Extracellular Vesicles Isolated from Familial Alzheimer’s Disease Neuronal Cultures Induce Aberrant Tau Phosphorylation in the Wild-Type Mouse Brain

Abstract

Extracellular vesicles (EVs) are a heterogeneous group of secreted particles consisting of microvesicles, which are released by budding of the cellular membrane, and exosomes, which are secreted through exocytosis from multivesicular bodies. EV cargo consists of a wide range of proteins and nucleic acids that can be transferred between cells. Importantly, EVs may be pathogenically involved in neurodegenerative diseases such as Alzheimer’s disease (AD). While EVs derived from AD neurons have been found to be neurotoxic in vitro, little is known about the pathological consequences of AD EVs in vivo. Furthermore, although all known familial AD (fAD) mutations involve either amyloid-β protein precursor (AβPP) or the machinery that processes AβPP, hyperphosphorylation of the microtubule associated protein tau appears to play a critical role in fAD-associated neurodegeneration, and previous reports suggest EVs may propagate tau pathology in the AD brain. Therefore, we hypothesized that fAD EVs may have a mechanistic involvement in the development of fAD-associated tau pathology. To test this, we isolated EVs from iPSC-derived neuronal cultures generated from an fAD patient harboring a A246E mutation to presenilin-1 and stereotactically injected these EVs into the hippocampi of wild-type C57BL/6 mice. Five weeks after injection, mice were euthanized and pathology evaluated. Mice injected with fAD EVs displayed increased tau phosphorylation at multiple sites relative to PBS and non-disease control EV injected groups. Moreover, fAD EV injected hippocampi contained significantly more tau inclusions in the CA1 hippocampal neuronal field than controls. In total, these findings identify EVs as a potential mediator of fAD-associated tau dysregulation and warrant future studies to investigate the therapeutic potential of EV-targeted treatments for fAD.

Keywords

Alzheimer’s disease C57BL/6 extracellular vesicles induced pluripotent stem cells tau tauopathy

INTRODUCTION

Early onset familial Alzheimer’s disease (fAD) is caused by mutations to the machinery that processes amyloid-β protein precursor (AβPP), such as presenilin-1 (PS1) and presenilin-2 (PS2), or by mutations to AβPP itself. PS1 and PS2 proteins form the catalytic core of γ-secretase enzymes that are required for amyloid-β (Aβ) production. Early pre-symptomatic accumulations of Aβ are a signature of fAD [1]. As we have previously reported, iPSC-derived neurons from fAD patients release increased amounts of Aβ in vitro [2]. Although the link between PS1 and AβPP mutations and increased amyloid production is relatively clear, the mechanisms that underlie the development of tauopathy in fAD patients are poorly understood. Aberrant aggregation of microtubule associated protein tau is a signature of many neurodegenerative diseases and both hyperphosphorylated tau and aggregated tau deposits are features of the fAD brain [3]. Moreover, tau concentrations are increased in the cerebrospinal fluid (CSF) of fAD patients [4, 5] and tau pathology may occur prior to cognitive decline [1]. These data suggest that tau hyperphosphorylation is an early feature of fAD and potential mediator of fAD-associated neurodegeneration.

A common motif amongst proteinopathies is neurodegeneration that progresses from region to region during the course of disease. Recent hypotheses posit that extracellular vesicles (EVs) are mechanistically involved in disease propagation via cell to cell trafficking of infectious peptides, including tau [6] and Aβ [7]. EVs are a heterogenous group of 50–1000 nm sized particles that encompasses exosomes, which are secreted from endosomal sorting pathways, and microvesicles, which are secreted through budding of the cytoplasmic membrane. Animal studies demonstrate that inhibiting EV secretion reduces pathology in mutant AβPP [8] and tau [9] expressing mice. Additionally, other studies have found that EVs isolated from mutant PS1 neurons are neurotoxic and contain pathological Aβ [10]. AD patient-derived EVs are sufficient to seed Aβ pathology [7] and, as we have recently demonstrated, EVs isolated from human tau-overexpressing neuronal cultures seed tau deposit when injected into the wild-type (Wt) mouse brain [11, 12]. Despite these findings, the effects of fAD EVs on tau pathology remain unknown. With this in mind, we sought to determine the effects of fAD EVs on tau dynamics in healthy brain tissue by infusing EVs isolated from fAD neuronal cultures into the brains of healthy mice and evaluating tau pathology.

MATERIALS AND METHODS

Cell culture

Neuronal cultures were produced from patient-derived iPSCs according to previously published methods [13]. For neuronal cultures, iPSC-derived neural stem cells (NSCs) were seeded at a density of 150,000 cells/cm² on Matrigel-coated (70μg/mL) plastic cell culture dishes. NSCs were grown to 80% confluency, at which time neuronal differentiation was initiated through withdrawal of basic fibroblast growth factor (bFGF) from the NSC media (DMEM-F12, 1% N-2, 2%, B-27, Pen-Strep, 20 ng/mL bFGF). NSCs were grown for 3 weeks in the absence of bFGF to achieve neuronal differentiation. Cell culture media was then collected every 3-4 days, centrifuged at 1690× g for 15 min to remove dead cells and debris and then frozen at –20°C until EV isolation. Media used for EV isolated was collected for up to 2 weeks (i.e., during weeks 3-4 and 4-5 post bFGF withdrawal). Media from individual cell cultures was pooled prior to EV isolation.

Isolation of EVs

Prior to EV extraction, cell media underwent PCR-based testing for infectious agents (mouse essential panel, Charles River). All tests were negative. EVs were then isolated from cell culture media using ExoQuick TC (System Biosciences, Cat # EXOTC50A-1) according to the manufacturer’s instructions with slight modifications. Briefly, after dead cells/debris was removed from cell culture media by centrifugation, the Exoquick TC/cell culture media mixture was rotated at 4°C for 48 h instead of 24 h as described by the manufacturer. We found that this modified protocol increased the yield of EVs isolated. All other steps were carried out according to the manufacturer’s instructions.

Nanoparticle tracking analysis (NTA) of EVs

To determine the size and abundance of isolated EVs, EVs were diluted in phosphate buffered saline (PBS) and analyzed using a NanoSight LM-10 instrument.

ELISA analysis of tau/pTau (231)

For analysis of phosphorylated tau (T231) and total tau contained within EVs and cell lysates, EVs were lysed in RIPA (50 mM Tris pH 7.4, 150 mM NaCl, 1% Triton-100X, 1 mM, EDTA pH 7.5, 0.1% SDS, 0.25% Na-deoxycholate, 1 mM EGTA) buffer and pelleted cells where lysed with MSD buffer provided in the kit used. Lysates were then analyzed for pTau (T231) and total tau content by ELISA (Meso Scale Diagnostics, Cat # K15121D-1).

Mice

For this study, male C57BL/6J mice were used (Jackson Labs Stock #00064). Mice were 8 weeks old at the time of stereotaxic injections. Because of the apparent effects of estrogen levels on tau phosphorylation [14 –16], only male mice were used in this study in order to avoid a potential confounding effect of fluctuating estrogen levels during the female reproductive cycle.

Stereotaxic injection of EVs

Mice were anesthetized with isoflurane and incisions made along the midline of the scalp. 2μl (0.67μg of protein/μl) of EVs suspended in PBS were injected into each hippocampi using the following coordinates (relative to the bregma): anteroposterior – 2 mm, mediolateral±1.75 mm, and dorsoventral – 1.75 mm. After surgery, scalp incisions were closed and mice returned to their home cages after the effects of anesthesia had worn off. Carprofen (5 mg/kg, SQ) was given up to 48 h post-operatively for pain control.

Collection of brain tissue

Mice were maintained for 5 weeks following stereotaxic injections. During the 4th week, mice underwent Morris water maze testing (MWM, see Supplementary Materials and Methods for a detailed description). After 5 weeks post-treatment, mice were placed under isoflurane anesthesia and euthanized by cervical dislocation. Brains were immediately dissected out and processed for immunohistochemistry (IHC) or frozen and saved for western blotting (WB).

WB analysis of brain tissue

Whole frozen hippocampi were homogenized in RIPA buffer and protein quantified with the Pierce BCA protein assay kit (ThermoFisher Cat # 23227). Lysates were separated by SDS-PAGE, transferred to PVDF membranes, and blocked with 5% skim milk in 0.1% tris buffered saline- tween (TBST). Primary antibodies (see Table 1) were diluted in either 5% skim milk or 5% BSA (in TBST) and added to blots overnight at 4°C while rocking. The following day, membranes were incubated with an appropriate species-specific secondary antibody, washed in TBST, and imaged using a chemidoc (Biorad) imaging system. Blots were quantified with Image lab (Biorad) software.

Table 1

Antibodies

Name	Epitope	Host	Source	Identifier	Dilution
AT8	Tau phosphorylated at a.a. S202/T205	Ms	Thermo Scientific	MN1020	WB: 1/1000
Flotillin-1	Mouse Flotillin a.a. 312–428	Ms	BD Biosciences	610821	WB: 1/200
HSP 90	N/A	Rb	Cell signaling	4874	WB: 1/1000
K9JA (total tau)	Tau a.a. 243–441	Rb	DAKO	A0024	IHC: 1/1000
PHF-1	Tau phosphorylated at a.a.S396/S404	Ms	Peter Davies	N/A	WB: 1/1000
pTau (T231)	Tau phosphorylated at a.a. T231	Rb	Abcam	ab151559	IHC: 1/200, WB: 1/2000
Tau5 (total tau)	Tau a.a. 210–241	Ms	Millipore	MAB 361	WB: 1/500
β-actin	Sequence within the N-terminus of β-actin	Ms	Sigma	A5316	WB: 1/2000

WB analysis of EVs

Isolated EVs were lysed in RIPA buffer and lysates analyzed as described above.

Fixation of brain tissue

Immediately after dissection, brain hemispheres were placed into 4% PFA (diluted in PBS) for 24 h. After 24 h, tissue was washed with PBS before paraffin embedding.

Processing and paraffin embedding of fixed tissue

Tissue was processed using a Leica TP 1020 processor and incubated in successive treatments of 70% ethanol (EtOH) 5 min, 70% EtOH 2 h, 95% EtOH 1 h (2X), 100% EtOH 1 h (3X), Xylene 1 h (3X), and finally paraffin 1.5 h (2X). Tissue was then placed into melted paraffin and rectangular molds and allowed to harden.

IHC analysis of brain tissue

Tissue embedded in paraffin was sliced into 6μm sections using a microtome (Leica RM 2035) and sections collected on to glass slides. Slides were then incubated in 100% Citrisolv (Decon laboratories cat. #1601) 2X for 15 min, 100% EtOH 2X for 3 min, 95% EtOH for 3 min, 70% EtOH for 3 min, 70% EtOH for 3 min and finally water 2X 3 min. After water incubation, slides were incubated in 0.5% H₂O₂ (diluted in methanol) for 15 min and then washed 2X 3 min in PBS. For antigen retrieval, slides were placed incubated in 9.0 pH solution (Vector H-3301) while in a water filled pressure cooker and heated to 90°C for 10 min. Slides were cooled and then placed in blocking buffer (1% FBS/0.1% triton-X diluted in PBS) for 25 min at room temperature. Slides were washed 2X 3 min with PBS. Slides were then removed from PBS and tissue was encircled with a hydrophobic pen. Primary antibodies diluted in 1% FBS (in PBS) were added to tissue overnight at 4°C. The next day, slides were allowed to equilibrate to room temperature for 30 min and then washed 2X 3 min PBS. Tissue was then blocked using 2% normal horse serum (Horse serum and secondary antibody solution: Impress reagent kit, VECTOR cat #MP-7401). Cells were incubated in horse serum for 20 min and then secondary antibody solution was added for 60 min at room temperature. After incubation in secondary antibodies, tissue was incubated in NovaRed (VECTOR SK-4800) solution and developed 1-2 min. After development, slides were immediately submerged into water to quench the reaction. After chromogenic reaction, slides were rinsed 2X with water and dried at room temperature for 30 min. Finally, slides were rinsed 2X 1 min in xylene and a glass coverslip added and adhered with Richard-Allan mounting media (ThermoFisher cat # 4112).

Imaging of IHC slides

Images in Figs. 3B-G and 4C were imaged using a NanoZoomer 2.0HT slide scanning microscope (Hamamatsu). Figure 3H and I were imaged using a BZ-X710 microscope (Keyence).

IHC analysis of tau aggregates

The number of K9JA positive cells in the CA1 layer of treated hippocampi were counted and standardized to the total number of cells (colored purple by Gill’s hematoxylin counterstain) in the field. Data is reported as the number of K9JA positive staining cells/total number of cells.

Antibodies

All antibodies and dilutions used can be found in Table 1. See citation [17] for information regarding the PHF-1 antibody used.

Ethics approval

All animal studies were performed in accordance with the University of California San Diego Institutional Animal Care and Use Committee rules (UCSD IACUC protocol # S18100).

RESULTS

Characterization of isolated EVs

For this study, we isolated EVs from iPSC-derived neuronal cultures generated from a NDC patient and an fAD patient harboring an A246E mutation to PS1 (See [13] for NDC cell line information (referred to as NDC1) and [2] for PS1 cell line information (referred to as PS1-1)). Hyperphosphorylation of tau is a feature of fAD neurons [18] which was reflected in our cultures (Fig. 1). To isolate EVs, cell culture media from NDC and PS1 neurons was collected every 2-3 days for 2 weeks following neuronal differentiation (i.e., weeks 3 to 5 following the withdrawal of bFGF from cultured NSCs) and EVs extracted as described in the Methods and Materials. We examined the size of isolated particles via NTA and found that the majority of NDC EVs were ∼158–218 nm in diameter (Fig. 2A) and PS1 EVs were slightly larger at ∼183–268 nm. Next, EVs were lysed and expression of flotillin-1, an EV marker, determined. Both NDC and PS1 EVs were flotillin-1 positive (Fig. 2C). Finally, we quantified the amount tau and pTau (T231) contained in isolated EVs. Surprisingly, both total tau and pTau levels were higher in NDC EVs (Fig. 2D/E) compared to PS1 EVs although the ratio of pTau 231 to total tau was higher in PS1 EVs (Fig. 2F).

Fig.1

Tau phosphorylation in fAD neuronal cultures. NDC and PS1 neuronal cultures lysates were analyzed by ELISA. Data is mean tau (A, pg of tau/mL of lysate), pTau T231 (B, A.U. or arbitrary units), and pTau/tau ratio (C, A.U.) of lysates±SEM. N = 6 wells for NDC group and 5 wells for PS1 group. Data was analyzed with a Student’s t-test. ^*p < 0.05.

Fig.2

Characterization of fAD EVs. A,B) Data is relative size distribution of EVs (determined by NTA, see Methods and Materials) isolated from NDC and PS1 neuronal cultures (X axis is diameter in nm). Blue numbers indicate the diameter of particles (in nm) at the various peaks. C) Isolated EVs were processed for WB and lysates probed from the EV marker flotillin-1. D-F) Relative total tau and pTau 231 levels in NDC and PS1 EVs were determined by ELISA. Data is mean tau (pg/μg of protein) (D), pTau T231 (arb. Units) (E), and pTau/Tau levels (arb. Units) (F) in lysates±SEM. N = 4 technical replicates for NDC and 3 technical replicates for PS1. ^*p < 0.05; ^**p < 0.01; ^***p < 0.001.

The effects of EVs in vivo

The overall goal of these experiments was to determine the effects of fAD derived EVs on healthy brain tissue. Therefore, for this study we chose healthy, young (8 weeks old) C57BL/6 mice as the recipient model for EV injections.

At 8-weeks of age, male C57BL/6 mice were stereotactically injected with either PBS, NDC EVs, or PS1 EVs bi-laterally into each hippocampus (Fig. 3A). Mice were maintained for 4 weeks following injection and then underwent cognitive testing in the Morris water maze (MWM); however, neither NDC nor PS1 EVs appeared to have an effect on cognition (Supplementary Figure 1). Three days after the final day of MWM testing mice were sacrificed and brains were dissected out for analysis.

Fig.3

Tau deposits in the hippocampi of PS1 treated mice. A) Timeline of study. 8-week-old (8 w.) mice were stereotactically injected with EVs and then maintained for 4 additional weeks before undergoing Morris water maze (MWM testing). After testing, mice were sacrificed (Sac.) and biochemical analyses performed. B-G) PBS, NDC EV, and PS1 EV treated hippocampi were sliced sagittally (6μm thick sections) and immunolabeled with anti-tau (K9JA) antibodies. NovaRed was used for chromonergic development and tau identified by the presence of red staining inclusions. Slices were counterstained with Gill’s hematoxylin (purple). Images are 5X (images B, D, and F. Scale bar = 20μm. Black arrows indicate tau inclusions) or 40X (images C, E, and G. Scale bar = 10μm) magnified view of the CA1 field. H) Image of a tau thread observed in PS1 treated hippocampus (80X magnification, scale bar = 5μm). I) Image of tau deposits localized to the nucleus observed in PS1 treated hippocampus (80X magnification, scale bar = 10μm) J) Data is mean # of K9JA positive cells/total cells in CA1 field,±SEM. Data was analyzed with a 1-way ANOVA, Tukey’s post-hoc test. N = 3 mice/group. ^*p < 0.05.

In our previous study [12], we found that EVs isolated from mutant tau overexpressing cultures were sufficient to cause tau aggregation in the Wt mouse brain and that these aggregates were detectable using K9JA anti-total tau antibodies (Table 1). Here, we used the same K9JA antibody to analyze tau deposits in the brains of NDC and PS1 EV treated mice. Initial analyses indicated that tau aggregates were localized almost entirely to the CA1 field of the hippocampus; therefore, only K9JA positive cells in CA1 field were quantified. PS1 EV treated mice displayed significantly more K9JA positive cells compared to controls while no differences in K9JA reactivity between PBS and NDC groups were detected (Fig. 3B-G, J). Additionally, a number of tau-thread like structures (Fig. 3H) and nuclear localized aggregates (Fig. 3I) were found in PS1 EV treated hippocampi.

Finally, in order to determine the effects of PS1 EVs on tau phosphorylation in the hippocampi of our mice, whole hippocampi were dissected out, homogenized, and lysates analyzed by WB. We first analyzed phosphorylation of tau at T231 in our lysates as phosphorylation of this site is an early phenomenon in AD [19 –21]. WB analysis revealed increased tau phosphorylation at T231 in PS1 EV treated lysates relative to controls (Fig. 4A, B). The differences in T231 phosphorylation were also reflected in IHC analysis of hippocampal slices using these antibodies (Fig. 4C), which revealed a robust, diffuse pattern of staining throughout the CA1 field of PS1 EV treated mice. Importantly, overall tau levels (Tau5 antibody) were unchanged between all three groups (Fig. 4A, B). Next, we further analyzed tau phosphorylation using two additional anti-pTau antibodies. Although phosphorylation was increased at S202/T205 (AT8 antibody) in PS1 EV treated lysates compared to both controls, no differences in phosphorylation at S396/S404 (PHF-1) were detected between groups (Fig. 5).

Fig.4

PS1 EVs increase tau phosphorylation at T231 in the naïve mouse brain. A) PBS, NDC EV (NDC) and PS1 EV (PS1) treated hippocampi were dissected out and processed for WB. Lysates were probed with antibodies that recognize tau phosphorylated at T231, total tau (Tau5) and β-actin. B) Data is mean phospho-tau/total tau ratios and total tau/actin ratio expressed as a percent of the PBS treated group±SEM. Data was analyzed with a 1-way ANOVA, Tukey’s post-hoc test. ^*p < 0.05. N = 3-4 mice/group. C) Sagittal hippocampal slices from PBS, NDC EV (NDC), and PS1 EV (PS1) treated mice were immunolabeled with pTau T231 antibodies (in red) and counterstained with Gill’s hematoxylin (purple). Images are 10X magnified view of the CA1 field. Scale bar = 200μm.

Fig.5

PS1-EV induced tau phosphorylation is site specific. EV and PBS treated lysates were probed with antibodies that recognize tau phosphorylated at (A) S202/T205 (AT8) and (B) S396/404 (PHF-1) plus total tau (Tau5) and heat shock protein 90 (HSP 90) as loading controls. C) Data is mean phospho-tau/tau ratio (expressed as a percentage of PBS)±SEM. ^*p < 0.05.

DISCUSSION

Mounting evidence suggests that Aβ accumulation results in dysregulation of tau dynamics [22] and that tau mediates the toxic effects of Aβ [23]. For instance, genetic depletion of tau prevents Aβ-induced reductions in long term potentiation [24] and prevents memory deficits in AβPP-overexpressing mice [25]. The primary finding of this study was the presence of tau pathology in the PS1 EV treated hippocampus, particularly in the CA1 field which is the first hippocampal region affected by AD [26 –28]. Whole hippocampal homogenates were used for WB analysis of hippocampal tau phosphorylation and follow up IHC analysis revealed that nearly every cell of the CA1 field was positive for pTau (T231). Although some dark staining deposits were visible within the PS1 EV, pTau (T231) immunolabeled CA1 field, the majority of staining was more diffuse suggesting that the increase in pTau levels in the PS1-EV treated hippocampi was a response by murine neurons to the toxic effects of PS1 EVs. Interestingly, PS1-EV treated hippocampi displayed a specific pattern of tau dysregulation with phosphorylation being increased at T231 and S202/T205 but not S396/S404. Both T231 and S202/T205 fall within the proline rich domain of tau far from the S396/S404 sites localized to the C-terminus which suggests that PS1 EVs may induce tau phosphorylation in a region-specific manner, although further analysis is needed. Moreover, while T231 has been identified as an early tau phosphorylation site in AD, recent work has revealed that hyperphosphorylation of S202/T205 does not occur until late in the disease process [29]. Furthermore, it is unclear why PS1 EVs failed to increase tau phosphorylation at S396/S404 in our mice as hyperphosphorylation of these sites is routinely associated with AD [30]. Interestingly, Wt murine brain tissue shows a different pattern of tau phosphorylation in response to okadaic acid, an inhibitor of the tau phosphatase PP2A, compared to mutant human AβPP expressing tissue particularly at residues recognized by the PHF-1 antibody [31]. Given that phosphorylation of tau at S396 is mediated by the tau kinase GSK3 [32], which is activated in response to Aβ [33], the hyperphosphorylation of tau we observed in PS1 EV treated tissue may be mediated by GSK3-independent mechanisms. Moreover, fAD EVs may increase PHF-1 immunoreactivity in the presence of Aβ pathology, although further studies are needed.

The main goal of this study was to determine whether or not fAD neuronally-derived EVs could induce tau pathology in healthy brain tissue. Although fAD EVs have been found to be acutely neurotoxic in vitro, the lasting effects of fAD EVs on AD-like pathology in vivo are unknown. In our study, the 5-week post-injection endpoint allowed us to analyze tau dynamics at a point in which tau hyperphosphorylation would not be a side effect of any potential, acute toxicity induced by EVs but rather a longer lasting pathology. However, the limitation of this study is that the effects of PS1 EVs on tau pathology beyond 5 weeks were not analyzed and it is unclear if fAD EVs are sufficient to recapitulate additional features of AD. Additionally, despite the effects of PS1 EVs on tau phosphorylation and aggregation, PS1 EVs failed to induce cognitive dysfunction in our mice. This result was not necessarily unexpected as cognitive impairment takes months to develop in genetic mouse models of tauopathy [34, 35]. Nevertheless, the lack of cognitive decline remains a limitation of our model and may be due to the delivery system used (i.e., a one-time EV injection as opposed to a constant infusion), the resistance of murine tau to tangle formation [36] and the duration of the treatment period. Therefore, future studies will utilize additional animal models and EV delivery systems in order to determine the degree of AD-like pathology that can be achieved with exogenously administered fAD EVs.

Although tau phosphorylation was increased in the hippocampi of PS1 EV treated mice, the EV associated cargo responsible for the observed hyperphosphorylation of tau was not identified. Two potential mechanisms that may underlie EV-mediated phosphorylation of tau in healthy brain tissue are: 1) corruption of endogenous tau dynamics by pathogenic proteins (i.e., tau) that are packaged into EVs or 2) dysregulation of intracellular signaling pathways that maintain tau homeostasis. Contrary to our initial hypothesis, we found that PS1-EVs contained less ptau/tau compared to NDC EVs. These disparate findings may indicate that fAD neurons have a reduced capacity to clear excess pTau via EV secretion although this does not address why PS1 EVs increased tau phosphorylation in our animals. We hypothesize that pathogenically modified tau seeds may still be present in PS1 EVs as these seeds have been reported to be released by cultured fAD neurons [6, 37]. Such seeds may be able to initiate tau phosphorylation/aggregation cascades although further characterization of fAD EV-associated tau is needed. Alternatively, the effects of PS1 EVs on tau phosphorylation in the naïve mouse brain may be independent of EV tau and may be the result of a toxic combination of other nucleic acids and proteins packaged in into EVs. For instance, I1PP2A, which inhibits PP2A, is upregulated in AD [38] and found in EVs released from several glioblastoma cell lines [39]. Moreover, miRNAs that may positively modulate tau phosphorylation such as miRNA-424 and miRNA-138 have been reported to be contained with EVs [40 –43]. Therefore, the effects of PS1 EVs on tau phosphorylation in vivo may be due dysregulated levels of tau modulating proteins/miRNAs such as these being packaged into AD neuronal-derived EVs. Given the number of proteins and miRNAs packaged into it is possible that multiple dysregulated proteins/miRNAs in EVs may alter the activity of tau kinases/phosphatases in recipient cells, or act on intracellular tau directly, in order to shift tau dynamics towards hyperphosphorylation. Therefore, future studies will first analyze AD-associated changes in EV cargo and later attempt to identify dysregulated proteins/miRNAs that may be mechanistically involved in fAD-EV mediated phosphorylation of tau.

CNS derived EVs are both an attractive therapeutic target as well as a potential disease biomarker given that EVs isolated from the CSF of human patients can predict the conversion of mild cognitive impairment to AD [44]. Although inhibition of EV release reduces pathology and prevents cognitive decline in AD animal models [8, 9], the viability of this strategy in human patients is unclear. Neuronal EV secretion may be critical for cell health by serving as a disposal system for intracellular waste. Furthermore, it is possible that not all EVs released by the AD brain are toxic as NSC-derived EVs have been found to have neuroprotective properties [45 –47] and the ramifications of depleting these EVs in vivo are unknown. Nevertheless, this work supports a broader hypothesis which posits that pathogenic cargo loaded into EVs facilitates the propagation of pathology in AD and related disorders.

Conclusion

The neurotoxic effects of dysregulated tau are well documented and multiple reports suggest that aberrant Aβ production is linked to the development of tau pathology. The present study demonstrates that EVs released by neurons that harbor fAD mutations are sufficient to induce tau hyperphosphorylation in healthy brain tissue. Although the specific mechanism in which fAD EVs induce tau phosphorylation remains unknown, this work nevertheless identifies neuronally-derived EVs as a potential link between fAD mutations and tau pathology.

Footnotes

ACKNOWLEDGMENTS

We would like to thank Dr. Peter Davies at Albert Einstein College of Medicine for providing PHF-1 antibody, and members of the Yuan lab for helpful discussion.

The LM 10 Nanosight imager is maintained and operated by the GT3 Core Facility of the Salk Institute with funding from NIH-NCI CCSG: P30 014195, an NINDS R24 Core Grant, and a grant from the National Eye Institute.

Both NanoZoomer 2.0HT and BZ-X710 microscopes are maintained and operated by the University of California San Diego Microscopy Core (NINDS grant #NS047101).

This study was supported by grants NIH AG057469 to RAR and SHY, AG057459, AG051848, AG018440, AG005131, VA BX003040 and BX004312 to RAR, and funds from the ITN to SHY.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Quiroz

, Sperling

, Norton

, Baena

, Arboleda-Velasquez

, Cosio

, Schultz

, Lapoint

, Guzman-Velez

, Miller

, Kim

, Chen

, Tariot

, Lopera

, Reiman

, Johnson

(2018) Association between amyloid and tau accumulation in young adults with autosomal dominant Alzheimer disease. JAMA Neurol 75, 548–556.

Liu

, Waltz

, Woodruff

, Ouyang

, Israel

, Herrera

, Sarsoza

, Tanzi

, Koo

, Ringman

, Goldstein

, Wagner

, Yuan

(2014) Effect of potent gamma-secretase modulator in human neurons derived from multiple presenilin 1-induced pluripotent stem cell mutant carriers. JAMA Neurol 71, 1481–1489.

Shepherd

, Gregory

, Vickers

, Brooks

, Kwok

, Schofield

, Kril

, Halliday

(2004) Positional effects of presenilin-1 mutations on tau phosphorylation in cortical plaques. Neurobiol Dis 15, 115–119.

Ringman

, Schulman

, Becker

, Jones

, Bai

, Immermann

, Cole

, Sokolow

, Gylys

, Geschwind

, Cummings

, Wan

(2012) Proteomic changes in cerebrospinal fluid of presymptomatic and affected persons carrying familial Alzheimer disease mutations. Arch Neurol 69, 96–104.

Rostgaard

, Waldemar

, Nielsen

, Simonsen

(2015) Cerebrospinal fluid biomarkers in familial forms of Alzheimer’s disease and frontotemporal dementia. Dement Geriatr Cogn Disord 40, 54–62.

Guix

, Corbett

, Cha

, Mustapic

, Liu

, Mengel

, Chen

, Aikawa

, Young-Pearse

, Kapogiannis

, Selkoe

, Walsh

(2018) Detection of aggregation-competent tau in neuron-derived extracellular vesicles. Int J Mol Sci 19, E663.

Sardar Sinha

, Ansell-Schultz

, Civitelli

, Hildesjo

, Larsson

, Lannfelt

, Ingelsson

, Hallbeck

(2018) Alzheimer’s disease pathology propagation by exosomes containing toxic amyloid-beta oligomers. Acta Neuropathol 136, 41–56.

Dinkins

, Dasgupta

, Wang

, Zhu

, Bieberich

(2014) Exosome reduction in vivo is associated with lower amyloid plaque load in the 5XFAD mouse model of Alzheimer’s disease. Neurobiol Aging 35, 1792–1800.

Asai

, Ikezu

, Tsunoda

, Medalla

, Luebke

, Haydar

, Wolozin

, Butovsky

, Kugler

, Ikezu

(2015) Depletion of microglia and inhibition of exosome synthesis halt tau propagation. Nat Neurosci 18, 1584–1593.

10.

Eitan

, Hutchison

, Marosi

, Comotto

, Mustapic

, Nigam

, Suire

, Maharana

, Jicha

, Liu

, Machairaki

, Witwer

, Kapogiannis

, Mattson

(2016) Extracellular vesicle-associated Abeta mediates trans-neuronal bioenergetic and Ca(2+)-handling deficits in Alzheimer’s disease models. NPJ Aging Mech Dis 2, 16019.

11.

Reilly

, Winston

, Baron

, Trejo

, Rockenstein

, Akers

, Kfoury

, Diamond

, Masliah

, Rissman

, Yuan

(2017) Novel human neuronal tau model exhibiting neurofibrillary tangles and transcellular propagation. Neurobiol Dis 106, 222–234.

12.

Winston

, Aulston

, Rockenstein

, Adame

, Prikhodko

, Dave

, Mishra

, Rissman

, Yuan

(2019) Neuronal exosome-derived human tau is toxic to recipient mouse neurons in vivo. J Alzheimers Dis 67, 541–553.

13.

Israel

, Yuan

, Bardy

, Reyna

, Mu

, Herrera

, Hefferan

, Van Gorp

, Nazor

, Boscolo

, Carson

, Laurent

, Marsala

, Gage

, Remes

, Koo

, Goldstein

(2012) Probing sporadic and familial Alzheimer’s disease using induced pluripotent stem cells. Nature 482, 216–220.

14.

Munoz-Mayorga

, Guerra-Araiza

, Torner

, Morales

(2018) Tau phosphorylation in female neurodegeneration: Role of estrogens, progesterone, and prolactin. Front Endocrinol (Lausanne) 9, 133.

15.

Steinmetz

, Ramos

, Campbell

, Morales

, Rissman

(2015) Reproductive stage and modulation of stress-induced tau phosphorylation in female rats. J Neuroendocrinol 27, 827–834.

16.

Gonzalez-Arenas

, Pina-Medina

, Gonzalez-Flores

, Gomora-Arrati

, Carrillo-Martinez

, Balandran-Ruiz

, Camacho-Arroyo

(2012) Expression pattern of tau in the rat brain during pregnancy and the beginning of lactation. Brain Res Bull 89, 108–114.

17.

Otvos

Jr. , Feiner

, Lang

, Szendrei

, Goedert

, Lee

(1994) Monoclonal antibody PHF-1 recognizes tau protein phosphorylated at serine residues 396 and 404. J Neurosci Res 39, 669–673.

18.

Ochalek

, Mihalik

, Avci

, Chandrasekaran

, Teglasi

, Bock

, Giudice

, Tancos

, Molnar

, Laszlo

, Nielsen

, Holst

, Freude

, Hyttel

, Kobolak

, Dinnyes

(2017) Neurons derived from sporadic Alzheimer’s disease iPSCs reveal elevated TAU hyperphosphorylation, increased amyloid levels, and GSK3B activation. Alzheimers Res Ther 9, 90.

19.

Cho

, Johnson

(2004) Primed phosphorylation of tau at Thr231 by glycogen synthase kinase 3beta (GSK3beta) plays a critical role in regulating tau’s ability to bind and stabilize microtubules. J Neurochem 88, 349–358.

20.

Luna-Munoz

, Chavez-Macias

, Garcia-Sierra

, Mena

(2007) Earliest stages of tau conformational changes are related to the appearance of a sequence of specific phospho-dependent tau epitopes in Alzheimer’s disease. J Alzheimers Dis 12, 365–375.

21.

Alonso

, Di Clerico

, Li

, Corbo

, Alaniz

, Grundke-Iqbal

, Iqbal

(2010) Phosphorylation of tau at Thr212, Thr231, and Ser262 combined causes neurodegeneration. J Biol Chem 285, 30851–30860.

22.

Tokutake

, Kasuga

, Yajima

, Sekine

, Tezuka

, Nishizawa

, Ikeuchi

(2012) Hyperphosphorylation of Tau induced by naturally secreted amyloid-beta at nanomolar concentrations is modulated by insulin-dependent Akt-GSK3beta signaling pathway. J Biol Chem 287, 35222–35233.

23.

Vergara

, Houben

, Suain

, Yilmaz

, De Decker

, Vanden Dries

, Boom

, Mansour

, Leroy

, Ando

, Brion

(2019) Amyloid-beta pathology enhances pathological fibrillary tau seeding induced by Alzheimer PHF in vivo. Acta Neuropathol 137, 397–412.

24.

Shipton

, Leitz

, Dworzak

, Acton

, Tunbridge

, Denk

, Dawson

, Vitek

, Wade-Martins

, Paulsen

, Vargas-Caballero

(2011) Tau protein is required for amyloid beta-induced impairment of hippocampal long-term potentiation. J Neurosci 31, 1688–1692.

25.

Ittner

, Ke

, Delerue

, Bi

, Gladbach

, van Eersel

, Wolfing

, Chieng

, Christie

, Napier

, Eckert

, Staufenbiel

, Hardeman

, Gotz

(2010) Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer’s disease mouse models. Cell 142, 387–397.

26.

Frisoni

, Ganzola

, Canu

, Rub

, Pizzini

, Alessandrini

, Zoccatelli

, Beltramello

, Caltagirone

, Thompson

(2008) Mapping local hippocampal changes in Alzheimer’s disease and normal ageing with MRI at 3 Tesla. Brain 131, 3266–3276.

27.

Padurariu

, Ciobica

, Mavroudis

, Fotiou

, Baloyannis

(2012) Hippocampal neuronal loss in the CA1 and CA3 areas of Alzheimer’s disease patients. Psychiatr Danub 24, 152–158.

28.

Wisse

, Biessels

, Heringa

, Kuijf

, Koek

, Luijten

, Geerlings

, Utrecht Vascular Cognitive Impairment Study Group (2014) Hippocampal subfield volumes at 7T in early Alzheimer’s disease and normal aging. Neurobiol Aging 35, 2039–2045.

29.

Neddens

, Temmel

, Flunkert

, Kerschbaumer

, Hoeller

, Loeffler

, Niederkofler

, Daum

, Attems

, Hutter-Paier

(2018) Phosphorylation of different tau sites during progression of Alzheimer’s disease. Acta Neuropathol Commun 6, 52.

30.

Gibbons

, Banks

, Kim

, Changolkar

, Riddle

, Leight

, Irwin

, Trojanowski

, Lee

VMY

(2018) Detection of Alzheimer disease (AD)-specific tau pathology in AD and nonAD tauopathies by immunohistochemistry with novel conformation-selective tau antibodies. J Neuropathol Exp Neurol 77, 216–228.

31.

Foidl

, Humpel

(2018) Differential hyperphosphorylation of Tau-S199, -T231 and -S396 in organotypic brain slices of Alzheimer mice. A model to study early tau hyperphosphorylation using okadaic acid. Front Aging Neurosci 10, 113.

32.

, Lu

, Tian

, Yang

, Wang

(2006) Activation of glycogen synthase kinase-3 induces Alzheimer-like tau hyperphosphorylation in rat hippocampus slices in culture. J Neural Transm (Vienna) 113, 93–102.

33.

Kirouac

, Rajic

, Cribbs

, Padmanabhan

(2017) Activation of Ras-ERK signaling and GSK-3 by amyloid precursor protein and amyloid beta facilitates neurodegeneration in Alzheimer’s disease. eNeuro 4, ENEURO.014916.2017.

34.

Sydow

, Hochgrafe

, Konen

, Cadinu

, Matenia

, Petrova

, Joseph

, Dennissen

, Mandelkow

(2016) Age-dependent neuroinflammation and cognitive decline in a novel Ala152Thr-Tau transgenic mouse model of PSP and AD. Acta Neuropathol Commun 4, 17.

35.

Harris

, Koyama

, Maeda

, Ho

, Devidze

, Dubal

, Yu

, Masliah

, Mucke

(2012) Human P301L-mutant tau expression in mouse entorhinal-hippocampal network causes tau aggregation and presynaptic pathology but no cognitive deficits. PLoS One 7, e45881.

36.

Ando

, Leroy

, Heraud

, Yilmaz

, Authelet

, Suain

, De Decker

, Brion

(2011) Accelerated human mutant tau aggregation by knocking out murine tau in a transgenic mouse model. Am J Pathol 178, 803–816.

37.

Muratore

, Rice

, Srikanth

, Callahan

, Shin

, Benjamin

, Walsh

, Selkoe

, Young-Pearse

(2014) The familial Alzheimer’s disease APPV717I mutation alters APP processing and Tau expression in iPSC-derived neurons. Hum Mol Genet 23, 3523–3536.

38.

Tanimukai

, Grundke-Iqbal

, Iqbal

(2005) Up-regulation of inhibitors of protein phosphatase-2A in Alzheimer’s disease. Am J Pathol 166, 1761–1771.

39.

Hurwitz

, Rider

, Bundy

, Liu

, Singh

, Meckes

Jr (2016) Proteomic profiling of NCI-60 extracellular vesicles uncovers common protein cargo and cancer type-specific biomarkers. Oncotarget 7, 86999–87015.

40.

Bellingham

, Coleman

, Hill

(2012) Small RNA deep sequencing reveals a distinct miRNA signature released in exosomes from prion-infected neuronal cells. Nucleic Acids Res 40, 10937–10949.

41.

Zhang

, Liu

, Pei

, Yang

, Xie

, Long

(2018) MicroRNA-322 cluster promotes tau phosphorylation via targeting brain-derived neurotrophic factor. Neurochem Res 43, 736–744.

42.

Wang

, Tan

, Lu

, Peng

, Zhu

, Zhang

, Sun

(2015) MicroRNA-138 promotes tau phosphorylation by targeting retinoic acid receptor alpha. FEBS Lett 589, 726–729.

43.

Zhang

, Wang

, Jia

, Su

, Zhang

(2019) Exosome-mediated microRNA-138 and vascular endothelial growth factor in endometriosis through inflammation and apoptosis via the nuclear factor-kappaB signaling pathway. Int J Mol Med 43, 358–370.

44.

Winston

, Goetzl

, Akers

, Carter

, Rockenstein

, Galasko

, Masliah

, Rissman

(2016) Prediction of conversion from mild cognitive impairment to dementia with neuronally derived blood exosome protein profile. Alzheimers Dement (Amst) 3, 63–72.

45.

Webb

, Kaiser

, Scoville

, Thompson

, Fatima

, Pandya

, Sriram

, Swetenburg

, Vaibhav

, Arbab

, Baban

, Dhandapani

, Hess

, Hoda

, Stice

(2018) Human neural stem cell extracellular vesicles improve tissue and functional recovery in the murine thromboembolic stroke model. Transl Stroke Res 9, 530–539.

46.

Rong

, Liu

, Wang

, Fan

, Luo

, Li

, Kong

, Chen

, Tang

, Cai

(2019) Neural stem cell-derived small extracellular vesicles attenuate apoptosis and neuroinflammation after traumatic spinal cord injury by activating autophagy. Cell Death Dis 10, 340.

47.

Webb

, Kaiser

, Jurgielewicz

, Spellicy

, Scoville

, Thompson

, Swetenburg

, Hess

, West

, Stice

(2018) Human neural stem cell extracellular vesicles improve recovery in a porcine model of ischemic stroke. Stroke 49, 1248–1256.