INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

HYPOKALEMIA is an extremely common medical condition that is known to have serious consequences in selected clinical situations (32). Its importance in the general population is still unclear (13, 15). It is likely that hypokalemia may become even more common because of changing treatment recommendations for essential hypertension, which affects nearly 20% of the adult population of the United States. Specifically, the recent Joint National Committee recommendations call for the use of a thiazide diuretic as initial treatment of essential hypertension (16). If we keep in mind that dietary sodium intake by Americans is extremely high and is difficult to modify with prescribed diet, and that the use of thiazides when dietary sodium is not restricted commonly produces ongoing potassium losses (2), it is not hyperbole to say that the public health implications of hypokalemia are potentially enormous. Against this background, it is clear that our understanding of how hypokalemia affects cardiac metabolism and function is still incomplete.

Our laboratory has previously demonstrated that acute, severe decreases in perfusate potassium concentrations caused marked decreases in NMR-visible tissue potassium and ATP concentrations in the isolated, perfused heart. We found that these energetic changes appeared to be independent of the development of ventricular fibrillation and were influenced by perfusate calcium concentrations (27). In short, our data were consistent with the hypothesis suggested by Hoerter and colleagues (14) that acute decreases in perfusate potassium impair cellular calcium extrusion through the Na/Ca exchanger by decreasing the sodium gradient across the plasmalemma. Returning to the clinical scenario, it appears that mild acute hypokalemia and prior diuretic use both impact negatively on patient outcome during myocardial infarction (11). With this in mind, we performed the present studies to examine the effects of moderate acute and chronic hypokalemia on hypoxic injury using the isolated, perfused heart model.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Male Sprague-Dawley rats (350-400 g) were used in these experiments. The rats were anesthetized with ketamine (1 ml/kg body wt, 50 mg/dl) and xylazine (0.6 ml/kg body wt, 20 mg/ml) before surgery. The technique used for removal of the heart was reported previously (27). Some animals were fed a potassium-deficient diet described by Linas and Marzec-Calvert (24). Specifically, an extremely low-potassium rat chow preparation (5 meq/kg; ICN Pharmaceuticals) was used for between 3 and 4 wk. This caused a reduction in plasma potassium values to ~2.0 mM (2.05 ± 0.02, N = 20 rats) in our animals before hearts were harvested.

Isolated heart perfusion methodology. A Langendorff heart preparation modified for serial NMR spectroscopy studies was used. Circuit design allowed for maintenance of constant perfusion pressure (usually 80 mmHg) and temperature (37°C). An isovolumic preparation was used. Ventricular pressure was monitored continuously using a latex balloon filled to achieve an end-diastolic pressure of 8-12 mmHg at the beginning of each experiment without changing balloon volume for the duration of the experiment. These data were sampled for 10 s every 2.5 min and digitized with a commercial A-D board (DAS8, Metrabyte) into a personal computer and manipulated with software written in our laboratory (30). Developed pressure refers to the difference between systolic and diastolic pressure, dP/dt refers to the maximum slope of the pressure increase, negative dP/dt refers to the maximum negative slope of the pressure decrease, and left ventricular end-diastolic pressure (LVEDP) refers to the minimum pressure obtained during diastole. Each of the reported physiological data values was averaged from the beats observed in the 10-s sampling windows from each experiment.

NMR spectroscopy. All NMR spectroscopy studies were performed on a 7.05-Tesla, 10-cm-bore cryomagnet with a AM-300 spectrometer (Bruker Instruments, Billerica, MA).

Fluorescence spectroscopy. Fluorescence spectroscopy was used to monitor cytosolic calcium during experimental perfusion. All experiments were performed using a Perkin-Elmer LS50B Luminescence Spectrofluorimeter equipped with a fiber-optic probe. The method of calcium determination was largely derived from work published by Brandes and co-workers (6-8). The specifics were as follows. The tip of the fiber-optic probe contained emission and excitation light fibers within a 1.5-mm-diameter area. This fiber-optic probe was placed adjacent to the isolated heart, and motion of the heart relative to the probe was restricted. Baseline autofluorescence values were determined using an excitation wavelength of 350 nm and emission wavelengths of 390 and 460 nm. Although indo 1 as an emission dye is well suited for rapid acquisitions so that calcium transients can actually be recorded (8), this was not possible with our instrumentation. Therefore, data were obtained averaged over the cardiac cycle using 2-s acquisition periods per emission wavelength. Excitation was gated off for two-thirds of the time to avoid photobleaching of the indicator. Hearts were loaded with indo 1 by perfusing for 20 min with perfusion media to which 3 µM indo 1 AM was added. Hearts were then perfused with normal perfusion media for an additional 30-min period before any experimental manipulations were performed to establish a baseline. Hearts were then exposed to 30 min of experimental perfusion (hypoxia ± alterations in perfusate potassium) and then 30 min of recovery perfusion. Some hearts were studied without indo 1 to establish changes in autofluorescence with these experimental manipulations. On the basis of these data (N = 3 in each group), autofluorescence values were then predicted for the course of the study, and the actual data were then corrected for these values. Results are reported as corrected ratios between the 390- and 460-nm wavelength values. Using simulated low- and high-calcium conditions with tissue homogenates as reported (6, 7), low- and high-calcium ratios were measured to be R_min = 0.05 and R_max = 0.83, respectively, on our system. For calculations of cytosolic calcium, the calculation where [Ca] is calcium concentration, K_d is dissociation constant, and R_obs is the observed autofluorescence ratio, was performed. Although pH_i was not constant throughout all experimental manipulations (see Table 2), we assumed a constant K_d of 844 nM for indo 1 reported by Bassani et al. (5) for calculations of cytosolic [Ca]. This approach probably led to a systematic underestimate of cytosolic [Ca] during hypoxia and reflow, because the K_d of indo 1 is known to increase dramatically with decreases in pH (4). Data for baseline, experimental, and recovery periods were obtained by averaging the last 2 min of recording during each measurement period.

Isolated mitochondria. In selected experiments mitochondria were isolated from perfused hearts in a manner identical to that which we previously reported (27). Specifically, studies were limited to hearts perfused with standard perfusate for 1 h or hearts perfused with standard, oxygenated perfusate for 30 min followed by 30 min of perfusion with hypoxic perfusate containing 2 or 4 mM potassium. The respiration of these isolated mitochondria was then studied using a computer-based respirometer with various substrates (malate/glutamate, succinate, and octanoate) both during and after coupling with ADP (states 3 and 4, respectively). Respiration studies were performed on each experimental preparation in either duplicate or triplicate. The average of these values for each experimental preparation was then used for statistical analysis.

Experimental groups. In hearts exposed to the hypoxia-recovery (physiology and ³¹P NMR spectroscopy or fluorescence spectroscopy) or hypoxia-alone (isolated mitochondria) protocols, we assigned the following group numbers. Hearts that were derived from normal animals and perfused with a perfusate potassium concentration of 4.0 mM are referred to as group I. Hearts that were derived from normal animals and perfused with a perfusate potassium concentration of 2.0 mM are referred to as group II. Hearts that were harvested from rats maintained on a potassium-deficient diet for 3-5 wk (serum potassium 2.0 ± 0.1 mM) and perfused with a perfusate potassium concentration of 4.0 or 2.0 mM are referred to as groups III and IV, respectively.

Statistical analysis. Data were compared using one- or two-way ANOVA and unpaired or paired Student's t-test with Scheffé's correction for multiple comparisons depending on the unpaired or paired nature of the data (33). Statistical analysis was performed using CRUNCH4 software.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Functional and ³¹P NMR studies. Exposure of normal animals (N = 6) to perfusion media with 4 mM potassium resulted in essentially stable functional performance (baseline dP/dt 1,440 ± 110 vs. 1,320 ± 123 mmHg/s at 90 min) and energetics (ATP = 0.94 ± 0.08 × initial value at 90 min of perfusion). Exposure of normal animals (N = 6) to perfusion media with 2 mM potassium resulted in stable functional performance (baseline dP/dt 1,539 ± 152 vs. 1,478 ± 165 mmHg/s at 90 min) and minor alterations in energetics (ATP 0.85 ± 0.12 at 90 min, P < 0.05). In chronically hypokalemic animals, function and energetics were also essentially stable for perfusion with 4 mM potassium perfusate (baseline dP/dt = 1,280 ± 156 vs 1,265 ± 143 mmHg/s at 90 min and ATP = 0.93 ± 0.07 initial at 90 min; N = 6 experiments). Perfusion of hearts isolated from chronically hypokalemic animals (N = 6) also demonstrated stable function and energetics for perfusion with 2 mM potassium perfusate (baseline dP/dt = 1,320 ± 120 vs. 1,240 ± 140 mmHg/s at 90 min of perfusion and ATP = 0.91 ± 0.10 initial at 90 min of perfusion).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Comparison of functional responses [maximum slope of pressure increase over time (dP/dt)] and metabolic responses [ATP concentration ([ATP])] to hypoxia-recovery protocol in isolated hearts. , group I; , group II; , group III; , group IV. Data are expressed as means ± SE. Statistical comparisons shown in Tables 1 and 2.

Fluorescence spectroscopy data. As shown by the fluorescence spectroscopy data, hearts isolated from both control and chronically hypokalemic rats appeared to maintain stable time-averaged cytosolic calcium ratios for 90 min of nonhypoxic perfusion (based on 390- to 450-nm emission ratio) when perfused with either 2 or 4 mM potassium media (N = 4 time controls in each group). These data were not significantly different among the four groups, nor were they different from the baseline data shown in Table 3. In response to the hypoxia-recovery protocol, group I hearts had stable cytosolic calcium values, whereas group II, III, and IV hearts demonstrated increases with hypoxia and recovery. These data are summarized in Table 3. Representative tracings of the corrected 390/450 ratios are shown in Fig. 2. It appeared that perfusion with 2 mM potassium perfusate allowed for greater increases in cytosolic calcium during hypoxia and recovery than 4 mM potassium perfusion in both normal and chronically hypokalemic animals (group II vs. group II and group IV vs. group III; both P < 0.05).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Representative tracings of corrected 390- to 460-nm ratio in 4 experimental groups. A: group I. B: group II. C: group III. D: group IV.

Isolated mitochondria data. The isolated mitochondria data are summarized in Table 4. Group II mitochondria displayed impaired respiratory function compared with group I mitochondria with octanoate or succinate as substrate. However, the mitochondria isolated from chronically hypokalemic animals (groups III and IV) demonstrated significantly worse respiratory function at either perfusate potassium concentration. Moreover, in the mitochondria isolated from chronically hypokalemic animals, we were not able to demonstrate a significant effect of acute lowering of perfusate potassium concentration in the hypoxia protocol. In other words, the respiratory function of group III and IV mitochondria were essentially the same.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We observed that both acute and chronic hypokalemia adversely affected metabolic and physiological responses to hypoxia. Although neither acute nor chronic hypokalemia of the degree used in this study (extracellular [K] = 2 mM) had adverse physiological effects under oxygenated perfusion conditions, both metabolic and physiological responses to hypoxia as well as recovery from hypoxic injury appeared to be impaired by both acute and chronic hypokalemia. Moreover, with respect to the physiological and energetic responses, a clear additive effect of both acute and chronic hypokalemia in this setting was discernible. In the experimental groups studied, energetic (as defined by tissue ATP concentrations) and physiological (defined by dP/dt) responses to hypoxia-recovery appeared to be closely coupled.

The mechanism by which acute hypokalemia impairs responses to hypoxia may be related to cellular calcium metabolism. On this note, we should point out that although time-averaged fluorescence values were not significantly altered by hypoxia alone in normal hearts, the slower heart rate noted in this setting could, by itself, produce lower time-averaged values even if end-diastolic calcium values were not lower. Koretsune and Marban (21) observed lower end-diastolic calcium values in the isolated heart subjected to hypoxia. Allen and Orchard (1) and Orchard et al. (26) reported no change in cytosolic calcium transients in ventricular strips subjected to hypoxia or metabolic inhibition of electron transport. These findings differ from what has been reported during ischemia, where rapid increases in time-averaged and end-diastolic cytosolic calcium concentrations have been consistently observed (10, 22). Although pH_i responses to hypoxia and ischemia may account for these differences, it is also possible that impairment of glycolysis during ischemia directly impairs sarcoplasmic reticulum calcium reuptake during diastole (1, 26).

Although the degree of acute hypokalemia used in this study by itself did not result in demonstrable increases in time-averaged cytosolic calcium, in the setting of hypoxia and recovery, profound increases in cytosolic calcium were observed. As a correlate, mitochondrial function appeared to be adversely affected in hearts removed from normal animals when perfusate potassium concentration was decreased. In the hearts isolated from chronically hypokalemic animals, the performance of the hypoxia-recovery protocol in the setting of low perfusate potassium also resulted in marked increases in cytosolic calcium, whereas only small differences in cytosolic calcium responses to hypoxia recovery could be discerned between the chronically hypokalemic and normal animals when perfusate potassium concentration was either 2 or 4 mM. In contrast, mitochondrial function appeared to be quite impaired after hypoxic perfusion of hearts harvested from chronically hypokalemic animals at either perfusate potassium concentration. In short, both functional and energetic responses to hypoxia-recovery perfusion correlated imperfectly with the degree of cytosolic calcium elevations observed or the degree of mitochondrial respiratory derangement viewed independently. These imperfect correlations underscore the complexity of factors involved in the regulation of cytosolic calcium as well as excitation-contraction coupling.

We anticipate that acute hypokalemia impairs Na-K-ATPase which, in turn, results in increases in cytosolic sodium (19). This would be expected to impair Na/Ca exchange and to allow cytosolic calcium to rise, as has been suggested by Hoerter and colleagues (14) as well as by our group (27). In the setting of hypoxia, the ability of the cardiac cell to maintain cytosolic calcium concentrations at normal levels through other extrusion pathways (e.g., the plasmalemmal Ca-ATPase) as well as mitochondrial uptake (9, 29) might be further impaired, and increases in cytosolic calcium would result, as was observed (23). Because chronic hypokalemia does not cause upregulation of Na-K-ATPase activity (3), we speculate that chronic adaptation to chronic hypokalemia might include upregulation of processes that might normalize cytosolic calcium in the setting of chronic increases in cytosolic sodium, specifically mitochondrial uptake of calcium and plasmalemmal Na/Ca exchange activity. These processes might result in an increase in mitochondrial injury during hypoxic conditions, because reversal of this upregulated plasmalemmal Na/Ca exchanger might allow for increased flux of calcium into the cell (21, 25). If mitochondrial calcium uptake capacity were increased, this calcium could then have greater access to the mitochondrial space to further injure respiratory function in this setting. Alternatively, some enzyme systems that are activated by calcium and that may exert deleterious effects in the setting of hypoxia, such as phospholipase A₂ (20), could be upregulated by chronic hypokalemia. However, these assertions remain highly speculative at this point, and further investigation must be pursued to test these hypotheses.

On a less speculative level, we observed that mitochondrial responses to hypoxia were comparably impaired when chronically hypokalemic hearts were perfused with either normal or reduced perfusate potassium concentrations. However, the increases in time-averaged cytosolic calcium concentrations were much more marked in those hearts perfused with a reduced perfusate potassium concentration. These data suggest that the additive injury of acute and chronic hypokalemia could be explained by the relatively increased energy demands produced by acute hypokalemia with decreased respiratory capacity during and after hypoxia that occurred with chronic potassium depletion. The increases in LVEDP and impairment in diastolic relaxation (negative dP/dt) detailed in Table 1 would appear to support the concept that acute hypokalemia increases energy demands in both normal and chronically hypokalemic hearts exposed to hypoxia.

For reasons discussed in the introduction, we feel that the clinical implications of these observations are considerable. Mild to moderate degrees of chronic hypokalemia are induced commonly with potassium-wasting diuretics. Moreover, acute hypokalemia either induced by sympathetic tone shifting potassium into cells in stress settings or artificially produced with pharmacological administration of -agonists or insulin or even acute hemodialysis procedures also occurs commonly (18). Finally, the coincidence of acute and/or chronic hypokalemia with acute myocardial hypoxic or ischemic insult may also be quite common, and the clinical consequences of this "coincidence" may be profound (11, 17, 28).

In summary, we found that acute hypokalemia worsened cardiac physiological and metabolic responses to hypoxia. Somewhat surprisingly, hearts subjected to chronic hypokalemia had worse physiological and metabolic responses to hypoxia than hearts obtained from control rats at both normal and reduced perfusate potassium concentrations. The physiological and metabolic responses to acute hypokalemia could be explained by differences in cytosolic calcium metabolism, whereas the differences between chronically hypokalemic and normal hearts probably involve other mechanisms that affect mitochondrial respiratory function.