Propofol-induced Modifications of Cardiomyocyte Calcium Transient and Sarcoplasmic Reticulum Function in Rats

Cardiac myocytes were obtained from the hearts of male Wistar rats (270–300 g). All procedures were performed according to the Guiding Principles in the Care and Use of Animals of the American Physiologic Society, Bethesda, Maryland, with the agreement of the French Ministry of Agriculture, Paris, France. Cardiomyocytes were isolated using the procedure of Powell and Twist, 14with slight modifications. 15After anesthetization of the rat (thiopental 100 mg/kg intraperitoneally), a median thoracotomy was performed. The heart was excised, the aorta was cannulated, and the heart was perfused under a Langendorff column, allowing a retrograde perfusion of the heart for 5 min by a Ca²⁺-free solution, followed by the infusion of the same buffer, to which 1.2 mg/ml collagenase A (Bohringer-Manheim, Meylan, France) was added.

After a 45-min perfusion period, the heart was taken from the column and cut into small (1-mm) pieces in the same Ca²⁺-free buffer. Cells were filtered and washed, and Ca²⁺(1.2 mM) was reintroduced. At the end of this procedure, approximately 80% of myocytes showed normal architecture, were quiescent, and could be electrically stimulated. Cardiomyocytes were then resuspended in a culture medium (BM86 Wisler; IMV Lab, L’Aigle, France) and placed on Petri dishes previously coated with laminin (30 μg/ml). Cells were incubated for 3 h at 37°C in the presence of 5% carbon dioxide on these dishes. At the end of the period, cells adhered to the bottom of the dish, and the culture medium was changed. Cells were used 12 h later for the study.

Before each experiment, the cells were preloaded for 15 min at room temperature with Indo-1 AM by incubation in 200 μl culture medium containing 10 μg Indo-1 AM, 180 μl bovine serum albumin, and 5 μl dimethylsulfoxide with 5 μg pluronic acid. The cells were then washed and maintained at room temperature for 40 min in the culture medium before the fluorescence measurements. The nutrient medium was replaced with a Krebs solution containing 117 mM NaCl, 5.7 mM KCl, 1.2 mM KH²PO⁴, 1.7 mM MgCl², 4.4 mM NaHCO³, 1.25 mM CaCl², 10 mM glucose, 10 mM HEPES, 10 mM creatine, and 20 mM taurine and was buffered to pH 7.4.

Cytosolic Ca²⁺measurements were performed by dual-emission microfluorometry (Hamamatsu, Massy, France) with the Indo-1 AM probe, as previously described. 15Cells loaded with the fluorescent probe were excited at 360 nm. Emission spectra were divided in two halves using interference filters at 405 and 480 nm. The fluorescence ratio 405:480 nm, which is independent of the probe concentration, was calculated directly from the two signals.

Electrical pacing was performed using a Harvard stimulator (Harvard, Les Ulis, France) delivering 40-V stimuli of a duration of 10 ms at a frequency of 1 Hz. Cells were perfused continuously at a rate of 2 ml/min by the Krebs solution at room temperature. The solution was bubbled previously by a gas mixture (95% oxygen, 5% carbon dioxide). We verified previously that if this procedure is used, pH is not modified for at least 2 h.

Each cell isolation allowed the seeding of approximately 30 Petri dishes. One to three cells could be observed in each field of the microscope. Data were recorded for 8.52 s (72 image acquisitions separated by 120-ms intervals). We verified previously that this procedure did not produce a significant attenuation of the measured peak of the Ca²⁺transient by comparing it with that obtained at the maximal rate of acquisition of the system (50 Hz; that is, 72 points separated by 20-ms intervals). In the time interval separating two recording periods, the electrical stimulation was interrupted and the pacing was initiated after 2 s of recording (fig. 1). After recording of the control state, the bath inside the Petri dish was aspirated and replaced by the solution of the drug to be tested, which was continuously infused for 5 min. Each Petri dish was used for either propofol or its solvent, 10% Intralipid (Kabi Pharmacia, Noisy le Grand, France) at cumulative drug concentrations of 10⁻⁷, 10⁻⁶, 10⁻⁵, and 10⁻⁴M. Intralipid contained 100 g purified soybean oil, 12 g purified egg phospholipids, 22.5 g glycerol USP, and 1,000 ml water. We called 10⁻⁷, 10⁻⁶, 10⁻⁵, and 10⁻⁴M doses of Intralipid the concentrations of Intralipid corresponding to those used for 10⁻⁷, 10⁻⁶, 10⁻⁵, and 10⁻⁴M doses of propofol. Drugs (propofol, 10% Intralipid) were infused in random order. A similar protocol using Krebs solution was also performed to verify the effect of repeated measurements on cytosolic Ca²⁺concentration. Lastly, it was verified that propofol and its solvent did not interfere with Indo-1 fluorescence. Doses of 10⁻⁴M of propofol, or Intralipid at corresponding doses, did not induce significant change in autofluorescence levels compared with those measured in the Krebs solution.

The mean ratio image of each cell was calculated by the average of all points included in the area of the cell. In each cell, three to five areas that included approximately 2,500 pixels were selected. The transformation of 405:480 ratio values into Ca²⁺concentrations takes into account dissociation constant (Kd) values and concentrations of Ca²⁺but also those of other ions, particularly H⁺. Because ion concentrations are different in different parts of the cells and in different conditions, the use of a fixed formula to obtain absolute values of Ca²⁺concentration may be misleading. Therefore, rather than trying to express cellular Ca²⁺concentration in absolute values, we preferred to present only the 405:480 ratio values as indices of Ca²⁺concentration.

If different cells, obtained from the same cell isolation, had the same drug infusion, the mean value for each measurement was calculated so that each cell isolation gave only one value for each intervention. Several time points were measured (fig. 1), including the end-diastolic ratio (at the bottom of the Ca²⁺rising) after at least 2 min of pacing interruption (D^rest) and the peak of the Ca²⁺ratio (systolic Ca²⁺, or S¹). The end-diastolic and end-systolic Ca²⁺ratios after the first stimulation were called D²through D⁵and S²through S⁵, respectively. The Ca²⁺transient (or Δ ratio) was defined as the difference between the systolic Ca²⁺ratio and the end-diastolic Ca²⁺ratio (systolic–diastolic ratio). We determined (fig. 1) the difference between D²and D^rest, which gives information about the ability of the cell to bring back diastolic Ca²⁺to normal resting values after a short pause and thus tests the SR pump or the Na⁺–Ca²⁺exchanger activity. The time course of the decrease of the Ca²⁺ratio after the peak of Ca²⁺transient was measured every 120 ms, corresponding to the sampling rate for image acquisitions (fig. 1). Four time points were measured during each Ca²⁺decrease, allowing an adequate analysis. The mean value of the Ca²⁺ratio at 480 ms, (corresponding to a Ca²⁺value close to resting Ca²⁺) was compared in a control situation and during propofol infusion (10⁻⁷, 10⁻⁶, 10⁻⁵, and 10⁻⁴M). The same analysis was performed for 10% Intralipid.

For biochemical experiments, rats were killed by heart excision after administration of the same anesthesia as used for cell isolation.

We studied the function of the SR Ca²⁺adenosine triphosphatase by measuring the rate of oxalate-stimulated Ca²⁺uptake. Crude homogenates were used because only a minor fraction of SR could be isolated by standard procedures and, moreover, marked loss of activity has been reported on isolated vesicles of SR. 16,17 

Ca²⁺uptake was measured, according to De la Bastie et al. , 18at 30°C in 0.5 ml of a medium containing 100 mM KCl, 5 mM adenosine triphosphate, 6 mM MgCl², 15 mM K-oxalate, 0.2 mM EGTA, 30 mM Tris–HCl buffer (pH = 6.85), and 0.15 mM ⁴⁵CaCl², giving a free Ca²⁺concentration of 1.25 μM as determined using the Fabiato program. 19Na-azide (5 mM) also was added to inhibit mitochondrial Ca²⁺uptake. The reaction was initiated by adding 20–50 μg total protein. After 4 min, the reaction was stopped by rapid dilution of the incubation mixture with a cold solution containing 100 mM KCl, 1 mM EGTA, and 10 mM histidine (pH = 6.4), followed by rapid filtration under vacuum through glass-fiber filters and radioactivity was counted in an LS 6000 SE Beckman scintillation counter (Beckman, Gagny, France).

For each animal, the value of Ca²⁺transport, expressed in nanomoles per minute per milligram of protein, was the mean value of two determinations.

Results are presented as the mean (SD). The significant dose-dependent effect of a drug (or time for Krebs infusion) in the same cell, used as its own control, was determined using one-way analysis of variance. If different groups of cells were compared (comparison of the effect of two drugs at different concentrations), two-way analysis of variance was performed. If the results of the analysis of variance were significant, multiple comparisons were performed using a post hoc  Scheffé test. For the 10⁻⁴M doses of propofol, the 405:480 ratios could not be obtained in all cells and were analyzed separately. A P  value < 0.05 was necessary to reject the null hypothesis.

The Ca²⁺transient measurements in the dose–response experiments with propofol (n = 8), 10% Intralipid (n = 8), and Krebs solution (n = 7) are summarized in table 1. There was no change in the diastolic 405:480 ratio during Krebs infusion, for the 10⁻⁷and 10⁻⁶M doses of propofol and for corresponding doses of 10% Intralipid. The 10⁻⁵M infusion of propofol increased the diastolic 405:480 ratios recorded at D^rest, D², D³, and D⁴. The 10⁻⁵M corresponding dose of 10% Intralipid also increased these diastolic ratios, but to a lower extent than propofol.

Neither Krebs solution nor 10% Intralipid at any concentration, nor propofol in doses less than 10⁻⁵M significantly changed the Δ ratios. In contrast, a 10⁻⁵M infusion of propofol induced a significant decrease in Ca²⁺transient as assessed by the Ca²⁺Δ ratios recorded at S², S³, and S⁴, but not at S¹(table 1, fig. 2).

In a group of seven cells analyzed separately, the 10⁻⁴-M dose of propofol significantly (P < 0.05) decreased, by 38%, the Ca²⁺transient (Δ ratio) recorded at S¹(from 0.88 ± 0.22 to 0.54 ± 0.15) and decreased, by 56%, the Ca²⁺transient recorded at S²(from 0.76 ± 0.15 to 0.34 ± 0.18). This effect was significantly different from that produced either by 10% Intralipid or by the Krebs solution. The changes in the Δ ratio in the S²beat are presented in figure 2.

The effects of propofol compared with those of its solvent, 10% Intralipid, are shown in figures 3 and 4.

Figure 3shows that propofol (upper) but not 10% Intralipid infusion (lower) induced a significant (P < 0.05) dose-dependent slowing of the slope of the decrease of the 405:480 ratio, measured after the systolic peak after a long rest.

The 10⁻⁴M dose of propofol and the corresponding dose of 10% Intralipid are not shown in figure 3because the peak value for this dose of propofol was lower than the control value (2.0 ± 0.18 vs.  2.15 ± 0.2 for the Krebs-solution value). The slope of the decrease was even slower than the 10⁻⁵M dose: The mean decrease in the 405:480 ratio 480 ms after the peak was 0.18, compared with 0.69 during control before propofol infusion.

As shown in figure 1in a typical example and in figure 4, cumulative concentrations (10⁻⁷, 10⁻⁶, 10⁻⁵M) of propofol increased the difference D²− D^rest. A 10⁻⁵M concentration of propofol increased the difference D²− D^rest(+0.13, ^P< 0.05) compared with the control value. Krebs solution or 10% Intralipid did not change the difference D²− D^restat any concentration.

The effects of propofol and 10% Intralipid on Ca²⁺transport are presented in figure 5.

Ca²⁺uptake was not modified by propofol between 10⁻⁷and 10⁻⁵M. A 10⁻⁴M concentration of propofol significantly decreased the rate of Ca²⁺uptake, from 11.8 ± 1.5 protein to 8.3 ± 2.6 nmol · min⁻¹· mg⁻¹protein (70 ± 5% of the control value). In contrast, 10% Intralipid did not induce changes in Ca²⁺transport at any concentration.

The principal results of this study may be summarized as follows:

• 1. Large concentrations of propofol decreased the slope of the decrease of the intracellular Ca²⁺after its systolic peak.

• 2. This effect produced a diastolic Ca²⁺elevation, after a short diastolic interval, compared with the Ca²⁺concentration at rest.

• 3. Propofol (10⁻⁴M) induced a decrease in Ca²⁺uptake capacity of the SR, which was probably responsible for these changes.

• 4. Propofol also decreased the Ca²⁺transient. This effect appeared for lower doses (10⁻⁵M) after short diastolic pauses rather than after a long (2- to 3-min) rest (appearing at 10⁻⁴M).

A propofol-induced impairment in Ca²⁺uptake capacity of the SR was suggested previously by an alteration of mechanical diastolic performance of left ventricular papillary muscles of rats. 4More recently, a prolonged time to 50% recovery for intracellular Ca²⁺and cell relengthening was observed on rat ventricular myocytes in response to supraclinical concentration of propofol, indicating a possible action of the drug on Ca²⁺handling by the SR. 8 

However, to our knowledge, the direct effect of propofol on SR Ca²⁺uptake never has been evaluated. Our results clearly show a rate-dependent alteration of the Ca²⁺transient and, in rat ventricular homogenates, a decrease of the rate of Ca²⁺uptake induced by propofol. There was a slight and nonsignificant tendency toward a decrease with a 10⁻⁵M dose, but the decrease induced by a 10⁻⁴M dose was significant (30%). This effect was not caused by the solvent because Intralipid did not induce any change in SR Ca²⁺uptake (fig. 5). Therefore, the current study shows that, as was suggested indi-rectly, 4,8SR function is indeed impaired by propofol.

The reduction in the slope of the systolic Ca²⁺decrease (fig. 3) may be caused by several different factors.

• 1. An increase in myofibrillar affinity for Ca²⁺could produce a slower myofibrillar Ca²⁺release during relaxation, as suggested by Kanaya et al.  8 

• 2. There may be a direct effect of propofol on the Na⁺–Ca²⁺exchanger because cytosolic Ca²⁺decreased during relaxation results from two different mechanisms: Ca²⁺extrusion out of the cell through the Na⁺–Ca²⁺exchanger, and Ca²⁺reuptake by the SR. These two mechanisms are balanced and may have variable relative importance according to the experimental conditions. 20Although a propofol-induced alteration of the Na⁺–Ca²⁺exchanger cannot be excluded in our study, such an effect has never been shown.

• 3. There may be a reduction in SR uptake capacity. This was found in our study and appears to be the major mechanism inducing the reduction in the slope of systolic Ca²⁺decrease during systole.

The slower SR Ca²⁺uptake is probably responsible for the increase in diastolic Ca²⁺values observed after a short diastolic pause (the difference D²− D^restin fig. 4). A slower Ca²⁺uptake does not have any influence on the level of diastolic Ca²⁺measured after a long pause (D^rest) but produces a larger D²value if the pause is short (1 s), not long enough for Ca²⁺to reach the basal value. The long duration of the diastolic pause (4 s) in the study of Cook and Housmans 7may explain why no SR abnormality was found. In that study, 7ryanodine injection was used to evaluate SR contribution to muscle contraction with and without propofol. The long rest (2 to 3 min) probably allowed the SR stores to be refilled with an unchanged releasable Ca²⁺, even in case of impaired Ca²⁺uptake.

The reduced Ca²⁺uptake induced by propofol probably also plays a role in the decrease in myocyte contractile function. The existence of a negative inotropic effect of the drug depends on species and pathophysiologic status. In isolated muscles, no negative inotropic effect 6,21or decreased contractility 3,7were found. In the current study, if the beat after a long pause was considered, a decrease in the Ca²⁺transient was observed only with the 10⁻⁴-M dose of propofol. After a long pause, SR loading is not altered if Ca²⁺uptake is slowed. It is thus likely that, in this case, the decrease in Ca²⁺transient can be attributed to a decreased trans-sarcolemmic Ca²⁺entry, which impairs myocyte contractility. A propofol-induced decreased contractility already has been attributed to a reduced trans-sarcolemmic Ca²⁺entry, 3,7especially by Ca²⁺L-type current inhibition. 10,11If the beat after a short pause (S², table 1) was considered, the decrease in Ca²⁺transient appeared for smaller doses of propofol (10⁻⁵M). It is likely that the increased diastolic Ca²⁺in this beat, associated with an incomplete filling of the Ca²⁺release stores, was responsible for the decreased Ca²⁺transient.

A possible limitation of the current study was that microfluorometric and biochemical studies were performed at different temperatures (22° and 30°C respectively). Obviously, the temperature at which experiments are performed may influence cellular Ca²⁺handling, and this may modify the results. We chose these procedures because pacing isolated cells at physiologic temperature may lead to an imbalance between energy supply and demand, and physiologic experiments in isolated cells usually are performed at room temperature. Ca²⁺uptake experiments have been performed at 30°C because this method was validated previously 18showing that Ca²⁺transport in the presence of oxalate is linear for several minutes at this temperature. Another potential limitation of the study is that cardiomyocytes contracted without load (preload and afterload). Changes in loading conditions affect myofilament affinity for Ca²⁺(high load) and relaxation (low load). 22Experiments with no load performed in this study allowed us to make more apparent relaxation abnormalities caused by SR.

In conclusion, this study showed that propofol decreases SR Ca²⁺uptake, inducing a slower decrease in systolic Ca²⁺after the peak, and, for relatively short diastolic pauses, this effect produces a diastolic increase in Ca²⁺at rest and a decrease in Ca²⁺transient. Clinical implications of this study must be drawn with caution because it is always difficult to extrapolate results obtained in vitro  in another species to humans. However, this study shows that propofol impairs SR function, with potential functional consequences. Although SR functional importance is lesser in humans than in rats, it represents at least 50% of the origin of systolic Ca²⁺23and is therefore not negligible for cardiac contraction. The doses of propofol that produced a negative inotropic effect in the study are larger than those used in clinical practice. Furthermore, propofol is highly bound (approximately 98%) to proteins in vivo , so that only a minor fraction of the drug is pharmacologically active if injected intravenously. As discussed by Hebbar et al. , 5concentrations in experimental solution may not reflect effective plasma concentration because propofol binding to proteins, lipid microsomes, and tissue is not taken into account in vitro . However, the free-drug concentration obtained during a bolus injection may locally reach 5.10⁻⁵M, 24close to the dose for which we show a depressant effect on Ca²⁺transient. In addition, if SR function is impaired, propofol clearly may be detrimental. A decreased SR function probably exists during heart failure, although this is still being debated. 25Indeed, propofol and Ca²⁺uptake could contribute to a more potent negative inotropic effect during heart failure compared with the normal state, particularly if blockade of Ca²⁺entry also contributes.