OUP user menu

Diazoxide causes early activation of cardiac sarcolemmal KATP channels during metabolic inhibition by an indirect mechanism

Glenn C Rodrigo, Noel W Davies, Nicholas B Standen
DOI: http://dx.doi.org/10.1016/j.cardiores.2003.10.004 570-579 First published online: 15 February 2004


Objective: We have used isolated myocytes to investigate the effects of diazoxide on sarcolemmal KATP channel (sarcoKATP) activity and action potential failure during metabolic inhibition, and the role of these channels in protection of functional recovery on reperfusion. Materials and methods: Isolated adult rat ventricular myocytes were exposed to metabolic inhibition (NaCN and iodoacetate) and reperfusion. Functional recovery was assessed from the ability of cells to contract on electrical stimulation and to recover calcium homeostasis, measured with fura-2. Action potentials and KATP currents were measured using patch clamp. Results: Pretreatment with diazoxide (100 μM, 5 min) increased the proportion of cells that recovered contractile function after MI and reperfusion from 16.8±2.4% to 65.0±2.2% (p<0.001) and the proportion of cells in which [Ca2+]i recovered to <250 nM. Pretreatment also accelerated action potential and contractile failure during MI. In cell-attached patches, MI activated sarcoKATP channels after 224±11 s, and diazoxide pretreatment decreased this to 145±24 s (p<0.01). However, diazoxide present in the patch pipette did not accelerate sarcoKATP channel activation. Intracellular Mg2+ rose earlier in diazoxide-pretreated cells. The sarcoKATP blocker HMR 1883 delayed action potential failure and reduced diazoxide protection. Conclusions: Diazoxide pretreatment increases recovery of function and [Ca2+]i following reperfusion. Protection is coupled with early action potential failure, due to early activation of sarcoKATP channels during metabolic inhibition (MI), which is likely to involve an indirect effect of diazoxide.

  • K-ATP channel
  • Ischemia
  • Myocytes
  • Preconditioning
  • Rat

1 Introduction

In 1986, Murry et al. [1] described ischaemic preconditioning (IPC), in which brief periods of conditioning ischaemia reduce the size of ischaemia/reperfusion-induced infarction. IPC also improves the recovery of ventricular function and decreases the incidence of arrhythmias [2]. The protective effects of IPC can be mimicked by pharmacological activators of ATP-sensitive K+ (KATP) channels such as cromakalim or diazoxide, while KATP channel blockers can prevent protection by IPC. Thus, the sarcolemmal KATP channel (sarcoKATP) became widely accepted as a mediator of cardioprotection, in agreement with the initial proposal of Noma [3] that the channel serves a protective function in hypoxia. Later, however, it was proposed that a different KATP channel, expressed in the mitochondrial inner membrane (mitoKATP), played a central role in protection [4–6]. Most of the evidence for the involvement of mitoKATP is pharmacological, based on the selectivities of the opener diazoxide and the blocker 5-hydroxydecanoate (5-HD) for mitoKATP over sarcoKATP. However, diazoxide can have channel-independent effects, blocking succinate dehydrogenase [7–9], and 5-HD also has metabolic effects in addition to channel inhibition [8,9]. Furthermore, the molecular composition of mitoKATP remains undetermined and even the existence of a diazoxide-sensitive KATP channel in mitochondria has recently been challenged on the basis of measurements of mitochondrial matrix volume [10].

Conversely, recent studies have provided renewed support for a role of sarcoKATP in ischaemic cardioprotection, and suggest that it may be especially important in functional recovery after reperfusion. In canine hearts, cromakalim and glibenclamide, which should respectively open and block both sarcoKATP and mitoKATP were more effective at mimicking or preventing the complete range of IPC protection than diazoxide or 5-HD alone [11]. In rabbit hearts, HMR 1883, a sulphonylurea selective for sarcoKATP, attenuated IPC [12]. Toyoda et al. [13] found that while glibenclamide blocked IPC protection measured both from infarct size and recovery of contractile function, 5-HD affected only infarct size and HMR 1883 only functional recovery. Finally, strong evidence for a major role of sarcoKATP in the mouse heart has come from knockout of its pore-forming subunit Kir6.2, which abolishes the protective effect of IPC [14].

These findings raise questions about the mechanism by which diazoxide exerts its cardioprotective action, and in particular whether it can cause activation of sarcoKATP channels during ischaemia. Although diazoxide is well known to reduce infarct size [6], less information exists about effects on functional recovery. However, diazoxide-pretreated rat hearts have lower left ventricular end-diastolic pressure and increased left ventricular developed pressure following ischaemia/reperfusion and this appears linked to a reduction in mitochondrial [Ca2+] [15]. As discussed above, functional recovery has been proposed to involve sarcoKATP channels. However, effects of diazoxide on sarcoKATP channel activity during ischaemia and reperfusion have not been reported so far.

We have shown previously that in a single cell model of ischaemia/reperfusion injury, diazoxide pretreatment protects against reperfusion-induced hypercontracture [16]. In the present study, we have investigated the protective action of diazoxide on functional recovery and Ca2+ homeostasis and sought to determine whether diazoxide affects sarcoKATP channel activity. Our results show that diazoxide pretreatment increases the proportion of myocytes that maintain low diastolic [Ca2+]i following reperfusion and recover contractile function. Protection is coupled with early contractile failure during metabolic inhibition (MI) resulting from action potential failure due to early sarcoKATP channel activation. Early channel activation is likely to result from a metabolic effect of diazoxide, rather than a direct effect on the sarcoKATP channel.

2 Materials and methods

2.1 Isolation of single ventricular myocytes

Adult male Wistar rats (around 300 g) were killed by cervical dislocation and single ventricular myocytes isolated as described previously [17]. The investigation conforms with the Guide for the Care and use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1966).

2.2 Measurement of contractile activity, intracellular Ca2+ and Mg2+

Myocytes were placed in a 500-μl chamber on the stage of an inverted microscope, continuously superfused with Tyrode solution at a rate of 5 ml/min, and stimulated at 1 Hz by electrical field stimulation. The washout time of the bath was about 10 s. Contractile activity was determined from video observation of a field containing 8–15 cells using a 20× objective. In some experiments, contraction of individual myocytes was determined from changes in cell length using a video-edge detection system (Crescent Electronics). To measure [Ca2+]i, myocytes were loaded with fura-2 and excited alternately at 340 and 380 nm with a monochromator. Emitted light was collected at >520 nm and fluorescence intensity was measured using a video-imaging system (Photon Technology International). Where shown, images of cells are the ratio of the 340:380-nm images. Free calcium values were estimated using a look up table generated with a fura-2 imaging calibration kit (Molecular Probes). Because this is an in vitro calibration and the KD of fura-2 may differ in the cell, the values give an indication of [Ca2+]i rather than representing precise intracellular concentrations. NADH autofluorescence intensity was <2% of the fura-2 signal, and so did not significantly affect our measurements. [Mg2+]i was estimated using the indicator Mg Green which was excited at 485 nm with emitted light >520 nm collected using a photomultiplier tube.

2.3 Electrophysiological measurements of action potentials and single channel currents

We used an Axopatch 200B patch-clamp amplifier (Axon Instruments) in whole-cell current-clamp mode to record action potentials and in cell-attached mode to record single channel activity. Pipettes were filled with a solution containing (in mM): KCl 140, ATP 0.1, ADP 0.1, MgCl2 1, 4-(2-hydroxyethyl)-1-piperazineethanesulphonic acid (HEPES) 10, titrated to pH 7.2 with KOH, and had resistances of 8–10 MΩ.

2.4 Drugs and experimental solutions

Tyrode solution contained (mM): NaCl 135, KCl 5, NaH2PO4 0.33, Na-pyruvate 5, glucose 10, MgCl2 1, CaCl2 2, HEPES 10, titrated to pH 7.4 with NaOH. MI Tyrode, which was used to induce metabolic inhibition, contained 2 mM NaCN and 1 mM iodoacetic acid (IAA) in substrate-free Tyrode (without glucose or pyruvate). Diazoxide (Sigma) and HMR 1883 (Aventis) were dissolved in dimethyl sulphoxide (DMSO) at 200 and 10 mM, respectively. The final DMSO concentration did not exceed 0.05%, which had no measurable effect on the cellular parameters measured. Fura-2 was dissolved in DMSO containing 5% pluronic acid.

2.5 Data acquisition and statistics

Electrophysiological data were recorded directly to hard disk using pClamp8 software. Statistical significance was calculated using ANOVA followed by Student–Newman–Keuls test or Student's t-test as appropriate, and p<0.05 was considered significant. Data are presented as mean±S.E.M. For experiments involving measurements from a number of cells in a field, we have given numbers of experiments and cells as n=experiments;cells. Each type of experiment used cells from at least three different hearts. Experiments were done at 34 °C.

3 Results

3.1 Diazoxide pretreatment improves functional recovery and calcium homeostasis on metabolic inhibition and reperfusion

We have reported previously that adult rat isolated ventricular cardiac myocytes subjected to metabolic inhibition (MI Tyrode for 7 min) cease to contract in response to electrical stimulation after several minutes, followed by a sustained rigor contraction and increase in [Ca2+]i. On reperfusion, these myocytes hypercontract, have an elevated [Ca2+]i, and most do not recover their ability to contract in response to stimulation [17]. Diazoxide pretreatment reduces the percentage of hypercontracted cells [16]. In the present study, we have assessed functional recovery from the ability of cells to contract in response to field stimulation 10 min after reperfusion with normal Tyrode. We have assessed Ca2+ homeostasis from the ability of cells to recover a low diastolic [Ca2+]i after MI and reperfusion.

Fig. 1A shows that pre-treatment with diazoxide (100 μM, applied for 5 min, 2 min before MI) substantially increased the percentage of cells that recovered contractile function after MI and reperfusion from 16.8±2.4% (n=21;183) in control cells to 65.0±2.2% (n=20;154) in diazoxide-pretreated cells (p<0.001). Diazoxide was just as effective if the time between its removal and MI was increased to 10 min; under these conditions, 67.5±3.8% of cells recovered (n=6;65). Diazoxide also improved Ca2+ homeostasis. Fig. 1D,E shows recordings and images from fields of cells exposed to MI and reperfusion without (control) and with diazoxide pretreatment. MI itself led to a progressive increase in [Ca2+]i to 100–400 nM in both control and diazoxide-pretreated cells. Following reperfusion, it can be seen that most control cells developed high [Ca2+]i, whereas in this experiment, 7/10 cells pretreated with diazoxide showed recovery of diastolic [Ca2+]i to levels close to those before MI. The mean [Ca2+]i measured 10 min after reperfusion in control and diazoxide-pretreated cells in many such experiments is shown in Fig. 1B. In effect, however, rather than reducing Ca2+ loading in all cells, diazoxide provides complete protection of [Ca2+]i in a proportion of cells, while a smaller number lose their ability to control [Ca2+]i. This difference between cells that recover Ca2+ homeostasis and those that do not is clear from the fura-2 images of diazoxide-pretreated cells, where protected cells appear green, while the phase contrast images show that more diazoxide-pretreated cells retain an essentially rectangular morphology (Fig. 1D,E). Fig. 1C summarizes the results from a large number of such experiments and shows that diazoxide pretreatment increased the proportion of cells in which [Ca2+]i, measured 10 min after reperfusion, was, <250 nM.

Fig. 1

Effect of diazoxide on functional recovery and Ca2+ homeostasis in isolated ventricular myocytes. (A–C) The effect of diazoxide pretreatment on the percentage of cells that had recovered the ability to contract in response to 1-Hz electrical stimulation, measured 10 min after removal of MI-Tyrode, intracellular [Ca2+], and the percentage of cells with [Ca2+]i <250 nM at the same time. C indicates control and DZX diazoxide-pretreated cells. Panels B and C show results from 439 control and 285 DZX cells. *p<0.001 vs. control. (D) Recordings of intracellular calcium in fields of control myocytes (below) exposed to metabolic inhibition and its removal and field stimulated at 1 Hz throughout. Each trace shows a recording from a different cell. Images were collected at 10-s intervals and the peaks in the early sections of the records arise from aliasing of the calcium transients associated with contraction as sampling and stimulation change their phase relationship slowly with time. The arrow shows the time at which the measurements shown in A–C were taken. The fura-2 image at the right shows a field of cells recorded 10 min after the removal of metabolic inhibition. The phase contrast image shows the morphology of cells at the same time recorded in a separate experiment. Arrows in the phase images indicate cells that had recovered contractile function in response to stimulation. The numbers above the colour calibration bar show [Ca2+] in nM. (E) Recordings and images as in D, but from diazoxide-pretreated myocytes.

3.2 Action potential and contractile failure during MI occur earlier in diazoxide-pretreated myocytes

Contractile failure during MI or hypoxia results from failure of the cardiac action potential due to an increase in K+ conductance as sarcoKATP channels open [18,19]. We have used simultaneous measurements of cell length and membrane potential to investigate the relationship between contractile activity and electrical excitability in control and diazoxide-pretreated myocytes exposed to MI. Membrane potential was recorded using whole-cell patch clamp and action potentials were stimulated using a 5-ms current pulse at 130% threshold. To minimise the possibility that delivery of ATP or washout of ADP by the pipette might delay the intracellular change in these nucleotides consequent on MI, we used high-resistance pipettes filled with a solution that contained 0.1 mM ATP and 0.1 mM ADP. The correspondence in time between action potential failure measured under these conditions and contractile failure measured in field stimulated cells which are not patch clamped (see below) provides strong evidence that patch recording does not affect the time course of the fall in ATP. Fig. 2A,B show records of membrane potential and cell length from a control and a diazoxide-pretreated myocyte exposed to MI. In both control and diazoxide-pretreated myocytes, loss of contractions during MI accompanied failure of the action potential, while rigor contraction occurred about 30s after these events. Fig. 2C shows data collected from experiments where contractile activity was recorded using video microscopy from field stimulated myocytes, and shows that the time from the start of MI to action potential failure (186±5 s, n=20 cells) was not significantly different from that to contractile failure (192±6 s, n=57). In diazoxide-pretreated myocytes, action potential and contractile failure also occurred at the same time (Fig. 2B,C) but both events occurred earlier than in control myocytes; the times to action potential failure and contractile failure were 142±9 s, n=9 and 138±4 s, n=23, respectively (p<0.001 vs. control in each case).

Fig. 2

Failure of the action potential and contraction in control and diazoxide-pretreated cells. (A) Recordings of membrane potential (above) and cell length (below) from a control myocyte exposed to metabolic inhibition (MI Tyrode) and stimulated at 1 Hz throughout. Action potential failure (APF) and the onset of rigor (R) are indicated by arrows. The stimulating depolarisation sets the upper limit of the trace after action potential failure. (B) Recordings as in A above, but from a cell pretreated with 100 μM diazoxide for 5 min, 2 min before MI. (C) Mean times from the start of metabolic inhibition to action potential failure and contractile failure in control (C) and diazoxide-pretreated (DZX) cells. Cell numbers were 32, 23, 19, 9, respectively. *p<0.001 vs. control.

3.3 Diazoxide pretreatment causes early activation of sarcolemmal KATP channels

Our finding that failure of the action potential occurs earlier in diazoxide-pretreated myocytes suggests that sarcoKATP channels should be activated earlier during MI in these myocytes. Indeed, preconditioning with brief hypoxia has been recently shown to accelerate the opening of sarcoKATP channels during subsequent prolonged hypoxia [20]. We therefore recorded sarcoKATP activity from cell-attached patches on myocytes during exposure to MI. These were the only experiments in this study in which myocytes were not stimulated. Fig. 3A,B shows simultaneous recordings of cell-attached patch current and cell length from a control myocyte and a diazoxide-pretreated myocyte. MI led to activation of sarcoKATP channels in both control and diazoxide-pretreated myocytes, but channels were activated earlier in diazoxide-pretreated myocytes. Mean results are shown in Fig. 3D; the delay to sarcoKATP activation in diazoxide-pretreated myocytes was 145±24 s (n=8) and in control myocytes was 224±11 s (n=17; p<0.01). Channel activation was followed by rigor contraction around 25 s later. The interval between channel activation and rigor was unaffected by diazoxide pretreatment (Fig. 3D), and was very similar to the interval between action potential failure and rigor contraction in stimulated cells (Fig. 2A). To confirm the identity of the channels activated by MI as KATP channels, we measured the single channel conductance by applying ramp voltages to cell-attached patches as described in the legend to Fig. 3C, which shows the single channel current–voltage relation in one such patch. Allowing for the cell resting potential, the reversal potential is close to 0 mV, as expected with 140 mM extracellular [K+], and the unitary conductance measured from six experiments was 79.0±6.5 pS, characteristic of the cardiac sarcolemmal KATP channel [21].

Fig. 3

KATP channel activation in cell-attached patches. (A) Recording of current from a cell-attached patch (above) and cell length (below) from a control myocyte exposed to MI as indicated. Pipette potential +40 mV. (B) Recording as in A, but from a diazoxide-pretreated myocyte. (C) Single channel current voltage relation for the channels activated by MI Tyrode. The relation was obtained from the ensemble average of the open-channel current measured in response to 100 ramps in which the voltage applied to the pipette was ramped from +100 to −100 mV. Since the resting potential of these cells is around −70 mV, the effective voltage change across the patch will be from about −170 to +30 mV. The relation was fitted with a polynomial function and its derivative gave a slope conductance at the reversal potential of 87 pS. (D) Mean times from the start of metabolic inhibition to KATP channel activation and time intervals between channel activation and rigor in control cells (C), diazoxide-pretreated (DZX) cells, and cells in which diazoxide (100 μM added to the pipette solution described in Materials and methods) was included in the cell-attached pipette, DZX(pip). Cell numbers were 17, 6, 8, respectively. *p<0.01 vs. control.

3.4 Local application of diazoxide does not induce sarcoKATP channel activity

It is possible that diazoxide leads to accelerated sarcoKATP channel activation during MI by a direct action on the channel, maybe causing activation when ADP levels rise during MI [22]. This would require that some diazoxide remains close to the channel following diazoxide pretreatment, which seems possible since diazoxide is highly lipid soluble. To investigate this possibility, we compared the times to sarcoKATP channel activation in cell-attached patches during MI in the presence and absence of diazoxide (100 μM) in the patch pipette. Fig. 3D shows that the continuous presence of diazoxide in the immediate vicinity of the channels in the patch did not lead to their early activation during MI, nor did it accelerate the time to rigor. We therefore consider it unlikely that diazoxide causes early sarcoKATP activation by a direct effect on the channel.

3.5 Intracellular Mg2+ rises earlier in diazoxide-pretreated cells

An alternative possibility is that diazoxide pretreatment activates sarcoKATP channels indirectly by changing cellular metabolism to alter the timing of changes in ATP levels on metabolic inhibition, an idea supported by our observation that rigor occurs sooner with diazoxide pretreatment. To obtain an indication of changes in intracellular ATP, we used the Mg2+-sensitive dye Mg Green. Since much of the total intracellular magnesium is present as MgATP, intracellular free Mg2+ rises when ATP is hydrolysed, so that fluorescent Mg2+ indicators can give a sensitive measure of rates of change of ATP [23,24]. Fig. 4A shows that in a control myocyte, Mg2+ initially rose slowly during MI, followed by a rapid rise, and rigor occurred after the Mg2+ signal had reached a plateau, as described previously [24]. Mean results from a number of cells show that the rapid rise in Mg2+ occurred earlier in cells pretreated with diazoxide (Fig. 4B), providing further evidence that diazoxide pretreatment accelerates the fall in ATP on MI.

Fig. 4

Changes in Mg Green fluorescence in response to metabolic inhibition. (A) Recording of Mg Green fluorescence in a control myocyte exposed to MI as indicated and stimulated at 1 Hz throughout. The arrows show the times at which contraction failed and the cell shortened into rigor. (B) Mean (±S.E.M.) results showing the change in Mg Green fluorescence measured in seven control (○) and seven diazoxide-pretreated myocytes (●). The change in fluorescence for each cell was normalised to its maximum value.

3.6 HMR 1883 delays action potential failure and reduces protection by diazoxide

To investigate the role of sarcoKATP channel activation in action potential and contractile failure, we used the sulphonylurea blocker HMR 1883, which shows selectivity for cardiac sarcoKATP channels [25,26]. Fig. 5A shows recordings of membrane potential and contraction from a myocyte exposed to MI in the presence of 10 μM HMR 1883. Under these conditions, action potentials and contractions were maintained until the cell developed rigor contraction, as can also be seen clearly from the recordings shown with an expanded timebase on the right. In contrast, in the absence of HMR 1883, there was a delay of around 30 s between failure of action potentials and contraction and the onset of rigor (see Fig. 2A). The mean data (Fig. 5B) show that action potential failure during MI was significantly delayed to 215±2 s (n=8) in the presence of HMR 1883 (p<0.05 vs. control).

Fig. 5

Effects of HMR 1883. (A) Recordings of membrane potential (above) and cell length (below) from a myocyte exposed to metabolic inhibition in the presence of 10 μM HMR1883 and stimulated at 1 Hz throughout. The panels on the right show on a faster time scale a series of action potentials and the corresponding contractions that span the period of AP and contractile failure, indicated by the interval between the vertical dashed lines on the left. (B) Mean time from the start of MI to action potential failure in control cells (C, from Fig. 2) and in the presence of HMR 1883 (n=20, 8, respectively, *p<0.05 vs. control). (C) Percentage of cells that had recovered the ability to contract in response to 1-Hz electrical stimulation, measured 10 min after removal of MI-Tyrode. DZX, diazoxide-pretreated cells; DZX+HMR 1883, diazoxide-pretreated cells exposed to 10 μM HMR 1883 throughout (n=20, 22, respectively, *p<0.001 vs. control).

The above results are consistent with sarcoKATP activation contributing to early failure of the action potential, and so of contraction, during MI. We therefore considered whether sarcoKATP activation underlies part or all of the functional protection that we observe in diazoxide-pretreated cells. Fig. 5C shows that HMR 1883 (10 μM), applied at the same time as diazoxide and present throughout the remainder of the experiment, reduced the proportion of cells that recovered contractile function from that seen with diazoxide pretreatment alone (p<0.001), but did not completely abolish the protective effect of diazoxide.

4 Discussion

Our results show that pretreatment with diazoxide can improve both contractile function and Ca2+ homeostasis in isolated ventricular myocytes exposed to simulated ischaemia in the form of metabolic inhibition and reperfusion. Further, diazoxide pretreatment leads to earlier activation of sarcoKATP channels during MI, together with earlier failure of action potentials and contraction. Block of these channels both delays contractile failure and reduces protection by diazoxide.

4.1 How does diazoxide cause sarcoKATP activation?

By measuring sarcoKATP channel activity directly in cell-attached patches, we found that the channels first opened about 4 min after the onset of metabolic inhibition in unstimulated myocytes in the absence of diazoxide pretreatment. Channel activity increased rapidly over the next 1–2 min, with peak activity corresponding to the time at which rigor shortening was complete, and then declined slowly. This pattern was the same in diazoxide-pretreated myocytes, but occurred earlier (Fig. 3). Diazoxide is generally considered not to activate cardiac sarcoKATP channels under normal conditions [6], and our failure to observe channel activation in cell-attached patches in response to acute application of diazoxide agrees with this. Although diazoxide has been reported to activate sarcoKATP channels when ADP levels are raised [22], in our experiments, inclusion of diazoxide in the pipette solution did not accelerate activation of KATP channels within the patch during MI, arguing against a direct effect of diazoxide on the sarcoKATP channel. It seems more likely that the effect of diazoxide results from a metabolic action at the level of mitochondria to cause an earlier fall in ATP levels during subsequent MI with corresponding earlier opening of sarcoKATP channels. Our measurements of [Mg2+]i provide strong evidence for such an early fall in ATP in diazoxide-pretreated cells, and this idea is also supported by our finding that rigor contraction occurs sooner, suggesting that the collapse of ATP levels around the myofibrils is accelerated. In intact rat hearts, IPC itself has also been reported to cause ATP to fall sooner during subsequent prolonged ischaemia [27]. It seems likely that the capacity to generate or defend ATP during MI is reduced by diazoxide pretreatment. Whether the initial effect of diazoxide on mitochondrial metabolism involves mitoKATP channels or results from other metabolic effects, for example, inhibition of succinate dehydrogenase, remains controversial at present [5,8–10,28]. Our previous finding that protection by diazoxide is partially blocked by 5-HD [16] does not resolve this problem since 5-HD can have channel-independent effects as well as blocking mitoKATP [8]. A definitive answer as to the role of mitoKATP probably awaits discovery of the molecular correlate of this channel. It has been suggested recently that mitochondrial energetic signals may be coupled to sarcoKATP channels by phosphotransfer mechanisms that involve adenylate kinase or creatine kinase [29,30], and such mechanisms might also contribute to sarcoKATP activation in ischaemia.

4.2 The role of sarcoKATP channels in cardioprotection by diazoxide

Ischaemic cardioprotection by diazoxide has generally been considered, largely on the basis of pharmacological evidence, to involve mitochondrial rather than sarcoKATP channels [6]. However, a recent study has shown that Langendorff-perfused hearts from Kir6.2-knockout mice, which lack sarcoKATP channels, also lack the protective effect of diazoxide, suggesting that sarcoKATP is essential for protection [31]. Diazoxide shortened the cardiac action potential in wild-type but not KO hearts, consistent with effects we observed in isolated rat myocytes, and sarcoKATP block also abolished diazoxide protection [31]. Our findings show that diazoxide can accelerate sarcoKATP channel activation in the metabolically inhibited rat myocyte. SarcoKATP channels contribute to protection, since the effect of diazoxide on functional recovery was reduced by HMR 1883, though not abolished as in the mouse heart. It is possible that sarcoKATP is particularly important in ischaemic protection in the mouse, maybe related to its very high heart rate [31]. It is also possible that HMR 1883 becomes less effective at blocking sarcoKATP channels during the severe MI used in our experiments, so that their role in protection is underestimated. The blocking effectiveness of sulphonylureas on cardiac sarcoKATP channels is reduced by MgADP [32]. Block by glimepiride is inhibited as ADP is raised [33], and our unpublished observations suggest that block by HMR 1883 behaves in a similar manner.

The mechanism by which sarcoKATP opening contributes to protection is likely to be essentially that proposed by Noma [3], (see also Ref. [31]): sarcoKATP channel opening in ischaemia causes action potential shortening and then failure, reducing Ca2+ overload and leading to improved recovery of function on reperfusion. This is consistent with our observations that action potential failure occurs earlier in diazoxide-pretreated cells, but is delayed by sarcoKATP channel blockade with HMR 1883, and is also consistent with the improvement in Ca2+ homeostasis we observed in diazoxide-pretreated cells. Reduced Ca2+ entry could in turn reduce Ca2+-induced Ca2+ release, and so mitochondrial Ca2+ loading. It is also possible that mitochondrial depolarization, which normally precedes rigor, occurs early in diazoxide-pretreated cells, and this could also reduce mitochondrial Ca2+ loading. The extent of the rise in mitochondrial Ca2+ during ischaemia is important in determining cellular fate on reperfusion [34], and preconditioning with both diazoxide and nitric oxide has been reported to reduce Ca2+ loading [15,35].


We thank Diane Everitt for expert technical assistance and the British Heart Foundation for support.


  • Time for primary review 26 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
View Abstract