Reducing Endogenous α-Synuclein Mitigates the Degeneration of Selective Neuronal Populations in an Alzheimer's Disease Transgenic Mouse Model

Abstract

Alzheimer's disease (AD) is characterized by the progressive accumulation of amyloid β (Aβ) and microtubule associate protein tau, leading to the selective degeneration of neurons in the neocortex, limbic system, and nucleus basalis, among others. Recent studies have shown that α-synuclein (α-syn) also accumulates in the brains of patients with AD and interacts with Aβ and tau, forming toxic hetero-oligomers. Although the involvement of α-syn has been investigated extensively in Lewy body disease, less is known about the role of this synaptic protein in AD. Here, we found that reducing endogenous α-syn in an APP transgenic mouse model of AD prevented the degeneration of cholinergic neurons, ameliorated corresponding deficits, and recovered the levels of Rab3a and Rab5 proteins involved in intracellular transport and sorting of nerve growth factor and brain-derived neurotrophic factor. Together, these results suggest that α-syn might participate in mechanisms of vulnerability of selected neuronal populations in AD and that reducing α-syn might be a potential approach to protecting these populations from the toxic effects of Aβ.

Significance statement: Reducing endogenous α-synuclein (α-syn) in an APP transgenic mouse model of Alzheimer's disease (AD) prevented the degeneration of cholinergic neurons, ameliorated corresponding deficits, and recovered the levels of Rab3a and Rab5 proteins involved in intracellular transport and sorting of nerve growth factor and brain-derived neurotrophic factor. These results suggest that α-syn might participate in mechanisms of vulnerability of selected neuronal populations in AD and that reducing α-syn might be a potential approach to protecting these populations from the toxic effects of amyloid β.

Keywords: Alzheimer's disease; amyloid β oligomer; cholinergic neuron; selective vulnerability; transgenic animal model; α-synuclein.

Figure 1.

Generation and characterization of Tg mice expressing APP or APP under the mThy1 promoter in the absence of endogenous α-syn. A, Schematic representation of the APP single Tg mice (line 41) and α-syn KO mice. B, Representative Western blot of total APP (doublet ∼110 kDa), Aβ (4 kDa), and total α-syn (monomer 14 kDa) in the membrane fraction in non-Tg, α-syn KO, APP Tg, and APP Tg/α-syn KO mice. C, D, Computer-aided analysis of the immunoblot for total APP and Aβ showing that the levels of endogenous APP protein were comparable between non-Tg and α-syn KO mice and between APP Tg and APP Tg/α-syn KO mice. E, Computer-aided analysis of the immunoblot for total α-syn showing that no endogenous α-syn was detected in either the α-syn KO or APP Tg/α-syn KO mouse lines compared with non-Tg mice. APP Tg mice had a significant increase in the level of endogenous α-syn protein. F, Representative photomicrographs of the hippocampus, frontal cortex, and magnified frontal cortex immunoreacted with an antibody against total APP in non-Tg, α-syn KO, APP Tg, and APP Tg /α-syn KO mice. G, Computer-aided image analysis indicated that total APP was significantly increased in APP Tg and APP Tg/α-syn KO mice compared with non-Tg mice. H, Representative photomicrographs of the hippocampus, frontal cortex, and magnified frontal cortex immunoreacted with an antibody against total α-syn in non-Tg, α-syn KO, APP Tg, and APP Tg /α-syn KO mice. Arrow indicates α-syn aggregate. I, Computer-aided image analysis indicated that total α-syn was undetectable in α-syn KO and APP Tg/α-syn KO mice compared with non-Tg mice. APP Tg mice had a significant increase in total α-syn compared with non-Tg mice. *p < 0.05 by one-way ANOVA and Dunnett's post hoc analysis. For analysis, 10 mice 4–6 months of age were used. Scale bars, 200 μm in low-power images and 40 μm in high-power images.

Figure 2.

Comparison of the patterns of Aβ and PK-resistant α-syn accumulation in the hippocampus of APP and APP Tg/α-syn KO. For these experiments, non-Tg, APP Tg, α-syn KO, and APP Tg/α-syn KO mice were used. A, Representative low-magnification photomicrographs of the hippocampus and neocortex using vibratome sections immunostained with antibody against Aβ from non-Tg, APP Tg, α-syn KO, and APP Tg/α-syn KO mice. Second row shows representative high-magnification photomicrographs of the hippocampus. B, Corrected optical densitometry analysis of the hippocampal CA3 region revealed that Aβ aggregates were comparable between APP Tg and APP Tg/α-syn KO mice. Aβ protein aggregates were not detected in the non-Tg and α-syn KO mouse lines. C, Representative low-magnification photomicrographs of the CA3 and CA1 regions of the hippocampus using vibratome sections pretreated with PK and immunostained with rabbit polyclonal antibody against full-length total α-syn (Millipore) from non-Tg, APP Tg, α-syn KO, and APP Tg/α-syn KO mice. Second row shows high-magnification photomicrographs from the hippocampus. Endogenous PK-resistant α-syn was observed in a punctate pattern in the neuropil of CA1 and CA3 in the non-Tg and was significantly increased in APP Tg mice. PK-resistant α-syn was undetectable in the α-syn KO and APP Tg/α-syn KO mice. D, Corrected optical densitometry analysis of the hippocampal CA3 region revealed that PK-resistant α-syn was significantly enhanced in the APP Tg mice compared with non-Tg mice α-syn and α-syn was undetectable in the α-syn KO and APP Tg/α-syn KO mouse lines. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*). For analysis, 10 mice 4–6 months of age were used. Scale bars, 200 μm in low-power images and 40 μm in high-power images.

Figure 3.

Immunocytochemical analysis of the cholinergic and dopaminergic systems in the APP Tg and APP Tg/α-syn KO mice. Vibratome sections were immunostained with an antibody against ChAT or TH. A, Representative low-magnification photomicrographs of the frontal cortex, CA3, and CA1 regions of the hippocampus and representative high-magnification photomicrographs of the molecular layer of the dentate gyrus and the nucleus basalis from non-Tg, APP Tg, α-syn KO, and APP Tg/α-syn KO mice immunoreacted with anti-ChAT. Endogenous ChAT was observed in a neuritic fiber pattern in the neuropil of the molecular layer of the dentate gyrus of the hippocampus in each mouse line. B, Corrected optical densitometry of ChAT immunoreactivity in the molecular layer of the dentate gyrus was significantly decreased in the APP Tg mice compared with non-Tg mice, whereas ChAT-immunoreactive levels were significantly increased in APP Tg/α-syn KO mice compared with APP Tg mice. C, Corrected optical densitometry analysis of ChAT immunoreactivity in the nucleus basalis revealed no significant differences between the mouse lines. D, Representative photomicrographs of TH immunoreactivity in the striatum at low and high magnification and the substantia nigra at high magnification. D, E, There was no significant difference between mouse lines in the corrected optical density in the striatum (E) or cell counts in the substantia nigra (F). Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*) and Tukey–Kramer post hoc analysis for comparison with α-syn Tg mice (#). For analysis, 10 mice 4–6 months of age were used. Scale bars, 200 μm in low-power images and 40 μm in high-power images.

Figure 4.

Immunocytochemical analysis of neuronal loss in the hippocampus of APP Tg and APP Tg/α-syn KO mice. A, Representative low- and high-magnification photomicrographs of the hippocampal CA3 region of vibratome-cut sections immunoreacted with an antibody against NeuN. B, Stereological assessment of every 12^th section revealed a significant decrease in the number of neurons in the CA3 in APP Tg mice compared with non-Tg mice. APP Tg/α-syn KO mice had a significant increase in the number of neurons in the CA3, which was statistically indistinguishable from non-Tg mice. C, Representative low- and high-magnification photomicrographs of the hippocampal CA3 region using vibratome-cut sections immunoreacted with an antibody against GFAP. D, Optical density assessment of GFAP immunoreactivity in the CA3 of the hippocampus indicated a statistically significant increase for APP Tg and APP Tg/α-syn KO mice compared with non-Tg mice. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*p < 0.05) and Tukey–Kramer for comparison with APP Tg mice (#p < 0.05). For analysis, 10 mice 4–6 months of age were used from each mouse type. Scale bars, 200 μm in low-power images and 40 μm in high-power images.

Figure 5.

Immunocytochemical analysis of NPY-, calbindin-, and parvalbumin-positive neurons in APP Tg and APP Tg/α-syn KO mice. A, NPY immunocytochemical analysis was performed in the molecular layer of the dentate gyrus and CA3 region of the hippocampus. Representative low- and high-magnification photomicrographs of NPY immunoreactivity in the hippocampus are shown. B, In the molecular layer of the dentate gyrus, NPY-immunoreactive fibers were significantly increased in APP Tg mice compared with non-Tg mice and significantly decreased in APP Tg/α-syn KO mice compared with APP Tg mice. C, Number of NPY-immunoreactive cells in the CA3 region of the hippocampus was unaffected by genotype. D, Representative photomicrographs of calbindin- and parvalbumin immunoreactivity in the molecular layer of the hippocampus. E, Optical density analysis of calbindin immunoreactivity revealed significant decreases in both APP Tg and APP Tg/α-syn KO mice compared with non-Tg mice. F, Number of parvalbumin-immunopositive cells was unaffected by mouse line. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*p < 0.05) and Tukey–Kramer test for comparison with APP Tg mice (#p < 0.05). For analysis, 10 mice 4–6 months of age were used from each mouse type. Scale bars, 200 μm in low-power images and 40 μm in high-power images.

Figure 6.

Morris water maze and open-field behavioral analysis in APP Tg and APP Tg/α-syn KO mice. A, Morris water maze test was performed in two phases. The first phase was the training portion conducted on days 1–3 and the second phase was with the platform hidden on days 4–7. During the hidden platform test, the APP Tg mice performed significantly worse in spatial learning portion of the test for comparison with non-Tg mice; however, the APP Tg/α-syn KO mice were similar to non-Tg mice. B, Probe test (day 8) was performed without the platform and the amount of time spent in the quadrant that used to contain the platform was analyzed. APP Tg and APP Tg/α-syn KO mice spent significantly less time in the target quadrant compared with non-Tg and α-syn KO mice. C–E, Time spent in nontarget quadrants. F, Open field behavioral test indicated a statistically significant increase in total activity (hyperactivity) for APP Tg mice compared with non-Tg mice, but APP Tg/α-syn KO mice were statistically decreased compared with APP Tg mice. G, Rearing was unaffected by mouse genotype. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*p < 0.05) and Tukey–Kramer test for comparison with APP Tg mice (#p < 0.05). For analysis, 10 mice 4–6 months of age were used from each mouse type.

Figure 7.

Linear regression analysis between markers of neurodegeneration and behavior. A, Correlation between ChAT immunoreactivity in the hippocampus and total activity. B, Correlation between ChAT immunoreactivity in the hippocampus and the probe test. C, Correlation between total rearing and ChAT immunoreactivity in the hippocampus. D, Correlation between neurons in the CA3 and total activity. E, Correlation between CA3 hippocampal neurons and the probe test. F, No correlation was found between total rearing and CA3 hippocampal neurons. For analysis, 10 mice 4–6 months of age were used from each mouse type.

Figure 8.

Western blot and immunocytochemical analysis of Rab expression pattern in APP Tg and APP Tg/α-syn KO mice. A, Western blots were performed using with the membrane fractions obtained from the hippocampus and cortex lysate and probed with antibodies against Rab 3 (25 kDa), 5 (24 kDa), 7 (23 kDa), and 11 (double band ∼ 22 kDa). B, Computer-aided analysis of Rab3a indicated a significant increase in α-syn KO and APP Tg/α-syn KO mice compared with non-Tg mice. In contrast, APP Tg mice displayed a significant reduction compared with non-Tg mice. Protein levels in APP Tg/α-syn KO mice were significantly increased compared with α-syn KO mice and statistically equivalent to non-Tg mice. C, Rab5 protein levels were significantly increased in APP Tg mice compared with non-Tg mice. α-syn ablation resulted in comparable levels of Rab5 in APP Tg/α-syn KO and α-syn KO mice. D, Rab7 protein levels were unchanged across mouse type. E, Rab11 protein levels were significantly increased in APP Tg and APP Tg/α-syn KO mice compared with non-Tg mice. F, Confocal image analysis of Rab3a and Rab5 and photomicrograph analysis of Rab7 and Rab11 expression in the hippocampus. G, Computer-aided image analysis indicated APP Tg mice had significantly decreased Rab3a immunoreactivity compared with non-Tg mice. APP Tg/α-syn KO mice Rab3a immunoreactivity was significantly increased compared with APP Tg mice and statistically equivalent to non-Tg mice. H, Computer-aided image analysis indicated that APP Tg mice had significantly increased Rab5 immunoreactivity compared with non-Tg mice. APP Tg/α-syn KO mice Rab5 immunoreactivity was significantly decreased compared with APP Tg mice and statistically equivalent to non-Tg mice. I, Computer-aided image analysis of Rab7 indicated no difference between mouse types. J, Computer-aided image analysis of Rab 11 immunoreactivity indicated a statistically significant increase in APP Tg mice compared with non-Tg mice. APP Tg/α-syn KO mice were statistically equivalent to non-Tg mice. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*p < 0.05) and Tukey–Kramer test for comparison with APP Tg mice (#p < 0.05). For analysis, 10 mice 4–6 months of age were used from each mouse type. Scale bar, 10 μm.

Figure 9.

Western blot analysis of pro-NGF, pro-BDNF, NGF, and BDNF. Brain lysates from the cortex and hippocampus were fractionated and the membrane fraction was used for immunoblot analysis. A, Representative Western blot of pro-NGF (∼ 32 kDa, double band) and pro-BDNF (∼ 30 kDa, single band). B, C, Computer-aided analysis of immunoblot of Pro-NGF (B) and Pro-BDNF signals normalized to actin signal (C) showing accumulation of the precursor growth factors in the APP Tg mice but normalization in the APP Tg/α-syn-KO mice. D, Representative Western blot of NGF (∼ 14 kDa, double band) and BDNF (∼ 20 kDa, single band). E, F, Densitometry analysis of immunoblot of NGF (E) and BDNF signals normalized to actin (F) showing decreased growth factors in the APP Tg mice but normalization of the signal in the APP Tg/ α-syn-KO mice. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with non-Tg mice (*p < 0.05) and Tukey–Kramer test for comparison with APP Tg mice (#p < 0.05). For analysis, 10 mice 4–6 months of age were used from each mouse type.

Figure 10.

Effects of LV-sh-α-syn and LV-Rab3a over expression on α-syn, ChAT, Rabs 3a and 5 immunoreactivity in the in vitro neuronal cholinergic cell line, N2A. Differentiated N2A neuronal cells were exposed to Aβ oligomer (5 nm) or vehicle for 24 h in the presence or absence of α-syn shRNA (knock-down α-syn) or LV-overexpressing Rab3a. A, Cells were immunoreacted with antibodies against α-syn (red), ChAT (DAB), Rab3a (green), or Rab5 (red). B, Coverslips were analyzed to determine levels of α-syn immunoreactivity expressed as pixel intensity. Computer-aided image analysis of α-syn-immunopositive pixel intensity indicated a low baseline expression of α-syn, which was significantly increased in N2A cells exposed to Aβ oligomers. Compared with N2A cells treated with Aβ oligomers, LV-sh-α-syn and LV-Rab3 both significantly reduced α-syn immunoreactivity. C, Coverslips were analyzed to by optical density to determine levels of ChAT. Computer-aided analysis of optical density revealed that ChAT immunoreactivity was significantly decreased in N2A cells exposed to Aβ oligomers. Both LV-sh-α-syn and LV-Rab3 showed significantly increased ChAT immunoreactivity in the presence of Aβ oligomers compared with Aβ-treated N2A cells. D, Coverslips were analyzed to determine levels of Rab3a immunoreactivity expressed as pixel intensity. Computer-aided analysis of Rab3a pixel intensity indicated that N2A cells exposed to Aβ oligomers was significantly reduced compared with vehicle-treated N2A cells. Both LV-sh-α-syn and LV-Rab3 significantly increased the Rab3a pixel intensity in N2A cells exposed to Aβ compared with Aβ-treated N2A cells alone. E, Computer-aided analysis of Rab5 pixel intensity showed that N2A cells exposed to Aβ oligomer had a significant increase in pixel intensity compared with vehicle-treated N2A cells. Both LV-sh-α-syn and LV-Rab3 reduced Rab 5 immunoreactivity in Aβ-treated N2A cells to vehicle-treated levels. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with vehicle-treated N2A control cells (*p < 0.05) and Tukey–Kramer test for comparison with Aβ oligomer-treated N2A cells (#p < 0.05). For analysis, n = 3.

Figure 11.

Double labeling and confocal analysis of the effects of LV-sh-α-syn and LV-Rab3a overexpression in the localization of NGF in N2A cells treated with Aβ. Differentiated N2A neuronal cells were exposed to either vehicle treatment or exposed to Aβ oligomer (5 nm) for 24 h and then infected with LV-sh-α-syn or LV-Rab3a grown on coverslips were double labeled with antibodies against tubulin and NGF. A, Laser scanning confocal microscopy of N2A cells immunoreacted with antibodies against tubulin (green) or NGF (red), as well as nuclei (DAPI, blue), to show NGF-positive processes. Inset box in merged panel is magnified in both the merged and NGF detailed panels. B, Computer-aided image analysis of the percentage of neuritis displaying NGF colocalization with tubulin. This study showed that N2A cells treated with Aβ oligomers had a significant decrease in NGF immunoreactivity in the neuritis. This effect was reversed by either LV-sh-α-syn or LV-Rab3a treatment. Statistical analysis was conducted using one-way ANOVA post hoc Dunnett's test for comparison with vehicle-treated N2A control cells (*p < 0.05) and Tukey–Kramer test for comparison with Aβ oligomer-treated N2A cells (#p < 0.05). For analysis, n = 3.