| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Published online before print November 30, 2007
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Cellular Neurology,* Hertie-Institute for Clinical Brain Research, University of Tübingen, Tübingen, Germany; the Department of Neuropathology, Institute of Pathology, University of Basel, Basel, Switzerland; Novartis Institutes for Biomedical Research, Basel, Switzerland; and the Department of Neuroscience, Mayo Clinic Jacksonville, Jacksonville, Florida
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Previously we have shown that intracerebral injection of dilute Aβ-containing human and murine brain extracts induces cerebral amyloidosis in young, predepositing amyloid precursor protein (APP23) transgenic mice.6 In the present study we examined whether the same Aβ-containing brain extracts would also induce tau pathology in young B6/P301L mutated tau transgenic mice and thus confirm and expand previous studies in which synthetic Aβ was intracerebrally infused in tau transgenic mice.4 However, in contrast to this latter study in which a high concentration of fibrillary synthetic Aβ was injected and gave rise to amyloid deposits in the host at the injection site, the Aβ in the brain extract was at low concentration and did not form amyloid deposits in the injected host.6 To complement this exogenous induction of tau pathology, we also crossbred B6/P301L tau transgenic mice with APP23 transgenic mice to study the induction of tau pathology by (genetic) expression of APP/Aβ. A similar cross breeding strategy has previously been reported to induce tau pathology in selected brain regions.3 Our results suggest that both the infusion of an Aβ-containing extract and the deposition of fibrillary Aβ can induce the formation of neurofibrillary pathology.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Heterozygous B6/P301L transgenic mice and nontransgenic wild-type littermate controls were used. B6/P301L mice were obtained by backcrossing the original P301L transgenic mice (JNPL3 mice)7 with C57BL/6J mice for at least seven generations. JNPL3 mice express the shortest four-repeat (4R0N) tau with the P301L mutation under control of the mouse prion promoter. B6/P301L transgenic mice were also crossbred with APP23 transgenic mice expressing human KM670/671NL mutant APP.8
Stereotaxic Surgery
Host mice were anesthetized with a mixture of ketamine (10 mg/kg body weight) and xylazine (20 mg/kg body weight) in saline. Unilateral stereotaxic injections of 2.5 µl of brain extracts were placed with a 5-µl Hamilton syringe into the right neocortex (A/P, –2.5 mm from bregma; L, –2.0 mm; D/V, –0.8 mm) and hippocampus (A/P, –2.5 mm from bregma; L, –2.0 mm; D/V, –1.8 mm). Injection speed was 1.25 µl/minute, and the needle was kept in place for an additional 3 minutes before it was slowly withdrawn. The surgical area was cleaned with sterile saline, and the incision was sutured. Operated animals were monitored weekly in their cages for signs of behavioral changes. No seizure activity was apparent during or after the infusion. All experiments were in compliance with protocols approved by the local animal care and use committee.
Preparation of Brain Extract
Brain extracts were prepared by first dissecting out the neocortex of a 24-month-old APP23 transgenic female mouse (APP23 extract) and an aged matched female nontransgenic littermate control. Brain extract was also prepared from the brainstem and diencephalon of a 21.5-month-old B6/P301L transgenic female mouse (B6/P301L extract). Tissue samples were frozen and stored at –80°C until use. The tissue pieces were then homogenized at 10% (w/v) in sterile phosphate-buffered saline (PBS), vortexed, sonicated for 5 seconds, and centrifuged at 3000 x g for 5 minutes as previously described.6 The supernatant was recovered and immediately frozen (–80°C).
Western Blot Analysis of Brain Homogenates
To determine Aβ levels in the brain homogenates,9 samples were diluted 1 to 4 in sample buffer [0.48 mol/L Bis-Tris, 0.21 mol/L Bicine, 1.32% (w/v) sodium dodecyl sulfate, 20% (w/v) sucrose, 3.33% (v/v) 2-mercaptoethanol, 0.0053% (w/v) bromophenol blue]. Samples were then subjected to 10% Bicine-Tris 8 mol/L urea sodium dodecyl sulfate-polyacrylamide gel electrophoresis.10 Synthetic Aβ1-40 and 1-42 (Bachem, Bubendorf, Switzerland) were used as controls. Proteins were transferred onto a polyvinylidene difluoride membrane and probed with monoclonal antibody 6E10 specific to human Aβ (diluted 1:500; Covance, Emeryville, CA). The secondary antibody was horseradish peroxidase-conjugated goat anti-mouse IgG (Chemicon, Temecula, CA). Bands were visualized using SuperSignal (Pierce, Rockford, IL) and developed onto Kodak X-OMAT AR film (Eastman-Kodak, Rochester, NY).
To determine tau levels in the brain homogenates, samples were dissolved in sample buffer and run on 10% sodium dodecyl sulfate-polyacrylamide Tris-tricine gels. Proteins were electrophoretically transferred onto a polyvinylidene difluoride membrane and probed with monoclonal antibody HT7 (diluted 1:1000, recognizes amino acids 159 to 163 of tau phosphorylated and nonphosphorylated epitopes; Pierce). The secondary antibody was horseradish peroxidase-conjugated sheep anti-rabbit IgG (NA931; Amersham-Biosciences, Uppsala, Sweden). Bands were visualized using SuperSignal and developed onto Kodak X-OMAT AR film.
Histology and Immunohistochemistry
Host mice were deeply anesthetized, and they were perfused transcardially with 4% paraformaldehyde in sodium phosphate buffer (pH 7.4) and processed for paraffin embedding. Immunohistological stainings were done on 5-µm coronal paraffin sections throughout areas of interest (see quantification), according to standard procedures.11 The following antibodies were used: monoclonal AT8 antibody (1:1000; Innogenetics, Gent, Belgium) specific to tau phosphorylated at ser-202 and thr-205; polyclonal anti-Aβ antibody NT12 (1:1000; courtesy of P. Paganetti, Basel, Switzerland) or DW6 (1:1000; courtesy of D. Walsh, Dublin, UK); polyclonal antibody to GFAP (1:20,000; Sigma Aldrich, St. Louis, MO); and polyclonal antibody to Iba-1 (1:1000; Wako, Neuss, Germany). Secondary antibodies were obtained from Vector Laboratories, Burlingame, CA (Vectastain ABC kits), for peroxidase staining. In addition sections were stained histologically with silver-impregnation by the Gallyas protocol.12
Stereological Assessment of Tau Deposition
For selection of sampling sites for the estimation of tau-immunopositive area fraction in the brain regions of interest stereological sampling rules were applied.13 Every 72nd section through the hippocampus (from positions AP –1.0 mm to AP –4.0 mm from bregma), entorhinal cortex (AP –2.5 mm to –4.0 mm), sensorimotor cortex (AP –1.0 mm to AP –3.5 mm), amygdala (AP –1.0 mm to AP –3.5 mm), and brainstem (AP –4.0 mm to AP –5.0 mm) was immunostained with AT8. Therefore, eight to nine sections through the hippocampus, four to five through the entorhinal cortex, six to seven through the sensorimotor cortex and amygdala, and two to three through the brain stem were analyzed per animal. The brain regions were delineated at low magnification according to cytoarchitectural criteria.14 Within a single section, a systematic random sampling scheme was performed to capture video images for measurement of AT8-positive staining (percent area). All images were captured with an x40 objective using a Zeiss Axioskop microscope (Zeiss, Göttingen, Germany) with a motorized x-y-z stage coupled to a Sony CCD-IRIS color camera. To analyze the individual images for AT8-positive burden, 24-bit color images (20 to 50 images per brain region) were converted to 8-bit gray scale images by using conversion algorithms associated with Adobe Photoshop CS (Adobe, San Jose, CA). Calculation of the percent area occupied by tau-immunopositive area fraction was performed using the public domain NIH ImageJ. Statistical analysis included analysis of variance and posthoc testing with the help of Statistica 5.0 (StatSoft, Tulsa, OK).
Stereological Assessment of Amyloid Deposition
To assess the burden of Aβ deposition in brains, we have used the same stereological sampling as described above for the assessment of AT8-positive deposits. Briefly, every 144th section through the hippocampus (from positions AP –1.0 mm to AP –4.0 mm from bregma), neocortex (AP –1.0 mm to AP –4.0 mm), amygdala (AP –1.0 mm to AP –3.5 mm), and brain stem (AP –4.0 mm to AP –5.5 mm) was immunostained with DW6. Therefore, four to five sections through the hippocampus and neocortex, three to four through the amygdala, and one to two through the brain stem were analyzed per animal. Using a systematic random sampling scheme, images were captured with a x20 objective and converted to 8-bit gray scale for calculating the percent area occupied by Aβ-immunoreactive pixels. Statistical analysis included analysis of variance and posthoc testing with the help of Statistica 5.0.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
On a mixed C57BL/DBA2/SW background, heterozygous P301L transgenic female mice develop tau pathology between 7 to 11 months of age, whereas disease onset in males is between 11 and 15 months.3,7 To minimize immunological consequences and to minimize genetic variability in brain extract infusion studies and cross-breeding (see below), P301L transgenic mice were backcrossed to a C57BL/6J genetic background for at least seven times (thereafter termed B6/P301L mice; see Materials and Methods). Surprisingly, B6/P301L mice revealed a substantial delay in the development of tau pathology compared to the original P301L mice.
In female B6/P301L mice intracellular tau deposition (assessed with phosphorylation-dependent AT8 antibody) was not apparent in a group of 11- to 12-month-old mice (mean age, 11.5 ± 0.3 months; n = 5; indicated is the SEM) (Figure 1, A–E)
. At 
18 months of age, female B6/P301L mice became symptomatic and were unable to spread their hindlimbs when lifted by the tail. Within 1 to 3 months of disease onset, mice showed weight loss, and there were signs of dystonic posture. At that stage another group of mice (mean age, 19.8 ± 0.6 months; n = 6) was sacrificed and analyzed. Neuropathologically, similar to the original report of P301L mice on the mixed background,7
a substantial amount of hyperphosphorylated tau protein accumulated in affected neurons of the spinal cord, brain stem nuclei, and also of the hippocampus, entorhinal cortex, and amygdala (Figure 1, F–J)
. In the spinal cord, amygdala and brain stem (Figure 1, I and J)
, the majority (80 to 90%) of these tau lesions were Gallyas-positive similar to neurofibrillary tangles found in AD brain. These changes were followed by axonal degeneration in descending fiber tracts of the spinal cord and by neurogenic atrophy in skeletal muscles (not shown), all overall similar as described previously in P301L mice on a mixed genetic background.7
|
Characterization of Murine Brain Extracts
Brain homogenates were prepared from the neocortex of aged, Aβ-depositing APP23 mice and an aged-matched nontransgenic control littermate (Figure 2, A and B)
. Western blotting of the APP23 brain extract revealed Aβ levels of 1 to 10 ng/µl with more Aβ1-40 compared to Aβ1-42 (Figure 2D)
consistent with previous findings.6,9
As an additional control, an extract from the brain stem and diencephalon of an aged B6/P301L transgenic mouse was used. By immunohistochemistry and Western blotting, the presence of tau lesions and of the shortest 4-repeat human tau isoform was demonstrated (Figure 2, C and D)
.
|
Extracts were unilaterally injected into the right hippocampus and overlying somatosensory cortex of young (5- to 6-month-old) heterozygous B6/P301L female mice. All mice were analyzed 6 months later, ie, when host mice were 11 to 12 months of age. At that age, uninjected B6/P301L female mice have not yet developed significant tau pathology (Figure 1, A–E)
.
Intracerebral injection of the Aβ-rich APP23 extract induced robust deposition of tau throughout the injected hippocampus (assessed by phosphorylation-dependent AT8 antibody; Figure 3, A and B
). In contrast injection of nontransgenic brain extract or PBS did not induce any tau lesions (Figure 3, C and E)
. Modest tau pathology was also induced with the B6/P301L extract, but this was not statistically significant (Figure 3D
and Figure 4A
). APP23 extracts injected into nontransgenic mice did not induce any tau deposits (Figure 3F)
, indicating that the expression of human tau was necessary for the induction of tau lesions by the Aβ-containing extracts. All these observations were supported by quantitative analysis (Figure 4A)
.
|
|
and Figure 4, B–E
). In most regions tau induction appeared confined to the injected hemisphere. However, in the amygdala the induction, albeit to a lesser degree, also occurred contralateral to the injection site (Figure 4C)
. Immunostaining for Aβ at any time after the infusion of the Aβ-containing extracts did not reveal any Aβ deposition (data not shown), consistent with previous studies.6
Gliosis was limited to the needle track and was similar among all mice independent of whether they were injected with PBS or brain extracts. Increased Tau Pathology in Double APP23 x B6/P301L Tau Transgenic Mice
B6/P301L tau transgenic mice were crossed with APP23 transgenic mice. Before amyloid deposition, double-transgenic B6/P301L x APP23 mice did not exhibit increased tau pathology compared to the single transgenic B6/P301L littermate control mice (results not shown). However, at the time of significant amyloid deposition, a robust increase of tau pathology (assessed by phosphorylation-dependent AT8 antibody) in the entorhinal cortex, hippocampus, and to a lower extent in the amygdala but not in brain stem regions was noted in the double-transgenic B6/P301L x APP23 mice compared to the single transgenic B6/P301L control mice. This was true for the female double-transgenic mice (21.6 ± 0.9 months old; data not shown) and particularly appreciable in male double-transgenic mice (27.6 ± 0.7 months old; Figure 5, A and C
) because single male B6/P301L tau mice do not develop tau pathology up to this age (Figure 5B)
. In the entorhinal cortex and hippocampus, a subset (50%) of these induced tau lesions were Gallyas-positive (Figure 5I)
. Stereological analysis of the amyloid load did not reveal any differences between double B6/P301L x APP23 and single transgenic APP23 male mice. Notably, however, there was a great region-specific variability of amyloid deposition with neocortex and hippocampus revealing the highest amyloid load, followed by the amygdala and the brain stem, with the latter showing virtually no amyloid deposition (Figure 5D)
. Tau lesions were typically induced in the vicinity of congophilic plaques in the entorhinal cortex and hippocampus (Figure 5, E and F)
; however, in the amygdala there was a clear separation with induced tau lesions in the basolateral part and amyloid deposition in the dorso- and ventrolateral parts (Figure 5H)
. In areas not prone to develop tau pathology in B6/P301L mice, such as the piriform and somatosensory cortex, there was no or only modest tau pathology induction in the double-transgenic mice despite a high plaque load (Figure 5G)
. In all mice congophilic amyloid deposits were surrounded by a robust activation of microglia and astroglia (not shown) similar to previous results in APP23 mice.8,15
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
A similar induction of tau pathology, albeit only in the amygdala, has been described after the infusion of synthetic Aβ1-42 into the somatosensory cortex and hippocampus of a tau transgenic mouse model.4 However, several important differences between the two studies emerge. 1) Gotz and colleagues4 injected fibrillary synthetic Aβ at a high concentration (1 µg/µl), and the injected Aβ was readily detectable as amyloid and persisted at the injection sites. In contrast, a highly dilute extract, with a 100- to 1000-fold lower Aβ concentration (1 to 10 ng/µl) was used in the present study. 2) Gotz and colleagues4 used as host a mouse model (albeit with the same P301L mutation) that develops tau pathology earlier (5 to 6 months of age) than the B6/P301L mouse model used in the present study. This may explain the induction of tau lesions already 3 weeks after the injection whereas in the present study analysis was done 6 months after infusion. 3) Gotz and colleagues4 reported that the induced tau lesions were Gallyas-positive whereas the majority of the induced tau lesions in the present study were Gallyas-negative and of pretangle stage. Again, this difference may be explained by the difference in susceptibility of developing neurofibrillary lesions of the two transgenic hosts or the concentration of the injected Aβ.
We have previously shown that extracts identical to the APP23 extracts have potent Aβ-amyloid-inducing activity whereas synthetic Aβ did not reveal similar activity.6 From these results we suggested that the amyloid-inducing activity in brain represents a brain-specific conformation or modification of Aβ that is different from synthetic Aβ. Thus from this latter study and the work of Gotz and colleagues,4 a dissociation of the Aβ-amyloid-inducing activity and tau lesion-inducing activity of Aβ species could be speculated, ie, synthetic Aβ is sufficient for the induction of the tau pathology but not for the induction of Aβ-amyloid.6 However, because of the above mentioned differences between the present study and that of Gotz and colleagues,4 the conclusion that synthetic Aβ and Aβ-containing brain extract are equally potent inducers of tau pathology appears premature. Thus, to rigorously test the hypothesis of a dissociation of the amyloid-inducing activity of Aβ from the neurofibrillary-inducing activity of Aβ, future studies will be necessary in which brain extract versus synthetic Aβ will be tested in the same transgenic model.
The observation that tau lesions were induced not only at the hippocampal injection site, but also in remote areas such as entorhinal cortex and amygdala, and at the contralateral sites might be explained by the axonal projections of these brain regions to the injection site.4,16 Various Aβ species are known to interact with synapses and to affect adversely their structure and function,17 which in turn may lead to retrograde signaling, damage and induction of tau lesions in the remote cell bodies. It is also possible that Aβ is taken up and transported retrogradely to the soma where Aβ interacts with tau and induces tau phosphorylation and aggregation.18-20 Independent of the exact mechanism, the idea of synaptic damage with retrogradely induced tau lesions might at least partly explain why induction of tau lesions was found in the entorhinal cortex and amygdala but not in the brain stem, which exhibits only diffuse projection to the injection sites. The lack of induction of tau pathology in the somatosensory cortex may simply be a result of low susceptibility of this region to form tau lesions in such tau transgenic mice.4,7,21 In addition, or alternatively to such a retrograde model of tau pathology propagation, the infused Aβ may have been transported via blood and/or via interstitial fluid pathways to the contralateral site and remote areas.22,23
Tau pathology was also induced in B6/P301L transgenic mice when crossed with APP23 mice, confirming a previous study in which the same tau transgenic mouse model, albeit on a mixed genetic background, was crossed with Tg2576 transgenic mice.3 Although in the infusion experiment, exogenously applied Aβ species have triggered the tau pathology, in the double B6/P301L x APP23 transgenic mice, it may be argued that the tau lesions were initiated by an intraneuronal (indirect) interaction of tau and Aβ, because both the Thy1 promoter (in case of the APP23 mice) and PrP promoter (in case of the B6/P301L mice) are expressed to a great extent in similar neuronal populations.24,25 However, because the induction of the tau lesions appeared concomitant with the appearance of significant cerebral amyloidosis, it is more likely that the extracellular deposition of amyloid is the trigger of the neurofibrillary pathology.
The idea of a local induction of the tau pathology by the amyloid deposits is supported by the current region-specific analysis of tau pathology induction versus amyloid load. The most robust induction of tau pathology was in the entorhinal cortex, followed by hippocampus, amygdala, and brain stem. Consistently, we found the highest plaque load in the entorhinal cortex, followed by the hippocampus, the amygdala, and brain stem. However, consistent with previous findings,3 neurofibrillary tangle-bearing neurons were not necessarily in close vicinity of the amyloid deposits exemplified in the basolateral amygdala, with a robust induction of tau pathology but virtually no amyloid deposition. These observations suggest that in the double-transgenic mice tau induction in areas without amyloid aggregates may be triggered from remote areas with amyloid deposits similar to the suggested mechanism with the Aβ-containing brain extracts.
Although the infusion experiment strongly suggests Aβ species as being the trigger of the induced tau pathology in remote areas, the mechanism of local tau induction in the crossed mice is more open for speculation. Congophilic amyloid deposition is always accompanied by a strong neuroinflammatory response15,26,27 and microglia activation has recently been linked to the progression of tau pathology in transgenic mice.28 It has also been shown that congophilic plaques induce local axonal dystrophy and synaptic abnormalities,16,29 which in turn may promote tau aggregation.
In conclusion, our results suggest that Aβ can induce neurofibrillary pathology by different mechanisms. In particular the observation that infused Aβ can trigger the tau pathology in the absence of amyloid deposits provides an explanation for the discrepancy between the neuroanatomical location of amyloid deposits, or absence thereof, and the development and spreading of tau lesions in AD brain.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the Alzheimer Association (grant ZEN-06-27341 to M.J.), the German National Genome Network (grant NGFN2 to M.J.), the German Competence Network in Degenerative Dementias (grant BMBF-01GI0705 to M.J.), and the National Institutes of Health (grant R01-NS46355 to J.L.).
T.B. and F.C. contributed equally to this study.
Accepted for publication September 11, 2007.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Oddo, A. Caccamo, B. Tseng, D. Cheng, V. Vasilevko, D. H. Cribbs, and F. M. LaFerla Blocking A{beta}42 Accumulation Delays the Onset and Progression of Tau Pathology via the C Terminus of Heat Shock Protein70-Interacting Protein: A Mechanistic Link between A{beta} and Tau Pathology J. Neurosci., November 19, 2008; 28(47): 12163 - 12175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. B. Paulson, M. Ramsden, C. Forster, M. A. Sherman, E. McGowan, and K. H. Ashe Amyloid Plaque and Neurofibrillary Tangle Pathology in a Regulatable Mouse Model of Alzheimer's Disease Am. J. Pathol., September 1, 2008; 173(3): 762 - 772. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
