Abstract
The mechanisms of Pb++ block and unblock of L-type Ca++ channel currents were measured using ventricular myocytes or the cloned channel. The cloned channel was expressed in either Xenopus laevis oocytes or human embryonic kidney cells (HEK 293, stable transfectants). The threshold for Pb++ block was 1 nM, and the apparent IC50value was 152 nM in oocytes and 169 nM in HEK 293 cells. Pb++ block was dependent on the composition of the external recording solution but not dependent on the subunit composition of the channel. Pb++ block was voltage dependent, with little block observed at negative test potentials using low concentrations of Pb++. Strong depolarizations (>+100 mV) reversed Pb++ block, allowing measurement of reblock kinetics. Reblock was fast (τ = 11 msec), as measured during a +20-mV test pulse. Simple washout did not completely reverse Pb++block, especially after exposure to concentrations of >100 nM. Full recovery could only be observed after treatment with heavy metal antidotes such as meso-2,3-dimercaptosuccinic acid, 2,3-dimercapto-1-propanesulfonic acid and EDTA. These results suggest that Pb++ blocks voltage-gated Ca++ channels by two mechanisms and that full reversal of lead block requires chelator treatment.
Lead continues to be an environmental hazard, contaminating food, air and water supplies (Davis et al., 1993). This raises public concern because what were once considered safe levels have been proved to cause neurotoxicity, especially in children (Angle, 1993). The threshold for toxicity is now considered to be when blood levels are > 10 μg/dl (0.5 μM) (Cory-Slechta, 1995; Davis et al., 1993). Lead toxicity can be successfully treated with chelators such as EDTA, DMSA and DMPS (Angle, 1993; Aposhian, 1983); DMSA appears to be the drug of choice due to its effectiveness and lower toxicity (Aposhianet al., 1995; Graziano et al., 1992).
The mechanism or mechanisms by which Pb++ causes neurotoxicity are poorly understood. Pb++ does not participate in redox reactions, but it can form complexes with sulfhydryl, amine and carboxyl groups, thereby allowing for high-affinity binding to proteins (Goering, 1993). Pb++ can bind to the metal binding sites of metallothioneins and to proteins with E-F hand motifs such as calmodulin. In many cases, Pb++ can mimic the action of Ca++ ; Pb++ can activate Ca++ -dependent protein kinase C (Markovac and Goldstein, 1988), troponin C (Chao et al., 1990) and Ca++ -activated K+ channels (Oortgiesen et al., 1993). Pb++ can also act as an antagonist, blocking Ca++ permeation through voltage-gated Ca++ channels. In contrast, Pb++does not block Na+ or K+ channels (Büsselberg et al., 1991; Reuveny and Narahashi, 1991). Because Ca++ is a key second messenger, it seems likely that Pb++ causes neurotoxicity by disrupting Ca++ homeostasis (Simons, 1993).
Lead block of voltage-gated Ca++ channels has been studied in a number of cell types (for a review, see Audesirk, 1993). Voltage-gated Ca++ channels are multisubunit complexes (Perez-Reyes and Schneider, 1994). The alpha-1 subunit contains the pore, the voltage-sensor, and most of the drug binding sites. The genes for six alpha-1 subunits have been cloned. On the basis of sequence identity, these alpha-1 subunits can be divided into two subfamilies: (1) L-type, alpha-1S (skeletal), alpha-1C (cardiac and brain) andalpha-1D (neuroendocrine); and (2) non-L-type,alpha-1A (brain P/Q-type), alpha-1B (brain N-type) and alpha-1E (brain R-type). Purification studies have shown that skeletal and cardiac muscle L-type and neuronal N-type channels also possess common alpha-2-deltasubunits and a specific beta subunit. Four betasubunit genes have been cloned (Perez-Reyes and Schneider, 1994). Coexpression studies have shown that beta subunits can increase the functional expression of alpha-1 and modulate the biophysical properties of the channel (Perez-Reyes and Schneider, 1994). Coexpression with alpha-2-delta also increases functional expression and modifies the pharmacological properties of the cloned channel (Mikami et al., 1989;Shistik et al., 1995; Wei et al., 1995).
Previous studies have established that voltage-gated Ca++channels can be blocked by low (μM) concentrations of Pb++. Washout-resistant block, termed “irreversible inhibition” (Audesirk, 1993), has been noted before but not studied in detail. Because Ca++ channels are highly regulated by protein kinases and Pb++ is capable of activating protein kinase C at very low concentrations (nM), it is possible that the observed block was not due to a direct effect of Pb++ on the channel. Because the cloned L-type Ca++ channel is not regulated by protein kinases (Zong et al., 1995), we were able to study the direct effects of Pb++ on Ca++ channel activity. We describe the toxicological effect of Pb++ on L-type Ca++ channels, pharmacological relief of block and some biophysical properties of Pb++ block and unblock and show that there are two types of block: one is reversed by washing with lead-free solutions, whereas the second type requires treatment with lead antidotes such as DMSA or EDTA.
Materials and Methods
Materials.
High purity lead acetate (99.999%) was purchased from Aldrich Chemical Co. (Milwaukee, WI). It was dissolved in water at a concentration of 10 mM. Serial dilutions were freshly prepared and then diluted 1:100 in the bath solution to give the indicated concentration. In contrast to lead chloride (Matthews et al., 1993), no white precipitate was observed in any of the lead acetate solutions. The concentration of free Pb++ in the bath solution was not determined, but it may either be lower than expected due to binding to ligands in the bath solution or higher due to Pb++ contamination of the reagents (Matthews et al., 1993; Simons, 1993). All other chemicals were purchased from Aldrich or Sigma (St. Louis, MO).
The cDNA encoding the alpha-1C subunit from rabbit heart was a gift from the late Dr. Xiangyang Wei (Wei et al., 1991). The alpha-1C cDNA used for oocyte expression was truncated such that the first 60 amino acids were deleted. A similar truncation has been described to occur naturally in human heartalpha-1C (Schultz et al., 1993). Thealpha-1C cDNA used for HEK 293 cell expression was the full-length clone. Cloning of rat brain beta-2 andbeta-4 subunits has been described previously (Castellanoet al., 1993; Perez-Reyes et al., 1992). The cDNA encoding alpha-2-delta skeletal muscle was a gift from Dr. Tsutomu Tanabe (Yale University). It was subcloned into pGEM-3 (Promega, Madison, WI).
Oocyte expression.
Capped cRNAs were synthesized in vitro using T7 RNA polymerase and the mMESSAGE mMACHINE kit (Ambion, Austin, TX). The concentration of cRNA was measured spectrophotometrically. The production of full-length cRNA transcripts was verified after electrophoretic separation on a denaturing agarose gel.
Oocytes were prepared from the South African clawed frog Xenopus laevis (Nasco, Fort Atkinson, WI) using standard techniques. Briefly, ovarian lobes were removed surgically from the frog. The lobes were torn into small clusters and then transferred into Ca++ -free OR solution composed of 82.5 mM NaCl, 2.5 mM KCl, 1 mM MgCl2 and 5 mM HEPES, pH 7.6. Defolliculation was performed by shaking in Ca++ -free OR solution containing 2 mg/ml collagenase (Type 1A, Sigma). Healthy defolliculated oocytes (stage V–VI) were manually selected and then allowed to recover (>2 hr) in SOS medium (100 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM HEPES, pH 7.6, 2.5 mM pyruvic acid and 50 μg/ml gentamicin). Each oocyte was injected with 50 nl of cRNA (5 ng for alpha-1, 2.5 ng for beta and 2.5 ng foralpha-2) diluted in 0.1 M KCl or water. Injection was performed using a Drummond Nanoject pipette injector (Parkway, PA) attached to a Narashige micromanipulator (Tokyo, Japan) under a dissecting microscope. To express alpha-1 alone, the amount of the cRNA was increased to 15 ng/oocyte. Oocytes were incubated at 19°C with SOS medium in glass petri dishes and were ready for electrophysiological recording after 2 days. L-type Ca++ channel currents were stable for ≤2 weeks.
Electrophysiological analysis of injected oocytes.
Oocytes were voltage-clamped using a two-microelectrode voltage-clamp amplifier (OC-725B, Warner Instrument Corp., Hampden, CT). Pipettes were made from glass capillaries (#6010, A-M Systems, Everett, WA), using a Model P-97 Flaming-Brown pipette puller (Sutter Instrument Co., Novato, CA). Voltage and current electrodes (1.8–2.6-MΩ tip resistance) were filled with 3 M KCl. The oocytes were impaled in SOS solution, and then the bath solution was exchanged with a solution of 10 mM Ba(OH)2, 80 mM NaOH, 1 mM KOH and 5 mM HEPES, adjusted to pH 7.4 with methanesulfonate (Lory et al., 1990). Because methanesulfonic acid may bind Pb++, experiments were repeated using 10 mM BaCl2, 90 mM NaCl, 2 mM KCl and 5 mM HEPES, pH 7.4. Similar results were obtained, and the data were pooled. Data were acquired at 2000 Hz using the pCLAMP system (Digidata 1200 and pCLAMP 6.0, Axon Instruments, Foster City, CA) and filtered at 400 or 1000 Hz (#902 Frequency Devices, Haverhill, MA).
Rundown of the Ca++ channel currents was minimized by reducing the amount of Ba++ influx and by the use of relatively high-resistance electrodes, which reduces dialysis of the oocyte with KCl. Ba++ influx was reduced by keeping the test pulses as short as possible (200 msec), by reducing the external Ba++ concentration from 40 to 10 mM and by reducing the amount of alpha-1C that was injected. Kinetics of reblock were not analyzed in oocytes due to the inherent voltage errors caused by clamping large currents (>2 μA) in a large cell (∼300 nF) with these electrodes. Precautions to reduce contamination by Pb++ between experiments included extensive washing of the chamber with detergent and 0.1% acetic acid, frequent replacement of perfusion tubing (IV solution set, Baxter Healthcare Corp., Deerfield, IL) and moving the agar bridges out of the well containing the cell and placing them downstream in a second well (RC-25 chamber, Warner Instrument Corp, Hamden, CT).
Generation of a stably transfected HEK 293 cell.
The cDNA insert encoding the alpha-1 subunit was subcloned into the expression vector pRC/CMV (InVitrogen, San Diego, CA). This vector contains a neomycin resistance gene, which allows for selection of transfected cells with geneticin (G418; GIBCO, Grand Island, NY). HEK 293 cells (1 × 106 cells in a 100-mm culture dish) were transfected with 10 μg of alpha-1 cDNA using lipofectamine (GIBCO). At 24 hr after transfection, the cells were suspended in Dulbecco’s modified Eagle’s medium supplemented with G418 (0.4 g/L) and fetal bovine serum (10%). Individual colonies were isolated and plated onto 24-well plates. The clones were expanded and then analyzed for alpha-1 protein expression using dihydropyridine binding assays with the ligand (+)-[3H]PN200–110 (Amersham, Arlington Heights, IL). Single-point assays were done as described previously (Perez-Reyes et al., 1992). Cell clone A20, which expressed 100 fmol of binding sites/mg of membrane protein, was selected for further study. Results were also obtained using cell clone LCa10; this cell line is a subclone of HCaα1β2-2 (Perez-Reyes et al., 1994). During passage in culture, HCaα1β2-2 lost expression ofbeta-2. A single cell was recloned, expanded and characterized. Polymerase chain reaction using beta-specific primers confirmed the lack of beta-2 mRNA expression. Results using the two alpha-1-transfected cell lines were identical and have been pooled.
Electrophysiological analysis of HEK 293-transfected cells.
HEK 293 cells were plated onto coverslips and cultured for ≥1 to 5 days before electrophysiological studies. Cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 0.4 g/L G418, 10% fetal bovine serum, 100 units/ml penicillin and 100 μg/ml streptomycin. To minimize dialysis of the cell and Ca++channel rundown, we used the perforated patch technique with amphotericin B (Rae et al., 1991). Amphotericin B was freshly dissolved (30 mg/ml) in dimethylsulfoxide. This stock was diluted to 0.24 mg/ml into a typical internal pipette solution that contained 55 mM CsCl, 75 mM CsSO4, 10 mM MgCl2, 0.1 mM EGTA and 10 mM HEPES, pH adjusted to 7.2 with CsOH (Perez-Reyeset al., 1994). Pipette tips were briefly dipped into this internal solution and then backfilled with the same solution plus amphotericin B. The external Tyrode solution contained 140 mM NaCl, 6 mM KCl, 2 mM CaCl2, 10 mM glucose and 5 mM HEPES, pH 7.4. The recording solution contained 10 mM BaCl2 solution (or 2 mM CaCl2 for ventricular myocytes), 140 mM TEA chloride, 5 mM CsCl, 1 mM MgCl2, 5 mM glucose and 10 mM HEPES, pH adjusted to 7.4 with TEA-OH (Perez-Reyes et al., 1994). Preliminary experiments used the following recording solution: 10 mM BaCl2, 130 mM aspartic acid, 130 mM NMG, 10 mM aminopyridine, 1 mM MgCl2, 10 mM glucose and 10 mM HEPES, pH adjusted to 7.3 with aspartic acid (de Leon et al., 1995).
Whole-cell currents were recorded from perforated patches using an Axopatch 200A amplifier, Digidata 1200 A/D converter and pCLAMP 6.0 software (Axon Instruments, Foster City, CA). Data were digitized at 2 kHz and filtered at 1 kHz or off-line. Capacitative transients were routinely subtracted using a P/4 subpulse routine, except for experiments using a prepulse. Pipettes were made from TW-150–6 capillary tubing (World Precision Instruments, Sarasota, FL). The pipette resistance was typically 1.5 to 2.5 MΩ. After gigaseal formation, perforation of the patch was monitored by following the increase in size of a capacitative transient induced by a 5-mV test pulse. The amplitude of this transient was used to calculate access resistance as previously described (Rae et al., 1991). Within 10 min, the pipette access resistance was usually 5 to 8 MΩ. During this period, the external solution was replaced with 10 mM BaCl2 solution. Cell capacitance was measured by integrating the charging current during a 10-mV hyperpolarizing pulse (holding potential, −80 mV). Data were not included from experiments in which the access resistance was >10 MΩ or when cells had processes or connections to other cells.
Rat ventricular myocytes were prepared as described previously (Lewet al., 1991). Freshly isolated myocytes were plated onto laminin-coated coverslips and maintained in modified essential medium until use. The ruptured patch method was used to record whole-cell currents from ventricular myocytes.
Data analysis.
All experiments began by recording of control currents (∼5 min). Only cells with stable currents were used for analysis. All test compounds were diluted in external solution and then perfused into the bath at a rate of 2 to 4 ml/min. The perfusion rate varied between experiments but was similar between solutions in any single experiment. The bath chamber had a volume of .15 ml. The bath was continuously perfused throughout the experiment. Experiments were conducted at room temperature, 22° to 24°C.
The effect of lead was calculated by averaging the plateau current from three or more traces taken after steady-state block had been achieved. Current amplitudes and exponential fits were calculated using the pCLAMP software program Clampfit (Axon Instruments). The average data were fit using a sigmoidal dose-response equation (bottom fixed at 0) using Prism software (GraphPAD, San Diego, CA). The IC50values were calculated from these fits. Pooled data are expressed as mean ± S.E.M. Statistical significance was evaluated with a paired Student’s t test (SigmaPlot, Jandel Scientific, San Rafael, CA).
Results
Figure 1 shows current traces recorded from cells before and after exposure to 1 and 10 μM Pb++ and then after washout. Inward Ca++ currents in rat ventricular myocytes are blocked by low concentrations of Pb++ (fig.1A). This current is through L-type Ca++ channels (Bean, 1989). The minimum subunit composition of these channels isalpha-1C, alpha-2 and beta-2 (Perez-Reyes and Schneider, 1994). Expression of the clonedalpha-1C subunit alone induces dihydropyridine-sensitive currents that can be measured with Ba++ as the charge carrier (Mikami et al., 1989). Pb++ also blocks inward Ba++ currents through channels composed ofalpha-1C alone, expressed in either X. laevisoocytes (fig. 1B) or HEK 293 cells (fig. 1C). Because it has been previously shown that currents from the cloned channel are not regulated by protein kinases (Zong et al., 1995), and hence more stable, we selected these for further study.
Lead dose response was measured by adding lead acetate to the bath solution and then measuring the block of cloned L-type currents expressed in oocytes (fig. 2A). Preliminary experiments suggested that block of alpha-1C expressed alone in HEK 293 cells was less sensitive to block thanalpha-1C-beta currents in oocytes (fig. 2B). Because beta subunits alter channel gating (Perez-Reyes and Schneider, 1994), we tested the hypothesis that subunit composition may alter Pb++ block. Oocytes were injected with cRNA foralpha-1C, alpha-1C-beta-2,alpha-1C-beta-4 oralpha-1C-alpha-2-delta-beta-2. Despite large differences in the resulting current amplitudes (Perez-Reyes et al., 1992; Wei et al., 1995), Pb++ blocked the currents with equal potency; therefore, pooled data were used for dose-response analysis (fig. 2A; IC50 = 152 ± 87 nM). L-type currents recorded from ventricular myocytes were also blocked to a similar extent.
Figure 2B shows Pb++ block of alpha-1-induced currents in HEK 293 cells. We tested the hypothesis that the bath solution used for HEK 293 cell experiments was chelating Pb++. Lead was a much more potent blocker in 10 mM BaCl2 solutions containing TEA (IC50 = 169 nM) than in solutions containing both NMG and aspartic acid (IC50 = 23 μM). Therefore, low concentrations of Pb++ blocked the cloned L-type channel in both expression systems and L-type channels in situ.
Two protocols were used to measure Pb++ dose responses: (1) where increasing concentrations of Pb++ were tested sequentially and (2) where there was a washout with control solution between Pb++ exposures (fig. 3). Figure 3shows that Pb++ block was fast and quickly reversed by washout, allowing for many concentrations to be tested. It is common for L-type Ca++ channel currents to rundown (McDonaldet al., 1994). In contrast, currents recorded fromalpha-1C-injected oocytes displayed very little rundown (0.23% per minute, n = 22; see figs. 5 and 6). We also noticed that the extent of recovery was dependent on Pb++concentration. Thus, a component of Pb++ block was resistant to washout, or “irreversible” under the time scale of these experiments (Audesirk, 1993). To quantify this component, we compared the currents after washout to the original pre-Pb++ control and then calculated the amount of washout resistant block (fig. 3B). Significant washout-resistant block could be measured after exposure to concentrations as low as 10 nM Pb++ (8.7 ± 2.7, n = 8).
Experiments were performed to test the efficacy of heavy metal poisoning antidotes, such as EDTA or DMSA (Aposhian, 1983; Angle, 1993;Aposhian et al., 1995), to reverse the Pb++block of L-type channels. To mimic the clinical situation, we first induced Pb++ block and then tried to reverse block using a solution containing both antidote and Pb++. Figure4 shows that DMSA, DMPS and EDTA could fully reverse Pb++ block, even when very high (>10 μM) concentrations of Pb++ were used. All three drugs caused a small stimulation relative to control activity (DMSA, 112 ± 2%,n = 14; EDTA, 117% ± 3%, n = 7; and DMSA, 118%, n = 2). This stimulation may be due to the presence of Pb++ in the reagents used (Simons, 1993).
We also tested the effect of the antidotes on the washout-resistant component of Pb++ block. Oocytes were treated with a high concentration (100 μM) of Pb++, washed with control solution and then treated with Pb++ plus EDTA (fig.5). As shown above, washout alone did not completely restore the currents to control levels; however, treatment with Pb++ plus EDTA did. The concentration of Pb++used in these experiments was >600-fold higher than the apparent IC50 value. This raises the possibility that a small contamination could be causing the observed block. For example, Pb++ may have accumulated in the cell, or on the recording chamber, and then slowly released during the control wash. To rule out these possibilities, we conducted experiments using 100 nM Pb++ (fig. 6). Representative current traces are shown in figure 6A. Because DMSA causes a small stimulation of currents, this experiment contains three control washes (representative experiment is shown in fig. 6B). The first control was used to establish the base-line current and to measure the stimulation by DMSA over control. The second control is the washout of Pb++, and the steady-state value was compared with control-1 to determine the washout-resistant inhibition. The third control is the washout after a short treatment (5 min) with 10 μM DMSA. If Pb++accumulation, or contamination, was causing the washout-resistant inhibition, then this third control would have returned to the value measured during control-2 (75 ± 4%, n = 4, fig.6C). This was not the case; control-3 activity returned to 91 ± 3% of control-1. The deviation from 100% is presumably due to channel rundown.
Figure 7 shows average current-voltage relationships measured in oocytes (fig. 7A) and transfected HEK 293 cells (fig. 7B) in either the absence or presence of Pb++. The percent inhibition observed at each test potential was averaged and then plotted (fig. 7, C and D). Very little block by 1 μM Pb++was observed when the test potential was near threshold (−20 mV). Pb++ was a more effective blocker during pulses to higher potentials. Although less pronounced, voltage-dependent block was also observed at higher Pb++ concentrations (10 μM; fig. 7C). Similar voltage-dependent block was observed in HEK 293 cells (fig.7D).
Strong depolarizations have been shown to enhance L-type channel gating (Pietrobon and Hess, 1990) and to reverse Cd++ block of Ca++ currents from frog sympathetic neurons (Thévenod and Jones, 1992). Our hypothesis was that Pb++ was an open channel blocker, similar to Cd++, and that strong depolarizations may remove block. Voltage protocols contained the following three depolarizations: (1) a control test pulse to +20 mV, (2) a depolarizing pulse of varying amplitude and (3) a second test pulse. Between the depolarizations was a short (10 msec) repolarization to −80 mV. Due to the large oocyte capacitance, these experiments were performed using only transfected HEK 293 cells. Figure8A shows that the middle pulse had little (slight inhibition) or no effect on control currents. Figure 8B shows the currents recorded in the presence of Pb++. Larger currents were measured in the second test pulse when the interpulse potential was >50 mV. To quantify this increase, peak currents were measured in the second test pulse and then divided by the current remaining at the end of the first test pulse. Figure 8C plots this postpulse-to-prepulse current ratio as a function of the depolarizing test potential.
The ability of strong depolarizations to remove Pb++ block allows for measurement of Pb++ blocking kinetics during the second test pulse. Figure 8D shows an overlay of currents measured during control (first test pulse, episode 1), in the presence of Pb++ (second test pulse, after +40 mV depolarization, episode 1) and after a strong depolarization in the presence of Pb++ (second test pulse, after +160 mV depolarization, episode 13). Control currents displayed very little decay over this time period. In contrast, currents taken in the presence of Pb++ did decay (currents in first test pulse, fig. 7B). This decay was greater after a strong depolarization. In analogy to Thévenod and Jones, we interpret this decay as due to Pb++ blocking of open channels. Exponential fits to the data indicate that Pb++ reblocks with a τ value of 11 msec. This rate was both voltage (during a 0- mV test pulse τ = 17 msec) and concentration dependent (at a 3-fold lower concentration of Pb++, the reblock τ was 17 msec at +20 mV and 42 msec at 0 mV).
Discussion
Lead is a potent blocker of L-type Ca++ channels, inhibiting both native and cloned channels with an IC50 of 150 nM. Similar total lead concentrations (500 nM) in blood are considered to be the threshold for toxicity (Cory-Slechta, 1995; Daviset al., 1993). It has been noted that the concentration of free Pb++ in plasma is much lower than the total blood concentration (Audesirk, 1993). Similarly, free Pb++concentrations in vitro may be either much lower than expected due to binding to buffer components or higher due to contamination of reagents (Matthews et al., 1993; Simons, 1993). For example, we found that the potency of Pb++ block depended on the composition of the bath solution. Both solutions contained 10 mM BaCl2, but one contained TEA and the other contained NMG and aspartic acid. It is likely that Pb++ was buffered by aspartic acid. Despite these uncertainties in the free Pb++ concentration, our results clearly show that Pb++ has a very high affinity for L-type channels.
Ca++ influx via voltage-gated Ca++channels is a highly regulated event. Cardiac L-type channels are regulated by both cAMP-dependent protein kinase and protein kinase C (McDonald et al., 1994). Pb++ can stimulate protein kinase activity (Markovac and Goldstein, 1988); therefore, direct effects of Pb++ on the channel can be obscured by its effects on kinases. We avoided this problem by using the clonedalpha-1C, which is not regulated by protein kinases as observed in situ (Zong et al., 1995).2
Simple washout of Pb++ did not completely reverse the block of L-type Ca++ channels. During our dose-response analysis, we noticed a difference between protocols in which Pb++ was added sequentially and protocols that included a wash. Washout-resistant block has been reported previously in studies using rat dorsal root ganglion neurons (Büsselberg et al., 1994) and snail neurons (Audesirk, 1993). This lack of recovery was termed “irreversible inhibition” and surprisingly was not dose dependent (Audesirk, 1993). In contrast, we found that washout-resistant block was dose dependent, following a similar dose response as the total block. We show for the first time that complete recovery of currents required treatment with heavy metal antidotes, such as DMSA, DMPS and EDTA. The ability of these antidotes to recover 100% of the original channel activity indicates that this washout-resistant inhibition is not simply due to channel rundown. A clinical implication of these findings is that some Pb++targets will continue to be blocked after removal of Pb++and that full reversal of lead poisoning requires treatment.
Numerous heavy metals, including Pb++, block voltage-gated Ca++ channels (Audesirk, 1993; McDonald et al., 1994). Heavy metal block has been studied because it provides (1) a toxicological description of heavy metal action, (2) a basis for classifying channel subtypes and (3) clues on the structure-function relationships of the channel. In many studies, the blocked currents are subtracted from the control currents to infer block of channel subtypes that differ in their inactivation rates (Audesirk and Audesirk, 1993). Our studies indicate that Pb++ block occurs during the pulse, thereby complicating this type of analysis. Studies with Pb++ have established that it is selective for voltage-gated Ca++ channels, with little effect on Na+ or K+ channels (Büsselberg et al., 1991; Reuveny and Narahashi, 1991). Unlike Ni++, Pb++ does not appear to be very selective among the various voltage-gated Ca++ channels, blocking low- and high-threshold channels with similar potency (Audesirk, 1993).
The mechanism by which Pb++ blocks Ca++channels is poorly understood. Largely based on the observation that increasing extracellular Ca++ decreased Pb++block, it was suggested that Pb++ was an open channel blocker (Büsselberg et al., 1991). Our results indicate that Pb++ block is very similar to the open channel blocker Cd++; block is voltage dependent (Chow, 1991; Swandulla and Armstrong, 1989) and can be relieved by strong depolarizations (Thévenod and Jones, 1992). Single-channel studies have allowed direct measurement of the rate of Cd++block and unblock (Lansman et al., 1986). These studies found that the rate of unblock was voltage dependent, whereas the rate of block was largely voltage independent. Because these changes occurred in the same voltage range as channel activation, we can infer that different states of the channel have different affinity for blocker. By analogy, we predict that the rate of Pb++unblock is voltage dependent. At potentials of <−20 mV, when most channels are in the closed state, we predict that Pb++unblock is faster than block, resulting in little tonic block. At potentials where channel opening is favored (>+20 mV), we predict that Pb++ block is faster than unblock, resulting in a time-dependent block. This would also explain the observed voltage dependence of Pb++ block. Similar voltage-dependent block was observed in studies using neurons from Aplysia and rat dorsal root ganglion (Büsselberg et al., 1991) but not rat hippocampus (Audesirk and Audesirk, 1993). Perhaps the voltage dependence of block was obscured by the fact that neurons contain multiple subtypes of Ca++ channels that differ in their voltage dependence.
Surprisingly, stronger depolarizations (>50 mV) also caused unblock. We took advantage of this unblock phenomenon to measure reblock kinetics. We found that the Pb++ block of open channels was fast (τ = 11 msec) and voltage dependent. The mechanism by which strong depolarizations reduce divalent cation block are not understood. Strong depolarizations have been shown to alter L-type channel gating (Pietrobon and Hess, 1990). Possibly, this depolarization is inducing a channel conformation with lower affinity for Pb++. A second possibility is that reverse current flow through the channel knocks Pb++ off its binding site in the pore. An implication of this result is that facilitation may be due to unblock of channels by contaminating divalent cations. In addition, these results provide another example of how the cloned channel is not regulated in the same manner as in situ channels.
In addition to blocking Ca++ channels, Pb++ and Cd++ may permeate these channels (Chow, 1991; Tomsig and Suszkiw, 1991). Perhaps Pb++ permeates the channel at negative test potentials where the driving force for Ba++ is high (Erev = +60).
Ca++ channels are multisubunit complexes that contain a large alpha-1 subunit that forms the pore and several auxiliary subunits (Perez-Reyes and Schneider, 1994). Molecular cloning has revealed that there is a family of at least six relatedalpha-1 subunits. One region that is particularly well conserved among these subtypes is the pore-lining region (Perez-Reyes and Schneider, 1994). This region contains glutamate residues that are thought to form a ring of negative charge. Mutagenesis studies have shown that mutation of these glutamates alters divalent cation permeation and block (Parent and Gopalakrishnan, 1995; Tang et al., 1993; Yang et al., 1993). Due to sequence conservation between the cloned alpha-1 subunits, we predict that Pb++ block will be very similar between these Ca++ channel subtypes. We suggest that the quickly reversible block represents Pb++ binding to this site. In contrast, washout-resistant block may be due to Pb++binding to a second site from which it only slowly dissociates. Pb++ can be removed from the second site by chelators, indicating that this site is accessible from the extracellular medium. Recent mutagenesis studies support the hypothesis that there are two divalent cation binding sites in the pore (Parent and Gopalakrishnan, 1995).
In addition to heart, L-type Ca++ channels are distributed throughout the brain. In particular, in situ hybridization has shown that alpha-1C is abundantly expressed in the hippocampus (Tanaka et al., 1995). Because our studies show that subunit composition does not affect Pb++ block, these results should be applicable to different neuronal subtypes. In conclusion, L-type Ca++ channels are so susceptible to Pb++ that they may be blocked during lead poisoning, contributing to the observed neurotoxicity.
Acknowledgments
We thank Tina Z. Hovance and Don Bers for providing healthy rat ventricular myocytes.
Footnotes
-
Send reprint requests to: Edward Perez-Reyes, Ph.D., Department of Physiology, Loyola University Medical Center, Maywood, IL 60153. E-mail eperez{at}luc.edu
-
↵1 This work was supported in part by the National Institutes of Health and the American Heart Association (E.P.R.). E.P.R. is an Established Investigator of the American Heart Association.
-
↵2 E. Perez-Reyes and L. L. Cribbs, unpublished observations.
- Abbreviations:
- DMSA
- meso-2,3-dimercaptosuccinic acid
- DMPS
- 2,3-dimercapto-1-propanesulfonic acid
- TEA
- tetraethylammonium
- NMG
- N-methyl-glucamine
- HEPES
- 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
- HEK
- human embryonic kidney
- Received September 4, 1996.
- Accepted March 7, 1997.
- The American Society for Pharmacology and Experimental Therapeutics