7B2 chaperone knockout in APP model mice results in reduced plaque burden

Abstract

Impairment of neuronal proteostasis is a hallmark of Alzheimer's and other neurodegenerative diseases. However, the underlying molecular mechanisms leading to pathogenic protein aggregation, and the role of secretory chaperone proteins in this process, are poorly understood. We have previously shown that the neural-and endocrine-specific secretory chaperone 7B2 potently blocks in vitro fibrillation of Aβ42. To determine whether 7B2 can function as a chaperone in vivo, we measured plaque formation and performed behavioral assays in 7B2-deficient mice in an hAPPswe/PS1dE9 Alzheimer's model mouse background. Surprisingly, immunocytochemical analysis of cortical levels of thioflavin S- and Aβ-reactive plaques showed that APP mice with a partial or complete lack of 7B2 expression exhibited a significantly lower number and burden of thioflavin S-reactive, as well as Aβ-immunoreactive, plaques. However, 7B2 knockout did not affect total brain levels of either soluble or insoluble Aβ. While hAPP model mice performed poorly in the Morris water maze, their brain 7B2 levels did not impact performance. Since 7B2 loss reduced amyloid plaque burden, we conclude that brain 7B2 can impact Aβ disposition in a manner that facilitates plaque formation. These results are reminiscent of prior findings in hAPP model mice lacking the ubiquitous secretory chaperone clusterin.

Figure 1

7B2 loss in hAPP transgenic mice does not affect memory. Mice were trained in a genotype-blind manner on a hidden platform Morris water maze for nine days. (a) Latency time. The average latency time to platform from Day 1 to Day 9 of training is shown. Mice expressing the hAPP transgene showed significantly longer latency times to platform, but the loss of 7B2 did not cause further increases in latency. Symbols above the line graph represent p-values from two-way ANOVA with Tukey Test for the corresponding genotype. *p < 0.05, **p < 0.01, ***p < 0.001 (b) Probe test. One day after training, the mice underwent a single probe test with the platform removed. Wild-type mice lacking the APP transgene spent significantly more time in the NE quadrant, which originally contained the platform, than did hAPP^tg7B2^+/− or hAPP^tg7B2^−/− mice. The red/dashed line denotes 25%, marking equal time in each quadrant. (c) Platform crossings. The average number of times each mouse crossed the former platform zone during the 60-sec probe test. hAPP^tg7B2^+/− and hAPP^tg7B2^−/− showed significantly less platform crossings than wild-type mice lacking the hAPP transgene. (**p < 0.01). (d) Passive avoidance test. One day after the initial shock training, the latency to cross from a brightly-lit compartment to a dark compartment provides an index of memory retention, with a longer latency representing a stronger memory association. While certain groups showed significant memory retention, no significant differences were found between any 7B2 genotype group. Results are given as the mean ± SEM., n = 12 per group (*p < 0.05; **p < 0.01). No significant differences were found between sexes; thus, all results were pooled.

Figure 2

7B2 content does not change as a function of hAPP transgene status; clusterin levels are not altered as a function of 7B2 genotype. Hemibrains were extracted in acid, and aliquots of the clarified supernatants dried and subject to radioimmunoassay for 7B2 (a). In (b), the ratio of total clusterin to total actin in aliquots of RIPA-buffer extracted hemibrains is shown. Results indicate the mean+/− the SD for 5–8 mice per group in panel (a) and 5 mice per group in panel (b).

Figure 3

7B2 loss does not impact total brain levels of insoluble and soluble Aβ. Levels of soluble Aβ-40 (a) and Aβ-42 (b) and insoluble Aβ-40 (c) and Aβ-42 (d) were assessed by ELISA of RIPA-extracted hemibrain homogenates. Data are shown as the mean ± SD; n = 12 mice per genotype. No significant differences were found between sexes, nor between genotypes.

Figure 4

7B2 loss does not alter the level of hAPP in hemibrain homogenates. (a) Composite Western blot images show examples of full-length APP, the CTFβ and CTFα fragments, and actin expression in hAPP^tg mice of various 7B2 genotypes (7B2^+/+, 7B2^+/− and 7B2^−/−). Raw corresponding Western blot images can be found in Supplemental Fig. 1. (b) FL-APP Analysis. Full-length APP was quantified using densitometric analysis and normalized to the actin loading control. No significant differences between the three genotypes were found. (c) CTF Analysis. The ratio between CTFβ and CTFα was compared between the three different genotypes using densitometric analysis. Results depict the mean± the S.E. of 12 mice per genotype. No significant differences were found between sexes, nor between genotypes.

Figure 5

Staining of mouse cortex reveals decreased plaque burden in 7B2 knockout mice. (a) Thioflavin-S staining and anti-β-amyloid immunostaining (4G8) were performed on sections of 12-month old mouse cortex expressing the hAPP transgene with the listed 7B2 genotypes. Representative sections from a rostral, dorsal portion of the cortex are shown for each genotype with a small subset (red box) enlarged below. Images are show in inverted contrast, with positive objects as black to enhance visibility. While all mice developed Thio-S and 4G8 positive staining, 7B2^+/− and 7B2^−/− mice showed a significantly decreased load for each stain as compared to 7B2^+/+. The percent area of the cortex positive for Thio-S or 4G8 staining was quantified in panels (b) and (c) respectively. The amount of cortical area covered by Thio-S or 4G8 staining was significantly greater in 7B2^+/+ mice as compared to 7B2^+/− and 7B2^−/− mice. (d) The ratio of the Thio-S load to the 4G8 load showed no significant change between genotypes. (e) The average size of Thio-S- and (f) 4G8-positive objects. The mean ± SEM, of the total load per slice (panels b and c) and of the average size (panels e and f) is shown. Data were analyzed using a one-way ANOVA, Newman-Keuls test; *p < 0.05, **p < 0.01,***p < 0.001). Bar = 300 µm.