Am J Physiol Cell Physiol Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Cell Physiol 283: C1441-C1453, 2002. First published July 3, 2002; doi:10.1152/ajpcell.00035.2002
0363-6143/02 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
283/5/C1441    most recent
00035.2002v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (18)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Clarke, A. L.
Right arrow Articles by Singer, J. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Clarke, A. L.
Right arrow Articles by Singer, J. J.
Vol. 283, Issue 5, C1441-C1453, November 2002

Modulation of BKCa channel activity by fatty acids: structural requirements and mechanism of action

Alison L. Clarke1,2, Steven Petrou1,2, John V. Walsh Jr.1, and Joshua J. Singer1

1 Department of Physiology, University of Massachusetts Medical School, Worcester, Massachusetts 01655; and 2 Department of Physiology, University of Melbourne, Victoria 3010, Australia


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

To determine the mechanism of fatty acid modulation of rabbit pulmonary artery large-conductance Ca2+-activated K+ (BKCa) channel activity, we studied effects of fatty acids and other lipids on channel activity in excised patches with patch-clamp techniques. The structural features of the fatty acid required to increase BKCa channel activity (or average number of open channels, NPo) were identified to be the negatively charged head group and a sufficiently long (C > 8) carbon chain. Positively charged lipids like sphingosine, which have a sufficiently long alkyl chain (C >=  8), produced a decrease in NPo. Neutral and short-chain lipids did not alter NPo. Screening of membrane surface charge with high-ionic-strength bathing solutions (330 mM K+ or 130 mM K+, 300 mM Na+) did not alter the modulation of the BKCa channel NPo by fatty acids and other charged lipids, indicating that channel modulation is unlikely to be due to an alteration of the membrane electric field or the attraction of local counterions to the channel. Fatty acids and other negatively charged lipids were able to modulate BKCa channel activity in bathing solutions containing 0 mM Ca2+, 20 mM EGTA, suggesting that calcium is not required for this modulation. Together, these results indicate that modulation of BKCa channels by fatty acids and other charged lipids most likely occurs by their direct interaction with the channel protein itself or with some other channel-associated component.

arachidonic acid; sphingosine; calcium-activated potassium channel


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

IT IS WELL ESTABLISHED that fatty acids are able to modulate the activity of a wide variety of ion channels including K+, Na+, and Ca2+ channels as well as channels activated by N-methyl-D-aspartate (NMDA) and GABA (19, 36, 40, 47, 48, 51, 52, 55, 58, 63, 66, 70). Previously, Kirber et al. (32) characterized large-conductance Ca2+-activated K+ (BKCa) channels from rabbit pulmonary artery (RPA) smooth muscle cells and demonstrated that they are activated by two fatty acids, the polyunsaturated 20-carbon arachidonic acid and the fully saturated 14-carbon myristic acid. Because myristic acid is not metabolized via the lipoxygenase, cycloxygenase, or cytochrome P-450 oxygenase pathways to produce bioactive metabolites (22, 50, 55, 61, 68), modulation of BKCa channel activity was determined to be a direct consequence of the fatty acid molecule itself.

Although the study by Kirber et al. (32) suggested that two fatty acids, which vary in chain length and conformation, are capable of activating the BKCa channel, a comprehensive understanding of the structural features required for this modulation as well as the mechanism of fatty acid action were not determined. For example, do all fatty acids activate this BKCa channel or are only certain fatty acids effective? In addition, is modulation of this channel limited to fatty acids or are other lipids effective? Answers to these questions would also provide information regarding the mechanism of action of fatty acids on the RPA smooth muscle BKCa channel. To date, the fatty acid modulation of a number of BKCa channels has been studied, and these channels show variation in effective lipids as well as the mechanisms of modulation (1, 5, 7, 11, 17, 20, 38, 71, 76, 78).

One possible mechanism of action is direct interaction of fatty acids with the channel protein or some other protein closely associated with the channel (1, 29, 57). Alternatively, fatty acids may be acting through a mechanism that involves alterations of the bulk lipid properties of the membrane, for instance, by acting as detergents to perturb the lipid membrane (47, 73), by altering membrane fluidity, bilayer stiffness and/or membrane curvature (2, 31, 39, 44, 65), or by changing the "protein-lipid interface" (6). In addition, it is possible that fatty acids affect channel behavior by altering membrane surface charge. Changes in membrane surface charge may cause alterations in the local concentration of counterions in the vicinity of the channel and may change the electric field in the membrane, even when the membrane potential is unchanged (23, 66, 72). Because both membrane potential and ions, particularly Ca2+ (32), affect channel behavior, it is possible that the changes in membrane surface charge brought about by fatty acids and other charged lipids are responsible for the changes observed in BKCa channel activity. Fatty acid modulation has also been shown to occur via channel blockade (28, 60), through other modulatory proteins like protein kinase C (PKC) (3, 8, 62) and protein phosphatases (56), and indirectly by bioactive fatty acid metabolites (for example, see Refs. 4 and 7).

In an attempt to understand the mechanism of fatty acid modulation of BKCa channel activity, we determined the structural features of the fatty acid molecule required for channel modulation by studying the effects of a variety of fatty acids and other charged and uncharged amphiphiles on this channel. Four structural features of the fatty acid molecule were considered in this study: 1) the carboxylate head group, 2) the negative charge on the carboxylate head group, 3) the length of the acyl chain, and 4) the structural conformation of the acyl chain. In addition, experiments were performed to address whether mechanisms of action that involve alterations in membrane surface charge or changes in the Ca2+ concentration in the vicinity of the channel are responsible for the modulation of channel activity by fatty acids. A brief account of some of this work has been reported elsewhere in abstract form (14).


    MATERIALS AND METHODS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

Recording conditions. New Zealand White rabbits weighing between 3 and 5 lb were anesthetized with a lethal dose of pentobarbital sodium (0.14 mg/g), and the pulmonary artery was then dissected away from the heart and lungs. Freshly isolated smooth muscle cells from the rabbit main pulmonary artery were obtained with the procedures of Clapp and Gurney (13). Single-channel currents were recorded from excised inside-out (I-O) and excised outside-out (O-O) patches with standard patch-clamp techniques (25). Single-channel recordings were usually carried out in symmetric solutions that were composed of (in mM) 130 K+, 1 Mg2+, 5 EGTA, 114.5 Cl-, and 10 HEPES-HCl at pH 7.4 (ionic strength I = 0.132; Ref. 74). Occasionally, recordings were carried out in nonsymmetric solutions where the external solution was composed of either (in mM) 127 Na+, 3 K+, 1 Mg2+, 5 EGTA, 114.5 Cl-, and 10 HEPES-HCl at pH 7.4 or 120 Na+, 20 K+, 1 Mg2+, 5 EGTA, 122.5 Cl-, and 10 HEPES-HCl at pH 7.4. Solutions containing 5 mM EGTA, zero calcium, and zero nucleotides were used to avoid changes in channel activity that could occur because of the involvement of calcium and other second messengers. In experiments in which we wished to shield membrane surface charge, recordings were carried out in symmetric solutions composed of (in mM) 130 K+, 300 Na+, 5 EGTA, 10 HEPES-HCl, 1 Mg2+, and 414.5 Cl- at pH 7.4 (I = 0.43) or 330 K+, 5 EGTA, 10 HEPES-HCl, 1 Mg2+, and 314.5 Cl- at pH 7.4 (I = 0.315). High-EGTA-containing solutions were composed of (in mM) 130 K+, 48 Cl-, 20 EGTA, 10 HEPES-HCl, and 1 Mg2+ at pH 7.4.

Preparation and application of compounds. The solubility of lipids used in this study varied. For example, tetradecanesulfonate (TDS) could be easily dissolved in water, whereas the water solubility of myristic acid was poor and it needed to be first dissolved in dimethyl sulfoxide (DMSO; Fluka). In general, all lipids were treated the same and were first dissolved in DMSO and then diluted (1:1,000 dilution) in bathing solution. Compounds were applied to the extracellular side of O-O patches and the cytosolic side of I-O patches by pressure ejection (Picospritzer II; General Value, Fairfield, NJ) from micropipettes ("puffer" pipette, 1- to 2-µm tip diameter) placed 50-100 µm from the patch electrode (35). We have defined the structural requirements with concentrations of 10-50 µM (typically 50 µM) that were determined previously to be effective at altering channel activity in this (32) and other (17, 47, 57) preparations and that did not greatly alter the resistance or integrity of the patch. Because compounds are diluted with the bathing solution as they exit the puffer pipette, their final concentration at the membrane surface is difficult to estimate but is likely to be less than that in the puffer pipette. The concentration given is, therefore, the maximum possible concentration. Bathing solutions containing DMSO were applied to patches and had no obvious effect on BKCa channel behavior (Table 1).

                              
View this table:
[in this window]
[in a new window]
  		Table 1.   Effects of fatty acids and other charged lipids on NPo and To

Applications were usually brief to avoid large changes in channel activity that could take many minutes to recover, making multiple applications of lipid compounds difficult. Lipids were normally applied to a patch more than once to ensure that the result observed was reproducible, and each patch was usually exposed to more than one lipid. Fatty acids and other charged lipids occasionally caused an unexplained shift in the baseline current. Occasionally a small, transient, and unexplained decrease in channel activity was observed on the initial application of fatty acids and other negatively charged lipids. Because these initial decreases in activity were only occasionally observed, it is unclear whether this is a real but occasional effect of negatively charged lipids or whether the decrease is a result of random fluctuations in channel activity.

To ensure that the changes in channel activity, observed around the time of the application of a lipid, were truly caused by the lipid itself and not by random fluctuations of channel activity, the activity of the channel was initially monitored over time (3-4 min). Although channel activity fluctuated with time, the changes produced by charged lipids were far greater than those seen in their absence. In a normal experiment, lipids were applied to the patch until they were seen to alter channel activity, at which time the application was terminated. The differences in the time course and strength of the responses produced by the lipid compounds were not necessarily indicative of the potency of the compounds, because many factors contribute to produce variations in these parameters. These factors include the puffer pipette tip size, the distance and geometry of the puffer pipette in relation to the patch, and the position of the membrane patch in the patch pipette. Any delay in the onset of the response did not appear to be a consequence of the lipid itself, because the same lipid (e.g., TDS) could increase activity immediately (in seconds) or instead could require up to 1 min to be effective. Neutral compounds and short-chain compounds, which produced no effect on channel activity, were applied for much longer. These compounds were applied to a patch for at least 1-2 min. The application was also repeated, in many cases at greater application strength (pressure) and volume (size of pipette tip), to ensure that they did not alter the activity of the channel.

Fatty acids and alcohols were obtained from Nu Check Prep (Elysian, MN), primary alkyl amines and alkyl sulfonates were obtained from Aldrich (Milwaukee, WI), lysophospholipids were obtained from Avanti Polar Lipids (Birmingham, AL), and sphingosine was obtained from Sigma (St. Louis, MO).

Data analysis and display. Recordings were made with a conventional patch-clamp amplifier (EPC 5; List). To obtain a similar background level of channel activity we recorded at a range of membrane potentials (usually +20 to +60 mV). However, channel activity varied greatly from patch to patch. The potential across the patch as well as other stimulus protocols were controlled by the software package pCLAMP 5 (Axon Instruments) and the laboratory interface TL1 (Axon Instruments). Data were filtered at 3 or 10 kHz and then digitally recorded onto videotape with a Sony PCM digital audio processor with a sampling frequency of 44 kHz. Data for figures were played back through the PCM to be converted back into an analog signal, filtered at 100 Hz, and then sampled at 300 Hz. Data for analysis were filtered at 300 Hz or 1 kHz and sampled at 1 or 3 kHz, respectively.

Channel activity in our case was defined as NPo (the average number of open channels), where N is the number of channels in the patch (unknown) and Po is the probability that an ion channel is in the open state. Qualitative changes in NPo were determined by visual inspection of the current record. Some records were analyzed quantitatively with the analysis packages described below. Patches chosen for analysis 1) showed low noise levels, 2) had very few or no other channel types evident in the patch that would have a major effect on the analysis, 3) could also show activation by a fatty acid control (when used), usually oleic acid, and 4) were representative of each compound. If many patches could fit these criteria, then patches used for analysis were chosen randomly from this group. The duration of the time period used for analysis before and during the application of a lipid was determined by visual inspection of the channel trace. Mean open times (To) were also determined over the same time periods as those used to determine NPo.

Single-channel current amplitudes (i) were estimated by inspection with a custom software package, Erwin, kindly supplied by Michel Vivaudou (CEA Grenoble, Grenoble, France), or by the commercially available software package pCLAMP 5 or 6. NPo was determined by dividing the average current by the unitary current amplitude with Erwin. Alternatively, the software package pCLAMP 6, which idealized the real channel records, was used to determine NPo and To as described previously (67). Data for Tables 1 and 2 were analyzed with pCLAMP 6. NPo and To were usually calculated for a period of time (10-120 s) before the application of the lipid and for 10-120 s during the time that the lipid exerted an effect. Because the activity of the channels in the patch was not always in a steady state (during application), the values of NPo and To represent an average over the time period used.

                              
View this table:
[in this window]
[in a new window]
  		Table 2.   Effects of charged lipids on NPo and To in presence of high-ionic-strength solutions or 20 mM EGTA

If we assume that all of the BKCa channels are identical and behave independently and that Po is low such that the mean closed time (Tc) is much greater than To, which is likely for the data presented here (32), then for lipids activating the channel, the fold increase in To multiplied by the fold decrease in Tc should equal the fold increase in NPo. Because all of the patches contained multiple channels and N was not known in these experiments, we could not obtain a measure of Tc, but we could obtain NPo and To. From the values of NPo and To before and after application of the lipid we could determine the fold change in Tc. When there was no significant change in To we could attribute all of the increase in NPo to a decrease in Tc. When there was a significant increase in To, we could determine whether this could explain the increase in NPo by dividing the fold increase in NPo by the fold increase in To to obtain a measure of the fold decrease in Tc. For lipids that caused a decrease in NPo we could determine the effect on Tc in a similar manner, in this case by dividing the fold decrease in NPo by the fold decrease in To to obtain the fold increase in Tc.

To determine whether fatty acids and other lipids significantly alter BKCa channel NPo and To, these parameters were compared before and during the application of a lipid. Because channel activity varied greatly from patch to patch, comparing mean changes in NPo was not useful as consistent increases or decreases in channel activity could be masked by this natural variation. Therefore, we used a paired t-test in which P < 0.05 was considered significant. The paired t-test compares the NPo of a patch before the application of the lipid with the NPo of the same patch during the application of a lipid, so the variability in channel activity of patches held at different membrane potentials is minimized. Therefore, a lipid will be found to have a significant affect on NPo if it produces the same change (i.e., an increase or decrease) on a patch-to-patch basis. In addition, data (Tables 1 and 2) presented as mean fold changes in NPo also remove patch-to-patch variability and therefore clearly illustrate the dramatic effects that these compounds can have on NPo. The paired t-test and the mean fold change were also used to compare any effects that various agents have on To. To determine whether lipid compounds produced a change in NPo and To in charge-screening solutions (330 mM K+, 5 mM EGTA or 130 mM K+, 300 mM Na+, 5 mM EGTA) and in high EGTA concentrations (130 mM K+, 20 mM EGTA) similar to those seen in normal bathing solutions (130 mM K+, 5 mM EGTA), the fold changes in To and NPo produced by application in screening conditions and in high EGTA concentrations were compared with the fold changes seen in normal bathing solutions by a t-test in which P < 0.05 was considered significant.


    RESULTS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

A range of fatty acids activate BKCa channels. The fatty acids oleic acid and myristic acid, which are not substrates for the arachidonic acid metabolic pathways that yield bioactive compounds, as well as arachidonic acid increased channel activity of RPA BKCa channels in both I-O and O-O membrane patches (Fig. 1, Table 1). Oleic acid increased channel activity in 15 of 17 I-O patches and 1 of 1 O-O patches, myristic acid increased activity in 10 of 12 I-O and 6 of 6 O-O patches, and arachidonic acid increased channel activity in 3 of 3 I-O and 2 of 2 O-O patches. Analysis of representative traces showed that the increases in NPo produced by these fatty acids was significant (Table 1). To, however, was not significantly altered, and i appeared unaffected by these fatty acids.


View larger version (42K):
[in this window]
[in a new window]
  		Fig. 1.   A variety of fatty acids activate large-conductance Ca2+-activated K+ (BKCa) channels from rabbit pulmonary artery. The fully saturated 14-carbon fatty acid myristic acid (20 µM), which is not a substrate of the arachidonic acid metabolic pathways that produce bioactive compounds, increased channel activity when applied to an inside-out (I-O) patch held at +60 mV (A). The cis-polyunsaturated 20-carbon fatty acid arachidonic acid (20 µM), when applied to an I-O patch held at +40 mV, produced an increase in channel activity (B). Caprylic acid (20 µM), the fully saturated 8-carbon fatty acid, did not significantly change BKCa channel activity in an I-O patch held at +30 mV (C). Trace in A was taken from an experiment carried out in asymmetric solutions (130 mM K+/3 mM K+), and traces in B and C were taken from experiments carried out in asymmetric solutions (130 mM K+/20 mM K+). Bar represents the time of application of these lipids to membrane patches. Average no. of open channels (NPo), single-channel current amplitude (i), and mean open time (To) (calculated with pCLAMP 6) are given in brackets above each trace for times before the application of the lipids and during the effects of the lipids. Arrows on left represent periods of time that are not shown but were used for analysis.

Thus a variety of fatty acids with different acyl chain lengths and chain conformations are able to activate the BKCa channel in RPA smooth muscle cells. Because To is not significantly altered by these fatty acids, the increase in NPo appears to result from a decrease in Tc.

The carboxylate group is not required for BKCa channel activation. To determine whether the carboxylate head group of the fatty acid was required for increasing NPo, the fatty acid analog TDS was applied to excised membrane patches. TDS, a 14-carbon, fully saturated lipid that is similar in structure to fatty acids but possesses a negatively charged sulfonate head group, increased channel activity in 16 of 16 I-O patches and 3 of 3 O-O patches. Analysis of representative traces showed that this increase was significant (Fig. 2, Table 1). TDS did not significantly alter To or i (Table 1). The naturally occurring, negatively charged 16-carbon palmitoyl lysophosphatidate (PLPA) also increased channel activity in 4 of 4 I-O and 6 of 7 O-O patches. This increase in NPo was also shown to be significant, but in this case, To was also minimally but significantly increased (Fig. 2, Table 1). However, the PLPA-induced changes in To could not account for the observed changes in NPo (i.e., NPo fold change To fold change). Therefore, it appears that the increase in NPo produced by these negatively charged lipids is primarily the result of a decrease in Tc.


View larger version (39K):
[in this window]
[in a new window]
  		Fig. 2.   Negatively charged lipids with a sufficiently hydrophobic carbon chain mimic the action of fatty acids by increasing BKCa channel NPo. Tetradecanesulfonate (50 µM), applied to an I-O patch held at +60 mV, produced a large increase in channel activity (A), as did the negatively charged palmitoyl lysophosphatidate (PLPA, 50 µM; B). In this case, PLPA was applied to an I-O patch held at +40 mV. Octanesulfonate (20 µM), a negatively charged lipid with an 8-carbon chain, was essentially ineffective when applied to an I-O patch held at +60 mV (C). Analysis and symbols are as in Fig. 1. Traces in A-C were taken from experiments carried out in asymmetric solutions (130 mM K+/3 mM K+).

Thus, although these various lipids bear quite different negatively charged head group structures, they are all capable of increasing the activity of the BKCa channel, suggesting that the carboxylate head group is sufficient but not necessary for channel activation by fatty acids.

Neutral lipids do not appear to alter NPo. To determine whether the negative charge per se was an important structural requirement for activation, uncharged or neutrally charged lipids were applied to excised patches. The 18-carbon monounsaturated neutral compound oleyl alcohol, the 10-carbon alcohol decanol, and the 12-carbon alcohol dodecanol had no effect on channel activity in 6 of 6 I-O and 2 of 2 O-O patches, 1 of 1 I-O patch, and 6 of 6 I-O and 5 of 5 O-O patches, respectively (Fig. 3). In these same patches fatty acids were seen to increase channel activity. Analysis of representative data for oleyl alcohol and dodecanol showed that they had no significant effect on channel NPo (Table 1). In addition, the fully saturated 16-carbon palmitoyl lysophosphatidylcholine (PLPC), a zwitterion that bears no net charge in the pH range used in this study, was also without any apparent effect in 3 of 4 I-O patches. Analysis showed that PLPC did not significantly change NPo and To (Table 1). Application of PLPC and oleyl alcohol did not alter i; however, dodecanol caused a significant decrease in To.


View larger version (40K):
[in this window]
[in a new window]
  		Fig. 3.   Neutral lipids neither increase nor decrease BKCa channel NPo. The neutral fatty alcohols dodecanol (50 µM; A) and oleyl alcohol (50 µM; B), applied to an outside-out (O-O) patch held at +40 mV and an I-O patch held at +30 mV, respectively, did not affect BKCa channel NPo. Analysis and symbols are as in Fig. 1. Traces in A and B were taken from experiments carried out in asymmetric solutions (130 mM K+/3 mM K+ and 130 mM K+/20 mM K+, respectively).

Neutrally charged lipids were unable to significantly alter BKCa channel NPo. Their inability to do so at the concentrations used here suggests that the presence of a charged head group is a necessary requirement for channel activation.

Positively charged lipids decrease NPo. To determine whether positively charged head groups could also increase NPo, compounds that are similar in structure to fatty acids but instead have positively charged head groups were applied to I-O and O-O patches. Two positively charged primary amines, the 14-carbon fully saturated compound tetradecylamine (TDA) and the 18-carbon monounsaturate oleylamine, both decreased channel activity, showing a decrease in activity in 5 of 6 I-O and 4 of 4 I-O patches and 5 of 5 O-O patches, respectively. These decreases in NPo were found to be significant (Table 1), whereas no obvious effect on i was observed (Fig. 4). Unlike TDA, however, oleylamine significantly decreased To (Table 1). In the case of oleylamine, the change in To alone could only partially account for the change in NPo, suggesting that the major reason for the observed change in NPo was an increase in Tc. The naturally occurring 14-carbon amino alcohol sphingosine, which at the pH used in this study should bear a positive charge, also decreased channel activity in 5 of 6 I-O and 2 of 2 O-O patches. This suppression of NPo was significant (Table 1). Sphingosine did not significantly alter To or have an apparent effect on i (Fig. 4, Table 1).


View larger version (34K):
[in this window]
[in a new window]
  		Fig. 4.   Positively charged lipids suppress BKCa channel NPo. When applied to an I-O patch at a holding potential of +50 mV, the positively charged amino alcohol sphingosine (50 µM) suppressed channel activity (A) as did the cis-monounsaturated oleylamine (50 µM), shown here applied to an I-O patch held at +40 mV (B). The short-chain octylamine (50 µM) produced a small but significant decrease in NPo when applied to I-O and O-O patches (C). This very small decrease is not obvious in raw data traces like the one shown in C, most likely because the mechanism is different (see text). In this I-O patch held at +30 mV, the reduction in i produced by the application of octylamine to both I-O and O-O patches can be seen. Analysis and symbols are as in Fig. 1. Traces in A-C were taken from experiments carried out in symmetric solutions (130 mM K+/130 mM K+).

Positively charged lipids are effective in altering BKCa channel activity, but unlike negatively charged lipids, they caused a decrease in NPo. These results suggest that the negatively charged head group is required for the increase in NPo produced by fatty acids. In addition, the positively charged head group is most likely responsible for the decrease in NPo produced by sphingosine and other positively charged lipids.

Short-chain lipids do not appear to alter NPo. Because a variety of negatively and positively charged lipids could effectively alter BKCa channel activity, we also wished to determine what structural features of the acyl and alkyl chains were required. As indicated in Figs. 1 and 4, saturated and unsaturated charged lipids could effectively alter BKCa channel NPo; however, a minimal chain length appears to be required. Octanesulfonate (4 of 4 I-O patches) and caprylic acid (4 of 4 I-O and 1 of 1 O-O patches) did not obviously change channel activity. Although small changes in channel activity were observed after analysis, the eight-carbon fully saturated fatty acid caprylic acid did not significantly alter NPo (Table 1). To and i were unaffected by its application (Fig. 1C, Table 1). Similarly, there was a lack of effect with the eight-carbon fully saturated alkyl sulfonate octanesulfonate (Fig. 2C, Table 1).

Octylamine appeared to produce no obvious effect on channel activity in 6 of 6 I-O and 1 of 1 O-O patches (Fig. 4C); however, on analysis, a small but significant decrease in NPo was observed (Table 1). Octylamine was much less effective than its longer-chain positively charged counterparts at decreasing NPo (Fig. 4C, Table 1). In addition, application of octylamine caused a decrease in i and a large and significant decrease in To in both patch configurations (Fig. 4, Table 1). Thus the decrease in channel NPo produced by octylamine, unlike most other lipids tested, is likely to result primarily from changes in To.

In summary, short-chain negatively charged lipids do not increase NPo at the concentrations used here. The eight-carbon octylamine, although capable of decreasing BKCa channel NPo, was not as effective as the longer-chain, positively charged lipids. Octylamine, in fact, appears to alter channel NPo through a mechanism unlike that of the longer-chain positively charged lipids. Its action, which decreases i, is consistent with it acting as a fast open channel blocker, although it may also have an allosteric effect on channel gating (21). Thus, to effectively change NPo through a mechanism that appears to involve an alteration of Tc, negatively charged and positively charged lipids appear to require a chain length of greater than eight carbons.

Fatty acids do not appear to affect NPo by altering the voltage dependence of channel activation. The activity of the RPA BKCa channel is strongly voltage dependent, showing, at low Po, an e-fold change in NPo for each 9-mV change in membrane potential (32). Therefore, the activation of BKCa channels by negatively charged lipids might be explained if we assume that fatty acids preferentially insert into the outer leaflet rather than into the inner leaflet of the cell membrane, thereby altering membrane surface charge. The addition of negatively charged lipids to the outer membrane leaflet would alter the electric field within the membrane in the direction expected for membrane depolarization, resulting in an increase in NPo. Similarly, preferential insertion of positively charged lipids into the outer leaflet would steepen the electric field within the membrane, producing an effective membrane hyperpolarization and a decrease in NPo (24).

To determine whether alterations in surface charge could explain the above results, fatty acids and other charged lipids were applied to I-O and O-O patches in the presence of high-ionic-strength bathing and pipette solutions (330 mM K+ or 130 mM K+, 300 mM Na+). High-ionic-strength solutions were previously used to shield membrane surface charge (41, 42, 49). It should also be pointed out, however, that shielding surface charge could also render these charged lipids less effective if they interact with a charged site on the channel protein that is exposed, or partially exposed, to the surrounding solutions.

High-ionic-strength solutions failed to alter the effects of fatty acids and other charged lipids on BKCa channel activity seen under normal ionic strength conditions. TDS (6 of 6 I-O and 3 of 3 O-O patches) and oleic acid (5 of 6 I-O and 8 of 8 O-O patches) were still able to increase channel activity in 330 mM K+-containing solutions. A similar result was found for solutions containing 130 mM K+, 300 mM Na+, where TDS increased NPo in 4 of 4 I-O and 2 of 2 O-O patches and myristic acid increased NPo in 3 of 3 I-O and 1 of 1 O-O patch. Data analysis showed that these negatively charged lipids significantly increased NPo in high-ionic-strength solutions (Table 2, Fig. 5). Similarly, positively charged lipids were also effective in high-ionic-strength solutions. TDA (2 of 2 I-O and 2 of 2 O-O patches) and sphingosine (4 of 4 I-O and 2 of 2 O-O patches) decreased NPo when they were applied to membrane patches in 330 mM K+-containing solutions. Data analysis showed that these positively charged lipids significantly decreased NPo in high-ionic-strength solutions (Table 2).


View larger version (35K):
[in this window]
[in a new window]
  		Fig. 5.   Charged lipids still effectively alter BKCa channel NPo in high-ionic strength bathing solutions. Tetradecanesulfonate (50 µM) and myristic acid (20 µM) produced increases in NPo in symmetric solutions containing 130 mM K+, 300 mM Na+ in an I-O patch held at +10 mV (A) and an I-O patch held at +20 mV (B), respectively. Note that in A, the application of tetradecanesulfonate causes a shift in the baseline. This was also occasionally observed with fatty acids and other charged lipids. Note also that i is reduced in these recording solutions. Analysis and symbols are as in Fig. 1.

The mean fold changes in NPo produced by both negatively and positively charged lipids in high-ionic-strength solutions were not significantly different from those seen in normal-ionic-strength solutions [mean fold changes for each lipid shown in Table 1 (130 mM K+) compared with data for each lipid shown in Table 2 (330 mM K+ or 130 mM K+, 300 mM Na+)]. To was significantly altered by the application of some of these compounds in high-ionic-strength solutions. The lipids that produced a significant change in To in high-ionic-strength solutions were not the same lipids that significantly altered To in normal ionic conditions and vice versa. In addition, TDS significantly increased To in 330 mM K+ but did not do so in 130 mM K+, 300 mM Na+-containing solutions.

Thus the modulation of BKCa channel activity by fatty acids and other charged lipids does not appear to involve a change in electric field brought about by alterations of membrane surface charge. This may be due to the fact that charged lipids may only produce small changes in membrane surface charge that do not alter channel gating or that charged lipids do not preferentially insert into a particular bilayer leaflet.

Fatty acids do not appear to affect NPo by altering concentration of calcium and other ions in the vicinity of the channel. Insertion of charged lipids into the inner leaflet of the membrane bilayer could attract or repel counterions in the vicinity of the channel (27). An increase in the internal concentration of the counterions Ca2+ and H+ near the channel could alter the activity of the BKCa channel in opposite ways. If the insertion of fatty acids into the inner membrane leaflet attracted Ca2+ to the channel, channel activity would be increased (54), whereas the attraction of H+ would cause a decrease in channel activity (33). Moreover, insertion of positively charged lipids would repel Ca2+ and H+, with the former causing a decrease in activity and the latter an increase. Experiments performed in high-ionic-strength solutions suggest that it is unlikely that channel activation by fatty acids and other charged lipids results from a change in the concentration of Ca2+ or H+ in the vicinity of the channel due to the alteration of membrane surface charge because responses were essentially unchanged in high-ionic-strength solutions.

However, calcium may still be involved in the charged lipid modulation of BKCa channel activity through a mechanism that does not involve an alteration of membrane surface charge. Because it has been suggested that calcium stores may exist in excised membrane patches (77) and that fatty acids can mobilize calcium from internal stores (12, 75), we further investigated the involvement of Ca2+. In this scheme we would have to assume that the positively and negatively charged lipids are acting to alter channel activity through two different mechanisms.

The ability of TDS to activate the BKCa channel was tested in different concentrations of the calcium chelator EGTA (5 and 20 mM). If fatty acids and other negatively charged lipids were acting to increase NPo through a Ca2+-dependent mechanism, it would be expected that they would cause a much smaller increase in NPo in the presence of 20 mM EGTA than in the presence of 5 mM EGTA. Oleic acid was able to increase NPo in the presence of 20 mM EGTA (2 patches; not shown). TDS was also able to significantly increase NPo and To in solutions containing 20 mM EGTA (Table 2, Fig. 6A). In one I-O patch in which the experiment was carried out, the increase in NPo produced by TDS in a 20 mM EGTA-containing solution was similar to that seen when the solution was changed to one with 5 mM EGTA (Fig. 7A), suggesting that this increase was not due to a calcium-dependent mechanism. In addition, the increases in NPo produced by TDS were similar in different patches, whether these experiments were carried out in 5 or 20 mM EGTA (compare Fig. 6, A and B). In fact, the mean fold change in NPo seen in 20 mM EGTA was not significantly different from that seen in 5 mM EGTA [i.e., 13.45 ± 7.80 (6 patches, Table 2) vs. 12.70 ± 9.90 (5 patches, Table 1)], suggesting that fatty acids and fatty acid analogs do not increase NPo by increasing the Ca2+ concentration ([Ca2+]) in the vicinity of the channel. Moreover, these results show that fatty acids can affect BKCa channel activity independently of the levels of internal [Ca2+]. BAPTA was also used in experiments to chelate calcium, and in these cases the results were similar to those seen with EGTA. TDS could still increase NPo in high concentrations of BAPTA (20 mM, 4 patches; not shown). Interestingly, in one I-O patch in which it was examined, the voltage dependence of NPo was essentially the same in 5 and 20 mM EGTA (Fig. 7B). In addition to the above, the fact that fatty acids and other lipids could repetitively alter channel activity in solutions containing no ATP also argues against a mechanism involving Ca2+ release from stores. In these high-EGTA, zero-calcium solutions, the stores would not be refilled after emptying by fatty acids and a rundown in the response would be predicted.


View larger version (37K):
[in this window]
[in a new window]
  		Fig. 6.   Fatty acids and other charged lipids increase NPo in different concentrations of the calcium chelator EGTA. Tetradecanesulfonate (50 µM; A) produces an increase in channel activity in solutions containing 130 mM K, 20 mM EGTA, and no added calcium. The increase produced by tetradecanesulfonate in 20 mM EGTA in an I-O patch held at +50 mV (A) is similar to the increase seen in another I-O patch held at +60 mV in 5 mM EGTA (B). Trace in A was taken from an experiment carried out in asymmetric solutions (130 mM K+, 20 EGTA/130 mM K+, 5 EGTA). Trace in B was taken from an experiment carried out in symmetric solutions (130 mM K+, 5 EGTA/130 mM K+, 5 EGTA). Analysis and symbols are as in Fig. 1.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 7.   Negatively charged lipids produce a similar increase in NPo in bathing solutions containing 5 or 20 mM EGTA. A comparable increase in channel activity (NPo) was observed when tetradecanesulfonate was applied to the same I-O patch (+40 mV) in the presence of 5 mM EGTA and 20 mM EGTA (A). In the presence of tetradecanesulfonate, the NPo vs. voltage plot in 5 mM EGTA overlies the plot produced in 20 mM EGTA (B).

All of these results, coupled with other evidence suggesting that fatty acids and other charged lipids alter NPo by acting from the extracellular surface (15), argue against a mechanism involving internal calcium. Moreover, these results also suggest that if fatty acids and other charged lipids directly interact with a charged site on the channel protein, it is likely that this site is protected from exposure to the bathing solution. If fatty acids interacted with an exposed charged site on the channel protein, then it would be expected that this site would be shielded in high-ionic-strength solutions, thus rendering fatty acids and charged lipids less effective under these conditions.

Membrane-bound protein kinases and phosphatases. Because the modulation of BKCa channel activity from the RPA by fatty acids and other charged lipids was studied in excised patches, in the absence of calcium and nucleotides, it is unlikely that they altered channel behavior by affecting the activity of protein kinases or protein phosphatases. Furthermore, the effects of fatty acids and other charged lipids could be obtained repeatedly in the same patch, and high concentrations of the kinase inhibitor staurosporin (70 nM) and the phosphatase inhibitor okadaic acid (2 mM) did not prevent fatty acid increases in NPo (4 patches; not shown). In addition, experiments that show that it is likely that charged lipids are acting from the external membrane surface (15) suggest that these lipids are not interacting directly with these enzymes to bring about an alteration in channel behavior.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

To elucidate the mechanism of fatty acid modulation of BKCa channel activity in RPA smooth muscle cells, we have studied the regulation of this channel by a variety of single-chain lipids. We identified certain structural features of the fatty acid molecule that are required for modulation of channel activity. While helping to address the mechanism of fatty acid action on this channel, these structural features should also provide clues as to the nature of the site with which the fatty acids interact. In addition, experiments were performed to address the mechanism by which fatty acids and other charged lipids modulate the activity of BKCa channels.

Structural requirements. The features of fatty acid compounds that were required for BKCa channel activation, at the concentrations used here, were found to be the negatively charged head group and a sufficiently long (C > 8) hydrophobic carbon chain. Thus not only could fatty acids increase NPo, but other negatively charged lipids could also do so. Interestingly, positively charged lipids were also able to affect the activity of the channel, but in this case NPo was decreased. Neutral and short-chain (C8) negatively charged lipids were apparently ineffective at the concentrations used. The short-chain, positively charged lipid octylamine produced a significant decrease in channel NPo; however, this decrease was much smaller than that produced by the longer-chain, positively charged lipids and appeared to be the result of a different mechanism. This suggests that chain length is also an important structural feature governing the effectiveness of positively charged lipids.

Chain conformation did not appear to be important, because saturates, monounsaturates, and polyunsaturates, as long as they were charged, could modulate channel activity. Fatty acids and other charged lipids, except octylamine, did not alter i and usually did not affect To. Therefore, all charged long-chain lipids affect NPo predominantly through an alteration in Tc. Changes in NPo produced by oleylamine are likely to result from changes in both Tc and To, whereas octylamine is likely to decrease NPo predominantly through changes in To. Fatty acids and charged lipids could effectively modulate channel activity when applied to both I-O and O-O patches, as would be expected for compounds that can traverse the lipid bilayer (26). The finding that fatty acid and lipid modulation of BKCa activity is not dependent on chain conformation is in agreement with the results from rabbit coronary artery smooth muscle cells obtained by Ahn et al. (1).

Fatty acids and charged lipids do not appear to alter NPo through a lipid mechanism. The fact that the longer-chain lipids were more effective at altering channel activity than the shorter-chain compounds supports the contention that lipids affecting BKCa channel activity associate with the membrane or, alternatively, with a hydrophobic binding pocket in the channel protein, to exert their effects on BKCa channel activity. If lipids must partition into the membrane to be effective, they could be altering channel activity by affecting the bulk lipid properties of the membrane. Insertion of lipids into the membrane may have a number of affects on membrane properties. First, these inserted lipids may act as detergents and disrupt the membrane; second, they may change membrane fluidity; and, finally, they may alter the organization of the lipid bilayer to affect membrane curvature.

Because there are specific structural features of the fatty acid molecule that appear to be required for modulating channel activity, it is unlikely that these lipids affect channel activity by acting on membrane fluidity or by acting as detergents. The effects of charged lipids on channel behavior are correlated with the length of the carbon chain and the charge of the head group and not with their ability to alter properties of the membrane. For example, negatively charged cis-unsaturates such as arachidonic acid and oleic acid, which are more likely to disrupt membrane order and, therefore, increase membrane fluidity, had the same effect on channel activity as the fully saturated, and hence membrane-ordering, myristic acid (9). PLPA and PLPC are both excellent detergents; however, the former increases BKCa channel activity whereas the latter produces no obvious effect. Moreover, the effects of charged lipids were reversible, making them unlikely to result from the disruption of the cell membrane, i.e., detergent effects (47).

Because this BKCa channel is mechanosensitive (32), intercalation of lipids into the membrane may affect membrane curvature, according to the "bilayer couple theory" (65), and thus produce stretch-induced alterations of channel activity, as suggested by Martinac et al. (44). According to the bilayer couple theory, charged amphipaths accumulate preferentially in one-half of the lipid bilayer, positively charged amphipaths in the inner layer and negatively charged amphipaths in the outer layer, and induce membrane curvature in opposite directions (concave or negative curvature and convex or positive curvature, respectively).

Martinac et al. (44) first used the bilayer couple theory to explain the effects of amphipathic compounds on the stretch-sensitive channels of bacteria. In their study, they found that cationic and anionic amphipaths mimicked the effect of stretch and activated the channels. Thus, to explain their results according to the bilayer couple theory, they had to assume that a convex or a concave curvature of the membrane creates a mechanical stress on the channels equivalent to stretch. Our results suggest that such a mechanism is unlikely, because all lipids (irrespective of the charge on their head group) should produce the same effect, channel activation. This, however, was not the case.

Our results could be consistent with the bilayer couple theory if we assume that the BKCa channels can differentiate between opposite curvatures, convex and concave. Thus, for lipids that can easily flip across the bilayer when applied from one side, preferential insertion of negatively charged lipids into the outer membrane leaflet would cause membrane "stretch" or positive bilayer curvature and thus channel activation, whereas preferential insertion of positively charged lipids into the inner leaflet would cause membrane "compression" or negative bilayer curvature and thus inhibit channel activity. However, this explanation is unlikely because in experiments in which we limited the application of a negatively charged lipid to one side of the membrane [palmitoyl coenzyme A, which does not flip across the bilayer (26)] under surface charge shielding conditions, channel activation was seen only when the compound was applied to the extracellular side. There was essentially no effect when the same lipid was applied to the intracellular surface (15). The bilayer couple hypothesis (where there was specific curvature dependence) would predict that for a lipid that only inserts into one side of the bilayer, insertion into one leaflet would cause channel activation whereas insertion into the other leaflet would cause channel inhibition.

Another mechanism by which lipids could affect channel activity is by shape-dependent membrane deformation (see Refs. 10 and 39 and references therein). Casado and Ascher (10) ascribed the effect of a number of lipid compounds (similar to those used in this study) on the stretch-sensitive NMDA receptor to the shape of the compound. Lysophospholipids, whose head group structures are larger than their tails (cones), were found, like membrane compression, to inhibit the NMDA receptor when applied to the outside membrane surface, whereas application of arachidonic acid, a compound whose head group structure is smaller than its tail (inverted cone), produced the same effect as membrane stretch when applied to the outside membrane surface, channel activation. Our findings are not consistent with such a hypothesis because there was no correlation between the shape of the compound and its effect on channel activity. Negatively charged cones (palmitoyl lysophosphatidic acid) and inverted cones (arachidonic acid) activated the BKCa channel, whereas cones of differing charge [palmitoyl lysophosphatidic acid (negative) and PLPC (neutral)] had different affects on channel activity. These results highlight the importance of the charge on the head group of the lipid compound, and not its shape, and raise the question of how positively charged cones and inverted cones would affect the NMDA channel, which was not addressed by Casado and Ascher (10).

Possible mechanisms of action of fatty acids and charged lipids. Our results show that single-chain charged lipids do not alter BKCa channel activity by altering membrane surface charge or by means of second messenger molecules like calcium or by altering the activity of protein kinases and phosphatases. Instead, our results are consistent with single-chain lipids having an effect on the channel protein complex itself or a closely associated membrane component. If the only role of the carbon chain is to attach the charged head group to the lipid bilayer or to a hydrophobic pocket in the channel, then this charged group must be in some way responsible for the alteration in channel activity. It is likely that the charged head group alters NPo by interacting with residues either on the channel itself or on some closely associated channel protein.

One possible way that this could happen is that the charged head groups interact with one, or some, of the positively charged residues found in the transmembrane spanning voltage sensor S4 (24, 34, 43). Kang and Leaf (30) previously proposed the involvement of a sequence within the voltage sensor in the fatty acid modulation of voltage-gated Na+, K+, and Ca2+ channels. Alternatively, these charged lipids may interact with other residues in the channel protein or in other membrane-bound proteins, and these residues may not be the same for positively charged and negatively charged lipids.

Interestingly, the lipid modulation of the activity of a number of other proteins, PKC (37, 46, 62, 64), the Na+/Ca2+ exchanger (59), and a small-conductance K+ channel (53, 57), show structural requirements similar to those required for the BKCa channel studied here. However, these structural requirements are different from those described for the BK channel of GH3 cells (a channel that is Ca2+ activated via phospholipase A2; Ref. 18), in which it was found that there was a significant correlation between the degree of fatty acid unsaturation and channel activation (and not with the compounds' ability to affect membrane fluidity) (17). Saturated and trans-unsaturated fatty acids were found to be ineffective. It is clear that such a correlation does not apply to the RPA smooth muscle BKCa channel because fatty acids and other lipids without double bonds in the carbon chains, e.g., myristic acid and TDS, were very effective at increasing channel activity.

Although our study identifies lipids that effectively modulate channel activity, limitations brought about by the method of application do not allow us to definitively compare the relative effectiveness of these lipids. Future studies that compare steady-state concentration-response curves for these lipids will enable such a comparison.

In summary, we have identified the structural features required for single-chain lipids to both activate and inhibit BKCa channel activity. BKCa channel activation requires a negatively charged head group with a chain length of greater than eight carbons. These negatively charged lipids increase NPo, primarily by decreasing Tc. Longer-chain positively charged lipids decrease NPo primarily by increasing Tc.

Together with the evidence indicating that these compounds do not appear to act by altering the properties of the lipid bilayer or membrane surface charge or the activity of second messenger molecules, this study suggests that it is likely that fatty acids and other charged lipids modulate BKCa channel activity by interacting with the channel protein itself, with some other channel-associated protein (e.g., beta -subunit) that is unlikely to be a kinase or phosphatase, or with some other membrane component closely associated with the channel. As with our earlier study (32), this study also provides evidence that BKCa channels from RPA smooth muscle cells show Ca2+-independent gating, because channel activity was observed in solutions containing essentially no Ca2+. Other studies have also determined that BKCa channels are capable of Ca2+-independent gating (16, 45, 69); however, in those studies BKCa channels required much more positive membrane potentials to open in the absence of Ca2+ than the BKCa channel studied here. Moreover, fatty acids and other charged lipids do not appear to require the presence of Ca2+ to affect channel behavior.


    ACKNOWLEDGEMENTS

We thank Paul Tilander, Rebecca McKinney, and Brian Packard for excellent technical assistance and Alejandro M. Dopico for helpful discussions.


    FOOTNOTES

This work was supported by National Institutes of Health Grants DK-31620 and HL-61297.

Address for reprint requests and other correspondence: A. L. Clarke, Dept. of Physiology, Univ. of Melbourne, VIC 3010, Australia (E-mail: alisonlc{at}unimelb.edu.au).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

July 3, 2002;10.1152/ajpcell.00035.2002

Received 22 January 2002; accepted in final form 24 June 2002.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES

1.   Ahn, DS, Kim YH, Lee BS, and Kang DH. Fatty acids directly increase the activity of Ca2+-activated K+ channels in rabbit coronary smooth muscle cells. Yonsei Med J 35: 10-24, 1994[Medline].

2.   Anel, A, Richieri GV, and Kleinfeld AM. Membrane partition of fatty acids and inhibition of T cell function. Biochemistry 2: 530-536, 1993.

3.   Asaoka, Y, Yoshida K, Oka M, Shinomura T, Ogita K, Kikkawa U, and Nishizuka Y. The family of protein kinase C in transmembrane signalling for cellular regulation. J Nutr Sci Vitaminol (Tokyo) SpecNo: 7-12, 1992.

4.   Barlow, RS, El-Mowafy AM, and White RE. H2O2 opens BKCa channels via the PLA2-arachidonic acid signaling cascade in coronary artery smooth muscle. Am J Physiol Heart Circ Physiol 279: H475-H483, 2000[Abstract/Free Full Text].

5.   Baron, A, Frieden M, and Beny JL. Epoxyeicosatrienoic acids activate a high-conductance, Ca2+-dependent K+ channel on pig coronary artery endothelial cells. J Physiol 504: 537-543, 1997[ISI][Medline].

6.   Barrantes, FJ. Structural-functional correlates of the nicotinic acetylcholine receptor and its lipid microenvironment. FASEB J 7: 1460-1467, 1993[Abstract].

7.   Benoit, C, Renaudon B, Salvail D, and Rousseau E. EETs relax airway smooth muscle via an EpDHF effect: BKCa channel activation and hyperpolarization. Am J Physiol Lung Cell Mol Physiol 280: L965-L973, 2001[Abstract/Free Full Text].

8.   Blobe, GC, Khan WA, and Hannun YA. Protein kinase C: cellular target of the second messenger arachidonic acid? Prostaglandins Leukot Essent Fatty Acids 52: 129-135, 1995[ISI][Medline].

9.   Carruthers, A, and Melchior DL. Role of bilayer lipids in governing membrane transport processes. In: Lipid Domains and the Relationship to Membrane Function, edited by Aloia RC, Curtain CC, and Gordon LM.. New York: Liss, 1988, p. 201-225.

10.   Casado, M, and Ascher P. Opposite modulation of NMDA receptors by lysophospholipids and arachidonic acid: common features with mechanosensitivity. J Physiol 513: 317-330, 1998[Abstract/Free Full Text].

11.   Chang, HM, Reitstetter R, and Gruener R. Lipid-ion channel interactions: increasing phospholipid headgroup size but not ordering acyl chains alters reconstituted channel behavior. J Membr Biol 145: 9-13, 1995.

12.   Chow, SC, and Jondal M. Polyunsaturated fatty acids stimulate an increase in cytosolic Ca2+ by mobilizing the inositol 1,4,5-trisphosphate-sensitive Ca2+ pool in T cells through a mechanism independent of phosphoinositide turnover. J Biol Chem 265: 902-907, 1990[Abstract/Free Full Text].

13.   Clapp, LH, and Gurney AM. Outward currents in rabbit pulmonary artery cells dissociated with a new technique. Exp Physiol 76: 677-693, 1991[Abstract].

14.   Clarke, AL, Petrou S, Walsh JV, Jr, and Singer JJ. Modulation of large conductance Ca2+-activated K+ channel activity by charged lipids: structural requirements (Abstract). Soc Neurosci Abstr 21: 1326, 1995.

15.  Clarke AL, Walsh JV Jr, and Singer JJ. Site of action of charged lipids on Ca2+-activated K+ (CAK) channels from rabbit pulmonary artery (Abstract). Abstract Book of the Fourth Annual Neuropharmacology Conference---Potassium Channels, 1996, p. 11.

16.   Cui, J, Cox DH, and Aldrich RW. Intrinsic voltage dependence and Ca2+ regulation of mslo large conductance Ca-activated K+ channels. J Gen Physiol 109: 647-673, 1997[Abstract/Free Full Text].

17.   Denson, DD, Wang X, Worrell RT, and Eaton DC. Effects of fatty acids on BK channels in GH3 cells. Am J Physiol Cell Physiol 279: C1211-C1219, 2000[Abstract/Free Full Text].

18.   Denson, DD, Worrell RT, Middleton P, and Eaton DC. Ca2+ sensitivity of BK channels in GH3 cells involves cytosolic phospholipase A2. Am J Physiol Cell Physiol 276: C201-C209, 1999[Abstract/Free Full Text].

19.   Devor, DC, and Frizzell RA. Modulation of K+ channels by arachidonic acid in T84 cells. II. Activation of a Ca2+-independent K+ channel. Am J Physiol Cell Physiol 274: C149-C160, 1998[Abstract/Free Full Text].

20.   Dumoulin, M, Salvvail D, Gaudreault SB, Cadieux A, and Rousseau E. Epoxyeicosatrienoic acids relax airway smooth muscles and directly activate reconstituted KCa channels. Am J Physiol Lung Cell Mol Physiol 275: L423-L431, 1998[Abstract/Free Full Text].

21.   Eghbali, M, Curmi JP, Birnir B, and Gage PW. Hippocampal GABAA channel conductance increased by diazepam. Nature 388: 71-75, 1997[Medline].

22.   Fitzpatrick, FA, and Murphy RC. Cytochrome P450 metabolism of arachidonic acid: formation and biological actions of "epoxygenase"-derived eicosanoids. Pharmacol Rev 40: 229-241, 1994[ISI][Medline].

23.   Fraser, DD, Hoehn K, Weiss S, and MacVicar BA. Arachidonic acid inhibits sodium currents and synaptic transmission in cultured striatal neurones. Neuron 11: 633-644, 1997.

24.   French, RJ, Prusak-Sochaczewski E, Zamponi GW, Becker S, Kularatna AS, and Horn R. Interactions between a pore-blocking peptide and the voltage sensor of the sodium channel: an electrostatic approach to channel gating. Neuron 16: 407-413, 1996[ISI][Medline].

25.   Hamill, OP, Marty A, Neher E, Sakmann B, and Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflügers Arch 391: 85-100, 1981[ISI][Medline].

26.   Hamilton, JA, and Cistola DP. Transfer of oleic acid between albumin and phospholipid vesicles. Proc Natl Acad Sci USA 83: 82-86, 1986[Abstract/Free Full Text].

27.   Hille, B. Ionic Channels of Excitable Membranes. Sunderland, MA: Sinauer, 1992.

28.   Honore, E, Barhanin J, Attali B, Lesage F, and Lazdunski M. External blockade of the major cardiac delayed-rectifier K+ channel (Kv1.5) by polyunsaturated fatty acids. Proc Natl Acad Sci USA 91: 1937-1941, 1994[Abstract/Free Full Text].

29.   Kang, JX, and Leaf A. Evidence that free polyunsaturated fatty acids modify Na+ channels by directly binding to the channel proteins. Proc Natl Acad Sci USA 93: 3542-3546, 1996[Abstract/Free Full Text].

30.   Kang, JX, and Leaf AL. Antiarrhythmic effects of polyunsaturated fatty acids. Circulation 94: 1774-1780, 1996[Free Full Text].

31.   Karnovsky, MJ, Kleinfeld AM, Hoover RL, and Klausner RD. The concept of lipid domains in membranes. J Cell Biol 94: 1-6, 1982[Free Full Text].

32.   Kirber, MT, Ordway RW, Clapp LH, Walsh JV, Jr, and Singer JJ. 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[ISI][Medline].

33.   Kume, H, Takagi K, Satake T, Tokuno H, and Tomita T. Effects of intracellular pH on calcium-activated potassium channels in rabbit tracheal smooth muscle. J Physiol 424: 445-457, 1990[Abstract/Free Full Text].

34.   Larsson, HP, Baker OS, Dhillon DS, and Isacoff EY. Transmembrane movement of the Shaker K+ channel S4. Neuron 16: 387-397, 1996[ISI][Medline].

35.   Lassignal, NL, Singer JJ, and Walsh JV, Jr. Multiple neuropeptides exert a direct effect on the same isolated single smooth muscle cell. Am J Physiol Cell Physiol 250: C792-C798, 1986[Abstract/Free Full Text].

36.   Lesage, F, and Lazdunski M. Molecular and functional properties of two-pore-domain potassium channels. Am J Physiol Renal Physiol 279: F593-F801, 2000[Abstract/Free Full Text].

37.   Limatola, C, Schaap D, Moolenaar WH, and van Blitterswijk WJ. Phosphatidic acid activation of protein kinase C-zeta overexpressed in COS cells: comparison with other protein kinase C isotypes and other acidic lipids. Biochem J 304: 1001-1008, 1994.

38.   Lu, T, Katakam PV, Van Rollins M, Weintraub NL, Spector AA, and Lee HC. Dihydroxyeicosatrienoic acids are potent activators of Ca2+-activated K+ channels in isolated rat coronary arterial myocytes. J Physiol 534: 651-667, 2001[Abstract/Free Full Text].

39.   Lundbaek, JA, Birn P, Girshman J, Hansen AJ, and Andersen OS. Membrane stiffness and channel function. Biochemistry 35: 3825-3830, 1996[Medline].

40.   Macica, CM, Yang Y, Hebert SC, and Wang WH. Arachidonic acid inhibition of cloned renal K+ channel, ROMK1. Am J Physiol Renal Fluid Electrolyte Physiol 271: F588-F594, 1996[Abstract/Free Full Text].

41.   MacKinnon, R, Latorre R, and Miller C. Role of surface electrostatics in the operation of a high-conductance Ca2+-activated K+ channel. Biochemistry 28: 8092-8099, 1989[Medline].

42.   MacKinnon, R, and Miller C. Functional modification of a Ca2+-activated K+ channel by trimethyloxonium. Biochemistry 28: 8087-8092, 1989[Medline].

43.   Mannuz