The neuritic plaque facilitates pathological conversion of tau in an Alzheimer's disease mouse model

Abstract

A central question in Alzheimer's Disease (AD) is whether the neuritic plaque is necessary and sufficient for the development of tau pathology. Hyperphosphorylation of tau is found within dystrophic neurites surrounding β-amyloid deposits in AD mouse models but the pathological conversion of tau is absent. Likewise, expression of a human tau repeat domain in mice is insufficient to drive the pathological conversion of tau. Here we developed an Aβ-amyloidosis mouse model that expresses the human tau repeat domain and show that in these mice, the neuritic plaque facilitates the pathological conversion of wild-type tau. We show that this tau fragment seeds the neuritic plaque-dependent pathological conversion of wild-type tau that spreads from the cortex and hippocampus to the brain stem. These results establish that in addition to the neuritic plaque, a second determinant is required to drive the conversion of wild-type tau.

Figure 1. Neuritic plaques stimulate the phosphorylation of tau.

(a) Brain sections of APP^swe;PS1ΔE9 mice (n=7) were detected by antibodies specific to Aβ (6E10), ubiquitin, Microtubule-associated protein 2 (Map2), neurofilament (Smi31, Smi312, and NF-H), and antibodies specific to phosphorylated tau: CP13 and PHF-1, Tau pT231, Tau pS262, Tau pS396, and Tau-pS422. Note the accumulation of phosphorylated tau surrounding the neuritic plaques. Scale bar, 50 μm. (b) Confocal microscopic analysis of Aβ and tau in cortex of APP^swe;PS1ΔE9 mice (n=6). Brain sections co-stained with antiserums specific to: Aβ (6E10) and tau (pS262) (upper panel); or Aβ (6E10) and tau (pS422) (lower panel). Scale bar, 50 μm. (c) Confocal microscopic analysis of Ubiquitin and tau in cortex of APP;PS1ΔE9 mice (n=6). Brain sections co-stained with antiserums specific to: Ubiquitin and tau (CP13) (upper panel); or Ubiquitin and PHF-1 (lower panel). Scale bar, 50 μm. (d) Accumulation of phosphorylated tau in dystrophic neurites surrounding the central Aβ core in 6-month-old APP^swe;PS1ΔE9 mice (n=9) as detected by antibodies specific to Aβ (6E10) and phosphorylated tau (pS422). Scale bar, 100 μm. (e) Accumulation of phosphorylated tau in dystrophic neurites surrounding the central Aβ core in 12-month-old APP^swe;PS1ΔE9 mice (n=11) as detected by antibodies specific to Aβ (6E10) and phosphorylated tau (pS422). Scale bar, 100 μm. (f) No Gallyas positive tau tangle was detected around the Aβ core of neuritic plaques (Congo red, arrows) in APP^swe;PS1ΔE9 mice (n=9), even up to 24 months. Scale bar, 25 μm.

Figure 2. Neuritic plaque is required for the pathological conversion of tau.

(a) Diagram depicting the expression construct of four-repeat domain of tau (TauRD). The top diagram represents the longest isoform of the human tau40 (441 residues). The bottom diagram shows the construct containing four-repeat domain of tau with (TauRD). (b) Protein blot using 77G7 antibody that recognized the repeat domain of tau showed the presence of exogenous (∼16 kDa) TauRD and endogenous tau protein from brain lysates of tau transgenic (Tau4R) mice (n=5). The expression level of tau in Tau4R mice is similar to that of non-transgenic mice. (c) Immunohistochemical analysis of brains of 20-month-old nTG (n=5), TTA (n=4), Tau4R (n=6) and Tau4R-AP mice (n=5) using antiserum specific to NeuN to detect neurons; sagittal sections of brains (upper panels; scale bar, 1,000 μm) and hippocampi (lower panels; scale bar, 200 μm). Note forebrain atrophy and reduction of neurons in the cortical and hippocampal area of Tau4R-AP mice. (d) Immunohistochemical analysis using antiserum specific to Aβ (6E10), microglia (IBA1) and reactive astrocytes (GFAP). Scale bar, 50 μm. Immunohistochemial analysis of brains of nTG (n=5), TTA (n=4), Tau4R (n=6) and Tau4R-AP mice (n=5) using antiserum specific to Aβ (6E10), microglial (IBA1) and reactive astrocytes (GFAP). Note increase of microglial and hypertrophic GFAP positive astrocytes in cortex of 20-month-old Tau4R-AP mice. Scale bar, 50 μm. (e) Neuronal cell count of dentate gyrus region from 18 months old TTA (n=5), AP (n=5), Tau4R (n=5), Tau4R-AP (n=5) mice using ImageJ analysis. (one-way analysis of variance (ANOVA), **P=0.0002) (f) Neuronal cell count of CA1 regions from 20 months old TTA (n=4), AP (n=6), Tau4R (n=6), Tau4R-AP (n=5) mice using ImageJ analysis. (one-way ANOVA, *P=0.0134).

Figure 3. Age-dependent spreading of endogenous tau aggregates in brain regions of Tau4R-AP mice.

(a) Gallyas-Braak silver staining of tau tangles in brain sections of Tau4R-AP mice at 12 (n=3), 18 (n=4), and 25 months of age (n=6). The sections were counterstained with fast red. The right panels are Tau4R mice at 25 months of age (n=6). The brain regions are cerebral cortex (CT); hippocampus (HP); olfactory bulb (OB); striatum (STR); thalamus (TH); hypothalamus (HY); midbrain (MB); Pons; Medulla (MY), and cerebellum (CB). Tau tangles could be detected in Tau4R-AP, but not Tau4R, mice at 18 months of ages, and the accumulation of tangles were dramatically increased with aging and spread to other brain regions (25 months of age, third column), except in cerebellum (bottom panel). Scale bar, 50 μm. (b) Immunohistochemical analysis showed age-dependent spreading of endogenous tau tangle in Tau4R-AP mice (n=13) using antibodies specific to endogenous phosphorylated S422 of tau (pS422). The brain regions are cerebral cortex (CT); hippocampus (HP); olfactory bulb (OB); striatum (STR); thalamus (TH); hypothalamus (HY); midbrain (MB); Pons; Medulla (MY), and cerebellum (CB). While no signal was detected in Tau4R mice (n=6) even at 25 months of age, tau tangles first appeared in cortical and hippocampal region (18 months of age, second column) of Tau4R-AP mice and spread to other brain region with aging (25 months of age, third column), except in cerebellum (bottom panel). Scale bar, 50 μm.

Figure 4. Gender difference in onset of neuritic plaques in APP^swe;PS1ΔE9 mice.

(a) Brain sections of 4-month-old female (n=5) APP^swe;PS1ΔE9 mice were detected by antibodies specific to Aβ (6E10). Only a few can be detected in the female APP^swe;PS1ΔE9 mice at 4 months of age. (b) Brain sections of 6-month-old female (n=6) and male (n=4) APP^swe;PS1ΔE9 mice were detected by antibodies specific to Aβ (6E10). Wide spread neuritic plaques are observed in female APP^swe;PS1ΔE9 mice, while only a few can be detected in the male APP^swe;PS1ΔE9 mice at 6 months of age. (c) Brain sections of 12-month-old female (n=9) and male (n=7) APP^swe;PS1ΔE9 mice were detected by antibodies specific to Aβ (6E10). Wide spread neuritic plaques are observed in both male and female APP^swe;PS1ΔE9 mice at this age.

Figure 5. Pathological conversion of tau induced by a mutant tau repeat domain.

(a) Protein blot using 77G7 antibody to detect exogenous (∼16 kDa) TauRDΔK and endogenous tau protein from brain lysates of mutant tau transgenic (Tau4RΔK) mice (n=9). (b) Representative brains of nTG, TTA and Tau4RΔK mice at 20 months of age. Note marked forebrain atrophy in Tau4RΔK mice (n=7). (c) The plot of brain weight of nTG (n=29), TTA (n=27) and Tau4RΔK (n=24) mice at different of ages. The brain weight of Tau4RΔK mice is progressively reduced with aging. (one-way analysis of variance (ANOVA), 9 M **P=0.02; 12 M ***P=0.0001; 15 M ***P=0.0003; 20 M ***P=0.0006). (d) Immunohistochemical analysis using antiserum specific to NeuN: sagittal sections (upper panels; scale bar, 1,000 μm) and hippocampi (lower panels; scale bar, 200 μm) of 18-month-old Tau4RΔK (n=8) and control mice (n=15). (e) Sizes of cortical (CT), hippocampal (HP) and cerebellum (CB) regions in brains of Tau4RΔK (n=24) at various ages. Note reduction of hippocampus and cortical region, but not cerebellum in Tau4RΔK mice. (f) Neuronal cell count of CA1 (T-Test, ***P=4.45E⁻⁰⁹) and CA2&3 (T-Test, ***P=0.00033) region from 12 months old TTA (n=7), Tau4RΔK (n=7) mice using ImageJ analysis. (g) Gallyas-Braak silver staining of brain sections of Tau4RΔK mice at various ages. The sections were counterstained with fast red. The left panel is overview image of the sagittal section of brains (Scale bar, 1,000 μm). The middle and right panels are hippocampal (HP) and cortical regions (CT). (Scale bar, 25 μm). Tau tangle could first be detected at 6 months of age, and the accumulation of tangles were dramatically increased while aging. (h) Total protein was extracted from different brain regions: olfactory bulb (OB), cortex (CT), hippocampus (HP), striatum (STR), midbrain (MB), brain stem (BS), cerebellum (CE) of 9 months old TTA (n=3) and Tau4RΔK (n=3) mice. Human tau fragment (∼16 kDa) detected using anti-human tau polyclonal antiserum KJ9A was only seen in frontal region, but not in MB, BS or CE. (i) Immunohistochemial analysis of brains of 20 months old Tau4RΔK mice (n=5) by antibodies specific to phosphorylated endogenous tau: PHF-1 (left panel), CP13 (middle panel), and tau-pS422 (right panel), respectively. Scale bar, 50 μm.

Figure 6. Neuritic plaque-dependent acceleration of pathological conversion of tau induced by a mutant tau repeat domain.

(a) Protein blot analysis of tau and APP in brain lysates of 3 month-old female TTA (n=6), tetO-tauRDΔK (n=6), AP (n=3), TTA-AP (n=7), Tau4RΔK (n=5), Tau4RΔK-AP (n=4) and nTG (n=7) mice using antisera K9JA and CT15. (b) Representative brains of 9 month-old female TTA, AP, TTA-AP, Tau4RΔK, Tau4RΔK-AP and nTG mice. Note marked forebrain atrophy in Tau4RΔK mice. (c) Immunostaining of brain sections of 6-month-old female TTA-AP (n=7) and Tau4RΔK-AP mice (n=5) with antibodies 6E10 specific to human Aβ. Scale bars, 200 μm. (d) Immunostaining of 6 month-old female brain sections of Tau4RΔK (n=4), and Tau4RΔK-AP mice (n=4) using antiserum tau-pS422. Scale bar, 50 μm. (e) Aβ plaques (arrows) and abundance of tau aggregates (arrowheads) were observed in Tau4RΔK-AP mice (n=5) at 9 months of age by Thioflavin-T staining. Scale bar, 50 μm. (f) Silver staining to detect Aβ plaques (arrows) and tau tangles (arrowheads) in brain of Tau4RΔK-AP mice (n=5). Scale bar, 50 μm. (g) Immunostaining of brain sections of 9 month-old female Tau4RΔK (n=5), and Tau4RΔK-AP mice (n=4) using tau-pS422. Scale bar, 100 μm. (h) Gallyas-Braak silver staining of brain sections of female Tau4RΔK-AP mice at 6 (top panel, n=4), 9 (middle panel, n=6) and 12 (bottom panel, n=3) months of age (counterstained with fast red). The left panel: sagittal section (Scale bar, 1,000 μm); middle and right panel are hippocampal (HP) and cortical regions (CT), respectively (Scale bar, 25 μm).

Figure 7. Acceleration of neuronal loss, astrocytosis and forebrain atrophy in Tau4RΔK-AP mice.

(a) Hematoxylin and eosin (H&E) staining of sagittal sections of brains (upper panels; scale bar, 1,000 μm) and hippocampi (lower panels; scale bar, 200 μm) of 9-month-old female TTA, APP^swe;PS1ΔE9(AP), TTA;APP^swe;PS1ΔE9 (TTA-AP), Tau4RΔK, Tau4RΔK;APP^swe;PS1ΔE9 (Tau4RΔK-AP) and non-transgenic (nTG) mice. Note forebrain atrophy and reduction of neurons in the cortical and hippocampal area of Tau4RΔK-AP mice. (b) Cresyl violet (CV) staining of sagittal sections of brains (upper panels; scale bar, 1,000 μm) and hippocampi (lower panels; scale bar, 200 μm) of 9-month-old female TTA, APP^swe;PS1ΔE9(AP), TTA;APP^swe;PS1ΔE9 (TTA-AP), Tau4RΔK, Tau4RΔK;APP^swe;PS1ΔE9 (Tau4RΔK-AP) and non-transgenic (nTG) mice. Note reduction of neurons in Tau4RΔK-AP mice. (c–e) Sizes of hippocampal (c), cortical (d) and cerebellum (e) regions in brains of 9 months old TTA (n=6), AP (n=7), TTA-AP (n=6), Tau4RΔK (n=5), Tau4RΔK-AP (n=4) and nTG (n=15). Note reduction of hippocampus (one-way analysis of variance (ANOVA), **P=0.00052) and cortical region (one-way ANOVA, *P=0.014), but not cerebellum in Tau4RΔK-AP mice (f) Immunohistochemical analysis of brains of 9 months old APP^swe;PS1ΔE9, TTA-AP, Tau4RΔK and Tau4RΔK-AP mice using antiserum specific to NeuN to detect neurons; sagittal sections of brains (upper panels; scale bar, 1,000 μm) and hippocampi (lower panels; scale bar, 200 μm). Note forebrain atrophy and reduction of neurons in the cortical and hippocampal area of Tau4RΔK-AP mice. (g) Neuronal cell count of CA1 region from 9 months old TTA (n=5), TTA-AP (n=6), Tau4RΔK (n=5), Tau4RΔK-AP (n=5) mice using ImageJ analysis. Note ∼80% reduction of neurons in Tau4RΔK-AP mice. (***, One-way ANOVA, P=2.04E-06) (h) Neuronal cell count of CA2 and CA3 regions from 9 months old mice using ImageJ analysis. Significant reduction (∼20%) of neurons are observed of Tau4RΔK-AP mice as compared with those of control mice. (one-way ANOVA, *P=0.0133) (i) Immunohistochemial analysis of brains of APP^swe;PS1ΔE99, TTA-AP, Tau4RΔK and Tau4RΔK-AP mice at 9 months of age using antiserum against GFAP: sagittal section (top panel, scale bar, 1,000 μm); hippocampi (second panel, scale bar, 200 μm); cortex (third panel, scale bar, 200 μm); higher power views of cortex (fourth panel, scale bar, 50 μm). Note hypertrophic GFAP positive astrocytes in cortex of Tau4RΔK-AP mice.

Figure 8. Acceleration of tau pathology in female Tau4RΔK-AP mice.

(a) Gallyas-Braak silver staining of brain sections of female Tau4RΔK (n=5) and Tau4RΔK-AP mice (n=4) at 9 months of age (counterstained with fast red). Cerebral cortex (CT); hippocampus (HP); olfactory bulb (OB); striatum (STR); thalamus (TH); hypothalamus (HY); midbrain (MB); Pons; Medulla (MY), and cerebellum (CB) were shown, respectively. While only sparse tau tangle was detected in frontal area of the Tau4RΔK mice, wide spread tau tangle and thread were observed in Tau4RΔK-AP mice, including brain stem. (Scale bar, 25 μm). (b) Gallyas-Braak silver staining of brain sections of female Tau4RΔK (n=4) and Tau4RΔK-AP mice (n=3) at 12 months of age (counterstained with fast red). Cerebral cortex (CT); hippocampus (HP); olfactory bulb (OB); striatum (STR); thalamus (TH); hypothalamus (HY); midbrain (MB); Pons; Medulla (MY), and cerebellum (CB) were shown, respectively. Note more tau tangles were detected in different brain regions of Tau4RΔK-AP mice as compare with that of Tau4RΔK mice. (Scale bar, 25 μm). (c) Semiquantificative score of tau tangle frequency in low power (200X) microscope field. Score was indicated as: 0, no tangle; 1, 1–5 tangles per field; 2, 6–15 tangles per field; and 3, >15 tangles per field. (Scale bar, 100 μm) (d) Score of tau tangle frequency in different brain regions of 9 months old female Tau4RΔK (n=5) and Tau4RΔK-AP (n=4) mice. Cerebral cortex (CT; T-Test, ***P=0.0006); hippocampus (HP; T-Test, ***P=0.0007); olfactory bulb (OB); striatum (STR; T-Test, ***P=0.0003); thalamus (TH; T-Test, ***P=0.0007); hypothalamus (HY; T-Test, ***P=0.0001); midbrain (MB; T-Test, ***P=0.0001); Pons(T-Test, *P=0.01); Medulla (MY; T-Test, ***P=0.0007), and cerebellum (CB) were shown. While tau tangles were only detected in frontal region in Tau4RΔK mice, widespread tau tangles were observed in whole brain of Tau4RΔK-AP mice, including brain stem. (e) Score of tau tangle frequency in different brain regions of 12 months old female Tau4RΔK (n=4) and Tau4RΔK-AP (n=3) mice. CT; HP; OB (T-Test, **P=0.002); STR; TH; HY; MB, (T-Test, *P=0.045); Pons; MY, and CB region were shown. Note more frequent tau tangles detected in different regions of Tau4RΔK-AP mice compared with that of Tau4RΔK mice, especially in olfactory bulb and brain stem.

Figure 9. Acceleration of onset of tau pathology is dependent on the Neuritic plaque.

(a) Brain weight of 12-month-old female nTG (n=11), AP (n=9), TTA (n=7), TTA-AP (n=7), Tau4RΔK (n=5) and Tau4RΔK-AP (n=4) mice. The brain weight of Tau4RΔK-AP mice is 25% smaller than that of the Tau4RΔK mice at 12 month of age. (one-way analysis of variance, **P=0.002). (b) Brain weight of 12 month-old male nTG (n=11), AP (n=6), TTA (n=6), TTA-AP (n=7), Tau4RΔK (n=5) and Tau4RΔK-AP (n=5) mice. No significant difference in brain weight is detected between the male Tau4RΔK-AP and Tau4RΔK mice at 12 month of age. (c) CV staining of brain sections of 11-month-old male TTA-AP, Tau4RΔK and Tau4RΔK-AP mice brains (upper panels; scale bar, 1,000 μm; and hippocampi, lower panels; scale bar, 200 μm). (d) Neuronal cell count of CA1, and CA2&3 region from 11 months old male Tau4RΔK (n=6) and Tau4RΔK-AP (n=4) mice. No significant difference was observed between the two mouse lines. (e) Immunostaining of brain sections of 11-month-old male TTA-AP (n=5), Tau4RΔK (n=5) and Tau4RΔK-AP mice (n=4) with antibodies (6E10) specific to human Aβ. Scale bars, 200 μm. (f) Immunostaining of brain sections of 11 month-old male TTA-AP (n=5), Tau4RΔK (n=5), and Tau4RΔK-AP (n=4) mice using tau-pS422. Scale bar, 100 μm. (g) Immunostaining of brain sections of 11-month-old TTA-AP (n=5), Tau4RΔK (n=5) and Tau4RΔK-AP (n=4) mice with antibodies specific to GFAP. Scale bars, 200 μm.

Figure 10. A multifactorial Model for LOAD.

Diagram depicting that the neuritic plaque is required, but insufficient for the pathological conversion of tau. A second ‘hit' involving a variety of risk alleles/factors, would be necessary to facilitate the neuritic plaque-dependent pathological conversion of the wild-type tau in LOAD.