The effect of intracerebroventricular (icv) injection of Aβ25–35 and/or intraperitoneal (ip) application of the L-type calcium channel (VDCC) blockers verapamil or diltiazem were examined in vivo. To by-pass possible systemic actions of these agents, their effects on long-term potentiation (LTP) in the CA1 region of the in vitro hippocampal slice preparation were also examined. Application of Aβ25–35 (10 nmol in 5 μl, icv) significantly impaired LTP in vivo, as did IP injection of verapamil (1 or 10 mg/kg) or diltiazem (1 or 10 mg/kg). In the in vitro slice preparation, LTP was also depressed by prior application of Aβ25–35 (500 nmol), verapamil (20 μM), or diltiazem (50 μM). Combined application of Aβ25–35 and verapamil in either the in vivo or in vitro preparation resulted in a significant reversal of the LTP depression observed in the presence of either agent alone. However, co-application of diltiazem and Aβ25–35 failed to attenuate the depression of LTP observed in the presence of either agent alone in vivo or in vitro. Since LTP is a cellular correlate of memory and Aβ is known to be involved in Alzheimer's disease (AD), these results indicate that verapamil, a phenylalkylamine, may be useful in the treatment of cognitive deficits associated with AD.
Deposition of beta-amyloid (Aβ) is recognized as an early (Lippa et al. 1998) and critical event in the pathogenesis of Alzheimer's disease (AD) (Selkoe 1997). Amyloid deposits are found in cortical regions including the hippocampus, an area known to play a role in memory processing. Aβ peptides have been shown to be neurotoxic in cultured cells (Yankner et al. 1990) and can lead to apoptosis in cultured hippocampal (Mattson et al. 1998) and cortical neurons (Yan et al. 1999). Aβ has also been shown to cause disruption of calcium homeostasis. Aβ1–40 can enhance calcium influx in rat cortical synaptosomes and cultured neurons through L- and N-type voltage-dependent calcium channels (VDCCs) (MacManus et al. 2000) and through L-type VDCCs in PC12 cells (Green and Peers 2001). Indirect activation of L-type VDCCs by Aβ25–35 has also been reported via activation of mitogen activated protein (MAP) kinases (Ekinci et al. 1999) or generation of free radicals (Ueda et al. 1997). Postmortem analysis of the hippocampus of AD patients has revealed an increase in the density of L-type VDCCs in both the dentate gyrus and area CA1 compared with aged-matched controls (Coon et al. 1999). These findings suggest that L-type VDCCs may play a role in the pathogenesis of AD. Aβ peptides have been shown to impair hippocampal synaptic plasticity in the form of long-term potentiation (LTP) in vitro and in vivo (Chen et al. 2000; Freir and Herron 2003; Freir et al. 2001). In addition, studies have shown a correlation between impaired synaptic plasticity and memory deficits following the generation of Aβ aggregates in the rat hippocampus (Stephan et al. 2001).
Here we have investigated whether the depression of hippocampal LTP reported previously (Chen et al. 2000; Freir and Herron 2003; Freir et al. 2001) may be linked to a disruption of postsynaptic calcium influx, a critical event in LTP induction (Malenka et al. 1988). We therefore examined the effect of reducing the activity of L-type VDCCs via application of VDCC blockers in the presence of Aβ25–35 in vivo and in vitro. Part of this work has been presented previously in abstract form (Costello and Herron 2002; Freir and Herron 2001).
In vivo preparation
All experiments in vivo were performed in accordance with guidelines and under license from the Department of Health, Ireland (86/609/EEC). Male Wistar rats, 175–200 g (8–10 wk old) were surgically prepared for acute recordings. Briefly, rats were anesthetized with an ip injection of 1.5 g/kg urethane (ethyl carbamate), and supplementary injections (0.2–0.5 g/kg) were given when necessary to ensure full anesthesia. Deep body temperature was recorded throughout the experiment, and heating pads (Braintree Scientific) were used to maintain the temperature of the animals at 36.5 ± 0.5°C. Small holes were drilled in the skull at the positions of the reference, stimulating, and recording electrodes. Additionally, in some experiments, a separate hole was drilled to introduce a guide cannula for icv injection of drug/vehicle. Animals were placed in a stereotactic frame for recording. The recording electrode was positioned in the stratum radiatum of area CA1 (3 mm posterior, 2 mm lateral to bregma). A bipolar electrode was placed in the Schaffer-collateral/commissural pathway distal to the recording electrode (4 mm posterior, 3 mm lateral to bregma). The cannula was positioned above the lateral ventricle in the opposite hemisphere to that of the electrodes (1 mm posterior, 1.2 mm lateral to bregma).
Stimulating (bi-polar stainless steel; 0.125 mm diam) and recording electrodes (mono-polar stainless steel; 0.125 mm diam) obtained from Plastics One were lowered through the cortex and into area CA1 of the hippocampus using both physiological and stereotactic indicators. Test stimuli were delivered to the Schaffer-collateral/commissural pathway every 30 s (0.033 Hz). Electrodes were positioned to record a maximal field excitatory postsynaptic potential (EPSP). Baseline EPSPs were recorded at 35–40% of maximal response. LTP was induced using a high-frequency stimulus protocol (HFS; 3 × 10 trains of 10 stimuli at 200 Hz) at a stimulus intensity that evoked a field EPSP of approximately 80% of maximum response. Field EPSPs were evoked in the CA1 region using low-frequency stimulation (0.033 Hz.). Extracellular field potentials were amplified (×10), filtered at 5 kHz, digitized, and recorded using the MacLab software acquisition system. Baseline recordings were taken for ≥30 min prior to injection of drug/vehicle to ensure a steady-state response. Following injection of drug/vehicle, baseline recordings were monitored for a further period of1hto monitor normal synaptic transmission. At a time of 1 h postinjection, a series of HFS were delivered to induce a potentiation of the synaptic response. Low-frequency stimulation was then used to evoke EPSPs for a further period of 1 h to record any changes in the synaptic response. Data points displayed on graphs are an average of four consecutive recordings.
Aβ25–35 (10 nmol in 5 μl distilled water) was injected icv (over a period of 2–3 min) using a microsyringe (Hamilton) 1 h prior to HFS. Rats were injected ip with diltiazem (1 or 10 mg/kg in 0.5 ml distilled water) or verapamil (1 or 10 mg/kg in 0.5 ml distilled water) 1 h prior to high-frequency stimulation. In experiments involving the co-application of Aβ25–35 with verapamil or diltiazem, both drugs were injected 1 h pretetanus.
Hippocampal slice preparation
Male Wistar rats, weighing 50–100 g (4–6 wk old) were decapitated under anesthesia. The brains were rapidly removed and immersed in chilled, oxygenated artificial cerebrospinal fluid (ACSF). Brains were dissected, and transverse slices, 350 μm in thickness, were cut using a vibrotome (Campden Instruments). The slices were transferred to a holding chamber and incubated at room temperature in oxygenated ACSF for ≥1 h. The ACSF was composed of (in mM) 120 NaCl, 2.5 KCl, 2.0 MgSO4 2.0 CaCl2, 26 NaCO3, 1.25 NaH2PO4, and 10 d-glucose and was oxygenated with 95%O2-5%CO2. Slices were transferred to a submerged recording chamber, superfused with oxygenated ACSF at a rate of 7 ml/min, and maintained at a temperature of 29–30°C.
Field EPSPs (fEPSPs) were evoked using stimulating (approximately 1 MΩ) and recording (approximately 2 MΩ) glass capillary microelectrodes (Harvard Apparatus) filled with ACSF. The stimulating electrode was placed in the Schaffer collateral/commissural pathway of the CA1, and recordings were made from the CA1 s. radiatum. Stimulus frequency was 0.033 Hz, duration was 0.1 ms, and intensity was 2–8 V. The stimulation intensity was set at approximately 40% of maximal EPSP amplitude as determined from an input-output curve for each experiment. Stable baseline recordings were made for ≥20 min prior to application of drug/LTP induction. EPSPs obtained were amplified ×100 using a Brownlee Precision (model 410) instrumentation amplifier, displayed on an Iso-tech ISP622 oscilloscope, and recorded and analyzed using software supplied by Dr. J. Dempster (Strathclyde University). LTP was induced by three bursts of high-frequency stimulation (10 trains of 10 pulses at 200 Hz) given at 20-s intervals, with no change in stimulation intensity. Drugs were applied via the perfusion media. Verapamil (20 μM) and diltiazem (50 μM) were added 20 min pre-HFS. Aβ25–35 (500 nM) was added to the perfusate 1 h pretetanus. Co-application of both Aβ25–35 and verapamil/diltiazem took place 1 h pretetanus. Drugs were maintained within the perfusate for the duration of each experiment.
The EPSP slope was used to measure synaptic efficacy. EPSPs are expressed as a percentage of the mean initial slope measured during the last 10 min of the baseline-recording period prior to LTP induction. LTP data were analyzed using a two-way ANOVA, which examined all data recorded between 55 and 60 min post-HFS. The significance level was set at P < 0.05. Error bars on the graphs shown represent SE. Data insets are an average of four consecutive EPSPs recorded at the time indicated on the graph. Control experiments in vitro and in vivo were performed between test experiments.
Aβ25–35 caused a depression of LTP
Administration of Aβ25–35 or vehicle (distilled water) had no significant effect on baseline synaptic transmission, in vivo or in vitro (Fig. 1, A and B). In the hippocampal slice preparation, prior administration of 500 nM Aβ25–35 caused a significant impairment of LTP (139 ± 5%, n = 12, P < 0.001, F = 110.5) with respect to vehicle controls (169 ± 7%, n = 9, Fig. 1A).
Injection of 10 nmol Aβ25–35 (icv) 1 h prior to HFS, however, caused a significant depression of LTP in vivo (128 ± 10%, n = 6, P < 0.001, F = 97.98) compared with control values (176 ± 10%, n = 6, Fig. 1B).
Verapamil impairs LTP
IN VITRO. In the slice preparation, administration of verapamil (20 μM) had no significant effect on baseline responses; however, following HFS in the presence of verapamil, LTP was depressed significantly (132 ± 4%, n = 6, P < 0.001, F = 125) compared with the control value (169 ± 7%; Fig. 2A).
IN VIVO. In control experiments, injection of distilled water vehicle (0.5 μl ip) had no effect on baseline synaptic transmission, and following a high-frequency stimulus, robust LTP was produced in the control group (178 ± 11%, n = 7). Injection of either 1 or 10 mg/kg verapamil (ip) also had no significant effect on baseline recordings when measured up to 1 h postinjection. Both concentrations, however, significantly depressed LTP when measured 1 h post-HFS compared with control (139 ± 5%, n = 5, P < 0.001, F = 62.06; and 127 ± 10, n = 6, P < 0.001, F = 96.10; respectively; Fig. 2B).
Diltiazem reduces LTP
IN VITRO. In the slice preparation, diltiazem (50 μM) had no effect on baseline synaptic transmission when administered 20 min prior to HFS; however, diltiazem caused a significant impairment of LTP (123 ± 4%, n = 7, P < 0.001, F = 236.3) compared with vehicle controls (169 ± 7%; Fig. 3A).
IN VIVO. Following injection (ip) of a second class of L-type calcium channel antagonist, diltiazem, at 1 or 10 mg/kg, there was no significant effect on baseline EPSPs. LTP was depressed significantly, however (1 mg: 146 ± 10, n = 6, P < 0.001, F = 34.72; and 10 mg: 130 ± 10%, n = 6, P < 0.001, F = 57.38) compared with control values (178 ± 11%, n = 7; Fig. 3B).
Verapamil attenuates the Aβ25–35-induced depression of LTP
IN VIVO. Aβ25–35 (10 nmol, icv) and verapamil (ip) at concentrations of 1 or 10 mg/kg were co-injected 1 h prior to LTP induction. Co-application of Aβ25–35 and verapamil did not alter baseline synaptic transmission (Fig. 4, A and B). Injection of Aβ25–35 with verapamil (1 mg/kg) produced a similar LTP (176 ± 7%, n = 6, P = 0.947, F = 0.0052) to that seen for controls (178 ± 11%; Fig. 4A). The level of LTP produced in this group was significantly different to that observed following application of verapamil (1 mg/kg; 139 ± 5%) or Aβ25–35 (128 ± 10%) alone. Co-administration of Aβ25–35 and a higher concentration of verapamil (10 mg/kg) also produced a smaller, yet significant, reversal of the Aβ25–35-induced depression of LTP (169 ± 11%, n = 5, P = 0.06, F = 4.33; Fig. 4B).
IN VITRO. Using the slice preparation, verapamil (20 μM) and Aβ25–35 (500 nM) were superfused for 1 h prior to the induction of LTP. In the presence of both agents, the level of LTP produced (159 ± 7%, n = 6) was similar to that recorded under control conditions (169 ± 7%, P = 0.0993, F = 2.759), yet significantly different to that produced following application of Aβ25–35 (139 ± 5%, n = 12, P < 0.001, F = 58.8) or verapamil (132 ± 4%, P < 0.001, F = 107) alone (Fig. 5)
Diltiazem failed to reverse the Aβ25–35 -induced impairment of LTP
IN VIVO. Diltiazem, at concentrations of 1 or 10 mg/kg, was co-applied with Aβ25–35 (10 nmol) 1 h pretetanus. However, unlike verapamil, co-administration of Aβ25–35 and 1 mg/kg diltiazem produced an LTP (149 ± 11%, n = 6) that was significantly different to that seen in controls (178 ± 11%, P < 0.001, F = 25.11; Fig. 6A). Similarly, co-application of the higher dose of diltiazem (10 mg/kg) failed to reverse the Aβ25–35-induced depression of LTP (135 ± 5%, n = 6, P < 0.001, F = 94.57; Fig. 6B).
IN VITRO. In the slice preparation, we also found that co-administration of Aβ25–35 and diltiazem produced a significant depression of LTP (98 ± 7%, n = 5, P < 0.001, F = 404) compared with control values (169 ± 7%). The degree of potentiation produced was, however, significantly lower than that seen following application of diltiazem (123 ± 4%, P < 0.001, F = 108.8) or Aβ25–35 (139 ± 5%, P < 0.001, F = 189.5) alone (Fig. 7).
Previous studies in vivo have demonstrated that icv injection of Aβ25–35 (Freir et al. 2001) and Aβ1–40 (Cullen et al. 1997) significantly impaired LTP. Similar studies in vitro have also shown that treatment of hippocampal slices with Aβ25–35 (Chen et al. 2000; Saleshando and O'Connor 2000) can cause an inhibition of LTP. More recently, soluble oligomers of Aβ25–35 have been reported to impair LTP in the hippocampus both in vitro (Wang et al. 2002) and in vivo (Walsh et al. 2002).
LTP is regarded by many as a cellular correlate for certain forms of learning and memory (Bliss and Collingridge 1993; Malenka 1994). A depression of LTP, with corresponding deficits in memory-based performance tasks, has also been documented in vivo in transgenic mice carrying the Swedish mutation of AD (Chapman et al. 1999). In our study, we report a significant depression of LTP in area CA1 in vivo and in vitro following administration of the 11 amino acid peptide Aβ25–35. An understanding of the mechanisms by which Aβ leads to impairment of LTP and perhaps memory formation is therefore critical to explain a role for Aβ in the pathology of AD.
Aβ has been reported to disrupt calcium homeostasis in cultured cells, which can lead to apoptosis in cultured cortical neurons (Huang et al. 2000). An intracellular rise in postsynaptic calcium has been shown to be a critical and necessary event in the induction of LTP (Lynch et al. 1983; Malenka et al. 1988). Disruption of calcium dynamics by Aβ peptide may explain the depression of LTP observed in this and previous studies. Evidence for a link between Aβ25–35 and altered intracellular calcium levels is suggested by reports showing a direct interaction of Aβ25–35 and calcium channels present on cell membranes. Aβ peptide can cause a potentiation of Ca2+ influx through L-type (Ueda et al. 1997) and both L- and N-type VDCCs (MacManus et al. 2000) in cultured neurons. A rise in intracellular calcium concentration mediated by activation of L-type VDCCs in microglial cells following Aβ25–35 administration has also been reported (Silei et al. 1999). Aβ fragments have also been shown to induce calcium influx through L-type channels via activation of mitogen-activated protein kinase (MAPK) (Ekinci et al. 1999). In addition, postmortem analysis of hippocampi from AD patients revealed a significant increase in the number of L-type VDDCs in area CA1 and the dentate gyrus compared with age-matched controls (Coon et al. 1999). In our study, we investigated whether reducing the activity of L-type VDCCs with two separate channel blockers, verapamil and diltiazem, could attenuate the Aβ25–35-induced impairment of LTP induced by a 200-Hz stimulus that we have reported here and previously (Freir et al. 2001).
Verapamil, a phenylalkylamine, and diltiazem, a benzothiazepine, have been shown to block L-type voltage-gated calcium channels by binding to separate domains of the α1 pore-forming subunit of these channels (for review, see Striessnig et al. 1998). Studies performed in vitro (Grover and Teyler 1990) and in vivo (Freir and Herron 2003; Morgan and Teyler 1999), using similar stimulation protocols as this present study, revealed a depression of LTP in area CA1 following prior application of L-type VDCC blockers. A similar role for L-type VDCCs in LTP in area CA3 (Kapur et al. 1998) and in the amygdala (Weisskopf et al. 1999) has also been documented. We found that administration of either diltiazem or verapamil produced a significant impairment of LTP in vitro and in vivo. This suggests that L-type VDCCs play a role in tetanically induced LTP in vitro and in vivo when using this high-frequency (200 Hz) protocol (Freir and Herron 2003; Grover and Teyler 1999; Morgan and Teyler 1990).
Since LTP was depressed following administration of Aβ25–35 or L-type VDCC blockers, the effects of co-applying Aβ25–35 and either diltiazem or verapamil were investigated. Co-application of Aβ25–35 and either 1 or 10 mg/kg verapamil in vivo produced an LTP that was similar to that seen in control animals. Similarly, in the slice preparation, perfusion of both agents reversed not only the Aβ25–35 but also the verapamil-induced depression of LTP observed previously. Surprisingly, however, we did not see a reversal in the depression of LTP caused by either Aβ or diltiazem following co-application of both compounds. This suggests that there may be a differential mode of action of these channel blockers with respect to Aβ25–35. One possibility is a potential interaction between Aβ25–35 and verapamil, which reduces the effect of each drug alone on LTP. Diltiazem, a structurally different compound, may not interact with Aβ25–35 in a similar manner, allowing each compound to produce a significant depression of LTP alone. Our in vitro study suggests that co-application of Aβ25–35 and diltiazem produces a depression of LTP significantly greater than that seen in the presence of either agent alone, although this was not observed in vivo. A second possibility is that, despite evidence of an interaction between Aβ25–35 and L-type VDCCs (Eknici et al. 1999; Ueda et al. 1997), the Aβ25–35-mediated depression of LTP may be due to an interaction with other channels/signaling mechanisms. If this is the case, some non-specific action of verapamil such as block of potassium channels (Rauer and Grismer 1996) may explain the differences in action of verapmil and diltiazem. Reports suggest that verapamil may also block N-type VDCCs (Kelley et al. 1996). A recent study in cultured cortical neurons suggested that Aβ mediates calcium influx through both L- and N-type VDCCs (MacManus et al. 2000). Verapamil may therefore reverse the Aβ-induced depression of LTP via an action on N-type channels. Finally, Aβ has been reported to form nonspecific cation channels in lipid membranes (Arispe et al. 1994) and in cultured neurons (Bhatia et al. 2000). It is possible therefore that verapamil may be more effective than diltiazem at reducing calcium influx through these Aβ-formed channels. Aside from its effects on L-type calcium channels, verapmail can also act as an antagonist of p-glycoprotein transporters (p-gps) or multidrug resistance proteins (MDRPs) (Pajeva and Wiese 2002). There is now evidence that p-gps act as Aβ efflux pumps (Lam et al. 2001). Deposition of β-amyloid has been shown to be inversely correlated with p-gp expression in the brains of elderly nondemented humans (Vogelgesang et al. 2002). p-gp transporters are known to be present in the endothelial cells that form the blood brain barrier (Jonker et al. 1999). Recently, mRNAs encoding p-gps have been demonstrated in cultured neurons and glial (Hirrlinger et al. 2002), suggesting that these transporters may play a significant role in membrane trafficking within the brain. It is possible therefore that verapamil may alter Aβ-mediated effects in the hippocampus via an interaction with p-gps.
There appears to be a balance between the concentration of verapamil used in vivo and the reversal of the Aβ-induced depression of LTP. Injection of 1 mg/kg verapamil, which caused a smaller depression of LTP when injected alone, reversed fully the Aβ effect on LTP. A higher concentration of 10 mg/kg verapamil also reversed the Aβ-mediated impairment of LTP but to a lesser extent. It is possible therefore that there is a fine balance between a potential activation of L-type VDCCs by Aβ and a block of these L-type calcium channels by verapamil leading to the production of LTP similar to that in controls. However, this hypothesis requires further examination.
To conclude, we have demonstrated a significant depression of LTP when L-type VDCC blockers were applied either in vitro or in vivo, in agreement with previous studies, suggesting a role for L-type calcium channels in LTP in area CA1 (Freir and Herron 2003; Grover and Teyler 1990; Morgan and Teyler 1999). We have also shown an impairment of LTP following administration of Aβ25–35 both in vivo and in vitro. Co-application of verapamil and Aβ was found to reverse both the Aβ and the verapamil-induced depression of LTP. Co-application of diltiazem with Aβ, however, did not result in a reversal of the effects of either agent alone. Although L-type calcium channel antagonists, including verapamil, are used in the treatment of cardiovascular disorders such as angina, hypertension, and cardiac arrhythmias, these results suggest that verapamil or similar compounds may be useful in the treatment of AD.
This research was supported by Enterprise Ireland, Health Research Board Ireland, Dept. of Physiology Presidents Research Award 2001, both from University College Dublin.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society