Abstract
Exosomes are small extracellular vesicles secreted by variety of cell types such as neurons, astrocytes, and oligodendrocytes. It is suggested that exosomes play essential role in the maintenance of the neuronal functions and also in the clearance of amyloid-β (Aβ) from the brain. Aβ is well known to cause neuronal cell death, whereas little is known about its effect on astrocytes. In this study, we examined the effect of Aβ on release of exosomes from astrocytes in culture. We analyzed release of exosomes and apoE, both of which are known to remove/clear Aβ from the brain, in the culture medium of astrocytes. We found that exosome and apoE-HDL were successfully separated by density gradient ultracentrifugation demonstrated by distribution of their specific markers, flotillin and HSP90, and cholesterol, and morphological analysis using electron microscopy. Exosome release was significantly reduced by Aβ1–42 treatment in cultured astrocytes accompanied by an increased JNK phosphorylation. Whereas, apoE-HDL release remained unchanged. A JNK inhibitor restored the decreased levels of exosome release induced by Aβ treatment to levels similar to those of control, suggesting that Aβ1–42 inhibits exosome release via stimulation of JNK signal pathway. Because exosomes are shown to remove Aβ in the brain, our findings suggest that increased Aβ levels in the brain may impair the exosome-mediated Aβ clearance pathway.
INTRODUCTION
Deposition of the amyloid β-protein (Aβ), a major component of senile plaques in the brains of Alzheimer’s disease (AD) patients, was hypothesized to initiate a pathologic cascade that eventually leads to AD [1]. Several lines of evidence have demonstrated that the major component of the amyloid deposits is the 39- to 42-aa peptide of Aβ [2, 3], and that the soluble oligomers of Aβ is responsible for inducing toxic effects on neurons and synaptic dysfunction in the brains of AD patients and AD animal models [4]. The underlying mechanism(s) and the molecular pathogenesis of AD to be addressed are, therefore, the issues to clarify how Aβ is synthesized from its precursor, amyloid-β protein precursor (AβPP), how Aβ is degraded in the brain and removed from the brain, and how Aβ transforms to its toxic oligomers, in addition to how to prevent its toxicity.
In this context, it is of note that recently studies have reported the new extracellular particles, exosomes, which are shown to binds Aβ on its surface to stimulate Aβ fibril formation and enhances Aβ uptake by the cells together with exosomes [5]. Exosomes are small membranous vesicles, which are formed by the invagination of the membrane of endosomal multivesicular bodies (MVBs), harboring proteins and RNAs [6]. Exosomes remain inside the lumen of the MVB until they are secreted into the extracellular space when the MVB fuses with the cytoplasmic membrane [7, 8]. Previous reports have shown that the exosome secretion promotes clearance of Aβ [9, 10], and upregulation of exosome secretion from neuronal cells by treatment with SMS2 siRNA enhanced Aβ uptake into microglial cells and significantly decreased extracellular levels of Aβ [9]. It has been shown that within exosomes, there are AβPP, β-secretase, and γ-secretase complexes and that the brain exosomes are enriched with AβPP CTFs [11]. In addition to Aβ pathologies, it has been shown that tau is secreted in association with exosomes in tauopathy model mouse and tau–associated exosomes are secreted in cerebrospinal fluid in early AD patients [12]. Other study has shown that exosome attenuates synaptic plasticity-disrupting activity of Aβ assemblies in vivo [13]. These results suggest that exosomes play key roles in the pathogenesis of AD, which are exosome has effects on AβPP/Aβ metabolism, its toxicity on neurons, and tauopathy. However, the effect of Aβ on exosome release and/or function remains unknown.
Exosomes are released from many types of cells including neurons and astrocytes [9, 15]. Aβ has a strong toxic effect on neurons; however, the effect of Aβ on astrocytes has not understood well. In this study, we examined the effect of Aβ on exosome secretion from astrocytes in culture. We found that exosome and apoE-HDL were successfully separated by density gradient ultracentrifugation, and exosome release was significantly reduced by Aβ1–42 treatment, whereas apoE-HDL release was not in cultured astrocytes. Aβ1–42-induced inhibition of exosome release was accompanied by an increased JNK phosphorylation. A JNK inhibitor recovered the decreased levels of exosome induced by Aβ treatment to levels similar to those of control, suggesting that Aβ1–42 inhibits exosome release via stimulation of JNK signal pathway. Because, exosome is shown to remove Aβ in the brain, our findings suggest that increased Aβ levels in the brain may impair the exosome-mediated Aβ clearance pathway.
MATERIALS AND METHODS
Cell culture
Astrocyte-enriched cultures were obtained from embryonic day 18 Sprague–Dawley rat embryos as described previously [16, 17]. In brief, isolated cortices of the fetuses were minced and the cortical fragments were incubated in 0.25% trypsin and 20μg/ml DNase I in phosphate-buffered saline (PBS) (8.1 mM Na2HPO4, 1.5 mMKH2PO4, 137 mM NaCl, and 2.7 mM KCl, pH 7.4) at 37°C for 15 min. The fragments were then dissociated into single cells by pipetting. The dissociated cells were seeded in 75-cm2 flask at a cell density of 1×107 in DMEM containing 10% FBS. After 10 days of incubation in vitro, astrocytes in the monolayer were trypsinized (0.1%) and reseeded onto 6-cm2 dishes. The astrocyte-rich cultures were maintained in DMEM containing 10% FBS until use.
Evaluation of cells survival
Cells were plated in 6-cm2 dishes and treated with various compounds, as indicated in the figure legends. At the end of the treatments, cells proliferation was evaluated using LDH-Cytotoxic test Wako (Wako pure chemical.) and taking image of cells (Olympus 1×70, Leica DC500 microsystem, Flex scan L568).
Density gradient ultracentrifugation
After incubation in DMEM for 4 days, the astrocyte culture medium was collected, centrifuged at 2,270 × g for 15 min in a 50-ml plastic tube to exclude cell debris, and adjusted to a discontinuous sucrose gradient. A discontinuous sucrose gradient was prepared in a 1.5 × 9.6-cm ultracentrifuge tube (S303276, 13PET tube, Hitachi, Japan) from the bottom to the top, with 2 ml of sucrose at a density of 1.30 g/ml, 3 ml at 1.21 g/ml, 3 ml at 1.06 g/ml, and 4 ml at 1.006 g/ml medium. The sample in the sucrose gradient was then centrifuged in an RPS40T-256 rotor (Himac CP85β Instruments, Hitachi, Japan) at 4°C for 65 h at 100,000 × g. Following density gradient centrifugation, 1.2 ml fraction were collected with a micropipette from the top of the gradient, and fraction flanking the interphase separating to neighboring sucrose layer were pooled together for a total of 11 fractions. The densities of the fractions were determined by measuring the weight of 100μl of each fraction using a micropipette. The final fraction was stirred to re-suspend the pellet. The cholesterol content in each fraction was determined as described below.
Cholesterol distribution in the sucrose gradient fraction
By the using of sucrose gradient fraction aliquots of 1.0 ml each of the fraction media were transferred to clean glass tubes containing 4.0 ml of chloroform/methanol (2 : 1 v/v). The organic phase was separated from the aqueous phase by centrifugation 3,000 rpm, 4°C for 15 min, after centrifugation cholesterol containing phase were remove from the bottom site to another tube and dried under N2 gas. The dried cholesterol were then dissolved in 300μl of isopropyl alcohol, each sample was transferred onto 96-well polypropylene plates (Corning Glass) and dried under air flow. The dried content of cholesterol was determined using Amplex® Red Cholesterol assay kit (Invitrogen, UK), respectively.
Exosome isolation and identification by electron microscopy
Astrocytes were culture in serum free medium in 25-cm2 flask for four days, and then the culture media were harvested and centrifuged at 2,000×
Cell treatment and immunoblotting
For the treatment of the culture, Aβ1 - 42 was added directly to the medium to produce final concentrations of Aβ1 - 42 at 2μM. Synthetic Aβ1 - 42 (Peptide Institute, Osaka, Japan) was dissolved in 0.1% NH3, and diluted with culture medium immediately before use. SP600125 (ab120065, JNK inhibitor) was added to the culture medium at final concentrations of 1μM, 10μM, and 20μM in the presence or absence of Aβ1–42. The cultures were incubated for the time required for each experiment, and the condition medium was collected. The cells were washed in cold PBS for twice and collected using scraper in 100μl of RIPA buffer containing cocktail of protease inhibitors (Roche Diagnostica, Indianapolis) and phosphatase inhibitors (Wako Co. Ltd., Japan). The cell lysates were transferred to a 1.5-ml microtube. Those lysates were homogenized on ice by glass homogenizer and resultant homogenate was centrifuged at 12,000×rpm for 5 min at 4°C to separate the solution from the pellet fraction. Equal volume of conditioned medium and equal amount of homogenate protein were mixed with the sampling buffer consisting of 100 mM Tris-HCl (pH 7.4), 10% glycerol, 4% SDS, 10% mercaptoethanol, and 0.01% bromophenol blue, and were analyzed by 12.5% Tris/Tricine SDS-PAGE. The separated proteins were electrophoretically transferred onto PVDF membranes (Immobilon, Millipore), using a transfer buffer (0.1 M Tris, 0.192 M glycine, and 20% methanol). Membranes were then incubated in a blocking solution consisting of 5% powdered milk in TBST [10 mmole/L Tris-HCl (pH 8.0), 150 mmole/L NaCl, and 0.1% Tween 20] for overnight at 4°C, followed by immunoblotting with the respective antibodies. The first antibodies used were rabbit anti-flotillin 1 polyclonal antibody (Sigma-Aldrich, USA), mouse anti-HSP90 monoclonal antibody (BD Bioscience, Franklin Lakes, USA), rabbit anti-apolipoprotein E polyclonal antibody (EMD Millipore, Billerica, USA), rabbit anti-SAPK/JNK polyclonal antibody (Cell Signaling, USA), mouse anti-phospho-SAPK/JNK (Thr183/Thr185) monoclonal antibody (Cell Signaling, USA), rabbit anti-p44/42 MAPK monoclonal antibody (Cell Signaling, USA), rabbit antiphospho-p44/p42 MAPK polyclonal antibody (Cell Signaling, USA), rabbit anti-Akt polyclonal antibody (Cell Signaling, USA), rabbit anti-phospho-Akt polyclonal antibody (Cell Signaling, USA), and mouse anti-a-tubulin monoclonal antibody (Sigma Aldrich, USA).
Statistical analysis
Immunopositive bands visualized by western blot analysis were quantified using image analysis software (ImageJ 1.46r; Java 1.6.0-20 [64 bit]). Results are expressed as mean±S.E.D.. Statistical analysis was performed using the student‘s t test, with p < 0.05 deemed as statistically significant. All experiments were performed at least three times.
RESULTS
The characteristics of the exosome and cholesterol particles released into the serum-free media from cultured astrocytes were examined. The results of analysis using density gradient ultracentrifugation of the cultured media, as shown in Fig. 1A and B, shows that the distribution of exosome (fractions 7, 8, and 9) and apoE-HDL (fractions 5, 6, and 7) as indicated by flotillin and apoE were different. The peak of exosome distribution was in fraction 8, and that of apoE-HDL was fractions 6. They show that most of the cholesterol are distributed similarly in the fractions with densities of 1.14–1.200 (fractions 5–7). They also show that smaller amounts of cholesterol are distributed in the fractions having densities of 1.200–1.300 g/ml (fractions 8 and 9). These results show that the major parts of cholesterol were present in the lighter density fractions 5–7, the densities of which corresponded to those of HDL, and the minor part of cholesterol peak was present in the heavier density fractions 8 and 9, which corresponded to those of exosome (Fig. 1B). We also performed immunoblot analysis of each fraction using anti-apoE antibody and antibodies against exosome specific marker protein flotillin and HSP90. ApoE was detected mainly in fractions 5, 6, and 7, which were separated by sucrose gradient ultracentrifugation, and these were consistent with the major peak of cholesterol distribution. Flotillin and HSP90 were detected in the fraction 7, 8, and 9, which is consistent with the minor peak of lipid distribution, both of are shown in Fig. 1A, respectively. These results demonstrated that exosome and apoE-HDL were successfully separated by density gradient ultracentrifugation, that is, apoE-HDL particles are recovered in lighter fractions, while exosome is recovered in heavier fractions. Next, we performed EM imaging of isolated exosome in the fraction 8 using the exosome isolation kit (Invitrogen, USA) from serum-free conditioned medium of astrocyte cultures (Fig. 1C).
We examined that the effect of Aβ1 - 42 on release of exosome from conditioned medium of astrocyte cultures. Astrocytes were incubated for four days with the treatment of Aβ1 - 42 at a concentration of 2 μM. Because effect of Aβ on flotillin release was saturated at 2 μM (Supplementary Figure 1). The treatment of Aβ1 - 42 decreased the release of exosome demonstrated by release of specific markers such as flotillin and HSP90 in the conditioned medium in astrocyte cultures without changing cellular levels of flotillin and HSP90 (Fig. 2A). Whereas, apoE-HDL release remained unchanged, which is shown in Fig. 2A. Quantitative analysis of intensities of the bands representing flotillin, HSP90, and apoE, shows that Aβ1 - 42 treatment significantly decreased the level of exosome release from astrocytes compare with control, as shown in Fig. 2B. We further analyzed cellular molecules, which may be involved in exosome release. We have observed that the treatment with Aβ1 - 42 significantly enhanced the levels of phosphorylated form of JNK (p-JNK) of p46, whereas the levels of p54 p-JNK remained unchanged. With respect to the total JNK (pan-JNK), Aβ1 - 42 had no effect on the levels of both p46 and p54 of pan-JNK (Fig. 3A and B). We also analyzed others molecules including pan-ERK, phospho-ERK, pan-Akt, and phospho-Akt; however, their levels remained unchanged (Fig. 3A and B).
To determine whether the increased levels of p-JNK induced by Aβ1 - 42 treatment reduced exosome release, we treated cells with JNK-specific inhibitor, SP600125, at varying concentrations of 1, 10, and 20μM, in the presence or absence of Aβ1 - 42. SP600125 treated astrocytes showed the deceased levels of exosome release demonstrated by flotillin release, which was induced by Aβ1 - 42 treatment, were significantly recovered when the p-JNK signals were attenuated by the inhibitor (Fig. 4A, B). A JNK inhibitor recovered the decrease levels of exosome induced by Aβ1 - 42 treatment to levels similar those of control, suggesting that Aβ1 - 42 inhibits exosome release via stimulation of JNK signal pathway. We have found that enhanced phosphorylation of JNK was induced by Aβ1 - 42 at early as 1 h after the commencement of treatment without changing the level of pan-JNK (Fig. 4C). In this case, flotillin level was too low to detect by western blot analysis (data not shown).
Next, we examined whether that SP600125, has cytotoxic effect on astrocytes. The LDH concentration in the cell culture media was determined after four days incubation. As shown in Fig. 5A, there are no morphological difference in cultured astrocytes with each treatment. Released LDH levels in the media of cells collected at 1 h and 4 days after the commencement of the treatment showed that there were no significant difference between the cultures examined (Fig. 5B, C).
DISCUSSION
In this study, we report the novel findings: 1) that exosome and apoE-HDL were successfully separated by density gradient ultracentrifugation, which was shown by distribution of their specific markers, flotillin and HSP90, and cholesterol, and morphological analysis using electron microscopy; 2) that exosome release was significantly reduced by Aβ1 - 42 treatment in cultured astrocytes; 3) that the reduction in the levels of exosome release caused by Aβ1 - 42 was accompanied by an increased JNK phosphorylation; and 4) that a JNK inhibitor restored the decreased levels of exosome release induced by Aβ treatment to levels similar to those of control. Whereas, apoE-HDL release remained unchanged. These lines of evidence suggest that Aβ1 - 42 inhibits exosome release via stimulation of JNK signal pathway in astrocytes. Because the effects of Aβ on astrocytes have not been understood well, our findings may provide novel molecular insight into Aβ-inducing impairments in the brain.
The previous studies have shown that cellular signal pathways are stimulated by Aβ treatment in culture. In particularly, Aβ stimulates MAP kinase family including ERK, JNK, and p38, in neuronal cells [18, 19] and also in astrocytes [20]. It has been shown that Aβ induced elevation of MAP kinase subfamily activity, including ERK, p38, and JNK in vivo [21–23]. Among the family, it has been shown that inhibition of JNK kinase activation reverses AD phenotypes in APPswe/PS1dE9 mice [24] and completely rescued memory impairments as well as the long-term potentiation deficits of TgCRND8 mice [25].
In support of these findings, our present study has demonstrated that Aβ1 - 42 stimulated JNK cascade leading to cause reduction of exosome release. However, other signal cascades including ERK or Akt was not activated by Aβ treatment in cultured astrocytes. It has been shown that exosome plays a key role in Aβ clearance [9–11, 26]; however, the effect of Aβ on exosome generation/release has not been examined. In the present study, we found that extracellular Aβ impairs exosome release by activating JNK phosphorylation. It may be possible that reduced exosome release induced by Aβ, in turn, enhances Aβ accumulation and its toxicity leading to exacerbation pathologies of AD. These lines of evidence suggest that the interaction of Aβ and exosomes is bi-directional.
The mechanism underlying formation and secretion of exosome is still unclear. However, there have been reports showing that exosome release is modulated by sphingolipid dependent manner [9] and activation of sphingosine 1-phosphate receptors mediate maturation of exosome [27]. Other lines of studies have shown that exosome release is regulated by wnt3 signaling [28], and that exosome secretion is regulated by cellular Ca(2)(+)-dependent manner [29, 30]. However, the involvement of JNK signaling in exosome release has not been reported. In correlated with previous findings that JNK cascade is activated in cultures and in AD brains [31, 32], here we found that JNK cascade activated by Aβ, would play a key role in reduction of exosome release in Aβ-treated astrocytes.
It is known that both exosomes and apoE-HDL have shown to be associated with Aβ and to be involved in Aβ removal from the extracellular space of brain [10, 33]. The present study has demonstrated that two Aβ-removal machines are isolated by ultra-centrifugation and their release was differently regulated, that is, apoE release is not affected by Aβ and JNK signaling (Fig. 2 and data not shown). These results suggest that there may be different approaches for development of AD therapy with different approaches by modulating exosome release and apoE-HDL generation.
