Coupling of ATP-sensitive K+ channels to A1 receptors by G proteins in rat ventricular myocytes G. E. KIRSCH, J. CODINA, L. BIRNBAUMER, AND A. M. BROWN Departments of Anesthesiology, Molecular Physiology and Biophysics, and Cell Biology, Baylor College of Medicine, Houston, Texas 77030
KIRSCH, G. E., J. CODINA, L. BIRNBAUMER, AND A. M. BROWN. Coupling of ATP-sensitive K+ channels to A, receptors by G proteins in rat ventricular rnyocytes. Am. J. Physiol. 259 (Heart Circ. Physiol. 28): H820-H826, 1990.-ATP-sensitive K’ (K+[ATP]) current is thought to be regulated by GTPbinding proteins (G proteins), but the pathways that couple receptor, G protein, and channel have not been defined. We studied regulation of tolbutamide-sensitive K’[ATP] current in neonatal rat ventricular myocytes. Application of 0.1 mM ATP to the intracellular side of membrane patches reduced K’ [ATP] channel activity, and addition of the nonhydrolyzable GTP analogue guanosine 5’-O-(3-thiotriphosphate) (GTPqS) at 0.1 mM restored activity. Application of 0.1 mM intracellular GTP plus 10 PM extracellular adenosine or 100 nM N”-cyclohexyladenosine had the same effect as GTP$S; hence K’[ATP] channels may be coupled to adenosine receptors via G proteins. To determine which G protein, we applied Gcu subunits, preactivated with GTPyS to the cytoplasmic side of membrane patches, and found that ail, ai2, and ai mimicked the effect of GTP+, but not cyOor G,, suggesting that Gia acts via a membrane-delimited pathway. Adenosine receptor coupling may be important for activating K’[ATP] channels in ischemic muscle.
cell experiments (cell-attached patch recording from whole cells partially disrupted by saponin) it is consistently fivefold higher (17, 24). 2) In excised patches a subgroup of channels with reduced ATP sensitivity and altered ATP-dependent kinetic properties has been described (26). 3) K+[ATP] channels exposed to ATP-free intracellular solutions show a gradual cessation of activity that can be partially restored by reexposure to Mg* ATP (10, 32). The ATP sensitivity of insulin-secreting cells is known to be modulated by pertussis toxin-sensitive G proteins, which are activated by galanin or somatostatin receptors (7,12). Other than this example, receptors that might be coupled to K+[ATP] channels via G proteins are unknown. For this reason we tested the possible involvement of G proteins in modulating cardiac K’[ATP] channels and found that three Gia proteins but not G,a or G, can reduce the sensitivity of the channel to block by ATP. Modulation takes place in a membrane-delimited pathway that requires Mg2+ and GTP as well as Gia. Purinergic receptors could provide input to the Gi proteins that act on K’[ATP] channels.
adenosine 5’-triphosphate; guanosine 5’-triphosphate-binding protein; heart; ion channels; ischemia; potassium; purinergic receptors; IV”-cyclohexyladenosine; tolbutamide
METHODS
K+ (K+[ATP]) channels, which activate when intracellular ATP levels drop, are found in cells from a wide variety of tissues including heart, skeletal muscle, brain, and pancreas (2). Whereas their pivotal role in controlling insulin release from pancreatic P-cells has been well established, much less is known about the function of K+[ATP] channels in heart cells. When activated pharmacologically (11, 30) these channels drastically reduce action potential duration, and it has been suggested that they may be responsible for action potential shortening in metabolically compromised ischemic muscle (9). However, the ATP concentration in metabolically blocked muscle remains above the level that inhibits channels in excised membrane patches. A possible explanation for this discrepancy is that the ATP sensitivity of the channels can be modulated by intracellular mechanisms. Several observations suggest the presence of modulatory mechanisms. 1) The mean inhibitory concentration for ATP block of K+ channels in excised patches has been reported to be 0.1 mM (17,23,33), whereas in open-
Cell culture. Ventricular myocytes were prepared from neonatal rat hearts using procedures described previously (20). Briefly, myocytes were isolated from hearts of 1- to 3-day-old neonatal rats by trypsin digestion. Dispersed cells were seeded on glass cover slips and incubated in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum at 37OC.Recordings were made from cells in culture l-2 days. Electrical recording. Membrane ionic currents were measured using the gigaseal patch-clamp method (14). Sylgard-coated micropipettes of 3-5 MQ were used for recording whole cell and single-channel currents. Currents were measured using a commercial patch-clamp amplifier (Axopatch-lC, Axon Instruments, Burlingame, CA), low-pass filtered at 10 kHz (-3 dB), digitized at 48 kHz, and stored on videotape using a PCM-VCR adaptor (PCM-2, Medical Systems, Greenvale, NY). Segments of the continuous single-channel recordings were analyzed using a laboratory computer (Compaq 386, Houston, TX) and IPROC (29) software (Axon Instruments). Transitions between closed and open states were identified using a half-maximum single-unit amplitude threshold, and histograms of unitary amplitude, open-time dura-
ATP-SENSITIVE
H820
0363-6135/90
$1.50 Copyright
0 1990 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpheart (193.093.192.151) on February 13, 2019.
G PROTEINS
MODULATE
tion, and closed-time duration were constructed. Amplitude histograms were fit to Gaussian functions and dwelltime histograms were fit to exponential probability density functions using a maximum likelihood estimate. Whole cell currents evoked by voltage ramps (20 mV/ s from -100 to 100 mV) were filtered at 15 Hz and digitized at 50 Hz. Records were not corrected for capacitative current (average cell capacitance was 43 t 26 pF, n= 12) estimated to be
KIRSCH, G. E., J. CODINA, L. BIRNBAUMER, AND A. M. BROWN. Coupling of ATP-sensitive K+ channels to A, receptors by G proteins in rat ventricular rnyocytes. Am. J. Physiol. 259 (Heart Circ. Physiol. 28): H820-H826, 1990.-ATP-sensitive K’ (K+[ATP]) current is thought to be regulated by GTPbinding proteins (G proteins), but the pathways that couple receptor, G protein, and channel have not been defined. We studied regulation of tolbutamide-sensitive K’[ATP] current in neonatal rat ventricular myocytes. Application of 0.1 mM ATP to the intracellular side of membrane patches reduced K’ [ATP] channel activity, and addition of the nonhydrolyzable GTP analogue guanosine 5’-O-(3-thiotriphosphate) (GTPqS) at 0.1 mM restored activity. Application of 0.1 mM intracellular GTP plus 10 PM extracellular adenosine or 100 nM N”-cyclohexyladenosine had the same effect as GTP$S; hence K’[ATP] channels may be coupled to adenosine receptors via G proteins. To determine which G protein, we applied Gcu subunits, preactivated with GTPyS to the cytoplasmic side of membrane patches, and found that ail, ai2, and ai mimicked the effect of GTP+, but not cyOor G,, suggesting that Gia acts via a membrane-delimited pathway. Adenosine receptor coupling may be important for activating K’[ATP] channels in ischemic muscle.
cell experiments (cell-attached patch recording from whole cells partially disrupted by saponin) it is consistently fivefold higher (17, 24). 2) In excised patches a subgroup of channels with reduced ATP sensitivity and altered ATP-dependent kinetic properties has been described (26). 3) K+[ATP] channels exposed to ATP-free intracellular solutions show a gradual cessation of activity that can be partially restored by reexposure to Mg* ATP (10, 32). The ATP sensitivity of insulin-secreting cells is known to be modulated by pertussis toxin-sensitive G proteins, which are activated by galanin or somatostatin receptors (7,12). Other than this example, receptors that might be coupled to K+[ATP] channels via G proteins are unknown. For this reason we tested the possible involvement of G proteins in modulating cardiac K’[ATP] channels and found that three Gia proteins but not G,a or G, can reduce the sensitivity of the channel to block by ATP. Modulation takes place in a membrane-delimited pathway that requires Mg2+ and GTP as well as Gia. Purinergic receptors could provide input to the Gi proteins that act on K’[ATP] channels.
adenosine 5’-triphosphate; guanosine 5’-triphosphate-binding protein; heart; ion channels; ischemia; potassium; purinergic receptors; IV”-cyclohexyladenosine; tolbutamide
METHODS
K+ (K+[ATP]) channels, which activate when intracellular ATP levels drop, are found in cells from a wide variety of tissues including heart, skeletal muscle, brain, and pancreas (2). Whereas their pivotal role in controlling insulin release from pancreatic P-cells has been well established, much less is known about the function of K+[ATP] channels in heart cells. When activated pharmacologically (11, 30) these channels drastically reduce action potential duration, and it has been suggested that they may be responsible for action potential shortening in metabolically compromised ischemic muscle (9). However, the ATP concentration in metabolically blocked muscle remains above the level that inhibits channels in excised membrane patches. A possible explanation for this discrepancy is that the ATP sensitivity of the channels can be modulated by intracellular mechanisms. Several observations suggest the presence of modulatory mechanisms. 1) The mean inhibitory concentration for ATP block of K+ channels in excised patches has been reported to be 0.1 mM (17,23,33), whereas in open-
Cell culture. Ventricular myocytes were prepared from neonatal rat hearts using procedures described previously (20). Briefly, myocytes were isolated from hearts of 1- to 3-day-old neonatal rats by trypsin digestion. Dispersed cells were seeded on glass cover slips and incubated in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum at 37OC.Recordings were made from cells in culture l-2 days. Electrical recording. Membrane ionic currents were measured using the gigaseal patch-clamp method (14). Sylgard-coated micropipettes of 3-5 MQ were used for recording whole cell and single-channel currents. Currents were measured using a commercial patch-clamp amplifier (Axopatch-lC, Axon Instruments, Burlingame, CA), low-pass filtered at 10 kHz (-3 dB), digitized at 48 kHz, and stored on videotape using a PCM-VCR adaptor (PCM-2, Medical Systems, Greenvale, NY). Segments of the continuous single-channel recordings were analyzed using a laboratory computer (Compaq 386, Houston, TX) and IPROC (29) software (Axon Instruments). Transitions between closed and open states were identified using a half-maximum single-unit amplitude threshold, and histograms of unitary amplitude, open-time dura-
ATP-SENSITIVE
H820
0363-6135/90
$1.50 Copyright
0 1990 the American
Physiological
Society
Downloaded from www.physiology.org/journal/ajpheart (193.093.192.151) on February 13, 2019.
G PROTEINS
MODULATE
tion, and closed-time duration were constructed. Amplitude histograms were fit to Gaussian functions and dwelltime histograms were fit to exponential probability density functions using a maximum likelihood estimate. Whole cell currents evoked by voltage ramps (20 mV/ s from -100 to 100 mV) were filtered at 15 Hz and digitized at 50 Hz. Records were not corrected for capacitative current (average cell capacitance was 43 t 26 pF, n= 12) estimated to be
