Propofol Modulates Na⁺–Ca²⁺Exchange Activity via  Activation of Protein Kinase C in Diabetic Cardiomyocytes

All experimental procedures and protocols were approved by Cleveland Clinic Institutional Animal Care and Use Committee, Cleveland, Ohio.

Adult, male, Sprague-Dawley rats (6 weeks old) were used for the study. Diabetes was induced by a single intraperitoneal injection of streptozotocin (55 mg/kg). Age-matched controls were injected with the vehicle only (0.1N sodium citrate, pH 4.5). The development of diabetes was assessed by biweekly measurements of urine glucose and ketone using Keto-Diastix (Baxter Scientific, McGaw Park, IL). Animals were maintained with free access to food and water for 12 weeks after streptozotocin administration. At the time of euthanasia, blood glucose levels were assessed using a glucometer (One Touch II, Lifescan, Milpitas, CA).

Freshly isolated adult ventricular myocytes from rat hearts were obtained as previously described.16,17Immediately after euthanasia, the hearts were rapidly removed and perfused in a retrograde manner at a constant flow rate (8 ml/min) with oxygenated (95% oxygen–5% carbon dioxide) Krebs-Henseleit buffer (37°C) containing the following: 118 mm NaCl, 4.8 mm KCl, 1.2 mm MgCl², 1.2 mm KH²PO⁴, 1.2 mm CaCl², 37.5 mm NaHCO³, and 16.5 mm dextrose, pH 7.35. After a 5-min equilibration period, the perfusion buffer was changed to a Ca²⁺-free Krebs-Henseleit buffer containing collagenase type II (309 U/ml). After digestion with collagenase (20 min), the ventricles were minced and shaken in Krebs-Henseleit buffer, and the resulting cellular digest was washed, filtered, and resuspended in HEPES-buffered saline (23°C) containing the following: 118 mm NaCl, 4.8 mm KCl, 1.2 mm MgCl², 1.25 mm CaCl², 11.0 mm dextrose, 25.0 mm HEPES, and 5.0 mm pyruvate, pH 7.35.

Simultaneous measurement of intracellular Ca²⁺concentration ([Ca²⁺]ⁱ) and cell shortening was performed, as previously described by our laboratory.18,19Ventricular myocytes exhibiting a rod-shaped appearance with clear striations were chosen for study. Myocytes (0.5 × 10⁶cells/ml) were incubated in HEPES-buffered saline containing 1 μm fura-2/acetoxy methylester at room temperature for 20 min. Fura-2–loaded ventricular myocytes were placed in a temperature regulated (37°C) chamber (Bioptechs, Inc., Butler, PA) mounted on the stage of an Olympus IX-70 inverted fluorescence microscope (Olympus America, Lake Success, NY). The cells were superfused continuously with HEPES-buffered saline at a flow rate of 2 ml/min and field-stimulated via  bipolar platinum electrodes at a frequency of 0.3 Hz with a 5 ms pulse using a Grass SD9 stimulator (Grass Instruments, West Warwick, RI).

Fluorescence measurements were performed on individual myocytes using a dual-wavelength spectrofluorometer (Deltascan RFK6002; Photon Technology International, Lawrenceville, NJ) at excitation wavelengths of 340 and 380 nm and an emission wavelength of 510 nm. Because calibration procedures rely on a number of assumptions, the ratio of the light intensities at the two wavelengths was used to measure qualitative changes in [Ca²⁺]ⁱ. Just before data acquisition, background fluorescence was measured and automatically subtracted from the subsequent experimental measurement. The fluorescence sampling frequency was 100 Hz, and data were collected using software from Photon Technology International.

To simultaneously monitor cell shortening, the cells were also illuminated with red light. A dichroic mirror (600-nm cutoff) in the emission path deflected the cell image through a charge-coupled device video camera (Phillips VC 62505T; Marshall Electronics, Culver City, CA) into a video-edge detector (Crescent Electronics, Sandy, UT) with 16-ms resolution. The video-edge detector was calibrated using a stage micrometer so that cell lengths during shortening and relengthening could be measured.

Before cessation of electrical stimulation, values representing peak [Ca²⁺]ⁱand shortening from at least five contractions were averaged to obtain the mean baseline peak value for each parameter. For analysis of data related to postrest potentiation (PRP) and SR Ca²⁺load, the change in shortening and/or [Ca²⁺]ⁱin response to resuming electrical stimulation or caffeine administration were plotted as the absolute percent change in [Ca²⁺]ⁱand/or shortening compared with the value obtained in normal cardiomyocytes, which was set at 100%. For analysis of data related to the time required for cytosolic Ca²⁺removal ([Ca²⁺]ⁱdecay), the time required to return to 50% of the baseline value (t^1/2) from the peak of the caffeine-induced Ca²⁺transient was calculated before and after treatment with the intervention.

A stock solution of propofol was obtained by dissolving the drug in dimethyl sulfoxide. Baseline measurements of [Ca²⁺]ⁱand shortening in response to electrical stimulation were collected from individual cardiomyocytes (normal and diabetic) for 20–30 s in the presence or absence of propofol (10 μm). Electrical stimulation was paused (15 s) and then resumed. The PRP achieved with a rest interval of 15 s represents approximately 50% of the maximum PRP achieved at a 60-s rest interval. Dimethyl sulfoxide (0.05% vol/vol) alone had no effect on [Ca²⁺]ⁱor shortening. Summarized data for the change in peak [Ca²⁺]ⁱand/or shortening in protocols 1–4 are expressed as absolute percent change.

Protocol 1 was repeated, except extracellular NaCl in the HEPES-buffered saline was replaced with 118 mm tetramethylammonium chloride (CH³)⁴NCl during the quiescent period, which changes the thermodynamic driving force on NCX to favor Ca²⁺influx. Alternatively, a selective inhibitor of reverse mode NCX (Ca²⁺in, Na⁺out), KB-R7943 (10 μm), was added to normal HEPES-buffered saline (containing NaCl) for 10 min before initiating electrical stimulation. KB-R7943 has no effect on forward mode NCX (Na⁺in, Ca²⁺out).20Experiments were performed in the presence or absence of propofol (10 μm).

Protocol 1 was repeated in cardiomyocytes, except caffeine (10 mm) was added after cessation of electrical stimulation (15 s) to stimulate SR Ca²⁺release.

Protocol 2 was repeated, except caffeine was added after cessation of electrical stimulation. Experiments were performed in the presence or absence of propofol (10 μm). In separate experiments, NaCl was again replaced with (CH³)⁴NCl, but now Ca²⁺was omitted and EGTA (1 mm) was added to HEPES-buffered saline (0 Na⁺, 0 Ca²⁺). The composition of this solution effectively blocks both forward and reverse mode NCX (Ca²⁺in, Na⁺out).20,21Alternatively, extracellular NaCl was elevated from 118 to 180 mm in the presence of extracellular Ca²⁺(1.2 mm) to favor forward mode NCX (Na⁺in, Ca²⁺out).

Protocol 1 was repeated in diabetic cardiomyocytes, except the PKC inhibitor bisindolylmaleimide I (10 μm) was added to normal HEPES-buffered saline for 5 min during the quiescent period. In separate experiments, caffeine (10 mm) was added after cessation of electrical stimulation (15 s) to stimulate SR Ca²⁺release. Experiments were performed in the presence or absence of propofol (10 μm). Summarized data are presented as absolute change compared with control values before PKC inhibition.

All experimental protocols were repeated in myocytes obtained from at least five different hearts. Results obtained from each heart were averaged so that all hearts were weighted equally. Within-group comparisons were made using one-way analysis of variance with repeated-measures and the Bonferroni post hoc  test. Comparisons between groups were made using two-way analysis of variance. Differences were considered statistically significant at P < 0.05. All results are expressed as mean ± SD.

Collagenase was obtained from Worthington Biochemical Corp. (Lakewood, NJ). Streptozotocin, propofol, bisindolylmaleimide I, and KB-R7943 were purchased from Sigma Chemical Co. (St. Louis, MO). Fura-2 AM was obtained from Texas Fluorescence Labs (Austin, TX).

Diabetic animals (n = 31; 12 weeks diabetic) had blood glucose levels of 447 ± 52 mg/dl (approximately 250 mm) and body weights of 304 ± 21 g at the time of euthanasia. Control animals injected with vehicle only (n = 18) had blood glucose levels of 101 ± 19 mg/dl (approximately 60 mm) and weighed 422 ± 26 g.

The overall goal of protocols 1 and 2 was to use PRP as a tool to investigate alterations in reverse mode NCX (Ca²⁺in, Na⁺out) in diabetic cardiomyocytes, alone and in combination, with propofol. Representative traces from an individual normal and diabetic cardiomyocyte demonstrating the effect of diabetes and propofol on PRP of peak [Ca²⁺]ⁱand shortening are shown in figures 1A and B. Summarized data are shown in figure 1C. After a 15-s rest duration, resuming electrical stimulation in normal cardiomyocytes resulted in PRP characterized by an increase in peak [Ca²⁺]ⁱand peak shortening that were 130 ± 4 and 236 ± 15%, respectively, of the value obtained before cessation of electrical stimulation. In diabetic cardiomyocytes, PRP values of peak [Ca²⁺]ⁱand peak shortening were 121 ± 6% (71 ± 5% compared with normal cardiomyocytes) and 133 ± 6% (58 ± 9% compared with normal cardiomyocytes), respectively, of the value obtained before cessation of electrical stimulation. In normal cardiomyocytes, propofol (10 μm) exerted no additional effect on peak [Ca²⁺]ⁱor shortening after cessation of electrical stimulation compared with untreated controls. In contrast, propofol increased peak [Ca²⁺]ⁱand peak shortening by 23 ± 7% (98 ± 4% of value obtained in normal cardiomyocytes) and 49 ± 6% (86 ± 6% of value obtained in normal cardiomyocytes), respectively.

Summarized data depicting the effect of Na⁺removal or inhibition of reverse mode NCX (Ca²⁺in, Na⁺out) with KB-R7943 on PRP of peak [Ca²⁺]ⁱand shortening in the presence or absence of propofol are shown in figure 2. Removal of Na⁺(tetramethylammonium substituted) during the quiescent period enhanced PRP of peak [Ca²⁺]ⁱand peak shortening in both normal and diabetic cardiomyocytes. In normal cardiomyocytes, PRP of peak [Ca²⁺]ⁱand shortening increased by 24 ± 4 and 29 ± 3%, respectively, whereas PRP of peak [Ca²⁺]ⁱand shortening increased by 21 ± 5 and 28 ± 5%, respectively, in diabetic cardiomyocytes. Exposure of normal cardiomyocytes to propofol had no additional effect on peak [Ca²⁺]ⁱor shortening in the absence of Na⁺but enhanced peak [Ca²⁺]ⁱand shortening in diabetic cardiomyocytes. Selective inhibition of reverse mode NCX (Ca²⁺in, Na⁺out) with KB-R7923 (10 μm) completely blocked PRP in both normal and diabetic cardiomyocytes. Propofol had no effect on PRP of peak [Ca²⁺]ⁱor shortening after inhibition of reverse mode NCX (Ca²⁺in, Na⁺out) in either normal or diabetic cardiomyocytes.

The overall goal of protocols 3 and 4 was to use SR Ca²⁺load as a tool to further investigate alterations in reverse mode NCX (Ca²⁺in, Na⁺out), as well as to use t^1/2for [Ca²⁺]ⁱdecay as a tool to explore alterations in forward mode NCX (Na⁺in, Ca²⁺out) in diabetic cardiomyocytes, alone and in combination with propofol. Representative traces depicting the effect of diabetes and propofol (10 μm) on SR Ca²⁺load and t^1/2for [Ca²⁺]ⁱdecay are shown in figures 3A and B. Summarized data are shown in figure 3C. Compared with normal cardiomyocytes, SR Ca²⁺load was reduced by 17 ± 4% in diabetic cardiomyocytes. Moreover, the t^1/2for [Ca²⁺]ⁱdecay was markedly prolonged in diabetic cardiomyocytes (3.5 ± 0.3 s) compared with normal cardiomyocytes (1.7 ± 0.2 s). Addition of propofol had no effect on SR Ca²⁺load in normal cardiomyocytes. However, propofol enhanced SR Ca²⁺load in diabetic cardiomyocytes by 14 ± 3% (97 ± 4% of value obtained in normal cardiomyocytes). In addition, propofol further prolonged the t^1/2for [Ca²⁺]ⁱdecay in diabetic cardiomyocytes (5.0 ± 0.3 ms) but had no effect on the t^1/2for [Ca²⁺]ⁱin normal cardiomyocytes (1.8 ± 0.2 s).

Summarized data depicting the effect of Na⁺removal or inhibition of reverse mode NCX (Ca²⁺in, Na⁺out) with KB-R7943 on SR Ca²⁺stores in normal and diabetic cardiomyocytes in the presence or absence of propofol are shown in figure 4. Similar to the data depicted in figure 2, Na⁺removal during the quiescent period caused an increase in SR Ca²⁺load in both normal and diabetic cardiomyocytes. Inhibition of reverse mode NCX (Ca²⁺in, Na⁺out) with KB-R7943 during the quiescent period reduced SR Ca²⁺load in both normal and diabetic cardiomyocytes. In the absence of extracellular Na⁺, propofol increased SR Ca²⁺load in diabetic cardiomyocytes by 26 ± 5%, but had no effect in normal cardiomyocytes when compared with the absence of extracellular Na⁺alone. KB-R7943 prevented the propofol-induced increase in SR Ca²⁺load. The [Ca²⁺]ⁱdecay was prolonged by Na⁺removal in both normal (t^1/2= 3.1 ± 0.4 s) and diabetic cardiomyocytes (t^1/2= 3.9 ± 0.4 s). In contrast, KB-R7943 had no effect on the [Ca²⁺]ⁱdecay in either normal (t^1/2= 1.8 ± 0.2 s) or diabetic (t^1/2= 3.2 ± 0.3 s) cardiomyocytes. In the absence of Na⁺, propofol further prolonged [Ca²⁺]ⁱdecay in diabetic cardiomyocytes (t^1/2= 5.9 ± 0.4 s), whereas the propofol-induced prolongation in [Ca²⁺]ⁱdecay was not affected by KB-R7943 (t^1/2= 5.1 ± 0.4 ms) compared with propofol alone.

Summarized data demonstrating the effect of removing both extracellular Na⁺and Ca²⁺(0 Na⁺, 0 Ca²⁺) or increasing extracellular Na⁺while Ca²⁺is held constant on [Ca²⁺]ⁱdecay in diabetic cardiomyocytes in the presence or absence of propofol are shown in figure 5. In the absence of both Na⁺and Ca²⁺, [Ca²⁺]ⁱdecay was markedly prolonged in both normal and diabetic cardiomyocytes, and propofol exerted no additional effect. In contrast, when extracellular Na⁺was elevated, the [Ca²⁺]ⁱdecay was shortened (faster) in both normal and diabetic control cells compared with 118 mm Na⁺controls. Similarly, the [Ca²⁺]ⁱdecay was significantly faster in propofol-treated cardiomyocytes.

The overall goal of protocol 5 was to investigate the role of PKC as a mediator of alterations in reverse mode NCX (Ca²⁺in, Na⁺out), SR Ca²⁺load, and forward mode NCX in diabetic cardiomyocytes, alone and in combination with propofol. Summarized data depicting the effect of PKC inhibition on PRP of peak [Ca²⁺]ⁱand peak shortening as well as SR Ca²⁺load and t^1/2[Ca²⁺]ⁱdecay in diabetic cardiomyocytes are shown in figures 6A and B, respectively. Addition of the PKC inhibitor bisindolylmaleimide I (10 μm) during the quiescent period (10 min) prevented the propofol-induced increase in PRP of peak [Ca²⁺]ⁱand shortening. Moreover, bisindolylmaleimide I also prevented the propofol-induced prolongation in SR Ca²⁺load and t^1/2[Ca²⁺]ⁱdecay. In the absence of propofol, bisindolylmaleimide I alone slightly restored PRP of peak [Ca²⁺]ⁱand peak shortening as well as SR Ca²⁺load and [Ca²⁺]ⁱdecay (data not shown).

This is the first study to assess the extent to which propofol alters the bimodal actions of the NCX on removal of cytosolic Ca²⁺(forward mode NCX, Na⁺in, Ca²⁺out) and SR Ca²⁺load (reverse mode NCX, Ca²⁺in, Na⁺out) in the setting of diabetes-induced cardiac dysfunction using freshly isolated ventricular myocytes. Previous studies suggest that altered Ca²⁺homeostasis is the primary abnormality contributing to cardiomyocyte dysfunction in diabetic hearts. Our key findings are that altered Ca²⁺homeostasis involving the bimodal actions of the NCX is a contributor to diabetic cardiomyocyte dysfunction. Moreover, propofol modulates the bimodal actions of the NCX in diabetic cardiomyocytes via  activation of PKC. These findings are summarized in figure 7.

The cellular mechanisms contributing to PRP are not entirely clear but likely result from Ca²⁺entry during rest via  reverse mode NCX causing an increase in the SR Ca²⁺load.21,22In the current study, PRP was attenuated in diabetic cardiomyocytes when compared with that observed in normal cardiomyocytes. Both diminished function and expression of the cardiac NCX have been observed in diabetic cardiomyocytes.23,24Therefore, it is likely that diminished function of NCX in the diabetic cardiomyocyte is responsible for the diminished PRP observed in the current study. Diabetic cardiomyocytes have been shown to have a higher baseline [Na⁺]ⁱthan normal cardiomyocytes,25and [Na]ⁱdrives the NCX in the reverse mode. This should facilitate an increase in SR Ca²⁺load and enhance PRP, but this was not observed in the current study. The lack of an increase in SR Ca²⁺load may be due to reports of diminished Na⁺–K⁺ATPase activity in diabetic cardiomyocytes.26Diminished activity would result in a more positive membrane potential such that the reversal potential of NCX would not favor Ca²⁺influx via  reverse mode (Na⁺out, Ca²⁺in). It is questionable whether diminished Na⁺–K⁺ATPase activity itself, independent of changes in NCX activity, could contribute to the alterations in Ca²⁺handling observed in this model. The reasons for this are that an increase in [Na]ⁱcan inhibit Na⁺–H⁺exchange via  a Na⁺-activated G protein resulting in an accumulation of protons (H⁺) in the cytoplasm, but acidosis increases both diastolic and peak [Ca²⁺]ⁱ,27which should increase SR Ca²⁺load. In contrast, increased [Na]ⁱcan also activate K⁺channels resulting in a shortening of action potential duration that would limit Ca²⁺influx and SR Ca²⁺load consistent with our findings; however, action potential prolongation is typically observed in diabetic cardiac myocytes28primarily via  decreases in transient outward and steady state K⁺current density.29If Na⁺levels are truly higher in our diabetic cardiomyocytes, our current findings could still be explained by reports of decreased expression and function of the cardiac ryanodine receptor30and/or decreased SR Ca²⁺uptake by the SR Ca²⁺ATPase.2,31,32Either of these could contribute to the decreased [Ca²⁺]ⁱand shortening observed in diabetic cardiomyocytes. We also cannot discount the potential of reports indicating increased expression of nonphosphorylated forms of phospholamban in diabetic cardiomyocytes, which would attenuate SR Ca²⁺loading and diminish PRP of peak [Ca²⁺]ⁱand shortening.2,31,33 

Propofol had no effect on PRP of peak [Ca²⁺]ⁱand shortening in normal cardiomyocytes but enhanced these values in diabetic cardiomyocytes. One potential explanation for our current findings may be related to our previous findings that propofol activates the Na⁺–H⁺exchanger in normal cardiomyocytes.17It is possible that a propofol-induced increase in [Na⁺]ⁱ(via  Na⁺–H⁺exchange) could serve to further facilitate and/or result in favoring reverse mode NCX in diabetic cardiomyocytes by increasing [Na⁺]ⁱ. Assuming a resting [Na]ⁱlevel of 12–14 mm and a [Ca²⁺]ⁱlevel of 100 nm in normal rat cardiomyocytes, the predicted NCX reversal potential would be −67 to −79 mV.20If [Na⁺]ⁱincreased by just 3 mm, and it is already elevated in diabetic cardiomyocytes,25perhaps by a decrease in Na⁺–K⁺ATPase activity,26the reversal potential for NCX would become −85 to −95 mV, negative to the resting membrane potential, and strongly favor net Ca²⁺influx at rest.20The lack of effect of propofol on PRP in normal cardiomyocytes, despite a propofol-induced activation of Na⁺–H⁺exchange,34suggests that other factors in diabetic cardiomyocytes, such as decreased Na⁺–K⁺ATPase activity,26abnormal K⁺channel activity,29prolonged action potential duration,28and/or decreased membrane potential,35may contribute to the propofol-induced alterations in Ca²⁺handling observed in diabetic cardiomyocytes. Although we previously demonstrated that propofol-induced activation of Na⁺–H⁺exchange causes an increase in myofilament Ca²⁺sensitivity in normal cardiomyocytes,17we recently demonstrated that propofol decreases peak [Ca²⁺]ⁱ, prolongs t^1/2[Ca²⁺]ⁱdecay, and decreases myofilament Ca²⁺sensitivity in diabetic cardiomyocytes.36Moreover, it should be noted that the increase in peak [Ca²⁺]ⁱrelease by propofol in diabetic cardiomyocytes is much larger than the percent increase in fractional shortening, again implying a propofol-induced decrease in myofilament Ca²⁺sensitivity, despite enhanced fractional Ca²⁺release.

The bimodal action of the NCX is primarily regulated by the concentration gradients of Na⁺and Ca²⁺across the sarcolemma. Therefore, removal of Na⁺from the perfusion buffer facilitates reverse mode NCX (Ca²⁺in, Na⁺out). KB-R7943 is a novel agent that preferentially blocks reverse mode NCX and has no effect on forward mode NCX.20,37In the current study, removal of Na⁺completely restored, and KB-R7943 completely blocked, PRP of peak [Ca²⁺]ⁱand shortening in both normal and diabetic cardiomyocytes. This indicates that the reduced PRP in diabetic cardiomyocytes is entirely due to a defective NCX. In addition, these data support findings that PRP in both cell types depends on activation of reverse mode NCX (Ca²⁺in, Na⁺out).21,22,38 

In diabetic cardiomyocytes, removal of Na⁺further enhanced the propofol-induced increase in PRP, whereas blockade of reverse mode NCX with KB-R7943 blocked propofol's effect on PRP. These data further indicate that the actions of propofol on PRP are mediated by a propofol-induced activation of the reverse mode NCX (Ca²⁺in, Na⁺out). A recent report demonstrated that propofol enhances PRP in rat cardiac trabeculae by increasing Ca²⁺influx via  the reverse mode NCX (Ca²⁺in, Na⁺out).13 

Caffeine-induced SR Ca²⁺release was used to assess the effect of diabetes and propofol on SR Ca²⁺load driven by reverse mode NCX (Ca²⁺in, Na⁺out) and t^1/2for [Ca²⁺]ⁱdecay driven by forward mode NCX (Na⁺in, Ca²⁺out) in cardiomyocytes. Although it is well documented that the SR Ca²⁺ATPase contributes approximately 90% to the decline of twitch [Ca²⁺]ⁱin rat cardiomyocytes, the SR only releases approximately 50% of the stored Ca²⁺during a twitch.39In the current study, peak twitch [Ca²⁺]ⁱand SR Ca²⁺release in response to caffeine were reduced in diabetic cardiomyocytes compared with normal cardiomyocytes. These data imply a reduced SR Ca²⁺load in diabetic cardiomyocytes as previously suggested, although not directly tested, in one recent study.30Moreover, the t^1/2[Ca²⁺]ⁱdecay was significantly prolonged in diabetic cardiomyocytes compared with normal cardiomyocytes, implying a defect in removal of Ca²⁺by forward mode NCX (Na⁺in, Ca²⁺out), and not SERCA2, because caffeine-induced activation of the SR Ca²⁺release channels is so strong in promoting SR Ca²⁺release that it prevents Ca²⁺accumulation by the SR.39–41Therefore, removal of Ca²⁺is dependent on other transport systems (e.g. , NCX, mitochondria, sarcolemmal Ca²⁺ATPase). These various transport systems and their relative roles in Ca²⁺removal may be altered in diabetic cardiomyocytes and could contribute to the “fast” and “slow” decay in Ca²⁺removal observed in diabetic cardiomyocytes after caffeine exposure, which is not observed in normal cardiomyocytes. Our data are consistent with studies demonstrating alterations in the expression and/or function of these key Ca²⁺regulatory proteins in diabetic hearts,2,31as well as functional studies demonstrating prolonged relaxation in cardiomyocytes and papillary muscles.4,6,31,36,42,43 

In diabetic cardiomyocytes, propofol increased peak [Ca²⁺]ⁱachieved in response to caffeine after a rest that resembled levels observed in normal cardiomyocytes. We believe this is due to a propofol-induced increase in reverse mode NCX (Ca²⁺in, Na⁺out) leading to an increase in the SR Ca²⁺load. More striking is the marked prolongation in the t^1/2[Ca²⁺]ⁱdecay by propofol. These data further suggest a defect in the forward mode NCX (Na⁺in, Ca²⁺out) rather than a defective SERCA2, as mentioned above. Our data are consistent with other studies demonstrating that propofol prolongs Ca²⁺transients and impairs myocardial relaxation,8,9,18,44suggesting that propofol exerts inhibitory effects on SERCA2 and/or forward mode NCX (Na⁺in, Ca²⁺out). There are numerous conflicting reports regarding the effects of propofol on SR Ca²⁺uptake by SERCA2,9,10,12,18,44and only one report of propofol exerting effects on the cardiac NCX, in particular, activation of reverse mode NCX in trabecular muscle.13There are no reports of propofol exerting effects on forward mode NCX (Na⁺in, Ca²⁺out). Therefore, we further examined the role of NCX as the target for propofol and the mediator of prolonged t^1/2[Ca²⁺]ⁱdecay in diabetic cardiomyocytes.

Under normal conditions, NCX is responsible for approximately 10% of the Ca²⁺removal from the cytosol, and SERCA2 is responsible for the remainder.45We suspected it was unlikely that prolongation in t^1/2[Ca²⁺]ⁱdecay involved SERCA2, because this would be inconsistent with our findings of enhanced PRP by propofol, which would require an active SERCA2 to increase Ca²⁺load of the SR. Our findings that Na⁺removal enhanced, and KB-R7943 reduced, fractional Ca²⁺release in response to caffeine were predictable based on data depicted in figure 2that have previously been discussed. Interestingly, we found that under conditions where both forward and reverse mode NCX (Ca²⁺in, Na⁺out) were blocked (0 Na⁺and 0 Ca²⁺), propofol no longer prolonged the t^1/2[Ca²⁺]ⁱdecay and restored values to control levels. These data provided strong evidence that propofol was inhibiting forward mode NCX (Na⁺in, Ca²⁺out) and not exerting effects on SERCA2 in diabetic cardiomyocytes. These novel data may help to explain the controversy regarding propofol's effects on cytosolic Ca²⁺removal, SR Ca²⁺handling, and impairment of myocardial relaxation in cardiac muscle. We next explored a possible cellular mechanism to explain the propofol-induced effects on NCX.

We recently demonstrated that several PKC isoforms are up-regulated in diabetic cardiomyocytes36and that propofol causes PKC-dependent phosphorylation of contractile proteins.16Our current data indicating that inhibition of PKC prevents the propofol induced increase in PRP of peak [Ca²⁺]ⁱand shortening as well as increased SR Ca²⁺load and t^1/2[Ca²⁺]ⁱdecay suggest that a PKC-dependent mechanism is involved. Not only could propofol-induced, PKC-dependent activation of Na⁺–H⁺exchange be involved in facilitating reverse mode NCX (Ca²⁺in, Na⁺out), as discussed above, but it may also be responsible for inhibiting forward mode NCX (Na⁺in, Ca²⁺out). Alternatively, PKC-dependent phosphorylation of the NCX may also work in parallel with changes in Na⁺and H⁺to regulate/modulate the NCX in diabetic cardiomyocytes. This would be consistent with a report indicating that reduced NCX activity resulting from diabetes may be related to changes in PKC activity but is not related to altered expression of the transporter.24 

As with all in vitro  studies, there are always some limitations to the model system and its extrapolation to the clinical situation. It should be noted that this model system represents type 1 diabetes, and that the lack of insulin may induce changes in cardiomyocytes different from those seen in insulin-resistance and a high or normal level of insulin as occurs in type 2 diabetes. However, a recent report indicates that cardiomyocytes bathed in high glucose exhibit all the same changes in excitation–contraction coupling that are observed in vivo  as well as in chemically induced models of type 1 diabetes. Moreover, patients with type 2 diabetes also exhibit a prolongation in action potential duration, and intracellular Ca²⁺clearing and mechanical relaxation are slowed. It is also well established that species differences can contribute to difficulties in extrapolating in vitro  data to the in vivo  situation, and in this model system, rat cardiomyocytes are know to have an abnormally high [Na⁺]ⁱ, which means that the contribution of NCX in clearing intracellular Ca²⁺is far greater than that observed in other species. In addition, this in vitro  study only deals with intrinsic myocardial function, whereas changes in cardiac function after propofol administration in vivo  also depend on a variety of other factors, including venous return, afterload, and neurohumoral compensatory mechanisms. However, we believe that the strengths of this model system far outweigh the limitations.

It is well established that the response to propofol is widely variable among patients given the same dose. Its binding to serum proteins exceeds 98%,46so small changes in protein concentrations can be amplified in the unbound fraction of the drug and in its effect. It is likely that that part of the variability in the response among patients is because of differences in protein levels among individuals, and particularly among those with pathologies such as liver disease. The clinical relevance of in vitro  studies using propofol is often questioned because the concentrations of propofol that cause changes in cell, tissue, or organ function are typically outside of our estimations of what we perceive as a clinically relevant concentration. However, estimations of the clinically relevant plasma concentrations of propofol in vivo  as well as relating these clinical plasma concentrations to free aqueous drug concentrations in vitro  are difficult for several reasons. First, the rate of exchange between propofol-containing liposomes, the aqueous phase, serum proteins, and cellular constituents is not precisely known, which could significantly affect plasma concentrations in vivo . In addition, protein binding in vivo  is unlikely to be instantaneous, so free drug concentrations with a bolus injection would probably be higher than the steady state value. Peak plasma levels after a bolus injection have been estimated at 50 μm, and stable levels of approximately 10–25 μm during maintenance infusion.47Additional factors such as speed of injection, volume of distribution, and pH are all factors that can affect plasma concentrations of anesthetics. Given the difficulty and uncertainty in estimating the in vivo  concentrations and the likelihood that these estimations may be different in pathologic conditions (hemodilution, liver disease, diabetes), we believe that the concentration of propofol used in this study is likely to be similar to that encountered in clinical practice.

A hallmark of cardiac dysfunction in diabetic patients is the depressed contractility accompanied by prolonged relaxation. Although our study identifies that propofol enhances PRP in diabetic cardiomyocytes, the clinical significance of this finding is uncertain because propofol also causes a decrease in myofilament Ca²⁺sensitivity resulting in an overall negative inotropic effect. However, a more important clinically related finding is that propofol exacerbates the already prolonged relengthening of the cardiomyocyte causing a negative lusitropic effect and suggests that propofol may increase the risk for additional diastolic dysfunction in diabetic patients. Moreover, this negative lusitropic effect can influence cardiac inotropy because diastolic function significantly influences systolic cardiac function, left ventricular filling, and coronary blood flow. Although difficulties arise when attempting to extrapolate results from in vitro  studies to the in vivo  situation, we believe that the results from this study support observations in the clinical setting. In light of our findings, we propose that caution should be used when administering propofol in patients with limited inotropic reserve and/or diastolic dysfunction. Further studies are required to assess the role of PKC-dependent modulation of NCX by propofol in the diabetic heart.