Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M705282200 on August 23, 2007

J. Biol. Chem., Vol. 282, Issue 43, 31755-31765, October 26, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/43/31755    most recent
M705282200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eckermann, K.
Right arrow Articles by Mandelkow, E.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Eckermann, K.
Right arrow Articles by Mandelkow, E.-M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

The beta-Propensity of Tau Determines Aggregation and Synaptic Loss in Inducible Mouse Models of Tauopathy*

Katrin Eckermann{ddagger}, Maria-Magdalena Mocanu{ddagger}, Inna Khlistunova{ddagger}, Jacek Biernat{ddagger}, Astrid Nissen{ddagger}, Anne Hofmann{ddagger}, Kai Schönig§, Hermann Bujard§, Andreas Haemisch, Eckhard Mandelkow{ddagger}, Lepu Zhou||, Gabriele Rune||, and Eva-Maria Mandelkow{ddagger}1

From the {ddagger}Max-Planck Unit for Structural Molecular Biology, Notkestrasse 85, 22607 Hamburg, Germany, the §Center of Molecular Biology (ZMBH), University of Heidelberg, Im Neuenheimer Feld, 69120 Heidelberg, Germany, and the ||Department of Neuroanatomy and the Central Animal Facility, University of Hamburg Medical School, Martinistrasse 20, 20146 Hamburg, Germany

Received for publication, June 27, 2007 , and in revised form, August 22, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Neurofibrillary lesions are characteristic for a group of human diseases, named tauopathies, which are characterized by prominent intracellular accumulations of abnormal filaments formed by the microtubule-associated protein Tau. The tauopathies are accompanied by abnormal changes in Tau protein, including pathological conformation, somatodendritic mislocalization, hyperphosphorylation, and aggregation, whose interdependence is not well understood. To address these issues we have created transgenic mouse lines in which different variants of full-length Tau are expressed in a regulatable fashion, allowing one to switch the expression on and off at defined time points. The Tau variants differ by small mutations in the hexapeptide motifs that control the ability of Tau to adopt a beta-structure conformation and hence to aggregate. The "pro-aggregation" mutant {Delta}K280, derived from one of the mutations observed in frontotemporal dementias, aggregates avidly in vitro, whereas the "anti-aggregation" mutant {Delta}K280/PP cannot aggregate because of two beta-breaking prolines. In the transgenic mice, the pro-aggregation Tau induces a pathological conformation and pre-tangle aggregation, even at low expression levels, the anti-aggregation mutant does not. This illustrates that abnormal aggregation is primarily controlled by the molecular structure of Tau in vitro and in the organism. Both variants of Tau become mislocalized and hyperphosphorylated independently of aggregation, suggesting that localization and phosphorylation are mainly a consequence of increased concentration. These pathological changes are reversible when the expression of Tau is switched off. The pro-aggregation Tau causes a strong reduction in spine synapses.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Tau is a microtubule-associated protein that was initially identified by its ability to drive microtubule (MT)2 assembly in vitro (1). It binds to tubulin via its repeat domain and adjacent sequences (Fig. 1) and thereby promotes MT nucleation and stabilization in neurons, particularly in axons. Thus Tau plays a crucial role in modulating MT dynamics and supporting axonal transport (2-4). The protein is encoded by a single gene on chromosome 17, alternative splicing leads to 6 isoforms ranging from 352 to 441 amino acids. They differ by the presence or absence of two N-terminal inserts (exons 2 and 3) and by the second of four pseudo-repeats of 31 amino acids in the C-terminal half (exon 10). The efficacy of Tau in binding to MTs is influenced by the degree of phosphorylation and the exclusion/inclusion of exon 10. The majority of phosphorylation sites are serine or threonine residues followed by prolines (SP/TP motifs) that lie in the domains flanking the repeats and can be phosphorylated by several proline-directed kinases. There are other non-SP/TP sites whose phosphorylation causes the detachment of Tau from microtubules. This includes Ser-214 and the KXGS motifs within the repeats (Ser-262, Ser-293, Ser-324, and Ser-356) (5). Most of these sites are "hyperphosphorylated" in Alzheimer disease (6). Tau belongs to the class of natively unfolded proteins, it contains very little secondary structure, has a hydrophilic and basic character, and is highly soluble (7, 8). Despite this, Tau aggregates into abnormal "paired helical filaments" in Alzheimer disease that can coalesce into "neurofibrillary tangles." The aggregation is based on short hexapeptide motifs in repeats R2 and R3 that are prone to form beta-structure (8). Tau is normally confined to axons, but in diseased neurons it becomes missorted to cell bodies and dendrites. Thus, the three most visible changes in Alzheimer Tau are its hyperphosphorylation, aggregation, and mislocalization. The progressive spreading of neurofibrillary tangles correlates well with cognitive deficits and neuronal loss in AD (9).

Whereas AD is characterized by the coexistence of two types of protein aggregates (amyloid fibers composed of Abeta, and PHFs composed of Tau), there are several neurodegenerative disorders dominated by Tau aggregates and hence termed "Tauopathies" (10, 11). The main link between AD and other Tauopathies is the presence of aggregated Tau protein. Several transgenic mouse models have been developed in the past to unravel the underlying mechanisms of Tau pathology in AD and other tauopathies (12-23). The principle behind most models was the overexpression of Tau, assuming that this would accelerate aggregation. An alternative was the overexpression or activation of kinases, assuming that this would lead to hyperphosphorylation and hence aggregation of Tau (18, 24, 25). The expression of human Tau isoforms met with limited success with regard to Tau aggregation, presumably because of the high solubility of Tau and the short lifetime of mice. However, the approaches could be improved by: (i) Tau mutants with a higher aggregation potential, e.g. P301L (20, 26, 27) or P301S (21, 22, 28); (ii) combining Tau expression with enhanced Abeta activity, by using APP or PS mutations (19, 29); or (iii) expressing several Tau isoforms analogous to human brain (16). With these tools, part of the Tauopathies and AD-like pathology can be approximated in mice, such as missorting, hyperphosphorylation, aggregation (ranging from pretangle accumulations of Tau to full neurofibrillary tangles), and age-dependent spreading of the pathology in the appropriate brain areas. These studies revealed that high concentrations of (mutant) Tau can generate Tauopathy, that Abeta exacerbates the degeneration, and that Tauopathy leads to synaptic dysfunction and behavioral deficits.

Nevertheless, several issues remain poorly understood. This includes the mode of Tau toxicity and the relationship between Tau levels, Tau conformation, phosphorylation, and aggregation. We set out to test these questions with new transgenic mouse models designed on the basis of our previous experience with Tau properties in vitro (1). We use a regulatable expression system (Tet-Off) that allows one to switch the Tau expression on and off at defined time points and avoids artifacts of expression at an early age (2). We use in parallel two variants of Tau that are similar with regard to microtubule interactions but opposite with regard to aggregation. The first variant is full-length Tau (hTau40, 2N4R) carrying the FTDP-17 mutation {Delta}K280 in the hexapeptide motif of the second repeat. The deletion of Lys-280 facilitates the formation of beta-structure and strongly promotes Tau aggregation ("pro-aggregation mutant") (8, 30). The second variant contains the same {Delta}K280 mutation and two additional proline mutations, one in each of the two hexapeptide motifs (I277P and I308P). They disrupt beta-structure and thus abrogate aggregation ("anti-aggregation" mutant). For both types of mutants we used a Tet-Off inducible system that combines the tetracycline-responsive transactivator and the calcium/calmodulin kinase II{alpha} (CaMKII{alpha}) promoter (31) for the temporal and regional expression of Tau. The temporal control of the Tet-Off system allows us to study age-dependent effects and reversibility of Tau pathology. The CaMKII{alpha} promoter enables us to achieve a preferential expression of the transgene in the entorhinal region and hippocampus so that the spatial distribution of the human Tau mutants in our model is similar to that in AD. With these mouse models we show that pathological conformation, aggregation, and loss of synapses depend mainly on the beta-propensity of the protein. By contrast, hyperphosphorylation and somatodendritic missorting are observed with both Tau mutants, indicating that this depends mainly on Tau concentration but not on aggregation. All changes can be reversed by switching off the Tau expression. Thus, strategies to prevent Tau aggregation should also protect neuronal integrity.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Transgenic Constructs—The bidirectional transcription unit encoding the human longest Tau central nervous system isoform with deletion of lysine 280 (hTau40/{Delta}K280) was generated by PCR-based site-directed mutagenesis and PCR amplification. This cDNA was inserted at the ClaI and SalI restriction sites into the pBI-5 bidirectional expression vector carrying the reporter luciferase gene (32). The resultant plasmid DNA was digested with XmnI and AseI enzymes and the 5.75-kb fragment containing the Tet-operon responsive element (tetO), bidirectional cytomegalovirus promoter (Ptet-bi), hTau40/{Delta}K280, and luciferase sequences, was used for the introduction by microinjection into fertilized DBA eggs.

Animals—The activator mouse line (CaMKII{alpha}-tTA mice) in which the tTA transgene is under control of the CaMKII{alpha} promoter was a gift from Dr. E. Kandel, Columbia University, New York (31). For the generation of responder mice we used the bidirectional plasmid carrying the bidirectional tetO responsive cytomegalovirus promoter (Ptet-bi), followed by both a TAU mutant in one direction and luciferase reporter sequences in the other. The double transgenic mice were generated by crossing the activator and responder mice, resulting in double transgenic progeny constitutively expressing both transgenes, the mutant TAU and luciferase. The expression can be switched on by withdrawal of doxycycline, and switched off by addition of doxycycline (200 µg/ml in the drinking water) (Tet-Off system). All animals were handled according to standards and guidelines based on the German Animal Welfare Act. The offspring of mice were screened by PCR using the primer pair 5'-AATGAGGTCGGAATCGAAGG-3'/5'-TAGCTTGTCGTAATAATGGCGG-3' for activator genes and the two primers pairs 5'-GACCTTCCGCGAGAACGCCAAA-3'/5'-AAGAACAATCAAGGGTCCCCA-3' and 5'-CCAGCCTGGAAGACGAAGCT-3'/5'-GCGGAAGACGGCGACTTGGG-3' for responder transgenes.

Luciferase Reporter Gene Assay—Luciferase activity was used to filter out positive founder animals to ensure that the integration site is not within a silencer region. To make sure that the effects in switched off animals are not due to the silenced integration site but due to the switched off TAU gene, fibroblasts from mouse ear tissue were cultivated, transfected with 400 ng of plasmid encoding the reverse transactivator protein, and induced with 1 mg/ml doxycycline. 18 h later cells were harvested, lysed, and luciferase activity was measured with the Promega kit according to the manufacturers protocol. Non-induced fibroblasts served as a reference.

Tissue Preparation—To quantify the species of Tau and the extent of Tau aggregation, two methods were used for the preparation of the tissue, the Sarkosyl extraction method (33) and the sequential extraction with "reassembly buffer" (RAB), "radioimmunoprecipitation assay buffer" (RIPA) and 70% formic acid (14). For both protocols, the brain was dissected into the cortex, hippocampus, and the rest of the brain.

RAB-RIPA-Formic Acid Extraction—The tissue was homogenized in 2 volumes of RAB (0.1 M MES, pH 7, 1 mM EGTA, 0.5 mM MgSO4, 0.75 M NaCl, 0.02 M NaF, 1 mM phenylmethylsulfonyl fluoride and protease inhibitors (Complete Mini, Roche Applied Science) using a DIAX 900 homogenizer (Heidolph) and centrifuged for 20 min at 50,000 x g and 4 °C. The RAB supernatants were collected. The pellets were suspended in 2 volumes of RIPA (50 mM Tris, pH 8, 150 mM NaCl, 1% Nonidet P-40, 5 mM EDTA, 0.5% sodium deoxycholate, 0.1% SDS) and centrifuged as above. The RIPA supernatants were also collected. The pellets were resuspended in 0.6 volumes of 70% formic acid and again centrifuged as above. The supernatants were dialyzed against 50 mM Tris, pH 7.4, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride and used for immunoblot analysis together with the RAB and RIPA supernatants.

Sarkosyl Extraction—The tissue was homogenized (DIAX 900, Heidolph) in 3 (for cortex) and 10 volumes (for hippocampus) of buffer H (10 mM Tris, pH 7.4, 0.8 M NaCl, 1 mM EGTA, 10% sucrose, 1 mM phenylmethylsulfonyl fluoride, protease inhibitors). The homogenate was kept on ice for 20 min and centrifuged for 20 min at 27,200 x g and 4 °C. The pellets were homogenized a second time in the same volumes of buffer H and centrifuged as above. The supernatants from both centrifugation steps were combined (S1 + S2). The pellets (P2) were resuspended in 3 volumes of buffer III (0.5 M NaCl, 5 mM EGTA, 50 mM Na-PIPES, 50 mM NaF, 1 mM Na3VO4, 5 mM dithiothreitol, 5 mM Na2H2PO7, 500 µM phenylmethylsulfonyl fluoride, 1 µM microcystin, 10 µg/ml leupeptin, aprotinin, and pepstatin A). The combined supernatants S1 + S2 were mixed with 1% N-lauroylsarcosinate (Sarcosyl) and 1% mercaptoethanol, incubated for at least 1 h at 37 °C, and centrifuged for 35 min at 150,000 x g at room temperature. The supernatants S3 were collected. The pellets (P3) were resuspended in 0.5 volumes of Tris-buffered saline (10 mM Tris, 154 mM NaCl). All fractions (S1 + S2, P2, S3, and P3) were used for immunoblot analysis.

Immunoblot Analysis—Fractionated tissue extracts were dissolved in sample buffer containing 2% beta-mercaptoethanol and 2.5% SDS and boiled. Equal volumes of the samples were separated on 8-10% SDS-PAGE gels and transferred onto polyvinylidene fluoride membranes using blotting buffer (38.4 mM Tris, 31.2 mM glycine, 0.024% SDS, 20% methanol) in a semi-dry chamber. After blocking with Tris-buffered saline + 5% nonfat milk + 0.1% Tween 20, the membranes were incubated with various primary antibodies overnight at 4 °C and subsequently incubated with peroxidase-conjugated secondary antibodies (DAKO, anti-mouse and anti-rabbit, 1:5,000) for 1 h at room temperature. Bound antibodies were detected using the enhanced chemiluminescence system ECL Plus (Amersham Biosciences). The intensity of immunolabeling was quantitatively analyzed with the imaging system LAS 3000 (Fuji). In some cases polyvinylidene fluoride membranes were stripped (62.5 mM Tris, pH 6.8, 2% SDS, 0.7% beta-mercaptoethanol for 30 min at 60 °C) and used several times with different primary antibodies.

Immunohistochemistry—Brains from hTau40{Delta}K280 transgenic mice and control animals were either immersion fixed in 3.5% phosphate-buffered formaldehyde (29 mM NaH2PO4·H2O, 45.8 mM Na2HPO4, pH 7.0), or they were from animals that had been perfused transcardially with 4% paraformaldehyde in PBS (Biochrom AG). Histological staining and immunohistochemistry were performed on 5-µm paraffin sections. Paraffin was removed from the sections by immersing them in xylole substitute (Fluka), followed by sequential incubation in a descending series of ethanol. Endogenous peroxidase activity was blocked using 0.6% H2O2 in 100% ethanol as an intermediate step. Epitope retrieval was done dependent on the primary antibodies and performed in citrate buffer (1.8 mM citric acid, 8.2 mM sodium citrate, pH 6) for 5 min at 95 °C. The sections were incubated with primary antibodies diluted in Tris-buffered saline/Triton (50 mM Tris, pH 7.6, 145 mM NaCl, 0.1% Triton X-100) for 1 h at room temperature. All following steps were done using the Vectastain Elite ABC Kit Universal (Vector Laboratories) according to the manufacturer's protocol. 3,3'-Diaminobenzidine plus solution (DAKO) served as substrate for the peroxidase. In some experiments the sections were counterstained with Mayer Hemealaun (MERCK). Sections were dehydrated in an ascending series of ethanol and xylol substitute and mounted in Permount (Fisher Scientific). Gallyas silver staining was done on 5-µm paraffin sections as described (9).

Antibodies—Antibodies used were K9JA (pan-Tau antibody from DAKO, A0024, recognizes human and mouse Tau), MC-1 (pathological Tau conformation) and PHF-1 (phospho-Ser-396 + Ser-404) (both gifts from P. Davies, Albert Einstein College), and 12E8 (phosphorylated KXGS motifs, Ser(P)-262, Ser(P)-356, gift from P. Seubert, Elan Pharma). Ser(P)-199, Thr(P)-231, Ser(P)-214, and Ser(P)-46 recognize phosphorylation at single SP or TP sites (custom-made, Eurogentec), HT-7 (PerBio), and T14 (Zymed Laboratories Inc.) are human Tau specific. AT180 (Thr(P)-231 + Ser(P)-235) and AT8 (Ser(P)-202 + Thr(P)-205) recognize dual phosphorylation sites (PerBio).

Synapse Count—Animals were perfused transcardially with 4% paraformaldehyde in phosphate buffer (50 mM NaH2PO4, 50 mM Na2HPO4, pH 7.4), and analysis of synapses was performed as described (34). Briefly, 400-µm hippocampal slices after fixation in 1% glutaraldehyde, 1% paraformaldehyde in 0.1 M phosphate buffer were postfixed in 1% OsO4, dehydrated in ethanol, and embedded in Epon 820. Blocks were trimmed to contain only the stratum pyramidale and radiatum of the CA1 region. Ultrathin sections were cut on a Reichert-Jung OmU3 ultramicrotome. The sections were stained with uranyl acetate, followed by lead citrate. The spine synapse density was calculated using stereological methods as described (35).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Inducible Transgenic Mice—We generated two inducible transgenic mouse models with animals expressing hTau40 (the longest isoform in human CNS, 2N4R) carrying the deletion mutation {Delta}K280 (Fig. 1A). This mutation is known from a case of FTDP-17 (36) and strongly increases the tendency of Tau to aggregate into PHFs in vitro because it enforces the beta-structure around the hexapeptide motif in repeat 2 (pro-aggregation mutant). By contrast the additional point mutations I277P (in R2) and I308P (in R3) abrogate the ability of Tau to form PHFs (anti-aggregation mutant) because the prolines act as beta-breakers (8, 30). We used a transgenic mouse where the tetracycline transactivator (tTA) gene (37) was placed downstream of the CaMKII{alpha} promoter to restrict the expression of tTA protein to the forebrain (31). The choice of the CaMKII{alpha} promoter, rather than the Thy-1 promoter used in earlier studies (12, 27, 38) was dictated by the need to avoid the expression of Tau in the spinal cord where it would impair axonal traffic and cause motor neuron defects (15, 39). The second mouse contained the responder Tau cDNA downstream of the bidirectional tTA responsive promoter Ptetbi-1 to enforce a tTA-dependent Tau expression. The two types of mice harboring the TAU or activator transgenes were bred to generate bigenic offspring. Expression was repressed by doxycycline during gestation and for 6 weeks postnatally to avoid any disturbances during development to adulthood. As expected, transgenic Tau protein expression in bigenic mice was restricted to the forebrain with highest levels in the hippocampus, cortex, striatum, and the olfactory bulb, and occurred predominantly in neurons (Fig. 1B).


Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		FIGURE 1.
Generation of inducible transgenic mice expressing the longest human Tau isoform with pro- and anti-aggregation mutations. A, illustration of Tau mutant transgenes hTau40{Delta}K280 (pro-aggregation) and hTau40{Delta}K280/PP (anti-aggregation). B, the CaMKII{alpha} promoter directed the transgene expression to the forebrain (mainly hippocampus, cortex, and striatum, shown on sagittal paraffin sections stained with pan-Tau antibody K9JA).

 


Figure 2
View larger version (22K):
[in this window]
[in a new window]

  		FIGURE 2.
Tau aggregation increases in pro-aggregation but not in anti-aggregation mice. Western blot analysis using the pan-Tau antibody K9JA shows soluble and Sarkosyl-insoluble Tau species. Human Tau (arrow, upper band) can be distinguished from mouse Tau (arrowhead, lower bands). A, hTau40{Delta}K280 accumulates in the Sarkosyl-insoluble fractions with increasing age of the animals. B, hTau40{Delta}K280/PP does not show up in the Sarkosyl-insoluble fractions. C, comparison of two aged mice (24 months gene expression), one with full tangle pathology (starred lanes, left) with an upward shift of human and mouse Tau in the SDS gel due to hyperphosphorylation, the other showing only pre-tangle aggregates and no shift in the SDS gel.

 
Tau Pre-tangle Aggregation in the Transgenic Mice—The use of the longest human Tau isoform as a transgene made it possible to discriminate between the human Tau protein (mass ~ 67 kDa) and the three endogenous mouse Tau isoforms (45-55 kDa, Fig. 2, A and C), which represent the three 4-repeat isoforms typical of adult rodent brain (40). This allowed the reliable quantification of the ratio of human to endogenous mouse Tau by densitometry of the bands on gels or Western blots with one pan-Tau antibody. The expression ratio of total human Tau over endogenous Tau in transgenic mice varied from 1- to 3-fold for both mutants. The relatively low overexpression ensures that other Tau-induced changes are minimized.

We investigated the solubility of Tau in the two transgenic mouse lines hTau40{Delta}K280 and hTau40{Delta}K280/PP (pro- and anti-aggregation mutants). Across all ages the majority of the human pro-aggregation Tau was recovered in the soluble fractions (Fig. 2, A and B). A minor part of Tau was detected in the Sarkosyl-insoluble fractions, but this fraction clearly increased with age, from 2% of the total human Tau at 4 months to 13% at 8 months. In very old mice (>24 months Tau expression) we found up to 18% (Fig. 2C). We next analyzed the endogenous mouse Tau in the different fractions (Sarkosyl soluble and insoluble). Like the human Tau, endogenous mouse Tau was present mainly in the soluble fractions, but small amounts (~1-9% of total mouse Tau) were detected in the Sarkosyl-insoluble pellets. As a control, non-transgenic animals had no detectable insoluble Tau (Fig. 2B). We conclude that the additional level of human Tau caused a time-dependent aggregation, and that both mouse and human Tau partially co-aggregated in the insoluble fractions of transgenic animals.

We compared these results with the solubility of Tau in transgenic mice expressing the anti-aggregation mutant hTau40{Delta}K280/PP. This protein was recovered entirely in the soluble fractions, even at 24 months of age, and there was no detectable Sarkosyl-insoluble fraction. The same was true for the endogenous mouse Tau (Fig. 2, A and B). This shows that elevation of Tau as such is not sufficient for inducing aggregation; the additional requirement is that Tau must allow a beta-conformation in the hexapeptide motifs to support self-propagating assembly. We further conclude that the endogenous mouse Tau, which contains the hexapeptide motifs in a wild-type form and therefore is in principle capable of aggregation (41), cannot persuade the exogenous anti-aggregation Tau to aggregate. It is remarkable that the behavior of Tau in the transgenic mice corresponds exactly to the behavior in vitro, except that the time scales are longer.

Similar results were obtained when Tau was sequentially extracted with the different buffers RAB, RIPA, and formic acid described previously (14). In the case of the pro-aggregation mouse line, most of the protein was soluble and detected in the first two fractions RAB and RIPA, but in older animals (>5.5 months) it began to appear in the insoluble fraction formic acid. Here, too, mouse Tau co-aggregated with human Tau in pre-tangles. By contrast, the anti-aggregation mutant remained soluble at all extraction stages (data not shown).

Early Pathological Hallmarks of Human Tau—The formation of neurofibrillary tangles in AD is preceded by a "pretangle" stage characterized by a translocation of Tau from the axon to the somatodendritic compartment, a "pathological" conformation of Tau detected by certain antibodies, e.g. MC1, Alz-50 (42), and elevated phosphorylation at several sites (6, 43). Analogous observations were made in the transgenic mice expressing the pro-aggregation mutant Tau. Elevated human Tau (detected by the pan-Tau antibody K9JA) was seen in the dendritic and cell body layers of the hippocampus such as stratum radiatum, molecular layer lacunosum, pyramidal cells of CA1 to CA3, granule cells of the dentate gyrus, hilus cells, and sub-iculum (Fig. 3, A and B). Human Tau was likewise expressed in the cortex of these animals (Fig. 3C). By contrast, in control animals, the mouse Tau appeared pronounced only in the stratum lucidum, the location of axonal mossy fibers (Fig. 3E). The anti-aggregation human Tau had generally a similar distribution as the pro-aggregation mutant, except that it was lower in the dendrites of the hippocampus, especially the stratum radiatum (Fig. 3D).


Figure 3
View larger version (136K):
[in this window]
[in a new window]

  		FIGURE 3.
Somatodendritic mislocalization of human Tau in pro- and anti-aggregation mice. Coronal paraffin sections of hippocampus stained with antibody K9JA recognizing human and mouse Tau. A, pro-aggregation hTau40{Delta}K280 animal. Tau is prominent in the stratum radiatum (sr), the molecular layer lacunosum (ml), the stratum lucidum (sl), and the hilus cells (hc). Less Tau is detected in the pyramidal cell (pc) layer of CA1 to CA3 and the granular cell layer (gc). B and C, magnifications of the CA1 region and the cortex emphasizing the dendritic compartmentalization of Tau. D, anti-aggregation hTau40{Delta}K280/PP animal. The somatodendritic localization of mutant Tau is less pronounced. E, endogenous Tau in a control animal is mainly visible in the mossy fibers of the CA3 region, indicating a normal axonal distribution. Scale bars in A, D, and E, 500 µm; B and C, 50 µm.

 


Figure 4
View larger version (60K):
[in this window]
[in a new window]

  		FIGURE 4.
Pathological conformation in pro-aggregation but not anti-aggregation mutant mice. Paraffin sections of hippocampus of pro-aggregation (A) and anti-aggregation mice (B) stained with conformation-dependent MC-1 antibody. A, Tau with pathological conformation in cell bodies and dendrites of pyramidal layer and stratum radiatum (MC-1 staining starting at 3 months). B, no MC-1 staining in anti-aggregation mice.

 


Figure 5
View larger version (74K):
[in this window]
[in a new window]

  		FIGURE 5.
Human Tau is highly phosphorylated at KXGS motifs in the repeat domain. A, paraffin sections of pro-aggregation mouse: weak phosphorylation at Ser-262/Ser-356 as seen by 12E8 antibody. B, at 6 months, phosphorylation is strongly enhanced and appears in most apical dendrites of CA1 (arrowheads) and in the cell body layer. C, the anti-aggregation mutant shows phosphorylation at Ser-262 to a lesser extent. Sections were counterstained with hemealaun.

 
Immunohistochemistry with antibody MC-1 revealed an abnormal conformation of Tau, especially in the hippocampus and cortex where it labeled apical dendrites and cell bodies (Fig. 4A). The MC-1 staining started very early at 3 months of gene expression and increased with advancing age. By contrast, no conformationally changed human Tau was detected in the brains of anti-aggregation mice (Fig. 4B). Similarly, we noticed MC-1-positive Tau species in the soluble fractions of the Sarkosyl extraction or in RAB/RIPA fractions but not in the insoluble fractions of the pro-aggregation mutant mice. No MC-1 positive species could be detected in anti-aggregation mice (which contain only soluble fractions).

For the pro-aggregation mutant mice, an increase in Tau phosphorylation was initially observed at 3 months after induction of Tau expression with the antibody 12E8 (Fig. 5). This antibody recognizes two KXGS motifs in the repeat domain (Ser(P)-262, Ser(P)-356) whose phosphorylation causes the detachment of Tau from microtubules (5, 44). On immunohistochemically stained sections of the pro-aggregation mutant mice we found slight phosphorylation at Ser-262 after 3 months of gene expression, increasing at later stages (6 months) with high levels in the apical dendrites of the hippocampal cells (Fig. 5, A and B). Only a few cell bodies were stained in this area. Phosphorylated Tau species were also detected in anti-aggregation mutant mice, although at lower levels (Fig. 5C).

In addition, in the pro-aggregation mutant mouse Tau was highly phosphorylated at many single SP or TP sites that occur mainly in regions flanking the repeat domain, including site Ser-214 (a non-SP site). The phosphorylation was initially detected at 3 months and remained stable with age. We observed phosphorylation at Ser-46, Ser-199, Ser-214, Thr-231, or at the AT180 epitope (Thr(P)-231 + Ser(P)-235) with different intracellular localization of the phosphorylated Tau species. An antibody against Thr(P)-231 showed that most of the phosphorylated Tau was localized in the cell bodies of the pyramidal layer (Fig. 6A). In some cases we found also the proximal processes of the cells stained. Control animals showed no staining with the phosphorylation-dependent antibodies. The dual phosphorylation at Thr-231/Ser-235 was investigated with antibody AT180. Similar to the single phosphorylation at Thr-231 we found most of the AT180-Tau species in the cell bodies of the pyramidal cells (Figs. 6C and 9A). Substantial staining was observed with an antibody recognizing Ser(P)-199. Most of the phosphorylated Tau was localized in the apical dendrites and the cell bodies of both the hippocampal and cortical neurons (Fig. 6E and 9C). The distribution of phosphorylated Tau (Ser(P)-199) was not restricted to certain areas but rather occurred in the whole hippocampus and cortex. By contrast, immunohistochemical staining of phosphorylated Ser-46 revealed a completely different pattern; it was almost exclusively present in the stratum lucidum of the hippocampus (CA3 mossy fibers), indicating a mostly axonal distribution of this phospho-Tau species (Fig. 6G). A similar axonal distribution in the mossy fibers was seen for Ser(P)-214 in the pro-aggregation mutant (Fig. 6I). This residue, together with Thr(P)-212, creates the epitope of antibody AT100, one of the most characteristic antibodies for Alzheimer Tau (45, 46), yet AT100 reactivity was not seen at all in our mouse models.


Figure 6
View larger version (142K):
[in this window]
[in a new window]

  		FIGURE 6.
Single site and double site phosphorylation at SP/TP motifs in pro- and anti-aggregation mice. Panels show paraffin brain sections stained with different antibodies against phosphorylated SP or TP motifs. Overall, phosphorylation of single sites is elevated at early stages and is more pronounced in pro-aggregation than in anti-aggregation mice. By contrast, dual-site phosphorylation (characterized by AD-diagnostic antibodies AT8 and PHF1) tends to be elevated at late stages and only in pro-aggregation mice. Special cases were AT180 (dual site but early stage in both types of mice) and MC-1 (conformation specific, early, but only in pro-aggregation mice). Comparison of pro-aggregation (A, C, E, G, and I) and anti-aggregation mice (B, D, F, H, and J). A and B, Thr(P)-231 is localized in the periphery of cell bodies and in proximal processes of hippocampal CA3 neurons in pro- and anti-aggregation mice. C and D, AT180 staining (Thr(P)-231 + Ser(P)-235) localizes to the cell bodies of pyramidal cells. This dual site is an "early" site, in contrast to others, see below. E and F, Ser(P)-199 staining reveals localization in the apical dendrites of pyramidal CA3 neurons. G and H, Ser(P)-46 staining of Tau is localized only in the stratum lucidum of the CA3 region, in contrast to other SP/TP sites. The pyramidal cell bodies are free of staining, indicating an axonal rather than somatodendritic distribution. I and J, the non-SP/TP site Ser-214 is distributed similar to Ser(P)-46 with most of the phosphorylated Tau species in the mossy fibers of CA3. K, L, M, and N, example of aged pro-aggregation mouse, demonstrating that prolonged gene expression (24 months) leads from pre-tangles to neurofibrillary tangle pathology. K, AT8 staining in CA3 region; L, PHF-1 staining; M, MC-1 staining; and N, gallyas staining in the cortex demonstrates that prolonged gene expression leads from pre-tangles to neurofibrillary tangle pathology.

 
We investigated the phosphorylation of Tau at many other sites that are considered to be relevant for later stages of Tau aggregation and tangle formation. In 1-18-month-old pro-aggregation mice no phosphorylation of Tau at dual sites Thr(P)-212/Ser(P)-214 (AT100), Ser(P)-202/Thr(P)-205 (AT8), or Ser(P)-396/Ser(P)-404 (PHF1), as well as the single site Ser-422. Immunohistochemical studies with these and other antibodies revealed that in most of our transgenic animals, Tau did not undergo neurofibrillary tangle formation. Consistent with that we found very little Tau species stainable by Gallyas silver or Thio-flavin S, even though we detected routinely Tau in the Sarkosyl pellet, considered to be mostly pre-tangle aggregates. Only in very old mice, after longer and higher Tau expression (>24 months) did we find phosphorylation of pro-aggregation mutant Tau at the epitopes of AT8 and PHF-1 as well as tangles stained by Gallyas silver (Fig. 6, K-N). This indicates that the pre-tangles eventually convert to tangle pathology if the concentration of mutant Tau becomes high enough.

We compared the phosphorylation of pro-aggregation and anti-aggregation Tau at different SP/TP sites. The anti-aggregation mutant showed a similar phosphorylation at Thr-231, Ser-199, Ser-214, Ser46, or the AT180 epitope although the phosphorylation was mostly weaker than in the pro-aggregation Tau mice (Fig. 6, B, D, F, H, and J). In contrast to these single phosphoepitopes we did not find phosphorylation at the dual sites of AT8 or PHF-1 in anti-aggregation mutant mice, nor Gallyas silver staining, even in very old mice.

Overall, these observations suggest that the elevation of Tau leads to increased early phosphorylation at various single SP/TP sites in the regions flanking the repeats, combined with a gradual increase of conformationally altered pre-tangle Tau. However, the appearance of silver-stained tangles correlates closely with hyperphosphorylation at the dual sites AT8 and PHF1 in very old mice (Fig. 6).

Reversibility of Tau Pathology—One of the major aims of this study was to determine whether and to what extent the Tau-induced pathological changes are reversible (pathological conformation, missorting, hyperphosphorylation, and aggregation). This was investigated by first switching on the Tau expression by withdrawal of doxycycline, then switching the expression off by addition of doxycycline. Switching off the TAU gene expression led to an almost complete reduction of human Tau levels in the mouse brains within 6 weeks (Fig. 7). Interestingly, this decrease was more pronounced in the soluble than the insoluble fractions. Human Tau disappeared from the dendrites in the stratum radiatum and stratum lacunosum-moleculare (Fig. 7, A and B). Cell layers became completely free of human Tau. Solely in the stratum lucidum, some Tau remained after switching off the human TAU gene, presumably due to endogenous mouse Tau.

Switching off the human TAU gene expression resulted in the complete reversibility of the pathological conformational change of the protein, the pre-tangle aggregation, as well as a reversal of the hyperphosphorylated state of human Tau, both within and outside the repeat domain (Figs. 8 and 9). MC1 positive Tau became undetectable after switching off the Tau gene for 6 weeks and disappeared independently at the age of the transgenic animals (Fig. 9, G and H). Similar results were obtained for the cases of the phosphorylation-dependent antibodies Ser(P)-46, Ser(P)-199, Ser(P)-214, Thr(P)-231, and AT180. Whereas the staining at these sites became pronounced after 6-8 months of Tau expression, the switching off of the TAU gene led to complete clearance of the phosphorylated Tau from dendrites and cell bodies at all ages of the mice (Fig. 9, A-F). The reversibility of the phosphorylation was, furthermore, independent of the previous intracellular localization of the Tau species (cell bodies, dendrites) detected with the different antibodies. This coincides with the overall clearance of Tau protein from the hippocampus after switching off the Tau gene.


Figure 7
View larger version (46K):
[in this window]
[in a new window]

  		FIGURE 7.
Somatodendritic mislocalization is reversed and Tau levels are decreased after switching off expression of pro-aggregation Tau. A and B, paraffin sections of a pro-aggregation mouse with 10-month Tau expression (A) and of a mouse with 8.5-month Tau expression and subsequent switch-off for 1.5 months (B). Brain sections after switching off Tau expression are devoid of hTau40{Delta}K280 protein. The somatodendritic localization of Tau (A) is fully reversible in 6 weeks. The remaining protein is localized mainly in the stratum lucidum of CA3 (inset in B). C, Western blots of soluble and insoluble fractions of the same animals show significant reduction in human pro-aggregation Tau expression, whereas endogenous mouse Tau levels remain stable. D, quantification of Western blots indicates the disappearance of hTau40{Delta}K280 to control levels.

 


Figure 8
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 8.
Phosphorylation of pro-aggregation Tau at Ser-262 is abolished after switching off Tau expression. A and B, paraffin sections of brains of pro-aggregation animals with continuous or interrupted Tau expression, stained with 12E8 antibody. A, pro-aggregation of Tau phosphorylated at Ser-262 shows a dendritic distribution within the hippocampus. B, suppression of the Tau gene expression for 6 weeks leads to clearance of dendrites from Ser(P)-262-Tau. C, biochemical analysis by differential extraction (RAB, RIPA, and formic acid (FA)) of Tau from the brain of a transgenic animal with 8 months of continuous expression is strongly phosphorylatedatSer-262 in all three fractions (lanes8). Interruption of the Tau expression for 6 weeks decreases the phosphorylation at Ser-262 to the limit of detection (lanes 8 off). D, the quantitative analysis demonstrates the Tau gene suppression and the reversibility of the phosphorylation in switched off animals.

 
Biochemical analysis of the MARK target site Se(P)-262 and other KXGS motifs showed a complete reversibility of human Tau phosphorylation at these sites (Fig. 8, C and D). The upper Tau band (corresponding to human Tau) disappeared independently of the length of TAU gene expression. However, the lower bands (corresponding to endogenous phosphorylated mouse Tau) did not vanish upon switching off the expression of human Tau, indicating that some of the mouse Tau still remained phosphorylated at KXGS motifs. This reversibility was confirmed in the immunohistochemical staining. Tau phosphorylated at the MARK target sites was no more detectable in neuronal cells after switching off the human Tau gene for 6 weeks (Fig. 8, A and B).

Loss of Synapses—The loss of synapses is known to precede neuronal degeneration in AD (47, 48). We therefore performed electron microscopy experiments to count the total number of synapses and the fraction of spine synapses in the CA1 region of the hippocampus (Fig. 10). Transgenic animals expressing the pro-aggregation mutant hTau40{Delta}280 showed a moderate but significant decline in the number of synapses, particularly of spine synapses by ~40% after 13 months. The effect was less pronounced with the anti-aggregation mutant mice (~20%). This suggests that the decline of synapses is due to a combination of Tau aggregation and other effects (possibly transport inhibition (49)).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Tau protein undergoes several pathological changes in Alzheimer disease and other Tauopathies, most notably aggregation, hyperphosphorylation, and missorting to the somatodendritic compartment. The linkage between these properties is not well understood at present. The purpose of this study was to separate the effects of aggregation. To this end, we created two new transgenic mouse models based on the structural properties of Tau. In the first model we expressed hTau40, the longest isoform in the human central nervous system, with the deletion mutation {Delta}K280 known from FTDP-17 (36). This mutation, located in the hexapeptide motif of repeat R2, is a potent enhancer of Tau aggregation because it increases the propensity of Tau for beta-structure (hence termed pro-aggregation mutant) (30). In the second variant (anti-aggregation mutant) we inserted two additional point mutations into the two hexapeptide motifs (I277P and I308P) to block beta-structure and thus aggregation (8). Apart from aggregation the two variants show no significant differences, e.g. they have similar affinities to microtubules and contain the same phosphorylatable residues. Therefore this pair of Tau variants makes it possible to observe effects that are solely due to aggregation.


Figure 9
View larger version (96K):
[in this window]
[in a new window]

  		FIGURE 9.
Reversibility of pathological conformation and hyperphosphorylation of Tau after switching off expression of pro-aggregation Tau. Paraffin hippocampal sections of transgenic animals expressing hTau40{Delta}K280 were stained with different antibodies, recognizing the phosphorylation state (Ser(P)-199, Thr(P)-231, AT180, Ser(P)-46) or a pathological conformation (MC1). A, C, E, and G, illustrates continuous gene expression; B, D, F, and H, shows switch-off for the last 6 weeks. A-F, Tau phosphorylation at sites AT180, Ser(P)-199, and Ser(P)-46 disappears from hippocampal cells after switch-off. G and H, the conformational change seen by the MC1 antibody and its somatodendritic localization is fully reversible.

 


Figure 10
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 10.
Reduction of spine synapses in transgenic mice. Brains of transgenic and control animals were embedded, cut, and stained for electron microscopy. The numbers of synapses and spine synapses were counted in 2 serial sections. Pro-aggregation mutant Tau mice showed a significant reduction in spine synapses (-40%, black bar on right), anti-aggregation mice show a 20% reduction (gray bar). The loss of total synapses was less conspicuous but showed a clear trend (-20 or -15%, respectively, left). Values are shown as mean ± S.E.

 
The transgenes were expressed via the CaMKII{alpha} promoter, for two reasons. First, the expression of Tau was directed to the forebrain to achieve a realistic model for AD and other tauopathies. The second aim was to avoid interference with spinal cord and motor neuron pathology. In some earlier mouse models, Tau was expressed under the Thy-1 or prion promoters (12, 14, 15, 26, 27, 38), which resulted in motor neuron defects and precluded the analysis of cognitive deficits. The likely cause is that the elevation of Tau tends to inhibit axonal traffic by interfering with microtubule-based traffic (39, 49), which becomes most noticeable in the case of long axons such as the motor neurons (50).

These features were combined with regulatable expression using the Tet-off system, whereby the transgene is expressed only in the absence of doxycycline (37). This served a 3-fold purpose. First, during brain development, Tau undergoes major changes, notably a strong up-regulation in neurons, redistribution from ubiquitous to axonal, reduction of the embryonically high level of phosphorylation, and diversification of isoforms by alternative splicing (2, 51). However, AD is a disease of advanced age, and therefore it is preferable to allow normal development into adulthood before switching on transgenic Tau. The second purpose is to allow Tau expression and abnormal changes to take place, and then to switch the expression off to observe recovery of the brain. Third, the Tet-off system enables controlling the level of exogenous Tau via the doxycycline concentration. This allows one to test how the age of onset of Tau expression or the Tau concentration affects pathology.

Generation of the Tau transgenic mouse models are valuable tools for the research of the neurodegenerative diseases. The principle behind most models was the overexpression of one of the Tau isoforms, based on the assumption that a higher level of Tau would accelerate aggregation. An alternative approach was the overexpression or activation of kinases, assuming that this would lead to hyperphosphorylation and hence aggregation of Tau (18, 24, 25). The expression of human Tau isoforms met with limited success with regard to Tau aggregation, presumably because of the high solubility of Tau, short lifetime of the animals (compared with the pathological process), and other factors. However, the approaches could be improved by: (i) using Tau mutants with a higher aggregation potential, e.g. P301L: (20, 26, 27), P301S (21, 22, 28); (ii) combining Tau expression with enhanced Abeta activity, by using APP or PS mutations (19, 29); or (iii) expressing several Tau isoforms analogous to human brain (16). Compared with other mouse models of tauopathy carrying the tau transgenes with P301L and P301S mutations our results can be summarized as follows.

Tau Conformation—A major parameter that decides whether or not Tau will aggregate is the propensity for beta-structure around the hexapeptide motifs in the repeat domain. Enhancing this conformation will lead to faster aggregation, whereas abrogating the beta-structure will prevent aggregation. The kinetics of aggregation in mouse brains is much slower than in vitro, but the behavior in vitro is surprisingly predictive of the behavior in vivo. With regard to the pro-aggregation mutation {Delta}K280, our results agree well with other mouse models that made use of other FTDP-17 mutations, notably P301L, because in both cases the rate of aggregation is enhanced in vitro and in vivo (19, 20, 26, 27, 50). On the other hand, aggregation can be blocked completely by destroying the beta-propensity of the hexapeptide motifs in Tau by inserting prolines. This illustrates that even in a complex organism the mechanistic explanation of aggregation is rather simple.

Tau Concentration—The rate of aggregation is modified by the concentration (as expected for a self-assembling system). Therefore, to observe Tau aggregation within the limited lifespan of a mouse, it is important to increase the Tau concentration (52, 53). However, the level of exogenous Tau can be kept lower if one uses a Tau species with a higher beta-propensity, as is the case with the {Delta}K280 mutation used here, or for the P301L or P301S mutations used by others. Compared with these mouse models (20, 22, 50), our level of Tau expression is lower. This would explain why our mice do not reach the full neurofibrillary tangle state (except for beginnings at very old age), why we do not observe pronounced AT8 and PHF1 staining (which accompanies tangles), and why there is synapse loss but no neuronal loss.

Tau Composition—A further modifier of the rate of aggregation is the type of expressed Tau. In general, shorter isoforms (e.g. the fetal form hTau23) aggregate more readily because the domains outside the repeats tend to slow down aggregation (3, 4). This is demonstrated by several mouse models (13, 14, 54), and our cell models expressing the repeat domain alone underscores this finding (55). The advantage of using the longest human isoform hTau40 (2N-4R), despite the slower rate of aggregation, is that it is more easily distinguished from the endogenous adult mouse isoforms where the corresponding 2N-4R form has a low abundance (50).

Stages of Aggregation—Aggregation becomes detectable in several stages, initially in the pre-tangle state by staining with conformation-dependent antibodies, e.g. MC1, Alz50 (42, 56) in stained sections, or by a Sarkosyl-insoluble pellet by biochemical extraction (33). At this point, PHFs may still be difficult to detect, and most aggregates are still in a pre-tangle oligomeric state. PHFs would be expected to gradually assemble and coalesce into neurofibrillary tangles (detectable by Gallyas silver staining); however, because we use the "slow" hTau40 isoform, the aggregation in our models proceeds only to the pre-tangle state in the case of the pro-aggregation Tau (with some exceptions for very old mice, >24 months gene expression), although no sign of any aggregation is seen with the anti-aggregation Tau at any time. To observe pronounced tangle formation at earlier time points one would require higher Tau concentrations (19-22) or more amyloidogenic Tau variants (e.g. the repeat domain only with pro-aggregation mutations).

Tau Co-aggregation—The question of whether endogenous mouse Tau co-assembles with exogenous human Tau has been a matter of debate (12, 18, 57). In our case, we find that mouse Tau is clearly incorporated into the pre-tangle aggregates, provided that the exogenous Tau suffices to initiate the aggregation (e.g. by carrying pro-aggregation mutations). This is consistent with our earlier observation that normal mouse Tau does not differ significantly from normal human Tau in terms of assembly properties, and that both types of Tau can be co-polymerized in suitable conditions (41).

Tau Missorting—Another controversial issue has been the relationship between missorting of Tau into the somatodendritic compartment and aggregation. In our models, we clearly observe missorting of both mutant Tau variants. The extent and the distribution of missorted Tau differs somewhat, which may depend on cell type and level of expression. But the dominant feature is that missorting occurs when Tau levels are increased, independently of aggregation. In other words, there is no evidence to assume that the missorting of Tau is caused by aggregation. This is best explained in terms of the general behavior of Tau in cells. In principle, Tau diffuses readily into the soma and dendrites when it becomes elevated (e.g. by expression or microinjection) (58) so that the normal axonal sorting machinery becomes overwhelmed (59).

Tau Phosphorylation—A major discussion point in the field is the question whether phosphorylation primes Tau for aggregation. Our data leave this possibility open: in both types of mice, Tau variants become phosphorylated at a number of sites, especially single sites (e.g. Ser-262 and Ser-214, both of which potently detach Tau from MT, as well as Ser-199, Ser-46, and Ser-231), but the effect is more pronounced with pro-aggregation mutants. The transition from pre-tangle to tangle pathology in old pro-aggregation mice correlates with a higher level and new quality of phosphorylation, notably the appearance of dual phosphorylation sites characteristic of AD-Tau, recognized by antibodies AT8 (epitope Ser(P)-202 + Thr(P)-205), AT100 (Thr(P)-212 + Ser(P)-214), and PHF1 (Ser(P)-396 + Ser(P)-404) (16, 60, 61). It appears that pairwise phosphorylation is indicative of advanced stages of aggregation (Fig. 6N). It is also noteworthy that the phosphorylation at KXGS sites (recognized by antibody 12E8) or Ser(P)-214 rises at an early time point, consistent with the notion that the cell responds to increased Tau by detaching it from microtubules via phosphorylation in the repeat domain and/or Ser-214. This could be followed by the (early) conformational change visualized by antibody MC1, and later by aggregation.

Synapses Versus Tau Aggregation—Loss of synapses is considered as one of the early signs of AD (47, 48), and therefore it is notable that there is a moderate but significant loss of synapses in the mice. It is particularly visible at the level of spine synapses (40%) in the CA1 region and appears to be partly related to aggregation, because it is most pronounced with the pro-aggregation Tau mutant. This suggest that there is toxicity caused by pre-tangle Tau species, perhaps oligomers, which precede outright tangle formation. An additional cause may be the inhibition of intraneuronal traffic, due to the obstruction of the microtubule tracks (49), or other toxic effects of Tau (62, 63). Note that both pro- and anti-aggregation mutants bind equally well to microtubules and therefore have the same capacity of inhibiting transport. This would explain why the anti-aggregation mutant also shows some synapse loss.

Reversibility of Tau-induced Changes—Nearly all Tau-induced pre-tangle changes are reversible when the expression of Tau is discontinued. This means that the pathways for Tau degradation, probably via proteasomes (64), are still intact. Therefore, in contrast to the inducible Tau mouse model of Santacruz et al. (20) we did not observe a time point beyond which the changes in Tau became irreversible. A possible explanation is that our Tau expression levels were too low so that the mice did not develop the irreversible tangle state within the lifespan tested so far. These observations are also consistent with those of Oddo et al. (29) where the Abeta-induced pre-tangle pathology disappears upon Abeta immunization, but fully formed tangles remain.

In conclusion, in this study we attempted to test to what extent the observations on Tau-transgenic mice can be explained by the biophysical properties of Tau. With regard to aggregation, the correlation is remarkably good if one allows for appropriate scaling of the reaction rates (from hours in vitro to days in cell models to months in mice). Other pathological changes (phosphorylation, missorting) correlate only partly with aggregation and could be explained by changes in Tau interaction partners (e.g. imbalance of the phosphorylation potential, or overwhelming the sorting machinery of the cell). However, Tau aggregation clearly enhances synaptic decay, the likely cause of cognitive deficits in AD and mice (23, 65). This reinforces the view that a reduction of abnormal protein aggregation is a means to combat the disease.


   FOOTNOTES
 
* This research was supported in part by grants from Max-Planck-Gesellschaft and Deutsche Forschungsgemeinschaft. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 To whom correspondence should be addressed. E-mail: mandelkow{at}mpasmb.desy.de.

2 The abbreviations used are: MT, microtubule; AD, Alzheimer disease; FTDP-17, frontotemporal dementia with parkinsonism linked to chromosome 17; PHF, paired helical filament; MES, 4-morpholineethanesulfonic acid; PIPES, 1,4-piperazinediethanesulfonic acid; tTA, tetracycline transactivator; Ca/CMKII{alpha}, calcium/calmodulin kinase II{alpha}. Back


   ACKNOWLEDGMENTS
 
We acknowledge the excellent technical assistance of O. Petrova, D. Drexler, and K. Skokann. We thank animal caretakers A. Kurzawa, N. Lüder, and S. Noster (University of Hamburg Medical School) for their dedicated help in breeding. We are grateful to Dr. P. Davies (Albert Einstein College, Bronx, NY) and Dr. P. Seubert (Elan Pharma, South San Francisco, CA) for providing antibodies MC1, PHF-1, and 12E8, and Dr. E. Kandel for the transactivator (CamKII{alpha}-tTA) construct.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Weingarten, M. D., Lockwood, A. H., Hwo, S. Y., and Kirschner, M. W. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 1858-1862[Abstract/Free Full Text]
  2. Garcia, M. L., and Cleveland, D. W. (2001) Curr. Opin. Cell Biol. 13, 41-48[CrossRef][Medline] [Order article via Infotrieve]
  3. Binder, L. I., Guillozet-Bongaarts, A. L., Garcia-Sierra, F., and Berry, R. W. (2005) Biochim. Biophys. Acta 1739, 216-223[Medline] [Order article via Infotrieve]
  4. Mandelkow, E., von Bergen, M., Biernat, J., and Mandelkow, E. M. (2007) Brain Pathol. 17, 83-90[CrossRef][Medline] [Order article via Infotrieve]
  5. Biernat, J., Gustke, N., Drewes, G., Mandelkow, E. M., and Mandelkow, E. (1993) Neuron 11, 153-163[CrossRef][Medline] [Order article via Infotrieve]
  6. Morishima-Kawashima, M., Hasegawa, M., Takio, K., Suzuki, M., Yoshida, H., Titani, K., and Ihara, Y. (1995) J. Biol. Chem. 270, 823-829[Abstract/Free Full Text]
  7. Schweers, O., Schonbrunn-Hanebeck, E., Marx, A., and Mandelkow, E. (1994) J. Biol. Chem. 269, 24290-24297[Abstract/Free Full Text]
  8. von Bergen, M., Friedhoff, P., Biernat, J., Heberle, J., Mandelkow, E. M., and Mandelkow, E. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 5129-5134[Abstract/Free Full Text]
  9. Braak, H., and Braak, E. (1991) Acta Neuropathol. (Berl.) 82, 239-259[CrossRef][Medline] [Order article via Infotrieve]
  10. Spillantini, M. G., Bird, T. D., and Ghetti, B. (1998) Brain Pathol. 8, 387-402[Medline] [Order article via Infotrieve]
  11. Lee, V. M., Goedert, M., and Trojanowski, J. Q. (2001) Annu. Rev. Neurosci. 24, 1121-1159[CrossRef][Medline] [Order article via Infotrieve]
  12. Götz, J., Probst, A., Spillantini, M. G., Schafer, T., Jakes, R., Burki, K., and Goedert, M. (1995) EMBO J. 14, 1304-1313[Medline] [Order article via Infotrieve]
  13. Brion, J. P., Tremp, G., and Octave, J. N. (1999) Am. J. Pathol. 154, 255-270[Abstract/Free Full Text]
  14. Ishihara, T., Hong, M., Zhang, B., Nakagawa, Y., Lee, M. K., Trojanowski, J. Q., and Lee, V. M. (1999) Neuron 24, 751-762[CrossRef][Medline] [Order article via Infotrieve]
  15. Spittaels, K., Van den Haute, C., Van Dorpe, J., Bruynseels, K., Vandezande, K., Laenen, I., Geerts, H., Mercken, M., Sciot, R., Van Lommel, A., Loos, R., and Van Leuven, F. (1999) Am. J. Pathol. 155, 2153-2165[Abstract/Free Full Text]
  16. Andorfer, C., Kress, Y., Espinoza, M., de Silva, R., Tucker, K. L., Barde, Y. A., Duff, K., and Davies, P. (2003) J. Neurochem. 86, 582-590[CrossRef][Medline] [Order article via Infotrieve]
  17. Noble, W., Olm, V., Takata, K., Casey, E., Mary, O., Meyerson, J., Gaynor, K., LaFrancois, J., Wang, L., Kondo, T., Davies, P., Burns, M., Veeranna, Nixon, R., Dickson, D., Matsuoka, Y., Ahlijanian, M., Lau, L. F., and Duff, K. (2003) Neuron 38, 555-565[CrossRef][Medline] [Order article via Infotrieve]
  18. Cruz, J. C., Tseng, H. C., Goldman, J. A., Shih, H., and Tsai, L. H. (2003) Neuron 40, 471-483[CrossRef][Medline] [Order article via Infotrieve]
  19. Oddo, S., Caccamo, A., Shepherd, J. D., Murphy, M. P., Golde, T. E., Kayed, R., Metherate, R., Mattson, M. P., Akbari, Y., and LaFerla, F. M. (2003) Neuron 39, 409-421[CrossRef][Medline] [Order article via Infotrieve]
  20. Santacruz, K., Lewis, J., Spires, T., Paulson, J., Kotilinek, L., Ingelsson, M., Guimaraes, A., DeTure, M., Ramsden, M., McGowan, E., Forster, C., Yue, M., Orne, J., Janus, C., Mariash, A., Kuskowski, M., Hyman, B., Hutton, M., and Ashe, K. H. (2005) Science 309, 476-481[Abstract/Free Full Text]
  21. Schindowski, K., Bretteville, A., Leroy, K., Begard, S., Brion, J. P., Hamdane, M., and Buee, L. (2006) Am. J. Pathol. 169, 599-616[Abstract/Free Full Text]
  22. Yoshiyama, Y., Higuchi, M., Zhang, B., Huang, S. M., Iwata, N., Saido, T. C., Maeda, J., Suhara, T., Trojanowski, J. Q., and Lee, V. M. (2007) Neuron 53, 337-351[CrossRef][Medline] [Order article via Infotrieve]
  23. Roberson, E. D., Scearce-Levie, K., Palop, J. J., Yan, F., Cheng, I. H., Wu, T., Gerstein, H., Yu, G. Q., and Mucke, L. (2007) Science 316, 750-754[Abstract/Free Full Text]
  24. Lucas, J. J., Hernandez, F., Gomez-Ramos, P., Moran, M. A., Hen, R., and Avila, J. (2001) EMBO J. 20, 27-39[CrossRef][Medline] [Order article via Infotrieve]
  25. Muyllaert, D., Terwel, D., Borghgraef, P., Devijver, H., Dewachter, I., and Van Leuven, F. (2006) Rev. Neurol. (Paris) 162, 903-907[Medline] [Order article via Infotrieve]
  26. Lewis, J., McGowan, E., Rockwood, J., Melrose, H., Nacharaju, P., Van Slegtenhorst, M., Gwinn-Hardy, K., Paul Murphy, M., Baker, M., Yu, X., Duff, K., Hardy, J., Corral, A., Lin, W. L., Yen, S. H., Dickson, D. W., Davies, P., and Hutton, M. (2000) Nat. Genet. 25, 402-405[CrossRef][Medline] [Order article via Infotrieve]
  27. Götz, J., Chen, F., Barmettler, R., and Nitsch, R. M. (2001) J. Biol. Chem. 276, 529-534[Abstract/Free Full Text]
  28. Allen, B., Ingram, E., Takao, M., Smith, M. J., Jakes, R., Virdee, K., Yoshida, H., Holzer, M., Craxton, M., Emson, P. C., Atzori, C., Migheli, A., Crowther, R. A., Ghetti, B., Spillantini, M. G., and Goedert, M. (2002) J. Neurosci. 22, 9340-9351[Abstract/Free Full Text]
  29. Oddo, S., Billings, L., Kesslak, J. P., Cribbs, D. H., and LaFerla, F. M. (2004) Neuron 43, 321-332[CrossRef][Medline] [Order article via Infotrieve]
  30. Barghorn, S., Zheng-Fischhofer, Q., Ackmann, M., Biernat, J., von Bergen, M., Mandelkow, E. M., and Mandelkow, E. (2000) Biochemistry 39, 11714-11721[CrossRef][Medline] [Order article via Infotrieve]
  31. Mayford, M., Bach, M. E., Huang, Y. Y., Wang, L., Hawkins, R. D., and Kandel, E. R. (1996) Science 274, 1678-1683[Abstract/Free Full Text]
  32. Baron, U., Freundlieb, S., Gossen, M., and Bujard, H. (1995) Nucleic Acids Res. 23, 3605-3606[Free Full Text]
  33. Greenberg, S. G., and Davies, P. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 5827-5831[Abstract/Free Full Text]
  34. Kretz, O., Fester, L., Wehrenberg, U., Zhou, L., Brauckmann, S., Zhao, S., Prange-Kiel, J., Naumann, T., Jarry, H., Frotscher, M., and Rune, G. M. (2004) J. Neurosci. 24, 5913-5921[Abstract/Free Full Text]
  35. Prange-Kiel, J., Rune, G. M., and Leranth, C. (2004) Eur. J. Neurosci. 19, 309-317[CrossRef][Medline] [Order article via Infotrieve]
  36. Rizzu, P., Van Swieten, J. C., Joosse, M., Hasegawa, M., Stevens, M., Tibben, A., Niermeijer, M. F., Hillebrand, M., Ravid, R., Oostra, B. A., Goedert, M., van Duijn, C. M., and Heutink, P. (1999) Am. J. Hum. Genet. 64, 414-421[CrossRef][Medline] [Order article via Infotrieve]
  37. Gossen, M., and Bujard, H. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 5547-5551[Abstract/Free Full Text]
  38. Probst, A., Gotz, J., Wiederhold, K. H., Tolnay, M., Mistl, C., Jaton, A. L., Hong, M., Ishihara, T., Lee, V. M., Trojanowski, J. Q., Jakes, R., Crowther, R. A., Spillantini, M. G., Burki, K., and Goedert, M. (2000) Acta Neuropathol. 99, 469-481[CrossRef][Medline] [Order article via Infotrieve]
  39. Stamer, K., Vogel, R., Thies, E., Mandelkow, E., and Mandelkow, E. M. (2002) J. Cell Biol. 156, 1051-1063[Abstract/Free Full Text]
  40. Kosik, K. S., Orecchio, L. D., Bakalis, S., and Neve, R. L. (1989) Neuron 2, 1389-1397[CrossRef][Medline] [Order article via Infotrieve]
  41. Kampers, T., Pangalos, M., Geerts, H., Wiech, H., and Mandelkow, E. (1999) FEBS Lett. 451, 39-44[CrossRef][Medline] [Order article via Infotrieve]
  42. Jicha, G. A., Bowser, R., Kazam, I. G., and Davies, P. (1997) J. Neurosci. Res. 48, 128-132[CrossRef][Medline] [Order article via Infotrieve]
  43. Kopke, E., Tung, Y. C., Shaikh, S., Alonso, A. C., Iqbal, K., and GrundkeIqbal, I. (1993) J. Biol. Chem. 268, 24374-24384[Abstract/Free Full Text]
  44. Seubert, P., Mawal-Dewan, M., Barbour, R., Jakes, R., Goedert, M., Johnson, G. V., Litersky, J. M., Schenk, D., Lieberburg, I., Trojanowski, J. Q., and Lee, V. Y.-M. (1995) J. Biol. Chem. 270, 18917-18922[Abstract/Free Full Text]
  45. Matsuo, E. S., Shin, R. W., Billingsley, M. L., Van deVoorde, A., O'Connor, M., Trojanowski, J. Q., and Lee, V. M. (1994) Neuron 13, 989-1002[CrossRef][Medline] [Order article via Infotrieve]
  46. Zheng-Fischhofer, Q., Biernat, J., Mandelkow, E. M., Illenberger, S., Godemann, R., and Mandelkow, E. (1998) Eur. J. Biochem. 252, 542-552[Medline] [Order article via Infotrieve]
  47. DeKosky, S. T., and Scheff, S. W. (1990) Ann. Neurol. 27, 457-464[CrossRef][Medline] [Order article via Infotrieve]
  48. Coleman, P. D., and Yao, P. J. (2003) Neurobiol. Aging 24, 1023-1027[CrossRef][Medline] [Order article via Infotrieve]
  49. Thies, E., and Mandelkow, E. M. (2007) J. Neurosci. 27, 2896-2907[Abstract/Free Full Text]
  50. Terwel, D., Lasrado, R., Snauwaert, J., Vandeweert, E., Van Haesendonck, C., Borghgraef, P., and Van Leuven, F. (2005) J. Biol. Chem. 280, 3963-3973[Abstract/Free Full Text]
  51. Andreadis, A. (2005) Biochim. Biophys. Acta 1739, 91-103[Medline] [Order article via Infotrieve]
  52. Götz, J., Streffer, J. R., David, D., Schild, A., Hoerndli, F., Pennanen, L., Kurosinski, P., and Chen, F. (2004) Mol. Psychiatry 9, 664-683[Medline] [Order article via Infotrieve]
  53. Spires, T. L., and Hyman, B. T. (2005) NeuroRx 2, 423-437[Abstract/Free Full Text]
  54. Higuchi, M., Ishihara, T., Zhang, B., Hong, M., Andreadis, A., Trojanowski, J., and Lee, V. M. (2002) Neuron 35, 433-446[CrossRef][Medline] [Order article via Infotrieve]
  55. Khlistunova, I., Biernat, J., Wang, Y., Pickhardt, M., von Bergen, M., Gazova, Z., Mandelkow, E., and Mandelkow, E. M. (2006) J. Biol. Chem. 281, 1205-1214[Abstract/Free Full Text]
  56. Weaver, C. L., Espinoza, M., Kress, Y., and Davies, P. (2000) Neurobiol. Aging 21, 719-727[CrossRef][Medline] [Order article via Infotrieve]
  57. Liou, Y. C., Sun, A., Ryo, A., Zhou, X. Z., Yu, Z. X., Huang, H. K., Uchida, T., Bronson, R., Bing, G., Li, X., Hunter, T., and Lu, K. P. (2003) Nature 424, 556-561[CrossRef][Medline] [Order article via Infotrieve]
  58. Hirokawa, N., Funakoshi, T., Sato-Harada, R., and Kanai, Y. (1996) J. Cell Biol. 132, 667-679[Abstract/Free Full Text]
  59. Konzack, S., Thies, E., Marx, A., Mandelkow, E.-M., and Mandelkow, E. (2007) J. Neurosci. 27, 9916-9927[Abstract/Free Full Text]
  60. Greenberg, S. G., Davies, P., Schein, J. D., and Binder, L. I. (1992) J. Biol. Chem. 267, 564-569[Abstract/Free Full Text]
  61. Otvos, L., Jr., Feiner, L., Lang, E., Szendrei, G. I., Goedert, M., and Lee, V. M. (1994) J. Neurosci. Res. 39, 669-673[CrossRef][Medline] [Order article via Infotrieve]
  62. Park, S. Y., and Ferreira, A. (2005) J. Neurosci. 25, 5365-5375[Abstract/Free Full Text]
  63. King, M. E., Kan, H. M., Baas, P. W., Erisir, A., Glabe, C. G., and Bloom, G. S. (2006) J. Cell Biol. 175, 541-546[Abstract/Free Full Text]
  64. Kosik, K. S., and Shimura, H. (2005) Biochim. Biophys. Acta 1739, 298-310[Medline] [Order article via Infotrieve]
  65. Shankar, G. M., Bloodgood, B. L., Townsend, M., Walsh, D. M., Selkoe, D. J., and Sabatini, B. L. (2007) J. Neurosci. 27, 2866-2875[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Neurol. Neurosurg. PsychiatryHome page
C Reitz, L Honig, J P Vonsattel, M-X Tang, and R Mayeux
Memory performance is related to amyloid and tau pathology in the hippocampus
J. Neurol. Neurosurg. Psychiatry, July 1, 2009; 80(7): 715 - 721.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurosci.Home page
M.-M. Mocanu, A. Nissen, K. Eckermann, I. Khlistunova, J. Biernat, D. Drexler, O. Petrova, K. Schonig, H. Bujard, E. Mandelkow, et al.
The Potential for -Structure in the Repeat Domain of Tau Protein Determines Aggregation, Synaptic Decay, Neuronal Loss, and Coassembly with Endogenous Tau in Inducible Mouse Models of Tauopathy
J. Neurosci., January 16, 2008; 28(3): 737 - 748.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/43/31755    most recent
M705282200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Eckermann, K.
Right arrow Articles by Mandelkow, E.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Eckermann, K.
Right arrow Articles by Mandelkow, E.-M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2007 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement