Gene delivery of apoE2 reduces brain amyloid burden in apoE4-target replacement APP.PS1 transgenic mice

Apolipoprotein E4 allele (APOE4) carriers have a markedly increased risk for developing Alzheimer disease (AD) whereas the APOE2 allele is protective, reducing AD risk by 50% and markedly delaying the age of onset. ApoE is also a major determinant of amyloid β-peptide (Aβ) deposition and brain amyloid burden in AD patients and transgenic mice (E4>>E3>E2). We have previously reported that lentiviral-mediated gene delivery of apoE2 significantly reduces brain Aβ levels and amyloid burden in APP transgenic mice. We have now extended this finding using adeno-associated viral (AAV) vectors and a new APP.PS1/TRE4 mouse model where Aβ/amyloid deposition is dependent on apoE4 expression. We administered 0.25X1010, 0.5X1010, or 1X1010 viral genomes (vg) of AAVrh10-CAG-E2 bilaterally into hippocampi of young (2.5-month old) APP.PS1/TRE4 mice (mild to moderate amyloid pathology in hippocampus). Using specific ELISAs to quantify apoE and Aβ levels in brain, we observed an apoE dose-dependent reduction in Aβ 8 weeks post injection; however, when 1X1010 vg virus were delivered via the same route to 5-month old APP.PS1/TRE4 mice ( ≥ 8 times more Aβ in their hippocampi), there was no statistically significant reduction in Aβ burden even with apoE2 levels that are comparable in both groups of mice. This suggests that the “efficacy” of apoE2 gene delivery in reducing brain Aβ/amyloid burden may depend on the degree of preexisting amyloid pathology. We therefore defined an apoE2/Aβ42 ratio (apoE2 levels in E2 injected mice/Aβ42 levels in control mice) to assess the effectiveness of treatment. We calculated that the apoE2/Aβ42 ratio in effectively treated young APP.PS1/TRE4 mice was 17.8 but was only 3.7 in older (5-month old) APP.PS1/TRE4 mice, where pathology was already quite advanced. We next administered AAV9-CAG-E2 intrathalamically in young (2.5-month old) APP.PS1/TRE4 mice. ApoE2 levels were increased 7.5-fold in the thalamus, 2.6-fold in the hippocampus, 1.7-fold in the cortex. Correspondingly, Aβ42 levels were decreased 70% (p < 0.001) in the thalamus, 45% (p < 0.05) in the hippocampus, and 14% (p > 0.05) in the cortex. The apoE2/Aβ42 ratios in these same three brain regions were 32.7, 7.2, and 2.4, respectively, demonstrating a strong positive correlation between the level of apoE2 expression and the reduction in tissue Aβ42 levels. Taken together, our study confirms that AAV-mediated apoE2 gene delivery reduces brain Aβ burden in transgenic mice where the latter is dependent on apoE4 expression. Importantly, the level of apoE2 expression and the degree of already existent amyloid pathology were important determinants of the “efficacy” of apoE2 gene therapy in this model.