Acta Neuropathologica

, Volume 123, Issue 1, pp 31–37

The presence of Aβ seeds, and not age per se, is critical to the initiation of Aβ deposition in the brain


  • Tsuyoshi Hamaguchi
    • Department of Cellular NeurologyHertie-Institute for Clinical Brain Research, University of Tübingen
    • DZNE, German Center for Neurodegenerative Diseases
  • Yvonne S. Eisele
    • Department of Cellular NeurologyHertie-Institute for Clinical Brain Research, University of Tübingen
    • DZNE, German Center for Neurodegenerative Diseases
  • Nicholas H. Varvel
    • Department of Cellular NeurologyHertie-Institute for Clinical Brain Research, University of Tübingen
    • DZNE, German Center for Neurodegenerative Diseases
  • Bruce T. Lamb
    • The Lerner Research Institute, The Cleveland Clinic Foundation
  • Lary C. Walker
    • Department of Neurology, Yerkes National Primate Research CenterEmory University
    • Department of Cellular NeurologyHertie-Institute for Clinical Brain Research, University of Tübingen
    • DZNE, German Center for Neurodegenerative Diseases
Original Paper

DOI: 10.1007/s00401-011-0912-1

Cite this article as:
Hamaguchi, T., Eisele, Y.S., Varvel, N.H. et al. Acta Neuropathol (2012) 123: 31. doi:10.1007/s00401-011-0912-1


The deposition of the β-amyloid (Aβ) peptide in senile plaques and cerebral Aβ-amyloid angiopathy can be seeded in β-amyloid precursor protein (APP)-transgenic mice by the intracerebral infusion of brain extracts containing aggregated Aβ. Previous studies of seeded β-amyloid induction have used relatively short incubation periods to dissociate seeded β-amyloid induction from endogenous β-amyloid deposition of the host, thus precluding the analysis of the impact of age and extended incubation periods on the instigation and spread of Aβ lesions in brain. In the present study using R1.40 APP-transgenic mice (which do not develop endogenous Aβ deposition up to 15 months of age) we show that: (1) seeding at 9 months of age does not induce more Aβ deposition than seeding at 3 months of age, provided that the incubation period (6 months) is the same; and (2) very long-term (12 months) incubation after a focal application of the seed results in the emergence of Aβ deposits throughout the forebrain. These findings indicate that the presence of Aβ seeds, and not the age of the host per se, is critical to the initiation of Aβ aggregation in the brain, and that Aβ deposition, actuated in one brain area, eventually spreads throughout the brain.


AlzheimerAmyloidPrionSeedingSenile plaquesTransgenic mouse


The deposition of β-amyloid (Aβ) peptide in senile plaques and cerebral Aβ angiopathy (CAA) can be induced in the brains of β-amyloid precursor protein (APP)-transgenic (tg) mice by the infusion of Aβ-rich brain extracts from patients with Alzheimer’s disease (AD) or from aged APP-tg mice [5, 11, 18, 29]. Mechanistically, the seeding of Aβ deposition resembles the transmission of prion disease in that a misfolded, multimeric form of the disease-associated protein appears to instigate aggregation by corruptive molecular templating [10, 25, 28]. Aβ deposition also has been induced in wild-type marmosets [1, 22], a nonhuman primate that produces human-type sequence Aβ [9].

Advancing age is the greatest risk factor for developing AD [21], but how age promotes the development of Aβ plaques and CAA remains uncertain [3]. Previous studies of the seeded induction of Aβ deposition in mice have employed Tg2576 and APP23 mouse models that spontaneously form amyloid plaques and CAA beginning sometime around 8–10 months of age [11, 18]. To unambiguously identify the seeded Aβ deposits in the absence of endogenous and spontaneously generated lesions, these studies typically used relatively short incubation periods of 3–6 months after intracerebral administration [5, 11, 16, 18], and 6–7 months after intraperitoneal administration [6]. After longer periods, distinguishing seeded deposits from those that normally form in these models becomes increasingly problematic, thus hampering the analysis of the impact of advancing age and extended incubation periods on the instigation and spread of Aβ lesions in the brain.

In the present study, we used a genomic, homozygous APP-tg mouse model (R1.40 mice) that begins to exhibit Aβ deposition only in late adulthood [13] to examine (1) whether seeding is more effective when initiated in older animals, compared to younger animals, and (2) the long-term effects of focally administered Aβ seeds on the evolution and spread of β-amyloid lesions in the brain.

Materials and methods


Male, hemizygous R1.40 APP-tg mice on a B6.129 background were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and bred with C57BL/6J mice. Homozygous mice were then generated by cross-breeding the hemizygous offspring. R1.40 mice carry a yeast artificial chromosome containing the genomic copy of human APP harboring the K670N-M671L mutation [15]. R1.40 APP-tg mice in the present study were homozygous and were maintained under specific pathogen-free conditions. The experimental procedures were carried out in accordance with the veterinary office regulations of Baden–Württemberg (Germany) and approved by the local Animal Care and Use Committees.

Preparation of brain extracts

Extracts were prepared from 22- to 25-month old, Abeta-depositing APP23 mice (tg extract), or from age-matched, non-transgenic (wild-type [Wt]) mice (Wt extract) as previously described [18]. The forebrains were obtained and frozen on dry ice and stored at −80°C. Samples were homogenized at 10% (w/v) in PBS, vortexed, sonicated 3 × 5 s and centrifuged at 3,000×g for 5 min. The supernatant was used as the seeding agent. Aβ levels were estimated with electrochemiluminescence-linked immunoassay, revealing Aβ concentrations of 10–20 ng/μl.

Stereotaxic injection of brain extracts

Mice were anaesthetized with a mixture of fentanyl (0.05 mg/kg body weight), midazolam (5 mg/kg body weight) and medetomidine (0.50 mg/kg body weight). Bilateral stereotaxic injections of extract were made with a Hamilton syringe into the hippocampus (2.5 μl) and overlying neocortex (1.0 μl). Extract was injected into the hippocampus (AP −2.5 mm, L ±2.0 mm, DV −1.8 mm) over a 2-min period, and the needle was kept in place for an additional 2 min. The needle was raised to the overlying neocortex (AP −2.5 mm, L ±2.0 mm, DV −1.0 mm) and the additional 1 μl of extract was injected over a 1-min period. After 2 more minutes, the needle was slowly withdrawn before suturing. After surgery an antidote (flumazenil 0.5 mg/kg body weight, atipamezole 2.5 mg/kg body weight, naloxone 1.2 mg/kg body weight) was administered. A total of 38 mice were used (see figure captions for details). Two mice were excluded from the analysis because the injections missed the dorsal hippocampus; one mouse was excluded because of hydrocephalus; two mice were excluded due to lack of seeding. The excluded mice were dispersed over the experimental groups.

Tissue processing, histological and stereological analysis

The mice were transcardially perfused with PBS under deep anesthesia (ketamine 400 mg/kg, xylazine 40 mg/kg). Brains were harvested, hemidissected, and the left hemispheres were processed as described [6]. The sections were immunostained using polyclonal antibody CN3 (1:1,000; raised against residues 1–16 of human Aβ) [6] and visualized using standard immunoperoxidase procedures with the Vectastain Elite ABC Kit (Vector Laboratories, Burlingame, CA, USA).

The areal density of Aβ-immunoreactive lesions was quantitated in random-systematic sets of every 12th section through the neocortex and hippocampus. Pia-associated vessels were not counted. Stereological analysis of CN3-positive staining was performed using a Zeiss microscope equipped with a motorized x–y–z stage coupled to a video-microscopy system and the Stereo Investigator software (MicroBrightField, Inc, Williston, VT, USA) as described [18].

Adjacent sets of sections were stained with Congo red according to standard protocols and viewed under cross-polarized light and analyzed. The following additional antibodies were used: rabbit polyclonal antibody against cow glial fibrillary acidic protein (GFAP) (Dako, Glostrup, Denmark); rabbit polyclonal antibody against ionized calcium-binding adapter molecule 1 (Iba1) (Wako; Richmond, VA, USA).

Immunoblotting and electrochemiluminescence-linked immunoassay

The right hemisphere of PBS-perfused mice was homogenized at 10% w/v in buffer (50 mM Tris [pH 8.0], 150 mM NaCl, 5 mM EDTA, and Complete Mini protease inhibitor cocktail [Roche Diagnostics GmbH, Penzberg, Germany]). To assess Aβ and APP levels, 4–12% NuPage Bis–Tris mini gels using NuPage LDS sample buffer were used (Invitrogen, Carlsbad, CA, USA), followed by protein transfer onto nitrocellulose membranes and probing with antibody 6E10 (Covance, Princeton, NJ, USA). As an internal control, monoclonal anti-GAPDH antibody (Hy Test Ltd, Turku, Finland) was used. Densitometric values of band intensities were analyzed using ImageJ 1.44 (

Statistical analysis

All values are expressed as means ± SEMs. Statistical analyses were performed using SPSS 16.0 software (SPSS Japan, Tokyo, Japan) and Statview 5.0.1 (SAS Institute Inc., Cary, NC, USA), with the significance threshold set at p ≤ 0.05, two tailed.


Human APP expression and the quantity of human Aβ were studied in the brains of R1.40 APP-tg mice at 3, 9, and 15 months of age. No significant differences among the three age groups were found (Fig. 1). Immunohistochemical analysis did not detect any Aβ deposits in the brains of the mice at 3 and 9 months, while in two of six 15-month-old mice, occasional Aβ deposits were found in the frontal cortex (Fig. 1).
Fig. 1

Steady-state levels of APP and Aβ are similar in pre-depositing R1.40 mice. a Human APP and Aβ levels in the brains of R1.40 mice at multiple ages. (3 months [n = 6; 3 male, 3 female], 9 months [n = 6; 3 male, 3 female] and 15 months [n = 6; 3 male, 3 female]). Shown are immunoblots, using antibody 6E10, of three mice/age group; GAPDH served as a loading control. b Quantification of protein levels in all mice revealed no difference among the groups (one-way ANOVA for APP and Aβ; both p > 0.05). Three samples in each age group were applied to one gel, and two gels were used in total. Densitometric values of band intensities were analyzed using ImageJ 1.44, and percentages relative to the amount in 3-month-old mice were calculated from the average value at 3 months of age. c No Aβ deposition was found in the brains of 3- and 9-month-old R1.40 mice by immunohistochemical analysis, while in two of six 15-month-old mice, infrequent Aβ deposits were present in the frontal cortex. Scale bar 500 μm

To determine whether the age of the host is important for the seeded induction of β-amyloidosis, R1.40 APP-tg mice received injections of Aβ-containing brain extracts (tg extract) or wild-type (Wt) extract into the hippocampus and overlying neocortex at 3 months of age and were analyzed at 9 months of age; a second group was seeded at 9 months of age and analyzed at 15 months of age (Fig. 2). As an additional control to study the long-term effects of seeding, a third group was injected at 3 months of age and sacrificed at 15 months of age (Fig. 2). The results revealed quantitatively similar induction of Aβ deposition in the hippocampus and overlying neocortex of the tg extract-seeded mice in groups 1 and 2, i.e., the degree of seeding after a 6-month incubation period was similar whether the mice were infused at 3 months of age or at 9 months of age (Fig. 3). In both groups, the amyloid was Congo red-negative and diffuse in nature. Wt extract-inoculated controls did not show Aβ induction (Fig. 3).
Fig. 2

Study design. Group 1 was intracerebrally seeded at 3 months of age, and analyzed at 9 months; group 2 was seeded at 9 months of age and analyzed at 15 months; group 3 was seeded at 3 months of age and analyzed at 15 months
Fig. 3

Older mice and younger mice show similar levels of induced Aβ deposition. However, long-incubation periods allow for generation of widespread Aβ aggregation and glial cell response. a Induction of Aβ deposition was evident in all three groups of R1.40 mice that were intracerebrally seeded with the tg extract, but not with the Wt extract. While the β-amyloid induction was similar in groups 1 and 2, Aβ load was the greatest in group 3. Adjacent sections were double-stained with Congo red and GFAP or Iba-1. While in groups 1 and 2, all induced β-amyloid was Congo red-negative, group 3 showed at least some Congo red-positive deposits surrounded by darkly stained and hypertrophic GFAP-positive astrocytes (inset, left) and Iba1-positive microglia (inset, right). b Quantitative stereological analysis. Group 1 (Wt extract: n = 5 [3 female, 2 male], tg extract: n = 7 [4 female, 3 male]); group 2 (Wt extract injected: n = 6 [2 female, 4 male], tg extract: n = 6 [2 female, 4 male]); group 3 (Wt extract: n = 5 [2 female, 3 male], tg extract: n = 4 [3 female, 1 male]). Three-way ANOVA for Gender × Group × Extract revealed significant main effects for Group and Extract (p < 0.001) but no significant effect of gender (p > 0.05). There was a significant Group × Extract interaction (F[2, 21] = 89.23; p < 0.001) and subsequent Scheffé post hoc analyses were used to pinpoint group differences (*p < 0.05, **p < 0.01, ***p < 0.001). Scale bar 500 and 10 μm

When R1.40 APP-tg mice were intracerebrally injected with tg extract at 3 months of age and allowed to incubate for 12 months, extensive Aβ deposition was apparent in the 15-month-old mice that exceeded the quantity of lesions after 6 months of incubation by approximately sixfold (Fig. 3). Although most of the induced amyloid again was Congo red-negative, some of the Aβ deposits in the vicinity of the injection site (dorsal hippocampus) were Congo red-positive and surrounded by darkly stained, GFAP-positive hypertrophic astrocytes and Iba1-positive enlarged microglia (Fig. 3). Strikingly, in 12-month incubation mice, Aβ deposition had spread throughout the entire neocortex and hippocampus, with additional deposition in the thalamus and septal nuclei, and was more widespread compared to that in the 6-month incubation groups (Fig. 4). Both parenchymal and vessel-associated Aβ deposits were observed (Fig. 4).
Fig. 4

Longer incubation times result in extensive spreading of Aβ deposition in the forebrain. a Twelve months after intracerebral injections of the tg extract, copious Aβ was deposited in the brain (shown are coronal sections of the mouse presented in Fig. 3; the coordinates from bregma AP are indicated). Note that Aβ deposition was widespread throughout the neocortex and hippocampus, although tg extract was injected only locally into the hippocampus and overlying cortex at AP −2.5 mm. b Aβ load in each section from anterior to posterior is shown for the 12-month incubation mice (group 3) and compared to the 6-month incubation mice (groups 1 and 2) (section 1 to section 11 approximate location from bregma AP is as follows, section 1 AP +2.2 mm, section 5 AP +0.1 mm, section 8 AP −1.8 mm, section 11 AP −3.3 mm). Note that the induced Aβ is the highest around the injection site (corresponding to sections 9 and 10) in all groups, and that the Aβ deposits have spread throughout the brain, including the most anterior parts. Scale bar 500 μm


Previous studies have shown that brain-derived Aβ aggregates can induce Aβ lesions in the brains of susceptible mice [5, 6, 11, 16, 18, 29]. In the present study, we used the genomic-based R1.40 APP-tg model as host. These animals exhibit endogenous Aβ deposition at approximately 15 months of age [13], thus allowing for the analysis of extended incubation periods in Aβ extract-inoculated mice. Our results reveal that intracerebral injections of Aβ-containing brain extract also induce Aβ deposition in this model, although the Aβ induction after a 6-month incubation period was less than that seen in APP23 mice after 4 months of incubation [16, 18]. This observation is consistent with the threefold overexpression of human APP in R1.40 mice compared to sevenfold overexpression in APP23 animals [15, 26], and supports the view that the induction of Aβ deposition is dependent on the concentration of Aβ in the inoculum as well its production by the host [8, 18].

We found that a 6-month incubation period induced similar levels of Aβ aggregation in the brains of young (3 months) and older (9 months) R1.40 APP-tg mice, indicating that the aged brain does not provide a more favorable environment for the induction of Aβ deposition once Aβ seeds are present. Our results, while surprising in light of the view that age is a prominent risk factor for amyloid lesions [3, 7, 27], may be attributed to the fact that the brain levels of Aβ are constant in pre-depositing R1.40 mice between 3 and 15 months of age [17]. Unchanging Aβ levels prior to Aβ pathology also have been reported in other AD mouse models [12, 14]. These findings suggest that age may increase the risk of AD by impairing the ability of cells to dispose pathogenic seeds [2, 20]. In addition, it cannot be excluded that a similar experiment in which seeding is compared in a young host versus a host at an age closer to the end of its mean life span would yield different results.

The most striking finding to emerge from the inoculations in R1.40 mice was the presence of Aβ deposits in much of the forebrain after the 12 month incubation period. Previous work has shown that the focal infusion of Aβ seeds into the hippocampus is sufficient to induce Aβ deposition throughout the entire hippocampus and even in connected brain areas; however, induction of lesions in widespread regions of the forebrain has not been previously observed [5, 10]. The mechanism of this extensive spreading of Aβ deposition is not clear, but implies that seeds not only migrate within brain structures and along defined neuronal pathways, but also may disseminate along perivascular fluid drainage channels, by vascular transport, or by simple diffusion through the brain parenchyma [10]. The recent finding that small, soluble Aβ species are particularly potent inducers of β-amyloidosis [16], similar to the strong infectivity of small, non-fibrillar prion particles [24], underscores potential mechanistic similarities in the induction and spread of Aβ and PrP aggregates in the brain.

Birefringence under crossed-polarizing filters is indicative of the amyloid-like nature of proteinaceous deposits after staining with the dye Congo red [30]. In seeded R1.40 APP-tg mice, the duration of the incubation period influenced the degree to which the resulting Aβ deposits were congophilic. Whereas the 6-month incubation period, starting at either 3 or 9 months of age, did not result in the induction of congophilic lesions, Congo red-positive plaques were observed after the 12-month incubation period. However, congophilic deposits were only observed near the injection site, indicating that diffuse pathology occurs first and eventually develops into Congo red-positive (i.e., amyloid) plaques. Because the cytological changes associated with Aβ deposition are most strongly associated with amyloid per se [4] rather than with diffuse deposits, the lack of congophilic amyloid induction after the 6-month incubation period did not allow us to determine whether the amyloid-associated cytopathology (neuritic dystrophy and glial activation) differs between young (9 months) and older (15 months) hosts [3, 7].

From a translational perspective, the present findings support the widely held notion that therapies directed at mitigating the formation of Aβ pathology should begin early in the pathogenic cascade that leads to AD [19, 23]. The exposure to Aβ seeds profoundly influences the onset and degree of β-amyloid pathology in APP-tg mouse models. While there is not yet direct evidence that exogenous seeds are involved in the initiation of AD, it is likely that endogenously generated seeds become increasingly prevalent with age due to the decline of proteostatic processes [3, 20]. The neutralization or removal of endogenous Aβ seeds thus remains a promising but elusive therapeutic goal for interfering with the pathogenesis of AD.


This work was supported by grants from the Competence Network on Degenerative Dementias (BMBF-01GI0705), the BMBF in the frame of ERA-Net NEURON (MIPROTRAN), NIH RR-00165, P50AG025688, the Alzheimer’s Association (NIRG-10-173099) and the CART Foundation. TH is recipient of a postdoctoral fellowship from the Alexander von Humboldt Foundation (Bonn, Germany). We gratefully acknowledge helpful discussions and experimental help of Franziska Langer, Götz Heilbronner, Ulrike Obermüller, Jörg Odenthal, Stephan Kaeser, Michael Hruscha and Andrea Bosch (Tübingen, Germany) and Harry LeVine III (Lexington, KY, USA).

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag 2011