P021

3xTg-AD (female, 3 months old at study initiation, hybrid 129/Sv x C57BL/6 genetic background, Alzheimer's disease transgenic model harboring APP_Swe, tau P301L, and PS1 M146V knock-in mutations)^[1]

60 nM P021/g feed (equivalent to ~162 nM P021/mouse/day)

p.o.; continuously; 18 months

Prevented neurodegeneration (measured by Fluorojade C staining percent area) in CA3, dentate gyrus, parietal cortex, and frontal cortex at 9 months post-treatment relative to vehicle-treated mice.
Prevented neurodegeneration in CA1, CA3, DG, and parietal cortex at 18 months post-treatment relative to vehicle-treated mice.
Reduced Aβ plaque load (percent area occupied by 4G8 staining) in the subiculum at 9 months post-treatment relative to vehicle-treated mice, with no significant effect in CA1.
Reduced Aβ load in the subiculum at 15 months post-treatment relative to vehicle-treated mice, with no effect in CA1.
Reduced Aβ load in both CA1 and subiculum at 18 months post-treatment relative to vehicle-treated mice.
Slowed Aβ load increase over time in the subiculum via regression analysis.
Reduced tau hyperphosphorylation (PHF1 staining percent area) in CA1 and subiculum, with a trend toward reduction in AT8 staining in CA1 at 9 months post-treatment relative to vehicle-treated mice.
Reduced AT8 staining in CA1 and subiculum, and reduced PHF1 staining in CA1 and subiculum at 15 months post-treatment relative to vehicle-treated mice.
Reduced AT8 staining in CA1 and subiculum, reduced PHF1 staining in CA1 and subiculum, and reduced hippocampal PHF1-positive tau on western blot at 18 months post-treatment relative to vehicle-treated mice.
Slowed the increase in AT8 and PHF1 staining over time in CA1 and subiculum via regression analysis.
Rescued episodic memory impairment in the Novel Object Recognition task at 16-17 months post-treatment, with treated mice showing a significantly higher discrimination index relative to vehicle-treated mice.
Increased survival rate from 41% (vehicle) to 87% by week 71 of treatment, and reduced hindlimb paralysis incidence (1/32 vs 5/34 vehicle-treated mice).