Expression of Tau Produces Aberrant Plasma Membrane Blebbing in Glial Cells Through RhoA-ROCK-Dependent F-Actin Remodeling

Plasmid constructs

Plasmid preparation was described in our previous report [18]. Prior to transfection experiments, the integrity of pcDNA3.1 Zeo (-) plasmids containing the sequence to express either Tau441 or Tau421, as well as the empty vector, was confirmed by double sequence analysis. Green fluorescent protein (GFP)-fused Tau was produced from the pVLGT42 plasmid (GFP-Tau441).

Cell culture and transfection

The C6-glial cell line obtained from American Type Culture Collection (ATCC) (Manassas, VA, USA) was grown in Dulbecco’s Modified Eagle Medium (DMEM-F12); glucose (1/1) medium supplemented with 5% fetal bovine serum (FBS) (GIBCO-Life Technologies, Grand Island, NY, USA), 2 mM L-glutamine, and 100 U/mL penicillin, 100μg/mL streptomycin, and was maintained under a humidified atmosphere of 5% CO₂ at 37°C. When this reached 50–60% confluence, C6-glial cells were changed to FBS-free Optimem specialized medium (GIBCO-Invitrogen) for 15 min and was then transiently transfected with the Tau constructs (plasmid pcDNA 3.1 Zeo (-), which contain the sequences for either Tau441 or Tau421, and plasmid pVLGT42, which contains the sequence for GFP-Tau441 by using the Lipofectamine^®-PLUS^trademark reagent following the manufacturer’s instructions (Invitrogen-Life Technologies, Carlsbad, CA, USA). Three h after transfection, the cells were post-incubated in fresh DMEM-F12 medium supplemented with 5% FBS and maintained for 48 h at 37°C. Thereafter, transfected cells were washed in ice-cold Phosphate buffered saline (PBS) pH 7.4 and processed for the different assays as described later. Under this protocol we obtained a transfection efficiency of 40%.

Cell viability

The 3- (4, 5-diMethylThiazol-2-yl)-2, 5-diphenyl Tetrazolium bromide (MTT) assay (Sigma-Aldrich, St. Louis, MO, USA) was employed to evaluate cell viability. This technique determines the ability of healthy cells to produce formazan from the cleavage of the MTT tetrazolium ring [36]. Briefly, enzyme-linked immunosorbent assay (ELISA) plates with both transfected and non-transfected cells were washed with fresh culture medium and then incubated in fresh medium containing MTT (0.5 mg/mL) for 3 h at 37°C. The MTT medium was discarded and the cells were incubated in Dimethyl sulfoxide (DMSO) to dissolve the formazan aggregates. The intensity of the MTT products was read at 570 nm utilizing an ELISA microplate reader (iMark; Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Immunofluorescence and confocal microscopy

For double, triple, and multiple immunofluorescence, transfected and non-transfected C6-glial cells were fixed with 4% paraformaldehyde in PBS (pH 7.4) at room temperature (RT) for 15 min and then permeabilized in 0.1% Triton X-100-PBS (PBSt). After blocking for 15 min at RT (in 0.5% gelatin and 1.5% FBS in PBS), cells were incubated with a variety of antibodies in pair-wise combinations, always mixing one mouse monoclonal-IgG with one rabbit polyclonal-IgG to avoid crosstalk and non-specific cross-reactions. Primary antibody incubations were conducted for 1 h at RT. Antibody specifications are described in Table 1. Pair-wise combinations, as well as multiple labeling with fluorescent markers, are depicted in the figures. Secondary antibodies corresponding to mouse-IgG and rabbit polyclonal IgG were tagged with either Fluorescein-IsoThioCyanate (FITC) (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) or TetramethylRhodamine-IsoThioCyanate (TRITC^®) (Jackson ImmunoResearch Laboratories, Inc.) and incubated simultaneously in PBSt for 1 h at RT (1 : 100 dilution) in a humidity chamber. In some experiments, multilabeling was performed by including the Hoechst-33258 fluorescent nuclear marker (Invitrogen Molecular Probes, Eugene, Oregon, USA) and Rhodamine-phalloidin a high-affinity F-Actin probe conjugated to the red-orange fluorescent dye, TRITC (Invitrogen Molecular Probes).

For C6-glial cells transfected with GFP-Tau441, fixation was carried out in 100% methanol at –20°C for 5 min and incubated with a pair-wise combination of a monoclonal antibody to α-Tubulin and one rabbit polyclonal antibody to GEF-H1 (GeneTex, Inc., Irvine, CA, USA) (see Table 1). Cyanine-5 (Cy5)-tagged anti-mouse IgG and TRITC-tagged anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc.) were employed as secondary antibodies (1 : 100). These cells were also counterstained with Hoechst-33258 to visualize the nuclei. In some experiments, Tau-expressing C6-glial cells were incubated with RhoA, ROCK, and Myosin II ATPase activity inhibitors (C3-transferase, Y-27632, and Blebbistatin, respectively) (Cytoskelenton Inc., Denver, CO, USA; Tocris Bioscience, Bristol, UK and Sigma Aldrich, respectively) [37, 38] and, based on Tau, F-Actin, Tubulin, and nuclei staining, some changes in cytoskeletal and PM organization were analyzed. Immunolabeled cells were commonly analyzed by epifluorescence through a 20X (Numerical aperture [NA]: 0.5) and a 40X (NA: 0.75) Nikon Plan-Fluor lens coupled to a Nikon Eclipse-80i Microscope (Nikon Corp., Tokyo, Japan). The images were obtained and collected by using a Nikon digital sight-DG-Ri1 camera controlled with the NIS-Elements AR-3.0-SP7 software included in the system (Nikon Corp.). Quantitative counting of fluorescent cells was conducted at low magnification in recorded images. For multiple labeling, optical sectioning, co-localization, and critical observations of subcellular components, images were collected in a TCS-SP8 confocal laser-scanning microscope by using a 63X (NA: 1.4) OIL PH3 CS2 HC PL APO lens (Leica Microsystems, Heidelberg, Germany).

Analysis of cells extracts by electrophoresis and western blotting

After cell transfection and treatments with some drugs, the cell cultures were washed twice with PBS, scraped and lysed in Radio-Immuno-Precipitation-Assay (RIPA) buffer containing a cocktail of proteases inhibitors (150 mM NaCl, 50 mM TRIS, pH 8.0, 1% Triton X-100, 0.5% sodium deoxycholate, 1 mM PMSF, 100 mM NaF, 1 mM Na₃VO₄, and 2μg/mL Complete; Roche, Indianapolis, IN, USA) and centrifuged for 10 min at 12, 000 g. In some experiments, non-transfected cells were previously treated with either 1μM staurosporine (STS) (Sigma-Aldrich) or 1μg/mL Actinomycin-D (Act-D) for 12 h to induce apoptosis, and then lysed and processed in the same manner. The supernatant was collected and the protein content in the samples was determined by means of the Mini-Bradford assay utilizing the Bio-Rad protein assay reagent (Bio-Rad Laboratories, Inc.). Total extracts of the cells containing 30μg of protein were mixed in 2X sample buffer (TRIS-HCL 100 mM pH 6.8, sodium dodecyl sulfate (SDS) 4%, bromophenol blue 0.2%, β-mercaptoethanol 5%, and glycerol 20%) and boiled at 95°C for 5 min. Proteins from the samples were separated by electrophoresis on 8–10% SDS-polyacrylamide gel (SDS-PAGE) and then transferred onto a nitrocellulose membrane for immunoblotting analysis. Membranes were blocked in 5% nonfat dried milk in PBS-tw (PBS-0.1% Tween 20) overnight at 4°C. All primary antibodies (see Table 1) were diluted in PBS-tw and incubated at RT for 1 h. After washing in PBS-tw, incubation with the corresponding peroxidase-conjugated secondary antibody to either mouse or rabbit (1 : 10, 000; ZYMED Invitrogen, Carlsbad, CA, USA) was carried out for 1 h in PBS-tw. Bands of immunoreactive proteins were visualized after membrane incubation in Western Lightning^® Plus-ECL Enhanced Chemiluminescence Substrate (PerkinElmer, Inc., Waltham, MA, USA) and developed on autoradiography films (Kodak Medical X-ray, general-purpose-blue; Eastman Kodak Co., Rochester, NY, USA) according to the manufacturer’s instructions.

TUNEL assay

To determine whether Tau affects the survival of C6-cells, the Terminal deoxynucleotidyl transferase-mediated dUTP Nick End Labeling (TUNEL) was performed by using In Situ Cell Death Detection Kit (ROCHE Applied Science, Mannheim, Germany). Briefly, cells were transfected with the empty vector and incubated for 12 h in the presence of STS (1μM) to induce chemical apoptosis. The cells were fixed with 4% formaldehyde for 1 h at RT, rinsed twice in PBS, and permeabilized with 0.1% Triton X-100 in freshly prepared 0.1% sodium citrate for 2 min on ice. After rinsed twice in PBS, the coverslips were incubated with TUNEL reaction mixture for 15 min at 37°C, and then incubated with Tau 46.1 antibody to detect Tau. The coverslips were rinsed with PBS and incubated with Hoechst-33258 for 5 min and mounted with Vectashield.

Scratch-wound assay

C6-glial cells were grown to confluence on 15.6 mm culture dishes, transfected for 48 h and treated for 2 h with 12μM mitomycin C to inhibit proliferation during the experiment. Cell cultures were scratch-wounded using a sterile 10μl pipette tip, washed twice with PBS and re-fed with DMEM-F12 with 10% FBS. After 48 h, C6-cells were fixed and the progress of cell migration into the wounded area was observed by immunofluorescence with antibodies to Tau (polyclonal) and α-Tubulin (mouse monoclonal).

Transient expression of Tau alters the normal morphology and the Tubulin cytoskeleton of C6-glial cells

Tau expression was induced in C6-glial cells to evaluate alterations in cell morphology and the disturbance of the intracellular organization of the cytoskeleton. The expression of either Tau441 or Tau421, which is a proteolytic product of Caspase-3 action [17], was confirmed by immunofluorescence (Fig. 1A, B) and immunoblot analysis (Fig. 1D) in 48 h post- and non-transfected C6-glial cells (transfection efficiency around 40%). A polyclonal antibody to either variant of Tau, in combination with a monoclonal antibody to α-Tubulin, was utilized for immunofluorescence analysis. C6 cells transfected with the empty vector were not recognized by the polyclonal antibody to Tau (left panel, Fig. 1A), which indicates the absence of endogenous expression of the Tau in these glial cells. Immunoblot analysis with the Tau-5 antibody, which recognizes residues 210–230 in the molecule of Tau (see Table 1) also confirmed the lack of endogenous expression of Tau in both normal cells and in cells only transfected with the empty vector (pcDNA3.1) (Fig. 1D). For extracts of cells expressing either Tau441 or Tau421, Tau-5 recognizes the corresponding bands of 65 kDa and 63 kDa, respectively. Expression of Tau441 and Tau421 was also corroborated by using Tau-46.1 and Tau-C3 antibodies, which specifically recognize the intact C-terminus and Asp⁴²¹-truncation in the molecule, respectively (data not shown).

In culture, C6-glial cells display a normal spindle-like morphology with radiating distribution of microtubules along the cytoplasm (white arrows, Fig. 1A, B). In contrast, when C6 cells expressed either Tau441 or Tau421, severe morphological changes were observed, such as the formation of PM blebs along the entire cell surface (yellow arrows, Fig. 1A, B). In the majority of Tau-expressing cells, morphological change was also accompanied by abnormal bundling of the Tubulin cytoskeleton (arrowheads, Fig. 1A, B). These microtubule bundles were commonly observed in the vicinity of the nuclear region (arrowheads, Fig. 1B), which we previously reported as also occurring in Tau-transfected SH-SY5Y cells [18]. We found that 84% (Standard deviation [SD], ±1.5%) of Tau-expressing cells develop PM blebs (Fig. 1C). Likewise, when the number of cells exhibiting PM blebs was compared and analyzed by Student t test, no difference was found between Tau441-expressing or Tau421-expressing cells (p = 0.4641; Not significant [NS]).

When Tau-expressing cells were viewed through differential interference contrast (DIC) microscopy, the manner in which these cells acquired deformed multilobular morphology was noticeable (arrows, Fig. 2A), in comparison with the flat shape displayed by non-transfected cells. The lack of induced apoptosis in Tau-expressing cells was evidenced by negative staining with the TUNEL assay (Fig. 2B, C). After 48 h of transfection, cell viability was evaluated by using the MTT assay and expression of either Tau441 or Tau421 in C6-glial cells did not decrease this parameter when compared with empty vector-transfected cells (p = 0.803 by one-way Analysis of variance [ANOVA]) (Fig. 2D). Even after 72 h of transfection, no difference was found between these groups.

To further discard the possibility that PM blebs formation was the result of Tau-induced apoptosis, we evaluated the formation of proteolytic products of the activity of the Caspases, such as Asp²¹⁴-cleaved Poly (ADP)-ribose polymerase (PARP) in Tau-expressing and non-transfected C6 cells.

By employing a specific antibody to Asp²¹⁴-cleaved PARP in western blotting (WB) analysis of cells extracts, only a faint band corresponding to this product was observed in both Tau441 and Tau421-expressing cell extracts (upper panel, Fig. 2E). This product is less detected in non-transfected cells. For comparative purpose, an increased band corresponding to Asp²¹⁴-cleaved PARP is observed in non-transfected cells induced to apoptosis with either STS (1μM) or Act-D (1μg/mL) (last two rows, upper panel, Fig. 2E). The expression level of Actin was not altered under any condition (lower panel, Fig. 2E). These results indicate that PM blebs do not originate as apoptotic consequences in Tau-expressing cells.

Plasma membrane blebbing in Tau-transfected C6-glial cells depends on F-Actin remodeling

Because the Actin cytoskeleton has been involved in dynamic changes of the PM in several types of cells [39], we evaluated whether this fibrillary component was affected in Tau-expressing C6-glial cells. For this purpose, these cells were processed for immunofluorescence with the polyclonal antibody to Tau in combination with Rhodamine-phalloidin, a specific fluorescent marker of F-Actin [40]. As illustrated in Fig. 3A, cells not expressing Tau display spindle-like morphology evidenced by a cortical array of the Actin cytoskeleton and Actin-conformed stress fibers (arrowheads). In contrast, in both Tau441- and Tau421-expressing cells, strong PM blebbing is produced in close association with changes in F-Actin distribution (red channel, Fig. 3B, C). Under this condition, Actin-stress fibers disappeared and the cortical pool of F-Actin now surrounded the large number of blebs emerging from the PM (arrows, Fig. 3B, C). It appears that Tau-induced remodeling of the Actin cytoskeleton is involved in PM blebbing. To evaluate this possibility, we treated Tau-expressing cells with 20μM Cytochalasin-D (a depolymerization agent of F-Actin) for 30 min, and both the integrity of this component and the formation of PM blebs were analyzed by immunofluorescence.

As presented in Fig. 3D, when Tau441-expressing cells were incubated with Cytochalasin-D, formation of PM blebs was prevented (arrow in green channel, Fig. 3D) in close association with disruption of the F-Actin cytoskeleton (red channel, Fig. 3D). In the majority of cells, only a positive pool of Actin remains in polarized clusters (yellow arrow, Fig. 3D). To further validate the contribution of the F-Actin cytoskeleton to the formation of PM blebs, Cytochalasin-D was washed away from treated Tau-expressing cells, and 30 min later; the Actin cytoskeleton was reorganized in close association with the formation of PM blebs (yellow arrows, Fig. 3E). In non-transfected cells, the Actin cytoskeleton is restored in normal cortical distribution (asterisk, Fig. 3E). Similar results were observed when Tau421-expressing cells were treated and washed with this drug (data not shown). Significant differences among Cytochalasin-D-treated, Cytochalasin-D-treated and washed, and non-treated C6 cells are depicted in Fig. 4A.

To further analyze the contribution of both Tubulin and Actin cytoskeleton components to PM blebbing, C6-glial cells not expressing Tau (Fig. 5) were treated with 10μM Taxol, a microtubule polymerization agent, to mimic the Tau-induced microtubule bundling effect. Taxol treatment for 3 h promoted the formation of thick microtubule bundles (arrows, Fig. 5B), but not the formation of PM blebs. The F-Actin cytoskeleton remains unaltered in these cells, with normal distribution along the cell cortex (asterisk, Fig. 5B). For comparative purpose, the normal cytoskeleton in empty-vector transfected cells is depicted in Fig. 5A.

In contrast, when the normal microtubule network was disrupted in these cells due to incubation with 5μM Nocodazole, PM blebbing, and changes in cortical F-Actin organization were also produced (arrows, Fig. 5C); likewise, this was observed in Tau-expressing cells (arrows, Fig. 5D). Quantitative analysis of these effects is shown in Fig. 4B. These results may indicate that abnormal bundling of microtubules specifically induced by Tau promotes a reassembly of the F-Actin that gives rise to the formation of PM blebs.

Tau-induced plasma membrane blebbing is associated with activation of the Rho-GTPase-ROCK pathway

In different eukaryotic cells, stabilization of the microtubule lattice is a regulated process that depends on the binding of several microtubule-associated proteins (MAPs) [41]; however, additional molecules may also bind to these structures with a double purpose as follows: i) contribution to microtubule stabilization, and ii) transient immobilization for self-inactivation. GEF-H1 is one of these molecules that, in turn, actives a small GTPase referred to as RhoA [42]. Bound to microtubules, GEF-H1 is inactivated; however, when the microtubule depolymerizes, GEF-H1 is released into the cytoplasm, wherein it actives GTPase-RhoA in a downstream sequence. Because GEF-H1 is a relevant factor for microtubule stabilization, we evaluated the expression and distribution of this factor in Tau-expressing C6-glial cells showing PM blebs.

As depicted in Fig. 6A, in C6 cells with no Tau expression, the majority of the GEF-H1 is observed in close association with the microtubule lattice (arrows). Bound to microtubules, GEF-H1 immunostaining appears to emulate a fibrillary appearance. In contrast, when Tau-expressing cells were analyzed under the same immunostaining protocol, the majority of GEF-H1 labeling is observed in association with thick rings of microtubule bundles (arrowheads, Fig. 6B); however, some amount of immunodecorated GEF-H1 also remains free in the cytoplasm and is distributed to the PM blebbing area (arrows, Fig. 6B). The latest effect was not produced by Taxol-induced microtubule bundling, and the entire pool of GEFH-1 remains bound to these polymerized structures (arrows in Fig. 6C-D).

Because these results indicate that some amount of GEF-H1 is displaced from the microtubule lattice to the cytoplasm under Tau expression, this Rho exchange factor might in turn catalytically activate distinct Rho GTPases, such as RhoA [43], which participates in a catalytic pathway that regulates membrane-Actin cortex interactions for cell motility [44–48]. To probe this, we assessed the role of RhoA in the formation of PM blebs in Tau-expressing cells by incubating these with the exoenzyme C3-transferase, an ADP-Ribosyl Transferase (ADPRT) that selectively ribosylates Rho proteins in the effector-binding domain on asparagine⁴¹, rendering them inactive [37]. In C6 cells with no expression of Tau, incubation with C3-transferase (0.5μg/mL) for 48 h did not show any alteration in the normal organization of the cytoskeleton components and cell morphology (Fig. 7A). In comparison with the membrane blebbing formation observed in Tau-expressing cells (Fig. 7B), when C3-transferase was incubated while Tau expression began in a separate group of transfected cells, no evidence of PM blebbing was observed during 48 h of treatment (green channel, Fig. 7C). Concomitantly in these C3-transferase-treated C6 cells, the F-Actin cytoskeleton did not show any cortical remodeling associated with the formation of PM blebs (red channel, Fig. 7C). Instead, this cytoskeletal component appears to adopt a new structure that is closely associated with microtubule bundles (arrows, Fig. 7C).

Because this result indicates that activated-RhoA participates in F-Actin and PM remodeling in Tau-expressing cells, we evaluated whether this GTPase acts through the downstream effector ROCK to increase PM blebbing. To accomplish this, while Tau expression began, transfected-C6 cells were co-incubated with Y-27632 ((+)- (R)-trans-4- (1-aminoethyl)-N- (4-pyridyl) cyclohexanecarboxamide+++dihydrochloride) (10μM), a widely used specific ROCK inhibitor [38], and both F-Actin remodeling and PM blebbing were evaluated and compared with non-treated Tau-expressing C6 cells.

In accordance with Fig. 7A, in cells with no expression of Tau, no changes in either F-Actin organization or in the general morphology of the cells were visualized during 48 h of incubation with Y-27632 (not shown). In contrast, both the PM blebbing and F-Actin remodeling observed in Tau-expressing cells (Fig. 7B) were reverted when Y-27632 was included in the incubation media (arrows, Fig. 7D). Note that cells with no expression of Tau are unaffected by this treatment, and both Tubulin and F-Actin cytoskeletons remain unaltered (asterisks, Fig. 7D). Under either C3-transferase or Y-27632 treatments, the Tau-promoted microtubule bundling effect is not disturbed (see blue channel, Fig. 7C, D), which emphasizes the specificity of these drugs in regulating F-Actin remodeling. In control experiments, the rescued effects of both C3-transferase and Y-27632 in avoiding PM blebbing and F-Actin remodeling were not sustained for long when these drugs were washed from the cells, thus reestablishing the abnormal phenotype produced by the expression of Tau (not shown). Quantitative data significantly representing the effect of the previously mentioned inhibitors are presented in Fig. 4C. Because Actin-myosin interactions may take place in the Tau-induced PM blebbing, Tau-expressing C6 cells were incubated with Blebbistatin (10μM), a non-muscle Myosin-II-specific inhibitor [49], to validate the participation of Myosin-II in this process. As it was expected, the Tau-induced PM blebbing was avoided (arrow, Fig. 7E), likewise the cortical distribution of F-Actin mediating the formation of PM blebs (Fig. 7E). Quantitative data representing the significance of this effect is presented in Fig. 4C.

All of the previously mentioned results indicate that PM blebbing and F-Actin cytoskeleton remodeling are produced by Tau expression through activation of the GEF-H1-RhoA-ROCK pathway.

To give evidence for physiological alterations of the abnormal phenotype observed in C6-glial cells expressing Tau, we evaluated their migration in vitro by using a scratch-wounded assay. As shown in Fig. 8A, after stimulation with high concentrations of FBS, significant amount of empty-vector-transfected cells migrate to the wounded area to occupy the available space (arrowheads). By contrast, when Tau441-expressing cells were also stimulated, only few amounts of these cells migrate to the wounded area (arrows). Quantitatively, less than 20% of the cells that express Tau migrated to this area. This result indicates that Tau-induced abnormalities in the cytoskeleton and PM, are also accompanied by physiological alteration in the motility of C6 glial cells.

PM blebbing in Tau-expressing cells depends on F-Actin cytoskeleton remodeling

Because PM blebbing represented a specific effect produced by Tau expression and due to that microtubule bundles were not localized in association with these blebs, we searched for alterations in the organization of the F-Actin cytoskeleton, which is known to regulate cell shape and dynamics [39]. In astrocytic cells, the Actin cytoskeleton is involved in additional functions, such as calcium mobilization, signaling, glutamate uptake, and the hormone modulation of cell growth [54–56]. We observed, in C6-glial cells, that the normal F-Actin cytoskeleton is distributed cortically and is localized in stress fibers and focal adhesions that attach the cell surface to the substrate [57]. Because this F-Actin meshwork was dramatically remodeled in Tau-expressing C6 cells and observed in close association with emerging PM blebs (Fig. 3), we conclude that this type of cytoskeleton regulates the formation of these abnormal protrusions of the PM. This statement was corroborated when Tau-expressing C6 cells were incubated with Cytochalasin-D, a drug that depolymerizes the F-Actin cytoskeleton [58], and no formation of PM blebs was observed (Fig. 3D). Moreover, the reversibility by which PM blebs were formed again when Cytochalasin-D was eliminated from the media of the cells clearly demonstrates the regulatory role of F-Actin in conducting the formation of these abnormal structures. This result is in agreement with early evidence proposing that the Actin-myosin system is the driving power that promotes cell contraction and the formation of PM blebs and apoptotic bodies under physiological and pathological conditions in response to different stimuli [59, 60].

Tau-induced PM blebs are formed by activation of RhoA-GTPase and ROCK

PM blebbing is a dynamic process implicated in cytokinesis and cell movement, but also in pathological conditions, such as apoptosis and cancer cell invasion. Several studies have demonstrated that PM blebbing depends on the promotion of F-Actin-myosin interactions, which is regulated by RhoGTPase-ROCK signaling [44–47]. In our study, we corroborated that PM blebbing in Tau-expressing C6 cells was also regulated by RhoA-GTPase and ROCK signaling, because the addition of their specific inhibitors, C3-transferase and Y-27632, respectively, avoided the formation of these structures (Figs. 7C, D). ROCK is a major effector of RhoA-GTPase, and it was initially characterized for its role in mediating the formation of RhoA-induced stress fibers and focal adhesions [61, 62]. ROCK is a serine/threonine kinase that can directly phosphorylate the Myosin light-chain (MLC) or indirectly increase MLC phosphorylation (MLC-pp) by inactivating the myosin phosphatase [63, 64]. In our Tau-expressing C6-glial cells, we would expect that active ROCK would contribute to F-Actin-myosin force generation, causing blebs to protrude through MLC-pp. Confirmation of F-Actin-myosin interaction during Tau-induced PM blebbing was obtained when cells were incubated with Blebbistatin, a specific inhibitor of Myosin-II activity [49], and neither F-Actin remodeling nor PM blebs formation were observed (Fig. 4C).

Because F-Actin remodeling is involved in the formation of distinct types of PM protrusions, such as lamellipodia and filipodia that depends on the activation of different GTPases, it was difficult to rule out the participation of Rac-1 and Cdc42 in the formation of PM blebs in the C6 cells. However, we discarded this possibility because the inclusion of NSC23766, a specific inhibitor of Rac1 [65], did not affect the formation of PM blebs (data not shown). With these experiments, we emphasize the role of RhoA-GTPase in the regulation the dynamics of F-Actin remodeling associated with the formation of PM blebs that, in our particular case, is not initiated by external stimuli [59, 67]. Instead, this is produced only by the presence of Tau in the cytoplasmic space. This comprises a relevant observation, and we can speculate that, even in the absence of external stimuli, only intracellular accumulation of Tau in glial and neuronal cells occurring in distinct tauopathies may render and initiate pathologic alterations in PM architecture and functioning. Despite the evidence that Tau may interact with Actin under several normal and pathological conditions [25, 32], in our control experiments, we did not find direct interactions among these proteins; however, under incubation with either C3-transferase or Y-27632, a close association between F-Actin and microtubule bundles, probably mediated by Tau, was observed. This result supports recent in vitro evidence that Tau can induce guided polymerization of Actin filaments along microtubule tracks [27, 68]. Whether Tau or other cytoskeletal-associated proteins are responsible for this interaction is another interesting issue for further investigation.

GEF-H1: the link between Tau-induced microtubule bundling and F-Actin-dependent PM bleb formation

Despite that F-Actin remodeling through activation of the RhoA-GTPase-ROCK pathway was responsible for PM blebbing formation in C6 cells, precisely how this cascade with no extracellular stimuli was initiated, represented a major issue to be addressed. How could Tau, with its increased microtubule-bundling properties, initiate activation of the RhoA-ROCK-Actin-myosin pathway, which mediates PM bleb formation? Given that RhoA does not exhibit clear co-localization to microtubules [69], direct interaction between these two participants in activating RhoA is unlikely. However, the participation of additional upstream candidates implicated in the RhoA-ROCK pathway, such as GEF, would address this issue. Of the several classes of GEF, the catalytic activity of GEF-H1 is uniquely regulated by its localization to the microtubules. Depolymerization of microtubules leads to GEF-H1 activation, while relocalization to microtubules inhibits its activity [43]. In this regard, when we analyzed GEF-H1 localization in Tau-expressing C6-glial cells, we found that some of this factor remains bound to the microtubule bundles; however, a free pool of this factor was also observed in the cytoplasm (Fig. 6B). We think that under Tau-expression, the induction of microtubule bundling also leads to the release of the microtubule-bounded GEF-H1 into the cytoplasm, which in turn may active RhoA-GTPase and consequently the previously mentioned F-Actin remodeling pathway. This result confronts the traditional statement that depolymerization of the microtubule lattice is the key factor for releasing GEF-H1 into the cytoplasm. Tau-induced microtubule bundling is a specific process for inducing the previously noted effects, because incubation with Taxol to stimulate the same effect was not sufficient to induce the release of GEF-H1 to the cytoplasmic space (Figs. 5B; 6C-D). Because Tau and GEF-H1 share localization in the microtubule bundles (Fig. 6B), it may be possible that the binding of Tau along the microtubule surface in some way displaces some amount of the microtubule-bounded GEF-H1 into the cytoplasm; however, more experiments are required to be conclusive.

Overall, it has been demonstrated that the regulated release of GEF-H1 from microtubules is a recurring mechanism for achieving localized activation of RhoA in several physiological processes [70–72]; however, this regulation is also implicated in the development of numerous pathophysiological conditions, including chronic diarrhea [73], cardiomyopathies [74], and cancer [75]. Tau-induced PM blebbing, responding to changes in the Tubulin and F-Actin cytoskeletons, may then result in a pathological condition that alters the normal functioning of glial cells, thus comprising a relevant factor that contributes to neuronal dysfunction.

In summary, the present study further contributes to the description a new molecular mechanism for Tau-induced neuropathology and degeneration in astrocytes. When expressed in astrocytic C6 cells, Tau toxicity was represented by the formation of abnormal blebs along the PM, associated with remodeling of the Tubulin and F-Actin cytoskeletons. This phenotype was produced by the activation of the RhoA-GTPase-ROCK pathway that regulates F-Actin-myosin interaction. We think that this cascade is initiated upstream by the release of GEF-H1 into the cytoplasm, due to the abnormal bundling of Tubulin promoted by Tau (Fig. 9). Because glial cells contribute to the neuronal maintenance and survival, Tau-induced alterations in these cells may also account for the neurodegenerative process observed in the brains of individuals suffering from tauopathies.

Authors’ disclosures available online (http://www.j-alz.com/manuscript-disclosures/15-0396r1.