P53 regulates disruption of neuronal development in the adult hippocampus after irradiation

Abstract

Inhibition of hippocampal neurogenesis is implicated in neurocognitive dysfunction after cranial irradiation for brain tumors. How irradiation results in impaired neuronal development remains poorly understood. The Trp53 (p53) gene is known to regulate cellular DNA damage response after irradiation. Whether it has a role in disruption of late neuronal development remains unknown. Here we characterized the effects of p53 on neuronal development in adult mouse hippocampus after irradiation. Different bromodeoxyuridine incorporation paradigms and a transplantation study were used for cell fate mapping. Compared with wild-type mice, we observed profound inhibition of hippocampal neurogenesis after irradiation in mice deficient in p53 despite the absence of acute apoptosis of neuroblasts. The putative neural stem cells were apoptosis resistant after irradiation regardless of p53 genotype. Cell fate mapping using different bromodeoxyuridine incorporation paradigms revealed enhanced activation of neural stem cells and their consequential exhaustion in the absence of p53 after irradiation. Both p53-knockout and wild-type mice demonstrated similar extent of microglial activation in the hippocampus after irradiation. Impairment of neuronal differentiation of neural progenitors transplanted in irradiated hippocampus was not altered by p53 genotype of the recipient mice. We conclude that by inhibiting neural progenitor activation, p53 serves to mitigate disruption of neuronal development after irradiation independent of apoptosis and perturbation of the neural stem cell niche. These findings suggest for the first time that p53 may have a key role in late effects in brain after irradiation.

Figure 1

Inhibition of hippocampal neurogenesis after irradiation is p53 dependent. There is loss of DCX+ (a and b, 0 Gy; c and d, 17 Gy; DCX, green; DAPI, blue) and calretinin+ cells (e–g, 0 Gy, h–j, 17 Gy; calretinin cells, arrow, green; NeuN, red; DAPI, blue) in SGZ at 9 weeks after irradiation. Arrowhead (e) denotes the normal band of calretinin+ nerve fibers at the inner molecular layer. Newborn neurons in dentate gyrus demonstrate BrdU (k, arrows, green) and NeuN immunostaining (l, red; m, merged). The p53 genotype has an independent effect on the number of BrdU+/NeuN+ cells at 9 weeks after single doses of cranial irradiation (n) or 20 Gy in 5 daily fractions (o). Mice were given BrdU daily for 7 consecutive days 4 weeks after irradiation. Data are expressed as mean±S.E.M. and analyzed with two-way ANOVA with three to five mice per dose per genotype.

Figure 2

Neuroblasts in SGZ undergo p53-dependent apoptosis after irradiation. DCX+ apoptotic cells are identified using TUNEL (a–d, arrows) and caspase-3 immunohistochemistry (e–h, arrows). There is a marked loss of DCX+ cells at 24 h after irradiation (i and j, 0 Gy; k and l, 17 Gy; DCX, green; DAPI, blue). The number of DCX+/TUNEL+ apoptotic cells observed at 8 h is radiation dose and p53 genotype dependent. Data are expressed as mean±S.E.M. and analyzed with a two-way ANOVA with three to five mice per experimental group.

Figure 3

Irradiation results in p53-dependent ablation of type-1 cells in mouse dentate gyrus. A representative newborn type-1 cell (a–d, arrow) demonstrates BrdU incorporation (a, green), and is positive for nestin (b, red) and GFAP (c, yellow; d, merged), and has a characteristic process that traverses the granule cell layer. A proliferating type-1 cell (e–h, arrow) demonstrates immunostaining for Ki67 (e, green), nestin (f, red) and GFAP (g, white; h, merged). At 9 weeks after irradiation, there is p53-dependent reduction of total (i), BrdU+ (j) and Ki67+ type-1 cells (k). Data are expressed as mean±S.E.M. and analyzed with two-way ANOVA with three to four mice per experimental group.

Figure 4

Deficiency in p53 alters neural stem cell and progenitor cell fate after irradiation. In non-irradiated mice, p53 genotype does not alter the decline of BrdU+ type-1 cells over time after BrdU (a). After 5 Gy, the decrease in the number of BrdU+ type-1 cells over time is p53 dependent (b). The decline of BrdU+ type-2 cells over time is independent of p53 genotype in non-irradiated mice (c) and is p53 genotype dependent after 5 Gy (d). The number of BrdU+/DCX+cells over time after BrdU is independent of p53 genotype in non-irradiated mice (e) but p53 genotype dependent after 5 Gy (f). A type-1 BrdU-doublet is observed in SGZ of a p53−/− mouse after irradiation (g, arrow; BrdU, green; nestin, red; GFAP, white). The number of BrdU doublets and type-1 BrdU doublets in SGZ at 2 days after BrdU is p53 genotype dependent following 5 Gy (h and i). BrdU was given at 4 weeks after 0 or 5 Gy, and cell populations determined at 2 h, 2 days, 1 and 4 weeks after BrdU. Data are represented as mean±S.E.M. and analyzed with two-way ANOVA and post hoc Bonferroni test, *P<0.05, **P<0.01, ***P<0.001, p53−/− versus p53+⧸+; ^†P<0.01, 5 Gy versus 0 Gy in p53−/− mice. There was a minimum of three to four mice per genotype per time point.

Figure 5

Deficiency in p53 does not alter microglial activation or inhibition of neuronal differentiation after irradiation. An activated microglia demonstrates nuclear BrdU incorporation and CD68+ (a–d, arrow) or Iba1+ (f–i, arrow). The increase in the number of BrdU+/CD68+ (e) and BrdU+/Iba1+ (j) cells in the dentate gyrus at 9 weeks after cranial irradiation is independent of p53 genotype. An eGFP+ neural progenitor cell transplanted in mouse hippocampus demonstrates immunoreactivity for DCX (k, arrow) and another one for Prox1 (m, arrow). The percentage of eGFP+ cells that expresses DCX or Prox1 is reduced in mice given cranial irradiation before transplantation, independent of p53 genotype of the recipient mice (l, DCX+/eGFP+ cells; n, Prox1+/eGFP+ cells). Data are expressed as mean±S.E.M. and analyzed with two-way ANOVA and post hoc Bonferroni test, *P<0.05, **P<0.01, ***P<0.001, 5 Gy versus 0 Gy; a minimum of three to five mice per experimental group (e and j) and four to seven mice per experimental group (l and n).