Alzheimer's-type amyloidosis in transgenic mice impairs survival of newborn neurons derived from adult hippocampal neurogenesis

Abstract

Alzheimer's disease (AD) is characterized by severe neuronal loss in several brain regions important for learning and memory. Of the structures affected by AD, the hippocampus is unique in continuing to produce new neurons throughout life. Mounting evidence indicates that hippocampal neurogenesis contributes to the processing and storage of new information and that deficits in the production of new neurons may impair learning and memory. Here, we examine whether the overproduction of amyloid-beta (Abeta) peptide in a mouse model for AD might be detrimental to newborn neurons in the hippocampus. We used transgenic mice overexpressing familial AD variants of amyloid precursor protein (APP) and/or presenilin-1 to test how the level (moderate or high) and the aggregation state (soluble or deposited) of Abeta impacts the proliferation and survival of new hippocampal neurons. Although proliferation and short-term survival of neural progenitors in the hippocampus was unaffected by APP/Abeta overproduction, survival of newborn cells 4 weeks later was dramatically diminished in transgenic mice with Alzheimer's-type amyloid pathology. Phenotypic analysis of the surviving population revealed a specific reduction in newborn neurons. Our data indicate that overproduction of Abeta and the consequent appearance of amyloid plaques cause an overall reduction in the number of adult-generated hippocampal neurons. Diminished capacity for hippocampal neuron replacement may contribute to the cognitive decline observed in these mice.

Figure 1.

Amyloid plaques are specific to APP/PS1 mice at 6 months of age. Immunohistochemistry against Aβ reveals widespread amyloid pathology in 6-month-old APP/PS1 mice (A), whereas no Aβ plaques are present in APP single-transgenic mice at this age (B). Scale bars, 150 μm.

Figure 2.

Proliferation of hippocampal progenitor cells is unchanged by transgenic expression of mutant APP and/or PS1. A–D, Ki67 immunostaining in the DG is similar in all four genotypes. Nuclear Fast Red was used as a counterstain to identify morphological boundaries of the GCL and SGZ. Scale bars, 30 μm. E, The number of Ki67+ cells in the DG was counted for each genotype (mean ± SEM; n = 4 for each group). The number of Ki67+ cells is similar across all four genotypes; overproduction of APP/Aβ had no significant effect on progenitor cell proliferation in the SGZ.

Figure 3.

Short-term survival of newborn cells in the adult DG is not affected by APP/Aβ overproduction. A–D, One day after the last of 12 daily BrdU injections, BrdU immunostaining in the DG is similar in all four genotypes. Nuclear Fast Red was used as a counterstain to identify morphological boundaries of the GCL and SGZ. Scale bars, 30 μm. E, The absolute number of BrdU+ cells in the DG is shown for each genotype (mean ± SEM; n = 9–12 for each group). Overproduction of APP/Aβ had no significant effect on the number of BrdU+ cells that were labeled 1–12 d earlier. Similarly, transgenic expression of mutant PS1 had no effect on early survival within the DG, either alone or when coexpressed with APPswe.

Figure 4.

Late survival of newborn cells in the adult DG is dramatically reduced in APP/PS1 mice. A–D, Thirty days after the final BrdU injection, the number of BrdU-immunoreactive cells is noticeably diminished in the DG of double-transgenic APP/PS1 mice. Nuclear Fast Red was used as a counterstain to identify morphological boundaries of the GCL and SGZ. Scale bars, 30 μm. E, The absolute number of BrdU+ cells in the DG is shown for each genotype (mean ± SEM; n = 10–12 per group). Overproduction of APP/Aβ reduced the number of labeled cells surviving 30 d after the final BrdU injection in APP/PS1 double-transgenic animals (***p < 0.001 vs NTG; ^###p < 0.001 vs PS1; ANOVA with Tukey's post hoc). Neither APP nor PS1 by themselves had any effect on the survival of newborn DG cells compared with NTG (APP vs PS1, ^#p < 0.05).

Figure 5.

Overproduction of APP/Aβ specifically diminishes survival of newborn neurons in APP and APP/PS1 mice. A–C, E–G, Confocal analysis was used to score the coexpression of NeuN (green; C, G) and S100β (blue; B, F) in BrdU+ cells (red; A, E) from each genotype. D, H, Arrows in the merged images indicate BrdU+/NeuN+ neurons (D, H), the asterisk identifies a BrdU+/S100β+ astrocyte (H), and arrowheads indicate BrdU+ cells coexpressing neither marker (D). I, Distribution of phenotypes in BrdU+ cells by genotype. Compared with NTG, a significantly smaller number of BrdU+ cells colabel with NeuN (green) in the hippocampus of both APP and APP/PS1 transgenic mice (***p < 0.01 vs NTG; ANOVA with Tukey's post hoc). In contrast, the number of newborn S100β+ astrocytes (blue) and newborn cells expressing neither neuronal nor glial markers (red) is unchanged by overproduction of APP/Aβ. Error bars indicate SEM.

Figure 6.

Overproduction of APP/Aβ exacerbates cell death of newborn neurons as they approach maturity. A–C, Confocal analysis was used to score the coexpression of NeuN (blue; C) and DCX (green; B) in BrdU+ cells (red; A) in APP/PS1 and NTG mice. D, The arrowhead in the merged image identifies a BrdU+/DCX+/NeuN− immature neuron. E, The number of BrdU+ cells coexpressing the immature neuronal precursor marker DCX (green) is smaller in APP/PS1 than in NTG mice (*p < 0.05; ANOVA with Tukey's post hoc). The decrease in BrdU+ cells coexpressing the postmitotic neuronal marker NeuN (blue) in APP/PS1 mice is even more dramatic than the loss of DCX+ cells (***p < 0.001; ANOVA with Tukey's post hoc). In contrast, a similar number of newborn cells express neither marker (red) in APP/PS1 and NTG mice. Error bars indicate SEM.