Cardiovascular Research 1992;26:1077-1086
Regulation of ATP sensitive potassium channel of isolated guinea pig ventricular myocytes by sarcolemmal monocarboxylate transport William A Coetzee
K' channel opening with low levels of cytosolic ATP has been described in cardiac tissue' and subsequently in a variety of other tissue types, including pancreatic p cells, skeletal muscle, neural tissue, and smooth muscle.' This channel seems to be important in a variety of physiological processes; in cardiac tissue, it seems only to open under pathological conditions such as metabolic impairment" or ischaemia.' There is, however, a great deal of controversy regarding its exact role during myocardial ischaemia, mostly because ATP levels during ischaemia (at the time when the channel is thought to open) are still quite high.' In contrast, the KATPchannel in excised patches only opens when the concentration of ATP is decreased to very low (micromolar) values." Thus there is a continuing search for possible modulatory factors of KATP activity during ischaemia. Some compounds have been found to increase KATPchannel activity, including ADP, GDP, and adenosine.' ' Others, however, like lysoph~spholipids,~ might decrease KATPchannel activity. In the present study, current through KATP channels (iK,ATP) was elicited using the metabolic inhibitors, 2,4-dinitrophenol (DNP) or cyanide. The DNP induced current was found to be modulated by sarcolemmal monocarboxylate transport. It appears that transport in the inward direction (as during normal physiological conditions) has an inhibitory action on
iK,ATP, whereas outward transport (as would occur during ischaemia) may have a stimulatory effect on iK.ATP. Some of these results have appeared in abstract form.'" " Methods Preparation of single isolated cardiac myocytes Single myocytes from guinea pig ventricle were isolated using slight modifications of the technique described by Mitra and Morad." Briefly, hearts from guinea pigs (250-300 g, killed by cervical dislocation) were retrogradely perfused at a hydrostatic pressure of 60-80 cm H.0 for sequential periods of 5 min each, using a nominally Ca" free solution, a solution containing the enzymes collagenase and protease, and finally a solution containing a low Ca" concentration (see later). The ventricles were cut into small (-30 mm3) pieces which were agitated in the low CaZt solution to allow living cells to be dispersed. After allowing sufficient time for the cells to sediment, the supernatant was removed and replaced with Tyrode's solution. Experimental techniques After placing the cells in the experimental chamber (a modified microscope slide), sufficient time (usually 5 min) was allowed for cells to settle to the glass bottom. Superfusion was started at a rate of 2-3 ml.min-' for 10 min at a temperature of 33-34"C, before any interventions or recordings were made. Patch type microelectrodes (2-4 MR) were made from filamented borosilicate thin walled glass capillaries ( I .5 mm outside diameter, TWI50F-6 WPI Inc, New Haven, CT, USA) on a vertical puller (model PB-7, Narashige Co, Tokyo, Japan). For single channel recordings, electrodes were fire polished and dipped in Sigmacote (Sigma Chemical Company, Poole, United Kingdom) before use. High resistance seals _______~
Cardiovascular Research, The Rayne Institute, St Thomas' Hospital, London SEI 7EH, United Kingdom: W A Coetzee.
Downloaded from by guest on March 25, 2016
Objective: The aim was to describe the effects of extracellular application of monocarboxylates (pyruvate, lactate, or acetate) on current through KATPchannels (iK.ATP) in isolated guinea pig ventricular myocytes. Methods: The iK.ATP was elicited during whole cell voltage damping by application of metabolic poisons, 2,4 dinitrophenol (150 pM) or glucose free cyanide ( 1 mM) and could be blocked by glibenclamide (3 pM). Results: Extracellular application of monocarboxylates, pyruvate (0.1- 10 mM), L-lactate (0.1- 10 mM), and acetate (10 mM) led to a rapid inhibition of iK,ATP - an effect which was fully reversible upon washout. Substances without any effect on iK.ATP were (10 mM each) gluconate, citrate, glutamate, creatine, succinate, and glycine. The mechanism underlying the effects of monocarboxylates on iK.ATP was unlikely to be related to an increased ATP production, since D-lactate (10 mM) essentially had the same effect on iK,ATP as the L-isomer of lactate. Furthermore, with intracellular dialysis of a-cyano-4-hydroxycinnamate (0.1-0.5 mM), which inhibits pyruvate uptake into mitochondria, extracellular pyruvate exerted the same inhibitory effect on iK,ATP. High concentrations of extracellular a-cyano-4-hydroxycinnamate (4 mM), which blocks the sarcolemmal monocarboxylate camer, prevented the effects on iK.ATP by pyruvate, L-lactate, D-lactate, and acetate. Furthermore, intracellular dialysis with D-lactate (10 mM) led to a more rapid onset of iK.ATP when activated by ATP free dialysis. Activity of isolated KATP channels, measured in isolated membrane patches in the inside out or outside out configuration, typically had a single channel conductance of around 80 pS and was blocked by glibenclamide (3-9 pM). No significant effect of pyruvate was observed in either patch configuration. Conclusions: In cardiac tissue there may be some modulatory role involving monocarboxylate transport on KATPchannel activity, the nature of which is unclear at present but which may involve cytosolic pH changes. Physiological and pathophysiological implications of these findings are discussed. Cardiovascular Research 1992;26: 1077-1086
were established (in the cell centre) between the cell membrane and the electrode tip by applying slight suction to the lumen of the pipette. Following the neutralisation of capacitive transients, one of three different procedures were followed. For inside out patches, the bathing solution was changed to the test (K'rich) solution before lifting the electrode from the cell surface. Often, the patch was briefly exposed to air or touched against the glass bottom of the bath before single channel appeared - presumably because of vesicle formation. For whole cell recording,'.' the membrane under the electrode tip was ruptured by an additional pulse of stronger suction ("breaking into the cell"). At least 5 min was allowed for intracellular dialysis to occur before any recordings began. The latter procedure was also followed to make outside out patches. Here, however, the bathing solution was changed to a K' rich solution before lifting the electrode from the cell. In the whole cell mode, current through KATPchannels was activated using the metabolic poisons DNP ( 100-150 pM) or cyanide ( I mM) in the absence of extracellular glucose. In whole cell mode, the standard A 3000
I - 100
Single cell isolation For cell isolation, the following solutions were used: nominal Ca" free solution contained (in mmol4tre-'): NaCl 130, KCI 5.4, KH2P04 1.2, MgS04 1.2, CaCh 0.76, HEPES 6. glucose 10, and EGTA 1 (pH adjusted to 7.2 at 34°C with NaOH). The Ca" activity in the latter solution was calculated as I pM. A low Ca2' (180 pM) solution was obtained by addition of CaClz to the nominal Ca" free solution. Collagenase (10-13 mg, Type 1, Sigma Chemical Company, St Louis, MO. USA) and protease (10 mg, Sigma) was added to 50 ml of the low Ca-' solution.
Whole cell recordings The Tyrode's solution used contained (in mmol.litre-'): NaCl 140, KC1 5.4, CaC12 1.8, MgCh 0.5, NaHZPOr 0.33, HEPES 5 , and glucose 5 or 10. The pH was corrected to 7.4 by addition of NaOH at 34°C. The pipette solution contained: K aspartate 110, KCI 20, EGTA 10, HEPES 10, Na2ATP 5, and MgClz 5.45 (pH adjusted to 7.2 at 34OC with KOH). In all cases when free [Ca"] and [Mg"] was calculated it was done using a computer program (CABUF) kindly supplied by Dr G Droogmans (Department of Physiology, University of Leuven, Belgium). Single channel recordings For experiments on single channels in the inside out or outside out configurations, a solution was prepared consisting of (mmol4tre-'): KCI 140, EGTA I , HEPES 10, and MgCh 1.05 (pH adjusted to 7.2 with KOH at 34°C). To part of this solution 2 mM KzATP and 1.78 mM MgClz were added. Different concentrations of ATP (at constant [Mg"]) could then be obtained by appropriate dilutions. Drugs and chemicals The following were added to experimental solutions as a Nat salt where possible. or as a free acid without any corrections for ionic composition or osmolarity. Pyruvate (BDH Laboratory supplies, Poole. United Kingdom), acetate (BDH), oh-lactate (Hopkins and Williams Ltd, Chadwell Heath, United Kingdom) L-lactate (Sigma), D-lactic acid (Sigma), D-gluconate (Sigma), succinate (Sigma), citrate (Sigma), and glutamate (Sigma). Other substances were glycine (Sigma) and creatine (Sigma). In all cases, pH was checked and corrected with NaOH when needed. 2.4-dinitrophenol (BDH) and glibenclamide (Sigma) were made up as a stock solution in DMSO. a-Cyano-4-hydroxycinnamate (Sigma) was solubilised by heating Tyrode's solution to about 50°C and stirring for 1-2 h. The pH was corrected with NaOH. Data collection and analysis Whole cell and single channel currents were recorded using a Dagan Model 8900 patch clamp amplifier (Dagan Corp, Minneapolis, MN, USA) with a 0.1 G f i headstage in whole cell mode or a 10 Gfi headstage in single channel mode. Signals were simultaneously displayed on a digital oscilloscope (Gould 1425, Cleveland, Ohio, USA) and a paper chart recorder (Gould 2400S), and also acquired on a PCM video system (Unitrade, Axon Instruments, Foster City. California, USA) and an IBM PC computer for later analysis (pClamp, Axon Instruments). Voltage pulses for whole cell voltage clamping were elicited using a custom built stimulator. Single channel currents were sampled at 3 kHz (Fetchex, Axon Instruments) after low pass filtering (-3 dB at 500-1000 Hz; Bessel response). Data were aquired using a commercial program (Fetchex, Axon Instruments) with subsequent analysis using a computer program (ASCD; Dr G Droogmans). Current out of the pipette was defined as negative current.
Results Figure 1 Effect of 2.4-dinitrophenol (DNP) and glibenclamide on membrane current of guinea pig ventricular myocytes. ( A ) A current-voltage relation (ramp pulse; +5 to -105 mV; ramp rate of -I0 m V Y ' ) is shown for control and in the presence of DNP (150 pM). ( B ) Examples of current traces (another preparation) as a result of step changes in membrane potential from a holding potential of -84 mV to a test potential of 0 mV ( 2 s duration, 0.2 Hz; n o symbol). The current scale bar represents 2 nA. The bottom of the current scale bar, in this and in subsequentfigures, represents zero current. Effects of DNP (150 pM; filled symbols) and subsequent addition of glibenclamide (3 @, empty symbol) is shown.
Activation of current through KATPchannels
The application of DNP leads to a large outward shift of membrane current at positive potentials. This is illustrated in fig 1A, where current-voltage relationships were obtained using ramp voltage pulses (+5 to -105 mV; ramp rate 10 mV.s-'). In control, total membrane current at positive potentials (ie, 0 mV) is small. This is due to the well known property of inward rectification. Application of DNP (150 kM) led to a marked outward shift of membrane current at positive potentials. The difference current reverses at a potential expected for a K+ current but with rectification
Downloaded from by guest on March 25, 2016
voltage clamp protocol throughout the study consisted of continuous step clamps (2 s duration) to 0 mV from a holding potential of -75 to -85 mV repeated at a rate of 0.2 Hz. In single channel experiments, recordings were made at a steady holding potential as indicated in the text.
Monocarboxylates and the KATPchannel
characteristics of current through iK.ATP. The reversal potential of the DNP induced current is close to -80 mV. In figure lB, this finding is again illustrated, but using a step voltage pulse from a holding potential of -84 mV to 0 mV. DNP (150 pM) caused a large shift of current during the clamp pulse, which could be blocked by glibenclamide (3 pM). Following the return to the holding potential, an inward tail current of unknown origin can be seen. One possibility is that this tail might be caused by local [K'], changes (ie, in t tubules) as a result of the large current flow.
0 r 5
Figure 2 A continuous recording at slow time scale of membrane
current of guinea pig ventricular myocyte as a result of the voltage protocol followed throughout this study, which is described in the text. (A) 2.4-Dinitrophenol (DNP) ( I S 0 pM) and pyruvate (10 mM) were added as indicated. This result is representative of 31 similar experiments. ( B ) Current changes as a result of glucose free cyanide ( I mM) superfusion. Glucose (50 mM) or pyruvate (10 mM) was added to the cyanide solution as indicated. The inward current in response to cyanide was cut offby the limit of the chart recorder: The holding potential was -80 mV in both traces and a 2 s depolarising pulse was applied to 0 mV (0.2 Hz).
Figure 3 Inhibition of 2,4-dinitrophenol (DNP) induced current by different concentrations (0.I , 0.3, I , 3, and I0 mM) of pyruvate and lactate. The numbers of cells per concentration are 3, 2, 7, 3, and 31 for pyruvate und I , I , 4, 4, and 2 for lactate. The inhibition
is defined as a percentage of maximum current (at 0 mV) obtained in the presence of DNP - usually following washout of pyruvate or lactate. Columns are means, bars=SEM, of results from individual cells. In the majority of experiments, only one compound was applied to a given cell. When more than one concentration was applied, suficient time was aIlowed for the effects of the first to wash 08
Downloaded from by guest on March 25, 2016
Effects of monocarboxylates on DNP or cyanide induced current Shown in fig 2A is a similar result as in fig 1B (using the same voltage protocol), but at a continuous slow time base. Within 30 s after the application of DNP (150 pM), there was a large outward shift of membrane current measured at 0 mV. The time of onset, rate of onset, and the magnitude of the DNP induced current differed greatly among different preparations and was therefore not systematically analysed. The DNP induced current often reached a plateau and was relatively stable until the onset of cell shortening and irreversible rounding up of the cell (hypercontracture). The exact reason for hypercontracture is unknown. The time to onset of hypercontracture was also highly variable from cell to cell, but usually allowed 1-20 min during which relatively stable recordings and interventions could be made. When cell shortening occurred following DNP application, it usually took the form of a slow and progressive shortening (while the cell remained relatively square) up to a point at which the rate of contracture accelerated and the cell took on a round form with the formation of blebs. The experiment was terminated when any sign of cell shortening was observed. The large outward shift of current at 0 mV caused by DNP was usually found to be fully reversible before the onset of cell shortening and hypercontracture (see fig 2).
The novel aspect of this study was that pyruvate was able to reverse the effect of DNP very rapidly and mostly completely (fig 2). In fig 2A, the effect of pyruvate (10 mM) on the DNP induced current is shown. Following the addition of pyruvate in the continued presence of DNP (150 pM), the DNP induced current was reduced within 30 s. Pyruvate was washed out after about 2 min, which led to a re-establishment of the DNP induced current. This rate of washout was variable between preparations, but usually occurred at a slower rate than the rate of onset. Similar effects were seen in a total of 31 cells using 10 mM pyruvate. A further addition of glibenclamide (3-9 pM) in the presence of pyruvate did not cause any further block (n=3). Pyruvate (10 mM) was also able to reverse to a large extent the shortening of action potential caused by DNP (n=3, not shown). In fig 2B, application of cyanide (1 mM) in the absence of extracellular glucose led to a similar predominantly outward shift of membrane current. Addition of glucose (50 mM) led to a decrease in the rate of development of cyanide induced current, whereas pyruvate (10 mM) had a very marked and rapid inhibitory effect. This effect of pyruvate was reversible. It was confirmed in three other experiments that 50 mM glucose (given as the sole intervention) did not cause the same degree of inhibition of the DNP induced current. Lower concentrations of pyruvate and lactate (0.1-3 mM) were still able to cause a significant inhibition of DNP induced current (fig 3), but these effects were often characterised by a transient (unsustained) inhibitory effect. An interesting observation was that the effect of pyruvate on DNP induced current seemed to decrease as a function of
time. An example of such a phenomenon can be seen in fig 4. Although this phenomenon was not studied in detail, a second exposure to the concentration of pyruvate (in the continued presence of DNP) often led to a slower rate and less inhibition than the first.
Effects of DNP induced current by other substances Substances with structures analogous to those of pyruvate and lactate were examined for possible effects on DNP induced current. In three cells, acetate (10 mM) led to a rapid inhibition of the DNP induced current (see fig 4)- an effect comparable with that of pyruvate or L-lactate. One of the molecular similarities between these three compounds is the presence of a carboxylic acid group. To test whether the effect is mediated via the presence of such a group, the effect on DNP induced current of other compounds having carboxylic acid groups were tested. Figure 4, A-D, shows the effects of pyruvate, gluconate, citrate, succinate, glutamate, and acetate (10 mM each) on DNP induced current in a single experiment. As before, pyruvate (10 mM, fig 4A) led to a rapid inhibition of the DNP induced current. This effect was fully reversible upon washout of pyruvate. Gluconate (10 mM) and citrate (10 mM, fig 4B) had no effect on the DNP induced current. This
Effects of different isomers of luctute The efficacy of the two isomers of lactate on DNP induced current was examined next. In fig 6, the DNP induced current was elicited as before, using the same voltage protocol. Immediately following the application of D-lactate (10 mM), there was a small and transient outward shift of membrane current at 0 mV, followed by a much larger inward shift to almost control conditions. The effect of D-lactate was fully reversible. A subsequent application of L-lactate (10 mM) led essentially to the same result. There
Figure 4 Effects of (10 mM each) ofp.vruvate, gluconate, citrate, succinate, glutamate, and acetate on 2.3-dinitrophenol (DNP) induced current. DNP (150 f l )wus already present at the start of the figure. The various panels (A, B, C. D) are a continuation of a single experiment with the various substances applied as indicated. The inward current dejections in the presence of pyruvate and acetate rejects the recovery of an inward current component (cut offin C by the limit ofthe chart recorder) and was not studied in detail.
Figure 5 Effects of glycine (10 mM). acetate (10 mM), and glibenclamide (3 pM) on 2,3-dinitrophenol (DNP) induced current. The effect ofglibenclumide was very slowly reversed on washout. Note that the inward current was cut off by the limit of the chart recorde,:
Downloaded from by guest on March 25, 2016
was followed by a second application of pyruvate (10 mM, fig 4C). illustrating the reproducibility of this phenomenon. Note, as mentioned before, the slower rate of inhibition resulting from a second exposure of pyruvate. Following washout of pyruvate, succinate (10 mM) or glutamate (10 mM, fig 4D) had no effect on DNP induced current. If anything, there was a slight increase of current magnitude, although this phenomenon was not studied in detail. Acetate (10 mM, fig 4D) led to an even more rapid inhibition of DNP induced current (compared with second pyruvate application) - suggesting that acetate may be more potent than pyruvate or lactate to inhibit the DNP induced current. As some of these compounds may bind Ca”, the effect of a lowered extracellular Ca” concentration was investigated. No significant effect on DNP induced current was observed when [Ca”], was lowered to 0.18 mM (fig 4A). In different experiments where not more than one intervention was made to any given cell, the following compounds were found to be without an inhibitory effect on DNP induced current: gluconate (10 mM, n=3), citrate ( 10 mM, n=2), creatine (I0 mM, n=3), succinate (10 mM, n=3). and glutamate (10 mM, n=3), thus ruling out the presence of a carboxylic acid group as the sole mechanism for the inhibitory effect on DNP induced current. The effect of glycine (a small 2 carbon amino acid) on DNP induced current was investigated. Figure 5 shows that the application of glycine (10 mM) had no effect on DNP induced current, whereas acetate (10 mM) led to a marked and rapid depression. Replacement of acetate with glycine again led to the development of the DNP induced current, which was blocked by a subsequent addition of glibenclamide (3 pM). Essentially the same result was obtained in another similar experiment. Glucose (20 mM, n=2; 50 mM, n=2) also had no major effects on DNP induced current. From these experiments it appears that the effect is limited to short (2-3 carbon) monocarboxylates and that the amine group interferes.
Monocarboxylates and the KATPchannel
Figure 6 The effect ($two isomers of Iuctute on 2,3-dinitrophenol (DNP) induced current at 0 mV measured as a response to voltage stepsfrom -80 mV to 0 mV u-lactate (10 m M ) or [.-lactate (10 mM) was added following the addition of DNP (150 pM) us indicated. The discontinuity of the current trace is due to datu nyuisition using ramp data pulses (not shown).
were subtle differences, however. Note the absence of the initial small outward shift of membrane current and the faster rate of onset and washout of L-lactate. In eight other similar experiments, the same result by D-lactate on DNP induced current was observed in one, while only partial (often transient) effects were observed in five, and no effects in two cells.
-P ’ I
Effects of pyruvateAactate transport inhibitor
current (n=7). Thus it seems that CIN was able to counteract the inhibitory effect of pyruvate on DNP induced current. In some experiments, CIN (0.1-0.5 mM) was included in the pipette solution. These cells were dialysed with CIN for at least 10 min before the application of DNP (150 pM).In these experiments, pyruvate (10 mM) was still able to inhibit the DNP induced current (n=3), results not shown). lntracellular dialysis with pyruvate o r D-lactate
Cells were internally dialysed for a minimum of 10 min before any interventions were made using a pipette filling solution containing 10 mM pyruvate. Despite the presence of pyruvate, DNP (150 pM) was still able to induce the same current as before. Furthermore, a subsequent addition of extracellular pyruvate (10 mM) led to the same inhibition of the DNP induced current (n=5, not shown). In a separate group of cells, iK.ATP was activated using a pipette solution not containing ATP.’‘ These results are illustrated in fig 8. The open symbols represent three control experiments using the same voltage protocol as before, where an outward shift of membrane current at 0 mV was observed between 45-70 min. This current could be blocked by glibenclamide (3 pM). With D-lactate (10 mM) present in the pipette solution (filled symbols), this outward shift in membrane current occurred sooner (7-25 min). Lack of effect ofpyruvate on KATPchannels
To test for a direct effect of pyruvate on the KATP channel, patch clamping was performed using excised patches. In the inside out configuration, the KATPchannel was identified by its sensitivity to [ATP] in the bath. The experimental protocol consisted of sequential exposure of the “cytosolic” face of the patch to the following solutions: 50 pM ATP, 50 pM ATP plus 10 mM pyruvate, washout of pyruvate (50 pM ATP), and finally with 2 mM ATP. The
Downloaded from by guest on March 25, 2016
The possibility was investigated that monocarboxylates mediated their effects on the DNP induced current via sarcolemmal entry. For these studies, the pyruvate/lactate transport inhibitor, a-cyano-4-hydroxycinnamate14 (CIN) was used. In the first series of experiments, cells were pretreated with CIN ( I or 4 mM) for 10 min. Although it cannot be excluded that CIN by itself affected membrane currents, no significant effects of CIN on membrane current as a response to voltage steps from -70 mV to 0 mV were observed. Following this period, DNP was applied. The presence of CIN did not prevent the activation of the DNP induced current (not shown), suggesting that CIN by itself did not have a major blocking effect on KATP channels. Subsequent addition of pyruvate in the presence of CIN (4 mM) had no effect on DNP induced current in 88% of cells studied (14/16), whereas a small but transient reduction of DNP induced current was observed in the other two cells. CIN (4 mM) was also able completely to prevent the effects of L-lactate (10 mM, n=4) and D-lactate ( 1 0 mM, n=6). Glucose (n=3) had no effect on DNP induced current in the presence of CIN (4 mM), whereas acetate (10 mM, n=9) only had a partial and transient effect in 5/9 cells, and no effect in the remainder. A similar partial and transient reduction of DNP induced current was observed by pyruvate (10 mM) in 3/5 cells in the presence of I mM CIN, with no effect by pyruvate in the other two cells. In a second approach, an attempt was made to reverse the pyruvate induced inhibition of DNP induced current using CIN (4 mM). In fig 7A, following the inhibition of DNP induced current by pyruvate (10 mM), CIN (4 mM) was added. The pyruvate induced inhibition of DNP induced current was reversed by CIN. After a brief washout of pyruvate, a second exposure of pyruvate (still in the presence of CIN) was without effect. In fig 7B, current at 0 mV is plotted in the presence of DNP ( I 50 pM) and pyruvate ( 10 mM). Shown are experiments with CIN (4 mM; filled symbols) and without (empty symbols). In experiments with CIN, the DNP induced current reactivated within 180(SEM 42) s, n=6. However when no CIN was present, the presence of pyruvate led to a long term inhibition of DNP induced
40 Time (rnin)
Figure 8 The effect of long term intracellular dialysis of single cells on membrane current at 0 mV using a pipette solution with no ATP The MgCh was adjusted to keep the calculated ionic activity of Mg2+ at I mM. The empty symbols represent control experiments, while the filled symbols represent experiments performed with 10 mM 0-lactate in the pipette solution. The voltage protocol was the same as before. Experiments were terminated with the application of glibenclamide or when signs of hypercontracture were observed.
[Mg”] was kept constant at 1 mM (see Methods). The top part of fig 9 shows randomly selected traces (from an experiment which is representative of five similar experiments) illustrating activity of KATPchannels. The pipette potential was +90 mV. Channel activity was abolished by 2 mM ATP (fig 9C), whereas pyruvate (10 mM) had no significant inhibitory effects on channel activity (fig 9B) also seen in the probability density functions at the bottom of the figure. Thus the marked effect of pyruvate on whole cell currents cannot be explained by a direct blocking effect of intracellular pyruvate of the KATPchannel. To test for possible extracellular effects of pyruvate on KATPchannel, two approaches were used. In the first, the experiments mentioned in the preceding paragraph were repeated, but with pyruvate (10 mM) present in the pipette (ie, on the extracellular side). In three such experiments (not shown), activity through K A Tchannel ~ was readily observed following patch excision, showing that extracellular pyruvate probably did not have a marked inhibitory effect. In a second approach, current through outside out patches was measured (using identical K’ rich solutions on either side of the patch, but with 50-90 FM ATP in the pipette). No significant depression of current through KATPchannels in the presence of pyruvate (10 mM) was observed. In the experiment depicted in fig 10, at least two open levels are present with approximate single channel conductances of 36 and 80 pS Downloaded from by guest on March 25, 2016
Figure 9 Effect of cytosolic pyruvate (10 mM) on ATP sensitive K‘ channel activity in an inside out patch. The “cytosolic” solution contained 150 pM ATP (in A), 50 pM ATP plus 10 mM pyruvate (in B). This was followed by a washout of pyruvate during which channel uctivity was unaltered (not shown) andfinally by the application of 2 mM ATP (C). The top of each part of the figure representsfour randomly selected sweeps of activity of ATP sensitive K* channel in inside out configuration, and the bottom illustrates probability density functions (0.1 ms bins) for the same conditions. The traces were low pass filtered (500 Hz)and sampled at 2.5 kHz. The single channel conductance was estimated at 69 pS (at a pipette potential of +90 mV). In four similar experiments, the single channel conductance was measured as 70-85pS at [r(l,,=140mM (pipette potentials ranging from +30 to +I10 mV).
Monocarboxylates and the KATPchannel Gli benclamide
(measured from the oscilloscope screen) respectively. The single channel conductance of the first corresponds with that of the inward rectifier" (40 pS) and was not blocked by glibenclamide (fig 10D). The other channel corresponded with a KATPchannel - both in single channel conductance and in its sensitivity to glibenclamide (fig 10 A,B). Pyruvate (10 mM) did not cause a significant blockade of the KATP channel. Although the presence of multiple KATP channels or inward rectifying channels complicated analysis, similar results were found in a total of five patches.
Discussion Current through KATPchannels In this study, metabolic poisons DNP or cyanide were used to activate an outward rectifying membrane current in guinea pig ventricular myocytes. The DNP induced current appeared despite the presence of high concentrations of ATP in the pipette (see also16 18). A similar current was previously observed following metabolic blockade'' 2" and can be attributed to the opening of KATPchannels I 3 4 2 1 22 for several reasons: first, the current-voltage relation of DNP sensitive current exhibits outward rectification, similar to that of the KATPchannels in physiological [K']:3; second, the reversal potential of the DNP sensitive current is close to the predicted equilibrium potential for K' ions (see fig 1); third, contribution of the DNP induced current (at 0 mV) is at least an order of magnitude larger than that of the delayed and inward rectifiers (fig 1); and finally, the DNP induced current is sensitive to glibenclamide, which is thought to be specific for KATP channel^.'^ Because extracellular application of some monocarboxylates partially blocked this current, one can conclude that KATPchannel activity was diminished under these conditions. It does not rule out, however, the possibility that other membrane currents can also be affected by these compounds.
Absence of effects from increased ATP production A major finding of this study is that current as a result of metabolic poisoning (ie, KATP channel activity) was inhibited by application of pyruvate, lactate, or acetate. The most obvious explanation would be an enhanced ATP production from energy metabolism (despite metabolic poisoning) to cause indirect block of KATPchannels. Thus DNP is known to uncouple oxidative phosphorylation, leading to an increased rate of turnover of the citric acid cycle, as is evident from an increased oxygen consumption in the presence of DNP.26 This would lead to guanosine triphosphate (GTP) formation in the citric acid cycle, which may be converted to ATP to cause subsequent blockade of KATPchannels. To exclude this possibility, experiments were performed using cyanide, which blocks oxidative phosphorylation by end product inhibition of the citric acid cycle.26Under these conditions, no indirect formation of ATP via GTP can occur. However, pyruvate still had the same inhibitory effect on KATP channels. This result argues against a possible metabolic effect of monocarboxylates, but also rules out a possible interaction between pyruvate and DNP, as the structure of cyanide is completely different. The application of glucose during cyanide poisoning (fig 2) stimulates ATP production from glycolytic pathways and partially prevents effects of metabolic blockers on the action potential.*" 27 Although the rate of development of CN induced current was slowed by glucose, increased [glucose] never led to the same speed and degree of inhibition of iK.ATP that was observed with monocarboxylates. These results suggest that even high rates ATP production from glyc~lysis'~were unable to mimic the effects of pyruvate. Furthermore, intracellular dialysis with CIN, in concentrations which should completely block pyruvate uptake into the mitochondria (0.1-0.5 mM),2ydid not prevent pyruvate from exerting an inhibitory effect on DNP induced current. Thus it is inconceivable that pyruvate enhanced
Downloaded from by guest on March 25, 2016
Figure I0 Lack of effect of pyruvate (10 mM) on ATP sensitive K ' channel in outside out conjguration. The pipette contained 90 ph4 ATP: and 140 mM K' was present on either side of the membrane. The pipette potential was -60 mV and current was low pass filtered at 500 Hz. ( A ) Slow time scale recording of current in outside out conjiguration. Shown is application ojpyruvate (10 m M ) and glibenclamide (3 pM). (B), (C), and (D): Current recordings at a higher time resolution (100 ms.division8) before application of pyruvate, during exposure to pyruvate, and during exposure to glibenclamide respectively. There appeared to be two conductance levels, corresponding to about 36 pS and 80 pS respectively. Vertical calibration in (B), (C), and ( D ) is 2 PA.
(see earlier), and thus may cause an intracellular concentration of pyruvate which may be lower than that in the pipette. Nevertheless, despite the presence of intracellular pyruvate, DNP still activated iK.ATP, suggesting that intracellular pyruvate had no major blocking effect on KATP channels. The influence of monocarboxylates on other membrane channels was not studied. Even if they were influenced, however, it seems unlikely that their effects on iK.ATP could be explained in this way. In both inside out and outside out excised patches, KATP channels were readily observed. These channels had single channel conductances of around 70-80 pS with 140 m M K' on both sides of $e membrane (cf 78 pS in Kakei et d 3 ;79 pS in Horie et al- ) and were blocked by glibenclamide and cytosolic ATP. The KATPchannel activity was not markedly influenced by pyruvate in either patch configuration. A similar observation was made by Lederer and Nichols? who found no significant effects of cytosolic lactate (20 mM) on KATPchannel activity. These investigators did not describe effects of lactate in outside out patches.
Sarcolemmal transport of monocarboxylates Cardiac sarcolemmal permeability to pyruvate and lactate is much higher than expected from simple passive diffusion33 because of the presence of a sarcolemmal monocarboxylate ~arrier.~" 34-37 As mentioned before, a similar carrier for pyruvate exists in mitochondrial inner membranes.14 Both these carriers are blocked by CIN,I43" although higher concentrations of CIN are needed to block the sarcolemmal carrier3"(90% inhibition occurs with 5 mM CIN). Thus high concentrations of CIN (4 mM) are normally used to block sarcolemmal transport of lactate or pyruvate in intact cells.' I' The present study shows that the inhibitory effects of pyruvate and lactate on DNP induced current were almost entirely prevented by high (4 mM) but not lower ( I mM or less) concentrations of CIN. These data suggest that pyruvate and lactate exerted their effects on DNP induced current (directly or indirectly) via a carrier mediated sarcolemmal transporter. Another argument in favour of this hypothesis is that the D-isomer of lactate also inhibited the DNP induced current (see earlier), which is in concordance with the low isomer specificity of the sarcolemmal lactate carrier." 3" In plasma membrane of erythrocytes, acetate is thou ht not to be transported by the same carrier as pyru~ate!~ In this study, however, the effect of acetate on DNP induced current seemed more pronounced than that of pyruvate (fig 4), and was only partially prevented by CIN. These results suggest that the transport of acetate in cardiac sarcolemma may occur via a different carrier (which is only partially blocked by CIN), or alternatively that acetate is transported via the same carrier as pyruvate, but that the carrier is not completely blocked by CIN. The lack of effects on iK.ATP by other compounds such as glutamate, gluconate, citrate, creatine, or succinate is consistent with the concept that sarcolemmal transport of these substances is low. Further, the mere presence of carboxylic groups as a cause of the observed effects on iK.ATP can be excluded, as many of these compounds share with pyruvate, lactate, and acetate the presence of one (or more) carboxylic groups.
Possible mechanisms? Because there does not seem to be any direct effect of intracellular or extracellular monocarboxylates on the activity of KATPchannels, the results may point to a possible link (direct or indirect) between the activity of the sarcolemmal monocarboxylate carrier and that of the KATPchannel. However, such a putative link cannot be the sole regulator of KATP channels, as evidenced from cell free recordings of these channels in the absence of intracellular or extracellular monocarboxylates.' 2s Rather, monocarboxylate transport may only play a modulatory role. From the data in fig 2, it appears that lactate/pyruvate transport into the cell exerts an inhibitory effect on the iK.ATP. Dialysis of cells with D-lactate led to a more rapid onset of iK.ATP. These data are consistent with those of Keung and Li,'9 who found that dialysis with 20-40 mM lactate and 2-5 mM ATP led to a rapid activation of iK.ATP in whole cell recordings. Although these results are not conclusive (D-lactate may have had other intracellular effects), it seems possible that outward transport of lactate/pyruvate may have a stimulatory effect on iK.ATP. The exact mechanism of the inhibitory effect of monocarboxylates on KATP channels is not known. An attractive candidate is inhibition of KATP channels by intracellular acidification. Thus it is known that extracellular application of weak acids (including pyruvate, lactate, and acetate) leads to intracellular acidification." Despite the well known mild stimulatory effect on KATP channels by intracellular acidosi~,~ a recent preliminary report suggests that these channels may be inhibited by stronger a~idification.~" The challenge in proving this hypothesis is that extracellular monocarboxylates only cause a modest acidification (about 0.3 pH units''), whereas inhibition of KATPchannels occurs at pH