► Alzheimer's disease. These drugs have limited efficacy and cannot be used ubiquitously across all disorders where cognitive enhancement might be beneficial (for example, they are inappropriate for the use in patients with ► Parkinson's disease as they can aggravate the motor disability). Therefore, the future development of a range of drugs with different mechanisms of action is likely to occur.
lonotrophic Glutamate Receptors
Ionotrophic ► glutamate receptors as a target for cognition enhancement have been of interest since the 1980s following the discovery of their involvement in ► Long-term potentiation (LTP; a persistent strengthening of synapses). This form of synaptic plasticity requires the activation of N-methyl-d-aspartate (NMDA)-receptors for its induction. Alpha-amino-3-hydroxy-5-methyl-4-isoxazoleproprionate (AMPA)-receptors can also promote the induction of LTP, and they play an important role in the expression of the potentiated response (Cooke and Bliss 2005). It has been presumed that LTP underlies some forms of learning and memory for two reasons. First, because the induction of LTP at any particular synapse requires coincident activity, it fulfils the criterion for a Hebb Synapse, predicted in 1949 to be necessary for the formation of associations. Second, compounds that act as antagonists at NMDA-receptors both impair the induction of LTP and produce clear cognitive deficits in animals and humans (see Robbins and Murphy 2006). Not surprisingly then, the enhancement of cognition with ligands that positively modulate NMDA- and AMPA-receptors has been a goal for some time. While it is not possible to use agonists that act by stimulating these receptors via the glutamate binding site, due to the risk of causing neurodegeneration or inducing seizures, in recent years, ligands that act through other binding sites to modulate receptor activity and which have pro-cognitive effects have emerged.
► NMDA-receptors are comprised offour protein subunits clustered around an ion channel. The channel is normally occupied by magnesium, but when the neuronal membrane is sufficiently depolarised, magnesium dissociates and calcium influx into the cell is permitted. Each heteromeric NMDA-receptor is believed to consist of two protein NR1 subunits plus two NR2 subunits. There are four variants of the NR2 subunit, denoted A,B,C, and D. While glutamate itself binds to sites on the NR2 subunits, there are a number of other binding sites located at different parts of the receptor, including those for the endogenous "co-agonists,"
glycine and d-serine which bind to the NR1 subunits, at GlycineB sites (see Fig. 1). Although glutamate and Gly-cineB sites are located on different subunits of the NMDA-receptor, both glycine and d-serine can enhance the action of glutamate. In turn, the effectiveness of glycine and d-serine at GlycineB sites is determined by the variant of NR2 subunit(s) present in the receptor (Millan 2005). In terms of putative nootropics acting at NMDA-receptors, compounds acting on receptors containing NR1 plus NR2A/NR2B subunits are of interest, as these subunits are distributed in brain areas such as the ► hippocampus and the frontal cortex. Ideas about enhancement of cognition via these particular NMDA-receptors have also been encouraged by the seminal publication of Tang et al. (1999), indicating that mice that are genetically engineered to over express receptors containing NR2B subunits showed superior performance in several tests of learning and memory. Although the GlycineB site was once thought to be saturated by its endogenous ligands, it is now recognized that it is, which opens up the possibility that this site might be targeted by drugs to produce procognitive actions via the NMDA-receptor.
Agonists acting directly at the GlycineB site of the NMDA-receptor include the amino acids glycine, d-serine, and the synthetic ► partial agonist d-cycloserine. In preclini-cal studies, the full agonists such as glycine and d-serine have been investigated as cognition enhancers. Large doses are required for efficacy, but they do reverse deficits in ► novel object recognition tests and the impairment seen in a developmental model of sensorimotor gating deficits. Human studies have also investigated the effects of full agonists such as glycine, d-serine, and d-alanine. In healthy humans, many studies report that glycine is ineffective in a variety of tests and that high doses may
Cognitive Enhancers: Role of the Glutamate System. Fig. 1. Schematic diagram of subunits of the NMDA-receptor showing binding sites for glutamate and the co-agonists glycine and D-serine.
even impair some aspects of cognition. One study employing a very low dose, however, has reported retrieval improvements in a test of word recall. The effects of the full agonists have also been investigated in ► schizophrenia where they appear to produce some improvement in negative symptoms. Their cognitive benefits appear less clear though.
The effects of the partial agonist d-cycloserine have been more studied than those of the full agonists. Preclinical experiments in rats have shown that it can enhance ► spatial memory, such as that measured in radial and water mazes, in addition to visuospatial memory and visual recognition in primates. Reversal of decrements in cognition can also be seen in ageing animals. In rats, some elegant studies by Davies and colleagues have demonstrated the ability of d-cycloserine to facilitate fear extinction. Fear extinction refers to the process of reducing a response to a cue previously paired with a fear-evoking event, by exposure to the cue in the absence of the associated event. This is the basis of exposure therapy, where a patient is repeatedly exposed to a feared object or situation, with support and without any adverse consequences. Importantly, extinction of the fear response occurs through new learning which inhibits the original fear, rather than by simple forgetting. As with some other forms of learning, extinction is a process that can be prevented by antagonists acting at NMDA-receptors. Davis and colleagues therefore predicted that ligands that positively modulate the NMDA-receptor should facilitate fear extinction and in preclinical studies they went on to show that d-cycloserine could do exactly that. These and later studies led to clinical trials of d-cycloserine in humans as an adjunct to improve exposure therapy for acrophobia (fear of heights). During these trials that were carried out in a virtual reality environment, patients were given d-cycloserine (or placebo) before two therapy sessions and retested 1 week and 3 months later. At the retests, patients who had received d-cycloserine during therapy sessions, showed enhanced fear reduction in the virtual reality environment (Davis et al. 2006). Therefore, by potentiating the action of glutamate through the GlycineB-site, extinction learning can be accelerated in both animals and humans.
In other studies, d-cycloserine has been found to improve different aspects of learning, memory and also performance in tests of cognitive flexibility in both healthy humans and in patients with schizophrenia or Alzheimer's disease. As in the fear extinction studies, these procognitive actions of d-cycloserine have invariably been seen at small doses only, perhaps because its efficacy as an agonist may be lost at larger doses (Priestley et al. 1995).
Another way to enhance the activity of NMDA-receptors via the GlycineB-site is to elevate the synaptic levels of endogenous glycine itself. Normally, the action of glycine would be terminated by its removal from the synapse by specialized transporters located on neuronal membranes and on the surrounding glial cell membranes. Although there are multiple subtypes of transporter, glycine-1 transporters (GLYT-1) located on the glial cells play a major role in the removal of glycine from the synapse and it is the activity of these transporters that is thought to be the reason why GlycineB-sites are not saturated in vivo. Drugs that prevent the activity of GLYT-1, glycine reuptake inhibitors (GRIs) should therefore increase glutamate action at NMDA-receptors.
The first generation of GRIs were based on sarcosine (N-methyl-glycine), a selective inhibitor of glycine uptake at GLYT-1 and include NFPS (N-[3-(4'-Fluorophenyl)-3-(4'-phenylphenoxy)propyl]sarcosine) and Org24461/ 24598. Second generation GRIs are non-aminoacid-based compounds, such as SSR504734. Accompanying the rise of the glutamatergic hypothesis of schizophrenia, our knowledge of the procognitive actions of GRIs comes mostly from preclinical studies aimed at modeling aspects of cognition that are impaired in the disorder. Pre-pulse inhibition, or PPI, describes the ability of a small pre-stimulus to inhibit the startle response to a larger stimulus. In animal models, deficits in PPI arise spontaneously in the DBA/2 strain of mouse or can be seen to develop in adult rats, following neonatal treatment with phencycli-dine. These are thought to model a schizophrenic abnormality in filtering information and studies in animals have consequently shown that GRIs can reverse these deficits. Other studies have shown their effectiveness at improving object recognition or reference memory, in tests where deficits have been pharmacologically induced with NMDA-receptor antagonists. In human schizophrenia patients, sarcosine and other GRIs can improve "cognitive symptoms'' but the nature of these improvements is not yet well described.
Overall, drugs targeting the GlycineB-site of the NMDA-receptor do appear to have procognitive effects on several aspects of learning and memory. There is little evidence for improvements in attention and their effectiveness for enhancing other cognitive domains remains to be fully investigated.
Most of the fast glutamatergic neurotransmission in the brain is mediated by ► AMPA-receptors. In addition, changes in synaptic plasticity are associated with
Was this article helpful?