Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The structural similarity of hexachlorocyclohexane (HCH) isomers, especially the  isomer, to the cytosolic second messenger inositol-1,4,5-trisphosphate (IP₃) led Pessah et al. (1992) to investigate the effects of HCH isomers on intracellular Ca²⁺ transport and signaling. Results from studies with isolated sarcoplasmic reticulum (SR) membranes from cardiac and skeletal muscle and brain endoplasmic reticulum (ER) revealed that -HCH showed pronounced selectivity over -, -, and -HCH toward modulation of ryanodine-sensitive Ca²⁺ channels. In rat atrial strips, -HCH-induced release of Ca²⁺ from SR was correlated with pronounced positive inotropy and contracture (Pessah et al., 1992). Subsequent studies with rat basophilic leukemia (RBL) cells indicated that -HCH mobilized Ca²⁺ from a thapsigargin-sensitive ER store and concomitantly inhibited depletion-activated Ca²⁺ entry (Mohr et al., 1995). Surprisingly, -HCH-stimulated Ca²⁺ efflux from ER appeared to proceed via a mechanism independent of the IP₃ receptor in the RBL cell. In contrast, Criswell et al. (1994), using myometrial smooth muscle cells, found that -HCH increased intracellular calcium through modulation of IP₃-sensitive, but not ryanodine-sensitive, stores. More recently, -HCH has been shown to stereoselectively mobilize Ca²⁺ from intracellular stores in cultured neural cells that appears to be mediated, at least in part, by interaction with ryanodine receptors (Rosa et al., 1997a,b). The relationship between the neurotoxic effects of - and -HCH on cultured cerebellar granule cells (as measured by lactate dehydrogenase leakage, trypan blue staining, and propidium iodide uptake) and their ability to change intracellular Ca²⁺ concentrations ([Ca²⁺]_i) exhibited complex pharmacology. -HCH was significantly more potent than -HCH at increasing [Ca²⁺]_i and inducing cytotoxicity at all concentrations and times tested (0-200 µM and 0-18 h, respectively). The results of Rosa et al. (1996, 1997b) with cerebellar granule cells indicated that the increase in [Ca²⁺]_i that occurred within 2 to 4 min of exposure to either HCH isomer was not dependent on Ca²⁺ release from ER stores via an IP₃-sensitive pathway. However, differences in Ca²⁺ mobilization induced by the addition of the RyR-specific Ca²⁺ channel blocker neomycin or by the Na⁺/Ca²⁺ exchange blocker amiloride revealed that the two HCH isomers increased [Ca²⁺]_i and induced cytotoxicity by different mechanisms.

To clarify the molecular mechanisms underlying HCH toxicity, we undertook studies to characterize the actions of - and -HCH on channel gating kinetics of RyR2 isolated from rat cardiac muscle and its relationship to altered cardiomyocyte function (see companion report). In the course of the study, an interesting new observation was made regarding the actions of -HCH on the ion permeability of biological membranes. The present study recognizes and characterizes a Ca²⁺-dependent increase in membrane permeability that is especially selective toward K⁺. The present results, taken together with the direct actions of -HCH on RyR function, provide insight into the complex pharmacology that has been seen with HCH isomers.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Preparation of SR Membrane Vesicles. Cardiac SR vesicles were isolated from Sprague-Dawley rats (Simonsen, Gilroy, CA) according to the method of Feher (Feher and Davis, 1991) as described in the companion paper (Buck et al., 1999).

Current Measurements Across Bilayer Lipid Membranes. A planar lipid bilayer membrane (BLM) was formed across a 200-µm hole in the side of a polystyrene cup using a 5:3:2 mixture of phosphatidylethanolamine (PE)/phosphatidylserine (PS)/phosphatidylcholine (PC) suspended (50 mg/ml) in decane. The BLM separated two chambers of 0.7 ml each. For records showing single-channel data, rat cardiac SR membranes enriched in RyR2 were reconstituted into the BLM by adding 5 µg of total protein to the cis chamber containing 500 mM CsCl, 200 µM CaCl₂, and 20 mM HEPES, pH 7.2, whereas the trans chamber contained 100 mM CsCl and 20 mM HEPES, pH 7.2. After fusion of a vesicle (as revealed by a steplike increase in membrane current or by the appearance of a rapidly gating channel), an excess of EGTA, pH 7.0, was added to halt the reaction, and the cis chamber was perfused with an identical buffer with no added Ca²⁺ or EGTA. Single-channel activity was measured with respect to the trans (ground) side in the presence or absence of HCH. For experiments designed to directly measure HCH-induced permeability of phospholipid membranes, lipid bilayers were formed in the presence of the specified concentrations of either CsCl, KCl, CaCl₂, or NaCl. HCH and/or Ca²⁺ was added as described, and membrane currents were monitored as a function of transmembrane holding potential.

Materials. -HCH and -HCH were obtained from Sigma Chemical Co. (St. Louis, MO). All other chemicals were of the highest quality commercially available.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

-HCH Alters Membrane Permeability to Cs⁺ in a Ca²⁺-Dependent Manner. Analysis of RyR2 single-channel gating kinetics in the presence of -HCH also revealed that this isomer produced a nonfluctuating steady-state membrane current that remained even after all RyR2 channels had been pharmacologically blocked. Figure 1A shows a representative trace that reveals the nature of this additional current in the presence of 50 µM -HCH cis, with a 500:100 mM CsCl gradient (cis/trans) and in the presence of 25 µM Ca²⁺. In the experiment shown, the bilayer membrane contains a single cardiac Ca²⁺ release channel. As expected, a step change in membrane potential (from 0 to +20 mV) produced a large amplitude capacitance current that reflects the charging of the bilayer membrane. The well characterized fluctuations of the RyR2 channel were clearly seen superimposed on the rapidly decaying capacitance current. A new and interesting observation was that the capacitance current failed to return to baseline but instead decayed to a steady-state current that was not seen in the absence of -HCH. By contrast, the addition of 50 µM -HCH under identical experimental conditions did not produce a steady-state current, and the capacitative transient rapidly decayed to the original baseline (Fig. 1B). This effect of 50 µM -HCH on phospholipid membrane permeability was observed in the absence of RyR2 and maintained isomer selectivity because 50 µM aliquots of -, -, or -HCH did not produce any significant steady-state current (Fig. 1C).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Action of HCH isomers on membrane permeability is independent of its action on single-channel function. A, representative trace showing the increase in phospholipid membrane permeability induced by 50 µM -HCH in the presence of a cardiac RyR channel (RyR2). Data were acquired in the presence of a 500:100 mM CsCl gradient (cis/trans) and 25 µM Ca2+. Dashed line shows the -HCH-induced membrane current in the absence of an externally applied transmembrane holding potential. The initial leakage current (at 0 mV holding potential) was nonzero under the conditions described due to the nonsymmetric CsCl gradient. B, representative trace showing that 50 µM -HCH does not induce an increase in permeability of a phospholipid membrane-containing RyR2 channel. Data were acquired in the presence of a 500:100 mM CsCl gradient (cis/trans) and 50 µM Ca2+. Dashed line shows the membrane current observed in both the absence and presence of -HCH. C, representative traces showing the -HCH-induced increase in phospholipidmembrane permeability in the absence of RyR2 channels. Data were acquired in the presence of symmetric 100 mM CsCl, 50 µM concentration of the specified HCH isomer, and 50 µM Ca2+. Capacitative membrane currents observed under the described conditions were induced by a 0- to 40-mV step change in holding potential. The small rapid fluctuations present in C are electrical noise artifacts.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   The magnitude of the -HCH-induced increase in membrane permeability is dependent on the species of monovalent ion present. Membranes were formed in the presence of the specified ion and in the absence of added protein, -HCH or Ca2+. Steady-state membrane currents were then measured at various holding potentials in the presence or absence of -HCH and Ca2+. Membrane currents are shown in the absence of -HCH or Ca2+ (), after the addition of 50 µM -HCH (), and after the addition of 50 µM Ca2+ (). Membranes formed in the presenceof the ions specified and in the absence of -HCH and Ca2+ are not permeant to the ions present. The exposure of membranes to 50 µM -HCH did not significantly increase membrane permeability, whereas a further addition of 50 µM Ca2+ dramatically increased permeability over that of control. A, data collected in the presence of symmetric 500 mM CsCl. B, data collected in the presence of symmetric 100 mM KCl. Note compressed the y-axis scale. C, data collected in the presence of symmetric 100 mM NaCl. Conductance slopes for these experiments are collected in Table 1 and are representative of at least 5 measurements each for A, B, and C.

View this table: [in this window] [in a new window]   TABLE 1 Calculated conductance slopes for Ca2+-activated leakage currents induced by -HCH

Ca²⁺-Dependent Changes in Membrane Permeability Induced by -HCH Show Selectivity for K⁺. We investigated whether physiologically relevant monovalent cations could also act as current carriers in the presence of -HCH and Ca²⁺. Figure 2B () revealed that membranes in the presence of symmetric 100 mM KCl and 50 µM -HCH produced little leak current in the absence of added Ca²⁺. The addition of 50 µM Ca²⁺ to the existing 50 µM -HCH cis induced a profound 8.7-fold increase in conductance to 4.25 pA/mV (Fig. 2B, ; note the compressed y-axis scale). By contrast, the addition of 50 µM Ca²⁺ and -HCH in combination to membranes formed in the presence of 100 mM symmetric NaCl showed significantly lower conductance (0.66 pA/mV) than that seen in the presence of either K⁺ or Cs⁺ ions (Fig. 2C, ). The conductance values obtained in the presence of symmetric Cs⁺, K⁺, and Na⁺ and fixed 50 µM Ca²⁺ and -HCH were not dependent on the concentration of monovalent ion from 50 to 500 mM. These results indicate that -HCH, in the presence of Ca²⁺, induced a monovalent cation selectivity following the order K⁺ Cs⁺ > Na⁺. An important observation is that once K⁺ ionophore activity is established, Ca²⁺ is no longer needed for maintenance of ionophore activity because the addition of EGTA after ionophore activity is established does not remove conductance to monovalent cations (not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   -HCH does not induce an increase in membrane permeability regardless of the species of monovalent ion present. Membranes were formed in the presence of 100 mM concentration of the specified ion and in the absence of added protein, -HCH or Ca2+, and steady-state membrane currents measured at various holding potentials. Membrane currents are shown in the absence of -HCH or Ca2+ (), after the additionof 50 µM -HCH (), and after the addition of 50 µM Ca2+ (). Membranes formed in the presence of the ions specified and in the absence of -HCH and Ca2+ are not permeant to the ions present. In all cases, exposure of the membranes to 50 µM -HCH in both the presence and absence of 50 µM Ca2+ did not significantly increase membrane permeability over that of control. A, data collected in the presence of symmetric 100 mM CsCl. B, data collected in the presence of symmetric 100 mM KCl. C, data collected in the presence of symmetric 100 mM NaCl. Conductance slopes for these experiments are collected in Table 1 and are representative of four, three, and three trials for A, B, and C, respectively.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   HCH-induced membrane permeability does not conduct Ca2+ ions and is specific for monovalent cations. A, leakage currents through membranes formed in the presence of symmetric CaCl2 (1 mM) are monitored as a function of transmembrane holding potential. The addition of 50 µM -HCH () did not significantly increase membrane permeability over that of control (). This experiment was repeated twice with similar results. B, large-amplitude -HCH-induced membrane currents are maintained in the presence of Cs+ and in the absence of Cl ions. Bilayer membranes are formed in the presence of symmetric 100 mM cesium methanesulfonate and in the absence of added Ca2+. -HCH (50 µM) () did not induce an appreciable increase in permeability over control (). The addition of 50 µM Ca(OH)2 dramatically increased membrane permeability (). This experiment was repeated twice using CsMeSO3 and once with Cs gluconate. Conductance slopes for these experiment are given in Table 1.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   -HCH dose dependently increases membrane permeability. Membranes formed in the presence of symmetrical 100 mM CsCl and 50 µM Ca2+ (cis) are probed for -HCH-induced currents at varying -HCH concentrations and transmembrane holding potentials. This experiment was repeated three times with similar results.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Membrane permeability increases induced by -HCH in both CsCl and KCl saturate at high Ca2+ concentrations. A, phospholipid membranes were formed in the presence of symmetric 100 mM CsCl and in the absence of Ca2+ and -HCH. Membranes in the presence () or absence () of 50 µM -HCH alone are not permeable to the ions present, whereas the addition of CaCl2 dose dependently increased membrane permeability. Total Ca2+ present is 50 µM (), 350 µM (), 650 µM (), 1150 µM (+), and 1650 µM (×). Conductance slopes for this experiment are given in Table 1. B, -HCH-induced membrane permeability in the presence of CsCl is saturable and does not depend on holding potential. Phospholipid membranes are formed in the presence of symmetric 100 mM CsCl, and leakage current is monitored as a function of holding potential and Ca2+concentration. The measurements in A and B were repeated twice with similar results. C, -HCH-induced membrane permeability in the presence of KCl is saturable and does not depend on holding potential. Phospholipid membranes are formed in the presence of symmetric 100 mM KCl, and leakage current was monitored as a function of holding potential and Ca2+ concentration. This measurement was repeated twice.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The ability of HCH isomers to disrupt intracellular Ca²⁺ signaling and to induce cytotoxicity has been recognized as a major mechanism in their toxicity. Although -HCH (lindane) has been extensively researched for the mechanisms underlying its neurotoxicity, especially its convulsant activity, the role of -HCH in modulating Ca²⁺ homeostasis in various cells has only recently been addressed. Compelling results have shown that -HCH stereoselectively activates the Ca²⁺ release pathway of SR from skeletal and cardiac muscle and brain ER microsomes (Pessah et al., 1992) and ER from RBL cells (Mohr et al., 1995). These observations have recently been extended by Rosa et al. (1996, 1997b) to include cultured cerebellar granule cells. The results provided in the companion paper (Buck et al., 1999) provide direct evidence that -HCH, but not -HCH, interacts with the ryanodine-sensitive Ca²⁺ release complex in a manner that leads to enhanced channel activation.

A striking new finding reported here is that -HCH induced an increase in the steady-state conductance of phospholipid bilayer membranes possessing or lacking protein components. Although we tested other HCH isoforms (, , and ), the  isoform was the only one that displayed this unique property of inducing a Ca²⁺-dependent membrane current that was highly selective for monovalent cations, especially physiologically relevant K⁺ and Na⁺. This conclusion was supported by the observation that the large membrane current was present in the absence of Cl ions and was not present in solutions lacking a monovalent cation. In addition, the lack of an appreciable membrane current when only CaCl₂ was present clearly indicated that although the presence of Ca²⁺ was required for -HCH to induce a membrane permeability, it was not the current carrier. Although Ca²⁺ was required for inducing membrane current mediated by -HCH, it nevertheless was not required for maintenance of the current.

One plausible explanation for this intriguing observation is that successful incorporation of the -HCH ionophore into membranes requires stabilization by Ca²⁺. Support for this interpretation comes from the observations that 1) at a fixed -HCH concentration, Ca²⁺ activates monovalent current in a saturable manner that can be described by a single K_a value regardless of holding potential, and that 2) once conductance for monovalent cations is established, Ca²⁺ is no longer required for maintaining the current. However, the finding that the K_a value for Ca²⁺ differs from those for Cs⁺ (442 µM) and K⁺ (267 µM) suggests that the species of monovalent ion present influences the efficiency of -HCH-mediated cation translocation.

Although -HCH has no structural similarity to any known ionophore, it nevertheless demonstrates pore-forming ionophore activity under the conditions described here. This is supported by the observation that the membrane current 1) increases linearly with increasing ion concentration, 2) increases linearly with increasing HCH concentration and does not appear to saturate (data not shown), and 3) does not appear to have any substrate dependence because the sequential replacement of each ion used in our experiments did not significantly modify the appearance of -HCH-induced membrane currents. A possible mechanism for this effect of -HCH on membrane permeability could involve ordering of the molecule within the lipid membrane because Verma and Rastogi (1990) showed that -HCH preferentially accumulates within the acyl chains of lipids, influences interchain interactions, and fluidizes lipids. This observation suggests the possibility that membrane lipid composition may influence the action of -HCH. We have performed experiments using mixtures of PE, PS, and/or PC suspended in decane from both natural and synthetic sources with no observable difference in the ionophoric behavior of -HCH. Whether -HCH influences the permeability of native cellular membranes in a manner similar to those described here awaits further studies. However, considering that PE, PS, and PE are the major lipid components of mammalian cell membranes, the present results should be predictive.

These rather unique properties of -HCH on membrane permeability would be expected to have profound actions on cell membrane potential, especially with regard to the contribution of K⁺ and Na⁺ currents. Because -HCH is highly lipophilic, it would be expected to distribute widely into all cellular membrane compartments. However, the physiological Ca²⁺ concentration outside the typical mammalian cell exceeds 1 mM, whereas the typical intracellular milieu is maintained <100 nM at rest; therefore, incorporation of -HCH ionophore activity into the plasma membrane would initially predominate. The biophysical findings reported here suggests that once the -HCH ionophore is incorporated into the plasma membrane, it would initially increase the influx of Na ⁺ into the cell, resulting in depolarization of cells. In support of this prediction, the addition of 10 µM -HCH to differentiated C₂C₁₂ in the presence of the membrane potential-sensitive dye bis-(1,3-dibutylbarbaturic acid)trimethine oxonol indicates it causes cellular depolarization (J. Fessenden and I. N. Pessah, unpublished observation). The demonstrated ability of -HCH to deplete ER stores yet inhibit depletion-activated Ca²⁺ entry in the RBL cell (Mohr et al., 1995) could also be explained by the rather unique ionophore activity of this chlorinated hydrocarbon because cell depolarization is known to inhibit depletion-activated Ca²⁺ entry (Mohr and Fewtrell, 1987; Hide and Beaven, 1991). However, the net contribution of -HCH toward altering membrane potential could be quite complex because it would depend not only on the relative conductances to K⁺ and Na⁺ but also on whether excitable cells were at rest or undergoing evoked or spontaneous electrical activity. It should be noted, however, that the increased permeability to monovalent cations induced by -HCH exposure would not be expected to occur at the SR/ER membranes of cells because these membranes maintain a high K⁺ and Na⁺ permeability in vivo (Garcia and Miller, 1984).

Acute effects of -HCH exposure would include mobilization of Ca²⁺ from ryanodine-sensitive intracellular stores and loss of tight regulation of cellular membrane potential. In this respect, the positive inotropy and contracture elicited by -HCH with atrial strips (Pessah et al., 1992) may be attributed primarily to enhanced RyR2 activity and tissue depolarization, which would be expected to discharge catecholamines. By contrast, reduction in the Ca²⁺ transient and contractility seen with isolated ventricular myocytes may be the result of both depletion of SR Ca²⁺ and altered control of surface membrane potential (Buck et al., 1999).

Recently, Rosa et al. (1996, 1997b) reported a complex pharmacology of HCH isomers on cytotoxicity and intracellular Ca²⁺ regulation in cultured cerebellar granule cells. Their results indicate that although both - and -HCH are cytotoxic at high (>150 µM) concentrations, -HCH induces lactate dehydrogenase leakage (a measure of cytotoxicity) at significantly lower (25 µM) concentrations. The nature of HCH-induced changes in intracellular Ca²⁺ was also shown to be significantly different for the  and  isomers in cerebellar granule cells (Rosa et al., 1996, 1997b), suggesting that different mechanisms underlie changes in cellular Ca²⁺ signaling. Results with cerebellar granule cells indicate that within 2 to 4 min of exposure, the major portion of -HCH-induced changes in [Ca²⁺]_i occur through a Ca²⁺ entry pathway, whereas that of -HCH is due to release of Ca²⁺ from intracellular stores. These findings are in complete agreement with our observation that -HCH directly stimulates the RyR2 channel complex (the major RyR isoform in cerebellar granule cells associated with intracellular Ca²⁺ stores) (Furuichi et al., 1994), whereas -HCH does not.

However, Rosa et al. (1996, 1997b) also showed that -HCH-induced increases in [Ca²⁺]_i were reduced ~50% by the addition of the Na⁺/Ca²⁺ exchange blocker amiloride. This result is curious because -HCH is shown to directly stimulate the Ca²⁺ release mechanism of SR/ER. The presence of amiloride would be expected to amplify the -HCH-induced increase in [Ca²⁺]_i because the Na⁺/Ca²⁺ exchange mechanism that removes Ca²⁺ from the cytosol under these conditions would be inhibited. However, a nonspecific mechanism for this decrease cannot be ruled out because amiloride at the concentrations used by Rosa et al. (1996, 1997b) (1 mM) has been shown to block L-type Ca²⁺ channels of the plasma membrane. This would inhibit Ca²⁺ entry via this pathway (Kleyman and Cragoe, 1990), leading to a net decrease in [Ca²⁺]_i.

Much of the complex pharmacology identified by Rosa and coworkers can be addressed by considering the unique effects of -HCH on RyR function and on the its ability to alter fluxes of monovalent cations across the plasma membrane in a Ca²⁺-dependent manner. For example, Rosa et al. (1996, 1997b) reported that the addition of the RyR channel blocker neomycin reduced both - and -HCH-induced increases in [Ca²⁺]_i seen within 2 to 4 min of exposure. This decrease, however, was more pronounced in the presence of -HCH than in the presence of -HCH (30% versus 20%, respectively). Because our results indicate that -HCH increases [Ca²⁺]_i by stimulation of RyR2, it would be expected to be more sensitive to neomycin blockade than -HCH. The decrease seen in the presence of -HCH is likely due to blockade of the ryanodine-sensitive Ca²⁺-induced Ca²⁺ release pathway initiated by -HCH-induced Ca²⁺ entry.

In addition to measuring short-term changes in Ca²⁺ regulation, Rosa et al. (1996, 1997b) determined cytotoxicity in cerebellar granule cells induced by long-term exposure of - and -HCH using a propidium iodide entry assay. Their results indicate that cytotoxicity induced by -HCH was reduced by the presence of the L-type channel blocker verapamil or the Ca²⁺ chelator EGTA in the extracellular medium, whereas neomycin had no effect. In contrast, -HCH-induced cytotoxicity was not affected by verapamil or EGTA, whereas the presence of neomycin in the bathing medium had a pronounced effect. These results can be interpreted by considering the differing mechanisms by which - and -HCH modulate the movement of ions. Because -HCH induces an increase in [Ca²⁺]_i, and presumably cytotoxicity, primarily via a Ca²⁺ entry pathway, it is likely that removal of extracellular Ca²⁺ or blockade of Ca²⁺ entry through L-type Ca²⁺ channels would result in a reduction in -HCH-induced cytotoxicity. In addition, although neomycin reduced -HCH-induced increases in [Ca²⁺]_i in the short term (2-4 min), it would not be expected to contribute significantly to long-term deregulation of Ca²⁺ homeostasis because -HCH increases [Ca²⁺]_i primarily via Ca²⁺ entry. In contrast, verapamil and EGTA had no effect on -HCH-induced cytotoxicity. This is not unexpected because -HCH acts directly on the intracellular Ca²⁺ release pathway. In addition, the altered membrane potential induced by -HCH exposure would alter the excitability of L-type Ca²⁺ channels and negate a simple additive effect of verapamil.