Mutant (R406W) Human Tau Is Hyperphosphorylated and Does Not Efficiently Bind Microtubules in a Neuronal Cortical Cell Model*

Frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17) is an autosomal dominant neurodegenerative disorder caused by mutations in the gene that encodes for tau, a microtubule-binding protein. Neuropathologically the disease is characterized by extensive neuronal loss in the frontal and temporal lobes and the filamentous accumulation of hyperphosphorylated tau. The R406W missense mutation was originally described in an American and a Dutch family. Although R406W tau is hyperphosphorylated in FTDP-17 cases, R406W tau expressed in cell model systems has not shown increased phosphorylation. The purpose of this study was to establish a neuronal model system in which the phosphorylation of R406W tau is increased and thus more representative of the in vivo situation. To accomplish this goal immortalized mouse cortical cells that express low levels of endogenous tau were stably transfected with human wild type or R406W tau. In this neuronal model R406W tau was more highly phosphorylated at numerous epitopes and showed decreased microtubule binding compared with wild type tau, an effect that could be reversed by dephosphorylation. In addition the expression of R406W tau in the cortical cells resulted in increased cell death as compared with wild type tau-expressing cells when the cells were exposed to an apoptotic stressor. These results indicate that in an appropriate cellular context R406W tau is hyperphosphorylated, which leads to decreased microtubule binding. Furthermore, expression of R406W tau sensitized cells to apoptotic stress, which may contribute to the neuronal cell loss that occurs in this FTDP-17 tauopathy.

Frontotemporal dementia (FTD) 1 refers to a group of neurological disorders that are characterized clinically by progressive behavioral changes and neuropathologically by neuronal loss in the frontal and temporal lobes and the presence of filamentous deposits of abnormally hyperphosphorylated tau protein in neurons and/or glial cells (1)(2)(3)(4). FTDs occur mainly as sporadic cases but also as familial forms with an autosomal dominant mode of inheritance and age-dependent penetrance (1). In 1998 it was demonstrated that several familial FTDs were due to mutations in the tau gene and have been referred to as frontotemporal dementia and Parkinsonism linked to chromosome 17 (FTDP-17) (5-7). Since these initial discoveries, additional tau mutations have been described in frontotemporal dementia families, and presently over 25 mutations in the tau gene have been identified (8,9). They include missense mutations in coding regions, amino acid deletions, and intronic mutations in the region following exon 10. The mutations may be divided into two main groups, those that affect alternative splicing of exon 10, leading to changes in the ratio of tau mRNAs with or without exon 10, and thus the proportion of tau with three microtubule binding repeats versus tau with four microtubule binding repeats, and missense mutations.
Tau is a microtubule-associated protein that facilitates microtubule assembly and stabilization (for review see Ref. 2). Alternative splicing of the tau gene generates the six tau isoforms that occur in adult human brain (10,11). The tau isoforms differ by the presence of three or four carboxyl-terminal repeats due to the splicing in or out of exon 10, which are involved in microtubule binding, and zero, one, or two aminoterminal inserts whose functions are still unknown (11,12). The function of tau is negatively regulated by site-specific phosphorylation (13,14) and in FTDP-17 brain tau is hyperphosphorylated and non-functional (3,15).
In vitro studies have shown that many of the FTDP-17 missense mutations likely impair the ability of tau to bind to microtubules (16 -19). The mutations can also impair the ability of tau to stimulate the formation of microtubules by increasing the lag time and reducing the rate of polymerization (16,17,19). However, these data are not unequivocal because identical in vitro assays have given differing results (16,19) and there is some discussion as to the extent to which the FTDP-17 missense mutations affect tau-microtubule interactions (18,20). The effects of the FTDP-17 mutations on the ability of tau to interact with microtubules and affect microtubule function in situ have also been examined with variable outcomes being reported (21)(22)(23). It should be noted that the majority of these previous studies were carried out using either cell-free systems or non-neuronal cells that had been transiently transfected with wild type or FTDP-17 mutant tau constructs (16, 19 -24).
The nature of the neuronal loss seen in FTDP-17 cases is still unknown. Furukawa and co-workers (25) showed that the N279K and V337M tau mutations sensitized transfected SH-SY5Y cells to apoptotic death compared with wild type tauexpressing cells. Furthermore, using a microarray technique to examine gene expression in a P301L tau mouse model of FTDP-17, it was reported that there was decreased expression of anti-apoptotic genes and altered expression of trafficking proteins (26), suggesting that apoptosis is occurring. This P301L tau mouse model showed evidence of gliosis and neuronal loss primarily in the brain stem and spinal cord (27). Furthermore, another P301L tau transgenic mouse expressing the longest tau isoform also showed evidence of astrocytosis and neuronal apoptosis (28). In contrast to these findings a transgenic mouse model of the P301S tau mutation did not show evidence of apoptotic death (29). Thus, the processes that contribute to the neuronal cell death in FTDP-17 cases have yet to be elucidated.
The FTDP-17 R406W tau mutation has been described in an American, a Dutch, and a Japanese family (30 -32). Neuropathological examination showed neuronal loss in the frontal and temporal lobes and the presence of neurofibrillary tangles and neuropil threads throughout the brains (30 -32). Interestingly, it was found that the morphological and biochemical characteristics of the tau filaments with the R406W mutation appeared indistinguishable from those found in Alzheimer's disease brains (19,31,33). In one report tau with the R406W mutation did not bind microtubules or promote microtubule assembly as effectively as wild type tau or the other FTDP-17 mutant taus that were examined (19). However in two other studies R406W mutant tau was found to be more efficient at promoting microtubule assembly than the other FTDP-17 tau mutants (16,18). Therefore the effects of the R406W tau mutation on microtubule binding and function remain to be fully established. In addition, stable or transient transfection of R406W tau into non-neural cell lines has resulted in R406W tau being less phosphorylated than the wild type tau (23,34), even though the tau from R406W FTDP-17 cases is hyperphosphorylated (30,31).
In this study we have used a novel model system to elucidate the effects of the FTDP-17 R406W tau mutation on tau phosphorylation and function. Because cortical neurons are the most vulnerable in FTDP-17 cases (1, 3), immortalized mouse cortical neurons (CN1.4) (35) were stably transfected with wild type or mutant R406W tau containing four microtubule repeats. Using this cell model the effects of wild type and R406W mutant tau on cell morphology were examined, and the effects of the R406W mutation on tau localization and interactions with the cytoskeleton and microtubules were evaluated further. Previous cell models in which R406W mutant tau was expressed (20,22,23) were unable to reproduce the increased tau phosphorylation that occurs in FTDP-17 brains with this mutation (30,31). However, in this study we show that in stably transfected cortical cells, mutant R406W tau is more highly phosphorylated than wild type tau, and therefore this cell model provides an appropriate system to examine the effects of this mutation on tau function. These studies also clearly demonstrate that in situ, R406W tau does not efficiently bind microtubules, in part due to the increased phosphorylation. Furthermore, R406W tau-expressing cortical cells were more vulnerable to an apoptotic stimuli compared with cells expressing wild type tau. Overall, our results indicate that when mutant R406W tau is expressed in an appropriate cellular context it is hyperphosphorylated and exhibits decreased microtubule binding. Furthermore, in this model, cells that express mutant R406W tau show increased sensitivity to apoptotic stress, processes which may contribute to the neuronal cell loss that occurs in FTDP-17.
Creation of Wild Type and Mutant tau Constructs-Mutant and wild type human tau constructs were a generous gift from Dr. M. Hutton, Mayo Clinic, FL and were subcloned into the inducible expression vector pBIG2i (36) at the unique restriction enzyme sites BamHI and NotI. Using a PCR-based approach, wild type (WT) and R406W mutant tau DNA with BamHI and NotI restriction sites was amplified using forward and reverse primers (forward: 5Ј-GAT CGC GGA TCC GAT GGC TGA GCC CCG CCA GGA GTT CG-3Ј; reverse: 5Ј-CAT ATA AAT TGC GGC CGC TCA CAA ACC CTG CTT GGC CAG GGA GGC-3Ј) under the following conditions: pre-denaturation at 94°C for 3 min followed by 30 cycles of denaturation at 94°C for 1 min, annealing at 70°C for 1 min, elongation at 72°C for 40 s, and amplification at 72°C for 10 min. The PCR products were purified and digested with BamHI and NotI and cloned into the same sites of the inducible expression vector pBig2i resulting in pBIG2i/WT and pBIG2i/R406W. The integrity of the wild type and mutant R406W tau constructs was confirmed by sequence analysis (UAB Sequencing Facility).
Stable Transfections-Plasmids pBIG2i/WT or pBIG2i/R406W were introduced into low passage CN1.4 cells using Effectene® reagent (Qiagen) as per the manufacturer's instructions. Cells were maintained in hygromycin B (1 mg/ml) selection media for 21 days before individual surviving colonies were selected and clonally expanded. Selected clones were screened for their expression levels of wild type or mutant tau by incubating cells in media containing doxycycline (2 g/ml) and immunoblotting.
Immunocytochemistry-CN1.4 cortical cells and stably transfected wild type tau and mutant R406W tau-expressing CN1.4 cells were replated in doxycycline containing media on poly-D-lysine (Sigma)coated coverslips in 6-well plates. Cells were processed using a combined fixation and extraction protocol as previously described (44,45). Briefly, cells were rinsed in pre-warmed extraction buffer minus Triton X-100, then incubated with extraction buffer (80 mM Pipes (pH 6.8), 2 mM GTP, 1 mM MgCl 2 , 1 mM EGTA, 0.5% Triton X-100, 30% glycerol) at 37°C for 30 -60 s. Extraction buffer was then removed, and cells were washed with extraction buffer minus Triton X-100. Cells were then fixed and permeabilized by incubating in extraction buffer containing 0.3% glutaraldehyde for 20 min at room temperature. Cells were then incubated in NaBH 4 (10 mg/ml in PBS) for 7 min, followed by incubation in 0.1 M glycine in PBS for 20 min. The cells were blocked with 4% BSA in PBS for 30 min, rinsed in PBS, and then incubated with a tau polyclonal antibody (Dako Corp.) and a monoclonal ␤ tubulin antibody 5H1 (IgM, a generous gift from Dr. L. Binder), diluted in 4% BSA for 30 min at room temperature. Cells were then rinsed with PBS and incubated with the appropriate secondary antibodies (fluorescein isothiocyanate-conjugated donkey anti rabbit IgG and Texas Red-conjugated donkey anti mouse IgM (Jackson ImmunoResearch Laboratories, Bar Harbor, ME)), also diluted in 4% BSA for 30 min at room temperature. Cells were rinsed in PBS and water before the coverslips were mounted. Cells were visualized with a Nikon Diaphot 200 epifluorescence microscope, and images were captured with a Digital spot camera (Diagnostic Instruments), digitally stored, and displayed using the accompanying software.
Osmotic Stress Protocol-CN1.4 cells and stably transfected wild type and R406W tau-expressing cortical cells were treated with 0.5 M sorbitol (Sigma) in serum-free Dulbecco's modified Eagle's medium (Irvine Scientific) for up to 4 h as described previously (46). Cells were collected at the indicated time points and used for immunoblotting or the caspase 3 assay.
Caspase 3 Assay-At the indicated times post sorbitol treatment, cells were collected and used to measure caspase 3 activity as previously described (47).
In Vitro Microtubule Binding Assay-The microtubule binding assay was carried out as previously described (48). Unbound (supernatant) and microtubule-bound (pellet) fractions were electrophoresed on a 10% SDS-polyacrylamide gel, transferred to nitrocellulose membrane, and probed with the total tau antibodies Tau5/5A6.
Lambda Protein Phosphatase Treatment and in Vitro Microtubule Binding Assay-50 g of whole cell lysates (wild type or R406W mutant tau) in a total volume of 100 l of modified binding buffer (80 mM Pipes (pH 6.8), 2 mM MnCl 2 , 1 mM EGTA, protease inhibitor mixture, and Ϯ0.1 M okadaic acid) were incubated in the absence or presence of 500 units of Lambda Protein Phosphatase (New England Biolabs) at 30°C for 1 h. Following incubation, microtubules were added to 30 l of the reaction mix (equivalent to 15 g of whole cell lysate), and the microtubule binding assay was carried out as described above. Microtubulebound (pellet) fractions were electrophoresed on an 8% SDS-polyacrylamide gel, transferred to nitrocellulose membrane, and probed with the total tau antibodies Tau5/5A6. Cellular Fractionation-Cells were separated into soluble and cytoskeletal insoluble fractions as described previously (49) with a few modifications. Cells were rinsed with pre-warmed PBS and collected in 100 l of PBS. Cells were spun, and the resulting pellet was resuspended in 40 l of pre-warmed extraction buffer (80 mM Pipes (pH 6.8), 1 mM MgCl 2 , 2 mM EGTA, 0.1 mM EDTA, 0.1% Triton X-100, 30% glycerol) containing a protease inhibitor mixture (Sigma) and a phosphatase inhibitor (0.5 M okadaic acid). Lysate was incubated at 37°C for 8 min prior to centrifugation (15 min at 15,000 ϫ g and 25°C). Equal amounts of pellet and supernatant fraction were electrophoresed and processed for immunodetection with anti-tau antibodies as described above.
Statistics-All data were analyzed using analysis of variance. The values were considered significantly different when p Ͻ 0.05. Results were expressed as mean Ϯ S.E.  (Fig. 1).

Cell Morphology of Wild Type and R406W
tau Expression and Phosphorylation State in Stably Transfected CN1.4 Cortical Cells-We next measured the expression levels of wild type and R406W mutant tau in the stably transfected CN1.4 cortical cells. CN1.4 cortical cells normally ex-press very low levels of tau (50), so cell lines overexpressing wild type or mutant R406W tau could be readily detected. Both the wild type and R406W mutant tau cell lines expressed almost equivalent amounts of total tau, and tau expression in untransfected CN1.4 cortical cells could not be readily detected (Fig. 2). The R406W mutant tau expressed in the CN1.4 cortical cells was more highly phosphorylated at both the PHF-1 (tau phosphorylated at Ser-396/404) and Tau-1 sites (Tau-1 recognizes tau only when Ser-195, Ser-198, Ser-199, Ser-202, and Thr-205 are not phosphorylated) compared with wild type tau (Fig. 2).
Cytoskeletal Association of Wild Type and R406W Mutant tau in Stably Transfected CN1.4 Cortical Cells-Since the mutant R406W tau was in an increased phosphorylation state, the extent of interaction of wild type and mutant R406W tau with the neural cytoskeleton was assessed. Cell lysates were separated into soluble and insoluble cytoskeletal fractions and immunoblotted. Wild type tau was distributed evenly between the soluble supernatant and cytoskeletal insoluble pellet fractions, but mutant R406W tau was localized almost exclusively in the soluble fraction (Fig. 3). Additionally, mutant R406W tau in the soluble fraction was more highly phosphorylated at the Tau-1, PHF-1, and AT180 sites (tau phosphorylated at Thr-231) compared with wild type tau (Fig. 3), indicating that mutant R406W tau is more phosphorylated and less associated with the neuronal cytoskeleton. It is interesting to note that AT180 immunoreactivity was only observed in the soluble fraction from the R406W mutant tau-expressing cells given that phosphorylation of Thr-231 likely plays a key role in decreasing the association of tau with microtubules (48).
Mutant R406W tau Does Not Efficiently Bind Microtubules-To determine the microtubule binding capacity of FTDP-17 mutant R406W tau, a microtubule binding assay was carried out. High speed supernatants from wild type and R406W mutant tau-expressing CN1.4 cells were incubated with taxol-stabilized microtubules, and the amount of tau bound to the microtubules in the pellet and the amount of tau that remained unbound were measured. Wild type tau was found primarily in the microtubule-bound pellet, but R406W mutant tau was enriched in the unbound supernatant (Fig.  4B). Total tau expression levels were also evaluated in cell lysates prepared from wild type and mutant R406W stably transfected CN1.4 cortical cells prior to use in the microtubulebinding assay and found to be equivalent (Fig. 4A). These results indicate that mutant R406W tau does not bind microtubules as efficiently as wild type tau.
Dephosphorylation of R406W Mutant tau Restores Its Ability to Bind Microtubules Efficiently-Because the R406W mutant tau was in an increased phosphorylation state and did not bind microtubules as efficiently as wild type tau, we wanted to determine if the microtubule-binding ability of R406W mutant tau could be restored by dephosphorylation. Treatment with lambda protein phosphatase resulted in a dephosphorylation of wild type and R406W mutant tau as indicated by the increased electrophoretic mobility shift of the tau-immunoreactive bands (Fig. 5A). Additionally, dephosphorylation of R406W mutant tau resulted in a change in presentation of the tau band from a doublet to a single immunoreactive band (Fig. 5A). Dephosphorylation of R406W mutant tau with lambda protein phosphatase almost fully restored the microtubule-binding capacity of R406W mutant tau, because dephosphorylated wild type and R406W mutant tau showed equivalent microtubule binding (Fig. 5B). Densitometric analysis indicated that for equivalent amounts of input tau there was a 2-fold increase in the amount of microtubule-bound R406W mutant tau following dephosphorylation with lambda protein phosphatase, whereas the amount of wild type tau that was bound in the absence or presence of lambda protein phosphatase treatment remained the same. These results indicate that dephosphorylation of R406W mutant tau greatly increased its ability to efficiently bind microtubules.

Cellular Localization of Wild Type and Mutant R406W
Tau-To determine the intracellular localization of wild type and R406W mutant tau, cells were grown on poly-D-lysinecoated coverslips and examined immunocytochemically. Untransfected CN1.4 cells had undetectable levels of tau, but the cytoplasmic microtubule network was evident after immunostaining with an anti-tubulin antibody (red in Fig. 6). The immunostaining with the anti-tau antibody was intense in the wild type tau-expressing CN1.4 cortical cells (green in Fig. 6). Tau was present throughout the cell body and extended into the cell processes (neurites). Co-localization of tau and tubulin was readily observed (yellow in Fig. 6). CN1.4 cortical cells expressing mutant R406W tau exhibited an altered tau distribution compared with wild type tau-expressing cells. The R406W tau-expressing cells had a more stellate appearance  5. Restoration of the microtubule binding capacity of R406W mutant tau by dephosphorylation. A, total tau expression levels (Input tau) were evaluated in cell lysates prepared from wild type (WT) and R406W stably transfected CN1.4 cortical cells incubated in the absence (Ϫ) or presence (ϩ) of lambda protein phosphatase (PPase) prior to use in the microtubule-binding assay. Actin is shown as a loading control. B, following incubation in the absence (Ϫ) or presence (ϩ) of PPase, wild type, and R406W mutant tau-expressing cortical cell lysates were incubated with taxol-stabilized microtubules. Microtubule-bound tau was determined by immunoblot analysis using the phospho-independent and total tau antibodies Tau5/5A6. Dephosphorylation of R406W mutant tau restores its ability to bind microtubules to almost the levels observed for WT tau.
with short projections, and tau immunostaining was predominantly perinuclear (green in Fig. 6). Furthermore, compared with wild type tau, mutant R406W tau did not show a total colocalization with the tubulin network (merged images in Fig. 6).
Increased Caspase 3 Activity in R406W Mutant tau-expressing CN1.4 Cortical Cells-To determine if R406W mutant tau could sensitize CN1.4 cortical cells to an apoptotic stimuli, untransfected CN1.4 cortical cells and cells stably transfected with wild type or R406W mutant tau were treated with 0.5 M sorbitol for up to 4 h (46). Cells were collected and used to measure caspase 3 activity as an indicator of apoptosis (51). Caspase 3 activity was significantly greater in R406W mutant tau-expressing CN1.4 cortical cells at the 4-h point compared with wild type tau-expressing CN1.4 cortical cells (Fig. 7). This result indicates that the expression of R406W mutant tau sensitizes the CN1.4 cortical cells to apoptotic stimuli.
Wild Type and R406W Mutant tau Phosphorylation in Response to Osmotic Stress-To examine alterations in the phosphorylation state of tau during apoptosis, lysates from wild type and R406W mutant tau-expressing CN1.4 cortical cells incubated in the presence of 0.5 M sorbitol for 0, 0.5, 2, and 4 h were immunoblotted with phospho-specific tau antibodies. Tau phosphorylation at the PHF-1 (tau phosphorylated at Ser-396/ 404) and Tau-1 sites (Tau-1 recognizes tau only when Ser-195, Ser-198, Ser-199, Ser-202, and Thr-205 are not phosphorylated) was increased by 30 min of sorbitol treatment and remained elevated for up to 2 h in wild type tau-expressing CN1.4 cortical cells. In R406W mutant tau-expressing CN1.4 cortical cells there is an increase in phosphorylation at the PHF-1 and Tau-1 epitopes within 30 min of sorbitol treatment then a decreased phosphorylation within 2 h (Fig. 8). The quantitated data is shown in the bar graphs below each panel of blots (Fig.  8). In addition, there was evidence of tau degradation in the R406W mutant tau-expressing CN1.4 cortical cells following osmotic stress treatment (Fig. 8). DISCUSSION The purpose of this study was to establish a cellular model system in which R06W mutant tau exhibited characteristics of tau with the R406W mutation in the human disease state. The results presented in this study clearly demonstrate that when R406W mutant tau is stably expressed in mouse cortical cells it is hyperphosphorylated and thus reproduces a key biochemical property of R406W mutant tau from FTDP-17 cases (30,31), unlike previous cell models of the R406W FTDP-17 mutation (22,23,34). Furthermore, the microtubule-binding capacity of R406W mutant tau expressed in cortical cells is impaired, but this loss of microtubule binding was reversed or "rescued" by dephosphorylating the tau. Additionally, we showed that

cortical cells and cells stably
transfected with wild type (WT) or R406W mutant tau were processed using a combined fixation and extraction protocol and immunostained with the phospho-independent total tau polyclonal antibody (Dako tau) (green) and ␤ tubulin antibody (red). CN1.4 cortical cells express very low levels (undetectable) of endogenous tau. WT tau-expressing CN1.4 cortical cells have a diffuse tau distribution extending into the cell processes, and there is total co-localization with ␤ tubulin (yellow). R406W mutant tau has a mainly perinuclear distribution and did not entirely co-localize with ␤ tubulin.
R406W mutant tau exerts a cytotoxic effect in cortical cells when the cells were exposed to an apoptotic stimulus.
Tau is known to be involved in process formation and neurite extension. Overexpression of tau in non-neuronal cells induces the formation of long neurite-like processes (52,53), whereas inhibiting tau expression using antisense tau resulted in the loss of neurite outgrowth and extension (54). In mouse cortical cells expression of wild type tau resulted in the formation of long processes/neurites, unlike the R406W mutant tau-expressing cells, which were much shorter, stubby projections. Thus, our results would indicate that R406W mutant tau is less effective in stimulating neurite outgrowth in the cortical cells.
A previous study showed that expression of R406W mutant tau in COS cells resulted in fewer processes than those observed with wild type tau. It was suggested that, although R406W tau could still bind tubulin, as evidenced by co-localization of R406W mutant tau and tubulin, R406W mutant tau was less efficient at inducing microtubule assembly (24). In contrast to these findings in COS cells, significantly less co-localization of R406W tau and tubulin was observed in the cortical cells, compared with wild type tau. However, it should be noted that in the previous study, R406W tau was in a more dephosphorylated state, and hence the tubulin binding capacity may not have been as affected (24). Additionally, in the cortical cells the R406W mutant tau showed a predominantly perinuclear localization, and a previous report showed a similar distribution for V337M mutant tau when it was expressed in COS cells (21). Interestingly, V337M mutant tau has also been reported to be less phosphorylated than wild type tau when expressed in neuroglioma cells, except for the PHF-1 site that was more phosphorylated (55).
Previously reported cell models of the R406W mutation have not been successful at reproducing the increased phosphorylation observed in vivo (22,23,34,55). A possible explanation for this could simply be that it is due to using non-neuronal cells such as Chinese hamster ovary and COS-1 cells (22)(23)(24)34). However, R406W mutant tau also was in a lower phosphorylation state than wild type tau when expressed in more neurallike cell systems such as neuroglioma cells (55) and neuroblastoma cells (20). Interestingly, the one reported R406W mutant transgenic mouse model did show evidence of hyperphosphorylated tau inclusions in forebrain neurons (56). Therefore, it can be hypothesized that immortalized mouse cortical cells provide the appropriate cellular context for increases in R406W mutant tau phosphorylation to occur and thus model this feature of FTDP-17.
It has been proposed that the FTDP-17 missense mutations that affect tau-microtubule interactions, such as the R406W mutation, exert a pathogenic effect by a partial loss of function i.e. a reduced ability to bind microtubules (57). In vitro studies have clearly demonstrated that the tau mutations reduce, but do not abolish, the ability of tau to promote microtubule assembly (16,18,19), and we have confirmed this finding in situ in our R406W mutant tau CN1.4 cortical cell model. Additionally, we have shown that if R406W mutant tau is dephosphorylated, microtubule binding is increased to approximately wild type tau levels. Given the reported subtle decreases in the effectiveness of recombinant R406W mutant tau to bind microtubules and promote microtubule assembly in vitro (18), it is not surprising that dephosphorylating R406W mutant tau increases its ability to bind microtubules. These findings would imply that the primary reason that mutant R406W tau does not efficiently bind microtu- Wild type (WT) and R406W mutant tau-expressing CN1.4 cortical cells were incubated with 0.5 M sorbitol for the times indicated, and samples were collected and immunoblotted with phospho-independent Tau5/5A6, phospho-dependent PHF-1, which recognizes tau phosphorylated at Ser-396/404, and phospho-dependent Tau-1, which recognizes dephosphorylated epitopes Ser-195, Ser-198, Ser-199, Ser-202, and Thr-205. Phosphorylation at the PHF-1 and Tau-1 epitopes was increased by 30 min of sorbitol treatment and remained elevated for up to 2 h in WT tau-expressing CN1.4 cortical cells. In R406W mutant tau-expressing CN1.4 cortical cells there is an increase in phosphorylation at the PHF-1 and Tau-1 epitopes within 30 min of sorbitol treatment then decreased phosphorylation by 2 h. Actin is shown as a loading control. Bar graphs below each panel of blots show the PHF-1 (black bars) and Tau-1 (white bars) relative optical density normalized to total tau (Tau5/5A6) levels.
bules is because it is hyperphosphorylated.
The mechanism leading to the increased phosphorylation state of R406W mutant tau is unknown. One possibility is that the mutation in tau results in a conformational change that increases its association with protein kinases or decreases protein phosphatase binding. Because glycogen synthase kinase 3␤ (GSK3␤) is a predominant tau kinase (58), we examined the interaction of R406W mutant tau and GSK3␤. But co-immunoprecipitation studies did not reveal any differences in the interaction of wild type or R406W mutant tau with GSK3␤ (data not shown), although increased phosphorylation at Thr-231, a primed GSK3␤ site that negatively impacts microtubule binding (48,59), was observed for R406W tau. Goedert et al. (60) reported that FTDP-17 mutant tau, including the R406W mutant, showed a decreased ability to bind protein phosphatase 2A (PP2A), a major tau phosphatase. So the increased phosphorylation state of R406W mutant tau in the cortical cell model could be due to a decreased association with PP2A. Indeed, it is clear that PP2A can efficiently dephosphorylate sites that are phosphorylated by GSK3␤ on other proteins (61).
The exact nature of the cell death observed in FTDP-17 tauopathies has not been determined. Animal models of the P301L tau mutation have shown gliosis, astrocytosis, and neuronal apoptosis (27,28). A pro-apoptotic effect of N279K and V337M mutant tau in transfected SH-SY5Y cells has also been reported (25). However, a P301S mutant tau mouse did not show evidence of apoptotic death (29). Our group has previously shown that osmotic stress of SH-SY5Y cells results in activation of an apoptotic cascade (46). Because the R406W tau mutation is reported to have a relatively mild phenotype in the families (30) and therefore under normal conditions may not be toxic to cells, the cells were subjected to osmotic stress to determine if R406W mutant tau-expressing CN1.4 cortical cells were more vulnerable compared with cells expressing wild type tau. The results clearly demonstrated that R406W mutant tau does in fact exert a toxic effect under stress conditions in our cortical cell model, because there was more caspase activation in cells expressing R406W mutant tau compared with cells expressing wild type tau. Interestingly, in a recent study it was demonstrated that pseudohyperphosphorylated tau was capable of inducing apoptotic cell death in neurons, implying that hyperphosphorylated tau may gain a toxic function in neurons (49). Because R406W mutant tau is hyperphosphorylated in the cortical cells, it can be speculated that the combined effects of the mutation and increased phosphorylation impair the function (microtubule binding) of tau, and this may directly or indirectly result in compromised cellular function, which increases vulnerability to stress-induced cell death. Further studies are required to determine the mechanisms by which mutant R406W sensitizes cortical cells to apoptotic stimuli. Overall, these studies demonstrate that immortalized cortical cells are a suitable model for the expression of wild type and mutant FTDP-17 tau and that in this cell model R406W mutant tau is hyperphosphorylated compared with wild type tau and does not efficiently bind microtubules due to the increased phosphorylation, thus reproducing features of the disease state.