Diazoxide Maintenance of Myocyte Volume and Contractility During Stress: Evidence for a non-Sarcolemmal K_ATP Channel Location

Abstract

Objectives

Animal and human myocytes demonstrate significant swelling and reduced contractility during exposure to stress (metabolic inhibition, hyposmotic stress, or hyperkalemic cardioplegia), and these detrimental consequences may be inhibited by the addition of diazoxide (adenosine triphosphate–sensitive potassium (K_ATP) channel opener) via an unknown mechanism. Both SUR1 and SUR2A subunits have been localized to the heart and mouse sarcolemmal K_ATP channels are composed of SUR2A/Kir6.2 subunits in the ventricle and SUR1/Kir6.2 subunits in the atria. This study was performed to localize the mechanism of diazoxide by directly probing sarcolemmal K_ATP channel current and by genetic deletion of channel subunits.

Methods

Sarcolemmal K_ATP channel current was recorded in isolated wild type ventricular mouse myocytes during exposure to Tyrode’s solution, Tyrode’s + 100µM diazoxide, hyperkalemic cardioplegia, cardioplegia + diazoxide, cardioplegia + 100µM pinacidil, or metabolic inhibition using whole cell voltage clamp. Ventricular myocyte volume was measured from SUR1(−/−) and wild type mice during exposure to control solution, hyperkalemic cardioplegia, or cardioplegia + 100µM diazoxide.

Results

Diazoxide did not increase sarcolemmal K_ATP current in wild type myocytes, although they demonstrated significant swelling during exposure to cardioplegia that was prevented by diazoxide. SUR1(−/−) myocytes also demonstrated significant swelling during exposure to cardioplegia but this was not altered by diazoxide.

Conclusions

Diazoxide does not open the ventricular sarcolemmal K_ATP channel, but provides volume homeostasis via an SUR1 – dependent pathway in mouse ventricular myocytes, supporting a mechanism of action distinct from sarcolemmal K_ATP channel activation.

Introduction

The adenosine triphosphate–sensitive potassium (K_ATP) channel opener diazoxide is cardioprotective and mimics ischemic preconditioning in animal models (1). Previously we have documented significant swelling of isolated myocytes and associated reduced contractility secondary to exposure to standard hypothermic hyperkalemic cardioplegia solution in human and other species (2,3). These detrimental consequences were ameliorated by the addition of diazoxide (2,3). As diazoxide is a K_ATP channel opener, we hypothesized that K_ATP channels may play a role in myocyte volume homeostasis and that myocyte swelling may be one mechanism of myocardial stunning (4). Myocyte swelling and reduced contractility are also observed following exposure to hyposmotic and ischemic stress (4,5). Interestingly, diazoxide ameliorates both the structural and functional derangements secondary to all three stresses: hyperkalemic cardioplegia, mild hyposmotic stress, and ischemic stress (2–5). The addition of 5-hydroxydecanoate, a claimed mitochondrial K_ATP (mK_ATP) channel blocker, or HMR 1098, a claimed sarcolemmal K_ATP (sK_ATP) channel blocker, to ischemic stress and hyperkalemic cardioplegia did not alter the beneficial observations noted with diazoxide alone (2,4). The cardioprotective mechanism of diazoxide remains unknown.

Diazoxide has been described as a mitochondrial K_ATP channel opener, with “selectivity” towards the mK_ATP channel and only weak sK_ATP channel activation at high doses (6). However, controversy regarding the specificity o f K_ATP channel openers and blockers suggests that pharmacologic manipulation of the K_ATP channel, using potassium channel openers or sulfonylurea receptor blockers, may be inadequate to definitively confirm ion flux across the sK_ATP channel (7–9), and it remains unclear whether diazoxide provides cardioprotection via the sK_ATP channel, the mK_ATP channel, or via a K_ATP channel-independent mechanism.

K_ATP channels are composed of a Kir inward rectifier channel forming subunit and a sulfonylurea sensitive regulatory subunit, SUR (Figure 1) (10). It is clear that sarcolemmal K_ATP channels require both Kir6.2 and SUR2A subunits in the ventricle, but Kir6.2 and SUR1 in the atria (11). Both SUR1 and SUR2A are expressed in mouse heart (11,12) and there is evidence that SUR2 and SUR1 subtypes are both present in cultured neonatal rat ventricular myocytes (13). Mice lacking the SUR1 subunit appear to tolerate ischemia/reperfusion injury better than wild type mice in a model of left anterior descending coronary artery ligation (12). While there is some evidence for SUR1 subunits in the ventricle of various species, K_ATP currents in mouse ventricle are unaltered in mice lacking the SUR1 subunit (11).

Figure 1.

Composition of the sarcolemmal K_ATP channel and location of various subunits in various tissues. The channel is a functional octamer of four Kir6.x subunits (generating the channel pore) and four sulfonylurea receptor (SUR) subunits that generate the regulatory subunit. TMD is transmembrane domain, NBF is nucleotide – binding fold, and M1 and M2 are transmembrane helices.

We initially hypothesized that the mechanism of action of diazoxide might involve opening the sarcolemmal K_ATP channel (SUR2A/Kir6.2). Previous work documented a beneficial effect of K_ATP channel opener diazoxide in ventricular myocytes that was unaltered by the pharmacologic inhibition of the K_ATP channel (2,4). The first part of this study was designed to definitively investigate the action of diazoxide by the direct measurement of sarcolemmal K_ATP channel current in wild type mice using whole cell voltage clamp.

After demonstrating that sK_ATP (SUR2A/Kir6.2) channel activity was not observed in wild type mice during exposure to diazoxide, and knowing that SUR1 subunits have been documented in ventricular tissue (11,12,13), we next hypothesized that diazoxide might act via SUR1 subunits of an alternative K_ATP channel or some other channel in ventricular cells. The second part of this study was therefore designed to determine if ventricular myocytes lacking the SUR1 subunit would be responsive to diazoxide during stress.

This study was designed to elucidate the location of action of diazoxide by the direct measurement sK_ATP channel activity (in wild type mice) and by response to stress in ventricular myocytes from (wild type mice and mice lacking the SUR1 subunit. The elucidation of diazoxide’s mechanism of action in this model will facilitate its future clinical use.

Materials and Methods

Myocyte Isolation

All animal procedures were approved by the Animal Studies Committee at Washington University School of Medicine and all animals received humane care in compliance with the “Guide to Care and Use of Laboratory Animals”.

Ventricular myocytes were utilized for all experiments and were isolated from adult mice (either wild type (WT) or SUR1 knockout (KO), either sex, 6 weeks to 5 months, 25 to 30 g body weight) as previously described (14). Rapid cardiectomy was performed in the anesthetized (2.5% Avertin) mouse and the aorta was cannulated using a 28-gauge needle. The heart was attached to a Langendorff apparatus, and solution A was perfused through the aorta for 5 minutes. The heart was then perfused at 37°C for 12 minutes with solution B. The left ventricle was removed and transferred into solution C, where it was gently dispersed by glass pipette at room temperature. The cells were allowed to centrifuge by gravity, and serial washings were performed every 10 minutes for a 30-minute period. Cells were used in experiments within 5 hours after isolation. A typical yield of viable myocytes was 65% to 75% per mouse.

Solution A consisted of (in mmol/L, except as noted) 116 NaCl; 5.36 KCl; 0.97 Na₂HPO₄; 1.47 KH₂PO₄; 21.10 HEPES (N-[2-hydroxyethyl]piperazine-N′-[4-butanesulfonic acid]); 11.65 glucose; 26.50 µmol/L phenol red (Sigma Chemical Co; St. Louis, MO); 3.72 MgCl₂; 4.40 NaHCO₃; essential vitamins (100×, 10 mL, GIBCO, Grand Island, NY); and amino acids (50×, 20 mL, GIBCO, Grand Island, NY). Solution B consisted of solution A plus 10 µmol/L CaCl₂; 1.2 mg/mL collagenase (Type 2, Worthington Biochemical Corporation; Freehold, NJ). Solution C consisted of solution A plus 5 mg/mL bovine serum albumin (Sigma); 1.25 mg/mL taurine; and 150 µmol/L CaCl₂.

The diazoxide (7-chloro-3-methyl-1,2,4-benzothiadiazine-1,1-dioxide [DZX]; Sigma, St. Louis, MO) dose of 100 µmol/L was utilized as it was effective in ameliorating cell swelling secondary to stress (hyperkalemic cardioplegia, hyposmotic stress, and metabolic inhibition) in pervious studies (2–5). A stock solution of DZX was made by dissolving the reagent in 0.1% dimethyl sulfoxide (DMSO), at which concentration DMSO has no effect on cell volume (15).

Cells were selected for viability using the following criteria: normal rod shape, smooth edges, sharp borders and clear striations, absence of vacuoles or blebbing, and lack of spontaneous beating (15). A maximum of two cells were utilized per each animal.

Electrophysiology in Wild Type Myocytes

Following isolation, wild type myocytes were placed in a recording chamber containing normal Tyrode’s solution. Macroscopic currents in isolated ventricular myocytes were recorded using standard whole cell voltage-clamp recording techniques (16). Patch-clamp electrodes (1–3MΩ when filled with electrode solution) were fabricated from soda lime glass microhematocrit tubes (Kimble 73813, Kimble Glass Co., Vineland, NJ). Electrode solution contained the following (in mmol/L): 140 KCl, 10 HEPES, and 10 EGTA; pH 7.3–7.4. Cell capacitance and series resistance were determined using a 5- to 10-mV hyperpolarizing square pulse from a holding potential of −70 mV after establishment of the whole cell recording configuration. PClamp 9.2 software and DigiData 1322 were used to generate command pulses and collect data. Data was filtered at 5 kHz. A 4 second ramp from −110 and 40 mV was used to isolate and detect sK_ATP channel current only (17).

Experimental protocol

Isolated wild type myocytes were exposed to normal Tyrode’s solution (NT) for baseline measurement for 1–2 minutes followed by exposure to test solution (5–10 min) followed by NT for 5–10 min. Test solutions included NT (n=8 cells), NT+100µM/L diazoxide (DZX) (n=7 cells), hyperkalemic cardioplegia in the form of St. Thomas’ solution (CPG, Plegisol, Abbott Laboratories, North Chicago, IL) (n=11 cells), CPG+100µM/L DZX (n=8 cells), CPG+100µM/L pinacidil (Sigma, St. Louis, MO) (n=12 cells), and metabolic inhibition (MI) (n=8 cells). NT consisted of the following (in mmol/L): 137 NaCl, 5.4 KCl, 25 NaH₂PO₄, 10 glucose, 0.5 MgCl₂, 5 HEPES, and 3 NaHCO₃; pH 7.3–7.4. St. Thomas’ solution consisted of: (in mmol/L): 110 NaCl, 10 NaHCO₃, 16 KCl, 32 MgCl₂, and 2.4 CaCl₂ and was equilibrated with 95% O₂ – 5% CO₂ and titrated to correct to pH 7.3. MI solution consisted of (in mmol/L): 137 NaCl, 5.4 KCl, 3 NaHCO₃, 0.16 NaH₂PO₄, 10 2-deoxyglucose, 0.5 MgCl₂, 5 HEPES, 5% oligomycin, and 3 NaHCO₃; pH 7.3–7.4. Isolated sarcolemmal K_ATP channel activity and cellular capacitance were recorded during exposure to NT and compared to channel activity during test solution exposure.

Metabolic inhibition was utilized as a positive control to document an increase in sarcolemmal K_ATP channel current. Myocytes which demonstrated an increase in potassium current were subsequently exposed to glibenclamide (10 µM/L, a known sK_ATP channel blocker, Sigma, St. Louis, MO) to confirm that the increase in potassium current during exposure to metabolic inhibition was through sK_ATP channels.

Myocyte Volume Imaging in Wild Type and SUR1 (−/−) Myocytes

An aliquot of isolated myocytes was placed in a glass-bottomed chamber on an inverted microscope stage (Leitz, Wetzlar, Germany) equipped with Hoffman modulation optics (Modulation Optics, Greenvale, NY). After a 5-minute stabilization period, the chamber was perfused at a rate of 3 mL/min with 37°C normal Tyrode’s solution (in mmol/L) 130 NaCl; 5 KCl; 2.5 CaCl₂; 1.2 MgSO₄; 24 NaHCO₃; 1.75 Na₂HPO₄; and 10 glucose; buffered to a pH of 7.4 using 95% O₂–5% CO₂. Chamber temperature was controlled by a waterbath system (Thermo Haake, Karlsruhe, Germany). Cell images were displayed on a video monitor using a charge-coupled device camera (KPM1U; Hitachi Denshi, Tokyo, Japan). Digital images of cells were captured at a rate of 120 frames per second using a video-frame grabber (Scion Corporation, Frederick, MD) and were manually traced using Scion Image software (Scion Corporation). Length, width, and area were measured and recorded. To calculate cell volume, it was assumed that changes in cell width and thickness were proportional, and relative cell volume change was determined by the following formula (15):

$${\text{volume}}_t/{\text{volume}}_c=({\text{area}}_t\times {\text{width}}_t)/({\text{area}}_c\times {\text{width}}_c)$$

where t and c refer to test and control, respectively. On the basis of repeated measurements of single images and measurements of multiple images of a cell, this methodology for estimating cell volume has been shown to be reproducible with an error of less than 1% (15).

SUR1(−/−) mice were created by removal of the 1-kbp gene segment containing both promoter and exon 1 sequences of SUR1 gene by cre-mediated recombination (18). Genotype was confirmed by PCR analysis (18). This deletion is not lethal, and the SUR1 (−/−) mice lack pancreatic β cell K_ATP channels (SUR1/Kir6.2) and lack glucose-stimulated insulin secretion and are mildly glucose intolerant (18).

Experimental Protocol

Myocytes were not subjected to ischemia or modified ischemia in this protocol to delineate changes caused by the stress of exposure to cardioplegia alone. Myocytes from WT and KO mice were perfused with 37°C normal Tyrode’s solution for 5 min to obtain baseline measurements. Any baseline changes in cell volume secondary to the isolation or imaging protocol would be evident during this period. Myocytes were then perfused for 5 min with test solution, followed by a 5-min reexposure period to 37°C normal Tyrode’s solution. Test solutions included NT (n=8 cells for KO; n=7 cells for WT), hyperkalemic cardioplegia (CPG) in the form of St. Thomas’ solution (n=10 cells for KO; n=7 cells for WT) or CPG+100µM/L DZX (n=8 cells for KO; n=7 cells for WT). Volume measurements were made after the end of each 5 minute period.

Statistical Analysis

Data were analyzed using SYSTAT 11 (SYSTAT software Inc., Point Richmond, CA). All data are presented as mean value ± standard error of the mean, with N equal to the number of cells in each group. For cell volume measurements a repeated-measures analysis of variance was used for sequential time-based measurements for each test solution against its own baseline value. Using Fisher’s least significant difference test, post hoc multiple comparisons between different test groups were made separately during the test solution and reexposure periods. For electrophysiology experiments, Student’s t test was used to compare data. Probability values less than 0.05 were considered significant.

Results

Sarcolemmal K_ATP Channel Current in Wild Type Myocytes

Wild type ventricular myocytes exposed to normal Tyrode’s solution exhibited no increase in sarcolemmal adenosine triphosphate–sensitive potassium current throughout the experiment (Figure 2A), and the addition of diazoxide to normal Tyrode’s solution was not associated with an increase in potassium current (Figure 2A). Sarcolemmal K_ATP current increased significantly in wild type cells exposed to metabolic inhibition (p=0.002 vs. Tyrode’s) (Figure 2B), and was inhibited by the addition of glibenclamide (Figure 2A).

Figure 2.

A. Diazoxide does not elicit sarcolemmal adenosine triphosphate–sensitive potassium current in wild type myocytes. The y-axis represents potassium current (nA) and the x-axis represents time. Wild type myocytes in were initially exposed to normal Tyrode’s solution for baseline potassium current measurement, followed by test solution, followed by reexposure to normal Tyrode’s solution. The first graph (far left) is a representative cell exposed to normal Tyrode’s solution throughout the experiment. The second graph (middle) is a representative cell exposed to normal Tyrode’s solution with diazoxide during the test period. The addition of diazoxide to normal Tyrode’s solution did not alter the potassium current. The third graph (far right) is a representative cell exposed to normal Tyrode’s solution followed by metabolic inhibition. Metabolic inhibition was associated with an increase in potassium current which is reversed by the addition of glibenclamide. DZX is diazoxide, GLIB is glibenclamide, MI is metabolic inhibition, and NT is normal Tyrode’s solution.

B. Diazoxide does not elicit sarcolemmal adenosine triphosphate-sensitive potassium current in wild type myocytes. Cell conductance is represented in nanoseimens/picofarads (nS/pF, y axis) and the test solution on the x axis. Test solutions included: normal Tyrode’s, normal Tyrode’s solution in addition to diazoxide, and metabolic inhibition. When correcting for cell conductance, there was a significant increase in potassium current in cells exposed to metabolic inhibition (*p= 0.002 vs. NT). DZX is diazoxide, MI is metabolic inhibition, NT is normal Tyrode’s solution, and N is number of myocytes.

Wild type myocytes exposed to cardioplegia alone after exposure to normal Tyrode’s solution exhibited no increase in sK_ATP current (Figure 3A), and the addition of diazoxide to cardioplegia had no effect on the potassium current when compared to cardioplegia alone (Figure 3A). The addition of pinacidil in cardioplegic solution was associated with an increase in sK_ATP current, which returned to baseline when cells were re-exposed to normal Tyrode’s solution (Figures 3A and 3 B). This increase in potassium current was comparable to the potassium current induced by metabolic inhibition.

Figure 3.

A. Pinacidil elicits an increase in adenosine triphosphate–sensitive potassium current in wild type cells exposed to cardioplegia while diazoxide does not. The y-axis represents potassium current (nA) and the x-axis represents time. Wild type myocytes were initially exposed to normal Tyrode’s solution for baseline potassium current measurement, followed by test solution, followed by reexposure to normal Tyrode’s solution. The first graph (far left) is a representative cell exposed to hyperkalemic cardioplegia (St. Thomas solution) alone during the test period. The second graph (middle) is a representative cell exposed to cardioplegia with diazoxide during the test period. The third graph (far right) is a representative cell exposed to cardioplegia with pinacidil during the test period. The addition of pinacidil to cardioplegia induced potassium current. CPG is hyperkalemic cardioplegia, DZX is diazoxide, NT is normal Tyrode’s solution, and PIN is pinacidil.

B. Pinacidil elicits an increase in adenosine triphosphate–sensitive potassium current in wild type cells exposed to cardioplegia while diazoxide does not. Cell conductance is represented in nanoseimens/picofarads (nS/pF, y axis) and test solution on the x axis. Test solutions included: hyperkalemic cardioplegia (St. Thomas solution), cardioplegia with diazoxide, and cardioplegia with pinacidil. When correcting for cell conductance, there was a significant increase in potassium current in cells exposed to pinacidil supplemented cardioplegia (*p= 0.025 vs. CPG). CPG is cardioplegia, DZX is diazoxide, PIN is pinacidil, and N is number of myocytes.

Myocyte Volume in Wild Type and SUR1(−/−) Mice

Wild type and SUR1(−/−) myocytes exposed to normal Tyrode’s solution had no significant change in volume throughout the experiment (Figure 4). As demonstrated previously, exposure to hyperkalemic cardioplegia resulted in significant (p<0.05 vs. control) wild type myocyte swelling (Figures 4 and 5), and diazoxide (100µM) significantly ameliorated the volume change secondary to cardioplegia (p<0.05). Although diazoxide is apparently without effect on mouse ventricular myocytes, and ventricular myocytes from SUR(−/−) animals contain normal sK_ATP density (11), there was a marked loss of diazoxide action on cardioplegic swelling in SUR(−/−) myocytes (Figures 4 and 5). As discussed below, this result indicates that diazoxide is indeed ameliorating cardioplegic swelling through the SUR1 subunit, but that this is independent of sarcolemmal K_ATP channels.

Figure 4.

Diazoxide requires the SUR1 subunit of the K_ATP channel to maintain myocyte volume homeostasis in isolated ventricular myocytes. Myocyte volume change is represented as relative percent increase in volume for wild type and SUR1(−/−) cells on the y axis. Myocytes were initially exposed to normal Tyrode’s solution for 5 minutes for baseline volume measurement. Myocytes were then exposed to hyperkalemic cardioplegia or hyperkalemic cardioplegia + 100µM diazoxide. Wild type myocytes exposed to cardioplegia significantly swelled, and this was prevented by the addition of diazoxide (*p<0.05 vs. control and cardioplegia+DZX). SUR1(−/−) myocytes also significantly swelled when exposed to cardioplegia (#p<0.05 vs. control), but the addition of diazoxide did not prevent the swelling secondary to cardioplegia exposure (#p<0.05 vs. control). N= 8 cells for KO NT, N= 7 cells for KO WT, N=10 cells for KO CPG, N= 7 cells for WT CPG, N= 8 cells for KO CPG + DZX, and N= 7 cells for WT CPG + DZX.

Figure 5.

Representative isolated wild type or SUR1 (−/−) myocytes at baseline volume and following exposure to test solution via inverted microscopy. Myocytes were initially exposed to normal Tyrode’s solution for 5 minutes for baseline volume measurement. Myocytes were then exposed to test solution (hyperkalemic cardioplegia or hyperkalemic cardioplegia + 100µM diazoxide). The red line drawn on the representative test cell represents the outline of that cell’s baseline volume. Significant volume increase was noted in response to cardioplegia in the wild type (WT) and SUR (−/−) cells, and this was eliminated with the addition of diazoxide in the WT cells only. WT is wild type, CPG is cardioplegia, and DZX is diazoxide.

Discussion

Diazoxide is not Associated with an Increase in sK_ATP Channel Current in Wild Type Myocytes

Diazoxide did not induce a sarcolemmal K_ATP channel current in ventricular myocytes alone or in the presence of hyperkalemic cardioplegia. These experiments indicate that under the observed conditions, diazoxide (DZX, 100ųM/L) does not open the sK_ATP channel in isolated mouse ventricular myocytes, consistent with recent findings also in mouse (11), rabbit (1), and rat ventricular myocytes (19).

Previous experiments documented significant (3.2% from baseline) myocyte swelling secondary to hyperkalemic cardioplegia in myocytes isolated from mice lacking the Kir6.2 subunit of the sK_ATP channel (3). This swelling was not altered by the addition of diazoxide. Together with the findings of the present study, these data could be consistent with an action of diazoxide on a unique ventricular K_ATP channel composed of Kir6.2/SUR1. Such a composition is clearly the major composition in pancreatic, neuronal, and mouse atrial cells (20), but there is clear evidence that SUR2A, and not SUR1, is the critical SUR subunit in generation of mouse ventricular K_ATP (11,21).

Both Pinacidil and Metabolic Inhibition Produce an Increase in sK_ATP Channel Current in Wild Type Myocytes

Pinacidil, a non-specific K_ATP channel opener, was associated with a significant increase in sK_ATP channel current, consistent with the findings of previous studies (22). Metabolic inhibition is known to induce a large sK_ATP channel current, and this is consistent with the results of the present study (23).

Mechanism of Myocyte Swelling Secondary to Stress

In the present study, the myocyte volume derangement following exposure to hyperkalemic cardioplegia alone was evaluated as in previous work (2). Wild type and SUR1(−/−) myocytes both demonstrated a significant increase in size following exposure to hyperkalemic cardioplegia (4–7% increase in size vs. baseline volume). Similar volume derangements have been observed following the stress of mild hyposmotic stress (6% increase in volume) and metabolic inhibition (10% increase in volume) (4,5). Each stress exposed the myocyte to a unique osmotic challenge: hyperkalemic cardioplegia exposed the myocyte to a hyposmotic extracellular environment, hyposmotic stress exposed the myocyte to a different hyposmotic extracellular environment, and metabolic inhibition exposed the myocyte to a hyperosmotic intracellular environment. The K_ATP channel opener diazoxide prevented the volume derangement in response to all 3 stresses independently. The mechanism remains unknown; however, these data suggest a role in cellular volume regulation.

Maintenance of Myocyte Volume Homeostasis by Diazoxide Requires SUR1

Myocytes from mice lacking the SUR1 gene also swelled significantly when exposed to cardioplegia, however, diazoxide failed to reverse the effect. This finding indicates that diazoxide is in fact ameliorating cardioplegic swelling through the SUR1 subunit, but that this is independent of sarcolemmal K_ATP channels. This raises the possibility that SUR1 is actually present in mouse ventricle, but is not required for, or part of the sK_ATP channel.

Location of Action of Diazoxide

The results of the present study suggest that the observed protective mechanism of diazoxide in isolated myocytes (maintenance of volume homeostasis and contractility during stress) is dependent upon the SUR1 subunit, but is independent of the sarcolemmal K_ATP channel. The mechanism of action may be at the mitochondrial (mK_ATP) level or at a K_ATP channel-independent location in the ventricle. Many investigators attribute the myocardial protection provided by K_ATP channel openers to the opening of the purported mitochondrial K_ATP (mK_ATP) channel rather than the sarcolemmal K_ATP (sK_ATP) channel (6,25,26). However, much of the data claimed as support for the existence of a mitochondrial K_ATP channel have been indirect (27–29), including by measurement of mitochondrial flavoprotein oxidation. Critics have noted that diazoxide inhibits succinate dehydrogenase activity, which in turn will inhibit the tricarboxylic acid cycle leading to oxidation of mitochondrial flavoproteins (28). Other evidence supported by patch clamping of the inner mitochondrial membrane documenting the presence of mK_ATP channels has been criticized because of its lack of reproducibility and potential contamination by other cellular membranes (7,28,30). The determination of the structure of a mitochondrial K_ATP channel would thus facilitate localization of diazoxide’s mechanism of action.

Recent studies have documented RNA expression for SUR1 in left ventricular tissue from failing human hearts (24). The elucidation of the composition of the K_ATP channel specifically in humans will also increase the knowledge of the exact site of cardioprotection offered by diazoxide.

Study Limitations

This study investigated the action of diazoxide at the cellular level and in one species. This species was chosen because of the availability of a genetic knockout. Genetically tractable larger animal models are not available and pharmacologic methods have limitations as discussed. Potassium current at the cellular level was measured in order to definitively observe any action of diazoxide at the sK_ATP channel in isolation. Extrapolation to the whole organism level should therefore be taken with caution.

Clinical Relevance

The inverse relationship previously demonstrated between myocyte volume derangement and contractility in isolated myocytes suggested loss of myocyte volume homeostasis as a potential mechanism of myocardial stunning (4). The ability of diazoxide to prevent myocyte swelling and resultant contractile dysfunction secondary to 3 independent stresses in three species suggests that its use may be exploited for the reduction of myocardial stunning. Elucidation of the mechanism of action of diazoxide in the mouse at the cellular and subcellular levels will subsequently facilitate its acceptance and use at the clinical level.