Mechanism of Cloned ATP-sensitive Potassium Channel Activation by Oleoyl-CoA*

Insulin secretion from pancreatic beta cells is coupled to cell metabolism through closure of ATP-sensitive potassium (KATP) channels, which comprise Kir6.2 and sulfonylurea receptor (SUR1) subunits. Although metabolic regulation of KATP channel activity is believed to be mediated principally by the adenine nucleotides, other metabolic intermediates, including long chain acyl-CoA esters, may also be involved. We recorded macroscopic and single-channel currents from Xenopusoocytes expressing either Kir6.2/SUR1 or Kir6.2ΔC36 (which forms channels in the absence of SUR1). Oleoyl-CoA (1 μm) activated both wild-type Kir6.2/SUR1 and Kir6.2ΔC36 macroscopic currents, ∼2-fold, by increasing the number and open probability of Kir6.2/SUR1 and Kir6.2ΔC36 channels. It was ineffective on the related Kir subunit Kir1.1a. Oleoyl-CoA also impaired channel inhibition by ATP, increasing the Ki values for both Kir6.2/SUR1 and Kir6.2ΔC36 currents by ∼3-fold. Our results indicate that activation of KATP channels by oleoyl-CoA results from an interaction with the Kir6.2 subunit, unlike the stimulatory effects of MgADP and diazoxide which are mediated through SUR1. The increased activity and reduced ATP sensitivity of KATP channels by oleoyl-CoA might contribute to the impaired insulin secretion observed in non-insulin-dependent diabetes mellitus.

Potassium channels that are inhibited by ATP (K ATP channels) 1 are found in many tissues, where they serve to couple the metabolic state of the cell to its electrical activity (1). In the pancreatic beta cell, for example, they provide the link between changes in blood glucose concentration and insulin secretion. The K ATP channel sets the beta cell resting membrane potential, and its closure in response to glucose metabolism elicits membrane depolarization, activation of voltage-gated Ca 2ϩ channels, and a rise in Ca 2ϩ influx that stimulates insulin release (2). In cardiac muscle, and in brain neurons, K ATP channels may be involved in the response to ischemia, whereas in smooth muscle they are important for the regulation of vascular tone (3,4).
K ATP channels are formed by the physical association of four inwardly rectifying K ϩ channel (Kir6.2) subunits with four sulfonylurea receptor (SUR) subunits (5)(6)(7). Kir6.2 serves as an ATP-sensitive pore that is common to many types of K ATP channels. SUR is a regulatory subunit that modulates the channel gating properties, enhances the apparent ATP-sensitivity, and acts as the target for sulfonylurea drugs, K-channel openers and intracellular Mg-nucleotides, which modulate K ATP channel activity (8 -12). It is currently believed that Kir6.2/SUR1 forms the beta cell K ATP channel, Kir6.2/SUR2A forms the cardiac K ATP channel, and Kir6.2/SUR2B forms the smooth muscle K ATP channel (13)(14)(15)(16)(17). It has recently been shown that long chain (LC) acyl-CoA esters are able to activate native K ATP channels in inside-out membrane patches excised from pancreatic beta cells (18,19). The intracellular concentrations of LC acyl-CoA esters are predicted to vary with the metabolic state of the cell. Because long-term exposure to nonesterified fatty acids increases cellular levels of LC acyl-CoA esters in the beta cell (18,20), this finding may have implications for the regulation of insulin secretion under conditions in which nonesterified fatty acids are increased, as in obese subjects (21,22). The elevated intracellular LC acyl-CoA ester concentration would be expected to activate K ATP channels, thereby hyperpolarizing the beta cell and inhibiting insulin secretion. It is possible, therefore, that increased K ATP channel activity, induced by LC acyl-CoA esters, may contribute to the impaired insulin secretion observed in obese non-insulin-dependent diabetics.
The mechanism of beta cell K ATP channel activation by oleoyl-CoA differs from that of the classical potassium channel opener diazoxide in several ways. In particular, the effects of oleoyl-CoA do not require Mg 2ϩ , nor are they abolished following mild proteolysis of the inner membrane surface (19). This suggests that oleoyl-CoA and diazoxide may act by different mechanisms. In this paper, we show that oleoyl-CoA stimulates the activity of the cloned beta cell K ATP channel (Kir6.2/SUR1). We further show this effect is mediated by interaction of the acyl-CoA ester with the Kir6.2 subunit of the channel and that it is associated with a reduced sensitivity to the inhibitory effects of ATP and a decreased rate of channel rundown. Activation by oleoyl-CoA is not observed for a related member of the inward rectifier K-channel family (Kir1.1a).  (12). For oocyte expression studies, Kir6.2 and SUR1 constructs were subcloned into the pBF expression vector that provides the 5Ј and 3Ј untranslated regions of the Xenopus ␤-globin gene. Synthesis of capped mRNA was carried out using the mMessage mMachine large scale in vitro transcription kit (Ambion, Austin, TX).
Oocyte Handling-Female Xenopus laevis were anaesthetized with MS222 (2 g/liter added to the water). One ovary was removed via a minilaparotomy, the incision was sutured, and the animal was allowed to recover. Once the wound had completely healed, the second ovary was removed in a similar operation, and the animal was then killed by decapitation while under anesthesia. Immature stage V-VI Xenopus oocytes were incubated for 75 min with 1.0 mg/ml collagenase (Boehringer Mannheim, type A) and manually defolliculated. Oocytes were injected with mRNA encoding Kir6.2⌬C36 (ϳ2 ng/oocyte) or Kir1.1a (ϳ0.2 ng/oocyte). For coexpression experiments, ϳ0.04 ng of Kir6.2 was coinjected with ϳ2 ng of SUR1 (giving a ratio of ϳ1:50). The final injection volume was ϳ50 nl/oocyte. Isolated oocytes were maintained in tissue culture and studied 1-4 days after injection (25).
Electrophysiology-Macroscopic currents were recorded from giant excised inside-out patches at a holding potential of 0 mV and at 20 -24°C (25). Patch electrodes were pulled from thick-walled borosilicate glass (GC150; Clark Electromedical Instruments) and had resistances of 250 -500 k⍀ when filled with pipette solution. Currents were evoked by repetitive 3-s voltage ramps from Ϫ110 mV to ϩ100 mV and recorded using an EPC7 patch-clamp amplifier (List Electronik, Darmstadt, Germany). They were filtered at 0.2 kHz, digitized at 0.5 kHz using a Digidata 1200 Interface, and analyzed using pClamp software (Axon Instruments, Burlingame, CA). The data was also stored on video tape (filter ϭ 10 kHz), and records for display in the figures were resampled at 20Hz. Single-channel currents were recorded from small inside-out membrane patches. They were filtered at 5 kHz using an 8-pole Bessel filter and sampled at 10 kHz.
The pipette solution contained (in mM) 140 KCl, 1.2 MgCl 2 , 2.6 CaCl 2 , 10 HEPES (pH 7.4 with KOH); and the internal (bath) solution contained (in mM) 110 KCl, 1.4 MgCl 2 , 30 KOH, 10 EGTA, 10 HEPES (pH 7.2 with KOH) and nucleotides as indicated. Oleoyl-CoA (ICN, Costa Mesa, CA) was made as a 1 mM stock solution in water and stored at Ϫ20°C. Solutions containing ATP were made up fresh each day, and the pH was readjusted after addition of the nucleotide. Bovine serum albumin (BSA) was dissolved directly in the bath solution. Rapid exchange of solutions was achieved by positioning the patch in the mouth of one of a series of adjacent inflow pipes placed in the bath. Bath solutions containing oleoyl-CoA were kept in a glass beaker, and the tubing was flushed with fresh solution before each application.
Data Analysis-The slope conductance was measured by fitting a straight line to the current-voltage relationship between Ϫ20 mV and Ϫ100 mV: the average of five consecutive ramps was calculated in each solution.
ATP dose-response relationships were measured by alternating each test solution with an ATP-free (control) solution, and the extent of inhibition was expressed as a fraction of the mean of the value obtained in the ATP-free solution before and after application of the test nucleotide. This procedure ensured that any change of channel activity in ATP-free solution (either because of rundown or reactivation) was easily detected and was corrected for. Experiments in which the control level of channel activity decreased by Ͼ25% during the course of measuring the ATP dose-response curve were discarded. ATP concentrations were applied in random order, and each solution was applied for 25-30 s (the time required to exchange the solution completely was Ͻ1 s). Because the stimulatory response to oleoyl-CoA was maintained following return of the patch to a solution lacking the acyl-CoA (unless BSA was added, see below), we actually measured the ATP dose-response curve immediately after return to oleoyl-CoA-free solution. Reapplication of fresh oleoyl-CoA after construction of the dose-response curve did not cause any further stimulation of channel activity.
Dose-response curves were fitted to the Hill equation, where G is the conductance in the presence of ATP, G c is the conductance in ATP-free solution, [ATP] is the ATP concentration, K i is the ATP concentration at which inhibition is half-maximal, and h is the slope factor (Hill coefficient).
Single channel currents were analyzed using a combination of pClamp and in-house software written by Dr. P. A. Smith (Oxford University). Single-channel current amplitudes were calculated from an all-points amplitude histogram. Channel activity (NP o ) was measured as the mean patch current (I) divided by the single-channel current amplitude (i), for segments of the current records of about 1 min duration. The number of channels in the patch (N) and the open probability (P o ) were estimated by binomial analysis. For analysis of the single-channel kinetics, events were detected using a 50% threshold level method.
All data are given as mean Ϯ S.E. The symbols in the figures indicate the mean and the vertical bars equal 1 S.E. (where this is larger than the symbol). Statistical significance was tested using Student's t test.

Effect of Oleoyl-CoA on Wild-type Kir6.2-SUR1 Currents-
Wild-type K ATP currents were measured in giant macropatches from oocytes coinjected with mRNAs encoding Kir6.2 and SUR1. In cell-attached patches, the Kir6.2/SUR1 conductance was very small and similar in magnitude to that of uninjected oocytes. A marked increase in current was observed, however, following excision of the patch into a nucleotide-free solution as reported previously (25). At Ϫ100 mV, the mean current amplitude in the inside-out patch immediately after patch excision was Ϫ3.7 Ϯ 1.0 nA (n ϭ 11). Subsequently, the current magnitude slowly declined, but it could be restored following exposure to 1 mM MgATP (Fig. 1A, panel a). A similar rundown and MgATP-dependent refreshment of channel activity has also been observed for native beta cell K ATP channels following patch excision (26,27).
Application of 1 M oleoyl-CoA to the intracellular surface of excised patches, caused an increase in conductance, which started after a delay of 10 -30 s and reached a peak 1-2 min later (Fig. 1B, panel a). The mean activation of Kir6.2/SUR1 currents by 1 M oleoyl-CoA was 2.4 Ϯ 0.3-fold (n ϭ 10; p ϭ 0.0005). The response to oleoyl-CoA was apparent at membrane potentials from Ϫ110 mV to ϩ100 mV, demonstrating that the effect is not voltage-dependent. Following removal of the acyl-CoA, the patch conductance remained high for more than 5 min unless 0.1% BSA was added to the solution, in which case the effect of oleoyl-CoA was rapidly reversed (Fig.  1B, panel a). BSA did not affect Kir6.2/SUR1 currents when applied to the patch prior to activation with oleoyl-CoA.
No effect of 1 M oleoyl-CoA was observed in patches excised from uninjected oocytes (n ϭ 4). This indicates that the current activated by oleoyl-CoA results from activation of a recombinant K ATP channel rather than a current endogenous to the oocyte.
Effects of Oleoyl-CoA on Kir6.2⌬C36 Currents-Previous studies have suggested that the mechanism of activation of native beta cell K ATP channels by oleoyl-CoA differs from that produced by MgADP or diazoxide; unlike diazoxide and MgADP, oleoyl-CoA is effective in the absence of Mg 2ϩ ions or following mild trypsinization of the intracellular membrane surface (19). Channel activation by MgADP or diazoxide is mediated by the SUR1 subunit of the K ATP channel complex (8 -10). The different behavior of oleoyl-CoA might be explained, therefore, if this compound were to interact with the Kir6.2 subunit.
To investigate this possibility, we studied the effect of oleoyl-CoA on a truncated form of Kir6.2 (Kir6.2⌬C36), which expresses functional channels in the absence of the sulfonylurea receptor subunit (12). Macroscopic currents in cell-attached patches on oocytes expressing Kir6.2⌬C36 were small in the cell-attached configuration, increased on patch excision into nucleotide-free solution, and subsequently slowly ran down (Fig. 1A, panel b). As reported previously (12), Kir6.2⌬C36 currents were inhibited by ATP and were refreshed following ATP removal. Application of 1 M oleoyl-CoA to the inner membrane surface enhanced Kir6.2⌬C36 currents by 2.6 Ϯ 0.6-fold (n ϭ 9; p ϭ 0.03), with a time course similar to that observed for Kir6.2/SUR1 currents (Fig. 1B, panel b). As observed for Kir6.2/SUR1 currents, this effect was maintained on return to control solution but reversed following addition of 0.1% BSA. These results suggest that K ATP channel activation by oleoyl-CoA is mediated either by a direct interaction with the Kir6.2 subunit or, possibly, via a third subunit endogenously expressed in Xenopus oocytes. In the latter case, a similar protein must also be expressed in pancreatic beta cells because native K ATP channels are also sensitive to oleoyl-CoA.
Interaction between Oleoyl-CoA and ATP-Since inhibition of K ATP channels by ATP results also from an interaction with the Kir6.2 subunit, we explored whether the inhibitory effect of the nucleotide is modified by oleoyl-CoA. We compared the ATP-sensitivity of Kir6.2⌬C36 currents measured in the absence of oleoyl-CoA with that obtained for patches in which Kir6.2⌬C36 channels were first activated by pretreatment with the acyl-CoA at 1 M. Fig. 2 shows that the inhibitory effect of ATP on both Kir6.2/SUR1 and Kir6.2⌬C36 currents was impaired following exposure to oleoyl-CoA. Half-maximal inhibition (K i ) of Kir6.2⌬C36 currents increased from a mean value of 130 Ϯ 8 M (n ϭ 7) in control solution to 381 Ϯ 37 M (n ϭ 6; p ϭ 0.0001) after application of 1 M oleoyl-CoA. The Hill coefficient was unaffected, being 1.0 Ϯ 0.1 and 1.2 Ϯ 0.1 in the absence and presence of oleoyl-CoA, respectively. Similar results were obtained for wild-type K ATP channels (Kir6.2/SUR1), where the mean K i for ATP increased from 31 Ϯ 5 M (n ϭ 9) to 76 Ϯ 17 M (n ϭ 8; p ϭ 0.0005). The Hill coefficient was 1.1 Ϯ 0.1 before and 1.0 Ϯ 0.1 after addition of oleoyl-CoA. A reduction in ATP sensitivity in the presence of oleoyl-CoA has also been reported for native K ATP channels although the effect was not quantified (18).
Effect of Oleoyl-CoA on the Single-channel Kinetics-To investigate further the mechanism by which oleoyl-CoA potentiates the macroscopic current, we recorded single-channel currents from oocytes expressing either Kir6.2⌬C36 or wild-type K ATP channels (Kir6.2/SUR1). Application of the acyl-CoA had no effect on the single-channel current amplitude at Ϫ60 mV (Fig. 3), which was 4.  (Table I).
The value of P o measured for Kir6.2/SUR1 currents in control solution (0.2; Table I) was similar to that previously reported for the recombinant channel (0.11; Ref. 12) and for the native K ATP channel in beta cells (0.1; Ref. 28). Likewise, the mean open and closed times were close to those found for recombinant and native K ATP channels (29,30).
A lack of effect of oleoyl-CoA on the single-channel current amplitude has also been reported for the native beta cell K ATP channel (19). Interestingly, however, the effect of oleoyl-CoA on the single-channel kinetics of native K ATP channels was different: a 2-fold increase in o was reported at 0 mV. It is possible that this reflects the different membrane potentials at which the experiments were carried out. We conducted our experiments at Ϫ60 mV because this is close to the resting potential of the pancreatic beta cell (2).

FIG. 2. Mean ATP dose-response relationships for Kir6.2/SUR1 currents (left) or Kir6.2⌬C36 currents (right) in the absence (q) or presence (E) of 1
M oleoyl-CoA. Test solutions were alternated with control solutions, and the slope conductance (G) is expressed as a fraction of the mean (G c ) of that obtained in control solution before and after exposure to ATP. The number of patches was (left) q ϭ 9, E ϭ 8; and (right) f ϭ 5, Ⅺ ϭ 6. The lines are the best fit of the data to the Hill equation using the mean values for K i and h given in the text. whether oleoyl-CoA is able to inhibit Kir1.1a, a related Kir channel that shares ϳ45% sequence identity with Kir6.2. As shown in Fig. 4, 1 M oleoyl-CoA did not cause a significant activation of Kir1.1a. Rather, the acyl-CoA induced a slow rundown of Kir1.1a currents. This is in marked contrast to the lack of rundown found for Kir6.2⌬C36 currents after exposure to oleoyl-CoA. The mean macroscopic conductance in the presence of the drug was 30.3 Ϯ 6.2% (n ϭ 4) of that in control solution after 3 min of exposure to 1 M oleoyl-CoA. DISCUSSION We report here that oleoyl-CoA activates the cloned beta cell K ATP channel (Kir6.2/SUR1) expressed in Xenopus oocytes. This suggests that the acyl-CoA interacts directly with the K ATP channel, or a channel regulator that is expressed endogenously in the Xenopus oocyte, rather than with a beta cellspecific membrane component.
Subunit Interaction-We found that oleoyl-CoA activated both Kir6.2/SUR1 channels and Kir6.2⌬C36 channels expressed in the absence of SUR1. This result indicates that the mechanism of acyl-CoA activation does not require the presence of the sulfonylurea receptor subunit and is thus very different from that of other K ATP channel activators, such as MgADP and diazoxide, which mediate their effects via SUR1 (8 -12). Our results provide an explanation for the findings of studies on the native beta cell K ATP channel which have shown that the response to oleoyl-CoA was both Mg 2ϩ -independent and unimpaired by trypsinization of the inner membrane surface (19), conditions that abolish the potentiatory action of ADP and diazoxide (28,31,32). Oleoyl-CoA is the first compound to be shown to mediate a stimulatory effect on K ATP channel activity via the Kir6.2 subunit.
Although we show that SUR1 is not required for K ATP channel activation by oleoyl-CoA, our results do not allow us to conclude that the compound interacts directly with the Kir6.2 subunit itself. It is possible, for example, that oleoyl-CoA interacts with the lipid membrane, or with a third protein endogenously present in both Xenopus oocytes and pancreatic beta cells that influences K ATP channel activity. Direct binding of oleoyl-CoA to other proteins has been demonstrated, including those within the metabolic pathway such as the ATP/ADP translocase and the uncoupling protein (33,34) and with transcription factors (35). Binding studies are therefore now required to determine whether oleoyl-CoA directly binds to Kir6.2.
Altered ATP Sensitivity-Our results demonstrate that oleoyl-CoA reduces the sensitivity of Kir6.2/SUR1 and Kir6.2⌬C36 channels to ATP. There are a number of possible explanations for this finding. First, the acyl-CoA might compete with ATP for its binding site. Second, oleoyl-CoA might allosterically affect the conformation, and thereby the affinity, of the ATP-binding site. Third, oleoyl-CoA might indirectly alter the apparent ATP sensitivity by affecting the singlechannel kinetics; for example, if the ATP-inhibited state were only accessible when the channel were closed, an increase in the open probability would necessarily decrease the channel ATP sensitivity (36,37). Because oleoyl-CoA does in fact alter the single-channel kinetics and shifts the gating of the channel toward the open state, the latter possibility may contribute, at least in part, to the reduced ATP-sensitivity.
Other K-channels and Lipid Effectors-The anionic phospholipid PIP 2 interacts with native cardiac and beta cell K ATP channels to enhance their activity (38,39). Activation of Kir6.2/ SUR1 currents by PIP 2 has also been reported (39). However, the properties of PIP 2 activation are not identical to those of oleoyl-CoA, because the effect of PIP 2 was instantaneous and the compound also activated Kir1.1a (39). This makes it less likely that the two compounds interact with the channel in exactly the same way.
In contrast to the native and cloned beta cell K ATP channel, the mitochondrial K ATP channel is inhibited by oleoyl-CoA (K i , 80 nM; Ref. 40). The molecular identity of this channel has not been determined, but our data suggest either that the poreforming subunit is unlikely to be Kir6.2 or that an additional inhibitory effect of oleoyl-CoA is mediated through a second subunit.
Physiological Implications-Our results indicate that oleoyl-CoA interacts with the Kir6.2, rather than the SUR1, subunit of the K ATP channel. Kir6.2 is expressed in a number of tissues, including heart, skeletal muscle, and certain brain regions (13,15,41). It is believed to serve as the pore-forming subunit in all these tissues. Our data therefore suggest that long chain acyl-CoA esters may also influence K ATP channel function, and thereby electrical activity, in tissues other than the beta cell.
The increased activity, and reduced sensitivity of Kir6.2/ SUR1 and Kir6.2⌬C36 to ATP, observed in the presence of oleoyl-CoA may have important physiological implications. Long-term exposure to elevated levels of circulating free fatty acids occurs in diabetes and obesity and has also been shown to increase the intracellular levels of long-chain acyl CoA esters (18,20). Our results suggest that this both enhances the activity of the K ATP channel and renders it less sensitive to ATP, thereby producing membrane hyperpolarization and decreasing glucose-induced insulin secretion. This might contribute to  the reduced sensitivity of the pancreatic beta cell to glucose observed in non-insulin-dependent diabetes mellitus.