INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE CONTROL OF INTRACELLULAR pH (pH_i) is important in virtually all tissues, and intracellular acidosis has profound effects on various cellular functions. In hearts, intracellular acidosis as a major component of myocardial ischemia or as a result of renal or respiratory failure disturbs the electrical and mechanical performances. In ischemic hearts, reperfusion is necessary for the functional recovery, but it is not always beneficial to the damaged myocardium and causes an additional myocardial injury. A phenomenon in which rapid return from acidotic to physiological pH_i produces undesirable secondary effects (termed the "pH paradox") has been recognized to be a factor for reperfusion cell injury (8, 21).

Intracellular acidosis exerts substantial effects on contractile performance and intracellular Ca²⁺ metabolism in cardiac myocytes. These include a decrease in twitch contraction, a prolongation of relaxation, an increase in the Ca²⁺ transient (CaT) amplitude, and a decrease in the rate of twitch intracellular Ca²⁺ concentration ([Ca²⁺]_i) decline (5, 15, 19, 35). Previous studies (12, 34) have indicated that the decrease in twitch contraction is caused principally by a decreased Ca²⁺ responsiveness of contractile proteins.

Although twitch contraction rapidly decreases in the initial minutes of acidosis, it slowly recovers (15, 19, 24, 35). The mechanism of the contractile recovery remains undefined, but the increase in CaT amplitude, which overcomes the decreased Ca²⁺ responsiveness of contractile proteins, has been suggested (15, 19). The increase in [Ca²⁺]_i during acidosis seems to be due to two mechanisms: displacement of Ca²⁺ from intracellular buffering sites (24, 35) and activation of Na⁺/H⁺ exchange by an increase in intracellular [H⁺] (15). The activation of Na⁺/H⁺ exchange increases [Na⁺]_i which, in turn, increases [Ca²⁺]_i via Na⁺/Ca²⁺ exchange (36).

In mammalian hearts, the sarcoplasmic reticulum (SR) plays a dominant role both in the contraction and relaxation processes (5). Previous in vitro studies have shown that acidosis directly inhibits the SR Ca²⁺-ATPase (12) and the SR Ca²⁺ release channels (ryanodine receptors) (1, 44). Moreover, acidosis decreases (or does not change) the Ca²⁺ entry via the L-type Ca²⁺ channels (11, 19, 20). These inhibitory effects are expected to reduce the total SR Ca²⁺ content, suppress the Ca²⁺-induced Ca²⁺ release (CICR) from the SR, and decrease the CaT amplitude (3).

Here we hypothesized that the reactivation of SR Ca²⁺-ATPase might be another mechanism for the increase in CaT amplitude during prolonged acidosis. The SR Ca²⁺-ATPase is activated through phosphorylation of phospholamban (PLB) and also by its direct phosphorylation (5, 28, 38, 40). Because [Ca²⁺]_i increases during intracellular acidosis, the Ca²⁺-calmodulin-dependent kinase II (CaMKII) is a possible candidate for the reactivation of SR Ca²⁺-ATPase. CaMKII is highly enriched in the nervous system and mediates many actions of Ca²⁺ (37). In cardiac myocytes, CaMKII phosphorylates several proteins and modulates the excitation-contraction coupling (4, 14, 27, 43, 45).

In this study, we examined the effects of two types (respiratory and metabolic) of prolonged acidosis on the twitch contraction and CaT in isolated rat myocytes, and tried to clarify the mechanism and physiological implication of the recovery of twitch contraction. Our results indicate that the reactivation of SR Ca²⁺ uptake caused by CaMKII activation during prolonged acidosis can reaccelerate the twitch [Ca²⁺]_i decline, increase the SR Ca²⁺ content and the CaT amplitude, and thereafter induce the contractile recovery.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Isolation of ventricular myocytes. This investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Ventricular myocytes were isolated from male Sprague-Dawley rats (220-300 g) and were stocked in the modified Kraftbrühe (KB) solution as described previously (16). Composition of the modified KB solution was as follows (in mM): 70 KOH, 40 KCl, 20 KH₂PO₄, 3 MgCl₂, 50 glutamic acid, 10 glucose, 10 HEPES, and 0.5 EGTA (pH 7.4 with KOH). Before the start of experiments, the modified KB solution was replaced with a Krebs solution containing (in mM) 113.1 NaCl, 4.6 KCl, 1.2 MgCl₂, 3.5 NaH₂PO₄, 21.9 NaHCO₃, 1.5 CaCl₂, and 10 glucose, or a HEPES-buffered solution that contained (in mM) 137 NaCl, 4 KCl, 1.2 MgSO₄, 10 glucose, 10 HEPES, and 1.5 CaCl₂, with pH adjusted to 7.4 with NaOH.

Loading of fluorescent indicators. Myocytes were loaded with indo-1-acetoxymethyl ester (indo-1-AM; 10 µM) for 10 min at room temperature. The stock solution of indo-1-AM was prepared by mixing 1 mg of dye in dry dimethyl sulfoxide and was kept frozen in aliquots until use. Immediately before use, the dyes were dissolved with KB solution containing 1% bovine serum albumin and 0.075% Pluronic F-127 (wt/vol, final concentration). The myocytes were then washed twice in a modified KB solution and were incubated for 30 min to complete hydrolysis of dyes. For the measurement of pH_i, the cells were loaded with 2',7'-bis (carboxyethyl)-5,6-carboxyfluorescein (BCECF)-AM (1 µM) for 30 min at room temperature and were then washed and incubated as mentioned above.

Apparatus. After fluorescent indicators were loaded, a small aliquot of myocytes were placed in an experimental chamber mounted on the stage of an inverted microscope (TMD, Nikon; Tokyo, Japan), and perfused with Krebs or HEPES solution at room temperature. The myocytes were field stimulated through platinum electrodes at 0.5 Hz, and were illuminated by a transmitted illuminator or ultraviolet light via an epifluorescence illuminator from a 100-W xenon lamp equipped with an interference filter.

Calibration of BCECF fluorescence. We conducted an in vivo calibration in the preliminary cell group, according to the method previously described (16). BCECF-loaded myocytes were perfused with the calibration solutions, which contained 10 µM nigericin in the following solution (in mM): 130 KCl, 15 MgCl2, 1 2-(N-morpholino)ethanesulfonic acid, and 15 HEPES. The pH was adjusted appropriately with KOH. Fluorescence ratios were linearly related to pH_i from 6.5 to 7.5.

Experimental protocol. The myocytes that showed potentiation of twitch after 30 s of rest (postrest potentiation) were selected, and those in which the diastolic [Ca²⁺]_i exceeded 200 nM or the spontaneous contractions during rest occurred were excluded from the analyses. In the present study, two types of acidosis were applied. For respiratory acidosis, increasing the percentage of CO₂ equilibrated with the bicarbonate buffer from 5 to 30% reduced pH_i by 0.32 pH units (Fig. 1A, n = 10). The time constant of pH_i change was ~1.0 min, and stable acidosis was achieved up to 15 min. On returning to the control perfusate gassed with 5% CO₂, pH_i recovered to its control value without exhibiting an overshoot. For metabolic acidosis, the myocytes were superfused with HEPES solution containing sodium propionate (20 mM) with pH 6.8. The pH_i decreased by 0.38 pH units and a stable acidosis was obtained as respiratory acidosis, but on returning to the control HEPES solution, pH_i exhibited an alkaline overshoot (Fig. 1A, n = 10).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Changes in intracellular pH (pHi) during respiratory and metabolic acidosis and the effect of pHi on indo-1 calibration curves. A: rapid and stable intracellular acidosis (pH: ~0.35) was obtained by increasing the percentage of CO2 in the bicarbonate buffer from 5 to 30% (respiratory acidosis;  on solid trace), or by adding sodium propionate (20 mM) in HEPES solution at pH 6.8 (metabolic acidosis;  on dashed trace). On returning to the control perfusate, pHi rapidly recovered to the control values and at metabolic acidosis, a transient alkaline overshoot was observed. Values are means ± SE from 10 experiments (2 rats). B: in vitro calibration curves of indo-1 fluorescence at pH 7.2 (solid line and ) and at pH 6.8 (dotted line and ). Free Ca2+ concentrations ([Ca2+]) in calibration solutions were corrected with the use of proper dissociation constants (Kd) of EGTA for Ca2+ (151 nM at pH 7.2 and 940 nM at pH 6.8). The fluorescence ratios at different levels of [Ca2+] were plotted and were fitted using the formula [Ca2+] = Kd × (Sf/Sb) × (R  Rmin)/(Rmax  R), where Rmin and Rmax are the ratios at zero and saturating [Ca2+], Sf/Sb is the ratio of fluorescence at 485 nm at zero and saturating [Ca2+]. The obtained values of Rmax and Rmin were almost identical between pH 7.2 and 6.8, but the Kd at pH 6.8 was 1.2-fold larger (see MATERIALS AND METHODS). C: in vivo calibration curve of indo-1 fluorescence at pH 7.2. The Kd at pH 7.2 was 786 nM, Sf/Sb was 3.3, Rmax and Rmin were 3.1 and 0.31, respectively. Under the acidic perfusion, the Kd was assumed to be 943 nM. [Ca2+]i, intracellular [Ca2+]. Values are means ± SE from 20 experiments (2 rats).

Calibration of indo-1 fluorescence. A major problem for [Ca²⁺]_i measurement in our experiments is the direct effect of pH_i on indo-1 fluorescence (25). We first compared in vitro calibration curves of indo-1 fluorescence between pH 7.2 and 6.8. Because the free [Ca²⁺] in calibration solutions (calcium calibration buffer kits; Molecular Probes) are also affected by the change in pH, we corrected them with the use of proper dissociation constants (K_d) of EGTA for Ca²⁺ (151 nM at pH 7.2 and 940 nM at pH 6.8). Thus the calibration curves were made with two series of different [Ca²⁺] (3.4, 27, 60, 237, and 951 nM and 1 mM at pH 7.2 and 1.7, 21, 106, 238, and 937 nM and 1 mM at pH 6.8) (Fig. 1B). The indo-1 fluorescence ratios at different levels of [Ca²⁺] were fitted using the formula [Ca²⁺] = K_d × (S_f/S_b) × (R  R_min)/(R_max  R), where R_min and R_max are the ratios at zero and saturating [Ca²⁺], (S_f/S_b) is the ratio of fluorescence at 485 nm at zero and saturating [Ca²⁺] (13). The obtained values of R_max and R_min were almost identical between pH 7.2 and 6.8 (R_max, 1.26 and 1.27, and R_min, 0.06 and 0.06), but the K_d at pH 6.8 was 1.2-fold larger (375 and 419 nM). Subsequently, we performed in vivo calibration, in the preliminary cell group using calibration solutions with different [Ca²⁺]_i (17, 150, and 600 nM, 1.35, 39.6 µM and 5 mM) at pH_i 7.2. The K_d was 786 nM, (S_f/S_b) was 3.3, and R_max and R_min were 3.10 and 0.31, respectively (Fig. 1C). Under the control perfusion with pH 7.4, we applied these values for calculation of [Ca²⁺]_i, and under acidic perfusion, the K_d was assumed to be 943 nM (786 nM × 1.2; from in vitro calibration).

Reagents and solutions. BCECF-AM and indo-1-AM were supplied from Molecular Probes. Thapsigargin, ryanodine, and cantharidin were purchased from Sigma. BAY K 8644 was obtained from Calbiochem. 2-[N-(2-hydroxyethyl)-N- (4-methoxybenzenesulfonyl)] amino-N-(4-chlorocinnamyl)-N-methyl benzylamine (KN-93) was purchased from Seikagaku and isoproterenol was purchased from Wako. These reagents were used from stock solutions in water or dimethyl sulfoxide, and were added to the perfusate immediately before use. The final concentration of dimethyl sulfoxide was not >0.1%, and each reagent produced neither change in cellular autofluorescence nor fluorescence artifact by itself.

Statistical analysis. Results were expressed as means ± SE for the indicated number (n) of myocytes. The number of animals that generated cell samples was also shown. Because several myocytes were derived from the same rat, it was verified by a nested ANOVA procedure that the heterogeneity of cellular responses was dependent on the individual myocytes rather than differences among rats. Changes in parameters of cell shortening and CaT were then estimated by repeated-measures ANOVA for overall effects of the treatments. The differences in each parameter within treatments were compared with the use of ANOVA, followed by Bonferroni's modified t-test. The proportions of recovered cells among treatments were compared using Fisher's exact probability test. The comparisons of baseline cell shortening and CaT between the cells that recovered and did not recover were performed with the use of an unpaired Student's t-test. The probability was considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effects of respiratory acidosis on twitch cell contraction and CaT. Figure 2A shows a representative example of the changes in steady-state (SS) twitch cell contraction during respiratory acidosis. The simultaneous records of CaT in Fig. 2B were obtained at the times indicated by the letters in Fig. 2A at 10 min of control perfusion, 4 min and 10 min during respiratory acidosis, and 2 min after reperfusion with control Krebs solution, respectively. Respiratory acidosis quickly decreased the twitch contraction but slightly increased both the diastolic [Ca²⁺]_i and CaT amplitude. Contraction reached a nadir at ~4 min, but thereafter slowly recovered. The recovery was accompanied by larger increases in both the diastolic [Ca²⁺]_i and CaT amplitude. After washout of the acidic solution, both the twitch contraction and [Ca²⁺]_i increased, but then spontaneous contractions and [Ca²⁺]_i oscillation occurred within 5 min, indicating cellular Ca²⁺ overload. The averaged changes of twitch contraction and CaT were analyzed in 14 myocytes (8 rats, Table 1). We defined the time at minimum contraction as depression phase (DP; 4.2 ± 0.3 min). Because the contractile recovery was visually apparent when contraction exceeded 1% of diastolic length, this value was chosen to determine whether the recovery occurred or not. The contractile recovery was generally prominent by 10 min, and we took this time point (10 min) as the recovery phase (RP). Both the diastolic [Ca²⁺]_i and CaT amplitude increased significantly at DP and increased further at RP, whereas the diastolic cell length or time to peak of CaT did not change. The SR Ca²⁺ contents were also examined by discharging SR Ca²⁺ with a rapid application of caffeine (10 mM) (2). In this experiment, the perfusate [Ca²⁺] was reduced to 0.8 mM to prevent Ca²⁺ overload. The amplitude of caffeine-induced CaT increased significantly at DP and increased further at RP (Table 1).

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Effects of respiratory acidosis on twitch cell shortening and Ca2+ transient (CaT) in a rat ventricular myocytes loaded with indo-1. A: continuous record of twitch cell shortening (0.5 Hz) from a typical experiment. Respiratory acidosis as indicated by bar was obtained by increasing the percentage of CO2 in bicarbonate buffer from 5 to 30%. B: CaT obtained at the times indicated by the lowercase letters in A at control perfusion (a), 4 min (b), and 10 min (c) during respiratory acidosis and 2 min after reperfusion (d) with control solution, respectively. Each time constant () of twitch [Ca2+]i decline is also shown. Arrowheads indicate electrical stimuli at 0.5 Hz.

Cell-to-cell heterogeneity in recovery of twitch contraction during respiratory acidosis. In our experimental protocol, the initial decrease in twitch contraction occurred in all myocytes, but the contractile recovery was evident in 9 of 14 cells (64%: 1.5 ± 0.3% of diastolic length at RP; P < 0.05 vs. that at DP). To investigate the underlying mechanism for this cell-to-cell heterogeneity, we compared possible parameters of twitch contraction and CaT between the cells that recovered (Rec: n = 9) and did not recover (NRec: n = 5). There was no significant difference in twitch contraction (2.4 ± 0.4% and 2.2 ± 0.6% of diastolic length), diastolic [Ca²⁺]_i (132 ± 15 and 141 ± 15 nM), or CaT amplitude (378 ± 66 and 323 ± 45 nM) at control perfusion. In Rec, the CaT amplitude continuously increased (Fig. 3A), whereas in NRec, the CaT amplitude did not increase (from 389 ± 60 nM at DP to 240 ± 67 nM at RP, P = not significant). On the other hand, the increase in diastolic [Ca²⁺]_i from DP to RP was more evident in NRec (300 ± 54 to 387 ± 69 nM, P < 0.05).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Acidosis induced changes in the profile of CaT in the cells that twitch contraction recovered (Rec). A: steady-state CaT amplitude () and diastolic [Ca2+]i () during control perfusion (CTL), at depression phase (DP), and recovery phase (RP) during respiratory acidosis. B: time constants of twitch [Ca2+]i decline obtained from single exponential fits. Values are means ± SE from 9 myocytes (6 rats). * P < 0.05 vs. the values at CTL,  P < 0.05 vs. those at DP by analysis of variance (ANOVA), followed by Bonferroni-modified t-test. C: plots of rates of [Ca2+]i decline (d[Ca2+]i/dt) against [Ca2+]i at CTL (solid line), DP (dashed line), and RP (dotted line) in a representative example. D: time constants of twitch [Ca2+]i decline against their peak [Ca2+]i at CTL (), DP (), and RP (). In (C) and (D), the [Ca2+]i values in the x-axis was limited <1 µM to make the difference between DP and RP more apparent. The plotted number of DP and RP therefore decreased to eight and six, respectively.

Effects of respiratory acidosis on rate of twitch [Ca²+]_i decline. The SR Ca²⁺ content is determined both by [Ca²⁺]_i and by Ca²⁺ uptake through the SR Ca²⁺-ATPase (5). Because the increase in CaT amplitude during acidosis was accompanied by the increase in the SR Ca²⁺ content but was not simply dependent on diastolic [Ca²⁺]_i, the increase in the SR Ca²⁺ content may be due to reactivation of SR Ca²⁺ uptake. In mammalian ventricular myocytes, the SR Ca²⁺ uptake can be estimated from the rate of twitch [Ca²⁺]_i decline (6, 19). Figure 3B shows the mean time constants () of twitch [Ca²⁺]_i decline obtained from single exponential fits. The [Ca²⁺]_i decline slowed in all cells at DP but reaccelerated at RP in Rec (see also Fig. 2B). In NRec, such a reacceleration was never seen. However, the rate of twitch [Ca²⁺]_i decline depends on the range of [Ca²⁺]_i over which it is measured (6). Because the CaT amplitude increased during acidosis, the observed change in the time constant might reflect the change in amplitude rather than that in SR Ca²⁺ uptake. To rule out this possibility, the rate of twitch [Ca²⁺]_i decline in a given [Ca²⁺]_i was investigated. The exponential curve fitted to the declining phase of CaT was differentiated to give the rate of change of [Ca²⁺]_i (d[Ca²⁺]_i/dt) and plotted against [Ca²⁺]_i (Fig. 3C, see also Ref. 19). The d[Ca²⁺]_i/dt decreased in any given [Ca²⁺]_i at DP, but partially recovered at RP. Similar results were obtained in all of Rec. Figure 3D also indicates that the time constants of twitch [Ca²⁺]_i decline recovered at RP for similar peak [Ca²⁺]_i. Thus the reacceleration of [Ca²⁺]_i decline at RP was not secondary to the increase in CaT amplitude.

Effects of metabolic acidosis on twitch cell contraction and CaT. Metabolic acidosis (with 20 mM propionate, pH 6.8) produced changes similar to those in respiratory acidosis. Seven of fourteen myocytes (10 rats) showed contractile recovery, whereas the remaining seven cells did not. In Rec, the CaT amplitude increased from 320 ± 73 nM at control to 464 ± 64 nM at DP and to 656 ± 129 nM at RP, whereas the increase in diastolic [Ca²⁺]_i was small (from 417 ± 76 nM at DP to 429 ± 101 nM at RP). The twitch [Ca²⁺]_i decline initially slowed in all of the cells but reaccelerated in Rec (: 442 ± 26 ms at DP and 310 ± 15 ms at RP, P < 0.05). In NRec, neither the large increase in CaT amplitude nor the reacceleration of the twitch [Ca²⁺]_i decline was observed, while the diastolic [Ca²⁺]_i continued to increase (data not shown). However, soon after returning to control HEPES solution, all the cells exhibited hypercontracture. This morphological change was possibly ascribed to both Ca²⁺ overload and augmented myofibrillar Ca²⁺ responsiveness by an alkaline overshoot.

Inhibition of the SR function and acidosis-induced changes in twitch cell contraction and CaT. To investigate further the correlation between the contractile recovery and the SR Ca²⁺ uptake, the twitch cell contraction and CaT were monitored in the cells where the SR function was abolished with thapsigargin (1 µM) and ryanodine (10 µM). For this experiment, Ca²⁺ influx via sarcolemmal Ca²⁺ channels was increased with BAY K 8644 (100 nM) to make twitch contractions and CaT comparable. BAY K 8644 acts directly to sarcolemmal L-type Ca²⁺ channels and enhances Ca²⁺ currents by increasing open time of the channels (17). The cells were preincubated for 30 min with these agents, and the decreases in twitch contraction and CaT amplitude were limited to ~50%. Figure 4 shows a typical example of myocytes that were electrically stimulated at 0.1 Hz while being perfused with control and acidic solution. The SR inhibition apparently slowed the twitch [Ca²⁺]_i decline and relaxation. The twitch contraction disappeared soon after the perfusion of acidic solution and never recovered. The diastolic [Ca²⁺]_i continued to increase, but the increase in CaT amplitude was small. The slow twitch [Ca²⁺]_i decline prolonged further during acidosis, and reacceleration did not occur. Figure 5 summarizes the changes in twitch contraction (Fig. 5A), diastolic [Ca²⁺]_i and CaT amplitude (Fig. 5B), and time constant of twitch [Ca²⁺]_i decline (Fig. 5C) during respiratory acidosis (n = 9, 3 rats). The twitch contraction decreased during acidosis and never recovered in all cells. The diastolic [Ca²⁺]_i continued to increase, and the CaT amplitude initially increased but soon reached plateau. Respiratory acidosis further slowed the [Ca²⁺]_i decline, but the reacceleration was never observed. Thus the functional SR was necessary for the reacceleration of twitch [Ca²⁺]_i decline, increase in CaT amplitude and contractile recovery during prolonged acidosis.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Effects of respiratory acidosis on twitch cell shortening and CaT in a myocytes in which the SR function was blocked. A representative myocyte preincubated with thapsigargin (Tg; 1 µM) and ryanodine (Ry; 10 µM) for 30 min was exposed to respiratory acidosis. The cell was electrically stimulated at 0.1 Hz and Ca2+ influx via L-type Ca2+ channels was increased with BAY K 8644 (BayK: 100 nM). The changes in twitch contraction (A) and CaT (B) during respiratory acidosis were shown. The twitch contraction decreased during acidosis and never recovered. Diastolic [Ca2+]i continued to increase and the CaT amplitude initially increased but soon reached plateau. SR inhibition largely slowed the twitch [Ca2+]i decline. The acidosis further slowed the [Ca2+]i decline but reacceleration was never observed. Dashed lines indicate the baselines of cell shortening and CaT.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of SR inhibition on acidosis-induced changes in twitch cell shortening and CaT. Summary data of the changes in twitch contraction (A), diastolic [Ca2+]i (), and CaT amplitude () (B), and time constant of twitch [Ca2+]i decline (C) during respiratory acidosis in myocytes where the SR function was blocked. Values are means ± SE from 9 myocytes (3 rats). * P < 0.05 vs. the values at time 0 by ANOVA, followed by Bonferroni-modified t-test.

Acidosis-induced changes in twitch cell contraction and CaT under phosphorylated state. From the experiments above, the reactivation of SR Ca²⁺ uptake was necessary to the reacceleration of twitch [Ca²⁺]_i decline. The SR Ca²⁺ uptake is activated through phosphorylation of PLB and also through direct phosphorylation of SR Ca²⁺-ATPase (5, 28, 38, 40). Therefore, the next question is whether the phosphorylation is involved in the reactivation of SR Ca²⁺ uptake during prolonged acidosis. To examine this, the myocytes in normal Krebs solution with 1.5 mM Ca²⁺ were treated with isoproterenol (10 µM) and cantharidin (10 µM), an inhibitor of protein phosphatases (7) (Fig. 6). Treatment with cantharidin slightly increased the SS CaT amplitude and accelerated the twitch [Ca²⁺]_i decline but did not change the twitch contraction. The addition of isoproterenol largely augmented these changes and also increased the twitch contraction (Fig. 6, A-C). However, application of cantharidin after isoproterenol treatment induced no additive effects on these parameters (Fig. 6, D-F). The doses of isoproterenol and cantharidin exhibited maximum effects because higher concentrations of either drug induced no additional changes. Decreasing extracellular [Ca²⁺] to 0.5 mM matched the SS contraction and CaT amplitude with the preexposed levels but the twitch [Ca²⁺]_i decline remained accelerated (Fig. 7 and Table 2). Under the condition, contractile recovery during respiratory acidosis never occurred (n = 12, 6 rats). The diastolic [Ca²⁺]_i continued to increase, but CaT amplitude did not change, significantly. The twitch [Ca²⁺]_i decline prolonged and never recovered.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Effects of isoproterenol (Iso) and cantharidin (Can) on twitch cell shortening and CaT. Changes in twitch contraction (A and D), diastolic [Ca2+]i () and CaT amplitude () (B and E), and time constant of twitch [Ca2+]i decline (C and F). A-C: the effects of Can (10 µM) alone and after addition of Iso (10 µM). D-F: the effects of Iso alone and after addition of Can were compared. Values are means ± SE from 8 and 9 myocytes (2 and 3 rats, respectively). * P < 0.05 vs CTL,  P < 0.05 vs. Can by ANOVA, followed by Bonferroni-modified t-test.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effects of phosphorylation on acidosis-induced changes in twitch cell shortening (A) and CaT (B). Myocytes were initially treated with Iso (10 µM) and Can (10 µM) and then exposed to respiratory acidosis. The pretreatment of a myocyte with Iso and Can increased twitch contraction and CaT amplitude, and shortened the twitch [Ca2+]i decline. The perfusate [Ca2+] was reduced to 0.5 mM to match the steady-state contraction and CaT amplitude with values at control (Iso + Can). The twitch contraction decreased during acidosis and never recovered. The diastolic [Ca2+]i continued to increase, and the CaT amplitude did not increase significantly. The acidosis prolonged the [Ca2+]i decline but reacceleration was never observed. [Ca2+]o, extracellular [Ca2+].

Effects of KN-93 on acidosis-induced changes in twitch cell contraction and CaT. The PLB and SR Ca²⁺ ATPase are mainly phosphorylated by cAMP-dependent protein kinase (PKA) and CaMKII. Because the increase in CaT amplitude and reacceleration of twitch [Ca²⁺]_i decline were preceded by the increase in diastolic [Ca²⁺]_i, it was hypothesized that the CaMKII-dependent phosphorylation contributed primarily to the recovery of SR Ca²⁺ uptake. Thus the twitch contraction and CaT were monitored during respiratory acidosis with a selective inhibitor of endogenous CaMKII, KN-93 (1 µM) (39). The perfusion of KN-93 decreased SS twitch contraction, CaT amplitude, and prolonged twitch [Ca²⁺]_i decline. Increasing extracellular [Ca²⁺] to 2.5 mM restored SS contraction and CaT amplitude toward the preexposed levels but the twitch [Ca²⁺]_i decline remained prolonged (Fig. 8 and Table 3). KN-93 did not alter these parameters at the cells in which the SR function was abolished with ryanodine and thapsigargin (data not shown). Contractile recovery during respiratory acidosis was observed in only 3 of 11 cells (27%; from 0.9 ± 0.8% of diastolic length at 4 min to 2.0 ± 0.3% at 10 min, 5 rats). The diastolic [Ca²⁺]_i continued to increase, but CaT amplitude did not change, significantly. The twitch [Ca²⁺]_i decline prolonged and never recovered. The effects of KN-93 were also identical at the cells exposed to metabolic acidosis (Table 3). The contractile recovery was seen in only 2 of 11 cells (4 rats).

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Effects of 2-[N-(2-hydroxyethyl)-N-(4-methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl) methyl benzamine (KN-93) on acidosis-induced changes in twitch cell shortening (A) and CaT (B). The pretreatment of a myocyte with KN-93 (1 µM) decreased twitch contraction and CaT amplitude, and prolonged the twitch [Ca2+]i decline. The perfusate [Ca2+] was increased to 2.5 mM to match the steady state contraction and CaT amplitude with values at control (KN-93). The twitch contraction decreased during acidosis and never recovered. The diastolic [Ca2+]i continued to increase, and the CaT amplitude did not increase significantly. The acidosis further slowed the [Ca2+]i decline, but reacceleration was never observed.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The major findings in the present study are the following: 1) with prolonged acidosis, there was recovery of twitch contractions, which was primarily due to stimulation of SR Ca²⁺ uptake, and 2) the recovery process was at least partly associated with CaMKII-dependent phosphorylation.

Measurements of [Ca²+]_i and estimation of SR Ca²⁺ uptake. We have to consider the possible factors that may perturb the present [Ca²⁺]_i measurement. The first factor was incomplete hydrolysis of indo-1-AM. The myocytes were incubated for 30 min after dye loading at room temperature to gain hydrolyzed fluorescent dye forms. The second factor was compartmentalization of indo-1. In rat ventricular myocytes, the residual fluorescence of indo-1 after treatment of digitonin was reported to be relatively high (40-50% of preexposed level) (33). Therefore, we performed in vivo calibration to get more precise values for [Ca²⁺]_i. The third factor was photobleaching of indo-1 fluorescence. Illumination from the xenon lamp was minimized by neutral density filters, which reduce photobleaching. The diastolic [Ca²⁺]_i estimated from our method was 75-135 nM, which was nearly identical to that previously reported using fura 2 or indo-1-free acid with patch-clamp technique (6, 15).

Inhibition of CaMKII by KN-93. In this study, we used KN-93, an isoquinoline sulfonamide derivative, at the concentration of 1 µM to inhibit CaMKII. KN-93 elicits a potent inhibition on CaMKII by competing with calmodulin [inhibitory constant (K_i); 0.37 µM]. The K_i values of KN-93 for other protein kinases, such as PKA, protein kinase C, and myosin light-chain kinase, are all >30 µM (39). Therefore, the dosage we used seemed to be specific to CaMKII, but the inhibition against the CaMKII-dependent reactivation of SR Ca²⁺ uptake (and/or SR Ca²⁺ release) might be incomplete. Actually, some myocytes showed contractile recovery during prolonged acidosis. In these cells, the CaT amplitude tended to increase from DP to RP (data not shown), although the prolonged twitch [Ca²⁺]_i decline never reaccelerated.

Contractile recovery during prolonged acidosis. During prolonged acidosis, the initially decreased twitch contraction slowly recovered, and this recovery was accompanied by a continuous increase in CaT amplitude. The diastolic cell lengths did not alter presumably because of reduced Ca²⁺ responsiveness of contractile proteins (12, 34). Hulme and Orchard (19) suggested that although acidosis inhibits both SR Ca²⁺ uptake and CICR from the SR, these effects are compensated by an increase in SR Ca²⁺ content secondary to a rise in [Ca²⁺]_i. Indeed, we documented that the SR Ca²⁺ content (evaluated with caffeine-induced CaT) continued to increase during acidosis, and the inhibition of SR function completely abolished the contractile recovery. However, there was a cell-to-cell heterogeneity in the contractile recovery both in respiratory and metabolic acidosis. In Rec, the CaT amplitude increased further and the twitch [Ca²⁺]_i decline reaccelerated from DP to RP, whereas in NRec these changes did not occur. In contrast, the increase in diastolic [Ca²⁺]_i was more extensive in NRec. Thus the increase in [Ca²⁺]_i is not the sole mechanism for the contractile recovery; however, reactivation of SR Ca²⁺ uptake seems to be crucial, although we could not compare the SR Ca²⁺ contents between Rec and NRec.

Reactivation of SR Ca²+ uptake during prolonged acidosis. The SR Ca²⁺ uptake is activated through phosphorylation of PLB and can also be stimulated by direct phosphorylation of SR Ca²⁺-ATPase. The PLB and SR Ca²⁺-ATPase are mainly phosphorylated by PKA (28, 38), CaMKII (4, 27, 41, 43), and are dephosphorylated by protein phosphatase-1 (PP1) (29, 32). Because acidosis increased [Ca²⁺]_i, and the reacceleration of twitch [Ca²⁺]_i decline was prevented under phosphorylated state, the CaMKII-dependent phosphorylation was thought to be primary for the reactivation of SR Ca²⁺ uptake. The importance of endogenous CaMKII for normal excitation-contraction coupling has been reported in intact mouse and ferret myocytes (26, 27). Actually, KN-93 prolonged the twitch [Ca²⁺]_i decline at control perfusion and abolished its reacceleration at prolonged acidosis (Fig. 7 and Table 3).

Physiological implications. Control of pH_i is important in all tissues for optimal functioning of metabolic enzymes, contractile elements, protein synthesis, and ion conductivity (21). Myocardial acidosis caused by ischemia, renal, or respiratory failure has generally been considered detrimental. Our finding that the CaMKII-induced reactivation of SR Ca²⁺ uptake increased the CaT amplitude and limited the contractile dysfunction proposes a compensatory reaction of myocytes against the decreased Ca²⁺ responsiveness of contractile protein. The prevention of extensive increase in diastolic [Ca²⁺]_i may also provide a protective mechanism against cellular Ca²⁺ overload.