New neurons in old brains: implications of age in the analysis of neurogenesis in post-mortem tissue

Abstract

Adult neurogenesis, the proliferation and integration of newly generated neurons, has been observed in the adult mammalian hippocampus of many species. Numerous studies have also found adult neurogenesis in the human hippocampus, but several recent high-profile studies have suggested that this process is considerably reduced in humans, occurring in children but not in adults. In comparison, rodent studies also show age-related decline but a greater degree of proliferation of new neurons in adult animals. These differences may represent biological species differences or could alternatively be explained by methodological differences in tissue handling and fixation. Here, we examine whether differences in the post-mortem interval between death and tissue fixation might impact subsequent detection of adult neurogenesis due to increased tissue degradation. Because there are fewer new neurons present in older subjects to begin with we hypothesized that, subject age might interact significantly with post-mortem interval in the detection of adult neurogenesis. We analyzed neurogenesis in the hippocampus of rats that were either perfusion-fixed or the brains extracted and immersion-fixed at various post-mortem intervals. We observed an interaction between animal age and the time delay between death and tissue fixation. While similar levels of neurogenesis were observed in young rats regardless of fixation, older rats had significantly fewer labeled neurons when fixation was not immediate. Furthermore, the morphological detail of the labeled neurons was significantly reduced in the delayed fixation conditions at all ages. This study highlights critical concerns that must be considered when using post-mortem tissue to quantify adult neurogenesis.

Fig. 1

Doublecortin labeling is reduced with increased post-mortem interval. a Following post-mortem intervals of 0, 6, or 12 h, brains from 4- and 9-month-old rats (n = 4 were fixed for 48 h. The effects of age and post-mortem interval interacted significantly in influencing the b density of doublecortin⁺ cells (Two-Factor ANOVA, Age × Post-Mortem Interval Interaction; F_2,18 = 5.448, p = 0.0141), c which we have also presented as a fold difference in doublecortin⁺ cell density relative to the mean density of the perfused groups of each age (Two-Factor ANOVA, Age × Post-Mortem Interval Interaction; F_2,18 = 4.478, p < 0.0264). d This change in density of doublecortin⁺ cells occurred without any differences in dentate gyrus volume. e Age and post-mortem interval also interacted significantly to influence the optical density of doublecortin labelling in the granule cell layer (Two-Factor ANOVA, Age × Post-Mortem Interval Interaction; F_2,18 = 12.34, p = 0.0004). Representative images of doublecortin⁺ cells in the subgranular zone in 4-month-old (f) and 9-month-old (g) rats with post-mortem intervals of 0, 6, and 12 h. Representative images of the morphology of doublecortin labelling in 4-month-old (h) and 9-month-old (i) rats with post-mortem intervals of 0, 6, and 12 h. Data are mean ± SEM. Post-hoc analyses of significant interactions in Two-Factor ANOVA used Newman Keuls. *p < 0.05