Cerebrovascular alterations in a mouse model of late-onset Alzheimer's disease

Abstract

Significance: Alzheimer's disease (AD) is an age-related neurodegenerative disorder with cerebrovascular alterations contributing to cognitive decline. Assessing cerebrovascular changes in mouse models that mimic the human condition of late-onset, sporadic AD is important for better human applicability.

Aim: To assess cerebrovascular changes in three mouse models: (1) 3xTg-AD; (2) the humanized amyloid-beta knock-in ( $\mathrm{hA}\beta$ -KI) mouse model of late-onset, sporadic AD; and (3) age-matched wild-type mice.

Approach: We measured resting-state cerebral blood flow (CBF) and neurovascular coupling (NVC) using laser speckle imaging (LSI) and performed ex vivo analyses of gene expression and cerebrovascular structure using bulk ribonucleic acid sequencing and confocal microscopy, respectively.

Results: Our study identifies specific cerebrovascular alterations in the $\mathrm{hA}\beta$ -KI mouse model, including increased resting-state CBF, a shift toward smaller blood vessel diameters, impaired NVC, and transcriptomic changes related to metabolism and inflammation. Notably, we found that the increased resting-state CBF was primarily associated with female $\mathrm{hA}\beta$ -KI mice.

Conclusions: Our findings demonstrate that the $\mathrm{hA}\beta$ -KI mouse model exhibits cerebrovascular alterations that warrant further investigation to uncover the underlying mechanisms. Expanding these studies could enhance our understanding of cerebrovascular alterations in AD and support the development of targeted therapeutic strategies.

Keywords: Alzheimer’s disease; RNA sequencing; cerebral blood flow; laser speckle contrast imaging; microscopy; neurovascular coupling.

Fig. 1

Experimental cohorts and timeline. (a) Mouse numbers used in cohort #1 to examine resting-state CBF in 12 and 18 m.o. WT and $\mathrm{hA}\beta$-KI mice from the same genetic background and 22 to 27 m.o. WT ($\mathrm{hA}\beta$-KI and 3xTg-AD mouse backgrounds), $\mathrm{hA}\beta$-KI, and 3xTg-AD mice. (b) Mouse numbers used in cohort #2 to examine RNA sequencing data in 23 to 25 m.o. WT ($\mathrm{hA}\beta$-KI background) and $\mathrm{hA}\beta$-KI mice. (c) Mouse numbers used in cohort #3 to examine NVC and cerebrovasculature in 24 m.o. WT ($\mathrm{hA}\beta$-KI background) and $\mathrm{hA}\beta$-KI mice. (d) Experimental timeline of the measurements performed for each cohort.

Fig. 2

$\mathrm{hA}\beta$-KI mice have increased resting-state CBF, whereas WT mice trend toward increased resting-state CBF with increasing age. Comparing resting-state CBF between 12, 18, and 22 to 27 m.o. (a) WT and (b) $\mathrm{hA}\beta$-KI mice (two-way ANOVA, Tukey post hoc test). Representative images from (c) wild type and (d) $\mathrm{hA}\beta$-KI mice, where the mice selected had the median resting-state CBF from each age and genotype. $P$-values are shown for each comparison, with a $p<0.05$ considered statistically significant (red text).

Fig. 3

Male and female $\mathrm{hA}\beta$-KI mice have differing resting-state CBF trajectories compared with male and female WT mice. (a)–(c) (left) Comparing resting-state CBF between 12, 18, and 22 to 27 m.o. for (a) male, WT and (b) female, WT mice (two-way ANOVA, Tukey post hoc test). (c) Comparing resting-state CBF at 12, 18, and 22 to 27 m.o. between male and female WT mice (two-way ANOVA, Sidak post hoc test). (d)–(f) (right) Comparing resting-state CBF between 12, 18, and 22 to 27 m.o. for (d) male, $\mathrm{hA}\beta$-KI and (e) female, $\mathrm{hA}\beta$-KI mice (two-way ANOVA, Tukey post hoc test). (f) Comparing resting-state CBF at 12, 18, and 22 to 27 m.o. between male and female $\mathrm{hA}\beta$-KI mice (two-way ANOVA, Sidak post hoc test). $P$-values are shown for each comparison, with a $p<0.05$ considered statistically significant (red text) and $p<0.15$ considered trending (blue text).

Fig. 4

Assessing resting-state CBF differences among $\mathrm{hA}\beta$-KI, WT, and 3xTg-AD mice. (a) Comparing resting-state CBF between WT and $\mathrm{hA}\beta$-KI mice at 12, 18, and 22 to 27 m.o. (two-way ANOVA, Sidak post hoc test). (b) Comparing resting-state CBF between WT, 3xTg-AD, and $\mathrm{hA}\beta$-KI mice at 22 to 27 m.o. (one-way ANOVA, Tukey post hoc test). (c) Representative images from (left) wild type, (middle) 3xTg-AD, and (right) $\mathrm{hA}\beta$-KI mice, where the mice selected had the median resting-state CBF from each genotype. (d) Summarizing the temporal evolution of CBF in WT, $\mathrm{hA}\beta$-KI, and 3xTg mice by fitting a logarithmic curve through the mean CBF of each group. Solid lines are to only guide the eye because we do not have data at all these ages. $P$-values are shown for each comparison, with a $p<0.05$ considered statistically significant (red text) and $p<0.15$ considered trending (blue text).

Fig. 5

Male and female 3xTg-AD mice have reduced resting-state CBF, whereas female $\mathrm{hA}\beta$-KI mice drive resting-state CBF trends compared to WT mice. (a)–(c) (left) Resting-state CBF in male mice. (a) Comparing resting-state CBF between WT and $\mathrm{hA}\beta$-KI male mice at 12, 18, and 22 to 27 m.o. (two-way ANOVA, Sidak post hoc test) (b). Comparing resting-state CBF between WT, 3xTg-AD, and $\mathrm{hA}\beta$-KI male mice at 22 to 27 m.o. (two-way ANOVA, Tukey post hoc test). (c) Summarizing the temporal evolution of CBF in WT, $\mathrm{hA}\beta$-KI, and 3xTg male mice by fitting a logarithmic curve through the mean CBF of each group. Solid lines are to only guide the eye because we do not have data at all these ages. (d)–(f) (right) Resting-state CBF in female mice. (d) Comparing resting-state CBF between WT and $\mathrm{hA}\beta$-KI female mice at 12, 18, and 22 to 27 m.o. (two-way ANOVA, Sidak post hoc test). (e) Comparing resting-state CBF between WT, 3xTg-AD, and $\mathrm{hA}\beta$-KI female mice at 22 to 27 m.o. (two-way ANOVA, Tukey post hoc test). (f) Summarizing the temporal evolution of CBF in WT, $\mathrm{hA}\beta$-KI, and 3xTg female mice by fitting a logarithmic curve through the mean CBF of each group. Solid lines are to only guide the eye because we do not have data at all these ages. $P$-values are shown for each comparison, with a $p<0.05$ considered statistically significant (red text) and $p<0.15$ considered trending (blue text).

Fig. 6

Transcriptomic differences between 23 and 25 m.o. WT and $\mathrm{hA}\beta$-KI mice. (a) Volcano plot and (b) heatmap showing differentially expressed genes with $p<0.05$ and $\mathrm{FDR}<0.1$. Blue shows downregulated gene expression, and red shows upregulated gene expression.

Fig. 7

Comparison of cerebrovascular structure between 24 m.o. WT and $\mathrm{hA}\beta$-KI mice. (a) Comparing the vascular density between WT (blue circle = male; pink diamond = female) and $\mathrm{hA}\beta$-KI (blue square = male; pink triangle = female) mice ($t$-test). (b) Distribution of vessel diameters in WT (red, $n=7$) and $\mathrm{hA}\beta$-KI (black, $n=6$) mice. Shaded areas represent the standard deviation across all mice within each respective group.

Fig. 8

$\mathrm{hA}\beta$-KI mice have altered neurovascular coupling (NVC). (a) Averaged percent change in CBF during hindpaw stimulation to assess NVC in WT (red, $n=7$) and $\mathrm{hA}\beta$-KI (black, $n=5$) mice. Shaded areas represent the standard deviation across all mice within each respective group. (b) Comparing the maximum percent increase in CBF due to hindpaw stimulation between WT (blue circle = male; pink diamond = female) and $\mathrm{hA}\beta$-KI (blue square = male; pink triangle = female) mice ($t$-test). (c) Comparing the area under the curve (AUC) of CBF during hindpaw stimulation between WT (blue circle = male; pink diamond = female) and $\mathrm{hA}\beta$-KI (blue square = male; pink triangle = female) mice ($t$-test). $P$-values are shown for each comparison, with a $p<0.05$ considered statistically significant (red text).