Triflusal reduces dense-core plaque load, associated axonal alterations and inflammatory changes, and rescues cognition in a transgenic mouse model of Alzheimer's disease

Abstract

Inflammation has been associated with the two classic lesions in the Alzheimer's (AD) brain, amyloid deposits and neurofibrillary tangles. Recent data suggest that Triflusal, a compound with potent anti-inflammatory effects in the central nervous system in vivo, might delay the conversion from amnestic mild cognitive impairment to a fully established clinical picture of dementia. In the present study, we investigated the effect of Triflusal on brain Abeta accumulation, neuroinflammation, axonal curvature and cognition in an AD transgenic mouse model (Tg2576). Triflusal treatment did not alter the total brain Abeta accumulation but significantly reduced dense-cored plaque load and associated glial cell proliferation, proinflammatory cytokine levels and abnormal axonal curvature, and rescued cognitive deficits in Tg2576 mice. Behavioral benefit was found to involve increased expression of c-fos and BDNF, two of the genes regulated by CREB, as part of the signal transduction cascade underlying the molecular basis of long-term potentiation. These results add preclinical evidence of a potentially beneficial effect of Triflusal in AD.

Fig. 1

(A–B) Untreated Tg+ mice showed significantly longer escape latencies on multiple days of the invisible platform training compared to Tg− untreated littermates (p=0.038). Escape latencies of Triflusal Tg+ mice did not significantly differ from those of Tg− mice (p=0.555) indicating a treatment related benefit. (C) In addition, Triflusal-treated Tg+ mice achieved a significantly higher number of crossings over the platform during the probe trials than untreated Tg+ mice (p=0.040). 5 Tg+ mice of the original groups (3 Triflusal and 2 untreated) had to be excluded from the analysis due to circling behavior. (D) Conditioned freezing (fear response) was measured immediately and 7 days after the training. After seven days, untreated Tg+ mice demonstrated significantly less freezing than Tg− mice (p=0.006), while no significant differences were found between Triflusal Tg+ mice and Tg− mice (p=0.344) indicating a treatment related benefit. *p<0.05 **p<0.01.

Fig. 2

(A–B) No significant effect of treatment was observed on the amount of brain amyloid deposition as labelled by Aβ immunostaining (4G8) in any of brain regions examined at 13 (n=12 per group) (p=0.232) and 18 months of age (n=6 per group) (p=0.680). (C) No significant differences were detected on brain levels of soluble and insoluble Aβ40 and Aβ42 as determined by ELISA at 18 months (N=6 per group) (p>0.05). DG = dentate gyrus, EC = entorhinal cortex, Cg = cingulate, MC = motor cortex, VC = visual cortex. *p<0.05 **p<0.01.

Fig. 3

(A) Triflusal treatment significantly decreased dense-core amyloid plaque load and number in several brain regions. (B) At 13 months, an overall 47% reduction on compact (fibrillar, Thioflavin-S reactive) plaque load was observed in the brain of Triflusal-treated Tg+ in comparison to untreated Tg+ animals (n=12 per group) (p=0.052). At 18 months, an overall 64% reduction in the amount of dense-core amyloid plaque load was detected in the brain of Triflusal-treated Tg+ mice in comparison to untreated Tg+ animals (n=6 per group) (p=0.002). (C) Representative photomicrographs of visual cortex and EC sections from untreated and Triflusal Tg+ mice stained with Thio-S. Calibration bars 400 μm. DG = dentate gyrus, EC = entorhinal cortex, Cg = cingulate, MC = motor cortex, VC = visual cortex. *p<0.05 **p<0.01.

Fig. 4

(A) Stereologically-based astrocyte counts showed a significant reduction in Triflusal Tg+ mice compared to untreated Tg+ mice at 13 months in the CA1 region of the hippocampus (n=12 per group) (p=0.004). Significant reductions in the number of astrocytes in the CA1 and the EC were also observed at 18 months in Triflusal Tg+ mice compared to untreated Tg+ mice (n=6 per group) (p=0.001 and p=0.002, respectively). (B) Stereologically-based microglial cell counts also showed a significant reduction in Trifusal Tg+ mice compared to untreated Tg+ mice in the CA1 region of the hippocampus both at 13 (n=12 per group) (p=0.001) and 18 months of age (n=6 per group) (p=0.004). (C) Representative photomicrographs of EC and CA1 sections from untreated and Triflusal Tg+ mice stained with GFAP and IBA-1. Calibration bars 400 μm (EC) and 100 μm (CA1). Calibration bars 150 μm and 40 μm for the insert. *p<0.05 **p<0.01.

Fig. 5

(A–B) Brain levels of IL-1β and TNF-α were significantly increased in untreated Tg+ compared to Tg− mice. A significant effect of treatment was observed both on levels of IL-1β (p=0.020) and TNF-α (p=0.004) as determined by ELISA when comparing Triflusal Tg+ to untreated Tg+ mice at 13 months (n=3–4 per group). *p<0.05, **p<0.01.

Fig. 6

(A) Representative image of SMI312-positive neurites immunostaining (red) and senile plaque (Thioflavin-S) (green). Curvature ratio was calculated by dividing the curvilinear length (d1) of the axon by the straight line length of the axon (d2). Calibration bar 10 μm. (B) Treatment with Triflusal significantly reduced axonal curvature in Tg2576 mice close to plaques (<50 μm) at 13 months (p=0.0006) and 18 months of age (p<0.0001). (B) Axons far (>50 μm) from plaques were significantly curvier in 18 month-old Tg+ mice compared to non-Tg littermate controls (p<0.001). Triflusal treatment significantly straightened axons further from plaques at 18 months of age (p <0.01) (N =88–135 axons analyzed per group). *p <0.05, **p <0.01, ***p<0.001.

Fig. 7

(A) Triflusal treatment increased hippocampal levels of c-fos expression in Tg+ mice compared to untreated mice as measured by RT-PCR (p=0.042). (B) The level of hippocampal BDNF expression was significantly decreased in untreated Tg+ compared to Tg− mice (p=0.022). BDNF levels of Triflusal Tg+ mice did not significantly differ from those of Tg− mice (p=0.263) indicating a treatment related benefit (N=9–10 per group at 13 months). *p<0.05.