The function of Maxi Cl− channels in cell physiology remains unresolved. Single-channel events were first described in skeletal muscle and subsequently in many different cell types, including epithelia (Hanrahan et al. 1985; Velasco et al. 1989; Vaca & Kunze, 1992; Brown et al. 1993; Riquelme et al. 1995), muscle (Blatz & Magleby, 1983; Saigusa & Kokubun, 1988), macrophages (Schwarze & Kolb, 1984), nerve cells (Forshaw et al. 1993; Bettendorff et al. 1993) and glial cells (Dermietzel et al. 1994). Almost all Maxi Cl− channel recordings have been made following excision of the membrane patch containing the channel. The need for membrane excision might indicate the involvement of intracellular inhibitory factors rather than artefactual activation by membrane disruption. An early indication of Maxi Cl− channel modulation came from experiments showing that protein kinase C regulates Maxi Cl− channels (Saigusa & Kokubun, 1988), and that excised Maxi Cl− channels can be modulated by GTP, GDP and some of their analogues (McGill et al. 1993). Subsequently, the observations that Maxi Cl− channels could be reversibly activated under whole-cell (Hardy & Valverde, 1994; Diaz et al. 1999) and cell-attached (Hardy & Valverde, 1994; Li et al. 2000) recording conditions by triphenylethylene antioestrogens such as toremifene or tamoxifen provided a method of studying these channels in intact cells.
Oestrogens exert most of their actions by binding and activating their receptors, which function as transcription factors (Beato, 1989). Antioestrogens are generally considered to act by competing with the binding of oestrogens to the oestrogen receptors (Jordan, 1984). In addition to the classical effect of oestrogen and antioestrogen, some of their actions are related to their interaction with binding sites, which generally determine a rapid effect (seconds to minutes) as opposed to the long-term genomic effect (> 30 min). In most cases, the rapid effects represent the interaction of oestrogens with a plasma membrane target and/or the generation of intracellular signals (Ropero et al. 1999; Falkenstein et al. 2000; Nadal et al. 2000, 2001).
Oestrogens have been found to exert rapid effects on the electrical activity of cells in the central nervous system (Minami et al. 1990), vascular system (Ruehlmann et al. 1998) and endocrine system (Nadal et al. 1998), and in non-excitable cells such as fibroblasts (Hardy & Valverde, 1994). Some of these actions can be either mimicked (Ruehlmann et al. 1998) or antagonised (Wong & Moss, 1991) by antioestrogens. However, the mechanisms underlying the rapid modulation of membrane excitability by oestrogens and antioestrogens are still poorly understood. Recently, both the direct interaction of steroids with the subunits forming channel structures (Valverde et al. 1999) and the generation of intracellular signals which, in turn, would modulate the activity of different ion channels have been described (Ropero et al. 1999).
In the study presented here we have demonstrated that triphenylethylene antioestrogens activate Maxi Cl− currents by binding to an extracellular plasma membrane site. Such activation depends upon the activity of an okadaic acid-sensitive protein phosphatase. On the other hand, the activation of Maxi Cl− channels by antioestrogens can be prevented by the extracellular application of 17β-oestradiol, a process that appears to be dependent upon a phosphorylation step.
- Top of page
In the present study we report the modulation of Maxi Cl− channels in C1300 mouse neuroblastoma cells by both oestrogens and antioestrogens. Antioestrogen-activated Maxi Cl− channels in C1300 cells exhibited a similar conductance (∼300 pS) and ionic selectivity (I− > Cl− > F−) to the Maxi Cl− channels activated by antioestrogens in fibroblasts (Hardy & Valverde, 1994), although the voltage-dependent inactivation differed slightly from one cell type to another. While the channel in the fibroblasts shows marked inactivation at positive and negative potentials (Hardy & Valverde, 1994), the channel in the neuroblastoma cell line (this study) only presented inactivation at negative potentials, although the inactivation constant varied greatly from cell to cell. A Maxi Cl− channel activated by antioestrogens has also been described in endothelial cells, although its voltage dependence was not reported (Li et al. 2000). Differences in the voltage-dependent inactivation of Maxi Cl− channels have been observed among different cell types (Blatz & Magleby, 1983; Schwarze & Kolb, 1984; Kolb et al. 1985; Forshaw et al. 1993; Riquelme et al. 1995). In this respect, it is interesting to note that the mitochondrial voltage-dependent anionic channel (VDAC), also known as mitochondrial porin, exhibits bell-shaped voltage dependence, a property associated with the presence of different protein modulators (Liu & Colombini, 1991) or polyamines (Horn et al. 1998). VDAC have also been identified in the plasma membrane (Dermietzel et al. 1994). Thus, although there is still much debate regarding the molecular identification of plasma membrane Maxi Cl− channels and whether they can be considered members of the VDAC family, it would be interesting to investigate whether the differences in voltage-dependent inactivation between fibroblast and neuroblastoma Maxi Cl− channels are due to the presence of different structural features, regulatory subunits, or associations with low molecular-weight molecules such as polyamines.
Maxi Cl− channel activity has been seen in intact cells only rarely, although it generally appears following membrane-patch excision. This observation has suggested the existence of regulatory mechanisms that keep the channel closed under non-stimulatory conditions. Early evidence of such intracellular regulation came from Saigusa & Kokubun (1988), who showed the activation of Maxi Cl− channels by inhibiting protein kinase activity with H7. That work suggested that phosphorylation of the Maxi Cl− channel or a regulatory subunit kept the channel closed. The activation of Maxi Cl− channels under the whole-cell and cell-attached patch-clamp configurations by the addition of extracellular antioestrogens is interesting for two reasons. First, it supports the hypothesis that activation of the channel following patch excision is most likely the result of altered intracellular regulatory mechanisms rather than merely a physical disruption of the membrane caused by the process of patch excision itself. Second, the fact that activation of the channels only occurs when triphenylethylene antioestrogens were added to the extracellular solution indicates the presence of an antioestrogen binding site on the plasma membrane that is accessible from the external side. This is supported by the use of the membrane-impermeant derivative EB-tamoxifen (Jarman et al. 1986), which was only effective when applied to the extracellular solution. The presence of both oestrogen (Pietras & Szego, 1977; Bression et al. 1986; Pappas et al. 1995; Nadal et al. 2000) and antioestrogen (Sudo et al. 1983; Klinge et al. 1992) binding sites in the plasma membrane has been characterised in several cell preparations. We have demonstrated previously that the activation of Maxi Cl− channels in fibroblasts is independent of the activity of the so-called antioestrogen binding site (Hardy & Valverde, 1994).
The possibility that the activation of the Maxi Cl− channels by antioestrogens is related to the alteration of the phosphorylation-dephosphorylation balance was also investigated. We hypothesised that if phosphorylation of the channel, or a regulatory factor, keeps the channel closed under non-stimulatory conditions, as suggested by Saigusa & Kokubun (1988), the activation of the channel may be linked to a dephosphorylation reaction. The activation of Maxi Cl− currents by antioestrogens appears to be related to the activity of S/T PP, since general inhibitors of S/T PP (20 mm NaF), but not tyrosine phosphatases (2 mm ZnCl2 or 100 μm vanadate), prevented it. Of the four major S/T PPs, PP1, PP2A, PP2B and PP2C (reviewed by Cohen, 1991), the involvement of the divalent cation-dependent phosphatases, 2B and 2C, was excluded because current activation was preserved in the absence of intracellular Ca2+ or Mg2+ or in the presence of cyclosporin A, an inhibitor of PP2B (Hunter, 1995). The phosphatases PP1 and PP2A are both divalent cation independent and can be distinguished by their respective sensitivity to okadaic acid. PP2A is inhibited by 1 nm okadaic acid (IC50 < 1 nm; Cohen, 1991, 1997), while PP1 is unaffected by 1 nm okadaic acid, but is completely inhibited at 1 μm (IC50≈ 60–600 nm; Cohen, 1991, 1997). The fact that 1 nm okadaic acid totally inhibited the activation of Maxi Cl− channels by antioestrogens is consistent with the participation of a PP2A-like phosphatase. In accordance with this hypothesis, we detected phosphatase activity in C1300 cells that was inhibited by 1 nm okadaic acid. The okadaic acid-sensitive phosphatase activity represented around 30 % of the total cation-independent phosphatase activity. However, we did not observe changes in okadaic acid-sensitive or -insensitive phosphatase activity with either oestrogen or antioestrogens. The okadaic acid-sensitive PP2A-like phosphatases are heterotrimeric enzymes that are composed of a catalytic subunit (C) and two regulatory subunits (A and B: Mumby & Walter, 1993; Mayer-Jaekel & Hemmings, 1994; Price & Mumby, 2000). The characteristics of PP2A phosphatase (e.g. localisation, substrate specificity and kinetic properties) can be modulated by changes in the regulatory subunits, without a concomitant change in the total PP2A activity (Goldberg, 1999; Price & Mumby, 2000), that may represent a more subtle modulation of the enzyme than its massive activation. We speculate that the activation of Maxi Cl− channels, which depends upon an okadaic acid-sensitive step, represents a change in the substrate specificity by PP2A-like phosphatases in response to antioestrogens, enabling them to dephosphorylate Maxi Cl− channels or as yet unknown regulatory subunits of the channel.
Many of the rapid, non-genomic effects of oestrogens, including the modulation of ion channels, have been associated with increases in protein kinase activity via the activation of a membrane G-protein-coupled receptor (Minami et al. 1990; Lagrange et al. 1997; Gu & Moss, 1998; Kelly & Wagner, 1999). The effect of oestradiol on the prevention of Maxi Cl− activation by antioestrogens also appears to be linked to protein phosphorylation events. Several pieces of evidence support this assumption. First, the oestradiol effect can be mimicked by cAMP. Second, the effect of both cAMP and oestradiol can be prevented in the presence of staurosporine, a broad-spectrum S/T protein kinase inhibitor. Third, the Maxi Cl− channel activation effected by toremifene can be greatly reduced by including 100 μm GDPβS, a non-hydrolysable GDP analogue, in the pipette solution, and slightly increased by including 100 μm GppNHp (guanosine 5′-(β-imido) triphosphate), a non-hydrolysable analogue of GTP (Diaz et al. 1999). Finally, the inhibitory effect of oestradiol can be prevented by loading cells with Mg2+-free (50–100 nm) solutions, a situation in which transfer of the γ phosphate from ATP to the receptor protein by the kinase(s) is impaired. However, since no specific kinase inhibitors have been used we have no indication as to whether oestrogens use more than one kinase pathway to prevent channel activation, or whether they induce phosphorylation of the channel protein or other elements involved in the activation of Maxi Cl− channels.
The facts that antioestrogen-induced activation of Maxi Cl− channels depends upon the presence of intracellular GTP and is modulated by GDPβS and GppNHp suggest the participation of a G-protein in the signalling cascade. However, channel activation by antioestrogens is retained in Mg2+-free intracellular solutions, despite the fact that G-protein activation requires Mg2+. The different Mg2+ requirements of these two processes (G-protein activation and protein phosphorylation) might explain this apparent contradiction. While the concentration of Mg2+ required for the activation of many G-proteins (GTP hydrolysis and GTP binding in the presence of a ligand) is in the range of 10−9-10−8m (Brandt & Ross, 1986; Gilman, 1987; Higashijima et al. 1987), the [Mg2+]free required by different protein kinases is typically > 10−5m (Pickett-Gies & Walsh, 1985; Sun & Budde, 1997; Vinals et al. 1997; Rodriguez-Zavala & Moreno-Sanchez, 1998; Harvengt et al. 2000; Srivenugopal et al. 2000). Therefore, it is plausible that the actual intracellular [Mg2+]free (∼ 10−7m) present in the Mg2+-free solution is sufficient to activate G proteins but insufficient to support a phosphorylation process.
In summary, we propose that under non-stimulatory conditions the Maxi Cl− channels are kept closed by phosphorylation of the channel protein or a regulator of its activity. Upon the addition of triphenylethylene antioestrogens, a PP2A-like phosphatase is activated, probably via a G-protein-coupled receptor, leading to the dephosphorylation of the channel (or a regulatory subunit) and channel opening. On the other hand, oestrogen prevents channel activation by promoting phosphorylation of the channel or a key element in the channel activation pathway.