APOE genotype-dependent pharmacogenetic responses to rapamycin for preventing Alzheimer's disease

Abstract

The ε4 allele of Apolipoprotein (APOE4) is the strongest genetic risk factor for Alzheimer's disease (AD), the most common form of dementia. Cognitively normal APOE4 carriers have developed amyloid beta (Aβ) plaques and cerebrovascular, metabolic and structural deficits decades before showing the cognitive impairment. Interventions that can inhibit Aβ retention and restore the brain functions to normal would be critical to prevent AD for the asymptomatic APOE4 carriers. A major goal of the study was to identify the potential usefulness of rapamycin (Rapa), a pharmacological intervention for extending longevity, for preventing AD in the mice that express human APOE4 gene and overexpress Aβ (the E4FAD mice). Another goal of the study was to identify the potential pharmacogenetic differences in response to rapamycin between the E4FAD and E3FAD mice, the mice with human APOE ε3 allele. We used multi-modal MRI to measure in vivo cerebral blood flow (CBF), neurotransmitter levels, white matter integrity, water content, cerebrovascular reactivity (CVR) and somatosensory response; used behavioral assessments to determine cognitive function; used biochemistry assays to determine Aβ retention and blood-brain barrier (BBB) functions; and used metabolomics to identify brain metabolic changes. We found that in the E4FAD mice, rapamycin normalized bodyweight, restored CBF (especially in female), BBB activity for Aβ transport, neurotransmitter levels, neuronal integrity and free fatty acid level, and reduced Aβ retention, which were not observe in the E3FAD-Rapa mice. In contrast, E3FAD-Rapa mice had lower CVR responses, lower anxiety and reduced glycolysis in the brain, which were not seen in the E4FAD-Rapa mice. Further, rapamycin appeared to normalize lipid-associated metabolism in the E4FAD mice, while slowed overall glucose-associated metabolism in the E3FAD mice. Finally, rapamycin enhanced overall water content, water diffusion in white matter, and spatial memory in both E3FAD and E4FAD mice, but did not impact the somatosensory responses under hindpaw stimulation. Our findings indicated that rapamycin was able to restore brain functions and reduce AD risk for young, asymptomatic E4FAD mice, and there were pharmacogenetic differences between the E3FAD and E4FAD mice. As the multi-modal MRI methods used in the study are readily to be used in humans and rapamycin is FDA-approved, our results may pave a way for future clinical testing of the pharmacogenetic responses in humans with different APOE alleles, and potentially using rapamycin to prevent AD for asymptomatic APOE4 carriers.

Keywords: APOE3; APOE4; Alzheimer's disease prevention; Amyloid-beta plaques; Blood brain barrier; Cerebral blood flow; Cerebrometabolic function; Cerebrovascular reactivity; Cognition; MRI; Neuroinflammation; Pharmacogentics; Rapamycin; Water content; White matter integrity; mTOR.

Figure 1.. Effect of Rapamycin on the blood-brain barrier function and Aβ clearance.

(A) Representative confocal images showing decreased luminal accumulation of N-ε(4-nitro-benzofurazan-7-yl)-D-Lys(8)-cyclosporin A (NBD-CSA) fluorescence (white) in brain capillaries isolated from the old mice compared to young mice, indicating reduced P-glycoprotein (P-gp) activity. (B) Corresponding quantitative fluorescence data; images are shown in arbitrary fluorescence units (scale 0–255). (C) Western blotting (WB) for P-gp from the cortical vasculature, β-Actin was used as loading control. ***p < 0.001. Data are mean ± SEM.

Figure 2.. Effect of Rapamycin on Aβ retention.

(A) Representative images of Aβ immunohistochemical staining from the E3-control, E3-Rapa, E4-Control and E4-Rapa. The FAD negative brain tissue was used as controls to verify the Aβ loading. (B) quantitation of the Aβ load. Data are presented as the mean ± SEM. n.s. = not significant; ***p < 0.001.

Figure 3.. Cerebral blood flow (CBF) measurements.

(A) CBF images at 2 months of age; the color code indicates the level of CBF on a linear scale. (B) Post-treatment quantitative CBF (ml/g/min) obtained from hippocampus. (C) Pre- and post-treatment percent change in CBF for both E3FAD and E4FAD rapamycin fed mice. (D) Percent change in CBF stratified by sex. *p < 0.05; **p < 0.01. Data are mean ± SEM.

Figure 4.. Brain metabolites determination in vivo using ¹H-MRS.

(A) The voxel replacement on the hippocampus and (B) the representative ¹H-MRS spectrum, showing lactate (Lac), N-acetyl-aspartate (NAA), glutamate (Glu) and glutamine (Gln), creatine (Cr), glycerophosphocholine (GPC) and phosphocholine (PCh), taurine (Tau) and myo-inositol (mI) in parts per million (ppm). Percent changes between pre-and post-treatment in (C) GPC and PCh, (D) Glu and Gln, and (E) NAA in E3-Control (E3-Con), E3-Rapa, E4-Control (E4-Con) and E4-Rapa groups. Data are Mean ± SEM. **p < 0.001.

Figure 5.. White matter integrity measurements.

(A) The maps showing fractional anisotropy (FA) and mean diffusivity (MD), and diffusion-encoded-color (DEC) of the four groups of mice. (B) Comparison of quantitative FA values in corpus callosum, cortex, hippocampus and thalamus between Control and Rapa groups. (C) Comparison of quantitative MD values in corpus callosum, cortex, hippocampus and thalamus between Control and Rapa groups. Data are Mean ± SEM. *p < 0.05. ***p < 0.0001.

Figure 6.. Water content measurements.

(A) Representative images of water content (indexed by BBPC) from each group. Yellow is representative of the highest water content, whereas red to black is low water content. (B) Mean and SEM for water content measured by brain-blood partition coefficient (BBPC). *p < 0.05.

Figure 7.. Hindpaw stimulation and cerebrovascular reactivity (CVR) responses.

(A) Maps (left) and percent of BOLD responses (right) under hindpaw stimulation in E3FAD and E4FAD mice. (B) (Top) CVR maps from the E4FAD-Control (left) and E4FAD-Rapa (right) mice; (Bottom left) an overall BOLD response curve between the two groups; (Bottom right) quantitative measures of CVR in somatosensory cortex, thalamus and hippocampus between the two group. (C) (Top) CVR maps from the E3FAD-Control (left) and E3FAD-Rapa (right) mice; (Bottom left) an overall BOLD response curve between the two groups; (Bottom right) quantitative measures of CVR in somatosensory cortex, thalamus and hippocampus between the two group. Data are Mean ± SEM. *p < 0.05.

Figure 8.. Behavior assessments in memory and anxiety.

(A) The Novel Object Recognition (NOR) test found the rapamycin-treated groups had a significantly higher recognition index, or D₂, than the control groups, independent of APOE-genotype. (B) The elevated plus maze (EPM) test found the E3FAD-Rapa mice had a significantly higher open arm duration compared to the E3FAD-control group; no differences were found between the E4FAD-Control and E4FAD-Rapa group. Data are Mean ± SEM. **p < 0.01.