(Received for publication, April 7, 1997)
From The Rolf Luft Center for Diabetes Research, Department of
Molecular Medicine, Karolinska Institute, S-171 76 Stockholm,
Sweden and Diabetes and Metabolism Unit, Evans Department
of Medicine, Boston University Medical Center,
Boston, Massachusetts 02118
The mechanism by which long chain acyl-CoA (LC-CoA) esters affect the ATP-regulated potassium channel (KATP channel) was studied in inside-out patches isolated from mouse pancreatic beta cells. Addition of LC-CoA esters dramatically increased KATP channel activity. The stimulatory effect of the esters could be explained by the induction of a prolonged open state of the channel and did not involve alterations in single channel unitary conductance. Under control conditions, absence of adenine nucleotides, the distribution of KATP channel open time could be described by a single exponential, with a time constant of about 25 ms. Exposing the same patch to LC-CoA esters resulted in the appearance of an additional component with a time constant of >150 ms, indicating a conformational change of the channel protein. LC-CoA esters were also able to potently activate channel activity at different ratios of ATP/ADP. Simultaneous additions of MgADP and LC-CoA esters resulted in a supra-additive effect on channel mean open time, characterized by openings of very long duration. Following modification of the KATP channel by a short exposure of the patch to the protease trypsin, the stimulatory effect of ADP on channel activity was lost while activation by LC-CoA esters still persisted. This indicates that LC-CoA esters and MgADP do not bind to the same site. We conclude that LC-CoA esters may play an important role in the physiological regulation of the KATP channel in the pancreatic beta cell by binding to a unique site and thereby inducing repolarization of the beta cell-membrane potential.
Potassium channels that are ATP-sensitive (KATP)1 are found in many types of cells and serve to couple metabolic state to electrical activity. In the pancreatic beta cell the KATP channel provide a critical link between changes in blood glucose concentration and insulin secretion (1, 2). The initial step in the stimulus-secretion-coupling in the beta cell is closure of the KATP channel subsequent to a rise in the ATP/ADP ratio, resulting in depolarization, activation of voltage-dependent Ca2+ channels and thereby triggering of insulin secretion (3). Stimulation of the beta cell with intermediate glucose concentrations results in a characteristic pattern of slow oscillations in membrane potential on which bursts of action potentials are superimposed (4). Intracellular free Ca2+ concentration ([Ca2+]i) oscillates in synchrony with electrical activity (5). We recently showed that fluctuations in the activity of the KATP channel underlie the oscillations in electrical activity and [Ca2+]i in single pancreatic beta cells (6). A possible mechanism underlying such oscillations in KATP channel activity could be metabolism-driven oscillations in the ATP/ADP ratio (7). Because of the close relation between KATP channel activity and beta cell electrical activity, it is essential to study mechanisms which control or modulate the activity of this channel.
We have recently shown that long-term exposure to free fatty acid increases cellular levels of LC-CoA esters in the beta cell and that these esters are able to directly stimulate KATP channel activity (8). This indicates that increased steady-state content of cytosolic LC-CoA esters could affect glucose-induced closure of the KATP channel. In the present study, we have investigated in detail the mechanisms by which LC-CoA esters exert their stimulatory action and to what extent they interact with ATP and ADP in modulating KATP channel activity in the pancreatic beta cell. These findings show that LC-CoA induces a distinct open state leading to increased channel activity, characterized by openings of long duration which does not require the presence of Mg2+. Thus, binding of LC-CoA induces a conformational change of the KATP channel protein. A potent stimulatory effect of LC-CoA esters also occurred in the presence of different ratios of ATP/ADP, indicating that the esters may play an important role in modulating the channel under physiological conditions.
Adult obese mice (gene ob/ob) of both sexes were obtained from a local noninbred colony (9). The mice were fasted for 24 h and then killed by decapitation. The islets of these mice contain more than 90% beta cells (10). Dispersed islets were isolated by a collagenase technique (11). Collagenase was obtained from Boehringer Mannheim GmbH, Germany. A cell suspension was prepared and washed essentially as described previously (12). The cells were resuspended in RPMI 1640 culture medium (Flow Laboratories, Scotland, UK), containing 11 mM glucose, supplemented with 10% fetal bovine serum, 100 IU/ml penicillin, 100 µg/ml streptomycin, and 60 µg/ml gentamycin. The cell suspension was seeded into Petri dishes (Corning Glass Works, Corning, NY) and incubated at 37 °C in 5% CO2 for 1-3 days.
SolutionsThe standard extracellular solution contained (in mM): 138 NaCl, 5.6 KCl, 1.2 MgCl2, 2.6 CaCl2, and 5 HEPES-NaOH at pH 7.4. The intracellular solution (i.e. the bath solution) consisted of (in mM): 125 KCl, 1 MgCl2, 10 EGTA, 25 KOH, and 5 HEPES-KOH at pH 7.15. ATP was added as the Mg2+ salt to the intracellular solution as shown in the text and figures. Cis-9-monounsaturated oleoyl-CoA (C18:1), myristoyl-CoA (C14:0), and malonyl-CoA (C3:0) were prepared as stock solutions with a concentration ranging from 1 to 5 mM in deionized water (Millipore) and then added to the intracellular solution with final concentrations as indicated in the figures. Nucleotides and CoA esters were all from Sigma. All reagents were of analytical grade.
ElectrophysiologyLC-CoA esters reach their target site by
partitioning into the lipid phase of the cell membrane. However, the
CoA moiety is hydrophilic and not able to flip across the membrane
(13). Like other endogenous modulators of KATP channel
activity, such as adenine nucleotides, the CoA esters are not effective
when applied to the extracellular face of the cell membrane (14).
Therefore, excised inside-out patches from pancreatic beta cells were
used to study the effects of the CoA esters on KATP channel
activity (15). This type of recording mode allows free access to the cytoplasmic side of the plasma membrane, making it easy to vary the
intracellular composition. Pipettes were pulled from alumina- or
borosilicate glass (Hilgenberg, Malsfeld, Germany), coated with Sylgard
near the tips to reduce electrical noise and then fire-polished. The
electrodes had resistances between 3-5 M, when filled
with standard extracellular solution. Recordings were made using an
Axon patch-clamp amplifier (Axopatch 200, Axon Instruments, Burlingame,
CA). During the experiment the current signal was stored on magnetic
tape using a VCR (Sony-200, Sony, Tokyo, Japan). Recordings of
KATP channel activity were made with the membrane potential
(Vm) of the patches clamped at 0 mV. With the solutions used, K+ currents are outward (i.e.
into the pipette) and channel records are displayed according to the
convention with upward deflections denoting outward ion current. The
KATP channel activity was identified based on the unitary
amplitude (1.5-2 pA) and the sensitivity to ATP. The experiments were
carried out at room temperature (22-24 °C). The zero current
potential of the pipette was adjusted with the electrode in the bath,
before establishment of the seal. Patches were excised into
nucleotide-free solution and 0.1 mM ATP was first applied
to test for channel inhibition. After channel inhibition, ATP was
removed and the patch was subsequently exposed to the test solutions
indicated in text and figures.
Records were filtered at 200 Hz (3 db
value, 8-pole Bessel filter, Frequency Devices, Haverhill, MA),
digitized at 800 Hz using an Axon Instrument analogue digital converter
(TL-1) and stored in a computer. For trace figures, digitized
recordings were exported into CorelDraw (Corel Inc., Ontario, Canada)
for final layout. Digitized segments of current records (30-40 s long) were also used to determine channel activity using in-house software. The mean current (iX) was calculated according to
the equation
![]() |
(Eq. 1) |
Recently, we discovered that a new group of substances, LC-CoA esters, can act as potent KATP channel openers and even counteract the blocking effect of ATP (8). Another important finding is that the esters appear to specifically affect the KATP channel, in that channel activity of at least two other K+ channels present in the beta cell are unaltered by the esters, namely the large conductance K+ channel (KBK) and the 8-pS K+ channel (8). However, little is known about the mechanisms underlying the stimulatory action of these esters or how LC-CoA esters interact with ADP, which also stimulates channel activity. The present study therefore focuses on effects of LC-CoA esters on single channel kinetics and on the interaction between the esters and adenine nucleotides.
Effects of LC-CoA Esters on KATP Channel KineticsFig. 1, A and B,
show channel activity following administration of a 3-carbon (malonyl),
14-carbon (myristoyl), and 18-carbon (oleoyl) CoA ester to inside-out
patches shortly after isolation. It is clear that KATP
channel activity, in the presence of myristoyl-CoA and oleoyl-CoA, was
increased compared with the activity under control conditions or with
malonyl-CoA containing solutions. In Fig. 1, C-F, we have
quantified this effect by analyzing the distribution of channel open
time during exposure of the patches to oleoyl-, myristoyl-, and
malonyl-CoA. In control solution, channel activity consisted of short
openings (Fig. 1, C, inset). The distribution of
the openings was best described by a single exponential with a time
constant () of 19.8 ms. Mean open time was estimated to be 29.3 ± 4.5 ms (n = 5). In the presence of oleoyl-CoA, there were two types of channel openings, short openings, similar to those
observed under control conditions, and long openings, occasionally lasting several hundred milliseconds (Fig. 1D). The
distribution of the openings was best described as the sum of two
exponentials with
values of 49.6 and 260.4 ms, respectively. The
slow component comprised 27.6% of the events and the mean open time
was increased approximately 3-fold to 84.6 ± 29.0 ms
(n = 5). Similar results were obtained when adding
myristoyl-CoA to the patch (Fig. 1E), resulting in
values of 46.1 and 219.1 ms, respectively, with 19.4% belonging to the
slow component. However, when applying malonyl-CoA (Fig.
1F), channel activity did not differ from control conditions
with a mean open time of 34.0 ± 9.4 ms (n = 5).
The distribution could be described by a single exponential with a time
constant of 28.9 ms.
We have previously reported comparable effects on KATP channel open time induced by ADP and diazoxide (17). Thus, ADP and diazoxide increase channel activity by promoting a similar long lasting open state. We now demonstrate that, when measured under physiological ionic conditions, LC-CoA esters also induced a conformational change of the KATP channel, leading to a prolonged open state. Another possibility could be that LC-CoA esters affect single channel unitary conductance. We therefore estimated KATP channel conductance in the absence and presence of LC-CoA esters. Under control conditions, the channel conductance was 18.0 ± 0.97 pS (n = 6). The conductance was not significantly changed in the presence of oleoyl- and myristoyl-CoA, 17.7 ± 0.73 pS (n = 7) and 18.1 ± 0.22 pS (n = 4), respectively. Exposing patches to malonyl-CoA was also without effect on channel conductance (17.9 ± 1.2 pS; n = 4). Thus, the stimulatory effect of LC-CoA esters on KATP channel mean current can not be explained by effects on single channel unitary conductance.
Do ADP and LC-CoA Bind to the Same Site?In many respects,
the LC-CoA esters seem to affect KATP channel activity in a
manner similar to ADP. Thus, the ability to counteract the blocking
effect of ATP, prevent channel run down (8), and increase channel open
time without affecting single channel conductance are characteristics
shared by the two compounds. In this context it should be noted that
the CoA moity has a close structural resemblance to ADP, suggesting the
possibility of competition for a common binding site. However, on a
molar basis, LC-CoA esters are considerably more potent than ADP, and
the esters also induce a significantly higher degree of channel
stimulation. To what extent this difference between ADP and LC-CoA
esters can be accounted for by an additional 3-phosphate group on the
CoA moity is not clear. Nevertheless, this may indicate that LC-CoA
esters interact at a site different from that of ADP on the
KATP channel complex. In Fig. 2A,
perifusions with 1 µM oleoyl-CoA enhanced channel activity 5-fold. An addition of 0.1 mM MgADP, in the
continuous presence of oleoyl-CoA, further increased mean current by
720 ± 220% (n = 4; p < 0.001).
Noteworthy is that channel openings were characterized by long
openings. As MgADP was withdrawn, the long openings of the channel
disappeared. In trace B, addition of 0.1 mM
MgADP led to an increase in mean current of 320 ± 100% (n = 3). Channel activity declined significantly as the
MgADP concentration was increased to 0.5 mM. Inhibition of
KATP channel activity at higher concentrations (>0.3
mM) of MgADP is a well documented effect (18-20). Exposing
the same patch to oleoyl-CoA, in the continuous presence of MgADP,
induced a dramatic augmentation in channel activity with a mean current
increase of 690 ± 140% (n = 4). In C,
we have quantified the effects by making amplitude-histograms under the
different conditions tested in the recording of trace B.
Compiled data on mean open time clearly show that a combination of
MgADP and LC-CoA ester led to a supra additive effect (Fig. 3).
The fact that a combination of ADP and LC-CoA esters activate the
channel to a larger extent than administration of either of the two
substances alone, suggests that they interact at distinct binding
sites. In an attempt to obtain more information on this matter, we have
tried modifying the KATP channel by applying a short pulse
of trypsin. This approach was used earlier in studies of various types
of ion channels including Na+ (21), Ca2+ (22),
and K+ channels (23). Although the technique seems crude,
it modifies the KATP channel in very specific ways (24).
There is a resulting complete loss of the stimulatory effect of ADP as
well as of the inhibitory effect of sulfonylurea on channel activity.
However, inhibition of channel activity by ATP remains intact, although with slightly decreased sensitivity (24). One interpretation is that
the binding sites for ADP and sulfonylurea are lost due to alterations
of the channel proteins as a result of proteolytic effects of trypsin,
which has a primary affinity for arginine and lysine residues (25).
Interestingly, in trypsin modified patches, where ADP was totally
ineffective in altering channel activity, addition of LC-CoA esters
induced a pronounced increase in channel activity. LC-CoA esters were
also able to potently counteract ATP-induced inhibition of channel
activity in modified patches (Fig. 4). A possible
explanation for these results is that trypsin alters or removes the
site to which ADP binds to exert activation, whereas the site involved
in LC-CoA-induced stimulation remains. These data further support the
notion that LC-CoA esters interact at a unique binding site, separate
from that of ADP.
To further study the interaction between LC-CoA esters and ADP, we
performed a series of experiments under Mg2+-free
conditions, since the ability of ADP to open channels requires Mg2+. As shown in Fig. 5A, LC-CoA
activated the KATP channel in the absence of
Mg2+, whereas ADP3 not only failed to
activate but had an inhibitory effect on KATP channel
activity (Fig. 5B) (17). This blocking effect of
ADP3
is well documented and it has been proposed that
ADP3
binds to the ATP site thereby explaining the
inhibitory effect on the channel (20). Addition of oleoyl-CoA to the
patch, in the continuous presence of ADP3
, still evoked a
dramatic elevation in channel activity (Fig. 5B). Analyzing
the effects of simultaneous additions of oleoyl-CoA and
ADP3
in the absence of Mg2+ in five patches
showed an increase of 590 ± 280% in mean currents (p < 0.01). Inclusion of Mg2+ in the
perifusion medium caused the KATP channel activity to display openings of long duration. Together these observations lend
strong support to the idea that simultaneous exposure of the
KATP channel to a combination of LC-CoA esters and MgADP, results in a unique activity pattern with extremely long open times,
not previously observed with either substance alone. All effects of the
LC-CoA esters were fully reversible upon withdrawal. Due to the extreme
pattern and the resulting high number of open channels in the patches,
a precise determination of channel open time was not possible.
Earlier studies suggested that the physiological regulation of the
KATP channel results from changes in the ATP/ADP ratio (3),
changes in ADP exerting the major influence (1). This implies that
ATP-induced blockade of the channel is potently counteracted by
intracellular ADP. We therefore assessed the extent to which LC-CoA
esters were able to further activate channel activity in the presence
of fixed ATP/ADP ratios. Fig. 6A shows the
effect of 100 µM ATP and ADP on channel activity, a ratio
which has been reported to give maximal stimulation of KATP
currents. Subsequent addition of 1 µM oleoyl-CoA, in the
continuous presence of nucleotides, resulted in an augmentation of the
KATP currents. In six out of six patches we found that
addition of oleoyl-CoA, in the presence of 100 µM ATP and
ADP, increased mean currents significantly (380 ± 140%;
p < 0.01; n = 5). Adding LC-CoA esters
to 500 µM ATP and ADP also induced an increase in
KATP channel activity (Fig. 6B). It should,
however, be pointed out that the most dramatic effects were seen when
the nucleotides and CoA ester were washed out. Thus, just after
withdrawal of the substances, we repeatedly observed the same channel
activity pattern as following the combination of MgADP and LC-CoA ester
(see Fig. 3), characterized by channel activity with very long
openings. A possible explanation for this phenomena is that ADP is
washed-out more slowly than ATP (26), leaving MgADP and LC-CoA at their
binding sites. Even at an ATP/ADP ratio of 10, administration of 1 µM oleoyl-CoA potently increased channel activity (Fig.
6C).
Concluding Remarks
The recent cloning of the rat
sulfonylurea receptor (SUR1) (27) combined with the reconstitution
of the KATP channel (28) has elegantly shown that the beta
cell KATP channel is encoded by the SUR1 and an inwardly
rectified K+ channel (Kir6.2) (Fig. 7) with
small intrinsic activity. The SUR1 belongs to a superfamily of
ATP-binding cassette proteins and comprises a 13-membrane spanning
segment and two cytosolic nucleotide binding folds (NBFs) (27). It is
not clear which physiological agonists interact with the two NBFs.
Recent data indicate that NBF-2 is involved in ADP binding, since
mutation of this site leads to a lack of stimulatory effect of ADP
which is no longer able to counteract the blocking effect of ATP (29). To what extent LC-CoA esters are interacting with the NBFs is at
present not known. Although it is generally believed that the main
physiological activator of the KATP channel in the
pancreatic beta cell is ADP, we now confirm that LC-CoA esters are
potent activators of the channel (8) and that the kinetic effects on
channel activity is similar to those of ADP. Furthermore, the LC-CoA
esters are able to activate the channel in the absence of nucleotides,
at various concentrations of ATP and ADP as well as to counteract the
blocking effect of ATP (8).
The fact that a combination of ADP and LC-CoA ester resulted in supra-additive effects on mean channel open time and that trypsin fully eliminates the stimulatory effect of ADP, under conditions where CoA esters still stimulated the channel, strongly suggest that ADP and LC-COA esters do not bind to the same site. Thus, LC-CoA esters form a class of substances which with high potency activates the beta cell KATP channel. The fact that most of our results were obtained by CoA esters derived from oleate, which is one of the predominant free fatty acid components in rodent and man plasma, supports the notion that these esters may serve the function of important modulators of beta cell electrical activity and thereby insulin release under physiological conditions.