We used mouse MN-like NSC-34 cells stably expressing wt and mSOD1, which were a kind gift from Júlia Costa (ITQB, Universidade Nova de Lisboa, Portugal). Mouse NSC-34 is a hybrid cell line produced by the fusion of MNs from the spinal cord embryos with N18TG2 neuroblastoma cells (Cashman et al., 1992) that exhibit properties of MNs after differentiation and maturation protocols. Thus, NSC-34 cells were grown in proliferation media [Dulbecco's modified Eagle's medium (DMEM) high glucose with glutamine, w/o pyruvate, supplemented with 10% of fetal bovine serum (FBS) and 1% of Penicillin/Streptomycin] and selection was made with Geneticin 418 sulfate (G418) at 0.5 mg/ml (Vaz et al., 2015). Medium was changed every 2–3 days. Culture plates were coated with Poly-d-Lysine (10 μg/ml) before plating the cells. Cells were seeded at a concentration of 5 × 10⁴ cells/ml and maintained at 37°C in a humidified atmosphere of 5% CO₂. After 48 h in proliferation medium, differentiation was induced by changing medium for DMEM-F12 plus 1% of FBS-exosome depleted, 1% of non-essential amino acids (NEAA), 1% of Penicillin/Streptomycin and 0.5 mg/ml of G418, as indicated by Cho et al. (2011). Cells were maintained in culture with differentiation medium for 4 days in vitro (DIV) to induce SOD1 accumulation (Vaz et al., 2015). DMEM, DMEM-F12, FBS, Penicillin/Streptomycin, and NEAA were purchased from Biochrom AG (Berlin, Germany). G418 was obtained from Gibco/Calbiochem (Darmstadt, Germany), and Poly-d-Lysine and RPMI were from Sigma-Aldrich (St. Louis, MO, USA).

Exosomes were isolated from the extracellular media of wt and mSOD1 NSC-34 cells, as currently in use in our lab (Cunha et al., 2016). Briefly, the culture media (20 ml) of the NSC-34 cells differentiated for 4 DIVs was centrifuged at 1,000 g for 10 min to remove cell debris. Then, the supernatant was transferred to another tube and centrifuged again at 16,000 g for 60 min, to separate microvesicles (size ~1,000 nm). The recovered supernatant was subsequently filtered in a 0.22 μm pore filter, and further centrifuged in the Ultra L-XP100 centrifuge (Beckman Coulter Inc., California, USA) at 100,000 g for 120 min to pellet exosomes (size ~100 nm). The pellet of exosomes was then resuspended in phosphate-buffered saline (PBS) and centrifuged one last time at 100,000 g for 120 min, in order to wash the pellet. All centrifugations were performed at 4°C. Characterization of exosomes in terms of size and concentration was performed by Nanoparticle tracking analysis (NTA) using the Nanosight, model LM10-HSBF (Malvern, UK) and the NTA software version 3.1. Transmission electron microscopy (TEM) technique used the Jeol JEM 1400 Transmission Electron Microscope (Peabody, MA, USA). Western blot analysis was performed as usual in our lab (Vaz et al., 2015) to evaluate the expression of alix, flotillin-1 and CD63 by using 20 μg of total protein and specific antibodies (mouse anti-Alix, Cell Signaling, #2171; mouse anti-flotillin-1, BD Biosciences, #6108; goat anti-CD63, Santa Cruz Biotechnology, #sc-31214). Normalization was made by using Amido Black staining as loading control. To evaluate total RNA and microRNA, the final pellet containing exosomes was resuspended in lysis buffer, and exosomal RNA extracted with the miRCURY Isolation Kit-Cell (Exiqon), as described below.

Mouse microglial N9 cell line, a popular retroviral-immortalized cell line resulting from the immortalization of microglia isolated from the cortex of CD1 mouse embryos (Righi et al., 1989), was a gift from Teresa Pais (Institute Gulbenkian de Ciência, Oeiras, Portugal). This cell line shows diverse features similar to microglia in primary cultures, such as migration, phagocytosis and inflammation-related features (Fleisher-Berkovich et al., 2010). Indeed, N9 cells were shown to respond similarly to primary microglial cells derived from the same mouse strain, when treated with LPS (Nikodemova and Watters, 2011), as recently demonstrated by us (Cunha et al., 2016). Cells were cultured in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% of FBS, 1% of L-glutamine (Biochrom AG) and 1% of Penicillin/Streptomycin. Cells were seeded at a concentration of 1 × 10⁵ cells/ml and maintained at 37°C in a humidified atmosphere of 5% CO₂.

NSC-34 cells differentiated for 4 DIV were cocultured with N9 cells in RPMI medium for 24 h, at 37°C in a humidified atmosphere of 5% CO2. We used the normal proportion of microglia and neurons in the CNS (ratio 1:1) (Silva et al., 2011).

In the coculture model, the N9 microglial cells were plated in coverslips with paraffin wax feet, as described by Phatnani et al. (2013). The coverslips containing microglia were placed inverted over the layer of wt or mSOD1 NSC-34 MNs and maintained separated from such layer by the paraffin spots, thus avoiding direct contact between the two types of cells. Cells were plated in the same proportion (1:1) and maintained in coculture for 24 h. At the end of incubation, exosomes in the supernatant of NSC-34 (wt or mSOD1) with N9 cocultures were isolated as described above. To obtain PKH67 labeled fluorescent exosomes, the isolated exosomes were resuspended in PBS and mixed with an equal volume of PKH67 probe solution for 5 min at room temperature, using the PKH67 Fluorescent Cell Linker Kit (#MINI67, Sigma-Aldrich), as described by Dutta et al. (2014). Then, isolated exosomes were resuspended in RPMI medium and added either to N9 cells + wt NSC-34 MNs, or to N9 cells + mSOD1 NSC-34-MNs, using the above described cocultured system, for an additional period of 24 h and in a ratio of 1:1 (v/v). After incubation, NSC-34 MNs and N9 microglia were collected separately, and fixed with 4% (w/v) paraformaldehyde in PBS and cell nuclei were stained with Hoechst 33258 dye. UV and fluorescence images (original magnification: 630X) were acquired per sample by using Zen 2012 (blue edition, Zeiss) software.

N9 cells were plated for 24 h before incubation with exosomes from wt and mSOD1 NSC-34 MNs, at a concentration of 1 × 10⁵ cells/ml. Exosomes from NSC-34 cells were resuspended in RPMI medium and incubated in N9 microglial cells, using a fixed ratio of 1:1. To assess exosome internalization by N9 cells, exosomes were labeled with PKH67, as previously described.

To evaluate the effects produced by exosomes on N9 microglia, we incubated the cells in RPMI medium in the absence (control), or in the presence of exosomes, either from wt NSC-34 MNs, or from mSOD1 NSC-34 MNs. N9 microglia responses were evaluated at 2, 4, and 24 h. These different time-points of incubation were accomplished to evaluate the effects produced by an early (2 and 4 h) and a lasting (24 h) period of exosome interaction with naïve N9 microglia. At the end of each incubation period, medium free of cellular debris was collected to assess the released soluble factors. Attached cells were: (i) fixed for 20 min with freshly prepared 4% (w/v) paraformaldehyde in PBS for immunocytochemical studies or to detect PKH67 labeled fluorescent exosomes; (ii) fixed with Fixing Solution for cellular senescence assays; (iii) used to extract total RNA using TRIzol® reagent, according to the manufacturer's instructions; or (iv) collected in Cell Lysis Buffer (Cell Signaling Beverly, MA, USA) plus 1 mM phenylmethylsulfonyl fluoride (PMSF, Sigma) for western blot analysis.

To evaluate the phagocytic ability of N9 microglia, cells were incubated with 0.0025% (v/v) fluorescent latex beads (#L1030, Sigma-Aldrich), diameter 1 μm, for 75 min at 37°C. For immunofluorescence detection, N9 cells were fixed for 20 min with freshly prepared 4% (w/v) paraformaldehyde in PBS and a immunocytochemical technique was performed as usually in our lab for microglial cells (Caldeira et al., 2014; Cunha et al., 2016). N9 microglia were immunostained with rabbit anti-ionized calcium-binding adaptor molecule 1 (Iba1) (1:250, #019-19741, Wako), and nuclei were counterstained with Hoechst 33258 dye (blue). UV and fluorescence images of ten random microscopic fields (original magnification: 400X) were acquired per sample using an AxioCam HR camera adapted to an AxioScope A1® microscope (Zeiss, Germany), and Zen 2012 (blue edition, Zeiss) software. The number of cells with ingested beads and total cells were counted using ImageJ software to determine the percentage of phagocytosing cells (Silva et al., 2010).

Activity of senescence-associated beta-galactosidase (SA-β-gal) was determined as a biomarker of microglia-like senescence by using the Cellular Senescence Assay Kit (#KAA002RF, Merck Millipore, Darmstadt, Germany), according to the manufacturer's instructions. Nuclei were counterstained with hematoxylin (Merck). The number of total cells was counted in 10 microscopic fields with ImageJ software (original magnification: 400X) acquired to observe the complete well using Leica IM50 software and Leica DFC490 camera (Leica Microsystems, Wetzlar, Germany), adapted to an AxioSkope HBO50 microscope (Zeiss). The number of turquoise stained microglia (SA-β-gal-positive cells) was counted to determine the amount of senescent cells relatively to the total cell number (percentage) (Caldeira et al., 2014).

For immunofluorescence detection of nuclear factor-kappa B (NF-κB) translocation, cells were fixed as above and a standard immunocytochemical technique (Fernandes et al., 2006) was carried out using a rabbit anti-p65 NF-κB subunit antibody (1:200, #sc-372, Santa Cruz Biotechnology®, CA, USA). N9 microglial nuclei were stained with Hoechst 33258 dye. UV and fluorescence images of ten random microscopic fields (original magnification: 400X) were acquired as indicated for phagocytosis assay. NF-κB positive cells were identified by the ratio between the mean gray value of the nucleus and the mean gray value of the whole cell, using ImageJ software. A threshold was defined for each individual experiment and cells above that value were considered positive for NF-κB. Positive cells and total cells were counted to determine the percentage of NF-κB positive nuclei.

Levels of nitric oxide (NO) were estimated by measuring the concentration of nitrites (NO2-), a product of NO metabolism, in the extracellular media of N9 cells incubated in the absence, or in the presence of exosomes released by wt and mSOD1 NSC-34 cells. Extracellular media, free from cellular debris, were mixed with Griess reagent [1% (w/v) sulfanilamide in 5% H₃PO₄ and 0.1% (w/v)N-1 naphtylethylenediamine, all from Sigma-Aldrich, in a proportion of 1:1 (v/v)] in 96-well tissue culture plates for 10 min in the dark, at room temperature, as we published (Vaz et al., 2015). The absorbance at 540 nm was determined using a microplate reader. A calibration curve was used for each assay. All samples were measured in duplicate.

Activities of MMP-9 and MMP-2 were determined in the N9 extracellular media after incubation in the absence and in the presence of exosomes released by wt and mSOD1 NSC-34 cells by performing a SDS-PAGE zymography in 0.1% gelatin-10% acrylamide gels, under non-reducing conditions, as usual in our lab (Silva et al., 2010). Briefly, after electrophoresis, the gels were washed for 1 h in a solution containing 2.5% Triton-X-100 to remove SDS and to renature the MMP species in the gel and then incubated at 37°C to induce gelatin lysis (buffer: 50 mM Tris pH 7.4, 5 mM CaCl₂, 1 μM ZnCl₂) overnight. Gels were then stained with 0.5% Coomassie Brilliant Blue R-250 (Sigma-Aldrich) and destained in 30% ethanol/10% acetic acid/H₂O (v/v). Gelatinase activity, detected as a white band on a blue background, was measured using computerized image analysis (Image Lab™, Bio-Rad).

Total RNA was extracted from exosome-treated N9 microglia using TRIzol® (LifeTechnologies, Carlsbad, CA, USA), according to manufacturer's instructions. Total RNA was quantified using Nanodrop ND-100 Spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and conversion to cDNA was performed with SensiFAST™ cDNA synthesis (#BIO-65054, BIOLINE, London, UK). Quantitative RT-PCR (qRT-PCR) was performed on a 7300 Real Time PCR System (Applied Biosystems) using a SensiFAST™ SYBR® (Hi-ROX, #BIO-92002/S, BIOLINE). qRT-PCR was accomplished under optimized conditions: 50°C for 2 min and 95°C also for 2 min, followed by 40 cycles at 95°C for 5 s and 62°C for 30 s. In order to verify the specificity of the amplification, a melt-curve analysis was performed, immediately after the amplification protocol. Non-specific products of PCR were not found in any case. Results were normalized to β-actin and expressed as fold change. The sequences used for primers are represented in Table S1 (Supplementary Material). Relative miRNA concentrations were calculated using the ΔΔCT equation. RNA inside exosomes was extracted using miRCURY Isolation Kit - Cell (#300110, Exiqon, Vedbaek, Denmark). For miRNA analysis, conversion of cDNA was achieved with the universal cDNA Synthesis Kit (#203301, Exiqon), as described by Cardoso et al. (2012) and currently implemented in our lab (Caldeira et al., 2014; Cunha et al., 2016), following manufacturer's recommendations. The miRCURY LNA™ Universal RT miRNA PCR system (#203403, Exiqon) was used in combination with predesigned primers (Exiqon), represented in Table S1 (Supplementary Data), using SNORD110 as reference gene. The reaction conditions consisted of polymerase activation/denaturation and well-factor determination at 95°C for 10 min, followed by 50 amplification cycles at 95°C for 10 s and 60°C for 1 min (ramp-rate 1.6°/s). Relative miRNA concentrations were calculated using the ΔΔCT equation. All samples were measured in duplicate.

Cells were collected in Cell Lysis Buffer, as usual in our lab (Vaz et al., 2015). Briefly, total cell extracts were lysed for 5 min on ice with shaking, collected with cell scrapper, and sonicated for 20 s. The lysate was then centrifuged at 14,000 g for 10 min at 4°C, and the supernatants were collected and stored at −80°C. Protein concentration was determined using a protein assay kit (Bio-Rad, Hercules, CA, USA) according to manufacturer's specifications. Then, equal amounts of protein were subjected to SDS-PAGE and transferred to a nitrocellulose membrane. After blocking with 5% (w/v) nonfat milk solution, membranes were incubated with primary antibody mouse anti-HMGB1 (1:200, from Biolegend) diluted in 5% (w/v) BSA overnight at 4°C, followed by the secondary antibody goat anti-mouse HRP-linked (1:5,000, sc-2005, Santa Cruz Biotechnology®) diluted in blocking solution. Chemiluminescence detection was performed by using LumiGLO® reagent (Cell Signaling) and bands were visualized in the ChemiDoc^TM XRS System (Bio-Rad). The relative intensities of protein bands were analyzed using the Image Lab^TM analysis software (Bio-Rad).

Results of at least seven independent experiments were expressed as mean ± SEM. Comparisons between the different parameters evaluated in wt and mSOD1 NSC-34 MNs were made via one-tailed Student's t-test for equal or unequal variance, as appropriate. In addition, we have performed unpaired t-test with Welch's correction when the variances were different between 2 groups. Comparison of more than two groups was done by one-way ANOVA followed by multiple comparisons Bonferroni post-hoc correction using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA). P-values of 0.05 were considered statistically significant.

Lately, miRNAs are emerging as potent fine-tuners of neuroinflammation and reported to be dysregulated in ALS (Koval et al., 2013; Butovsky et al., 2015). However, the contribution of individual miRNAs to neurodegeneration and neuroinflammation in ALS disease remains to be elucidated. We decided to investigate alterations on specific inflamma-miRs in the mSOD1 NSC-34 MNs and in their derived exosomes, using for comparison wt cells and their correspondent exosomes.

Cellular differentiation was induced and exosomes were isolated from extracellular media after 4 DIV, to induce SOD1 accumulation, as previously reported (Vaz et al., 2015). Recent data from our group, obtained by dynamic light scattering and TEM, indicated that the nanoparticle size of exosomes in the extracellular media of NSC-34 cells was approximately of ~140 nm (Vaz et al., 2016). We have additionally performed a more detailed characterization by NTA analysis, having observed an average diameter size of ~130 nm and a concentration of ~5.3 × 10⁸ particles/ml (Figures 1A,B), independently of being originated from wt or mSOD1 NSC-34 MNs. Western blot analysis showed the presence of exosomal marker proteins, namely Alix, Flotillin-1, and the tetraspanin CD63 in lysates obtained after the exosomal isolation procedure (Figure 1C). In addition, exosomes isolated from wt and mSOD1 cells exhibited the characteristic cup-shape morphology by TEM (Figure 1D). It is important to refer that similar total RNA concentrations were found for exosomes isolated from wt or mSOD1 cells (Figure 1E). When compared to their wt counterpart, the mSOD1 MNs revealed an increased cargo in miR-124 (Figure 1F), which is the predominant miRNA in the brain, namely in neurons (Sun et al., 2015). No detectable amounts of miR-146a or miR-155 were found in either of them.

Since exosomes are known to contain miRNAs that can be transferred into recipient cells causing changes in their functionality (Valadi et al., 2007; Alexander et al., 2015), we decided to evaluate the miRNA expression in the MN-derived exosomes. Interestingly, exosomes from mSOD1 MNs evidenced to be enriched in miR-124, recapitulating their cells of origin, and again no expression of miR-146a or miR-155 was detected in either type of exosomes.

NSC-34 cells were demonstrated to uptake extracellular vesicles from a muscle cell line (C2C12) (Madison et al., 2014) and toxic exosomes were shown to be transferred from mSOD1 astrocytes to MNs (Basso et al., 2013). To determine whether exosomes released from mSOD1 and wt NSC-34 cells after 24 h incubation were similarly transferred into N9 microglial cells, we fluorescently labeled exosomes with PKH67, as previously described (Figure 2A). No differences were found in such distribution. To further understand whether mixed exosomes in the supernatant of NSC-34(wt)+N9 and of NSC-34(mSOD1)+N9 cocultures were preferentially captured by MNs or by N9 microglia, we isolated exosomes from the culture medium, labeled them with PKH67, incubated the exosomes with matched NSC-34(wt)+N9 and NSC-34(mSOD1)+N9 cocultures, and assessed the distribution of PKH67-labeled exosomes in either one of the cells. In Figure 2B, it is clearly shown that N9 microglia are the preferential recipient cells for the mixed exosomes released from both donor cell types. Indeed, intracytoplasmic green exosomes are only visible in N9 microglia, indicating that these cells are more likely to incorporate and to be functionally influenced by exosomes, as compared to NSC-34 MNs.

HMGB1 is a ubiquitous nuclear protein that is increasingly expressed and released by injured neurons and activated microglia (Gao et al., 2011; Brites and Vaz, 2014; Cunha et al., 2016). To evaluate whether mSOD1 MNs that were shown to be dysfunctional (Vaz et al., 2015) expressed increased HMGB1 and influenced the expression of HMGB1 in N9 microglia, we assessed its expression levels in both wt and mSOD1 NSC-34 MNs, as well as in N9 microglia when in coculture with NSC-34 MNs, in the absence and in the presence of a surcharge of exosomes isolated from the coculture supernatant (Figure 3).

We observed up-regulated HMGB1 gene and protein expression in mSOD1 NSC-34 MNs, as compared to wt cells (Figures 3A,B). To note, however, that the exosomes per se did not produce noticeable alterations in the N9 microglia HMGB1 gene expression in the absence of mSOD1 NSC-34 MNs (data not shown). The decrease observed in N9 microglia at 24 h incubation with mSOD1 exosomes (Figure 3C), suggests either degradation/cleavage of the protein or its release into the cell supernatant. In addition, although not significant, we observed a slight elevation in HMGB1 mRNA levels in the N9 microglia exposed to mSOD1 NSC-34 MNs. Significant increase in the HMGB1 gene expression was, however, obtained in N9 cells cocultured with mSOD1 MNs (without cell contact) surcharged with exosomes isolated from a 24 h matching coculture system (Figure 3D, p < 0.05), emphasizing secretome relevance in the signaling mechanisms underlying HMGB1-induced microglial activation.

Therefore, we next decided to evaluate if the internalization of wt and mSOD1 NSC-34 MN-derived exosomes in N9 microglia caused changes in the cell dynamic properties and function, determined by specific cell polarization phenotypes.

The proinflammatory transcription factor NF-κB has been shown to activate numerous molecules and factors, and to be critical in the regulation of neuroinflammation-associated disease pathogenesis, (Shih et al., 2015). Thus, we evaluated whether mSOD1 exosomes were able to activate NF-κB in N9 cells, a process implicated in MN death in ALS (Frakes et al., 2014).

We observed that although a slight effect was produced by exosomes derived from wt NSC-34 MNs on the NF-κB translocation into the nucleus, only those from mSOD1 NSC-34 MNs activated significantly and persistently (from 2 to 24 h incubation) the NF-κB signaling pathway (Figures 4A,B). This early and lasting NF-κB activation (Sen and Smale, 2010) suggest that distinct sets of genes are activated in N9 microglia upon interaction with exosomes released from ALS NSC-34 MNs.

We and others have previously shown that NO is a key player in MN degeneration in ALS (Drechsel et al., 2012; Vaz et al., 2015) and that increased generation of redox molecules (NO) and iNOS activation occurs in M1 polarized N9 microglia (Cunha et al., 2016). Increased NO production was observed after 2 h of incubation with exosomes only from mSOD1 MNs (Figure 4C). Such effect disappeared after 4 h incubation and even an inhibitory effect was produced by MN-derived exosomes at 24 h of incubation.

Activation of MMPs is another marker of neuroinflammation and elevation of MMP-9 and MMP-2 expression was observed in the spinal cord of SOD1G93A mice (Fang et al., 2010). Exosomes revealed to induce the MMP-2 activation whenever released from wt or mSOD1 MNs (Figure 4D). Intriguingly, only those from mSOD1 NSC-34 MNs were able to activate MMP-9 (Figure 4E), in accordance with our prior data showing such activation in mSOD1 NSC-34 MNs (Vaz et al., 2015). However, similarly to NO, this increase ceased over time, returning to basal levels.

Finally, we observed that the expression of the pro-inflammatory cytokines TNF-α and IL-1β was significantly upregulated and maintained until 4 h interaction, differently from the above mentioned inflammatory mediators, in cells exposed to exosomes from mSOD1 NSC-34 MNs, also disappearing after 24 h incubation (Figures 4F,G).

Based on these data we may assume that exosomes from the mSOD1 NSC-34 MNs transiently switch N9 microglia into a M1 polarized cell (Durafourt et al., 2012; Chhor et al., 2013). Since early or late NF-κB activation was shown to induce different sets of genes, by respectively encoding TNF-α, IL-8, MMP-9, or cell surface receptors, adhesion molecules and signal adapters (Tian et al., 2005), we next evaluated the effects produced on the expression of cell surface receptors.

To determine whether late NF-κB activation in microglia treated with mSOD1 exosomes was associated with the increased expression of membrane surface receptors, like TREM2, RAGE, and TLR4, we evaluated their gene expression levels in a time-dependent manner. Indeed, microglia was shown to express multiple receptors able to efficiently respond to external stimuli (Pocock and Kettenmann, 2007).

TREM2 receptor has been identified as a potential regulator of the microglial phenotype (Stefano et al., 2009) and found elevated in the spinal cord of ALS patients and SOD1G93A mice (Cady et al., 2014). As depicted in Figure 5A, increased expression of TREM2 gene in N9 microglia was evident after 24 h incubation with both wt NSC-34 MNs and mSOD1 MNs-derived exosomes, although some fluctuations were observed overtime. TREM2 overexpression has been related with suppression of neuroinflammation and microglia M2 polarization associated with increased phagocytic ability (Painter et al., 2015; Jiang et al., 2016).

RAGE is also a receptor found elevated in association with mSOD1 (Shibata et al., 2002). In the present study, it is clear its net elevation only in the N9 microglia treated for 24 h with exosomes from mSOD1 MNs (Figure 5B, p < 0.05 vs. wt NSC-34 MNs, and p < 0.01 vs. non-treated N9 microglia).

Besides RAGE, elevation of TLR4 was also identified in the spinal cord of sALS patients, mainly in glial cells (Casula et al., 2011). Both receptors play an important role in the regulation of innate and adaptive immunity during neuroinflammation. RAGE was recently indicated as enhancing TLR responses through binding and internalization of RNA (Bertheloot et al., 2016). Therefore, it was not surprising to find the same pattern of increased gene expression of TLR4 only in cells incubated for 24 h with exosomes released by mSOD1 NSC-34 MNs (Figure 5C).

At this point, our data indicate that exosomes from mSOD1 NSC-34 MNs determine an early inflammatory response on N9 microglia, which by releasing inflammatory mediators trigger the activation of RAGE/TLR signaling mechanisms and a second delayed stage of activation.

In order to fully understand the effect of mSOD1 NSC-34-derived exosomes in N9-microglia phenotypic diversity, we searched for pro- and anti-inflammatory markers expressed in M1 and M2 microglial phenotypes (Freilich et al., 2013; Brites and Vaz, 2014; Cunha et al., 2016), respectively.

Data showed that exosomes from mSOD1 NSC-34 cells trigger upregulation of the M1-associated markers iNOS and MHC-II after 2 and 4 h incubation, but not after 24 h interaction (Figures 6A,B), suggesting that N9 microglial cells switch their polarization after continued interaction with the mSOD1 exosomes. Attenuated immune response with reduced MHC-II levels was observed at 24 h incubation, indicating that later, after activation, N9 microglial cells may downregulate MHC-II synthesis, as observed for dendritic cells (Villadangos et al., 2001). Indeed, the gene expression of M2-related markers, such as IL-10 and Arginase 1 (Figures 6C,D), was found substantially enhanced at this time after treatment with mSOD1 exosomes.

To study the role of exosomal miR-124, and other cargo contents, in producing microglia dynamic changes we evaluated the expression of two anti-inflammatory miRNAs (miR-146a and miR-124) and the proinflammatory miR-155, a recognized inducer of the M1 polarization found increased in ALS patients and models (Koval et al., 2013; Liu and Abraham, 2013; Butovsky et al., 2015) in N9 microglial cells after the transfer of mSOD1 exosomes.

We observed that a prompt reduction of calming miRNAs by NSC-34 MN-derived exosomes (Figures 7A,B–2 h incubation) was followed by a marked and moderate selective elevation of miR-124 and miR-155, respectively, by mSOD1 exosomes (Figures 7A,C–24 h incubation). Surprisingly, both wt and mSOD1 exosomes produced a delayed increase in miR-146a expression.

The immediate decrease in the N9 microglial miR-124 and miR-146 upon interaction with exosomes, indicative of M1 (pro-inflammatory) in opposite to M2 (alternative) microglia subtype, may justify the acute upregulation of inflammatory mediators previously observed (Figures 4, 6) for both wt (not significant) and mSOD1 NSC-34 MN-derived exosomes (at least p < 0.05). In contrast, the marked elevation of miR-124 at 24 h incubation in the N9 microglia treated with mSOD1 exosomes may derive, at least in part, from its increased content in MNs and in their derived exosomes that are collected by the cells, thus skewing M1 to M2a polarization (Veremeyko et al., 2013).

The upregulation of both calming and inflammatory miRNAs at 24 h, subsequent to the transfer of mSOD1 exosomes into the N9 cells, is indicative of induction of different polarized microglia subtypes, representing heterogeneous classes of activated N9 microglia, including both M1/M2 phenotypes. Influence of these diverse and simultaneous states on the variable rate of ALS progression surely deserves further investigation.

After identifying that exosomes released by mSOD1 NSC-34 MNs induce phenotypical alterations, we asked whether the N9 microglial cells were disturbed in one of their most relevant neuroprotective functions, the phagocytic ability. Also based on the elevation of miR-146a at 24 h incubation produced by both sorts of exosomes in the N9 microglial expression, we questioned if increased cell senescence was produced. Actually, miR-146a was already suggested as a marker of a senescence-associated proinflammatory status (Olivieri et al., 2013; Caldeira et al., 2014).

As depicted in Figure 8, we observed a sudden and sustained decreased in the number of engulfed beads by N9 microglia upon interaction with mSOD1 exosomes, not noticed with those from wt NSC-34 MNs. Increased release of cytokines at 2 and 4 h incubation followed by upregulation of TLR4 after treatment with mSOD1 exosomes for 24 h, may account to this compromised N9 microglia function.

We have previously used the quantitative assay of SA-β-gal activity to evaluate the increase in the number of senescent-like microglia (Caldeira et al., 2014), as similarly proposed by other Authors (Debacq-Chainiaux et al., 2009). Results revealed that the percentage of positively stained cells increased after 24 h of incubation with exosomes from mSOD1 MNs, when compared with non-treated cells (Supplementary data, Figure S1). Therefore, and despite of causing less than 10% enhancement in the number of senescent-like N9 microglia, such occurrence may additionally contribute to microglia impaired function.

Overall these findings suggest that besides inducing microglia M1/M2 activation, exosomes from mSOD1 NSC-34 MNs also generate harmful effects by triggering the loss of microglia protective functions.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.