Structural analogues of glutamate, which are neurotoxic (see NEUROTOXINS) when injected into the CENTRAL NERVOUS SYSTEM in appropriate concentrations. They include AMPA, DOMOIC ACID, IBOTENIC ACID, KAINIC ACID, N-methyl-D-aspartate (NMDA), QUINOLINIC ACID and QUISQUALIC ACID—all of which are naturally acidic, but must be titrated to a pH of approximately 7, by adding concentrated sodium hydroxide, an alkali, before they are suitable for injection into the central nervous system. Excitotoxins bind to GLUTAMATE RECEPTORS on neuronal dendrites and excite the neurons to death. Typical histological evidence (see HISTOLOGY) of excitotoxic action includes intense REACTIVE GLIOSIS and neuronal loss in the region of the injection, as seen in sections stained with cresyl violet, and in some cases large calcium deposits around the edges of the lesion, identified with the stain Alizarin red. When excitotoxins were first discovered, they were unique lesionmaking tools, because of their ability to make FIBRE-SPARING LESIONS. Other, cruder, lesionmaking methodologies included ASPIRATIONLESIONS, ELECTROLYTIC LESIONS, KNIFE CUT LESIONS and RADIOFREQUENCY LESIONS.
Kainic acid, AMPA and NMDA have their greatest effects at these particular IONOTROPIC receptors; quinolinic acid has part of its action through NMDA receptors; ibotenic acid activates both NMDA and METABOTROPIC receptors. The exact mechanisms which lead to neuronal death differ according to which receptor is involved. Action of excitotoxins at NMDA receptors leads to massive influx of calcium ions and over-excitation of the neuron; activation at AMPA/kainate receptors involves influx sodium ions; metabotropic receptors initiate second messenger systems such as the INOSITOL phosphate pathway. In all instances, the neurotoxic process itself is relatively fast, and is complete in less than 24 hours. During this period, the animal usually shows signs of SEIZURE activity. In some instances, typically following kainate infusions, evidence has been found of lesions at sites distant from the infusion. The HIPPOCAMPUS is most susceptible to such damage, leading to the hypothesis that it occurs as a result of ANOXIA during seizure activity. Pretreatment with a BENZODIAZEPINE such as DIAZEPAM is an effective means of dampening seizures and reducing the risk of remote damage, without decreasing the excitotoxic action at the site of interest.
Although excitotoxins can make fibre-sparing lesions in the central nervous system, careful histological analyses at different stages post-lesion have shown that the inflammatory response (reactive gliosis) which is part of the excitotoxic process actually leads to breakdown of the BLOOD-BRAIN BARRIER in the injected region. This can result in loss of MYELIN sheaths in lesioned areas which contain diffuse fibre systems. Although initiation of remyelination can be seen after 2–3 weeks, this side-effect has clear implications for studies which involve behavioural testing following an excitotoxic lesion. The size of an excitotoxic lesion is governed by many factors, such as the volume of the infusion, the concentration of the excitotoxic agent and TISSUE TORTUOSITY/receptor profile in the targeted region. Factors specific to a particular excitotoxin may also come into play. For instance, ibotenic acid undergoes enzyme-catalysed decarboxylation to MUSCIMOLin vivo, which may act at GABA RECEPTORS to dampen the sensitivity of local neurons to further excitation. This is likely to be the reason why, at well-titrated doses, ibotenate can be used to make small, well-defined lesions. In some instances, ANAESTHETICS such as SODIUM PENTOBARBITONE or CHLORAL HYDRATE, Can modify the neurotoxicity of excitotoxins. This is also thought to be related to their action at GABA-A receptors.
An excitotoxin is most effective when neurons of interest in a given region are more susceptible to its toxic effects than others, thereby inducing a relatively selective lesion. Generally, the pattern of damage caused by different excitotoxins is related to the types of glutamate receptors found in the target area, and perhaps even the specific RECEPTOR SUBUNITS which form these receptors. A good example of this is the use of AMPA in the rat BASAL FOREBRAIN, where it can be used to target CHOLINERGIC neurons, if the concentration and volume are titrated correctly. Histochemical and lesion data suggest that this is linked to the high expression of the GluR4 receptor subunit in AMPA receptors in this area. Unfortunately, excitotoxins are never generally more than partially selective neurotoxic agents, and any selectivity which is obtained is critically dependent on the concentration of the infusion. The more recent immunotoxin approach, which uses ribosomeinactivating proteins such as SAPORIN, conjugated to specific receptor antibodies, is a more selective means of making brain lesions.
