Age-related changes of neuron numbers in the frontal cortex of a transgenic mouse model of Alzheimer's disease

Abstract

Alzheimer's disease (AD) is a neurodegenerative disorder, characterized by amyloid plaque accumulation, intracellular tangles and neuronal loss in selective brain regions. The frontal cortex, important for executive functioning, is one of the regions that are affected. Here, we investigated the neurodegenerative effects of mutant human amyloid precursor protein (APP) and presenilin 1 (PS1) on frontal cortex neurons in APP/PS1KI mice, a transgenic mouse model of AD, expressing two mutations in the human APP, as well as two human PS1 mutations knocked-in into the mouse PS1 gene in a homozygous (ho) manner. Although the hippocampus is significantly affected in these mice, very little is known about the effects of these mutations on selective neuronal populations and plaque load in the frontal cortex. In this study, cytoarchitectural changes were characterized using high precision design-based stereology to evaluate plaque load, total neuron numbers, as well as total numbers of parvalbumin- (PV) and calretinin- (CR) immunoreactive (ir) neurons in the frontal cortex of 2- and 10-month-old APP/PS1KI mice. The frontal cortex was divided into two subfields: layers II-IV and layers V-VI, the latter of which showed substantially more extracellular amyloid-beta aggregates. We found a 34% neuron loss in layers V-VI in the frontal cortex of 10-month-old APP/PS1KI mice compared to 2-month-old, while there was no change in PV- and CR-ir neurons in these mice. In addition, the plaque load in layers V-VI of 10-month-old APP/PS1KI mice was only 11% and did not fully account for the extent of neuronal loss. Interestingly, an increase was found in the total number of PV-ir neurons in all frontal cortical layers of single transgenic APP mice and in layers II-IV of single transgenic PS1ho mice between 2 and 10 months of age. In conclusion, the APP/PS1KI mice provide novel insights into the regional selective vulnerability in the frontal cortex during AD that, together with previous findings in the hippocampus, are remarkably similar to the human situation.

Fig. 1

Mean and standard error of the mean (SEM) of volumes of layers II–IV (a), layers V–VI (b), and the entire frontal cortex (c) of 2-month-old (open bars) and 10-month-old (closed bars) APP mice, PS1he mice, PS1ho mice, and APP/PS1KI mice. Statistical analysis did not reveal significant differences among the groups of mice (p < 0.05)

Fig. 2

Mean and standard error of the mean (SEM) of numbers of NeuN-ir neurons (a, b), CR-ir neurons (c, d), and PV-ir neurons (e, f) within layers II–IV (a, c, e) and layers V–VI (b, d, f) of 2-month-old (open bars) and 10-month-old (closed bars) APP mice, PS1he mice, PS1ho mice, and APP/PS1KI mice. Statistical analysis revealed significant (p _ANOVA < 0.05) differences among the groups of mice in b (numbers of NeuN-ir neurons within layers V–VI), e (numbers of PV-ir neurons within layers II–IV), and f (numbers of PV-ir neurons within layers V–VI), respectively. Results of post-hoc Bonferroni’s multiple comparison test for pair-wise comparisons are indicated by asterisks (*p < 0.05)

Fig. 3

Representative photomicrographs of coronal sections showing the frontal cortex of 2- (a) and 10-month-old (b) APP/PS1KI mice, immunoprocessed for the detection of NeuN. Note the age-related increase in the plaque load (holes, indicated by arrows). The scale bar represents 100 μm

Fig. 4

Representative photomicrograph of CV-ir neurons (red) and PV-ir neurons (green) in the frontal cortex of a 2-month-old PS1ho mouse. The picture was created by montaging a total of 573 high-resolution fields-of-view into a single two-dimensional virtual slide. The scale bar represents 200 μm

Fig. 5

Mean and standard error of the mean (SEM) of the plaque load within layers II–IV (a, c) and layers V–VI (b, d) of the frontal cortex of 2-month-old (open bars) and 10-month-old (closed bars) APP mice and APP/PS1KI mice, analyzed by immunohistochemical detection of amyloid-beta (a, b) or staining of the plaque cores with thioflavin S (c, d), respectively. Statistical analysis revealed significant (p _ANOVA < 0.05) differences among the groups of mice in a–d. Results of post-hoc Bonferroni’s multiple comparison test for pair-wise comparisons are indicated by asterisks (**p < 0.01; ***p < 0.001). PS1he and PS1ho mice did not exhibit amyloid plaques. The SEM of APP mice (all graphs) and 10-month-old APP/PS1 mice are not shown because the interindividual variation was too small

Fig. 6

Representative photomicrographs of plaque load in layers V–VI of the frontal cortex of 2- and 10-month-old APP (a, b) and APP/PS1KI mice (c, d). Amyloid-beta is shown in red, thioflavin S in green and Hoechst in blue. Since PS1he and PS1ho mice do not have plaques, these mice were not included in the figure. Each image is the maximum intensity projection of a complete confocal image stack of 15 μm thick on average, with a distance between the single images of 0.5 μm. The scale bar represents 50 μm