The choice of anaesthesia used during excitotoxic surgery of the striatum is an important factor to consider as it can influence the lesion outcome. For example, ketamine, a non-competitive NMDA receptor antagonist and widely used anaesthetic agent, shares its target with QA, and hence can mitigate the excitotoxic damage by exerting a neuroprotective effect. It is known for over two decades that ketamine can inhibit and modulate the neurotoxic effect of QA lesions in a dosage-dependent manner (36, 37). Therefore, it is remarkable that it is still used today as an anaesthetic for QA lesion experiments where neurotoxicity is the primary objective, for example, in rodent models of HD (see (38) for review).

Early work that established the ketamine/QA combination in rodent models of HD did not recognise this as an issue when high dose of QA (e.g., 240 nmol) were being used (39). Due to the dose-dependent relationship between ketamine and QA, the neuroprotective interference of the two ligands can be overcome by either increasing the dose of QA or reducing the dose of ketamine. Controversially, the previously described profile of QA relating to its ability to achieve selective neuronal death akin to HD pathology, such as killing GABAergic medium spiny neurones while ­sparing somatostatin, neuropeptide Y, and NADPH-diaphorase immunoreactive striatal interneurons (30, 31) could be related to the lesion modifying aspect of ketamine and not to the inherent properties of the excitotoxin itself. Furthermore, ketamine is commonly used along with xylazine (Rompun), an alpha-2-adrenoreceptor agonist that works as a sedative and muscle relaxant (40), minimising side effects like tremor and muscle rigidity, while increasing the duration of anaesthesia by reducing its renal elimination (41, 42). Doing so, it may increase the bioavailability of ketamine thereby indirectly enhancing its neuroprotective effect, as has been suggested in a recent study where it was directly compared with isoflurane (38) (see Fig. 2).

Isoflurane is a gaseous anaesthetic that provides for a more predictable alternative and has become the anaesthesia of choice in many labs working with rodent excitotoxic lesion paradigms (43, 44). Although isoflurane has been linked with neuropro­tective effects as well, the only study – known to the authors of this chapter – that directly investigated this issue found no interaction between isoflurane and NMDA receptors (45). Isoflurane provides a more rapid induction, greater control and rapid recovery from the anaesthetised state, which means less trauma and danger for the experimental animals. Ketamine, on the other hand, is cheaper, easily applied and does not require additional equipment. The control over the depth of the anaesthesia, however, is far less, the risk of ­overdosing is higher, and of course, there is the issue of inappropriate use in the context of models of excitotoxic neuronal death.

Most common excitotoxic lesions in the striatum are unilateral and this consistently produces impairments in contralateral paw use, or deficits in responding in the contralateral space. Bilateral lesions can be too debilitating, especially if behavioural testing of the animals is planned. The somatotopic organisation of the cortex, and the topographically organised afferent and efferent connections of the striatum mirror the diverse nature of deficits observed following striatal lesions in the rat (46). For example, regional striatal lesions have demonstrated the differential effects of the lateral and medial striatal contribution to motor functions (47). Dorsolateral striatal lesions selectively produce skilled forelimb deficits, while dorsoventral striatal lesions impair both skilled forelimb use and tongue reaching. Medial striatal lesions, receiving inputs mainly from the auditory and visual cortex, do not affect either measurement (48–51). In a choice reaction-time task, lateral and medial striatal lesions induce ipsilateral response bias and increase latency in response initiation, respectively (52).

Non-motor, or cognitive behaviours can also be affected by appropriate targeting of the striatal excitotoxic lesion (53, 54). The striatum receives topographically organised projections from the whole of the neocortex, in particular the prefrontal cortex, providing a major system for the selection and initiation of cortically derived action plans. For example, habit, or procedural learning is characterised by gradual and stable acquisition of an association between a stimulus and a response, and is a form of memory believed to be mediated by the striatum where the sensory information is associated with a learned motor output (55–58). Lesions of the anteromedial area of the striatum, a region that receives its afferents predominantly from associative, auditory and visual cortices, induces deficits in delayed alternation and spatial navigation tasks, for example, but leaves visual discrimination performance intact; however, lateral, particularly ventrolateral striatal lesions destroying cortical afferents from sensorimotor areas, disrupt performance on complex visual stimulus–response habit tasks (24, 59).

The excitotoxic lesion of the rodent striatum relies on the excessive stimulation of the glutamate receptors found on the distal portions of the dendritic spines of the medium spiny striatal neurones. Studies have shown that the extent of the QA-induced lesion can be influenced not only by the availability of the glutamate receptors – e.g., in the presence of a ketamine, a non-competitive NMDA receptor ligand – but also by the state of other striatal afferents such as the corticostriatal glutamatergic or the nigrostriatal ­dopaminergic inputs that form synapses equally on the distal areas of the dendritic spines (60).

The level of endogenous glutamate present in the striatum has a bearing on the extent of striatal lesion induced by excitotoxicity. For example, decortication, leading to the removal of the corticostriatal glutamatergic projection, protects the striatal cells from quinolinic acid-induced necrosis (14, 28). Conversely, glutamate levels within the striatum – and the sensorimotor cortex – have been shown to be higher on the side contralateral to an active, compared to an immobilised limb, and studies have indicated that the exclusive use of a limb can exaggerate the extent of tissue damage in rodent stroke model (61, 62), or following unilateral striatal lesion (44).

Similarly, removal of the nigrostriatal dopaminergic input by 6-OHDA lesion has been shown to significantly reduce the vulnerability of the striatum to subsequent QA lesion (63), and this is particularly important to recognise when trying to establish the double lesion animal model for Multiple System Atrophy (see Kuzdas and Wenning, this volume, ch. 3). Dopamine acting on striatal postsynaptic D1 and D2 receptors can modulate ­glutamate signalling and excitatory currents by, respectively, increasing or decreasing NMDA and AMPA currents in the ­striatal neurons (64, 65), which will impact on the overall cell activity and subsequent excitotoxic damage.