The study of the molecular pathogenesis of epilepsy in tuberous sclerosis has taken on a new dimension with the identification of the TSC1 and TSC2 genes. While the development of seizures is ultimately related to mutations in one of the two genes, the mechanism underlying the genotype-phenotype relationship remains a puzzle. This chapter, presented arguments in favor of the hypothesis that abnormal cortical excitability originates in and around focal areas of structural malformations (i.e., cortical tubers and dysplasia) and that these "lesions" are the biologic consequences of tuberin and/or hamartin dysfunction. This model relies on the concept of a multistep process occurring early in cortical development whereby certain progenitor cells in the germinal layer of the ventricular zone destined for the cortex undergo inactivation of the TSC1 or TSC2 locus (Fig. 2). Immature neuroepithelial cells carrying "two-hit" mutations at either locus are believed to proliferate, migrate, and differentiate abnormally, resulting in the formation of "dysplastic" cells that are heterotopic in distribution. The pathology of the classic tuber suggests a clonal expansion of the bizarre-appearing giant cells that display incomplete, multilineage, and often ambiguous phenotype. Further, they infiltrate the six-layered structure of the cortex to form a poorly circumscribed area containing a mixture of cell types to create a highly disorganized region of a neuronal and glial network. Whether arising from the dysplastic "two-hit" target cells themselves or adjacent "innocent" bystander neurons as a result of aberrant cell-cell interaction, abnormal epileptic discharges originate from these structural abnormalities. The mechanism of how TSC1 and TSC2 inactivation causes tuber to develop is not known, but emerging experimental evidence suggests a disruption of the hamartin-tuberin "haloenzyme" in the regulation of cell size and number via the insulin signaling pathway and a p27/CDK-dependent mechanism. Biochemically, TSC1/TSC2 may associate with cytoskeletal components and vesicular adaptors in regulating sorting and trafficking of newly synthesized and recycling proteins in the post-Golgi compartments. As such, spatial and temporal localization of proteins may be affected in tuberin or hamartin-deficient neuronal cells where proper synaptic delivery of neurotransmitters plays an important role in normal cerebral function. We are in the earliest stages of understanding the role of TSC genes in epileptogenesis. To test the hypothesis outlined earlier, there is a need to create in vitro and in vivo models, as direct human experimentation is not feasible. To date, there are several rodent models of TSC, both spontaneous and recombinant strains. Unfortunately, none has consistently developed spontaneous cortical tubers, although one example was reported in an otherwise asymptomatic Eker rat (Mizuguchi et al., 2000). If the "two-hit" hypothesis is operational in tubers, as seen in other TSC lesions, it follows that radiation and chemical carcinogens should have a quantitative and qualitative effect on the development of these cerebral malformations. In preliminary experiments, we have found evidence of areas of cortical dysplasia in Eker rats irradiated early in life (Fig. 3). These dysplastic [figure: see text] cells stained positively with NeuN, consistent with the immunophenotype of cells in tubers. Alternatively, one can analyze the in vivo and in vitro characteristics of neuroprogenitor cells that are deficient of hamartin or tuberin. While homozygous mutants of TSC1 and TSC2 are lethal during midgestation, one of several techniques can be used to derive mutant neuroepithelial cells, including the procurement of -/- cells prior to embryonic deaths and subsequent cortical transplantation into syngeneic animals, development of conditional "knock outs," or chimeric mutants. These approaches, with their unique advantages and disadvantages, will be helpful in gaining insights into the development of cortical tubers and their electrophysiologic consequences.