Hybrid assemblies of ATP-sensitive K+ channels determine their muscle-type-dependent biophysical and pharmacological properties
- Domenico Tricarico*,†,
- Antonietta Mele*,†,
- Andrew L. Lundquist‡,
- Reshma R. Desai‡,
- Alfred L. George, Jr.‡, and
- Diana Conte Camerino*,§
- *Department of Pharmacobiology, Faculty of Pharmacy, University of Bari, via Orabona no. 4, 70120 Bari, Italy; and ‡Division of Genetic Medicine, Department of Medicine, Vanderbilt University, 529 Light Hall, Nashville, TN 37232-0275
-
Edited by Lily Y. Jan, University of California School of Medicine, San Francisco, CA, and approved November 15, 2005 (received for review July 18, 2005)
Abstract
ATP-sensitive K+ channels (KATP) are an octameric complex of inwardly rectifying K+ channels (Kir6.1 and Kir6.2) and sulfonylurea receptors (SUR1 and SUR2A/B), which are involved in several diseases. The tissue-selective expression of the subunits leads to different channels; however, the composition and role of the functional channel in native muscle fibers is not known. In this article, the properties of KATP channels of fast-twitch and slow-twitch muscles were compared by combining patch-clamp experiments with measurements of gene expression. We found that the density of KATP currents/area was muscle-type specific, being higher in fast-twitch muscles compared with the slow-twitch muscle. The density of KATP currents/area was correlated with the level of Kir6.2 expression. SUR2A was the most abundant subunit expressed in all muscles, whereas the vascular SUR2B subunit was expressed but at lower levels. A significant expression of the pancreatic SUR1 was also found in fast-twitch muscles. Pharmacological experiments showed that the channel response to the SUR1 agonist diazoxide, SUR2A/B agonist cromakalim, SUR1 antagonist tolbutamide, and the SUR1/SUR2A/B-antagonist glibenclamide matched the SURs expression pattern. Muscle-specific KATP subunit compositions contribute to the physiological performance of different muscle fiber types and determine the pharmacological actions of drugs modulating KATP activity in muscle diseases.
ATP-sensitive K+ channels (KATP) are a class of K+ channels widely distributed in the tissues involved in several pathophysiological conditions (1–3). Loss-of-function mutations of the pancreatic KATP channel are responsible for a rare inherited disease of infancy known as persistent hyperinsulinemia hypoglycemia of infancy, whereas gain-of-function mutations have been associated with some forms of diabetes type II mellitus (1–4). In skeletal muscle, the KATP channels have a protective role buffering ATP levels during fatigue and tetanus; regulate the extracellular K+ concentration through vasodilation, thereby enhancing muscle performance; and modulate glucose uptake (1, 3). Reduced expression or activity of sarcolemmal KATP channels is a common mechanism in the primary and secondary forms of hypokalemic periodic paralysis (hypoPP), a neuromuscular disorder associated with abnormal insulin responses leading to fiber depolarization, transient weakness, and hypokalemia (5–7).
KATP channels are octameric complexes of inwardly rectifying K+ channels (Kir6.1 and Kir6.2) and sulfonylurea receptor subunits (SUR1, SUR2A, and SUR2B) in 1:1 stoichiometry. The Kir6.2/SUR1 complex is the β-cell KATP channel, which is activated by diazoxide (diazo), a K+-channel opener (KCO), whereas it is inhibited with low affinity by tolbutamide (tolb) and with high affinity by glibenclamide (glib). The Kir6.2/SUR2A complex is the sarcolemmal (cardiac and skeletal muscle) KATP channel activated by cromakalim (crom), which is unable to activate the pancreatic channel. This channel is also inhibited by glib but with low affinity. The Kir6.1/SUR2B complex is the vascular form activated by crom and diazo and is also inhibited by sulfonylureas (2, 8). In recent years, however, the findings that the pancreatic SUR1 and the vascular SUR2B subunits were found in cardiac tissue and SUR2B, having near ubiquitous expression in other nonvascular tissues, raised a question regarding the composition of native channels. Furthermore, studies on the composition of the functional KATP channels in native cells are lacking. This investigation is an emerging point in the field because properties of recombinant channels often do not match with those of the native vascular and sarcolemmal KATP channels (8–10).
Moreover, the role of KATP channels in skeletal muscle function and metabolism is under debate. It is known that slow-twitch and fast-twitch muscles can be distinguished on the basis of their contractile proteins, cellular metabolism, hormonal regulation, and drug responses. Differences between muscle phenotypes in the expression and activity of sarcolemmal ion channels are now recognized. Down-regulation of the sarcolemmal voltage-dependent Na+ channel, the ClC1 chloride channel, and the aquaporin-4 channels has been observed in slow-twitch muscles (11–13). Up-regulation of maxi Ca-activated K+ (BKCa) channels has been observed in slow-twitch muscle as a result of the activity of a slow-twitch muscle-specific BKCa channel, which is resistant to acetazolamide, a BKCa opener in fast-twitch muscle (14).
In the present work, the biophysical properties, molecular composition, and drug responses of the functional KATP channels of fast- and slow-twitch rat muscle fibers were compared by combining patch-clamp experiments with real-time RT-PCR experiments. The drug responses of the channels of native fibers were investigated by testing the muscle-specific effects of diazo, crom, tolb, and glib. On the basis of these studies, we conclude that KATP channels in skeletal muscle exhibit fiber-specific differences in subunit composition.
Results
Muscle-Specific Biophysical Properties of KATP Channels. Patch excision into the ATP-free solution produced a dramatic increase in inward currents in 90%, 76%, 36%, and 17% of macropatches from TA, FDB, EDL, and SOL fibers, respectively. The mean inward current recorded at –60 mV (V m) immediately after excision was –470.1 ± 23 pA (n = 21 patches per 4 muscles), –401.44 ± 56 pA (n = 38 patches per 7 muscles), –290.22 ± 55 pA (n = 31 patches per 4 muscles), and –110.47 ± 13 pA (n = 42 patches per 10 muscles) for TA, FDB, EDL, and SOL fibers, respectively (Fig. 1A). Exposure of macropatches excised from all fibers to intracellular ATP (5 × 10–3 M) reduced the currents, confirming that these currents flowed through KATP channels. The ATP-sensitive current was –403.32 ± 24 pA, –359.06 ± 72 pA, –250.22 ± 36 pA, and –90.06 ± 11 pA in TA, FDB, EDL, and SOL fibers, respectively (Fig. 1B).
Sample KATP channel current traces of patches excised from TA, FDB, EDL, and SOL muscle fibers of the rat and current density from each muscle. The currents flowing through macropatches were recorded during voltage steps going from 0 to –60 mV (V m) in the presence of 150 × 10–3 M KCl on both sides of the membrane. (A) ATP applied on the intracellular side of the patches inhibited the currents in all patches. O, open-channel levels; C, closed channel levels. (B) KATP channel current densities recorded from all muscles. The numbers below the bars are the numbers of patches sampled.
No differences were observed in the single-channel conductance between muscle preparations. The single-channel conductance was 71 ± 3pS(n = 8 patches), 70 ± 11 pS (n = 11 patches), 72 ± 8pS(n = 7 patches), and 68 ± 7pS(n = 6 patches) in TA, FDB, EDL, and SOL muscles, respectively (Fig. 2A).
Single-channel recordings of KATP channels of TA, FDB, EDL, and SOL muscles of the rat and expression levels of KATP channel subunits. (A) The currents flowing through micropatches were recorded in excised patches at –60 mV (V m) in the presence of 150 mM KCl on both sides of the membrane. The amplitude of the single-channel current was –4.1, –3.91, –4.2, and –4.0 pA for the channels of TA, FDB, EDL, and SOL, respectively. O, open-channel levels; C, closed channel levels. (B Left) Kir6.2 subunit was the main expressed subunit in all muscles. The expression levels of Kir6.2 of FDB and TA were significantly higher with respect to that of EDL and SOL. *, P < 0.05. (B Right) A linear relationship has been observed between the mean KATP currents flowing through macropatches and expression levels of Kir6.2 subunit in different muscles; the correlation coefficient was 0.8 (y = b 1 x + b 0; fitting parameters: b 0 = 20.57; b 1 =–1,441). (C) SUR2A was the main subunit expressed in all muscles; SUR1 was significantly expressed in FDB (*, P < 0.05) with respect to the other muscles, whereas SUR2B was expressed in all muscles at lower levels. Expression of the subunits was evaluated by RT-PCR experiments on contralateral muscles of the same rats used for patch-clamp experiments.
ATP applied on the intracellular side of the patches dose-dependently reduced the KATP currents in all muscles but with different potencies. The IC50 for ATP calculated by a fitting routine was 13.1 ± 8 × 10–6 M (slope, 0.7) (n = 6 patches), 32.3 ± 2 × 10–6 M (slope, 1) (n = 3 patches), 45.3 ± 9 × 10–6 M (slope, 1.1) (n = 4 patches), and 37.2 ± 9 × 10–6 M (slope, 1.2) (n = 5) for FDB, TA, EDL, and SOL muscles, respectively.
Expression of KATP Subunits in TA, FDB, EDL, and SOL Muscles. Quantitative real-time RT-PCR analysis showed that the relative expression value of Kir6.2 was significantly higher than that of Kir6.1 in all muscles (Fig. 2B Left). The expression of Kir6.2 was also significantly different among muscles (FDB ≥ TA ≫ EDL > SOL) and was linearly correlated with the mean KATP currents recorded in different muscles (Fig. 2B Right).
We found that the expression patterns of SURs varied between muscles, with SUR2A being the most highly expressed subunit in all muscles (Fig. 2C). A significant expression of SUR1 subunit was found in FDB muscle, whereas its expression in TA, EDL, and SOL muscles was negligible. The SUR2B subunit was also expressed in all muscles, although no significant differences among muscles were observed.
Muscle-Specific Effects of KCO and Sulfonylureas on KATP. crom and diazo in the presence of internal ATP, in excised patches, caused a dose-dependent activation of KATP channels in all muscles but with different efficacies, potencies, and muscle specificities. crom was more effective and potent than diazo in activating KATP channels in all muscles. The order of efficacy of crom (n = 23 patches from all muscles) as a Kir6.2/SUR2A-B agonist was TA > FDB > EDL > SOL, and the order of efficacy of diazo (n = 24 patches from all muscles) as a Kir6.2/SUR1/2B agonist was TA > FDB > SOL ≫ EDL (Fig. 3 and Table 1). The crom concentration–response data were fitted with the sum of two stimulatory site functions, whereas the diazo data were fitted with the sum of one stimulatory site function and one inhibitory site function in all muscles (Fig. 3 and Table 1). The down-turn of responses in the concentration–responses data of diazo was observed at a concentration >150 × 10–6 M.
Concentration–response relationships of crom, diazo, glib, and tolb performed in FDB, EDL, TA, and SOL muscles and fitting parameters. crom and glib were more effective and potent than diazo and tolb, respectively, in all muscles, and their effects were muscle specific.
glib and tolb, in the absence of internal ATP, caused a dose-dependent inhibition of KATP channels in all muscle types but with different efficacies, potencies, and muscle specificities. glib was more effective and potent than tolb in all muscles (Fig. 3 and Table 1). The order of efficacy of glib (n = 19 patches from all muscles) as a Kir6.2/SUR1/2B-A antagonist was FDB > TA > EDL > SOL, and the order of efficacy of tolb (n = 15 patches from all muscles) as a Kir6.2/SUR1 antagonist was FDB ≫ TA > SOL ≥ EDL (Fig. 3 and Table 1). The concentration–response data of glib were fitted with the sum of two inhibitory site functions, whereas the tolb data were fitted with one inhibitory site in all muscles (Fig. 3 and Table 1).
Selective drug experiments showed differences in drug responses within a muscle. Three different drug responses were observed (Fig. 4). First, in FDB muscle (no. of drug-sensitive patches per no. of patches = 5/12), crom leads to full channel activation. Nevertheless, the crom-sensitive channels were insensitive to diazo and tolb, suggesting the contribution of the SUR2A subunit to the functional channels (Fig. 4A). This drug response was also observed in TA (no. of drug-sensitive patches per no. of patches = 10/12), EDL (no. of drug-sensitive patches per no. of patches = 10/12), and SOL (no. of drug-sensitive patches per no. of patches = 8/10) muscles with a high frequency (Fig. 4B). Second, in FDB (no. of drug-sensitive patches per no. of patches = 4/12), we found channels that were crom-insensitive but were fully activated by diazo and inhibited by tolb, suggesting the contribution of SUR1 (Fig. 4A). Third (no. of drug-sensitive patches per no. of patches = 3/12), both crom and diazo partially activated the same channels. The coincubation with both drugs leads to full channel activation that was antagonized by tolb, suggesting the contribution of two independent complexes of incorporating SUR1 and SUR2A subunits into the functional channels (Fig. 4A). Additional drug responses were observed, however, with less frequency. In TA (no. of drug-sensitive patches per no. of patches = 2/12), crom and diazo fully activated the same channels; in EDL (no. of drug-sensitive patches per no. of patches = 2/12) and SOL (no. of drug-sensitive patches per no. of patches = 2/10), both drugs partially activated the channels. The coincubation of these patches with both drugs leads to channel inhibition, suggesting an interaction with a common subunit, e.g., SUR2B. These channels were less responsive to tolb (Fig. 4).
Selective drug experiments and distribution of drug responses for each muscle. (A) Sample traces of KATP currents flowing through macropatches from FDB (models 1, 2, and 3), TA (model 4), and SOL (model 5) muscles representative of the drug experiments performed in all muscles. The EDL drug response, not shown, was similar to that of SOL muscle. diazo (10–4 M) and crom (10–4 M) were tested alone or in combination on the same excised patches in the presence of internal ATP (10–4 M) at –60 mV (V m). tolb (1.5 × 10–3 M) and glib (100 × 10–6 M) were tested in the absence of ATP. Three different drug responses were observed in FDB, and two different responses were observed in TA, SOL, and EDL muscles. crom-sensitive channels insensitive to diazo and tolb (model 1). This finding was the most frequently observed drug response in all muscles. crom-insensitive channels fully activated by diazo and inhibited by tolb found in FDB (model 2). crom- and diazo-sensitive channels fully activated by coincubation with both drugs and antagonized by tolb found in FDB (model 3). crom- and diazo-sensitive channels fully inhibited by coincubation with both drugs and less responsive to tolb found in TA, SOL, and EDL muscles (models 4 and 5).(B) Drug effects vs. patch frequency in different muscles. The data are expressed as mean ± SE.
The lack of agonist effects observed after the coincubation of the patches with both diazo and crom can be interpreted as the channel responses to over-saturating concentrations of drugs interacting on the common SUR2B subunit. This finding was in line with previous observations showing that KCO, in the presence of ATP, induces channel activation at low concentrations and inhibition at concentrations >200 × 10–6 M (15).
Discussion
Molecular Composition of Functional KATP Channels in Native Fibers. In this article, the biophysical properties and molecular composition of functional skeletal muscle KATP channels were investigated in native fibers. This study was achieved by combining patch-clamp and selective drug experiments with RT-PCR experiments in fast-twitching TA, FDB, and EDL muscles and slow-twitching SOL muscles.
Patch-clamp experiments showed that the significant differences in the density of the KATP currents between muscles were TA > FDB ≫ EDL ≫ SOL. This finding is probably related to an increase of the number of functional channels multiplied by the open probability because the single-channel conductance did not change within muscles. RT-PCR experiments showed that the Kir6.2 subunit was the main subunit expressed in all muscles (FDB ≥ TA ≫ EDL ≥ SOL), and the expression levels of this subunit were correlated with the KATP currents recorded in the contralateral muscles of the same rats. We also confirmed that the main β-subunit of the KATP channel complex is the SUR2A subunit, which was abundantly expressed in all muscles (1–3). However, it is intriguing that other SUR subunits were expressed in skeletal muscle. In fact, a significant and muscle-specific expression of the pancreatic SUR1 subunit has been found in FDB. In addition, the vascular SUR2B subunit was expressed in all muscles, although the contamination of the muscle samples with vascular tissues in our samples could have affected the results.
The observed differences in drug responses among muscles and within a specific muscle opened the question on the composition of the functional channels in native fibers. Drug experiments indicated that the KATP channels in skeletal muscles are hybrid assemblies of Kir6.2/SUR2A and Kir6.2/SUR1 subunits organized as homomeric complexes with the possible contribution of SUR2B to the functional channels. This hypothesis is demonstrated by several findings. First, crom and glib acted by binding to multiple sites of action showing different affinity, whereas one site of interaction was sufficient for diazo and tolb in all muscles, as demonstrated by the concentration–response relationships analysis. If we assume that one binding site × subunit is responsible for the pharmacological action of these drugs, then it is likely that the high-affinity binding site for glib is on the SUR1 subunit, whereas the low-affinity sites are on the SUR2A and -B subunits. In the case of crom, the high-affinity binding site may be located on the SUR2B subunit, whereas the low-affinity site is on the SUR2A subunit. However, the contribution of multiple sites × subunit to the drug–receptor interaction cannot be excluded. In fact, binding experiments performed on cloned subunits have shown one binding site × subunit for most of the KCOs, although in some cases two binding sites were also observed (2). Second, selective drug experiments have shown that the channels were selectively responsive to diazo, crom, and tolb, which are known for their different subunit selectivity, whereas all channels sampled were responsive to glib, which is an unselective channel blocker. Third, the observed muscle-specific drug responses of the channels were related to the expression pattern of the SUR subunits and to the current levels recorded in the muscles. For example, FDB muscle expressing both SUR1 and SUR2A subunits was highly sensitive to all drugs, whereas the muscles expressing mostly SUR2A were crom-sensitive and less responsive to diazo, tolb, and glib. Furthermore, muscles showing a high current density such as FDB and TA were also highly responsive to drugs, indicating that the drug efficacy and drug specificity depend on the drug binding sites available and the type of subunit expressed. Fourth, our observation that the IC50 for ATP of the FDB channels was significantly lower than that of TA, EDL, and SOL channels is in line with the fact that the channels composed of SUR1 are more sensitive to ATP inhibition than channels composed of SUR2A and -B subunits (16).
Although our drug experiments suggested that SUR2B participates in the formation of the channel complex, the contribution of this subunit to the functional channels in native fibers is masked by the possible contamination in our muscle samples from vascular tissues, which could have affected the expression levels of SUR2B. The homomeric complex hypothesis is also in agreement with the fact that SUR1 and SUR2A cannot form functional channels (9).
Physiological Implications. Our data indicate that KATP channel properties vary depending on the muscle phenotype. Despite the proposed protective role of KATP channels against fatigue, the phenotype-dependent KATP activity observed in our study corroborates the involvement of KATP channels in enhancing muscle performance. High KATP activity in fast-twitch muscle contributes to regulating the extracellular K+ concentration, which in turn stimulates the pressor reflex leading to vasodilation (17). The phenotype-dependent KATP activity also leads to a better use of glucose among muscles in proportion to their metabolic needs. For instance, high KATP activity would reduce the glucose uptake in fast-twitch muscles, making it more available to slow-twitch muscles, which are characterized by a high glucose demand during contraction. The low sarcolemmal KATP channel activity recorded in slow-twitch muscle can be related to the slow kinetics of contraction and relaxation and is consistent with the high rate of glucose uptake, which characterizes this muscle phenotype (18, 19).
Unexpectedly, we observed differences in KATP channel properties among muscles. The KATP channels of FDB were indeed composed of SUR1 and SUR2 subunits, whereas KATP channels of TA, EDL, and SOL muscles were composed of only SUR2 subunits. It is known that channel complexes composed of SUR2 subunits are more responsive to metabolic stresses compared with channel complexes of SUR1 (20). The differences in SURs composition of the channels among muscles may serve specific functions. For example, TA, EDL, and SOL muscles are exposed more often to hypoxia and fatigue than FDB muscle, which shows a different morphology and function.
Pharmacological Implications. The findings that skeletal muscle expresses SUR1 and SUR2B subunits have important implications. The high-affinity interaction of sulfonylureas with the SURs of skeletal muscle, vascular, and cardiac tissues would cause side effects (21). For example, the interaction of sulfonylureas with the sarcolemmal KATP channels at therapeutic concentrations may enhance the insulin sensitivity of skeletal muscle, contributing to the characteristic sulfonylurea-dependent hypoglycemia. Furthermore, the observation that the drug responses varied with muscle types may have a role in those diseases associated to changes in the muscle morphology and/or phenotypic transitions as occurring in response to muscle disuse. Phenotypic transitions of the muscle lead to the expression of different subunits showing abnormal drug responses (14, 22).
Materials and Methods
Rat Housing (Animal Care) and Muscle Biopsy. Animal care was performed in accordance with the Guide for Care and Use of Laboratory Animals prepared by the National Research Council.
The flexor digitorum brevis (FDB), tibialis anterioris (TA), extensor digitorum longus (EDL), and soleus (SOL) muscles were dissected from adult male Wistar rats (260–300 g) while they were urethane anesthesia (1.2 g/kg). After dissection, the animals were rapidly killed with an overdose of urethane. Single muscle fibers were prepared from FDB, TA, EDL, and SOL muscles by enzymatic dissociation for patch-clamp experiments. The contralateral FDB, TA, EDL, and SOL muscles were removed from the same rats and frozen in liquid nitrogen promptly after surgical removal for mRNA analysis.
Real-Time Quantitative PCR. For each muscle sample, the total RNA was isolated by using TRIzol reagent and treated with DNase I (4 units, 37°C, 1 h). RNA was quantified by using a spectrophotometer (Beckman DU 530), and 3 μg was used for reverse transcription. Synthesis of cDNA was performed by using random hexamers (annealed 10 min, 25°C) and Superscript II reverse transcriptase (Invitrogen–Life Technologies) incubated at 42°C for 50 min.
We identified the available rat sequence for Kir6.1 (NM_017099), Kir6.2 (NM_031358), SUR1 (NM_013039), SUR2A (D83598), SUR2B (AF019628), and β-actin (NM_031144). Fluorescently labeled TaqMan (Applied Biosystems) probes for Kir6.1, SUR1, SUR2A, SUR2B, and β-actin were designed by using Primer Express (Applied Biosystems) to amplify 124- to 141-bp products encompassing each probe annealing site. Specific primer and probe sequences for each gene are reported in Table 2. To achieve a high level of specificity and to avoid detection of genomic DNA, we designed a probe to span exon–exon junctions for each gene except Kir6.2 (Kir6.2 gene is intronless). For Kir6.2, a TaqMan probe was designed within a region having substantial sequence divergence with other genes, and control amplifications of RNA without reverse transcription were performed to exclude genomic DNA contamination. None of the SUR and Kir primer and probe sets crossreacted with nonspecific SUR or Kir sequences after 40 cycles of PCR, and no amplification was observed after 45 cycles of PCR in control reactions containing no DNA template. Triplicate reactions were carried out in parallel for each individual muscle sample. The results were compared with a gene-specific standard curve and normalized to expression of the housekeeping gene β-actin in the same sample. The template used for determining standard curves consisted of plasmid DNA containing the expected target sequence evaluated by Pico Green fluorescence (Molecular Probes).
Drugs and Solutions. The normal Ringer's solution used during muscle biopsy and for preparation of isolated fibers contained 145 × 10–3 M NaCl, 5 × 10–3 M KCl, 1 × 10–3 M MgCl2, 0.5 × 10–3 M CaCl2, 5 × 10–3 M glucose, and 10 × 10–3 M 3-(N-morpholino)propanesulfonate (Mops) sodium salt and was adjusted to pH 7.2 with Mops acid. The patch-pipette solutions contained 150 × 10–3 M KCl, 2 × 10–3 M CaCl2, and 1 × 10–2 M Mops (pH 7.2). The bath solution contained 150 × 10–3 M KCl, 5 × 10–3 M EGTA, and 1 × 10–2 M Mops (pH 7.2). Stock solutions of crom (5 × 10–2 M), diazo (5 × 10–2 M), glib (6 × 10–3 M), and tolb (6 × 10–3 M) (Sigma) were prepared by dissolving the drugs in DMSO. Microliter amounts of the KCO (crom and diazo) and sulfonylureas (glib and tolb) stock solutions were then added to the bath solution in the presence and in the absence of K2ATP, respectively. DMSO applied at the maximal concentration tested, which was 0.05%, did not affect the channel currents in the absence or in the presence of ATP (solvent control).
Patch-Clamp Experiments. Experiments were performed in inside-out configurations by using the standard patch-clamp technique. Channel currents were recorded during voltage steps going from 0 mV of holding potential to different voltages (from –70 to +70 mV) and from 0 to –60 mV voltage membrane (V m) immediately after excision, at 20–22°C, in the presence of KCl on both sides of membrane patches in the absence (control) or presence of ATP in the bath. The currents were recorded at a 1-kHz sampling rate (filter = 0.2 kHz) by using an Axopatch-1D amplifier equipped with a CV-4 headstage (Axon Instruments, Union City, CA).
Macropatches having an average pipette area of 8.3 ± 1 μm2 (n = 320 patches) were used to measure the mean KATP currents and the pharmacological responses of the channels, whereas micropatches with mean pipette area of 1.3 ± 0.1 μm2 (n = 32 patches) were used for single-channel recordings.
The mean currents were calculated by subtracting the baseline level from the open-channel level of each current trace and then digitally averaging all generated files by using clampfit (Axon Instruments). The base-line level for the KATP current was measured in the presence of internal ATP (5 × 10–3 M). Macropatches containing voltage-dependent K+ channels or other Kir and showing loss of channel currents during the time of observation were excluded from the analysis. Current amplitude was measured by using clampfit. We made no correction for liquid junction potential, which was approximately < 2 mV in our experimental conditions.
The concentration–response relationships were constructed by applying increasing concentrations of diazo (1 × 10–9 to 6 × 10–4 M), crom (1 × 10–10 to 2 × 10–4 M), tolb (50 × 10–6 to 1.5 × 10–3 M), and glib (1 × 10–11 to 1 × 10–4 M) on macropatches excised from FDB, TA, EDL, and SOL muscles, as described in ref. 15.
The selective drug experiments have been performed by testing the effects of saturating concentrations of crom (100 × 10–6 M)/ATP (100 × 10–6 M) (step 1), diazo (100 × 10–6 M)/ATP (100 × 10–6 M) (step 3), crom (100 × 10–6 M) + diazo (100 × 10–6 M)/ATP (100 × 10–6 M) (step 5), and tolb (1.5 × 10–3 M) or glib (100 × 10–6 M) (step 7) on the same channels of excised macropatches. A 60-s washout period followed each step of drug application periods (steps 2, 4, and 6) (see Table 3).
Statistical Analysis. The data were expressed as mean ± SE. Statistical comparisons between groups of data were made by using Student's t test, and significance was assumed for P < 0.05. The fitting of the concentration–response relationships data was performed by functions on the basis of one stimulatory component or one inhibitory component, as described in ref. 15.
Data from channels not responsive to drug applications were excluded from the mean. The algorithms of the fitting procedures used are based on Levenberg–Marquardt least-squares fitting routines. Linear regression analyses was used to test for correlation between variables (23).
Acknowledgments
This work was supported by Telethon–Conte Grant GGP04140.
Footnotes
-
↵ § To whom correspondence should be addressed. E-mail: conte{at}farmbiol.uniba.it.
-
↵ † D.T. and A.M. contributed equally to this work.
-
Conflict of interest statement: No conflicts declared.
-
This paper was submitted directly (Track II) to the PNAS office.
-
Abbreviations: KATP, ATP-sensitive K+ channel; SUR, sulfonylurea receptor; Kir, inwardly rectifying K+ channel; KCO, K+-channel opener; FDB, flexor digitorum brevis; TA, tibialis anterioris; EDL, extensor digitorum longus; SOL, soleus; diazo, diazoxide; crom, cromakalim; tolb, tolbutamide; glib, glibenclamide.
- Copyright © 2006, The National Academy of Sciences
