Regulation of the ATP-sensitive Potassium Channel Subunit, Kir6.2, by a Ca2+-dependent Protein Kinase C*

The activity of ATP-sensitive potassium (KATP) channels is governed by the concentration of intracellular ATP and ADP and is thus responsive to the metabolic status of the cell. Phosphorylation of KATP channels by protein kinase A (PKA) or protein kinase C (PKC) results in the modulation of channel activity and is particularly important in regulating smooth muscle tone. At the molecular level the smooth muscle channel is composed of a sulfonylurea subunit (SUR2B) and a pore-forming subunit Kir6.1 and/or Kir6.2. Previously, Kir6.1/SUR2B channels have been shown to be inhibited by PKC, and Kir6.2/SUR2B channels have been shown to be activated or have no response to PKC. In this study we have examined the modulation of channel complexes formed of the inward rectifier subunit, Kir6.2, and the sulfonylurea subunit, SUR2B. Using a combination of biochemical and electrophysiological techniques we show that this complex can be inhibited by protein kinase C in a Ca2+-dependent manner and that this inhibition is likely to be as a result of internalization. We identify a residue in the distal C terminus of Kir6.2 (Ser-372) whose phosphorylation leads to down-regulation of the channel complex. This inhibitory effect is distinct from activation which is seen with low levels of channel activity.

ATP-sensitive potassium (K ATP ) 2 channels couple cell metabolism to membrane potential in many cell types and control insulin release, vascular smooth muscle tone, and excitability in neurons and muscle (1). Inhibition by ATP and activation by nucleotide diphosphates allow the metabolic state of the cell to control membrane potential and cell excitability. Additionally, activation of K ATP channels as a result of metabolic stress such as ischemia and hypoxia has been shown to protect muscle, heart, and brain (2). K ATP channels are composed of an octomeric complex consisting of four pore-forming subunits of the Kir6.x subfamily of inwardly rectifying potassium channels (Kir6.1 or Kir6.2) and four regulatory sulfonylurea receptor subunits (SUR1, SUR2A, or SUR2B), belonging to the ATP-binding cassette superfamily of proteins (1,3,4). The assembly of one pore-forming subunit (Kir6.1 or Kir6.2) with a particular SUR generates currents with distinct nucleotide sensitivities and pharmacological properties (4,5). For example, the pancreatic ␤-cell K ATP channel is composed of Kir6.2 and SUR1 (6, 7), Kir6.1 and SUR2B are thought to form the vascular smooth muscle K ATP channel (8 -10), Kir6.2 and SUR2B are present in nonvascular smooth muscle and portal vein (11)(12)(13)(14)(15), and Kir6.2 and SUR2A comprise the cardiac K ATP channel (16). Characteristically, the vascular smooth muscle K ATP channel has a lower single-channel conductance and has an absolute dependence on nucleotide diphosphates for activity (8). In contrast, in portal vein and colonic smooth muscle cells the single-channel conductance and nucleotide regulation are more compatible with a channel complex containing Kir6.2 (11,15,17).
Regulation of proteins by intracellular signals through protein kinase-mediated phosphorylation is an important mechanism by which the activity of many ion channels, including K ATP channels, can be modulated (18,19). Phosphorylation of ion channels is most commonly catalyzed by protein kinase A (PKA) and protein kinase C (PKC), where phosphorylation of serine or threonine residues leads to an alteration in channel properties by modifying kinetics and/or the number of channels at the membrane (20). Regulation of K ATP channels in the vasculature by protein kinases is of particular importance as vasoconstrictors, such as angiotensin II, for example, modulate channel activity by activating PKC (21). A number of studies have shown that the K ATP channel composed of Kir6.1/SUR2B is inhibited by PKC (11,(22)(23)(24). A short motif containing specific sites for PKC phosphorylation has been found on the poreforming Kir6.1 subunit (23). There are two main subfamilies of PKC enzymes regulated by G q/11 -coupled receptors: conventional and novel. Conventional PKCs require Ca 2ϩ and diacylglycerol to function, whereas novel PKCs require diacylglycerol but are Ca 2ϩ -independent. Several groups have demonstrated the role of a Ca 2ϩ -independent isoform of PKC, PKC⑀, in the regulation of Kir6.1 (22,25,26). PKC has also been shown to regulate K ATP channels containing the Kir6.2 subunit (24,(27)(28)(29)(30). Interestingly, PKC has a dual effect on Kir6.2, both up-regulating and down-regulating channel activity by phosphorylating threonine 180 (29). In addition, trafficking studies have revealed that PKC initiates internalization of the channel complex leading to decreased channel activity (28,30). To date Thr-180 is the only specific PKC site to be identified on the Kir6.2 subunit, and the mechanism behind its phosphorylation by PKC has not been investigated. In this study, we identify a new site for PKC-mediated phosphorylation leading to channel inhibition and additionally show Ca 2ϩ dependence of PKC phosphorylation at this site.

EXPERIMENTAL PROCEDURES
Molecular Biology and Cell Culture-Single-point mutations were introduced into the mouse Kir6.2 clone and the maltosebinding protein (MBP)-Kir6.2C vector by PCR using the Stratagene XL Mutagenesis kit according to the manufacturer's instructions. Human embryonic kidney (HEK) 293 cells stably transfected with Kir6.1/SUR2B and Kir6.2/SUR2B were maintained in G418 and Zeocin selective media as described previously (9,31). Transfection of Kir6.2⌬C26 or the mutant Kir6.2 constructs with SUR2B was carried out using FuGENE HD (Roche Diagnostics) as per the manufacturer's instructions. GFP (100 ng) was co-transfected to enable transfection success and efficiency to be assessed, and cells were patched 48 h after transfection.
MBP Fusion Proteins and in Vitro Phosphorylation Assays-Expression and purification of the wild-type and mutant MBP-Kir6.2C were carried out as described previously (22). The protein was judged to be approximately 90% pure based on SDS-PAGE analysis. In vitro phosphorylation assays using the wild-type and mutant MBP-Kir6.2C were carried out using the method described previously (22,32). Briefly, 1 mg of protein was bound to 100 l of amylose resin (New England Biolabs). After subsequent washes with a HEPES wash buffer the amylose-bound protein was incubated in a mix containing [␥-32 P]ATP (2 Ci) in the presence or absence of the catalytic subunit of PKC (20 ng to give a final concentration of 8 nM; Calbiochem) for 2 h at 37°C. The sample was washed a further five times to remove excess radioactivity before subjecting it to SDS-PAGE (10%) and autoradiography. Determination of the band intensity was carried out using ImageJ (National Institutes of Health), and statistical analysis using an unpaired Student's t test was carried out using Prism v4 (GraphPad).
Phosphopeptide Mapping of PKC-phosphorylated Residues-The MBP-Kir6.2C protein was phosphorylated as above with the following alterations. [␥-32 P]ATP was omitted from the assay, and the following wash buffer was used (50 mM HEPES, pH 7.4, 1 mM EDTA, 10% glycerol, 5 mM sodium pyrophosphate, 50 mM sodium fluoride, 1 mM sodium vanadate, 10 mM glycerol phosphate, 1 mM DTT, 1% Triton X-100). After incubation with PKC and further washing the sample was resuspended in 2 ϫ SDS loading buffer before denaturation for 4 min at 85°C. Samples were loaded onto a 10% Novex gel (Invitrogen) and subsequently stained with Coomassie Blue. The relevant bands were excised and subjected to phosphopeptide mapping at the Fingerprints Proteomics Facility, University of Dundee.
In Cell Western Assay-The In Cell Western assay was carried out as described by Meimaridou et al. (33). In brief, Chinese hamster ovary-K1 (CHO-K1) cells were seeded into 24-well dishes and transfected with Kir6.2-HA/SUR2B using Turbofect (Fermentas). After 48 h, cells were washed with PBS and incubated with DMSO, 1 M PdBu, or 1 M 4␣-phorbol for 30 min at 37°C. Cells were then fixed with 3.7% formaldehyde for 20 min before half of the wells were permeabilized with 0.1% Triton X-100. The other wells were left unpermeabilized to measure the amount of protein expressed at the cell surface. Cells were blocked with PBS ϩ 5% nonfat milk solution for 1.5 h at room temperature followed by overnight incubation (4°C) with anti-HA antibody (1:500 dilution; Sigma) in blocking solution. Cells were washed with PBS before incubation with IRDye 800CW goat anti-mouse secondary antibody for 1 h. After washing with PBS the fluorescent signal was measured using a LI-COR Odyssey plate reader. Data are presented as cell surface protein over total protein, normalized to DMSO (control).
Cell Surface Biotinylation-Surface expression of Kir6.2 in the presence and absence of PdBu was investigated using the Pierce Cell Surface Protein Isolation kit (Thermo Scientific). In brief, CHO-K1 cells were seeded into 10-cm dishes and transfected with Kir6.2-HA or Kir6.2-HA-S372A/SUR2B, and enhanced GFP for analysis of transfection efficiency. 48 h after transfection cells were washed twice with ice-cold PBS supplemented with 1 mM CaCl 2 and 1 mM MgCl 2 . Cells were then incubated with 1 M PdBu or DMSO for 30 min at 37°C. Following this, cells were incubated with PBS containing EZ-Link Sulfo-NHS-SS-Biotin for 2 h at 4°C. Cells were again washed with PBS and the reaction quenched with Pierce Quenching Solution as supplied. Cells were harvested and resuspended in TDET-S buffer (1% Triton X-100, 0.4% deoxycholic acid, 5 mM EDTA, 25 mM Tris, pH 7.4, 150 mM NaCl, 10% glycerol) containing protease inhibitors (Complete Mini EDTA-free; Roche Applied Science) and phosphatase inhibitors (PhosStop; Roche Applied Science). The resuspended cells were sonicated (5 ϫ 1 s) and left for 30 min on ice with periodic vortexing every 5 min. The lysate was obtained by centrifuging at 10,000 ϫ g for 2 min, and this was then incubated with avidin overnight at 4°C to allow the biotinylated protein to bind. Following several washes with TDET-S buffer, labeled protein was eluted into 2 ϫ SDS loading buffer ϩ DTT at 95°C for 5 min. Western blotting was used to detect labeled protein using a rat monoclonal anti-HA antibody (Roche Applied Science) or a rabbit anticalnexin antibody (Santa Cruz Biotechnology) and the ECL detection kit (GE Healthcare).
Rubidium Flux-An 86 Rb ϩ assay was used to determine the amount of flux passing through K ATP channels in response to various drugs as described previously (34,35). HEK293 cells stably expressing either Kir6.1/SUR2B or Kir6.2/SUR2B were used. The cells (in 6-well dishes) were incubated for 24 h with 86 RbCl (0.037 MBq/ml) before being washed three times with HBS assay medium (10 mM HEPES, pH 7.4, 10 mM glucose, 130 mM NaCl, 7 mM KCl, 2 mM CaCl 2 , and 1 mM MgCl 2 ). Cells were then preincubated with 2 ml of HBS Ϯ PKC inhibitors, 1 M staurosporine or 3 M GF109203X, or the calcium chelator, 20 nM BAPTA-AM for 15 min at 25°C. PdBu or 4␣-phorbol was then added to each well to give a final concentration of 1 M (concentration commonly used to activate PKC), and the cells were incubated for 5 min at 37°C before channel activators/ inhibitors were added as follows (DMSO, 10 M pinacidil, 10 M pinacidil ϩ 10 M glibenclamide). After 15 min further incubation at 25°C, the supernatant was aspirated into vials, and the cells were lysed using HBS ϩ 2% Triton X-100 solution and collected. All vials were assayed for 86 Rb content by measurement of Cherenkov radiation in a liquid scintillation counter (TriCarb, Packard 2000CA). Efflux was expressed as a percentage relative to the total amount of radioactivity incorporated. One-way ANOVA followed by a Bonferroni test FEBRUARY 24, 2012 • VOLUME 287 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 6197 was used to determine statistical significance as indicated in the text.

Kir6.2 Subunit Regulation by PKC
Electrophysiology-Whole cell patch clamp recordings were performed as described previously (22). Capacitance transients and series resistance in whole cell recordings were compensated electronically by using amplifier circuitry (Axopatch 200B). Data were filtered at 1 kHz using the filter provided with the Axopatch 200B (four-pole Bessel) and sampled at 5 kHz using a Digidata 1440 (Axon Instruments). Currents were acquired and analyzed using pClamp 10 (Axon Instruments). The intracellular (pipette) solution contained 140 mM KCl, 1.2 mM MgCl 2 , 1 mM CaCl 2 , 10 mM EGTA, and 5 mM HEPES, 1 mM MgATP, and 0.5 mM NaUDP, pH 7.2, using KOH. The free Ca 2ϩ concentration was estimated using Webmaxc standard software. An additional 5.57 mM CaCl 2 was added to the standard pipette solution to give a higher free intracellular Ca 2ϩ concentration of ϳ300 nM. The bath solution contained 140 mM KCl, 2.6 mM CaCl 2 , 1.2 mM MgCl 2 and 5 mM HEPES, pH 7.4. For inside-out recordings the bath and pipette solutions contained 140 mM KCl, 1.4 mM MgCl 2 , 1 mM EGTA, 10 mM HEPES, pH 7.2. Pipette resistances were between 2 and 4 megohms for whole cell recordings and 5 and 8 megohms for single-channel recordings. All chemicals were obtained from Sigma-Aldrich. Agents were applied to the bath using a gravity-driven perfusion system.
Data Analysis-To analyze the rate of current inhibition, time course data were fitted using the following equation (in Origin 6), where x is the time at any given point, x 0 is the time at no inhibition, and k is the time constant. A 1 is the initial Y value, and A 2 is the final Y value. Statistical analysis was carried out using one-way ANOVA with a Bonferroni post hoc test or a paired Student's t test as appropriate. Statistical significance is indicated in the figure legends. Data are presented as mean Ϯ S.E. Where appropriate, current density is presented normalized to the control. PdBu ϩ pinacidil, 47.19 Ϯ 2.63%; PdBu ϩ pinacidil ϩ glibenclamide, 23.77 Ϯ 5.51%, n ϭ 9, p Ͻ 0.01 compared with pinacidil) (Fig. 1B). The inactive phorbol ester 4␣-phorbol was used to rule out nonspecific phorbol effects. Activation of flux by pinacidil through both Kir6.1/SUR2B and Kir6.2/SUR2B was unaffected by 1 M 4␣-phorbol. To ascertain whether the PdBu-induced inhibition of pinacidil-activated flux is due to PKC, we used PKC inhibitors in the presence of PdBu. Pinacidil-induced flux through both Kir6.1/SUR2B and Kir6.2/ SUR2B was restored to pre-PdBu levels in the presence of the PKC inhibitors, 1 M staurosporine and 3 M GF109203X, suggesting that inhibition of flux through both channels was due to activation of PKC ( Fig. 1). Additionally, we tested whether there was a Ca 2ϩ dependence of the PKC effect on the flux through either channel. To this end, the cells were preincubated in 20 nM BAPTA-AM, a Ca 2ϩ chelator. In the case of Kir6.1/SUR2B, the presence of BAPTA-AM had no effect on the inhibition of flux by PdBu (BAPTA-AM ϩ PdBu, 17.81 Ϯ 0.63%; BAPTA-AM ϩ PdBu ϩ pinacidil, 19.03 Ϯ 0.44%, n ϭ 9, p Ͼ 0.05, Fig. 1A). However, in cells expressing Kir6.2/SUR2B, BAPTA-AM prevented the inhibition previously seen in the
PdBu Inhibits Whole Cell Kir6.2/SUR2B Currents with High Intracellular Ca 2ϩ Concentrations-In our previous studies we failed to see the inhibition of the Kir6.2/SUR2B channel current upon PKC activation using the whole cell configuration of the patch clamp (22). However, because the flux data described above showed that PKC does inhibit Kir6.2/SUR2B, possibly via a Ca 2ϩ -dependent mechanism, we investigated whether the lack of PKC inhibition seen previously may have been due to the low free Ca 2ϩ concentration present in the pipette. The standard whole cell pipette solution normally used for such recordings had a free intracellular Ca 2ϩ concentration of ϳ20 nM. To test the Ca 2ϩ dependence of the PKC effect on Kir6.2/SUR2B we subjected stably transfected HEK 293 cells to whole cell patch clamping using concentrations of free intracellular Ca 2ϩ in the pipette ranging from 20 to 300 nM. Current was elicited by a repetitive 1-s voltage ramp from Ϫ100 mV to ϩ100 mV every 15 s from a holding potential of 0 mV. Whole cell currents evoked through Kir6.2/SUR2B channels by 10 M pinacidil at Ϫ50 mV (Ϫ193.48 Ϯ 6.15 pA/pF, n ϭ 5) were not significantly affected by the application of 1 M PdBu (Ϫ188.23 Ϯ 12.05 pA/pF, n ϭ 5, p Ͼ 0.05) when the intracellular Ca 2ϩ concentration was buffered to 20 nM (Fig. 2, A and E). 10 M glibenclamide (Ϫ20.47 Ϯ 12.71 pA/pF, n ϭ 5) completely inhibited the pinacidil-evoked current. Increasing the free intracellular Ca 2ϩ to 50 nM did not lead to a significant PdBu-induced inhibition of KATP currents (pinacidil, Ϫ160.5 pA/pF Ϯ 41.21; PdBu, Ϫ156.06 Ϯ 43.7 pA/pF, n ϭ 6, p Ͻ 0.05, Fig. 2, B and E). In contrast, when the intracellular solution was buffered to ϳ100 nM free Ca 2ϩ , PdBu significantly inhibited the pinacidil-evoked current (pinacidil, Ϫ240.07 Ϯ 17.62 pA/pF; PdBu, Ϫ152.8 Ϯ 12.24 pA/pF, n ϭ 5, p Ͻ 0.01) through Kir6.2/SUR2B channels (Fig. 2, C and E). The PdBu-induced inhibition was not reversed upon washing with pinacidil (wash, Ϫ147.57 Ϯ 12.38 pA/pF, n ϭ 5, p Ͼ 0.05 compared with PdBu). Glibenclamide further reduced the current to control levels (control, Ϫ44.02 Ϯ 3.66 pA/pF; glibenclamide, Ϫ59.64 Ϯ 3.22, n ϭ 5, p Ͼ 0.05 compared with control). When the concentration of free Ca 2ϩ was further increased to ϳ300 nM, PdBu inhibited the current (pinacidil, Ϫ117.23 pA/pF Ϯ 29.8; PdBu, Ϫ37.61 Ϯ 15.83 pA/pF, n ϭ 12, p Ͻ 0.001) through Kir6.2/SUR2B channels (Fig.  2, D and E). Glibenclamide further reduced the current to control levels (control, Ϫ18.32 Ϯ 9.14 pA/pF; glibenclamide, Ϫ19.49 Ϯ 5.50, n ϭ 12, p Ͼ 0.05 compared with control). PdBu inhibited ϳ50% (50.46 Ϯ 10.89%, n ϭ 12) of the glibenclamidesensitive current in high intracellular Ca 2ϩ (300 nM) compared with Ͻ1% (0.27 Ϯ 1.28%, n ϭ 5) inhibition by PdBu when the Ca 2ϩ concentration was low (20 nM) (Fig. 2F). To confirm that the inhibition observed with PdBu was due to PKC as suggested by the flux data, we incubated cells containing Kir6.2/SUR2B channels with 1 M staurosporine, a protein kinase inhibitor, and subjected them to whole cell patch clamp. Fig. 2 (Fig. 3). This indicates that a nonspecific effect of phorbol ester treatment was not responsible for the inhibition of the current in the presence of PdBu.
Kir6.2⌬C26 Truncation Mutant-We next investigated the location of the potential site of action of PKC on Kir6.2 by using a truncated mutant of Kir6.2 where the last 26 amino acids have been deleted. This mutant is functional with or without SUR (36). In this case, we transfected Kir6.2⌬C26 with SUR2B to allow better expression and also to allow regulation by the K ATP channel opener pinacidil and blocker glibenclamide. The proteomic data revealed several potential sites and one residue in particular, a serine at position 372 of Kir6.2 was found to be modified by a phosphate group and therefore a potential phosphorylation site for PKC (Fig. 5A).
To determine the effect of the proposed PKC phosphorylation of the Ser-372 residue, it was mutated to an alanine residue within the MBP-Kir6.2C protein. Both the wild-type and the S372A proteins were then subjected to in vitro phosphorylation using [␥-32 P]ATP as the phosphate substrate. In the presence of PKC, MBP-Kir6.2C is phosphorylated (Fig. 5Bii). MBP-Kir6.2C Ser-372 also shows some phosphorylation; however, this is considerably less than that of the wild-type protein (Fig. 5Bii). The amount of protein loaded on the gel is similar after staining with Coomassie Blue (Fig. 5Bi), indicating that the reduction in phosphorylation is most likely due to the presence of the S372A mutation. Densitometric analysis of the intensity of the autoradiography bands clearly shows a significant reduction of ϳ75% in the amount of phosphorylation of the mutant protein compared with the wild type (Fig. 5C). It would therefore appear that the Ser-372 residue is one of the main phosphorylation sites in the Kir6.2 C terminus.
PdBu Reduces Surface Expression of Kir6.2-To investigate whether phosphorylation of Kir6.2 also affects the amount of protein at the cell surface we used an In Cell Western assay. We used CHO-K1 cells because we have found in HEK293 cells that there is some leak, and "control" intracellular proteins are often labeled by the apparent surface treatment. This assay uses an anti-HA antibody to look at cell surface expression of proteins (Fig. 7A). In the presence of 1 M PdBu, there was a significant reduction in surface expression of Kir6.2 (68% reduced compared with control, p Ͻ 0.01, Fig. 7, A and B). When the inactive phorbol, 4␣-phorbol, was added, a small reduction in surface expression occurred (13%, not significant, Fig. 7, A and B), indicating that PdBu-induced activation of PKC reduces the amount of channel present at the surface. This was further confirmed using cell surface biotinylation (Fig. 7, C and D). Kir6.2 can be detected in both the total protein and in the cell lysates in the presence and absence of PdBu (Fig. 7C). In the absence of PdBu treatment, Kir6.2 was visible in the eluted sample, sug-gesting that it is present at the cell surface (Fig. 7C). In the presence of PdBu, there was a clear reduction in the amount of protein detected in the eluted sample, signifying that the protein is no longer at the cell surface (Fig. 7C). When the Kir6.2 S372A mutant was investigated, no apparent change was visible between the surface-labeled protein in the presence and absence of PdBu (Fig. 7D). Calnexin was used as a control intracellular protein and was only found in the cell lysate samples and not in the surface-labeled samples (Fig. 7E).
Phosphomimetic of Ser-372 Does Not Alter ATP Sensitivity of Kir6.2/SUR2B-Earlier studies have shown that phosphorylation of Kir6.2 alters the sensitivity of the channel to ATP (29,37). To investigate whether there is a correlation between ATP sensitivity and potential PKC phosphorylation of Ser-372, we tested the relative ATP sensitivity of the WT (Kir6.2/SUR2B) channel and the phosphomimetic S372E mutant channel using the inside-out patch clamp configuration (Fig. 9, A and B). Under these experimental conditions the patches contained large numbers of single channels, and a mean current was measured to assess channel activity. Both channels exhibited similar levels of sensitivity to varying concentrations of internal ATP, with IC 50 values of ϳ57 M for WT and 55 M for the S372E mutant. The S372E mutation had no effect on the Hill coefficient for the ATP dose-response curve of Kir6.2 (1.52 and 1.5, respectively) (Fig. 9C). The data suggest that in the case of  FEBRUARY 24, 2012 • VOLUME 287 • NUMBER 9

JOURNAL OF BIOLOGICAL CHEMISTRY 6203
Ser-372, a phosphomimetic of PKC phosphorylation does not change ATP sensitivity of the Kir6.2/SUR2B K ATP channel.

DISCUSSION
Activation of K ATP channels in the sarcolemma and mitochondria protects against the metabolic insult of ischemia in a number of tissues but particularly the heart (2, 38 -40). Furthermore, K ATP channel function can be regulated by cell surface receptors and by intracellular signaling molecules such as PKC (21,41). PKC has been shown to act on several isoforms of K ATP channels in different ways (11, 22-24, 28, 29). In particular, the sites and mechanisms responsible for PKC regulation of the predominant vascular smooth muscle channel subtype, Kir6.1/SUR2B, are well defined (11,22,23,42,43). PKC activation leads to inhibition of the channel, and this effect is thought to be a result of direct phosphorylation of Kir6.1 (22)(23)(24). The regulation of K ATP channels containing the Kir6.2 subunit is not as clear. PKC has been shown to both activate and inhibit the Kir6.2/SUR1 and Kir6.2/SUR2A channels depending on the intracellular ATP concentration (29). When expressed with SUR2B, the Kir6.2 channel is activated by PKC or shows no response, depending on the recording conditions (22)(23)(24). Single-channel inside-out patch recordings have shown that PKC has a dual action on K ATP channels containing the Kir6.2 subunit (co-expressed with either SUR1 or SUR2A), activating when ATP is high and inhibiting when ATP is low (29). When Kir6.2 is co-expressed with SUR2B, inside-out patch recordings in relatively high ATP (0.5 mM) showed that PKC activation with PdBu activates the channel (24). Conversely, we and others have shown, using whole cell patch clamp studies, that PKC has no effect on recombinant Kir6.2/SUR2B currents (22,23).
Our 86 Rb ϩ efflux studies on Kir6.1/SUR2B and Kir6.2/ SUR2B showed that pinacidil-induced efflux through both channels was reduced in the presence of PdBu, and this effect was reversed by the PKC inhibitors staurosporine and GF109203X (Fig. 1). The apparent inhibition of flux through Kir6.2/SUR2B was surprising in our hands, as we had not observed this effect in our previous whole cell patch clamp experiments (22). Interestingly, when cells expressing Kir6.2/ SUR2B channels were incubated with a Ca 2ϩ chelator, PdBu failed to inhibit the pinacidil-induced efflux, suggesting that the PKC effect on this isoform of the K ATP channel is Ca 2ϩ -dependent. Kir6.1/SUR2B, on the other hand, was still inhibited by PdBu in the presence of the chelator, agreeing with the consensus that a novel Ca 2ϩ -independent PKC phosphorylates this channel (22,26,43). These data led us to hypothesize whether there may be a Ca 2ϩ -dependent component to the regulation of Kir6.2 by PKC. The lack of an effect by PdBu in whole cell patch clamp experiments may well have been due to a low intracellular free Ca 2ϩ concentration (ϳ20 nM) in the pipette (22,23). In fact, we found this to be the case, there was no effect with PdBu with low intracellular Ca 2ϩ (ϳ20 nM and ϳ50 nM); however, upon increasing the free Ca 2ϩ concentration (ϳ100 or 300 nM) in the pipette, application of PdBu led to ϳ40 -50% inhibition of the glibenclamide-sensitive current. Previous studies have used at least 100 nM free Ca 2ϩ to study Ca 2ϩ -dependent PKC effects on ion channels (44,45). Our data imply that the PdBuinduced inhibition of Kir6.2/SUR2B requires a typical resting level of 100 nM free Ca 2ϩ or higher. Although we have not investigated the identity of the specific PKC isoform responsible for inhibition of Kir6.2-containing K ATP channels in this study, several Ca 2ϩ -dependent PKC isoforms have been shown to be expressed in vascular smooth muscle and the heart, including PKCs ␣, ␤, and ␥ (43, 46, 47). Furthermore, angioten- The excised patches were held at Ϫ50 mV in symmetrical 140 mM K ϩ solution and subsequently exposed to different intracellular concentrations of ATP. C, relative ATP sensitivity of Kir6.2/SUR2B and Kir6.2-S372E/ SUR2B. Data were grouped from four to eight inside-out patches and fitted to the following equation: Relative current ϭ 1/(1 ϩ ([ATP]/IC 50 ) k ) where relative current is the current relative to the current measured in the absence of ATP and k is the Hill coefficient.

Kir6.2 Subunit Regulation by PKC
sin II-evoked inhibition of inwardly rectifying K ϩ channels in rabbit coronary arterial cells has been shown to be mediated by PKC␣ (48). Recent studies have shown a clear translocation of PKC␣ from the cytosol to the membrane following stimulation with a phorbol ester in aortic and arterial smooth muscle cells (43,49). More specifically for our studies, the Ca 2ϩ -dependent isoforms PKC␣, ␤I, and ␤II show a strong endogenous presence in HEK293 cells (44). The inhibition of Kir6.2/SUR2B current as a result of PKC activation was slow, taking ϳ20 min to reach steady state at room temperature, suggesting that the inhibition we see with PdBu is likely to be due to PKC-induced internalization rather than a direct effect on plasma membrane channel activity and reduction in open probability. Indeed, our biochemical studies using cell surface expression assays (Fig. 7) show significantly (ϳ68%) reduced cell surface expression of Kir6.2 in the presence of the PKC activator PdBu. This would concur with recent studies implicating PKC activation in the stimulation of endocytosis of K ATP channels containing the Kir6.2 subunit (28,30). The inhibitory effect was also irreversible within the time course of our experiments. However, this is not surprising as it is known that phorbol ester activation of PKC promotes an irreversible and active membrane-bound complex of the kinase (50).
PKC phosphorylation of Kir6.2 has been shown by a number of groups; however, to date only one specific phosphorylation target (Thr-180) has been elucidated (24, 28 -30). It has been shown that this site was not responsible for the PKC-mediated down-regulation of Kir6.2 (28). Phosphopeptide mapping identified the amino acid residue Ser-372 as a target for PKC phosphorylation, and this was confirmed using in vitro phosphorylation assays. There was ϳ75% reduction in the phosphorylation of MBP-Kir6.2C-S372A compared with the wild-type protein. Additionally, there was no clear reduction in the cell surface expression of the Kir6.2-S372A protein in the presence of PdBu (Fig. 7D). Furthermore, our whole cell patch clamp data strongly suggest that Ser-372 is a likely site for down-regulation of the channel. In support of this, the Kir6.2-S372A/S372E and Kir6.2⌬C26 (Ser-372 lies within the last 26 amino acids of the C-terminal mutant) channels were not inhibited by PKC activation. The Kir6.2-S372E mutant showed typically lower basal expression and a smaller response to pinacidil, an effect expected as this mutant is a phosphomimetic. Given the precedent in the literature (28,30), it is likely that PKC promotes internalization through endocytosis of the Kir6.2 subunit; however, the phosphorylation site is adjacent to the RKR motif known to influence forward trafficking of the KATP channel complex (1). Thus, it is also a possibility that forward trafficking is inhibited and these are topics for future investigation.
Our observations that Kir6.2⌬C26/SUR2B and in particular Kir6.2-S372A/SUR2B basal currents were activated by PKC support another PKC phosphorylation site such as Thr-180; and given the fairly quick increase in current, this is most likely to be due to an increase in channel activity. Interestingly, the potentiating effect on Kir6.2 does not appear to be Ca 2ϩ -dependent, as even at low Ca 2ϩ we see a substantial increase in current (Fig. 8); however, this does not completely rule out the possibility of a conventional PKC also acting on Kir6.2 to increase current through it. Patch clamp studies of guinea pig ventricular myocytes suggest that this could be true. K ATP channels in these cells, thought to contain Kir6.2 as the poreforming subunit, have been shown to be regulated by both conventional and novel isoforms of PKC (51). Rainbow et al. have also shown that two isoforms of PKC (PKC␣ and PKC⑀) regulate voltage-gated K ϩ channels in rat mesenteric arteries depending on which receptor pathway is activated (45). Our data imply that the activation and inhibition of current through Kir6.2 by PKC may be similarly isoform-specific. Previous studies have shown that phosphorylation of Thr-180 leads to a decrease in ATP sensitivity of Kir6.2 (29,37). In our hands, the relative ATP sensitivity of the WT channel and the phosphomimetic Kir6.2-S372E mutant channel did not change, suggesting that by implication phosphorylation of Ser-372 does not affect ATP binding to Kir6.2. Thr-180 is thought to line the hydrophobic pocket where the adenine ring of ATP sits explaining why modification of this residue reduces ATP sensitivity (52). Conversely, Ser-372 is at the distal C terminus of Kir6.2 and has not been shown to be involved in ATP binding. The Ser-372 residue has also been implicated as a site for PKA phosphorylation (27). However, it is not uncommon for both PKA and PKC to act on the same site on a protein (53)(54)(55). Indeed, on the Kir6.1 subunit one site (Ser-385) has been implicated as a target for both PKC and PKA (23,56). PKA and PKC phosphorylation have opposing effects on K ATP channels in smooth muscle, with the former leading to activation and vasodilation and the latter inhibition and subsequent vasoconstriction (21,25,57). Thus, other factors must be at play to give context to the net signaling response, and for example these might be additional phosphorylation sites on SUR2B or other proteins in the complex and/or particular intracellular nucleotide or Ca 2ϩ conditions.
In this study, using protein biochemistry and electrophysiology, we have shown a relatively slow inhibition of Kir6.2/ SUR2B by PKC and shown that this is Ca 2ϩ -dependent consistent with the action of a conventional PKC. In addition, we have identified a novel site (Ser-372) for Ca 2ϩ -dependent PKC phosphorylation on the Kir6.2 subunit.