M. J. Dunne

Intracellular ADP activates K+ channels that are inhibited by ATP in an insulin-secreting cell line

the effect of ADP on ATP‐sensitive K+ channels in the insulin‐secreting RINm5F cell line has been investigated with the help of single‐channel current recording from saponin‐permeabilized cells. ADP (100–500 μM) markedly activates K+ channels when added to the bath solution in contact with the membrane inside. ADP‐β‐S cannot mimick this effect. During sustained ATP (500 μM)‐evoked inhibition of K+ channel opening, 500 μM ADP markedly and reversibly activates the channels. Conversely ATP markedly reduces the opening probability of ADP‐activated channels. It is suggested that the physiological control of K+ channel opening in the insulin‐secreting cells is mediated by changes in ATP/ADP ratio rather than being solely determined by the ATP concentration.

ATP-sensitive inward rectifier and voltage- and calcium-activated K+ channels in cultured pancreatic islet cells

SummaryK+ channels in cultured rat pancreatic islet cells have been studied using patch-clamp single-channel recording techniques in cell-attached and excised inside-out and outside-out membrane patches. Three different K+-selective channels have been found. Two inward rectifier K+ channels with slope conductances of about 4 and 17 pS recorded under quasi-physiological cation gradients (Na+ outside, K+ inside) and maximal conductances recorded in symmetrical K+-rich solutions of about 30 and 75 pS, respectively. A voltage- and calcium-activated K− channel was recorded with a slope conductance of about 90 pS under the same conditions and a maximal conductance recorded in symmetrical K+-rich solutions of about 250 pS. Single-channel current recording in the cell-attached conformation revealed a continuous low level of activity in an apparently small number of both the inward rectifier K+ channels. But when membrane patches were excised from the intact cell a much larger number of inward rectifier K+ channels became transiently activated before showing an irreversible decline. In excised patches opening and closing of both the inward rectifier K+ channels were unaffected by voltage, internal Ca2+ or externally applied tetraethyl-ammonium (TEA) but the probability of opening of both inward rectifier K+ channels was reduced by internally applied 1–5mm adenosine-5′-triphosphate (ATP). The large K+ channel was not operational in cell-attached membrane patches, but in excised patches it could be activated at negative membrane potentials by 10−7 to 10−6m internal Ca2+ and blocked by 5–10mm external TEA.

High-conductance K+ channel in pancreatic islet cells can be activated and inactivated by internal calcium

SummaryThe Ca2+-activated K+ channel in rat pancreatic islet cells has been studied using patch-clamp single-channel current recording in excised inside-out and outside-out membrane patches. In membrane patches exposed to quasi-physiological cation gradients (Na+ outside, K+ inside) large outward current steps were observed when the membrane was depolarized. The single-channel current voltage (I/V) relationship showed outward rectification and the null potential was more negative than −40 mV. In symmetrical K+-rich solutions the single-channelI/V relationship was linear, the null potential was 0 mV and the singlechannel conductance was about 250 pS. Membrane depolarization evoked channel opening also when the inside of the membrane was exposed to a Ca2+-free solution containing 2mm EGTA, but large positive membrane potentials (70 to 80 mV) were required in order to obtain open-state probabilities (P) above 0.1. Raising the free Ca2+ concentration in contact with the membrane inside ([Ca2+]i) to 1.5×10−7m had little effect on the relationship between membrane potential andP. When [Ca2+]i was increased to 3×10−7m and 6×10−7m smaller potential changes were required to open the channels. Increasing [Ca2+]i further to 8×10−7m again activated the channels, but the relationship between membrane potential andP was complex. Changing the membrane potential from −50 mV to +20 mV increasedP from near 0 to 0.6 but further polarization to +50 mV decreasedP to about 0.2. The pattern of voltage activation and inactivation was even more pronounced at [Ca2+]i=1 and 2 μm. In this situation a membrane potential change from −70 to +20 mV increasedP from near 0 to about 0.7 but further polarization to +80 mV reducedP to less than 0.1. The high-conductance K+ channel in rat pancreatic islet cells is remarkably sensitive to changes in [Ca2+]i within the range 0.1 to 1 μm which suggests a physiological role for this channel in regulating the membrane potential and Ca2+ influx through voltage-activated Ca2+ channels.

ATP maintains ATP-inhibited K+ channels in an operational state.

In patch-clamp records of K+ATP channels in an insulin-secreting cell line (RINm5F) inhibition evoked by exposing the internal surface of the membrane to ATP is followed not just by the recovery of K+ATP channel activity when the ATP is removed but by a marked activation of K+ATP channels. This phenomenon is not a direct consequence of channel closure as inhibition induced by quinidine and quinine is followed upon the removal of the drug only by the recovery of K+ATP channel activity and not by post-inhibitory activation. If ATP is applied to the exposed internal surface of a membrane patch when all of its K+ATP channel have run down subsequent removal of the ATP causes their activation. The magnitude and duration of the reactivation of K+ATP channels is shown to depend upon both the concentration of ATP and the length of time for which the membrane is exposed to ATP. We therefore have a paradoxical situation in that K+ channels which are inhibited by intracellular ATP require intracellular ATP to retain the ability to open.

Quinine inhibits Ca2+-independent K+ channels whereas tetraethylammonium inhibits Ca2+-activated K+ channels in insulin-secreting cells

The effects of quinine and tetraethylammonium (TEA) on single‐channel K+ currents recorded from excised membrane patches of the insulin‐secreting cell line RINm5F were investigated. When 100 μM quinine was applied to the external membrane surface K+ current flow through inward rectifier channels was abolished, while a separate voltage‐activated high‐conductance K+ channel was not significantly affected. On the other hand, 2 mM TEA abolished current flow through voltage‐activated high‐conductance K+ channels without influencing the inward rectifier K+ channel. Quinine is therefore not a specific inhibitor of Ca2+‐activated K+ channels, but instead a good blocker of the Ca2+‐independent K+ inward rectifier channel whereas TEA specifically inhibits the high‐conductance voltage‐activated K+ channel which is also Ca2+‐activated.

Interaction of diazoxide, tolbutamide and ATP4− on nucleotide-dependent K+ channels in an insulin-secreting cell line

SummaryThe single-channel current recording technique has been used to study the effects of diazoxide, tolbutamide and ATP, separately and combined, on the gating of nucleotide-regulated K+ channels in the insulin-secreting cell line RINm5F. The effects of diazoxide, tolbutamide and ATP4− were studied at the intracellular membrane surface, using, the open-cell membrane patch configuration. Alone diazoxide was found only inconsistently to evoke channel stimulation, 57% of all applications of the drug (72 times in 48 separate patches) having no effect at concentrations between 0.02 and 0.4mm. In the presence of ATP, however, diazoxide consistently evoked channel activation (seen 87 times in 49 patches, 95% of all applications). The interactions of diazoxide and ATP seemed competitive. Stimulation of channels by diazoxide in the presence of 1mm ATP was suppressed if the concentration of ATP was elevated to 2 or 5mm. In solutions in which Mg2+ had been chelated with EDTA, diazoxide failed to activate channels closed by 1mm ATP; however, this was not due to a direct effect on the channels caused by the absence of Mg2+, but could be explained by the enhanced ATP4− concentration after Mg2+ removal. When the total ATP concentration was lowered to give the same [ATP4−] in the absence of Mg2+ to that present in the control experiments, diazoxide was able to evoke full activation. Channel inhibition evoked by tolbutamide, 0.01 to 1.0mm, did not require the presence of either ATP or Mg2+. In the presence of ATP tolbutamide further reduced the number of channel openings. Diazoxide was able to compete with tolbutamide for control of channel activity, an effect that was augmented by the presence of ATP. In the presence of 0.1mm tolbutamide, diazoxide was unable to stimulate channel openings; however, if the dose of tolbutamide was lowered or ATP made available to the inside of the membrane, channel stimulation occurred.

The gating of nucleotide-sensitive K+ channels in insulin-secreting cells can be modulated by changes in the ratio ATP4−/ADP3− and by nonhydrolyzable derivatives of both ATP and ADP

SummaryThe31P-NMR technique has been used to assess the intracellular ratios and concentrations of mobile ATP and ADP and the intracellular pH in an insulin-secreting cell line, RINm5F. The single-channel current-recording technique has been used to investigate the effects of changes in the concentrations of ATP and ADP on the gating of nucleotide-dependent K+ channels. Adding ATP to the membrane inside closes these channels. However, in the continued presence of ATP adding ADP invariably leads to the reactivation of ATP-inhibited K+ channels, even at ATP4−/ADP3− concentration ratios greater than 7∶1. Interactions between ATP4− and ADP3− seem competitive. An increase in the concentration ratio ATP4−/ADP3− consistently evoked a decrease in the open-state probability of K+ channels; conversely a decrease in ATP4−/ADP3− increased the frequency of K+ channel opening events. Channel gating was also influenced by changes in the absolute concentrations of ATP4− and ADP3−, at constant free concentration ratios. ADP-evoked stimulation of ATP-inhibited channels did not result from phosphorylation of the channel, as ADP-β-S, a nonhydrolyzable analog of ADP, not only stimulated but enhanced ADP-induced activation of K+ channels, in the presence of ATP. Similarly, ADP was able to activate K+ channels in the presence of two nonhydrolyzable derivatives of ATP, AMP-PNP and βγmethylene ATP.

ATP-sensitive K+ channels in an insulin-secreting cell line are inhibited byd-glyceraldehyde and activated by membrane permeabilization

SummaryThe control of K+ channels in the insulin-secreting cell line RINm5F has been investigated by patch-clamp singlechannel current recording experiments. The unitary current events recorded from cell-attached patches are due to large and small inwardly rectifying ATP-sensitive K+ channels with conductance properties similar to the two channels previously identified in primary cultured rat islet cells (Findlay, I., Dunne, M.J., & Petersen, O. H.J. Membrane Biol.88:165–172, 1985). Cell permeabilization through brief exposure to 10 μm digitonin or 0.05% saponin (outside the isolated membrane patch area) results in a dramatic increase in current through the cell-attached patch due to opening of many large and small K+-selective channels. These channels are inhibited in a dose-dependent manner by ATP applied to the bath (near-complete inhibition by 5mm ATP). During prolonged ATP exposure (1–5 min) the initial inhibition is followed by partial recovery of channel activity, although further activation does occur when ATP is subsequently removed. From the maximal number of coincident channel openings in the permeabilized cells (in the absence of ATP), it is estimated that there are on average 12 large ATP-sensitive K+ channels per membrane patch, but in the intact cells less than 5% of the membrane patches exhibited three or more coincident K+ channel openings, indicating the degree to which the channels are inhibited in the resting condition by endogenous ATP. Stimulation of RINm5F cells to secrete insulin was carried out by challenging intact cells with 10mm d-glyceraldehyde.d-glyceraldehyde induced depolarization of the membrane from about −70 to −20 mV and evoked a marked reduction in the open-state probability of both the large and small ATP-sensitive channels.d-glyceraldehyde also induced action potentials in a number of cases. All effects of stimulation were largely transient, lasting about 100 sec. The two ATP-sensitive K+ channels are probably responsible for the resting potential and play a crucial role in coupling metabolism to membrane depolarization.

Protein phosphorylation is required for diazoxide to open ATP‐sensitive potassium channels in insulin (RINm5F) secreting cells

The patch‐clamp open‐cell recording configuration has been used to investigate the effects of non‐hydrolyzable analogues of ATP on the diazoxide‐activation of KATP channels in the insulin‐secreting cell line RINm5F. K+ channels inhibited by 0.1, 0.5 and 1.0 mM ATP were consistently activated by 200 μM diazoxide. During sustained activation of channels, exchange of ATP for either AMP‐PNP, AMP‐PCP or ATPγS abolished the effects of diazoxide. If diazoxide was added to the membrane in the continued presence of AMP‐PNP, AMP‐PCP or ATPγS either no effects were observed or alternatively a small transient activation of channels occurred. This study suggests that protein phosphorylation is necessary for diazoxide to activate ATP‐sensitive potassium channels in insulin‐secreting cells.

Vasopressin directly closes ATP-sensitive potassium channels evoking membrane depolarization and an increase in the free intracellular Ca2+ concentration in insulin-secreting cells.

The effects of arginine‐vasopressin (AVP) (0.01‐1 microM) on membrane potential, [Ca2+]i and ATP‐sensitive potassium channels have been studied in the insulin‐secreting cell line RINm5F. In whole cells, with an average spontaneous cellular transmembrane potential of ‐64 +/‐ 3 mV (n = 33) and an average basal [Ca2+]i of 102 +/‐ 6 nM (n = 40), AVP evoked: (i) membrane depolarization, (ii) voltage‐dependent Ca2+ spike‐potentials and (iii) a sharp rise in [Ca2+]i. Single‐channel current events recorded from excised outside‐out membrane patches show that AVP closes potassium channels that are also closed by tolbutamide (100 microM) and opened by diazoxide (100 microM). AVP acts on KATP channels specifically from the outside of the membrane and a soluble cytosolic messenger appears not to be involved, since there is no channel activation in cell‐attached membrane patches when the peptide is added to the bath solution.