Antimicrobial responses of peripheral and central nervous system glia against Staphylococcus aureus

Staphylococcus aureus infections of the central nervous system are serious and can be fatal. S. aureus is commonly present in the nasal cavity, and after injury to the nasal epithelium it can rapidly invade the brain via the olfactory nerve. The trigeminal nerve constitutes another potential route of brain infection. The glia of these nerves, olfactory ensheathing cells (OECs) and trigeminal nerve Schwann cells (TgSCs), as well as astrocytes populating the glia limitans layer, can phagocytose bacteria. Whilst some glial responses to S. aureus have been studied, the specific responses of different glial types are unknown. Here, we compared how primary mouse OECs, TgSCs, astrocytes and microglia responded to S. aureus. All glial types internalized the bacteria within phagolysosomes, and S. aureus-conjugated BioParticles could be tracked with subtle but significant differences in time-course of phagocytosis between glial types. Live bacteria could be isolated from all glia after 24 h in culture, and microglia, OECs and TgSCs exhibited better protection against intracellular S. aureus survival than astrocytes. All glial types responded to the bacteria by cytokine secretion. Overall, OECs secreted the lowest level of cytokines, suggesting that these cells, despite showing strong capacity for phagocytosis, have immunomodulatory functions that can be relevant for neural repair.


Results
Intracellular survival of S. aureus differs between glial types. To investigate the susceptibility of different glial cell types to S. aureus infection, we compared the adhesion and invasion capacity of live S. aureus between primary OECs and TgSCs (PNS glia from the olfactory and trigeminal nerve, respectively) with astrocytes and microglia (CNS glia). We first exposed the cells to S. aureus for 1 h, and then applied antibiotics to the medium to inhibit extracellular survival of the bacteria after which cells/bacteria were incubated for a further 6 h and 24 h. At 1 h after commencement of the assay, OECs and TgSCs that were not inoculated with S. aureus exhibited typical bipolar morphology, while astrocytes had multi-branched broad morphology and microglia showed a resting or ramified morphology ( Fig. 1A-D). When the cells were inoculated with S. aureus, bacteria were present along the length of the processes of the OECs and TgSCs (Fig. 1E-F). While there appeared to be minor changes to the cell morphology, with some cells becoming broader while other became more elongated. Measurements of cell length revealed that the presence of S. aureus did not alter overall cell length of OECs, TgSCs and microglia at 1 h and 6 h (Fig. 2D). At 24 h, however, TgSCs exposed to S. aureus were significantly longer than control cells (Fig. 2D). In contrast, astrocytes underwent considerable morphological changes when exposed to the bacteria; their overall lengths were significantly greater at all timepoints (Figs. 1G, 2D). Microglia showed uptake of bacteria localised in the perinuclear area at 1 h (Fig. 1H). At 24 h though a significant decrease in size of infected microglia was observed (Fig. 2D). After the antibiotic protection media was added to prevent extracellular survival of the bacteria, the bacteria were clearly internalised within the cells at 6 h post exposure for all glial cells. Some bacteria were present in the processes of the cells (arrows, Fig. 1I-K), but majority of bacteria appeared to be localised in the perinuclear region (arrows with tails). At 24 h, the bacteria were mainly in the perinuclear region of OECs, TgSCs and microglia ( Fig. 1M-N) with few bacteria localised in the processes of OECs, TgSCs and microglia ( Fig. 1M-N,P). In contrast, while bacteria were in the perinuclear region of astrocytes, there appeared to be more bacteria present along the branches (arrows, Fig. 1O).
To quantify the amount of bacteria that adhered or were internalised into cells, cells were washed and then lysed at different times and lysates were plated onto selective BHI agar plates. We observed significant differences between types of glia regarding S. aureus adherence to the cells. Significantly more bacteria adhered to astrocytes than to TgSCs and microglia after 1 h ( Fig. 2A). OECs showed significantly higher bacterial adherence than microglia too ( Fig. 2A). Following antibiotic treatment at 1 h to prevent extracellular survival of bacteria, and subsequent analysis at 6 h post exposure to S. aureus, live bacteria could be isolated from all three glial types. There was, however, no difference in the amount of bacteria between the glial types (Fig. 2B). In contrast, at 24 h post exposure, there were significantly more bacteria in astrocytes as compared to OECs, TgSCs and microglia, suggesting a decrease in intracellular survival or a more efficient bacterial killing by OECs, TgSCs and microglia www.nature.com/scientificreports/ as compared to astrocytes (Fig. 2C). We analysed the invasion frequency (the percentage of the initially adhered bacteria that were inside cells at 6 h); there was no significant difference in invasion frequency between cell types (Fig. 2E). We also determined the percentage of bacteria that survived intracellularly (amount of bacteria isolated from inside cells at 24 h compared to 6 h) and found no significant difference (Fig. 2F). To account for changes in cell numbers during the assay (proliferation or death), a cell count was performed and intracellular survival was calculated on the cell numbers at 24 h (OECs, TgSCs and astrocytes showed no significant change in cell S. aureus attaching to and being internalised into OECs, TgSCs, astrocytes and microglia. Panels show primary cultures of OECs (left panels), TgSCs (second panels), astrocytes (third panels) and microglia (right panels) from S100β-DsRed mice, in which all glia express the DsRed protein. For OECs and TgSCs, red labelling shows DsRed; for astrocytes, red labelling shows GFAP (these cells express low levels of DsRed); for microglia, red labelling shows Iba-1 (these cells express low level of DsRed) and Hoechst for nucleus staining. (I-L) S. aureus in glia at 6 h post exposure, some bacteria were present in the processes (arrows) while larger amounts accumulated in the perinuclear region (arrows with tail). (M-P) At 24 h post exposure, most bacteria were present in the perinuclear region (arrows with tail) while some bacteria were also present in the processes (arrows). Scale bar: 20 μm. www.nature.com/scientificreports/ numbers; only microglia showed a significant decrease in cell number over the assay). Thus the effect observed in the results is due to glial responses to bacteria and not due to cell death occurring post infection.

Phagocytosis of S. aureus and pHrodo S. aureus BioParticles by glial cell types.
To determine whether the glia phagocytosed S. aureus and internalised the bacteria in lysosomes, we stained the cells with a lysosomal membrane protein 2 marker (LAMP-2) which is a protein component of the lysosome membrane. The lysosomes function to fuse to foreign particles to form a phagolysosome which then aids in their acidification and degradation. We observed that all glial cells showed positive LAMP-2 staining around the bacterial stains of S. aureus (Fig. 3) demonstrating that the bacteria was internalized or colocalized inside lysosomes.
To determine the time course of glial phagocytosis of S. aureus (resulting in the bacteria being internalized into lysosomes) we exposed the glia to pH-sensitive (pHrodo) S. aureus BioParticles, which exhibit fluorescence only when in an acidic environment (i.e. in phagolysosomes). The phagocytosis assay was performed using the four types of glia for increasing durations ranging from 30 min to 8 h, with imaging of the cells every 30 min (Fig. 4). The cells internalized an increasing amount of BioParticles over time (Fig. 4M). In microglia, a significantly higher proportion of cells internalized the BioParticles at 30 min, compared to astrocytes, TgSCs or OECs. Significantly more astrocytes and TgSCs internalised the BioParticles than OECs. BioParticle internalization into acidic cellular compartment differed between astrocytes and OECs after 3.5 h (Fig. 4M). At 4.5 h post exposure, TgSCs contained more phagocytosed particles than OECs (Fig. 4M) .
Glia produce multiple cytokines and chemokines after exposure to S. aureus. To   www.nature.com/scientificreports/ interferon γ (IFN-γ) and tumour necrosis factor α (TNF-α), with levels of IFN-γ being high in microglia than TNF-α and lower than those for OECs; TgSCs and astrocytes ( Fig. 5A-H). The cells also responded to S. aureus with secretion of interleukin 6 (IL-6), and the anti-inflammatory and immune-regulatory cytokine interleukin 10 (IL-10). The amount of IL-6 was consistently higher than the amount of IL-10 ( Fig. 5) in OECs, TgSCs and astrocytes; in contrast microglia expressed higher IL-10 than IL-6. Microglia produced significantly higher levels of IFN-γ and IL-10, TgSCs showed significantly higher levels of TNF-α and IL-6, than the other glial types, with some variations between time-points ( Fig. 5 and Table 1). Overall, production of TNF-α by OECs, TgSCs and astrocytes peaked at 6 h post exposure and for microglia at 1 h. The levels of IFN-γ and IL-10 were highest at 1 h for microglia as compared to OECs, TgSCs and astrocytes and continued till 24 h. Whilst the levels of IFN-γ, IL-6 and IL-10 were highest at 24 h for OECs, TgSCs and astrocytes. The glia also produced chemokines, including high levels of CXCL1, monocyte chemoattractant protein 1 (MCP-1), macrophage inflammatory proteins 1α and β (MIP-1α, MIP-1β), regulated upon activation-normal www.nature.com/scientificreports/ T cell expressed and presumably secreted (RANTES) and low levels of eotaxin, in response to S. aureus challenge for OECs, TgSCs and astrocytes (Fig. 6). The levels of all these chemokines increased gradually over time from 1 to 24 h in OECs, TgSCs and astrocytes (Fig. 6). Overall, TgSCs produced higher levels of chemokines than astrocytes, which in turn secreted higher levels than OECs (with individual variations depending on chemokine and/or time-point) ( Table 1). Microglia showed high levels of eotaxin as compared to OECs, TgSCs and astrocytes all throughout the time points. Eotaxin, MIP-1 β and MCP-1 all were significantly high at 1 h for microglia then decreased at 6 h and increased again at 24 h (Fig. 6).

Discussion
Certain pathogens can enter the CNS via the nerves extending between the nasal cavity and the brain (the olfactory and trigeminal nerves) 2 . S. aureus is one of these microbes and can rapidly invade the brain (olfactory bulb) after injury to the nasal epithelium 12,13 . Previous studies have shown that the glia of the olfactory nerve, olfactory ensheathing cells (OECs), respond to S. aureus by nuclear translocation of NFκB accompanied by NO and nitrite production 12 . Capacity for intracellular survival inside OECs and/or TgSCs has previously been shown for Streptococcus pneumoniae 31 , B. pseudomallei 24 and Neisseria meningitidis 32 , which can also invade the CNS via the olfactory and/or trigeminal nerves. The aim of the current study was to build on these previous findings to better understand how glia in the olfactory and trigeminal nerve, as well as astrocytes and microglia are affected by S. aureus infection. We showed that S. aureus could adhere to and become internalized into all four glial types, over 24 h, but found differences between the capacities for intracellular survival between glia. Intracellular survival was significantly higher in astrocytes than in microglia, OECs and TgSCs, suggesting that peripheral nerve glia and microglia (a professional phagocyte) show better capacity for killing intracellular bacteria than astrocytes. Microglia internalised the S. aureus into the perinuclear region within 1 h of exposure to the bacteria. In contrast, OECs, TgSCs and astrocytes took longer to accumulate the bacteria in the perinuclear region. However, even at 24 h astrocytes appeared to have bacteria still within cell processes. This was reflected in the intracellular survival www.nature.com/scientificreports/ with astrocytes retaining significantly more viable bacteria at 24 h compared to the other cell types. The different intracellular distribution of bacteria between astrocytes and peripheral glia may hold clues to the higher capacity for intracellular survival of S. aureus in astrocytes 33 .
We also compared the time-course of phagocytosis between the different types of glia by exposing the cells to pHrodo-green S. aureus BioParticles, which become fluorescent after internalization into phagolysosomes/ lysosomes, and imaging the cells over time. The results showed that microglia and astrocytes were the first cells to show significant uptake of the BioParticles in phagolysosomes/lysosomes, followed by TgSCs and then OECs. Whilst microglia, the resident macrophage of the brain 34 , were the fastest to take up the BioParticles (30 min to start their immune defence function 35 ), astrocytes were faster than TgSCs or OECs. Astrocytes have previously been reported to rapidly (within 2 h) phagocytose damaged cells and synaptosomes 36,37 . Internalization of bacteria (Streptococcus agalactiae) by astrocytes in vitro has been shown to be slower (~ 9 h) and variable between cells 38 . Thus, it is possible that the time-course of BioParticle internalization into astrocytes may vary depending on the type of cargo (cell debris versus bacteria, as well as bacterial species). Previous studies have shown that OECs respond to the presence of cell debris by extending filopodia within 15 min of exposure and to internalize axonal debris within 4 h of exposure 11 . OECs and TgSCs can internalize E. coli bacteria 6 h post exposure 16 . The timing of these responses is relatively similar to the time-course for phagocytosis of S. aureus by OECs and TgSCs reported in the current study. To the best of our knowledge, the invasion frequency (percentage of attached bacteria that end up inside the cell) has not been compared between glia for any bacterial species, previously. These findings show that OECs, TgSCs, astrocytes and microglia all can respond to bacteria and bacteria-conjugated BioParticles, which are internalized into phagolysosomes within 1-4 h, with some differences between cell types regarding the percentages of cells internalizing the cargo. This time-course is similar to what has previously been reported for microglia/macrophages, which respond to bacteria and PAMPs 14 and phagocytose bacteria-conjugated BioParticles in less than 1 h 35 as shown in our results too.
All glia responded to S. aureus with secretion of multiple cytokines and chemokines. Interestingly, production of these was overall highest in microglia and TgSCs, followed by astrocytes and then OECs. This does not  www.nature.com/scientificreports/ match the fact that S. aureus exhibits stronger capacity for intracellular survival in astrocytes than in OECs/ TgSCs, suggesting that other inflammatory mediators than cytokines are involved in differential innate immune responses between peripheral glia and astrocytes. We found all four glial cell types rapidly responded to S. aureus by secretion of TNF-α (with the highest levels being produced by TgSCs). Production of TNF-α by these cells after S. aureus exposure is in alignment with a previous study, in which bacterial invasion of the olfactory nerve and bulb by S. aureus after epithelial injury resulted in increased levels of TNF-α in primary olfactory nervous system tissue 13 . One previous study has also shown that astrocytes can produce TNF-α in response to S. aureus 39 . TNF-α is a critical component of the innate immune response against S. aureus brain abscess 39 , but as it is a pro-inflammatory cytokine, it can also cause damage to brain tissue and death of neurons 40 . TNF-α receptors exist both in neurons and glia, and play an important role in cell death 41 . Due to their location, OECs are often likely exposed to pathogens, and therefore mechanisms must exist that limit damage induced by pro-inflammatory cytokines. OECs and supporting cells of the primary olfactory nervous system produce pituitary adenylate cyclase activating peptide (PACAP) 42 , which protects against TNF-α-mediated death of neurons in both the olfactory nerve 43 and brain 44 , and it has previously been suggested that PACAP may be counteracting potential harmful effects of TNF-α as a response to bacteria in the primary olfactory nervous system 13 . PACAP and its receptors are also expressed in the trigeminal nerve 45 . TNF-α expression in microglia was enhanced as early as 1 h following infection and was elevated up to 24 h which follows a similar trend as shown in a previous study 21 . We also found that the glia secreted another pro-inflammatory cytokine, IFN-γ, in response to S. aureus, at very high levels in microglia and at low levels for the other glial cells. In CNS injury, low levels of IFN-γ can induce neuroprotective functions by microglial cells. At high concentrations of IFN-γ, however, this neuroprotective effect decreases 46 . This was the case in microglia which produced high levels of IFN-γ and positive results for all other 23 cytokines (Supplementary Table 2).
All glia also produced regulatory cytokines IL-6 and IL-10 in response to S. aureus; this response was significantly delayed compared to TNF-α. Production of IL-6 by OECs 13 and astrocytes 39 responding to S. aureus has previously been shown. IL-6 plays a central role in cellular responses to nerve injury and is important for regeneration and cell survival 47,48 . IL-6 has been suggested to prevent cell death of OECs 13 as activation of IL-6 can stimulate anti-apoptotic pathways and the IL-6 receptor is upregulated in OECs after neuronal injury 49 . B. pseudomallei, another bacterium that can invade the olfactory and trigeminal nerves, also stimulates production of IL-6 and TNF-α in OECs 50 . IL-10 is a potent anti-inflammatory cytokine 51,52 that can inhibit TNF-α production, thus regulating potential damaging effects of TNF-α on tissue 53 . This is observed in microglia with high level of IL-10 production corresponding to low level of TNF-α at different time points. At 24 h post exposure to S. aureus, the high level of IL-10 in OECs, TgSCs, astrocytes and microglia, correlated with significantly reduced levels of TNF-α compared to different time points, perhaps suggesting that IL-10 reduced TNF-α secretion.
We found that S. aureus triggered production of several chemokines (chemotactic cytokines), CXCL-1, MCP-1, MIP-1α, MIP-1β and RANTES, and Eotaxin. OECs, but not TgSCs, have previously been demonstrated to respond to E. coli and PAMPs by CXCL-1 secretion 14 , which is demonstrated to be critical for neutrophildependent bacterial elimination via induction of reactive oxygen species and reactive nitrogen species 54 . Astrocytes have also been shown to secrete MIP-1, MCP-2 and MIP-1β in response to S. aureus 39 . These chemokines are part of the main group of cytokines attracting different populations of leukocytes, (preferentially monocytes, macrophages, eosinophils and subsets of lymphocytes) and play important roles in inflammatory responses against pathogens 55,56 . The level of MCP-1 has been shown to be increased in cerebrospinal fluid during pyogenic and tuberculous meningitis and may thus be a common responder of CNS cells to bacterial infection 57 . MCP-1 levels are also increased in plasma in meningococcal disease 58 . Microglia under non-activated conditions produce numerous cytokines such as MIP-1α, MIP-1β and MCP-1 59 and we observed this at 24 h in the non-infected microglia.
It has previously been shown 59 that microglia release a powerful immune response when activated. OECs and astrocytes, but not Schwann cells, mount a similar immune response to E. coli and PAMPs 14 . OECs and astrocytes have also been found to express higher levels of mRNA for innate immune factors than Schwann cells in a microarray study 60 . In contrast to these findings, we found that all glial cells (TgSCs, OECs, astrocytes and microglia) could all internalize and phagocytose S. aureus / S. aureus BioParticles, and that all four glial types responded to S. aureus with secretion of many cytokines and chemokines; in fact, microglia and TgSCs consistently secreted higher amounts of cytokines and chemokines than the other glia. In the previous study 60 , the Schwann cells were derived from the sciatic nerve and brachial plexus. Here, we used trigeminal nerve Schwann cells, which may have evolved to exhibit a more powerful immune response to pathogens than other Schwann cells, as the trigeminal nerve is likely to be more often exposed to microbes than other peripheral nerves.
Due to their unique growth-promoting properties, and because they can be relatively easy isolated from the roof of the nasal cavity, transplantation of OECs is emerging as a promising therapy for spinal cord injury repair 2,61,62 . The innate immune functions of OECs are very important in this context, due to the inflammatory environment of the spinal cord injury site (which in turn varies depending on time post-injury). Depending on the activation state of OECs, they may secrete cytokines that are pro-inflammatory or regulatory and thus have both detrimental and beneficial effects 12,14 . The fact that OECs secreted lower levels of pro-inflammatory cytokines than the other glia indicate that OECs could have immunomodulatory functions, as has been suggested previously 63 . The capacity for phagocytosis by transplanted OECs is also likely beneficial, as the cells can help clear cell debris present at the injury site. Compounds that can stimulate the phagocytic activity of OECs without causing a strong pro-inflammatory response have been suggested to potentially increase the therapeutic potential of these cells [64][65][66][67][68] . Thus, it is important to characterise which cytokines are expressed by OECs under various conditions.
In conclusion, these results have demonstrated that OECs, TgSCs, astrocytes and microglia can phagocytose S. aureus. Whilst the glia mounted an innate immune response, live bacteria could still be isolated from cells after www.nature.com/scientificreports/ 24 h. OECs, TgSCs (PNS) and microglia (CNS) showed stronger capacity for killing of intracellular S. aureus than astrocytes, however, OECs secreted the lowest amounts of both pro-and anti-inflammatory cytokines in response to bacteria, potentially suggesting an immunomodulatory function of these cells.

Materials and methods
Primary glia culture. Glia cultures were obtained from S100β-DsRed transgenic mice according to previously described methods 69 . Briefly, the olfactory bulb and trigeminal nerve were dissected out for preparations of OECs and TgSCs, respectively, from postnatal day seven pups. Tissue explants were plated into the wells of polystyrene 24-well plates, pre-coated with Matrigel basement membrane matrix (Corning Matrigel Basement Membrane Matrix, FAL354234). The explants were maintained in glial medium, constituting of Dulbecco's Modified Eagle Medium (DMEM), 10% foetal bovine serum (FBS) and gentamicin (Gibco 50 mg/mL), supplemented with GlutaMAX and G5 (both Gibco, added according to the manufacturer's instructions), at 37 °C with 5% CO 2 . Cells were cultured to 80% confluency after which they were trypsinized (using Gibco TrypLE Express, 1X) and used for experiments. Primary OEC cultures were > 70% pure and TgSCs cultures were > 80% pure (based on DsRed expression) (Supplementary Fig. 1 and Supplementary Table 1). Primary astrocytes were obtained from postnatal day three pups following a previous published protocol 70 . The brain was removed from the cranium, the olfactory bulb and cerebellum were removed. Careful removal of meninges was performed to avoid contamination of fibroblasts and meningeal cells. Forebrain was carefully separated from midbrain containing major cerebral vessels to avoid endothelial contamination. It was then cut into four smaller pieces followed by trypsinization. The cell suspension was plated in a poly-D-lysine hydrobromide (Sigma-Aldrich P6407)-coated T75 flask. After seven to eight days (90% confluency), astrocytes were separated from microglia (as overlaying microglia sit exposed) by shaking on an orbital shaker at 180 rpm for 30 min. Oligodendrocyte precursor cells were next removed by shaking the flask at 240 rpm for 6 h. The remaining astrocytes were trypsinized (using Gibco TrypLE Express) and used for experiments. Primary astrocyte cultures were > 70% pure (based on GFAP immunostaining) (Supplementary Fig. 1 and Supplementary Table 1).
Primary microglia were prepared from postnatal day 3 (P3) S100ß-DsRed transgenic mice following a previous published protocol 71 . The entire brain cell population was isolated from the brain tissue by enzymatic digestion and mechanical dissociation using Neural Tissue Dissociation Kit with GentleMACS (Miltenyi Biotec,130-093-231). The cell pellet consisting of a mixture of all brain cells was further subjected to magnetic cells sorting for microglia enrichment using CD11b/c microbeads (Miltenyi Biotec,130-093-636) according to manufacturer protocol. The different glial preparations were separately plated in plastic 24-well plates and maintained in glial medium containing Dulbecco's Modified Eagle Medium with 10% foetal bovine serum (FBS), gentamycin (Gibco, 50 mg/mL) and GlutaMAX at 37 °C with 5% CO 2 for 5 days. Cells were replated into T-25 flasks and allowed to proliferate to 80% confluency then trypsinized (using Gibco TrypLE Express) and used for experiments. Primary microglia cultures were > 85% pure (based on Iba-1 immunostaining) (Supplementary Fig. 1  Bacterial strain and culture conditions. Staphylococcus aureus (ATCC 29213) cultures were grown, from a sterile loop inoculum from a glycerol stock, in liquid Brain Heart Infusion (BHI) broth, at 37 °C in a shaking incubator (180-200 rpm) for 14-18 h 72 . After overnight incubation, the bacterial culture was centrifuged at 10,000g for 10 min at 20 °C. The supernatant was removed, and the bacterial pellet was washed with sterile phosphate buffered saline (PBS). The washing step was repeated twice before resuspending the bacteria in antibioticfree medium (DMEM, 10% FBS and GlutaMAX) for experiments. Bacterial concentration was determined by plating on BHI agar plates overnight, after which the number of colony-forming units (CFU) was determined.
In vitro infection assay. To analyse the interaction of S. aureus with glia, we exposed primary OEC, TgSCs, astrocytes and microglia to the bacteria. Glia were seeded in 96-well plates at 4000 cells per well and incubated at 37 °C in 5% CO 2 until approximately 80% confluence. Monolayers were then washed and infected with S. aureus diluted in antibiotic-free medium at a multiplicity of infection (MOI) of 100:1 73 or to medium alone (control) for 1 h. Adhesion of S. aureus to glia was determined at 1 h post exposure by washing monolayers with PBS to remove unattached bacteria, after which adherent bacteria was enumerated by CFU counts on BHI agar 74 . To determine the number of S. aureus CFU that have been internalised into the cells, cells were lysed using 0.05% Triton X (Sigma-Aldrich Triton X-100 laboratory grade) for 2-3 min followed by mixing with PBS and serial ten-fold dilution of the lysate. After which bacterial counts were determined on BHI agar.
To determine invasion and survival frequency of the cells, the cells were exposed to 1 h bacterial inoculation after which the cells were washed at 1 h, 6 h or 24 h with PBS and antibiotic protection media to kill extracellular bacteria. Antibiotic protection media contained penicillin 250 U/mL, streptomycin 250 U/mL (from stock Gibco Penicillin-Streptomycin, 10,000 U/mL) and gentamicin (Gibco, 50 mg/mL). Bacterial load was determined by lysing the cells and determining CFUs on BHI agar. Invasion frequency in % was determined by comparing bacterial invasion of each glial cell type between 6 h post exposure and 1 h post exposure using the following formula: Similarly, data from bacterial load at 24 h post exposure and 6 h post exposure was used to calculate intracellular survival, according to the following formula: www.nature.com/scientificreports/ Phagocytosis assay. To compare the capability for phagocytosis between the different types of glia, we exposed primary OEC, TgSCs, astrocytes and microglia to pHrodo Green S. aureus Bioparticles Conjugate (Invitrogen) and studied internalization into lysosomes over time using live-cell imaging. Glia were seeded in 96-well plates at 4000 cells per well (in normal culture medium, DMEM, 10% FBS, gentamicin, GlutaMAX) and incubated at 37 °C in 5% CO 2 . The bioparticles were added to each well (final concentration: 10 µg/mL) from a stock solution of 1 mg/mL and live cell imaging was performed (red and green channels for DsRed cells with green bioparticles).
Immunofluorescence. Glia  Imaging. Lower power images were captured on a Nikon Eclipse Ti2 inverted microscope. Higher magnification images were taken using an Olympus FV3000 confocal microscope. Images were colour-balanced using Adobe Photoshop CS5 (Adobe Systems Incorporated) with the entire field of view being altered uniformly. Cytokine assay. A set of 23 cytokines were analysed using a highly sensitive antibody-based multiplex cytokine assay kit, the Mouse Cytokine 23-Plex Group 1 kit (Bio-Rad Laboratories). Glia were seeded in 6-well plates at 100,000 cells per well (in antibiotic free culture medium) and incubated at 37 °C in 5% CO 2 . Then the in-vitro assay bacterial protocol was followed till 24 h. Cell culture supernatants were collected, filtered, and centrifuged at 1000 g for 15 min at 4 °C. The samples were stored at − 80 °C until use. On the day of experiment, samples were thawed on ice and diluted with Assay Diluent (from the kit) as directed by the manufacturer. Preparation of standards and assay techniques were followed as per the manufacturer-recommended protocol. All incubation steps were performed on a shaker and washing steps were done using a magnetic plate wash station. The beads were resuspended by shaking the plate vigorously on an orbital shaker, and immediately analysed using a Bio-Plex 200 Multiplex Reader instrument (Bio-Rad Laboratories), following the manufacturersuggested settings. The Bio-Plex Manager Software (Bio-Rad Laboratories) was used for instrument control, data acquisition, and data analysis. The acquired data and graphs (standard curves) were exported into Microsoft Excel and then GraphPad PRISM version 7 (www. graph pad. com/ scien tific-softw are/ prism) was used for further analysis and preparation of graphs for figures.