Behavioural effects of interference with the postnatal acquisition of hippocampal granule cells.

MULTIPLYING cells of the nervous system, unlike differentiating and mature cells, are extremely radiosensitivel-3 , which allows selective elimination of the postnatally fonning basket, stellate and granule cells in the cerebellar cortex',". The proliferating precursors of these cells are killed by one or two doses of 150­ 200 r X-irradiation, but the cerebelIum must be exposed to up to 8 repeated doses during early infancy to prevent regenera­ tion 6 .'. Such schedules of low-level irradiation do not produce pathological changes in the prenatally-formed, differentiating Purkinje cells, as shown by electron microscopy·. As most granule cells of the dentate gyrus of the hippocampus originate postnatally in altricial species·-12 , irradiation of the infant rat hippocampus using a schedule derived from previous cerebellar autoradiographic and irradiation studies·,' should reduce the population of granule cells without directly harming the pre­ natally-fonned pyramidal cells of Ammon's horn. Purdue-Wistar male rat pups, cross fostered and raised eight to a litter, were used. To determine the location of the hippocampus, pups of two to eighteen days were decapitated, sagittal slices were made through the head 2 mID from the midline and distances from the snout to the anterior and posterior borders of the hippocampus were measured. A curve was constructed to serve as a guide for establishing hippo­ campal position as a function of age allowing a 2 mID error margin. Pups from twelve litters were wrapped in plastic tubes as described previously13 and placed in a lead-shielded lucite block holder with a slit allowing only that portion of the head containing the hippocampus to be exposed to X-rays from a 'Maxitron' 300 kV unit. At the rate of 50 r min, 200 r was delivered on days 2 and 3, followed by 150 r on days 5,7, 9, 11, 13 and 15. Pups from 14 control litters were wrapped in the same manner as the irradiated group, pups from 7 non­ wrapped control litters were left undisturbed. The 15 irradiated and 15 wrapped control animals on which we report quantitative histology were killed at 30, 60 and 90 days by cardiac perfusion with 10 % buffered formalin, and the brains were postfixed in Bouin's solution. As there were behavioural differences between the wrapped and nonwrapped control groups, 5 nonwrapped control animals were included in the 9O-day study. The brains were embedded in 'Paraplast'