APOE4 exacerbates α-synuclein seeding activity and contributes to neurotoxicity in Alzheimer's disease with Lewy body pathology

Abstract

Approximately half of Alzheimer's disease (AD) brains have concomitant Lewy pathology at autopsy, suggesting that α-synuclein (α-SYN) aggregation is a regulated event in the pathogenesis of AD. Genome-wide association studies revealed that the ε4 allele of the apolipoprotein E (APOE4) gene, the strongest genetic risk factor for AD, is also the most replicated genetic risk factor for Lewy body dementia (LBD), signifying an important role of APOE4 in both amyloid-β (Aβ) and α-SYN pathogenesis. How APOE4 modulates α-SYN aggregation in AD is unclear. In this study, we aimed to determine how α-SYN is associated with AD-related pathology and how APOE4 impacts α-SYN seeding and toxicity. We measured α-SYN levels and their association with other established AD-related markers in brain samples from autopsy-confirmed AD patients (N = 469), where 54% had concomitant LB pathology (AD + LB). We found significant correlations between the levels of α-SYN and those of Aβ40, Aβ42, tau and APOE, particularly in insoluble fractions of AD + LB. Using a real-time quaking-induced conversion (RT-QuIC) assay, we measured the seeding activity of soluble α-SYN and found that α-SYN seeding was exacerbated by APOE4 in the AD cohort, as well as a small cohort of autopsy-confirmed LBD brains with minimal Alzheimer type pathology. We further fractionated the soluble AD brain lysates by size exclusion chromatography (SEC) ran on fast protein liquid chromatography (FPLC) and identified the α-SYN species (~ 96 kDa) that showed the strongest seeding activity. Finally, using human induced pluripotent stem cell (iPSC)-derived neurons, we showed that amplified α-SYN aggregates from AD + LB brain of patients with APOE4 were highly toxic to neurons, whereas the same amount of α-SYN monomer was not toxic. Our findings suggest that the presence of LB pathology correlates with AD-related pathologies and that APOE4 exacerbates α-SYN seeding activity and neurotoxicity, providing mechanistic insight into how APOE4 affects α-SYN pathogenesis in AD.

Keywords: Alzheimer’s disease; Apolipoprotein E; Lewy body dementia; RT-QuIC; Seeding; α-synuclein.

Fig. 1

Levels of α-SYN in the brain samples from patients with AD and controls. The brain samples from the superior temporal lobe of patients with AD (with or without LB co-pathologies) a-c and controls (Ctrl) d-f were lysed sequentially in TBS (buffer-soluble fraction), TBSX (TBS with 1% Triton X-100, detergent-soluble fraction), and FA (70% formic acid, insoluble fraction) buffer. The levels of total α-SYN in the TBS (a and d), TBSX (b and e), and FA (c and f) fractions were measured by ELISA and compared between AD and AD + LB patients with or without APOE4 gene (E4⁻ and E4⁺) a–c, or controls with or without APOE4 gene d–f. Data represent mean ± SEM. N = 469 samples in total for AD and AD with LB co-pathologies. Comparisons of α-SYN between E4⁻ and E4⁺ subjects, and between AD and AD + LB subjects, were made using linear regression models that were adjusted for age at death, sex, CAA score, Braak stage, Thal phase, and number of APOE4 alleles (only for AD vs. AD + LB comparisons), where each α-SYN outcome was examined on the square root scale. N = 16 samples in total for control cases. Mann–Whitney U tests were used for statistical analyses. **p < 0.01; ****p < 0.0001; N.S. not significant

Fig. 2

Effects of LB pathology and APOE4 gene allele on α-SYN seeding activities. Samples of TBS brain lysate from patients with AD, AD + LB, LBD, or controls, with or without APOE4 gene allele, were subjected to α-SYN RT-QuIC assay. a and d The extent of aggregation was monitored by ThT fluorescence and the aggregation curve is shown. The protein aggregation rate (PAR, b and e) and the maximum fluorescence values (A.U., measured at plateau of aggregation, c and f) are shown. Each dot represents the average value of an individual biological sample measured in triplicate. Data are mean ± SEM. N = 21–24 samples/group for AD and AD + LB. Comparisons of RT-QuIC outcomes between E4⁻ and E4⁺ subjects, and between AD and AD + LB subjects, were made using proportional odds logistic regression models that were adjusted for age at death, sex, CAA score, Braak stage, Thal phase, and number of APOE4 alleles (only for AD vs. AD + LB comparisons), where each α-SYN outcome was examined as an ordered categorical variable. N = 8 samples/APOE genotype in controls and 9 samples/APOE genotype in LBD. Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses. *p < 0.05, **p < 0.01, ****p < 0.0001, N.S. not significant

Fig. 3

The size distribution of α-SYN in TBS brain lysates from AD and control brains. TBS-soluble human brain lysates from AD, AD + LB, and control brains were fractioned by size exclusion chromatography (SEC) using AKTA FPLC with tandem Superose 6, 10/300 GL columns. The fractions from #18 to #56 were collected and the levels of α-SYN were measured by ELISA and Western blotting. a and b The size distribution of α-SYN in the AD and AD + LB brains, without or with APOE4 gene allele, were quantified and compared. Molecular weight markers (kDa) for the peak fractions are indicated. N = 6 samples/group. Each sample was mixed with equal amount of TBS lysate from 3–4 brains with the same APOE genotype and the same pathological diagnosis (AD or AD + LB). Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses. c The distribution of the α-SYN in different fractions was validated using Western blotting with AD samples. d and e The size distribution of α-SYN in the control brains, without or with APOE4 gene allele, were quantified and compared. N = 4–5 samples/group. The student t tests were used for statistical analyses. Data are shown as mean ± SEM. *p < 0.05; N.S. not significant. Only the significantly changed fractions were labeled

Fig. 4

The size distribution of tau in TBS brain lysates from AD and control brains. TBS-soluble human brain lysates from AD, AD + LB, and control brains were fractioned by SEC as described in Fig. 3. The levels of tau were measured by ELISA (a, b, d, and e) and Western blotting (c). The size distribution of tau in the AD and AD + LB brains (a–c), control brains (d and e), without or with APOE4 gene allele, were quantified and compared. N = 6 samples/group in AD cases. Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses (b). N = 4–5 samples/group in control cases. The student t tests were used for statistical analyses (e). Data are shown as mean ± SEM. Only the significantly changed fractions were labeled

Fig. 5

The size distribution of APOE in TBS brain lysates from AD and control brains. TBS-soluble human brain lysates from AD, AD + LB, and control brains were fractioned by SEC as described in Fig. 3. The levels of APOE were measured by ELISA (a, b, d, and e) and Western blotting (c). The size distribution of APOE in the AD and AD + LB brains (a–c), control brains (d and e), without or with APOE4 gene allele, were quantified and compared. N = 6 samples/group in AD cases. Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses (b). N = 4–5 samples/group in control cases. The student t tests were used for statistical analyses (e). Data are shown as mean ± SEM. Only the significantly changed fractions were labeled

Fig. 6

The contribution of different α-SYN species on the seeding activities. TBS-soluble human brain lysates from AD and AD + LB brains were fractioned by SEC as described in Fig. 3. The fractions from #40 to #47 containing α-SYN proteins were subjected to α-SYN RT-QuIC assays. (a, c, e, g, i, k, m, and o) The extent of aggregation was monitored by ThT fluorescence and the aggregation curves from fraction #40 to #47 are shown. (b, d, f, h, j, l, n, and p) The maximum fluorescence values (A.U., measured at plateau of aggregation) from fraction #40 to #47 are shown. Each dot represents the average value of an individual biological sample measured in duplicate. N = 11–12 samples/group. Data are mean ± SEM. The Mann–Whitney U tests were used for statistical analyses. *p < 0.05; **p < 0.01; N.S. not significant. q–t The α-SYN RT-QuIC aggregation curve (q and s) and the maximum fluorescence values (r and t) from the fraction #40 and #42 are shown by further separating APOE genotypes (without or with APOE4 gene allele). N = 5–6 samples/group. Data are mean ± SEM. The Mann–Whitney U tests were used for statistical analyses. *p < 0.025; N.S. not significant

Fig. 7

Protease resistance of α-SYN aggregates derived from the brains of patients with AD. α-SYN-RT-QuIC products derived from TBS lysates of brains from patients with AD (APOE4⁻, APOE4⁺) and AD + LB (APOE4⁻, APOE4⁺) were incubated without (input) or with 2.5 μg/ml of protease K at 37 °C for 30 min. Samples were then subjected to Western blotting to detect α-SYN (a). The ratio of the undigested α-SYN fragment and the total α-SYN (input) were quantified and compared among groups (b). Molecular weight markers (kDa) are indicated on the right of each blot. N = 5 samples per group. Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses. **p < 0.01; N.S. not significant

Fig. 8

The seeding activity of α-SYN aggregates derived from human AD brain samples in α-SYN biosensor cells. The α-SYN biosensor cells expressing α-SYN (A53T) mutant protein fused to cyan and yellow fluorescent proteins (α-SYN-CFP/YFP) were treated with vehicle (vehicle control, VC), 0.35 μM α-SYN monomer, or 0.35 μM amplified α-SYN aggregates from the TBS brain lysates of patients with AD (APOE4⁻, APOE4⁺) or AD + LB (APOE4⁻, APOE4⁺). a–d After 72 h of incubation, cells were harvested and subjected to fluorescence resonance energy transfer (FRET)-based assay by fluorescence microscopy. Representative images of each experimental group are shown with a merged channel of CFP, FRET and YFP, and individual channel of FRET (a). Inclusions in FRET channel were subjected to quantification b–d, the yellow arrowhead shows the example of inclusion). The percentage of inclusion-positive cells (b), the intensity of the inclusions (c), and the size of inclusions (d) were quantified. Scale bars: 10 μm. Experiments were carried out with three independent experiments. Images from 4–5 fields per group were captured, each dot represents an individual field. Data are shown as mean ± SEM. Two-way ANOVA with Sidak’s multiple comparison tests were used for statistical analyses. *p < 0.05; ***p < 0.001; N.S. not significant

Fig. 9

The cytotoxicities of α-SYN aggregates derived from human AD brain samples in iPSC-derived neurons. The human-induced pluripotent stem cells (iPSC)-derived neurons at DIV 12 were treated with vehicle (vehicle control, VC) or 0.35 μM amplified α-SYN aggregates from the TBS brain lysates of patients with AD (APOE4⁻, APOE4⁺) or AD + LB (APOE4⁻, APOE4⁺) (a), VC or 0.35 μM α-SYN monomers (b). After 48 h of incubation, cells were harvested and the cell viability was determined by MTT assay. Experiments were carried out in duplicate with three independent experiments, each dot represents an individual replicate. For the measurement of neurite length (c and d), the cells treated with VC or 0.35 μM amplified α-SYN aggregates were fixed and stained with TUJ1 antibody (red). The nuclei were stained with DAPI (blue). The average length of TUJ1 positive neurites per nuclei were quantified (c). Representative images of each experimental group are shown (d). Scale bars: 50 μm. Experiments were carried out with three independent experiments. Images from 2 to 3 fields per group were captured, each dot represents an individual field. Data are shown as mean ± SEM. One-way ANOVA with Holm-Sidak’s multiple comparison tests were used for statistical analyses in panel a and c. The student t test was used for statistical analysis in panel b. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; N.S. not significant