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.
Alzheimer’s disease (AD), the most common cause of dementia, is pathologically characterized by the accumulation of neuritic plaques and neurofibrillary tangles (NFT), with widespread gliosis, loss of synapses and degeneration of neurons1,2. The core of neuritic plaques consists of extracellular deposits of Aβ3 surrounded by astrocytes and microglia. The NFT is comprised of intracellular paired helical filaments derived from hyperphosphorylated microtubule binding protein, tau4,5. Genetic studies of early-onset familial AD (fAD), which accounts for 5–10% of cases of AD6,7, have fuelled the notion that abnormal accumulation of Aβ in the brain would trigger the aggregation of tau leading to neurodegeneration and dementia2,8. While previous studies showed that accumulation of Aβ could facilitate tau aggregation in FTDP-17 linked mutant Tau mice9,10,11,12,13, no such evidence have been demonstrated using a non-mutant tau model. Furthermore, it is not known whether the neuritic plaque is necessary and sufficient to drive the conversion of wild-type tau. In contrast to mutations in APP and Presenilins linked to fAD14,15, genetic risk alleles, including ApoE16,17,18,19 and Trem2 (refs 20, 21), are associated with late-onset AD (LOAD), which constitutes ∼90% of AD cases. The notion that accumulation of Aβ is not sufficient to drive cognitive decline in LOAD is strongly supported by the following: a high Aβ burden is found in brains of some elderly in their eighties or nineties who remain cognitively normal22,23,24; and there is no correlation of Aβ plaques with cognition in LOAD25. In contrast, the aggregation of tau correlates with cognitive decline and is thought to drive neurodegeneration in AD26,27. This raises the intriguing notion that in addition to the Aβ burden, other risk factors help provoke the pathological conversion of tau to drive neuron loss and cognitive decline in LOAD. Resolution of this question will impact our view regarding disease mechanisms and identification of targets, and validate and test novel therapeutic strategies for LOAD.
Animal model systems have been instrumental in clarifying the molecular mechanisms and testing therapeutic strategies for AD28,29,30,31,32,33. Notwithstanding advances made using current rodent models10,34,35,36,37,38,39,40, a major limitation in the field is the lack of a rodent model that faithfully reproduce the tau pathology seen in humans with AD. Because no mutation in the tau gene has been identified in fAD or LOAD patients and expression of human wild-type tau failed to elicit tau pathology41, current mouse models of tau pathology are primarily based on transgenes expressing tau mutants linked to FTDP-17 (refs 9, 10, 11, 35, 36, 37, 42). However, this raises major concerns regarding their relevance to the disease context. It is well recognized that the tau pathology occurring in AD invariably depends on the conversion of the wild-type tau, a prion strain that is distinct from other human tauopathies such as that associated with FTDP-17 (ref. 43). Moreover, the tau pathology occurring in these FTDP-17-linked tau models is sufficient to drive cell death independent of Aβ plaques44, supporting the view that these mice may be good models of FTDP-17, but not of AD. To address this critical unmet need, we developed a mouse model in which wild-type tau is converted into pathological tau aggregates and NFT that are propagated through neuronal circuits to drive neuron loss in a neuritic plaque-dependent manner. Our results offer novel mechanistic insight for the pathogenesis of AD.
Neuritic plaque is not sufficient for the conversion of tau
Although it is well recognized that endogenous tau can be hyperphosphorylated in mouse models of β-amyloidosis40, whether the neuritic plaque is necessary and sufficient to facilitate the pathological conversion of wild-type tau is not known. To address this issue, we examined first the biochemical alterations of endogenous tau occurring in APPswe;PS1ΔE9 mice (a well-established mouse model of β-amyloidosis) during aging. Correlated with the presence of neuritic plaques was hyperphosphorylation of tau occurring within dystrophic neurites (Fig. 1a). The hyperphosphorylated tau was seen surrounding the central Aβ core but never co-localized with it (Fig. 1b); while closely associated, there is no direct physical interaction between the central Aβ core and the hyperphosphorylated tau. Moreover, the hyperphosphorylated tau was localized adjacent to ubiquitinated proteins within dystrophic neurites (Fig. 1c). Because tau was not hyperphosphorylated before the formation of neuritic plaques and its biochemical alteration persisted during aging (Fig. 1d,e), we infer that the neuritic plaque serves as the crucible that facilitates the hyperphosphorylation of tau. However, this persistent neuritic plaque-dependent biochemical alteration of tau failed to convert to tangle-like aggregates even in aged APPswe;PS1ΔE9 mice that were beyond 24 months as judged by Gallyas preparation (Fig. 1f), a silver stain method for detecting conformational changes in tau. While the neuritic plaque is capable of stimulating the phosphorylation of tau, it is not sufficient to alter the conformation of tau to form tangles and drive the loss of neurons.
Tau repeat domain is insufficient for the conversion of tau
The insufficiency of the neuritic plaque to drive the pathological conversion of tau suggested that a second-risk determinant may be required. That accumulating evidence indicated that other types of tau modification, such as fragmentation45 or acetylation46,47, facilitate its aggregation encouraged us to determine whether a tau repeat domain could facilitate the pathological conversion of wild-type tau. We used the moPrP-tet vector48 to express a human four-repeat domain of tau (Q244-E372, TauRD) in the nervous system of mice under the control of the tetracycline transactivator (tTA) (Fig. 2a). From multiple lines, we selected for subsequent studies one Tau4R line, which accumulated the human tau fragment to ∼1.1-fold of endogenous tau (Fig. 2b, Supplementary Fig. 1). No difference can be detected in the overall structure and weight of brains of Tau4R mice even at 25 months of age (Fig. 2c, third column). Histological analysis confirmed no evidence of activation of astrocytes/microglia or neuronal loss in Tau4R mice (Fig. 2d, third column). We also failed to observe hyperphosphorylation of tau or formation of tau tangle (Fig. 3a,b, forth column), indicating that the tau repeat domain alone is insufficient to drive the pathological conversion of wild-type tau.
Tau fragment seeds neuritic plaque-dependent tau conversion
Although the human tau repeat domain was not sufficient to alter the conformation of endogenous tau, we asked whether such fragmentation of tau could serve as the second-risk determinant in the presence of neuritic plaques to drive tau pathology. To address this question, we crossbred Tau4R mice with APPswe;PS1ΔE9 mice49 to generate compound Tau4R;APPswe;PS1ΔE9 (Tau4R-AP) mice. Tau4R-AP mice exhibited marked forebrain atrophy (Fig. 2c) with significant neuronal loss in dentate gyrus or CA1 region of the hippocampus, respectively, at 18 (Fig. 2e) or 20 (Fig. 2f) month of age; the loss of neurons in the cortex and hippocampus is associated with the accumulation of phosphorylated tau and NFT (Fig. 3a,b). Taken together with our finding that the neuritic plaque is required for the phosphorylation of tau, these results indicate that the fragmentation of tau could seed the neuritic plaque-dependent pathological conversion of wild-type tau.
To access the evolution of tau tangles in Tau4R-AP mice, we performed a time series focusing on the pattern of distribution of mouse tau aggregates. Whereas tau aggregates were undetectable before 1 year of age, they were first detected in hippocampus and cortical regions by 15 months of age and significantly increased by 18 months of age (Fig. 3a). By 2 years, tau tangles were readily observed throughout the whole brain except the cerebellum (Fig. 3a), suggesting that the tau tangle initiated from the frontal area is able to spread to the rest of the brain through neuronal circuits. Because the pattern and intensity of Gallyas stain paralleled (Fig. 3a) those of the immunocytochemical staining of misfolded mouse tau (Fig. 3b), these tau tangles would be principally derived from the endogenous tau, especially in the brain stem and midbrain where the human tau fragment was not expressed. These findings indicate that while neither the neuritic plaque nor the human tau repeat domain is sufficient to drive the pathological conversion of wild-type tau, they both are required for the seeding and aggregation of the tau repeat domain, which facilitate the pathological conversion of mouse wild-type tau that spread from the cortex and hippocampus to the brain stem.
Neuritic plaque is necessary for the conversion of tau
To test directly whether the neuritic plaque is required for the pathological conversion of wild-type tau, we took advantage of the gender difference in onset of neuritic plaques occurring in APPswe;PS1ΔE9 mice; onset for females or males was, respectively, at 4 or 6 months of age (Fig. 4). We would anticipate that if formation of the neuritic plaque precedes tau pathology, the pathological conversion of wild-type tau should be accelerated; otherwise, no such effect would be expected. For such a test, a mouse model in which the onset of tau pathology occurred after the onset of neuritic plaque in female APPswe;PS1ΔE9 mice, but similar to that of males, would be necessary. We therefore generated mice expressing the four-repeat domain of tau (Q244-E372) harbouring a pro-aggregation ΔK280 mutation (Tau4RΔK), previously shown to induce tau pathology in mice50. From multiple lines, we selected one line of Tau4RΔK, in which the exogenous tau fragments accumulated to 0.7-fold of endogenous tau (Fig. 5a, Supplementary Fig. 1). Brains of Tau4RΔK mice showed a progressive and marked atrophy of their forebrains (Fig. 5b) accompanied over time by a decline in the weight of brains (Fig. 5c). Histopathological analysis confirmed a marked decrease in size of the hippocampus (Fig. 5d,e) and cortex (Fig. 5e) and a marked reduction in the number of neurons by 12 months of age (Fig. 5f). Tau tangles first appeared at 6 months of age (Fig. 5g, upper panel), which were dramatically increased by 12 months (Fig. 5g) and except for the cerebellum, spread throughout the brain (Fig. 5g, lower panel) by 18 months. Moreover, aggregations derived from the endogenous wild-type tau exhibited a similar pattern (Fig. 5i), including the midbrain and brain stem. However, the presence of the human tau fragment was restricted to frontal area of the brain (Fig. 5h, Supplementary Fig. 2), suggesting that the exogenous tau fragment serves as a seed to convert the wild-type tau to tangles, which spread in a prion mechanism to other regions of the brain.
Because the onset of tau pathology in Tau4RΔK mice occurs after the onset of neuritic plaque formation in the female APPswe;PS1ΔE9 mice (∼4 months of age, Fig. 4a), but earlier or at the same time as that of males (∼6 months of age, Fig. 4b), we took advantage of this sex difference to determine whether neuritic plaques could accelerate the conversion of wild-type tau to tangles. We crossbred Tau4RΔK mice with APPswe;PS1ΔE9 mice49 to generate compound Tau4RΔK;APPswe;PS1ΔE9 (Tau4RΔK-AP) mice. The expression level of exogenous tau in Tau4RΔK-AP mice was comparable to that of the Tau4RΔK mice (Fig. 6a, Supplementary Fig. 3), indicating that the Aβ burden did not influence the expression of tau. Compared with female APPswe;PS1ΔE9 mice, the amount of Aβ deposit in 6-month-old female Tau4RΔK-AP mice remained unchanged (Fig. 6c). As expected, we observed marked increase in accumulation of phosphorylated tau in female Tau4RΔK-AP mice by 6 months of age (Fig. 6d). By 9 months of age, female Tau4RΔK-AP mice exhibited marked reduction in brain size (Fig. 6b) accompanied by widespread Aβ and tau pathologies (Fig. 6e,f). Histological analysis confirmed accelerated atrophy of forebrains in 9-month-old female Tau4RΔK-AP mice(Fig. 7a,b); specifically, sizes of hippocampus and cortex were significantly reduced (Fig. 7c,d). As expected, the cerebellum (Fig. 7e) of these Tau4RΔK-AP mice remained normal as tau pathology was absent in this region. To determine the impact of neuritic plaques, we examined neuron loss in brains of 9 month-old female Tau4RΔK-AP mice (Fig. 7f). As expected, marked reduction in number of neurons was observed in the cortical and hippocampal area of 9-month-old female Tau4RΔK-AP, but not female Tau4RΔK mice (Fig. 7g,h). An 80% and 25% reduction, respectively, in number of neurons were observed in CA1 (Fig. 7g) as well as CA2 and CA3 regions (Fig. 7h). Moreover, the development of phosphorylated tau aggregates (Fig. 6g) and tau tangles (Fig. 6h) was greatly accelerated in female Tau4RΔK-AP mice. Using Gallyas silver staining, we observed as early as 9 months of age tau tangles that already spread to the brain stem of Tau4RΔK-AP, but not Tau4RΔK, mice (Fig. 8a,c,d); tau tangles in Tau4RΔK mice were seen only at a later time point and by 12 months of age, a greater tau burden was found throughout the brain of Tau4RΔK-AP mice (Fig. 8b,c,e). Hypertrophic GFAP positive astrocytes were observed in cortical regions of Tau4RΔK-AP, but not Tau4RΔK, mice (Fig. 7i), indicating that reactive astrogliosis is correlated with the loss of neurons. These results indicate that neuritic plaques accelerate the onset of tau aggregation.
However, tau pathology was not accelerated in male Tau4RΔK-AP mice as compared with that of Tau4RΔK mice. At 12 months of age, while marked acceleration of brain atrophy was observed in female Tau4RΔK-AP mice (Fig. 9a), no such accelerated atrophy of forebrains occurred in male Tau4RΔK-AP mice (Fig. 9b). Histological analysis showed similar level of brain atrophy (Fig. 9c) and neuronal loss in male Tau4RΔK-AP as compared with that of male Tau4RΔK mice (Fig. 9d). No significant changes in accumulation of Aβ plaques (Fig. 9e) or activated astrocytes (Fig. 9g) were observed in male Tau4RΔK-AP mice. The accumulation of tau aggregates was not significantly altered in male Tau4RΔK-AP mice as compared with that of male Tau4RΔK mice (Fig. 9f), indicating that neuritic plaques have little effect on the rate of tau aggregation after its onset. Taken together, these results establish that the neuritic plaque is required for the pathological conversion of wild-type tau.
Clarification of the key factors contributing to the pathological conversion of wild-type tau, a prion strain or conformation that is perhaps unique to AD43, could be instrumental for developing effective therapy for this devastating disease of the elderly. Our findings here establish that the neuritic plaque is required but not sufficient to drive the pathological conversion of wild-type tau. While previous effort primarily employed transgenes expressing tau mutants linked to FTDP-17 (refs 9, 10, 11, 35, 36, 37, 42) that are sufficient to aggregate independent of the neuritic plaque, we developed a mouse model expressing human tau repeat domain that is capable of inducing the pathological conversion of mouse wild-type tau only in the presence of the neuritic plaque. Our discovery provides a fundamental paradigm shift and a framework towards identification of risk alleles/factors that play essential roles in the pathological conversion of tau in AD.
On the basis of our discovery, we propose here a model (Fig. 10) whereby the β-amyloid dependent formation of neuritic plaque during aging provides the molecular environment to facilitate the biochemical modification of tau within dystrophic neurites. While the neuritic plaque is necessary, it is not sufficient to convert the wild-type tau to a pathological conformation that drives neurodegeneration with ensuing cognitive decline and atrophy of the brain. In addition to the neuritic plaque, we envision a second-risk determinant (here we illustrate with the fragmentation of tau) would be required to drive the pathological conversion of wild-type tau. Thus, we propose that a combination of risk alleles/factors may facilitate the neuritic plaque-dependent pathological conversion of the wild-type tau in LOAD. Genetic studies has identified a series of risk alleles, in addition to ApoE4 allele, that are associated with LOAD51,52,53,54. That subsets of individuals whose brains are littered with neuritic plaques remain cognitive normal22,23,24 can be explained by our ‘two-hit’ hypothesis because they fail to harbour another risk allele that is required to drive the conversion of tau. Moreover, our model is also consistent with the long pre-symptomatic phase seen in LOAD55,56 during which time neuritic plaques accumulate because until the second-risk determinant is activated, tau remains unconverted. Likewise, our model predicts that cognitive decline in LOAD would be correlated with tau pathology rather than with neuritic plaques because once initiated, the tau aggregates are sufficient to propagate and spread throughout the brain via neuronal circuits and drive neuronal loss with ensuing cognitive decline. Thus, we believe that the pathogenesis of LOAD is a multifactorial problem initiated by the age-dependent formation of the neuritic plaque and activation of a variety of potential risk determinants to drive the pathological conversion of tau (Fig. 10).
From a therapeutic perspective, our model first predicts that therapies designed to prevent the conversion of tau would be most beneficial because the presence of neuritic plaques, in the absence of tau aggregates will not be sufficient to drive neuron loss. Our novel mouse model exhibiting age- and neuritic plaque-dependent tau pathology will now provide an invaluable platform for testing a variety of risk alleles/factors that may be relevant for the pathological conversion of tau. While a three-dimensional human neural cell culture system displaying Aβ and tau pathologies57 is a very useful tool for in vitro drug screening, our model exhibiting age-dependent development of neuritic plaques and tau pathologies that drives the progressive loss of neuron will be an essential in vivo tool for drug discovery programs designed to attenuate neurodegeneration for LOAD. Second, our model predicts that anti-amyloid therapies directed at early pre-symptomatic AD long before the pathological conversion of tau would slow disease by delaying the onset of tau pathology. In retrospect, perhaps it was not too surprising that outcomes of clinical trials with anti-amyloid agents in humans have been disappointing58,59,60,61 when delivered during a time when the conversion of tau has already initiated. Finally, our model predicts that a combinatorial therapy designed to prevent both neuritic plaque formation as well as pathological conversion of tau may provide optimal benefit for LOAD.
Generation of transgenic mice
To generate tau transgenic mice, DNA fragments encoding wild-type and ΔK280 mutant four-repeat domains of tau (TauRD, that is,Q244-E372) were subcloned into a previously characterized moPrP-tetP vector48, which expresses the gene consistently in neurons under the control of tTA proteins. The DNA was microinjected into C57BL/6 X SJL F2 mouse embryos (University of Michigan transgenic facility) to produce transgenic mice carrying wild-type (TetO-TauRD) or mutant (TetO-TauRDΔK) Tau fragment with regulatory element, moPrP-tetP promoter. Subsequently, these mice were crossbred with CamKII-tTA mice to bring the Tau transgene under the control of tet-off CamKII promoter62 (Tau4R and Tau4RΔK mice).
Doxycycline hydrochloride, dissolved in water to final concentration of 20 mg ml−1 and sterilized by filtration with 0.22 μm filter, was supplied in the drinking water to suppress the exogenous Tau expression during early development. During gestation and before weaning, the mice were raised in the presence of 50 μg ml−1 doxycycline hydrochloride.
We crossbred CamKII-tTA;TetO-TauRD or CamKII-tTA;TetO-TauRDΔK mice with APPswe;PS1ΔE9 mice49 to generate APPswe;PS1ΔE9;CamKII-tTA;TetO-TauRD (Tau4R-AP ) and APPswe;PS1ΔE9;CamKII-tTA;TetO-TauRDΔK (Tau4RΔK-AP) mice that develop both Tau pathology and Aβ amyloidosis; Tau4R or Tau4RΔK, CamKII-tTA (TTA), APPswe;PS1ΔE9;CamKII-tTA (TTA-AP), APPswe;PS1ΔE9 (AP) and non-transgenic littermates were used as controls. The brains of mice at various ages were collected, weighed and processed for biochemical, histological and immunohistochemical analysis.
All mice were maintained under 12L:12D cycle in centralized animal housing facility programme managed by Research Animal Resources (RAR) at the Johns Hopkins University. All animal procedures were in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Johns Hopkins University Animal Care and Use Committee.
Immunoblots and antibodies
Immediately after killing, mouse brains were dissected63 in the specific brain regions olfactory bulb, hippocampus, striatum, cortex, midbrain, brain stem, cerebellum and the rest. Proteins of each region were extracted with RIPA buffer (10 mM Tris-Cl (pH 8.0), 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS and 140 mM NaCl) containing 1 × complete protease inhibitor cocktail (Roche, Indianapolis, IN). The protein concentrations in the supernatants were determined by the BCA method (Pierce Chemical Co., Rockford, IL) and equal amounts of protein lysates (∼20 μg per lane) resolved on 4–12% Bis-Tris SDS–PAGE gels with MES running buffer, then transferred to polyvinylidene difluoride (Invitrogen, Carlsbad, CA) membranes, and probed with following antibodies: anti-human tau polyclonal antiserum KJ9A (1:5,000; A0024, Dako Cooperation, Carpinteria, CA); Purified anti-Tau 316–355 monoclonal antibody 77G7 (1:1,000; BioLegend, San Diego, CA); rabbit anti-APP C-terminal (1:2,000; AB5352, Chemicon); monoclonal antibody against synaptophysin (1:1,000; AB8049, Abcam, Cambridge, MA); rabbit anti-GAPDH antiserum (1:5,000; G9545, Sigma); and monoclonal anti-β-tubulin III antiserum (1:10,000; T2200, Sigma). Immunoblots were developed using enhanced chemiluminescence method (Millipore Corp., MA).
Histology and immunohistochemical analysis
For the histological and immunohistochemical analysis mice were anaesthetised and killed by decapitation at different ages64; brains were removed and weighed. Hemibrains were fixed by submerging into 4% paraformaldehyde (PFA) in PBS, embedded into the paraffin, sectioned in sagittal plane and processed. For histological analyses, 10 μm brain sections were stained with hematoxylin and eosin or Cresyl violet. Hirano silver stain utilized Hirano’s modification of the Bielschowsky method65,66. Briefly, 200 ml of ammonium hydroxide were evaporated under vacuum for 30 min. Sections were deparaffinized for 10 min and placed in 20% sliver solution for 30 min. After a rinse in dH2O, slides were transferred to ammoniacal solution (200 ml 20% silver nitrate solution with drops of evaporated ammonium hydroxide) for 15 min. Slides were washed in dH2O with four drops of ammonium hydroxide, transferred to ammoniacal solutions with three drops of developer, rinsed in dH2O and fixed in 1% sodium thiosulfate for 1 min. Finally the slides were washed with dH2O, dehydrated, cleared and mounted.
Gallyas silver staining was modified from Braak67 and performed as follows. Slides were deparaffinized and dipped in 5% periodic acid. After a brief wash with water, the slides were incubated for 1 min in silver iodide solution (290 ml dH2O, 12 g sodium hydroxide, 30 g potassium iodide, 10 ml 1%silver nitrate). After a wash in 0.5% acetic acid (2 × 5 min), slides were rinsed with dH2O and developed for 10–20 min in physical developer solution. After a 5 min wash in 0.5% acetic acid and dH2O, slides were incubated in 0.1% gold chloride for 5 min. Next, the slides were rinsed in dH2O and fixed in 1% sodium thiosulfate for 5 min. For counterstaining, the slides were dipped in 0.1% nuclear fast red for 2 min before getting washed in tap water, dehydrated and mounted.
The Congo Red stain was performed as followed. Slides were deparaffinized, stained in congo red solution (50 ml 80% alcohol, 0.15 g Congo red, 0.15 g sodium chloride, 0.5 ml 1% sodium hydroxide) for 10 min and rinsed in dH2O. The stain was differentiated by dipping the slides 5–10 times in alkaline alcohol solution (100 ml 50% alcohol, 1 ml 1% sodium hydroxide), rinsed in water, dehydrated and mounted.
For the Thioflavin-T staining slides were deparaffinized and hydrated. Slides were incubated in 0.25% potassium permanganate solution for 5 min, transferred to 1% potassium-diulfate and oxcalic acid for 5 min and incubated in 0.02% Thioflavin-T for 8 min. Slides were dehydrated and mounted with aqueous mounting medium for fluorescence.
The area size of hippocampus, cortex and cerebellum in sagittal sections at 2 mm from the midline of the brains were measured by the ImageJ program. Semiquntitative score of Tau tangles was evaluated in low power ( × 200) microscope field as described previously68. 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.
For immunohistochemical analysis, antigen retrieval was performed with 10 mM citrate buffer (pH 6.0) for an efficient epitope exposure to the antibodies; endogenous peroxidase was quenched by treating the paraffin sections with 0.3% H2O2; and nonspecific binding of antibodies was eliminated using blocking buffer (10% normal goat serum in PBS with 0.3% Triton-X) for 1 h at room temperature. The primary antibody prepared in blocking buffer was applied for overnight at 4 oC, followed by a secondary antibody for 30 min incubation at room temperature. For the secondary antibody and avidin-biotinylated peroxidase system, we used the Vectastain Universal Elite ABC kit (Vector Laboratories). Brain sections were stained with: antiserum against Aβ peptides 6E10 (1:1,000; SIG-39300, Covance); antiserum against pan-tau antibody K9JA (1:1,000; A0024, Dako Cooperation, Carpinteria, CA); mouse monoclonal antiserum against phosphorylated tau CP13 (1:1,000; Tau pS202) and PHF-1 (1:1,000, Tau pS396/S404) (gifts of P. Davis, Albert Einstein College of Medicine); rabbit antiserum against phosphorylated T231(1:1,000; 44746G, Invitrogen, Carlsbad, CA), phosphorylated S262 (1:1,000; 44705G, Invitrogen, Carlsbad, CA) phosphorylated S396 (1:1,000; 44752, Invitrogen, Carlsbad, CA) and S422 of tau (1:2,000; 44764G, Invitrogen, Carlsbad, CA); ubiquitin, Microtubule-associated protein 2 (Map2, 1:1,000; AB5622, Millipore), phosphorylated neurofilament (Smi31, 1:1,000; BioLegend, San Diego, CA), neurofilaments in axons (Smi312, 1:1,000; BioLegend, San Diego, CA), Neurofilament triplet H protein (NF-H; 1:1,000, Sigma, Saint Louis, MO, USA), polyclonal antiserum against GFAP (Z0334, Dako Cooperation, Carpinteria, CA); polyclonal antiserum against microglial (IBA1, CP290, Biocare Medical, CA), and monoclonal antibody against NeuN (MAB377, Millipore). Sections for immunofluorescence were examined under a Zeiss LSM 510 laser scanning fluorescence confocal microscope. Z-stack projections were made from serial scanning every 1 μm to reconstruct the Aβ plaques.
All data were analysed statistically by unpaired Student’s two-tailed t-test or one-way analysis of variance with Tukey’s correction for multiple comparisons using GraphPad Prism version 6.0 f for Mac, GraphPad Software, La Jolla, CA, USA. In all tests, values of P<0.05 were considered to indicate significance.
The authors declare that all data supporting the findings of this study are available within the article and its Supplementary Information files or from the authors on request.
How to cite this article: Li, T. et al. The neuritic plaque facilitates pathological conversion of tau in an Alzheimer’s disease mouse model. Nat. Commun. 7:12082 doi: 10.1038/ncomms12082 (2016).
This work was supported in part by the Ellison Medical foundation, a grant from the Brain Science Institute at Johns Hopkins, JHU Neuropathology Pelda fund and the Johns Hopkins Alzheimer’s Disease Research Center (NIH P50AG05146). We thank V. Nehus, F. Davenport, J. Yu, R. Patel, Y. Lin, A. Wei and B.T. Kale for technical support; A. Savonenko, J. Troncoso and D.L. Price for discussions and critical reading of the manuscript.
Supplementary Figures 1 – 3
About this article
European Journal of Epidemiology (2018)