Binding of Tau to Heat Shock Protein 27 Leads to Decreased Concentration of Hyperphosphorylated Tau and Enhanced Cell Survival*

Pathological hyperphosphorylated tau is the principal component of paired helical filaments, a pathological hallmark of Alzheimer disease (AD) and a strong candidate for a neurotoxic role in AD and other neurodegenerative disorders. Here we show that heat shock protein 27 (Hsp27) preferentially binds pathological hyperphosphorylated tau and paired helical filaments tau directly but not non-phosphorylated tau. The formation of this complex altered the conformation of pathological hyperphosphorylated tau and reduced its concentration by facilitating its degradation and dephosphorylation. Moreover, Hsp27 rescues pathological hyperphosphorylated tau-mediated cell death. Therefore, Hsp27 is likely to provide a neuroprotective effect in AD and other tauopathies. m the degradation of pathological hyperphosphorylated recombinant His-tau incubated in the by the degradation PHF tau phosphorylated dephosphorylated dephosphorylation dephosphorylation

Pathological hyperphosphorylated tau is the principal component of paired helical filaments, a pathological hallmark of Alzheimer disease (AD) and a strong candidate for a neurotoxic role in AD and other neurodegenerative disorders. Here we show that heat shock protein 27 (Hsp27) preferentially binds pathological hyperphosphorylated tau and paired helical filaments tau directly but not non-phosphorylated tau. The formation of this complex altered the conformation of pathological hyperphosphorylated tau and reduced its concentration by facilitating its degradation and dephosphorylation. Moreover, Hsp27 rescues pathological hyperphosphorylated tau-mediated cell death. Therefore, Hsp27 is likely to provide a neuroprotective effect in AD and other tauopathies.
Tau is a microtubule-associated protein expressed mainly in neurons where it has a role in the assembly and stability of the microtubule network. Tau also is strongly linked to several neurodegenerative disorders as the principal pathological component of the filamentous inclusions found in the brains of AD 1 patients and in a variety of diseases collectively called "tauopathies," including progressive supranuclear palsy, Pick disease, and corticobasal degeneration (1). Although these diseases often are sporadic, the tau gene can harbor mutations that lead to diverse clinical phenotypes collectively grouped under the designation frontotemporal dementia with parkinsonism linked to chromosome 17 (2). Considerable evidence supports the view that hyperphosphorylated tau represents the pathologic entity in the tauopathies, including the observation that increased hyperphosphorylated tau can cause neuronal cell death (3). Because phosphorylation releases tau from microtubules and because tau in the PHF is highly phosphorylated, kinases have been viewed suspiciously as having a possible role in pathogenesis. In vivo evidence for an interaction with tau exists for cyclin-dependent kinase-5 and glycogen synthase kinase-3␤ (GSK-3␤). Previous work has shown that tau hyper-phosphorylation by GSK-3␤ in the fly (4) and cyclin-dependent kinase-5 in the mouse (5) can cause or accelerate neurofibrillary tangle formation in vivo with an attendant worsening of neurodegeneration. Cruz et al. (6) showed that mice transgenic for p25, the truncated cyclin-dependent kinase-5 activator, exhibited neuronal loss in the cortex and hippocampus accompanied by forebrain atrophy, astrogliosis, and caspase-3 activation. Moreover, endogenous tau was hyperphosphorylated, aggregated tau accumulated, and neurofibrillary pathology developed progressively in these animals. The importance of dephosphorylation in neurofibrillary disease is highlighted by the ability of the phosphorylation-dependent prolyl isomerase, Pin1, to provide relative protection from age-dependent neurodegeneration (7). Pin1 recognizes specific phosphorylated serine or threonine residues in tau that are followed by a proline and catalyzes a critical conformational change that allows dephosphorylation at these residues. PP2A and PP1A are major phosphatases for phosphorylated tau (8). PP2A is abundant in the brain and is associated with microtubules (9). Sun et al. (10) showed that inhibition of protein phosphatase 2A and protein phosphatase 1 by calyculin A induced tau hyperphosphorylation and impairment of spatial memory retention in rats. Gong et al. (11) showed that the PP2A inhibition by okadaic acid resulted in increased phosphorylated tau in the axons in metabolically active rat brain slices.
Therefore, the cellular mechanisms that decrease the amount of hyperphosphorylated tau or inhibit the cell death signals conveyed by hyperphosphorylated tau may slow AD progression. Here we show that Hsp27, one of the small heat shock proteins, directly associated with hyperphosphorylated recombinant tau and PHF tau from the AD brain and can rescue cell death caused by pathological hyperphosphorylated tau.
PHF Tau Purification-PHF tau was purified by the method of Greenberg and Davies (16). PHF preparations were precipitated with * 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.
10% trichloroacetic acid. Proteins were resuspended in 1% SDS, 0.1 M phosphate-buffered saline, and the buffer was exchanged to 0.01 M phosphate-buffered saline for further experiments.
Hsp27 Purification-Temporal cortex specimens (100 g) from a normal human brain were homogenized using lysis buffer (50 mM MES, 0.5 mM MgCl 2 , and 2 mM EGTA, pH 6.5, containing 1% Triton X-100, 1 mM sodium orthovanadate, 1 mM sodium pyrophosphate, and 20 mM NaF in the presence of EDTA-free protease inhibitors (Roche Applied Science)) and sedimented at 20,000 ϫ g for 1 h at 4°C. The supernatants were passed over a phosphorylated His-tau (200 g) column followed by a wash in phosphate-buffered saline. The supernatants then were fractionated and eluted with 1 M NaCl. Eluted proteins were separated by SDS-PAGE and stained with Coomassie Brilliant Blue (CBB). CBBstained gels were sliced and analyzed by mass spectrometry.
Protein Delivery-PHF tau, Hsp27, pHsp27, Hsp27 antibody, pHsp27 antibody, and control IgG were delivered into HCN2A cells using the BioPORTER protein delivery reagent according to the manufacturer's protocol. We used 1 mg of PHF tau, 2 mg of Hsp27, 2 mg of pHsp27, and 20 mg of Hsp27 or pHsp27 antibody per 6-well plate.
In Vitro Phosphorylation, Dephosphorylation, and Proteasomal Degradation-Active recombinant MAPKAP kinase 2, protein kinase A, or GSK-3␤ was incubated in a reaction mixture containing 2 mM ATP, 20 mM HEPES, 10 mM MgCl 2 , 10 mM MnCl 2 either with Hsp27, histone H3, or tau. Active recombinant PP2A was incubated in a reaction mixture containing 5 mM Tris-HCl, pH 7.0, 0.1 mM EDTA, 0.1% 2-mercaptoethanol, and 1 mg/ml bovine serum albumin with either PHF tau or histone H3 for 10 min at 22°C. 1 mg of PHF tau from AD brain or 1 mg of recombinant tau in 100 ml containing 50 mM Tris-HCl, pH 7.5, 0.01% Tween 20, and 10 mg of human erythrocyte was incubated with 26 S proteasome for 1 h at 37°C.
Immunopurification and Immunoblotting-Temporal cortex specimens (2 g) from human brain and HCN2A cells were homogenized using lysis buffer containing 1% Triton X-100 and sedimented at 20,000 ϫ g for 1 h at 4°C. Supernatants were incubated with protein A (Pierce) bound to polyclonal tau antibodies or polyclonal Hsp27 antibodies for 2 h at 4°C. 4ϫ SDS in a sample buffer was added to the lysates and immunoprecipitated. PAGE and immunoblots were performed as described (17). For confocal microscopy, the cells were washed with phosphate-buffered saline and fixed with 4% formaldehyde.
Pulldown Assay-His-tau or phosphorylated His-tau-coupled nickel beads or PHF tau or dephosphorylated PHF tau cross-linked to polyclonal tau antibody bound to protein A beads were added to either human brain homogenates, Hsp27, or pHsp27 and incubated at 4°C for 1 h with shaking. We added 0.2 M glycine buffer, pH 2.8, to release precipitated proteins from the beads. The eluted fraction was transferred to a new tube after centrifugation and 0.1 N Tris-HCl buffer, pH 8.5, was added.
Other Methods-Dead and live cells were distinguished by the LIVE/ DEAD Viability/Cytotoxicity Kit (Molecular Probes, Eugene, OR). Stained cells were analyzed with a laser-scanning confocal microscope system. Data are shown as dead cell counts from six randomly selected microscopic fields in each of three replicate cultures (n ϭ 18), and statistical significance was assessed using a one-way analysis of variance with Student-Neuman-Kuells post hoc tests between each group (*, p Ͻ 0.05). For detection of caspase-3 activity, cells were scraped off the dish, collected by centrifugation, distributed in aliquots of 10 6 cells, and assayed using a colorimetric kit (Clontech, Cambridge, UK) according to the manufacturer's protocol.

RESULTS
We sought to identify the molecular complexes associated with tau when it is phosphorylated at sites (Ser-202, Thr-205, Ser-396, Ser-404) that are phosphorylated in AD. Human brain homogenates were passed over an affinity column with nickel beads coupled to recombinant four-repeat His-tau phosphorylated by GSK-3␤, and associated proteins were eluted with 1 M NaCl. Eluted proteins were separated by SDS-PAGE, and several bands were stained with CBB. A 27-kDa CBB band was identified by mass spectrometry to contain the sequence VSLD-VNHFAPDELTVK, which exactly matched the sequence of Hsp27 (Fig. 1A). This 27-kDa molecule was immunoreactive for anti-Hsp27 antibody. To determine the specificity of the inter-action between phosphorylated tau and Hsp27 or pHsp27, we incubated nickel beads coupled to recombinant His-tau that was either GSK-3␤-phosphorylated or a control that was nonphosphorylated with human brain homogenates, purified recombinant Hsp27, or pHsp27 solution, which was phosphorylated by constitutively active MAPKAP kinase 2. The precipitated proteins were immunoblotted with Hsp27 and pHsp27 antibodies (Fig. 1B, left). Considerably larger amounts of Hsp27 and pHsp27 were precipitated by phosphorylated tau compared with dephosphorylated tau. No Hsp27 was present in precipitants with nickel beads alone. This result indicates that Hsp27 binds directly to phosphorylated tau.
Next we examined whether Hsp27 can bind to AD-type pathological hyperphosphorylated tau known as PHF tau. PHF tau was purified from AD brain, and some of it was dephosphorylated by PP2A for use as a negative control. PHF tau or dephosphorylated PHF tau was coupled to anti-tau antibodies conjugated to protein A beads. Human brain homogenates and Hsp27 or pHsp27 solution was added to the beads, and the precipitated proteins were immunoblotted with Hsp27 and pHsp27 antibodies. The amount of Hsp27 and pHsp27 in the precipitates of PHF tau was more than that of dephosphorylated PHF tau, indicating that phosphorylation of PHF tau is a recognition signal for both Hsp27 and pHsp27 (Fig. 1B, right).
To examine the interaction between Hsp27 and phosphorylated tau in human brain tissue, we carried out co-immunoprecipitations on homogenates of frozen temporal cortex from three AD brains and three normal brains. We used polyclonal tau or Hsp27 antibodies for immunoprecipitation followed by immunoblotting with AT8 (15), a monoclonal antibody that recognizes pathological tau phosphorylated at residues Ser-202 and Thr-205; PHF1 (13,14), a monoclonal antibody that recognizes pathological tau phosphorylated at residues Ser-396 and Ser-404; tau1 (15), a monoclonal antibody that recognizes residues Ser-199, Ser-202, and Thr-205 when non-phosphorylated; 5E2 (12), a monoclonal antibody that recognizes all isoforms of tau; and monoclonal antibodies against Hsp27 and pHsp27 (18). Multiple tau bands at 30 -60 kDa were recognized by 5E2 in all cases, whereas pathological hyperphosphorylated tau bands were recognized by AT8 (Fig. 1C) and PHF1 (data not shown) only in AD cases. Hsp27 co-precipitated with anti-tau antibody from homogenates of AD brain but not from normal brain. In all cases, Hsp27 was specifically precipitated with anti-Hsp27 antibody, whereas pre-absorbed Hsp27 antibody failed to do so (data not shown). Pathological hyperphosphorylated tau co-precipitated with Hsp27 from the homogenates of AD brain but not normal brain. These co-immunoprecipitations and reverse co-immunoprecipitations strongly suggest that Hsp27 preferentially binds to pathological hyperphosphorylated tau in human brain tissue.
To determine the role of Hsp27 in a complex with pathological hyperphosphorylated tau, the human cortical neuronal cell line HCN2A was used to create conditions that reproduced the formation of this complex. We delivered soluble PHF tau from AD brain, Hsp27, pHsp27, anti-Hsp27 antibody (goat IgG), anti-pHsp27 antibody (mouse IgG), and fluorescein isothiocyanate-IgG or TRITC-IgG as a positive control for delivery into HCN2A cells with the BioPORTER reagent (18). We conjugated TRITC to various proteins and used them to confirm successful delivery by confocal microscopy and then used either TRITCconjugated or non-TRITC-conjugated proteins for further studies. After 24 h, cell lysates were used for immunoprecipitation with Hsp27 or tau antibodies and immunoblotted ( Fig. 2A). In the absence of calyculin A and okadaic acid, only a small amount of AT8-positive pathological hyperphosphorylated tau was present, and therefore HCN2A cells were incubated with 5 mM calyculin A and 3 mM okadaic acid, PP1 and PP2A inhibitors, respectively, to inhibit dephosphorylation of pathologically phosphorylated tau after delivery to the intracellular compartment. As shown previously in rat brain, PHF tau was rapidly dephosphorylated (19). As expected, the amount of tau, Hsp27, or pHsp27 increased in those cell lysates to which the protein was delivered ( Fig. 2A). Consistent with the dephosphorylation of the PHF tau, TRITC-PHF tau distributed in a typical microtubule pattern, whereas calyculin A and okadaic acid treatment was distributed more diffusely (Fig. 2, B and C). Interestingly, delivery of Hsp27 and pHsp27 reduced the amount of pathological hyperphosphorylated tau and increased dephosphorylated tau (Fig. 2A). On the other hand, Hsp27 and pHsp27 antibody increased the amount of hyperphosphorylated tau. This suggests that Hsp27 and pHsp27 facilitated the degradation and/or dephosphorylation of pathological hyperphosphorylated tau.
To address these hypotheses, we delivered various combinations of pathological phosphorylated tau, Hsp27, and Hsp27 antibody into HCN2A cells and treated the cells with calyculin A, okadaic acid, and 5 M of the proteasome inhibitor, lactacystin, or Me 2 SO control. Pathological hyperphosphorylated tau was increased in the lactacystin-treated cells compared with Me 2 SO-treated cells (Fig. 2D). These data suggest that pathological phosphorylated tau may be degraded by the proteasome in HCN2A cells. To examine whether Hsp27 facilitates the degradation of PHF tau, we performed in vitro assays using a 26 S proteasome fraction. We incubated PHF tau or dephosphorylated tau with the 26 S proteasome fraction in the presence or absence of Hsp27 or lactacystin. In the presence of a FIG. 1. Hsp27 interacts with phosphorylated tau. A, human brain homogenates were passed over nickel beads coupled to His-tau that had been phosphorylated by GSK-3␤. Eluates were separated by SDS-PAGE and stained by CBB or immunoblotted with monoclonal Hsp27 antibody. The peptide sequence (bold) obtained from the 27-kDa band was consistent with the Hsp27 protein sequence (NCBI accession number AAH12768). B, the ability of Hsp27 to interact with phosphorylated tau was examined by a pulldown assay in which human brain homogenates and 100 ng of Hsp27 or 100 ng of pHsp27 were incubated with 20-ml nickel beads coupled to His-phosphorylated tau or His-tau, or protein A beads conjugated to polyclonal tau antibody (Sigma) coupled to PHF tau or dephosphorylated PHF tau. Precipitated proteins were analyzed by Hsp27 or pHsp27 antibody. The immunoblot shows that Hsp27 mainly interacted with phosphorylated tau and PHF tau. C, brain homogenates of temporal cortex from three AD and three normal control cases were immunoprecipitated with Hsp27 or polyclonal tau antibodies. The precipitates were analyzed by 5E2, AT8, Hsp27, or pHsp27 antibodies. WB, Western blot; IP, immunoprecipitate. proteasomal fraction, PHF tau was further reduced by Hsp27, indicating that Hsp27 facilitates the degradation of pathological hyperphosphorylated tau. On the other hand, Hsp27 facilitates the degradation of only a small amount of dephosphorylated tau (Fig. 2E). As expected, lactacystin inhibited the degradation of both PHF tau and dephosphorylated tau.
Hsp27 also facilitated the dephosphorylation of PHF tau in addition to its degradation. To demonstrate this result, we incubated PHF tau or protein kinase A-phosphorylated histone H3 with PP2A in the presence or absence of Hsp27. Hsp27 increased the dephosphorylation of PHF tau but had no effect on the phosphorylation state of histone H3 (Fig. 2F). Therefore, Hsp27 facilitated the dephosphorylation specifically of PHF tau without causing a general increase in PP2A activity.
Hsp27 appeared to facilitate the degradation and dephosphorylation of PHF tau by altering the conformation of tau. To show conformational effects, PHF tau was incubated with Hsp27 and subsequently immunoprecipitated with Alz50, FIG. 2. Hsp27 decreases the amount of pathological hyperphosphorylated tau. A, PHF tau, Hsp27, pHsp27, Hsp27 antibody, or pHsp27 antibody was delivered into HCN2A cells. After delivery, cells were incubated with 5 mM calyculin A and 3 mM okadaic acid or with Me 2 SO. Cell lysates were immunoprecipitated with tau antibody or Hsp27 antibody. The precipitates and cell lysates were analyzed with the indicated antibodies. B and C, confocal microscopy of TRITC-PHF tau in the cells treated with or without calyculin A and okadaic acids. D, PHF tau, Hsp27, or Hsp27 antibody was delivered into HCN2A cells that were treated with calyculin A and okadaic acid in the presence or absence of 10 mM lactacystin. Cell lysates were analyzed by AT8. Lactacystin treatment inhibited the proteasome-mediated degradation of pathological hyperphosphorylated tau. E, PHF tau or recombinant His-tau was incubated with the 26 S proteasomal fraction in the presence or absence of Hsp27, Hsp27 antibody, or lactacystin. Tau was detected by AT8 or Tau 1. Hsp27 facilitated the degradation of PHF tau compared with tau. F, PHF tau or phosphorylated histone H3 was dephosphorylated by PP2A with or without Hsp27. PHF tau dephosphorylation was detected by AT8 and tau 1. Histone H3 dephosphorylation was detected by histone H3 and pHistone H3 antibody. Hsp27 facilitated the dephosphorylation of PHF tau but not histone H3. G, PHF tau was incubated with or without Hsp27 and immunoprecipitated with Alz50. The precipitates were separated with SDS-PAGE, and a gel was stained with CBB. WB, Western blot; IP, immunoprecipitate; Ab, antibody.
which recognizes a conformational epitope specific for PHF tau (20,21). Hsp27 treatment decreased the amount of Alz50precipitated PHF tau (Fig. 2G), suggesting that PHF tau undergoes a change in its conformation when associated with Hsp27.
Finally, we sought to determine the effects of Hsp27 on PHF tau-induced cell death. Various combinations of PHF tau, Hsp27, or anti-Hsp27 antibody were delivered to HCN2A cells, some of which also were treated with lactacystin or calyculin A and okadaic acid. The cell lysates were prepared, and the amount of pathological hyperphosphorylated tau and Hsp27 was analyzed with AT8 and anti-Hsp27 antibody (Fig. 3A). Dead cells were counted using a LIVE/DEAD Viability/Cytotoxicity Kit (Fig. 3B). The amount of pathological hyperphosphorylated tau paralleled the cell death as indicated by calyculin A-, okadaic acid-, and lactacystin-induced increases in the amount of hyperphosphorylated tau along with a corresponding increase in cell death (p Ͻ 0.05) when [1 Ϫ 10] was compared with [2 Ϫ 11] or [3 Ϫ 12] (Fig. 3B). This observation suggested that pathological hyperphosphorylated tau enhances cell death. Interestingly, Hsp27 decreased cell death and the amount of hyperphosphorylated tau (p Ͻ 0.05) when [2 Ϫ 11] was compared with [5 Ϫ 14], whereas anti-Hsp27 antibody increased both (p Ͻ 0.05) when [2 Ϫ 11] was compared with [8 Ϫ 17]. Taken together, pathological hyperphosphorylated tau caused cell death, and Hsp27 attenuated the cell toxicity of pathological hyperphosphorylated tau.
Several studies have shown that hyperphosphorylated tau can cause apoptosis, and apoptotic cell death has been observed in postmortem AD brain and AD animal models (22)(23)(24). To examine whether pathological hyperphosphorylated tau causes apoptosis, activation of an effector caspase (caspase-3) was analyzed in a colorimetric assay. The delivery of pathological phosphorylated tau significantly increased caspase-3 activity compared with TRITC-IgG-delivered cells (Fig. 3B). Calyculin A and okadaic acid or lactacystin treatment also increased caspase-3 activity; however, the presence of PHF tau produced a further increase (p Ͻ 0.05) when [1 Ϫ 10] was compared with [2 Ϫ 11] or [3 Ϫ 12] (Fig. 3B). Delivery of Hsp27 decreased the caspase-3 activity, which was up-regulated by pathological hyperphosphorylated tau (compare [2 Ϫ 11] to [5 Ϫ 14]). Delivery of anti-Hsp27 antibody increased the activity (compare [2 Ϫ 11] to [8 Ϫ 17]). Thus, one of the mechanisms of PHF tau-induced cell death is apoptosis, and Hsp27 prevents it. DISCUSSION Using a cellular model of tauopathy and AD postmortem brain tissue, we found that Hsp27 directly bound to patholog- FIG. 3. Hsp27 attenuates the toxicity of PHF tau. PHF tau, Hsp27, or Hsp27 antibody was delivered into HCN2A cells. The cells were treated with or without calyculin A, okadaic acids, and lactacystin after the delivery. A, cell lysates were analyzed by AT8 or Hsp27 antibody. B, blue bars, quantification of dead cells in HCN2A cells; red bars, caspase-3 activity in HCN2A cells. To eliminate any possible complicating effect of calyculin A, okadaic acid, lactacystin, Hsp27, or anti-Hsp27 antibody that might occur independently of PHF tau, we subtracted the number of dead cells in the absence of PHF tau from the number in the presence of PHF tau, e.g. subtract the number of set 10 from set 1. Likewise, caspase-3 activity in the absence of PHF tau was subtracted from the value in the presence of PHF tau, e.g. subtract the number of set 10 from set 1. Data are presented as mean S.D. *, a significant difference with p Ͻ 0.05.WB, Western blot; Ab, antibody. ical hyperphosphorylated tau but not dephosphorylated tau. Hsp27 also protected against cell death induced by pathological hyperphosphorylated tau. Among the likely mechanisms to explain this effect is the ability of Hsp27 to facilitate the degradation and/or dephosphorylation of pathological hyperphosphorylated tau by altering its conformation. We also observed that pathological hyperphosphorylated tau induced apoptosis, and Hsp27 prevented it. The mechanism by which Hsp27 prevents apoptosis may be caused not only by the interaction between Hsp27 and pathological hyperphosphorylated tau but also by Hsp27-mediated inhibition of pro-caspase-9 and caspase-3 (25).
Interestingly, Hsp27 facilitated the degradation of hyperphosphorylated tau without ubiquitination. Notably, treatment with the proteasome inhibitor lactacystin did not induce the accumulation of ubiquitinated tau in the presence of Hsp27. Ubiquitin-independent degradation of hyperphosphorylated tau may be another of the many parallels with ␣-synuclein (26). Both proteins are natively unfolded and therefore could bypass the need for ubiquitination and unfolding to enter the 20 S proteasome. Denatured, non-ubiquitinated proteins can be degraded by the proteasome (27). Other proteins such as p21 WAF1/CIP1 (28) and ornithine decarboxylase (29) can be degraded in a ubiquitin-independent manner. Phosphorylated tau may have at least two mechanisms leading to its degradation: a ubiquitin-independent pathway that can be facilitated by Hsp27 and a ubiquitin-dependent pathway mediated by the carboxyl terminus of the Hsc70 interacting protein-Hsc70 complex (30).
The known cytoprotective effects of Hsp27 include its role as a molecular chaperone, the inhibition of caspase activation, the prevention of stress-induced disruption of the cytoskeleton, and the modulation of the intracellular redox potential (25). The induction and expression of Hsp27 have been studied widely in the mammalian central nervous system (31). Hsp27 is highly inducible in cortical astrocytes after seizure activity (32) and ischemic injury (33). Hsp27 also is present normally in many neurons of the rat and mouse brainstem and spinal cord (34). Hsp27 is induced and expressed at very high levels for extended periods of time in vagal trans-section (35). These results suggest that Hsp27 plays a critical role in neuronal metabolism and neuronal survival. Hsp27 also prevents neuronal cell death in simulated ischemia (36), peripheral nerve trans-section (37,38), and polyglutamine repeat expansion of huntingtin (39). David et al. (40) showed that unfolded native tau protein is degraded by the proteasome, but they did not test pathological hyperphosphorylated tau. In our study, Hsp27 facilitated the degradation of pathological hyperphosphorylated tau, which is otherwise not efficiently degraded by the proteasome in the absence of Hsp27. Therefore, Hsp27 activation may serve as the basis for a new therapeutic approach to AD and other tauopathies.