- Email: [email protected]
Secretion of Tau via an Unconventional Non-vesicular Mechanism
Report
Secretion of Tau via an Unconventional Nonvesicular Mechanism Graphical Abstract
Authors Maria Merezhko, Cecilia A. Brunello, Xu Yan, Helena Vihinen, Eija Jokitalo, Riikka-Liisa Uronen, Henri J. Huttunen
Correspondence [email protected]
In Brief Merezhko et al. show that Tau protein is released from cells by an unconventional secretory mechanism that can be manipulated by changing plasma membrane properties. Phosphorylated oligomeric Tau clusters at the plasma membrane, allowing its escape from cells directly through the plasma membrane.
Highlights d
ATP-independent unconventional secretion of Tau directly through the plasma membrane
d
Phosphorylated oligomeric Tau clusters in plasma membrane microdomains
d
Tau secretion can be manipulated by changing plasma membrane properties
d
Heparan sulfate proteoglycans facilitate unconventional secretion of Tau
Merezhko et al., 2018, Cell Reports 25, 2027–2035 November 20, 2018 ª 2018 The Author(s). https://doi.org/10.1016/j.celrep.2018.10.078
Cell Reports
Report Secretion of Tau via an Unconventional Non-vesicular Mechanism Maria Merezhko,1,3 Cecilia A. Brunello,1,3 Xu Yan,1 Helena Vihinen,2 Eija Jokitalo,2 Riikka-Liisa Uronen,1 and Henri J. Huttunen1,4,* 1Neuroscience
Center, HiLIFE, University of Helsinki, 00014 Helsinki, Finland Microscopy Unit, Institute of Biotechnology, HiLIFE, University of Helsinki, 00014 Helsinki, Finland 3These authors contributed equally 4Lead Contact *Correspondence: [email protected] https://doi.org/10.1016/j.celrep.2018.10.078 2Electron
SUMMARY
Tauopathies are characterized by cerebral accumulation of Tau protein aggregates that appear to spread throughout the brain via a cell-to-cell transmission process that includes secretion and uptake of pathological Tau, followed by templated misfolding of normal Tau in recipient cells. Here, we show that phosphorylated, oligomeric Tau clusters at the plasma membrane in N2A cells and is secreted in vesicle-free form in an unconventional process sensitive to changes in membrane properties, particularly cholesterol and sphingomyelin content. Cell surface heparan sulfate proteoglycans support Tau secretion, possibly by facilitating its release after membrane penetration. Notably, secretion of endogenous Tau from primary cortical neurons is mediated, at least partially, by a similar mechanism. We suggest that Tau is released from cells by an unconventional secretory mechanism that involves its phosphorylation and oligomerization and that membrane interaction may help Tau to acquire properties that allow its escape from cells directly through the plasma membrane. INTRODUCTION Accumulation of misfolded Tau into paired helical filaments and further into insoluble aggregates and neurofibrillary tangles is a hallmark of tauopathies, including Alzheimer’s disease (Lee et al., 2001). Tau is a natively unfolded neuronal microtubulestabilizing protein that, in pathological conditions, is hyperphosphorylated and detaches from microtubules, leading to misfolding and formation of pathological Tau aggregates in the cytosol. Pathological Tau is released from cells, even in the absence of cell death or membrane leakage, and internalized by neighboring cells, where the misfolded Tau templates misfolding of healthy Tau molecules, propagating Tau pathology (Goedert et al., 2017). Although this pathological spreading of Tau has gained much attention, the secretion mechanism remains unknown, although several routes, such as secretion in
extracellular vesicles (EVs), have been suggested (Saman et al., 2012; Simo´n et al., 2012; Dujardin et al., 2014). Recently, some groups have reported that only a minority of Tau is localized in extracellular vesicles, whereas the vast majority of secreted Tau is found as a vesicle-free protein (Yan et al., 2016; Wegmann et al., 2016; Wang et al., 2017). Here we show that Tau uses an unconventional secretion route that does not require ATP or Ca2+ but can be regulated by manipulating plasma membrane properties or cell surface heparan sulfate proteoglycans (HSPGs). Tau clusters in microdomains at or near the plasma membrane, independent of vesicles, and the presence of cholesterol and sphingomyelin support Tau secretion. This unconventional secretion process shares several characteristics with both overexpressed human Tau in N2A neuroblastoma cells and endogenous rat Tau in cultured cortical neurons. Based on these data, we suggest a membrane microdomain- and HSPG-dependent unconventional mechanism for Tau secretion directly through the plasma membrane. RESULTS AND DISCUSSION Overexpressed Tau Is Localized to the Plasma Membrane and Secreted Independent of Vesicles and ATP For studying the secretion mechanisms of pathological forms of Tau, we used a model where human Tau (isoform 0N4R) is overexpressed in the mouse N2A neuroblastoma cell line, leading to increased levels of free hyperphosphorylated Tau in the cytosol. We first isolated EVs from N2A/Tau cell media and found no Tau present in EVs by western blot (Figure S1A). A protein fragment complementation assay (PCA), a highly sensitive detection method for Tau dimers and oligomers (Figures S2A and S2B), showed that more than 99% of Tau oligomers were secreted in a vesicle-free form (Figure S1B). To characterize the secreted Tau species in more detail, we used a combination of size-based filter centrifugation and dot blot (Figure S2C). The majority of secreted Tau (80%) was found to be in the form of dimers, trimers, or tetramers, as shown in Figures S2D and S2E, which is in line with previous reports (Yan et al., 2016; Brunello et al., 2016). Next, we tested the behavior of secreted Tau in an in vitro Thioflavin S aggregation assay. The Tau-conditioned medium alone showed higher Thioflavin S fluorescence compared with mock-conditioned
Cell Reports 25, 2027–2035, November 20, 2018 ª 2018 The Author(s). 2027 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
(legend on next page)
2028 Cell Reports 25, 2027–2035, November 20, 2018
medium, indicative of the presence of b sheet aggregates in the conditioned medium (Figure S2F). Moreover, the Tau-conditioned medium was capable of seeding aggregation of recombinant human Tau (0N4R). These results show that Tau secreted by N2A cells has several pathological characteristics. We then tested whether Tau is secreted by classical secretory processes. Depletion of cellular ATP by 2-deoxyglucose for 30 min had no effect on Tau secretion, strongly indicating that no vesicular transport or membrane fusion processes are involved in the secretion process (Figure 1A). No effect was seen in lactate dehydrogenase (LDH) release, demonstrating that this short treatment did not affect the membrane integrity of the cells. Treatment of cells with the canonical inducer of exocytosis, the Ca2+ ionophore ionomycin, did not induce Tau secretion but, instead, decreased it (Figure 1B), further excluding classical exocytosis as a mediator of Tau secretion. Using light microscopy, total Tau staining with Tau-5 antibody showed cytosolic staining with distinct puncta on the cell surface of N2A/Tau cells (Figure 1C). The T22 antibody, which specifically recognizes Tau oligomers (Lasagna-Reeves et al., 2012), and the phosphospecific antibodies AT8 (Ser202/Thr205) and PHF13 (Ser396) predominantly stained small puncta near the cell boundaries (Figures 1D–1F). Thioflavin S did not stain these sub-plasma membrane Tau structures (Figures 1G–1J), suggesting that they represent the phosphorylated oligomeric forms of Tau not yet converted into aggregates with b sheet structures. Using a super-resolution microscopy method, direct stochastic optical reconstruction microscopy (dSTORM) Tau-5 staining showed numerous small puncta at the edges of the cells together with some cytosolic signal (Figures 1K and 1L). Analysis of particle size showed a mean diameter of 23 nm and that more than 90% of all detected particles were smaller than 40 nm (Figure 1M). To visualize whether Tau-containing puncta were membranebound organelles, large protein aggregates, or something else, we used immunoelectron microscopy to localize wild-type untagged Tau stained with Tau-5 antibody and Tau-GLuc using anti-HA antibody. In both setups, Tau localized to the plasma membrane in microdomain-like structures but not in vesicles, multivesicular bodies, or any other membrane-bound compartments (Figures 1N–1S; Figure S3). Cytosolic Tau was observed
in both diffuse and dense aggregated forms. At the plasma membrane, we observed large (200 nm) membrane patches associated with diffuse cytosolic Tau clusters just below the plasma membrane (<100 nm) (Figures 1P and 1Q) and smaller membrane microdomains (50–80 nm) with very dense staining (Figures 1R and 1S). The only endomembrane that showed prominent Tau localization was the nuclear membrane, where Tau was localized in nuclear pores (Figures 1N and 1O). Some phenotypic variability was seen between cells, as can be expected in transient expression (quantified in Table S1). These results suggest a non-vesicular secretion mechanism for Tau, likely involving Tau localization in specific microdomains at the plasma membrane. This is in line with previous studies reporting Tau localization to membranes (Brandt et al., 1995; Kawarabayashi et al., 2004; Pooler et al., 2012) and further suggest that the interaction of Tau with the plasma membrane within specific microdomains may be a prerequisite for non-vesicular release of Tau. Structural compaction upon interaction of Tau with membranes (Elbaum-Garfinkle et al., 2010; Jones et al., €nze et al., 2012; Georgieva et al., 2014; Flach et al., 2012; Ku 2012) and/or formation of pore-like structures in membranes (Lasagna-Reeves et al., 2014; Patel et al., 2015) could mediate membrane penetration of Tau. Manipulation of Plasma Membrane Lipids Affects Tau Secretion If Tau is secreted directly through the plasma membrane, then manipulating plasma membrane properties should influence this process. We first extracted membrane cholesterol from N2A/Tau cells using methyl-b-cyclodextrin (Mahammad and Parmryd, 2015). Treatment with 1 mM methyl-b-cyclodextrin decreased Tau secretion by 47.0% ± 7.2% (Figure 2A), suggesting that cholesterol or membrane order in general affects Tau release. Loading more cholesterol to cells using preformed cholesterol:methyl-b-cyclodextrin complexes increased Tau secretion by 75.8% ± 8.1%, further supporting this view (Figure 2B). Second, we depleted the cells of sphingolipids with two different strategies. Plasma membrane sphingomyelin was degraded by adding exogenous sphingomyelinase (SMase) to the medium, or, alternatively, all sphingolipids were depleted with the serine palmitoyl-coenzyme A (CoA) transferase inhibitor myriocin.
Figure 1. Tau Is Secreted Independent of ATP or Ca2+ and Localized to the Plasma Membrane (A and B) Tau PCA and LDH release in N2A/Tau cells and media. Data are represented as mean ± SEM. (A) ATP depletion by 2-deoxyglucose treatment (30 min). ATP levels were measured by CellTiter-Glo ATP assay; n = 4–6. (B) Ionomycin treatment (30 min); n = 3. (C–J) Immunofluorescence micrographs of N2A cells transfected with Tau-GLuc1/2. Shown is staining with Tau-5 (C), T22 (D), AT8 (E), and PHF13 (F) antibodies. White arrowheads indicate punctate Tau staining near the cell surface (C–F). Also shown are (G) Tau-5, (H) AT8, (I) Thioflavin S staining, and (J) a merged picture of a single cell. White arrows indicate differential staining between different forms of Tau (G–J). (K) A dSTORM image of Tau-5 staining. The image was acquired by first finding the total internal reflection fluorescence (TIRF) plane and then focusing just above that. (L) Higher magnification of the area indicated by a white box in (K). The cell surface is indicated by a dotted white line. (M) Histogram of mean diameter of Tau-5-positive clusters observed in dSTORM. (N–S) Transmission electron microscopy. (N, P, and R) Transfection with native Tau (0N4R sequence without tags or fusions) stained with Tau-5 antibody-immunogold. (P) and (R) show higher magnification of Tau close to the plasma membrane. (O, Q, and S) Transfection with Tau-GLuc1/2 (containing an hemagglutinin [HA] tag) stained with an anti-HA antibody-immunogold. (Q) and (S) show higher magnification of Tau close to the plasma membrane. White arrowheads in (N) and (O) point to plasma membrane microdomains and black arrowheads to nucleopores. Scale bars in (C)–(K), 10 mm; in L, 100 nm, in (N) and (O), 1 mm, in (P)–(S), 100 nm. N, nucleus; Mi, mitochondrion; MVB, multivesicular body. See also Figures S1–S3 and Table S1.
Cell Reports 25, 2027–2035, November 20, 2018 2029
Figure 2. Modulation of Membrane Properties Alters Tau Secretion Shown is Tau PCA and LDH release in N2A/Tau cells and media. (A) Cholesterol extraction by methyl-b-cyclodextrin (30-min pretreatment followed by 30-min medium collection); n = 3–4. (B) Cholesterol loading by cholesterol:methyl-b-cyclodextrin (mbCD) complexes (1-hr treatment); n = 3. (C and D) Plasma membrane sphingomyelin reduction by SMase (4 hr) (C) and inhibition of sphingolipid synthesis by myriocin (16-hr pretreatment followed by 4-hr medium collection in the continued presence of myriocin) (D); n = 3–4. (E) Increased membrane fluidity by DHA (16-hr pretreatment with DHA followed by 4-hr medium collection in the continued presence of DHA); n = 3–4. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (ANOVA).
Both treatments reduced Tau secretion maximally by 36% (Figures 2C and 2D). Third, N2A/Tau cells were treated with docosahexaenoic acid (DHA), a polyunsaturated fatty acid that reduces membrane order and disrupts lipid raft-like membrane microdomains and whose consumption has been associated
2030 Cell Reports 25, 2027–2035, November 20, 2018
with reduced risk of Alzheimer’s disease (Yurko-Mauro et al., 2010; Zhang et al., 2016). Upon DHA treatment, Tau secretion was concentration-dependently reduced, with a maximal inhibition of 93.8% ± 0.9% at 50 mM DHA. Tau secretion was already significantly reduced by 1 mM DHA (Figure 2E), close to the DHA
Figure 3. HSPGs and Tau Aggregation Inhibitors Modulate Tau Secretion and Uptake (A–C) Tau PCA and LDH release in N2A/Tau cells and media. (A) Cells treated with NaClO3 (16-hr pretreatment followed by 4-hr medium collection in the continued presence of NaClO3); n = 3–5. (B) Heparinase I and heparinase III treatments (4 hr); n = 3–4. (C) Chondroitinase ABC treatment (4 hr); n = 3. (D) A representative dot blot of cells and conditioned media from N2A/Tau cells treated with emodin, BSc3094, PcTS, and EGCG at 10 mM for 16 hr, stained with Tau-5 antibody. (E) Quantification of dot blot data (n = 4). Medium dots were normalized by the corresponding cell lysate dots. (F) Tau PCA from N2A/Tau cells treated with 10 mM EGCG for 16 hr. LDH release was monitored as a control. (G–I) Tau uptake assay. (G) Medium conditioned with untransfected or Tau-transfected N2A cells, with or without 10 mM EGCG for 16 hr, was fed to naive N2A cells for 16 hr. Cells were washed extensively and then stained with Tau-5 antibodies and Hoechst 33342. Representative images are shown. Scale bar, 20 mm. (H) Quantification of Tau-5 fluorescence intensity in N2A cells in the Tau uptake assay. Mean fluorescence intensity was measured from 18 images of EGCGtreated cells and 17 images of control-treated cells (average, n = 350 cells/condition). The values were normalized by the number of cells per image. (I) Tau uptake PCA. Tau medium was collected from N2A cells expressing Tau-GLuc1/2 with or without EGCG as in (F). Naive non-transfected N2A cells were exposed to Tau-conditioned medium for 16 hr and washed extensively to remove any non-internalized Tau and EGCG. Then the level of internalized Tau oligomers was measured by PCA. (J) Tau-conditioned medium was generated as in (F) without EGCG. Naive, non-transfected N2A cells were then exposed to the Tau-conditioned medium in the presence of 10 mM EGCG or vehicle. n = 3–4. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (ANOVA).
concentration found in normal human cerebrospinal fluid (Freund Levi et al., 2014). Cultured neuroblastoma cells typically have a significantly lower level of DHA in their membranes compared with neurons in vivo (Reynolds et al., 2001; Akbar et al., 2005), which may make them particularly sensitive to the effects of DHA.
In summary, these data indicate that the rigid liquid-ordered phase of the membrane enriched in specific lipid species, such as cholesterol and sphingomyelin, are centrally involved in Tau release from cells. Somewhat surprisingly, Tau is more efficiently released from ordered membranes, and Tau secretion can be reduced by increasing the fluidity of cellular membranes.
Cell Reports 25, 2027–2035, November 20, 2018 2031
Figure 4. Secretion of Endogenous Tau in Primary Cortical Neurons Is Modulated by Levels of Plasma Membrane Cholesterol and HSPGs and by EGCG (A–D) Primary cultures of rat cortical neurons were treated at 14–25 DIV with increasing concentrations of (A) methyl-b-cyclodextrin for 16 hr, (B) cholesterol:methyl-b-cyclodextrin for 1 hr, (C) NaClO3 for 72 hr, and (D) EGCG for 16 hr. Cells and conditioned media were analyzed for Tau using ELISA and cell viability and membrane integrity using an LDH assay. n = 3–4. Data are represented as mean ± SEM. **p < 0.01, ***p < 0.001 (ANOVA).
Interestingly, a similar interdependence of membrane composition and rigidity and membrane permeation was recently reported for arginine-rich cell-penetrating peptides (Wallbrecher et al., 2017). Cell Surface HSPGs and the Aggregation State of Tau Modulate Secretion Efficiency Because Tau binds glycosaminoglycans (GAGs) (Goedert et al., 1996; Hasegawa et al., 1997) that are abundantly present at the plasma membrane, we hypothesized that these interactions may play a role in Tau secretion. HSPGs play a central role in the unconventional secretion of FGF2 (Steringer and Nickel, 2018). Moreover, while this manuscript was under review, Katsinelos et al. (2018) published a paper suggesting that Tau secretion may share some similarities with unconventional secretion of FGF2. NaClO3 is widely used to inhibit synthesis of PAPS (30 -phosphoadenosine 50 -phosphosulphate), the general sulfate donor in GAG biosynthesis (Venkatachalam, 2003). Treatment of N2A/Tau cells with NaClO3 resulted in a dose-dependent decrease in Tau secretion (maximal decrease, 37.6% ± 5.4% 2032 Cell Reports 25, 2027–2035, November 20, 2018
at 50 mM NaClO3) (Figure 3A). To reduce the amount of cell surface HSPGs more specifically, we treated cells with heparinase I or heparinase III, two enzymes that cleave heparin and heparan sulfate-type sulfated glycans with different specificities. Both treatments led to a decrease in both intracellular and secreted Tau, with heparinase III having a more consistent effect of reducing Tau secretion by 25.0% ± 3.5% (Figure 3B). Treatment of N2A/Tau cells with chondroitinase ABC to reduce cell surface levels of chondroitin and dermatan sulfate GAGs did not decrease Tau secretion (Figure 3C). These data suggest that cell surface proteoglycans, specifically HSPGs, facilitate Tau secretion and are in line with a previous report showing similar results (Katsinelos et al., 2018). Because oligomerization plays an important role in unconventional secretion of FGF2 (Steringer and Nickel, 2018), and because at least some of the plasma membrane-localized Tau appears to be hyperphosphorylated and oligomerized, we next tested the effect of inhibition of Tau aggregation on secretion. To narrow down the Tau species that penetrate the membrane, we used four different Tau aggregation inhibitors
(TAIs), emodin (Pickhardt et al., 2005), BSc3094 (Pickhardt et al., 2007), phthalocyanine tetrasulfonate (PcTS) (Akoury et al., 2013), and epigallocatechin gallate (EGCG) (Ehrnhoefer et al., 2008; Wobst et al., 2015), which inhibit Tau aggregation at different stages (Bulic et al., 2013; Cisek et al., 2014). Because some of the TAIs are colored, they prevent the use of PCA. Instead, we used a dot blot to study Tau secretion. All four compounds reduced Tau secretion, with effects ranging from 35% to 94% reduction (Figures 3D and 3E). This suggests that formation of oligomeric aggregation intermediates, possibly occurring at the membrane interface, is a prerequisite for unconventional secretion of Tau. EGCG caused the strongest reduction in Tau secretion in the dot blot assay, and because it is a non-colored compound, it was also tested in the PCA-based secretion assay. EGCG reduced secreted Tau oligomers, as detected by PCA, by 63.4% ± 3.6% (Figure 3F). Furthermore, we found that Tau secreted in the presence of EGCG is taken up significantly less effectively compared with the control (Figures 3G–3I). Notably, EGCG did not reduce Tau uptake in recipient cells when added to the medium post-secretion; i.e., after the conditioned medium was removed from the donor cells (Figure 3J). Because EGCG not only inhibits Tau aggregation but also modulates membrane properties (Sun et al., 2009; Patra et al., 2008), both of these mechanisms could contribute to the robust overall effect of EGCG on Tau secretion in N2A cells. Overall, together with the finding that phosphorylated oligomeric forms of Tau localize in clusters in sub-plasma membrane microdomains (Figure 1), these results suggest that certain forms of oligomeric aggregation intermediates may be the Tau species capable of penetrating the plasma membrane and exiting cells. Endogenous Tau from Primary Neurons Is Also Secreted via a Similar Unconventional Mechanism To test whether endogenous neuronal Tau is also secreted via the same or a similar mechanism as overexpressed pathological Tau, we analyzed media from cultures of rat primary cortical neurons (21 days in vitro [DIV]) using Tau ELISA. Similar to N2A/Tau cells, Tau secretion was inhibited by decreasing plasma membrane cholesterol with methyl-b-cyclodextrin, with a maximal decrease of 47.8% ± 0.6% at 1 mM methylb-cyclodextrin (Figure 4A). Increasing neuronal plasma membrane cholesterol content by cholesterol:methyl-b-cyclodextrin complexes resulted in a 105% ± 15.6% increase in Tau secretion (Figure 4B). Reduction of sulfated GAGs on the cell surface by treatment of cortical neurons with 50 mM NaClO3 decreased Tau secretion by 45.0% ± 5.3% (Figure 4C). EGCG reduced neuronal secretion of Tau by 35.7% ± 7.1% (Figure 4D). LDH release was not significantly altered by any of these treatments, suggesting that plasma membrane integrity was not affected. These results indicate that not only overexpressed pathological forms of human Tau in N2A cells but also endogenous rat Tau in primary cortical neurons are secreted by an unconventional secretion mechanism. In summary, our current data demonstrate that Tau uses an unconventional protein secretion pathway to exit cells. Importantly, overexpressed pathological Tau and also endogenous Tau from primary neurons appear to be secreted via a similar
mechanism. This process is highly dependent on plasma membrane lipid content and biophysical properties, especially in the presence of cholesterol and sphingomyelin. A number of proteins are released from eukaryotic cells under physiological conditions directly through the plasma membrane in an unconventional secretion process (Rabouille, 2017). We propose that Tau acquires properties, likely associated with its phosphorylation and oligomerization, that promote its plasma membrane interaction and permeation, resulting in secretion. Suppression of plasma membrane interaction and penetration of Tau could be an effective therapeutic target for slowing down disease progression in tauopathies. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d d d
d
d
KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS B N2A cell culture B Primary rat cortical neuron culture METHOD DETAILS B Protein-fragment complementation assay (PCA) B Media fractionation B Cholesterol:mbCD complex preparation B ELISA B Immunofluorescence microscopy B dSTORM B Immunoelectron microscopy B Western blot and dot blot assay B Thioflavin S assay B Size-exclusion dot blot assay QUANTIFICATION AND STATISTICAL ANALYSIS
SUPPLEMENTAL INFORMATION Supplemental Information includes four figures and two tables and can be found with this article online at https://doi.org/10.1016/j.celrep.2018.10.078. ACKNOWLEDGMENTS We thank Prof. Stephan Michnick (Universite´ de Montre´al, Canada) for providing the humanized GLuc plasmid. Dr. Veijo Salo, Amr Abou Elezz, Mikko Liljestro¨m, and the Biomedicum Imaging Unit, University of Helsinki are thanked for technical support with dSTORM imaging. This study was funded by the Academy of Finland (grant 296409) and the Finnish Cultural Foundation. AUTHOR CONTRIBUTIONS Conceptualization, M.M., C.A.B., X.Y., R.-L.U., and H.J.H.; Methodology, M.M., C.A.B., X.Y., H.V., E.J., and H.J.H.; Investigation, M.M., C.A.B., and H.V.; Writing – Original Draft, R.-L.U. and H.J.H.; Writing – Review & Editing, M.M., C.A.B., R.-L.U., and H.J.H.; Funding Acquisition, R.-L.U. and H.J.H.; Supervision, E.J., R.-L.U., and H.J.H. DECLARATION OF INTERESTS H.J.H. is an employee and shareholder of Herantis Pharma Plc, which is unrelated to this study.
Cell Reports 25, 2027–2035, November 20, 2018 2033
Received: April 18, 2018 Revised: September 26, 2018 Accepted: October 19, 2018 Published: November 20, 2018 SUPPORTING CITATIONS The following references appear in the Supplemental Information: Ercan et al. (2017); Hoffmann et al. (1997); LoPresti et al. (1995); Mercken et al. (1992); Niman et al. (1983); Wilson et al. (1984). REFERENCES Akbar, M., Calderon, F., Wen, Z., and Kim, H.Y. (2005). Docosahexaenoic acid: a positive modulator of Akt signaling in neuronal survival. Proc. Natl. Acad. Sci. USA 102, 10858–10863. Akoury, E., Gajda, M., Pickhardt, M., Biernat, J., Soraya, P., Griesinger, C., Mandelkow, E., and Zweckstetter, M. (2013). Inhibition of tau filament formation by conformational modulation. J. Am. Chem. Soc. 135, 2853–2862. Brandt, R., Le´ger, J., and Lee, G. (1995). Interaction of tau with the neural plasma membrane mediated by tau’s amino-terminal projection domain. J. Cell Biol. 131, 1327–1340. Brunello, C.A., Yan, X., and Huttunen, H.J. (2016). Internalized Tau sensitizes cells to stress by promoting formation and stability of stress granules. Sci. Rep. 6, 30498. Bulic, B., Pickhardt, M., and Mandelkow, E. (2013). Progress and developments in tau aggregation inhibitors for Alzheimer disease. J. Med. Chem. 56, 4135–4155. Cisek, K., Cooper, G.L., Huseby, C.J., and Kuret, J. (2014). Structure and mechanism of action of tau aggregation inhibitors. Curr. Alzheimer Res. 11, 918–927. Dujardin, S., Be´gard, S., Caillierez, R., Lachaud, C., Delattre, L., Carrier, S., Loyens, A., Galas, M.C., Bousset, L., Melki, R., et al. (2014). Ectosomes: a new mechanism for non-exosomal secretion of tau protein. PLoS ONE 9, e100760. Ehrnhoefer, D.E., Bieschke, J., Boeddrich, A., Herbst, M., Masino, L., Lurz, R., Engemann, S., Pastore, A., and Wanker, E.E. (2008). EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers. Nat. Struct. Mol. Biol. 15, 558–566. Elbaum-Garfinkle, S., Ramlall, T., and Rhoades, E. (2010). The role of the lipid bilayer in tau aggregation. Biophys. J. 98, 2722–2730. Ercan, E., Eid, S., Weber, C., Kowalski, A., Bichmann, M., Behrendt, A., Matthes, F., Krauss, S., Reinhardt, P., Fulle, S., and Ehrnhoefer, D.E. (2017). A validated antibody panel for the characterization of tau post-translational modifications. Mol. Neurodegener. 12, 87.
Hasegawa, M., Crowther, R.A., Jakes, R., and Goedert, M. (1997). Alzheimerlike changes in microtubule-associated protein Tau induced by sulfated glycosaminoglycans. Inhibition of microtubule binding, stimulation of phosphorylation, and filament assembly depend on the degree of sulfation. J. Biol. Chem. 272, 33118–33124. Hoffmann, R., Dawson, N.F., Wade, J.D., and Otvo¨s, L., Jr. (1997). Oxidized and phosphorylated synthetic peptides corresponding to the second and third tubulin-binding repeats of the tau protein reveal structural features of paired helical filament assembly. J. Pept. Res. 50, 132–142. Jones, E.M., Dubey, M., Camp, P.J., Vernon, B.C., Biernat, J., Mandelkow, E., Majewski, J., and Chi, E.Y. (2012). Interaction of tau protein with model lipid membranes induces tau structural compaction and membrane disruption. Biochemistry 51, 2539–2550. €ller, H.M., Nachman, Katsinelos, T., Zeitler, M., Dimou, E., Karakatsani, A., Mu E., Steringer, J.P., Ruiz de Almodovar, C., Nickel, W., and Jahn, T.R. (2018). Unconventional Secretion Mediates the Trans-cellular Spreading of Tau. Cell Rep. 23, 2039–2055. Kawarabayashi, T., Shoji, M., Younkin, L.H., Wen-Lang, L., Dickson, D.W., Murakami, T., Matsubara, E., Abe, K., Ashe, K.H., and Younkin, S.G. (2004). Dimeric amyloid beta protein rapidly accumulates in lipid rafts followed by apolipoprotein E and phosphorylated tau accumulation in the Tg2576 mouse model of Alzheimer’s disease. J. Neurosci. 24, 3801–3809. Kowal, J., Arras, G., Colombo, M., Jouve, M., Morath, J.P., Primdal-Bengtson, B., Dingli, F., Loew, D., Tkach, M., and The´ry, C. (2016). Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. Proc. Natl. Acad. Sci. USA 113, E968–E977. Kuja-Panula, J., Kiiltoma¨ki, M., Yamashiro, T., Rouhiainen, A., and Rauvala, H. (2003). AMIGO, a transmembrane protein implicated in axon tract development, defines a novel protein family with leucine-rich repeats. J. Cell Biol. 160, 963–973. €nze, G., Barre´, P., Scheidt, H.A., Thomas, L., Eliezer, D., and Huster, D. Ku (2012). Binding of the three-repeat domain of tau to phospholipid membranes induces an aggregated-like state of the protein. Biochim. Biophys. Acta 1818, 2302–2313. Kysenius, K., Muggalla, P., Ma¨tlik, K., Aruma¨e, U., and Huttunen, H.J. (2012). PCSK9 regulates neuronal apoptosis by adjusting ApoER2 levels and signaling. Cell. Mol. Life Sci. 69, 1903–1916. Lasagna-Reeves, C.A., Castillo-Carranza, D.L., Sengupta, U., Sarmiento, J., Troncoso, J., Jackson, G.R., and Kayed, R. (2012). Identification of oligomers at early stages of tau aggregation in Alzheimer’s disease. FASEB J. 26, 1946– 1959. Lasagna-Reeves, C.A., Sengupta, U., Castillo-Carranza, D., Gerson, J.E., Guerrero-Munoz, M., Troncoso, J.C., Jackson, G.R., and Kayed, R. (2014). The formation of tau pore-like structures is prevalent and cell specific: possible implications for the disease phenotypes. Acta Neuropathol. Commun. 2, 56.
€ger, M., Leonhardt, M., Flach, K., Hilbrich, I., Schiffmann, A., Ga¨rtner, U., Kru Waschipky, H., Wick, L., Arendt, T., and Holzer, M. (2012). Tau oligomers impair artificial membrane integrity and cellular viability. J. Biol. Chem. 287, 43223–43233.
Lee, V.M., Goedert, M., and Trojanowski, J.Q. (2001). Neurodegenerative tauopathies. Annu. Rev. Neurosci. 24, 1121–1159.
Freund Levi, Y., Vedin, I., Cederholm, T., Basun, H., Faxe´n Irving, G., Eriksdotter, M., Hjorth, E., Schultzberg, M., Vessby, B., Wahlund, L.O., et al. (2014). Transfer of omega-3 fatty acids across the blood-brain barrier after dietary supplementation with a docosahexaenoic acid-rich omega-3 fatty acid preparation in patients with Alzheimer’s disease: the OmegAD study. J. Intern. Med. 275, 428–436.
Mahammad, S., and Parmryd, I. (2015). Cholesterol depletion using methylb-cyclodextrin. Methods Mol. Biol. 1232, 91–102.
Georgieva, E.R., Xiao, S., Borbat, P.P., Freed, J.H., and Eliezer, D. (2014). Tau binds to lipid membrane surfaces via short amphipathic helices located in its microtubule-binding repeats. Biophys. J. 107, 1441–1452. Goedert, M., Jakes, R., Spillantini, M.G., Hasegawa, M., Smith, M.J., and Crowther, R.A. (1996). Assembly of microtubule-associated protein tau into Alzheimer-like filaments induced by sulphated glycosaminoglycans. Nature 383, 550–553. Goedert, M., Eisenberg, D.S., and Crowther, R.A. (2017). Propagation of Tau Aggregates and Neurodegeneration. Annu. Rev. Neurosci. 40, 189–210.
2034 Cell Reports 25, 2027–2035, November 20, 2018
LoPresti, P., Szuchet, S., Papasozomenos, S.C., Zinkowski, R.P., and Binder, L.I. (1995). Functional implications for the microtubule-associated protein tau: localization in oligodendrocytes. Proc. Natl. Acad. Sci. USA 92, 10369–10373.
McLean, I.W., and Nakane, P.K. (1974). Periodate-lysine-paraformaldehyde fixative. A new fixation for immunoelectron microscopy. J. Histochem. Cytochem 22, 1077–1083. €bke, U., Six, J., Boons, J., Van de Voorde, Mercken, M., Vandermeeren, M., Lu A., Martin, J.J., and Gheuens, J. (1992). Monoclonal antibodies with selective specificity for Alzheimer Tau are directed against phosphatase-sensitive epitopes. Acta Neuropathol. 84, 265–272. Niman, H.L., Houghten, R.A., Walker, L.E., Reisfeld, R.A., Wilson, I.A., Hogle, J.M., and Lerner, R.A. (1983). Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. Natl. Acad. Sci. USA 80, 4949–4953.
Patel, N., Ramachandran, S., Azimov, R., Kagan, B.L., and Lal, R. (2015). Ion Channel Formation by Tau Protein: Implications for Alzheimer’s Disease and Tauopathies. Biochemistry 54, 7320–7325.
Sun, Y., Hung, W.C., Chen, F.Y., Lee, C.C., and Huang, H.W. (2009). Interaction of tea catechin (-)-epigallocatechin gallate with lipid bilayers. Biophys. J. 96, 1026–1035.
Patra, S.K., Rizzi, F., Silva, A., Rugina, D.O., and Bettuzzi, S. (2008). Molecular targets of (-)-epigallocatechin-3-gallate (EGCG): specificity and interaction with membrane lipid rafts. J. Physiol. Pharmacol. 59 (Suppl 9), 217–235.
Thery, C., Amigorena, S., Raposo, G., and Clayton, A. (2006). Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr. Protoc. Cell Biol. 30, 3.22.1–3.22.29.
Pickhardt, M., Gazova, Z., von Bergen, M., Khlistunova, I., Wang, Y., Hascher, A., Mandelkow, E.M., Biernat, J., and Mandelkow, E. (2005). Anthraquinones inhibit tau aggregation and dissolve Alzheimer’s paired helical filaments in vitro and in cells. J. Biol. Chem. 280, 3628–3635.
Venkatachalam, K.V. (2003). Human 30 -phosphoadenosine 50 -phosphosulfate (PAPS) synthase: biochemistry, molecular biology and genetic deficiency. IUBMB Life 55, 1–11.
Pickhardt, M., Larbig, G., Khlistunova, I., Coksezen, A., Meyer, B., Mandelkow, E.M., Schmidt, B., and Mandelkow, E. (2007). Phenylthiazolyl-hydrazide and its derivatives are potent inhibitors of tau aggregation and toxicity in vitro and in cells. Biochemistry 46, 10016–10023. Pooler, A.M., Usardi, A., Evans, C.J., Philpott, K.L., Noble, W., and Hanger, D.P. (2012). Dynamic association of tau with neuronal membranes is regulated by phosphorylation. Neurobiol. Aging 33, 431.e27–431.e38. Rabouille, C. (2017). Pathways of Unconventional Protein Secretion. Trends Cell Biol. 27, 230–240. Remy, I., and Michnick, S.W. (2006). A highly sensitive protein-protein interaction assay based on Gaussia luciferase. Nat. Methods 3, 977–979. Reynolds, L.M., Dalton, C.F., and Reynolds, G.P. (2001). Phospholipid fatty acids and neurotoxicity in human neuroblastoma SH-SY5Y cells. Neurosci. Lett. 309, 193–196. Salo, V.T., Belevich, I., Li, S., Karhinen, L., Vihinen, H., Vigouroux, C., Magre´, J., Thiele, C., Ho¨ltta¨-Vuori, M., Jokitalo, E., and Ikonen, E. (2016). Seipin regulates ER-lipid droplet contacts and cargo delivery. EMBO J. 35, 2699–2716. Saman, S., Kim, W., Raya, M., Visnick, Y., Miro, S., Saman, S., Jackson, B., McKee, A.C., Alvarez, V.E., Lee, N.C., and Hall, G.F. (2012). Exosome-associated tau is secreted in tauopathy models and is selectively phosphorylated in cerebrospinal fluid in early Alzheimer disease. J. Biol. Chem. 287, 3842–3849. Seemann, J., Jokitalo, E.J., and Warren, G. (2000). The role of the tethering proteins p115 and GM130 in transport through the Golgi apparatus in vivo. Mol. Biol. Cell 11, 635–645. Simo´n, D., Garcı´a-Garcı´a, E., Royo, F., Falco´n-Pe´rez, J.M., and Avila, J. (2012). Proteostasis of tau. Tau overexpression results in its secretion via membrane vesicles. FEBS Lett. 586, 47–54. Steringer, J.P., and Nickel, W. (2018). A direct gateway into the extracellular space: Unconventional secretion of FGF2 through self-sustained plasma membrane pores. Semin. Cell Dev. Biol. 83, 3–7.
Wallbrecher, R., Ackels, T., Olea, R.A., Klein, M.J., Caillon, L., Schiller, J., Bove´e-Geurts, P.H., van Kuppevelt, T.H., Ulrich, A.S., Spehr, M., et al. (2017). Membrane permeation of arginine-rich cell-penetrating peptides independent of transmembrane potential as a function of lipid composition and membrane fluidity. J. Control. Release 256, 68–78. €ger, L., Irsen, S., Tepper, K., ChanduWang, Y., Balaji, V., Kaniyappan, S., Kru patla, R., Maetzler, W., Schneider, A., Mandelkow, E., and Mandelkow, E.M. (2017). The release and trans-synaptic transmission of Tau via exosomes. Mol. Neurodegener. 12, 5. Wegmann, S., Nicholls, S., Takeda, S., Fan, Z., and Hyman, B.T. (2016). Formation, release, and internalization of stable tau oligomers in cells. J. Neurochem. 139, 1163–1174. Wilson, I.A., Niman, H.L., Houghten, R.A., Cherenson, A.R., Connolly, M.L., and Lerner, R.A. (1984). The structure of an antigenic determinant in a protein. Cell 37, 767–778. Wobst, H.J., Sharma, A., Diamond, M.I., Wanker, E.E., and Bieschke, J. (2015). The green tea polyphenol (-)-epigallocatechin gallate prevents the aggregation of tau protein into toxic oligomers at substoichiometric ratios. FEBS Lett. 589, 77–83. Yan, X., Nyka¨nen, N.P., Brunello, C.A., Haapasalo, A., Hiltunen, M., Uronen, R.L., and Huttunen, H.J. (2016). FRMD4A-cytohesin signaling modulates the cellular release of tau. J. Cell Sci. 129, 2003–2015. Yurko-Mauro, K., McCarthy, D., Rom, D., Nelson, E.B., Ryan, A.S., Blackwell, A., Salem, N., Jr., and Stedman, M.; MIDAS Investigators (2010). Beneficial effects of docosahexaenoic acid on cognition in age-related cognitive decline. Alzheimers Dement. 6, 456–464. Zhang, Y., Chen, J., Qiu, J., Li, Y., Wang, J., and Jiao, J. (2016). Intakes of fish and polyunsaturated fatty acids and mild-to-severe cognitive impairment risks: a dose-response meta-analysis of 21 cohort studies. Am. J. Clin. Nutr. 103, 330–340.
Cell Reports 25, 2027–2035, November 20, 2018 2035
STAR+METHODS KEY RESOURCES TABLE
REAGENT or RESOURCE
SOURCE
IDENTIFIER
Tau-5
Invitrogen
AHB0042, RRID: AB_2536235
AT8
Thermo Scientific
MN1020, RRID: AB_223647
PHF-13
Cell Signaling
9632, RRID: AB_2266237
T22
Sigma
ABN454
HA
Sigma
H3663, RRID: AB_262051
Tau 0N2R
rPeptide
T-1005-1
Heparinase I and III
Iduron
EC 4.2.2.8 https://enzyme.expasy.org/ EC/ 4.2.2.8 EC 4.2.2.7 https://enzyme.expasy. org/EC/4.2.2.7
DHA
Cayman Chemicals
6217-54-5
Chondroitinase ABC
Sigma
C3667
Methyl-b-cyclodextrin
Sigma
C4555
Antibodies
Chemicals, Peptides, and Recombinant Proteins
2-deoxyglucose
Sigma
D8375
Heparin
Sigma
H3149
Thioflavin S
Sigma
T1892
Neutral sphingomyelinase
Sigma
C8633
Myriocin
Sigma
M1177
Cholesterol
Sigma
C3045
Ionomycin
Sigma
407953
Epigallocatechin-3-gallate (EGCG)
Sigma
50299
Emodin
Sigma
E7881
BSC3094 monohydrobromide
Sigma
B7937
Phthalocyanine tetrasulfonate hydrate (PcTS)
Sigma
P4374
NaClO3
Sigma
7775-09-9
Cytotox 96 Assay
Promega
G1780
Mouse ELISA kit
Thermo Scientific
KMB7011
CellTiter-Glo Assay
Promega
G7570
Critical Commercial Assays
Experimental Models: Cell Lines Mouse Neuroblastoma cells Neuro-2A (N2A) p. 3-19
ATCC
CCL-131 (59538655)
Rat primary cortical neurons DIV 15-25
Neuroscience Center, University of Helsinki
N/A
Tau 0N4R cDNA
Thermo Scientific
GenBank: BC114948 https://www.ncbi. nlm.nih.gov/nuccore/BC114948
phGLuc1 and phGLuc2 plasmids
Prof. Stephen Michnick, University of Montreal; Remy and Michnick, 2006
N/A
phGLuc1-Tau and phGLuc2-Tau plasmids
Yan et al., 2016
N/A
ImageJ version 1.49
NIH
https://imagej.nih.gov/ij/
GraphPad Prism version 5.08
GraphPad Software
https://www.graphpad.com/scientificsoftware/prism/
Recombinant DNA
Software and Algorithms
(Continued on next page)
e1 Cell Reports 25, 2027–2035.e1–e4, November 20, 2018
Continued REAGENT or RESOURCE
SOURCE
IDENTIFIER
SPSS Statistics version 25
IBM Analytics
https://www.ibm.com/support/ knowledgecenter/en/SSLVMB_25.0.0/ statistics_kc_ddita/spss/product_landing.html
ImageStudioLight version 5.2.5.
LI-COR Biosciences
https://www.licor.com/bio/products/ software/image_studio_lite/
CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Henri Huttunen ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS N2A cell culture Neuro-2a (N2A) mouse neuroblastoma cells (RRID:CVCL_0470) were maintained in DMEM (Corning) supplemented with 10% (v/v) of fetal bovine serum (Invitrogen) and 1% (v/v) streptomycin, penicillin and L-glutamine (Lonza) at 37 C in 5% CO2. Transfection of N2A cells was performed 24 hours after plating using JetPEI reagent (Polyplus) according to manufacturer’s instructions. Primary rat cortical neuron culture Primary cortical neurons were prepared from rat embryos as previously described (Kuja-Panula et al., 2003), and maintained in Neurobasal media (Invitrogen) supplemented with 2% (v/v) B27 (Invitrogen) and 1% (v/v) streptomycin, penicillin and L-glutamine (Lonza) at 37 C in 5% CO2. Experiments were performed between DIV14 and DIV25. METHOD DETAILS Protein-fragment complementation assay (PCA) The PCA system based on split humanized Gaussia princeps luciferase (hGLuc) used in this study has been previously described (Remy and Michnick, 2006). The hGLuc plasmids, with or without HA tag, were cloned on the C terminus of human Tau (isoform 0N4R) in a pcDNA3.1/zeo backbone (Figure S2A). PCA studies were conducted as already described (Yan et al., 2016; Brunello et al., 2016). Briefly, N2A cells on 96-well plates were transfected with phGLuc/Tau plasmids. Luminescence-based detection of PCA signal was done 48 hours post transfection, following injection of native coelenterazine, the substrate for the luciferase with Varioskan plate reader (Thermo Scientific). The detection of the PCA signal is directly proportional to the amount of Tau molecules interacting at a given time in the cells. For Tau secretion assays, Tau-Gluc expressing N2A cell media was changed to PRF-DMEM with and without treatments for indicated times before PCA measurements. For reading of PCA signal of the conditioned media, plates were spun at 400 g for 5 min in swing bucket rotor (Eppendorf) and media was collected to another plate for signal detection. Cells were placed in fresh media for PCA measurement of intracellular dimers. Both media and cell lysate were also used for measurement of lactate dehydrogenase (LDH) release to assess cell membrane integrity and cell viability. Absorbance was detected at 490 nm with Varioskan plate reader (Thermo Scientific). For Tau uptake assays, Tau-GLuc expressing N2A cells were treated with 10 mM TAIs for 16 h. Tau-conditioned media was centrifuged at 1500 g for 10 min to remove any cellular debris. Naive N2A cells (with no endogenous Tau expression) were incubated with this Tau-conditioned media for 16 hours. Tau-conditioned media was removed, cells were carefully washed twice with PBS to remove extracellular Tau and placed in fresh media for detection of PCA signal as described above and immunofluorescence analysis of internalized Tau-GLuc. For quantification, background values were subtracted from all images. Mean fluorescence intensity was measured from 18 images of EGCG-treated cells and 17 images of control treated cells (on average, 19 cells per image). The values were normalized by the number of cells per image. Media fractionation Separation of conditioned media into vesicle-free, ectosomal and exosomal fractions was performed as previously described (Thery et al., 2006; Yan et al., 2016). Briefly, conditioned media was first cleared from debris by centrifugation at 3000 x g for 30 min, and then centrifuged at 20 000 x g for 1 hour (Sorvall WX Floor Ultra centrifuge) to pellet the ectosomal fraction. While the supernatant containing the exosomal fraction was transferred to a clean tube, the pellet was washed once with fresh media and spun for one additional hour at 20 000 x g, and resuspended into 100 mL of media. In a similar way, the exosomal fraction was pelleted following two centrifugation steps at 100 000 x g for 70 min (Beckman Coulter ultracentrifuge, SW41 Ti rotor), with one washing of the pellet in
Cell Reports 25, 2027–2035.e1–e4, November 20, 2018 e2
between, before resuspension in 100 mL of fresh media. Flotillin-2 (rabbit monoclonal, Cell Signaling Technology) was used as a marker for the microvesicle-containing fractions (Kowal et al., 2016). The supernatant of the exosomal fraction, corresponding to the vesicle-free fraction of the media, was concentrated with Amicon filters (30 kDa cut-off; Millipore) according to manufacturer’s instructions. Cholesterol:mbCD complex preparation Cholesterol was first dissolved in chloroform and dried under nitrogen gas. Solution of mbCD in water (300 mM) was added to cholesterol (weight ratio of 1 to 10), vortexed and sonicated for 1 hour at RT until complete dissolution. The complex was stored at 20 C in single use aliquots. The final calculated concentration of cholesterol added to cells was 0.1 mM. ELISA For the measurement of endogenous Tau secretion to the extracellular space, the media of primary cortical neurons was changed to 250 mL of fresh Neurobasal with or without treatments for 1 hour up to 72 hours before the collection and centrifugation at 400 g for 5 min in swing bucket rotor (Eppendorf) to remove cellular debris. 50 mL of each media sample was used for the measurement of secreted Tau by a mouse-specific ELISA kit (Thermo Fisher) according to manufacturer’s instructions. The sequence identity between the 2N4R isoforms of mouse and rat Tau is 96%. The mouse Tau ELISA kit was found to detect both mouse and rat Tau. LDH release assay was performed as described above for PCA experiments. Immunofluorescence microscopy Immunofluorescence microscopy was performed as previously described (Kysenius et al., 2012). Briefly, cells were grown on glass coverslips coated with poly-L-lysine and laminin (Sigma), fixed with 4% PFA in PBS for 20 minutes, and washed with PBS before incubation for 1 hour in blocking buffer (1% BSA, 0.1% gelatin, 5% goat or donkey serum, 0.1% Triton X-100 and 0.05% Tween20 in PBS). Coverslips were then incubated overnight at +4 C with the primary antibodies diluted 1:500 in 1% BSA, 0.1% gelatin and 5% serum. The primary antibodies used in this study were Tau-5 (mouse monoclonal, Invitrogen), HA (rabbit or mouse monoclonal, Cell Signaling Technology), AT8 (mouse monoclonal, Thermo Scientific), PHF13 (mouse monoclonal, Invitrogen) and T22 (rabbit polyclonal, Sigma-Aldrich). Additional information on antibody validation can be found in Table S2 and Figure S4. The secondary antibodies Cy5 (Jackson Immunoresearch) and Alexa Fluor-conjugated (488-goat-anti-mouse and 568-donkey-antirabbit; Invitrogen) were diluted 1:2000 in PBS and incubated 1 hour at room temperature. Coverslips were finally incubated in Hoechst 33342 (Invitrogen) for nuclear staining before being mounted on microscope slides with Prolong Gold antifade reagent (Invitrogen). For the Thioflavin S staining, Thioflavin S was applied 0.01% (w/v) in 50% EtOH for 5 min, preceded and followed by two washes in 70% EtOH, prior the Hoechst 33342 incubation. Pictures were acquired with a Zeiss AxioImager M1 epifluorescence microscope and analyzed with ImageJ version 1.49 and compiled with Adobe Photoshop CS4. dSTORM Direct stochastic optical reconstruction microscopy (dSTORM) was performed as previously described (Salo et al., 2016). Briefly, cells were grown on MatTek 35 mm dishes coated with poly-L-lysine, and fixed as for regular immunofluorescence imaging. Primary antibodies Tau-5 and HA (Invitrogen) were diluted 1:200 and incubated overnight at +4 C. The secondary antibody Alexa Fluor 647anti-mouse was diluted 1:500 and incubated at room temperature for 1 hour. During imaging, cells were incubated in blinking buffer (0.1 M Tris buffer, 10% glucose, 0.07% cysteamine, 0.75 mg/ml glucose oxidase and 0.04 mg/ml catalase). The buffer was refreshed approximately every hour. Imaging was performed with Nikon Eclipse Ti-E N-STORM microscope and 30 000 images per cell were acquired. Analysis of the localization of single fluorophores activation was done with N-STORM software (Nikon), and images were further analyzed with ImageJ version 1.49. Immunoelectron microscopy Cells growing on poly-L-lysine coated coverslips were fixed with PLP fixative (McLean and Nakane, 1974) for 2 hours, washed with NaPO3 buffer and permeabilized with 0.01% saponin for 8 minutes, and incubated 1 hour in room temperature with the primary antibodies Tau-5 and HA 1:50 in NaPO3. Labeling with nano-gold-conjugated anti mouse FAB-fragments for 1 hour (Nanoprobes) was followed by quenching with 50 mM glycine and enhancement of the nano-gold particles with the HQ Silver enhancement kit (Nanoprobes) and gold toning with 2% sodium acetate, 0.05% HAuCl4 and 0.3% Na2S2O3. Epon embedding and processing of the samples for EM were performed as described (Seemann et al., 2000). Sections were cut with a Leica UCT6 microtome and images were acquired with a Jeol JEM-1400 microscope. Western blot and dot blot assay Western blotting of media fractions was performed as previously described (Yan et al., 2016). Briefly, fractions were boiled at 80 C for 10 min in NuPAGE loading buffer (Invitrogen) and 0.25% mercaptoethanol before loading on 4%–12% Bis-Tris gel (Invitrogen). Resolved proteins were transferred to a PVDF methanol-activated membrane (GE Healthcare) by semi-dry blotting (Bio-Rad). For dot blot assays, 2.3 million cells were seeded on 10-cm plates and transfected with 9 mg of phGluc-Tau/HA plasmid per well 24 hours after seeding. Culture media was changed to serum- and phenol red-free (PRF) DMEM (Invitrogen) 24 hours
e3 Cell Reports 25, 2027–2035.e1–e4, November 20, 2018
post-transfection. Conditioned media was collected 24 hours later and cleared from debris by a centrifugation at 1500 3 g. The cells were washed twice with ice-cold PBS followed by extraction on ice for 30 min in a buffer containing 10 mM Tris-HCl, pH 6.8, 1 mM EDTA, 150 mM NaCl, 0.25% Nonidet P-40, 1% Triton X-100, PhosStop phosphatase inhibitor, and protease inhibitor mixture tablets (Thermo Scientific). Cell debris was removed by a centrifugation at 13 000 3 g. 50 mL of lysate and 100 mL of media were pipetted to a 96-well Manifold Spot-Blot unit (Whatman / Schleicher and Schuell), and protein were trapped by filtration through a methanolactivated PVDF membrane (GE Healthcare). Chemiluminescence signal was detected following incubation of the membrane with primary antibodies and horseradish peroxidase-conjugated secondary antibodies and ECL Western Blotting reagent (Thermo Scientific). ImageStudioLite software was used for quantitative analysis of filter trap assay blots and the ratio between the extracellular and intracellular protein was determined. Thioflavin S assay Tau conditioned media and mock conditioned media from N2A cells were concentrated 100x with Amicon filters (30 kDa MWCO, Millipore) according to manufacturer’s instructions. 100 mL of samples were incubated with or without 0.5 mM recombinant 0N4R Tau (rPeptide), 1% protease inhibitor cocktail (Thermo Scientific) and DTT for 5 days at 37 C with shaking at 250 rpm. As a positive control, 50 mM heparin was incubated with recombinant Tau in the same conditions. For fluorescence measurement, 10 mM Thioflavin S was added to each sample, incubated 30 min and fluorescence was measured at 440 excitation, 550 emission with a Varioskan Flash multiplate reader (Thermo Scientific). Size-exclusion dot blot assay Tau-GLuc conditioned media from N2A cells was concentrated in parallel using different molecular weight cut-off filters [30 kDa, 50 kDa, 100 kDa (Amicon/Millipore) and 300 kDa (Sartorius)] by centrifugation. Concentrated retained fraction (that did not pass through the filter) was diluted with PRF-DMEM to match the volume of the pass-through fraction. Both fractions were dot blotted on a PVDF membrane and processed as described above under ‘‘Western blot and dot blot assay‘‘. The presence of Tau in the media fractions was assessed by Tau-5 staining. QUANTIFICATION AND STATISTICAL ANALYSIS All values were expressed as mean ± SEM. All statistical analyses were performed in IBM SPSS Statistics software (version 25) and the graphs were compiled in GraphPad Prism software (version 5.0a). The statistical details of experiments can be found in the figure legends. Statistical analyses were performed with unpaired, two-tailed t test or one-way ANOVA followed by Bonferroni post-test (ThioS data) or Dunnett post-test (all other experiments). The choice of one post-test or the other was determined by the need to compare all of the data point to each other, or with the control only. If not otherwise specified, ‘‘n’’ represents the number of independent replication of the entity–interventions. In PCA, each independent replica was the average of four technical replicas, while in ELISA each independent replica was the average of two technical replicas. A minimum of 3 independent replicas was used for the statistical analysis. All p values below 0.05 were considered significant. *p < 0.05, **p < 0.01, ***p < 0.001.
Cell Reports 25, 2027–2035.e1–e4, November 20, 2018 e4
Secretion of Tau via an Unconventional Nonvesicular Mechanism Graphical Abstract
Authors Maria Merezhko, Cecilia A. Brunello, Xu Yan, Helena Vihinen, Eija Jokitalo, Riikka-Liisa Uronen, Henri J. Huttunen
Correspondence [email protected]
In Brief Merezhko et al. show that Tau protein is released from cells by an unconventional secretory mechanism that can be manipulated by changing plasma membrane properties. Phosphorylated oligomeric Tau clusters at the plasma membrane, allowing its escape from cells directly through the plasma membrane.
Highlights d
ATP-independent unconventional secretion of Tau directly through the plasma membrane
d
Phosphorylated oligomeric Tau clusters in plasma membrane microdomains
d
Tau secretion can be manipulated by changing plasma membrane properties
d
Heparan sulfate proteoglycans facilitate unconventional secretion of Tau
Merezhko et al., 2018, Cell Reports 25, 2027–2035 November 20, 2018 ª 2018 The Author(s). https://doi.org/10.1016/j.celrep.2018.10.078
Cell Reports
Report Secretion of Tau via an Unconventional Non-vesicular Mechanism Maria Merezhko,1,3 Cecilia A. Brunello,1,3 Xu Yan,1 Helena Vihinen,2 Eija Jokitalo,2 Riikka-Liisa Uronen,1 and Henri J. Huttunen1,4,* 1Neuroscience
Center, HiLIFE, University of Helsinki, 00014 Helsinki, Finland Microscopy Unit, Institute of Biotechnology, HiLIFE, University of Helsinki, 00014 Helsinki, Finland 3These authors contributed equally 4Lead Contact *Correspondence: [email protected] https://doi.org/10.1016/j.celrep.2018.10.078 2Electron
SUMMARY
Tauopathies are characterized by cerebral accumulation of Tau protein aggregates that appear to spread throughout the brain via a cell-to-cell transmission process that includes secretion and uptake of pathological Tau, followed by templated misfolding of normal Tau in recipient cells. Here, we show that phosphorylated, oligomeric Tau clusters at the plasma membrane in N2A cells and is secreted in vesicle-free form in an unconventional process sensitive to changes in membrane properties, particularly cholesterol and sphingomyelin content. Cell surface heparan sulfate proteoglycans support Tau secretion, possibly by facilitating its release after membrane penetration. Notably, secretion of endogenous Tau from primary cortical neurons is mediated, at least partially, by a similar mechanism. We suggest that Tau is released from cells by an unconventional secretory mechanism that involves its phosphorylation and oligomerization and that membrane interaction may help Tau to acquire properties that allow its escape from cells directly through the plasma membrane. INTRODUCTION Accumulation of misfolded Tau into paired helical filaments and further into insoluble aggregates and neurofibrillary tangles is a hallmark of tauopathies, including Alzheimer’s disease (Lee et al., 2001). Tau is a natively unfolded neuronal microtubulestabilizing protein that, in pathological conditions, is hyperphosphorylated and detaches from microtubules, leading to misfolding and formation of pathological Tau aggregates in the cytosol. Pathological Tau is released from cells, even in the absence of cell death or membrane leakage, and internalized by neighboring cells, where the misfolded Tau templates misfolding of healthy Tau molecules, propagating Tau pathology (Goedert et al., 2017). Although this pathological spreading of Tau has gained much attention, the secretion mechanism remains unknown, although several routes, such as secretion in
extracellular vesicles (EVs), have been suggested (Saman et al., 2012; Simo´n et al., 2012; Dujardin et al., 2014). Recently, some groups have reported that only a minority of Tau is localized in extracellular vesicles, whereas the vast majority of secreted Tau is found as a vesicle-free protein (Yan et al., 2016; Wegmann et al., 2016; Wang et al., 2017). Here we show that Tau uses an unconventional secretion route that does not require ATP or Ca2+ but can be regulated by manipulating plasma membrane properties or cell surface heparan sulfate proteoglycans (HSPGs). Tau clusters in microdomains at or near the plasma membrane, independent of vesicles, and the presence of cholesterol and sphingomyelin support Tau secretion. This unconventional secretion process shares several characteristics with both overexpressed human Tau in N2A neuroblastoma cells and endogenous rat Tau in cultured cortical neurons. Based on these data, we suggest a membrane microdomain- and HSPG-dependent unconventional mechanism for Tau secretion directly through the plasma membrane. RESULTS AND DISCUSSION Overexpressed Tau Is Localized to the Plasma Membrane and Secreted Independent of Vesicles and ATP For studying the secretion mechanisms of pathological forms of Tau, we used a model where human Tau (isoform 0N4R) is overexpressed in the mouse N2A neuroblastoma cell line, leading to increased levels of free hyperphosphorylated Tau in the cytosol. We first isolated EVs from N2A/Tau cell media and found no Tau present in EVs by western blot (Figure S1A). A protein fragment complementation assay (PCA), a highly sensitive detection method for Tau dimers and oligomers (Figures S2A and S2B), showed that more than 99% of Tau oligomers were secreted in a vesicle-free form (Figure S1B). To characterize the secreted Tau species in more detail, we used a combination of size-based filter centrifugation and dot blot (Figure S2C). The majority of secreted Tau (80%) was found to be in the form of dimers, trimers, or tetramers, as shown in Figures S2D and S2E, which is in line with previous reports (Yan et al., 2016; Brunello et al., 2016). Next, we tested the behavior of secreted Tau in an in vitro Thioflavin S aggregation assay. The Tau-conditioned medium alone showed higher Thioflavin S fluorescence compared with mock-conditioned
Cell Reports 25, 2027–2035, November 20, 2018 ª 2018 The Author(s). 2027 This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
(legend on next page)
2028 Cell Reports 25, 2027–2035, November 20, 2018
medium, indicative of the presence of b sheet aggregates in the conditioned medium (Figure S2F). Moreover, the Tau-conditioned medium was capable of seeding aggregation of recombinant human Tau (0N4R). These results show that Tau secreted by N2A cells has several pathological characteristics. We then tested whether Tau is secreted by classical secretory processes. Depletion of cellular ATP by 2-deoxyglucose for 30 min had no effect on Tau secretion, strongly indicating that no vesicular transport or membrane fusion processes are involved in the secretion process (Figure 1A). No effect was seen in lactate dehydrogenase (LDH) release, demonstrating that this short treatment did not affect the membrane integrity of the cells. Treatment of cells with the canonical inducer of exocytosis, the Ca2+ ionophore ionomycin, did not induce Tau secretion but, instead, decreased it (Figure 1B), further excluding classical exocytosis as a mediator of Tau secretion. Using light microscopy, total Tau staining with Tau-5 antibody showed cytosolic staining with distinct puncta on the cell surface of N2A/Tau cells (Figure 1C). The T22 antibody, which specifically recognizes Tau oligomers (Lasagna-Reeves et al., 2012), and the phosphospecific antibodies AT8 (Ser202/Thr205) and PHF13 (Ser396) predominantly stained small puncta near the cell boundaries (Figures 1D–1F). Thioflavin S did not stain these sub-plasma membrane Tau structures (Figures 1G–1J), suggesting that they represent the phosphorylated oligomeric forms of Tau not yet converted into aggregates with b sheet structures. Using a super-resolution microscopy method, direct stochastic optical reconstruction microscopy (dSTORM) Tau-5 staining showed numerous small puncta at the edges of the cells together with some cytosolic signal (Figures 1K and 1L). Analysis of particle size showed a mean diameter of 23 nm and that more than 90% of all detected particles were smaller than 40 nm (Figure 1M). To visualize whether Tau-containing puncta were membranebound organelles, large protein aggregates, or something else, we used immunoelectron microscopy to localize wild-type untagged Tau stained with Tau-5 antibody and Tau-GLuc using anti-HA antibody. In both setups, Tau localized to the plasma membrane in microdomain-like structures but not in vesicles, multivesicular bodies, or any other membrane-bound compartments (Figures 1N–1S; Figure S3). Cytosolic Tau was observed
in both diffuse and dense aggregated forms. At the plasma membrane, we observed large (200 nm) membrane patches associated with diffuse cytosolic Tau clusters just below the plasma membrane (<100 nm) (Figures 1P and 1Q) and smaller membrane microdomains (50–80 nm) with very dense staining (Figures 1R and 1S). The only endomembrane that showed prominent Tau localization was the nuclear membrane, where Tau was localized in nuclear pores (Figures 1N and 1O). Some phenotypic variability was seen between cells, as can be expected in transient expression (quantified in Table S1). These results suggest a non-vesicular secretion mechanism for Tau, likely involving Tau localization in specific microdomains at the plasma membrane. This is in line with previous studies reporting Tau localization to membranes (Brandt et al., 1995; Kawarabayashi et al., 2004; Pooler et al., 2012) and further suggest that the interaction of Tau with the plasma membrane within specific microdomains may be a prerequisite for non-vesicular release of Tau. Structural compaction upon interaction of Tau with membranes (Elbaum-Garfinkle et al., 2010; Jones et al., €nze et al., 2012; Georgieva et al., 2014; Flach et al., 2012; Ku 2012) and/or formation of pore-like structures in membranes (Lasagna-Reeves et al., 2014; Patel et al., 2015) could mediate membrane penetration of Tau. Manipulation of Plasma Membrane Lipids Affects Tau Secretion If Tau is secreted directly through the plasma membrane, then manipulating plasma membrane properties should influence this process. We first extracted membrane cholesterol from N2A/Tau cells using methyl-b-cyclodextrin (Mahammad and Parmryd, 2015). Treatment with 1 mM methyl-b-cyclodextrin decreased Tau secretion by 47.0% ± 7.2% (Figure 2A), suggesting that cholesterol or membrane order in general affects Tau release. Loading more cholesterol to cells using preformed cholesterol:methyl-b-cyclodextrin complexes increased Tau secretion by 75.8% ± 8.1%, further supporting this view (Figure 2B). Second, we depleted the cells of sphingolipids with two different strategies. Plasma membrane sphingomyelin was degraded by adding exogenous sphingomyelinase (SMase) to the medium, or, alternatively, all sphingolipids were depleted with the serine palmitoyl-coenzyme A (CoA) transferase inhibitor myriocin.
Figure 1. Tau Is Secreted Independent of ATP or Ca2+ and Localized to the Plasma Membrane (A and B) Tau PCA and LDH release in N2A/Tau cells and media. Data are represented as mean ± SEM. (A) ATP depletion by 2-deoxyglucose treatment (30 min). ATP levels were measured by CellTiter-Glo ATP assay; n = 4–6. (B) Ionomycin treatment (30 min); n = 3. (C–J) Immunofluorescence micrographs of N2A cells transfected with Tau-GLuc1/2. Shown is staining with Tau-5 (C), T22 (D), AT8 (E), and PHF13 (F) antibodies. White arrowheads indicate punctate Tau staining near the cell surface (C–F). Also shown are (G) Tau-5, (H) AT8, (I) Thioflavin S staining, and (J) a merged picture of a single cell. White arrows indicate differential staining between different forms of Tau (G–J). (K) A dSTORM image of Tau-5 staining. The image was acquired by first finding the total internal reflection fluorescence (TIRF) plane and then focusing just above that. (L) Higher magnification of the area indicated by a white box in (K). The cell surface is indicated by a dotted white line. (M) Histogram of mean diameter of Tau-5-positive clusters observed in dSTORM. (N–S) Transmission electron microscopy. (N, P, and R) Transfection with native Tau (0N4R sequence without tags or fusions) stained with Tau-5 antibody-immunogold. (P) and (R) show higher magnification of Tau close to the plasma membrane. (O, Q, and S) Transfection with Tau-GLuc1/2 (containing an hemagglutinin [HA] tag) stained with an anti-HA antibody-immunogold. (Q) and (S) show higher magnification of Tau close to the plasma membrane. White arrowheads in (N) and (O) point to plasma membrane microdomains and black arrowheads to nucleopores. Scale bars in (C)–(K), 10 mm; in L, 100 nm, in (N) and (O), 1 mm, in (P)–(S), 100 nm. N, nucleus; Mi, mitochondrion; MVB, multivesicular body. See also Figures S1–S3 and Table S1.
Cell Reports 25, 2027–2035, November 20, 2018 2029
Figure 2. Modulation of Membrane Properties Alters Tau Secretion Shown is Tau PCA and LDH release in N2A/Tau cells and media. (A) Cholesterol extraction by methyl-b-cyclodextrin (30-min pretreatment followed by 30-min medium collection); n = 3–4. (B) Cholesterol loading by cholesterol:methyl-b-cyclodextrin (mbCD) complexes (1-hr treatment); n = 3. (C and D) Plasma membrane sphingomyelin reduction by SMase (4 hr) (C) and inhibition of sphingolipid synthesis by myriocin (16-hr pretreatment followed by 4-hr medium collection in the continued presence of myriocin) (D); n = 3–4. (E) Increased membrane fluidity by DHA (16-hr pretreatment with DHA followed by 4-hr medium collection in the continued presence of DHA); n = 3–4. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (ANOVA).
Both treatments reduced Tau secretion maximally by 36% (Figures 2C and 2D). Third, N2A/Tau cells were treated with docosahexaenoic acid (DHA), a polyunsaturated fatty acid that reduces membrane order and disrupts lipid raft-like membrane microdomains and whose consumption has been associated
2030 Cell Reports 25, 2027–2035, November 20, 2018
with reduced risk of Alzheimer’s disease (Yurko-Mauro et al., 2010; Zhang et al., 2016). Upon DHA treatment, Tau secretion was concentration-dependently reduced, with a maximal inhibition of 93.8% ± 0.9% at 50 mM DHA. Tau secretion was already significantly reduced by 1 mM DHA (Figure 2E), close to the DHA
Figure 3. HSPGs and Tau Aggregation Inhibitors Modulate Tau Secretion and Uptake (A–C) Tau PCA and LDH release in N2A/Tau cells and media. (A) Cells treated with NaClO3 (16-hr pretreatment followed by 4-hr medium collection in the continued presence of NaClO3); n = 3–5. (B) Heparinase I and heparinase III treatments (4 hr); n = 3–4. (C) Chondroitinase ABC treatment (4 hr); n = 3. (D) A representative dot blot of cells and conditioned media from N2A/Tau cells treated with emodin, BSc3094, PcTS, and EGCG at 10 mM for 16 hr, stained with Tau-5 antibody. (E) Quantification of dot blot data (n = 4). Medium dots were normalized by the corresponding cell lysate dots. (F) Tau PCA from N2A/Tau cells treated with 10 mM EGCG for 16 hr. LDH release was monitored as a control. (G–I) Tau uptake assay. (G) Medium conditioned with untransfected or Tau-transfected N2A cells, with or without 10 mM EGCG for 16 hr, was fed to naive N2A cells for 16 hr. Cells were washed extensively and then stained with Tau-5 antibodies and Hoechst 33342. Representative images are shown. Scale bar, 20 mm. (H) Quantification of Tau-5 fluorescence intensity in N2A cells in the Tau uptake assay. Mean fluorescence intensity was measured from 18 images of EGCGtreated cells and 17 images of control-treated cells (average, n = 350 cells/condition). The values were normalized by the number of cells per image. (I) Tau uptake PCA. Tau medium was collected from N2A cells expressing Tau-GLuc1/2 with or without EGCG as in (F). Naive non-transfected N2A cells were exposed to Tau-conditioned medium for 16 hr and washed extensively to remove any non-internalized Tau and EGCG. Then the level of internalized Tau oligomers was measured by PCA. (J) Tau-conditioned medium was generated as in (F) without EGCG. Naive, non-transfected N2A cells were then exposed to the Tau-conditioned medium in the presence of 10 mM EGCG or vehicle. n = 3–4. Data are represented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (ANOVA).
concentration found in normal human cerebrospinal fluid (Freund Levi et al., 2014). Cultured neuroblastoma cells typically have a significantly lower level of DHA in their membranes compared with neurons in vivo (Reynolds et al., 2001; Akbar et al., 2005), which may make them particularly sensitive to the effects of DHA.
In summary, these data indicate that the rigid liquid-ordered phase of the membrane enriched in specific lipid species, such as cholesterol and sphingomyelin, are centrally involved in Tau release from cells. Somewhat surprisingly, Tau is more efficiently released from ordered membranes, and Tau secretion can be reduced by increasing the fluidity of cellular membranes.
Cell Reports 25, 2027–2035, November 20, 2018 2031
Figure 4. Secretion of Endogenous Tau in Primary Cortical Neurons Is Modulated by Levels of Plasma Membrane Cholesterol and HSPGs and by EGCG (A–D) Primary cultures of rat cortical neurons were treated at 14–25 DIV with increasing concentrations of (A) methyl-b-cyclodextrin for 16 hr, (B) cholesterol:methyl-b-cyclodextrin for 1 hr, (C) NaClO3 for 72 hr, and (D) EGCG for 16 hr. Cells and conditioned media were analyzed for Tau using ELISA and cell viability and membrane integrity using an LDH assay. n = 3–4. Data are represented as mean ± SEM. **p < 0.01, ***p < 0.001 (ANOVA).
Interestingly, a similar interdependence of membrane composition and rigidity and membrane permeation was recently reported for arginine-rich cell-penetrating peptides (Wallbrecher et al., 2017). Cell Surface HSPGs and the Aggregation State of Tau Modulate Secretion Efficiency Because Tau binds glycosaminoglycans (GAGs) (Goedert et al., 1996; Hasegawa et al., 1997) that are abundantly present at the plasma membrane, we hypothesized that these interactions may play a role in Tau secretion. HSPGs play a central role in the unconventional secretion of FGF2 (Steringer and Nickel, 2018). Moreover, while this manuscript was under review, Katsinelos et al. (2018) published a paper suggesting that Tau secretion may share some similarities with unconventional secretion of FGF2. NaClO3 is widely used to inhibit synthesis of PAPS (30 -phosphoadenosine 50 -phosphosulphate), the general sulfate donor in GAG biosynthesis (Venkatachalam, 2003). Treatment of N2A/Tau cells with NaClO3 resulted in a dose-dependent decrease in Tau secretion (maximal decrease, 37.6% ± 5.4% 2032 Cell Reports 25, 2027–2035, November 20, 2018
at 50 mM NaClO3) (Figure 3A). To reduce the amount of cell surface HSPGs more specifically, we treated cells with heparinase I or heparinase III, two enzymes that cleave heparin and heparan sulfate-type sulfated glycans with different specificities. Both treatments led to a decrease in both intracellular and secreted Tau, with heparinase III having a more consistent effect of reducing Tau secretion by 25.0% ± 3.5% (Figure 3B). Treatment of N2A/Tau cells with chondroitinase ABC to reduce cell surface levels of chondroitin and dermatan sulfate GAGs did not decrease Tau secretion (Figure 3C). These data suggest that cell surface proteoglycans, specifically HSPGs, facilitate Tau secretion and are in line with a previous report showing similar results (Katsinelos et al., 2018). Because oligomerization plays an important role in unconventional secretion of FGF2 (Steringer and Nickel, 2018), and because at least some of the plasma membrane-localized Tau appears to be hyperphosphorylated and oligomerized, we next tested the effect of inhibition of Tau aggregation on secretion. To narrow down the Tau species that penetrate the membrane, we used four different Tau aggregation inhibitors
(TAIs), emodin (Pickhardt et al., 2005), BSc3094 (Pickhardt et al., 2007), phthalocyanine tetrasulfonate (PcTS) (Akoury et al., 2013), and epigallocatechin gallate (EGCG) (Ehrnhoefer et al., 2008; Wobst et al., 2015), which inhibit Tau aggregation at different stages (Bulic et al., 2013; Cisek et al., 2014). Because some of the TAIs are colored, they prevent the use of PCA. Instead, we used a dot blot to study Tau secretion. All four compounds reduced Tau secretion, with effects ranging from 35% to 94% reduction (Figures 3D and 3E). This suggests that formation of oligomeric aggregation intermediates, possibly occurring at the membrane interface, is a prerequisite for unconventional secretion of Tau. EGCG caused the strongest reduction in Tau secretion in the dot blot assay, and because it is a non-colored compound, it was also tested in the PCA-based secretion assay. EGCG reduced secreted Tau oligomers, as detected by PCA, by 63.4% ± 3.6% (Figure 3F). Furthermore, we found that Tau secreted in the presence of EGCG is taken up significantly less effectively compared with the control (Figures 3G–3I). Notably, EGCG did not reduce Tau uptake in recipient cells when added to the medium post-secretion; i.e., after the conditioned medium was removed from the donor cells (Figure 3J). Because EGCG not only inhibits Tau aggregation but also modulates membrane properties (Sun et al., 2009; Patra et al., 2008), both of these mechanisms could contribute to the robust overall effect of EGCG on Tau secretion in N2A cells. Overall, together with the finding that phosphorylated oligomeric forms of Tau localize in clusters in sub-plasma membrane microdomains (Figure 1), these results suggest that certain forms of oligomeric aggregation intermediates may be the Tau species capable of penetrating the plasma membrane and exiting cells. Endogenous Tau from Primary Neurons Is Also Secreted via a Similar Unconventional Mechanism To test whether endogenous neuronal Tau is also secreted via the same or a similar mechanism as overexpressed pathological Tau, we analyzed media from cultures of rat primary cortical neurons (21 days in vitro [DIV]) using Tau ELISA. Similar to N2A/Tau cells, Tau secretion was inhibited by decreasing plasma membrane cholesterol with methyl-b-cyclodextrin, with a maximal decrease of 47.8% ± 0.6% at 1 mM methylb-cyclodextrin (Figure 4A). Increasing neuronal plasma membrane cholesterol content by cholesterol:methyl-b-cyclodextrin complexes resulted in a 105% ± 15.6% increase in Tau secretion (Figure 4B). Reduction of sulfated GAGs on the cell surface by treatment of cortical neurons with 50 mM NaClO3 decreased Tau secretion by 45.0% ± 5.3% (Figure 4C). EGCG reduced neuronal secretion of Tau by 35.7% ± 7.1% (Figure 4D). LDH release was not significantly altered by any of these treatments, suggesting that plasma membrane integrity was not affected. These results indicate that not only overexpressed pathological forms of human Tau in N2A cells but also endogenous rat Tau in primary cortical neurons are secreted by an unconventional secretion mechanism. In summary, our current data demonstrate that Tau uses an unconventional protein secretion pathway to exit cells. Importantly, overexpressed pathological Tau and also endogenous Tau from primary neurons appear to be secreted via a similar
mechanism. This process is highly dependent on plasma membrane lipid content and biophysical properties, especially in the presence of cholesterol and sphingomyelin. A number of proteins are released from eukaryotic cells under physiological conditions directly through the plasma membrane in an unconventional secretion process (Rabouille, 2017). We propose that Tau acquires properties, likely associated with its phosphorylation and oligomerization, that promote its plasma membrane interaction and permeation, resulting in secretion. Suppression of plasma membrane interaction and penetration of Tau could be an effective therapeutic target for slowing down disease progression in tauopathies. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d d d
d
d
KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS B N2A cell culture B Primary rat cortical neuron culture METHOD DETAILS B Protein-fragment complementation assay (PCA) B Media fractionation B Cholesterol:mbCD complex preparation B ELISA B Immunofluorescence microscopy B dSTORM B Immunoelectron microscopy B Western blot and dot blot assay B Thioflavin S assay B Size-exclusion dot blot assay QUANTIFICATION AND STATISTICAL ANALYSIS
SUPPLEMENTAL INFORMATION Supplemental Information includes four figures and two tables and can be found with this article online at https://doi.org/10.1016/j.celrep.2018.10.078. ACKNOWLEDGMENTS We thank Prof. Stephan Michnick (Universite´ de Montre´al, Canada) for providing the humanized GLuc plasmid. Dr. Veijo Salo, Amr Abou Elezz, Mikko Liljestro¨m, and the Biomedicum Imaging Unit, University of Helsinki are thanked for technical support with dSTORM imaging. This study was funded by the Academy of Finland (grant 296409) and the Finnish Cultural Foundation. AUTHOR CONTRIBUTIONS Conceptualization, M.M., C.A.B., X.Y., R.-L.U., and H.J.H.; Methodology, M.M., C.A.B., X.Y., H.V., E.J., and H.J.H.; Investigation, M.M., C.A.B., and H.V.; Writing – Original Draft, R.-L.U. and H.J.H.; Writing – Review & Editing, M.M., C.A.B., R.-L.U., and H.J.H.; Funding Acquisition, R.-L.U. and H.J.H.; Supervision, E.J., R.-L.U., and H.J.H. DECLARATION OF INTERESTS H.J.H. is an employee and shareholder of Herantis Pharma Plc, which is unrelated to this study.
Cell Reports 25, 2027–2035, November 20, 2018 2033
Received: April 18, 2018 Revised: September 26, 2018 Accepted: October 19, 2018 Published: November 20, 2018 SUPPORTING CITATIONS The following references appear in the Supplemental Information: Ercan et al. (2017); Hoffmann et al. (1997); LoPresti et al. (1995); Mercken et al. (1992); Niman et al. (1983); Wilson et al. (1984). REFERENCES Akbar, M., Calderon, F., Wen, Z., and Kim, H.Y. (2005). Docosahexaenoic acid: a positive modulator of Akt signaling in neuronal survival. Proc. Natl. Acad. Sci. USA 102, 10858–10863. Akoury, E., Gajda, M., Pickhardt, M., Biernat, J., Soraya, P., Griesinger, C., Mandelkow, E., and Zweckstetter, M. (2013). Inhibition of tau filament formation by conformational modulation. J. Am. Chem. Soc. 135, 2853–2862. Brandt, R., Le´ger, J., and Lee, G. (1995). Interaction of tau with the neural plasma membrane mediated by tau’s amino-terminal projection domain. J. Cell Biol. 131, 1327–1340. Brunello, C.A., Yan, X., and Huttunen, H.J. (2016). Internalized Tau sensitizes cells to stress by promoting formation and stability of stress granules. Sci. Rep. 6, 30498. Bulic, B., Pickhardt, M., and Mandelkow, E. (2013). Progress and developments in tau aggregation inhibitors for Alzheimer disease. J. Med. Chem. 56, 4135–4155. Cisek, K., Cooper, G.L., Huseby, C.J., and Kuret, J. (2014). Structure and mechanism of action of tau aggregation inhibitors. Curr. Alzheimer Res. 11, 918–927. Dujardin, S., Be´gard, S., Caillierez, R., Lachaud, C., Delattre, L., Carrier, S., Loyens, A., Galas, M.C., Bousset, L., Melki, R., et al. (2014). Ectosomes: a new mechanism for non-exosomal secretion of tau protein. PLoS ONE 9, e100760. Ehrnhoefer, D.E., Bieschke, J., Boeddrich, A., Herbst, M., Masino, L., Lurz, R., Engemann, S., Pastore, A., and Wanker, E.E. (2008). EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers. Nat. Struct. Mol. Biol. 15, 558–566. Elbaum-Garfinkle, S., Ramlall, T., and Rhoades, E. (2010). The role of the lipid bilayer in tau aggregation. Biophys. J. 98, 2722–2730. Ercan, E., Eid, S., Weber, C., Kowalski, A., Bichmann, M., Behrendt, A., Matthes, F., Krauss, S., Reinhardt, P., Fulle, S., and Ehrnhoefer, D.E. (2017). A validated antibody panel for the characterization of tau post-translational modifications. Mol. Neurodegener. 12, 87.
Hasegawa, M., Crowther, R.A., Jakes, R., and Goedert, M. (1997). Alzheimerlike changes in microtubule-associated protein Tau induced by sulfated glycosaminoglycans. Inhibition of microtubule binding, stimulation of phosphorylation, and filament assembly depend on the degree of sulfation. J. Biol. Chem. 272, 33118–33124. Hoffmann, R., Dawson, N.F., Wade, J.D., and Otvo¨s, L., Jr. (1997). Oxidized and phosphorylated synthetic peptides corresponding to the second and third tubulin-binding repeats of the tau protein reveal structural features of paired helical filament assembly. J. Pept. Res. 50, 132–142. Jones, E.M., Dubey, M., Camp, P.J., Vernon, B.C., Biernat, J., Mandelkow, E., Majewski, J., and Chi, E.Y. (2012). Interaction of tau protein with model lipid membranes induces tau structural compaction and membrane disruption. Biochemistry 51, 2539–2550. €ller, H.M., Nachman, Katsinelos, T., Zeitler, M., Dimou, E., Karakatsani, A., Mu E., Steringer, J.P., Ruiz de Almodovar, C., Nickel, W., and Jahn, T.R. (2018). Unconventional Secretion Mediates the Trans-cellular Spreading of Tau. Cell Rep. 23, 2039–2055. Kawarabayashi, T., Shoji, M., Younkin, L.H., Wen-Lang, L., Dickson, D.W., Murakami, T., Matsubara, E., Abe, K., Ashe, K.H., and Younkin, S.G. (2004). Dimeric amyloid beta protein rapidly accumulates in lipid rafts followed by apolipoprotein E and phosphorylated tau accumulation in the Tg2576 mouse model of Alzheimer’s disease. J. Neurosci. 24, 3801–3809. Kowal, J., Arras, G., Colombo, M., Jouve, M., Morath, J.P., Primdal-Bengtson, B., Dingli, F., Loew, D., Tkach, M., and The´ry, C. (2016). Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. Proc. Natl. Acad. Sci. USA 113, E968–E977. Kuja-Panula, J., Kiiltoma¨ki, M., Yamashiro, T., Rouhiainen, A., and Rauvala, H. (2003). AMIGO, a transmembrane protein implicated in axon tract development, defines a novel protein family with leucine-rich repeats. J. Cell Biol. 160, 963–973. €nze, G., Barre´, P., Scheidt, H.A., Thomas, L., Eliezer, D., and Huster, D. Ku (2012). Binding of the three-repeat domain of tau to phospholipid membranes induces an aggregated-like state of the protein. Biochim. Biophys. Acta 1818, 2302–2313. Kysenius, K., Muggalla, P., Ma¨tlik, K., Aruma¨e, U., and Huttunen, H.J. (2012). PCSK9 regulates neuronal apoptosis by adjusting ApoER2 levels and signaling. Cell. Mol. Life Sci. 69, 1903–1916. Lasagna-Reeves, C.A., Castillo-Carranza, D.L., Sengupta, U., Sarmiento, J., Troncoso, J., Jackson, G.R., and Kayed, R. (2012). Identification of oligomers at early stages of tau aggregation in Alzheimer’s disease. FASEB J. 26, 1946– 1959. Lasagna-Reeves, C.A., Sengupta, U., Castillo-Carranza, D., Gerson, J.E., Guerrero-Munoz, M., Troncoso, J.C., Jackson, G.R., and Kayed, R. (2014). The formation of tau pore-like structures is prevalent and cell specific: possible implications for the disease phenotypes. Acta Neuropathol. Commun. 2, 56.
€ger, M., Leonhardt, M., Flach, K., Hilbrich, I., Schiffmann, A., Ga¨rtner, U., Kru Waschipky, H., Wick, L., Arendt, T., and Holzer, M. (2012). Tau oligomers impair artificial membrane integrity and cellular viability. J. Biol. Chem. 287, 43223–43233.
Lee, V.M., Goedert, M., and Trojanowski, J.Q. (2001). Neurodegenerative tauopathies. Annu. Rev. Neurosci. 24, 1121–1159.
Freund Levi, Y., Vedin, I., Cederholm, T., Basun, H., Faxe´n Irving, G., Eriksdotter, M., Hjorth, E., Schultzberg, M., Vessby, B., Wahlund, L.O., et al. (2014). Transfer of omega-3 fatty acids across the blood-brain barrier after dietary supplementation with a docosahexaenoic acid-rich omega-3 fatty acid preparation in patients with Alzheimer’s disease: the OmegAD study. J. Intern. Med. 275, 428–436.
Mahammad, S., and Parmryd, I. (2015). Cholesterol depletion using methylb-cyclodextrin. Methods Mol. Biol. 1232, 91–102.
Georgieva, E.R., Xiao, S., Borbat, P.P., Freed, J.H., and Eliezer, D. (2014). Tau binds to lipid membrane surfaces via short amphipathic helices located in its microtubule-binding repeats. Biophys. J. 107, 1441–1452. Goedert, M., Jakes, R., Spillantini, M.G., Hasegawa, M., Smith, M.J., and Crowther, R.A. (1996). Assembly of microtubule-associated protein tau into Alzheimer-like filaments induced by sulphated glycosaminoglycans. Nature 383, 550–553. Goedert, M., Eisenberg, D.S., and Crowther, R.A. (2017). Propagation of Tau Aggregates and Neurodegeneration. Annu. Rev. Neurosci. 40, 189–210.
2034 Cell Reports 25, 2027–2035, November 20, 2018
LoPresti, P., Szuchet, S., Papasozomenos, S.C., Zinkowski, R.P., and Binder, L.I. (1995). Functional implications for the microtubule-associated protein tau: localization in oligodendrocytes. Proc. Natl. Acad. Sci. USA 92, 10369–10373.
McLean, I.W., and Nakane, P.K. (1974). Periodate-lysine-paraformaldehyde fixative. A new fixation for immunoelectron microscopy. J. Histochem. Cytochem 22, 1077–1083. €bke, U., Six, J., Boons, J., Van de Voorde, Mercken, M., Vandermeeren, M., Lu A., Martin, J.J., and Gheuens, J. (1992). Monoclonal antibodies with selective specificity for Alzheimer Tau are directed against phosphatase-sensitive epitopes. Acta Neuropathol. 84, 265–272. Niman, H.L., Houghten, R.A., Walker, L.E., Reisfeld, R.A., Wilson, I.A., Hogle, J.M., and Lerner, R.A. (1983). Generation of protein-reactive antibodies by short peptides is an event of high frequency: implications for the structural basis of immune recognition. Proc. Natl. Acad. Sci. USA 80, 4949–4953.
Patel, N., Ramachandran, S., Azimov, R., Kagan, B.L., and Lal, R. (2015). Ion Channel Formation by Tau Protein: Implications for Alzheimer’s Disease and Tauopathies. Biochemistry 54, 7320–7325.
Sun, Y., Hung, W.C., Chen, F.Y., Lee, C.C., and Huang, H.W. (2009). Interaction of tea catechin (-)-epigallocatechin gallate with lipid bilayers. Biophys. J. 96, 1026–1035.
Patra, S.K., Rizzi, F., Silva, A., Rugina, D.O., and Bettuzzi, S. (2008). Molecular targets of (-)-epigallocatechin-3-gallate (EGCG): specificity and interaction with membrane lipid rafts. J. Physiol. Pharmacol. 59 (Suppl 9), 217–235.
Thery, C., Amigorena, S., Raposo, G., and Clayton, A. (2006). Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr. Protoc. Cell Biol. 30, 3.22.1–3.22.29.
Pickhardt, M., Gazova, Z., von Bergen, M., Khlistunova, I., Wang, Y., Hascher, A., Mandelkow, E.M., Biernat, J., and Mandelkow, E. (2005). Anthraquinones inhibit tau aggregation and dissolve Alzheimer’s paired helical filaments in vitro and in cells. J. Biol. Chem. 280, 3628–3635.
Venkatachalam, K.V. (2003). Human 30 -phosphoadenosine 50 -phosphosulfate (PAPS) synthase: biochemistry, molecular biology and genetic deficiency. IUBMB Life 55, 1–11.
Pickhardt, M., Larbig, G., Khlistunova, I., Coksezen, A., Meyer, B., Mandelkow, E.M., Schmidt, B., and Mandelkow, E. (2007). Phenylthiazolyl-hydrazide and its derivatives are potent inhibitors of tau aggregation and toxicity in vitro and in cells. Biochemistry 46, 10016–10023. Pooler, A.M., Usardi, A., Evans, C.J., Philpott, K.L., Noble, W., and Hanger, D.P. (2012). Dynamic association of tau with neuronal membranes is regulated by phosphorylation. Neurobiol. Aging 33, 431.e27–431.e38. Rabouille, C. (2017). Pathways of Unconventional Protein Secretion. Trends Cell Biol. 27, 230–240. Remy, I., and Michnick, S.W. (2006). A highly sensitive protein-protein interaction assay based on Gaussia luciferase. Nat. Methods 3, 977–979. Reynolds, L.M., Dalton, C.F., and Reynolds, G.P. (2001). Phospholipid fatty acids and neurotoxicity in human neuroblastoma SH-SY5Y cells. Neurosci. Lett. 309, 193–196. Salo, V.T., Belevich, I., Li, S., Karhinen, L., Vihinen, H., Vigouroux, C., Magre´, J., Thiele, C., Ho¨ltta¨-Vuori, M., Jokitalo, E., and Ikonen, E. (2016). Seipin regulates ER-lipid droplet contacts and cargo delivery. EMBO J. 35, 2699–2716. Saman, S., Kim, W., Raya, M., Visnick, Y., Miro, S., Saman, S., Jackson, B., McKee, A.C., Alvarez, V.E., Lee, N.C., and Hall, G.F. (2012). Exosome-associated tau is secreted in tauopathy models and is selectively phosphorylated in cerebrospinal fluid in early Alzheimer disease. J. Biol. Chem. 287, 3842–3849. Seemann, J., Jokitalo, E.J., and Warren, G. (2000). The role of the tethering proteins p115 and GM130 in transport through the Golgi apparatus in vivo. Mol. Biol. Cell 11, 635–645. Simo´n, D., Garcı´a-Garcı´a, E., Royo, F., Falco´n-Pe´rez, J.M., and Avila, J. (2012). Proteostasis of tau. Tau overexpression results in its secretion via membrane vesicles. FEBS Lett. 586, 47–54. Steringer, J.P., and Nickel, W. (2018). A direct gateway into the extracellular space: Unconventional secretion of FGF2 through self-sustained plasma membrane pores. Semin. Cell Dev. Biol. 83, 3–7.
Wallbrecher, R., Ackels, T., Olea, R.A., Klein, M.J., Caillon, L., Schiller, J., Bove´e-Geurts, P.H., van Kuppevelt, T.H., Ulrich, A.S., Spehr, M., et al. (2017). Membrane permeation of arginine-rich cell-penetrating peptides independent of transmembrane potential as a function of lipid composition and membrane fluidity. J. Control. Release 256, 68–78. €ger, L., Irsen, S., Tepper, K., ChanduWang, Y., Balaji, V., Kaniyappan, S., Kru patla, R., Maetzler, W., Schneider, A., Mandelkow, E., and Mandelkow, E.M. (2017). The release and trans-synaptic transmission of Tau via exosomes. Mol. Neurodegener. 12, 5. Wegmann, S., Nicholls, S., Takeda, S., Fan, Z., and Hyman, B.T. (2016). Formation, release, and internalization of stable tau oligomers in cells. J. Neurochem. 139, 1163–1174. Wilson, I.A., Niman, H.L., Houghten, R.A., Cherenson, A.R., Connolly, M.L., and Lerner, R.A. (1984). The structure of an antigenic determinant in a protein. Cell 37, 767–778. Wobst, H.J., Sharma, A., Diamond, M.I., Wanker, E.E., and Bieschke, J. (2015). The green tea polyphenol (-)-epigallocatechin gallate prevents the aggregation of tau protein into toxic oligomers at substoichiometric ratios. FEBS Lett. 589, 77–83. Yan, X., Nyka¨nen, N.P., Brunello, C.A., Haapasalo, A., Hiltunen, M., Uronen, R.L., and Huttunen, H.J. (2016). FRMD4A-cytohesin signaling modulates the cellular release of tau. J. Cell Sci. 129, 2003–2015. Yurko-Mauro, K., McCarthy, D., Rom, D., Nelson, E.B., Ryan, A.S., Blackwell, A., Salem, N., Jr., and Stedman, M.; MIDAS Investigators (2010). Beneficial effects of docosahexaenoic acid on cognition in age-related cognitive decline. Alzheimers Dement. 6, 456–464. Zhang, Y., Chen, J., Qiu, J., Li, Y., Wang, J., and Jiao, J. (2016). Intakes of fish and polyunsaturated fatty acids and mild-to-severe cognitive impairment risks: a dose-response meta-analysis of 21 cohort studies. Am. J. Clin. Nutr. 103, 330–340.
Cell Reports 25, 2027–2035, November 20, 2018 2035
STAR+METHODS KEY RESOURCES TABLE
REAGENT or RESOURCE
SOURCE
IDENTIFIER
Tau-5
Invitrogen
AHB0042, RRID: AB_2536235
AT8
Thermo Scientific
MN1020, RRID: AB_223647
PHF-13
Cell Signaling
9632, RRID: AB_2266237
T22
Sigma
ABN454
HA
Sigma
H3663, RRID: AB_262051
Tau 0N2R
rPeptide
T-1005-1
Heparinase I and III
Iduron
EC 4.2.2.8 https://enzyme.expasy.org/ EC/ 4.2.2.8 EC 4.2.2.7 https://enzyme.expasy. org/EC/4.2.2.7
DHA
Cayman Chemicals
6217-54-5
Chondroitinase ABC
Sigma
C3667
Methyl-b-cyclodextrin
Sigma
C4555
Antibodies
Chemicals, Peptides, and Recombinant Proteins
2-deoxyglucose
Sigma
D8375
Heparin
Sigma
H3149
Thioflavin S
Sigma
T1892
Neutral sphingomyelinase
Sigma
C8633
Myriocin
Sigma
M1177
Cholesterol
Sigma
C3045
Ionomycin
Sigma
407953
Epigallocatechin-3-gallate (EGCG)
Sigma
50299
Emodin
Sigma
E7881
BSC3094 monohydrobromide
Sigma
B7937
Phthalocyanine tetrasulfonate hydrate (PcTS)
Sigma
P4374
NaClO3
Sigma
7775-09-9
Cytotox 96 Assay
Promega
G1780
Mouse ELISA kit
Thermo Scientific
KMB7011
CellTiter-Glo Assay
Promega
G7570
Critical Commercial Assays
Experimental Models: Cell Lines Mouse Neuroblastoma cells Neuro-2A (N2A) p. 3-19
ATCC
CCL-131 (59538655)
Rat primary cortical neurons DIV 15-25
Neuroscience Center, University of Helsinki
N/A
Tau 0N4R cDNA
Thermo Scientific
GenBank: BC114948 https://www.ncbi. nlm.nih.gov/nuccore/BC114948
phGLuc1 and phGLuc2 plasmids
Prof. Stephen Michnick, University of Montreal; Remy and Michnick, 2006
N/A
phGLuc1-Tau and phGLuc2-Tau plasmids
Yan et al., 2016
N/A
ImageJ version 1.49
NIH
https://imagej.nih.gov/ij/
GraphPad Prism version 5.08
GraphPad Software
https://www.graphpad.com/scientificsoftware/prism/
Recombinant DNA
Software and Algorithms
(Continued on next page)
e1 Cell Reports 25, 2027–2035.e1–e4, November 20, 2018
Continued REAGENT or RESOURCE
SOURCE
IDENTIFIER
SPSS Statistics version 25
IBM Analytics
https://www.ibm.com/support/ knowledgecenter/en/SSLVMB_25.0.0/ statistics_kc_ddita/spss/product_landing.html
ImageStudioLight version 5.2.5.
LI-COR Biosciences
https://www.licor.com/bio/products/ software/image_studio_lite/
CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Henri Huttunen ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS N2A cell culture Neuro-2a (N2A) mouse neuroblastoma cells (RRID:CVCL_0470) were maintained in DMEM (Corning) supplemented with 10% (v/v) of fetal bovine serum (Invitrogen) and 1% (v/v) streptomycin, penicillin and L-glutamine (Lonza) at 37 C in 5% CO2. Transfection of N2A cells was performed 24 hours after plating using JetPEI reagent (Polyplus) according to manufacturer’s instructions. Primary rat cortical neuron culture Primary cortical neurons were prepared from rat embryos as previously described (Kuja-Panula et al., 2003), and maintained in Neurobasal media (Invitrogen) supplemented with 2% (v/v) B27 (Invitrogen) and 1% (v/v) streptomycin, penicillin and L-glutamine (Lonza) at 37 C in 5% CO2. Experiments were performed between DIV14 and DIV25. METHOD DETAILS Protein-fragment complementation assay (PCA) The PCA system based on split humanized Gaussia princeps luciferase (hGLuc) used in this study has been previously described (Remy and Michnick, 2006). The hGLuc plasmids, with or without HA tag, were cloned on the C terminus of human Tau (isoform 0N4R) in a pcDNA3.1/zeo backbone (Figure S2A). PCA studies were conducted as already described (Yan et al., 2016; Brunello et al., 2016). Briefly, N2A cells on 96-well plates were transfected with phGLuc/Tau plasmids. Luminescence-based detection of PCA signal was done 48 hours post transfection, following injection of native coelenterazine, the substrate for the luciferase with Varioskan plate reader (Thermo Scientific). The detection of the PCA signal is directly proportional to the amount of Tau molecules interacting at a given time in the cells. For Tau secretion assays, Tau-Gluc expressing N2A cell media was changed to PRF-DMEM with and without treatments for indicated times before PCA measurements. For reading of PCA signal of the conditioned media, plates were spun at 400 g for 5 min in swing bucket rotor (Eppendorf) and media was collected to another plate for signal detection. Cells were placed in fresh media for PCA measurement of intracellular dimers. Both media and cell lysate were also used for measurement of lactate dehydrogenase (LDH) release to assess cell membrane integrity and cell viability. Absorbance was detected at 490 nm with Varioskan plate reader (Thermo Scientific). For Tau uptake assays, Tau-GLuc expressing N2A cells were treated with 10 mM TAIs for 16 h. Tau-conditioned media was centrifuged at 1500 g for 10 min to remove any cellular debris. Naive N2A cells (with no endogenous Tau expression) were incubated with this Tau-conditioned media for 16 hours. Tau-conditioned media was removed, cells were carefully washed twice with PBS to remove extracellular Tau and placed in fresh media for detection of PCA signal as described above and immunofluorescence analysis of internalized Tau-GLuc. For quantification, background values were subtracted from all images. Mean fluorescence intensity was measured from 18 images of EGCG-treated cells and 17 images of control treated cells (on average, 19 cells per image). The values were normalized by the number of cells per image. Media fractionation Separation of conditioned media into vesicle-free, ectosomal and exosomal fractions was performed as previously described (Thery et al., 2006; Yan et al., 2016). Briefly, conditioned media was first cleared from debris by centrifugation at 3000 x g for 30 min, and then centrifuged at 20 000 x g for 1 hour (Sorvall WX Floor Ultra centrifuge) to pellet the ectosomal fraction. While the supernatant containing the exosomal fraction was transferred to a clean tube, the pellet was washed once with fresh media and spun for one additional hour at 20 000 x g, and resuspended into 100 mL of media. In a similar way, the exosomal fraction was pelleted following two centrifugation steps at 100 000 x g for 70 min (Beckman Coulter ultracentrifuge, SW41 Ti rotor), with one washing of the pellet in
Cell Reports 25, 2027–2035.e1–e4, November 20, 2018 e2
between, before resuspension in 100 mL of fresh media. Flotillin-2 (rabbit monoclonal, Cell Signaling Technology) was used as a marker for the microvesicle-containing fractions (Kowal et al., 2016). The supernatant of the exosomal fraction, corresponding to the vesicle-free fraction of the media, was concentrated with Amicon filters (30 kDa cut-off; Millipore) according to manufacturer’s instructions. Cholesterol:mbCD complex preparation Cholesterol was first dissolved in chloroform and dried under nitrogen gas. Solution of mbCD in water (300 mM) was added to cholesterol (weight ratio of 1 to 10), vortexed and sonicated for 1 hour at RT until complete dissolution. The complex was stored at 20 C in single use aliquots. The final calculated concentration of cholesterol added to cells was 0.1 mM. ELISA For the measurement of endogenous Tau secretion to the extracellular space, the media of primary cortical neurons was changed to 250 mL of fresh Neurobasal with or without treatments for 1 hour up to 72 hours before the collection and centrifugation at 400 g for 5 min in swing bucket rotor (Eppendorf) to remove cellular debris. 50 mL of each media sample was used for the measurement of secreted Tau by a mouse-specific ELISA kit (Thermo Fisher) according to manufacturer’s instructions. The sequence identity between the 2N4R isoforms of mouse and rat Tau is 96%. The mouse Tau ELISA kit was found to detect both mouse and rat Tau. LDH release assay was performed as described above for PCA experiments. Immunofluorescence microscopy Immunofluorescence microscopy was performed as previously described (Kysenius et al., 2012). Briefly, cells were grown on glass coverslips coated with poly-L-lysine and laminin (Sigma), fixed with 4% PFA in PBS for 20 minutes, and washed with PBS before incubation for 1 hour in blocking buffer (1% BSA, 0.1% gelatin, 5% goat or donkey serum, 0.1% Triton X-100 and 0.05% Tween20 in PBS). Coverslips were then incubated overnight at +4 C with the primary antibodies diluted 1:500 in 1% BSA, 0.1% gelatin and 5% serum. The primary antibodies used in this study were Tau-5 (mouse monoclonal, Invitrogen), HA (rabbit or mouse monoclonal, Cell Signaling Technology), AT8 (mouse monoclonal, Thermo Scientific), PHF13 (mouse monoclonal, Invitrogen) and T22 (rabbit polyclonal, Sigma-Aldrich). Additional information on antibody validation can be found in Table S2 and Figure S4. The secondary antibodies Cy5 (Jackson Immunoresearch) and Alexa Fluor-conjugated (488-goat-anti-mouse and 568-donkey-antirabbit; Invitrogen) were diluted 1:2000 in PBS and incubated 1 hour at room temperature. Coverslips were finally incubated in Hoechst 33342 (Invitrogen) for nuclear staining before being mounted on microscope slides with Prolong Gold antifade reagent (Invitrogen). For the Thioflavin S staining, Thioflavin S was applied 0.01% (w/v) in 50% EtOH for 5 min, preceded and followed by two washes in 70% EtOH, prior the Hoechst 33342 incubation. Pictures were acquired with a Zeiss AxioImager M1 epifluorescence microscope and analyzed with ImageJ version 1.49 and compiled with Adobe Photoshop CS4. dSTORM Direct stochastic optical reconstruction microscopy (dSTORM) was performed as previously described (Salo et al., 2016). Briefly, cells were grown on MatTek 35 mm dishes coated with poly-L-lysine, and fixed as for regular immunofluorescence imaging. Primary antibodies Tau-5 and HA (Invitrogen) were diluted 1:200 and incubated overnight at +4 C. The secondary antibody Alexa Fluor 647anti-mouse was diluted 1:500 and incubated at room temperature for 1 hour. During imaging, cells were incubated in blinking buffer (0.1 M Tris buffer, 10% glucose, 0.07% cysteamine, 0.75 mg/ml glucose oxidase and 0.04 mg/ml catalase). The buffer was refreshed approximately every hour. Imaging was performed with Nikon Eclipse Ti-E N-STORM microscope and 30 000 images per cell were acquired. Analysis of the localization of single fluorophores activation was done with N-STORM software (Nikon), and images were further analyzed with ImageJ version 1.49. Immunoelectron microscopy Cells growing on poly-L-lysine coated coverslips were fixed with PLP fixative (McLean and Nakane, 1974) for 2 hours, washed with NaPO3 buffer and permeabilized with 0.01% saponin for 8 minutes, and incubated 1 hour in room temperature with the primary antibodies Tau-5 and HA 1:50 in NaPO3. Labeling with nano-gold-conjugated anti mouse FAB-fragments for 1 hour (Nanoprobes) was followed by quenching with 50 mM glycine and enhancement of the nano-gold particles with the HQ Silver enhancement kit (Nanoprobes) and gold toning with 2% sodium acetate, 0.05% HAuCl4 and 0.3% Na2S2O3. Epon embedding and processing of the samples for EM were performed as described (Seemann et al., 2000). Sections were cut with a Leica UCT6 microtome and images were acquired with a Jeol JEM-1400 microscope. Western blot and dot blot assay Western blotting of media fractions was performed as previously described (Yan et al., 2016). Briefly, fractions were boiled at 80 C for 10 min in NuPAGE loading buffer (Invitrogen) and 0.25% mercaptoethanol before loading on 4%–12% Bis-Tris gel (Invitrogen). Resolved proteins were transferred to a PVDF methanol-activated membrane (GE Healthcare) by semi-dry blotting (Bio-Rad). For dot blot assays, 2.3 million cells were seeded on 10-cm plates and transfected with 9 mg of phGluc-Tau/HA plasmid per well 24 hours after seeding. Culture media was changed to serum- and phenol red-free (PRF) DMEM (Invitrogen) 24 hours
e3 Cell Reports 25, 2027–2035.e1–e4, November 20, 2018
post-transfection. Conditioned media was collected 24 hours later and cleared from debris by a centrifugation at 1500 3 g. The cells were washed twice with ice-cold PBS followed by extraction on ice for 30 min in a buffer containing 10 mM Tris-HCl, pH 6.8, 1 mM EDTA, 150 mM NaCl, 0.25% Nonidet P-40, 1% Triton X-100, PhosStop phosphatase inhibitor, and protease inhibitor mixture tablets (Thermo Scientific). Cell debris was removed by a centrifugation at 13 000 3 g. 50 mL of lysate and 100 mL of media were pipetted to a 96-well Manifold Spot-Blot unit (Whatman / Schleicher and Schuell), and protein were trapped by filtration through a methanolactivated PVDF membrane (GE Healthcare). Chemiluminescence signal was detected following incubation of the membrane with primary antibodies and horseradish peroxidase-conjugated secondary antibodies and ECL Western Blotting reagent (Thermo Scientific). ImageStudioLite software was used for quantitative analysis of filter trap assay blots and the ratio between the extracellular and intracellular protein was determined. Thioflavin S assay Tau conditioned media and mock conditioned media from N2A cells were concentrated 100x with Amicon filters (30 kDa MWCO, Millipore) according to manufacturer’s instructions. 100 mL of samples were incubated with or without 0.5 mM recombinant 0N4R Tau (rPeptide), 1% protease inhibitor cocktail (Thermo Scientific) and DTT for 5 days at 37 C with shaking at 250 rpm. As a positive control, 50 mM heparin was incubated with recombinant Tau in the same conditions. For fluorescence measurement, 10 mM Thioflavin S was added to each sample, incubated 30 min and fluorescence was measured at 440 excitation, 550 emission with a Varioskan Flash multiplate reader (Thermo Scientific). Size-exclusion dot blot assay Tau-GLuc conditioned media from N2A cells was concentrated in parallel using different molecular weight cut-off filters [30 kDa, 50 kDa, 100 kDa (Amicon/Millipore) and 300 kDa (Sartorius)] by centrifugation. Concentrated retained fraction (that did not pass through the filter) was diluted with PRF-DMEM to match the volume of the pass-through fraction. Both fractions were dot blotted on a PVDF membrane and processed as described above under ‘‘Western blot and dot blot assay‘‘. The presence of Tau in the media fractions was assessed by Tau-5 staining. QUANTIFICATION AND STATISTICAL ANALYSIS All values were expressed as mean ± SEM. All statistical analyses were performed in IBM SPSS Statistics software (version 25) and the graphs were compiled in GraphPad Prism software (version 5.0a). The statistical details of experiments can be found in the figure legends. Statistical analyses were performed with unpaired, two-tailed t test or one-way ANOVA followed by Bonferroni post-test (ThioS data) or Dunnett post-test (all other experiments). The choice of one post-test or the other was determined by the need to compare all of the data point to each other, or with the control only. If not otherwise specified, ‘‘n’’ represents the number of independent replication of the entity–interventions. In PCA, each independent replica was the average of four technical replicas, while in ELISA each independent replica was the average of two technical replicas. A minimum of 3 independent replicas was used for the statistical analysis. All p values below 0.05 were considered significant. *p < 0.05, **p < 0.01, ***p < 0.001.
Cell Reports 25, 2027–2035.e1–e4, November 20, 2018 e4