| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Cell Biology and Physiology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261
 |
ABSTRACT |
|---|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
We used
single-channel recording techniques to identify and characterize a
large-conductance,
Ca2+-independent
K+ channel in the colonic
secretory cell line T84. In symmetric potassium gluconate, this channel
had a linear current-voltage relationship with a single-channel
conductance of 161 pS. Channel open probability
(Po) was
increased at depolarizing potentials. Partial substitution of bath
K+ with
Na+ indicated a permeability ratio
of K+ to
Na+ of 25:1. Channel
Po was reduced by
extracellular Ba2+. Event-duration
analysis suggested a linear kinetic model for channel gating having a
single open state and three closed states: C3
C2C1O.
Arachidonic acid (AA) increased the
Po of the
channel, with an apparent stimulatory constant
(Ks)
of 1.39 µM. Neither channel open time (O) nor the fast closed time
(C1) was affected by AA. In
contrast, AA dramatically reduced mean closed time by decreasing both
C3 and
C2. The
cis-unsaturated fatty acid linoleate increased Po
also, whereas the saturated fatty acid myristate and the
trans-unsaturated fatty acid elaidate
did not affect
Po. This channel
is activated also by negative pressure applied to the pipette during
inside-out recording. Thus we determined the effect of the
stretch-activated channel blockers amiloride and Gd3+ on the
K+ channel after activation by AA.
Amiloride (2 mM) on the extracellular side reduced single-channel
amplitude in a voltage-dependent manner, whereas
Gd3+ (100 µM) had no effect on
channel activity. Activation of this K+ channel may be important during
stimulation of Cl
secretion
by agonists that use AA as a second messenger (e.g., vasoactive
intestinal polypeptide, adenosine) or during the volume regulatory
response to cell swelling.
potassium channel; chloride secretion; fatty acids; intestine
| |
INTRODUCTION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
INTESTINAL CHLORIDE SECRETION is modulated by changes
in cellular adenosine 3',5'-cyclic monophosphate (cAMP)
[e.g., vasoactive intestinal polypeptide (VIP)], guanosine
3',5'-cyclic monophosphate (cGMP) (e.g., heat-stable
enterotoxin of Escherichia coli),
and Ca2+ (e.g., acetylcholine) in
response to secretory agonists (3, 4, 7, 24). Adenosine is released by
invading neutrophils and eosinophils during an inflammatory response
and thus may act as a stimulus for diarrhea in intestinal inflammatory
diseases (31, 32, 41). Although adenosine was shown to be a potent Cl secretagogue, changes in
these classical second messengers were not detected (6, 12, 32).
However, adenosine was shown to increase cellular arachidonic acid (AA)
levels in T84 cells, and its secretory effects could be blocked by
inhibitors of both phospholipase
A2 and 1,2-diacylglycerol (DAG)
lipase (5, 6); both enzymes catalyze AA generation (36). More recently,
VIP (cAMP) and E. coli heat-stable
enterotoxin (cGMP) have been shown to increase AA in T84 cells (5), in
addition to their effects on cAMP production. Inhibition of DAG lipase
resulted in a partial inhibition of these
Cl secretory responses (5).
These results suggest that AA may be an important second messenger in
modulating Cl secretion in
intestinal diarrheal disease. However, the apical and basolateral ion
channels responsible for carrying the secretory current in response to
AA have not been identified.
AA is a well-known modulator of
K+,
Na+,
Ca2+, and
Cl channels in a variety of
tissues (33, 38). Importantly, fatty acids have been shown to modulate
Cl channels in secretory
epithelia. AA was shown to inhibit an outwardly rectifying
Cl channel in airway
epithelia (2, 17) as well as a volume-sensitive Cl conductance in
intestinal epithelium (23). Also, cAMP-mediated Cl secretion is inhibited
by AA in airway epithelia (34). However, these inhibitory effects are
incongruent with the secretory response associated with elevated AA. In
our companion paper (10), we demonstrate that the basolateral membrane
K+ channel activated by
Ca2+-dependent agonists
(KCa) is potently inhibited by
AA. To our knowledge, neither an intestinal
K+ conductance
(GK) nor
Cl conductance has been
shown to be upregulated directly by AA, as would be required if
elevated AA stimulates Cl
secretion under some conditions. During the course of our studies on
the role of AA in regulating KCa,
we identified a large-conductance K+ channel that is activated by
AA. We have characterized this channel's K+/Na+
selectivity, blocker sensitivity, and its regulation by fatty acids. We
speculate that this channel may be important in carrying K+ across the basolateral membrane
during Cl secretory
responses in which AA serves as a stimulatory second messenger.
| |
METHODS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
Cell culture. T84 cells were grown in Dulbecco's modified Eagle's medium and Ham's F-12 (1:1) supplemented with 15 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 14 mM NaHCO3, and 10% fetal bovine serum. The cells were incubated in a humidified atmosphere containing 5% CO2 at 37°C. Patch-clamp experiments were performed on single cells plated onto glass coverslips 18-48 h before use.
Solutions.
During inside-out patch-clamp recordings, the bath contained (in mM)
145 potassium gluconate, 5 KCl, 1 MgCl2, 1 ethylene
glycol-bis(
-aminoethyl ether)-N,N,N',N'-tetraacetic
acid (EGTA), and 10 HEPES (pH adjusted to 7.2 with KOH). A free
Ca2+ concentration of <10 nM was
chosen for these experiments to eliminate the activity of the
KCa we previously described in
these cells (9). The pipette solution contained (in mM) 140 potassium
gluconate, 5 KCl, 1 CaCl2, 1 MgCl2, and 10 HEPES (pH adjusted
to 7.2 with KOH). For outside-out recordings, the bath contained 1 mM
CaCl2 in the absence of any added
EGTA, whereas the pipette solution Ca2+ was buffered to <10 nM with
EGTA (1 mM).
Single-channel recording.
Single-channel currents were recorded using a List EPC-7 amplifier
(Medical Systems) and recorded on videotape for later analysis as
described previously (9). Pipettes were fabricated from KG-12 glass
(Willmad Glass). All recordings were done at a holding voltage of
100 mV unless otherwise noted. The voltage is referenced to the
extracellular compartment, as is the standard method for membrane
potentials. Inward currents are defined as the movement of positive
charge from the extracellular compartment to the intracellular compartment and are presented as downward deflections from baseline in
all recording configurations.
o) and fast closed time
(c1) were constructed by
binning at the sampling rate (125 µs) from 1.0 to 6-15 ms (i.e., >4 times the time constant). To determine the burst duration
(burst) and the long closed
times (c2 and
c3), the channel recordings were filtered at 100 Hz and sampled at 500 Hz to eliminate the contribution of c1 from the
recordings. An open-closed transition was considered valid if it
remained in the state for at least two sample periods (4 ms). For
burst, the data were binned at 4-ms intervals for construction of the event-duration histogram. Because of the limited number of long closed events under control conditions, an exponential fit of the data could not be achieved from a
single record. Therefore, the event-duration histogram for
c2 and
c3 was constructed after the
concatenation of eight separate recordings with a
Po >3%
(average Po = 0.08). The additional five single-channel recordings had
Po
<3% such that these recordings would contribute very few long
closed events. The event-duration histogram was constructed by binning
at 10-ms intervals. These event-duration histograms were then fit to
exponential functions to determine the open
(o) and closed
(c) time constants.
Chemicals.
All fatty acids were obtained from Sigma Chemical, made as
>1,000-fold stock solutions in dimethyl sulfoxide (DMSO), and stored under N2 at 80°C. The
fatty acids were dissolved to the final working concentration just
before use, and all solutions were continuously bubbled with
N2 during perfusion through the
patch-clamp chamber. Charybdotoxin (CTX) was made as a 10 µM stock
solution in our bath solution for outside-out recording.
4-Aminopyridine (4-AP), quinine, glibenclamide, and
trans-6-cyano-4-(N-ethylsulfonyl-N-methylamino)-3-hydroxy-2,2-dimethyl-chromane (293B) were made as a 1,000-fold stock solution in DMSO. 293B was a
generous gift from Dr. Rainer Greger (Albert-Ludwigs-Universtat, Freiberg, Germany). Indomethacin and nordihydroguaiaretic acid (NDGA)
were obtained from Biomol. Cell culture medium was obtained from GIBCO.
Data analysis. All data are presented as means ± SE, where n indicates the number of experiments. Statistical analysis was performed using the Student's t-test. A value of P < 0.05 was considered statistically significant.
| |
RESULTS |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
After excision of membrane patches from T84 cells into symmetric
K+ in the absence of bath
Ca2+ (<10 nM), a
large-conductance channel was observed in ~10-15% of
recordings. Single-channel currents for one such patch are shown in
Fig. 1 in the absence of AA
(A). The average current-voltage (I-V)
relationship for eight recordings is shown in Fig.
1B. The channel displayed a linear
I-V
relation with an average conductance of 161 ± 3 pS
(n = 8). The
K+/Na+
selectivity of this channel was assessed by equimolar substitution of
Na+ for
K+ in the bath solution;
K+-to-Na+
permeability ratio
(PK/PNa)
was calculated from the Goldman-Hodgkin-Katz relation. Substitution of
100 meq Na+ for
K+ resulted in a shift in the
reversal potential of 26 ± 1 mV
(n = 3). A shift of 27 mV is predicted
for a perfectly K+-selective
electrode. These data yield a calculated
K+/Na+
selectivity of ~25:1. As is apparent from Fig. 1, this channel exhibited modest voltage dependence, with the
Po increasing at depolarizing potentials. In five patches, the
Po at 100
mV was 0.17 ± 0.02, and this increased to 0.59 ± 0.05 at +100 mV.
|
Effect of AA on K+ channel activity. The effect of AA on this large-conductance, Ca2+-independent K+ channel is shown in Fig. 2A. Under control conditions, the channel opened infrequently. However, superfusion of AA (3 µM) resulted in a rapid increase in channel activity. The average concentration-Po curve is shown in Fig. 2B. These data were fit to a Michaelis-Menten function with an apparent stimulatory constant (Ks) of 1.39 µM. At a maximally effective concentration (10 µM), AA increased the Po from 0.04 ± 0.01 (n = 13) to 0.66 ± 0.03 (n = 4). On the basis of this AA-dependent activation, this channel will be referred to as KAA throughout this paper.
|
o) is defined by a single
exponential (2.4 ms), indicating a single open state (Fig.
3A). In 13 patches,
o averaged 2.3 ± 0.2 ms.
The closed time (c) of
KAA was described by three states:
a short-lived closed state
(c1) and two long-lived
closed states (c2 and
c3). As illustrated for one
patch (Fig. 3B), the short-lived
closed state was defined by a single exponential
(c1 = 1.0 ms).1
In 13 patches, c1 averaged 1.3 ± 0.2 ms. Because of the limited number of long closed events in
any one single-channel recording, we are unable to determine
c2 and
c3 from a single patch.
Therefore, we concatenated eight recordings that had a
Po of >3%
(average Po = 0.08; see METHODS). These data were
fit by two exponentials (Fig. 3C),
where c2 = 14.9 ms and
c3 = 106 ms. These results indicate that KAA is kinetically
described by one open and three closed states in the absence of AA (see
DISCUSSION).
|
o (Fig.
3D) of
KAA (2.6 ms). In 11 experiments, the o averaged 2.2 ± 0.2 ms
after activation by AA (3 µM), which is not different from control.
In contrast, AA induced a concentration-dependent decrease in mean
closed time from 266 ± 125 ms in control solutions to 2.6 ± 1.0 ms in the presence of 3 µM AA (n = 11; P < 0.05). As shown for one
patch in Fig. 3E,
c1 was independent of AA
concentration (1.2 ms). In 11 patches,
c1 averaged 1.5 ± 0.1 ms,
which is not different from control. Because of the limited number of
long closed events in the presence of AA, we are unable to determine c2 and
c3 in these experiments. These
data demonstrate that AA increases the
Po of
KAA by reducing the long closed
times (c2 and
c3) of the channel, thereby
increasing the channel opening rate (see
DISCUSSION).
The effect of AA on burst duration
(burst) was determined after
filtering of the data at 100 Hz to eliminate the fast closed events
(c1) that punctuate the open
channel burst (see METHODS). In one
patch, burst was 16.7 ms (Fig.
4A), and
this increased to 35.2 ms in the presence of 3 µM AA (Fig.
4B). In eight experiments, burst averaged 19.8 ± 4.3 ms, and this was increased to 46.7 ± 4.3 ms
(P < 0.01) in the presence of AA (3 µM).
|
Effect of cyclooxygenase, lipoxygenase, and cytochrome P-450 inhibitors. AA is rapidly metabolized by the cyclooxygenase-, lipoxygenase-, and cytochrome P-450-dependent oxidative pathways. Thus we determined whether inhibitors of these pathways would alter the ability of AA to activate KAA. Neither phenidone (250 µM), an inhibitor of both cyclooxygenase and lipoxygenase, a combination of cyclooxygenase (indomethacin; 1 µM) and lipoxygenase (NDGA; 2 µM) inhibitors, nor clotrimazole (1 µM), an inhibitor of cytochrome P-450, altered the ability of AA to active the channel (data not shown). Thus these oxidative metabolites are not responsible for the increased gating observed.
Effects of additional fatty acids.
We next determined whether the observed stimulatory effect of AA on
KAA was specific for AA or whether
additional fatty acids would also modulate its activity. For these
experiments, we used an additional
cis-unsaturated fatty acid, linoleic
acid (C18; cis,cis-
9,12),
the trans-unsaturated fatty acid
elaidic acid (C18;
trans-9),
and a saturated fatty acid, myristic acid
(C14). The results of one
experiment are shown in Fig.
5A.
Elaidic acid (3 µM) failed to increase
Po above control
levels. The subsequent addition of linoleic acid (3 µM) induced a
distinct increase in channel
Po, and the
further addition of AA (3 µM) resulted in an additional activation.
The data for these fatty acids plus myristic acid are summarized in
Fig. 5B. Similar to elaidic acid,
myristic acid failed to activate
KAA, whereas linoleic acid was
~40% as effective as AA in activating the channel.
|
Effects of K+
channel blockers.
Although AA has been implicated in modulating
Cl secretion in T84 cells,
no blocker pharmacology has been defined for the GK that
participates in this secretory response. Because of the very low
Po of this
channel under control conditions, all of our blocker studies were
carried out in the presence of 3 µM AA to activate the channel. In
initial experiments, we determined whether Ba2+ would block
KAA from the cytoplasmic side in
an inside-out patch. The results of one experiment are shown in Fig.
6. Initially, the patch voltage was held at
+80 mV. Application of Ba2+ (1 mM)
to the inside of the channel resulted in a nearly complete inhibition
of activity, which was subsequently relieved by changing the clamp
potential to 80 mV. This voltage-dependent block is typical for
Ba2+ inhibition of
K+ channel currents (47).
|
100 mV while
having no apparent effect at +60 mV. In four patches, 3 mM
Ba2+ reduced mean current
(I) by 90 ± 1% at 100 mV
(P < 0.01). Although cytoplasmic
Ba2+ blocked
KAA by inducing a long-lived
closed state (Fig. 6), the inhibition of
KAA by extracellular
Ba2+ was accompanied by an
apparent reduction in single-channel amplitude (i; Fig. 7), although this could not
be clearly resolved at this high concentration of
Ba2+. Therefore, we determined the
effect of 1 mM Ba2+ in three
additional patches. Ba2+ (1 mM)
reduced Po from
0.54 ± 0.02 to 0.25 ± 0.02 at 100 mV (n = 3), and this was accompanied by
an apparent reduction in i from 17.6 ± 0.6 to 14.9 ± 1.0 pA (n = 3;
P < 0.05). These results suggest
that Ba2+ blocks
KAA by interacting with two
distinct binding sites from the intra- and extracellular side of the
channel.
|
100 mV, i was reduced 21.8 ± 0.2% (P < 0.01),
whereas at +100 mV, this inhibition was 7.2 ± 1.1%
(n = 4;
P < 0.01). We previously demonstrated that CTX was a potent blocker of KCa
(inhibition constant = 3.5 nM) in both single-channel and Ussing
chamber studies (9, 11). In contrast to this property of
KCa, CTX (50 nM) failed to
inhibit KAA (data not shown).
Lohrmann et al. (28) characterized a novel inhibitor of the
cAMP-dependent K+ channel in
rabbit colon, 293B. We recently demonstrated that this
compound inhibited the cAMP-mediated
Cl secretion across T84
monolayers while not affecting
Cl secretion dependent on
KCa (11). In three outside-out
patches, 293B (30 µM) failed to inhibit
KAA, suggesting it is not the
cAMP-activated K+ channel.
Additionally, 4-AP (4 mM; n = 2),
glibenclamide (300 µM; n = 3), and
quinine (300 µM; n = 5)
all failed to inhibit channel activity.
Effect of membrane stretch on K+ channel activity. During the course of our studies, we observed that negative pressure applied to the pipette during inside-out recordings resulted in the activation of a large-conductance channel that appeared to have identical characteristics to the channel activated by AA (Fig. 8, top). Therefore, we wished to determine whether this stretch-activated channel was in fact the same as the AA-dependent K+ channel we have characterized. Application of a submaximal concentration of AA (2 µM) to the same patch that was shown to possess the stretch-activated channel resulted in the activation of KAA (Fig. 8, bottom), suggesting they are the same channel. In the presence of AA, application of negative pressure resulted in the further activation of the same channel rather than an additional conductance state. This result demonstrates that both AA and stretch are capable of activating this large-conductance, Ca2+-independent K+ channel.
|
Effect of stretch-activated channel blockers on
KAA.
The above results led us to determine whether the known inhibitors of
stretch-activated ion channels, amiloride (25, 44) and
Gd3+ (49), would block
KAA in excised, outside-out
patches. For these experiments, we used AA to first activate the
channel. Gd3+ (100 µM) had no
effect on KAA
(n = 3; data not shown). In contrast, amiloride induced an apparent reduction in
i at both +100 and 100 mV. The
effect of 2 mM amiloride on KAA in
an outside-out patch is shown in Fig. 9. In
this patch, i was reduced from 15.2 to
7.0 pA at +100 mV and from 17.0 to 3.6 pA at 100 mV. In four patches, amiloride (2 mM) reduced i at
100 mV from 18.5 ± 0.6 to 3.6 ± 0.1 pA
(P < 0.001). This inhibition of
i resulted in an 80 ± 3.5%
reduction in I with no apparent
reduction in channel Po (control = 0.43 ± 0.04; amiloride = 0.43 ± 0.06). These results suggest
that amiloride induces a voltage-dependent reduction in i of
KAA. Unfortunately, we were not
able to routinely maintain patch seal integrity at positive voltages,
so the voltage dependence of this inhibition could not be
quantitatively determined.
|
| |
DISCUSSION |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
We have characterized a large-conductance (160 pS), voltage-dependent, highly K+-selective channel from the colonic cell line T84. We demonstrate that this channel gates independently of changes in cytoplasmic Ca2+. However, we cannot rule out the possibility that Ca2+ might modulate this channel's activity. Our experiments were done in a nominally free Ca2+ solution to inhibit the activity of the inwardly rectifying KCa that we previously described (9). Because KCa is observed in >90% of all patches, we could not obtain patches in which KAA was present in the absence of KCa. Several investigators have described a "maxi" K+ channel from the surface cells of intestinal epithelium (22, 35, 46). Although these channels were inhibited by CTX, the channel described here is not. Large-conductance (130-190 pS) K+ channels have also been described in crypts from rabbit distal colon (8, 29). Similar to our findings, the 190-pS channel described by Burckhardt and Gogelein (8) from distal colonic crypts was observed in only 10% of patches and exhibited a similar blocker pharmacology; the channel was not sensitive to block by CTX, although extracellular Ba2+ did inhibit the channel. Thus a channel in native epithelium studied in vitro expresses many of the same characteristics as described here.
Mechanism of AA-induced activation. Our results demonstrate that the activation of KAA by AA is not due to an oxidative metabolite of AA; neither phenidone, indomethacin, NDGA, nor clotrimazole inhibited the AA-induced activation. Rather, we speculate that AA itself activates KAA. In addition, the levels of AA required to activate KAA (Ks = 1.4 µM) are likely to be physiologically relevant, since both cyclooxygenase and lipoxygenase have Michaelis constants of 3-5 µM for AA oxidation in vitro (36).
Based on event-duration analysis data, we demonstrate that KAA is minimally described by a single open state (O), a fast closed state (C1), and two long-lived closed states (C2, C3). We predict a linear state diagram of the form
|
C1 transition
(k0,1) can be
directly determined from the inverse of the time constant
(1/o = 435 s1). An initial
approximation of the additional rate constants was obtained from their
respective time constants. Two of these can be reliably approximated.
The opening rate
k1,0 (769 s1) can be determined
from c1, since this transition
constitutes the majority of closed events. This is especially true in
the presence of AA, where
Po is defined by
the open time divided by the open plus fast closed times
(o/o + c1).
c1 is independent of AA
concentration. In addition, we can approximate the
C1C2 transition rate from the burst duration analysis. In this case, C1 was eliminated by filtering the
data at a low frequency such that the transition rate of interest,
OC2, can be determined from 1/burst (50.5 s1). Using channel
simulation software (Biopatch, version 3.11), we determined that these
first approximations did not reliably predict channel behavior under
control conditions; these rate constants predicted a
Po of 0.35 (the
channel records used to determine
c2 and
c3 had a
Po of 0.08). With
k0,1,
k1,0, and k1,2 held
constant, the additional rate constants were determined based on their
ability to reliably predict the observed channel behavior. The rate
constants shown above predict a
Po of 0.08, a
o of 2.6 ms, and
c1,
c2, and
c3 of 1.3, 15.8, and 200 ms, respectively. Eliminating events shorter than 4 ms (3 times
c1) from the simulated data
indicated a burst of 31 ms.
Thus this kinetic scheme reliably predicts the observed channel
behavior.
We used this kinetic scheme to predict which state was being affected
by AA. If AA interacts with C3 to
increase the opening rate
k3,2, then we can
approximate the AA-dependent on-rate
(Kd = koff/kon)
at 18 µM1 · s1.
At 10 µM AA, this would predict a
Po of 0.38. Similarly, with the assumption that AA interacts with
C2 to increase
k2,1 with a
predicted on-rate of 36 µM1 · s1,
our model predicts a
Po of 0.42 at 10 µM AA. Neither of these accurately reflects the observed gating of
KAA. Because AA does not affect
C1 or the open state of the
channel, this suggests that AA interacts with both
C3 and
C2. Using our model, we evaluated this possibility. If AA induced a 10-fold increase in both
k3,2 and
k2,1, this would
predict a Po of
0.61, similar to that observed at 10 µM AA. In addition, this model
accurately predicts the increased burst duration observed during AA
stimulation (Fig. 4). In the simulated model, with 10 µM AA
interacting with both C3 and
C2, burst increased to 45 ms.
Finally, our results suggest that the on-rates for these two reactions
must be different. Identical on-rates would predict a
concentration-Po
curve exhibiting negative cooperativity, whereas our data were best fit
using a Hill coefficient of 1.4, i.e., slight positive cooperativity.
Thus we conclude that AA interacts with both long-closed states of the
channel to increase the opening rate and hence
Po. Clapham et
al. (48), studying an AA-activated
K+ channel in cardiac tissue,
predicted a similar kinetic scheme, comprising one open and three
closed states. Similar to our findings, the open time of this channel
was not affected by AA (48).
Although we are able to identify which kinetic state of the channel is
being modulated by AA, our results do not allow us to discern whether
AA is acting at the channel protein itself or interacting with a
closely associated protein or with the membrane bilayer. Fatty acids
are known to interact directly with purified protein kinase C (43) as
well as with fatty acid binding proteins. Indeed, a putative fatty acid
binding domain has been identified in the
N-methyl-D-aspartate
receptor (40). Thus a similar direct interaction with
KAA is possible. We demonstrate
also that stretch activates KAA.
Although a kinetic analysis has not been undertaken on the
stretch-dependent activation, it appears qualitatively to be similar to
that induced by AA, i.e., the channel clearly displays bursting
behavior that is consistent with an open state punctuated by fast
closed events (Fig. 8). This, coupled with an increased
Po, suggests that
stretch decreases the long-closed time(s) of the channel. The
KAA described in cardiac tissue is also activated by stretch; in both cases a similar kinetic scheme describes this activation (18). A similar kinetic has previously been
described for stretch-activated channels of chick skeletal muscle (13)
and amphibian proximal tubule (42) as well. In both cases, it was the
interburst interval that was shortened to increase
Po with no effect
on o or the fast closed time(s) (13, 42). This suggests a similar mechanism of action for these two
modulators or that a common modulator, AA, activates the channel in
both instances.
Effects of K+ channel blockers on KAA. We evaluated the effects of several known K+ channel blockers on KAA activity, including CTX, 4-AP, 293B, TEA, glibenclamide, quinine, and Ba2+. Of these, only TEA and Ba2+ blocked the channel from the extracellular side. However, neither of these blocked with high affinity. TEA blocked in a voltage-dependent fashion, similar to its effect on KCa (9). Our results suggest KAA expresses two binding sites for Ba2+. From the cytoplasmic side of the channel, Ba2+ induced a long-lived, voltage-dependent closed state (Fig. 6) as previously described (47). In contrast, application of Ba2+ to the extracellular side of the channel resulted in an apparent reduction in single-channel amplitude that was also voltage dependent (Fig. 7). Recently, Gray and co-workers (45) demonstrated a similar blocker pharmacology for Ba2+ in epithelial cells from human vas deferens, i.e., cytoplasmic Ba2+ induced long closed events, whereas extracellular Ba2+ caused a voltage-dependent flickery block. Similarly, Hurst et al. (16) demonstrated that extracellular Ba2+ induced both long- and short-lived blocked states in the Shaker K+ channel. In both cases, these results were interpreted as indicating the channel expressed two distinct Ba2+ binding sites.
Effect of stretch-activated ion channel blockers on KAA. Our results demonstrate that Gd3+, a known inhibitor of nonselective, stretch-activated cation channels (49), had no effect on KAA, suggesting it is unrelated to these channels. Amiloride has also been shown to inhibit both stretch-activated K+ channels (44) and nonselective cation channels (25). We demonstrate that amiloride (2 mM) dramatically reduced the i of KAA (Fig. 9). This reduction in i by amiloride is similar to what has previously been reported for amiloride block of both the Xenopus oocyte nonselective cation channel and Lymnaea neuron K+ channel (25, 44). Although our results on the voltage dependence of this block are preliminary, they suggest that the amiloride-induced reduction in i is voltage dependent, i.e., the block is partially relieved at positive voltages (Fig. 9). The amiloride block of the mechanosensitive channel in Xenopus oocytes was highly voltage dependent, being completely relieved by positive holding potentials (25), whereas the amiloride block of the K+ channel of Lymnaea neurons was voltage independent (44). Thus our results fall between these two extremes. Our results with amiloride further support the notion that KAA and Kstretch are indeed the same protein.
The elucidation of the role of KAA in either transepithelial Cl secretion or volume
regulation (see below) will require the identification of high-affinity
blockers that distinguish between
KAA,
KCa, and the cAMP-activated
K+ channel,
KcAMP. Lane et al.
(25) have demonstrated a structure-activity relationship
for the block of the Xenopus oocyte
mechanosensitive channel by amiloride analogs. It will be important to
determine whether high-affinity analogs can be identified for
KAA in T84 cells, thus allowing
its physiological role to be elucidated.
Does AA modulate intestinal Cl
secretion?
Cl secretion requires the
coordinate regulation of both an apical
Cl conductance and
basolateral GK.
Minimally, if AA is to modulate Cl secretion, it must
increase basolateral
GK, thus
hyperpolarizing both membranes and increasing the driving force for
Cl exit across the apical
membrane through constitutively open
Cl channels, a paradigm
proposed previously for the mechanism of action of
Ca2+-dependent agonists (4). We
have now characterized a K+
channel that is potently activated by AA,
KAA. Recently, Barrett (5)
demonstrated that the response of T84 cells to VIP and forskolin was
attenuated by a DAG lipase inhibitor and that VIP increased AA. These
results suggest that AA may be an important modulator of
Cl secretion in intestinal
tissue. Although our results demonstrate that
KAA has a distinct pharmacological
profile from the cAMP-activated GK (it is not
blocked by 293B, an inhibitor of
KcAMP), a selective blocker for
KAA will be required to further
explore its role in this response.
secretory response (6,
12, 32), suggesting a novel signaling pathway. Subsequently, Barrett
and Bigby (6) demonstrated a tight correlation between changes in
cellular AA generation and the resultant
Cl secretory response to
apical adenosine, suggesting that AA may serve this role.
Ulcerative colitis is also characterized by elevated active kallikrein,
which will release kinin (50). The kinins are known to activate
phospholipase A2 and therefore
increase AA levels (1). Also, the proinflammatory cytokine tumor
necrosis factor-
, which is thought to play a role in the
pathogenesis of inflammatory bowel disease, was recently shown to
potentiate the phospholipase A2-stimulated release of AA (14,
27). Clearly then, the inflamed intestine is characterized by elevated
levels of secretory agonists that likely increase cellular AA levels.
In fact, the inflamed intestine is primed for an exaggerated response;
the AA composition of phospholipids is increased in both Crohn's
disease and ulcerative colitis (37, 39). The oxidative conversion of AA
to the eicosanoids is known to be a key step in the etiology of
inflammatory bowel disease. Because the initial step in this cascade is
the liberation of AA from phospholipids, and AA is a well-known
modulator of ion channels in a variety of tissues (33, 38), including
epithelia (2, 17, 34), it is likely that AA itself may play an
important role in the diarrhea associated with inflammatory bowel
disease. We speculate that KAA is
the GK
responsible for carrying K+
current across the basolateral membrane during an AA-mediated Cl secretory response.
Does KAA play a role in volume regulation? In addition to being activated by AA, we demonstrate that KAA is similarly activated by changes in membrane tension. Thus KAA may be activated by the mechanical stimuli associated with haustral contractions required to move the colonic contents toward the rectum or with the regulatory volume decrease associated with cell swelling. The regulatory volume decrease response of jejunum (30) is associated with activation of a Ba2+-sensitive GK. In our studies, Ba2+ inhibited the channel while other known K+ channel blockers did not. Both AA and stretch have previously been shown to activate K+ channels in smooth muscle (21), neurons (20), and cardiac tissue (18, 19). One hypothesis that has been proposed to link these two separate modulators is that the physical deformation of the patch during stretch activates a phospholipase causing the release of free AA and activation of the channel (18).
Differential modulation of two
K+ channels in
the T84 cell line.
In our companion paper we characterized the effect of AA on the
basolateral membrane K+ channel
activated by Ca2+-dependent
agonists, KCa, in intestinal
epithelia (10). Although KCa was
potently inhibited by AA (inhibition constant = 425 nM), KAA is potently activated
(Ks = 1.39 µM).
Although KCa was inhibited by
fatty acids in general, the activation of
KAA was very selective for
cis-unsaturated fatty acids, with AA
being more effective than linoleic acid at the same concentration. The
differential modulation of two distinct
K+ channels in the same cells has
previously been described in cardiac myocytes (19) and stomatal guard
cells (26). Thus the role that AA will play in the
Cl secretory response of
the cell will depend on the preexisting regulatory context in which it
acts.
secretion
by agonists employing AA as a second messenger (e.g., VIP, adenosine)
during both physiological and pathophysiological responses. As such,
this channel may represent a novel pharmacological target in the
treatment of diarrheal disease. The development of selective blockers
of this channel (perhaps amiloride analogs) will be required to confirm
its role in the Cl
secretory response, particularly in inflammatory states.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Dr. Bruce Schultz for help in data simulation and in interpreting the kinetic data. We gratefully acknowledge the excellent technical assistance of Cheng Zhang Shi in tissue-culture work.
| |
FOOTNOTES |
|---|
This work was supported by Cystic Fibrosis Foundation Grant DEVOR 96PO (to D. C. Devor) and National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-31091 (to R. A. Frizzell).
1 If the closed-time histograms are fit starting at 0.625 ms (the limit of resolution using a filter cut-off frequency of 2 kHz), an additional exponential was required to fit the data that averaged 0.21 ± 0.02 ms (n = 19). Because this is below the limit of resolution filtering at 2 kHz, the first three bins were excluded from the fit such that we began binning at 1 ms. Because this is five times the rapid time constant, it will contribute little to the remaining fit.
Address for reprint requests: D. C. Devor, Dept. of Cell Biology and Physiology, S312 BST, 3500 Terrace St., University of Pittsburgh, School of Medicine, Pittsburgh, PA 15261 (E-mail: dd2+{at}pitt.edu).
Received 7 February 1997; accepted in final form 20 September 1997.
| |
REFERENCES |
|---|
|
Top
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
1.
Abdel-Latif, A. A.
Calcium-mobilizing receptors, polyphosphoinositides, and the generation of second messengers.
Pharmacol. Rev.
38:
227-272,
1986[Medline].
2.
Anderson, M. P.,
and
M. J. Welsh.
Fatty acids inhibit apical membrane chloride channels in airway epithelia.
Proc. Natl. Acad. Sci. USA
87:
7334-7338,
1990
3.
Ashkenazi, S.,
T. G. Cleary,
and
L. K. Pickering.
Bacterial toxins associated with diarrheal disease.
In: Textbook of Secretory Diarrhea, edited by E. Liebentahl,
and M. E. Duffey. New York: Raven, 1990, p. 255-272.
4.
Barrett, K. E.
Positive and negative regulation of chloride secretion in T84 cells.
Am. J. Physiol.
265 (Cell Physiol. 34):
C859-C868,
1993
5.
Barrett, K. E.
Effect of the diglyceride lipase inhibitor, RG80267, on epithelial chloride secretion induced by various agents.
Cell. Signal.
7:
225-233,
1995[Medline].
6.
Barrett, K. E.,
and
T. D. Bigby.
Involvement of arachidonic acid in the chloride secretory response of intestinal epithelial cells.
Am. J. Physiol.
264 (Cell Physiol. 33):
C446-C452,
1993
7.
Brzuszczak, I.,
J. Zhao,
C. Bell,
D. Stiel,
I. Fielding,
J. Percy,
R. Smith,
and
E. O'Loughlin.
Cyclic AMP-dependent anion secretion in human small and large intestine.
J. Gastroenterol. Hepatol.
11:
804-810,
1996[Medline].
8.
Burckhardt, B.-C.,
and
H. Gogelein.
Small and maxi K+ channels in the basolateral membrane of isolated crypts from rat distal colon: single channel and slow whole-cell recordings.
Pflügers Arch.
420:
54-60,
1992[Medline].
9.
Devor, D. C.,
and
R. A. Frizzell.
Calcium-mediated agonists activate an inwardly rectified K+ channel in colonic secretory cells.
Am. J. Physiol.
265 (Cell Physiol. 34):
C1271-C1280,
1993
10.
Devor, D. C.,
and
R. A. Frizzell.
Modulation of K+ channels by arachidonic acid in T84 cells. I. Inhibition of the Ca2+-dependent K+ channel.
Am. J. Physiol.
274 (Cell Physiol. 43):
C138-C148,
1998
11.
Devor, D. C.,
A. K. Singh,
R. A. Frizzell,
and
R. J. Bridges.
Modulation of Cl secretion by benzimidazolones. I. Direct activation of a Ca2+-dependent K+ channel.
Am. J. Physiol.
271 (Lung Cell. Mol. Physiol. 15):
L775-L784,
1996
12.
Dho, S.,
K. Stewart,
and
J. K. Foskett.
Purinergic receptor activation of Cl secretion in T84 cells.
Am. J. Physiol.
262 (Cell Physiol. 31):
C67-C74,
1992
13.
Guhary, F.,
and
F. Sachs.
Stretch-activated single ion channel currents in tissue-cultured embryonic chick skeletal muscle.
J. Physiol. (Lond.)
352:
685-701,
1984
14.
Gustafson-Svard, C.,
I. Lilja,
R. Sjodahl,
and
C. Tagesson.
Cytosolic phospholipase A2 and cyclooxygenase-2 mediate release and metabolism of arachidonic acid in tumor necrosis factor--primed cultured intestinal epithelial cells (INT 407).
Scand. J. Gastroenterol.
30:
1000-1007,
1995[Medline].
15.
Hallgren, R.,
J. F. Colombel,
R. Dahl,
K. Fredens,
A. Kruse,
N. O. Jacobsen,
P. Venge,
and
J. C. Rambaud.
Neutrophil and eosinophil involvement of the small bowel in patients with celiac disease and Crohn's disease: studies on the secretion rate and immunohistochemical localization of granulocyte granule constituents.
J. Med.
86:
56-64,
1989.
16.
Hurst, R. S.,
R. Latorre,
and
E. Stefani.
External barium block of shaker potassium channels: evidence for two binding sites.
J. Gen. Physiol.
106:
1069-1087,
1995
17.
Hwang, T.-C.,
S. E. Guggino,
and
W. B. Guggino.
Direct modulation of secretory chloride channels by arachidonic and other cis unsaturated fatty acids.
Proc. Natl. Acad. Sci. USA
87:
5706-5709,
1990
18.
Kim, D.
A mechanosensitive K+ channel in heart cells: activation by arachidonic acid.
J. Gen. Physiol.
100:
1021-1040,
1992
19.
Kim, D.,
and
R. A. Duff.
Regulation of K+ channels in cardiac myocytes by free fatty acids.
Circ. Res.
67:
1040-1046,
1990
20.
Kim, D.,
C. D. Sladek,
C. Aguado-Velasco,
and
J. R. Mathiasen.
Arachidonic acid activation of a new family of K+ channels in cultured rat neuronal cells.
J. Physiol. (Lond.)
484:
643-660,
1995
21.
Kirber, M. T.,
R. W. Ordway,
L. H. Clapp,
J. V. Walsh, Jr.,
and
J. J. Singer.
Both membrane stretch and fatty acids directly activate large conductance Ca2+-activated K+ channels in vascular smooth muscle cells.
FEBS Lett.
297:
24-28,
1992[Medline].
22.
Klaerke, D. A.,
H. Wiener,
T. Zeuthen,
and
P. L. Jorgensen.
Ca2+ activation and pH dependence of a maxi K+ channel from rabbit distal colon epithelium.
J. Membr. Biol.
136:
9-21,
1993[Medline].
23.
Kubo, M.,
and
Y. Okada.
Volume-regulatory Cl channel currents in cultured human epithelial cells.
J. Physiol. (Lond.)
456:
351-371,
1992
24.
Kuhn, M.,
K. Adermann,
J. Jahne,
W. G. Forssmann,
and
G. Rechkemmer.
Segmental differences in the effects of guanylin and Escherichia coli heat-stable enterotoxin on Cl secretion in human gut.
J. Physiol. (Lond.)
479:
433-440,
1994
25.
Lane, J. W.,
D. W. McBride, Jr.,
and
O. P. Hamill.
Structure-activity relations of amiloride and its analogues in blocking the mechanosensitive channel in Xenopus oocytes.
Br. J. Pharmacol.
106:
283-286,
1992[Medline].
26.
Lee, Y.,
H. J. Lee,
R. C. Crain,
A. Lee,
and
S. J. Korn.
Polyunsaturated fatty acids modulate stomatal aperture and two distinct K+ channel currents in guard cells.
Cell. Signal.
6:
181-186,
1994[Medline].
27.
Lilja, I.,
K. Smedh,
G. Olaison,
R. Sjodahl,
C. Tagesson,
and
C. Gustafson-Svard.
Phospholipase A2 gene expression and activity in histologically normal ileal mucosa and in Crohn's ileitis.
Gut
37:
380-385,
1995
28.
Lohrmann, E.,
I. Burhoff,
R. B. Nitschke,
H. J. Lang,
D. Mania,
H. C. Englert,
M. Hropot,
R. Warth,
W. Rohm,
M. Bleich,
and
R. Greger.
A new class of inhibitors of cAMP-mediated Cl secretion in rabbit colon, acting by the reduction of cAMP-activated K+ conductance.
Pflügers Arch.
429:
517-530,
1995[Medline].
29.
Loo, D. D. F.,
and
J. D. Kaunitz.
Ca2+ and cAMP activate K+ channels in the basolateral membrane of crypt cells isolated from rabbit distal colon.
J. Membr. Biol.
110:
19-28,
1989[Medline].
30.
Macleod, R. J.,
and
J. R. Hamilton.
Separate K+ and Cl transport pathways are activated for regulatory volume decrease in jejunal villus cells.
Am. J. Physiol.
260 (Gastrointest. Liver Physiol. 23):
G405-G415,
1991
31.
Madara, J. L.,
C. Parkos,
S. Colgan,
R. J. MacLeod,
S. Nash,
J. Matthews,
C. Delp,
and
W. Lencer.
Cl secretion in a model intestinal epithelium induced by a neutrophil-derived secretagogue.
J. Clin. Invest.
87:
1474-1477,
1991.
32.
Madara, J. L.,
T. W. Patapoff,
B. Gillece-Castro,
S. P. Colgan,
C. A. Parkos,
C. Delp,
and
R. J. Mrsny.
5'-Adenosine monophosphate is the neutrophil-derived paracrine factor that elicits chloride secretion from T84 intestinal epithelial cell monolayers.
J. Clin. Invest.
91:
2320-2325,
1993.
33.
Meeves, H.
Modulation of ion channels by arachidonic acid.
Prog. Neurobiol.
43:
175-186,
1994[Medline].
34.
Mochizuki, H.,
A. C. Chao,
and
J. H. Widdicombe.
Arachidonic acid inhibits Cl secretion in cultures of dog tracheal epithelium.
Life Sci.
50:
1129-1135,
1992[Medline].
35.
Morris, A. P.,
D. V. Gallacher,
and
J. A. C. Lee.
A large conductance, voltage- and calcium-activated K+ channel in the basolateral membrane of rat enterocytes.
FEBS Lett.
206:
87-92,
1986[Medline].
36.
Needleman, P.,
J. Turk,
B. A. Jakschik,
A. R. Morrison,
and
J. B. Lefkowith.
Arachidonic acid metabolism.
Annu. Rev. Biochem.
55:
69-102,
1986[Medline].
37.
Nishida, T.,
H. Miwa,
A. Shigematsu,
M. Yamamoto,
M. Iida,
and
M. Fujishima.
Increased arachidonic acid composition of phospholipids in colonic mucosa from patients with active ulcerative colitis.
Gut
28:
1002-1007,
1987
38.
Ordway, R. W.,
J. J. Singer,
and
J. V. Walsh, Jr.
Direct regulation of ion channels by fatty acids.
Trends Neurosci.
14:
96-100,
1991[Medline].
39.
Pacheco, S.,
K. Hillier,
and
C. Smith.
Increased arachidonic acid levels in phospholipids of human colonic mucosa in inflammatory bowel disease.
Clin. Sci.
73:
361-364,
1987[Medline].
40.
Petrou, S.,
R. W. Ordway,
J. J. Singer,
and
J. V. Walsh, Jr.
A putative fatty acid-binding domain of the NMDA receptor.
Trends Biochem. Sci.
18:
41-42,
1993[Medline].
41.
Resnick, M. B.,
S. P. Colgan,
T. W. Patapoff,
R. J. Mrsny,
C. S. Awtrey,
C. Delp-Archer,
P. F. Weller,
and
J. L. Madara.
Activated eosinophils evoke chloride secretion in model intestinal epithelia primarily via regulated release of 5'-AMP.
J. Immunol.
151:
5716-5723,
1993[Abstract].
42.
Sackin, H.
A stretch-activated K+ channel sensitive to cell volume.
Proc. Natl. Acad. Sci. USA
86:
1731-1735,
1989
43.
Shinomura, T.,
Y. Asaoka,
M. Oka,
K. Yoshida,
and
Y. Nishizuka.
Synergistic action of diacylglycerol and unsaturated fatty acids for protein kinase C activation: its possible implications.
Proc. Natl. Acad. Sci. USA
88:
5149-5153,
1991
44.
Small, D. L.,
and
C. E. Morris.
Pharmacology of stretch-activated K channels in Lymnaea neurones.
Br. J. Pharmacol.
114:
180-186,
1995[Medline].
45.
Sohma, Y.,
A. Harris,
C. J. C. Wardle,
B. E. Argent,
and
M. A. Gray.
Two barium binding sites on a maxi K+ channel from human vas deferens epithelial cells.
Biophys. J.
70:
1316-1325,
1996[Medline].
46.
Turnheim, K.,
J. Costantin,
S. Chan,
and
S. G. Schultz.
Reconstitution of a calcium-activated potassium channel in basolateral membranes of rabbit colonocytes into planar lipid bilayers.
J. Membr. Biol.
112:
247-254,
1989[Medline].
47.
Vergara, C.,
and
R. Latorre.
Kinetics of Ca2+-activated K+ channels from rabbit muscle incorporated into planar bilayers. Evidence for a Ca2+ and Ba2+ blockade.
J. Gen. Physiol.
82:
543-568,
1983
48.
Wallert, M. A.,
M. J. Ackerman,
D. Kim,
and
D. E. Clapham.
Two novel cardiac atrial K+ channels, IK,AA and IK,PC.
J. Gen. Physiol.
98:
921-939,
1991
49.
Yang, X. C.,
and
F. Sachs.
Block of stretch-activated ion channels in Xenopus oocytes by gadolinium and calcium ions.
Science
243:
1068-1071,
1989
50.
Zeitlin, I. J.,
and
A. N. Smith.
Mobilization of tissue kallikrein in inflammatory disease of the colon.
Gut
14:
133-138,
1973
This article has been cited by other articles:
|
|
 |
|
  R. T. Worrell, J. Oghene, and J. B. Matthews Ammonium effects on colonic Cl- secretion: anomalous mole fraction behavior Am J Physiol Gastrointest Liver Physiol, January 1, 2004; 286(1): G14 - G22. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. L. Clarke, S. Petrou, J. V. Walsh Jr., and J. J. Singer Modulation of BKCa channel activity by fatty acids: structural requirements and mechanism of action Am J Physiol Cell Physiol, November 1, 2002; 283(5): C1441 - C1453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R. Cobb, F. Ruiz, C. M. King, J. Fortenberry, H. Greer, T. Kovacs, E. J. Sorscher, and J. P. Clancy A2 adenosine receptors regulate CFTR through PKA and PLA2 Am J Physiol Lung Cell Mol Physiol, January 1, 2002; 282(1): L12 - L25. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. McKay and J. F. Worley III Linoleic acid both enhances activation and blocks Kv1.5 and Kv2.1 channels by two separate mechanisms Am J Physiol Cell Physiol, October 1, 2001; 281(4): C1277 - C1284. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Devor, R. J. Bridges, and J. M. Pilewski Pharmacological modulation of ion transport across wild-type and Delta F508 CFTR-expressing human bronchial epithelia Am J Physiol Cell Physiol, August 1, 2000; 279(2): C461 - C479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bang, Y. Kim, and D. Kim TREK-2, a New Member of the Mechanosensitive Tandem-pore K+ Channel Family J. Biol. Chem., June 2, 2000; 275(23): 17412 - 17419. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) or after
equimolar substitution of 100 meq bath
K+ with
Na+ (
). Anion was gluconate
throughout.
