Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Apoptosis is a form of programmed cell death that occurs under physiological conditions as a mechanism of cell/tissue homeostasis (Kerr et al., 1972; Raff et al., 1993) and under certain pathological conditions (Thompson, 1995; Choi, 1996). A reduction in cell volume and activation of caspases are fundamental features of apoptosis (Kerr et al., 1972; Armstrong et al., 1997). Recent work from several groups suggests that changes in ionic content, primarily K⁺, play a pivotal role in the progression of apoptosis (Bortner and Cidlowski, 1999; Dallaporta et al., 1999; Orlov et al., 1999). In immune and neuronal cells, apoptotic insults trigger considerable intracellular K⁺ loss (Bortner et al., 1997; Yu et al., 1997, 1999a), which may consequently cause caspase-3 activation and apoptosis (Hughes et al., 1997; Yu et al., 1999b). The excessive K⁺ efflux can be mediated by overactivation of voltage-gated K⁺ channels. The delayed rectifier K⁺ channel is up-regulated during certain stages of several apoptotic insults in cortical neurons (Yu et al., 1997, 1998, 1999b), myeloblastic leukemia cells (Wang et al., 1999), and cholinergic septal cells (Colom et al., 1998). Tetraethylammonium (TEA) and other K⁺ channel blockers attenuate apoptotic cell death in cultured cortical neurons (Yu et al., 1997, 1999b; Yu and Choi, 1999), cholinergic septal cells (Colom et al., 1998), liver cells (Gantner et al., 1995), and in rat's cerebral cortex of transient focal ischemia (Choi et al., 1998).

TEA blocks K⁺ channels and apoptosis at millimolar concentrations (Yu et al., 1997). Dallaporta et al. (1999) recently showed that the TEA analog TPeA, at micromolar concentrations, blocked all features of apoptosis in thymocyte induced by dexamethasone, etoposide, -irradiation, or ceramide. TPeA, like TEA, inhibits several voltage-gated K⁺ channels from inside and outside of the membrane (Kirsch et al., 1991; Im and Quandt, 1992; Carl et al., 1993), including Ca²⁺-activated K⁺ channels (Langton et al., 1991; Carl et al., 1993), the inward-rectifier K⁺ channel (Spassova and Lu, 1993), and the ATP sensitive K⁺ channels (Davies et al., 1989). TPeA blocks Kv3.1 and Kv2.1 channels in a fast, reversible, and time-dependent manner. Unlike the voltage-dependent effects of TPeA and TEA from the cytoplasmic side of the membrane, the external binding of TPeA appears to be voltage independent (Carl et al., 1993; Jarolimek et al., 1995). TPeA may possess nonspecific effects on other ion channels. Intracellular-applied TPeA was shown to inhibit human cardiac Na⁺ channels expressed in a mammalian cell line (O'Leary and Horn, 1994). Intracellular-applied TPeA and other quaternary ammonium ions can block chloride channels in rat cortical neurons (Sanchez and Blatz, 1995). TPeA at high concentrations (>10-20 µM) reduced the contractions in smooth muscle, whereas TEA had no such effect; it was assumed that TPeA might have an inhibitory effect on nifedipine-sensitive Ca²⁺ channels (Kwok et al., 1998).

We tested the antiapoptotic actions of TPeA and another long-chain quaternary ammonium ion tetrahexylammonium (THA) in cultured cortical neurons. To understand the mechanism of their antiapoptotic actions, we studied effects of TPeA and THA on the membrane currents carried by Ca²⁺, Na⁺, and K⁺ that are believed to be important in neuronal cell death. Our data suggest that TPeA and THA are neuroprotective against staurosporine-induced apoptosis; as potent K⁺ channel blockers, they also show strong inhibitory effects on several ion channels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cortical Cultures. Mixed cortical cultures, containing neurons and a confluent glia bed, were prepared as described previously (Rose et al., 1993). Mice of 15 to 17 days gestation were anesthetized with halothane. Dissociated cortical cells were plated onto a previously established glial monolayer at a density of 3.5 to 4.0 hemispheres/10 ml in 35-mm culture dishes or 24-well plates (Falcon, Primaria, Lincoln Park, NJ), in Eagle's minimal essential medium (MEM, Earle's salts) supplemented with 20 mM glucose (final concentration = 25 mM), 5% fetal bovine serum, and 5% horse serum (HS). Medium was changed after 1 week to MEM containing 25 mM glucose and 10% HS, as well as 10 µM cytosine arabinoside to inhibit cell division. Subsequently, cultures were fed once weekly with MEM supplemented with 20 mM glucose. Cultures were kept in a 37°C, humidified incubator in a 5% CO₂ atmosphere. All experiments were performed between 10 and 15 days in vitro. Glial cultures were prepared from dissociated neocortices of postnatal day 1 to 3 mice. Cells were plated at a density of 0.65 hemisphere/10 ml, in Eagle's MEM containing 25 mM glucose, 10% fetal bovine serum, 10% HS, and 10 ng/ml epidermal growth factor; confluent glial bed was formed in 1 to 2 weeks. Neuronal identity has been previously confirmed by Nissl staining and electrophysiological characterization, whereas the glial bed is immunoreactive for glial fibrillary acidic protein (Choi et al., 1987; Rose et al., 1993).

Electrophysiology. Whole-cell voltage clamp was performed on cortical neuronal cultures in 35-mm dishes on the stage of an inverted microscope (Nikon, Tokyo, Japan) using an EPC-7 amplifier (List Electronics, Darmstadt, Germany); patch electrodes had tip resistance of 8 to 14 M (fire-polished). Current was digitally sampled at 100 µs (10 kHz). The current signals were filtered by a 3-kHz, 3-pole Bessel filter. Current and voltage traces were displayed and stored on a Macintosh computer (Quadra 950; Apple Computer Corp., Cupertino, CA) using the data acquisition/analysis program package PULSE (HEKA Electronics, Lambrecht/Pfalz, Germany).

Neuronal Cell Death. Neuronal cell death was assessed in 24-well plates by cell counts after staining with 0.4% trypan blue dye, and by measuring lactate dehydrogenase released into the bathing medium (Koh and Choi, 1987).

Caspase Activity Assay. Caspase activity was measured as described previously by Armstrong et al. (1997). Briefly, cultures were washed three times with phosphate-buffered saline and lysed in 80 µl of buffer A (10 mM HEPES, 42 mM KCl, 5 mM MgCl₂, 1 mM dithiothreitol, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, pH 7.4). Lysate (10 µl) was combined in a 96-well plate with 90 µl of buffer B (10 mM HEPES, 42 mM KCl, 5 mM MgCl₂, 1 mM dithiothreitol, 1% Triton X-100, 10% sucrose, pH 7.4) containing fluorometric substrate (final concentration 30 µM) and incubated for 45 min at room temperature in the dark. Formation of fluorogenic product was determined in a Cytofluor fluorometric plate reader by measuring emission at 460 nm with 360-nm excitation. Caspase-3 like activity was defined as N-acetyl-Asp-Glu-Val-Asp-7-amino-4-methylcoumarin cleavage (Thornberry et al., 1997).

Cell Volume Assay. For cell volume assay, the cell maximum cross-sectional area was measured using the MetaMorph imaging system (Universal Imaging Corporation, West Chester, PA) based on the assumption that cell soma swells and shrinks in a symmetrical manner as if it were a sphere. This assumption was validated in cortical cultures by measuring cell volume changes directly using optical sectioning techniques; and there was no difference between cell volume changes measured by optical sectioning and those calculated from cross-sectional area (Churchwell et al., 1996).

Chemicals. The caspase inhibitor Z-Val-Ala-Asp(OMe)-fluoromethyl ketone (Z-VAD) was purchased from Enzyme Systems Products (Dublin, CA). TPeA, THA, and other chemicals were purchased from Sigma-Aldrich (St. Louis, MO).

Statistics. Changes were identified as significant if P value in Student's t test was less than .05; multiple comparisons were done using one-way ANOVA test. Mean values were reported together with the S.E.M. unless otherwise specified.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

TPeA and THA Attenuated Staurosporine-Induced Neuronal Apoptosis. Staurosporine, the nonselective protein kinase inhibitor, is a typical inducer of apoptosis (Bertrand et al., 1994; Koh et al., 1995); 0.2 µM staurosporine added into the culture medium induced about 20% cell body shrinkage and 50% of neuronal death in cortical cultures in 24 h (Fig. 1, A and B). Staurosporine stimulated caspase-3 like protease activation (Fig. 1D); consistently, the broad-spectrum caspase inhibitor Z-VAD (100 µM) blocked 87 ± 3% of the death (Fig. 1B). These caspase-mediated events confirmed the apoptotic nature of staurosporine toxicity. TPeA (0.1-1.0 µM) or THA (1.0 µM) coapplied with staurosporine blocked cell shrinkage, attenuated caspase activation, and reduced neuronal death (Fig. 1, A-D). High concentrations of TPeA alone (10 µM), however, showed toxic effects on cortical neurons (Fig. 1B). The protective effect by TPeA was persistent; significant attenuation of neuronal death remained 72 h after coapplication of 0.2 µM staurosporine and 1.0 µM TPeA (57 ± 3% protection; n = 12 cultures, P < .05 compared with staurosporine alone). The L-type Ca²⁺ channel antagonist nifedipine (2 µM; n = 4) and Na⁺ channel blocker tetrodotoxin (TTX) (1 µM; n = 4) had no significant protection against staurosporine-induced neuronal death (data not shown).

View larger version (40K): [in this window] [in a new window]   Fig. 1.   TPeA and THA were protective against staurosporine-induced neuronal apoptosis. A, phase-contract photos of cortical cultures taken 24 h after sham wash, 0.2 µM staurosporine, or 0.2 µM staurosporine + 1 µM TPeA, respectively. Staurosporine exposure caused cell body shrinkage and cell death in 24 h; 1 µM TPeA coapplied with staurosporine prevented the apoptotic cell body shrinkage and increased cell viability. Scale bar, 40 µm. B, staurosporine (0.1 or 0.2 µM) exposure of 24 h induced significant neuronal death in cortical cultures measured by lactate dehydrogenase release. Cell viability is shown as the percentage of the total cell death induced by NMDA (500 µM) after 24 h. The caspase inhibitor Z-VAD (100 µM; n = 12) almost fully blocked the staurosporine-induced neuronal death. TPeA (n = 8) attenuated the cell death in a concentration-dependent manner; higher concentration of TPeA (10 µM) alone, however, was toxic to cells by itself. THA (1.0 µM; n = 8) also reduced staurosporine-induced cell death. *, significant difference from staurosporine alone (P < .05). C, staurosporine-induced cell body shrinkage in 24 h was blocked by coapplied 1 µM TPeA. The relative cell volume was calculated from cross-sectional area as described by Churchwell et al. (1996). *, significant difference from sham control (P < .05). D, expose to staurosporine for 20 h induced a marked enhancement of caspase-3 like activity; TPeA (1 µM; n = 3) and THA (1 µM; n = 3), each coapplied with staurosporine, largely suppressed the caspase activation. For sham control, n =6, and for staurosporine alone, n = 4. *, significant difference from staurosporine group (P < .05).

TPeA and THA Blocked Delayed Rectifier K⁺ Current in Use-Dependent and -Independent Manners. Apoptotic cell body shrinkage is believed mainly to be due to excessive K⁺ efflux followed by water loss (Ojcius et al., 1991; Deckers et al., 1993; Duke et al., 1994; Beauvais et al., 1995; Bortner et al., 1997). The effects of TPeA and THA on staurosporine-induced cell body shrinkage suggested a blockage of K⁺ channel-mediated excessive K⁺ efflux. To characterize the K⁺ channel blocking activities, we studied effects of TPeA and THA on the major K⁺ currents I_K and I_A in cortical neurons. Bath-applied TPeA and THA blocked I_K at low micromolar concentrations. Measured by a voltage step of 300 ms from 70 to +40 mV applied every minute, TPeA and THA in 15 min suppressed I_K steady-state current at IC₅₀ of 2.7 ± 0.2 and 1.9 ± 0.2 µM, respectively (Fig. 2, A-C). The inhibitory effect of TPeA was partially reversible upon washing (Fig. 2A). The effect of TPeA and THA was not dependent on the membrane potential; I_K was blocked at positive and negative potentials after the membrane potential was transiently jumped to these levels (Fig. 2D). Alternatively, when the membrane potential was persistently held at 70 or 20 mV, the I_K current activated by depolarizing pulses to +40 mV was blocked by 76 ± 7% (n = 5) and 67 ± 3% (n = 8), respectively (P = .19).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Suppression of IK currents by TPeA and THA. The outward delayed rectifier IK current was suppressed by TPeA and THA in concentration-dependent manners. A, IK was gradually blocked by 10 µM TPeA; TPeA preferably blocked the steady-state current, suggesting a use-dependent blocking mechanism. The TPeA effect was partially reversed after 10 min of washing in control solution. IK was activated by voltage steps from 70 to +40 mV every minute (same in B and C). B, concentration-dependent block of IK by 0.1 µM (; n = 5), 0.5 µM (1167; n = 5), 1.0 µM (; n = 8), 10 µM (; n = 8), and 20 µM (; n = 9) TPeA during 15-min exposures. IK subjected to sham wash () did not show any reduction. Response curves were constructed by nonlinear exponential curve fitting (Sigmaplot; Jandel Scientific, San Rafael, CA; same as in C for THA). C, concentration-dependent block of IK by THA at 0.1 µM (), 1.0 µM (), 5.0 µM (), and 10 () µM. Each point was compared with sham controls () (*P < .05 by one-way ANOVA). Of note that the inhibitory effect of THA developed faster than that of TPeA especially at high concentrations (see text). D, I-V curves before (control; n = 5) and after 15 min in 20 µM TPeA (n = 5). IK was blocked at negative and positive potentials (P < .05 for IK in TPeA at 40 mV and more positive potentials; Student's t test). The membrane holding potential was 70 mV and one after the other it was transiently shifted to various potential levels as shown in the figure.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Use-dependent block of IK by TPeA but not by THA. A, block of IK was accelerated by increasing the frequency of IK activation from 0 Hz to once a minute ( Hz; n = 5) and to every 10 s ( Hz; n = 7) (voltage step = 70 to +40 mV for 300 ms). The effect of THA, on the contrary, was not dependent on the activation rate; IK was blocked by more than 90% after 5 min of exposure without IK activation (n = 3). In sham control experiment, -Hz activation of IK during the first 5 min showed no alteration on the amplitude of IK (n = 5). The symbols in the figure show only selective time points, they do not necessarily correspond to the number of IK activation. B, IK was drastically blocked by 10 µM THA even if the channels were not preactivated, THA, however, preferably blocked the steady-state current, suggesting that opening of the channel facilitated the THA binding to a site accessible after channel opening. Similar results were obtained with or without the A-type K+ channel blocker 5 mM 4-aminopyridine, suggesting that the initial peak current in the presence of THA was not IA current.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Inhibitory effects of TPeA and THA on IA. A, IA, activated by voltage steps from 100 to 20 mV, was suppressed by 20 µM TPeA. B, concentration-dependent block of IA by TPeA (10-15-min exposure; n = 5, 4, 7, 8, and 5 for 0.1, 0.5, 1.0, 10, and 20 µM, respectively). C, concentration-dependent block of IA by THA (10-15-min exposure; n = 5, 6, 4, and 5 for 0.1, 1.0, 5.0, and 10 µM, respectively). *, significant difference from sham control (P < .05 by one-way ANOVA).

Effects of TPeA and THA on Ca²⁺ and Na⁺ Currents. TPeA and THA were potent antagonists at HVA Ca²⁺ currents; HVA currents were suppressed by 1 to 10 µM TPeA or THA in concentration-dependent manners (Fig. 5). Up to 20 µM TPeA did not affect the fast inward Na⁺ current; however, 10 µM THA drastically blocked the Na⁺ current (Fig. 6).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   TPeA and THA suppressed the HVA Ca2+ currents. A, HVA Ca2+ currents were activated by voltage steps from 70 to 0 mV. The current was blocked by 20 µM TPeA in 15 min. B, TPeA blocked HVA currents at 1.0 µM (; n = 4), 10 µM (; n = 7), and 20 µM (; n = 8) µM. C, THA suppressed HVA Ca2+ currents at 1.0 µM (; n = 3), and 10 µM (; n = 5). *, significant different from sham controls () (P < .05 by one-way ANOVA).

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effects of TPeA and THA the fast inactivating Na+ current. A, TPeA showed no significant effect on the inward Na+ current activated by voltage steps from 70 to 0 mV. B, during 15 min of recording, no difference was seen between sham control Na+ current (n = 17) and the current exposing to 20 µM TPeA (n = 5). C, compared with the sham wash control (; n = 11), the Na+ current was not affected by 1 µM THA (; n = 3), however, 10 µM THA markedly suppressed the Na+ current (; n = 5). *, significant difference from sham control (P < .05 by one-way ANOVA).

Effects of TPeA and THA on the NMDA Subtype of Glutamate Receptor Currents. TEA and its derivatives may have inhibitory effects on NMDA receptor channels (Wright et al., 1991). Extracellularly applied 10 µM (n = 3) or 20 µM (n = 5) TPeA did not show significant effect on NMDA currents recorded at 70 or +40 mV (Fig. 7, A and B). THA, on the other hand, suppressed NMDA current at 10 µM (n = 3) (Fig. 7B), so THA, although a potent K⁺ channel blocker, is less selective than TPeA.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Effects of TPeA and THA on NMDA receptor currents. A, NMDA receptor current triggered by 200 µM NMDA + 10 µM glycine was recorded at 70 mV (inward currents) and +40 mV (outward current), respectively. After 10 min in 20 µM TPeA, there was no change in the outward current; the inward current showed a trend of reduction but it was not significant compared with sham wash control (B). B, TPeA (20 µM; n = 5) had not significant effect on NMDA current recorded at 70 mV; THA of 10 µM (n = 3), however, inhibited the current. *, significant difference from sham control (P < .05 by one-way ANOVA).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study shows that, as K⁺ channel blockers, TPeA and THA are 1000-fold stronger than TEA in central neurons. Consistently, TPeA and THA attenuate neuronal apoptosis at 0.1 to 1.0 µM, concentrations that inhibit I_K currents and are anticipated to prevent the extra K⁺ efflux triggered by an apoptotic insult (Yu et al., 1997). In addition, TPeA and THA are potent blockers at I_A, HVA Ca²⁺, and Na⁺ currents at similar or higher concentrations; on the other hand, the ligand-gated NMDA receptor channel was only affected by THA.

Although K⁺ channel blockers such as TEA attenuate apoptotic cell death, the mechanism of action is under debate. It was proposed that the protective effect by K⁺ channel blockers is due to an increase in [Ca²⁺]_i as a result of activation of voltage-activated Ca²⁺ channels (Colom et al., 1998). A contribution from a Ca²⁺ channel-mediated [Ca²⁺]_i increase, however, was excluded in the protective effects of TEA and high K⁺ medium in cortical neurons, based on the fact that complete blockages of HVA Ca²⁺ currents and [Ca²⁺]_i increases did not eliminate the TEA- or high K⁺-induced protection (Yu et al., 1997, 1999b). In the present study, TPeA and THA, as potent K⁺ and Ca²⁺ channel blockers, attenuated staurosporine-induced caspase activation and neuronal apoptosis. These results further support the idea that an increase in [Ca²⁺]_i may not contribute to the protective effect of K⁺ channel blockers in cortical neurons. In addition, our data are consistent with a recent report that TPeA does not have an inhibitory effect on NMDA receptors (Sobolevsky et al., 1999). Although TPeA and THA block the fast inward Na⁺ current, it may not be the mechanism of protection because TTX showed no protection against apoptotic death; instead, TTX slightly increased the staurosporine-induced death (20% increase; n = 4, P = .052), perhaps due to consequently diminished activity of Na⁺/K⁺-ATPase and reduced K⁺ uptake.

As derivatives of TEA, TPeA and THA displayed distinctive mechanisms of block on I_K channels. The block of I_K by TPeA was strongly use dependent, suggesting its binding site was only accessible when the channels were in open state. Effect of TPeA on I_K was time dependent, and this may partly reflect a possible act on the putative internal quaternary ammonium site because of its higher lipid solubility than TEA (Snyders and Yeola, 1995). In contrast to TPeA, block of I_K by THA did not require preopening of I_K channels, probably suggesting a different biding site for THA located outside of the pore region. Activation of I_K channels, however, did accelerate the blocking effect of THA, implying that THA might also act on the TPeA site or a second binding site for THA. Bath-applied TPeA and THA blocked I_K in a voltage-independent manner, suggesting that the extracellular binding sites for TPeA and THA were conceivably not deeply inside of the channel pore region, consistent with the larger sizes of these compounds.

This study demonstrates that although TPeA and THA are potent K⁺ channel blockers, their ion channel selectivity is limited and the intricate blocking of several channels may explain the toxic effects at higher concentrations. More selective channel blockers specifically targeting at I_K channels will be needed as useful tools for blocking proapoptotic excessive K⁺ efflux and apoptotic death.