Assistant Research Professor of Neurology Assistant Research Professor of Neuroscience. Molecular-level research studies and diagnostics on neurogenetic ...
Toxicology Letters ELSEVIER
Toxicology Letters 82183 (1995) 255-261
Calcium channels as target sites of heavy metals Dietrich Biisselberg* Heinrich He&e-University
Diisseldorf, Physiology Il. Moorenstrafie 5, 40225 Diisseldorf. Germany
Abstract Zinc (Zn), aluminium (Al), mercury (Hg), methylmercury (MeHg) and lead (Pb) extracellulary applied reduce voltage-activated calcium channel currents (VACCCs); Pb and Al also reduce N-methyl-o-aspartate (NMDA)activated channel currents (NACCs). Pb is most effective in reducing VACCCs, with an IC,, of 0.46 yM, followed by Hg (IC,, = 1.1 PM) and MeHg (IC,, = 2.6 PM). Zn and Al were less potent (IC,, = 69 and 84 PM, respectively). Al acts on channels in the open state; its effect is pH dependent. The effects of Pb were specific for VACCCs and NACCs. Hg, Al and Zn had only minor effects on voltage-activated potassium and sodium channels, while MeHg reduced potassium channel currents (IC,, = 2.2 PM) and, at higher concentrations, sodium channel channel currents. These results demonstrate currents (IC,,, = 12.3 PM). Al also reduced other receptor-activated that a variety of metal species produce different actions at the level of the cell membrane. Keywords:
Voltage-activated calcium channel currents (VACCCs); NMDA-activated Lead (Pb); Mercury (Hg); Methylmercury (MeHg); Aluminum (Al): Zinc (Zn)
1. Introduction Learning and memory processes are triggered by a rise of the intracellular calcium concentration. While the extracellular calcium concen-
tration is in the millimolar range, the intracellular calcium concentration in neurons is very low (10-7-10-6 M) and closely regulated. Extracellular calcium enters neurons through calcium permeable ‘gates’. These calcium permeable channels are opened by 2 entirely different mechanisms: the first is the receptor-activated type, opened by the agonist N-methyl-o-aspartate (NMDA), while the second type is opened
* Corresponding
author.
channel
currents
(NACCs):
by depolarization of the membrane potential. Hence, the first type is named the NMDA-activated channel (NAC) and the second the voltage-activated channel (VACC). NACs have a magnesium and a zinc binding site and are modulated by different metals [l]. For VACCCs several subtypes have been described [2], all of them are highly selective for calcium. Zinc (Zn) is an essential metal, while other metals and metal compounds such as mercury (Hg), methylmercury (MeHg), aluminum (Al) and lead (Pb) are toxic and interfere with cognitive functions. We examined the effects of these metals using the whole-cell patch clamp technique with either cultured rat dorsal root ganglion (DRG) neurons for recording voltage-activated calcium channel currents (VACCCs), or
0378-4274/95/$09.50 @ 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0378-4274(95)03559-7
256
I Toxicology
D. Btisselberg
acutely isolated NMDA-activated
Letters
rat hippocampal neurons for channel currents (NACCs).
teasing of the slices using a pair of glass needles. In most cases CA1 neurons were used for recording. For whole cell patch-clamp recordings the solution was changed to the recording solution as indicated in Table 1.
2. Materials and methods
2.2. Preparation and recordings of DRG
Cultures of neurons from the DRG of young rats were used to study the actions of metals on VACCCs. Since DRG neurons do not express NMDA receptors, acutely isolated hippocampal neurons were used to study NACCs. 2.1. Preparation and recordings neurons
neurons
DRG were removed from 2-4-day-old rat pups and stored in phosphate-buffered saline. The ganglia were incubated in 0.9 ml F14 medium (DUN, Germany) containing 10% horse serum (GIBCO, USA) and 0.1 ml collagenase (12.5 mg/ml; Sigma Type II), for 13 min at 37°C. After the incubation the collagenase was removed and the ganglia were washed in F14 medium 3-5 times. Trypsin stock (0.1 ml, from 25 mg/ml; Sigma Type IX) and 0.9 ml F14 were added and the ganglia were incubated for 6 min at 37°C. After removing the trypsin-containing solution, the ganglia were washed with F14 and transferred to a plastic test tube containing 2 ml F14 medium with 10% horse serum, and DNAase (1 mg/ 10 ml; Sigma Type II).
of hippocampal
Hippocampal slices from 2-3-week-old rats were prepared and incubated for 40 min in Kreb’s Ringer solution (Table 1; bubbled with carbogen). The slices were transferred to a low calcium medium (Table 1) for 10 min and treated with Aspergillus oryzae protease (Sigma; 3.5 units/mg). These slices could be maintained for up to 10 h at room temperature. The neurons were isolated by mechanical Table 1 Ionic composition
of external Tyrodes
and internal
NaCl KC1 HEPES Glucose CaC& MgCt, TEA-Cl BaCl, TTX CsCl EGTA Na-ATP’ KH,PO, NaHCO,
145.0 2.5 10.0 10.0 1.5 1.2
solutions
Ca-currents External
82183 (1995) X5-261
Enzyme-Solution
External
Internal
10.0 10.0
10.0
1.0 130.0 10.0 0.0004
4.0
140.0 10.0 2.0
AsplGly
125.0 3.7 5.0 10.0 1.0 0.5
Internal
Krebs-Ringer
AsplGly
125.0 3.7 5.0 10.0 1.8 1.3
140.0 5.0 10.0 10.0 1.8
1.0 1.2 25.0
1.2 25.0 loo.0 30.0
KF Tri-Cl PH
7.4
7.2
All concentrations are given in mM. ’ To avoid rundown of the calcium channel
7.2
current
7.4
due to a reduction
7.4
of the energy
pool.
7.3
7.2
D. Busselberg
I Toxicology Letters 82i8_3 (199-f) XT-261
The ganglia were triturated with a fire-polished Pasteur pipette until they were dispersed and the medium appeared opaque. The debris was removed by filtering through a nylon mesh (4200 pm). The cell suspension (50-100 ~1) was placed in small petri dishes (Falcon, ‘easygrip’) and incubated for 2 h at 37°C with 5% CO,. Then the dishes were filled with 1 ml F14 containing 10% horse serum. The neurons were used within 3 days after preparation. 2.3. Recording
technique and analysis
The neurons were patch-clamped in the whole cell configuration using a HEKA EPC-9 patchclamp amplifier controlled by a computer. Electrode resistance was between 2 and 5 MR. The compositions of the internal and external solutions (Tyrode and calcium current solutions) are shown in Table 1. All experiments were conducted at room temperature. Cells were clamped at -80 mV (DRG neurons) or -60 mV (hippocampal neurons). All solutions used for recordings are shown in Table 1. The dose-response relations for the effects of the various metals on VACCCs and NACCs of the neurons were determined by fitting mean currents to the equation:
where Ica: -,x, is the calcium current measured in the presence of a given concentration of a metal, ICa’ *(conlrol) is the calcium current without the metal, K,,, is the apparent dissociation constant, and n is the Hill coefficient. Metals were added in different concentrations to the external solution immediately before application to the neurons.
3. Results 3.1. Actions of Pb and Al on NACCs
Both Pb and Al reduce currents activated through the NMDA receptor. Application of the agonists aspartate (500 ,uM) or NMDA (1 mM)
257
simultaneously with glycine (20 PM) resulted in an inward current. Application of Pb or Al together with the agonist results in a partially reversible, dose-dependent reduction of the current through the channel/receptor complex [35]. The current was reduced to half at Pb concentrations between 20 and 50 PM, and at Al concentrations of less than 50 PM (i. e. 1.4 pg/ml Al). Both metals reduced the receptoractivated currents over the whole voltage range without changing the reversal potential; no voltage dependence was found. Preincubation with Pb increased the reduction of the receptor-activated current [5]. We did not test the action of Al with preincubation, but we have shown that it reduces glutamate and AMPA-activated currents in a similar concentration range [4]. As Alkondon et al. [6] have proven, the effect of Pb is specific for NMDA-activated currents. Only minor actions on kainate- or quisqualate-activated currents, with the same concentrations of Pb which block NMDA-activated currents, have been shown. No interactions of Pb with the metal binding sites of Zn or Mg of the NMDA-receptor/channel complex have been reported [7].
3.2. Actions of Pb, Al. Hg. MeHg and Zn on VA CCCs 3.2.1. Actions on calcium channel subtypes 3.2.1.1. Actions on L-IN-type channels.
Depolarizing from the holding membrane potential of -80 mV to 0 mV for 75 ms results in activation of high VACCCs. (We have not tested for P-type calcium channel currents). All metals tested (Pb, Al, Hg, MeHg and Zn) reduced the peak of VACCCs. Pb was most effective in reducing these currents with an IC,,, of 0.46 ,LLM[3], followed by Hg (IC,,, = 1.1) [8], MeHg (IC,,, = 2.6 ,uM) [9], Zn (IC,,, = 69 ,uM) [lo] and Al (IC,,, = 83 PM) [ll]. The Hill coefficient was close to 1 for Pb, Hg, MeHg and Zn. When Al was applied the Hill coefficent varied with pH, from 2.8 (pH 7.7) to 2.2 (pH 7.3) to nearly 1.4 (pH 6.7) [12]. While the reduction of the calcium channel current by Pb was reversible up to 60%, the
258
D. Biisselberg
I Toxicology Letters 82183 (1995) 255-261
actions of Al and Zn were less reversible (up to 30%). The blocking effect of Hg or MeHg was not more than 10% reversible. 3.2.1.2. Actions on T-type calcium channel currents. In about 5% of the neurons tested low VACCCs (through the voltage-activated T-type calcium channels) could be found. These currents were activated maximally by a voltage step to -30 mV Pb, Hg, MeHg and Al reduced Tchannel currents in a similar concentration range as currents through high VACCCs. Zn was different: a concentration of this metal which reduced the peak current through the L-/N-type channels by less than 10% blocked the current through the low VACCs almost completely (>80%) [13]. Due to the low number of recordings with T-type channel currents, we were not able to determine a concentration response relationship. 3.2.2. Specificity We tested the actions of Pb, Al, MeHg and Zn on voltage-activated sodium and potassium channels. (We have not tested the specificity of Hg). All metals reduced these currents to some degree, but Pb, Al or Zn had only a slight action (~10% reduction) on voltage-activated potassium or sodium channels at concentrations that reduce the currents through VACCs by more than 80% [14]. MeHg was about as effective in reducing the currents through potassium channels (I& = 2.6 ,uM) as it was in VACCCs, while sodium channel currents were less sensitive to MeHg (IC,, = 12 PM) [15].
3.2.3. Time course and use dependence After application of Pb, Hg, MeHg or Zn, a new and lower steady state was reached within a few minutes [14J5]. With Al the time to reach a steady state was about twice as long [ll]. The actions of Pb and Zn on VACCCs did not depend on an open channel state. With Hg or MeHg the VACCCs were partly reduced when the channels were activated. In the presence of Al the VACCCs were not reduced when the protocol of channel activation was resumed after several minutes (up to 6 min) [ll].
3.2.4. Current-voltage relation All metals tested reduced the VACCCs over the entire voltage range. With Pb, the maximal current was generated with exactly the same depolarisation step (to -5 mV) as without Pb. All other metals shifted the maximum of the current voltage relation curve to more depolarized potentials. This shift depended upon the concentration of the metal used and was more pronounced at higher concentrations of the metals. The shift was most obvious with Al, smaller with Zn, Hg or MeHg [14,15]. 3.2.5. Simultaneous application of different metals and internal application
When Pb, Zn or Al were applied simultaneously in the range of their IC,, values, additive actions on VACCCs were found, which were independent of the order of application [16]. With 2 cations in the external solution,VACCCs were reduced by 75% (2 9%) and were even further reduced when a third metal was added. Al, applied extracellularly or intracellularly, on VACCCs reduced VACCCs independently [17]. When Al was applied simultaneously both inside and outside, the 2 effects were additive, suggesting that Al has both an external and an internal binding site. Besides the direct effects on NACs and VACCs we found some other effects, which are not directly related to their actions on NACCs or VACCCs: 3.2.5.1. Changes of membrane currents. While Pb, Al and Zn did not change the membrane current when used in the above-mentioned concentration range (in which the VACCCs were reduced), the application of higher concentrations of Hg (22 PM) or MeHg (a10 PM) resulted in an unidentified membrane current. Hg caused an inward current, while MeHg generated a biphasic current with a transient inward and a long-lasting outward component [9,15]. 3.2.5.2. Effects on long-term potentiation. Pb and Al reduce the generation and maintenance of long-term potentiation (LTP) in vitro [l&19] and in vivo [19,20]. In a rat brain slice preparation, Pb and Al reduced LTP in a concentration-
D. Biuselberg
I Toxicology Letters 82183 (1995) 255-261
dependent fashion. LTP was abolished with concentrations of 10 PM Pb or 100 PM Al.
259
Actions of Metals on VACCCs
4. Discussion AI(
VACCCs and NACCs are sensitive to the metals tested. Both Pb and Al reduce NACCs through the receptor/channel complex; however, the concentrations of metal needed were relatively high and most of the effects were reversible. Preincubation with Pb resulted in a more pronounced reduction of the current and there was always a small, but relevant, portion of the current which was irreversibly blocked. This irreversible effect might be more relevant in regard to the neurotoxicity of Pb. The metals Pb, Al, Hg, MeHg and Zn reduce the currents through VACCCs, but their actions differed with respect to time course, affected calcium channel subtypes, use dependence, effective concentration range, pH dependence, and screening effects at the surface of the membrane (seen in a shift of the current voltage relations). How could this variety of actions on VACCCs be explained? Why do all metal cations not act in the same, or at least in a similar way? For the explanation the following possible mechanisms have to be considered (Fig. 1): (1) chemical peculiarities of the various metal species; (2) unspecific effects at the membrane (screening of surface charges); (3) specific effects: (a) at the membrane, (b) at the entrance to the channel, (c) within the channel; (4) intracellular changes (interactions with second messengers, etc.). 4.1. Chemical peculiarities of various metal species The active forms of Pb and Zn in physiological solutions (pH = 7.2-7.3 and a chloride concentration of about 120 mM) are most likely Pbzf and Zn”. More than or other metals the form of Al in aqueous solution depends on pH: the amount of A13’ increased when the pH decreased (down to 6.7) and the concentration of AI(O increased when the pH decreased.
Fig. 1. Possible calcium channel
IAI(OH)~AI(OH),’
interactlons of metals on voltage-activated currents. For details see text.
Assuming, that Al”+ is the active form in reducing VACCCs, the pH dependence of the effect is easily explained. HG and MeHg probably exist in physiological solutions as uncharged complexes (HgCl, or CH,HgCl), which might pass through the neuronal membrane [21]. But we do not know how these compounds react with the surface charge of the cell membrane. There they might very well exist as singly or doubly charged cations. 4.2. Unspecific effects at the cell membrane Except for Pb, the metals tested shifted the maximum of the current-voltage relation for the
260
D. Busselberg
I Toxicology Letters 82l83 (1995) 255261
VACCCs concentration dependently to more depolarized voltages, the degree of shift depending on concentrations. Such a shift is typical of a charge screening effect. But since all concentrations used in our studies were in the micromolar range (even Al was never used in concentrations over 300 ,uM), a general screening of the surface charges at the cell membrane seems unlikely. Furthermore, a charge screening effect should change the currents through all channels in a similar fashion.
produces a concentration-dependent shift of the current voltage-relation to more depolarized voltages, which indicates an additional screening of membrane charges due to binding at unspecific or specific sites. Binding of Hg and MeHg within the channel is also conceivable, because at least a part of the effects of these 2 compounds on VACCCs needed an open channel and the effects were not reversible.
4.3. Specific effects
Although we have no direct evidence for such mechanisms, we cannot exclude the possibility that some of the metals tested had internal actions which might have changed the currents through NACs or VACCs. Externally applied Pb reduces the rise of the internal calcium concentration without passing through the cell membrane [22]. In the case of Al we have a different reason for believing that it does not pass through the cell membrane; the additional reduction of the current when Al was applied intracellulary suggests not only a second effect, which is triggered intracellulary, but also demonstrates that there was probably no metal within the cell before it was added intracellularly. Hg and MeHg are uncharged compounds in physiological solutions which are able to pass through the cell membrane and raise the internal calcium concentration [23], however, the time scale of such action is not known. Nevertheless, the first site of action of acutely applied Hg compounds is the surface of the cell membrane. We do not know to what extent the membrane currents we have seen at higher concentrations of Hg are related to intracellular effects. The current through the NMDA receptor channel complex is reduced by Pb and Al. Pb, Hg, MeHg, Al and Zn reduce VACCCs. While the rise of intracellular calcium is most important for learning and memory, a change in the rise time and/or amount of calcium in the neuron might explain some of the long-lasting effects of these metals. For a specific metal ion we may be able to determine a mode of action which is more likely than other interactions, but we are not able to exclude most of the other possible external or internal mechanisms discussed above.
4.3.1. At the cell membrane The shift of the current-voltage relation might be explained by specific binding sites at the surface of the membrane. But this is also improbable because this explanation suggests different specific binding sites for the different metals otherwise the additive effects could not be explained - and the effect should be similar for all types and subtypes of voltage-activated channels (like a general charge screening effect), which is clearly not true. 4.3.2. At the entrance to the channel Assuming that metal cations screen specific charges at the entrance of the channel, a specific metal binding site at this location could explain both the shift of the current voltage relation and the low concentrations needed. But such an explanation is only valid for such metal cations as Pb or Zn which do not need an open channel state for their action. The hypothesis that the binding site of these 2 cations might be at the entrance of the channel is underlined by another fact: compared to the other metals tested, the effects of these 2 cations were - at least partially - reversible. 4.3.3. Within the channel For actions within the channel the metals must enter the channel. A binding site within the channel might be the main location of action of Al: Al needs an open channel to reduce VACCCs and its effects were not reversible. These facts indicate a strong binding site within the channel, which is not easily accessible. However, Al also
4.4. Intracellular changes
D. Biisselherg
I Toxicology Letters 82183 (199.5) X-161
Overall, it is likely that the neurotoxicity of metal ions is mediated by a variety of different mechanisms. These different mechanisms and the simultaneous action of different metals at the same time are the reason for the complex neurotoxicity of metals on NACCs or VACCCs.
Due to restrictions of space not all important work is cited. This paper basically reviews some of our own data. Important contributions on the actions of metals on calcium channels have been made by other groups and I apologize to all my colleagues who are not mentioned. For critical reading I thank Dr. S. Cleveland and for technical assistance C. Wittrock, T. Kordela and P. Schwarz.
References
[‘I
Busselberg. D. (1995) Voltage-activated calcium channel currents of rat DRG neurons are reduced by mercuric chloride (HgCL) and methylmercury (CH,HgCI). Neurotoxicology (in press). [IO] Busselberg, D.. Pekel, M.. Michael. D. and Platt. B (1994) Mercury (Hg” ) and zmc (Zn’ ): two divalent cations with different acttons at voltage-activated calcium channel currents. Cell. Mol. Neurobtol. 14. 67% 687. [II]
Acknowledgements
Busselberg. D.. Platt. B.. Haas. H.L. and Carnenter. D.O. (1993) Voltage gated calcium channel currents of rat dorsal root ganglion (DRG) cells are blocked by Al’ ’ Brain Res. 622. 163- 168.
I121 Platt. B., Haas. H. and Busselberg,
D. (1993) Extracellular pH modulates aluminum-blockade of mammalian voltage-activated calcium channel currents. NeuroReport 4, 125ll1254.
[I31 Busselberg.
D.. Michael. D., Evans. M.L., Carpenter. D.O. and Haas. H.L. (1992) Zinc (Zn’ ’ ) blocks voltage gated calcium channels m cultured rat dorsal root ganghon cells. Bram Res. 593. 77-81.
[I41 Btisselberg,
D.. Platt. B., Michael. D., Haas, H.L. and Carpenter, D.O. (1994) Mammahan voltage-activated calcium channel currents arc blocked by Pb’ ’ . Zn.’ ’ and Al”. J. Neurophystol. 71, 1491-1497
[I51 Leonhardt.
Mayer. M.L. and Westbrook. G.L. (1987) Permeation and block of N-methyl-aspartc receptor channels by divalent cations in mouse cultured central neurons. J. Physiol. Lond. 394. 501-527. A.P.. Nowycky, M.C. and Tsien, R.W. (1987) Kmetic and pharmacological properties distmguishing three types of calcium currents in chick sensory neurones. J. Phystol. Lond. 394. 149-172.
R.. Haas. H.L. and Busselberg. D. (1995) Voltage-gated calcium-, sodium-. and potassium channel currents of rat DRG neurons are reduced by methylmercury (CH,Hg ’ ) Proc. Gottinger Neurobtol. Conf. (Abst.).
[IhI Platt. B. and Busselberg.
D. (1994) Combmed actions of and AI’ on voltage-activated calcium channel currents. Cell. Mol. Neurobiol. 14, 831-840.
Ph”. Zn’
PI Fox.
D., Michael, D. and Platt, B. (1994) Pb” reduces voltage- and NMDA-activated calcium channel currents. Cell. Mol. Neurobiol. 14. 711-722. [41 Platt, B.. Haas, H. and Busselberg, D. (1994) Aluminum reduces glutamate-activated currents of rat hippocampal neurons NeuroReport 5. 2329-2332. I51 Uteshev, V, Busselberg, D. and Haas. H.L. (1993) Pb’* modulates the NMDA-receptor-channel complex. Naunyn Schmiedebergs Arch. Pharmacol. 347. 209-213.
I’71 Platt, B. and Biisselberg.
D. (1994) Actions of aluminum on voltage-activated calcium channel currents. Cell. Mol. Neurobiol. 14. 819-829.
I31 Busselberg.
bl Alkondun.
M., Costa. A.C.S., Radhakrishnan. V., Aronstam. R.S. and Albuquerque, E.X (1990) Selective blockade of NMDA-activated channel currents may be implicated m learning deficits caused by lead. FEBS Lett. 261, 123-130.
[71 UJthara.
H. and Albuquerque. E.X. (1992) Developmental change of the inhibition by lead of NMDA-activated currents m cultured hippocampal neurons. J. Pharmacol. Exp. Ther. 263, 868-875.
PI Pekel. M.. Platt, B. and Busselberg,
D. (1993) Effects of mercury (Hg’ ’ ) on voltage-activated calcium channel currents m Aplysia and cultured rat neurons. Bram Res. 632, 121-126.
[91 Leonhardt.
161
R..
Pekel.
M.. Platt.
B.. Haas.
H.L.
and
[I81 Hori. N.. Busselberg.
D.. Matthews. R., Parson, P. and Carpenter, D.O. (1993) Lead blocks LTP by an action not at NMDA receptors J. Exp. Neurol. 119. 192-197.
[I91 Platt.
B.. Carpenter. D.0, Busselberg. D.. Reymann, K.G. and Rtedel, G. (1995) Aluminum blocks hippocampal long-term potentiation in rats 111i’nro and m ~‘rt’o. Exp. Neurol. 134. 73-86.
PI
Altmann, L.. Svemsson. K. and Wiegand, H (1991) Longterm potentiation in rat hippocampal shces is impaired following acute lead perfusion. Neurosci Lett. 128. 1099112.
J (1981) Inorgamc mercury (Hg.’ ) transport through hptd btlayer membranes. J. Membr BIoI. 61, 61-66. WI Busselberg, D., Wunder. L.. Domann. R. and Haas. H.L. (1994) Lead reduces calcium entry wtthout passing the cell membrane of mammahan neurons: FURA 2 measurements. Proc. Neurosci. Meeting m Miami 623. 15. [231 Hare, M.F.. McGinms. K.M. and Atchison. W.D. (1993) Methylmercury increases mtracellular concentrations of and heavy metals m NGIOB-15 cells. J. Pharmacol. Ca’ Exp. Ther. 266, 1626-1635
PII Gutknecht,
