Fyn knock-down increases Aβ, decreases phospho-tau, and worsens spatial learning in 3×Tg-AD mice

Abstract

Fyn kinase phosphorylates tau and exacerbates amyloid beta (Aβ)-mediated synaptic dysfunction. However, Fyn also increases the nonpathological cleavage of amyloid precursor protein (APP), suggesting opposing roles for Fyn in the pathogenesis of Alzheimer's disease (AD). To determine the effect of Fyn on both Aβ and tau pathologies, we crossed homozygous Alzheimer's disease triple transgenic (3×Tg) mice harboring mutations in amyloid precursor protein, presenilin-1, and tau with wild-type or Fyn knockout mice to generate Fyn(+/+)3×Tg(+/-) or Fyn(+/-)3×Tg(+/-) mice. We found that Fyn(+/-)3×Tg(+/-) mice had increased soluble and intracellular Aβ, and these changes were accompanied by impaired performance on the Morris water maze at 18 months. Fyn(+/-)3×Tg(+/-) mice had decreased phosphorylated tau at 15-18 months (as did Fyn knockout mice), but Fyn(+/-)3×Tg(+/-) mice had increased phosphorylated tau by 24 months. In addition, we observed that Fyn(+/-)3×Tg(+/-) males were delayed in developing Aβ pathology compared with females, and displayed better spatial learning performance at 18 months. Overall, these findings suggest that loss of Fyn at early stages of disease increases soluble Aβ accumulation and worsens spatial learning in the absence of changes in tau phosphorylation.

Figure 1. Endogenous tau phosphorylation is decreased in Fyn knock-out mice

A. Brain lysates from wild-type or Fyn homozygous knock-out mice were Western blotted for phosphorylated tau epitopes Ser202/Thr205 (AT8), Thr231 (AT180), or Thr181 (AT270) (top 3 panels) or total tau (bottom panel). B. Quantification of data in A. shows Fyn knock-out mice had significantly decreased tau phosphorylation as detected by AT8 (36%, p<0.05), AT180 (quantification of both bands: 47%, p<0.05), and AT270 (56%, p<0.001).

Figure 2. Fyn levels are reduced in Fyn^+/−3xTg^+/− mice

A. Brain lysates from 15, 18, 21, and 24 month old female Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were Western blotted for Fyn. Female Fyn^+/−3xTg^+/− mice showed a 34% decrease at 15 months (p<0.001), a 39% decrease at 18 months (p<0.05), a 60% decrease at 21 months (p<0.01), and a non-significant decrease at 24 months. B. 18, 21, and 24 month old male Fyn^+/+3xTg^+/− and Fyn^+/− 3xTg^+/− mice were Western blotted for Fyn. Fyn^+/−3xTg^+/− mice had a 74% decrease in Fyn levels at 18 months (p<0.001), a 56% decrease at 21 months (p<0.05), and a 55% decrease at 24 months (p<0.01).

Figure 3. Soluble Aβ1–40 is increased in Fyn^+/−3xTg^+/− mice

A. Soluble Aβ1–40 was extracted from 15, 18, 21, and 24 month old female Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mouse brain homogenates and measured by ELISA. Soluble Aβ1–40 was significantly increased in Fyn^+/−3xTg^+/− mice at 15 (28%, p<0.05), 18 (60%, p<0.01), and 21 months (39%, p<0.05), but not at 24 months. B. Insoluble Aβ1–40 was extracted with formic acid (FA), and was significantly increased in 21 month old Fyn^+/−3xTg^+/− mice by 46% (p<0.05). C. Soluble Aβ 1–42 levels were not significantly different at any age. D. The ratio of soluble Aβ42/40 was calculated, and was unchanged at all ages except at 18 months, when it was significantly decreased (37%, p<0.01). E. Soluble Aβ1–40 was measured from 18, 21, and 24 month old male Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice. 21 month old Fyn^+/−3xTg^+/− mice had significantly increased soluble Aβ1–40 by 46% (p<0.05). 24 month old Fyn^+/−3xTg^+/− mice had a trend towards increased soluble (35%, p=0.19) Aβ1–40. F. Insoluble (FA) Aβ1–40 was measured from 18, 21, and 24 month male Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice. There were no significant increases in insoluble Aβ1–40 at any age, but a trend towards increased insoluble Aβ1–40 at 24 months (46%, p=0.12). G. Soluble Aβ1–42 levels were measured from male mice and not found to be significantly different at any age. H. Aβ42/40 ratios were calculated and found to be not significantly different at any age.

Figure 4. 6E10 positive cells are increased in female Fyn^+/−3xTg^+/− mice

A–B. Brain sections from 18 month old female Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were stained with antibody 6E10 to detect APP/Aβ. Fyn^+/−3xTg^+/− mice had greater intracellular 6E10 immunoreactivity in the hippocampus compared to controls. Scale bar= 20µm. C. Stereological quantification of 6E10 positive cells in the CA1 and subiculum regions of the hippocampus in Fyn^+/+3xTg^+/− (n=2) and Fyn^+/−3xTg^+/− (n=3) mice. Fyn^+/−3xTg^+/− mice had a significant increase in 6E10 positive cells by 75% (p<0.05).

Figure 5. Tau phosphorylation is decreased in Fyn^+/−3xTg^+/− mice at 15 months and increased at 24 months

A. Brain homogenates from 15, 18, 21, and 24 month old female Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were Western blotted for phosphorylated tau by antibody AT8 or AT270. Fyn^+/−3xTg^+/− mice showed significantly decreased phospho-tau at 15 months (34%, p<0.05), no significant change at 18 or 21 months, and significantly increased phospho-tau at 24 months (122%, p<0.05). All values were normalized to total tau levels. B. Brain homogenates from 15, 18, 21, and 24 month old female Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were Western blotted for total tau. There were no significant differences in total tau at any age. C. Brain homogenates from 18, 21, and 24 month old male Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were Western blotted for phosphorylated tau with antibody AT8 or AT270. 24 month old Fyn^+/−3xTg^+/− mice had significantly increased phospho-tau by 148% (p<0.05). D. Brain homogenates from 18, 21, and 24 month old male Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice were Western blotted for total tau. There were no differences in total tau at any age.

Figure 6. Fyn^+/−3xTg^+/− mice are impaired in spatial learning in the Morris water maze

A. 18 month old female Fyn^+/+3xTg^+/− (n=15) and Fyn^+/−3xTg^+/− (n=17) mice were tested in the Morris water maze. Animals were trained on 5 consecutive days with 2 120-second trials per day. The average latencies to find the hidden platform of the two trials per day for each animal in each group are depicted above. Fyn^+/−3xTg^+/− mice were significantly impaired compared to controls on day 3 (p<0.05), 4 (p<0.05), and 5 (p<0.01) of learning. B. Percent time spent in the target quadrant during the probe trial was not statistically different. C. The number of target platform crossings during the probe trial was not statistically significant, but showed a trend towards decreased number of crossings in Fyn^+/−3xTg^+/− mice. D. Average swim velocity was calculated for each animal and averaged per group. Swim speeds were not significantly different between Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice. E. 18 month old male Fyn^+/+3xTg^+/− (n=10) and Fyn^+/−3xTg^+/− (n=8) mice were subjected to the Morris water maze. Fyn^+/−3xTg^+/− mice were significantly impaired compared to controls on day 3 of learning (p<0.05); however, were not significantly different on days 4 or 5. F. Percent time spent in the target quadrant during the probe trial was not statistically different. G. The number of target platform crossings during the probe trial was not statistically significant. H. Average swim velocity was not different between Fyn^+/+3xTg^+/− and Fyn^+/−3xTg^+/− mice.