Potency, voltage-dependency, agonist concentration-dependency, blocking kinetics and partial untrapping of the uncompetitive N-methyl-d-aspartate (NMDA) channel blocker memantine at human NMDA (GluN1/GluN2A) receptors
Introduction
Memantine (1-amino-3,5-dimethyl-adamantane) is registered in over 60 countries worldwide, amongst them the USA and Europe, for the treatment of moderate to severe Alzheimer's disease. Both the clinical tolerability and the symptomatic effects of memantine have been attributed to its moderate affinity (IC50 around 1 μM at −70 mV) for NMDA receptor channels and associated fast double exponential blocking/unblocking kinetics and strong voltage-dependency (Rogawski, 1993, Parsons et al., 1993, Parsons et al., 1999, Johnson and Kotermanski, 2006). These functional properties have been characterized and confirmed by numerous groups using whole-cell patch clamp recordings from primary cultures of rat hippocampal and rat cortical neurones as well as from rat NMDA receptors expressed heterologously in HEK-293 cells (Parsons et al., 1993, Parsons et al., 1995, Parsons et al., 1996, Parsons et al., 1998, Bresink et al., 1996, Chen and Lipton, 1997, Blanpied et al., 1997, Sobolevsky and Koshelev, 1998, Sobolevsky et al., 1998, Losi et al., 2006).
How these biophysical properties account for the better therapeutic efficacy/safety of memantine compared to other channel blockers such as (+)MK-801, phencyclidine and ketamine has been a matter of considerable scientific debate and there are several theories, which will be discussed in more detail later. However, the reported underlying biophysical properties underlying all such theories are consistent and are detailed below.
Memantine blocks and unblocks open NMDA receptor channels with double exponential kinetics in a voltage- and use-dependent manner, meaning that it can only gain access to the channel in the presence of agonist and remains largely trapped in the channel following removal of agonist (Parsons et al., 1993). The amplitude and speed of the fast component of block increase with memantine concentration. In contrast, the speed of fast unblock remains constant but its weight (relative to the slow component) decreases with memantine concentration (Frankiewicz et al., 1996, Bresink et al., 1996, Blanpied et al., 1997, Sobolevsky et al., 1998, Sobolevsky and Koshelev, 1998). Moreover, the predominant effect of depolarization is to increase dramatically the weight of the faster recovery time-constant due to the voltage-dependence of the blockade (Frankiewicz et al., 1996, Bresink et al., 1996, Parsons et al., 1998). These data indicate that memantine binds to at least two sites within the channel (Sobolevsky and Koshelev, 1998, Sobolevsky et al., 1998).
It has further been suggested that the equilibrium blockade of NMDA receptor channels is dependent upon the concentration of agonist (Lipton, 2006). This assumption is supported neither by the blocking model proposed by Blanpied et al. (1997) nor by our own recent data which highlighted, for us, the technical problems in performing such studies, in particular regarding the kinetics of blockade (Gilling et al., 2007).
In addition to strong voltage-dependency and fast blocking kinetics, memantine appears to have a lesser tendency to be trapped in NMDA receptor channels than do ketamine, phencyclidine or (+)MK-801 (Blanpied et al., 1997, Sobolevsky and Koshelev, 1998, Johnson and Kotermanski, 2006). This difference has been partly attributed to the faster kinetics of memantine. Receptors blocked by memantine retain agonist and thereby open and release memantine following removal of both agonist and memantine from the extracellular solution. This partial untrapping was reported to be less pronounced for higher affinity compounds as their slower unblocking kinetics do not allow them to leave the channel quickly enough following agonist removal.
One major caveat for the translation of such in vitro functional findings to the in vivo human situation is that almost all studies were performed on cultures, clones or slices from rodent receptors/CNS tissue. There are very limited data available for human NMDA receptors. Memantine did indeed displace [3H]MK-801 from NMDA receptor channels in post-mortem human cortical tissue with a similar potency to that observed in these studies in receptors/tissue from rodents (Kornhuber et al., 1989). However, these studies provided no information on the functional biophysical properties of this interaction with human NMDA receptors and also bear the unfortunate caveat that such studies were made on receptors under no influence of membrane potential because they were performed in isolated membrane fragments. This fact begs the question, was it perhaps fortuitous that the potency of a voltage-dependent NMDA receptor channel blocker was similar at human NMDA receptors examined in this manner?
The only data that we are aware of on the functional blockade of human NMDA receptors by memantine is the study of Ferrer-Montiel et al. (1998) where memantine blocked human GluN1/GluN2A expressed in Xenopus oocytes with an IC50 of 220 nM at −100 mV. This fivefold greater potency of memantine than observed in the binding experiments (Kornhuber et al., 1989) on human receptors indicates that these caveats in the interpretation of binding data might be justified.
In vitro data are often used to predict in vivo effects of compounds, especially those to be used in a clinical situation, and it is therefore important that the assays used represent the therapeutic situation as closely as possible. Most in vitro experiments are performed using cells and tissue derived from rodent sources for practical reasons, but the use of cloned receptors has led to the increased possibility of testing substances upon human proteins. Some substances show large species-specific differences, so using human rather than rodent receptors and tissue may highlight important differences in the effects of drugs.
Here, we wished to confirm certain in vitro functional/biophysical properties of memantine, ketamine and (+)MK-801 upon human NMDA receptors, characteristics which have previously been reported by this and other groups on rodent receptors (for examples, see Chen et al., 1992, Parsons et al., 1993, Parsons et al., 1995, Bresink et al., 1996).
In this study, we were able to obtain human brain tissue for use in radioligand binding assays to confirm the findings reported by Kornhuber et al. (1989). Furthermore, we used human GluN1/GluN2A NMDA receptor subunits expressed in the human cell line, HEK-293 for functional studies. By using these cells, it was possible to measure the effect of antagonist upon NMDA-induced Ca2+ influx into the cells using a FLIPR device by detecting changes in intracellular Ca2+ levels. In addition, the effects on human GluN1/GluN2A receptors were studied using electrophysiological patch clamp techniques allowing faster perfusion of the cells compared to two electrode voltage clamp recordings made from Xenopus oocytes and therefore better resolution of the current blocking kinetics.
Specifically, several issues were addressed: 1, the potency of channel blockers in binding to NMDA receptors in post-mortem human cortical tissue expressing NMDA receptors; 2, their potency to antagonize intracellular Ca2+ responses of human GluN1/GluN2A receptors expressed in HEK-293 cells; 3, the potency, kinetics and voltage-dependency of block by memantine and ketamine at human GluN1/GluN2A receptors using patch clamp experiments; 4, partial trapping of memantine/ketamine by human GluN1/GluN2A receptors and 5, the agonist concentration-dependency of blockade of human GluN1/GluN2A receptors by memantine using different concentrations of the endogenous agonist glutamate.
Some of this data has previously been published in poster form (Gilling et al., 2006).
Section snippets
Binding experiments
Male Sprague–Dawley rats (200–250 g) were decapitated and their brains removed rapidly. Frozen human cortical tissue was kindly provided by Prof. P. Riederer of the University of Würzburg, Germany, and was treated in the same manner as the rodent tissue.
The cortex was dissected and homogenized in 20 volumes of ice-cold 0.32 M sucrose using a glass-Teflon homogenizer. The homogenate was centrifuged at 1000 × g for 10 min. The pellet was discarded and the supernatant centrifuged at 20,000 × g for 20 min.
Results
The results of binding experiments performed using the [3H](+)MK-801 assay in rat and human cortical tissues are shown in the graphs in Fig. 1. (+)MK-801 had a Ki of 0.005 ± 0.001 μM, and memantine and ketamine displaced the radioligand with similar potency when rat cortical tissue was used (Ki values of 1.19 ± 0.08 μM and 1.35 ± 0.43 μM, respectively). The absolute values agree well with those previously reviewed by Parsons et al. (1999). When the experiments were performed using human cortical
Discussion
The results from experiments using memantine, ketamine and (+)MK-801 presented here agree very well with most of those previously reported (Parsons et al., 1999). The main difference between previous studies and the present study was the use of human tissue/receptors in place of rodent. Although electrophysiological experiments have previously been performed to test memantine upon human GluN1/GluN2A receptors expressed in Xenopus oocytes (Ferrer-Montiel et al., 1998), the HEK-293 cells used
Summary
The use of transfected HEK-293 cells has allowed us to further investigate the in vitro effects of memantine and ketamine on human GluN1/GluN2A receptors. The present data closely match previously reported data from studies using rodent receptors and confirm that memantine is a fast, voltage-dependent open-channel blocker of human NMDA receptors, in accordance with its proposed mechanism of action in Alzheimer's disease. However, it should be noted that various factors such as secondary
References (49)
- et al.
Intra- and extracellular changes of amino acids in the cerebral cortex of the neonatal rat during hypoxic–ischemia
Dev. Brain Res.
(1991) - et al.
Study of potency, kinetics of block and toxicity of NMDA receptor antagonists using fura-2
Eur. J. Pharmacol.
(1996) - et al.
Pharmacokinetics and analgesic effect of ketamine in man
Br. J. Anaesth.
(1981) - et al.
Bioavailability, pharmacokinetics, and analgesic activity of ketamine in humans
J. Pharm. Sci.
(1982) - et al.
Low-frequency activation of the NMDA receptor system can prevent the induction of LTP
Neurosci. Lett.
(1989) - et al.
Memantine restores long term potentiation impaired by tonic N-methyl-d-aspartate (NMDA) receptor activation following reduction of Mg2+ in hippocampal slices
Neuropharmacology
(1999) - et al.
Reduction of voltage-dependent magnesium block of N-methyl-d-aspartate receptor-mediated current by in vivo axonal injury
Neuroscience
(2000) - et al.
Agonist concentration-dependency of blocking kinetics but not equilibrium block of N-methyl-d-aspartate receptors by memantine
Neuropharmacology
(2007) - et al.
Excitotoxic index – a biochemical marker of selective vulnerability
Neurosci. Lett.
(1991) - et al.
Mechanism of action of memantine
Curr. Opin. Pharmacol.
(2006)