Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The vascular endothelium modulates tone of the underlying smooth muscle by releasing a number of different contracting and relaxing factors. Among these, endothelin, nitric oxide, and prostacyclin have been particularly well characterized (Furchgott and Vanhoutte, 1989). Recent studies suggest that an additional factor [i.e., epoxyeicosatrienoic acid (EET)], which is a product of the cytochrome P450 pathway, is also synthesized and released from the endothelium (Bauersachs et al., 1994; Hecker et al., 1994; Campbell et al., 1996). Endothelial cells possess cytochrome monooxygenase activity (Abraham et al., 1985; Pinto et al., 1987; Rosolowsky et al., 1990), and several cytochrome P450 isoforms have been described in endothelial cells. In vitro studies have shown that EETs relax coronary, pial, cerebral, caudal, and renal arteries; in some studies, membrane hyperpolarization has been observed (Hecker et al., 1994; Campbell et al., 1996; Fukao et al., 1997; Eckman et al., 1998). These results suggest that EETs contribute to endothelium-dependent relaxation and hyperpolarization in some blood vessels.

The vasodilatory action of EET seems to be due in large part to activation of large conductance Ca²⁺-activated K channels (BK_Ca). Patch-clamp studies have shown that EETs increase the open probability of BK_Ca channels in native cells (Campbell et al., 1996; Li and Campbell, 1997; Hayabuchi et al., 1998). Membrane potential measurements in intact blood vessels have reported that EET-induced hyperpolarization is blocked by the BK_Ca channel blocker iberiotoxin (Eckman et al., 1998). Finally, EET-induced relaxation can be reduced or abolished by either iberiotoxin or tetraethylammonium (Campbell et al., 1996; Li and Campbell, 1997; Eckman et al., 1998; Li et al., 1999). Previous studies of native cells suggest that EETs enhance BK_Ca activity by activating the G protein Gs (Li and Campbell, 1997) via ADP-ribosylation (Li et al., 1999). Some studies suggest that this is a direct membrane-delimited action of Gs on the channel (Campbell et al., 1996; Li and Campbell, 1997).

Many details concerning the mechanism by which EETs modulate BK_Ca channel activity remain unclear. For example, a specific receptor for EETs has yet to be positively identified. Furthermore, the mechanism by which the G protein Gs leads to activation of the BK_Ca channel activity is unknown. It is possible that G protein subunits interact specifically with the -subunit of the BK_Ca channel. Alternatively, the -subunit may be involved in the process. Expression systems are particularly useful to address these questions, because it is possible to express a known isoform of the BK_Ca -subunit in the absence of the -subunit and ultimately to manipulate with molecular biology techniques the predicted components of the pathway. The goal of the present study was therefore to determine whether the pathway previously described for EET-induced modulation of BK_Ca channels in native cells (Campbell et al., 1996; Li and Campbell, 1997; Li et al., 1999) is present when a known isoform of the -subunit of the BK_Ca channel is expressed in a mammalian expression system (i.e., HEK293 cells). In previous studies, we have shown that BK_Ca -subunits (cslo-) expressed in HEK293 cells give rise to voltage-gated, Ca²⁺-sensitive currents with electrophysiological and pharmacological features similar to those of native BK_Ca (Adelman et al., 1992; Esguerra et al., 1994; Perez et al., 1994; Fukao et al., 1999). Specific experiments in this study were designed to determine whether: 1) EETs enhance cslo- channel activity in this expression system, 2) the G protein s and/or subunit is involved, 3) activation involves ADP-ribosylation, and 4) activation of cslo- involves the adenylyl cyclase/protein kinase A (PKA) pathway or a direct membrane-delimited pathway. Our results reveal a striking similarity in the pathway identified previously from experiments in native vascular smooth muscle cells and the pathway present in HEK293 cells expressing cslo-. These results suggest that all of the elements necessary for the EET pathway are present in HEK293 cells expressing cslo- and that the -subunit of BK_Ca is not required for activation. This expression system represents a promising model for future studies of this important regulatory pathway.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Expression of cslo- Channels. The cDNA encoding the -subunit of the canine colonic BK_Ca channel (cslo-) was expressed in HEK293 cells as described previously (Fukao et al., 1999).

Electrophysiological Recording. The patch-clamp technique was used to measure membrane currents in whole-cell and isolated patch configurations as previously described (Fukao et al., 1999). Data acquisition and analysis were performed with pClamp software (version 6.0.4; Axon Instruments, Burlingame, CA). Channel open probability (NP_o) in patches was determined from recordings of more than 3 min by fitting the sum of Gaussian functions to an all-points histogram plot at each potential. Single-channel conductance was determined from all-point amplitude histograms using Fechan and P-stat programs (Axon Instruments). Capacitance compensation and series resistance compensation (80%) were performed.

Solutions and Drugs. For whole-cell recordings of HEK293 cells, the bath solution contained 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, and 10 mM glucose, pH 7.4, and the pipette solution contained 50 mM KCl, 70 mM KAsp, 8 mM NaCl, 0.826 mM CaCl₂, 1 mM MgCl₂, 2 mM MgATP, 0.1 mM NaGTP, 10 mM HEPES, and 1 mM HEDTA, pH 7.2. For single-channel recordings in the inside-out mode, the bath solution contained 140 mM KCl, 1 mM MgCl₂, 10 mM HEPES, and 1 mM HEDTA, pH 7.2. The concentration of free Ca²⁺ in the bath solution was varied (range, 10⁸ to 10⁴ M) to determine the Ca²⁺ sensitivity of BK_Ca channels. The Ca²⁺ concentration was estimated by a computer program (Bers et al., 1994), and the appropriate amounts of CaCl₂ were added. The ionized Ca²⁺ concentration was confirmed using a Ca²⁺-sensitive electrode. The pipette solution contained 140 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, pH 7.4. For single-channel recordings in the cell-attached mode, the bath solution contained 140 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, and 10 mM glucose, pH 7.4, and the pipette solution contained 140 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES, pH 7.4. All patch-clamp experiments were performed at room temperature (22°C). 11,12-EET was purchased from Cayman Chemical (Ann Arbor, MI). KT5720, SQ22536, cholera toxin, and all antibodies were obtained from Calbiochem (San Diego, CA). Anti-G is a polyclonal antibody raised against brain G (catalog no. 371821). Anti-Gi/o is a mixture of two antibodies: 1) Anti-Gi1 and Gi2 subunit antibody (catalog no. 371723) generated by using a synthetic peptide antigen corresponding to a C-terminal decapeptide sequence (345-354) found in both Gi1 and Gi2. 2) Anti-Gi3 and Go-subunit antibody (catalog no. 371726) generated by using a synthetic peptide antigen corresponding to the C-terminal sequence 345-354 of Gi3. Anti-Gs antibody (catalog no. 371732) generated by using a synthetic peptide antigen corresponding to the C-terminal sequence 385-394 of Gs. 3-Aminobenzamide (3-AM) and m-iodobenzylguanidine (MIBG) were obtained from Sigma Chemical Co. (St. Louis, MO). 11,12-EET was dissolved in ethanol and KT5720, forskolin in dimethyl sulfoxide. Solvent per se had no effect on channel activity at final concentration (ethanol, 0.03%; dimethyl sulfoxide, 0.1%).

Total RNA Isolation and Reverse Transcriptase-Polymerase Chain Reaction. Total RNA was prepared from human jejunum smooth muscle and cultured HEK293 cells using the SNAP Total RNA isolation kit (Invitrogen, San Diego, CA) as per the manufacturer's instructions. First-strand cDNA was prepared from the RNA preparations using the Superscript II reverse transcriptase kit (Life Technologies, Inc., Gaithersburg, MD), 500 µg/µl of oligo(dT) primers were used to reverse transcribe the 1-µg RNA sample. The cDNA reverse transcription product was amplified with -slo-specific primers by polymerase chain reaction (PCR) (Epperson et al., 1999). The amplification profile for these primer pairs was: 95°C for 10 min to activate the amplitaq polymerase (PE Biosystems, Foster City, CA), 95°C for 15 s, and 60°C for 1 min; each for 40 cycles. The amplified products (5 µl) were separated by electrophoresis on a 4% agarose/1× TAE (Tris, acetic acid, EDTA) gel, and the DNA bands were visualized by ethidium bromide staining. The reverse transcriptase control used a cDNA reaction as template for which the reverse transcriptase was not added, controlling for genomic DNA contamination in the source RNA. The no template control was a PCR amplification for which the template was not added, controlling for nonspecific amplification and spurious primer dimer fragments. These negative controls were subjected to a second round of amplification to assure specificity of the reactions and the quality of the reagents.

Primer Design. The following PCR primers were used (the GenBank accession numbers are given in parentheses for the reference nucleotide sequences used): human -slo (U25138) nucleotides 199-218 and 376-397; amplicon = 199 base pairs (bp), -actin (V01217) nucleotides 2204-2224 and 2384-2402; amplicon = 198 bp.

Statistics. Data are expressed as mean ± S.E. Statistical significance for repeated measures was determined using analysis of variance. P < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The endogenous currents of HEK293 cells and the currents recorded in cells expressing cslo- were previously characterized by this group using the whole-cell and inside-out excised patch configurations. Endogenous currents are much smaller in amplitude than those recorded from cells expressing cloned cslo- channels and thus do not significantly contaminate recordings (Fukao et al., 1999).

Native HEK293 cells were examined for endogenous expression of BK_Ca  subunit mRNA and compared with freshly isolated human jejunal smooth muscle cells. Using primers designed to hybridize to conserved regions of -slo and reverse-transcriptase PCR, no detectable product was observed in HEK293 cells, whereas an abundant message was detected in jejunal cells (Fig. 1). These results suggest that the actions of EET on cslo- subunits can be examined in HEK293 cells in the absence of the cslo- subunit.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   Lack of slo amplification products in cultured HEK293 cells. Shown is a representative gel containing amplification products from reverse transcription-PCR using human jejunum smooth muscle and cultured HEK293 cells as the source of RNA. Although abundant slo message was detected in human jejunal cells, no message was apparent in HEK293 cells. A single 40-cycle amplification was used to generate the products. The amplicon for slo was 199 bp and, for -actin, 198 bp. Reverse transcription and no template control represent negative controls testing the quality of the reagents used (see Materials and Methods).

Effect of 11,12-EET on Whole-Cell cslo- Currents. Experiments were undertaken to determine the action of 11,12-EET on cells expressing cslo- using the whole-cell, patch-clamp mode. Cytosolic Ca²⁺ concentration was buffered at 10 µM with HEDTA in these experiments. Addition of 11,12-EET (1 µM) to the bathing solution led to a significant increase in whole-cell outward current (Fig. 2A). Steady-state current was obtained after approximately 5 min (Fig. 2B). A representative voltage-current relationship of cslo- current before and after treatment of 11,12-EET is shown in Fig. 2C. In eight cells tested, 11,12-EET significantly (P < 0.05) increased outward current amplitude 2-fold at +50 mV (Fig. 2D).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Effect of 11,12-EET on whole-cell cslo- current expressed in HEK293cells. A, representative traces of cslo- current expressed in HEK293 cells in the absence and presence of 11,12-EET (1 µM). Membrane potential was held at 70 mV and stepped at 15-s intervals to potentials between 50 mV and +80 mV in 10-mV increments for 200 ms and then held at 30 mV for 20 ms. B, time course of the steady-state current amplitude before and after treatment of 11,12-EET (1 µM) at a membrane potential of +50 mV. C, current-voltage relationship of cslo- currents before () and after () application of 11,12-EET (1 µM). D, averaged cslo- currents of control and after application of 11,12-EET (1 µM). Data shows averaged peak currents at voltage clamp steps from 70 to +50 mV at 30-s intervals. Values are means ± S.E. (n = 8). *P < 0.05 compared with control.

Effect of 11,12-EET on cslo- Channel Activity in Cell-Attached Patches. Additional experiments were performed to determine whether changes also occur in single-channel activity recorded in cell-attached patches. 11,12-EET increased cslo- channel activity in a concentration-dependent manner (Fig. 3A). 11,12-EET at concentrations between 0.1 and 1 µM produced a 2- to 3.5-fold increase in NP_o of cslo- channels (Fig. 3B) but had no effect on the unitary conductance [control, 238 ± 9.5; 11,12-EET (1 µM), 246 ± 10; n = 7, P > 0.05].

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Effect of 11,12-EET on cslo- channel activity in cell-attached patches. A, representative recording of cslo- channel under control condition (top trace) and after application of 0.3 µM 11,12-EET (middle trace) and 1 µM 11,12-EET (bottom trace) to the bath solution. Cells were held at +40 mV. c, the closed state. B, summary of the effect of 11,12-EET (0.1, 0.3, and 1 µM) on NPo of the cslo- channels (n = 6-13). Values are means ± S.E. *P < 0.05 compared with control.

Effect of 11,12-EET on cslo- Channel Activity in Inside-Out Patches. In contrast to the stimulatory effect of 11,12-EET on cslo- channel in cell-attached patches, 11,12-EET did not significantly affect cslo- channel activity in inside-out patches when applied to the cytosolic surface of the membrane (n = 12). The unitary conductance of cslo- channels was also unchanged by 11,12-EET in inside-out patches (control, 247 ± 19; EET, 247 ± 10; n = 11; P > 0.05). Because activation of channels by 11,12-EET may involve phosphorylation, additional experiments were undertaken with ATP. Application of ATP (1 mM) to the bath solution had no effect on cslo- channel activity (n = 8). Likewise 11,12-EET was without effect in the presence of ATP (n = 4). These results suggest that ATP alone is insufficient to support activation of cslo- channels by 11,12-EET in isolated patches.

Effect of GTP and GTPS on cslo- Channel Activity in Inside-Out Patches. There is evidence from native cell experiments that EET can lead to activation of BK_Ca channels via Gs (Li et al., 1997). To explore the role of G proteins in our expression system, additional experiments were undertaken with GTP and the nonhydrolyzable analogs GTPS and GDPS. GTP (100 µM) significantly increased cslo- channel activity in inside-out patches without affecting the single-channel conductance (Fig. 4, A and B). In the presence of 100 µM GTP, addition of 11,12-EET (1 µM) led to a further increase in cslo- channel activity (Fig. 4, A and B). These data suggest that 11,12-EET activates cslo- via a GTP-dependent mechanism. GDPS (200 µM), which inhibits GTP-dependent pathways, did not affect the basal cslo- channel activity in inside-out patches (Fig. 4D), but completely blocked the stimulatory effect of GTP as well as GTP plus 11,12-EET (Fig. 4B). The GTP analog GTPS (10 µM) also increased channel activity of cslo- in inside-out patches (Fig. 4, C and D). This effect was also blocked by GDPS (Fig. 4D).

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Effect of GTP and GTPS on cslo- channel activity in inside-out patches. A, representative recording of cslo- channel in inside-out patches at membrane potential of +40 mV. GTP (100 µM) increased the cslo- channel activity in inside-out patches. In the presence of GTP, 11,12-EET gave rise to a further increased the channel activity. Cytosolic Ca2+ concentration was maintained at 106 M. c, the closed state. B, summary of the effect of GTP (n = 16), GTP + 11,12-EET (n = 5), GDPS (200 µM) + GTP (n = 6), and GDPS + GTP + 11,12-EET (n = 5) on cslo- channel activity. C, representative recordings of cslo- channel activity before (top trace) and in the presence of GTPS (10 µM). GTPS increased the cslo- channel activity in inside-out patches (second trace). In a different patch, GTPS was without effect on cslo- channel activity when added in the presence of GDPS (200 µM). D, summary of the effect of GTPS (n = 6), GDPS (n = 6), and GDPS + GTPS (n = 5) on cslo- channel. Responses under the various conditions were compared with the control condition (normalized to 1). Values are means ± S.E. *P < 0.05 compared with control. **P < 0.05 compared with GTP.

Effect of Anti-Gs Antibody on the Stimulatory Effect of 11,12-EET on cslo- Channels. To further investigate the nature of the GTP-dependent response, antibodies to various G proteins were tested in experiments using the whole-cell configuration. Specific antibodies against Gs, Gi/o, and G were included in the pipette solution to inhibit the effects of these G protein subunits. After obtaining a stable whole-cell current, 11,12-EET was added to the bathing solution to activate cslo- channels. There was no significant difference in the ability of 11,12-EET to stimulate cslo- current in the presence of anti-Gi/o or anti-G antibodies. In contrast, 11,12-EET was without effect on cslo- current in the presence of anti-Gs antibody (Fig. 5A), suggesting that 11,12-EET activated cslo- via Gs.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Effect of various agents that modify G protein function on responses to 11,12-EET in HEK293 cells. A, effect of G protein antibodies. The bar graph shows a summary of the effect of anti-Gs, anti-Gi/o, and anti-G antibodies on the stimulatory effect of 11,12-EET of cslo- current in whole-cell, patch-clamp mode. Antibodies were included in the pipette solution and dialyzed into cells. Peak current attained during a voltage step from 70 to +50 mV in the presence of 11,12-EET was normalized to the current obtained in the absence of EET. Only the anti-Gs antibody produced a significant reduction in the stimulatory effect of 11,12-EET on cslo- channel activity. Antibodies were diluted to produce a final concentration that was two times greater than that necessary for a Western blot (1:500) (n = 6-10). B, effect of the Gs activator cholera toxin (CTX). The bar graph shows a summary of the effect of 11,12-EET (1 µM, n = 13), CTX (100 ng/ml, n = 8), and CTX + 11,12-EET (1 µM, n = 5) on cslo- channel activity in cell-attached patches. Channel activity following 15-min exposure to cholera toxin was compared with the control channel activity (normalized to 1). The membrane potential was clamped at +40 mV. C, effect of ADP-ribosyltransferase inhibitors on the stimulatory effect of 11,12-EET on cslo- channel activity in cell-attached patches. Channel activity in the presence of inhibitor plus 11,12-EET was compared with the activity obtained with 11,12-EET in the absence of inhibitor (normalized to 1). The ADP-ribosyltransferase inhibitors, 3-aminobenzamide (3-AM, n = 6) and m-iodobenzylguanidine (MIBG, n = 7) both blocked the stimulatory effect of 11,12-EET on cslo- channel. Membrane potential was clamped at +40 mV. Values are means ± S.E. *P < 0.05 compared with control.

Effect of Cholera Toxin on cslo- Channel Activity in Cell-Attached Patches. To provide further evidence for coupling of Gs to cslo- channels, we tested cholera toxin (CTX), which activates Gs by ADP-ribosylation (Hopkins et al., 1988). Inclusion of CTX in the bathing solution gave rise to a significant increase in cslo- channel activity in cell-attached patches (Fig. 5B) without a change in single-channel conductance (control, 248 ± 13 pS; cholera toxin, 250 ± 12 pS; n = 8). In the presence of CTX, 11,12-EET did not produce a further increase cslo- channel activity (Fig. 5B).

Effect of Mono-ADP-ribosyltransferase Inhibitors on the Stimulatory Effect of 11,12-EET on cslo- Channels. A recent study by Li et al. (1999) has suggested that activation of native BK_Ca channels by EET involves ADP-ribosylation of Gs. To determine whether this same pathway is present in HEK293 cells, we investigated the effect of two different inhibitors of mono-ADP-ribosyltransferase, 3-aminobenzamide (Purnell and Whish, 1980) and m-iodobenzylguanidine (Smets et al., 1990). 3-AM (1 mM) did not affect basal channel activity in cell-attached patches. However, in the presence of 3-AM the stimulatory effect of 11,12-EET (1 µM) was abolished. Likewise, MIBG (100 µM) was without effect on basal channel activity but blocked the stimulatory effect of 11,12-EET on cslo- channels (Fig. 5C).

Effect of KT5720 and SQ22536 on the Stimulatory Effect of 11,12-EET on cslo- Channels. Our results suggest that 11,12-EET activates BK_Ca via ADP-ribosylation of the G protein Gs. Gs is a well known activator of the adenylyl cyclase/PKA pathway. To investigate the role of this pathway in the actions of Gs and 11,12-EET, additional experiments were undertaken with blockers of this pathway. The PKA inhibitor KT5720 (200 nM) was without effect on basal cslo- channel activity in cell-attached patches (Fig. 6). In addition, the stimulatory effect of 11,12-EET on cslo- channel was not inhibited by pretreatment with KT5720 (Fig. 6, A and B). In contrast, KT5720 completely abolished the stimulatory effect of the adenylyl cyclase activator forskolin on cslo- channels (Fig. 6B). The adenylate cyclase inhibitor SQ22536 (200 µM) was also without affect on basal cslo- channel activity (Fig. 6C) in cell-attached patches. Stimulation of BK_Ca channels by either 11,12-EET or cholera toxin was unchanged in the presence of SQ22536 (Fig. 6C). These results indicate that the adenylyl cyclase/PKA pathway is present in HEK293 cells but that this pathway cannot represent the predominant mechanism by which 11,12-EET and Gs activate cslo- channels.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Effect of blockers of the adenylyl cyclase/PKA pathway on the stimulatory effect of 11,12-EET on cslo- channel activity in cell-attached patches. A, representative recording of cslo- channel recorded at a membrane potential of +40 mV (top trace). The PKA inhibitor KT5720 alone (200 nM) was without effect on channel activity (middle trace), and 11,12-EET (1 µM) still enhanced cslo- channel activity in the presence of KT5720 (bottom trace). Cytosolic Ca2+ concentration was maintained at 106 M. c, the closed state. B, summary of the effect of KT5720 (n = 16), KT5720 + 11,12-EET (n = 9), forskolin (10 µM, n = 7), and KT5720 + forskolin (n = 6) on cslo- channel activity. C, effect of the adenylyl cyclase inhibitor SQ22536 on the actions of 11,12-EET and cholera toxin on HEK293 cells. The bar graph shows a summary of the effects of SQ22536 (200 µM, n = 17), 11,12-EET (1 µM, n = 13); SQ22536 + 11,12-EET (n = 7) and SQ22536 + CTX (100 ng/ml; n = 6) on cslo- channel activity in cell-attached patches. Effects on channel activity were compared with the control channel activity (normalized to 1). Values are means ± S.E. (n = 6-10) *P < 0.05 compared with control.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

11,12-EET is a cytochrome P450 product of the arachidonic acid cascade that is synthesized and released by the vascular endothelium and may serve as one of several different endothelium-derived factors, which relax and hyperpolarize the adjacent smooth muscle (Bauersachs et al., 1994; Hecker et al., 1994; Campbell et al., 1996). In the present study we found that 11,12-EET leads to activation of the cloned -subunit of the BK_Ca channel (cslo-) when expressed in HEK293 cells. This activation involves a novel pathway in which 11,12-EET leads to ADP-ribosylation of the G protein Gs. Gs in turn activates cslo- via a membrane-delimited pathway that is independent of PKA and may involve a direct action of Gs on the channel.

A number of studies have suggested that EETs relax blood vessels by activating BK_Ca channels in the smooth muscle (Hecker et al., 1994; Campbell et al., 1996; Li et al., 1997; Eckman et al., 1998). In the present study, 11,12-EET stimulated outward cslo- current in whole-cell recordings and enhanced the activity of cslo- channels in cell-attached patches without a change in single-channel conductance. The sensitivity of cloned cslo- channels to 11,12-EET (i.e., 0.1 µM EET) was between that reported for native BK_Ca channels of large arteries (i.e., 0.3-10 µM EET) (Gebremedhin et al., 1992; Hu and Kim, 1993; Eckman et al., 1998) and that of small arteries (1 nM EET) (Li and Campbell, 1997). At present, it is not known what factors contribute to these differences in sensitivity.

The -slo subunit has been cloned from a variety of tissue sources and species. All are derived from the same gene and exhibit very similar amino acid and nucleotide sequences across species. Northern blot analysis using a cDNA probe containing the conserved core region of cslo- has revealed that cslo- transcripts are ubiquitously expressed in a number of canine vascular muscles, including portal vein, renal artery, and pulmonary artery (Vogalis et al., 1996). Various splice forms have been detected at the transcriptional level, some of which are expressed with tissue specificity. It is not clear to what extent the complete form and the various splice forms are utilized translationally to form functional BK_Ca channels in any smooth muscle. In the only complete description of -slo in artery to date, Salkoff and coworkers detected two alternatively spliced forms of -slo in human aorta and umbilicus as well as the complete -slo sequence (McCobb et al., 1995). However, no functional differences between the three forms of -slo were observed when expressed in oocytes or Chinese hamster ovary cells. Neither of the splice variants described by McCobb et al. (1995) were detected in canine colon (Vogalis et al., 1996). The cslo- clone used in the present study is equivalent to the full-length hslo 1.1 form from McCobb et al. (1995). Using this molecular form we were able to mimic the actions of 11,12-EET and Gs previously described for native vascular BK_Ca channels, suggesting that cslo- is functionally indistinguishable from the form(s) expressed in vascular muscles, which give rise to the EET response.

GTP per se, as well as the nonhydrolyzable analog GTPS, gave rise to a significant increase in cslo- channel activity that was blocked by GDPS. These data suggest that the predominant action of G proteins on cslo- channels in HEK293 cells is stimulatory. In the presence of GTP, 11,12-EET caused a significant increase in channel activity, and this effect was also blocked by GDPS. In contrast, in the absence of GTP, 11,12-EET was without effect. Thus, 11,12-EET seems to activate channels via a GTP-dependent mechanism. Because anti-Gs antibody but not anti-Gi/o or anti-G antibodies blocked the actions of 11,12-EET, this suggests that 11,12-EET stimulates cslo- via the GTP-binding protein Gs. The known Gs activator cholera toxin also enhanced BK_Ca channel activity, providing additional evidence for coupling between Gs and cslo- channels. This conclusion is in agreement with previous studies of 11,12-EET in native bovine coronary artery cells (Li and Campbell, 1997).

Recently it has been suggested that 11,12-EET can activate mono-ADP-ribosyltransferase, which leads to the transfer of ADP-ribose to the 52-kDa G protein Gs, resulting in activation of BK_Ca channels in small bovine coronary arteries (Li et al., 1999). A similar result was previously reported for EET in the rat liver (Seki et al., 1992). In agreement with these data, we observed that the stimulatory effect of 11,12-EET on cslo- was blocked by two different mono-ADP-ribosyltransferase inhibitors, 3-AM and MIBG. This suggests that ADP-ribosylation of Gs is also important in the regulation of the cloned -subunit of BK_Ca by 11,12-EET. The pathway by which 11,12-EET leads to activation of mono-ADP-ribosyltransferase remains unclear. A high-affinity binding site for 14(R),15(S)-EET in guinea pig mononuclear membranes has been reported, suggesting that a receptor for EET may exist (Wong et al., 1993). Thus, 11,12-EET may stimulate specific receptors that activate mono-ADP-ribosyltransferase, leading to ADP-ribosylation of Gs. Interestingly, this action mimics cholera toxin, which is an exogenous ADP-ribosyltransferase that also activates Gs by ADP-ribosylation (Hopkins et al., 1988).

In both native cell experiments and in expression systems there is good evidence that PKA activation leads to an increase in BK_Ca channel activity (Standen and Quayle, 1998) via phosphorylation of serine 869 (Nara et al., 1998). Indeed, BK_Ca channel activity was increased in the present study by the adenylyl cyclase activator forskolin. Inhibition of this effect by the PKA inhibitor KT5720 implies the existence of a functional adenylyl cyclase/PKA pathway in HEK293 cells, and we considered the possibility that this pathway might contribute to the Gs-dependent responses to 11,12-EET. However, in cell-attached patches, the stimulatory effect of 11,12-EET was not blocked by either the adenylyl cyclase inhibitor SQ22536 nor the PKA inhibitor KT 5720, providing direct evidence that the actions of 11,12-EET are independent of the adenylyl cyclase/PKA pathway. This conclusion is in agreement with a study by Campbell et al. (1996) in which 11,12-EET was shown to relax the bovine coronary artery without a significant change in tissue levels of either cAMP or cGMP. In studies of native cells, EET activates BK_Ca channels through a PKA-dependent mechanism in renal arteries (Imig et al., 1999), whereas in porcine (Hayabuchi et al., 1998) and bovine coronary arteries (Campbell et al., 1996) a PKA-independent pathway has been proposed. Multiple isoforms of adenylate cyclase and PKA exist (Houslay and Milligan, 1997). The variable role of PKA in the actions of EET may be related to: 1) the presence of different isoforms of adenylyl cyclase and PKA in different cells, 2) the quantity of isoforms present, and 3) the degree of coupling between Gs and adenylyl cyclase. In HEK293 cells the membrane-delimited actions of Gs seem to far outweigh those of the adenylyl cyclase/PKA pathway, because SQ22536 was also without effect on cholera toxin, which activates all Gs within the cell. This suggests very poor coupling between Gs and adenylyl cyclase in these cells. Thus, in HEK293 cells, 11,12-EET seems to stimulate cslo- via a direct membrane-delimited action of Gs. The nature of this interaction between channel and G protein requires further investigation but seems to involve ADP-ribosylation of Gs.

BK_Ca channels play a fundamental role in the regulation of membrane potential in smooth muscle, particularly under circumstances where intracellular calcium is elevated (Brayden and Nelson, 1992). In recent years it has become apparent that the activity of these channels can be importantly modulated by a variety of different physiological stimuli, including EET. Native BK_Ca channels are composed of pore-forming -subunits (i.e., -slo) plus a regulatory -subunit [predominantly 1 in smooth muscle (Jiang et al., 1999)] raising the possibility that the -subunit plays a role in regulation of BK_Ca channel activity by Gs. HEK293 cells transfected with specific BK_Ca subunits provide an excellent system to investigate this issue, because endogenous currents in general are small and there are no endogenous BK_Ca currents (Yu and Kerchner, 1998; Fukao et al., 1999) or message encoding the cslo- subunit. Accordingly, we interpret our results as indicating the regulation of cslo- channels expressed in the absence of -subunits. Further evidence of the lack of -subunits is that the voltage for half-maximal activation of the expressed BK_Ca currents with 10 µM free Ca²⁺ is +20 mV (Fukao et al., 1999), similar to values reported by others for activation of the -slo subunits in the absence of the -slo subunits (Toro et al., 1998)). Thus, the present study suggests that 11,12-EET activates BK_Ca channels through a direct action on the -subunit independent of the -subunit. However, we do not rule out a modulatory role for the  subunit in this process.

In summary, we have shown that the cslo- expressed in HEK293 cells is activated by 11,12-EET in both the whole-cell and single-channel configuration. Activation involves ADP-ribosylation of Gs but is independent of the adenylyl cyclase/PKA pathway, suggesting a direct, membrane-delimited pathway. These results agree well with previous studies of native cells (Campbell et al., 1996; Li and Campbell, 1997; Li et al., 1999) and suggest that the -subunit of BK_Ca is not required for this pathway. The HEK293 expression system seems to be a promising model for future studies of this important regulatory pathway.