INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

DURING MYOCARDIAL ISCHEMIA, an acidification of the myocardium takes place, caused by increased anaerobic glycolysis, ATP hydrolysis, and carbon dioxide retention (5, 25). This leads to a decline in intracellular pH (pH_i) and a subsequent decline of extracellular pH (pH_e). This lowered pH seems to protect the myocardium during the metabolic insult, as has been shown for isolated myocardial cells (1, 4), isolated trabeculae (27), and Langendorff perfused hearts (6). Increasing the pH_e by restoring coronary flow or, in the case of isolated cells, changing media leads to the so-called pH paradox (4) and cell death. The mechanism(s) underlying the pH paradox and the protective effect of acidosis are unknown. It has been proposed that acidosis might inhibit Ca²⁺ overload and/or degradative enzymes (1, 4).

It is generally accepted that Ca²⁺ overload finally leads to irreversible cell damage during ischemia-reperfusion and metabolic inhibition. Previous work from our group has demonstrated that, in a cultured neonatal rat heart cell model of metabolic inhibition (at neutral pH_e), the cells start to accumulate Ca²⁺ within 15 min after ATP levels have fallen to 10% of control (2, 17). This initial Ca²⁺ uptake has been shown to be proportional to an increase in intracellular H⁺ concentration/extracellular H⁺ concentration gradient during metabolic inhibition. This increased gradient stimulates Na⁺/H⁺ exchange and increases intracellular Na⁺ concentration and subsequent uptake of Ca²⁺ via Na⁺/Ca²⁺ exchange. This Ca²⁺ accumulation proceeds for 30-40 min before the cells begin to leak lactate dehydrogenase (LDH), a sign of beginning major sarcolemmal structural damage. Lowering pH_e might reduce Ca²⁺ influx during metabolic insult and thereby reduce irreversible damage. We have proposed that, at the sarcolemmal level, this irreversible damage involves the Ca²⁺-induced expression of nonbilayer behavior of phosphatidylethanolamine (PE; see Ref. 20). Indeed a reduction of the sarcolemmal PE content reduced cell lysis during metabolic inhibition, as well as during ischemia (16). Furthermore, it was shown that irreversible sarcolemmal damage is preceded by a loss of the asymmetric distribution of PE (14, 17) and seems to be a prerequisite for cell lysis to occur (20). Therefore, we studied whether changing pH during metabolic inhibition affects the change in PE distribution.

Another mechanism that might be involved in the effect of pH_e on irreversible cell damage during metabolic inhibition is effecting the activation of phospholipase A₂ (PLA₂; see Refs. 1, 4, 10, 13, 22). We therefore studied the release of radiolabeled fatty acid from cellular phospholipids during metabolic inhibition at various pH_e and studied the effect of PLA₂ inhibitors on cell lysis during metabolic inhibition. The present study was set up to gain more detailed knowledge of the relation between pH_e, pH_i, net Ca²⁺ influx, possible PLA₂ activation, and cell lysis during metabolic inhibition. Therefore, this study combines techniques for on-line measurement of cellular Ca²⁺ and pH_i as correlated with measurement of LDH release and PLA₂ activation during metabolic inhibition.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cell culture. Culturing of the cells was done according to a modification of the method of Harary and Farley (8) and has been described previously (14, 17). Briefly, hearts of 1- to 2-day-old neonatal rats were excised, minced, and digested by trypsin. Myocytes were purified by successive fractionation steps and plated on either on Primaria-treated (Falcon Plastics), scintillant-containing disks (Bicron; for the ⁴⁵Ca studies) or on 60-mm Primaria culture dishes (for all other studies). Within 3 days, a confluent monolayer of spontaneously beating cells was formed. The growth medium was changed every other day.

Metabolic inhibition. Before the different experimental protocols, the cells were washed in buffer W containing (in mM) 133 NaCl, 3.6 KCl, 1.0 CaCl₂, 0.3 MgCl₂, and 5 tris(hydroxymethyl)aminomethane maleate at the pH at which the metabolic inhibition would be performed. The cells were then incubated for various periods, at 37°C, in the same buffer to which 10 mM 2-deoxyglucose (Sigma) and 1 mM iodoacetic acid (Sigma) were added as described previously (2, 17, 16). All of the incubations were done in a fixed volume (3 ml/dish or disk) to be able to directly compare the results of different experimental protocols.

High-energy phosphates. At the end of the 30 min of metabolic inhibition, the incubation fluid was removed, and the high-energy phosphates were extracted by the addition of 1 ml ice-cold 0.5 N perchloric acid for 30 min on ice. The extract was spun to remove any particulates (15 min, 900 g, 4°C), and 50 µl of the extract were analyzed the same day by injection in a Waters high-performance liquid chromatograph, using a Waters Radial-Pak Cartridge 8C₁₈ 5m column, and eluted isocratically by a buffer containing 0.3 M NH₄H₂PO₄ (pH 4.2) at 2 ml/min (9). The separate ATP, ADP, and AMP peaks were detected by absorbance at 214 nm. Standard curves for each metabolite were made at the beginning of each analysis. High-energy phosphate contents were expressed as nanomoles per milligram protein. Protein content was determined by using the procedure described by Lowry, as described previously (17, 19).

LDH release. LDH activity was measured by the decrease in absorbance at 340 nm, due to the conversion of NADH to NAD by LDH in the presence of pyruvate, as described previously (14, 16). During the metabolic inhibition, a sample of the incubation medium was withdrawn at different time points to measure the amount of LDH released into the medium. Finally, the cells were incubated in a 0.1% Triton X-100 solution (in buffer W) to release the LDH remaining in the cells. The release of LDH during the course of metabolic inhibition was expressed as percentage of total cellular LDH. LDH activity measurements were done immediately after taking the sample. In some experiments, the occurrence of LDH release was determined in the presence of trinitrobenzenesulfonic acid (TNBS). With the use of purified LDH (Sigma), it was shown that TNBS did not affect the activity of the enzyme.

⁴⁵Ca experiments. The ⁴⁵Ca binding and uptake by the cells was monitored by a scintillation flow cell technique described previously (12, 17). In short, the plastic disks on which the cells were grown contained a scintillant material. The disks were mounted in a flow cell, which then was placed in the well of a specially designed scintillation counter. To maintain the temperature at 37°C, the entire system was placed in a temperature-controlled chamber. The cells were first perfused with buffer W to remove any loosely adherent material. Subsequently, the cells were superfused with buffer W containing ⁴⁵Ca (1 µCi/mmol; ICN) for 25 min, at which time point asymptotic labeling was reached. The cells were then incubated in a fixed volume of the metabolic inhibition buffer (no flow) containing the same amount of ⁴⁵Ca. The ⁴⁵Ca signal of the cells in close proximity to the scintillation disk counts with a much higher efficiency (39%) than the ⁴⁵Ca in the bulk solution (<5%). Because the signal of the bulk solution does not change significantly, the effect of metabolic inhibition on ⁴⁵Ca labeling of the cells can be continuously followed on-line. The washout characteristics of the cell-associated Ca²⁺ pool after different periods of metabolic inhibition of the cells was obtained by superfusion of the cells with nonisotopic buffer (at 26 ml/min) at the end of the metabolic inhibition. At the completion of the experiment, the cells were scraped onto preweighed pieces of filter paper and dried overnight at 100°C, and the dry weight was obtained. Changes in Ca²⁺ content were then expressed as millimoles Ca²⁺ per kilogram dry weight.

[³H]arachidonic acid experiments. To measure release of ³H label from cells labeled with [³H]arachidonic acid, the cells were labeled on the second day of culture (day of isolation is day 0) with [³H]arachidonic acid (0.5 µCi/ml; Du Pont) in growth medium, after a modification (21). After 24 h, the medium was replaced by fresh medium without [³H]arachidonic acid, and cells were used on day 4 or 5. The cells were washed extensively with buffer W and incubated in buffer W (appropriate pH) in the presence or absence of metabolic inhibitors. The ³H release due to metabolic inhibition was obtained by subtracting the release in the absence of metabolic inhibitor at the same pH.

pH_i measurements. pH_i was measured by using carboxy-seminaphthorhodafluor-1 (SNARF-1), as described previously (3) with some modifications. Acetoxymethyl ester of SNARF-1 was dissolved in dimethyl sulfoxide (DMSO) and added to the incubation buffer to a final concentration of 10 µM SNARF-1 and 0.2% vol/vol DMSO. Cells grown on a coverslip were loaded with SNARF-1 for 20 min at 37°C. After washing, the cells were excited at 530 nm, and fluorescence was measured at 600 and 640 nm. The autofluorescence of the cells and instrument at these wavelengths was not higher than 5% of the total signal, and this value was always subtracted. SNARF-1 intracellular calibration was performed as described previously (3). Briefly, SNARF-1-loaded cells were incubated in a high-K buffer containing nigericin (20 µM), carbonyl cyanide p-trifluoromethoxyphenylhydrazone (1 µM) and valinomycin (1 µM). Under these conditions, pH_i was equal to pH_e. By varying pH_e, one can obtain the pK_a from SNARF-1 in situ calibration curves.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

LDH release versus pH_e. LDH release was used as a marker of sarcolemmal integrity during metabolic inhibition, using 10 mM 2-deoxyglucose and 1 mM iodoacetic acid, over a range of pH_e. Figure 1A shows that varying the pH_e during metabolic inhibition has a very pronounced effect on cell lysis, with hardly any LDH release at pH 5.2 and nearly 100% LDH release at pH 8.2, with release graded at the intermediate pH values. Total LDH (cellular plus released) was not affected by the various pH_e values. Analysis of the data in Fig. 1A is shown in Fig. 1B, where the amount of LDH release during 90 min of metabolic inhibition is plotted against the percentage of LDH released at different pH values. Clearly, there is a sigmoidal relationship between the two parameters, and 50% of the LDH is released at pH 7.0 (pH₅₀). A second buffer system (NaHCO₃/CO₂) was studied. This gave essentially the same results as the Trizma/maleate system, except that the pH dependency was shifted to a higher pH, resulting in a pH₅₀ of 7.5 (results not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effect of varying extracellular pH (pHe) on lactate dehydrogenase (LDH) release during 90 min of metabolic inhibition. A: cumulative LDH release at various pHe, which are indicated at the right (for the sake of clarity, SD have been omitted but were never >10%). B: total LDH released during 90 min of metabolic inhibition plotted against pHe. Nonlinear curve fitting showed that LDH release was 50% of maximal at pH of 7.02 (pH50).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Effect of varying pHe on adenosine nucleotides (Nucl.) during 30 min of metabolic inhibition. Left: data for control cells. Lowering pHe does not modify the decline of ATP, although it increases the AMP content somewhat (n = 3, mean + SD). pr, Protein.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Effect of varying pHe on sarcolemmal phosphatidylethanolamine (PE) distribution after 30 min of metabolic inhibition (met. inh.). Metabolic inhibition clearly results in an increase in the percentage of PE that can be labeled, e.g., is present in the extracellular leaflet of the sarcolemma. No significant difference is observed among the 3 different pHe values (8  n  13, mean + SD).

Ca²⁺ uptake versus pH_e. Many studies indicated that Ca²⁺ accumulation in myocardial cells is a precursor to irreversible damage. Therefore, we followed cellular Ca²⁺ uptake during metabolic inhibition. Figure 4A shows the increase in cellular ⁴⁵Ca content after 45 min of metabolic inhibition, at which time point very little or no LDH released occurred as a function of pH. A sigmoidal relationship between the two parameters is observed, with a pH₅₀ of ~7.3. For example, Fig. 4B shows ⁴⁵Ca uptake and LDH release during metabolic inhibition at a pH_e of 7.7, and it can be clearly seen that the increase of ⁴⁵Ca precedes cell lysis, as was observed for all pH_e values tested (not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   A: effect of varying pHe on net uptake of Ca2+ after 45 min of metabolic inhibition. Nonlinear curve fitting resulted in a pH50 of 7.3. B: increase in cellular Ca2+ () and LDH release () during metabolic inhibition at pHe of 7.7. It can be clearly seen that Ca2+ uptake precedes LDH release (n = 3, mean ± SD).

LDH release versus pH_e and external Ca²⁺ concentration. Figure 5 shows the relation between cell lysis and external Ca²⁺ concentration ([Ca²⁺]_e) at various pH_e during 90 min of metabolic inhibition. At low pH_e (6.2), very little LDH release occurred and was unaffected by [Ca²⁺]_e. High pH_e (8.0) resulted in maximal LDH release, which again was unaffected by varying [Ca²⁺]_e. In contrast, metabolic inhibition at pH_e 7.2 led to graded cell lysis as [Ca²⁺]_e was increased. Thus, at intermediary pH, LDH release is clearly dependent on the level of [Ca²⁺]_e.

View larger version (10K): [in this window] [in a new window]   Fig. 5.   Effect of varying pHe and Ca2+ concentration on LDH release during 90 min of metabolic inhibition. Note that only at pHe 7.2 does LDH release increase with increasing [Ca2+]e. By contrast, at pHe 6.2 and 8.0, no graded effect of varying [Ca2+]e is seen. At pHe 6.2, LDH release remains minimal; at pHe 8.0, LDH release remains maximal (n = 3, mean ± SD).

Alteration of pH_e during metabolic inhibition. LDH release was followed while changing pH_e during metabolic inhibition (Fig. 6). When metabolic inhibition was started at a pH_e of 6.2, little lysis occurred for 60 min. However, upon increase of pH_e, a rapid LDH release was observed, which was clearly dependent on the pH_e increment introduced (Fig. 6A). Analysis of the data obtained 45 min after the pH switch and plotting the total LDH release versus pH_e shows a sigmoid curve, with a pH₅₀ of 7.0 (Fig. 6B). Lysis during metabolic inhibition at pH_e 7.7 could be strongly attenuated by reducing pH_e to 6.2, provided the switch occurs before cell lysis starts (Fig. 6C). Inclusion of 50 µM monensin in the incubation medium strongly increased the rate of LDH release upon increasing pH_e from 6.2 to 7.7 (Fig. 6D).

View larger version (25K): [in this window] [in a new window]   Fig. 6.   Effect of switching pHe during metabolic inhibition on LDH release. A: cells were metabolically inhibited at pHe 6.2. After 60 min, pHe was switched, and cell lysis commenced, dependent on pHe. B: total LDH released 45 min after pHe switch plotted against pHe; nonlinear curve fitting resulted in extracellular pH50 of 6.97 ± 0.03. C: cells were metabolically inhibited at pHe 8.0 for 30 min. At this point, pHe was switched to either 6.2 () or maintained at 8.0 (). Switching to pHe 6.2 clearly reduced LDH release. D: cells were metabolically inhibited at pHe 6.2 for 60 min at which time pHe was switched to 8.0. Presence of 50 µM monensin () significantly increased the rate of LDH release compared with control ().

pH_i during metabolic inhibition. pH_i was measured as it responded to the different experimental protocols of metabolic inhibition. With the fluorescent probe SNARF-1, pH_i of untreated cells was found to be about 7.1. Alteration of pH_e to 6.2 led to a rapid decline and stabilization of pH_i at 6.5. Subsequent metabolic inhibition led to a further decrease of pH_i to 6.3. Note that pH_i then remained unchanged for the next 80 min (Fig. 7A). At the start of the metabolic inhibition at pH_e of 7.2, pH_i rapidly falls to 6.8 but remains stable for only 20 min. It then rises progressively to >7.0 over the next 50 min (Fig. 7B). In Fig. 7C, it is shown that increasing pH_e to 8.2 results in an increase of pH_i to ~7.7. During subsequent metabolic inhibition, pH_i rapidly falls to 7.1 and remains there for <15 min, after which pH_i rapidly increases to >7.8. In Fig. 7, A and B, we show the effect of increasing pH_e during metabolic inhibition. Note that pH_i rises to ~7.8 in both cases. In summary, we have shown that, at pH_e 6.2, pH_i remains low and stable in the face of metabolic inhibition. At pH_e 7.2, pH_i rises slowly during metabolic inhibition, and, at pH_e 8.2, pH_i falls only to 7.2 for a brief period after metabolic inhibition and then rises rapidly to >7.8.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of varying pHe on intracellular pH (pHi) during both control and metabolic inhibition. Results shown are representative of 5 experiments using each experimental protocol. Switches in pHe and start of metabolic inhibition [iodoacetic acid (IAA)] are indicated by arrows and the pHe. A: lowering pHe from 7.2 to 6.2 results in a decrease in pHi to 6.6, which further falls to 6.3 upon start of metabolic inhibition, but then stays stable for the next 75 min. Subsequent increase of pHe results in a rapid rise of pHi to 7.8. B: start of metabolic inhibition at pHe of 7.2 produces lowering of pHi to 6.8, after which a gradual increase is observed. Subsequent increase of pHe results in a rapid rise of pHi to 7.8. C: increasing pHe from 7.2 to 8.0 results in an increase of pHi, which stabilizes at ~7.7. Subsequent metabolic inhibition results in a rapid fall of pHi to 7.2, where it remains for 15 min, followed by a rapid increase, which again stabilizes at pHi ~7.8.

Activation of PLA₂. The results thus far suggest that LDH release (cell lysis) during metabolic inhibition resulted from the interaction of pH_i and intracellular Ca²⁺ to affect a mechanism or mechanisms responsible for the lysis. This further suggested to us the possible involvement of a Ca²⁺-sensitive PLA₂, with an optimal activity at alkaline pH. PLA₂ activity was measured as the release of ³H-labeled products during the course of metabolic inhibition. Upon labeling of the cells with [³H]arachidonic acid, 75% of total cellular radioactivity was present in phospholipids. Figure 8A shows the release of ³H caused by metabolic inhibition at pH_e of 7.2, 8.0, and 6.2. Up to 60 min, no significant difference is observed between the three groups. After longer periods, a significantly increased ³H release was observed at pH_e 7.2 and 8.0 compared with 6.2 (P < 0.05). However, as can be seen in Fig. 1, during these prolonged periods of metabolic inhibition, cell lysis occurred. No difference in ³H release was observed between pH_e 7.2 and 8.0. Analysis of the distribution of the cellular radioactivity as distributed among (lyso)phospholipids, neutral lipids, and free fatty acids at the end of the experiments did not show a significant difference between the various groups (not shown). Another approach to study the possible involvement of PLA₂ was to investigate whether PLA₂ inhibitors affected cell lysis during metabolic inhibition as a function of pH. As can be seen in Fig. 8, B-D, the inhibitors mepacrine and 4-bromophenacylbromide, used at concentrations known to inhibit PLA₂ activity (14a, 28, 29), did not modulate cell lysis during metabolic inhibition.

View larger version (25K): [in this window] [in a new window]   Fig. 8.   Effect of varying pHe on phospholipase A2 (PLA2) during 90 min of metabolic inhibition. A: 3H release during metabolic inhibition of cells previously labeled with [3H]arachidonic acid. Data have been corrected for release during control incubations at various pHe. Significant increase in the release at pHe of 7.2 () and 8.0 () vs. 6.2 () is observed after 75 and 90 min. B-D: effect of PLA2 inhibitors on LDH release during metabolic inhibition at pHe of 6.2 (B), 7.2 (C), and 8.0 (D). Neither mepacrine (100 µM, hatched bars) nor 4-bromophenacylbromide (20 µM, solid bars) significantly reduced LDH release compared with control (open bars; n = 4, mean ± SD).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

There has, thus far, been no study that has correlated the progression of sarcolemmal destruction during metabolic inhibition with two factors recognized to be of importance in this progression: pH and Ca²⁺. Cell lysis (LDH release) during metabolic inhibition is clearly dependent on pH_e as can been seen in Fig. 1A. This observation agrees with several other studies (1, 4). Because, in the present study, a wide range of pH_e values was applied, we were able to analyze the cell lysis data versus pH_e. A sigmoidal relationship was obtained (Fig. 1B). This relationship suggests the activation or inhibition of a process or "factor" with a half-maximal effect at about pH_e 7.0.

An altered usage of ATP during metabolic inhibition at different pH_e might be involved in the evolution of the loss of sarcolemmal integrity. Lowering pH_e, which dramatically reduced LDH release during prolonged metabolic inhibition (Fig. 1A), did not attenuate the decline in ATP during the first 30 min of metabolic inhibition (Fig. 2). Increasing pH_e, leading to increased LDH release upon prolongation of the metabolic inhibition, did not affect the level of ATP either (Fig. 2). Thus it seems unlikely that the effect of changing pH_e on cell lysis is mediated by an effect on ATP levels during metabolic inhibition. pH_e seems to have some effect on AMP content, which is significantly higher at lower pH_e (Fig. 2). This indicates that, at lower pH_e, the breakdown of AMP is decreased. In the case of reversible metabolic compromise, this might be beneficial for the cells since it would allow a faster regeneration of ATP during the recovery phase. Nevertheless, it must be concluded that changing pH_e does not affect cellular ATP content during metabolic inhibition.

Several sarcolemmal changes have been observed before cell lysis occurs during (simulated) ischemia and metabolic inhibition. One of them is a loss of asymmetric distribution of PE (14, 16, 17), and we proposed that this change might be involved in dysfunction of the sarcolemma (20). Therefore, we tested whether varying pH_e during metabolic inhibition would affect the loss of PE asymmetry. As can be seen in Fig. 3, it must be concluded that the effects of varying pH_e during metabolic inhibition on cell lysis are not mediated by an effect on the mechanism by which sarcolemmal PE asymmetry is lost. Furthermore, this indicates that a loss of PE asymmetry does not, per se, lead to cell lysis. This agrees with our recent observation that PE asymmetry loss upon simulated ischemia can be reversed, provided reoxygenation takes place before cell lysis starts (15).

The level of cellular Ca²⁺ is generally believed to play a key role in the transition from reversible to irreversible damage of the heart myocyte. Ca²⁺ is proposed to activate Ca²⁺-dependent proteases and lipases (1, 4, 25) and to affect the physicochemical properties of sarcolemmal phospholipids (20). Figure 4A shows that, during metabolic inhibition, a clear relationship exists between the net uptake of Ca²⁺ and pH_e. This suggests a relationship between the size of the transsarcolemmal H⁺ gradient and the amount of Ca²⁺ taken up and is in line with experimental evidence suggesting involvement of Na⁺/H⁺ and Na⁺/Ca²⁺ exchange in cellular Ca²⁺ accumulation (1, 3, 4). A causal relationship between H⁺ production, the generated transsarcolemmal pH gradient, and cellular Ca²⁺ overload during metabolic inhibition has recently been established (P. Korge, S.-Y. Wang, and G. A. Langer, unpublished observation). These researchers were able to diminish or even prevent the net Ca²⁺ uptake by dissipation of the H⁺ gradient without allowing Na⁺ to accumulate in the cell. Thus a clear relationship between net increase of cellular Ca²⁺ and pH_e would be expected during metabolic inhibition. Such is shown in Fig. 4A. This relationship, pH_e versus Ca²⁺ uptake, is similar to that between pH_e and LDH release (Fig. 1B). This suggests a possible relationship between net accumulated Ca²⁺ and LDH release. This is supported by the data presented in Fig. 4B, where it is seen that an increase in cellular Ca²⁺ precedes cell lysis.

In the experiments discussed thus far, [Ca²⁺]_e was maintained at a concentration of 1 mM. To study further the possible Ca²⁺ dependency of LDH release, we varied [Ca²⁺]_e at three different pH_e (Fig. 5). The effect of varying [Ca²⁺]_e was very critically dependent on pH_e during metabolic inhibition. There was no graded dependency at either high or low pH_e, but there was a clear dependency at moderate pH_e. This indicates that the process(es) involved in cell lysis during metabolic inhibition at high pH_e are already maximally active at low Ca²⁺ concentration, whereas, at low pH_e, the process(es) are inhibited and insensitive to increasing [Ca²⁺]_e. That pH_e is a major controlling factor is also shown in Fig. 6A, where an abrupt increase of pH_e during metabolic inhibition rapidly activates LDH release. The opposite also holds; lowering pH_e before the onset of lysis prevents LDH release (Fig. 6B).

The previous part of this discussion finding assumes that changes in pH_e induce relatively rapid and substantial changes of pH_i. To further understand the process(es) responsible for cell lysis, it is pertinent to know how pH_i reacts to the changes in pH_e. Changing pH_e from 7.2 to either 6.2 or 8.2, in the absence of metabolic inhibitors, leads to a relatively rapid change (10-20 min) of pH_i to 6.5 or 7.6, respectively. Subsequent application of metabolic inhibition led to a very rapid (2-6 min) decrease of pH_i to 6.3, 6.8, and 7.2 for pH_e of 6.3, 7.2, and 8.2, respectively. During continued metabolic inhibition, pH_i remained absolutely stable (pH_e 6.2), showed a small delayed increase (pH_e 7.2), or increased relatively rapidly (pH_e 8.2; Fig. 7). The size and time course of the increase in pH_i correlate with the observed increase in Ca²⁺ uptake and LDH release during the course of metabolic inhibition (Figs. 1 and 4). Furthermore, Fig. 7, A and C, shows that increasing pH_e during metabolic inhibition results in a very rapid increase of pH_i, which again precedes a rapid release of LDH (Fig. 6A). Thus the data on measurement of pH_i show that the assumption that changes in pH_e induce relatively rapid and substantive changes of pH_i is correct and indicate a relationship between pH_e, pH_i, Ca²⁺ uptake, and subsequent cell lysis.

This relationship suggested to us that activation of a factor dependent on increase of pH_i and intracellular Ca²⁺ could be responsible for the onset of cell lysis. The literature suggests that this factor may be cytosolic PLA₂. However, the data presented in Fig. 8A show that different degrees of activation of PLA₂ cannot explain the dramatic effect of varying external (and subsequently internal) pH on cell lysis during metabolic inhibition. This observation is confirmed by the data presented in Fig. 8, B-D, in which it is shown that application of two well-known inhibitors of PLA₂, mepacrine and 4-bromophenacylbromide, do not reduce cell lysis during metabolic inhibition at the various pH_e values tested.

Data by Bond et al. (4) indicated that changes in pH_i are involved in cell damage during the so-called pH paradox and in the protective effect of lowered pH_e. The data presented in the present paper also show that an increase in pH_i during metabolic inhibition precedes cell lysis. Furthermore, increasing pH_e after 60 min of metabolic inhibition at pH 6.2 resulted in a rapid LDH release, and this release is greatly increased and accelerated by the presence of 50 µM monensin, a Na-H ionophore (Fig. 6D), which results in an even more rapid increase of pH_i upon the pH switch (4).

The mechanism involved in cell death during the pH paradox also seems to be involved in cell lysis during metabolic inhibition at constant pH_e. This is indicated by the fact that the relationship between pH_e and the degree of cell lysis is identical for 90 min of metabolic inhibition at constant pH_e (Fig. 1B) and for the pH paradox [60 min metabolic inhibition at pH 6.2, followed by 30 min of metabolic inhibition at varying pH_e (Fig. 6B)]. Interestingly, besides an identical pH₅₀ of 7.0, the degree of cell lysis under both 90-min protocols was also in the same range. This indicates that the pH_e before cell lysis begins hardly affects cell lysis at the various pH_e values later in the protocol. This is also shown in Fig. 6C, where it is shown that lowering pH_e from 8.0 to 6.2 before cell lysis starts almost completely prevents cell lysis.

With regard to the mechanisms involved in cell death in the present model, we conclude that, despite the slightly alkaline pH optima of PLA₂, this group of enzymes is not directly involved in cell death. Another possible mechanism involves the activation of proteases upon increase of pH_i and [Ca²⁺]_i, although no direct experimental proof is available. The very rapid LDH release during the pH paradox, especially in the presence of monensin, indicates that, if proteases are involved, they have to be very rapidly and massively activated to cause such an amount of cell lysis.

Therefore, another explanation for the observed effects of pH_e on cell lysis during metabolic inhibition must be put forward. In our previous work, we proposed, based on morphological and biochemical data on myocardial tissue, other biomembranes, and isolated membrane components, that an increase of cytosolic Ca²⁺ would result in a binding of Ca²⁺ to the negatively charged PS in the cytoplasmic leaflet of the sarcolemma (14, 16, 20). This binding of Ca²⁺ to PS head groups leads to a cross-linking of the PS molecules, thereby leading to a phase segregation within the cytoplasmic leaflet of the sarcolemma. Due to this phase segregation, the bilayer stabilizing effect of PS on PE is lost, resulting in destabilization of the sarcolemma and a disruption of the membrane. H⁺ are also able to bind to the negatively charged PS molecules but will not induce cross-linking of PS and will prevent or attenuate the binding of Ca²⁺ to PS (26). Indeed, a strong reduction of Ca²⁺ binding to both synthetic phospholipids (23) and to isolated sarcolemma (11) is produced by lowered pH. The effect of pH on Ca²⁺ binding is mediated by the extent of ionization of the sarcolemmal Ca²⁺ binding sites, of which at least 75% is of phospholipid origin (11). These are predominantly present in the cytoplasmic leaflet of the sarcolemma (18). The extent of ionization of these binding sites could be fitted by a Henderson-Hasselbach relation, with pK of the putative sites between 6.60 and 7.15 (11). This is in the range found for the LDH release (Fig. 1B) and Ca²⁺ uptake (Fig. 4A) responses. The above proposed model predicts that a strongly reduced sarcolemmal Ca²⁺ binding capacity and a strongly reduced influx of Ca²⁺ will attenuate or even prevent the destabilization of the sarcolemma. This alternative mechanism clearly needs further experimental proof and is currently under investigation.