mg2+]o) on cell Ca2+ during hypoxia and reoxygenation. Cells were exposed to 1 h of hypoxia and subsequent reoxygenation. [Mg2+]o was increased from 0.8 to 5 mM either at the start of experiments or at the onset of reoxygenation. A study group with low [Mg2+]o (0.3 mM, throughout experiment) was also included. Each column represents the mean ± SE of 6-12 determinations from 2 different cell isolations. *P < 0.05 vs. hypoxia and reoxygenation by ANOVA and the least-significant difference (LSD) method.

Effect of increased [Mg²+]_o on Ca²⁺ influx during reoxygenation. Trace amounts of ⁴⁵Ca²⁺ were added to cells exposed to 60 min of hypoxia and 30 min of reoxygenation. [Mg²⁺]_o increased to 5 mM at the onset of reoxygenation significantly reduced the initial rate of ⁴⁵Ca²⁺ uptake (up to 5 min), reflecting a reduced influx velocity of Ca²⁺ (Fig. 3A). Because ⁴⁵Ca²⁺ uptake was apparently not linear for 5 min (Fig. 3A, inset), a separate set of experiments was performed (Fig. 3B) with sample collections at 2 and 3 min as well. These experiments confirmed that the initial rate of ⁴⁵Ca²⁺ uptake was reduced by elevated [Mg²⁺]_o. The Na⁺/Ca²⁺ exchange inhibitor DCB (10 µM, added at the onset of reoxygenation) also significantly reduced the influx velocity of ⁴⁵Ca²⁺ (Fig. 3, A, inset, and B).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effect of increased [Mg2+]o and 3,4-dichlorobenzamil (DCB) on Ca2+ influx during reoxygenation. Cells were exposed to 1 h of hypoxia and subsequent reoxygenation with no supplements () or with 5 mM Mg2+ () or 10 µM DCB (X) added at the onset of reoxygenation. 45Ca2+ was added at 90 min, and samples were collected subsequently. In normoxic control cells (), 45Ca2+ was added from the start of experiments. A: whole time course of the experiments (inset, time points from 90 to 95 min). B: separate set of experiments (only time points from 90 to 95 min). Each point represents the mean ± SE of 6-12 and 4-8 determinations from 3 and 2 different cell isolations in A and B, respectively. *P < 0.05 vs. hypoxia-reoxygenation () by ANOVA and the LSD method.

Effect of increased [Mg²+]_o on Ca²⁺ efflux during reoxygenation. To study the effect of elevated [Mg²⁺]_o (5 mM, added at reoxygenation) on Ca²⁺ efflux, cells were loaded for 90 min with ⁴⁵Ca²⁺ (60 min of hypoxia and 30 min of reoxygenation). The buffer was then diluted ~100-fold by centrifugation and resuspension. The fractions of cellular ⁴⁵Ca²⁺ (of start value) remaining after 95, 120, and 180 min were 84, 74, and 40% and 96, 73, and 41% in the cells exposed to hypoxia-reoxygenation in the absence and presence of increased [Mg²⁺]_o, respectively, showing similar efflux kinetics (Fig. 4A). Normoxic control cells loaded with ⁴⁵Ca²⁺ for 90 min and subjected to buffer dilution also showed the same pattern of decline of isotope content in the presence of increased [Mg²⁺]_o (5 mM) added at 60 min as in the presence of normal [Mg²⁺]_o (0.8 mM; Fig. 4B).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of increased [Mg2+]o at the onset of reoxygenation on Ca2+ efflux. Cells were exposed to 1 h of hypoxia and subsequent reoxygenation. Cells were loaded with 45Ca2+ from start of experiments, and the content of 45Ca2+ was determined after diluting extracellular 45Ca2+ at 30 min of reoxygenation. A: cells exposed to 1 h of hypoxia and subsequent reoxygenation with 0.8 () and 5 mM [Mg2+]o (). B: normoxic control cells with 0.8 () vs. 5 mM [Mg2+]o (). Values are given as the cellular content of 45Ca2+ in counts per minute (cpm) per milligram of protein. Each point represents the mean of 7-14 determinations from 4 different cell isolations in A and 2-4 determinations from 2 different cell isolations in B.

[Ca²+]_i during hypoxia and reoxygenation and effect of increased [Mg²⁺]_o. The relationship between [Ca²⁺]_i and fluorescence intensity of fluo 3 in calcimycin-permeablized cells is shown in Fig. 5. The relationship was time dependent, approaching a stable level after 180 min of permeabilization. Figure 5A, inset, shows that fluo 3 fluorescence in calcimycin-permeablized cells was not influenced by varying [Mg²⁺]_o. Compared with normoxia, fluo 3 fluorescence intensity was increased 4.5-fold at the end of 60 min of hypoxia, corresponding to an increase in [Ca²⁺]_i from 60 to 300 nM (Fig. 6A). The fluorescence signal of fluo 3 rapidly fell at reoxygenation and by the end of experiments was ~30% higher than control cells, corresponding to a [Ca²⁺]_i of 125 nM in reoxygenated cells and 90 nM in normoxic cells, respectively. Increasing [Mg²⁺]_o to 5 mM at the onset of reoxygenation did not affect the fluo 3 signal during reoxygenation (Fig. 6B). The hypoxic group, which received DCB at the onset of reoxygenation, tended to have ~30-40% lower fluo 3 signal than control hypoxia-reoxygenation; however, the difference was not statistically significant. Ouabain-treated cells exposed to hypoxia and reoxygenation showed a similar increase in fluo 3 fluorescence signal at the end of 60 min of hypoxia as without ouabain, but no significant fall in fluo 3 fluorescence was achieved during reoxygenation. Elevated [Mg²⁺]_o during reoxygenation of hypoxic and ouabain-treated cells had no effect on fluo 3 fluorescence.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Calibration of the fluo 3 signal. Fluo 3 fluorescense [in arbitrary units (a.u.)] was measured in cells loaded with fluo 3-AM for 1 h in normoxic condition before permeabilization with calcimycin (50 µM) in the presence of different Ca2+ concentrations for 20 min (), 1 h (), 2 h (×), and 3 h (). Each point represents the mean of 3 determinations from 3 different cell isolations. Inset: fluo 3 measurements in cells loaded with fluo 3-AM for 1 h in normoxic conditions before permeabilization with calcimycin (50 µM) in the presence of different Mg2+ concentrations for 3 h (). Each point represents the mean of 2 determinations from 2 different cell isolations.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   A: free cystolic Ca2+ concentration ([Ca2+]i) during hypoxia and reoxygenation. Shown is the fluo 3 fluorescence measured in cells exposed to 1 h of hypoxia and 2 h of reoxygenation () compared with normoxic control cells (). Each point represents the mean ± SE of 6 determinations from 4 different cell isolations. *P < 0.05 vs. normoxic control by paired Student's t-test. B: effect of elevated [Mg2+]o and DCB on [Ca2+]i during reoxygenation. Fluo 3 fluorescence was measured in cells exposed to 1 h of hypoxia and 2 h of reoxygenation with no supplements (), with 5 mM [Mg2+]o (), and with 10 µM DCB (X) as a percentage of normoxic control cells (). Fluo 3 fluorescence was also measured in cells exposed to 1 h of hypoxia and 2 h of reoxygenation in the presence of ouabain (1 mM, added from start) with no other supplements () or with 5 mM [Mg2+]o () added at the onset of reoxygenation. Each point represents the mean of 3-4 determinations from 4 different cell isolations.

[Mg²+]_i in hypoxia and reoxygenation and effect of increased [Mg²⁺]_o. The relationship between [Mg²⁺]_i and fluorescence intensity of MgG in calcimycin-permeablized cells is shown in Fig. 7A. The relationship approached a stable level after 180 min of permeabilization. Figure 7A, inset, shows that MgG fluorescence in calcimycin-permeablized cells is less sensitive to increasing Ca²⁺ than to increasing Mg²⁺ concentrations. MgG fluorescence intensity was increased twofold compared with normoxic control cells at the end of hypoxia (60 min), showing an increased [Mg²⁺]_i at the end of the hypoxic period (Fig. 7B). The fluorescence signal of MgG started to fall at the onset of reoxygenation and by the end of the experiments was ~20% higher than in normoxic control cells. In cells exposed to hypoxia-reoxygenation in the presence of 5 mM [Mg²⁺]_o, mean MgG fluorescence intensity tended to be higher, but no significant difference was found.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   A: calibration of the magnesium green (MgG) signal. MgG fluorescence was measured in cells loaded with MgG-AM for 1 h in normoxic condition before permeabilization with calcimycin (50 µM) in the presence of different Mg2+ concentrations for 3 h (). Each point represents the mean of 3 determinations from 3 different cell isolations. Inset: MgG measurements in cells loaded with MgG-AM for 1 h in normoxic conditions before permeabilization with calcimycin (50 µM) in the presence of different Ca2+ concentrations for 3 h (). Each point represents the mean of 2 determinations from 2 different cell isolations. B: free cystolic Mg2+ concentration in hypoxia and reoxygenation and the effect of increased [Mg2+]o. MgG fluorescence was measured in cells exposed to 1 h of hypoxia and 2 h of reoxygenation with no supplements () and with 5 mM [Mg2+]o () as a percentage of normoxic control cells (). Each point represents the mean ± SE of 10 determinations from 5 different cell isolations. *P < 0.05 vs. normoxic cells by ANOVA and the LSD method.

Effect of increased [Mg²+]_o in combination with clonazepam or cyclopiazonic acid on Ca²⁺ accumulation. The effect of clonazepam and cyclopiazonic acid on cell Ca²⁺ was tested to investigate the possible role of the mitochondria and SR during reoxygenation. Clonazepam (100 µM, added at start) reduced the elevation in cell Ca²⁺ by 85% at 2 h and 70% at 3 h. Increased [Mg²⁺]_o significantly increased the inhibitory effect on Ca²⁺ accumulation when combined with clonazepam (Fig. 8A). The effect of cyclopiazonic acid (SR Ca²⁺-ATPase inhibitor) on hypoxia-reoxygenation-mediated Ca²⁺ accumulation was also studied. Surprisingly, cyclopiazonic acid (20 µM, added at start) significantly amplified the reoxygenation-induced increase in cell Ca²⁺ by 45% at 180 min compared with hypoxia-reoxygenation without cyclopiazonic acid. This cyclopiazonic acid-induced increase in cell Ca²⁺ was abolished by increased [Mg²⁺]_o (5 mM, added at reoxygenation). Actually, the cells treated with the cyclopiazonic acid and Mg²⁺ had significantly lower values than cells exposed to hypoxia-reoxygenation without treatment (Fig. 8B).

View larger version (23K): [in this window] [in a new window]   Fig. 8.   A: effect of increased [Mg2+]o, clonazepam, and their combination on Ca2+ accumulation. Cell Ca2+ in normoxic control cells () and cells exposed to 1 h of hypoxia and 2 h of reoxygenation with no supplements () or with 5 mM [Mg2+]o (), 100 µM clonazepam (), and 5 mM [Mg2+]o in combination with 100 µM clonazepam () are shown. [Mg2+]o was increased at the onset of reoxygenation, and clonazepam was added from start. Each point represents the mean ± SE of 22-24 determinations from 4 different cell isolations. *P < 0.05 vs. hypoxia-reoxygenation () and P < 0.05 vs. hypoxia-reoxygenation with clonazepam () at the respective incubation times by ANOVA and the LSD method. B: effect of increased [Mg2+]o, cyclopiazonic acid, and their combination on Ca2+ accumulation. Cell Ca2+ in control cells () and cells exposed to 1 h of hypoxia and 2 h of reoxygenation with no supplements () or with 5 mM [Mg2+]o (), 20 µM cyclopiazonic acid (), and 5 mM [Mg2+]o in combination with 20 µM cyclopiazonic acid () are shown. [Mg2+]o was increased at the onset of reoxygenation, and cyclopiazonic acid was added from start. Each point represents the mean ± SE of 10-12 determinations from 3 different cell isolations. *P < 0.05 vs. hypoxia-reoxygenation () at the respective incubation times by ANOVA and the LSD method.

Effect of increased [Mg²+]_o, clonazepam, and their combination on _m during reoxygenation. The ratio of red to green fluorescence of JC-1 showed linear dose dependence up to 20 µM JC-1 after 10 min of normoxic incubation (R² = 0.953, linear regression). Cyanide (5 mM) reduced markedly the red-to-green fluorescence ratio of JC-1 after 10 and 60 min of normoxic incubation with JC-1, as shown in Fig. 9. The effect of clonazepam (100 µM, added at start) and increased [Mg²⁺]_o (supplemented at reoxygenation) on _m was investigated in cells exposed to 1 h of hypoxia and 1 h of reoxygenation using the JC-1 fluorescence ratio. Reoxygenated cells had a 24% lower red-to-green fluorescence ratio than normoxic cells, indicating a fall in _m. Cells exposed to hypoxia and reoxygenation in the presence of 5 mM [Mg²⁺]_o showed a significantly higher (17 ± 3%) red-to-green fluorescence ratio of JC-1 than control hypoxia-reoxygenation-treated cells. Reoxygenated cells in the presence of 100 µM clonazepam (added at start of experiments) showed 40% higher and the combination of clonazepam and elevated [Mg²⁺]_o showed a 50% higher red-to-green fluorescence ratio of JC-1, respectively, compared with control reoxygenated cells (Fig. 10A). The red-to-green fluorescence ratio of JC-1 was inverse correlated (R² = 0.975, P < 0.01) to cell Ca²⁺ in hypoxia-reoxygenation alone or after treatment with Mg²⁺, clonazepam, or the combination of both agents as shown in Fig. 10B.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Effect of cyanide (CN; 5 mM) on the red-to-green fluorescence ratio of JC-1 after 10 and 60 min of incubation (cells were incubated with cyanide throughout the indicated time). Each point represents the mean of 2 determinations from 2 different cell isolations.

View larger version (22K): [in this window] [in a new window]   Fig. 10.   A: effect of increased [Mg2+]o, clonazepam, and their combination on mitochondrial membrane potential (m) during reoxygenation. Cells were incubated with JC-1 for 10 min. The ratio of red to green fluorescence emitted from the cells was measured and normalized to values from cells exposed to 1 h of hypoxia and 1 h of reoxygenation without treatment. Dashed line, values obtained from normoxic control cells at the same time. Each column represents the mean ± SE of 12-14 determinations from 7 different cell isolations. *P < 0.05 vs. hypoxia-reoxygenation by ANOVA and the LSD method. B: correlation between the red-to-green fluorescence ratio of JC-1 and cell Ca2+ in cells exposed to 1 h of hypoxia and 1 h of reoxygenation without treatment or with 5 mM [Mg2+]o, 100 µM clonazepam, and 5 mM [Mg2+]o in combination with 100 µM clonazeopam. (Values are taken from Fig. 10A and 8A, respectively.)

Cell size and granularity. Forward scatter (FSC) and side scatter (SSC) were determined by flow cytometry to characterize the effect of [Mg²⁺]_o and clonazepam on cell size and granularity. Elevated [Mg²⁺]_o did not alter FSC (Fig. 11A) or SSC signals (Fig. 11B); however, clonazepam and the combination of clonazepam and elevated [Mg²⁺]_o showed significant attenuation of the reduction of FSC and the increase in SSC induced by hypoxia and reoxygenation, the combination being most effective. Clonazepam also reduced lactate dehydrogenase (LDH) release by 70% (7 determinations from two different cell isolations, P < 0.05) measured 1 h after reoxygenation.

View larger version (22K): [in this window] [in a new window]   Fig. 11.   Effect of increased [Mg2+]o, clonazepam, and their combination on flow cytometrically determined forward scatter (FSC; A) and side scatter (SSC; B) of cells exposed to 1 h of hypoxia and 2 h of reoxygenation with no supplements () or with 5 mM [Mg2+]o (), 100 µM clonazepam (), 5 mM [Mg2+]o in combination with 100 µM clonazepam (), and normoxic control cells (). [Mg2+]o was increased at the onset of reoxygenation, and clonazepam was added from start. Each point represents the mean ± SE of 6 determinations from 2 different cell isolations. *P < 0.05 vs. hypoxia-reoxygenation () at the respective incubation times by ANOVA and the LSD method.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study was designed to elucidate the mechanism(s) behind the effect of increased [Mg²⁺]_o on Ca²⁺ accumulation during reoxygenation of hypoxic cardiomyocytes. By increasing [Mg²⁺]_o to 5 mM at reoxygenation, Ca²⁺ accumulation was reduced by 40-50%. Ca²⁺ influx was significantly reduced, whereas Ca²⁺ efflux kinetics were unaltered, and [Ca²⁺]_i and [Mg²⁺]_i were not detectably affected. Increased [Mg²⁺]_o also reduced the reoxygenation-induced fall in _m assessed by JC-1 fluorescence and potentiated the effect of clonazepam on Ca²⁺ accumulation, _m, FSC, and SSC.

Cell Ca²+ in hypoxia and reoxygenation. The marked increase in cell Ca²⁺ after hypoxia and reoxygenation observed in this study is in agreement with previous reports (8, 25, 33). The present results showing rapidly attained equilibrium of ⁴⁵Ca²⁺ in normoxic cells within 2-5 min of addition to the extracellular buffer also are in agreement with the literature (3, 32). The similarity in equilibration pattern of normoxic cells with ⁴⁵Ca²⁺ added either 10 min after start of incubation or 2 h after start of incubation indicates that Ca²⁺ influx and efflux mechanisms remained intact at normoxia during the experiment. Also, uptake of ⁴⁵Ca²⁺ added to cells after exposure to 1 h of hypoxia and 1 h of reoxygenation approached the same value of equilibrium (exchangeable) cell Ca²⁺ (at 3 h) as in cells exposed to the same protocol of hypoxia and reoxygenation and with ⁴⁵Ca²⁺ added at the start of the experiment. This finding indicates that ⁴⁵Ca²⁺ equilibrates with the same exchangeable pool of Ca²⁺ also when added after increasing the size of the pool by hypoxia and reoxygenation and that the availability of the Ca²⁺ pool for exchange with ⁴⁵Ca²⁺ was maintained. The longer time needed to equilibrate exchangeable Ca²⁺ with ⁴⁵Ca²⁺ in hypoxia- and reoxygenation-treated cells than in normoxia could be due partly to the increase in size of the pool, partly to slower sarcolemmal influx and efflux rates of Ca²⁺, and possibly also to slower exchange of Ca²⁺ between the cytosol and intracellular stores.

Effect of increased [Mg²+]_o on cell Ca²⁺ accumulation, Ca²⁺ influx, and Ca²⁺ efflux. The observed 40-50% reduction in reoxygenation-induced Ca²⁺ accumulation by increased [Mg²⁺]_o is in agreement with our previous data (34). The effect of increased [Mg²⁺]_o on Ca²⁺ accumulation during reoxygenation was similar irrespective of whether [Mg²⁺]_o was increased from the start of hypoxia or increased at the onset of reoxygenation, showing that the effect of Mg²⁺ was exerted at reoxygenation and not during hypoxia. Our results showing significant reduction of influx velocity of Ca²⁺ and no effect on the pattern of decline of ⁴⁵Ca²⁺ in cells exposed to hypoxia and subsequent reoxygenation indicate that the reduction of Ca²⁺ accumulation by increased [Mg²⁺]_o during reoxygenation was due to reduced influx of Ca²⁺. Reduction in Ca²⁺ accumulation during reoxygenation might conceivably be caused by inhibition of sarcolemmal Ca²⁺ influx mechanisms such as reverse Na⁺/Ca²⁺ exchange, extensively discussed in our recent report (34), and/or inhibition of Ca²⁺ accumulation in intracellular pool(s), most probably the SR and mitochondria.

Methodological aspects in measurements of [Ca²+]_i and [Mg²⁺]_i. A central aspect in the present work is the selectivity of intracellular fluorescence probes for detecting the free ion concentration of Ca²⁺ and Mg²⁺ and the actual contamination of signal by other ions than those intended to detect. Fluo 3 is essentially nonfluorescent unless bound to Ca²⁺, with the maximum fluorescence signal at 1,400 nM Ca²⁺. Fluo 3 has a K_d of 390 nM for Ca²⁺ according to the manufacturer (Molecular Probes) in 100 mM KCl, 10 mM MOPS (pH 7.2), and 0-10 mM Ca²⁺-EGTA at 22°C. This value agrees well with the value found based on our calibration curve in permeabelized cells, which was 305 nM. Fluo 3 exhibits an at least 100-fold Ca²⁺-dependent fluorescence enhancement (10). Fluo 3 can also bind Mg²⁺. The relative magnitude of the Ca²⁺-dependent fluorescence increase of fluo 3 is about eightfold greater than the Mg²⁺-dependent fluorescence increase in 100 µM Ca²⁺ and 10 mM Mg²⁺, respectively (guideline data provided by Molecular Probes). We therefore conclude that Mg²⁺ was not likely to significantly influence fluo 3 fluorescence in the present experiments. Similarly, MgG can bind both Mg²⁺ and Ca²⁺. However, typical physiological Ca²⁺ concentrations (10 nM-1 µM) do not usually interfere with MgG fluorescence because the affinity of MgG for Ca²⁺ is low. Leyssens et al. (22) showed that [Mg²⁺]_i can be assessed using MgG and found that the MgG signal was not affected by changes in [Ca²⁺]_i in the range of 100-400 nM evoked by 10 mM caffeine. Additionally, the increase in the MgG signal by the mitochondrial uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) was neither significantly affected by removing external Ca²⁺ nor buffering intracellular Ca²⁺ with 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid. Because our measurements of [Ca²⁺]_i showed values of 60 and 300 nM for normoxic and hypoxic cells, respectively, we conclude that our measurement of MgG fluorescence was likely not contaminated by Ca²⁺ (K_d of MgG for Ca²⁺ is 6 µM, data reported by Molecular Probes).

Effect of increased [Mg²+]_o on [Ca²⁺]_i during hypoxia and reoxygenation. The increase in fluo 3 fluorescence at the end of hypoxia, demonstrating an increase in [Ca²⁺]_i, is in agreement with known changes in [Ca²⁺]_i in cardiac tissue during hypoxia (16) and ischemia (37), probably due to ATP depletion and subsequent inability to maintain Ca²⁺ gradients across sarcolemma and intracellular stores like the SR. The observed level of fluo 3 fluorescence corresponding to 60 nM in normoxic control cells after 1 h of incubation is in accordance with the value from normoxic cells in data presented by Nishida et al. (27) for end-diastolic [Ca²⁺]_i of 72 ± 6 nM in isolated adult rat cardiomyocytes electrically stimulated at 2 Hz. Increased [Ca²⁺]_i at the end of hypoxia and the subsequent fall in [Ca²⁺]_i starting at reoxygenation is in agreement with reported effects of reoxygenation on [Ca²⁺]_i recovery during reoxygenation (21, 30). The fluo 3 fluorescence stabilized at a level 60% above the normoxic control, indicating permanent disturbance of cellular Ca²⁺ homeostasis in reoxygenated cells in the present model, compatible with irreversible cell damage also demonstrated by the rounded shape of the cells and the higher proportion of trypan blue-positive cells. Because increased [Mg²⁺]_o did not alter [Ca²⁺]_i, the reduction of Ca²⁺ accumulation by increased [Mg²⁺]_o indicates reduced intracellular Ca²⁺ storage by the SR and/or mitochondria.

Effect of increased [Mg²+]_o on [Mg²⁺]_i during hypoxia and reoxygenation. To our knowledge, this is the first report on whether increased [Mg²⁺]_o alters [Mg²⁺]_i during reoxygenation of isolated hypoxic cardiomyocytes. Our results showing a large increase in [Mg²⁺]_i at the end of hypoxia were probably a consequence of ATP depletion and reduction of the Mg²⁺-ATP complex, as shown previously by others (22). Because cells treated with increased [Mg²⁺]_o did not show higher MgG fluorescence during reoxygenation than cells in the presence of normal [Mg²⁺]_o (0.8 mM) (and not lower [Ca²⁺]_i as discussed in Effect of increased [Mg²⁺]_o on [Ca²⁺]_i during hypoxia and reoxygenation), it appears that the primary effect of increased [Mg²⁺]_o was not exerted by elevating [Mg²⁺]_i or lowering [Ca²⁺]_i. We speculate that the effect of increased [Mg²⁺]_o could be exerted on extracellular sites such as the sarcolemmal transporters of Ca²⁺ and/or intracellular organelles like the mitochondria and SR, possibly influenced by the extracellular concentration of divalent cations like Mg²⁺ and Ca²⁺ via signal transduction pathway(s).

Effect of Mg²+ in combination with clonazepam on Ca²⁺ accumulation during reoxygenation. The large inhibitory effect of the selective mitochondrial Na⁺/Ca²⁺ exchange inhibitor clonazepam on reoxygenation-induced Ca²⁺ uptake indicates that mitochondrial reverse Na⁺/Ca²⁺ exchange might be responsible for a significant part of the cellular Ca²⁺ accumulation in the present model. Reperfusion of the myocardium has been shown to be associated with marked increase in mitochondrial Ca²⁺ content (26). Likewise, the inhibitory effect of increased [Mg²⁺]_o on Ca²⁺ accumulation might be due to inhibition of mitochondrial reverse-mode Na⁺/Ca²⁺ exchange. The additive effect of increased [Mg²⁺]_o and clonazepam on Ca²⁺ accumulation may suggest distinct modes of action. We cannot explain why or how cyclopiazonic acid actually augmented reoxygenation-mediated Ca²⁺ accumulation instead of reducing it as we had expected. Possibly, cyclopiazonic acid might lead to increased Ca²⁺ accumulation by inhibiting not only Ca²⁺ uptake by the SR but also Ca²⁺ transport (extrusion) by sarcolemma. The latter situation could make excess Ca²⁺ available to mitochondria. Because increased [Mg²⁺]_o markedly attenuated Ca²⁺ accumulation induced by cyclopiazonic acid (to values lower than in cells exposed to only hypoxia and reoxygenation), it appears that the effect of Mg²⁺ may be explained either by the effect on sarcolemmal Ca²⁺ transport and/or reduced Ca²⁺ accumulation by mitochondria.

Effect of Mg²+ in combination with clonazepam on _m during reoxygenation. The maintenance of _m is essential for cell survival and function in all aerobic cell types. Disruption of _m has been shown to provoke mPT, which is an early event in apoptotic and necrotic cell death, and mPT itself can actually cause further disruption of _m in a self-amplifying manner (20). Because reoxygenated cells in the presence of increased [Mg²⁺]_o, clonazepam, and combination of the two agents compared with nontreated reoxygenated cells had 17, 40, and 50% higher red-to-green fluorescence ratios of JC-1, respectively, we suggest a protective effect of these agents by diminishing the fall in _m. Increased [Mg²⁺]_o (5 mM) during reoxygenation did not alter [Mg²⁺]_i significantly in the present data. This is in agreement with data from Headrik et al. (12), showing improvement of myocardial energy metabolism and function in both normoxic and ischemic rat hearts in the presence of increased [Mg²⁺]_o (>8.0 mM) without alterations in [Mg²⁺]_i. The role of an extracellular locus of action for increased [Mg²⁺]_o and/or involvement of a signal transduction cascade stimulated by [Mg²⁺]_o therefore needs further investigation.

Interrelation between Ca²+ accumulation and _m. Our data showing an inverse correlation between Ca²⁺ accumulation with _m in cells exposed to hypoxia and reoxygenation suggest that the reduction of Ca²⁺ accumulation and preservation of _m by Mg²⁺ and clonazepam could be interrelated and involved in protecting cardiomyocytes against damage by hypoxia and reoxygenation. Kroemer et al. (20) reviewed the mechanisms by which mitochondria may regulate life and death. The three major pathways were briefly as follows: 1) signal transduction pathways via, e.g., caspase activation and calcium; 2) redox catastrophy because of glutathione depletion and augmented reactive oxygen species production; and 3) bioenergetic catastrophy, which includes ATP depletion and _m disruption. The mechanism of action of increased [Mg²⁺]_o and clonazepam alone and in combination on mitochondria cannot be determined based on the present results and needs further investigation.

Effect of Mg²+ and clonazepam on FSC and SSC. Ischemia (hypoxia) and reperfusion (reoxygenation) can result in apoptotic and necrotic cell death in myocardium (7, 15). The proportion of dead cells (trypan blue-positive cells) and LDH release were increased significantly within 10 min of reoxygenation in the present model (33). We (34) have previously shown that increased [Mg²⁺]_o at the onset of reoxygenation reduced LDH release, and the present study showed that clonazepam does the same. Therefore, increased [Mg²⁺]_o and clonazepam attenuate disruption of the cell membrane associated with cell necrosis and possibly late apoptosis. FSC and SSC reflect cell size and internal granularity and surface roughness, respectively (31). The decrease in FSC (cell shrinkage) paralleled either by no change or an increase in SSC (membrane blebbing) represents early changes during apoptosis (4). Reoxygenated cells in our model showing reduced FSC and increased SSC had therefore important characteristics of apoptotic cell death. These changes may be related to hypercontracture of cardiomyocytes as an early manifestation of irreversible cell injury (38). This could be due to activation of actin-myosin cycling by restoring of oxygen supply, ATP synthesis at still highly elevated [Ca²⁺]_i, and mitochondrial Ca²⁺ uptake (35). The samples taken at the onset of reoxygenation and analyzed during the following 2-3 min showed a small reduction in FSC and increase in SSC compared with normoxic control cells; however, we assume that this is an immediate response to reoxygenation and not to the hypoxia. Our previous study (33) supports this view by showing that if cells were fixed with paraformaldehyde (1%) at the onset of reoxygenation, they had the same FSC and SSC as normoxic control cells. Therefore, it appears that reoxygenation in this model induces apoptosis in agreement with other reports (7, 15). Apoptosis has been shown to be a dominant cause of cell death in rats and humans (2, 28) after myocardial infarction. To our knowledge, there is no data in the literature on the effect of increased [Mg²⁺]_o on apoptotic cell death during reoxygenation or reperfusion of the myocardium. Clonazepam and the combination of clonazepam and Mg²⁺ attenuated the reoxygenation-induced reduction in FSC and increase in SSC. The combination of clonazepam and Mg²⁺ tended to give larger effects on both FSC and SSC. Kang et al. (15) showed that apoptosis during reoxygenation was through a mitochondrion-dependent pathway. As described in Interrelation between Ca²⁺ accumulation and _m, the decrease in _m that has been associated with apoptosis (20) was attenuated by clonazepam and increased [Mg²⁺]_o in the present model. These results on FSC/SSC and _m support each other and indicate an antiapoptotic effect of clonazepam and increased [Mg²⁺]_o in this model. Internucleosomal fragmentation of genomic DNA that still can be defined as a hallmark of apoptosis (29) and activation of caspases during reoxygenation in the presence of increased [Mg²⁺]_o need to be addressed in further studies.