Anatomy and connectivity of the SNR

The substantia nigra (SN) is divided into three parts, the pars compacta (SNC), SNR, and pars lateralis (SNL) (33). The SNC contains the majority of dopaminergic neurons that project to the striatum via the nigrostriatal pathway and release dopamine in the striatum. The SNC dopaminergic neurons receive tonic inhibitory γ-aminobutyric acid (GABA) input from SNR cells (30). The SNR contains mainly GABAergic neurons (70, 72), which receive GABA inputs from the striatum (18, 43, 44, 67) and the globus pallidus (44, 82) and send inhibitory GABAergic projections to the thalamus (17, 42, 53), superior colliculus (10, 17, 96), tegmental reticular formation (11), and SNC (30). The SNR neurons projecting to the tectum are concentrated in the anterior part of the SNR (22, 54). The nigrotegmental cells are located in the posterior part of the SNR (15). The nigrothalamic cells are scattered throughout the SNR and consequently are intermixed with the other groups (15, 16, 22). There is also a set of neurons in the SNR projecting to the striatum; these cells are found within the ventromedial region of the posterior SNR. They correspond in their topographic distribution to the SNR dopamine-containing neurons (68). The neurons in the SNR, which exert an inhibitory influence on SNC cells, are located just ventral to the SNC (30). In addition, the SNL neurons project mainly to the superior colliculus (54).

Early electrophysiologic studies indicated that in adult rats, electrographic seizure activity propagates into the SNR (36, 97). Pharmacologic studies using site-specific drug microinjections have offered significant insight into the role of the SNR in seizures in adult rats. Initial studies by Iadarola and Gale (39) demonstrated that the SNR is a key site of GABA-mediated anticonvulsant action, since microinjections of muscimol (an agonist of GABA receptor site) into the SNR decreased susceptibility to electroshock and chemically induced seizures (39). Later studies in other experimental seizure models demonstrated that intranigral infusions of GABA_A agonists can suppress motor and electrographic seizures (25, 28, 62, 70, 85). In amygdala kindling seizures, bilateral infusions of muscimol into rat SNR markedly reduced the duration of motor seizures and decreased the duration of electrical afterdischarges (46, 57, 58).

During prolonged SE in rats, extensive lesions may develop in the SNR, whereas the SNC remains intact (4, 38, 66, 89). These lesions may arise from a massive metabolic derangement and hyperexcitation in the activated SNR. Therefore, within the SN the SNR plays a critical role in the control of seizures (24). The neurons in the SNR have a much greater sensitivity to muscimol than neurons in the SNC (100). Because dopaminergic agents have no effect on a variety of experimental seizures (41), it has been suggested that the anticonvulsant effect induced by GABA_A agonist infusions into the SNR is not mediated by the nigral dopaminergic pathway (2, 24), although the evidence is still inconclusive (64, 65). The SNR receives GABA inputs, which have been demonstrated to inhibit the SNR projection sites, based on both electrophysiologic (30, 53, 100) and biochemical studies (59). Because the projections of SNR are also GABAergic, the net effect of GABA activity in nigral synapses is disinhibition of SNR target neurons. The infusion studies just discussed indicate that the increased GABA activity in the SNR inhibits nigral outflow pathways and has anticonvulsant effects. Destruction of the SNR neurons should also block the nigral output systems and thus reduce seizure susceptibility. However, the findings of SNR lesion studies are contradictory. Bilateral destruction of the SNR suppressed seizures induced by bicuculline, electroshock (25), and kindling (57). Nevertheless, more selective and specific SNR lesions either had no effect or facilitated rather than inhibited kindling development in rats (80, 98, 99). It is possible that the extent of the SNR lesion may account for the different effects. Within the adult SNR, there are two functional regions that mediate opposite effects on seizures (88, 92–94). In our original study, we found that infusions of muscimol in the anterior SNR (SNR_anterior) have anticonvulsant effects, whereas infusions of muscimol in the posterior SNR (SNR_posterior) have proconvulsant effects on flurothyl seizures (64). This finding may provide another hint to explain the contradictory results from SNR lesion studies. Lesions of different regions of the SNR or lesions of only a part rather than the entire SNR may have different effects on seizures.