Abstracts

Rationale: Abnormalities of hippocampal dentate granule cells may contribute to the development of temporal lobe epilepsy, and recent studies have demonstrated that aberrant integration of adult-generated granule cells underlies several abnormalities of the epileptic brain, including formation of aberrant basal dendrites. These studies also revealed, however, significant heterogeneity in the responses of adult-generated cells to epileptogenesis. Specifically, while almost 50% of these young cells developed basal dendrites, the remaining 50% did not. Whether these populations exhibit other differences, however, is not known. Granule cells without basal dendrites, for example, may exhibit a variety of other changes indicative of maladaptive integration into the hippocampal circuit. Alternatively, these granule cells may integrate appropriately, or even adaptively, while those cells with basal dendrites may reflect a population of “bad actors” exhibiting a host of additional maladaptive changes. Distinguishing among these possibilities is important for determining the net effect of neurogenesis on the epileptic brain. Methods: Thy1-GFP expressing transgenic mice were injected with BrdU to birthdate adult-generated granule cells one week before pilocarpine treatment to induce status epilepticus and the development of epilepsy. Control mice were treated with saline and did not develop status. Animals were sacrificed eleven weeks after status, and twelve-week-old BrdU-labeled, GFP-expressing cells were identified in tissue from control and epileptic animals. Labeled neurons were reconstructed from confocal image stacks, and spine density was determined in the inner, middle and outer molecular layers. Results: Relative to age-matched granule cells from control animals, granule cells from epileptic animals with basal dendrites exhibited distorted apical dendritic trees. In particular, a subset of these cells failed to develop the characteristic, fanlike, projections typical of granule cells, and the portion of the dendritic tree contained within the granule cell body layer was significantly increased. Notably, however, adult-generated granule cells lacking basal dendrites were also abnormal, exhibiting an increase in the number of apical dendritic branches and a decrease in dendritic spine density in the inner molecular layer. Granule cells with basal dendrites did not exhibit significantly decreased spine density, but did exhibit a significant positive correlation between spine density and basal dendrite length, suggesting that aberrant hilar input to these cells was associated with increased input into the apical dendritic field. Conclusions: The present morphological findings suggest that functionally distinct populations of adult-generated granule cells exist in the epileptic brain, and support the hypothesis that granule cells with basal dendrites may be particularly disruptive. By contrast, spine loss among granule cells without basal dendrites may reduce excitability among this population.