Animals
The EFAD mice used in this study are hemizygous for 5xFAD transgenes and homozygous for targeted replacement of mouse APOE with human APOE3 (E3FAD) or APOE4 (E4FAD) (Youmans et al., 2012). Mice were euthanized at 6 months old, an age characterized by significant AD-related neuropathology (Tai et al., 2017). After mice were perfused with 4°C PBS, the brains were extracted and fixed in 4% paraformaldehyde for 48 hours. Four groups of mice were studied (n = 6 per group): male and female E3FAD, male and female E4FAD. This study was performed under an institutionally approved animal protocol and in accordance with the guidelines of the National Institutes of Health.
Histochemistry
Fixed brains were sectioned (40 μm) in the sagittal plane with at least three medial, equidistantly spaced sections per brain stained using modifications of previously described protocols (Stephen et al., 2019). Staining batches were balanced across groups. In brief, sections were permeabilized in Triton X-100 for 15 minutes, followed by incubation at 4°C with primary antibodies (diluted in blocking buffer) against glial fibrillary acid protein (GFAP, DAKO, 1:500), lysosomal-associated membrane protein 1 (LAMP1) (DHSB, 1:250), and or Ab (MOAB-2, Sigma-Aldrich, 1:100). After subsequent washing, sections were incubated with Alexa fluorophore-conjugated secondary antibodies (Invitrogen; anti-rabbit, anti-mouse, anti-rat) diluted 1:500 in blocking buffer. To label amyloidogenic plaques, immunostained sections were incubated with 0.5 % THK-265 (THK; Sigma-Aldrich) for 20 minutes, washed with PBS, then mounted with ProLong Gold Antifade medium (Vectashield).
Microscopy and image analyses
Image collection and analysis was performed as described in Stephen et al., (2019) except as noted. In brief, a Zeiss-780 upright confocal microscope with ZEN imaging software (Zeiss) was used for image capture with researcher-blinded acquisition. Laser and detector settings were maintained across imaging sessions and high-resolution z-stack images were collected with optimal section depths (~0.35 µm). A 63x oil immersion objective (1.4 NA) was used to acquire regions of interest (ROI, 192.8 µm x 192.8 µm, 512 x 512 pixels, 16 bit) in the subiculum and cornu ammonis 1 (CA1) stratum radiatum fields of hippocampus, sampling areas with individual plaques. Analyses were performed using a custom ImageJ blinding plugin (Stephen et al., 2015). Images were de-noised and average projections were used for analysis.
Plaque coverage, size, and compaction were quantified from all plaques >4 µm in diameter that were fully contained within the ROIs (> 3 per section, 3 sections per animal) and did not overlap with other plaques; 2-10 plaques were analyzed per animal. Plaque coverage was defined as the contact area between astrocyte processes and THK-265+ plaques (within 2 µm), calculated by summing arcs of plaque perimeters across three-dimensional stacks. Plaque area was manually determined in ImageJ and ranged from 10-108 µm2. Plaque circularity (a measure of compaction) was determined using the formula 4π x area / (perimeter)2.
Neuronal dystrophy was determined as a ration quantifying of LAMP1 lysosomal staining density normalized to the corresponding THK-265+ plaque area. Analyses of individual astrocytes (3-41 cells/animal) included all non-overlapping GFAP-immunoreactive cells fully within the ROIs and within a 100 µm radius of THK-265+ plaques. Soma size was measured by manually identifying, outlining, and measuring GFAP-immunoreactive cell bodies using ImageJ. Astrocyte primary process number was manually determined as the number of processes emanating directly from GFAP-labeled somas as previously described (Sun and Jakobs, 2012).
Statistical Analyses
Two-way analysis of variance, with APOE genotype and sex as independent variables was performed using Prism (GraphPad Software, Inc. version 9), followed by Tukey post-hoc test to account for multiple comparisons. Kolmogorov-Smirnov and Shapiro-Wilk test was used to test sample normality distribution. Kruskal-Wallis test was used to compare differences between groups whose distributions did not pass normality testing. Data are presented as box (mean and 25th and 75th quartiles) and whisker (minimum and maximum values) plots or as means (+SEM). For all statistical tests, p values less than 0.05 were considered significant.