HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 292/6/C2239    most recent 00326.2006v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Mao, X. Articles by Lin, Y.-F. Search for Related Content PubMed PubMed Citation Articles by Mao, X. Articles by Lin, Y.-F.

MEMBRANE TRANSPORTERS, ION CHANNELS, AND PUMPS

Dual regulation of the ATP-sensitive potassium channel by caffeine

Departments of ¹Physiology and Membrane Biology and ²Anesthesiology and ³Biochemistry and Molecular Biology Graduate Group, University of California, Davis, California

Submitted 14 June 2006 ; accepted in final form 7 February 2007

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
ATP-sensitive potassium (K_ATP) channels couple cellular metabolic status to changes in membrane electrical properties. Caffeine (1,2,7-trimethylxanthine) has been shown to inhibit several ion channels; however, how caffeine regulates K_ATP channels was not well understood. By performing single-channel recordings in the cell-attached configuration, we found that bath application of caffeine significantly enhanced the currents of Kir6.2/SUR1 channels, a neuronal/pancreatic K_ATP channel isoform, expressed in transfected human embryonic kidney (HEK)293 cells in a concentration-dependent manner. Application of nonselective and selective phosphodiesterase (PDE) inhibitors led to significant enhancement of Kir6.2/SUR1 channel currents. Moreover, the stimulatory action of caffeine was significantly attenuated by KT5823, a specific PKG inhibitor, and, to a weaker extent, by BAPTA/AM, a membrane-permeable Ca²⁺ chelator, but not by H-89, a selective PKA inhibitor. Furthermore, the stimulatory effect was completely abrogated when KT5823 and BAPTA/AM were co-applied with caffeine. In contrast, the activity of Kir6.2/SUR1 channels was decreased rather than increased by caffeine in cell-free inside-out patches, while tetrameric Kir6.2LRKR368/369/370/371AAAA channels were suppressed regardless of patch configurations. Caffeine also enhanced the single-channel currents of recombinant Kir6.2/SUR2B channels, a nonvascular smooth muscle K_ATP channel isoform, although the increase was smaller. Moreover, bidirectional effects of caffeine were reproduced on the K_ATP channel present in the Cambridge rat insulinoma G1 (CRI-G1) cell line. Taken together, our data suggest that caffeine exerts dual regulation on the function of K_ATP channels: an inhibitory regulation that acts directly on Kir6.2 or some closely associated regulatory protein(s), and a sulfonylurea receptor (SUR)-dependent stimulatory regulation that requires cGMP-PKG and intracellular Ca²⁺-dependent signaling.

phosphodiesterase; protein kinase; calcium; single channel; patch clamp

The K_ATP channel is an octamer (8, 16, 35, 74). It is composed of four pore-forming -subunits (Kir6.2 or Kir6.1) (33, 66), members of the inwardly rectifying potassium (Kir) channel family, and four regulatory -subunits (the sulfonylurea receptor SUR1, SUR2A, or SUR2B) (3, 34, 37), members of the ATP-binding cassette (ABC) protein superfamily. Partial complexes with fewer than eight subunits do not reach the cell membrane because of the exposure of an endoplasmic reticulum (ER) retention/retrieval signal that is present in each subunit (92). Truncation or mutation of the ER retention/retrieval signal located at the COOH terminus of Kir6.2 permits the functional expression of tetrameric Kir6.2 channels on the cell surface in the absence of SUR (83, 92). Native K_ATP channels in different tissues are composed of Kir6.x subunits in combination with different SUR subunits. For example, K_ATP channels in pancreatic -cells and most central neurons contain Kir6.2 and SUR1 (2, 33). Cardiac and skeletal muscle K_ATP channels are composed of Kir6.2 and SUR2A (34, 56). K_ATP channels in nonvascular smooth muscle and certain central neurons in the substantia nigra (SN) consist of Kir6.2 and SUR2B (37, 47, 89). The sites mediating ATP inhibition are located in the Kir6.2 subunit, but the SUR subunit can further modulate the ATP sensitivity of K_ATP channels (23, 39, 53, 73, 82, 83). Moreover, the SUR subunit confers the channel sensitivity to sulfonylurea drugs (inhibition) as well as to MgADP and synthetic K_ATP channel openers (stimulation) (4, 8, 38, 52, 70, 83).

Caffeine (1,2,7-trimethylxanthine) is a natural compound found in cocoa beans, cola nuts, coffee, and tea and is highly membrane permeable. The actions of caffeine include the release of calcium from intracellular stores (14, 64), inhibition of phosphodiesterase (PDE) (46, 63), and antagonism of adenosine receptors (51). It has also been demonstrated that caffeine at high concentrations acts as an inhibitor of several potassium channels, including depolarization-activated potassium channels in sympathetic neurons, secretory cells (59), taste receptor cells (93), and ventricular myocytes (67); calcium-activated potassium (K_Ca) channels in GH₃ pituitary cells (41), sympathetic neurons (59), secretory cells (59), vascular smooth muscle cells (54), and taste receptor cells (93); human ether a-go-go-related gene potassium (hERG) channels in transfected mammalian cells (17); Kir channels in GH₃ rat anterior pituitary cells (9), ventricular myocytes (15, 67, 85), and taste receptor cells (93); K_ATP channels in pancreatic -cells (36) and smooth muscle cells (80); and recombinant two-pore domain potassium channel TREK-1 in Chinese hamster ovary (CHO) cells (26). On the other hand, caffeine has also been shown to increase potassium efflux in skeletal muscle (86) and to activate K_Ca channels in adrenal chromaffin cells (55). How caffeine modulates the K_ATP channels in intact cells, however, was unclear, as the patch configuration (i.e., inside-out mode) used in previous studies (36, 80) may not allow detection of channel modulation involving intracellular signaling; in addition, the ion channel conductance inhibited by caffeine in urethra (80) seems too small for nonvascular smooth muscle K_ATP channels (37, 89).

To understand how caffeine regulates the function of K_ATP channels, we performed single-channel recordings of recombinant Kir6.2/SUR1, Kir6.2/SUR2B, and Kir6.2LRKR368/369/370/371AAAA (i.e., Kir6.2FL4A, a trafficking mutant; Ref. 92) channels in transiently transfected human embryonic kidney (HEK)293 cells as well as the endogenous K_ATP channels expressed in rat insulin-secreting Cambridge rat insulinoma G1 (CRI-G1) cells in both cell-attached and inside-out patches. The cell-attached patch configuration was included to preserve the cell integrity for the determination of potential signaling pathways, whereas the inside-out patch configuration was used to detect any direct effect that does not require diffusible cytosolic messengers. Here we report for the first time that caffeine exerts dual regulation on the function of K_ATP channels. Caffeine stimulates the K_ATP channel via suppression of PDE activity, which hydrolyzes cGMP and consequently leads to PKG activation, along with activation of some Ca²⁺-dependent mechanism, while it exhibits a relatively weak inhibition of channel function via direct interaction with the Kir6.2 subunit or some closely associated regulatory protein(s). The potent stimulatory action appears to predominate over the inhibitory effect of caffeine on the K_ATP channel in intact cells. These findings also raise caution against the simple interpretation of experiments employing caffeine to study Ca²⁺-dependent transduction pathways in cells expressing K_ATP channels.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Construction of cDNAs

To reconstitute the pancreatic/neuronal-type and smooth muscle-type K_ATP channels, cDNAs encoding SUR1 (hamster), SUR2B (rat) and Kir6.2 (mouse) were used. In addition, cDNAs encoding Kir6.2FL4A, a trafficking mutant that can be functionally expressed without SUR, were also used. All cDNA constructs were verified by DNA sequencing and subcloned into expression vector pcDNA3 (Invitrogen, Carlsbad, CA). Plasmids prepared with Qiagen maxipreps (Qiagen, Valencia, CA) were used for transient transfection.

Cell Culture and Transient Transfection of HEK293 Cells

HEK293 cells (American Type Culture Collection, Manassas, VA) were maintained in DMEM/F12 (Mediatech, Herndon, VA; supplemented with 2 mM L-glutamine, 10% fetal bovine serum, 100 IU/ml penicillin, and 100 µg/ml streptomycin) at 37°C in humidified 5% CO₂. Cells were transiently transfected using the FuGENE 6 reagent (Roche, Indianapolis, IN) mixed with expression plasmids containing cDNAs of interest in serum-free medium. A marker gene encoding the green fluorescent protein (pEGFP-1; Clontech, Mountain View, CA) was co-transfected with cDNAs of interest in a ratio of 1.5:10. Transfection was carried out according to the manufacturer's protocols. The cells were replated the following day at a density of 5,000–20,000 cells/dish onto 12-mm glass coverslips precoated with 1.5 µg/ml fibronectin (Sigma-Aldrich, St. Louis, MO) to be recorded 48–72 h after transfection (45).

Cell Culture of Insulin-Secreting Rat Pancreatic Islet Cell Line CRI-G1

Insulinoma CRI-G1 cells (ECACC, Wiltshire, UK) were maintained in DMEM (supplemented with 2 mM L-glutamine, 5% fetal bovine serum, 50 IU/ml penicillin, and 50 µg/ml streptomycin) at 37°C in humidified 5% CO₂ as described by Carrington et al. (13). Cells were plated onto 12-mm glass coverslips precoated with fibronectin at a density of 10,000 cells/dish and were used for recordings for a few days.

Electrodes, Recording Solutions, and Single-Channel Recordings

The recording electrodes were pulled from thin-walled borosilicate glass with an internal filament (MTW150F-3; World Precision Instruments, Sarasota, FL), using a P-97 Flaming Brown puller (Sutter Instrument, Novato, CA), and were fire-polished to a resistance of 5–10 M. The intracellular (bath) solution consisted of (in mM) 110 KCl, 1.44 MgCl₂, 30 KOH, 10 EGTA, 10 HEPES, pH to 7.2. The extracellular (intrapipette) solution consisted of (in mM) 140 KCl, 1.2 MgCl₂, 2.6 CaCl₂, 10 HEPES, pH to 7.4. The same combination of recording solutions was used for recordings performed in both HEK293 and CRI-G1 cells. For both inside-out and cell-attached single-channel recordings (24), the recording chamber (RC26; Warner Instruments, Hamden, CT) was filled with the intracellular solution, and the recording pipette was filled with the extracellular solution. The use of symmetrical recording solutions (140 mM K⁺) in cell-attached as well as inside-out patch recordings resulted in an equilibrium potential for potassium (E_K) and a resting membrane potential (V_m) 0 mV in both cell types, as determined from the I-V relationship of the K_ATP channel. All recordings were carried out at room temperature, and all patches were voltage clamped at –60 mV intracellularly (i.e., with +60-mV intrapipette potentials) unless otherwise specified. Single-channel currents were recorded with an Axopatch 200B patch-clamp amplifier (Molecular Devices-Axon Instruments, Sunnyvale, CA) and were low-pass filtered (3 dB, 2 kHz). Single-channel data were acquired and digitized at 20 kHz online using Clampex 9 software (Axon) via a 16-bit A/D converter (Digidata-acquisition board 1322A, Axon).

Preparations of Drugs

Caffeine (0.6, 6, and 30 mM) was prepared fresh daily from powder with bath recording solutions. Stock solutions of all other drugs were prepared in DMSO and stored at –80°C in aliquots. Working solutions of 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetra-acetic acid tetrakis(acetoxymethyl ester) (BAPTA/AM; 50 µM), thapsigargin (Tg; 0.1–0.2 µM), 3-isobutyl-1-methylxanthine (IBMX; 0.5 mM), 1,4-dihydro-5-(2-propoxyphenyl)-7H-1,2,3-triazolo[4,5-d]pyrimidine- 7-one (zaprinast; 50 µM), KT5823 (1–2 µM), 4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone (Ro-20-1724; 20 µM), N-[2-((p-bromocinnamyl)amino)ethyl]-5-isoquinolinesulfonamide, 2HCl (H-89; 0.1 µM), and glibenclamide (5 µM) were diluted from aliquots before use with bath recording solutions. These inhibitors were chosen based on their relatively high selectivity for specific target proteins and were used at working concentrations within the range showing little or no cross-reaction. All working drug solutions were put on ice and kept away from light. A small recording chamber was used, and drugs were applied through a pressure-driven perfusion system (BPS-8; ALA Scientific Instruments, Westbury, NY) via a micromanifold positioned closely to cell-attached or inside-out patches.

Data Analysis

Single-channel currents. Digitized single-channel records were detected with Fetchan 6.05 (events list) of pCLAMP (Axon) using 50% threshold crossing criterion and analyzed with Intrv5 (Dr. Barry S. Pallotta, University of North Carolina, Chapel Hill, NC; and Dr. Janet Fisher, University of South Carolina, Columbia, SC). Analysis was performed at the main conductance level (70 pS). Only patches with infrequent multiple-channel activity were used for single-channel analysis. Duration histograms were constructed as described by Sigworth and Sine (75), and estimates of exponential areas and time constants were obtained using the method of maximal likelihood estimation. The number of exponential functions required to fit the duration distribution was determined by fitting increasing numbers of functions until additional components could not significantly improve the fit (31, 49). Events with duration <1.5 times the system dead time were not included in the fit. Mean durations were corrected for missed events by taking the sum of the relative area (a) of each exponential component in the duration frequency histogram multiplied by the time constant () of the corresponding component.

Multiple-channel currents. In patches where multiple-channel activities of K_ATP channels were observed for >10% of recording time, the digitized current records were analyzed using Fetchan 6.05 (browse) of pCLAMP to integrate currents in 120-s segments. The current amplitude (I) values (current amplitude = integrated current/acquisition time) were then normalized to the corresponding controls obtained from the same patches to yield normalized open probability (NPo) (control as 1), as the normalized current amplitude was equivalent to the normalized NPo obtained from single-channel analysis when the single-channel conductance remains the same (see Current normalization).

Current normalization. Normalized current amplitude I_drug/I_control = (NPo·i)_drug/(NPo·i)_control = NPo_drug/NPo_control = normalized NPo, where i represents the single-channel conductance and I represents the macroscopic current amplitude. The normalized NPo values obtained from both single-channel and multiple-channel patches were then pooled and are presented as bar graphs (see Figs. 1B, 2D, 3F, 4F, 5B, and 6F) and in Table 1, the only exception being the NPo values of endogenous K_ATP channels obtained from CRI-G1 cells in the cell-attached patch configuration (see Fig. 6, A and B), which were compared before and during caffeine treatment in individual patches using absolute values (pairwise comparison) without normalization, as not all cell-attached patches obtained from these cells exhibited detectable K_ATP channel activity in the control condition. In contrast, data obtained from the inside-out patches excised from CRI-G1 cells as well as all data obtained from HEK293 cells were normalized to the corresponding controls obtained in individual patches (taken as 1), as all these patches exhibited K_ATP channel activity in the control condition.

View larger version (18K): [in this window] [in a new window]   Fig. 1. Caffeine enhances Kir6.2/SUR1 channel activity in intact cells in a dose-dependent manner. Recombinant Kir6.2/SUR1 channels were expressed in human embryonic kidney HEK293 cells by transient transfection. Recordings were performed in symmetrical high potassium (140 mM) solutions at room temperature in cell-attached configurations, and voltage was clamped at –60 mV. SUR, sulfonylurea receptor. A: single-channel current traces of Kir6.2/SUR1 channel obtained from a cell-attached patch before (top) and during caffeine (30 mM) application (bottom). Upward deflections represent openings from closed states. For all current trace figures, segments of raw recordings underlined are shown in successive traces at increasing temporal resolution, revealing singular openings (*) and bursts of openings (**). Horizontal scale bar represents 1 s, 300 ms, and 100 ms for traces from top to bottom, and the vertical scale bar represents 4 pA. B: normalized open probability (NPo) of Kir6.2/SUR1 channels obtained during application of caffeine at different concentrations: 0.6 mM (open column), 6 mM (hatched column), and 30 mM (solid column). NPo values were normalized to corresponding control (taken as 1) obtained before caffeine application in individual patches. Horizontal line indicates the control level. Data are presented as means ± SE of 4–9 patches (patch nos. in parentheses). Significance levels are as follows: *P < 0.05 and ***P < 0.001 (two-tailed one-sample t-tests).

View larger version (13K): [in this window] [in a new window]   Fig. 2. Enhancement of Kir6.2/SUR1 channel activity by caffeine is mimicked by phosphodiesterase (PDE) inhibition. Currents were obtained in cell-attached configurations from transiently transfected HEK293 cells. A: single-channel current traces of Kir6.2/SUR1 channels in a cell-attached patch before (top) and during bath perfusion of IBMX (0.5 mM), a nonselective PDE inhibitor (bottom). B: single-channel current traces of Kir6.2/SUR1 channels in another cell-attached patch before (top) and during application of zaprinast (50 µM), a cGMP-specific PDE inhibitor (bottom). C: single-channel current traces of Kir6.2/SUR1 channels in a cell-attached patch before (top) and during application of Ro-20-1724 (20 µM), a cAMP-specific PDE inhibitor (bottom). Scale bars are the same as described in Fig. 1. D: normalized NPo values of Kir6.2/SUR1 channels obtained during applications of IBMX, zaprinast, Ro-20-1724, and caffeine (30 mM; same data as shown in Fig. 1B, included here for comparison). Data are presented as means ± SE of 5–10 patches (patch nos. in parentheses). Drug effect was compared with the corresponding control obtained from the same patch and normalized as described in Fig. 1B (control as 1, horizontal line). Significance levels are as follows: *P < 0.05, **P < 0.01, and ***P < 0.001 (two-tailed one-sample t-tests).

View larger version (27K): [in this window] [in a new window]   Fig. 3. Enhancement of Kir6.2/SUR1 channel activity by caffeine is attenuated by intracellular Ca2+ buffering or inhibition of PKG but not by PKA inhibition. All currents were obtained in cell-attached configurations from transfected HEK293 cells. A: single-channel current traces of Kir6.2/SUR1 channels in a cell-attached patch obtained from a dish pretreated with BAPTA/AM (50 µM), a membrane-permeable Ca2+ chelator, before (top) and during application of caffeine (30 mM) in the continuous presence of BAPTA/AM (50 µM) (bottom). B: single-channel current traces of Kir6.2/SUR1 channels in a different cell-attached patch pretreated with H-89 (0.1 µM), a selective PKA inhibitor, before (top) and during application of caffeine (30 mM) in the continuous presence of H-89 (0.1 µM) (bottom). C: single-channel current traces of Kir6.2/SUR1 channels in a patch pretreated with KT5823 (2 µM), a selective PKG inhibitor, before (top) and during application of caffeine (30 mM) in the continuous presence of KT5823 (2 µM) (bottom). D: single-channel current traces of Kir6.2/SUR1 channels in a patch before (top) and during application of caffeine (30 mM) in the continuous presence of KT5823 (2 µM) and BAPTA/AM (50 µM) (bottom). E: single-channel current traces of Kir6.2/SUR1 channels in a cell-attached patch before (top) and during application of caffeine (30 mM) in the continuous presence of KT5823 (2 µM), BAPTA/AM (50 µM), and H-89 (0.1 µM) (bottom). Scale bars are the same as described in Fig. 1. F: normalized NPo of Kir6.2/SUR1 channels obtained from individual experimental groups described in A-E. Caffeine data (black column) were the same as shown in Fig. 1B and are displayed here for reference. NPo values were normalized to the corresponding control (taken as 1) obtained before drug perfusion in individual patches. Data are presented as means ± SE of 4-12 patches (patch nos. in parentheses). Significance levels are as follows: *P < 0.05, **P < 0.01, and ***P < 0.001 (two-tailed one-sample t-tests within individual groups, or Dunnett's multiple comparison tests between groups).

View larger version (33K): [in this window] [in a new window]   Fig. 4. Inhibition of Kir6.2/SUR1 channel activity by caffeine, as revealed in excised patches, is Kir6.2-dependent. Currents were obtained in cell-attached or inside-out configurations (as specified) from transfected HEK293 cells. A, C, and E: single-channel current traces of Kir6.2/SUR1 channel obtained from an inside-out patch (A) or Kir6.2FL4A channels obtained, respectively, from an inside-out patch (C) and a cell-attached patch (E) before (top) and during caffeine (30 mM) application (bottom). The bath contained no ATP. B and D: single-channel current traces of Kir6.2/SUR1 channel (B) or Kir6.2FL4A channel (D) obtained from inside-out patches during continuous (ATP-free) bath perfusion. Continuous bath perfusion in these time control experiments was used to provide estimation of current rundown rate (i.e., time control) over the same time course as was used for caffeine application in caffeine-treated groups (8-min intervals between 2 records). Scale bars are the same as described in Fig. 1. F: normalized NPo of Kir6.2/SUR1 and Kir6.2FL4A channels obtained from individual experimental groups described in A–E. Patch configurations in which data were obtained: iso, inside-out; ca, cell-attached. NPo values were normalized in individual patches as described in Fig. 1B and are presented as means ± SE of 4–9 patches (patch nos. in parentheses). Significance levels are as follows: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 (two-tailed one-sample t-tests within individual groups, or unpaired t-tests between caffeine-treated and time control groups).

View larger version (17K): [in this window] [in a new window]   Fig. 5. Caffeine enhances Kir6.2/SUR2B channel activity in intact cells. Single-channel currents were obtained in cell-attached configurations from transfected HEK293 cells. A: single-channel current traces of Kir6.2/SUR2B channel obtained from a cell-attached patch before (top) and during bath perfusion of caffeine (30 mM) (bottom). Scale bars are the same as described in Fig. 1. B: normalized NPo of Kir6.2/SUR2B channels obtained during application of caffeine (30 mM; open column). Data obtained from Kir6.2/SUR1 channels (solid column, 30 mM; same data as shown in Fig. 1B) were included here for the purpose of comparison. NPo values were normalized to the corresponding controls in individual patches (control taken as 1), and the average data are presented as means ± SE of 10 patches (for Kir6.2/SUR2B channels). Two-tailed one-sample t-tests and unpaired t-tests were performed to detect difference within and between Kir6.2/SUR1 and Kir6.2/SUR2B isoforms, respectively. Significance levels are as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

View larger version (24K): [in this window] [in a new window]   Fig. 6. Caffeine exerts dual modulation on the function of the native ATP-sensitive potassium (KATP) channel present in insulin-secreting Cambridge rat insulinoma G1 (CRI-G1) cells. Single-channel currents were obtained in both cell-attached (A) and inside-out (C–E) configurations. A: single-channel current traces obtained from a cell-attached patch before (top) and during caffeine (30 mM) application (bottom) in regular ATP-free bath. B: absolute NPo values of KATP channels obtained from cell-attached patches before and during caffeine treatment. Data were averaged without normalization from 6 patches and compared using two-tailed paired t-tests. C–E: single-channel current traces obtained from an inside-out patch before (top) and during application of caffeine (30 mM; C), prolonged bath perfusion (D), or glibenclamide (5 µM; E) (bottom) in the 10 µM MgATP-containing bath. Continuous bath perfusion was used to provide time control of the rundown rate over the same time course as was used for caffeine-treated groups. Scale bars are the same as described in Fig. 1. F: normalized NPo values of these native KATP channels obtained from inside-out patches treated with caffeine, prolonged bath perfusion (time control), or glibenclamide. NPo values were normalized in individual patches as described in Fig. 1B (control taken as 1) and are presented as means ± SE of 4–5 patches. Significance levels and statistical methods are the same as described in Fig. 4F.

Data were averaged and are presented as means ± SE. Statistical comparisons were made using Student's two-tailed one-sample, paired or unpaired, t-tests or one-way ANOVA followed by Dunnett's multiple comparison test. Significance was assumed when P < 0.05. Statistical comparisons were performed using Prism (GraphPad Software, San Diego, CA).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
The regulation of K_ATP channel function by caffeine was examined in transiently transfected HEK293 cells expressing octameric Kir6.2/SUR1, Kir6.2/SUR2B, or tetrameric Kir6.2FL4A channels as well as in rat insulinoma CRI-G1 cells expressing native K_ATP channels, using electrophysiological recordings. Both cell-attached and inside-out patch configurations were employed. To determine the potential signaling mechanisms involved in caffeine-induced K_ATP channel modulation, we monitored the effects of several PDE inhibitors, including IBMX (nonselective), zaprinast (selective for cGMP-specific PDE), and Ro-20-1724 (selective for cAMP-specific PDE), on the function of Kir6.2/SUR1 channels. We also examined the impact of intracellular buffering (by BAPTA/AM, a membrane-permeable Ca²⁺ chelator) as well as inhibition of PKA (by H-89, a selective PKA inhibitor) and PKG (by KT5823, a selective PKG inhibitor) on the stimulatory effects of caffeine. Our results suggest a novel caffeine-induced stimulation of K_ATP channels through cGMP-PKG and Ca²⁺-dependent signaling in intact cells, in addition to an inhibitory action directly on the Kir6.2 subunit or some closely associated regulatory protein(s).

Caffeine Enhanced the Single-Channel Currents of Recombinant Kir6.2/SUR1 Channels, a Neuronal/Pancreatic K_ATP Channel Isoform, in a Concentration-Dependent Manner

Caffeine, a membrane-permeable compound, has been suggested to alter the function of several potassium channels, including K_ATP channels (36, 80). In some of the previous studies investigating how caffeine modulates ion channel function, intracellular signaling components might have been lost because of the use of cell-free or conventional whole cell modes of recordings. To prevent this potential problem, cell-attached patch recordings were employed in the present study to elucidate the signaling mechanism underlying K_ATP channel modulation induced by caffeine. Recombinant Kir6.2/SUR1 channels were expressed in HEK293 cells by transient transfection, and the currents were recorded in symmetrical 140 mM potassium solutions. Both the E_K and the V_m were 0 mV, as determined from the I-V relationship of the currents. The single-channel currents obtained from positively transfected cells exhibited an inwardly rectifying I-V relationship, ATP sensitivity (i.e., current increase on patch excision to an ATP-free bath), and single-channel conductance at 70 pS (at hyperpolarizing membrane potentials), all characteristic of the expression of K_ATP channels. To determine how caffeine modulated K_ATP channel function in intact cells, we monitored changes in the single-channel currents of recombinant Kir6.2/SUR1 channels before and during caffeine application in cell-attached patches. All patches were voltage clamped at –60 mV intracellularly, and the single-channel openings of K_ATP channels are displayed as upward deflections throughout this study. At increasing temporal resolution (Fig. 1A, Control, top to bottom traces), single-channel openings of Kir6.2/SUR1 channels could be resolved into singular openings and bursts of openings. Application of caffeine (30 mM) via a closely positioned perfusion tubing induced marked enhancement of Kir6.2/SUR1 single-channel currents within 2 min, and the response reached a plateau within 6 min after initiating drug perfusion. The apparent opening frequency and burst duration were increased, whereas the single-channel conductance remained the same, during caffeine application (Fig. 1A, Caffeine). The caffeine effects were partially reversible with bath wash (data not shown); hence, only the responses obtained during the first caffeine application to individual cells were included for statistical comparisons.

To quantify how caffeine alters the function of K_ATP channels, the single-channel properties of Kir6.2/SUR1 channels in patches with suitable activity level (i.e., <10% of total openings exhibiting multiple-channel opening pattern) were obtained by performing single-channel event detection and analysis (see MATERIALS AND METHODS for details). The NPo, opening frequency, and corrected mean open time and mean closed time obtained during caffeine application were then normalized to the corresponding controls obtained in individual patches. The normalized values in the presence of 30 mM caffeine were 15.24 ± 1.34 (P < 0.01), 11.61 ± 2.03 (P < 0.05), 1.38 ± 0.13 (not significant), and 0.07 ± 0.03 (P < 0.001) for NPo, opening frequency, and corrected mean open time and mean closed time, respectively (control as 1) (Table 1; mean ± SE, 4 patches, two-tailed one-sample t-test). It appeared that caffeine application resulted in significant increases in NPo and the opening frequency while reducing the mean closed time of Kir6.2/SUR1 channels. The mean open time was not altered. These changes in the single-channel properties formed the basis of the functional enhancement of Kir6.2/SUR1 currents in the presence of caffeine.

In addition, we also analyzed the functional changes of channels in patches with higher activity (i.e., exhibiting multiple-channel openings for >10% of the total openings), which was not suitable for the aforementioned single-channel analysis. In these patches, integration of total currents was performed in lieu of single-channel event detection. The integrated current values were divided by the acquisition time (duration) to yield mean current amplitudes (pA), followed by normalization to the corresponding controls obtained in the same patches. These values were then pooled with the normalized NPo data obtained from single-channel patches and are presented as "normalized NPo" (see MATERIALS AND METHODS). Through such analyses, it was apparent that the stimulation of Kir6.2/SUR1 channels by caffeine was concentration-dependent (Fig. 1B): the normalized NPo values were 2.12 ± 0.20 (Fig. 1B; P < 0.05, 4 patches), 6.74 ± 1.30 (Fig. 1B; P < 0.05, 4 patches), and 25.17 ± 4.67 (Fig. 1B; P < 0.001, 9 patches) during application of caffeine at 0.6, 6, and 30 mM, respectively (two-tailed one-sample t-test). It should also be noted that the increase in the normalized NPo pooled from all patches (25.17 ± 4.67, including both single-channel and multiple-channel patches; Fig. 1B) was larger than that estimated solely from single-channel patches (15.24 ± 1.34; Table 1) by 30 mM caffeine. This was due to the contribution of the patches exhibiting stronger responses (and hence inducing multiple-channel current responses) to caffeine in the former but not the latter group. In either case, our results revealed a novel finding that caffeine was a strong stimulant for recombinant Kir6.2/SUR1 channels in intact HEK293 cells.

Nonselective PDE Inhibitor IBMX Mimicked the Stimulatory Effects of Caffeine on Recombinant Kir6.2/SUR1 Channels

Caffeine is a potent, nonselective PDE inhibitor. Inhibition of PDE leads to accumulation of intracellular cyclic nucleotides, such as cGMP and cAMP. Indeed, accumulation of cyclic nucleotides induced by caffeine has been demonstrated in several cell models, such as mouse papillae tissue (63) and opossum gallbladder muscle (46). Accumulation of cyclic nucleotides is also obtained using other PDE inhibitors, including those chemically related (such as theophylline; Ref. 63) and unrelated to caffeine (84). Would inhibition of PDE be responsible for the caffeine-induced stimulation of Kir6.2/SUR1 channels? Caffeine and theophylline as well as IBMX are classic methylxanthines that nonselectively inhibit PDE activity. We thus examined the activity of Kir6.2/SUR1 channels before and during bath application of IBMX to cell-attached patches obtained from positively transfected HEK293 cells. IBMX (0.5 mM) increased the single-channel currents of Kir6.2/SUR1 channels from the control level in the same patch (Fig. 2A, Control vs. IBMX); the apparent opening frequency and burst duration were increased, while the single-channel conductance remained unchanged. The averaged normalized NPo of Kir6.2/SUR1 currents obtained from a total of seven patches during application of IBMX was 9.61 ± 2.92 (control as 1) (Fig. 2D; P < 0.05, two-tailed one-sample t-test). These data indicated that nonselective inhibition of PDE activity in intact HEK293 cells resulted in enhancement of recombinant Kir6.2/SUR1 channel function, an action that may account for the stimulation of K_ATP channels by caffeine. The normalized NPo data obtained from patches treated with 30 mM caffeine (same data as shown in Fig. 1B) are also displayed in Fig. 2D for the purpose of comparison.

Inhibition of cGMP-Specific PDE Increased Kir6.2/SUR1 Single-Channel Currents

Activation of Kir6.2/SUR1 channels by the nonselective PDE inhibitors caffeine and IBMX suggests that accumulation of cyclic nucleotides such as cGMP and/or cAMP may positively modulate K_ATP channel function. To determine whether the increase of intracellular cGMP levels was sufficient to enhance K_ATP channel function, zaprinast, a selective inhibitor of cGMP-specific PDEs (i.e., PDE-5 and PDE-9), was applied to cell-attached patches obtained from transfected HEK293 cells expressing Kir6.2/SUR1 channels. Bath application of zaprinast (50 µM) increased Kir6.2/SUR1 single-channel currents from the control level obtained before drug treatment in the same patch (Fig. 2B, Control vs. Zaprinast); the apparent opening frequency and burst duration were increased, whereas the single-channel conductance level was unaltered. The normalized NPo of Kir6.2/SUR1 channels averaged from a group of five patches was 7.95 ± 1.40 (control as 1), which represented significant enhancement caused by zaprinast (Fig. 2D; P < 0.01, two-tailed one-sample t-test). These results indicated that accumulation of cGMP by PDE inhibition effected the activation of Kir6.2/SUR1 channels in intact HEK293 cells. Caffeine has been demonstrated to increase the tissue cGMP content (46, 63); a similar action of caffeine may occur in HEK293 cells, thereby stimulating Kir6.2/SUR1 channels through some cGMP-dependent signal transduction.

Inhibition of cAMP-Specific PDE Effected Mild Enhancement of Kir6.2/SUR1 Single-Channel Currents

Since caffeine is a nonselective PDE inhibitor, it may cause accumulation of cAMP in addition to cGMP. It has been suggested that simultaneous increases of intracellular cAMP and cGMP by caffeine inhibit the agonist-evoked intracellular Ca²⁺ response in pancreatic acinar cells (12). We thus performed the following experiments to determine whether accumulation of cAMP by inhibition of cAMP-hydrolyzing PDE affects the K_ATP channel function. We and others (10, 45) have previously demonstrated that PKA activates Kir6.2/SUR1 channels, and the activation is mediated by direct phosphorylation of the pore-forming Kir6.2 subunit. Provided there was basal endogenous cAMP-hydrolyzing PDE activity in HEK293 cells, a logical prediction along this line would be a positive effect resulting from the application of a cAMP-specific PDE inhibitor on the function of the Kir6.2/SUR1 channels through cAMP accumulation and consequently the activation of PKA. Indeed, we found that application of Ro-20-1724 (20 µM), a membrane-permeable inhibitor selective for cAMP-specific PDE-4, increased the single-channel currents of Kir6.2/SUR1 channels in cell-attached patches (Fig. 2C). The apparent opening frequency was higher during the perfusion of Ro-20-1724 compared with the control obtained before drug treatment in the same patch (Fig. 2C, Control vs. Ro-20-1724). The normalized NPo averaged from a total of 10 patches treated with Ro-20-1724 was significantly increased to 3.80 ± 1.17 (control as 1) (Fig. 2D; P < 0.05, two-tailed one-sample t-test), although the effect of Ro-20-1724 seemed relatively mild compared with those evoked by caffeine (30 mM; Figs. 1 and 2D), IBMX (0.5 mM; Fig. 2, A and D), and zaprinast (50 µM; Fig. 2, B and D). These data indicated that inhibition of cAMP-specific PDE and the subsequent increase in the intracellular cAMP level enhanced the function of Kir6.2/SUR1 channels in intact HEK293 cells, presumably through activation of endogenous PKA. The positive effect of Ro-20-1724 was in line with our earlier findings that direct application of the catalytic subunits of PKA in inside-out patches increases Kir6.2/SUR1 channel activity (45). Whether the cAMP-PKA signaling mechanism contributed to the caffeine-induced K_ATP channel stimulation was to be determined.

Buffering of Intracellular Ca²⁺ by BAPTA/AM Attenuated the Stimulation of Kir6.2/SUR1 Channels Induced by Caffeine

One of the actions of caffeine is to induce the release of calcium from intracellular ryanodine-sensitive stores through activation of the ryanodine receptor (RyR) (14, 64). Would the stimulatory effects of caffeine on Kir6.2/SUR1 channels (Fig. 1) result from an increase in the intracellular Ca²⁺ concentration following Ca²⁺ mobilization or, alternatively, be mediated solely by Ca²⁺-independent actions of caffeine, such as suppression of the PDE activity? We found that caffeine was able to enhance the currents of Kir6.2/SUR1 channels in both Ca²⁺-free (see Fig. 1) and Ca²⁺-containing (data not shown) bath solutions. However, to avoid refilling of intracellular calcium stores after caffeine-induced store depletion and to prevent sustained elevation of the intracellular Ca²⁺ concentration during prolonged caffeine application (48), a Ca²⁺-free bath solution was used throughout the present study. To determine whether the stimulation of Kir6.2/SUR1 channels by caffeine depended on the release of Ca²⁺ ions from intracellular stores, a group of cells were pretreated with thapsigargin (Tg; 0.1 µM), a cell-permeable inhibitor of ER Ca²⁺-ATPase, for at least 15 min at 37°C to empty intracellular calcium stores. Cells were then subjected to recordings at room temperature in the continuous presence of Tg. Subsequent co-application of caffeine (30 mM) and Tg (0.1–0.2 µM) resulted in a marked increase in the Kir6.2/SUR1 single-channel currents in cell-attached patches (traces not shown). The averaged normalized NPo of Kir6.2/SUR1 currents was 23.29 ± 6.20 (control as 1) (P < 0.01, 10 patches, two-tailed one-sample t-test) during co-application of caffeine and Tg, not different from the effects produced by treatment with 30 mM caffeine alone (Dunnett's multiple comparison test following one-way ANOVA). These data seemed to indicate that caffeine's stimulatory effect on the K_ATP channel was not disrupted by depletion of intracellular Ca²⁺ stores and therefore may not be dependent on an increase in the intracellular Ca²⁺ concentration. However, considering that the inhibition of ER Ca²⁺-ATPase by Tg might not completely prevent the initial, transient increase in intracellular Ca²⁺ concentration induced by 30 mM caffeine, we added another experimental group in which BAPTA/AM, a membrane-permeable and high-affinity Ca²⁺ chelator, was used to help elucidating the role of intracellular Ca²⁺ concentration in the stimulation of Kir6.2/SUR1 channels induced by caffeine. A group of HEK293 cells were pretreated with BAPTA/AM (50 µM) for at least 15 min at room temperature. Single-channel currents in cell-attached patches obtained from positively transfected HEK293 cells were then monitored in the continuous presence of BAPTA/AM (50 µM; Fig. 3A). Subsequent co-application of caffeine (30 mM) and BAPTA/AM (50 µM) to the same patch produced a smaller increase in the Kir6.2/SUR1 single-channel currents (Fig. 3A, Caffeine plus BAPTA/AM) compared with the effects caused by caffeine application alone (see Fig. 1A). The averaged normalized NPo of Kir6.2/SUR1 currents was 7.80 ± 1.68 (control as 1) (Fig. 3F; P < 0.01, 8 patches, one-sample t-test) during co-application of caffeine and BAPTA/AM, which was significantly different from the changes obtained with 30 mM caffeine applied alone (Fig. 3F; same data as shown in Fig. 1B, displayed herein for the purpose of comparison; Dunnett's multiple comparison test following one-way ANOVA). The single-channel conductance remained the same. These BAPTA/AM data indicated that intracellular Ca²⁺ was involved in mediating the stimulation of Kir6.2/SUR1 channels induced by caffeine; some Ca²⁺-dependent mechanism might be activated following the mobilization of Ca²⁺ from intracellular stores and contribute to channel stimulation during caffeine treatment.

Inhibition of PKA Did Not Affect the Stimulatory Action of Caffeine on Kir6.2/SUR1 Channels

In addition to inducing calcium release from intracellular stores, caffeine is also a potent, nonselective PDE inhibitor. Inhibition of the cAMP-specific PDE led to a mild increase of the single-channel currents of Kir6.2/SUR1 channels (see Fig. 2, C and D), presumably through activation of the cAMP-PKA signaling cascade. To determine whether the stimulation of Kir6.2/SUR1 channels induced by caffeine involved activation of endogenous PKA, we examined the effect of PKA inhibition on the stimulatory action of caffeine. Transfected HEK293 cells were treated with H-89 (0.1 µM), a specific membrane-permeable PKA inhibitor, for at least 15 min at room temperature. The cells were then transferred to the recording chamber with H-89 continuously present in the bath. The single-channel currents of Kir6.2/SUR1 channels recorded in the cell-attached configuration were markedly increased during subsequent application of caffeine (30 mM) plus H-89 to the same patch, resulting in higher opening and bursting frequencies as well as more frequent occurrence of multiple-channel openings (Fig. 3B; H-89 vs. Caffeine plus H-89). The average normalized NPo obtained from a total of 10 patches was 27.25 ± 7.61 (control as 1) (Fig. 3F; P < 0.01, one-sample t-test), which was not different from the effects obtained from cells treated with caffeine alone (Fig. 3F; same data shown in Fig. 1B, Dunnett's multiple-comparison test following one-way ANOVA). These results indicated that the stimulatory effects of caffeine on Kir6.2/SUR1 channels did not require PKA activation, suggesting that mechanisms other than cAMP-PKA signaling were responsible for the caffeine-induced K_ATP channel stimulation.

Inhibition of PKG Activity Significantly Attenuated the Stimulatory Effects of Caffeine on Kir6.2/SUR1 Channels

Accumulation of cGMP resulting from inhibition of cGMP-specific PDE may lead to activation of PKG. A potential role of PKG in mediating the caffeine-induced K_ATP channel activation was inferred based on the enhancement of Kir6.2/SUR1 channel function caused by zaprinast (see Fig. 2, B and D). PKG has been suggested to regulate K_ATP channels in heart and pancreatic -cells; however, earlier studies have yielded controversial results (25, 62). To determine whether caffeine stimulated the K_ATP channel via PKG-mediated phosphorylation, we pretreated a group of HEK293 cells with KT5823 (1–2 µM), a membrane-permeable specific PKG inhibitor, for at least 15 min at room temperature. Cells exhibiting Kir6.2/SUR1 channel activities were then recorded in the cell-attached configuration in the continuous presence of KT5823, followed by co-application of caffeine (30 mM) and KT5823. The perfusion of caffeine together with KT5823 resulted in a small yet clear activation of these channels (Fig. 3C, KT5823 vs. Caffeine plus KT5823); the normalized NPo averaged from a total of 12 patches was 4.95 ± 1.22 (control as 1) (Fig. 3F; P < 0.01, one-sample t-test). Compared with the strong activation obtained during application of caffeine alone (Fig. 3F), the stimulatory effect of caffeine on the normalized NPo of Kir6.2/SUR1 channels was significantly attenuated by KT5823 (Fig. 3F; P < 0.01, Dunnett's multiple comparison test following 1-way ANOVA). These results indicated that the stimulation of Kir6.2/SUR1 channels in intact HEK293 cells by caffeine was primarily mediated by PKG activation, possibly due to inhibition of cGMP-specific PDE and the resultant accumulation of cGMP. Nonetheless, it should be noted that a small portion of caffeine-induced K_ATP channel stimulation was sustained even in the presence of the PKG inhibitor KT5823 (Fig. 3, C and F), implying the potential involvement of some PKG-independent signaling mechanism(s) in mediating caffeine's stimulatory action on the K_ATP channel. Since inhibition of PKA did not affect the stimulatory effects of caffeine, whereas Ca²⁺ buffering did, it appeared reasonable to postulate that the Ca²⁺-dependent mechanism may mediate the PKG-independent component of the channel stimulation induced by caffeine. Would this be the case? (See below.)

Simultaneous Calcium Buffering and Inhibition of PKG Activity Completely Abrogated the Stimulation of Kir6.2/SUR1 Channels Induced by Caffeine

The observation that neither Ca²⁺ buffering by BAPTA/AM (see Fig. 3, A and F) nor PKG inhibition by KT5823 (see Fig. 3, C and F) could completely abolish the caffeine-induced K_ATP channel stimulation made us wonder whether it was the combined activation of the Ca²⁺- and PKG-dependent mechanisms that mediated the strong Kir6.2/SUR1 channel stimulation during caffeine application (Fig. 1). If this is the case, simultaneous Ca²⁺ buffering and inhibition of PKG should eliminate the stimulatory effect of caffeine. To examine this possibility, we recorded Kir6.2/SUR1 single-channel currents in the cell-attached configuration in the continuous presence of KT5823 (2 µM) and BAPTA/AM (50 µM), following the pretreatment with KT5823 and BAPTA/AM for at least 15 min at room temperature. We found that under such conditions, bath perfusion of caffeine (30 mM) did not enhance the single-channel currents of Kir6.2/SUR1 channels (Fig. 3D); the normalized NPo averaged from a total of five patches was 2.78 ± 1.24 (control as 1) (Fig. 3F; not significantly different, one-sample t-test), which was significantly different from the stimulatory effect obtained from cells treated with caffeine alone (Fig. 3F; P < 0.05, Dunnett's multiple comparison test following 1-way ANOVA). These results indicated that the stimulation of Kir6.2/SUR1 channels in intact HEK293 cells by caffeine was mediated by combined activation of PKG- and Ca²⁺-dependent signaling mechanisms. We also performed experiments in a separate group of cells in which all the potential signaling components downstream of caffeine, including PKG, PKA and Ca²⁺, were simultaneously inhibited by co-application of the highly specific PKG inhibitor KT5823 (2 µM), the selective PKA inhibitor H-89 (0.1 µM), and the Ca²⁺ chelator BAPTA/AM (50 µM). We found that caffeine (30 mM) was unable to increase the single-channel currents of Kir6.2/SUR1 channels in the continuous presence of these inhibitors, following a 15-min pretreatment at room temperature (Fig. 3E). The normalized NPo averaged from a total of four patches was 1.87 ± 0.67 (control as 1) (Fig. 3F; not significantly different, one-sample t-test); similar to what happened in the presence of KT5823 and BAPTA/AM (Fig. 3, D and F), the stimulatory effect of caffeine on the normalized NPo of Kir6.2/SUR1 channels was significantly and completely abolished in the presence of KT5823, BAPTA/AM, and H-89 (P < 0.05, Dunnett's multiple comparison test following one-way ANOVA). These results were not surprising, since H-89 itself (Fig. 3F), unlike BAPTA/AM or KT5823, was unable to alter the stimulatory effect of caffeine, and, in addition, the stimulatory action of caffeine was already diminished by the combined use of KT5823 and BAPTA/AM (Fig. 3F). On the basis of these findings, it appeared that PKG and Ca²⁺, but not PKA, were responsible for mediating the K_ATP channel stimulation induced by caffeine. In other words, our data indicated that the Ca²⁺-dependent signaling mechanism might account for the "PKG-independent" component of the caffeine-induced K_ATP channel stimulation.

Caffeine Suppressed the Single-Channel Activity of Recombinant Kir6.2/SUR1 Channels in Excised Inside-Out Membrane Patches

Our data as described thus far implied that the stimulation of Kir6.2/SUR1 channels by caffeine was a result of multiple-step signal transduction, involving PDE inhibition followed by PKG activation and phosphorylation of either the channel or other proteins that activated the channel, and mobilization of intracellular Ca²⁺ from the ER store followed by activation of some Ca²⁺-dependent mechanism that stimulated the channel. However, we could not exclude the possibility that caffeine may also modulate K_ATP channels directly. To determine whether caffeine exerted any direct modulation of K_ATP channel function, caffeine (30 mM) was applied to the cytoplasmic surface of cell-free inside-out membrane patches excised from transfected HEK293 cells expressing functional Kir6.2/SUR1 channels. The basal activity of Kir6.2/SUR1 channels in inside-out patches was generally higher than in cell-attached patches (e.g., Controls in Fig. 4A vs. Fig. 1A), owing to the relief of ATP inhibition after patch excision into the ATP-free bath. We found that application of caffeine resulted in clear inhibition of the single-channel currents of Kir6.2/SUR1 channels recorded in the inside-out patch configuration (Fig. 4A). The apparent opening frequency was reduced (Fig. 4A, Control vs. Caffeine) without changes in the single-channel conductance. The pooled normalized NPo during caffeine application was 0.38 ± 0.07 (control as 1) (see Fig. 4F, Kir6.2/SUR1; P < 0.0001, 9 patches, one-sample t-test), suggesting that caffeine may inhibit the K_ATP channel by direct interaction with the channel or some closely associated regulatory protein of the channel in addition to its stimulatory action mediated through the Ca²⁺- and PKG-dependent intracellular signaling mechanisms (see Figs. 1–3).

To ensure that the current reduction in the Kir6.2/SUR1 channel during caffeine application was due to a specific action of caffeine on the channel rather than the coincidental current rundown frequently observed in the inside-out patch recordings (in the absence of ATP), we monitored the temporal change of Kir6.2/SUR1 single-channel currents during continuous bath perfusion in a separate group of inside-out patches. There was current rundown during continuous bath perfusion (Fig. 4, B and F, Kir6.2/SUR1; P < 0.05, 4 patches, one-sample t-test) observed over a time course (8 min of continuous perfusion after obtaining the control record) as used for the experiments with caffeine application (Fig. 4A), possibly because there was no MgATP in the bath solution. However, compared with the data obtained during caffeine application from the caffeine-treated group (Fig. 4F, Kir6.2/SUR1), the normalized NPo obtained during continuous bath perfusion over the same time course from the "time control" group was significantly larger (0.67 ± 0.06) (Fig. 4F, Caffeine vs. Time control, Kir6.2/SUR1; P < 0.05, unpaired t-test), indicating that caffeine did inhibit the Kir6.2/SUR1 channel activity, which became evident in the cell-free condition. In contrast to the stimulatory action of caffeine, which required diffusible intracellular signals (Figs. 1 and 3), the inhibitory action of caffeine might result from a direct interaction between caffeine and the channel (or some closely associated regulatory protein of the channel). The bidirectional modulation of Kir6.2/SUR1 channels caused by caffeine suggests a complex model of K_ATP channel functional regulation.

Caffeine Reduced the Single-Channel Currents of Tetrameric Kir6.2 Channels Expressed in the Absence of SUR

K_ATP channels in different tissues differ in their SUR subunits (e.g., SUR1 in pancreas, SUR2A in myocardium, SUR2B in smooth muscle, and SUR1 or SUR2B in brain) but contain the same inwardly rectifying channel subunit, Kir6.2 (40). To determine whether caffeine-induced K_ATP channel modulation exhibited dependence on the pore-forming Kir6.2 and/or the regulatory SUR subunit, we examined the effect of caffeine on the function of channels formed by Kir6.2 subunits alone. Transfection with cDNAs encoding wildtype Kir6.2 in the absence of SUR fails to produce functional channels on the cell surface, because of the exposure of the ER retention/retrieval signal located in the COOH terminus of the Kir6.2 subunit (92). However, surface expression of Kir6.2 channels can be recovered after removal of the ER retention signal by alanine substitutions (i.e., RKRR368/369/370/371AAAA mutation; Ref. 92) or deletion of the COOH terminus containing the ER retention/retrieval signal (83). Kir6.2FL4A channels were expressed in HEK293 cells, and their single-channel currents were monitored in both cell-attached and inside-out patches before and during caffeine application. No functional K_ATP channels were observed in a parallel control experiment where only empty vectors were transfected (data not shown), indicating that the K_ATP channel-like currents obtained from HEK293 cells transfected with cDNAs encoding for the Kir6.2FL4A channel were specific to the heterologous expression of these channels. The single-channel opening pattern of Kir6.2FL4A channels (Fig. 4E) in the cell-attached patch configuration exhibited more singular and short-duration openings in contrast to that of Kir6.2/SUR1 channels (Figs. 1 and 2), a feature shared by the Kir6.2C36 channel (45), another Kir6.2 mutant that is capable of functional expression without the SUR subunit. Instead of evoking potent stimulation as seen on Kir6.2/SUR1 channels in cell-attached patches, bath application of caffeine (30 mM) inhibited the Kir6.2FL4A single-channel currents in the same patch configuration (Fig. 4E), resulting in significant changes in the single-channel properties (Table 1). The normalized NPo, opening frequency, corrected mean open time, and mean closed duration values were 0.40 ± 0.11 (Fig. 4F, Kir6.2FL4A; P < 0.01), 0.40 ± 0.09 (P < 0.01), 0.93 ± 0.06 (not significant), and 3.06 ± 0.50 (P < 0.05), respectively (control as 1) (Table 1; 5 patches, one-sample t-test). It was obvious that caffeine exerted only an inhibitory action on the function of K_ATP channels expressed in the absence of the SUR1 subunit. These results revealed that the SUR1 subunit was responsible for caffeine-induced K_ATP channel stimulation in intact cells, whereas the Kir6.2 subunit was accountable for the caffeine-induced K_ATP channel inhibition.

To further elucidate whether the caffeine-induced inhibition of tetrameric Kir6.2 channels was due to direct interaction between caffeine and the channel, we examined the effect of caffeine on Kir6.2FL4A channels in the inside-out patch configuration. Bath perfusion of caffeine (30 mM) to the cytoplasmic surface of inside-out membrane patches excised from transfected HEK293 cells clearly inhibited the Kir6.2FL4A single-channel currents: the apparent opening frequency was much lower in the presence than in the absence of caffeine perfusion (Fig. 4C). The averaged normalized NPo value was reduced to 0.22 ± 0.05 during caffeine application (control as 1) (Fig. 4F, Kir6.2FL4A; P < 0.0001, 7 patches, one-sample t-test), indicating that the function of the Kir6.2 subunit was suppressed by caffeine via direct interaction of caffeine with the Kir6.2 subunit or some closely associated regulatory protein(s) of the channel. In addition, we also performed the time control experiments to monitor the temporal changes of Kir6.2FL4A single-channel currents during continuous bath perfusion (Fig. 4D) in a separate group of inside-out patches (as described for Kir6.2/SUR1 channels; see Fig. 4B). The extent of current rundown was moderate (during 8 min of continuous perfusion after obtaining the control record) (Fig. 4D); the normalized NPo was 0.63 ± 0.05 (control as 1) (Fig. 4F, Kir6.2FL4A; P < 0.01, 4 patches, one-sample t-test). However, compared with the data obtained during caffeine application from the caffeine-treated group (Fig. 4, C and F, Caffeine), the normalized NPo obtained from the time control group was significantly larger over the same time course (Fig. 4F, Caffeine vs. Time control, Kir6.2FL4A; P < 0.001, unpaired t-test), supporting our working model wherein caffeine exerts an inhibitory action on the Kir6.2 subunit of the channel that requires no diffusible intracellular messengers. These data further suggest that caffeine exerts bidirectional modulation on the Kir6.2 and SUR subunits of the K_ATP channels.

Caffeine Increased the Single-Channel Activity of Recombinant Kir6.2/SUR2B Channel, a Nonvascular Smooth Muscle K_ATP Channel Isoform, Although the Effect Was Weaker Compared with Kir6.2/SUR1 Channels

It appeared that the SUR1 subunit was responsible for caffeine-induced stimulation of Kir6.2/SUR1 channels in intact cells; it was intriguing to find out whether the K_ATP channel encompassing the SUR2 subunit responded to caffeine in a similar fashion. We thus assessed the effect of caffeine on the function of recombinant K_ATP channels composed of Kir6.2 and SUR2B subunits, a K_ATP channel isoform present in the nonvascular smooth muscle and some central neurons. Caffeine application (30 mM) increased the single-channel currents of Kir6.2/SUR2B channels recorded in the cell-attached patch configuration (Fig. 5A); the normalized NPo, opening frequency, corrected mean open time, and mean closed duration values averaged from 10 patches were 7.09 ± 1.71 (Fig. 5B; P < 0.01), 5.45 ± 1.41 (P < 0.05), 1.47 ± 0.15 (not significant), and 0.29 ± 0.07 (P < 0.001), respectively (control as 1) (Table 1; 10 patches, one-sample t-test). These data indicated that Kir6.2/SUR2B channels, like Kir6.2/SUR1 channels, were stimulated by caffeine in intact cells. It is worth mentioning that, although both isoforms were positively modulated by caffeine, the enhancement of Kir6.2/SUR1 channels was significantly stronger than that of Kir6.2/SUR2B channels (Fig. 5B; P < 0.05, two-tailed unpaired t-tests). These results imply that different subtypes of SUR might confer the K_ATP channel with differential sensitivity to caffeine for channel stimulation.

Caffeine Modulated the Single-Channel Activity of Endogenous K_ATP Channels in Rat Insulinoma CRI-G1 Cells

To determine whether caffeine was capable of exerting dual modulation on the function of endogenous K_ATP channels in native cells in addition to the recombinant channels heterologously expressed in HEK293 cells (see Figs. 1 and 5), we performed single-channel recordings to monitor the effects of caffeine in an insulin-secreting pancreatic islet cell line, CRI-G1 (ECACC) (28, 78, 79), in both cell-attached and inside-out patch configurations. Pancreatic -cells, including CRI-G1 cells, express endogenous K_ATP channels composed of Kir6.2 and SUR1 subunits. The native K_ATP channels recorded in the cell-attached patch configuration from CRI-G1 cells opened at low frequencies under the control condition (Fig. 6A, Control), and they exhibited an I-V relationship of inward rectification that reversed at 0 mV (E_K) and a single-channel conductance of 70 pS at the inward direction. Application of caffeine (30 mM) markedly increased the apparent opening and bursting frequencies of these endogenous K_ATP channels in the same patches (Fig. 6A, Caffeine). The NPo values of endogenous K_ATP channels obtained in this set of experiments were compared before and during caffeine treatment in individual patches using absolute values without normalization (by paired t-test; Fig. 6B), as not all cell-attached patches obtained from these cells exhibited K_ATP channel activity in the control condition. The absolute NPo value averaged from a total of six patches was 0.08 ± 0.04% in control (Fig. 6B) and was significantly increased to 1.76 ± 0.50% during caffeine application (Fig. 6B; P < 0.01, two-tailed paired t-test). These results indicated that the endogenous K_ATP channels in intact pancreatic -cells (CRI-G1) were positively modulated by caffeine, as were the recombinant Kir6.2/SUR1 or Kir6.2/SUR2B channels expressed in transfected HEK293 cells.

Next, we applied caffeine (30 mM) to the cytoplasmic surface of inside-out membrane patches excised from insulinoma CRI-G1 cells to determine whether caffeine inhibited the activity of endogenous K_ATP channels in native cells under the cell-free condition. Because of the fact that the K_ATP channel current density was lower in native cells compared with transfected mammalian cells, the activity level after current rundown (in ATP-free condition) in inside-out patches sometimes became too low to allow reliable evaluation of any potential inhibitory effect. To circumvent this, 10 µM MgATP was included in the bath recording solution to help maintain and stabilize the K_ATP channel currents (10) in all inside-out patches obtained from CRI-G1 cells. Application of caffeine (in the continuous presence of 10 µM MgATP) resulted in marked reduction of the single-channel currents of K_ATP channels compared with the control activity acquired before caffeine application (Fig. 6C, Control vs. Caffeine); the apparent opening frequency was decreased, whereas the single-channel conductance remained unaltered. The normalized NPo during caffeine application was 0.10 ± 0.06 (control as 1) (Fig. 6F; P < 0.001, 4 patches, one-sample t-test), indicating that caffeine was capable of inhibiting endogenous K_ATP channels in insulin-secreting CRI-G1 cells in the cell-free condition, besides its stimulatory action on these K_ATP channel in intact cells (see Fig. 6, A and B).

In addition, time control experiments similar to those performed in HEK293 cells (see Fig. 4D) were carried out in CRI-G1 cells, recorded in the inside-out configuration. It appeared that K_ATP channel activity was maintained rather stable during continuous perfusion of the 10 µM MgATP-containing bath solution over the same time course as was used for the caffeine-treated group (i.e., 8 min after obtaining the control record) (Fig. 6D), resulting in an averaged normalized NPo of 1.03 ± 0.23 (control as 1) (Fig. 6F; 4 patches). Compared with the data obtained during caffeine perfusion from the caffeine-treated group (Fig. 6, C and F, Caffeine), the normalized NPo obtained from the time control group over the same time course was significantly higher (Fig. 6F, Caffeine vs. Time control; P < 0.01, unpaired t-test). These control results provide support for our findings that, aside from channel activation, caffeine also inhibits the endogenous K_ATP channel function in CRI-G1 cells; the caffeine-induced inhibition might result from an interaction between caffeine and the channel (or some closely associated regulatory protein of the channel) rather than nonspecific current fluctuation. Last, to assure the identity of these endogenous channels present in CRI-G1 cells as K_ATP channels, glibenclamide, a K_ATP channel blocker, was applied by bath perfusion in a separate group of inside-out patches. Glibenclamide (5 µM) significantly suppressed the single-channel currents with a single-channel conductance of 70 pS (Fig. 6E), resulting in an averaged normalized NPo of 0.39 ± 0.13 (Fig. 6F; 5 patches, one-sample t-test). The glibenclamide-induced suppression of K_ATP channel activity was significantly different from the time control group when the NPo values obtained over a similar time course were compared (Fig. 6F, Glibenclamide vs. Time control; P < 0.05, unpaired t-test), further confirming the identity of these endogenous channels in CRI-G1 cells as glibenclamide-sensitive K_ATP channels. Our findings thus suggest that caffeine exerts dual modulation on the function of the K_ATP channel endogenously expressed in clonal pancreatic -cells.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
K_ATP channels regulate insulin secretion, vascular tone, heart rate, and neuronal excitability by responding to transmitters as well as the internal metabolic state. These channels are also involved in cytoprotection in the brain and heart (11, 38, 50, 52, 61, 72, 77, 87, 91). K_ATP channels in native cells are composed of four pore-forming Kir6.x subunits and four regulatory SUR subunits. Although the sites responsible for ATP inhibition, pH, and phosphatidylinositol 4,5-bisphosphate (PIP₂) regulation reside on the Kir6.2 subunit, the SUR subunit of K_ATP channels appears to be the target on the channel for a few clinically important drugs, such as sulfonylureas and synthetic K_ATP channel openers. Caffeine, a naturally occurring alkaloid present in beverages such as tea and coffee, is one of the most consumed drugs. Caffeine is closely related to other methylxanthines. Its known actions include mobilization of intracellular calcium from ryanodine-sensitive stores, inhibition of PDEs, and antagonism of adenosine receptors (51). To understand how caffeine modulates the function of K_ATP channels, we first determined how caffeine changed the function of Kir6.2/SUR1 channels. We then evaluated the involvement of several potential signaling mechanisms in mediating the caffeine-induced K_ATP channel modulation, including inhibition of PDEs, release of intracellular Ca²⁺, and activation of protein kinases. We also compared the effects of caffeine on different K_ATP channel isoforms, including tetrameric Kir6.2 and octameric Kir6.2/SUR1 and Kir6.2/SUR2B channels, to define the respective roles played by the Kir6.2 and SUR subunits in caffeine-induced K_ATP channel modulation. Last, we determined whether caffeine exerted dual modulation on the function of native K_ATP channels expressed in rat insulinoma CRI-G1 cells.

Here we report that caffeine activated the Kir6.2/SUR1 channels in intact cells and that the stimulation of Kir6.2/SUR1 channels evoked by caffeine was mimicked by IBMX, a nonselective PDE inhibitor, and by zaprinast, a selective inhibitor of cGMP-specific PDE-5 and PDE-9. Ro-20-1724, a selective inhibitor of cAMP-specific PDE-4 inhibitor, was able to mildly activate Kir6.2/SUR1 channels; however, co-application of caffeine and H-89, a selective PKA inhibitor, did not abrogate caffeine-induced K_ATP channel stimulation. In contrast, application of caffeine in the presence of a membrane-permeable Ca²⁺ chelator, BAPTA/AM, or of a selective PKG inhibitor, KT5823, resulted in marked attenuation of the stimulatory effects of caffeine. Furthermore, complete abolishment of caffeine-induced stimulation of Kir6.2/SUR1 channels was achieved when inhibition of PKG and buffering of Ca²⁺ were administered simultaneously, regardless of whether PKA activity was inhibited. Interestingly, although caffeine stimulated Kir6.2/SUR1 channels in cell-attached patches, it suppressed these channels in excised inside-out patches, indicating that caffeine may in fact inhibit the Kir6.2/SUR1 channel in the absence of intracellular signaling. Furthermore, the single-channel currents of Kir6.2FL4A channels (expressed in the absence of the SUR subunit) were inhibited by caffeine in both intact cells and cell-free membrane patches. Caffeine also stimulated recombinant Kir6.2/SUR2B channels in intact cells, but the enhancement was significantly weaker than that of Kir6.2/SUR1 channels. Moreover, caffeine was also capable of exerting dual modulation on the function of endogenous K_ATP channels present in rat insulinoma CRI-G1 cells. Taken together, these data suggest that the inhibitory effect of caffeine on K_ATP channels may involve direct interaction of caffeine with the Kir6.2 subunit or some closely associated regulatory protein, while the stimulatory effect may be mediated through PDE inhibition, which consequently activates PKG as well as intracellular Ca²⁺ mobilization that triggers Ca²⁺-dependent signaling, and the intracellular signaling may use the SUR subunit as a final target. In the presence of both Kir6.2 and SUR subunits, the stimulatory effect of caffeine appears to surpass and mask its inhibitory one on K_ATP channels in intact cells. Our data also revealed that SUR1-containing K_ATP channels may be more sensitive than the SUR2B-containing channels to caffeine for functional stimulation.

K_ATP Channel Stimulation by Caffeine Relies on Intracellular Signaling

In this report, we have presented the first piece of evidence that caffeine is a potent K_ATP channel activator; this novel finding was obtained by performing mechanistic dissection of signaling pathways in intact HEK293 cells, using cell-attached patch recordings (Figs. 1, 3 and 5), and was confirmed in insulin-secreting pancreatic islet CRI-G1 cells (Fig. 6, A and B). Several groups have reported an inhibitory role of millimolar concentrations of caffeine on K_ATP channel activity in pancreatic -cells (36) and smooth muscle cells (80). As mentioned earlier, the intracellular signaling components may have been lost in some of the earlier studies investigating how caffeine modulates K_ATP channels, because of the use of cell-free patch configurations for recordings (36); in other cases, the single-channel conductance of the caffeine-sensitive channel in urethral muscle cells was too small for Kir6.2/SUR2 channels (80). The use of single-channel recordings in the cell-attached patch configuration allowed us to examine the effects of caffeine on the K_ATP channel with the potential intracellular signaling components preserved. A calcium-free bath solution was used throughout the present study to prevent Ca²⁺ store refilling and interference possibly arising from prolonged elevation of the intracellular Ca²⁺ level, as caffeine is known to induce Ca²⁺ mobilization from intracellular stores. Nevertheless, it should be pointed out that the enhancement of Kir6.2/SUR1 channel activity by millimolar concentrations of caffeine did occur in both Ca²⁺-free (Fig. 1) and Ca²⁺-containing bath solutions (data not shown). It should also be noted that there were no K_ATP channel-like currents observed in HEK293 cells when cDNAs encoding for the wild-type Kir6.2 or the SUR1 subunit were transfected alone, or when empty vectors were transfected (data not shown), indicating the absence of detectable endogenous K_ATP channel subunits in these cells. Here we demonstrated that caffeine activated Kir6.2/SUR1 channels only in intact cells (Fig. 1; Table 1) but not in cell-free membrane patches (Fig. 4, A, B, and F), suggesting that caffeine-induced K_ATP channel activation is mediated by some intracellular signaling mechanism involving cytosolic messengers rather than by a direct interaction of caffeine with the channel.

Caffeine-Induced K_ATP Channel Stimulation Involves PDE Inhibition

In the present study, we showed that caffeine significantly increased the single-channel currents of recombinant as well as native K_ATP channels in intact cells (Figs. 1, 5, and 6). Caffeine is a potent, nonselective PDE inhibitor, and accumulation of cyclic nucleotides induced by caffeine has been demonstrated in several cell types (46, 63). The PDE superfamily encompasses 11 different families, classified based on their substrates and inhibitor profiles as well as structural characteristics (76). Inhibition of PDE leads to accumulation of intracellular cyclic nucleotides. It has been shown that inhibition of PDE mimics the effect of caffeine on the agonist-evoked Ca²⁺ response in pancreatic acinar cells (12) and on the contractility of gallbladder muscle (46). In the present study, we showed that several nonselective as well as cGMP- or cAMP-specific PDE inhibitors were capable of reproducing the caffeine-induced enhancement of Kir6.2/SUR1 single-channel currents in intact HEK293 cells (Fig. 2), suggesting that the stimulatory effect of caffeine on K_ATP channels may be mediated by PDE inhibition. It is conceivable that multiple PDEs are present and active in HEK293 cells; therefore, changes in PDE activities may be transduced to different levels of K_ATP channel activity through altering intracellular cyclic nucleotide concentrations and the consequent cyclic nucleotide-dependent protein phosphorylation. Other cell types, such as hippocampal cells (84) and pancreatic -cells (27), also possess high PDE activity, and thus manipulating PDE activity with specific agents may provide a useful way to modulate K_ATP channel function in intact cells. Caffeine also acts as a nonselective, competitive adenosine receptor antagonist (21, 22; reviews). In the present study, we did not pursue adenosine receptor antagonism as one probable mechanism underlying caffeine's stimulatory action on K_ATP channels. Activation of A_2A receptors dilates retinal arterioles (30) or cerebral arterioles (57; review) by opening rather than inhibiting K_ATP channels. It is therefore unlikely that inhibition of G_s-coupled A_2A receptor-induced signaling by caffeine can account for K_ATP channel stimulation, even though some endogenous adenosine receptors may be present in HEK293 cells (A_2B type; Ref. 18).

Caffeine-Induced K_ATP Channel Stimulation is in Part Mediated by Intracellular Ca²⁺

In addition to acting as a nonselective PDE inhibitor, caffeine also induces release of calcium from intracellular stores through activation of the RyR (14, 64) and subsequently causes an increase in the intracellular Ca²⁺ concentration. Even though a calcium-free bath was used throughout the present study, significant enhancement of Kir6.2/SUR1 channel activity (depicted as the normalized NPo in all bar graphs throughout) was brought about by the application of caffeine in both intact HEK293 and insulin-secreting CRI-G1 cells (Figs. 1 and 6, A and B). It is possible that the stimulatory action of caffeine on the activity of Kir6.2/SUR1 channels may result, at least in part, from an increase in the intracellular Ca²⁺ concentration. Indeed, the stimulatory effects of caffeine on K_ATP channels were attenuated when intracellular Ca²⁺ was buffered with a membrane-permeable Ca²⁺ chelator, BAPTA/AM (Fig. 3, A and F). It should be noted that BAPTA/AM did not completely abolish the stimulatory effects of caffeine on Kir6.2/SUR1 channels (Fig. 3, A and F); the residual enhancement of the Kir6.2/SUR1 channel activity remained significant (Fig. 3F). Our data thus suggest that the stimulation of K_ATP channel function by caffeine involves in part some Ca²⁺-dependent mechanism, presumably triggered by the release of Ca²⁺ from intracellular stores. On the other hand, selective inhibition of ER Ca²⁺-ATPase with Tg (0.1–0.2 µM) did not change the magnitude of caffeine-induced K_ATP channel stimulation (see Buffering of Intracellular Ca²⁺ by BAPTA/AM Attenuated the Stimulation of Kir6.2/SUR1 Channels Induced by Caffeine). On the surface, the Tg data seem to contradict the BAPTA/AM results; however, it is more likely that prolonged inhibition of ER Ca²⁺-ATPase with Tg (which suppresses Ca²⁺ uptake by ER) simply failed to prevent the initial release of Ca²⁺ from ER stores (and hence an increase in the intracellular Ca²⁺ level) during high-concentration caffeine application. A role of Ca²⁺ ions in modulating K_ATP channels has been suggested (20, 29, 42). Although most of the Ca²⁺ effects on the K_ATP channel described thus far have been limited to the inhibitory type, a positive effect has been implicated in frog skeletal muscle (42). In line with this, our present study suggests that the K_ATP channel is stimulated by caffeine, partially via activation of some Ca²⁺-dependent process. The detailed mechanism underlying Ca²⁺-dependent modulation of the K_ATP channel, which appears to constitute the BAPTA/AM-sensitive component of the caffeine-induced K_ATP channel stimulation, requires further investigation.

Caffeine Stimulates K_ATP Channels Partially via the cGMP-PKG Signaling Mechanism

Accumulation of cGMP by inhibition of cGMP-hydrolyzing PDE activity can lead to activation of cGMP-dependent effector systems. In intact cells, cGMP effector systems include the protein families of PKGs, cGMP-sensitive PDEs, and cGMP-gated cation channels. In the present study, we showed that the stimulation of Kir6.2/SUR1 channels by caffeine was reproduced by inhibition of cGMP-specific PDEs (i.e., PDE-5 and PDE-9) with zaprinast (Fig. 2, B and D) and that inhibition of PKG with KT5823 during caffeine treatment significantly attenuated the extent of caffeine-induced Kir6.2/SUR1 channel stimulation (Fig. 3, C and F). Our data thus suggest a role of the cGMP-PKG signaling pathway, in addition to the Ca²⁺-dependent mechanism as described above, in mediating K_ATP channel stimulation evoked by caffeine in intact cells. PKG may enhance the K_ATP channel function via direct phosphorylation of the K_ATP channel or, alternatively, via phosphorylation of other intracellular proteins that regulate K_ATP channel stimulation. Previous studies investigating how PKG modulates K_ATP channel function have yielded contradictory results. It has been shown that PKG mediates the inhibition of K_ATP channels induced by 17-estradiol in intact mouse pancreatic -cells (62), whereas direct PKG phosphorylation enhances K_ATP channel activity elicited by pinacidil (a K_ATP channel opener) in rabbit ventricular myocytes (25). On the other hand, evidence obtained in the present study suggests that PKG activation mediates in part the K_ATP channel stimulation induced by caffeine in intact cells (Fig. 3). Intriguingly, although K_ATP channel activity was enhanced by cGMP-PKG signaling in intact cells (Figs. 2B and 3), we found that direct application of purified PKG (in the presence of MgATP and cGMP) to the cytoplasmic surface of excised inside-out patches failed to stimulate the K_ATP channel (Y. Chai and Y.-F. Lin, unpublished data). Our data thus suggest that caffeine stimulates the K_ATP channel partially through PKG activation; PKG seems to activate the K_ATP channel via indirect phosphorylation, accomplished by phosphorylating intracellular protein(s) rather than the channel per se.

Our data also revealed that inhibition of PKG was able to significantly attenuate, but not completely abolish, the stimulatory effects of caffeine on Kir6.2/SUR1 channels in intact cells (Fig. 3, C and F); a small yet significant increase of Kir6.2/SUR1 channel activity was sustained (Fig. 3F). These results suggest that PKG partially, rather than exclusively, mediates the stimulatory effects of caffeine. The same is applicable in understanding the role of intracellular calcium in mediating caffeine's stimulatory action on the K_ATP channel: calcium buffering also caused incomplete nullification of caffeine-induced K_ATP channel stimulation (Fig. 3, A and F). One point to make here is that the attenuation of caffeine-induced K_ATP channel stimulation was more pronounced when PKG activity was inhibited than when intracellular Ca²⁺ was buffered (Fig. 3F; Dunnett's multiple comparison test), implying differential contributions of PKG and intracellular Ca²⁺ in mediating caffeine-induced K_ATP channel stimulation. Moreover, complete abolishment of caffeine-induced K_ATP channel stimulation was achieved when inhibition of PKG and intracellular Ca²⁺ buffering were simultaneously effected (Fig. 3, D and F), suggesting that cGMP-PKG, together with Ca²⁺-dependent signaling mechanisms, accounts for the functional enhancement of K_ATP channels induced by caffeine. If this is the case, what role would cAMP or PKA play in mediating the stimulatory action of caffeine? (See below.)

Caffeine-Induced K_ATP Channel Stimulation Does Not Require the cAMP-PKA Signaling Pathway

Our observation that inhibition of a cAMP-specific PDE-4 resulted in a mild yet significant increase of Kir6.2/SUR1 single-channel currents (Fig. 2, C and D) suggests that the K_ATP channel is activated by cAMP-dependent signaling, which is in agreement with earlier findings by others and us (10, 45) that PKA-mediated protein phosphorylation modulates K_ATP channel function by direct phosphorylation of the channel protein. However, inhibition of PKA activity during caffeine application did not attenuate the magnitude of caffeine-induced enhancement of K_ATP channel currents (Fig. 3, B and F). In addition, complete abrogation of caffeine-induced K_ATP channel stimulation was achieved by simultaneous Ca²⁺ buffering and PKG inhibition (Fig. 3, D and F), regardless of whether PKA activity was inhibited (Fig. 3, E and F). These results thus suggest that K_ATP channel stimulation induced by caffeine may not involve the cAMP-PKA signaling pathway at all, or, alternatively, the contribution of cAMP-PKA signaling may be negligible compared with the PKG- or calcium-dependent mechanism in mediating the stimulatory action of caffeine on the K_ATP channel in intact cells.

Stimulation of the K_ATP Channel Induced by Caffeine is SUR-Dependent

Our present study suggests that both PKG- and Ca²⁺-dependent signaling mechanisms may be triggered by caffeine to mediate (octameric) K_ATP channel activation. Unlike Kir6.2/SUR1 channels (Fig. 1; Table 1), the tetrameric Kir6.2FL4A channel was not stimulated by caffeine in intact cells (Fig. 5, B and C; Table 1), implying, first, that caffeine-induced stimulation of octameric K_ATP channels is SUR dependent and, second, that the SUR subunit of K_ATP channels may be the final target protein for the aforementioned intracellular signaling induced by caffeine. In addition, the SUR2B-containing K_ATP channels exhibited milder stimulation with caffeine treatment than did the SUR1-containing channels in intact cells (Figs. 1 and 5; Table 1), suggesting that the various SUR subtypes may confer the K_ATP channel with different sensitivities to the stimulation induced by caffeine. K_ATP channels containing different SUR subunits have been shown to exhibit different sensitivity to compounds such as synthetic K_ATP channel openers, sulfonylureas, and ATP (4, 8, 38, 52, 70, 83). In line with this, the differential responsiveness to caffeine among Kir6.2/SUR1, Kir6.2/SUR2B, and Kir6.2FL4A channels may be accounted for by similar "SUR dependence."

Caffeine Inhibits the K_ATP Channel by Interacting with the Kir6.2 Subunit

Aside from the SUR subunit-dependent stimulatory action mediated through the intracellular signaling in intact cells (Figs. 1–3), here we also demonstrated that caffeine inhibited Kir6.2/SUR1 channel activity in excised inside-out patches (Fig. 4, A, B, and F). The inhibitory effect of caffeine on Kir6.2/SUR1 channels revealed in inside-out patches suggests that caffeine may inhibit the function of K_ATP channels in the absence of intracellular signaling. Moreover, although caffeine significantly inhibited octameric Kir6.2/SUR1 K_ATP channels only in cell-free patches, it suppressed tetrameric Kir6.2FL4A channels in both intact cells and cell-free patches (Fig. 4, C–F; Table 1). Our data thus suggest that caffeine functionally inhibits the K_ATP channel by interacting directly with the Kir6.2 subunit or some closely associated regulatory protein(s) of the channel, a process that requires neither the SUR subunit nor intracellular signaling. In line with this, Teramoto et al. (80) have demonstrated that caffeine at millimolar concentrations inhibits Kir6.2C36 channels in COS7 cells. We thus suggest that caffeine produces dual modulation of the function of K_ATP channels in intact cells: a strong potentiation through (indirect) interaction with SUR and a relatively weak inhibition through (direct) interaction with Kir6.2. It is conceivable that caffeine may also exert inhibitory action on octameric K_ATP channels via interaction with the Kir6.2 subunit (or some closely associated regulatory protein) in intact cells; however, the Kir6.2-dependent inhibitory action of caffeine appears to be much weaker than, and is therefore masked by, the SUR-dependent stimulatory action (Table 1).

Effects of Caffeine on the Activity of K_ATP Channels Are Due to Changes in the Single-Channel Opening and Closing Properties

Channel function and its modulation are determined by the conformational changes that the channel undergoes to enable opening or closing of the ion-permeating pore. Here we showed that caffeine stimulated the octameric K_ATP channel in intact cells (Figs. 1, 5, and 6; Table 1). The enhancement of K_ATP single-channel currents was due to changes in the single-channel properties, including increases in NPo and opening frequency and reduction of mean closed duration, yet neither the single-channel conductance nor the corrected mean open time was altered (Table 1). In contrast, caffeine inhibited tetrameric Kir6.2FL4A channels in both intact cells and cell-free patches (see above; Fig. 4, C–F; Table 1). Analysis of the single-channel properties of Kir6.2FL4A channels in intact cells revealed that caffeine decreased NPo and opening frequency and increased mean closed duration, without changing the single-channel conductance or the corrected mean open time (Table 1). The opposite functional modulation exerted by caffeine on the octameric and tetrameric K_ATP channels (Figs. 1, 4–6; Table 1) is reflected in the very distinctive changes of the single-channel properties, which may be isoform/subunit- dependent. We suggest that caffeine exerts stimulatory effects on octameric K_ATP channels in intact cells by destabilizing the closed states and by increasing the frequency of closed-to-open transitions, whereas it interacts with and suppresses the function of the pore-forming subunit Kir6.2 by stabilizing the closed states and reducing the transitions to opening. As discussed above, the stimulation of SUR-containing octameric K_ATP channels in intact cells represents a summated outcome of the bidirectional modulatory actions of caffeine, where channel stimulation greatly exceeds inhibition (Figs. 1, 5, and 6; Table 1).

Caffeine Exerts Dual Modulation on the K_ATP Channel Present in Insulin-Secreting Pancreatic -Cells

The dual modulation of the K_ATP channel function caused by caffeine was uncovered not only in the heterologous expression system (Figs. 1–5) but also in clonal pancreatic islet CRI-G1 cells expressing endogenous K_ATP channels (Fig. 6). The endogenous K_ATP-like channels in CRI-G1 cells exhibited single-channel conductance characteristic of the K_ATP channel (70 pS; Fig. 6A) and were sensitive to the K_ATP channel antagonist glibenclamide (Fig. 6, E and F). These channels were activated by caffeine in intact cells (Fig. 6, A and B), and their currents were suppressed by caffeine in inside-out patches (Fig. 6, C, D, and F). The K_ATP channel expressed in pancreatic -cells is of the Kir6.2/SUR1 isoform (1), which was also the major recombinant K_ATP channel type studied here (Figs. 1–3 and 4, A and B). The dual modulation of the endogenous K_ATP channels in CRI-G1 cells by caffeine, including activation in intact cells and inhibition in excised membrane patches, was similar to that observed on the recombinant Kir6.2/SUR1 channels in transfected HEK293 cells (Figs. 1 and 4). The underlying regulatory mechanisms may include the aforementioned cGMP-PKG and Ca²⁺-dependent signaling through interaction with the SUR subunit for K_ATP channel activation, plus a direct interaction with the pore-forming subunit Kir6.2 (or some closely associated regulatory protein of the channel) for K_ATP channel inhibition. The inhibitory effect of caffeine appears to involve no diffusible messengers, as the effect was present in cell-free membrane patches in both HEK293 and CRI-G1 cells. The presence of dual functional modulation of K_ATP channels by caffeine in different cell types suggests that caffeine activates some common regulatory mechanisms to alter the function of K_ATP channels.

Caffeine and Ion Channel Modulation

In the present study, we showed that caffeine stimulated Kir6.2/SUR1 and Kir6.2/SUR2B but inhibited Kir6.2FL4A channels in intact HEK293 cells (Figs. 1, 4, E and F, and 5; Table 1); moreover, caffeine stimulated the native K_ATP channels in intact insulin-secreting CRI-G1 cells (Fig. 6, A and B). Caffeine has been demonstrated to directly inhibit potassium channels and calcium channels in a number of cells, including mammalian ventricular myocytes (15, 19, 67, 85), guinea pig vascular smooth muscle (54), pig urethral smooth muscle cells (80), dissociated chick autonomic ganglion neurons and pineal cells (59), taste receptor cells (93), rat anterior pituitary cells (41), and outer hair cells of guinea pig cochlea (90). The channels previously reported to be suppressed by caffeine include TREK-1 channels (26), hERG channels (17), K_Ca channels (41), and K_ATP channels (36) plus outwardly and inwardly rectifying potassium and calcium currents in taste receptor cells (93). Among these channels, direct block of hERG channels (17), K_ATP channels (36), delayed rectifier channels (59), (I_A-like) transient outward potassium channels (54, 59), some inwardly rectifying (9) and outward potassium currents (15, 19, 85), and L-type calcium channels (41) by caffeine has been suggested. In contrast, several studies have suggested that caffeine activates the intermediate-conductance K_Ca (IK) channel currents (but not large- or small-conductance K_Ca channels) in transfected HEK293 cells (68) and K_Ca channels in guinea pig adrenal chromaffin cells (55) and increases potassium efflux in frog skeletal muscle (86). Here we report that caffeine exerts both activation (that is SUR-dependent and requires cGMP-PKG and Ca²⁺ signaling; Figs. 1, 3, 5, and 6) and inhibition (that is Kir6.2-dependent and requires no intracellular signaling; Fig. 4) of K_ATP channel function. The caffeine-induced stimulation (Figs. 1, 5, and 6) appears to predominate over the direct inhibition of Kir6.2 by caffeine (Fig. 4). Therefore, the bidirectional regulation of octameric K_ATP channels induced by caffeine summates to strong channel stimulation in intact cells (Figs. 1, 5, and 6). Our results suggest that the presence of Kir6.2 and SUR subunits in functional K_ATP channels may confer on the channel higher fine-tuning capability, as the two subunits can be differentially (or bidirectionally) modulated by the same agent, via distinct mechanisms. The regulatory mechanism uncovered in this study may be of general interest in understanding how the K_ATP channel is modulated under physiological and pathophysiological conditions.

In conclusion, in the present study, we investigated how caffeine modulates the function of K_ATP channels in transfected HEK293 cells and in rat insulinoma CRI-G1 cells. Here we report that caffeine exerts dual regulation on the function of K_ATP channels and that the caffeine-induced stimulation predominates over its inhibitory action in intact cells. The regulatory subunit SUR and intracellular signaling were both required for the caffeine-induced K_ATP channel stimulation, which represents a novel effect of caffeine. In contrast, a direct action on the pore-forming subunit Kir6.2 or some closely associated protein(s) was responsible for the channel inhibition by caffeine. Further examination of the signaling mechanisms underlying caffeine-induced K_ATP channel stimulation revealed that it involved not only intracellular Ca²⁺ mobilization but also the activation of PKG, and the effect could be mimicked by inhibition of nonselective or cGMP-specific PDEs. In contrast, although inhibition of cAMP-specific PDE mildly enhanced Kir6.2/SUR1 channel currents, inhibition of PKA during caffeine treatment did not affect the stimulatory action of caffeine. In fact, the stimulatory effects of caffeine were completely abolished by simultaneous PKG inhibition and Ca²⁺ buffering, suggesting that caffeine stimulates K_ATP channels through cGMP-PKG signaling and intracellular Ca²⁺ mobilization but requires no PKA activation following accumulation of cAMP. Moreover, the Kir6.2/SUR2B channel, a nonvascular smooth muscle K_ATP channel isoform, was also stimulated by caffeine, but its sensitivity to caffeine seems much lower than that of the Kir6.2/SUR1 channel. We thus suggest that the Kir6.2 and SUR subunits of K_ATP channels are bidirectionally modulated by caffeine: the SUR subunit may serve as the target for caffeine-mediated K_ATP channel stimulation and the common pore-forming subunit Kir6.2 (or some closely associated regulatory protein) as the target for channel inhibition by caffeine. The presence of dual functional modulation of K_ATP channels by caffeine in different cell types implies that the signaling processes elucidated in the present study may represent common mechanisms important for K_ATP channel regulation. The keen sensitivity of plasma-membrane K_ATP channels to PDE inhibition may provide another level of therapeutic intervention, particularly since PDE-5 inhibition has been suggested as a cardiac preconditioning strategy (43). Interestingly, chronic use of caffeine at low dosage has been reported to be neuroprotective (60, 69). Since the K_ATP channel functions as an important player in cytoprotection in the heart and brain, and these channels can be potently stimulated by caffeine, the neuroprotective effect of chronic caffeine use may, at least in part, involve K_ATP channel activation. Last, since caffeine is widely used as an experimental tool to study intracellular calcium dynamics and Ca²⁺-dependent transduction pathways, the effects of caffeine on K_ATP channels should be taken into consideration in the interpretation of experiments using caffeine in cell preparations that express K_ATP channels.

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
This study was supported by University of California-Davis (UCD) seed fund (Y.-F. Lin), UCD Health System Research Award (Y.-F. Lin), and Harrison Endowed Chair in Diabetes Research Award (Y.-F. Lin) and was conducted in a facility constructed with support from Research Facilities Improvement Program Grant C06-RR-12088-01 from the National Center for Research Resources.

	ACKNOWLEDGMENTS

 
We thank Dr. Susumu Seino (Kobe University, Chuo-ku, Japan) and Dr. Joseph Bryan (Baylor College of Medicine, Houston, TX) for the kind gifts of cDNA clones of hamster SUR1, rat SUR2B, and mouse Kir6.2. The cDNA of Kir6.2FL4A was a kind gift from Dr. Blanche Schwappach (University of Heidelberg, Heidelberg, Germany). The single-channel analysis program Intrv5 was generously provided by Drs. Barry Pallotta (University of North Carolina, Chapel Hill, NC) and Janet Fisher (University of South Carolina, Columbia, SC).

	FOOTNOTES

 

Address for reprint requests and other correspondence: Y.-F. Lin, Dept. of Physiology and Membrane Biology, Univ. of California-Davis, Rm. 4144, Tupper Hall, One Shields Ave., Davis, CA 95616 (e-mail: yflin{at}ucdavis.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
1. Aguilar-Bryan L, Bryan J. Molecular biology of adenosine triphosphate-sensitive potassium channels. Endocr Rev 20: 101–135, 1999.[Abstract/Free Full Text]

2. Aguilar-Bryan L, Clement JP 4th, Gonzalez G, Kunjilwar K, Babenko A, Bryan J. Toward understanding the assembly and structure of K_ATP channels. Physiol Rev 78: 227–245, 1998.[Abstract/Free Full Text]

3. Aguilar-Bryan L, Nichols CG, Wechsler SW, Clement JP 4th, Boyd AE 3rd, Gonzalez G, Herrera-Sosa H, Nguy K, Bryan J, Nelson DA. Cloning of the -cell high-affinity sulphonylurea receptor: a regulator of insulin secretion. Science 268: 423–426, 1995.[Abstract/Free Full Text]

4. Ashcroft FM. Mechanisms of the glycaemic effects of sulfonylureas. Horm Metab Res 28: 456–463, 1996.[Web of Science][Medline]

5. Ashcroft FM, Gribble FM. Correlating structure and function in ATP-sensitive K⁺ channels. Trends Neurosci 21: 288–294, 1998.[CrossRef][Web of Science][Medline]

6. Ashcroft FM, Rorsman P. Electrophysiology of the pancreatic -cell. Prog Biophys Mol Biol 54: 87–143, 1989.[CrossRef][Medline]

7. Ashford ML, Boden PR, Treherne JM. Glucose-induced excitation of hypothalamic neurones is mediated by ATP-sensitive K⁺ channels. Pflügers Arch 415: 479–483, 1990.[CrossRef][Web of Science][Medline]

8. Babenko AP, Aguilar-Bryan L, Bryan J. A view of SUR/Kir6.X, K_ATP channels. Annu Rev Physiol 60: 667–687, 1998.[CrossRef][Web of Science][Medline]

9. Barros F, del Camino D, Pardo LA, de la Pena P. Caffeine enhancement of electrical activity through direct blockade of inward rectifying K⁺ currents in GH3 rat anterior pituitary cells. Pflügers Arch 431: 443–451, 1996.[CrossRef][Web of Science][Medline]

10. Beguin P, Nagashima K, Nishimura M, Gonoi T, Seino S. PKA-mediated phosphorylation of the human K_ATP channel: separate roles of Kir6.2 and SUR1 subunit phosphorylation. EMBO J 18: 4722–4732, 1999.[CrossRef][Web of Science][Medline]

11. Brown DA, Chicco AJ, Jew KN, Johnson MS, Lynch JM, Watson PA, Moore RL. Cardioprotection afforded by chronic exercise is mediated by the sarcolemmal, and not the mitochondrial, isoform of the K_ATP channel in the rat. J Physiol 569: 913–924, 2005.[Abstract/Free Full Text]

12. Camello PJ, Petersen OH, Toescu EC. Simultaneous presence of cAMP and cGMP exert a coordinated inhibitory effect on the agonist-evoked Ca²⁺ signal in pancreatic acinar cells. Pflügers Arch 432: 775–781, 1996.[CrossRef][Web of Science][Medline]

13. Carrington CA, Rubery ED, Pearson EC, Hales CN. Five new insulin-producing cell lines with differing secretory properties. J Endocrinol 109: 193–200, 1986.[Abstract/Free Full Text]

14. Cheek TR, Moreton RB, Berridge MJ, Stauderman KA, Murawsky MM, Bootman MD. Quantal Ca²⁺ release from caffeine-sensitive stores in adrenal chromaffin cells. J Biol Chem 268: 27076–27083, 1993.[Abstract/Free Full Text]

15. Chorvatova A, Hussain M. Effects of caffeine on potassium currents in isolated rat ventricular myocytes. Pflügers Arch 446: 422–428, 2003.[CrossRef][Web of Science][Medline]

16. Clement JP 4th, Kunjilwar K, Gonzalez G, Schwanstecher M, Panten U, Aguilar-Bryan L, Bryan J. Association and stoichiometry of K_ATP channel subunits. Neuron 18: 827–838, 1997.[CrossRef][Web of Science][Medline]

17. Cockerill SL, Mitcheson JS. Direct block of human ether-a-go-go-related gene potassium channels by caffeine. J Pharmacol Exp Ther 316: 860–868, 2006.[Abstract/Free Full Text]

18. Cooper J, Hill SJ, Alexander SP. An endogenous A2B adenosine receptor coupled to cyclic AMP generation in human embryonic kidney (HEK 293) cells. Br J Pharmacol 122: 546–550, 1997.[CrossRef][Web of Science][Medline]

19. Cui Y, Terrar DA. Effects of caffeine on background potassium current in isolated guinea pig ventricular myocytes. J Cardiovasc Pharmacol 25: 691–695, 2005.

20. Findlay I. Calcium-dependent inactivation of the ATP-sensitive K⁺ channel of rat ventricular myocytes. Biochim Biophys Acta 943: 297–304, 1988.[Medline]

21. Fisone G, Borgkvist A, Usiello A. Caffeine as a psychomotor stimulant: mechanism of action. Cell Mol Life Sci 61: 857–872, 2004.[CrossRef][Web of Science][Medline]

22. Fredholm BB, Battig K, Holmen J, Nehlig A, Zvartau EE. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 51: 83–133, 1999.[Abstract/Free Full Text]

23. Gribble FM, Tucker SJ, Ashcroft FM. The essential role of the Walker A motifs of SUR1 in K-ATP channel activation by Mg-ADP and diazoxide. EMBO J 16: 1145–1152, 1997.[CrossRef][Web of Science][Medline]

24. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflügers Arch 391: 85–100, 1981.[CrossRef][Web of Science][Medline]

25. Han J, Kim N, Kim E, Ho WK, Earm YE. Modulation of ATP-sensitive potassium channels by cGMP-dependent protein kinase in rabbit ventricular myocytes. J Biol Chem 276: 22140–22147, 2001.[Abstract/Free Full Text]

26. Harinath S, Sikdar SK. Inhibition of human TREK-1 channels by caffeine and theophylline. Epilepsy Res 64: 127–135, 2005.[CrossRef][Web of Science][Medline]

27. Harndahl L, Jing XJ, Ivarsson R, Degerman E, Ahren B, Manganiello VC, Renstrom E, Holst LS. Important role of phosphodiesterase 3B for the stimulatory action of cAMP on pancreatic beta-cell exocytosis and release of insulin. J Biol Chem 277: 37446–37455, 2002.[Abstract/Free Full Text]

28. Harvey J, Hardy SC, Ashford ML. Dual actions of the metabolic inhibitor, sodium azide on K_ATP channel currents in the rat CRI-G1 insulinoma cell line. Br J Pharmacol 126: 51–60, 1999.[CrossRef][Web of Science][Medline]

29. Hehl S, Moser C, Weik R, Neumcke B. Internal Ca²⁺ ions inactivate and modify ATP-sensitive potassium channels in adult mouse skeletal muscle. Biochim Biophys Acta 1190: 257–263, 1994.[Medline]

30. Hein TW, Yuan Z, Rosa RH Jr, Kuo L. Requisite roles of A_2A receptors, nitric oxide, and K_ATP channels in retinal arteriolar dilation in response to adenosine. Invest Ophthalmol Vis Sci 46: 2113–2119, 2005.[Abstract/Free Full Text]

31. Horn R. Statistical methods for model discrimination. Applications to gating kinetics and permeation of the acetylcholine receptor channel. Biophys J 51: 255–263, 1987.[Web of Science][Medline]

32. Hu K, Li GR, Nattel S. Adenosine-induced activation of ATP-sensitive K⁺ channels in excised membrane patches is mediated by PKC. Am J Physiol Heart Circ Physiol 276: H488–H495, 1999.[Abstract/Free Full Text]

33. Inagaki N, Gonoi T, Clement JP 4th, Namba N, Inazawa J, Gonzalez G, Aguilar-Bryan L, Seino S, Bryan J. Reconstitution of I_KATP: an inward rectifier subunit plus the sulphonylurea receptor. Science 270: 1166–1169, 1995.[Abstract/Free Full Text]

34. Inagaki N, Gonoi T, Clement JP 4th, Wang CZ, Aguilar-Bryan L, Bryan J, Seino S. A family of sulfonylurea receptors determines the pharmacological properties of ATP-sensitive K⁺ channels. Neuron 16: 1011–1017, 1996.[CrossRef][Web of Science][Medline]

35. Inagaki N, Gonoi T, Seino S. Subunit stoichiometry of the pancreatic -cell ATP-sensitive K⁺ channel. FEBS Lett 409: 232–236, 1997.[CrossRef][Web of Science][Medline]

36. Islam MS, Larsson O, Nilsson T, Berggren PO. Effects of caffeine on cytoplasmic free Ca²⁺ concentration in pancreatic -cells are mediated by interaction with ATP-sensitive K⁺ channels and L-type voltage-gated Ca²⁺ channels but not the ryanodine receptor. Biochem J 306: 679–686, 1995.[Web of Science][Medline]

37. Isomoto S, Kondo C, Yamada M, Matsumoto S, Higashiguchi O, Horio Y, Matsuzawa Y, Kurachi Y. A novel sulfonylurea receptor forms with BIR (Kir6.2) a smooth muscle type ATP-sensitive K⁺ channel. J Biol Chem 271: 24321–24324, 1996.[Abstract/Free Full Text]

38. Isomoto S, Kondo C, Kurachi Y. Inwardly rectifying potassium channels: their molecular heterogeneity and function. Jap J Physiol 47: 11–39, 1997.[CrossRef][Web of Science][Medline]

39. John SA, Monck JR, Weiss JN, Ribalet B. The sulphonylurea receptor SUR1 regulates ATP-sensitive mouse Kir6.2 K⁺ channels linked to the green fluorescent protein in human embryonic kidney cells (HEK293). J Physiol 510: 333–345, 1998.[Abstract/Free Full Text]

40. Karschin C, Ecke C, Ashcroft FM, Karschin A. Overlapping distribution of K_ATP channel-forming Kir6.2 subunit and the sulfonylurea receptor SUR1 in rodent brain. FEBS Lett 401: 59–64, 1997.[CrossRef][Web of Science][Medline]

41. Kramer RH, Mokkapatti R, Levitan ES. Effects of caffeine on intracellular calcium, calcium current and calcium dependent potassium current in anterior pituitary GH₃ cells. Pflügers Arch 426: 12–20, 1994.[CrossRef][Web of Science][Medline]

42. Krippeit-Drews P, Lonnendonker U. Dual effects of calcium on ATP-sensitive potassium channels of frog skeletal muscle. Biochim Biophys Acta 1108: 119–122, 1992.[Medline]

43. Kukreja RC, Ockaili R, Salloum F, Yin C, Hawkins J, Das A, Xi L. Cardioprotection with phosphodiesterase-5 inhibition–a novel preconditioning strategy. J Mol Cell Cardiol 36: 165–173, 2004.[CrossRef][Web of Science][Medline]

44. Light PE, Bladen C, Winkfein RJ, Walsh MP, French RJ. Molecular basis of protein kinase C-induced activation of ATP-sensitive potassium channels. Proc Natl Acad Sci USA 97: 9058–9063, 2000.[Abstract/Free Full Text]

45. Lin YF, Jan YN, Jan LY. Regulation of ATP-sensitive potassium channel function by protein kinase A-mediated phosphorylation in transfected HEK293 cell. EMBO J 19: 942–955, 2000.[CrossRef][Web of Science][Medline]

46. Lindaman BA, Hinkhouse MM, Conklin JL, Cullen JJ. The effect of phosphodiesterase inhibition on gallbladder motility in vitro. J Surg Res 105: 102–108, 2002.[CrossRef][Web of Science][Medline]

47. Liss B, Bruns R, Roeper J. Alternative sulfonylurea receptor expression defines metabolic sensitivity of K-ATP channels in dopaminergic midbrain neurons. EMBO J 18: 833–846, 1999.[CrossRef][Web of Science][Medline]

48. Luo D, Sun H, Xiao RP, Han Q. Caffeine induced Ca²⁺ release and capacitative Ca²⁺ entry in human embryonic kidney (HEK293) cells. Eur J Pharmacol 509: 109–115, 2005.[CrossRef][Web of Science][Medline]

49. McManus OB, Magleby KL. Kinetic states and modes of single large-conductance calcium activated potassium channels in cultured rat skeletal muscle. J Physiol 402: 79–120, 1988.[Abstract/Free Full Text]

50. Mercuri N, Bonci A, Johnson S, Stratta F, Calibresi P, Bernadi G. Effects of anoxia on rat midbrain dopamine neurons. J Neurophysiol 71: 1165–1173, 1994.[Abstract/Free Full Text]

51. Nehlig A, Daval JL, Debry G. Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Res Brain Res Rev 17: 139–170, 1992.[CrossRef][Medline]

52. Nichols CG, Lopatin AN. Inward rectifier potassium channels. Annu Rev Physiol 59: 171–191, 1997.[CrossRef][Web of Science][Medline]

53. Nichols CG, Shyng SL, Nestorowicz A, Glaser B, Clement JP 4th, Gonzalez G, Aguilar-Bryan L, Permutt MA, Bryan J. Adenosine diphosphate as an intracellular regulator of insulin secretion. Science 272: 1785–1787, 1996.[Abstract]

54. Noack T, Deitmer P, Golenhofen K. Features of a calcium independent, caffeine sensitive outward current in single smooth muscle cells from guinea pig portal vein. Pflügers Arch 416: 467–469, 1990.[CrossRef][Web of Science][Medline]

55. Ohta T, Wakade AR, Nakazato Y, Ito S. Ca²⁺-dependent K⁺ current and exocytosis in responses to caffeine and muscarine in voltage-clamped guinea-pig adrenal chromaffin cells. J Neurochem 78: 1243–1255, 2001.[CrossRef][Web of Science][Medline]

56. Okuyama Y, Yamada M, Kondo C, Satoh E, Isomoto S, Shindo T, Horio Y, Kitakaze M, Hori M, Kurachi. The effects of nucleotides and potassium channel openers on SUR2A/Kir6.2 complex K⁺ channel expressed in a mammalian cell line, HEK293T cells. Pflügers Arch 435: 595–603, 1998.[CrossRef][Web of Science][Medline]

57. Phillis JW. Adenosine and adenine nucleotides as regulators of cerebral blood flow: roles of acidosis, cell swelling, and K_ATP channels. Crit Rev Neurobiol 16: 237–270, 2004.[CrossRef][Medline]

58. Quayle JM, Nelson MT. ATP-sensitive and inwardly rectifying potassium channels in smooth muscle. Physiol Rev 77: 1165–1232, 1997.[Abstract/Free Full Text]

59. Reiser MA, D'Souza T, Dryer SE. Effects of caffeine and 3-isobutyl-1-methylxanthine on voltage-activated potassium currents in vertebrate neurones and secretory cells. Br J Pharmacol 118: 2145–2151, 1996.[Web of Science][Medline]

60. Rigoulot MA, Leroy C, Koning E, Ferrandon A, Nehlig A. Prolonged low-dose caffeine exposure protects against hippocampal damage but not against the occurrence of epilepsy in the lithium-pilocarpine model in the rat. Epilepsia 44: 529–535, 2003.[CrossRef][Web of Science][Medline]

61. Roper J, Ashcroft FM. Metabolic inhibition and low central ATP activate K-ATP channels in rat dopaminergic substantia nigra neurons. Pflügers Arch 430: 44–54, 1995.[CrossRef][Medline]

62. Ropero AB, Fuentes E, Rovira JM, Ripoll C, Soria B, Nadal A. Non-genomic actions of 17-oestradiol in mouse pancreatic -cells are mediated by a cGMP-dependent protein kinase. J Physiol 521: 397–407, 1999.[Abstract/Free Full Text]

63. Rosenzweig S, Yan W, Dasso M, Spielman AI. Possible novel mechanism for bitter taste mediated through cGMP. J Neurophysiol 81: 1661–1665, 1999.[Abstract/Free Full Text]

64. Rousseau E, Ladine J, Liu QY, Meissner G. Activation of the Ca²⁺ release channel of skeletal muscle sarcoplasmic reticulum by caffeine and related compounds. Arch Biochem Biophys 267: 75–86, 1988.[CrossRef][Web of Science][Medline]

65. Sachs D, Cunha FQ, Ferreira SH. Peripheral analgesic blockade of hypernociception: activation of arginine/NO/cGMP/protein kinase G/ATP-sensitive K⁺ channel pathway. Proc Natl Acad Sci USA 101: 3680–3685, 2004.[Abstract/Free Full Text]

66. Sakura H, Ämmälä C, Smith PA, Gribble FM, Ashcroft FM. Cloning and functional expression of the cDNA encoding a novel ATP-sensitive potassium channel expressed in pancreatic -cells, brain, heart and skeletal muscle. FEBS Lett 377: 338–344, 1995.[CrossRef][Web of Science][Medline]

67. Sanchez-Chapula J. Caffeine inhibits depolarization-activated outward currents in rat ventricular myocytes. Eur J Pharmacol 229: 163–169, 1992.[CrossRef][Web of Science][Medline]

68. Schroder RL, Jensen BS, Strobaek D, Olesen SP, Christophersen P. Activation of the human, intermediate-conductance, Ca²⁺-activated K⁺ channel by methylxanthines. Pflügers Arch 440: 809–818, 2000.[CrossRef][Web of Science][Medline]

69. Schwarzschild MA, Xu K, Oztas E, Petzer JP, Castagnoli K, Castagnoli N Jr, Chen JF. Neuroprotection by caffeine and more specific A_2A receptor antagonists in animal models of Parkinson's disease. Neurology 61: S55–S61, 2003.[Abstract/Free Full Text]

70. Seino S, Inagaki N, Namba N, Wang CH, Kotake K, Nagashima K, Miki T, Aguilar-Bryan L, Bryan J, Gonoi T. Molecular basis of functional diversity of ATP-sensitive K⁺ channel. Jap J Physiol 47, Suppl 1: S3–S4, 1997.[Medline]

71. Seino S, Miki T. Physiological and pathophysiological roles of ATP-sensitive K⁺ channels. Prog Biophys Mol Biol 81: 133–176, 2003.[CrossRef][Web of Science][Medline]

72. Seutin V, Shen KZ, North R, Johnson S. Sulphonylurea-sensitive potassium current evoked by sodium loading in rat midbrain dopamine neurons. Neuroscience 71: 709–719, 1996.[CrossRef][Web of Science][Medline]

73. Shyng SL, Ferrigni T, Nichols CG. Regulation of K_ATP channel activity by diazoxide and MgADP: distinct functions of the two nucleotide binding folds of the sulphonylurea receptor. J Gen Physiol 110: 643–654, 1997.[Abstract/Free Full Text]

74. Shyng SL, Nichols CG. Octameric stoichiometry of K_ATP channel complex. J Gen Physiol 110: 655–664, 1997.[Abstract/Free Full Text]

75. Sigworth FJ, Sine SM. Data transformations for improved display and fitting of single-channel dwell time histograms. Biophys J 52: 1047–1054, 1987.[Web of Science][Medline]

76. Soderling SH, Beavo JA. Regulation of cAMP and cGMP signaling: new phosphodiesterases and new functions. Curr Opin Cell Biol 12: 174–179, 2000.[CrossRef][Web of Science][Medline]

77. Stanford I, Lacey M. Regulation of potassium conductance in rat midbrain dopamine neurons by intracellular adenosine triphosphate (ATP) and the sulphonylureas tolbutamine and glibenclamide. J Neurosci 15: 4651–4657, 1995.[Abstract]

78. Sturgess NC, Ashford ML, Carrington CA, Hales CN. Single channel recordings of potassium currents in an insulin-secreting cell line. J Endocrinol 109: 201–207, 1986.[Abstract/Free Full Text]

79. Sturgess NC, Kozlowski RZ, Carrington CA, Hales CN, Ashford ML. Effects of sulphonylureas and diazoxide on insulin secretion and nucleotide-sensitive channels in an insulin-secreting cell line. Br J Pharmacol 95: 83–94, 1988.[Web of Science][Medline]

80. Teramoto N, Yunoki T, Tanaka K, Takano M, Masaki I, Yonemitsu Y, Sueishi K, Ito Y. The effects of caffeine on ATP-sensitive K⁺ channels in smooth muscle cells from pig urethra. Br J Pharmacol 131: 505–513, 2000.[CrossRef][Web of Science][Medline]

81. Thorneloe KS, Maruyama Y, Malcolm AT, Light PE, Walsh MP, Cole WC. Protein kinase C modulation of recombinant ATP-sensitive K⁺ channels composed of Kir6.1 and/or Kir62 expressed with SUR2B. J Physiol 541: 65–80, 2002.[Abstract/Free Full Text]

82. Trapp S, Tucker SJ, Ashcroft FM. Activation and inhibition of K_ATP currents by guanine nucleotides is mediated by different channel subunits. Proc Natl Acad Sci USA 94: 8872–8877, 1997.[Abstract/Free Full Text]

83. Tucker SJ, Gribble FM, Zhao C, Trapp S, Ashcroft FM. Truncation of Kir6.2 produces ATP-sensitive K-channels in the absence of the sulphonylurea receptor. Nature 387: 179–183, 1997.[CrossRef][Medline]

84. van Staveren WC, Markerink-van Ittersum M, Steinbusch HW, de Vente J. The effects of phosphodiesterase inhibition on cyclic GMP and cyclic AMP accumulation in the hippocampus of the rat. Brain Res 888: 275–286, 2001.[CrossRef][Web of Science][Medline]

85. Varro A, Hester S, Papp JG. Caffeine-induced decreases in the inward rectifier potassium and the inward calcium currents in rat ventricular myocytes. Br J Pharmacol 109: 895–897, 1993.[Web of Science][Medline]

86. Venosa RA, Hoya A. Effect of caffeine on K⁺ efflux in frog skeletal muscle. Pflügers Arch 437: 417–22, 1999.[CrossRef][Web of Science][Medline]

87. Watts A, Hicks G, Henderson G. Putative postnatal pre- and postsynaptic ATP-sensitive potassium channels in the rat substantia nigra in vitro. J Neurosci 15: 3065–3074, 1995.[Abstract]

88. Wellman GC, Quayle JM, Standen NB. ATP-sensitive K⁺ channel activation by calcitonin gene-related peptide and protein kinase A in pig coronary arterial smooth muscle. J Physiol 507: 117–129, 1998.[Abstract/Free Full Text]

89. Yamada M, Isomoto S, Matsumoto S, Kondo C, Shindo T, Horio Y, Kurachi Y. Sulphonylurea receptor 2B and Kir 6.1 form a sulphonylurea-sensitive but ATP-insensitive K⁺ channel. J Physiol 499: 715–720, 1997.[Abstract/Free Full Text]

90. Yamamoto T, Kakehata S, Yamada T, Saito T, Saito H, Akaike N. Caffeine rapidly decreases potassium conductance of dissociated outer hair cells of guinea-pig cochlea. Brain Res 677: 89–96, 1995.[CrossRef][Web of Science][Medline]

91. Yokoshiki H, Sunagawa M, Seki T, Sperelakis N. ATP-sensitive K⁺ channels in pancreatic, cardiac, and vascular smooth muscle cells. Am J Physiol Cell Physiol 274: C25–C37, 1998.[Abstract/Free Full Text]

92. Zerangue N, Schwappach B, Jan YN, Jan LY. A new ER trafficking signal regulates the subunit stoichiometry of plasma membrane K_ATP channels. Neuron 22: 537–548, 1999.[CrossRef][Web of Science][Medline]

93. Zhao FL, Lu SG, Herness S. Dual actions of caffeine on voltage-dependent currents and intracellular calcium in taste receptor cells. Am J Physiol Regul Integr Comp Physiol 283: R115–R129, 2002.[Abstract/Free Full Text]

This article has been cited by other articles:

				F. S. Jones, J. Jing, A. H. Stonehouse, A. Stevens, and G. M. Edelman Caffeine Stimulates Cytochrome Oxidase Expression and Activity in the Striatum in a Sexually Dimorphic Manner Mol. Pharmacol., September 1, 2008; 74(3): 673 - 684. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 292/6/C2239    most recent 00326.2006v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Mao, X. Articles by Lin, Y.-F. Search for Related Content PubMed PubMed Citation Articles by Mao, X. Articles by Lin, Y.-F.