The pure anti-oestrogen ICI 182,780 (Faslodex™) activates large conductance Ca²⁺-activated K⁺ channels in smooth muscle

Abstract

Oestrogen and tamoxifen activate large conductance Ca²⁺-activated K⁺ (BK_Ca) channels in smooth muscle through a non-genomic mechanism that depends on the regulatory β1 subunit and an extracellular binding site. It is unknown whether a ‘pure' anti-oestrogen such as ICI 182,780 (Faslodex™), that has no known oestrogenic properties, would have any effect on BK_Ca channels. Using single channel patch clamp techniques on canine colonic myocytes, the hypothesis that ICI 182,780 would activate BK_Ca channels was tested. ICI 182,780 increased the open probability of BK_Ca channels in inside-out patches with an EC₅₀ of 1 μM. These data suggest that molecules with the ability to bind nuclear oestrogen receptors, regardless of oestrogenic or anti-oestrogenic nature, activate BK_Ca channels through this nongenomic, membrane-delimited mechanism. The identity and characteristics of this putative binding site remain unclear; however, it has pharmacological similarity to oestrogen receptors α and β, as ICI 182,780 interacts with it.

Introduction

Oestrogen (17-β oestradiol) has nongenomic, or alternative, effects unrelated to the nuclear events (Falkenstein et al., 2000) and some anti-oestrogens share this property. These compounds exert effects outside the classical signalling pathway of altering gene transcription. For example, oestrogen and tamoxifen activate BK_Ca channels in smooth muscle (Valverde et al., 1999; Dick et al., 2001). BK_Ca channels have a large unitary conductance and are Ca²⁺/voltage-sensitive. In smooth muscle cells, BK_Ca channels contain a regulatory β1 subunit that alters channel function in many ways (McManus et al., 1995). Appreciation for the physiological importance of the β1 subunit has recently increased, as genetically engineered mice lacking this regulatory subunit are hypertensive and demonstrate altered smooth muscle reactivity (Brenner et al., 2000). An extracellular binding site and the β1 subunit are essential for sensitivity of BK_Ca channels to oestrogen and tamoxifen (Valverde et al., 1999; Dick et al., 2001; 2002; Dick & Sanders, 2001). The identity of the extracellular binding site is not known (Nadal et al., 2001).

Tamoxifen is used clinically as an oestrogen receptor antagonist for the treatment and prevention of breast cancer. However, tamoxifen has mixed oestrogenic and anti-oestrogenic properties, ranging from full agonist to full antagonist depending on the tissue and endpoint examined. Tamoxifen inhibits the binding of oestrogen to nuclear receptors preventing cell proliferation. In contrast, tamoxifen has oestrogenic properties in liver, bone, and uterus, where it alters blood lipids, bone density, and endometrial thickness (Jordan, 2000). Because of oestrogenic properties, tamoxifen is associated with adverse chemotherapeutic reactions including tumour flare, hot flashes, and the stimulation of liver and myometrial carcinomas (Jordan, 2000; Howell et al., 2000). Thus, the search for oestrogen receptor antagonists without oestrogenic properties has been an area of intense interest. ICI 182,780 is a highly specific oestrogen receptor antagonist marketed under the trade name Faslodex (Howell et al., 2000). ICI 182,780 is used for treating oestrogen receptor-dependent tumours, and, in contrast to tamoxifen, has no known oestrogenic properties.

The question remains as to whether the ability of tamoxifen to activate BK_Ca channels is due to its partial oestrogenic properties. This was determined by assessing the effect of a pure anti-oestrogen, ICI 182,780. The results delineate a novel mechanism of action and indicate that the putative extracellular binding site has pharmacology in common with oestrogen receptors α or β, as a pure anti-oestrogen interacts with it. This interaction is translated, by an unknown mechanism, into changes in BK_Ca channel activity.

Methods

Canine colonic smooth muscle cells were isolated by enzymatic dispersion described previously (Dick et al., 2001; 2002). The methods were approved by an Animal Care and Use Committee and were consistent with NIH guidelines. Myocytes were suffused at 3 ml min⁻¹ with (mM) KCl 140, HEPES 10, Tris 5; pH 7.1. Additionally, this solution contained either 1 mM EGTA or 1 mM HEDTA (N-[2-hydroxy ethyl]-ethylenediamine-triacetic acid) and between 0.01–10 μM free Ca²⁺ (Maxchelator 2.05; Pacific Grove, CA, U.S.A.).

Data were analysed with the Analysis of Single Channel Data program (Guy Droogmans; University of Leuven, Belgium). NP_o (number of channels multiplied by the open probability) was measured before, during, and after exposing the patches to ICI 182,780 (Tocris Cookson Inc., Ballwin, MO, U.S.A.). ICI 182,780 was made as a 10 mM stock in DMSO and diluted to the final concentration in bathing solutions. DMSO, at concentrations of 0.1% and less, has no effect on BK_Ca channels (Dick et al., 2001). NP_o and unitary conductance were determined from all-points amplitude histograms. Data are presented as the mean±s.e. from n cells. Statistical analyses were made with Student's paired t-test or one-way analysis of variance (with repeated measures and Bonferroni post-hoc analysis as necessary). The threshold for statistical significance was P<0.05. Statistical tests were run in SigmaStat (version 2; Jandel Scientific Software; San Rafael, CA, U.S.A.).

Results

ICI 182,780 increases the NP_o of BK_Ca channels by a nongenomic mechanism

BK_Ca channels were studied in inside-out patches of membrane taken from canine colonic myocytes as described previously (Dick & Sanders, 2001; Dick et al., 2001, 2002). ICI 182,780 (10 μM) increased NP_o when applied to the bath (Figure 1A–D). In contrast to the effect of tamoxifen, ICI 182,780 increased NP_o without affecting the unitary conductance of BK_Ca channels (Figure 1E).

Figure 1.

ICI 182,780 increases BK_Ca channel NP_o in excised membrane patches. Representative 1 min recordings of BK_Ca channel activity in an inside-out patch before (panel A; control; C) and after a 5 min exposure to 10 μM ICI 182,780 (ICI; panel B). Patch potential was +80 mV; 100 nM free Ca²⁺. ICI 182,780 increased NP_o without effect on the unitary conductance. Scale bar represents 20 pA and 3 s. Panel C group data (n=13) demonstrating the effect of 10 μM ICI 182,780 on BK_Ca channel NP_o. The asterisk indicates a significant effect of ICI 182,780 on NP_o (P=0.001 by Student's paired t-test). Panel D NP_o in the presence of 10 μM ICI 182,780 normalized to control. Panel E ICI 182,780 had no effect on the unitary conductance of BK_Ca channels (P=0.60 by Student's paired t-test).

Effects of ICI 182,780 on BK_Ca channels are rapid, but not readily reversible

The increase of BK_Ca channel NP_o by ICI 182,780 was observed in cell-free patches of membrane, indicating it is a nongenomic effect. Additionally, a hallmark of nongenomic steroid effects is the speed of the effect. Figure 2 shows that ICI 182,780 (10 μM) rapidly increased BK_Ca channel NP_o. The time from ICI 182,780 (10 μM) application to reach a steady-state NP_o was 4.8±0.7 min (n=13). The effect of ICI 182,780, however, was not readily reversible. As can be seen in Figure 2A, brief exposure to ICI 182,780 caused an increase in NP_o that persisted for approximately 30 min. In six cells, NP_o was 306±38% higher than control 21±2 min after 10 μM ICI 182,780 was washed out of the bath.

Figure 2.

Rapid, but not readily reversible, effect of ICI 182,780 on BK_Ca NP_o. Panel A shows a representative recording of BK_Ca channel activity in an inside-out patch. Patch potential was +100 mV; 100 nM free Ca²⁺. Current was recorded for 2 min prior to exposing the patch to 10 μM ICI 182,780 for 3 min, which increased NP_o. NP_o remained elevated after ICI 182,780 was washed out; an effect that persisted after a 15 min wash. Scale bar represents 20 pA and 1 min Panel B is a plot of NP_o vs time from another patch under the same recording conditions. NP_o was steady after patch excision and increased with the application of 10 μM ICI 182,780.

The increase in NP_o by ICI 182,780 is concentration-dependent and saturable

Increasing concentrations of ICI 182,780 (0.1–10 μM) were added to solutions bathing the cytoplasmic face of inside-out patches. ICI 182,780 increased NP_o in a concentration-dependent manner. The EC₅₀ was 1.0±0.2 μM (n=10; Figure 3). The unitary conductance under control conditions was 260±4 pS (n=10) and did not change with exposure to ICI 182,780 (P=0.69 by one-way repeated measures analysis of variance).

Figure 3.

Concentration-dependent effect of ICI 182,780 on BK_Ca channel NP_o. Group data (n=10) showing the concentration-dependent effect of ICI 182,780 on BK_Ca channel NP_o at +80 mV and 100 nM free Ca²⁺. NP_o was measured at least 5 min after each concentration of ICI 182,780 was applied. The EC₅₀ was 1 μM. Asterisks indicate a difference from control (C) by one-way repeated measures analysis of variance.

ICI 182,780 shifts the voltage-dependence of BK_Ca gating to more negative potentials

In all the experiments described thus far, the bath free Ca²⁺ concentration was 100 nM, near the lower limit where the influence of the β1 subunit on BK_Ca channel gating is thought to decline. To determine the effect of ICI 182,780 on channel gating over a wider range of free Ca²⁺ concentrations, and thus NP_o, two different solutions were used. In the first, the bath contained 1 mM EGTA and no added Ca²⁺, setting the free Ca²⁺ concentration near 10 nM. Under these low Ca²⁺ conditions, ICI 182,780 increased NP_o 420±95% (n=8). The second solution contained 1 mM HEDTA and 10 μM free Ca²⁺. Under these high Ca²⁺ conditions, ICI 182,780 increased NP_o 431±107% (n=7). Additionally, an activation curve was made in 100 nM free Ca²⁺. ICI 182,780 (10 μM) shifted the voltage of half activation from 106±6 to 86±5 mV (n=7; P<0.0001 by Student's paired t-test; Figure 4).

Figure 4.

ICI 182,780 shifts the voltage-dependence of BK_Ca gating. Group data (n=7) demonstrating a negative shift in the voltage of half activation by 10 μM ICI 182,780. The patch potential was held at 0 mV and stepped to potentials from +10 to +200 mV; 100 nM free Ca²⁺. Conductance was calculated from current, normalized to the maximum, and fit with a Boltzmann sigmoidal function.

Discussion

ICI 182,780 activates BK_Ca channels in smooth muscle, a previously unrecognized action that is rapid, concentration-dependent, and similar to that of oestrogen and tamoxifen (Valverde et al., 1999; Dick & Sanders, 2001; Dick et al., 2001). This is a nongenomic, or alternative, effect as it does not involve changes in gene transcription (Falkenstein et al., 2000). The activation of BK_Ca channels by [xeno]oestrogens is not strictly due to oestrogenic properties, as a pure anti-oestrogen, ICI 182,780, has similar effects. The extracellular binding site has pharmacology in common with oestrogen receptor α and β, as oestradiol, tamoxifen, and ICI 182,780 interact with it. This binding is translated, by an unknown mechanism, into changes in BK_Ca channel activity.

ICI 182,780 (AstraZeneca, Cheshire, U.K.) is a steroidal oestrogen receptor antagonist (Howell et al., 2000). ICI 182,780 is thought to be a pure anti-oestrogen (i.e., it does not have any known oestrogen agonist activity). There are few data regarding nongenomic mechanisms of action for ICI 182,780. One example is inhibition of L-type Ca²⁺ channels (i.e., Ba²⁺ current) by ICI 182,780 in A7r5 cells (Ruehlmann et al., 1998). ICI 182,780 also increased K⁺ current in three of four cells tested; however, the nature of the current, and the concentration-dependence and magnitude of the ICI 182,780 effect were not determined. ICI 182,780 also reduces coronary vascular tone in isolated perfused rat hearts (Ruehlmann et al., 1998). This could be due, in part, to inhibition of L-type Ca²⁺ channels or activation of K⁺ channels. Increased outward current could be due to diminished inward current, and not the activation per se, of an outward current. These results suggest that previous observations could be explained by activation of BK_Ca channels.

The activation of smooth muscle BK_Ca channels by [xeno]oestrogens depends on the presence of the regulatory β1 subunit (Dick & Sanders, 2001; Dick et al., 2001; Valverde et al., 1999). There are four known regulatory β subunits, but only β1 and β4 (Behrens et al., 2000) confer oestrogen sensitivity upon BK_Ca channels. Studies using membrane impermeant [xeno]oestrogens suggest an extracellular binding site exists, perhaps within the extracellular loop of the BK_Ca channel β1 subunit (Valverde et al., 1999; Dick et al., 2002). However, preliminary data indicate that a chimeric β subunit containing the β1 extracellular loop and the β2 intracellular and transmembrane portions does not respond to oestrogen or tamoxifen (Orio et al., 2002). It remains to be determined whether this chimeric β1/2 subunit binds [xeno]oestrogens or whether additional portions of the β1 subunit are necessary for transducing the response to BK_Ca channels. It is possible that a yet unidentified mediator, such as oestrogen receptor α or β, may be a necessary molecular component.

Gene transcripts for both the α and β isoforms of the oestrogen receptor are present in smooth muscle (Hodges et al., 2000); however, the hypothesis that one or both of these receptor isoforms is involved in this signalling pathway to BK_Ca channels has not been tested. As attractive as the possibility seems, there are difficulties with suggesting that either oestrogen receptor α or β is a transducing element in the membrane (Nadal et al., 2001). Neither oestrogen receptor α nor β has a predicted transmembrane spanning domain. This limitation might be overcome by post-translational modification; however, neither isoform has an obvious site. Additionally, while Xenopus oocytes (Valverde et al., 1999) and human embryonic kidney cells (Dick et al., 2001) serve as competent expression systems for studying the β1-dependent effects of [xeno]oestrogens on BK_Ca channels, neither is thought to possess endogenous oestrogen receptors, and are commonly used to express them (Kahlert et al., 2000; Watson, 1991). Whether the plasma membrane receptor for [xeno]oestrogens is identical to, related to, or completely distinct from oestrogen receptor α or β remains a matter of debate.

Regardless of the molecular machinery involved, ICI 182,780 activates smooth muscle BK_Ca channels in a non-genomic fashion with an EC₅₀ near 1 μM. This effect on BK_Ca channels in smooth muscle would not be expected to have clinical significance, as the therapeutic serum concentration in humans is approximately 12.6 ng ml⁻¹ (≈21 nM; (Howell et al., 1996)). Thus, the effects of ICI 182,780 on BK_Ca channel NP_o, and L-type Ca²⁺ channels (Ruehlmann et al., 1998), in vitro are achieved only at much higher concentrations. However, the data suggest that molecules with the ability to bind nuclear oestrogen receptors, regardless of oestrogenic or anti-oestrogenic nature, activate BK_Ca channels through this membrane-delimited mechanism. Further research is required to identify the putative extracellular binding site for [xeno]oestrogens.

Abbreviations

BK_Ca

large conductance Ca²⁺-activated K⁺ channel

tamoxifen

[Z]-1-[p-dimethylaminoethoxy-phenyl]-1,2-diphenyl-1-butene

ICI 182,780

(7α-[9-[(4,4,5,5,5-pentafluoropentyl)sulphinyl]nonyl]-estra-1,3,5(10)-triene-3,17β-diol)

HEDTA

N-[2-hydroxyethyl]-ethylenediamine-triacetic acid