- Email: [email protected]
Inhibition of TEA-Induced LTP by Aluminum
EXPERIMENTAL NEUROLOGY ARTICLE NO.
141, 240–247 (1996)
0158
Inhibition of TEA-Induced LTP by Aluminum BETTINA PLATT1
AND
KLAUS G. REYMANN
Federal Institute for Neurobiology, Department Neurophysiology, P. O. Box 1860, D-39008 Magdeburg, Germany
Brief application of tetraethylammonium (TEA) to hippocampal slices causes long-term potentiation (TEA LTP) at synapses of CA1 pyramidal neurons characterized by a long-lasting increase of field excitatory postsynaptic potential (fEPSP) slope and population spike (PS) amplitude. Since this kind of potentiation requires the activation of voltage-dependent calcium channels, we examined the effect of the inorganic calcium channel blocker aluminum, which has been shown to impair tetanus-induced LTP (eLTP). We found that Al inhibited in a concentration-dependent manner both fEPSP slope and PS amplitude potentiation by TEA; 0.68 mg/ml Al attenuated TEA LTP, while a complete block of long-lasting potentiation was obtained for 2.7 mg/ml Al. Occlusion experiments revealed that both concentrations of Al allowed the induction of eLTP 60 min after TEA/Al exposure. However, longer application (15 min) of 2.7 mg/ml Al before the induction of TEA LTP prevented the subsequent induction of eLTP although no significant differences concerning the action on TEA LTP were observed. This indicates a general loss of neuronal plasticity which might be due to progressive neuronal cell damage. Since the effective concentration range of Al is directly comparable to the action of Al on eLTP, our data provide evidence for shared mechanisms of both potentiations. Although based on different induction mechanisms, Ca21 is assumed to be a general intracellular trigger for both forms of LTP and thus it can be hypothesized that the neurotoxic action of Al is due to interference with Ca21-dependent processes by inhibition of calcium conductances. r 1996 Academic Press, Inc.
1. INTRODUCTION
Aluminum has been reported to be a possible etiologic factor in Alzheimer’s disease (21, 25) and other neurological disorders such as Parkinson dementia of Guam and dialysis syndrome (10, 11). Furthermore, exposure to Al causes learning and memory deficits in
1 To whom correspondence should be addressed at present address: Department of Physiology, University of Leeds, Worsley Medical and Dental Building, Leeds LS2 9NQ, Great Britain, Fax: (144) 1132334228, E-mail: [email protected].
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
animals and humans (5, 17, 35). Experimental evidence for a direct interference of Al with mechanisms underlying neuronal excitability and plasticity has been provided by means of the model of long-term potentiation (LTP). LTP evoked by electrical stimulation of afferent fibers (termed eLTP) is a model frequently used to study mechanisms underlying synaptic plasticity (3). There are at least two forms of eLTP distinguished by the involvement of N-methyl-D-aspartate (NMDA) receptors, i.e., NMDA receptor-dependent and NMDA receptor-independent eLTP. The contribution of NMDA receptors appears to depend on the brain area and on the tetanus used for eLTP induction. The involvement of voltage-dependent calcium channels (VDCCs) has been proven for NMDA receptor-independent eLTP (12, 18). In addition to this ‘‘classical’’ LTP model, a variety of investigations were performed on chemically induced LTP, elicited for instance by application of high Ca21 and/or K1 concentrations, or by means of K1 channel blockers (2, 9, 13, 32, 34). Although the applied substances determine the nature of the elicited potentiation, all models are assumed to be based on postsynaptic depolarization in conjunction with enhanced presynaptic transmitter release and it is generally accepted that an increase of postsynaptic Ca21 is the major trigger for LTP (22). In a variety of investigations, the K1 channel blocker tetraethylammonium (TEA) has been used to characterize this kind of synaptic enhancement (termed TEA LTP) and also to compare it with the mechanisms underlying eLTP (1, 15, 16). Brief application of TEA causes a depolarization based on the inhibition of various K1 conductances, of which the delayed rectifier is assumed to be the crucial one (1). As pointed out by Petrozzino and Connor (26), application of TEA induces a more prolonged and pronounced increase of intracellular Ca21 compared to eLTP. Using the model of eLTP to study Al neurotoxicity, it has been shown that Al inhibits in vitro eLTP after in vivo exposure (8), and recently, we demonstrated a direct impairment of in vivo and in vitro eLTP by Al before any obvious morphological changes occur and even before basic electrophysiological parameters were affected (28, 29). In the present study, we examined the action of Al on TEA LTP as a comparative approach. Furthermore, Al acts as an inhibitor of both high- and
240
241
ALUMINUM INHIBITS TEA LTP
low-voltage activated calcium channels of cultured dorsal root ganglion neurons (7, 27) and the contribution of VDCCs in TEA LTP has been demonstrated by previous reports (1, 15, 16). The action of Al on TEA LTP is therefore of particular interest with respect to answering two major questions. First, using basically the same preparation and Al concentrations as in our previous study on eLTP, we may be able to detect further similarities and differences of TEA LTP versus eLTP. Second, the action of Al on TEA LTP may provide useful insight into the mechanisms causing cognitive dysfunctions due to Al exposure since an interaction with TEA LTP would support the hypothesis that Al interferes mainly with the Ca21 homeostasis of neurons and Ca21-dependent processes underlying neuronal plasticity. 2. MATERIALS AND METHODS
Experiments were performed on hippocampal slices (400 µm) from adult male Wistar rats (7–8 weeks old) after equilibration in a standard Ringer for at least 1 h at 32°C. Slices were placed in a submersion chamber and superfused (3 ml/min) with a modified Ringer solution for 30 min (composition in mM : NaCl 126, KCl 5, MgCl2 1.3, CaCl2 2.5, Tris–Cl 6, glucose 10, NaHCO3 15; saturated with 95% O2/5% CO2, pH 7.4, 32°C). The use of a modified Ringer was essential in order to prevent precipitation of Al as sulfate and phosphate compounds (23). Field potentials were elicited by stimulation of the Schaffer collateral/commissural fibers and extracellularly recorded in both stratum radiatum and stratum pyramidale of area CA1 using ,1 MV glass electrodes filled with modified Ringer. Every 3 min test stimuli (three bipolar pulses, 0.2-ms duration, 10-s interval) were applied via a monopolar stainless steel electrode. The initial slope of the rising phase of the averaged field excitatory postsynaptic potential (fEPSP) and the amplitude of the population spike (PS) were estimated. The test stimulus intensity was set at 30% of fEPSP maximum, which elicited a PS of about 20% of the maximum. All compounds except Al were obtained from Sigma; Al was taken from an ‘‘aluminum atomic absorption standard solution’’ (984 µg/ml of Al in 1 wt% HCl, Aldrich). Two different concentrations of Al, i.e., 0.68 and 2.7 µg/ml Al, were investigated. The concentrations of Al are given here in units of micrograms per milliliter, since Al forms a variety of species in solution and the relative molarities are not known. Assuming a complete dissociation of the trivalent cation, 0.68 µg/ml corresponds to 25 µM and 2.7 µg/ml corresponds to 100 µM Al31. These concentrations were used to allow a comparison with our previous work on eLTP (28, 29). Here we performed two different application schedules with both concentrations. The first series was con-
ducted without preapplication of Al alone; TEA and Al were applied simultaneously (7 min) in order to allow the analysis of the action of Al on the induction of TEA LTP without interference with general neurotoxic processes. In a second series of experiments, a preapplication of Al (15 min) was performed followed by the application of Al in conjunction with TEA (7 min). This schedule was chosen to investigate whether a longer exposure to Al induces a more pronounced inhibition of potentiation by Al. Accordingly, control experiments on the reversibility of baseline effects caused by Al application were performed by applying 2.7 µg/ml Al for 22 min, and the recovery after wash was monitored for 60 min. Sixty minutes after TEA application, a reset of the stimulation strength was performed to obtain fEPSPs comparable to baseline responses. As a subsequent test for plasticity and occlusion of eLTP by TEA LTP, eLTP was induced by three high-frequency stimuli (HFS; 100 Hz for 1 s; interburst interval, 20 s). All data reported here are given as means 6 SEM and potentiation is presented as percentage change relative to baseline. Statistical significance was revealed by analysis of variance (ANOVA) and subsequent t test on the basis of P # 0.05 for significant and #0.01 for highly significant differences. Values obtained 60 min after TEA application were taken as a measure of potentiation and were used for comparison between groups. 3. RESULTS
In control experiments (n 5 14), superfusion of 25 mM TEA for 7 min induced a sustained increase of the PS amplitude (Fig. 1A) and fEPSP slope (Fig. 1B) after washout of TEA. During TEA application, we observed an initial increase of the fEPSP slope, followed by a feigned transient decrease due to interference of the prolonged fiber volley with the measurement of slope function. Sixty minutes after TEA application, PS amplitude and fEPSP were potentiated up to 266 6 8 and 138 6 5%, respectively. Concerning the PS data, it should be mentioned that all reports published so far on TEA LTP were based only on fEPSP data in the stratum radiatum so that a comparison with data of others is not possible. Thus, this is the first report on PS amplitude potentiation by TEA. Since our last report on the action of Al on eLTP was performed using the PS amplitude as a measure (30), both parameters were recorded and analyzed in the present study. To obtain responses in the range of baseline values a subsequent reset of the stimulus strength was performed. HFS application resulted in a further potentiation (PS amplitude: 178 6 7%; fEPSP: 113 6 5%, 12 min after HFS). In general, the obtained amount of eLTP was weaker than that observed in naive slices
242
PLATT AND REYMANN
FIG. 1. Application of 25 mM TEA for 7 min (indicated by a solid bar) induces an increase of the population spike (PS) amplitude (A) and field excitatory postsynaptic potential (fEPSP) slope function (B). Averaged values 6 SEM (n 5 14) are plotted versus time relative to baseline. After reset of the stimulation strength (single arrow) and HFS application (double arrow), a further potentiation could be induced. Sample traces at time points indicated in the graphs are given on the right.
(Ref. 29, PS amplitude: 250%, 15–60 min after HFS), indicating a partial occlusion of eLTP by TEA LTP. Application of 0.68 µg/ml Al in conjunction with TEA (n 5 7) led to a stronger and longer lasting initial depression of the fEPSP slope (Fig. 2B) compared to
FIG. 2. Action of 0.68 µg/ml Al on TEA LTP (n 5 7). Time-course of mean values calculated relative to baseline is given. Application of TEA and 0.68 µg/ml Al (solid bar) caused a significantly attenuated potentiation of both PS amplitude (A) and fEPSP slope (B) compared to control TEA LTP. Subsequent reset of the stimulus strength (single arrow) and HFS (double arrow) potentiated both components. On the right, sample traces for the time points indicated in the graphs are illustrated.
controls. Other than a general inhibition there was no alteration observed of the time course of the PS amplitude during TEA application (Fig. 2A). A significantly attenuated potentiation (Table 1) developed, and after washout PS amplitude and fEPSP slope were potentiated to 189 6 8 and 119 6 6% (t 5 60 min), respectively. Reset of the stimulus strength and subsequent HFS
TABLE 1 Statistical Comparison at t 5 60 min t values Control 0.68 µg/ml 0.68 µg/ml (pre) 2.7 µg/ml
0.68 µg/ml PS ampl./EPSP slope
0.68 µg/ml (pre) PS ampl./EPSP slope
2.7 µg/ml PS ampl./EPSP slope
2.7 µg/ml (pre) PS ampl./EPSP slope
2.54(*)/2.37(*)
2.19(*)/2.24(*) 0.33(ns)/0.21(ns)
5.94(**)/5.48(**) 2.95(**)/2.92(*) 3.31(**)/3.24(**)
3.84(**)/4.72(**) 2.38(*)/2.27(*) 2.01(*)/2.56(*) 1.37(ns)/1.64(ns)
Note. pre, preapplication (15 min) of the respective Al concentration before induction of TEA LTP. * P , 0.05; ** P , 0.01; ns, not significant.
ALUMINUM INHIBITS TEA LTP
243
values (PS amplitude: t 5 2.53, P , 0.05; fEPSP: t 5 1.63, P . 0.05). Since no significant differences to the results shown in Fig. 2 were found (see Table 1), it appears that the action of 0.68 µg/ml Al on TEA LTP was not influenced by the preapplication of 0.68 µg/ml Al. When 2.7 µg/ml Al was applied simultaneously with TEA (n 5 8), a similar time course as for 0.68 µg/ml Al was obtained during application, but the initial depression was found to be prolonged for the fEPSP slope function (Fig. 4). A stronger inhibition of potentiation occurred with a significant decline below baseline values for both PS amplitude and fEPSP slope (t 5 60 min: PS amplitude: t 5 2.6, P , 0.05; fEPSP: t 5 2.5, P , 0.05). The reduction was highly significant compared to controls and compared to the action of 0.68 µg/ml Al at t 5 60 min (Table 1) with a PS amplitude of 69 6 4% and a mean fEPSP slope of 90 6 6%. In order to
FIG. 3. Preapplication of 0.68 µg/ml Al had no influence on the action of Al on TEA LTP (n 5 9). Averaged data are shown relative to baseline and plotted versus time. A slight but not significant inhibition of the baseline was observed during Al application alone (hatched bar). After coapplication with TEA (solid bar) potentiation of PS amplitude (A) and fEPSP slope (B) were highly significantly attenuated compared to controls. eLTP could be induced after reset (single arrow) and HFS (double arrow). Note the similarity to data shown in Fig. 2. Sample traces for time points in the graphs are illustrated on the right.
caused a clear potentiation of both PS amplitude (273 6 11%) and fEPSP slope (120 6 5%) 12 min after tetanus, which was more pronounced than under control conditions but only significantly different for PS amplitude data (PS amplitude: t 5 2.67; fEPSP: t 5 0.64, P . 0.05; P , 0.05). Al (0.68 µg/ml) applied 15 min before superfusion coincident with TEA (n 5 9) caused a slight but not significant reduction in baseline responses (Fig. 3). The potentiation obtained 60 min after TEA application was significantly reduced compared to controls (Table 1) with values of 175 6 20% (PS amplitude) and 121 6 6% (fEPSP slope). Subsequent reset and tetanization resulted in a potentiation of the PS amplitude (248 6 20%) and fEPSP slope (128 6 5%) 12 min after HFS. Again, the amount of eLTP obtained was significantly greater than under control conditions for the PS amplitude
FIG. 4. 2.7 µg/ml Al blocks TEA LTP (n 5 8). Time-courses of mean data calculated relative to baseline are shown. Application of TEA and Al (solid bar) caused a severe depression of TEA LTP, which was highly significantly reduced in comparison with controls. Both PS amplitude (A) and fEPSP slope (B) declined below baseline. Subsequent HFS (double arrow) induced a potentiation which was much greater for the fEPSP slope function compared to control data. On the right, representative raw traces are shown for the respective time points indicated in the graphs.
244
PLATT AND REYMANN
keep the time-course comparable to control experiments, HFS was applied after t 5 69 min, resulting in a clear potentiation of PS amplitude (179 6 18%) and fEPSP slope (152 6 10%) 12 min after tetanus. Compared to control conditions, the induced eLTP was significantly stronger in terms of the fEPSP slope, whereas the PS amplitude was potentiated in the same range as controls (fEPSP: t 5 3.774, P , 0.05; PS amplitude: t 5 0.005, P . 0.05). It appears that the application of 2.7 µg/ml Al may have a stronger and longer lasting effect on the PS which is supported by the stronger decline after TEA superfusion. For the fEPSP, it seems that stronger inhibition of TEA LTP by 2.7 µg/ml Al allowed a stronger subsequent eLTP. Application of 2.7 µg/ml Al alone (n 5 7; Fig. 5) led to a significant baseline effect; i.e., the PS amplitude and fEPSP slope were reduced by 12 6 3 and 8 6 1% after 15 min, respectively (PS amplitude: t 5 2.58, P , 0.05; fEPSP: t 5 3.6 P , 0.05). As already pointed out in our report on eLTP, the action of Al on the PS is not reversible after wash (28). Since we obtained differences in the action of Al on the fEPSP and PS (see above), we performed a series of control experiments in order to investigate the reversibility of the baseline effect on the fEPSP slope (n 5 8; see inset in Fig. 5B). In these experiments, the depression of the fEPSP slope by application of 2.7 µg/ml Al for 22 min (11 6 2%) recovered slightly after wash; an inhibition of 8 6 4% remained. This finding suggests that the action of 2.7 µg/ml Al causes a slight but sustained suppression of the fEPSP. When TEA was applied in conjunction with 2.7 µg/ml Al after preexposure, no initial increase in the fEPSP slope was obtained (Fig. 5B, main graph) when TEA application started and the enhancement of the PS amplitude was comparably small. The emerging potentiation after wash was weak and declined back to baseline within 45 min. At t 5 60 min, a mean PS amplitude of 107 6 8% and fEPSP slope of 106 6 5% were found. A comparison with control data revealed a highly significant difference of both PS amplitude and fEPSP slope data (see Table 1) and no significant difference from baseline data (PS amplitude: t 5 1.9, P . 0.05; fEPSP: t 5 2.1, P . 0.05). Nevertheless, comparison with the baseline values in the presence of 2.7 µg/ml Al indicated a significant potentiation of the fEPSP slope, but not of the PS (PS amplitude: t 5 1.9, P . 0.05; fEPSP slope: t 5 2.5, P , 0.05). Again, the effect on the PS appears to be stronger compared to the effect on the fEPSP. Group comparison between data obtained in the experiments with preapplication of 2.7 µg/ml Al versus those without preapplication gave no significant differences (see Table 1). Only in those experiments where a slight potentiation remained at t 5 60 min, was a reset of the stimulation strength performed. Unlike all other experiments, HFS did not result in a further potentiation.
FIG. 5. Preapplication of 2.7 µg/ml Al (n 5 7) inhibited synaptic transmission slightly but significantly as shown by the baseline effect on both PS amplitude (A) and fEPSP slope (B). Control experiments (n 5 8) revealed that the inhibition of the fEPSP slope by 2.7 µg/ml Al is only slightly reversible after wash (inset in B). Simultaneous superfusion of TEA and Al led to a highly significant reduction of potentiation after wash compared to control experiments. No eLTP could be induced after reset of the stimulus intensity (single arrow) and HFS (double arrow), indicating irreversible and/or damaging actions after prolonged superfusion of 2.7 µg/ml Al. Sample traces are given on the right for the time points indicated in the graphs.
Fifteen minutes after HFS, PS amplitude (98 6 6%) and slope values (101 6 5%) were found to be not significantly different from baseline data (PS amplitude: t 5 1.15, P . 0.05; fEPSP: t 5 1.57, P . 0.05) but significantly different from the data obtained without preapplication of 2.7 µg/ml Al (PS amplitude: t 5 2.28, P , 0.05; fEPSP: t 5 5.5, P , 0.01). Thus, it is likely that the longer application of 2.7 µg/ml Al resulted in an irreversible inhibition of components essential for eLTP induction. In Table 2 the absolute amount of potentiation in the presence of 0.68 and 2.7 µg/ml Al relative to baseline values and the relative inhibition of potentiation com-
ALUMINUM INHIBITS TEA LTP
TABLE 2 Inhibition of TEA LTP by 0.68 and 2.7 µg/ml Al: Absolute Values of Potentiation Relative to Baseline and Inhibition of Potentiation Relative to Control Data (in %) EPSP slope
PS amplitude
30 min 60 min 30 min 60 min absolute absolute absolute absolute pot./relative pot./relative pot./relative pot./relative inhibition inhibition inhibition inhibition 0.68 µg/ml 0.68 µg/ml (pre) 2.7 µg/ml 2.7 µg/ml (pre)
127/45 133/32 111/77 119/61
119/50 121/45 89/129 106/84
239/53 209/63 110/96 136/88
189/46 175/52 69/142 107/95
pared to control data at t 5 30 and t 5 60 min are summarized. Since the control potentiation of the PS amplitude reached much higher values than the fEPSP slope function, the absolute inhibition of potentiation appears to be much higher. Nevertheless, as revealed by the comparison of the relative percentage of inhibition, this holds true only for 2.7 µg/ml Al, for which even the relative inhibition by Al is much larger. For 0.68 µg/ml Al, the relative inhibition of potentiation is in the same range. Statistical analysis (overall ANOVA for t 5 3 to t 5 60 min) revealed highly significant differences between groups for both PS amplitude [F(4, 40) 5 10.5, P , 0.01] and fEPSP slope [F(4, 40) 5 12.5, P , 0.01]. A post hoc unpaired t test was performed, the results of which are given in Table 1 for t 5 60 min, showing the concentration dependence of the action of Al on TEA LTP. There was no influence of the different application schedules, but there were significant differences between the two concentrations used. 4. DISCUSSION
Our data demonstrate that Al inhibits TEA LTP in a concentration-dependent manner. For the lower concentration (0.68 µg/ml Al) the potentiation was attenuated equally whether Al was preapplied or was only coapplied with TEA. In both cases, a stronger eLTP compared to controls could be induced subsequently. We therefore conclude that the inhibition of the induction of TEA LTP by 0.68 µg/ml Al allowed a stronger subsequent eLTP since the corresponding synapses were not as strongly prepotentiated by TEA. Obviously, the presence of the lower Al concentration did not severely and irreversibly inhibit components necessary for eLTP, even when Al was applied for more than 20 min, since the amount of potentiation (TEA LTP and eLTP) was similar for both application regimes. It can therefore be concluded that mechanisms underlying neuronal plasticity are temporarily and reversibly influ-
245
enced by low Al concentrations before effects on basic synaptic transmission are observed and before any profound alterations of neuronal properties occur. Similar results have been obtained in our previous report on the action of Al on eLTP (29). In this study, we found that in vivo eLTP is already inhibited by Al before any effects on baseline responses were observed and before any morphological alterations are obtained histologically. When the higher Al concentration (2.7 µg/ml) was superfused, a weak but significant inhibition of baseline responses was obtained. This inhibition was found to be irreversible, which provides evidence for the interaction of Al with VDCCs, since the action of Al on these channels has been demonstrated to be only partially reversible (maximum 25%; Refs. 7, 27). Thus, it must be taken into account that a general reduction of excitability may have contributed to the depression of both TEA LTP and eLTP. It seems, however, unlikely that this is the major reason for the inhibition, since a reduced potentiation was already observed for 0.68 µg/ml Al, and this concentration did not cause a significant baseline effect. Nevertheless, the nonspecific depression of the fEPSP may have additionally contributed to the stronger inhibition of TEA LTP by 2.7 µg/ml Al. Interestingly, the reduction of TEA LTP was more pronounced when no preapplication of Al was performed. In this case, PS amplitude and fEPSP slope data even declined significantly below baseline whereas data remained at baseline levels when Al preapplication was conducted. This suggests that the preapplication of Al allowed an equilibration of the slices in the Al-containing solution, which caused an adjustment of the neurons and, as a consequence, a weaker inhibition of TEA LTP. It is possible that a partial inhibition of VDCCs during equilibration had protected the neurons. The action of Al may thus be stronger without preapplication since in these circumstances the massive activation of ion channels due to TEA application may provide better access for Al to enter the respective channels. When VDCCs are partially inhibited by preapplication, only the remaining channels may be activated during TEA application. Although this causes a weaker depolarization and thus an attenuated potentiation, it may protect the neurons from subsequent damaging processes. This hypothesis is in line with our finding about the use-dependent inhibition of VDCCs by Al (6, 27), since during preapplication Al can affect only the channels activated by the control pulses but during TEA-induced depolarization a much larger number of VDCCs are susceptible to Al. While preapplication of Al partially depressed the strong depolarization caused by TEA, glutamate release is reduced, leading to a weaker neurotoxic action of Al. Correspondingly, it has been shown that Al decreases Ca21-dependent glutamate release in hippocampal slices (31).
246
PLATT AND REYMANN
We suggest that both the baseline effect and the inhibition of TEA LTP may be caused by the action of Al on VDCCs. Although it must be considered that the data concerning the action of Al on VDCCs were obtained using a different preparation (cultured dorsal root ganglion neurons), the action of Al as a general calcium channel blocker is likely to account for the inhibition of TEA LTP. For TEA LTP—rather than for eLTP—there is profound evidence for the contribution of VDCCs (1, 14–16). Thus, the inhibition of calciumdependent processes is the most probable mechanism for Al neurotoxicity. Additional mechanisms including interactions with the cholinergic system (36, 24) may also have contributed to the observed action of Al. Concerning the comparison of the action of Al on the fEPSP slope versus PS amplitude, there is evidence that the PS is more susceptible to Al than the fEPSP: First, both the baseline effect and the decline below baseline after TEA LTP induced in the presence of 2.7 µg/ml Al were more pronounced for the PS amplitude. Second, the expression of the subsequent eLTP of the PS amplitude was already much weaker after the brief exposure to 2.7 µg/ml Al and TEA LTP (see Fig. 4). For the fEPSP slope it appeared that the stronger inhibition of TEA LTP allowed a stronger expression of the subsequent eLTP; only preapplication of 2.7 µg/ml Al depressed eLTP. Third, as revealed by the comparison of the inhibition with both the baseline in the absence and in the presence of 2.7 µg/ml Al, this concentration caused a stronger inhibition of PS amplitude potentiation compared to fEPSP potentiation. We therefore conclude that the more pronounced inhibition of the PS amplitude is based not only on the reduced depolarization due to inhibited fEPSPs but also on other mechanisms necessary for spike generation. The prolonged exposure to 2.7 µg/ml Al preventing the induction of eLTP of PS amplitude and suppressing eLTP of the fEPSP slope indicates that additional slowly developing neurotoxic processes have occurred, for example via accumulation of Al within the cells. Such mechanisms may have caused irreversible damages disabling the induction of eLTP. Additionally, the irreversible block of VDCCs may have caused a general depression of excitability sufficient to suppress eLTP. Our studies on VDCCs provided evidence for a slow time-course of equilibration and evidence for at least one Al binding site within the channel pores of VDCCs (6, 7, 27). These characteristics may well account for the inhibition of eLTP 60 min after exposure to Al and for long-lasting neurotoxic actions of Al in the nervous system. Nevertheless, a variety of other neurochemical actions of Al have to be considered, such as its influence on second messenger systems like inositolphosphate, cycliy-adenosine-monophosphate and the phosphorylation of microtubule-associated protein MAP-2. These actions have been demonstrated both in vivo after
chronic exposure (20) and in vitro (19). Since secondmessenger cascades are crucial for eLTP (3, 33), it is likely that an interference of Al with these components may have prevented eLTP induction after longer application of 2.7 µg/ml Al. Compared to the results of our previous study on eLTP (28, 29), the general characteristics of Al in blocking TEA LTP are similar to those obtained for eLTP, i.e., 0.68 µg/ml caused an attenuation, whereas the inhibition of eLTP by 2.7 µg/ml was more pronounced and potentiation declined back to baseline values within 60 min. The observation that Al inhibits TEA LTP in the same concentration range and with a similar time-course as eLTP provides further evidence that both models may share common mechanisms. In this context, the most obvious candidate is Ca21 as an intracellular trigger for plastic events. Although the contribution of VDCCs to eLTP in CA1 induced by 100 Hz HFS is not as obvious as it is for TEA LTP, we have shown that Al also inhibits NMDA-receptor mediated currents of acutely isolated hippocampal neurons in the same concentration range (30). Thus, an interference of Al with the Ca21 influx through NMDA receptors is a more likely mechanism for eLTP inhibition. The action of Al on TEA LTP provides clear evidence that Al interferes with mechanisms underlying neuronal plasticity which require Ca21 as a major intracellular trigger. We therefore suggest that although Al inhibits eLTP via a different mechanism than TEA LTP, the Ca21-dependent processes downstream from the induction mechanisms are affected in a comparable way. Such mechanisms may be crucial for many cognitive dysfunctions observed after Al exposure and for neuronal diseases in which Al is a etiologic or contributing factor. ACKNOWLEDGMENTS The authors thank Ms. K. Bo¨hm for excellent technical assistance and Dr. Gernot Riedel for critical comments on an earlier version of the manuscript.
REFERENCES 1.
ANIKSZTEJN, L., AND Y. BEN-ARI. 1991. Novel form of long-term potentiation produced by K1 channel blocker in the hippocampus. Nature 349: 67–69. 2. BERNARD, J., LAHSAINI, A., AND G. MASSICOTTE. 1994. Potassiuminduced long-term potentiation in area CA1 of the hippocampus involves phospholipase activation. Hippocampus 4(4): 447–453. 3. BLISS, T. V. P., AND G. L. COLLINGRIDGE. 1993. A synaptic model for memory: Long-term potentiation in the hippocampus. Nature 361: 31–39. 4. BLISS, T. V. P., ERRINGTON, M. L., AND LYNCH, M. A. 1988. Calcium-induced long-term potentiation in the dentate gyrus is accompanied by a sustained increase in glutamate release. In Excitatory Amino Acid Transmission (T. P. Hicks, D. Lodge, and H. McLennan, Eds.), pp. 337–340. A. R. Liss, New York.
ALUMINUM INHIBITS TEA LTP 5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20. 21.
22.
BOLLA, K. I., BRIEFEL, G., SPECTOR, D., SCHWARTZ, B. S., WIELER, L., HERRON, J., AND L. GIMNENEZ. 1992. Neurocognitive effects of aluminum. Arch. Neurol. 49: 1021–1026. BU¨ SSELBERG, D., PLATT, B., HAAS, H. L., AND D. O. CARPENTER. 1994. Mammalian voltage-activated calcium channel currents are blocked by Pb21, Zn21 and Al31. J. Neurophysiol. 71(4): 1491–1497. BU¨ SSELBERG, D., PLATT, B., HAAS, H. L., AND D. O. CARPENTER. 1993. Voltage-gated calcium channel currents of rat dorsal root ganglion (DRG) are blocked by Al31. Brain Res. 622: 163–168. FARNELL, B. J., U. DE BONI, AND D. R. CRAPPER MCLACHLAN. 1982. Aluminum neurotoxicity in the absence of neurofibrillary degeneration in CA1 hippocampal pyramidal neurons in vitro. Exp. Neurol. 78: 241–258. FLECK, M. W., PALMER, A. M., AND G. BARRIONUEVO. 1992. Potassium-induced long-term potentiation in rat hippocampal slices. Brain Res. 580: 100–105. GANROT, P. O. 1986. Metabolism and possible health effects of aluminum. Environ. Health Perspect. 65: 363–441. GARRUTO, R. M. 1991. Pacific paradigms of environmentallyinduced neurological disorders: Clinical, epidemiological and molecular perspectives. Neurotoxicology 12: 347–377. GROVER, L. M., AND T. J. TEYLER. 1990a. Two components of long-term potentiation induced by different patterns of afferent activation. Nature 347: 477–479. GROVER, L. M., AND T. J. TEYLER. 1990b. Effects of extracellular potassium concentrations and postsynaptic membrane potential on calcium-induced potentiation in area CA1 of rat hippocampus. Brain Res. 506: 53–61. HANSE, E., AND B. GUSTAFSSON. 1994. TEA elicits two distinct potentiations of synaptic transmission in the CA1 region of hippocampal slices. J. Neurosci. 14: 5028–5034. HUANG, Y.-Y., AND R. C. MALENKA. 1993. Examination of TEAinduced synaptic enhancement in area CA1 of the hippocampus: The role of voltage-dependent Ca21 channels in the induction of LTP. J. Neurosci. 13(2): 568–576. HUBER, K. M., MAUK, M. D., AND P. T. KELLY. 1995. Distinct LTP induction mechanisms: Contribution of NMDA receptors and voltage-dependent calcium channels. J. Neurophysiol. 73: 270– 279. JEFFREY, E. H. Biochemical mechanisms of aluminum toxicity. In Handbook of Experimental Pharmacology (R. A. Goyer and M. G. Cherian, Eds.), pp. 139–161. Springer, Berlin. JOHNSTON, D., WILLIAMS, S., JAFFE, D., AND R. GRAY. 1992. NMDA-receptor-independent long-term potentiation. Annu. Rev. Physiol. 54: 489–505. JOHNSON, G. V., COGDILL, K. W., AND R. S. JOPE. 1990. Oral aluminum alters in vitro protein phosphorylation and kinase activities in rat brain. Neurobiol. Aging 11: 209–216. JOPE, R. S., AND G. V. JOHNSON. 1992. Neurotoxic effects of dietary aluminum. Cibia Found. Symp. 169: 254–267. KRAUS, A. S., AND W. F. FORBES. 1992. Aluminum, fluorid and the prevention of Alzheimer’s disease. Can. J. Public Health 83(2): 97–100. MALENKA, R. C., KAUER, J. A., ZUCKER, R. S., AND R. A. NICOLL.
23. 24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
247
Postsynaptic calcium is sufficient for potentiation of hippocampal synaptic transmission. Science 242: 81–84. MARTIN, R. B. 1986. The chemistry of aluminum as related to biology and medicine. Clin. Chem. 32(10): 1797–1806. PENG, J. H., XU, Z. C., XU, Z. X., PARKER, J. C., FRIEDLANDER, E. R., TANG, J. P., AND S. MELETHIL. 1992. Aluminum-induced acute cholinergic neurotoxicity in rat. Mol. Chem. Neuropathol. 17(1): 79–89. PERL, P. D., AND A. K. BRODY. 1980. X-ray spectometric evidence of aluminum accumulation in neurofibrillary tangle-bearing neurons. Science 208: 297–299. PETROZZINO, J. J., AND J. A. CONNOR. 1994. Dendritic Ca21 accumulations and metabotropic glutamate receptor activation with an N-methyl-D-aspartate receptor-independent long-term potentiation in hippocampal CA1 neurons. Hippocampus, 4(5): 546–558. PLATT, B., AND D. BU¨ SSELBERG. 1994. Actions of aluminum on voltage-activated calcium channel currents. Cell. Mol. Neurobiol. 14(6): 819–829. PLATT, B., BU¨ SSELBERG, D., AND D. O. CARPENTER. 1994. Actions of aluminum on synaptic transmission and long-term potentiation of CA1 neurons in hippocampal slices. Soc. Neurosci. Abstr. 1516. PLATT, B., CARPENTER, D. O., BU¨ SSELBERG, D., REYMANN, K. G., AND G. RIEDEL. 1995. Aluminium impairs hippocampal longterm potentiation in rats in vitro and in vivo. Exp. Neurol. 134: 73–86. PLATT, B., HAAS, H., BU¨ SSELBERG, D. 1994. Aluminium reduces glutamate activated currents of rat hippocampal neurons. NeuroReport 5(17): 2329–2332. PROVAN, S. D., AND R. A. YOKEL. 1992. Aluminium inhibits glutamate release from transverse rat hippocampal slices: role of G proteins, Ca channels and protein kinase C. Neurotoxicology 13: 413–420. REYMANN, K. G., MATTHIES, H. K., FREY, U., VOROBYEV, V. S., AND H. MATTHIES. 1986. Calcium-induced long-term potentiation in the hippocampal slice: Characterization of the time course and conditions. Brain Res. Bull. 17: 291–296. REYMANN, K. G., AND S. STAAK. 1994. Molecular mechanisms underlying long-term potentiation: Postsynaptic glutamate receptors and protein kinase C. In Protein Kinase C in the CNS Focus on Neuronal Plasticity (P. L. Canonico, U. Scapagnini, U. Pamparana, and A. Routtenberg, Eds.), pp. 7–32. Masson Press, Milan, Italy. TURNER, R. W., BAIMBRIDGE, K. G., AND J. J. MILLER. 1982. Calcium-induced long-term potentiation in the hippocampus. Neuroscience 7(6): 1411–1416. YOKEL, R. A. 1994. Aluminum exposure produces learning and memory deficits: A model of Alzheimer’s disease. In Toxininduced Models of Neurological Disorders (M. L. Woodruff and A. J. Nonneman Eds.), pp. 301–318. Plenum, New York. ZATTA, P., ZAMBENEDETTI, P., BRUNA, V., AND B. FILIPPI. 1994. Activation of acetylcholinesterase by aluminum (III): the relevance of the metal species. NeuroReport 5(14): 1777–1780.
141, 240–247 (1996)
0158
Inhibition of TEA-Induced LTP by Aluminum BETTINA PLATT1
AND
KLAUS G. REYMANN
Federal Institute for Neurobiology, Department Neurophysiology, P. O. Box 1860, D-39008 Magdeburg, Germany
Brief application of tetraethylammonium (TEA) to hippocampal slices causes long-term potentiation (TEA LTP) at synapses of CA1 pyramidal neurons characterized by a long-lasting increase of field excitatory postsynaptic potential (fEPSP) slope and population spike (PS) amplitude. Since this kind of potentiation requires the activation of voltage-dependent calcium channels, we examined the effect of the inorganic calcium channel blocker aluminum, which has been shown to impair tetanus-induced LTP (eLTP). We found that Al inhibited in a concentration-dependent manner both fEPSP slope and PS amplitude potentiation by TEA; 0.68 mg/ml Al attenuated TEA LTP, while a complete block of long-lasting potentiation was obtained for 2.7 mg/ml Al. Occlusion experiments revealed that both concentrations of Al allowed the induction of eLTP 60 min after TEA/Al exposure. However, longer application (15 min) of 2.7 mg/ml Al before the induction of TEA LTP prevented the subsequent induction of eLTP although no significant differences concerning the action on TEA LTP were observed. This indicates a general loss of neuronal plasticity which might be due to progressive neuronal cell damage. Since the effective concentration range of Al is directly comparable to the action of Al on eLTP, our data provide evidence for shared mechanisms of both potentiations. Although based on different induction mechanisms, Ca21 is assumed to be a general intracellular trigger for both forms of LTP and thus it can be hypothesized that the neurotoxic action of Al is due to interference with Ca21-dependent processes by inhibition of calcium conductances. r 1996 Academic Press, Inc.
1. INTRODUCTION
Aluminum has been reported to be a possible etiologic factor in Alzheimer’s disease (21, 25) and other neurological disorders such as Parkinson dementia of Guam and dialysis syndrome (10, 11). Furthermore, exposure to Al causes learning and memory deficits in
1 To whom correspondence should be addressed at present address: Department of Physiology, University of Leeds, Worsley Medical and Dental Building, Leeds LS2 9NQ, Great Britain, Fax: (144) 1132334228, E-mail: [email protected].
0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
animals and humans (5, 17, 35). Experimental evidence for a direct interference of Al with mechanisms underlying neuronal excitability and plasticity has been provided by means of the model of long-term potentiation (LTP). LTP evoked by electrical stimulation of afferent fibers (termed eLTP) is a model frequently used to study mechanisms underlying synaptic plasticity (3). There are at least two forms of eLTP distinguished by the involvement of N-methyl-D-aspartate (NMDA) receptors, i.e., NMDA receptor-dependent and NMDA receptor-independent eLTP. The contribution of NMDA receptors appears to depend on the brain area and on the tetanus used for eLTP induction. The involvement of voltage-dependent calcium channels (VDCCs) has been proven for NMDA receptor-independent eLTP (12, 18). In addition to this ‘‘classical’’ LTP model, a variety of investigations were performed on chemically induced LTP, elicited for instance by application of high Ca21 and/or K1 concentrations, or by means of K1 channel blockers (2, 9, 13, 32, 34). Although the applied substances determine the nature of the elicited potentiation, all models are assumed to be based on postsynaptic depolarization in conjunction with enhanced presynaptic transmitter release and it is generally accepted that an increase of postsynaptic Ca21 is the major trigger for LTP (22). In a variety of investigations, the K1 channel blocker tetraethylammonium (TEA) has been used to characterize this kind of synaptic enhancement (termed TEA LTP) and also to compare it with the mechanisms underlying eLTP (1, 15, 16). Brief application of TEA causes a depolarization based on the inhibition of various K1 conductances, of which the delayed rectifier is assumed to be the crucial one (1). As pointed out by Petrozzino and Connor (26), application of TEA induces a more prolonged and pronounced increase of intracellular Ca21 compared to eLTP. Using the model of eLTP to study Al neurotoxicity, it has been shown that Al inhibits in vitro eLTP after in vivo exposure (8), and recently, we demonstrated a direct impairment of in vivo and in vitro eLTP by Al before any obvious morphological changes occur and even before basic electrophysiological parameters were affected (28, 29). In the present study, we examined the action of Al on TEA LTP as a comparative approach. Furthermore, Al acts as an inhibitor of both high- and
240
241
ALUMINUM INHIBITS TEA LTP
low-voltage activated calcium channels of cultured dorsal root ganglion neurons (7, 27) and the contribution of VDCCs in TEA LTP has been demonstrated by previous reports (1, 15, 16). The action of Al on TEA LTP is therefore of particular interest with respect to answering two major questions. First, using basically the same preparation and Al concentrations as in our previous study on eLTP, we may be able to detect further similarities and differences of TEA LTP versus eLTP. Second, the action of Al on TEA LTP may provide useful insight into the mechanisms causing cognitive dysfunctions due to Al exposure since an interaction with TEA LTP would support the hypothesis that Al interferes mainly with the Ca21 homeostasis of neurons and Ca21-dependent processes underlying neuronal plasticity. 2. MATERIALS AND METHODS
Experiments were performed on hippocampal slices (400 µm) from adult male Wistar rats (7–8 weeks old) after equilibration in a standard Ringer for at least 1 h at 32°C. Slices were placed in a submersion chamber and superfused (3 ml/min) with a modified Ringer solution for 30 min (composition in mM : NaCl 126, KCl 5, MgCl2 1.3, CaCl2 2.5, Tris–Cl 6, glucose 10, NaHCO3 15; saturated with 95% O2/5% CO2, pH 7.4, 32°C). The use of a modified Ringer was essential in order to prevent precipitation of Al as sulfate and phosphate compounds (23). Field potentials were elicited by stimulation of the Schaffer collateral/commissural fibers and extracellularly recorded in both stratum radiatum and stratum pyramidale of area CA1 using ,1 MV glass electrodes filled with modified Ringer. Every 3 min test stimuli (three bipolar pulses, 0.2-ms duration, 10-s interval) were applied via a monopolar stainless steel electrode. The initial slope of the rising phase of the averaged field excitatory postsynaptic potential (fEPSP) and the amplitude of the population spike (PS) were estimated. The test stimulus intensity was set at 30% of fEPSP maximum, which elicited a PS of about 20% of the maximum. All compounds except Al were obtained from Sigma; Al was taken from an ‘‘aluminum atomic absorption standard solution’’ (984 µg/ml of Al in 1 wt% HCl, Aldrich). Two different concentrations of Al, i.e., 0.68 and 2.7 µg/ml Al, were investigated. The concentrations of Al are given here in units of micrograms per milliliter, since Al forms a variety of species in solution and the relative molarities are not known. Assuming a complete dissociation of the trivalent cation, 0.68 µg/ml corresponds to 25 µM and 2.7 µg/ml corresponds to 100 µM Al31. These concentrations were used to allow a comparison with our previous work on eLTP (28, 29). Here we performed two different application schedules with both concentrations. The first series was con-
ducted without preapplication of Al alone; TEA and Al were applied simultaneously (7 min) in order to allow the analysis of the action of Al on the induction of TEA LTP without interference with general neurotoxic processes. In a second series of experiments, a preapplication of Al (15 min) was performed followed by the application of Al in conjunction with TEA (7 min). This schedule was chosen to investigate whether a longer exposure to Al induces a more pronounced inhibition of potentiation by Al. Accordingly, control experiments on the reversibility of baseline effects caused by Al application were performed by applying 2.7 µg/ml Al for 22 min, and the recovery after wash was monitored for 60 min. Sixty minutes after TEA application, a reset of the stimulation strength was performed to obtain fEPSPs comparable to baseline responses. As a subsequent test for plasticity and occlusion of eLTP by TEA LTP, eLTP was induced by three high-frequency stimuli (HFS; 100 Hz for 1 s; interburst interval, 20 s). All data reported here are given as means 6 SEM and potentiation is presented as percentage change relative to baseline. Statistical significance was revealed by analysis of variance (ANOVA) and subsequent t test on the basis of P # 0.05 for significant and #0.01 for highly significant differences. Values obtained 60 min after TEA application were taken as a measure of potentiation and were used for comparison between groups. 3. RESULTS
In control experiments (n 5 14), superfusion of 25 mM TEA for 7 min induced a sustained increase of the PS amplitude (Fig. 1A) and fEPSP slope (Fig. 1B) after washout of TEA. During TEA application, we observed an initial increase of the fEPSP slope, followed by a feigned transient decrease due to interference of the prolonged fiber volley with the measurement of slope function. Sixty minutes after TEA application, PS amplitude and fEPSP were potentiated up to 266 6 8 and 138 6 5%, respectively. Concerning the PS data, it should be mentioned that all reports published so far on TEA LTP were based only on fEPSP data in the stratum radiatum so that a comparison with data of others is not possible. Thus, this is the first report on PS amplitude potentiation by TEA. Since our last report on the action of Al on eLTP was performed using the PS amplitude as a measure (30), both parameters were recorded and analyzed in the present study. To obtain responses in the range of baseline values a subsequent reset of the stimulus strength was performed. HFS application resulted in a further potentiation (PS amplitude: 178 6 7%; fEPSP: 113 6 5%, 12 min after HFS). In general, the obtained amount of eLTP was weaker than that observed in naive slices
242
PLATT AND REYMANN
FIG. 1. Application of 25 mM TEA for 7 min (indicated by a solid bar) induces an increase of the population spike (PS) amplitude (A) and field excitatory postsynaptic potential (fEPSP) slope function (B). Averaged values 6 SEM (n 5 14) are plotted versus time relative to baseline. After reset of the stimulation strength (single arrow) and HFS application (double arrow), a further potentiation could be induced. Sample traces at time points indicated in the graphs are given on the right.
(Ref. 29, PS amplitude: 250%, 15–60 min after HFS), indicating a partial occlusion of eLTP by TEA LTP. Application of 0.68 µg/ml Al in conjunction with TEA (n 5 7) led to a stronger and longer lasting initial depression of the fEPSP slope (Fig. 2B) compared to
FIG. 2. Action of 0.68 µg/ml Al on TEA LTP (n 5 7). Time-course of mean values calculated relative to baseline is given. Application of TEA and 0.68 µg/ml Al (solid bar) caused a significantly attenuated potentiation of both PS amplitude (A) and fEPSP slope (B) compared to control TEA LTP. Subsequent reset of the stimulus strength (single arrow) and HFS (double arrow) potentiated both components. On the right, sample traces for the time points indicated in the graphs are illustrated.
controls. Other than a general inhibition there was no alteration observed of the time course of the PS amplitude during TEA application (Fig. 2A). A significantly attenuated potentiation (Table 1) developed, and after washout PS amplitude and fEPSP slope were potentiated to 189 6 8 and 119 6 6% (t 5 60 min), respectively. Reset of the stimulus strength and subsequent HFS
TABLE 1 Statistical Comparison at t 5 60 min t values Control 0.68 µg/ml 0.68 µg/ml (pre) 2.7 µg/ml
0.68 µg/ml PS ampl./EPSP slope
0.68 µg/ml (pre) PS ampl./EPSP slope
2.7 µg/ml PS ampl./EPSP slope
2.7 µg/ml (pre) PS ampl./EPSP slope
2.54(*)/2.37(*)
2.19(*)/2.24(*) 0.33(ns)/0.21(ns)
5.94(**)/5.48(**) 2.95(**)/2.92(*) 3.31(**)/3.24(**)
3.84(**)/4.72(**) 2.38(*)/2.27(*) 2.01(*)/2.56(*) 1.37(ns)/1.64(ns)
Note. pre, preapplication (15 min) of the respective Al concentration before induction of TEA LTP. * P , 0.05; ** P , 0.01; ns, not significant.
ALUMINUM INHIBITS TEA LTP
243
values (PS amplitude: t 5 2.53, P , 0.05; fEPSP: t 5 1.63, P . 0.05). Since no significant differences to the results shown in Fig. 2 were found (see Table 1), it appears that the action of 0.68 µg/ml Al on TEA LTP was not influenced by the preapplication of 0.68 µg/ml Al. When 2.7 µg/ml Al was applied simultaneously with TEA (n 5 8), a similar time course as for 0.68 µg/ml Al was obtained during application, but the initial depression was found to be prolonged for the fEPSP slope function (Fig. 4). A stronger inhibition of potentiation occurred with a significant decline below baseline values for both PS amplitude and fEPSP slope (t 5 60 min: PS amplitude: t 5 2.6, P , 0.05; fEPSP: t 5 2.5, P , 0.05). The reduction was highly significant compared to controls and compared to the action of 0.68 µg/ml Al at t 5 60 min (Table 1) with a PS amplitude of 69 6 4% and a mean fEPSP slope of 90 6 6%. In order to
FIG. 3. Preapplication of 0.68 µg/ml Al had no influence on the action of Al on TEA LTP (n 5 9). Averaged data are shown relative to baseline and plotted versus time. A slight but not significant inhibition of the baseline was observed during Al application alone (hatched bar). After coapplication with TEA (solid bar) potentiation of PS amplitude (A) and fEPSP slope (B) were highly significantly attenuated compared to controls. eLTP could be induced after reset (single arrow) and HFS (double arrow). Note the similarity to data shown in Fig. 2. Sample traces for time points in the graphs are illustrated on the right.
caused a clear potentiation of both PS amplitude (273 6 11%) and fEPSP slope (120 6 5%) 12 min after tetanus, which was more pronounced than under control conditions but only significantly different for PS amplitude data (PS amplitude: t 5 2.67; fEPSP: t 5 0.64, P . 0.05; P , 0.05). Al (0.68 µg/ml) applied 15 min before superfusion coincident with TEA (n 5 9) caused a slight but not significant reduction in baseline responses (Fig. 3). The potentiation obtained 60 min after TEA application was significantly reduced compared to controls (Table 1) with values of 175 6 20% (PS amplitude) and 121 6 6% (fEPSP slope). Subsequent reset and tetanization resulted in a potentiation of the PS amplitude (248 6 20%) and fEPSP slope (128 6 5%) 12 min after HFS. Again, the amount of eLTP obtained was significantly greater than under control conditions for the PS amplitude
FIG. 4. 2.7 µg/ml Al blocks TEA LTP (n 5 8). Time-courses of mean data calculated relative to baseline are shown. Application of TEA and Al (solid bar) caused a severe depression of TEA LTP, which was highly significantly reduced in comparison with controls. Both PS amplitude (A) and fEPSP slope (B) declined below baseline. Subsequent HFS (double arrow) induced a potentiation which was much greater for the fEPSP slope function compared to control data. On the right, representative raw traces are shown for the respective time points indicated in the graphs.
244
PLATT AND REYMANN
keep the time-course comparable to control experiments, HFS was applied after t 5 69 min, resulting in a clear potentiation of PS amplitude (179 6 18%) and fEPSP slope (152 6 10%) 12 min after tetanus. Compared to control conditions, the induced eLTP was significantly stronger in terms of the fEPSP slope, whereas the PS amplitude was potentiated in the same range as controls (fEPSP: t 5 3.774, P , 0.05; PS amplitude: t 5 0.005, P . 0.05). It appears that the application of 2.7 µg/ml Al may have a stronger and longer lasting effect on the PS which is supported by the stronger decline after TEA superfusion. For the fEPSP, it seems that stronger inhibition of TEA LTP by 2.7 µg/ml Al allowed a stronger subsequent eLTP. Application of 2.7 µg/ml Al alone (n 5 7; Fig. 5) led to a significant baseline effect; i.e., the PS amplitude and fEPSP slope were reduced by 12 6 3 and 8 6 1% after 15 min, respectively (PS amplitude: t 5 2.58, P , 0.05; fEPSP: t 5 3.6 P , 0.05). As already pointed out in our report on eLTP, the action of Al on the PS is not reversible after wash (28). Since we obtained differences in the action of Al on the fEPSP and PS (see above), we performed a series of control experiments in order to investigate the reversibility of the baseline effect on the fEPSP slope (n 5 8; see inset in Fig. 5B). In these experiments, the depression of the fEPSP slope by application of 2.7 µg/ml Al for 22 min (11 6 2%) recovered slightly after wash; an inhibition of 8 6 4% remained. This finding suggests that the action of 2.7 µg/ml Al causes a slight but sustained suppression of the fEPSP. When TEA was applied in conjunction with 2.7 µg/ml Al after preexposure, no initial increase in the fEPSP slope was obtained (Fig. 5B, main graph) when TEA application started and the enhancement of the PS amplitude was comparably small. The emerging potentiation after wash was weak and declined back to baseline within 45 min. At t 5 60 min, a mean PS amplitude of 107 6 8% and fEPSP slope of 106 6 5% were found. A comparison with control data revealed a highly significant difference of both PS amplitude and fEPSP slope data (see Table 1) and no significant difference from baseline data (PS amplitude: t 5 1.9, P . 0.05; fEPSP: t 5 2.1, P . 0.05). Nevertheless, comparison with the baseline values in the presence of 2.7 µg/ml Al indicated a significant potentiation of the fEPSP slope, but not of the PS (PS amplitude: t 5 1.9, P . 0.05; fEPSP slope: t 5 2.5, P , 0.05). Again, the effect on the PS appears to be stronger compared to the effect on the fEPSP. Group comparison between data obtained in the experiments with preapplication of 2.7 µg/ml Al versus those without preapplication gave no significant differences (see Table 1). Only in those experiments where a slight potentiation remained at t 5 60 min, was a reset of the stimulation strength performed. Unlike all other experiments, HFS did not result in a further potentiation.
FIG. 5. Preapplication of 2.7 µg/ml Al (n 5 7) inhibited synaptic transmission slightly but significantly as shown by the baseline effect on both PS amplitude (A) and fEPSP slope (B). Control experiments (n 5 8) revealed that the inhibition of the fEPSP slope by 2.7 µg/ml Al is only slightly reversible after wash (inset in B). Simultaneous superfusion of TEA and Al led to a highly significant reduction of potentiation after wash compared to control experiments. No eLTP could be induced after reset of the stimulus intensity (single arrow) and HFS (double arrow), indicating irreversible and/or damaging actions after prolonged superfusion of 2.7 µg/ml Al. Sample traces are given on the right for the time points indicated in the graphs.
Fifteen minutes after HFS, PS amplitude (98 6 6%) and slope values (101 6 5%) were found to be not significantly different from baseline data (PS amplitude: t 5 1.15, P . 0.05; fEPSP: t 5 1.57, P . 0.05) but significantly different from the data obtained without preapplication of 2.7 µg/ml Al (PS amplitude: t 5 2.28, P , 0.05; fEPSP: t 5 5.5, P , 0.01). Thus, it is likely that the longer application of 2.7 µg/ml Al resulted in an irreversible inhibition of components essential for eLTP induction. In Table 2 the absolute amount of potentiation in the presence of 0.68 and 2.7 µg/ml Al relative to baseline values and the relative inhibition of potentiation com-
ALUMINUM INHIBITS TEA LTP
TABLE 2 Inhibition of TEA LTP by 0.68 and 2.7 µg/ml Al: Absolute Values of Potentiation Relative to Baseline and Inhibition of Potentiation Relative to Control Data (in %) EPSP slope
PS amplitude
30 min 60 min 30 min 60 min absolute absolute absolute absolute pot./relative pot./relative pot./relative pot./relative inhibition inhibition inhibition inhibition 0.68 µg/ml 0.68 µg/ml (pre) 2.7 µg/ml 2.7 µg/ml (pre)
127/45 133/32 111/77 119/61
119/50 121/45 89/129 106/84
239/53 209/63 110/96 136/88
189/46 175/52 69/142 107/95
pared to control data at t 5 30 and t 5 60 min are summarized. Since the control potentiation of the PS amplitude reached much higher values than the fEPSP slope function, the absolute inhibition of potentiation appears to be much higher. Nevertheless, as revealed by the comparison of the relative percentage of inhibition, this holds true only for 2.7 µg/ml Al, for which even the relative inhibition by Al is much larger. For 0.68 µg/ml Al, the relative inhibition of potentiation is in the same range. Statistical analysis (overall ANOVA for t 5 3 to t 5 60 min) revealed highly significant differences between groups for both PS amplitude [F(4, 40) 5 10.5, P , 0.01] and fEPSP slope [F(4, 40) 5 12.5, P , 0.01]. A post hoc unpaired t test was performed, the results of which are given in Table 1 for t 5 60 min, showing the concentration dependence of the action of Al on TEA LTP. There was no influence of the different application schedules, but there were significant differences between the two concentrations used. 4. DISCUSSION
Our data demonstrate that Al inhibits TEA LTP in a concentration-dependent manner. For the lower concentration (0.68 µg/ml Al) the potentiation was attenuated equally whether Al was preapplied or was only coapplied with TEA. In both cases, a stronger eLTP compared to controls could be induced subsequently. We therefore conclude that the inhibition of the induction of TEA LTP by 0.68 µg/ml Al allowed a stronger subsequent eLTP since the corresponding synapses were not as strongly prepotentiated by TEA. Obviously, the presence of the lower Al concentration did not severely and irreversibly inhibit components necessary for eLTP, even when Al was applied for more than 20 min, since the amount of potentiation (TEA LTP and eLTP) was similar for both application regimes. It can therefore be concluded that mechanisms underlying neuronal plasticity are temporarily and reversibly influ-
245
enced by low Al concentrations before effects on basic synaptic transmission are observed and before any profound alterations of neuronal properties occur. Similar results have been obtained in our previous report on the action of Al on eLTP (29). In this study, we found that in vivo eLTP is already inhibited by Al before any effects on baseline responses were observed and before any morphological alterations are obtained histologically. When the higher Al concentration (2.7 µg/ml) was superfused, a weak but significant inhibition of baseline responses was obtained. This inhibition was found to be irreversible, which provides evidence for the interaction of Al with VDCCs, since the action of Al on these channels has been demonstrated to be only partially reversible (maximum 25%; Refs. 7, 27). Thus, it must be taken into account that a general reduction of excitability may have contributed to the depression of both TEA LTP and eLTP. It seems, however, unlikely that this is the major reason for the inhibition, since a reduced potentiation was already observed for 0.68 µg/ml Al, and this concentration did not cause a significant baseline effect. Nevertheless, the nonspecific depression of the fEPSP may have additionally contributed to the stronger inhibition of TEA LTP by 2.7 µg/ml Al. Interestingly, the reduction of TEA LTP was more pronounced when no preapplication of Al was performed. In this case, PS amplitude and fEPSP slope data even declined significantly below baseline whereas data remained at baseline levels when Al preapplication was conducted. This suggests that the preapplication of Al allowed an equilibration of the slices in the Al-containing solution, which caused an adjustment of the neurons and, as a consequence, a weaker inhibition of TEA LTP. It is possible that a partial inhibition of VDCCs during equilibration had protected the neurons. The action of Al may thus be stronger without preapplication since in these circumstances the massive activation of ion channels due to TEA application may provide better access for Al to enter the respective channels. When VDCCs are partially inhibited by preapplication, only the remaining channels may be activated during TEA application. Although this causes a weaker depolarization and thus an attenuated potentiation, it may protect the neurons from subsequent damaging processes. This hypothesis is in line with our finding about the use-dependent inhibition of VDCCs by Al (6, 27), since during preapplication Al can affect only the channels activated by the control pulses but during TEA-induced depolarization a much larger number of VDCCs are susceptible to Al. While preapplication of Al partially depressed the strong depolarization caused by TEA, glutamate release is reduced, leading to a weaker neurotoxic action of Al. Correspondingly, it has been shown that Al decreases Ca21-dependent glutamate release in hippocampal slices (31).
246
PLATT AND REYMANN
We suggest that both the baseline effect and the inhibition of TEA LTP may be caused by the action of Al on VDCCs. Although it must be considered that the data concerning the action of Al on VDCCs were obtained using a different preparation (cultured dorsal root ganglion neurons), the action of Al as a general calcium channel blocker is likely to account for the inhibition of TEA LTP. For TEA LTP—rather than for eLTP—there is profound evidence for the contribution of VDCCs (1, 14–16). Thus, the inhibition of calciumdependent processes is the most probable mechanism for Al neurotoxicity. Additional mechanisms including interactions with the cholinergic system (36, 24) may also have contributed to the observed action of Al. Concerning the comparison of the action of Al on the fEPSP slope versus PS amplitude, there is evidence that the PS is more susceptible to Al than the fEPSP: First, both the baseline effect and the decline below baseline after TEA LTP induced in the presence of 2.7 µg/ml Al were more pronounced for the PS amplitude. Second, the expression of the subsequent eLTP of the PS amplitude was already much weaker after the brief exposure to 2.7 µg/ml Al and TEA LTP (see Fig. 4). For the fEPSP slope it appeared that the stronger inhibition of TEA LTP allowed a stronger expression of the subsequent eLTP; only preapplication of 2.7 µg/ml Al depressed eLTP. Third, as revealed by the comparison of the inhibition with both the baseline in the absence and in the presence of 2.7 µg/ml Al, this concentration caused a stronger inhibition of PS amplitude potentiation compared to fEPSP potentiation. We therefore conclude that the more pronounced inhibition of the PS amplitude is based not only on the reduced depolarization due to inhibited fEPSPs but also on other mechanisms necessary for spike generation. The prolonged exposure to 2.7 µg/ml Al preventing the induction of eLTP of PS amplitude and suppressing eLTP of the fEPSP slope indicates that additional slowly developing neurotoxic processes have occurred, for example via accumulation of Al within the cells. Such mechanisms may have caused irreversible damages disabling the induction of eLTP. Additionally, the irreversible block of VDCCs may have caused a general depression of excitability sufficient to suppress eLTP. Our studies on VDCCs provided evidence for a slow time-course of equilibration and evidence for at least one Al binding site within the channel pores of VDCCs (6, 7, 27). These characteristics may well account for the inhibition of eLTP 60 min after exposure to Al and for long-lasting neurotoxic actions of Al in the nervous system. Nevertheless, a variety of other neurochemical actions of Al have to be considered, such as its influence on second messenger systems like inositolphosphate, cycliy-adenosine-monophosphate and the phosphorylation of microtubule-associated protein MAP-2. These actions have been demonstrated both in vivo after
chronic exposure (20) and in vitro (19). Since secondmessenger cascades are crucial for eLTP (3, 33), it is likely that an interference of Al with these components may have prevented eLTP induction after longer application of 2.7 µg/ml Al. Compared to the results of our previous study on eLTP (28, 29), the general characteristics of Al in blocking TEA LTP are similar to those obtained for eLTP, i.e., 0.68 µg/ml caused an attenuation, whereas the inhibition of eLTP by 2.7 µg/ml was more pronounced and potentiation declined back to baseline values within 60 min. The observation that Al inhibits TEA LTP in the same concentration range and with a similar time-course as eLTP provides further evidence that both models may share common mechanisms. In this context, the most obvious candidate is Ca21 as an intracellular trigger for plastic events. Although the contribution of VDCCs to eLTP in CA1 induced by 100 Hz HFS is not as obvious as it is for TEA LTP, we have shown that Al also inhibits NMDA-receptor mediated currents of acutely isolated hippocampal neurons in the same concentration range (30). Thus, an interference of Al with the Ca21 influx through NMDA receptors is a more likely mechanism for eLTP inhibition. The action of Al on TEA LTP provides clear evidence that Al interferes with mechanisms underlying neuronal plasticity which require Ca21 as a major intracellular trigger. We therefore suggest that although Al inhibits eLTP via a different mechanism than TEA LTP, the Ca21-dependent processes downstream from the induction mechanisms are affected in a comparable way. Such mechanisms may be crucial for many cognitive dysfunctions observed after Al exposure and for neuronal diseases in which Al is a etiologic or contributing factor. ACKNOWLEDGMENTS The authors thank Ms. K. Bo¨hm for excellent technical assistance and Dr. Gernot Riedel for critical comments on an earlier version of the manuscript.
REFERENCES 1.
ANIKSZTEJN, L., AND Y. BEN-ARI. 1991. Novel form of long-term potentiation produced by K1 channel blocker in the hippocampus. Nature 349: 67–69. 2. BERNARD, J., LAHSAINI, A., AND G. MASSICOTTE. 1994. Potassiuminduced long-term potentiation in area CA1 of the hippocampus involves phospholipase activation. Hippocampus 4(4): 447–453. 3. BLISS, T. V. P., AND G. L. COLLINGRIDGE. 1993. A synaptic model for memory: Long-term potentiation in the hippocampus. Nature 361: 31–39. 4. BLISS, T. V. P., ERRINGTON, M. L., AND LYNCH, M. A. 1988. Calcium-induced long-term potentiation in the dentate gyrus is accompanied by a sustained increase in glutamate release. In Excitatory Amino Acid Transmission (T. P. Hicks, D. Lodge, and H. McLennan, Eds.), pp. 337–340. A. R. Liss, New York.
ALUMINUM INHIBITS TEA LTP 5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20. 21.
22.
BOLLA, K. I., BRIEFEL, G., SPECTOR, D., SCHWARTZ, B. S., WIELER, L., HERRON, J., AND L. GIMNENEZ. 1992. Neurocognitive effects of aluminum. Arch. Neurol. 49: 1021–1026. BU¨ SSELBERG, D., PLATT, B., HAAS, H. L., AND D. O. CARPENTER. 1994. Mammalian voltage-activated calcium channel currents are blocked by Pb21, Zn21 and Al31. J. Neurophysiol. 71(4): 1491–1497. BU¨ SSELBERG, D., PLATT, B., HAAS, H. L., AND D. O. CARPENTER. 1993. Voltage-gated calcium channel currents of rat dorsal root ganglion (DRG) are blocked by Al31. Brain Res. 622: 163–168. FARNELL, B. J., U. DE BONI, AND D. R. CRAPPER MCLACHLAN. 1982. Aluminum neurotoxicity in the absence of neurofibrillary degeneration in CA1 hippocampal pyramidal neurons in vitro. Exp. Neurol. 78: 241–258. FLECK, M. W., PALMER, A. M., AND G. BARRIONUEVO. 1992. Potassium-induced long-term potentiation in rat hippocampal slices. Brain Res. 580: 100–105. GANROT, P. O. 1986. Metabolism and possible health effects of aluminum. Environ. Health Perspect. 65: 363–441. GARRUTO, R. M. 1991. Pacific paradigms of environmentallyinduced neurological disorders: Clinical, epidemiological and molecular perspectives. Neurotoxicology 12: 347–377. GROVER, L. M., AND T. J. TEYLER. 1990a. Two components of long-term potentiation induced by different patterns of afferent activation. Nature 347: 477–479. GROVER, L. M., AND T. J. TEYLER. 1990b. Effects of extracellular potassium concentrations and postsynaptic membrane potential on calcium-induced potentiation in area CA1 of rat hippocampus. Brain Res. 506: 53–61. HANSE, E., AND B. GUSTAFSSON. 1994. TEA elicits two distinct potentiations of synaptic transmission in the CA1 region of hippocampal slices. J. Neurosci. 14: 5028–5034. HUANG, Y.-Y., AND R. C. MALENKA. 1993. Examination of TEAinduced synaptic enhancement in area CA1 of the hippocampus: The role of voltage-dependent Ca21 channels in the induction of LTP. J. Neurosci. 13(2): 568–576. HUBER, K. M., MAUK, M. D., AND P. T. KELLY. 1995. Distinct LTP induction mechanisms: Contribution of NMDA receptors and voltage-dependent calcium channels. J. Neurophysiol. 73: 270– 279. JEFFREY, E. H. Biochemical mechanisms of aluminum toxicity. In Handbook of Experimental Pharmacology (R. A. Goyer and M. G. Cherian, Eds.), pp. 139–161. Springer, Berlin. JOHNSTON, D., WILLIAMS, S., JAFFE, D., AND R. GRAY. 1992. NMDA-receptor-independent long-term potentiation. Annu. Rev. Physiol. 54: 489–505. JOHNSON, G. V., COGDILL, K. W., AND R. S. JOPE. 1990. Oral aluminum alters in vitro protein phosphorylation and kinase activities in rat brain. Neurobiol. Aging 11: 209–216. JOPE, R. S., AND G. V. JOHNSON. 1992. Neurotoxic effects of dietary aluminum. Cibia Found. Symp. 169: 254–267. KRAUS, A. S., AND W. F. FORBES. 1992. Aluminum, fluorid and the prevention of Alzheimer’s disease. Can. J. Public Health 83(2): 97–100. MALENKA, R. C., KAUER, J. A., ZUCKER, R. S., AND R. A. NICOLL.
23. 24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
247
Postsynaptic calcium is sufficient for potentiation of hippocampal synaptic transmission. Science 242: 81–84. MARTIN, R. B. 1986. The chemistry of aluminum as related to biology and medicine. Clin. Chem. 32(10): 1797–1806. PENG, J. H., XU, Z. C., XU, Z. X., PARKER, J. C., FRIEDLANDER, E. R., TANG, J. P., AND S. MELETHIL. 1992. Aluminum-induced acute cholinergic neurotoxicity in rat. Mol. Chem. Neuropathol. 17(1): 79–89. PERL, P. D., AND A. K. BRODY. 1980. X-ray spectometric evidence of aluminum accumulation in neurofibrillary tangle-bearing neurons. Science 208: 297–299. PETROZZINO, J. J., AND J. A. CONNOR. 1994. Dendritic Ca21 accumulations and metabotropic glutamate receptor activation with an N-methyl-D-aspartate receptor-independent long-term potentiation in hippocampal CA1 neurons. Hippocampus, 4(5): 546–558. PLATT, B., AND D. BU¨ SSELBERG. 1994. Actions of aluminum on voltage-activated calcium channel currents. Cell. Mol. Neurobiol. 14(6): 819–829. PLATT, B., BU¨ SSELBERG, D., AND D. O. CARPENTER. 1994. Actions of aluminum on synaptic transmission and long-term potentiation of CA1 neurons in hippocampal slices. Soc. Neurosci. Abstr. 1516. PLATT, B., CARPENTER, D. O., BU¨ SSELBERG, D., REYMANN, K. G., AND G. RIEDEL. 1995. Aluminium impairs hippocampal longterm potentiation in rats in vitro and in vivo. Exp. Neurol. 134: 73–86. PLATT, B., HAAS, H., BU¨ SSELBERG, D. 1994. Aluminium reduces glutamate activated currents of rat hippocampal neurons. NeuroReport 5(17): 2329–2332. PROVAN, S. D., AND R. A. YOKEL. 1992. Aluminium inhibits glutamate release from transverse rat hippocampal slices: role of G proteins, Ca channels and protein kinase C. Neurotoxicology 13: 413–420. REYMANN, K. G., MATTHIES, H. K., FREY, U., VOROBYEV, V. S., AND H. MATTHIES. 1986. Calcium-induced long-term potentiation in the hippocampal slice: Characterization of the time course and conditions. Brain Res. Bull. 17: 291–296. REYMANN, K. G., AND S. STAAK. 1994. Molecular mechanisms underlying long-term potentiation: Postsynaptic glutamate receptors and protein kinase C. In Protein Kinase C in the CNS Focus on Neuronal Plasticity (P. L. Canonico, U. Scapagnini, U. Pamparana, and A. Routtenberg, Eds.), pp. 7–32. Masson Press, Milan, Italy. TURNER, R. W., BAIMBRIDGE, K. G., AND J. J. MILLER. 1982. Calcium-induced long-term potentiation in the hippocampus. Neuroscience 7(6): 1411–1416. YOKEL, R. A. 1994. Aluminum exposure produces learning and memory deficits: A model of Alzheimer’s disease. In Toxininduced Models of Neurological Disorders (M. L. Woodruff and A. J. Nonneman Eds.), pp. 301–318. Plenum, New York. ZATTA, P., ZAMBENEDETTI, P., BRUNA, V., AND B. FILIPPI. 1994. Activation of acetylcholinesterase by aluminum (III): the relevance of the metal species. NeuroReport 5(14): 1777–1780.