Ovarian Cycle Stages Modulate Alzheimer-Related Cognitive and Brain Network Alterations in Female Mice

Altered ovarian cycling in female hAPP mice. A, Representative images of vaginal cytology classified as proestrus (nucleated epithelial cells), estrus (cornified squamous epithelial cells), metestrus (mix of cell types), or diestrus (leukocytes). Scale bar, 50 μm. B, Ratio of estradiol (pg/ml) to progesterone (ng/ml; E/P) at each stage of the mouse estrous cycle, adapted from characterized hormone levels (Walmer et al., 1992; Fig. 1-1). C, Quantification of the number of ovarian cycles spanning ∼3 weeks in NTG and hAPP mice. D, Quantification of cycle length in days for NTG and hAPP mice during this same period. E, Percentage time spent in stages of the cycle with a High E/P ratio (proestrus, estrus). **p < 0.01^c. F, Number of pups per litter born to NTG and hAPP female mice. G, Representative estrous cycles in female NTG and hAPP mice. The dashed line represents a demarcation between High E/P and Low E/P cycle stages (n = 12–18 mice per genotype for C–F, age 2.5–4 months). Data are mean ± SEM. See also Figure 1-1.

Proestrus increases spontaneous epileptiform activity in female hAPP mice. NTG and hAPP female mice were cycled to determine ovarian cycle stage (Pro, Est, Met, Di) and another group was Gnx. All groups underwent network activity analysis with EEG. A, Representative EEG traces from NTG and hAPP female mice during Low E/P (metestrus/diestrus), High E/P (proestrus/estrus), and Gnx states. B, Number of spikes graphed by hour from 15:00 to 05:00 in proestrus (red), Low E/P (metestrus/diestrus; blue), and Gnx (black) hAPP mice. Mixed-model ANOVA: ovarian status by hour interaction p < 0.01^e1, ovarian status effect ***p < 0.0001^e2; p < 0.0001^e3 High E/P (proestrus) vs Low E/P (metestrus/diestrus), p < 0.05^e4 High E/P (proestrus) versus Gnx. C, Number of spikes during representative hours before (15:00) and after (03:00) E₂ surge. One-way ANOVA: after E₂ surge, p = 0.01^f3, *p < 0.05 versus Low E/P (metestrus/diestrus)^f4 and Gnx^f5 or as indicated by bracket^f1. (n = 3–5 mice per genotype, age 2.5–4 months). Data are mean ± SEM.

High E/P ovarian cycle stages, proestrus and estrus, worsen fear memory in female hAPP but not NTG mice. NTG and hAPP female mice were cycled to determine ovarian cycle stage (Pro, Est, Met, Di) and another group was Gnx. All groups were tested in the passive avoidance task. A, Latency to enter the dark chamber during training in NTG and hAPP female mice by ovarian status: High E/P (proestrus/estrus), Low E/P (metestrus/diestrus), and Gnx. B, Latency to enter the dark chamber during testing (minus latency during training) in NTG and hAPP female mice by ovarian status. Two-way ANOVA: ovarian status by genotype interaction, p < 0.05^h1, *p < 0.05 versus NTG^h5 or as indicated by bracket^h4. Dashed line indicates average of Gnx group. C, Correlation between percentage of time spent in High E/P (proestrus/estrus) over a 3 week period and latency to enter the dark chamber in intact mice; R ² = 0.714. (n = 4–9 mice per group for A and B; n = 9–12 mice per group for C; age 2.5-4 months). Data are mean ± SEM.

Proestrus and estrus, High E/P ovarian cycle stages, worsen spatial and working memory in female hAPP but not NTG mice. NTG and hAPP female mice were cycled to determine ovarian cycle stage (Pro, Est, Met, Di). All groups were then tested in the two-trial Y-maze that assesses exploration of a familiar and novel arm of the maze. A, Distance (meters) traveled in novel (black bars) and familiar (white bars) arms of the two-trial Y-maze during testing. Total distance did not differ across groups (Fig. 4-1). B, Percentage of time spent in the novel arm. **p < 0.01^j1-2, *p < 0.05^j5 as indicated by brackets (A) or versus NTG values (two-tailed one-sample t tests) in B. Dashed gray line is NTG average of percentage time in novel arm (n = 5–13 mice per group, age 3.5–5.5 months). Data are mean ± SEM. See also Figure 4-1.

Proestrus and estrus, High E/P cycle stages, worsen locomotor hyperactivity in female hAPP but not NTG mice as detected in the context of an open field. NTG and hAPP female mice were cycled to determine ovarian cycle stage (Pro, Est, Met, Di) and another group was Gnx. All groups underwent testing in the open-field apparatus. Total number of movements during exploration of an open field in NTG and hAPP female mice by ovarian status: High E/P (proestrus/estrus), Low E/P (metestrus/diestrus), and Gnx. Two-way ANOVA: ovarian status effect, p < 0.05^k2; genotype effect, p < 0.0001^k3. *p < 0.05 as indicated by bracket^k4. (n = 3–11 mice per group, age 2.5–4 months). Data are mean ± SEM.

Levels of Aβ_1-42 surge in the hippocampus of hAPP female mice during proestrus. NTG and hAPP female mice were cycled to determine ovarian cycle stage (Pro, Est, Met, Di) and another group was Gnx. Hippocampal homogenates were assessed for hAPP and pathogenic proteins related to AD. A, Aβ_1-42 levels determined by ELISA in hAPP mice at each stage of the ovarian cycle. B, Aβ_1-42 levels during metestrus/diestrus compared to proestrus (n = 4–13 mice per group, age 2.5–4 months). **p < 0.01 versus metestrus/diestrus.^l C, Relative hAPP mRNA levels determined by qPCR in hAPP mice at each stage of the ovarian cycle. D, Relative hAPP mRNA levels during metestrus/diestrus compared to proestrus (n = 4–14 mice per group, age 2.5–4 months). E, Representative Western blot showing hippocampal levels of hAPP, total mouse tau, phosphorylated mouse tau, and the loading control actin, in metestrus/diestrus, proestrus, and Gnx NTG and hAPP mice. F–H, Relative hAPP (F), total Tau (G), and phosphorylated Tau (H) protein levels determined by Western blot in hAPP mice during metestrus/diestrus compared to proestrus. All values are relative to Gnx hAPP levels as represented by dashed gray lines. (n = 7–13 mice per group for F–H; age 2.5–4 months). Data are mean ± SEM.